Waste Materials

EDITOR IN CHIEF Rudy J. M. Konings European Commission, Joint Research Centre, Institute for Transuranium Elements, Karlsruhe, Germany

SECTION EDITORS Todd R. Allen Department of Engineering Physics, University of Wisconsin, Madison, WI, USA Roger E. Stoller Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA Shinsuke Yamanaka Division of Sustainable Energy and Environmental Engineering, Graduate School of Engineering, Osaka University, Osaka, Japan

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK 225 Wyman Street, Waltham, MA 02451, USA Copyright © 2012 Elsevier Ltd. All rights reserved The following articles are US Government works in the public domain and not subject to copyright: Radiation Effects in UO2 TRISO-Coated Particle Fuel Performance Composite Fuel (cermet, cercer) Metal Fuel-Cladding Interaction No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (þ44) (0) 1865 843830; fax (þ44) (0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein, Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Catalog Number: 2011929343 ISBN (print): 978-0-08-056027-4 For information on all Elsevier publications visit our website at books.elsevier.com Cover image courtesy of Professor David Sedmidubsky´, The Institute of Chemical Technology, Prague Printed and bound in Spain 12 13 14 15 16 10 9 8 7 6 5 4 3 2 1

Editorial : Gemma Mattingley Production: Nicky Carter

EDITORS BIOGRAPHIES Rudy Konings is currently head of the Materials Research Unit in the Institute for Transuranium Elements (ITU) of the Joint Research Centre of the European Commission. His research interests are nuclear reactor fuels and actinide materials, with particular emphasis on high temperature chemistry and thermodynamics. Before joining ITU, he worked on nuclear fuel-related issues at ECN (the Energy Research Centre of the Netherlands) and NRG (Nuclear Research and Consultancy Group) in the Netherlands. Rudy is editor of Journal of Nuclear Materials and is professor at the Delft University of Technology (Netherlands), where he holds the chair of ‘Chemistry of the nuclear fuel cycle.’

Roger Stoller is currently a Distinguished Research Staff Member in the Materials Science and Technology Division of the Oak Ridge National Laboratory and serves as the ORNL Program Manager for Fusion Reactor Materials for ORNL. He joined ORNL in 1984 and is actively involved in research on the effects of radiation on structural materials and fuels for nuclear energy systems. His primary expertise is in the area of computational modeling and simulation. He has authored or coauthored more than 100 publications and reports on the effects of radiation on materials, as well as edited the proceedings of several international conferences.

Todd Allen is an Associate Professor in the Department of Engineering Physics at the University of Wisconsin – Madison since 2003. Todd’s research expertise is in the area of materials-related issues in nuclear reactors, specifically radiation damage and corrosion. He is also the Scientific Director for the Advanced Test Reactor National Scientific User Facility as well as the Director for the Center for Material Science of Nuclear Fuel at the Idaho National Laboratory, positions he holds in conjunction with his faculty position at the University of Wisconsin.

v

vi

Editors Biographies

Shinsuke Yamanaka is a professor in Division of Sustainable Energy and Environmental Engineering, Graduate School of Engineering, Osaka University since 1998. He has studied the thermophysics and thermochemistry of nuclear fuel and materials. His research for the hydrogen behavior in LWR fuel cladding is notable among his achievements and he received the Young Scientist Awards (1980) and the Best Paper Awards (2004) from Japan Atomic Energy Society. Shinsuke is the program officer of Japan Science and Technology Agency since 2005 and the visiting professor of Fukui University since 2009, and he is also the associate dean of Graduate School of Engineering, Osaka University since 2011.

PREFACE There are essentially three primary energy sources for the billions of people living on the earth’s surface: the sun, radioactivity, and gravitation. The sun, an enormous nuclear fusion reactor, has transmitted energy to the earth for billions of years, sustaining photosynthesis, which in turn produces wood and other combustible resources (biomass), and the fossil fuels like coal, oil, and natural gas. The sun also provides the energy that steers the climate, the atmospheric circulations, and thus ‘fuelling’ wind mills, and it is at the origin of photovoltaic processes used to produce electricity. Radioactive decay of primarily uranium and thorium heats the earth underneath us and is the origin of geothermal energy. Hot springs have been used as a source of energy from the early days of humanity, although it took until the twentieth century for the potential of radioactivity by fission to be discovered. Gravitation, a non-nuclear source, has been long used to generate energy, primarily in hydropower and tidal power applications. Although nuclear processes are thus omnipresent, nuclear technology is relatively young. But from the moment scientists unraveled the secrets of the atom and its nucleus during the twentieth century, aided by developments in quantum mechanics, and obtained a fundamental understanding of nuclear fission and fusion, humanity has considered these nuclear processes as sources of almost unlimited (peaceful) energy. The first fission reactor was designed and constructed by Enrico Fermi in 1942 in Chicago, the CP1, based on the fission of uranium by neutron capture. After World War II, a rapid exploration of fission technology took place in the United States and the Union of Soviet Socialist Republics, and after the Atoms for Peace speech by Eisenhower at the United Nations Congress in 1954, also in Europe and Japan. A variety of nuclear fission reactors were explored for electricity generation and with them the fuel cycle. Moreover, the possibility of controlled fusion reactions has gained interest as a technology for producing energy from one of the most abundant elements on earth, hydrogen. The environment to which materials in nuclear reactors are exposed is one of extremes with respect to temperature and radiation. Fuel pins for nuclear reactors operate at temperatures above 1000  C in the center of the pellets, in fast reactor oxide fuels even above 2000  C, whereas the effects of the radiation (neutrons, alpha particles, recoil atoms, fission fragments) continuously damage the material. The cladding of the fuel and the structural and functional materials in the fission reactor core also operate in a strong radiation field, often in a dynamic corrosive environment of the coolant at elevated temperatures. Materials in fusion reactors are exposed to the fusion plasma and the highly energetic particles escaping from it. Furthermore, in this technology, the reactor core structures operate at high temperatures. Materials science for nuclear systems has, therefore, been strongly focussed on the development of radiation tolerant materials that can operate in a wide range of temperatures and in different chemical environments such as aqueous solutions, liquid metals, molten salts, or gases. The lifetime of the plant components is critical in many respects and thus strongly affects the safety as well as the economics of the technologies. With the need for efficiency and competitiveness in modern society, there is a strong incentive to improve reactor components or to deploy advanced materials that are continuously developed for improved performance. There are many examples of excellent achievements in this respect. For example, with the increase of the burnup of the fuel for fission reactors, motivated by improved economics and a more efficient use of resources, the Zircaloy cladding (a Zr–Sn alloy) of the fuel pins showed increased susceptibility to coolant corrosion, but within a relatively short period, a different zirconium-based alloy was developed, tested, qualified, and employed, which allowed reliable operation in the high burnup range.

xxi

xxii

Preface

Nuclear technologies also produce waste. It is the moral obligation of the generations consuming the energy to implement an acceptable waste treatment and disposal strategy. The inherent complication of radioactivity, the decay that can span hundreds of thousands of years, amplifies the importance of extreme time periods in the issue of corrosion and radiation stability. The search for storage concepts that can guarantee the safe storage and isolation of radioactive waste is, therefore, another challenging task for materials science, requiring a close examination of natural (geological) materials and processes. The more than 50 years of research and development of fission and fusion reactors have undoubtedly demonstrated that the statement ‘technologies are enabled by materials’ is particularly true for nuclear technology. Although the nuclear field is typically known for its incremental progress, the challenges posed by the next generation of fission reactors (Generation IV) as well as the demonstration of fusion reactors will need breakthroughs to achieve their ambitious goals. This is being accompanied by an important change in materials science, with a shift of discovery through experiments to discovery through simulation. The progress in numerical simulation of the material evolution on a scientific and engineering scale is growing rapidly. Simulation techniques at the atomistic or meso scale (e.g., electronic structure calculations, molecular dynamics, kinetic Monte Carlo) are increasingly helping to unravel the complex processes occurring in materials under extreme conditions and to provide an insight into the causes and thus helping to design remedies. In this context, Comprehensive Nuclear Materials aims to provide fundamental information on the vast variety of materials employed in the broad field of nuclear technology. But to do justice to the comprehensiveness of the work, fundamental issues are also addressed in detail, as well as the basics of the emerging numerical simulation techniques. R.J.M. Konings European Commission, Joint Research Centre, Institute for Transuranium Elements, Karlsruhe, Germany T.R. Allen Department of Engineering Physics, Wisconsin University, Madison, WI, USA R. Stoller Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA S. Yamanaka Division of Sustainable Energy and Environmental Engineering, Graduate School of Engineering, Osaka University, Osaka, Japan

FOREWORD ‘Nuclear materials’ denotes a field of great breadth and depth, whose topics address applications and facilities that depend upon nuclear reactions. The major topics within the field are devoted to the materials science and engineering surrounding fission and fusion reactions in energy conversion reactors. Most of the rest of the field is formed of the closely related materials science needed for the effects of energetic particles on the targets and other radiation areas of charged particle accelerators and plasma devices. A more complete but also more cumbersome descriptor thus would be ‘the science and engineering of materials for fission reactors, fusion reactors, and closely related topics.’ In these areas, the very existence of such technologies turns upon our capabilities to understand the physical behavior of materials. Performance of facilities and components to the demanding limits required is dictated by the capabilities of materials to withstand unique and aggressive environments. The unifying concept that runs through all aspects is the effect of radiation on materials. In this way, the main feature is somewhat analogous to the unifying concept of elevated temperature in that part of materials science and engineering termed ‘high-temperature materials.’ Nuclear materials came into existence in the 1950s and began to grow as an internationally recognized field of endeavor late in that decade. The beginning in this field has been attributed to presentations and discussions that occurred at the First and Second International Conferences on the Peaceful Uses of Atomic Energy, held in Geneva in 1955 and 1958. Journal of Nuclear Materials, which is the home journal for this area of materials science, was founded in 1959. The development of nuclear materials science and engineering took place in the same rapid growth time period as the parent field of materials science and engineering. And similarly to the parent field, nuclear materials draws together the formerly separate disciplines of metallurgy, solid-state physics, ceramics, and materials chemistry that were early devoted to nuclear applications. The small priesthood of first researchers in half a dozen countries has now grown to a cohort of thousands, whose home institutions are anchored in more than 40 nations. The prodigious work, ‘Comprehensive Nuclear Materials,’ captures the essence and the extensive scope of the field. It provides authoritative chapters that review the full range of endeavor. In the present day of glance and click ‘reading’ of short snippets from the internet, this is an old-fashioned book in the best sense of the word, which will be available in both electronic and printed form. All of the main segments of the field are covered, as well as most of the specialized areas and subtopics. With well over 100 chapters, the reader finds thorough coverage on topics ranging from fundamentals of atom movements after displacement by energetic particles to testing and engineering analysis methods of large components. All the materials classes that have main application in nuclear technologies are visited, and the most important of them are covered in exhaustive fashion. Authors of the chapters are practitioners who are at the highest level of achievement and knowledge in their respective areas. Many of these authors not only have lived through a substantial part of the history sketched above, but they themselves are the architects. Without those represented here in the author list, the field would certainly be a weaker reflection of itself. It is no small feat that so many of my distinguished colleagues could have been persuaded to join this collective endeavor and to make the real sacrifices entailed in such time-consuming work. I congratulate the Editor, Rudy Konings, and

xxiii

xxiv

Foreword

the Associate Editors, Roger Stoller, Todd Allen, and Shinsuke Yamanaka. This book will be an important asset to young researchers entering the field as well as a valuable resource to workers engaged in the enterprise at present. Dr. Louis K. Mansur Oak Ridge, Tennessee, USA

Permission Acknowledgments The following material is reproduced with kind permission of Cambridge University Press Figure 15 of Oxide Dispersion Strengthened Steels Figure 15 of Minerals and Natural Analogues Table 10 of Spent Fuel as Waste Material Figure 21b of Radiation-Induced Effects on Microstructure www.cambridge.org The following material is reproduced with kind permission of American Chemical Society Figure 2 of Molten Salt Reactor Fuel and Coolant Figure 22 of Molten Salt Reactor Fuel and Coolant Table 9 of Molten Salt Reactor Fuel and Coolant Figure 6 of Thermodynamic and Thermophysical Properties of the Actinide Nitrides www.acs.org The following material is reproduced with kind permission of Wiley Table 3 of Properties and Characteristics of SiC and SiC/SiC Composites Table 4 of Properties and Characteristics of SiC and SiC/SiC Composites Table 5 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 5 of Advanced Concepts in TRISO Fuel Figure 6 of Advanced Concepts in TRISO Fuel Figure 30 of Material Performance in Supercritical Water Figure 32 of Material Performance in Supercritical Water Figure 19 of Tritium Barriers and Tritium Diffusion in Fusion Reactors Figure 9 of Waste Containers Figure 13 of Waste Containers Figure 21 of Waste Containers Figure 11 of Carbide Fuel Figure 12 of Carbide Fuel Figure 13 of Carbide Fuel Figure 4 of Thermodynamic and Thermophysical Properties of the Actinide Nitrides Figure 2 of The U–F system Figure 18 of Fundamental Point Defect Properties in Ceramics Table 1 of Fundamental Point Defect Properties in Ceramics Figure 17 of Radiation Effects in SiC and SiC-SiC Figure 21 of Radiation Effects in SiC and SiC-SiC Figure 6 of Radiation Damage in Austenitic Steels Figure 7 of Radiation Damage in Austenitic Steels Figure 17 of Ceramic Breeder Materials Figure 33a of Carbon as a Fusion Plasma-Facing Material Figure 34 of Carbon as a Fusion Plasma-Facing Material i

ii

Permission Acknowledgments

Figure 39 of Carbon as a Fusion Plasma-Facing Material Figure 40 of Carbon as a Fusion Plasma-Facing Material Table 5 of Carbon as a Fusion Plasma-Facing Material www.wiley.com The following material is reproduced with kind permission of Springer Figure 4 of Neutron Reflector Materials (Be, Hydrides) Figure 6 of Neutron Reflector Materials (Be, Hydrides) Figure 1 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 3 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 4 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 5 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 6 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 7 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 8 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 9 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 10 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 11 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 12 of Properties and Characteristics of SiC and SiC/SiC Composites Figure 22d of Fission Product Chemistry in Oxide Fuels Figure 3 of Behavior of LWR Fuel During Loss-of-Coolant Accidents Figure 14a of Irradiation Assisted Stress Corrosion Cracking Figure 14b of Irradiation Assisted Stress Corrosion Cracking Figure 14c of Irradiation Assisted Stress Corrosion Cracking Figure 25a of Irradiation Assisted Stress Corrosion Cracking Figure 25b of Irradiation Assisted Stress Corrosion Cracking Figure 1 of Properties of Liquid Metal Coolants Figure 5b of Fast Spectrum Control Rod Materials Figure 3 of Oxide Fuel Performance Modeling and Simulations Figure 8 of Oxide Fuel Performance Modeling and Simulations Figure 10 of Oxide Fuel Performance Modeling and Simulations Figure 11 of Oxide Fuel Performance Modeling and Simulations Figure 14 of Oxide Fuel Performance Modeling and Simulations Figure 5 of Thermodynamic and Thermophysical Properties of the Actinide Nitrides Figure 51 of Phase Diagrams of Actinide Alloys Figure 6 of Thermodynamic and Thermophysical Properties of the Actinide Oxides Figure 7b of Thermodynamic and Thermophysical Properties of the Actinide Oxides Figure 9b of Thermodynamic and Thermophysical Properties of the Actinide Oxides Figure 35 of Thermodynamic and Thermophysical Properties of the Actinide Oxides Table 11 of Thermodynamic and Thermophysical Properties of the Actinide Oxides Table 13 of Thermodynamic and Thermophysical Properties of the Actinide Oxides Table 17 of Thermodynamic and Thermophysical Properties of the Actinide Oxides Figure 18 of Radiation Damage of Reactor Pressure Vessel Steels Figure 7 of Radiation Damage Using Ion Beams Figure 9b of Radiation Damage Using Ion Beams Figure 28 of Radiation Damage Using Ion Beams Figure 34 of Radiation Damage Using Ion Beams Figure 35 of Radiation Damage Using Ion Beams Figure 36d of Radiation Damage Using Ion Beams Figure 37 of Radiation Damage Using Ion Beams Table 3 of Radiation Damage Using Ion Beams

Permission Acknowledgments

Figure 5 of Radiation Effects in UO2 Figure 9a of Ab Initio Electronic Structure Calculations for Nuclear Materials Figure 9b of Ab Initio Electronic Structure Calculations for Nuclear Materials Figure 9c of Ab Initio Electronic Structure Calculations for Nuclear Materials Figure 10a of Ab Initio Electronic Structure Calculations for Nuclear Materials Figure 23 of Thermodynamic and Thermophysical Properties of the Actinide Carbides Figure 25 of Thermodynamic and Thermophysical Properties of the Actinide Carbides Figure 26 of Thermodynamic and Thermophysical Properties of the Actinide Carbides Figure 27 of Thermodynamic and Thermophysical Properties of the Actinide Carbides Figure 28a of Thermodynamic and Thermophysical Properties of the Actinide Carbides Figure 28b of Thermodynamic and Thermophysical Properties of the Actinide Carbides Figure 2 of Physical and Mechanical Properties of Copper and Copper Alloys Figure 5 of Physical and Mechanical Properties of Copper and Copper Alloys Figure 6 of The Actinides Elements: Properties and Characteristics Figure 10 of The Actinides Elements: Properties and Characteristics Figure 11 of The Actinides Elements: Properties and Characteristics Figure 12 of The Actinides Elements: Properties and Characteristics Figure 15 of The Actinides Elements: Properties and Characteristics Table 1 of The Actinides Elements: Properties and Characteristics Table 6 of The Actinides Elements: Properties and Characteristics Figure 25 of Fundamental Properties of Defects in Metals Table 1 of Fundamental Properties of Defects in Metals Table 7 of Fundamental Properties of Defects in Metals Table 8 of Fundamental Properties of Defects in Metals www.springer.com The following material is reproduced with kind permission of Taylor & Francis Figure 9 of Radiation-Induced Segregation Figure 6 of Radiation Effects in Zirconium Alloys Figure 1 of Dislocation Dynamics Figure 25 of Radiation Damage Using Ion Beams Figure 26 of Radiation Damage Using Ion Beams Figure 27 of Radiation Damage Using Ion Beams Figure 4 of Radiation-Induced Effects on Material Properties of Ceramics (Mechanical and Dimensional) Figure 7 of The Actinides Elements: Properties and Characteristics Figure 20 of The Actinides Elements: Properties and Characteristics Figure 18a of Primary Radiation Damage Formation Figure 18b of Primary Radiation Damage Formation Figure 18c of Primary Radiation Damage Formation Figure 18d of Primary Radiation Damage Formation Figure 18e of Primary Radiation Damage Formation Figure 18f of Primary Radiation Damage Formation Figure 1 of Radiation-Induced Effects on Microstructure Figure 27 of Radiation-Induced Effects on Microstructure Figure 5 of Performance of Aluminum in Research Reactors Figure 2 of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 3 of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 5 of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 10a of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 10b of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 10c of Atomic-Level Dislocation Dynamics in Irradiated Metals

iii

iv

Permission Acknowledgments

Figure 10d of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 12a of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 12b of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 12c of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 12d of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 16a of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 16b of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 16c of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 16d of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 16e of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 17a of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 17b of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 17c of Atomic-Level Dislocation Dynamics in Irradiated Metals Figure 17d of Atomic-Level Dislocation Dynamics in Irradiated Metals www.taylorandfrancisgroup.com

5.01

Corrosion and Compatibility

S. Lillard Los Alamos National Laboratory, Los Alamos, NM, USA

ß 2012 Elsevier Ltd. All rights reserved.

2

5.01.1

Theory

5.01.1.1 5.01.1.2 5.01.1.3 5.01.1.4 5.01.1.5 5.01.1.6 5.01.2 5.01.2.1 5.01.2.2 5.01.2.3 5.01.2.4 References

Introduction Half Cell Reactions Cell Potentials and the Nernst Equation Reference Electrodes and Their Application to Nuclear Systems The Thermodynamics of Corrosion from Room Temperature to the PWR Kinetics of Dissolution and Passive Film Formation Analytical Methods Introduction Potentiodynamic Polarization Electrochemical Impedance Spectroscopy Mott–Schottky Analysis

Abbreviations BWR CNLS

Boiling water reactor Complex nonlinear least squares fitting of the data EC Electrical equivalent circuit EIS Electrochemical impedance spectroscopy EPBRE External pressure-balanced reference electrode FFTF Fast Flux Test Facility HIC Hydrogen-induced cracking HIFER Hi-Flux Isotope Reactor IG Intergranular LBE Lead–bismuth eutectic PWR Pressurized water reactor SCC Stress corrosion cracking SS Stainless steel

Symbols A ai C ci CR E EW Ecorr f

Surface area Activity of species i Capacitance Concentration of species i Corrosion rate Potential Equivalent weight Corrosion potential Mass fraction

ƒ F i icorr ji k L M MM n ND Q r R Rp RV S t ti T Vo z Z Z0 Z00 jZj b ba bc d

2 2 3 4 6 8 10 10 11 12 14 16

Fugacity Faraday’s constant Current density Corrosion current density Square root of
1 Rate constant Oxide thickness Molecular weight Metal cation Number of electrons Donor concentration Reaction quotient Rate of reaction Gas constant Polarization resistance Solution resistance Entropy of transport Time Transport number of species i Temperature Oxygen vacancy Charge Impedance Real part of the impedance Imaginary part of the impedance Magnitude of the impedance Symmetry factor Anodic Tafel slope Cathodic Tafel slope Double layer thickness

1

2

Corrosion and Compatibility 0

DCp Change in standard partial molar heat capacity DE0 Standard reduction potential DG Change in Gibbs energy DG0 Standard Gibbs energy DS0 Standard entropy change « Electronic charge Permittivity of space «0 f Applied potential h Overpotential u Phase angle r Material density v Frequency

5.01.1 Theory 5.01.1.1

Introduction

Mars Fontana identified eight forms of corrosion in his book Corrosion Engineering1 and it is quite easy to find examples of almost all of these in nuclear reactors in both the primary and secondary cooling water systems. For example, galvanic corrosion in zirconium
stainless steel couples,2,3 crevice corrosion in tube sheets4 and former baffle bolts,5 and pitting corrosion in alloy 600 steam generator tubes.6,7 Perhaps the most infamous form of corrosion observed in nuclear reactors is stress corrosion cracking (SCC), or environmental fracture, as we shall refer to it here, which has numerous examples in the literature. Environmental fracture includes both intergranular SCC (IG), such as that which occurs in austenitic stainless steel, and hydrogen-induced cracking (HIC), frequently observed in nickel base alloys. Failure by one of these mechanisms results from an interplay between stress, microstructure, and the environment (e.g., the electrochemical interface). The goal of this chapter is not to address each of the corrosion mechanisms outlined by Fontana individually, that will be accomplished in the following chapters. Rather, this chapter is meant to provide the reader with the fundamental electrochemical theory necessary to critically evaluate the data and discussions in the corrosion chapters that follow. In this section, we will review the fundamental theory of the electrochemical interface. In the first three subsections, we review Half Cell Reaction, Cell Potentials and the Nernst Equation, and Reference Electrodes in Nuclear Systems. In these sections, we develop the theory necessary to understand the role of electrochemical potential in environmental fracture

and corrosion mechanisms. For example, intergranular stress corrosion cracking (IGSCC) is only observed at potentials more positive than a critical value while HIC is only observed at potentials more negative than a critical value. In the remaining two sections, we review the Thermodynamics from Room Temperature to the pressurized water reactor (PWR) and Kinetics of Dissolution and Passive Film Formation. These sections should help the reader to understand the role of the passive film in the corrosion mechanism and the competition that occurs between film formation and metal dissolution rate. As the fundamental role of irradiation in corrosion and environmental fracture mechanism is far from well established, in each section, we incorporate empirical irradiation data as examples and discuss concepts that are more broadly important to nuclear systems. 5.01.1.2

Half Cell Reactions

The electrochemical interface is characterized by an electrode (in this case a metal such as a cooling pipe) and an electrolyte (e.g., the cooling water in a reactor). While the bulk electrolyte contributes to variables such as solution chemistry and ohmic drop (solution resistance is discussed later in this chapter), it is the first nanometer of electrolyte that plays the most important role in electrochemistry. In this short distance, referred to as the electrochemical double layer, a separation of charge occurs. It is this separation of charge that provides the driving force (potential drop) for corrosion reactions. For example, a 100 mV-applied potential across a typical double layer will result in an electric field on the order of 106 V cm
2. In the model proposed by Helmholtz,8 the double layer may be thought of as capacitor, with positive charge on the metal electrode and the adsorption of negatively charged cations on the solution side (Figure 1). The capacitance of the double layer is equal to that in its electrical analog e0D/d, where e0 is the permittivity of space, D is the dielectric, and d is the thickness of the layer. For most electrochemical double layers, C is on the order of 10
6 F cm
2. Electrochemical reactions that take place in the double layer are reactions in which a transfer of charge (electrons) occurs. There are two different types of cells in which electrochemical reactions may occur9: Electrolytic cells in which work, in the form of electrical energy, is required to bring about a nonspontaneous reaction.

Corrosion and Compatibility

Metal

Because the system cannot store charge, the electrons produced during the anodic reaction must be used. This occurs at the cathode where typical reactions may include oxygen reduction:

Excess negative charge

Excess positive charge

Bulk solution

+

-

+

-

+

-

+

-

+

-

+

+

-

+

Acid: O2 þ 4Hþ þ 4e
) 2H2 O

-

Base: O2 þ 2H2 O þ 4e
) 4OH
+

-

3

½II ½III

or hydrogen reduction: 2Hþ þ 2e
) H2

fmetal

½IV

From eqns [I] and [III], the general corrosion of an Fe surface in basic solution may then be written as: 2Fe þ O2 þ 2H2 O ) 2FeðOHÞ2

fsolution Double layer Figure 1 A diagram depicting the separation of charge at the electrochemical double layer and the associated potential drop (f).

H2O 2H+

ClH2

H2O H+

H2O

ClH2O

Cl-

H+

Oxide Fe

For any chemical reaction the driving force, the Gibbs energy, may be written as10:

Figure 2 Diagram of what the anodic and cathodic reactions may look like on an iron surface depicting the separation of reactions and ionic conduction.

DG ¼ DG 0 þ RT lnQ

Voltaic cells in which a spontaneous reaction occurs resulting in work in the form of electrical energy. Electrolytic cells cover a fairly large number of electrochemical reactions but may generally be thought of as ‘plating’ or ‘electrolysis’ type reactions and will not be treated here. Corrosion reactions are voltaic cells and will be the focus of this chapter. As in an electrolytic cell, voltaic cells are characterized by two separate electrodes, an anode and a cathode. In corrosion, reactions at the anode take the form of metal dissolution, the formation of a soluble metal cation: Fe ) Feþ2 þ 2e


where Fe(OH)2 is the corrosion product. An example of what the anodic and cathodic reactions on Fe electrode might look like is presented in Figure 2. Though the anodic and cathodic reactions occur at physically separate locations, as shown in this figure, the reactions must be connected via an electrolyte (aqueous solution). Figure 2 also suggests that corrosion reactions are controlled by variables such as mass transport (diffusion, convection, migration), concentration, and ohmic drop (resistivity of the electrolyte). These variables will be considered in our discussion of corrosion kinetics. 5.01.1.3 Cell Potentials and the Nernst Equation

Fe2+

e-

½V

½I

½1

where DG0 is the standard Gibbs energy, Q is the reaction quotient equal to the product of the activities (assumed to obey Raoult’s Law for dilute solutions and, thus, equal to the concentration) of the products divided by the reactants, R is the gas constant, and T is temperature. The electrical potential, E, is related to the Gibbs energy of a cell by the relationship: nFE ¼
DG

½2

where n is the number of electrons participating in the reaction and F is Faraday’s constant. For the reduction of hydrogen on platinum: 1 þ þ e
ðPtÞ , H2ðgÞ Haq 2

½VI

4

Corrosion and Compatibility

The reaction quotient, Q (starting conditions) becomes: Q ¼

½fH2 1=2 ½Hþ 

½3

where fH2 is the fugacity of hydrogen gas. Substituting eqns [2] and [3] into eqn [1], we find for the reduction of hydrogen on platinum that: " # RT ½fH2 1=2 0 ½4 ln E ¼ DE þ ½Hþ  F where F is Faraday’s constant and DE0 is the standard reduction potential for the reaction in eqn [VI]. Equation [10] is commonly referred to as the Nernst equation and defines the equilibrium reduction potential of the half cell and is pH dependent. The Nernst equation is commonly expressed in its generalized form as: E ¼ DE 0 þ

RT ln½Q  F

½5

5.01.1.4 Reference Electrodes and Their Application to Nuclear Systems In Equation [4], all of the parameters are easily calculated with one exception, DE0. Therefore, we define DE0 ¼ 0 in eqn [4] for a set of specific parameters and refer to this cell as the standard hydrogen electrode (SHE): H2 pressure of 1 atm, a pH ¼ 0, and a temperature of 25  C. This provides a reference from which we can calculate the standard potentials for all other reduction reactions using eqn [5]. These are referred to as standard reduction potentials and a few examples are provided in Table 1. While the SHE is the accepted standard, from a practical standpoint, this reference electrode is difficult to construct and maintain. As such, experimentalists have taken advantage of a number of other reduction reactions to construct reference electrodes for laboratory use. The reaction selected typically depends on the application. One common reference electrode is the silver–silver chloride electrode (Ag/AgCl) which is based on the reduction of Agþ in a solution of potassium chloride: Agþ þ e
, AgðsÞ

½VII

Agþ þ Cl
, AgClðsÞ

½VIII

and the overall reaction being: AgðsÞ þ Cl
, AgClðsÞ þ e


½XI

Table 1 Standard reduction potentials for several reactions important to the nuclear power industry Reduction reaction

Standard reduction potential (V)

Au3þ þ 3e
⇄ Au Cl2 þ 2e
⇄ 2Cl
O2 þ 4Hþ þ 4e
⇄ 2H2O Agþ þ e
⇄ Ag Fe3þ þ e
⇄ Fe2þ O2 þ 2H2O þ 4e
⇄ 4OH
AgCl þ e
⇄ Ag þ Cl
2Hþ þ 2e
⇄ H2 (NHE) Ni2þ þ 2e
⇄ Ni Fe2þ þ 2e
⇄ Fe Cr3þ þ 3e
⇄ Cr Zr4þ þ 4e
⇄ Zr Al3þ þ 3e
⇄ Al Liþ þ e
⇄ Li

1.52 1.36 1.23 0.80 0.77 0.4 0.22 0.0
0.25
0.44
0.74
1.53
1.66
3.04

The Nernst equation for eqn [XI] is equal to:   ½aAgCl  RT 0 E ¼ DEAg=AgCl þ ln ½aAg ½aCl
 F 0 ¼ DEAg=AgCl
ln½aCl


½6

where aCl
is the activity of chloride and for which the concentration (mCl
) in molal (mol kg
1) is frequently substituted. In the corrosion lab, the reference electrode is constructed by electrochemically depositing an AgCl layer onto a silver wire. This wire is then placed in a glass capillary filled with a solution of potassium chloride the concentration of which then defines the cell potential (aCl
in eqn [6]). One end of the capillary is sealed using a porous frit (typically a porous polymer) that acts as a junction between the solution of the reference electrode and the environment of the corrosion experiment. While the Ag/AgCl reference electrode construction described above is straightforward for the lab, there are several obstacles that must be overcome before it can be used in a nuclear power plant setting, namely, radiation flux, pressure, and temperature. As it turns out, the primary impact of ionizing radiation on laboratory reference electrodes relates to damage of the cotton wadding and polymer frits used in their construction and no change in cell potential occurs.11 As such, two approaches based on the Ag/AgCl reference electrode have been used to measure electrode potential in nuclear power reactors. In the first approach, an internal reference electrode operates in the same high-temperature environment as the reactor. In this case, one must consider the solubility of

Corrosion and Compatibility

Sapphire lid AgCl pellet

5

Rulon adapter Compression fitting

Sapphire container

Pt cap Ni wire Alumina insulators

Ceramic to metal braze Restrainer

Ag/AgCl

Kovar TIG weld 304SS

Seal

Coaxial cable Figure 3 Diagram of an internal Ag/AgCl reference electrode used in BWRs. Top end is inserted into the cooling loop, while the coaxial cable provides electrical connection. Reprinted from Indig, M. E. In 12th International Corrosion Congress, Corrosion Control for Low-Cost Reliability; NACE International: Houston, TX, 1993; p 4224, with permission from NACE International.

Ag/Cl complexes that form as a function of temperature eqn [6].12,13 That is, reactions in addition to eqns [VIII] and [XI] must be considered. From a construction viewpoint, the internal reference electrode consists of a silver chloride pellet on a platinum foil (Figure 3).14 External potential measurement is made via contact with a nickel wire which is connected to an electrometer via a coaxial cable. The electrode is housed in a sapphire tube that is sealed via a porous sapphire cap. In this configuration, there is no internal electrolyte per se. Upon placing the electrode in a boiling water reactor (BWR), the porous cap allows the cooling water to penetrate the electrode and the potential is determined from eqn [6] and the solubility of AgCl in high purity water as a function of temperature.15 In the second approach, an external pressurebalanced reference electrode is used (EPBRE). In the EPBRE, the reference electrode is maintained at room temperature and pressure and the corresponding constants are used in eqn [6]. The reference is connected to the high-temperature environment via a nonisothermal salt bridge sealed with a porous zirconia plug (Figure 4).16 As a result of this configuration, the EPBRE is not susceptible to

Compression fitting 1/4 NPT Pure water or 0.01 M KCl Glass wick 1/4 OD SS tube

Heat-shrinkable PTFE tube

SS nut Rulon sleeve Zirconia plug

Figure 4 A diagram of a pressure-balanced reference electrode is used in BWRs. Bottom of figure is sealed into pressure vessel via compression fitting while Ag/AgCl electrode (top) remains at room temperature and pressure. Reprinted from Oh, S. H.; Bahn, C. B.; Hwang, I. S. J. Electrochem. Soc. 2003, 150, E321, with permission from The Electrochemical Society.

potential deviations owing to the solubility of AgCl and its complexes as a function of temperature. However, the temperature gradient between the reactor and the reference electrode results in a junction potential that must be subtracted from eqn [6]. The corresponding thermal liquid junction potential (ETLJ)17 is given by:  ð  1 T2 tMþ SMþ tCl
SCl
dT ½7 ETLJ ¼
þ zMþ zCl
F T1 where t, S, and z represent the transport number, the entropy of transport, and the charge on the cation, respectively. The symbol M in eqn [7] represents the metal in the chloride salt, MCl, and is commonly Li, Na, or K. In addition to ETLJ, there is also the isothermal liquid junction potential, EILJ, which arises due to the differences in cation and anion mobilities through the porous frits and the fact that the electrolyte in the external reference (typically KCl) is vastly different from the reactor cooling water in which it is immersed17: ð RT T2 ti EILJ ¼
dln½ai  ½8 F T1 zi

6

Corrosion and Compatibility

where the subscript i denotes a species that may be transported through the zirconia plug and for a nuclear power reactor may include species ions as Agþ, Cl
, Hþ, OH
, Kþ, and B(OH4)
. It has been shown that both ETLJ and EILJ each increase by as much as 0.15 V over the temperature range of 25–350  C. The result is a decrease in the measured potential of 0.30 V at 350  C. While these junction potentials can be calculated and used to correct eqn [7], it has been shown that there is some deviation at higher temperatures (>200  C) and an experimental fitting procedure is the preferred method for calibration of the reference electrode.

reactions with two soluble species

 3þ  Fe ½11 Fe2þ ¼ Fe3þ þ e
E 0 ¼ 0:771 þ 0:059log Fe2þ

solubility of iron and its oxides Fe ¼ Fe2þ þ 2e
E 0 ¼
0:440 þ 0:0295logðFe2þ Þ

2Fe2þ þ 3H2 O ¼ Fe2 O3 þ 6Hþ þ 2e
E 0 ¼ 0:728
0:177pH
0:059logðFe2þ Þ

An atlas of electrochemical equilibria has been created by M. Pourbaix for metals in aqueous solution at room temperature.18 This atlas contains potential– pH diagrams, so-called Pourbaix diagrams, which define three equilibrium thermodynamic domains for metals in aqueous solutions: immunity, passivity, and corrosion. Immunity is defined as the state where the base metal is stable while corrosion is defined as the formation of soluble metal cations and passivity the formation of a stable oxide film. Pourbaix’s derivation requires that the values of the standard chemical potential, m0, for all of the reacting substances are known for the standard state at the temperature and pressure of interest. For chemical reactions at room temperature, the equilibrium conditions are defined by the relationship18: P 0 nm ½9 logK ¼ 5708 and for electrochemical reactions at room temperature (Table 1) equilibrium is defined by: P 0 nm 0 ½10 E ¼ 96485n where K is the equilibrium constant for the reaction, m0 is in Joules per mole, v is the stoichiometric coefficient for the species, n is the number of electrons, 5708 is a conversion constant equal to RT/(log10e) where T is temperature (298.15 K) and R the ideal gas constant (8.314472 J (K mol)
1), and 96 485 is Faraday’s constant in J (mol V)
1. As an example of these diagrams, consider the iron–water system and the solid substances Fe, Fe3O4, and Fe2O3. Pourbaix18 defined the relevant equations for this system as:

½13

þ
Fe þ 2H2 O ¼ HFeO
2 þ 3H þ 2e

E 0 ¼ 0:493
0:089pH þ 0:0295log½HFeO
2

5.01.1.5 The Thermodynamics of Corrosion from Room Temperature to the PWR

½12

½14

þ
3HFeO
2 þ H ¼ Fe3 O4 þ 2H2 O þ 2e

E 0 ¼
1:819 þ 0:029pH
0:088log½HFeO
2

½15

reaction of two solid substances 3Fe þ 4H2 O ¼ Fe3 O4 þ 8Hþ þ 8e
E 0 ¼
0:085
0:059pH

½16

2Fe3 O4 þ H2 O ¼ 3Fe2 O3 þ Hþ þ 2e
E 0 ¼ 0:221
0:059pH

½17

An example of a simplified Pourbaix diagram for Fe at room temperature based on the reactions in eqns [11]–[17] is presented in Figure 5, where Eq. [12] corresponds to figure line 23, [13] to line 28, [14] to line 24, [15] to line 27, [16] to line 13 and [17] to line 17. Note that Eq. [11] is the boundary between Fe2þ and Fe3þ and was not drawn in the original figure. In addition to the lines separating the domains for Fe, Pourbaix diagrams will typically include the domains associated with water stability (oxidation and reduction) represented by the dashed lines marked a and b in Figure 5. Upon inspection of this diagram one would conclude what is know from experience with Fe: that iron is passive in alkaline solutions and at higher applied potentials owing to oxide film formation while at more acidic solutions Fe is susceptible to corrosion owing to Fe2þ. It is worth noting again that these potential–pH domains are defined solely by the thermodynamic stability of the species within them and these diagrams do not consider kinetics which will be addressed later in this chapter. This is important as while a species/reaction may be thermodynamically stable it may be kinetically hindered. While the use of Pourbaix diagrams to characterize room temperature corrosion reactions is

Corrosion and Compatibility

1.5

Fe3+

1.5

Fe3+ 20

7

20

1.0

1.0 b

Fe2O3

Fe2+

0

EH2 (200 C) (V)

28

2

EH (25 C) (V)

0.5

a 26

-0.5

17 Fe3O4

23

0.5

0

Fe2+

26

13

-1.0

HFeO-2

17

Fe3O4

23 27 24

Fe

Fe2O3 a

-0.5

13

-1.0

b

28

Fe

HFeO22-

-1.5

-1.5 0

5

10

15

pH

widespread, these diagrams and the method for generating them as presented thus far cannot be used at the higher temperatures associated with nuclear power reactors. This is due to the lack of standard potentials at elevated temperature as required by eqns [9] and [10] (e.g., the application of Table 1 to higher temperature). In the absence of these high-temperature thermodynamic data, Townsend19 used an extrapolation method introduced by Criss and Coble (the correspondence principle). The method allows for empirical entropy data of ionic species at 25  C to be extrapolated to higher temperatures. In this method, the standard Gibbs free energy is calculated from the relationship: ð


T

ðT 250

0

T

250

DC p ðT ÞdlnT

0

DC p ðT ÞdT ½18 0

where DS is the standard entropy change and DC p is the change in standard partial molar heat capacity. The potential–pH diagram for the Fe–H2O system and the solid substances Fe, Fe3O4, and Fe2O3 at 200  C calculated by Townsend is presented in Figure 6. In comparison with the diagram at 25  C (Figure 5), the Fe2O3 and Fe3O4 regions are extended to lower pH and potentials. As a result the area associated with corrosion at lower solution pH is decreased. However, 0

24

5

10

15

pH

Figure 5 A simplified potential–pH diagram for the Fe–H2O system and the solid substances Fe, Fe3O4, and Fe2O3 at 25  C based on the reactions in eqns [11]–[17]. Reprinted from Townsend, H. E. Corrosion Sci. 1970, 10, 343, with permission from Elsevier.

DðDG 0 Þ ¼
DT DS 0 ð250 Þ þ

0

29

27

Figure 6 The potential–pH diagram for the Fe–H2O system and the solid substances Fe, Fe3O4, and Fe2O3 at 200  C. Most dramatic influence of increased temperature is the presence of a large region of soluble species (corrosion) at high pH. Reprinted from Townsend, H. E. Corrosion Sci. 1970, 10, 343, with permission from Elsevier.

the most notable change in the diagram is at high pH where the area associated with corrosion owing to the soluble HFeO
2 has increased dramatically. The Criss and Coble method is limited, however, to the 150–200  C range and, to extend the Pourbaix to the temperatures of power reactors, Beverskog used a Helgeson–Kirkham–Flowers model to extend the heat capacity data to 300  C.20 Thus far, we have described a method for generating electrochemical equilibria diagrams and regions of passivity, corrosion, and immunity for pure metals from 25 to 300  C. From an engineering standpoint, we would like to know this information for structural alloys such as austenitic stainless steels and super nickel alloys. At temperatures near 25  C, the predominant oxide responsible for passivity is Cr2O3 and it is sufficient to rely only on the Cr potential–pH diagram for alloys with a high Cr content. However, at higher temperatures other oxides form such as Fe(Fe,Cr)2O4, (Cr,Fe)2O3, (Cr,Fe,Ni)3O4, and (Cr,Fe,Ni)2O3, and it is desirable to know the thermodynamic stability of the alloy. Beverskog has developed the ternary potential– pH diagrams for the Fe–Cr–Ni–H2O–H2 system for temperatures up to 300  C using heat capacitance data and the revised Helgeson–Kirkham–Flowers model described above.21 However, Fe–Cr–Ni phases lack thermodynamic data and the ternary oxides were,

8

Corrosion and Compatibility

npH 2

H2CrO4(aq) HCrO4-

Potential (VSHE)

1

CrO42Cr(OH)2+ 0 NiCr2O4(cr) Cr2+

FeCr2O4(cr)

-1

Cr2O3(cr) Cr(cr) -2

0

2

4

6 pH300 C

8

10

Figure 7 Potential–pH diagram for chromium species in Fe–Cr–Ni at 300  C. Concentration of aqueous species is 10
6 molal. Reprinted from Beverskog, B.; Puigdomenech, I. Corrosion 1999, 55, 1077, with permission from NACE International.

thus, not considered. The diagrams assumed that the metallic elements in the alloy had unit activity, that is, equal amounts of iron, chromium, and nickel. An example of the potential–pH diagram for chromium species in Fe–Cr–Ni at 300  C and aqueous species with a concentration of 10
6 molal is presented in Figure 7. Unlike the Fe diagram, where the presence of soluble HFeO2
2 species increased with temperature (Figure 6), the diagram for Cr in Fe–Cr–Ni is dominated by passive region where the stable oxides of Cr2O3, FeCr2O4, and NiCr2O4 are formed. 5.01.1.6 Kinetics of Dissolution and Passive Film Formation The study of dissolution kinetics, corrosion rate, attempts to answer the question: ‘‘What are the relationships that govern the flow of current across a corroding interface and how is this current flow related to applied potential?’’ Consider the anodic dissolution of a metal with an activation barrier equal to G1a ¼ nFE (eqn [2]). If we increase the driving force (potential) from its equilibrium condition, E0, to a new value, f, the new barrier is given by the relationship22:

decreasing the barrier, that is, not all of the applied potential is dropped across the electrochemical double layer. The rate (ra) of this reaction is expressed in the same, Arrhenius, form as for chemical reactions:   DGa ½20 ra ¼ ia =nF ¼ ka co exp
RT where ia is the anodic current density, ka is the rate constant, co is the concentration of oxidized species, and DGa is the change in free energy for the anodic reaction. Substituting G2a
G1a in eqn [20] for DGa in eqn [19], we express the anodic reaction rate as22:   ð1
bÞnF  ½21 ia ¼ nFka cR exp RT where  (the overpotential) represents a departure from equilibrium and is equal to f
E0. We can derive a similar expression for the cathodic reaction22:   bnF  ½22 ic ¼ nFkc co exp
RT where ic is the cathodic current density, kc is the rate constant, and co is concentration of oxidized species. Combining eqns [21] and [22] and rearranging them, we can write an expression for the total current, i:      ð1
bÞnF bnF 
exp
 ½23 i ¼ io exp RT RT where io is the exchange current density and is equal to
b
b nFkcc1
b o ka cR . This expression is commonly referred to as the Butler–Volmer equation. To apply eqn [23] to corrosion reactions, we need to be able to relate  to the corrosion potential, Ecorr , that is, as it stands the Butler–Volmer equation is derived for equilibrium conditions. Returning to our definition of the overpotential  ¼ f
E0, by both subtracting and adding Ecorr from the right side of this definition, inserting the resulting expression back into eqn [23] and rearranging we find22:   ð1
bÞF ðEcorr
Ea Þ icorr ¼ ia exp RT   bF ¼ ic exp
ðEcorr
Ec Þ ½25 RT

½19

For small applied potentials around Ecorr , the Stern– Geary approximation of eqn [24] is used23:   ba þ bc ðf
Ecorr Þ ½26 i ¼ 2:303icorr ba bc

where b is the symmetry factor and reflects the fact that not all of the increase in potential goes to

where ba and bc are defined as the anodic and cathodic Tafel slopes (discussed in Section

Ga2 ¼ Ga1
ð1
bÞnF ðf
E0 Þ

9

0.4

0.1

0.35

0 Current density (A m-2)

Potential (V vs. SCE)

Corrosion and Compatibility

0.3 0.25 0.2 Beam on at 100 nA ~540 s

0.15 0.1

0

500

1000 Time (s)

1500

2000

5.01.2.2) having units of volts and are empirical factors related to the symmetry factor by the relationships22: RT ð1
bÞnF

bc ¼
2:303

RT bnF

-0.1 -0.2 Increasing radiation flux Increasing cathodic reaction rate

-0.3 -0.4 -0.5

Figure 8 Influence of proton irradiation on the Ecorr of a SS 304L electrode in dilute sulfuric acid, pH ¼ 1.6. The increase is caused by the production of water radiolysis products.

ba ¼ 2:303

Beam = 35 na Beam = 62 na Beam = 100 na

½27 ½28

As it relates to the nuclear power industry, eqn [24] not only relates the corrosion rate (icorr) to the applied potential, f, but it can also help us to rationalize other processes such as the influence of water radiolysis products on corrosion rate. For example, it is generally observed that ionizing radiation (g, neutron, proton, etc.) increases Ecorr potential (Figure 8) and corrosion rate at Ecorr .24 The flux of ionizing radiation on the cooling water results in radiolysis, the breaking of chemical bonds. During the course of water radiolysis, a wide variety of intermediate products are formed, such as O2
, eaq, and the OH radical.25 The vast majority of these species have very fast reaction rates so that the end result is a handful of stable species. These stable products are typically oxidants, such as O2, H2, and H2O2. That is, these products readily consume electrons (eqns [II]–[IV]) and increase cathodic reaction rate (Figure 9). From eqn [25] we see that an increase in the cathodic reaction rate, ic, necessarily results in an increase in corrosion rate, icorr , consistent with the observation described. While the development of dissolution kinetics is straightforward, the kinetics associated with passive

-0.6 -0.2

-0.1

0 0.1 0.2 0.3 Potential (V vs. SCE)

0.4

0.5

Figure 9 Influence of proton irradiation on the cathodic reactions on a Au electrode in dilute sulfuric acid, pH ¼ 1.6. The increase is caused by the production of water radiolysis products.

film formation and breakdown are less well understood yet equally as important to our understanding of corrosion mechanisms. One such example is the case of localized corrosion where the probability for a pit to transition from a metastable to a stable state is governed by the ability of the active surface to repassivate. Another example is the initiation of SCC where passive film rupture results in very high dissolution rates and, correspondingly, crack advance rate which is controlled by the activation kinetics described above as the bare metal dissolves.26–28 During the propagation stage of SCC, the crack tip must propagate faster than (1) the oxide film can repassivate the surface and (2) the corrosion rate on the unstrained crack sides so that dissolution of the walls does not result in blunt notch. To evaluate the role of repassivation kinetics in SCC and corrosion mechanisms in general, investigators set about measuring three critical experimental parameters, namely film: ductility,29,30 bare surface dissolution rates,31–33 and passive film formation rates34,35 for various alloys. Each of these techniques involves the depassivation of a metal electrode using a tensile frame or a nano-indenter (in the case of ductility studies) or scratching/breaking an electrode (bares surface current density and repassivation studies) and measuring the resulting current transient as a function of time. An example of a current transient for SS 304L in chloride solution is presented in Figure 10. The surface was under potentiostatic control and was bared using a diamond scribe. The data were collected using a high-speed oscilloscope.

10

Corrosion and Compatibility

7.0  10−4 6.0  10−4

Current (A)

5.0  10−4

td

tr

0

0.002 0.004 0.006 0.008 Time (s)

4.0  10−4 3.0  10−4 2.0  10−4 1.0  10−4 0.0 0.01

0.012

Figure 10 Scratch test current transient from a SS 304L electrode in 0.1 M NaCl. The transient is characterized by a growth period, td, and a repassivation period, tr.

The transient is characterized by two separate processes, anodic dissolution and repassivation represented by td and tr in Figure 10. To analyze the repassivation rates, the period tr is typically fit to an expression and evaluated as a function of solution pH or electrode potential. The most prolific work in this field is probably on the alloy SS 304L. For this alloy, it has been proposed that the kinetics of film growth are controlled by ion migration under high electric field.36–38 The kinetics of high-field film growth were first proposed by Cabrera and Mott39 to obey the kinetic relationship:   BV ½29 i ¼ Aexp L where i is the current density, V is the voltage, L is the oxide thickness, and A and B are constants.

5.01.2 Analytical Methods 5.01.2.1

Introduction

In this section, we will review the principle analytical methods used to probe the electrochemical interface. In the Section 5.01.2.2 Potentiodynamic Polarization, we discuss linear polarization resistance and the practical application of corrosion kinetics, eqns [25] and [26]. In that section, we also describe the salient points of the anodic polarization curve. In the Section 5.01.2.3 Electrochemical Impedance Spectroscopy, we introduce an ac method for interrogating the electrochemical interface. This technique is

probably the most versatile experimental method available to scientists and researchers. As it relates to nuclear reactors, this technique has the ability to subtract out the contribution of the solution resistance to polarization resistance measurements which, if not accounted for in highly resistive cooling water measurements will result in nonconservative corrosion rates. In the final section, we introduce a more seldom used technique, Mott–Schottky analysis. While this is by no means a common experimental method, it provides a conduit for the reader to become familiar with defects in the oxide film, their transport and ways to quantify it. This has particular interest here as ionizing radiation may promote corrosion rates by increasing transport of these defects through the passive oxide film. Regretfully, the scope of this chapter is limited, and we are not able to discuss the step-by-step details of the experimental methods that are used to make corrosion measurements. A comprehensive guide to experimental methods in corrosion has been published by Kelly et al.40 as well as Marcus and Mansfeld41 while a more broad description of electrochemical methods has been published by Bard and Faulkner.42 The reader is also encouraged to become familiar with the equipment that is used to make electrochemical measurements and a good introductory chapter on this topic has been presented by Schiller.43 The most important instrument is, no doubt, the potentiostat. While this instrument is the cornerstone of corrosion science, it does have its pitfalls including bandwidth limitations and the potential for ground loop circuits when used in conjunction with other equipment such as load frames, autoclaves and cooling loops. The latter can be overcome using proper instrumentation such as potentiostat with a floating ground, or isolation amplifiers. To investigate the influence of the neutron flux on corrosion rates and mechanisms, real-time in-situ corrosion measurements are often made ‘in-reactor’ or at neutron facilities such as Oak Ridge National Lab’s Hi-Flux Isotope Reactor (HIFER) and Argonne National Lab’s Fast Flux Test Facility (FFTF). Alternately, neutron damage can be simulated using ion beams. As it relates to ion irradiation, this method provides opportunities for studying the interaction of the components of reactor environments (radiation, stress, temperature, aggressive media) that are not possible with in-reactor or neutron irradiation facilities. For a full discussion of this topic, see Chapter 1.07, Radiation Damage Using Ion Beams. To summarize these experiments, controlled

11

Corrosion and Compatibility

environmental cells are coupled to accelerator beamlines to study the interaction of the environment and irradiation on structural materials. Corrosion at the substrate–environment interface is studied in real time by numerous electrochemical techniques including those described below. With respect to dose, light ions can be used to reach doses up to 10 dpa in several days. However, the depth of penetration is low (tens or micrometers), which puts unrealistic limitations on electrochemical cell construction. Heavy ion irradiation can reach several hundred displacements per atom in a matter of days but the penetration depth is much less. 5.01.2.2

Potentiodynamic Polarization

During our development of Butler–Volmer reaction kinetics, we introduced the Stern–Geary approximation and defined Tafel slopes within the context of the symmetry factor without much further explanation. To understand the empirical source of the Tafel slopes, we rearrange eqn [26] to define the polarization resistance Rp in units of O cm
2: Rp ¼

ba bc DE ¼ 2:303icorr ðba þ bc Þ Di

½30

The Tafel slopes can be obtained from a plot of potential as a function of the logarithm of current density as shown for the cathodic curve in Figure 11 where bc has units of volts. A similar anodic plot can be generated to obtain ba. It is important to realize that these slopes are frequently not equal as they are related to separate mechanisms on the electrode surface. From a plot of both the anodic and cathodic Tafel slopes, we can also obtain icorr (Figure 11). With knowledge of icorr , ba, and bc, the polarization iOH2 EH

+

+

H

2

20

10 Slope =

Applied current curve

Ecorr +

M

20

20

iapp(cathodic)

icorr

E iOM

h(mv)

40 H+

/H2

resistance can then be determined from eqn [30]. For small voltage perturbations, the slope of a plot of applied potential (DE ) versus the change in current density (Di ) is equal to Rp (Figure 12), often referred to as the linear polarization resistance as the slope is only linear for small voltage perturbations around Ecorr . Errors not accounted for in eqn [30] include uncompensated solution resistance (RO) and choosing a scan rate that is too fast. From a reactor standpoint, the value of RO in the cooling water or a simulant where the resistivity is high may be a significant contribution to the measured Rp and, therefore, result in a nonconservative corrosion rate, that is, the calculated corrosion rate will be too low. For a complete standardized method of conducting potentiodynamic polarization resistance measurements and data analysis, the reader is referred to the appropriate ASTM standards G3, G59, and G102.44–46 Potentiodynamic polarization curves can also be used to assess corrosion rate as well as determine if a material is passive, active, or susceptible to pitting corrosion in a given environment. Consider the curve in Figure 13. It plots the applied potential as a function of the log of the absolute value of the current density. The corrosion potential and corrosion current density are shown at the intersection of the cathodic and anodic curves. Mass loss (m, in grams) for a given period of exposure can then be

DE Diapp

40 iapp(anodic)

-10

Tafel region

M

- EM/M+ log iapplied

Figure 11 Diagram depicting the cathodic polarization of an electrode near Ecorr. Tafel extrapolation showing the determination of icorr is also presented. Reprinted from Fontana, M. G. Corrosion Engineering; McGraw Hill: New York, 1986, with permission from McGraw Hill.

-20 h(mv) Figure 12 Diagram depicting the linear polarization for an electrode about Ecorr. Slope, DE/Di, is equal to the polarization resistance. Reprinted from Fontana, M. G. Corrosion Engineering; McGraw Hill: New York, 1986, with permission from McGraw Hill.

12

Corrosion and Compatibility

positive hysteresis in the reverse portion of the curve that is not observed in the case of solution oxidation or transpassivity.

Epit Potential

Erepass

5.01.2.3 Electrochemical Impedance Spectroscopy Eflade

ba ipass

icorr

bc

Ecorr

Log current density Figure 13 Diagram depicting the potentiodynamic polarization of an electrode far from Ecorr. Relevant potentials and currents are defined in the text.

determined from the corresponding icorr using Faraday’s Law, for a pure metal40: icorr t EW ½31 FA where t is time in seconds, icorr has units of A (cm
2), A is surface area, and EW is equivalent weight and is equal to molecular weight (M) divided by the number of electrons in the reaction. Similarly, for an alloy: m¼

EW ¼ P

1 n i i fi =Mi

½32

where the subscript i denotes the alloying element of interest and f is the mass fraction of that element in the alloy. From Faraday’s law, it is also possible to calculate corrosion rate, the penetration depth owing to corrosion over a period of time in units of mm year
1 (CR): CRmm year
1 ¼

3:27  103 icorr EW r

½33

where r is the material density in g cm
3. Other critical parameters in Figure 13 include the Flade potential (EFlade) which marks the critical potential necessary for passive film formation, the passive current density (ipass), the pitting potential (Epit), and the repassivation potential (Erepass). With respect to EFlade, this active to passive transition is not observed for materials that are spontaneously passive in a given environment such as SS in a BWR. In such a case, the current would be limited by the film dissolution current (ipass). As it relates to Epit, the onset of localized corrosion is characterized by a sharp increase in current at a given potential. As such, materials that are more susceptible to pitting have lower Epit values. Pitting is also characterized by a

While potentiodynamic polarization is typically considered a destructive technique, that is, it alters the surface of the corrosion sample, electrochemical impedance spectroscopy (EIS) is a powerful nondestructive technique for obtaining a wealth of data including Rp.47 Further, this technique has the ability to subtract out the uncompensated solution resistance (RO) from the measurement which, for highly resistive reactor cooling water environments, is a significant advantage. In EIS, a small sinusoidal voltage perturbation (10 mV) is applied across the electrode interface over a broad frequency range (mHz to MHz). By measuring the transfer function of the applied voltage to the system current, the system impedance may be obtained. For corrosion systems, the impedance (Z) is a complex number and may be represented in Cartesian coordinates by the relationship: ZðoÞ ¼ Z0 þ Z00

½34

where o is the applied frequency in radians, Z0 is the real part of the impedance, and Z00 is the imaginary part of the impedance, and the magnitude of the impedance jZj is given by: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ½35 jZj ¼ ðZ0 Þ2 þ ðZ00 Þ2 In its simplest form, the electrochemical interface can be thought of as an electrical equivalent circuit (EC): a resistor (R) with an impedance Z(o) ¼ R and a capacitor (C) with Z(o) ¼ 1/joC where j is the square root of
1.48 Thus, the impedance of a resistor is purely real and independent of frequency while the impedance of a capacitor is purely imaginary and inversely proportional to frequency. With respect to the electrochemical interface, the polarization resistance is in parallel with the double layer capacitance (Cdl, owing to adsorption of charged anions/cations in the electrolyte). These two components act in series with the solution resistance, RO as seen in the EC in Figure 14. This circuit is referred to as a simple Randles circuit and represents an ideal interface. Commonly, however, Cdl does not behave as an ideal capacitor and its impedance is better represented by the expression

13

Corrosion and Compatibility

-6  104

Rp RW

-5  104 -4  104

Figure 14 Simplified Randles equivalent circuit of an electrochemical interface where Rp is the polarization resistance, Cdl is the double layer capacitance, and RO is the geometric resistance associated with the solution resistance.

Z (W)

Cd1

-3  104

-2  104 -1  104

0

105

0

1  104 2  104 3  104 4  104 5  104 6  104 Z (W)

Rp + RW

Figure 16 Nyquist format for data in Figure 15 where Rp ¼ 5  104 O, Cdl ¼ 4  10
6, and RO ¼ 200 O.

104 |Z| (W)

ZðoÞ ¼ 1=Cð j oÞa 103

RW 102 10-3 10-2 10-1 100 101 102 (a) Frequency (Hz)

103

104

105

0 -10

Q ()

-30 -40 -50

-70

wmax

-80 -90 10-3 10-2 10-1 100 101 102 (b) Frequency (Hz)

where a is typically found to be between 0.5 and 1. The element that represents this behavior is known as a constant phase element (Ccpe).49 The response of a simple Randle’s circuit as a function of frequency is shown in Figure 15(a) and 15(b). These plots are referred to as the Bode magnitude plot (eqn [35]) and the Bode phase plot50 where the phase angle, y, is equal to:  00  Z ½37 y ¼ tan
1 Z0 This phase angle is a result of the double layer capacitance where the current leads the applied ac voltage perturbation as is the case for pure capacitors. The parameters Rp and RO may be determined graphically from the Bode magnitude plot as shown in Figure 15(a) while Cdl is determined from the graphical parameter omax (converted to radians, 2pf) in Figure 15(b) and the relationship40: 1 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 þ Rp =RO ½38 Cdl ¼ omax Rp

-20

-60

½36

103

104

105

Figure 15 (a) Bode magnitude and (b) Bode phase plots. Data were generated from an electrical equivalent of a simplified Randles circuit (Figure 13) where Rp ¼ 5  104 O, Cdl ¼ 4  10
6, and RO ¼ 200 O.

Alternately, the data may be presented using the Nyquist format which plots the imaginary impedance as a function of the real impedance as seen in Figure 16 (sometimes referred to as a Cole–Cole plot). As graphical analysis is somewhat imprecise, commercially available complex nonlinear least squares fitting of the data (CNLS) is commonly used to obtain these parameters (Figure 17).

14

Corrosion and Compatibility

|Z| exper. |Z| fit

102

ROX

0

RW

-10

|Z| (W m2)

-30

Q exper. Q fit

-40 -50

100

Q (degrees)

-20 101

-60

COX Figure 19 Equivalent circuit of an oxide-covered metal in liquid PbBi eutectic (LBE) where Rox is the resistance of the passive film, Cox is the double layer capacitance of the oxide, and RO is the geometric resistance associated with the LBE.

-70 10-1 10-3

10-2

10-1

100

101

102

-80 103

Frequency (Hz) Figure 17 Bode magnitude and Bode phase plots for SS 304L during proton irradiation in a pressurized deionized cooling water loop at 125  C showing experimental data and complex nonlinear least squares fit of equivalent circuit in Figure 14.

Corrosion rate (mm year −1)

6.0

5.0

4.0

3.0 Photons Neutrons 2.0

1.0

Protons

0

20

100 120 40 60 80 Flux (particles m–2 per proton)

140

Figure 18 Corrosion rate as a function of particle flux for a SS 304L electrode during proton irradiation in a pressurized deionized cooling water loop at 125  C.

EIS has been used successfully to investigate the passive films on Zr alloys,51–53 SS 304L,54 and nickel base alloys55 in environments that simulate reactor cooling water systems. In addition, it has also been used to measure the real-time corrosion rates of materials during irradiation. Lillard et al.56,57 measured the corrosion rate of materials as a function of immersion time in a deionized water cooling loop during proton irradiation. In that work, radiationhardened probes made from Alloy 718, SS 304L, and Al 60 601 were exposed to a proton beam at

various current densities. The impingement of the beam on the probes resulted in a mixed neutron, photon, and proton flux. It was shown that corrosion rate was almost linear with photon and neutron flux as compared to proton flux where anomalies existed at intermediate fluxes (Figure 18). The data were ultimately used to extrapolate lifetimes for accelerator materials. In other studies, EIS has also been applied to investigate passive films on metals in sodium (Na)cooled and lead–bismuth (LBE) systems that simulate reactor environments. The equivalent circuit (EC) used to model the data is similar to the simplified Randles circuit; however, in this case, there is no electrochemical double layer, only the impedance and capacitance associated with the oxide (Figure 19). In this EC, Rox is the dc resistance of the oxide, Cox is the oxide capacitance, and RO is the geometric resistance associated with the liquid metal (Na or LBE). In one such study, Isaacs investigated the capacitance of anodized films on Zr in liquid Na. In that study, it was shown that both Rox and Cox were a function of Na temperature between 50 and 400  C. Lillard et al.58 reported similar trends for HT-9 in LBE. At constant temperature, Rox and Cox for HT-9 in LBE were a function of immersion time (Figure 20). The data were used to calculate oxide thickness as a function of time. In addition, Rox was related to ionic transport through the film and corrosion rates were calculated using Wagner’s oxidation theory. Upon irradiation in a proton beam, this rate fell even further.59 Additional information about lead
eutectic coolants may be found in Chapter 5.09, Material Performance in Lead and Lead-bismuth Alloy. 5.01.2.4

Mott–Schottky Analysis

The formation and growth of passive oxide films is driven, fundamentally, by interfacial reactions and defect, electron and ion transport processes. Yet the

Corrosion and Compatibility

105

ROX

15

5.5  1010

101

5.0  1010

104

Prior to irradiation

Fit

Beam on

Fit

10–1

101

1 C-2 (cm4 F-2)

ROX (W cm−2)

Rp = 0.9 ´ t1.7 102

COX (nF cm–2)

100 103

4.5  1010 4.0  1010 3.5  1010

10–2 COX

100 10–1 –50

0

50 100 Immersion time (h)

150

3.0  1010 10–3 200

2.5  1010 -0.2

0

0.2

0.4

0.6

0.8

1

Potential (V vs. SCE)

Figure 20 Impedance capacitance for the oxide on an HT-9 steel as a function of immersion in liquid PbBi eutectic. Reprinted from Lillard, R. S.; Valot, C.; Hanrahan, R. J. Corrosion 2004, 60, 1134, with permission from the author.

Figure 21 Mott–Schottky plots for a SS 304L electrode in dilute sulfuric acid, pH ¼ 1.6. Before and after proton irradiation. Reprinted from Lillard, R. S.; Vasquez, G. J. Electrochem. Soc. 2008, 155, C162, with permission from the author.

nature and relative importance of these processes are still far from being understood. Key to oxide growth, and, therefore, passivation, is the mobility of these defects, specifically vacancies. Under irradiation, however, the defect properties of the oxide are undoubtedly changed, and the extent of corrosion is related both to the microstructure and transport properties of defects. One way to probe the transport properties of the oxide film is by using Mott–Schottky theory. Returning to our discussion of EIS above, the oxide capacitance may be obtained at high frequency from the relationship

depends on what type of semiconductor the oxide is (p vs. n) and this effects the sign of the slope. Lillard used Mott–Schottky analysis to examine the influence of proton irradiation on defect generation and transport in the oxide film on SS 304L.61 The passive film on SS 304L is an n-type semiconducting oxide and the major defect is the oxygen vacancy which acts as an electron donor. According to the Point Defect Model62 for oxide growth, oxygen vacancies are produced at the metal–film interface by the injection of a metal atom into the oxide lattice:

Z00 ðoÞ ¼ 1=oC or C ¼ 1=oZ00 ðolim Þ

½39

By measuring Z00 as a function of applied dc voltage at the high frequency limit (olim) and calculating the film capacitance from eqn [39], it is possible to evaluate donor concentration (ND) in the oxide film from the well-known Mott–Schottky relationship:   1
kT ½40 ¼ 2=ee N U
U 0 D fb C2 q where e0 is the permittivity of space and is equal to 8.854  10
14 F cm
1, e is the electronic charge and is equal to 1.602  10
19 C, U is the applied potential in V, Ufb is the flatband potential in V, kT/q is equal to 25 mV at 25  C, and ND is the donor concentration (oxygen vacancies) in cm
3.60 ND is the slope of a plot of 1/C2 versus applied potential. The type of defect

m ! MM þ ðx=2ÞVo þ xe


½41

where m is a metal atom, MM is a metal cation, Vo is a oxygen vacancy, e
is an electron, and x is the stoichiometric coefficient. In pH 1.6 H2SO4, it was found that the cation vacancy concentration in the oxide increased from 2.94  1021 in the absence of irradiation to 3.41  1021 during irradiation (Figure 21). It was proposed that the film on SS 304L is composed of an inner Cr-rich p-type semiconductor and an outer Fe-rich n-type semiconductor. This bilayer film results in a p–n junction where the two layers meet. On the p side of the junction, there is a surplus of holes, while on the n side of the junction, a surplus of electrons exists. The energy bands are such that it is ‘uphill’ (an increase in energy) for electrons moving across the junction (from the n side to the p side). On a related topic, a discussion of defects in bulk oxides can be found in Chapter 1.02, Fundamental Point Defect Properties in Ceramics.

16

Corrosion and Compatibility

References 1. 2. 3. 4. 5.

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

28. 29. 30. 31. 32.

Fontana, M. G. Corrosion Engineering; McGraw Hill: New York, 1986. Cox, B. J. Nucl. Mater. 2005, 336, 331. Lysel, G.; Nystrand, A. C.; Ullberg, M. J. ASTM Int. 2005, 2, 355. Abella, J.; Balachov, I.; Macdonald, D. D.; Millet, P. J. Corrosion Sci. 2002, 44, 191. Scott, P. M.; Meunier, M. C.; Deydier, D.; Silvestre, S.; Trenty, A. In American Society for Testing and Materials Special Technical Publication; Kane, R. D., Ed.; ASTM: Conshohocken, PA, 2000; p 210. Hur, D. H.; Choi, M. S.; Lee, D. H.; Song, M. H.; Han, J. H. Corrosion 2006, 62, 905. Hwang, S. S.; Kim, H. P.; Kim, J. S. Corrosion 2003, 59, 821. Helmholtz, H. Pogg. Ann. 1853, LXXXIX, 211. Whitten, K. W.; Gailey, K. D. General Chemistry; Saunders College: Philadelphia, PA, 1981. Castellan, G. W. Physical Chemistry; Addison-Wesley: Reading, MA, 1983. Danielson, M. J. Corrosion 1995, 51, 450. Macdonald, D. D. Corrosion 1978, 34, 75. Oijerholm, J.; Forsberg, S.; Hermansson, H. P.; Ullberg, M. J. Electrochem. Soc. 2009, 156, P56. Indig, M. E. In 12th International Corrosion Congress, Corrosion Control for Low-Cost Reliability; NACE International: Houston, TX, 1993; p 4224. Indig, M. E.; Nelson, J. L. Corrosion 1991, 47, 202. Oh, S. H.; Bahn, C. B.; Hwang, I. S. J. Electrochem. Soc. 2003, 150, E321. Lvov, S. N.; Macdonald, D. D. J. Electroanal. Chem. 1995, 403, 25. Pourbaix, M. Atlas of Electrochemical Equilibria in Aqueous Solutions; NACE/Cebelcor: Houston, TX, 1974. Townsend, H. E. Corrosion Sci. 1970, 10, 343. Beverskog, B.; Puigdomenech, I. Corrosion Sci. 1996, 38, 2121. Beverskog, B.; Puigdomenech, I. Corrosion 1999, 55, 1077. Prentice, G. Electrochemical Engineering Principles; Prentice Hall International Series; Prentice-Hall: Upper Sadddle River, NJ, 1991. Stern, M.; Geary, A. L. J. Electrochem. Soc. 1957, 104, 56. Marsh, G. P.; Taylor, K. J.; Bryan, G.; Worthington, S. E. Corrosion Sci. 1986, 26. Lewis, M. B.; Hunn, J. D. J. Nucl. Mater. 1999, 265, 325. Hoar, T. P.; Hines, J. G. J. Iron Inst. 1956, 182, 124. Diegle, R. B.; Boyd, W. K. In Stress Corrosion Cracking – The Slow Strain-Rate Technique; Ugiansky, G. M., Payer, J. H., Eds.; ASTM: West Conshohocken, PA, 1979; p 26. Ford, F. P. Corrosion 1996, 52, 375. Bubar, S. F.; Vermilyea, D. A. J. Electrochem. Soc. 1966, 113, 892. Bubar, S. F.; Vermilyea, D. A. J. Electrochem. Soc. 1967, 114, 882. Hoar, T. P.; Scully, J. C. J. Electrochem. Soc. 1964, 111, 348. Hoar, T. P.; West, J. M. Nature 1958, 181, 835.

33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

57. 58. 59. 60. 61. 62.

Frankel, G. S.; Jahnes, C. V.; Brusic, V. V.; Davenport, A. J. J. Electrochem. Soc. 1995, 142, 2290. Ford, F. P.; Burstein, G. T.; Hoar, T. P. J. Electrochem. Soc. 1980, 127, 1325. Marshal, P.; Burstein, G. T. Corrosion Sci. 1983, 23, 1219. Burstein, G. T.; Davenport, A. J. J. Electrochem. Soc. 1989, 136, 936. Burstein, G. T.; Marshall, P. I. Corosion Sci. 1984, 24, 449. Davenport, A. J.; Burstein, G. T. J. Electrochem. Soc. 1990, 137, 1496. Cabrera, N.; Mott, N. F. Rep. Prog. Phys. 1948, 12, 163. Kelly, R. G.; Scully, J. R.; Shoesmith, D. W.; Buchheit, R. G. Electrochemical Techniques in Corrosion Science and Engineering; Marcel Dekker: New York, 2003. Marcus, P.; Mansfeld, F. Analytical Methods in Corrosion Science and Engineering; Taylor & Francis: New York, 2006. Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundementals and Applications; Wiley: New York, 2001. Schiller, C. A. In Analytical Methods in Corrosion Science and Engineering; Marcus, P., Mansfeld, F., Eds.; Taylor & Francis: New York, 2006; p 361. ASTM, G 3. Standard practice for conventions applicable to electrochemical measurements in corrosion testing; ASTM International; 1989. ASTM, G 102. Standard practice for calculation of corrosion rates and related information from electrochemical measurements; ASTM International; 1989. ASTM, G 59. Standard test method for conducting potentiodynamic polarization resistance measurements; ASTM International; 1997. MacDonald, J. R. Impedance Spectrocopy; Wiley: New York, 1987. Mansfeld, F. In Analytical Methods in Corrosion Science and Engineering; Marcus, P., Mansfeld, F., Eds.; Taylor & Francis: Boca Raton, FL, 2006; p 463. Hsu, C. H.; Mansfeld, F. Corrosion 2001, 57, 747. Mansfeld, F. Corrosion 1988, 44, 558. Ai, J.; Yingzi, C.; Uriquidi-Macdonald, M.; Macdonald, D. D. J. Nucl. Mater. 2008, 379, 162. Barberis, P.; Frichet, A. J. Nucl. Mater. 1999, 273, 182. Nagy, G.; Kerner, Z.; Battistig, G.; Csordas, A. P.; Balogh, J. J. Nucl. Mater. 2001, 297, 62. Kim, Y. J. Corrosion 2000, 56, 389. Fulger, M.; Ohai, D.; Mihalache, M.; Pantiru, M.; Malinovschi, V. J. Nucl. Mater. 2009, 385, 288. Lillard, R. S.; Gac, F.; Paciotti, M.; et al. In Effects Radiation on Materials; American Society for Testing and Materials Special Technical Publication #1045; Rosinski, S. T., Grossbeck, M. L., Allen, T. R., Kumar, A. S., Eds.; ASTM: West Conshohocken, PA, 2001. Lillard, R. S.; Willcutt, G. J.; Pile, D. L.; Butt, D. P. J. Nucl. Mater. 2000, 278, 277. Lillard, R. S.; Valot, C.; Hanrahan, R. J. Corrosion 2004, 60, 1134. Lillard, R. S.; Paciotti, M.; Tcharnotskaia, V. J. Nucl. Mater. 2004, 335, 487. Sikora, E.; Sikora, J.; Macdonald, D. D. Electrochim. Acta 1996, 41, 283. Lillard, R. S.; Vasquez, G. J. Electrochem. Soc. 2008, 155, C162. Macdonald, D. D. J. Electrochem. Soc. 1992, 139, 3434.

5.02

Water Chemistry Control in LWRs

C. J. Wood Electric Power Research Institute, Palo Alto, CA, USA

ß 2012 Elsevier Ltd. All rights reserved.

5.02.1

Introduction

18

5.02.2 5.02.2.1 5.02.2.2 5.02.2.3 5.02.2.4 5.02.2.5 5.02.3 5.02.3.1 5.02.3.2 5.02.3.3 5.02.3.4 5.02.4 5.02.4.1 5.02.4.2 5.02.4.3 5.02.4.4 5.02.5 5.02.6 References

BWR Chemistry Control Evolution of BWR Chemistry Strategies Mitigating Effects of Water Chemistry on Degradation of Reactor Materials Radiation Field Control Fuel Performance Issues Online Addition of Noble Metals PWR Primary Water Chemistry Control Evolution of PWR Primary Chemistry Strategies Materials Degradation PWR Radiation Field Control Fuel Performance PWR Secondary System Water Chemistry Experience Evolution of PWR Secondary Chemistry Strategies Chemistry Effects on Materials Degradation of SGs Control of Sludge Fouling of SGs Lead Chemistry Chemistry Control for FAC in BWRs and PWRs Water Chemistry Control Strategies

19 19 20 23 26 27 27 27 29 33 35 37 37 40 43 44 45 45 46

Abbreviations AO

AVT

BOP BRAC

BWR CGR CRUD DMA DZO EBA ECP

Axial offset, referring to localized flux depression in reactor core caused by buildup of boroncontaining deposits. Originally called AOA for axial offset anomaly. All-volatile treatment, suing ammonia for pH control in steam generators Balance of plant BWR radiation and control, referring to designated standard points in BWR reactors for radiation field measurements Boiling water reactor Crack growth rate Corrosion product deposits on fuel element surfaces Dimethylamine Depleted zinc oxide (BWRs) Enriched boric acid (PWRs) Electrochemical corrosion potential

ETA FAC GE

HWC HWC-L HWC-M IGA IGSCC LWR MOX MPA MRC NDE NMCA NWC OD IGA/SCC

Ethanolamine Flow-assisted corrosion General electric, the vendor for BWRs in the United States and some other countries Hydrogen water chemistry HWC (low) with 0.2–0.5 ppm hydrogen HWC (moderate) with 1.6–2.0 ppm hydrogen Intergranular attack Intergranular stress corrosion cracking Light water reactor Mixed oxide fuel 3-methoxypropylamine Molar ratio control (PWR secondary side) Nondestructive examination Noble metal chemical addition Normal water chemistry (BWRs) Outside diameter IGA/SCC in steam generator tubes

17

18

Water Chemistry Control in LWRs

OLNC OTSG PAA PbSCC PWR PWSCC SCC SG SHE

On-line noble chemistry Once through steam generator Poly acrylic acid Lead assisted stress corrosion cracking Pressurized water reactor Primary water stress corrosion cracking Stress corrosion cracking Steam generator Standard hydrogen electrode (for ECP measurements)

5.02.1 Introduction Other chapters of this comprehensive describe the various degradation processes affecting the structural materials used in the construction of nuclear power plants (see Chapter 5.04, Corrosion and Stress Corrosion Cracking of Ni-Base Alloys; Chapter 5.05, Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels; and Chapter 5.06, Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels). This chapter describes the influence of water chemistry on corrosion of the most important materials in light water reactors (LWRs). In particular, alloys susceptible to intergranular attack (IGA) and stress corrosion cracking (SCC) are significantly impacted by water chemistry, most notably, sensitized 304 stainless steel in boiling water reactors (BWRs) and nickelbased alloys in pressurized water reactors (PWRs). Excellent water quality is essential if material degradation is to be controlled. In the early days of nuclear power plant operation, impurities in the coolant water were a major factor in causing excessive corrosion. Chlorides and sulfates are particularly aggressive in increasing intergranular stress corrosion cracking (IGSCC) and other corrosion processes. Transient increases of impurities in the coolant that occur during fault conditions (e.g., condenser leaks and ingress of oil or ion exchange resins) proved to be particularly damaging. Thus, water chemistry was traditionally regarded as a key cause of material degradation. Initial efforts to improve water quality brought about a slow but steady reduction in impurities through improved design and operation of purification systems. Not only were the average concentrations of impurities reduced over time, but the frequency and magnitude of impurity ‘spikes’ from transient fault conditions were also diminished.

However, excellent water chemistry alone was not sufficient to control corrosion. Hence, programs to modify water chemistry were introduced, including minimizing oxygen to reduce the electrochemical corrosion potential (ECP) in BWRs, and oxygen and pH control in PWRs. More recently, additives to further inhibit the corrosion process have been developed and are now in widespread use. As a result, water chemistry advances are now an important part of the overall operating strategy to control material degradation. Primary system water chemistry also affects fuel performance through the deposition of corrosion products on fuel pin surfaces, and influences radiation fields outside the core. Core uprating through increased fuel duty has reduced margins for tolerating corrosion products (CRUD) on BWR fuel pin surfaces. In PWRs, increasing fuel cycle duration has increased the challenge of controlling pH within the optimum range. At the same time, regulatory limits on worker radiation exposure are tending to be tightened worldwide, putting pressure on the operators to reduce radiation dose rates. Successful operation of PWR steam generators (SGs) and the remainder of the secondary system demand strict water chemistry control in secondary side systems if corrosion problems are to be avoided. Other operating parameters also influence the optimization process, for example, life extension (to 60 years) has emphasized the importance of controlling degradation of circuit materials. Therefore, although control of structural material degradation remains the highest priority, water chemistry must be optimized between the sometimes-conflicting requirements affecting other parts of the reactor. Advances in water chemistry have enabled plant operators to respond successfully to these technical challenges, and the overall performance has steadily improved in recent years.1 Plant-specific considerations sometimes influence or indeed limit the options for controlling water chemistry, so we see different chemistry specifications at different plants. This is especially true internationally and significant differences between countries are noted. The US industry started developing water chemistry guidelines 25–30 years ago, and these now provide the technical basis for guidelines in many other countries. The early editions of these guidelines presented impurity specifications and required action if limits were exceeded. When advanced water chemistries were developed and qualified, the guidelines evolved into a menu of options within an envelope of specifications that should not be

Water Chemistry Control in LWRs

exceeded. Guidance is now provided on how to select a plant-specific water chemistry strategy.2 The basis for water chemistry control was discussed in detail by Cohen.3 The remainder of this chapter describes more recent water chemistry developments for BWRs, PWR primary systems, and PWR secondary systems including SGs, with a short section on flow-assisted corrosion (FAC) in both BWRs and PWRs.

5.02.2 BWR Chemistry Control 5.02.2.1 Evolution of BWR Chemistry Strategies BWR water chemistry has to be optimized between the requirements to minimize material degradation, avoid fuel performance issues, and control radiation fields. These factors are depicted in Figure 1,4 which also includes the main chemistry changes involved in the optimization process. Plant-specific considerations sometimes influence or indeed limit the options for controlling water chemistry, so we see different chemistry specifications at different plants. This is especially true internationally and significant differences in chemistry strategies between countries are noted. Design features are an important reason for these different chemistry regimes, to which must be added the effects of different operational strategies in recent years. For example, a key issue facing BWRs in the United States concerns IGSCC of reactor internals, as discussed in other chapters. The occurrence of IGSCC resulted in the

Clad corrosion crud deposition: Limits on feedwater zinc

Impurity control: Monitoring/analysis required

implementation of hydrogen water chemistry, with or without noble metal chemical addition (NMCA), to ensure that extended plant lifetimes are achieved. German plants use 347 stainless steel, which is less susceptible to IGSCC than sensitized 304 stainless steel used originally in US-designed plants. Some Swedish and Japanese plants have replaced 304 stainless steel reactor internals with 316 nuclear grade material to minimize potential problems, as this material is less susceptible to IGSCC. As a result, many of these plants continue to use oxygenated normal water chemistry, whereas all US plants control IGSCC through the use of hydrogen water chemistry (HWC) with or without normal metal chemical addition to improve the efficiency of the hydrogen in reducing ECP. Second, BWRs in United States undoubtedly have greater cobalt sources than plants in most other countries, despite strong efforts to replace cobalt sources. This resulted in higher out-of-core radiation fields, leading all US plants to implement zinc injection to control fields, whereas only a small number of plants of other designs use zinc. Third, the move to longer fuel cycles and increased fuel duty at US plants, while having major economic benefits, has led to new constraints on chemistry specifications in order to avoid fuel performance issues. Figure 2 depicts the changing chemistry strategies over the past 30 years, showing the focus on improving water quality in the early 1980s and the move to educing chemistry to control IGSCC, which in turn resulted in increased radiation fields, subsequently controlled by zinc injection.

Materials degradation and mitigation

Water chemistry guidelines

Fuel performance

Chemistry control issues

Figure 1 Boiling water reactor chemistry interactions.

19

BWR internals IGSCC, IASCC: HWC or NMC required

Radiation exposure

Radiation fields crud bursts: Zinc required

20

Water Chemistry Control in LWRs

Increasing concerns about core internals cracking led to the need to increase hydrogen injection rates, which in turn resulted in the introduction of NMCA to reduce operating radiation fields from N-16. Figure 3 shows the rate of implementation of HWC, zinc and NMCA, and online noble metal addition (OLNC). The rationale and implications of these developments are discussed in greater detail in subsequent sections. The goal for BWRs is therefore to specify chemistry regimes that, together with the improved materials used in replacement components (e.g., 316 nuclear grade stainless steel), will ensure that the full extended life of the plants will be achieved without the need for further major replacements. At the same time, radiation dose rates, and hence worker radiation exposure, must be closely controlled, and fuel performance must not be adversely affected by chemistry changes.

The first requirement of plant chemistry is to maintain high-purity water in all coolant systems, including the need to avoid impurity transients, which are beyond the scope of this paper. The performance of all plants has improved steadily over the years, as shown by the trend for reactor water conductivity for GE-designed plants, given in Figure 4. This figure shows that conductivity now approaches the theoretical minimum for pure water. In fact, deliberately added chemicals, such as zinc (discussed in the following section), account for much of the difference between measured values and the theoretical minimum. The conductivity data are consistent with the reactor water concentrations for sulfate and chloride. In fact, sulfate is the most aggressive impurity from the viewpoint of IGSCC, and much effort has gone into reducing it. 5.02.2.2 Mitigating Effects of Water Chemistry on Degradation of Reactor Materials

1977: Neutral, oxygenated water

Corrosion, radiation buildup issues

1980s: Purer is better

IGSCC was first observed in small bore piping using sensitized 304 stainless steel fairly soon after BWRs started operation. Laboratory studies showed that impurities increased IGSCC rates, and in fact water quality in BWRs gradually improved in the early 1980s. However, the same studies found IGSCC in high-purity oxygenated water typical of good BWR operations. The key parameter affecting IGSCC was found to be ECP, as shown in Figure 5. In this laboratory test, the change from oxidizing conditions typical of normal water chemistry (NWC) operation

Chemistry guidelines

Late 1980s–1990s: HWC, zinc

Controlling IGSCC, radiation buildup

2000s: Noble metal chemical addition

Core internals cracking control with lower fields

Promising new option

2006–2008: Online Noblechem

Figure 2 Evolution of Boiling water reactor chemistry options from 1977 to 2008.

40

Number of BWRs

35

Zn injection

NMCA

HWC (no NMCA)

OLNC

30 25 20 15 10 5 0 1983

1988

1993

1998

2003

2008

Figure 3 Implementation of zinc injection, hydrogen water chemistry, noble metals chemical addition, and online noble metal at US boiling water reactors.

21

Water Chemistry Control in LWRs

0.40 0.35 EPRI action level 1

Conductivity ( µS cm–1)

0.30 0.25 0.20 0.15 0.10 0.05 Theoretical conductivity limit, 25 ºC

0.00 1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

Figure 4 Boiling water reactor mean reactor water conductivity at US boiling water reactor.

250

0.4950

0.4945 200

2.7 ⫻ 10−8 mm s–1 1 ⫻ 10−6 mm s–1

150

0.4935

0.4930

0.4925

100 CT2 #7-304SS 4 dpa Constant load, 19 ksi√in.

Dissolved O2

Outlet cond: 0.30 μS cm–1

50

Inlet cond: 0.27 μS cm–1 Na2SO4

0.4920

0.4915 1488

Dissolved oxygen (ppb)

Crack length (in.)

0.4940

1508

1528

1548

1568

1588

0 1608

Test time (h) Figure 5 Laboratory results showing the effect of reducing oxygen concentration on crack growth of 304 stainless steel.

to reducing conditions greatly reduced the rate of crack growth. Furthermore, hydrogen injection was effective at reducing the ECP in BWRs, as shown in Figure 6. In this figure, it can be seen that crack growth rates (CGR) for Alloy 182 were low in hydrogen water

chemistry (HWC), but increased greatly when the plant reverted to normal water chemistry (NWC). These results indicated that continuous hydrogen injection was required to fully mitigate cracking. Examination of extensive inspection data from several plants indicated that no IGSCC was observed with an

22

Water Chemistry Control in LWRs

901.00 900.00

HWC ECP = −510 mV (SHE)

NWC ECP = +110 mV (SHE)

Crack length

174 miles year-1

HWC

< 5 miles year-1

899.00 898.00 897.00 < 5 miles year -1

Alloy 182

896.00 895.00 800

900

1000

1100

1200

1300 Time (h)

1400

1500

1600

1700

Figure 6 Effect of hydrogen water chemistry on crack growth of Alloy 182.

ECP of
230 mV or lower, using a standard hydrogen electrode (SHE). This is the basis for the
230 mV requirement used by US plants for IGSCC control. In BWRs, the radiation field in the core decomposes water to hydrogen and oxygen species, most of which immediately recombine back to water. But some remain as oxygen or hydrogen peroxide, because some hydrogen is stripped into the steam phase before it can recombine. These same radiolysis reactions cause hydrogen to react with oxygen or peroxide to reduce ECP. These reactions occur mainly in the downcomer, and relatively low hydrogen concentrations are effective at lowering ECP in out-of-core regions of the system. More than half the BWRs in the United States adopted low hydrogen injection rates of 0.2–0.5 ppm (called HWC-L), which, coupled with the replacement of recirculation piping using 316 stainless steel, mitigated IGSCC of recirculation piping. In the 1990s, concerns about the cracking of core internals increased, but the low concentrations of hydrogen used to protect out-of-core regions were not sufficient to reduce ECP enough to mitigate IGSCC of in-core materials, because of the radiolysis of water occurring in the core. As a result, it was necessary to increase hydrogen concentrations to 1.6–2.0 ppm to lower the in-core ECP sufficiently to provide protection in the reactor vessel (termed HWC-M for moderate concentrations of hydrogen). Although this approach was effective in protecting core internals, it also increased radiation fields in the steam side of the circuit, including the turbines, as a result of carryover of nitrogen-16 under reducing chemistry. (Under the oxidizing conditions of NWC, most of the N-16 remains in the water as soluble

species such as nitrate, and only a small percent is transported with the steam.) In some plants, local shielding of turbine components has reduced the impact of the gamma radiation to acceptable levels, but the projected 4–6-fold increase did in fact curtail plans for increased hydrogen injection rates at many plants. Note that these N-16 radiation fields are a problem only when the plant is at power, as rapid decay occurs at shutdown because of the short halflife of N-16. (By contrast, out-of-core radiation fields from Cobalt-60 persist after shutdown and impact on maintenance work during outages.) NMCA was developed to increase the efficiency of hydrogen in BWR cores, to avoid high N-16 fields. In this process, a nanolayer of platinum þ rhodium is deposited on the wetted surfaces of the reactor. These treated surfaces catalyze the hydrogen redox reaction, converting oxygen back to water. When the addition of hydrogen to the feedwater raises the molar ratio of H2 to O2 to 2 or higher, the ECP of the treated surfaces drops to the hydrogen/oxygen redox potential, which is about
450 mV (SHE). This can be achieved with hydrogen concentrations of only about 0.2 ppm, and under these conditions, the main steam radiation level is not increased to an unacceptable level. The first plant used NMCA successfully in 1997, and over 25 plants have already followed, with excellent results. Field measurements show that NMCA has been effective in providing mitigation against IGSCC by lowering the ECP below the
230 mV (SHE) threshold with relatively low hydrogen injection rates. The NMCA process is typically applied at refueling outage, before new fuel is inserted into the core,

Water Chemistry Control in LWRs

additional benefit with NMCA on the upper, outer shroud regions, as indicated by the additional shading in the left-hand side of the figure5. It is estimated that noble metals protect slightly more of the outer core region than does moderate HWC (HWC-M), but the difference is not significant. Figure 8 shows the dramatic benefit of noble metals in reducing the rate of stub tube cracking at Nine Mile Point 1 since the application in 2000. Before 2000, several stub tubes had to be repaired or replaced at each outage, but since the application, only one tube leaked, and this was believed to have already cracked before NMCA. Recently, attention has been focused on the online application of noble metals, with the first application at the KKM plant in Switzerland. By April 2008, there were four applications in the United States. This is discussed in a later section.

HWC protected regions

NMCA protected regions

and is effective for about three fuel cycles, before reapplication is necessary. The regions of the reactor vessel internals that are protected by HWC-M or NMCA are shown in Figure 7. While both techniques offer significant areas of mitigation, there is an

5.02.2.3

Radiation Field Control

Corrosion products deposited on the fuel become activated, are released back into the coolant, and may be deposited on out-of-core surfaces. Both soluble and insoluble species may be involved, the latter tending to deposit in stagnate areas (‘crud traps’). The chemistry changes to control IGSCC resulted in increased out-of-core radiation fields, and the implementation by most plants of depleted zinc injection to

Figure 7 Mitigated regions of the boiling water reactor core.

Number of stub tubes identified with IGSCC throughwall cracking based on leakage

12

10

8

Noble metal applied mid cycle may 2000

6

4

2

0 1984–1985 1986–1987 1988–1990

23

1991

1993

1995

1997

1999

2001

2003

2005

2007

RFO-11

RFO-12

RFO-13

RFO-14

RFO-15

RFO-16

RFO-17

RFO-18

RFO-19

Year Figure 8 Mitigation of stub tube cracking at Nine Mile Point Unit 1.

24

Water Chemistry Control in LWRs

control dose rates, as discussed later in this section. During shutdowns, the major radiation source for personnel exposure is activated corrosion products, deposited on primary system surfaces. Exposures are generally accumulated at high-radiation field locations where maintenance work is frequently needed. Although improvement of maintenance equipment and procedures, reduction of maintenance requirements, increased hot-spot shielding, and control of contamination dispersion have significantly reduced total exposure, further reduction of radiation fields is a major goal in programs for minimizing occupational radiation exposure. The primary source of radiation field buildup on out-of-core surfaces in BWRs is 60Co, which in mature plants usually accounts for 80–90% of the total dose. 60Co has a relatively long half-life of 5.27 years. The higher the soluble 60Co concentration in the coolant, the more 60Co is incorporated and deposited on out-of-core systems and components, resulting in higher dose rates on recirculation piping, the reactor water cleanup system, dead legs, and other crud traps in the system. Other activated transition metals such as 54Mn, 58Co, 59Fe, and 65Zn contribute the remainder of the dose. 51Cr also contributes significantly to the piping dose in some NMCA plants. The radiation fields commonly measured in a BWR at the straight vertical section of recirculation pipes are considered to be more representative for the purposes of radiation buildup trending and comparison with other plants. These measurements are done in a prescribed manner developed under the EPRI BWR Radiation and Control program and are called BRAC point measurements. These measurements represent primarily the incorporation of soluble 60Co into the corrosion film on the piping surfaces and tend to be a fairly good predictor of drywell dose rates. The deposition of particulate oxides that contain 60Co and other activated species can also contribute significantly to outof-core radiation levels in BWRs, especially in hot spots. The particulate oxides, which vary in size, originate primarily from corrosion of the steam/condensate system and are introduced via the feedwater. The sole precursor of the gamma-emitting 60Co isotope is 59Co. 59Co is present as an impurity in the nickel in structural alloys used in BWRs (e.g., Type 304 stainless steel) and is the main constituent of wear-resistant alloys (e.g., Stellite), used as hard facing in valves and other applications requiring outstanding wear resistance. Corrosion and wear lead to release of 59Co into the coolant from these sources,

which is transported to the core and incorporated into the crud that deposits on the fuel rods. The 59 Co is activated to 60Co by neutron activation, released back into the coolant, and incorporated as a minor constituent into the passive films that form on components that are inspected, repaired, and replaced by maintenance personnel. Components in the neutron flux (e.g., the control blades) directly release 60Co. Cobalt source removal is clearly important if radiation fields are to be minimized. Another gamma-emitting isotope, 58Co, is formed by the activation of nickel from stainless steel and nickel-based alloys. 58Co has a shorter half-life and is not as major a contributor to radiation fields as 60Co in BWRs, but is much more significant in PWRs. Shutdown drywell dose rates increase when coolant chemistry is changed for the first time from oxidizing (NWC) to reducing (HWC) conditions. This results from a partial restructuring of the oxides formed under the oxidizing conditions of NWC (Fe2O3 type) to a more reducing spinel type oxide compound (Fe3O4 type). The oxides affected are the fuel deposits, the corrosion films on stainless steel piping, and out-of core deposits. This results in an increase in the chemical cobalt (and 60Co) concentration in the oxide because of the higher solid-state solubility of transition metals in the spinel structure. The presence of a higher soluble reactor 60Co concentration released from fuel crud while this conversion is occurring only aggravates the situation. The processes are depicted in Figure 9. The net result at most plants is a temporary increase in reactor water 60 Co, both soluble and insoluble forms, which leads to significantly increased shutdown dose rates because of both the increased reactor water concentrations and the increased capacity for transition metal uptake by the spinel phases.6

Oxide stable under normal water chemistry Fe2O3 (containing 60Co, 58Co, 54Mn, etc.)

• Corrosion films • Vessel crud • Fuel crud

Restructuring under HWC conditions

Fe3O4 form of oxide

Small insoluble particles containing 60Co, 54Mn, etc. Soluble 60Co, etc. released during restructure

Figure 9 Boiling water reactor oxide behavior under reducing conditions.

Water Chemistry Control in LWRs

25

0.8 Before Zn addition After Zn addition

RxW 60Co (Ci kg −1)

0.6

0.4

0.2

0 Brunswick-1

Brunswick-2

Dresden-2

Figure 10 Hydrogen water chemistry plant RxW

60

Duane Arnold

Fitz patrick

Monticello

Pilgrim

Co response to zinc addition.

As mentioned earlier, zinc addition reduces radiation field buildup. The mechanism of the zinc ion effect is complex, as release of 60Co from fuel crud is reduced, and deposition out-core is also reduced. Overall, reactor water 60Co is decreased significantly after zinc addition, as shown by plant data in Figure 10. Aqueous zinc ion promotes the formation of a more protective spinel-structured corrosion film on stainless steel, especially when reducing conditions are present. Second, both cobalt and zinc favor tetrahedral sites in the spinel structure, but the site preference energy favors zinc incorporation. Thus, the available sites have a higher probability of being filled with a zinc ion than a cobalt ion (or 60Co ion), and hence the uptake of 60Co into the film will be significantly less if zinc ion is present in the water. The 60 Co remains longer in the water and is eventually removed by the cleanup system. The zinc was originally added to the feedwater as ZnO, but it was quickly found that the 65Zn that was created by activation of the naturally occurring 64Zn isotope in natural zinc created problems. With the use of zinc oxide depleted in the 64Zn isotope, called depleted zinc oxide (DZO), this drawback was eliminated. Because of the high cost of DZO, feedwater zinc injection was not implemented widely until HWC shutdown dose issues emerged. For the case of plants treated with NMCA and injecting hydrogen, the oxidant concentration on the surface of the stainless steel is zero (due to the Pt

and Rh catalyzing the reaction of any oxidant with the surplus hydrogen). The net result is that the ECP is at or very near the hydrogen redox potential, typically about –490 mV (SHE) for neutral BWR water. This low potential causes a much more thorough restructuring of the oxides to the spinel state than observed under moderate hydrogen water chemistry (HWC-M). Feedwater iron ingress has a significant influence on the effectiveness of zinc injection. As discussed in the next section, deposits on fuel cladding surfaces (called ‘CRUD’) are mainly composed of iron oxides, with other constituents. Therefore, reducing iron ingress from the feedwater has the benefit of minimizing crud buildup, which is important for fuel reliability (next section). For these reasons, extensive efforts have been made to reduce iron ingress, with significant success. Furthermore, fuel crud has a large capacity for incorporating zinc and is in fact where most of the zinc ends up. The lower the amount of crud on the fuel, the greater the proportion of zinc that remains in solution and can subsequently be incorporated in out-of-core surfaces. Therefore, at plants with low feedwater iron, less zinc is captured by the crud on the fuel, so a relatively greater amount remains in solution and is available to control out-of core radiation fields. This is very important, as zinc injection rates are limited by fuel performance concerns, and hence lowering feedwater iron is essential for maintaining lower radiation fields.

26

Water Chemistry Control in LWRs

5.02.2.4

Fuel Performance Issues

Fuel durability has improved over the years, and failures have declined, helped by improvements in water purity. In operation, zircaloy fuel cladding develops a thin oxide layer (ZrO2), which typically does not adversely affect performance. However, an increase of deposition of corrosion product deposits (‘crud’) on this oxide film is undesirable because it can reduce heat transfer and increase fuel pin temperatures, with resultant increased corrosion of the fuel cladding, ultimately increasing the risk of fuel failure. Moreover, the addition of additives to control corrosion may increase the risk of crud buildup on the fuel. For example, zinc and noble metals in BWRs tend to increase the adherence of crud deposits on the fuel, which can result in undesirable oxide spalling in higher-rated cores. In fact, corrosion-related fuel failures occurred at four plants in the United States between 1999 and 2003. Although the precise root cause of fuel failures is often difficult to determine, it is clear that excessive crud buildup played a role in these failures. With progressive uprating of fuel duty in both PWRs (and BWRs), the margin to tolerate crud has been reduced and additional care has to be taken in specifying the water chemistry to avoid undesirable fuel performance issues. Despite these more demanding conditions, fuel failures have decreased in recent years. Concern about the possibility of adverse effects of NMCA on fuel has prompted imposition of a strict limit on the amount of noble metal that can end up on the fuel and guidance on the injection of zinc. Plant data indicate that spalling of the corrosion layer from

fuel cladding, which is often regarded as a precursor to cladding failure, is prevented if the cycle average feedwater zinc is maintained below 0.4 ppb in NMCA plants (0.6 ppb for non-NMCA plants). More recent data indicate that quarterly averages may be as high as 0.5 ppb for NMCA plants, without occurrence of spalling.5 These feedwater zinc data are the basis for limits in the water chemistry guidelines. The 2008 chemistry guidelines7 retain the cycle average feedwater zinc limit of 0.4 ppb (0.6 ppb for non-NMCA plants) but enable a slight increase in the quarterly average to 0.5 ppb, which may allow flexibility in controlling radiation buildup in parts of the cycle. The tighter control of water chemistry in recent years has been successful in controlling crud formation on fuel cladding, and Figure 118 shows failures from pellet–clad interaction causing SCC, fabrication defects, debris, and crud/corrosion. Note that there have been zero crud/cladding related fuel failures in US BWRs since 2004 (although assessment of 2007 failures is not yet complete, crud/corrosion is not believed to be a factor here). Analysis of recent plant data confirms that control of feedwater iron ingress has the positive benefit of reducing the amount of crud on the fuel. Control of copper, which generally originates from admiralty brass alloys, is also beneficial; not only can copper have detrimental effects on the fuel, but it also limits the ability of hydrogen to reduce the ECP, and it also leads to higher radiation fields. As a result, most US plants have replaced condensers containing brass tubing.

Number of failed assemblies

30 25 20

PCI-SCC Unknown Fabrication Debris Crud/corrosion

15 10 5 0 2000

2001

2002

2003 2004 EOC year

2005

2006

Figure 11 US boiling water reactor fuel failures by mechanism for each end-of-cycle (EOC) year.

2007

Water Chemistry Control in LWRs

5.02.2.5

Online Addition of Noble Metals

As discussed earlier, the classic NMCA process is generally applied during refueling outages before the new fuel is loaded into the core. Reapplication after about three cycles of operation takes approximately 2 days, while the plant maintains 107–154  C as it enters the refueling outage. To reduce this outage time, GE-Hitachi developed OLNC, first demonstrated at KKM (a GE design of plant in Switzerland) in 2005, with several more additions subsequently. Preliminary results indicate that there have been no unexpected chemistry effects during the first OLNC applications, and shutdown radiation fields actually decreased at KKM after OLNC.5 Subsequently, CGR of susceptible welds decreased significantly, as shown by the decrease in slopes in Figure 12 after OLNC initiation for two welds that have been monitored for several years. The effects of OLNC on fuel have been extensively studied in fuel removed from KKM, and no adverse effects have been observed. The jury is still out on this concern, but the general assessment is that OLNC will have no more impact than the classic application, and may well prove to be of less concern. More IGSCC and fuel measurements are planned, but with no issues emerging to date, it appears that OLNC applications about every 12 months would be effective and economical, avoiding the critical path time necessitated for the classic NMCA application during refueling outages. Initial OLNC applications have been carried out at plants that had previously applied noble metals in the classic off-power manner.

However, the first OLNC application at a plant that has not used noble meals previously occurred in late 2008, but no results are available.

5.02.3 PWR Primary Water Chemistry Control 5.02.3.1 Evolution of PWR Primary Chemistry Strategies In the very early days of PWR operation, heavy crud buildup on fuel cladding surfaces was caused by the transport of corrosion products from the SGs into the reactor core. As a result, activated corrosion products caused high-radiation fields on out-of-core surfaces (Figure 13), fuel performance was compromised, and even coolant flow issues were observed in some plants. These problems were initially mitigated by imposing a hydrogen overpressure on the primary system, to reduce the ECP, and raising the primary chemistry pH. Materials degradation in primary systems was then not a major concern, with most of the emphasis focused on secondary side corrosion issues in the SGs. Commercial PWR power plants use a steadily decreasing concentration of boric acid as a chemical shim (for reactor control) throughout the fuel cycle, which results in the use of lithium hydroxide to control pH. Some 30 years ago, the concept of ‘coordinated boron and lithium’ was developed, whereby the concentration of LiOH was gradually reduced in line with the boric acid reduction to maintain a constant pH.

300 NC appl.

HWC

Indication length (mm)

250 Ind 9

Ind10

Not inspected in 00, 01,04

200 150 100 50

27

OLNC 37 g OLNC 98 g OLNC 198 g OLNC 199 g

Indications 9,10 may be seeing mitigation by OLNC

0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Year Figure 12 Ultrasonic inspection results after online noble metal chemical addition.

28

Water Chemistry Control in LWRs

Corrosion products deposit out of core Pressurizer Steam generator Corrosion products activated in reactor core

Corrosion products released from SG tubing Coolant pump Reactor Primary loop

Figure 13 Transport and activation of corrosion products in pressurized water reactor primary systems.

Corrosion products released from the steam generator tubes are transported, dissolved, or deposited by the coolant on the basis of solubility differences. The solubility of nickel and iron depend on pH, temperature, and redox potential, all of which vary with location around the nonisothermal system. Originally, a constant at-temperature pH of 6.9 was recommended, based on the minimum temperature coefficient of solubility of magnetite. In fact, it was determined that heavy fuel crud buildup was avoided if a constant pH of at least 6.9 was maintained. This was possible with 12-month fuel cycles, but fuel cladding corrosion concerns limited the maximum LiOH concentration to 2.2 ppm. Consequently, plants often started the fuel cycle with pH below 6.9, which resulted in deposition of corrosion products on the fuel, activation of cobalt and nickel, and subsequent transport to out-of-core surfaces, resulting in radiation fields remaining relatively high. Even though detailed studies of fuel crud showed that the prime constituent of the crud was nickel ferrite (for which the optimum pH is 7.4), this coordinated chemistry had remained the standard for many years, until higher pHs became the norm in the 1990s. Although research and plant demonstrations showed that the 2.2 ppm limit was excessively conservative, the move to higher Li concentrations has been slow. However, detailed fuel examinations from a recent plant demonstration (that will be discussed later) have indicated that Li can be raised to as high as 6 ppm.

About 25 years ago, primary water stress corrosion cracking (PWSCC) of Alloy 600 SG tubes was observed in a few plants, leading to studies on mitigating this effect. Following successful demonstration of zinc injection in BWRs, initial field tests at PWRs showed that radiation fields were reduced, and laboratory studies indicating that PWSCC was reduced were eventually confirmed. As a result, zinc injection is being implemented at an increasing rate, although concerns about fuel performance at highduty plants have not been completely resolved. Most recently, buildup of boron-containing crud in areas of subcooled nucleate boiling leading to localized flux depression has encouraged the use of higher Li concentrations to minimize corrosion product transport. Concerns about the potential adverse effects of zinc deposited in high-crud regions have resulted in several highly rated plants applying in situ ultrasonic fuel cleaning before implementing zinc injection. Although zinc injection was developed for radiation field control, laboratory studies showed that it also inhibited SCC under PWR conditions. The identification of PWSCC in reactor vessel penetrations in the last 15–20 years has encouraged the use of zinc injection, but has also focused attention on the effects of dissolved hydrogen, for which the recommended range has remained 25–50 ml kg
1 for 30 years. It now appears that raising hydrogen will reduce PWSCC rates, while lowering it may delay initiation of PWSCC. The interactions of materials, radiation fields, and fuels in PWR primary

Water Chemistry Control in LWRs

PWSCC: pH (Li, B) minimal effect Zn beneficial dissolved H2 effect

Plant operations

29

Dissolved H2 control range

Materials degradation

PWR chemistry control

Fuel performance

Plant dose rates Radiation fields: pH (Li, B), Zn beneficial

Crud deposition: Zn concern for highly rated cores

Figure 14 Pressurized water reactor primary chemistry optimization. Reproduced from Fruzzetti, K.; Perkins, D. PWR chemistry: EPRI perspective on technical issues and industry research. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008.

Dissolved (H2) range changes

Elevated constant pH (7.3/7.4) Ultrasonic fuel cleaning Elevated constant pH (7.1/7.2) Zinc injection Modified elevated lithium program EPRI water chemistry guidelines Elevated lithium program Constant pH 6.9 1975

1980

1985

1990

1995

2000

2006 2008

Figure 15 Pressurized water reactor primary chemistry changes at US plants.

chemistry and optimization issues covered in the water chemistry guidelines, which are discussed later, are depicted in Figure 14. The evolution of water chemistry control in PWR primary systems in the United States over the last 30 years is shown in Figure 15. The following sections address the three main factors – pH control, zinc injection, and dissolved hydrogen control – that have dominated PWR primary chemistry strategies in the past and continue to do so today.9 Each of these factors is considered from the viewpoint of materials degradation, radiation field control, and fuel performance concerns.

5.02.3.2

Materials Degradation

Materials degradation has been covered in detail in Chapter 5.04, Corrosion and Stress Corrosion Cracking of Ni-Base Alloys and Chapter 5.05, Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels, and here only the specific effects of water chemistry variables on materials in PWR primary systems will be reviewed, particularly those that may affect the chemistry of optimization process. Recent papers by Andresen et al.10,11 provide detailed results of a comprehensive study of the effects of PWR primary water chemistry on PWSCC of nickel-based alloys. Extensive studies have been carried out to determine the effect of lithium, boron, and pH on PWSCC, and the generally held conclusion is that any effects are minimal, especially compared to material susceptibility, stress state and temperature, and other operational issues. Crack initiation tests using the most reliable types of reverse U-bend specimens indicate that pH has a relatively small effect on crack initiation (generally less than a factor of 2). Although the most rapid crack initiation occurred at pH310  C 7.25, with slower rates at higher or lower pHs, CGR tests generally confirm that pH has minimal effect. The effect of lithium is even smaller than the pH effect, and the influence of boron is minor or nonexistent. Andresen et al. concluded that the effects of relevant variations in PWR primary water chemistry (B, Li,

30

Water Chemistry Control in LWRs

and pH) have little effect on the SCC growth rate in Alloy 600, and thus provide little opportunity for mitigation of PWSCC. Plant data have found no adverse effects from increasing lithium and pH in primary systems. As a result, it is considered that adjusting pH, lithium, or boron to minimize crack initiation may be of minimal value. The 2007 edition of the PWR Primary Water Chemistry Guidelines12 reviewed the most recent data and concluded that pH strategy changes based on PWSCC considerations are not warranted. This means that plants have the flexibility to pursue B/Li/pHt chemistry adjustments to minimize crud transport and radiation buildup without concern for negative effects on PWSCC susceptibility of nickelbased alloys, although of course chloride and sulfate impurities should continue to be minimized. Following good experience in BWRs, zinc injection has been implemented in the primary systems of PWRs, both to reduce primary side cracking of nickelbased alloys and to control dose rates. The qualification work for BWRs showed that zinc inhibited SCC, but the benefit was not sufficient to avoid the need for hydrogen water chemistry to mitigate IGSCC. Thus, the motivation for BWR zinc injection was exclusively radiation field control. The situation in PWRs is different, as laboratory work13 showed that initiation of PWSCC was significantly delayed by zinc injection, and hence the motivation for the initial applications of zinc in most US PWRs at the 10–30 ppb level was to control PWSCC of SG tubing. Additionally, German-designed PWRs and a few US plants used 5 ppb depleted zinc for radiation control.

Figure 16 shows the rate of introduction of zinc injection at PWRs worldwide. Zinc injection produces thinner, more protective oxides on stainless steel and Alloy 600, with zinc displacing Co2þ, Ni2þ, and Fe2þ from normal spinels to give ZnCr2O4, which is very stable. The benefits of PWR zinc injection have been clearly demonstrated in reducing PWSCC degradation (especially growth rate) of Alloy 600 SG tubes, and in controlling radiation fields. Evaluation of currently available laboratory data2 indicates that PWSCC initiation will be reduced, and PWSCC CGR may be reduced in thicker cross-section components, depending upon other factors such as the stress intensity factor of the specimen. Andresen et al.11 concluded that crack growth mitigation by adding Zn requires further study, although two of four tests show a decrease in growth rate of >3. Molander et al.14 also found that the effect of zinc on CGR was minor. Hence, more work is needed before making definitive conclusions from laboratory studies regarding the benefit of zinc in mitigating CGR. SG tube nondestructive examination (NDE) data from eight plants injecting zinc indicated reduction in the incidence of PWSCC by a factor of 2–10.9 An example from a 2-unit PWR showing the effect of zinc on SG tubing over successive cycles is given in Figure 17. The largest effect of zinc appears to be on initiation of cracking, with a smaller effect on CGR, with the data indicating a factor of 2–10 reduction for initiation and about a factor of 1.5 reduction in CGR, consistent with the extensive laboratory work,11

Application of zinc in world PWRs

Number of plants and percentage of PWR injecting zinc

50 45

Percent of PWRs injecting

Number of units injecting

40 35 30 25 20 15 10 5 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Year

Figure 16 Application of zinc injection in pressurized water reactors worldwide.

Water Chemistry Control in LWRs

31

140 X indicates last refueling outage before start of zinc injection

Number of new tubes affected

120 100 80

Unit 1 Unit 2 60 40 20 0

X−3

X−2

X−1

X+1 X Refueling outage

X+2

X+3

X+4

Figure 17 Effect of zinc on steam generator tube degradation at a US pressurized water reactor.

70 60 Number of plants

indicating that zinc inhibits mainly by delaying the initiation of PWSCC. However, the SG NDE data also showed that zinc reduced the rate of crack propagation (depth) by 17–60%. These results are consistent with initial laboratory test data indicating that zinc reduced crack propagation by a factor of approximately 3 at low stress intensities, but had no effect at higher stress intensities. In addition, the lack of cracking in the Farley PWR pressure vessel head penetrations (exposed to zinc for over 12 years), compared to PWSCC indications in similar pressure vessel heads in other plants, suggests that zinc addition is beneficial for Alloy 600 (and possibly Alloy 82/182) thick-section components under PWR primary service conditions. Recent work has studied the influence of dissolved hydrogen on PWSCC. In the early days of PWR operation, the lower limit on hydrogen was set at 25 ml kg
1, to provide adequate margin against radiolysis and heavy crud formation. Plant tests in France showed that this limit was excessively conservative and that less than 10 ml kg
1 would be satisfactory, provided good control of oxygen was maintained in makeup water. Several workers have found that the maximum in PWSCC CGR occurs close to the ECP corresponding to the Ni/NiO thermodynamic equilibrium condition.15 Although this potential is unaffected by lithium/boron/pH (consistent with the fact that these do not greatly influence PWSCC over the range of practical relevance), the equilibrium potential is significantly affected by the dissolved hydrogen

50 40 30 20 10 0 25–30 30–35 35–40 40–45 45–50 Cycle average hydrogen concentration (cm3 kg−1)

Figure 18 US plant data for dissolved hydrogen.

concentration. Andresen et al. found that the peak in SCC growth rate versus H2 fugacity was temperature dependent, but generally fell within the hydrogen concentration range used in PWRs. This provides an opportunity for mitigation, by perhaps a factor of 2 in Alloy 600 and a factor of 5 in Alloys 182, 82, and X750, as the median value of the dissolved hydrogen concentration for US plants is approximately 35 ml kg
1. US PWRs currently operate within dissolved hydrogen within the recommended 25–50 ml kg
1 range, with the majority in the 30–40 ml kg
1 range, but none with more than 44 ml kg
1 (Figure 18). The lower limit is set conservatively to provide an operating margin over the level of hydrogen required

32

Water Chemistry Control in LWRs

to suppress water radiolysis in the reactor core. Somewhat lower concentrations are used in other countries. The dissolved hydrogen concentrations corresponding to the peak CGR for a typical range of PWR primary operating temperatures are 4.3 ml kg
1 at 290  C, 10.4 ml kg
1 at 325  C, and 16.5 ml kg
1 at 343  C.15 Andresen et al.11 published Figure 19, which indicates the proposed factors of improvement on changing from an initial hydrogen concentration of 25 ml kg
1. It can be seen that raising the hydrogen provides benefit, but lowering it is detrimental below 330  C.

In response to the data showing the benefit of increasing hydrogen in reducing CGR, the US industry program in progress focuses on the extent to which dissolved hydrogen can be increased without adverse consequences to other parts of the system. Other countries, including Japan, are also investigating lowering hydrogen, because laboratory data suggest that the initiation of cracking is delayed at lower hydrogen concentrations. This is depicted in Figure 20, as discussed by Molander.14 The lower line in this figure shows the time to initiate cracking, based on laboratory tests using

Factor of improvement from H2

4.0 Based on Alloy 182, a current H2 level of 25 cm3 kg−1

3.5

70 cm3 kg-1 H2

25

3.0

45 cm3 kg-1 H2

25

2.5 2.0 1.5

Good 1.0 25 0.5 0.0 270

280

290

1 cm3 kg-1 H2

300 310 320 Temperature (⬚C)

4 cm3 kg-1 H2

25

330

340

Bad

350

Figure 19 Effect of dissolved H2 on primary water stress corrosion cracking crack growth rate at different temperatures.

ml H2/kg H2O (330 ˚C) 10

15

20

25

30

Crack initiation time (h)

Jenssen data on Alloy 600

35

1E−07

Growth 8E−08

20 000

6E−08

15 000

4E−08

10 000 Initiation

2E−08

5000

0 0

5

10 Hydrogen activity (kPa)

15

Crack growth rate (mm s–1)

5 25 000

0E+00 20

Figure 20 Dependencies between the dissolved hydrogen content in pressurized water reactor primary coolant on the crack initiation time observed on initially smooth surfaces and on the crack propagation rate.

Water Chemistry Control in LWRs

reverse U-bend specimens, whereas the upper line shows crack growth data over a similar concentration range. Thus, the lowering of hydrogen appears feasible. However, the relative importance of crack initiation and crack propagation is very dependent on material and plant conditions. In the United States, concern about increased crack propagation at low hydrogen and low temperatures, as shown in Figure 19, has resulted in moving to higher hydrogen being preferred to the alternative of reducing hydrogen. Several factors combine to make higher H2 the preferred way to mitigate SCC, including the importance of bottom-head penetrations (which are exposed to 290  C water) and the recent observation that the CGR in coldworked Alloy 600 is not mitigated at low H2.11 The preferred strategy in the United States is to gradually increase hydrogen to the upper end of the existing range, with the potential to move higher (say to 60 ml kg
1) when the ongoing qualification work is completed. This will include evaluation of the effects of dissolved hydrogen on radiation fields and fuel performance, although any such effects are expected to be minimal.16 5.02.3.3

PWR Radiation Field Control

Corrosion products released from out-of-core materials (primarily SG tubing) deposit on the fuel and become activated, are released back into the coolant, and may be deposited on out-of-core surfaces. Both soluble and insoluble species may be involved, with the latter tending to deposit in stagnate areas (‘crud traps’). In addition to the chemistry items discussed later in this section, it must be stressed that other factors are important to the goal of reducing radiation fields. In particular, the success of the later German-designed plants in eliminating cobalt sources in hardfacing alloys, thereby achieving very low radiation fields, demonstrates the benefits of cobalt source reduction. With many plants replacing SGs, a correlation between recontamination rates and surface finish of the new SG tubing has been noted by Hussey et al.17 Typical PWR fuel cycles start with a relatively high boric acid concentration, which gradually reduces to zero at the end of the cycle. Lithium hydroxide is added to maintain an approximately constant pH. As the duration of fuel cycles increased, more boric acid was required at the start of cycle, which in turn necessitated increased LiOH to maintain the desired pH (Figure 21). As mentioned earlier, radiation field buildup can be controlled by minimizing corrosion product

33

transport and activation. Initially, coordination of lithium hydroxide with boron to maintain a constant at-temperature pH of 6.9 was recommended, based on the minimum solubility of magnetite. In fact, the prime constituent of the crud turned out to be nickel ferrite, requiring a pH of 7.4 for minimum solubility. Fruzzetti et al.15 have recently reviewed the data on elevated pH, which provides a number of benefits including decreased general corrosion (and thus reduced corrosion product transport to the core). Field-tests of pHs greater than 6.9 confirmed that radiation fields were lower. Although no adverse effects were observed on the fuel, many plants were slow to abandon a 2.2 ppm limit, established to avoid excessive zircaloy corrosion. However, there were indications of heavier crud formation after long periods operating below pH 6.9, and as fuel concerns relaxed, a gradual move toward a maximum of 3 ppm lithium resulted. Moreover, pHs in the range 7.1–7.2 became more popular in the late 1990s, with 7.3–7.4 eventually gaining favor. Figure 22 shows the maximum lithium concentrations reported by US PWRs in recent years. It can be seen that 95% are now using greater than 3 ppm at full power: a significant change from earlier in the decade. A demonstration of elevated Lithium/pH is in progress at Comanche Peak PWR.18 The goal was to reduce radiation fields and reduce susceptibility to the Axial Offset Anomaly (AOA) by reducing crud buildup. This test involved increasing the primary system pH from 7.1/7.2 to 7.3 and then two cycles at 7.4. No significant adverse trends have been noted, either in the area of chemistry or core performance. Radiation fields measured have shown a modest but continued improvement. On the basis of the positive trends and absence of any negative effects, Comanche Peak has established elevated constant pHTave 7.4 as the primary chemistry regime for both units. Without the increases in pH/lithium that have taken place, radiation fields would have been expected to increase significantly for longer fuel cycles. The increase in boiling in localized regions of the core (called subcooled nucleate boiling) in PWRs resulting from power uprating has resulted in higher crud buildup on the upper fuel surfaces, and there is growing evidence from US PWRs that radiation fields are indeed higher for the highest rated cores. Enriched boric acid (EBA), that is boric acid enriched with B-10, enables a given pH to be achieved with less lithium hydroxide, as the required concentration of B-10 can be obtained with less total

34

Water Chemistry Control in LWRs

Constant pH 7.2 6

Lithium ‘Li high limit’ ‘Li low limit’

5 Li target = 6.0 E−7 B2 + 0.0023B + 0.4413

Lithium (ppm)

4 3.5 ppm limit 3 2.2 ppm limit 2

Start of 18-month cycle

1

Start of 12-month cycle

0

20

80

140

200

260

320

380

440

500

560

620

680

740

800

860

920

980

1040

1100

1160

1220

1280

1340

1400

1460

1520

1580

1640

1700

0

Boron (ppm) Figure 21 Lithium concentrations required to maintain pH 7.2 for different fuel cycle lengths.

boric acid. EBA is used at several plants in Europe, typically to increase shutdown margin when using mixed oxide fuel (MOX), but has not been applied to date in the United States. However, consideration is being given to using EBA at some plants that will use MOX fuel in the future. Despite the transition to the use of EBA in operating plants, designing for it in new plants is recommended.19 As discussed earlier, the motivation for the initial applications of zinc in most US PWRs was to control PWSCC of SG tubing. However, German-designed PWRs and a few US plants used 5 ppb depleted zinc for radiation control, mostly with depleted zinc to avoid zinc-65 formation. A recent paper ‘‘Understanding the zinc behavior in PWR primary coolant: a comparison between French and German experience’’ by Tigeras et al.20 provides a European perspective on this topic. This paper concludes that ‘zinc injection seems to present the most positive and clearest results: in all the units injecting zinc, a dose rate reduction has been detected after a certain period of exposure without leading to any negative impact on plant

systems, components, and operation.’ Thus ‘zinc injection should be considered as a strategy with benefits in short, medium, and long term. Its application as soon as possible in the life of nuclear power plants and especially before SG replacement and fuel cycles modifications seems to be an excellent decision to contribute to ensuring the passivation process of new components, the fuel performance, the full power operation of the units, and the long life of materials and components.’ Figure 23 shows the effect of zinc in reducing radiation dose rates at several plants. It can be seen that the reduction factor approximately correlates with the cumulative zinc exposure in ppb months (the product of the average zinc concentration and the duration of zinc addition). As little as 5 ppb zinc has been shown to reduce radiation fields by 35–50% at operating plants, based on zinc exposures of 700 ppb months. There is relatively little difference between plants with Alloy 600/690 SG tubing and those with Alloy 800 tubing, but plants using depleted zinc show greater benefit than those using natural zinc, as shown in the figure.

Water Chemistry Control in LWRs

35

Percentage of units within range

70

<3 ppm 3.0−3.5 ppm >3.5 ppm

60 50 40 30 20 10 0

2000

2001

2002

2003

2004

2005

2006

2007

EOC year Figure 22 Maximum reported coolant lithium (full power) at US pressurized water reactors.

Cumulative dose rate reduction fraction

1.2 Alloy 800 w/depleted zinc Alloy 600 and 690 w/depleted zinc Alloy 600 and 690 w/natural zinc Log Alloy 800 plants Log Alloy 600 and 690 w/depleted zinc Log Alloy 600 and 690 w/natural zinc

1

0.8

0.6

0.4

0.2

0 0

200

400

600 800 1000 1200 1400 1600 Cumulative zinc exposure (ppb months)

1800

2000

Figure 23 Effect of zinc injection on radiation dose rates.

5.02.3.4

Fuel Performance

With progressive uprating of fuel duty, the margin to tolerate crud has been reduced and additional care has to be taken in specifying the water chemistry to avoid undesirable fuel performance issues. Figure 24 shows the root causes of PWR fuel failures since 2000, including failures from pellet–clad interaction causing SCC, fabrication defects, debris, grid fretting, and crud/corrosion. In contrast to the BWR

situation, shown in Figure 11, very few failures in recent years have been attributed to crud/corrosion (the exceptions to this comment are discussed in a following section). A phenomenon called axial offset (AO) has caused concern over the past 10 years.21 AO is a measure of the relative power produced in the upper and lower parts of the core and is normally expressed as a percent, with a positive percent indicating that

36

Water Chemistry Control in LWRs

Number of failed assemblies

120 Unknown Debris Crud/corrosion

100

Fabrication PCI-SCC Grid fretting

80 60 40 20 0 2000

2001

2002

2003

2004

2005

2006

2007

EOC year Figure 24 US pressurized water reactor fuel failures by mechanism.

more power is produced in the upper part of the core. AOA occurs when boron concentrates in corrosion product deposits (crud) on the upper spans of fuel assemblies undergoing subcooled nucleate boiling, causing a reduction in neutron flux. AOA has affected at least 20 PWRs in the United States, as well as several in other countries. Clearly, fuel crud is involved in the AO phenomenon, and water chemistry effects must be considered in controlling AO. Besides their axial asymmetry, the composition of fuel deposits in boiling cores is different from nonboiling fuel. The nickel-rich deposits on boiling cores tend to be removed much less effectively by conventional chemistry shutdown evolutions than the nickel-ferrite deposits on nonboiling cores. Alternative methods are therefore required for removing corrosion product deposits from reload fuel from highduty cores, including ultrasonic fuel cleaning. An important difference exists between plants with Alloy 600 or 690 SG tubing and those (such as German-designed plants) with Alloy 800 tubing. The latter have a much lower proportion of nickel in fuel crud and have not experienced the AO phenomenon.22 Early work showed that lithium increased zircaloy oxidation rates, although the adverse effects were reduced in the presence of boric acid. As a result, a limit of 2.2 ppm lithium was generally imposed to reduce zircaloy corrosion, although excessive crud formation at low pHs was likely to be more detrimental to the cladding than higher lithium concentrations, especially as the resistance

to corrosion of zircaloy improved. This was confirmed by one of the few failures in recent years that was uniquely attributed to crud buildup. In this example, a move to a longer fuel cycle necessitated increasing the boron concentration at start of cycle; however, the 2.2 ppm lithium limit was retained, resulting in the pH being well below 6.9 for the initial period of the cycle. This in turn caused heavy crud formation, to which subsequent fuel failures were attributed. The move in the past ten years to greater fuel duty, with operation of fuel at higher temperatures (with localized subcooled nucleate boiling), has caused crud-related problems to reappear, particularly the localized flux depression as a result of buildup of boron-containing crud, which were discussed earlier. This in turn has renewed interest in elevated pH/ lithium to minimize corrosion product transport, the use of EBA and the more immediate mitigation that can be obtained from fuel cleaning. Fuel performance is always a concern with changes in water chemistry, such as zinc injection. On the basis of current experience, the impact appears to be minimal for the majority of plants, but insufficient data exist for plants with the highest fuel duties to allow application without postexposure fuel inspections. Data from US plants suggest little or no fuel concerns for coolant zinc levels up to 40 ppb for plants with less-highly rated cores. Extended experience at these plants, over at least 10 years of operation, indicates no adverse effects on fuel at zinc concentrations from 15 to 25 ppb. However, there have been no data

Water Chemistry Control in LWRs

available until recently for higher zinc concentrations in higher duty cores where significant subcooled nucleate boiling occurs on the fuel clad surface.23 Perkins et al.24 comment that fuel performance must be considered prior to injecting zinc and additional monitoring and fuel surveillances to understand and evaluate the impact and the role of zinc may be required in some circumstances.

5.02.4 PWR Secondary System Water Chemistry Experience 5.02.4.1 Evolution of PWR Secondary Chemistry Strategies The objectives of PWR secondary water chemistry control are to maximize secondary system integrity and reliability by minimizing impurity ingress and transport, minimizing SG fouling, and minimizing corrosion damage of SG tubes. Since secondary side corrosion damage of SG tubes is primarily caused by impurities in boiling regions, where high concentrations of impurities occur in occluded regions of the SG formed by corrosion product deposits, new approaches are continually sought to control corrosion product transport to and fouling within the SGs.25 PWRs have experienced IGA on both the primary and secondary sides of the Alloy 600 SG tubing, which has been a major contributing cause of the replacement of most of the SGs with mill-annealed tubing, not only in the United States but internationally. Figure 25 illustrates the various corrosion processes found in different locations in a recirculating SG.26 PWR secondary system water chemistry has evolved through many changes over the years, largely in response to emerging technical issues associated with this degradation of structural materials in SGs. In the early days of PWR operation, wastage became a problem in the secondary side of PWR SGs, resulting in a switch from the use of sodium phosphate inhibitor to all-volatile treatment (AVT) using ammonia, which in turn brought about the denting phenomenon. Tighter control of impurities, oxidizing potential, and pH were necessary to mitigate the denting problem. Despite continued chemistry improvements, many plants have had to replace SGs of earlier designs (e.g., those tubed with Alloy 600MA), as shown in Figure 26. Newer generation SGs are performing well, although there remain concerns about the adverse effects of lead impurity, causing Pb-assisted stress corrosion cracking (PbSCC), which is discussed later.

37

Lead has been observed in various flow streams (final feedwater, heater drains, etc.) in the secondary systems of PWRs. Lead is detected at some concentration in nearly all deposit analyses (SG and other locations). Lead is present in trace concentrations in secondary system materials of construction, as well as in chemical additives such as hydrazine.15 Figure 27 shows the worldwide causes of SG repairs through 2004. It can be seen that IGA is currently the most prevalent form of degradation. Figure 28 compares the behavior of three types of SG tubing, Alloy 600MA (mill-annealed material used in early plants, Alloy 600TT (thermally treated material used in later plants), and Alloy 690TT (an improved alloy used in most replacement SGs). This diagram is taken from the 2008 PWR Secondary Water Chemistry Guidelines,27 which contains a much more detailed account of corrosion processes. 600TT has reduced susceptibility under mildly oxidizing highalkaline conditions, that is, SCC is not observed until higher pH than for 600MA, and 600TT has approximately the same susceptibility as 600MA under acidic conditions. 690TT is indicated as having a still smaller region of susceptibility in the high-alkaline region and as having no susceptibility in the acid region except under highly oxidizing conditions that are unlikely to occur in plants. However, other work indicates that SCC can occur in 690TT at an acidic pH, especially if lead is present. Also, SCC occurs in both 600MA and 600TT in the mid pH region if lead is present. In the 1990s, improved pH control using amines became a regular practice, and fine-tuning, including using mixtures of different amines to control pH throughout the circuit and coordination with resin utilization, continues today. Hydrazine is used to remove oxygen from the system. Hydrazine levels have continually been reviewed and ‘optimized,’ with due regard to any impact on FAC in secondary systems, as FAC rates increase at very low oxygen concentrations. Molar ratio control (MRC) describes a control strategy that adjusts the bulk water chemistry, generally sodium and chloride, such that the solution that is developed in the flow-occluded region is targeted to be near neutral. MRC can involve the addition of chloride ions to ‘balance’ the cations that cannot be reduced via source term reduction programs. MRC was widely practiced to minimize SCC concerns, but has not been actively employed at plants replacing to SGs tubed with Alloy 690TT. With more plants replacing their SGs, less plants are adopting the MRC program. Only ten plants were doing MRC in 2007,28 and they are all with original SGs with

38

Water Chemistry Control in LWRs

U-bend cracks (PWSCC)

Fatigue

Free span ODSCC IGA

ODSCC

PWSCC

Expansion transition

PWSCC

PWSCC or ODSCC

ODSCC Denting

Fretting, wear, corrosion, thinning

Tubesheet Pitting IGA

Expansion transition

Tubesheet

ODSCC Sludge

Tubesheet

PWSCC tube-end cracking

Tubesheet

Figure 25 Corrosion processes in recirculating steam generators, showing primary water stress corrosion cracking and outside diameter stress corrosion cracking on the secondary side.

Water Chemistry Control in LWRs

39

140 Operating plants Plants w/replacement SGs 120

134

84 88

81

134 134

134 72

64 67

59

52

45 51

37

22 29

17

10

7

7

7

7

3 5 7

2

78

134

134

132

131 132

134 131 131

132

132 133

134

12

132

134

12

133

133

130

121

63 68

55

46

1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

1

22

20

31

41

40

0

11

93 99

60

108 115

127

80

79

Number of plants

100

Year Figure 26 Steam generator replacement status worldwide.

100 90 80

Percent

70 60 50 40 30 20 10 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

0 Year IGA Impingement Pitting Other Wear Thinning

Fatigue Unknown SCC Preventive

Figure 27 Worldwide causes of steam generator tube repair.

600MA and 600TT tubing. Currently, no plants with replaced SGs are believed to be using MRC. Titanium-based inhibitors to minimize corrosion are also employed at some plants. Boric acid

treatment (BAT) involves the addition of boric acid to feedwater. Such approaches are worthy of consideration, on the basis of plant-specific degradation mechanisms, operational considerations, and

40

Water Chemistry Control in LWRs

1.0

TT690

Potential (V vs. Ec )

0.8 Some tests indicate that 690 TT may be susceptible in the low pH region, especially if lead is 690 TT U-bend cracked present in near neutral AVT with lead and oxidizing sludge TT690 TT600 TT600 MA600

0.6

0.4

MA600 0.2 600 MA and 600 TT can be susceptible in mid pH range if lead or reduced sulfur is present.

0

2

3

4

5

6 7 8 pH 300 °C (572 °F)

9

10

11

12

Figure 28 Corrosion mode diagram for Alloys 600MA, 600TT and 690TT (based on Constant Extension Rate Tensile Tests at 300  C), showing regions where materials are susceptible to attack.

interactions. The most recent developments are aimed at reducing deposit buildup in crevices, including the use of dispersants, such as polyacrylic acid (PAA), that is discussed in more detail later. The historical trends in PWR secondary chemistry are shown in Figure 29. 5.02.4.2 Chemistry Effects on Materials Degradation of SGs Corrosion of SG tubes has been the major issue affecting selection of secondary water chemistry parameters. However, corrosion and FAC of SG internals and other secondary system components are also important concerns. Corrosion of SG tube materials is mainly affected by the following water chemistry related factors, in addition to nonwater chemistry factors such as material susceptibility, temperature, and stress:  pH – Corrosion of several different types, including IGA/SCC and pitting, are strongly affected by the local pH. High pH (caustic conditions) and low pH (acidic conditions) accelerate the rates of IGA/SCC.  ECP – The ECP is a measure of the strength of the oxidizing or reducing conditions present at the

metal surface. The rate of corrosion processes are strongly affected by the ECP. Secondary side SCC in tube alloys tends to be accelerated by increases in ECP, that is, by the presence of oxidizing conditions.  Specific species – Some impurity species accelerate corrosion of tubing alloys as a result of their effects on pH and ECP. In addition, lead and reduced sulfur species (e.g., sulfides) appear to interfere with formation of protective oxide films on the tube metal surfaces, and thereby increase risks of IGA/SCC, independent of influences on pH or potential. Similarly, chlorides tend to increase the probability of pitting. These factors have been most thoroughly explored for mill-annealed Alloy 600 (600MA). As discussed in Chapter 5.04, Corrosion and Stress Corrosion Cracking of Ni-Base Alloys, tests indicate that the other tubing alloys, that is, stress-relieved Alloy 600 (600SR), thermally treated Alloy 600 (600TT), nuclear grade Alloy 800 (800NG), and thermally treated Alloy 690 (690TT), exhibit similar tendencies, but have increased resistance to corrosive attack, in the order listed, with 690TT having the highest resistance. Laboratory tests and plant experience indicate that 690TT has very high resistance to IGA/SCC on the outside diameter on tubing (OD IGA/SCC) in

Water Chemistry Control in LWRs

41

Pb remediation Dispersants Titanium Molar ratio control MPA, DMA ETA chemistry Morpholine chemistry Boric acid addition EPRI water chemistry guidelines Ammonia chemistry Phosphate 1975

1980

1985

1990

1995

2000

2005

Figure 29 Evolution of water chemistry for pressurized water reactor secondary systems.

normally expected crevice conditions, but OD IGA/ SCC could possibly occur as a result of upsets or as a result of long-term fouling and accumulation of aggressive species in deposit-formed crevices. Alloy 800NG also has high resistance to OD IGA/SCC, but laboratory tests indicate that it is about twice as susceptible as Alloy 690TT, and it has experienced limited amounts of IGA/SCC in plants, while no operation-related corrosion of 690TT has been reported. Laboratory tests and some plant experience indicate that 600TT is significantly more resistant than 600MA but less resistant than 800NG and 690TT. Water chemistry selected to protect SG tubes appears to be satisfactory for most balance-of-plant (BOP) components such as turbines. The main corrosion concerns in the BOP that affect secondary system water chemistry are FAC of carbon steel piping, tubing, and heat exchanger internals and shells, and ‘ammonia’ attack of copper and copper alloy tubes. In addition, FAC has also affected some recirculating SG internal components (e.g., feedrings, swirl vanes). FAC is mainly influenced by the at-temperature pH and oxygen content around the secondary system. ‘Ammonia’ attack of copper alloys is mainly influenced by the concentrations of ammonia and oxygen at the copper alloy locations, but is also accelerated by increases in concentrations and pH associated with other amines, although not as strongly as by increases in ammonia. Once-through steam generators (OTSGs) have different thermal hydraulics and (in original SGs)

tube materials than recirculating steam generators (RSGs). These differences have led to OTSGs having somewhat different tube corrosion experience than RSGs of the same vintage. For the most part, OTSGs have experienced somewhat lower rates of tube degradation. However, significant IGA/IGSCC has been detected in the upper bundle free spans of several units, especially at scratches, and SG replacement has been performed or is planned at all units. The locations in SGs that are most affected by IGA/IGSCC are those where free circulation of secondary water is impeded by the local geometry, for example, in crevices formed by tube support plates or by sludge piles that can accumulate on the tube sheet. Impurities in the secondary water can concentrate in these locations by boiling and evaporation in a process called ‘hideout.’ The key issue influencing water chemistry regimes in PWR secondary system is to minimize SG degradation by controlling sludge buildup, reducing (and balancing, e.g., MRC) the concentration of impurities (i.e., sodium, chloride and sulfate) in deposits at the tube-tubesheet and tube-tube support plate interfaces. The use of advanced amines to control pH has increased significantly in the past few years, as discussed in a following section. Figure 30 shows the main approaches used in typical chemistry control strategies. Impurities are removed from SGs by blowdown of the coolant. Over the past 20 years or so, average

42

Water Chemistry Control in LWRs

Key issue: Mitigating IGA/IGSCC in concentrating regions

Approach: Control local chemistry

Molar ratio control

Reduce Na increase Cl

Reduce iron

Redox potential

Amines dispersants

Reduce Cu increase N2H4

Inhibitors

Boric acid TiO2

Figure 30 Pressurized water reactor secondary chemistry control strategies.

blowdown impurity concentrations in US SGs have been reduced from several ppb to the sub-ppb range. Many PWRs today have SG blowdown concentrations near or below the analytical detection limit. Minimization of impurities is recommended but has been insufficient to prevent or completely mitigate IGA/IGSCC at most plants with susceptible tube material and design, as it can result in sodium-rich feedwater. Cations such as sodium can be more effectively retained by boiling in a crevice than chloride. Hence, excess cations over anions or anions over cations result in specific corrosion issues because of concentration processes in local environments. The original all-volatile treatment used ammonia to control pH, but a less-volatile chemical than ammonia would improve pH control throughout the circuit. Early work employed morpholine, but now several other amines are used. Since the initial application of advanced amine chemistry about 15 years ago, there has been tremendous success in reducing the transport of corrosion products to the SGs by improving the attemperature pH around the BOP, especially in the two-phase regions. This has resulted in mitigation of FAC and thus reduced generation of corrosion products that ultimately get transported to the SGs. Ethanolamine (ETA) remains the most used amine at US plants, with 75% of the US plants using ETA or ETA with other amines, such as dimethylamine (DMA) or 3-methoxypropylamine (MPA), to control secondary cycle pH, as shown in Figure 31.17 Several plants now use a mixture of amines to achieve the optimum pH throughout the secondary system, with 25% of the US plants using MPA or MPA with other amines while 12% of the plants use morpholine or morpholine with other amines.

0%

4% 5%

4%

16% 56%

2% 7% 6%

ETA MPA

ETA/DMA ETA/MPA MPA/DMA MPA/Morph Morph/DMA

ETA/Morph Morph

Figure 31 Amines used in the secondary systems of US pressurized water reactors.

The proper control of oxygen in pressurized water reactor (PWR) secondary feedwater, using an oxygen scavenger such as hydrazine and/or carbohydrazide, has been an enduring issue. The requirements for oxygen concentration necessitate that some optimization take place. Maintaining reducing conditions – that is, low electrochemical potential – in the SG is essential to minimize SCC. On the other hand, some oxygen in the feedwater counteracts corrosion of

Water Chemistry Control in LWRs

1985

< 1992

1993

1994

1995

1996

1998

43

2005

% of US PWRs using advanced water chemistries 120 99

100 80 60

76 62

73

40 58

40 68

44

44

20 23

0

3

Using advanced amines

57

30

23

30 30

50

41

31 33

28

0

3

On molar ratio control

Using >100 ppb FW N2H4

35

19 37 41

17

Using boric acid treatment

Figure 32 Pressurized water reactor secondary chemistry trends.

carbon steel surfaces and the transport of corrosion products to the SG. Recent work has investigated the effect of hydrazine and oxygen on the ECP of SG tubing materials (Alloys 600 and 690) as well as stainless steel (304 and 316) and carbon steel during PWR startup conditions. These laboratory studies have shown that changes in the concentration of hydrazine, used to ensure a reducing potential in the SG, within the typical range allowed and employed (e.g., 20–150 ppb) have no discernable effect on FAC at feedwater temperatures (e.g., 180–235  C). Figure 32 shows the trends in using advanced water chemistry regimes in the secondary systems at US plants. 5.02.4.3

Control of Sludge Fouling of SGs

Corrosion products in the secondary side of PWR SGs primarily deposit on the SG tubes. These deposits can inhibit heat transfer, lead to thermal–hydraulic instabilities through blockage of tube supports, and create occluded regions where corrosive species can concentrate along tubes and within tube-to-tube support plate crevices. The performance of the SGs can be compromised not only through formation of an insulating scale, but also through the removal of tubes from service due to corrosion. Although the application of various amines to control the at-temperature pH (pHT) in specific locations of PWR secondary systems has been

successful in reducing the corrosion of BOP metals and thus reducing corrosion product transport to SGs, a complementary strategy now exists for significantly reducing SG fouling through online application of dispersant, which inhibits deposition. By inhibiting deposition of the corrosion products, the dispersant facilitates more effective removal from the SGs via blowdown. This strategy has been employed at fossil boilers for many decades. However, due to the use of inorganic polymerization initiators (containing sulfur and other impurities), polymeric dispersants had not been utilized in the nuclear industry. Only recently has a PAA dispersant been available that meets the criteria for nuclear application, and progress has been made in reducing SG fouling by application of an online dispersant to substantially improve the efficiency of blowdown iron removal. Dispersant application is proving to be a highly promising technology for markedly decreasing SG fouling, delaying (or possibly eliminating) the need for expensive chemical cleaning and effectively reducing the frequency of sludge lancing for SG maintenance. Online application of PAA to the feedwater system has been successfully demonstrated to greatly increase the efficiency of the blowdown system in eliminating feedwater corrosion products from fouling the SGs. The first application occurred at Arkansas Nuclear One Unit 2 for a three-month trial in early 2000, which demonstrated a significant improvement in the blowdown iron removal efficiency from 2% to 60% with 4–6 ppb PAA in

44

Water Chemistry Control in LWRs

With dispersant

Iron removal efficiency

100

10 Without dispersant

1 0

0.5 1 1.5 2 Dispersant concentration (ppb)

Figure 33 Iron removal efficiency during dispersant test at McGuire pressurized water reactor.

the feedwater. The second application, in 2005–2006, at McGuire Unit 2 for a 6–9 month trial in their replacement SGs tubed with Alloy 690TT showed a similar significant improvement., as shown in Figure 33.15 The following conclusions are evident from the McGuire 2 demonstration described in the above reference:  PAA is an effective dispersant. A feedwater PAA concentration approximately equal to the feedwater iron concentration (2 ppb in this case) appears to effectively remove approximately 50% of the influent feedwater iron under steady-state operating conditions.  Although blowdown copper spikes with initial PAA application (albeit to a much lesser extent than iron), it quickly returns to normal levels and remains there.  Filter element consumption is manageable.  Blowdown cation conductivity and ammonia behavior changed during the trial, but these changes are believed to be mainly due to changes in plant configuration and not PAA.  The SG thermal performance level has improved slightly with dispersant application, most likely due to slight beneficial changes in the tube deposit thermal properties. The long-term trial at McGuire 2 demonstrated the significant improvement in blowdown iron removal efficiency with application of PAA dispersant (a follow up to the successful short-term trial at ANO-2 in 2000). Based on the success of the McGuire 2 long-term trial, evaluations are in progress with SG vendors looking toward technical concurrence for long-term use in their fleet of recirculating SGs.

5.02.4.4

Lead Chemistry

PbSCC is a serious concern that can affect all SG tubing materials currently employed. A better understanding of lead behavior is needed at SG and feedwater temperatures before possible mitigation techniques can be successfully developed. It is well known that soluble lead at very low concentrations can contribute to SCC of nickel alloys. Likewise, it is well known that some locations on the secondary side of PWR SGs will accumulate lead in the solid state (i.e., deposit) with local concentrations considerably in excess of those observed to accelerate cracking in laboratory testing. Recent investigations using analytical transmission electron microscopy15 have identified lead in the cracks of many tubes pulled from PWRs. However, the absence of extensive operating SG tube failures at rates comparable to what might be predicted based on laboratory studies of PbSCC indicates that some mitigating phenomenon could be present.15 EPRI has published a sourcebook on lead29 that summarizes the state-of-knowledge regarding PbSCC and its effects on PWRs. It incorporates PbSCC laboratory testing, the current understanding of lead transport and other physical chemistry aspects of lead, and the accumulated industry experience regarding PbSCC and its mitigation. Three clearly understood and accepted facts regarding lead in PWR secondary water systems became clear as this sourcebook was being put together:  In laboratory testing, the presence of lead accelerates SCC of mill-annealed 600MA, stress-relieved 600SR, and thermally treated 600TT stainless steels as well as thermally treated Alloy 690 (690TT).  In operating PWRs, lead is present in the secondary system.  In two cases, a large ingress of lead to the secondary system has occurred as a result of lead blankets having been left behind in SGs; the tubes in the affected SGs cracked sooner and faster than in the other SGs at the same units. Set against these known facts are the following four points:  The mechanisms by which lead is transported from its ultimate source to the SG tube and into a crack are not well understood, and a comprehensive evaluation of possible mechanisms has not been performed.

Water Chemistry Control in LWRs

 The threshold concentration at which lead will accelerate SCC in SGs is not well defined.  No definitive indicator of PbSCC is available.  There is no well-characterized mechanism by which lead accelerates SCC. Recent work has shown that adsorption/desorption of Pb on corrosion products and SG tubing surfaces could potentially be a major sink/source, respectively, for Pb microscopy.15 There is no direct evidence of adsorption in SGs; however, there is sufficient potential for this mechanism that direct high-temperature measurements under SG conditions have been performed. As a result of ongoing laboratory studies, microscopy15 speculates that formation of a lead layer slows repassivation, after a passive film at the crack tip is disrupted, potentially to an extent to which a crack can initiate and propagate.

5.02.5 Chemistry Control for FAC in BWRs and PWRs FAC causes wall thinning of carbon steel piping, vessels, and components, as discussed in Chapter 5.06, Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels. The wall thinning is caused by an increased rate of dissolution of the normally protective oxide layer, for example, magnetite, that forms on the surface of carbon and low-alloy steels when exposed to highvelocity water or wet steam. The oxide layer reforms and the process continues. If the thinning is not detected in time, the reduced wall cannot withstand the internal pressure and other applied loads. The result can be either a leak or a complete rupture. The rate of wall loss (wear rate) of a given component is affected by temperature, fluid bulk velocity, the effect of component geometry on local hydrodynamics, the at-temperature pH, the liquid phase dissolved oxygen concentration, and the alloy composition. The addition of chromium to steels decreases the rate of FAC. Materials used to replace piping damaged by FAC include low-alloy steels containing chromium and molybdenum (P11, 1.25% Cr–0.5% Mo and P22, 2.25% Cr–1% Mo) and carbon or lowalloy steels clad with stainless steel. Corrosion models are used to estimate wall thinning and determine where monitoring is required. An example of the approach commonly used in the United States is described by Chexal and Horowitz.30

45

The main chemistry factors that affect the rate of FAC are pH and dissolved oxygen concentration. FAC is not an issue for PWR primary systems. As indicated earlier, laboratory studies have shown that changes in the concentration of hydrazine in the PWR secondary system feedwater, used to ensure a reducing potential in the SG, have no discernable effect on FAC at feedwater temperatures, within the typical range allowed and employed. The chemistry parameter that a BWR plant has some degree of control over is dissolved oxygen. Oxygen affects the form and solubility of the oxide layer, the dissolution of which is inherent in FAC. Several plants inject oxygen into the system, as the rate of FAC increases dramatically if the oxygen concentration is less than about 25 ppb. Plant data are shown in Figure 34. Use of HWC in a BWR can significantly reduce the amount of oxygen in the main steam, extraction steam, and heater drain systems, thus potentially increasing the FAC rates in these areas of the plant. The effect of NMCA on the corrosion behavior of carbon steel in 550  F (288  C) water containing various amounts of oxygen and hydrogen has been studied and the data confirm that there is no adverse effect of NMCA on FAC.7 The carbon steel segments of the BWR vessel bottom-head drain line have been identified as being FAC susceptible because of the flow conditions and the potential for low dissolved oxygen concentrations. However, a significant number of inspections have been performed recently at US plants and little thinning has been observed. The 2008 edition of the BWR Water Chemistry Guidelines7 recommends that feedwater oxygen should be maintained above 30 ppb to minimize FAC of carbon and low-alloy steels.

5.02.6 Water Chemistry Control Strategies Sometimes, step changes in chemistry strategy are unavoidable, as with the move to reducing chemistry in BWRs. In these cases, the operators must be prepared to deal with adverse effects. Some BWRs adopting reducing conditions experienced a large jump in out-of-core radiation fields, which may be avoided with prior zinc injection. Addition of new chemicals requires extensive qualification. For example, the successful demonstrations of BWR online noble

Water Chemistry Control in LWRs

Relative FAC wear rate (expressed as percentage of the average wear rate of components at 10 pbb oxygen)

46

160 140 Plant A 120

Plant A

100

Plant B Plant C

80

Plant D 60

Average

40 20 0 0

10

20

30

40

50

60

70

80

Dissolved oxygen (ppb) Figure 34 Plant data showing the relationship between flow-assisted corrosion and dissolved oxygen. (Oxygen values are localized, calculated by the CHECWORKS codes from measured values at condensate or feedwater locations.)

chemistry and PAA dispersants in PWR SGs resulted from detailed monitoring and evaluation during the first injections. If possible, changes in chemistry should be made in baby steps, with monitoring at each step, before further changes are implemented. Examples of this strategy are the gradual increases in lithium/pH and dissolved hydrogen in PWR primary systems. These incremental changes minimize adverse side-effects and allow a planned approach to the optimum plant-specific chemistry control program. The US nuclear power industry produces guidance documents to assist plant personnel in determining a plant-specific chemistry control strategy. The early versions of these documents, developed in the 1980s, listed water chemistry specifications and actions to be taken if the limits were exceeded. As more chemistry options became available, the guidelines evolved into providing guidance on selecting the most appropriate chemistry for a specific plant. Thus, the 2008 BWR Water Chemistry Guidelines7 offers recommendations on controlling ECP, zinc injection, and feedwater iron control. Likewise, the 2007 PWR Primary Water Chemistry Guidelines12 provides guidance on pH control and zinc injection, and the 2008 PWR Secondary Water Chemistry Guidelines27 discusses impurity control, amines, and dispersants. Theses documents are used by all US nuclear power plants and provide the technical basis for similar guidelines used in many other countries. Development of a strategic water chemistry plan, as discussed in these documents, is seen as crucial to controlling material degradation in the future.

References 1.

2.

3. 4. 5. 6.

7. 8. 9.

10.

11.

12.

Swan, T.; Wood, C. J. In Developments in Nuclear Power Plant Water Chemistry, VIIIth International Conference on Water Chemistry of Nuclear Reactor Systems, Oct 23–26, 2000; BNES: Bournemouth, UK, 2000. Fruzzetti, K.; Wood, C. J. In Developments in Nuclear Power Plant Water Chemistry. International Conference on Water Chemistry of Nuclear Reactor System, Jeju Island, Korea, Oct 23–26, 2006. Cohen, P. Water Coolant Technology of Power Reactors; Gordon and Breach: New York, 1969. Jones, R. L. In International Water Chemistry Conference, San Francisco, Oct 11–15, 2004; EPRI: Palo Alto, CA, 2004. Garcia, S.; Wood, C. Recent advances in BWR water chemistry. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. Cowan, R.; Hussey, D. Radiation field trends as related to chemistry in United States BWRs. In 2006 International Conference on Water Chemistry of Nuclear Reactor Systems, Jeju Island, Korea, Oct 23–26, 2006. EPRI. Boiling Water Reactor Water Chemistry Guidelines – 2008 Revision; EPRI: Palo Alto, CA, 2008. Edsinger, K. In Nuclear News; Tompkins, B., Ed.; 2008; pp 34–36. Fruzzetti, K.; Perkins, D. PWR chemistry: EPRI perspective on technical issues and industry research. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. Andresen, P.; Ahluwalia, A.; Hickling, J.; Wilson, J. Effects of PWR primary water chemistry on PWSCC of Ni alloys. In 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Whistler, Canada, Aug 19–23, 2007. Andresen, P.; Ahluwalia, A.; Wilson, J.; Hickling, J. Effects of dissolved H2 and Zn on PWSCC of Ni alloys. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. EPRI. Pressurized Water Reactor Primary Water Chemistry Guidelines: Revision 6; EPRI: Palo Alto, CA, 2007.

Water Chemistry Control in LWRs 13. Pathania, R.; Yagnik, S.; Gold, R. E.; Dove, M.; Kolstad, E. Evaluation of zinc addition to PWR primary coolant. In 7th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems, Breckenridge, CO, NACE: Houston, TX, 1995; pp 163–176. 14. Molander, A.; Jenssen, A.; Norring, K.; Ko¨nig, M.; Andersson, P.-O. Comparison of PWSCC initiation and crack growth data for Alloy 600. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. 15. Fruzzetti, K.; Rochester, D.; Wilson, L.; Kreider, M.; Miller, A. Dispersant application for mitigation of steam generator fouling: Final results from the McGuire 2 long-term trial and an industry update and EPRI perspective for long-term use. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. 16. Haas, C.; Ahluwalia, A.; Kucuk, A.; Perkins, D. PWR operation with elevated hydrogen. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. 17. Hussey, D.; Perkins, D.; Choi, S. Benchmarking radioactivity transport and deposition in PWRs. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. 18. Stevens, J.; Bosma, J. Elevated RCS pH program at Comanche peak. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. 19. Nordmann, F. Worldwide chemistry objectives and solutions for NPP. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008.

20.

21.

22. 23. 24.

25. 26. 27. 28. 29. 30.

47

Tigeras, A.; Stellwag, B.; Engler, N.; Bretelle, J.; Rocher, A. Understanding the zinc behavior in PWR primary coolant: A comparison between French and German experience. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. Frattini, P. L.; Blok, J.; Chauffriat, S.; Sawicki, J.; Riddle, J. In VIIIth International Conference on Water Chemistry of Nuclear Reactor Systems, Oct 23–26, 2000; BNES: Bournemouth, UK, 2000. Riess, R. Personal communication, 2008. Byers, W.; Wang, G.; Deshon, J. Limits of zinc addition in high duty PWRs. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. Perkins, D.; Ahluwalia, A.; Deshon, J.; Haas, C. An EPRI perspective and overview of PWR zinc injection. In VGB NPC’08 Water Chemistry Conference, Berlin, Sept 14–18, 2008. Millett, P. J; Hundley, F. Nucl. Energ. 1997; 36, pp 251–258. EPRI. Personal communication from K. Fruzzetti, 2009. EPRI. Pressurized Water Reactor Secondary Water Chemistry Guidelines – Revision 6; EPRI: Palo Alto, CA, 2008. Choi, S. Personal communication, 2009. EPRI. Pressurized Water Reactor Lead Sourcebook; EPRI: Palo Alto, CA, 2006. Chexal, V.; Horowitz, J. Chexal–Horowitz flowaccelerated corrosion model – Parameters and influences. In ASME PVP-Vol B, Current Perspectives of International Pressure Vessels and Piping Codes and Standards, Book No. H0976B, 1995.

5.03

Corrosion of Zirconium Alloys

T. R. Allen University of Wisconsin, Madison, WI, USA

R. J. M. Konings European Commission, Joint Research Centre, Institute for Transuranium Elements, Karlsruhe, Germany

A. T. Motta The Pennsylvania State University, University Park, PA, USA

ß 2012 Elsevier Ltd. All rights reserved.

5.03.1 5.03.2 5.03.2.1 5.03.2.2 5.03.2.3 5.03.3 5.03.3.1 5.03.3.2 5.03.3.3 5.03.3.4 5.03.3.4.1 5.03.3.4.2 5.03.3.4.3 5.03.4 5.03.5 5.03.5.1 5.03.5.2 5.03.5.3 5.03.5.4 5.03.5.5 5.03.6 5.03.7 References

Introduction General Considerations Oxidation Hydrogen Uptake Controlling Factors for Corrosion Uniform Oxidation Mechanism Temperature and Heat Flux Coolant Chemistry Irradiation Effects Radiolysis Irradiation effects in the oxide layer Changes in the metallurgical state of the metal Nodular Oxidation Hydrogen Embrittlement Hydrogen Production During Aqueous Corrosion of Zirconium-Base Materials Hydrogen Absorption Hydride Formation Hydride Formation Rates Formation of Hydride Rim Delayed Hydride Cracking Summary and Outlook

Abbreviations BWR CANDU CRUD DHC IAEA M5TM PWR tHM VVER ZIRLOTM Zry

Boiling water reactor Canadian Deuterium Uranium Chalk River unidentified deposits Delayed hydride cracking International Atomic Energy Agency Zirconium alloy material with niobium (AREVA) Pressurized water reactor Ton heavy metal Voda Voda Energy Reactor Zirconium alloy material with niobium, tin, and iron (Westinghouse) Zircaloy

49 50 50 51 52 53 53 57 57 59 59 60 60 61 61 62 62 62 63 64 65 66 66

5.03.1 Introduction Zirconium alloys are widely used for fuel cladding and in pressure tubes, fuel channels (boxes), and fuel spacer grids in almost all water-cooled reactors: light water reactors such as the pressurized water reactor (PWR) and the boiling water reactor (BWR) as well as the Canadian designed Canadian Deuterium Uranium (CANDU) heavy water reactor. Since its employment in the first commercial nuclear power plant (Shippingport) in the 1960s, Zircaloy, a zirconium–tin alloy, has shown satisfactory behavior during many decades. However, degradation due to waterside corrosion can limit the in-reactor design life of the nuclear fuel. The critical phenomenon is the

49

50

Corrosion of Zirconium Alloys

hydrogen ingress into the cladding during corrosion, which can cause cladding embrittlement. As utilities are striving to achieve higher fuel burnups, the nuclear industry has made several efforts to understand the mechanisms of corrosion and to mitigate its effects. In striving for increased burnup of the nuclear fuel from 33 000 to 50 000 MWd/tHM and beyond in PWRs, associated studies have shown that the corrosion of the Zircaloy-4 cladding accelerates under these higher burnup conditions. Although alloys that are more modern have not yet shown evidence of this high-burnup acceleration, this is a potential concern. Also, the efforts to increase the thermalcycle efficiency in PWRs by operating at higher temperatures (power uprates), combined with the more aggressive chemistry (introduction of B and Li for example) related to the use of high-burnup fuel, have resulted in increased fuel duty,1 and in increased corrosion rates. This has led to the introduction of cladding tubes of new zirconium alloys such as zirconium–niobium, which are much more corrosion resistant.2,3 With the introduction of these materials, the nuclear industry aims at zero tolerance for fuel failure in the future.4 Many reviews on the corrosion of zirconium alloys both out- and in-reactor, have been published.5–11 The extensive reviews made by an international expert group of the International Atomic Energy Agency (IAEA) and published as IAEA-TECDOCs 684 and 99612,13 are major references in this respect. As mentioned by Cox,6,7 ‘‘the number of publications on this topic is so enormous that it is impossible for a short review to be comprehensive.’’ This also applies to the current chapter, which therefore focuses on the main issues, naturally relying on the above-mentioned existing reviews and updating the information where possible with new results and insights.

protective, thus limiting the access of oxidizing species to the bare metal. Much evidence exists to indicate that Zr oxidation occurs by inward migration of oxygen ions through the oxide layer, either through grain boundaries or through the bulk.5,12,13 Zr þ O2 ¼ ZrO2 As shown in Figure 1, the growth of the oxide layer on the metal surface depends on the kinetics of the oxygen diffusion through this layer. Because the corrosion kinetics slow down as the oxide thickness increases, it has been argued that the rate controlling step in the oxidation process is the transport of atomic species in the protective oxide, by either oxygen diffusion through the oxide film14,15 or diffusion of electrons through the oxide film. These processes are necessarily coupled to maintain electroneutrality. Electron transport is, however, difficult in zirconium dioxide, as it is an electrical insulator when undoped. Although this is not positively confirmed, it is likely that the role of doping elements in the determination of corrosion kinetics is done through their influence on the electron or oxygen transport in the oxide layer. Several types of corrosion morphologies have been observed in nuclear reactors and in autoclave experiments, of which the most important are 1. Uniform: The formation of a thin uniform layer of zirconium dioxide on the surface of a zirconium alloy component (see Figure 2). 2. Nodular : The formation of local, small, circular zirconium oxide blisters (see Figure 3). 3. Shadow: The formation of local corrosion regions that mirror the shape (suggestive of a shadow) of other nearby noble reactor core components (Figure 4).

H2O ® O2− + 2H+

Coolant

5.03.2 General Considerations 5.03.2.1

Oxidation

Corrosion of zirconium alloys in an aqueous environment is principally related to the oxidation of the zirconium by the oxygen in the coolant, dissolved or produced by radiolysis of water. A small amount of oxygen can be dissolved in the metal, but once the thermodynamic solubility limit is exceeded, ZrO2 is formed on the metal. (All zirconium components normally have a thin oxide film (2–5 nm) on their surface in their as-fabricated state.) The oxide formed is

H+ O2−

H+

Oxide H+ + e

®

Zr + 2O2− ® ZrO2 + 4e−

H0

Metal

Figure 1 Schematic presentation of the corrosion of the zirconium alloys. Corrosion of zirconium alloys in nuclear power plants; TECDOC-684; International Atomic Energy Agency, Vienna, Austria, Jan 1993.

Corrosion of Zirconium Alloys

51

Zr + H2O = ZrO2 + H2

ZrH2−x ZrO2

Figure 2 Uniform oxide layer formation and hydride precipitation in Zircaloy cladding. © European Atomic Energy Commission.

The occurrence of these morphologies is strongly dependent on the reactor operating conditions and chemical environment (particularly the concentration of oxygen in the coolant), which are distinctly different in PWRs, BWRs, and CANDU (Table 1). In both BWRs and PWRs, a uniform oxide layer is observed, although its thickness is normally greater in PWR than in BWR, primarily because of the higher operating temperature. Nodular corrosion occurs occasionally in BWRs because a much higher oxygen concentration occurs in the coolant because of water radiolysis and boiling. Shadow corrosion is also occasionally observed in BWRs and is a form of galvanic corrosion.

1 mm

5.03.2.2

100 μm

Figure 3 General appearance of nodules formed on zirconium alloy following a 500  C steam test at 10.3 MPa. In the bottom, a cross-section view of a nodule is shown, exhibiting circumferential and vertical cracks. Photo courtesy of R. Ploc and NFIR (Nuclear Fuel Industry Research Group). Reproduced from Lemaignan, C.; Motta, A. T. Zirconium Alloys in Nuclear Applications, Materials Science and Technology, Nuclear Materials Pt. 2; VCH Verlagsgesellschaft mbH, Weinheim, Germany, 1994.

Hydrogen Uptake

The formation of an oxide layer would not bring severe consequences to cladding behavior were it not for the fact that in parallel with the corrosion process, a fraction of the hydrogen, primarily produced by the oxidation reaction as well as by radiolysis of water, diffuses through the oxide layer into the metal. Zirconium has a very low solubility for hydrogen (about 80 wt ppm at 300  C and 200 wt ppm at 400  C) and once the solubility limit is exceeded, the hydrogen precipitates as a zirconium hydride phase (Figure 2): ZrðH; slnÞ þ H2 ¼ ZrH1:6 or ZrH2 As a result, the following effects have been reported (although not all confirmed) to occur in the cladding: hydrogen embrittlement due to excess hydrogen or its localization into a blister or rim,16,17 loss of

52

Corrosion of Zirconium Alloys

fracture toughness, delayed hydride cracking (DHC), and acceleration of corrosion and of irradiation growth. Hydrogen embrittlement impacts the mechanical resistance of the Zircaloy cladding to

failure and it is thus of key importance to understand its underlying mechanisms. The ductility reduction due to hydrogen embrittlement is dependent on the volume fraction of hydride present, the orientation of the hydride precipitates in the cladding, and their degree of agglomeration.18,19

Oxide

5.03.2.3

Oxide

(b)

(a) Figure 4 Zirconium oxides near (b) and away from (a) a stainless steel control blade bundle, showing the effect of shadow corrosion. Reproduced from Adamson, R. B.; Lutz, D. R.; Davies, J. H. Hot cell observations of shadow corrosion phenomena. In Proceedings Fachtagung der KTG-Fachgruppe, Brennelemente und Kernbautelle, Forschungszentrum Karlsruhe, Feb 29–Mar 1, 2000. Table 1

Controlling Factors for Corrosion

The oxidation and hydrogen uptake of Zircaloy is of course determined by many factors. First of all, the chemical and physical state of the material: composition, metallurgical condition, and surface condition. These conditions are often specific to the material and sometimes batch-specific and also related to the fabrication process, as discussed in detail in Chapter 2.07, Zirconium Alloys: Properties and Characteristics. This is evident from the different behavior of Zircaloy and Zr–Nb alloys, as shown in Figure 5 for two different zirconium alloys employed in the French PWRs, Zircaloy and Zr1% Nb (M5). The peak oxide layer thickness of Zircaloy-4 (oxide thickness at the hottest fuel grid span) increases significantly with burnup (i.e., residence time in the reactor), whereas that of Zr1%Nb shows a moderate increase. In addition, a number of environmental factors affecting the corrosion of zirconium alloys must be considered: 1. Coolant Chemistry: It is obvious that the dissolved oxygen and hydrogen play a major role in the corrosion process, but other dissolved species must also be taken into account. To control the pH of the coolant at slightly alkaline conditions,

Typical reactor environments to which the zirconium alloys are exposed

Coolant Inlet temperature ( C) Outlet temperature ( C) Pressure (MPa) Neutron fluxa (n cm2 s1) Coolant chemistry [O2] (ppb) [H2] (ppm) pH B (as H3BO3) (ppm) Li (as LiOH) (ppm) Na (as NaOH) (ppm) K (as KOH) (ppm) NH3 (ppm)

BWR

PWR

VVER

CANDU

H2O 272–278 280–300 7 4–7  1013

H2O 280–295 310–330 15 6–9  1013

H2O 290 320 15 5–7  1013

D2O 255 300 10 2  1012

200 0.03 7 – – – – –

<0.05 2–5 6.9–7.4 0–2200 0.5–5 – – –

< 0.1 –

<5 0.5–1 10.2–10.8 – 1 – – –

0–140016 0.05–0.6 0.03–0.35 5–20 6–30

a E > 1 MeV. Corrosion of zirconium alloys in nuclear power plants; TECDOC-684; International Atomic Energy Agency, Vienna, Austria, Jan 1993.

Corrosion of Zirconium Alloys

53

60 50 Oxide thickness (mm)

M5 Zirc-4

40 30 20 10 0 0

20

40

60

80

Burnup (MWd kgU–1) Figure 5 Peak oxide layer thickness as a function of burnup for Zircaloy-4 and Zr1%Nb (M5). Reproduced from Bossis, P.; Peˆcheur, D.; Hanifi, K.; Thomazet, J.; Blat, M. J. ASTM Int. 2006, 3(1), Paper ID JAI12404.

LiOH is added and H3BO3 (boric acid) is added for reactivity control in PWRs. Furthermore, impurities (Cl, F) and coolant-borne species (Cu, Ni, etc.) must be considered. 2. Radiation: In reactor, the Zircaloy and the coolant are subjected to the effects of energetic particles. The principal effect is the production of oxidizing species such as O2 in the coolant. 3. Temperature : In the range of water reactor operation (240–330  C), the combined effect of temperature and radiation on zirconium alloy oxidation and hydriding have been characterized extensively, varying from almost no effect to acceleration of oxidation by factors of up to two orders of magnitude, depending on environment and radiation level. 4. In addition, the presence of boiling and CRUD (the term CRUD stands for Chalk River unidentified deposits, the nuclear power plant in which the effect was observed for the first time) deposition in PWR can enhance corrosion.

5.03.3 Uniform Oxidation 5.03.3.1

Mechanism

Uniform corrosion is defined as a process that occurs approximately with the same speed on the entire surface of an object (ISO 8044). It can be considered as an electrochemical cell process, in which the metal is anodically oxidized: Zr þ 2O2 ¼ ZrO2 þ 2Vo þ 4e O

o where VO indicates a lattice vacancy in the ZrO2 layer. The corresponding cathodic reaction at the oxide/coolant interface can be the reduction of water:

2H2 O þ 4e þ 2Vo ¼ 2O2 þ 4H O

or, when the water contains dissolved oxygen: 2H2 O þ O2 þ 2VOo þ 4e ¼ 4OH The oxygen ions diffuse preferentially via the oxide crystallite boundaries to the oxide/metal interface, whereas the vacancies diffuse in the opposite direction. The hydrogen can combine with electrons to form atomic/molecular hydrogen that dissolves in the coolant water or diffuses to the metal. Uniform corrosion is a passivating event since a protective layer of zirconium oxide is formed as a result of the reaction with the O2 ions or the OH radicals. Electron microscopy shows that the oxide layer is microcrystalline, initially equiaxed, later growing into columnar grains that are formed in a dense packing, of which the mean crystallite size increases as the oxide thickens.15 Figure 6 shows typical microstructures of the oxide layer for several zirconium-based cladding materials. Figure 6(c), in particular, shows the columnar grains extending right near the oxide/metal interface. The corrosion kinetics have been studied extensively. As mentioned above, because the corrosion rate slows down with oxide thickness, the rate controlling step is thought to be the transport of oxidizing species in the layer.15,20 During corrosion,

54

Corrosion of Zirconium Alloys

Zircaloy-4

ZIRLO

(a)

Zr–2.5Nb

(b)

(c)

150 nm

150 nm

150 nm

Figure 6 Grain size, shape, and orientation comparison near the oxide/metal interface of (a) Zircaloy-4, (b) ZIRLO, and (c) Zr–2.5Nb alloy oxides formed in 360  C pure water environments. The hand-drawn sketch below each bright-field image shows oxide crystallite grain boundaries. Black arrows indicate oxide growth direction. Reproduced from Yilmazbayhan, A.; Breval, E.; Motta, A. T.; Comstock, R. J. J. Nucl. Mater. 2006, 349, 265–281.

a potential develops across the oxide layer. The negative potential at the oxide/metal interface accelerates the electron migration process and retards the O2 diffusion until both operate at the same rate. Bossis et al.22 argue that the surface reactions are rate-determining in some Nb-containing alloys. The measurements of the weight gain kinetics for zirconium and its alloys (the weight gain is due to oxygen ingress and follows the overall corrosion kinetics) were found to fall into two stages, referred to as pre- and posttransition. For constant temperature and pressure, the pretransition corrosion kinetics are independent of pH between about 1 and 13 (if no specifically aggressive species such as LiOH are present) and of the source of the oxygen. The kinetics of the pretransition oxide layer formation, as measured from weight gain (DW ), have been found to approximately follow a cubic rate law21: 3

ðDW Þ ¼ k1 t

½1

where k1 is the preexponential factor and t is time. More recent results have shown that the rate law

depends on the alloys according to (DW)n ¼ kt, with n between 2 and 5.22 The temperature dependence of k1 follows an Arrhenius-type equation:   Q1 ½2 k1 ¼ B1 exp RT where B1 is an empirical constant, R is the universal gas constant, T is the absolute temperature, and Q1 is the activation energy for pretransition oxidation. The values for B1 are Q1 are obtained empirically from fitting of experimental data, for example, B1 ¼ 6.36  1011 (mg dm2)3 per day and Q1/R ¼ 13640 K, as found by Kass21 for Zry-2 and Zry-4. The posttransition kinetics, on the contrary, are approximately linear (n ¼ 1) in time23: DW ¼ k2 t þ C with k2 ¼ B2 exp

  Q2 RT

½3 ½4

and C the weight gain at transition. B2 is the empirical constant and Q2 the activation energy for posttransition

Corrosion of Zirconium Alloys

oxidation. Hillner et al.23 discussed the results of numerous analyses of experimental corrosion studies on Zircaloy with varying time and temperature to derive B2 and Q2. As discussed by these authors, most studies suffer from paucity of data for extended exposures. Their own results for Zry-2 and Zry-4 cover a wide range of time and weight gain and the posttransition kinetics were interpreted to consist of two linear stages (with a change at about 400 mg dm2 or about 30 mm) with B2 ¼ 2.47  108 mg dm2 day1 and Q2/R ¼ 12880 K for stage 1, and B2 ¼ 3.47  107 mg dm2day1 and Q2/R ¼ 11452 K for stage 2. Whether or not Hillner’s interpretation of a change in mechanism is correct, certainly the data is best described by a two-stage empirical fit. A schematic representation of these pre- and posttransition kinetics is shown in Figure 7 as the dashed lines. Also shown in this graph is the more recent view that three stages can be discriminated for zirconium alloy corrosion processes23: 1. The early pretransition regime, characterized by the formation of a thin, black, tightly adherent corrosion film that grows thicker in accordance with a nearly cubic rate law. 2. The intermediate stage that lies between the preand posttransition stages. As initially shown by Bryner,24 this region appears to comprise a series of successive cubic curves, similar to the initial cubic kinetic curve. This linear rate results from the superposition of various regions of the oxide layer following pretransition growth rate but slightly out of phase with each other.

Oxide thickness

Posttransition (linear) Transitory (cyclic)

Pretransition (cubic)

Time

Figure 7 Schematic representation of the zirconium alloy corrosion showing the pretransition, transitory, and posttransition regions. The dashed lines indicate early models that recognized only the pre- and posttransition regimes. Reproduced from Hillner, E.; Franklin, D. G.; Smee, J. D. J. Nucl. Mater. 2000, 278, 334.

55

3. The linear posttransition kinetic regime. In the very early stages of the oxide formation, the layer is dense and composed of grains that have a predominantly tetragonal or cubic structure. As the grains grow, columnar grain growth is established and the tetragonal grains tend to transform to monoclinic oxide, which constitutes the majority of the oxide formed.20 Although the tetragonal phase has often been associated with protective behavior, this correlation is noncausal and in fact, oxides with lower tetragonal fraction have been found to be more protective.26,27 The diffusion of oxygen takes place along the grain boundaries in the oxide layer,4 the kinetics of which are given by eqn [1]. The size of the columnar grains and their grain-to-grain misorientation (Figure 6) have been related to the transition thickness. Studies of Zircaloy corrosion in autoclaves clearly reveal the cyclic corrosion kinetics,20,24 the oxide layer appearing to be composed of successive layers of 2–3 mm thickness (Figures 8–10), for which the oxidation kinetics progressively decrease as a result of the growth of the oxide layer, in accordance with eqn [1]. The cycles are separated by transitions during which the kinetics appears to accelerate. The transitions are caused by the destabilization of the oxide layer, as a result of which the passivating layer becomes porous and fractured at the end of the cycle, losing its protective role, and reopening for rapid oxidation. A new oxidation cycle then starts. Several processes have been suggested for the destabilization of the oxide layer, such as7,25–27: (a) Cracking of the oxide as a result of the accumulation of compressive stresses in the oxide from imperfect accommodation of the volume expansion attendant upon oxide formation. (b) Cracking of the oxide as a result of the transformation of initially tetragonal ZrO2 to the monoclinic modification,10 or as a result of the oxidation of intermetallic precipitates initially incorporated in metallic form, both of which result in a volume increase. (c) The porosity formed in the oxide reaches a percolation condition, leading to easy access of the coolant to the underlying metal. The first factor is normally considered to be the main driver, although the other factors have also been proposed to contribute. The levels of stress accumulation depend on the phase transformation tensor (various levels of accommodation of the PillingBedworth strains in the in-plane directions), which

56

Corrosion of Zirconium Alloys

Oxide thickness (μm)

15 12 Zircaloy-4 9 M5

6 3 0

0

200

400

600

800

Time (days) Figure 8 Results of oxidation tests of Zircaloy-4 and of M5™ in autoclaves, at 360  C, with 10 ppm Li and 650 ppm B, showing the cyclic nature of the oxidation. Redrawn from Bataillon, C.; Fe´ron, D.; Marchetti, L.; et al. E-DEN Monograph ‘‘Corrosion’’ Commissariat a` l’E´nergie Atomique; 2008.

Zr4

Zr4

20 mm

20 mm

ZIRLO

ZIRLO

Figure 9 Optical micrographs of oxide layers formed in Zircaloy-4 and in ZIRLO™, in reflected (left) and transmitted light (right). The regular periods formed during the cyclic corrosion process correspond to the oxide transitions in the two alloys. Photo courtesy of G. Sabol, Westinghouse Electric Co.

has been shown to vary from alloy to alloy, thus likely causing the consistent differences seen among the oxide thicknesses at transition for various alloys. Thus, each alloy has a reproducible transition thickness in a given environment. This cyclic process has been shown to reproduce itself with remarkable regularity upward of 17 transitions,26,27 as shown in Figure 9. This can also be seen in the SEM micrograph in Figure 10 which suggests that cracking occurs at transition.

As discussed by Battaillon et al.,25 the kinetics of the cyclic process can be described by a succession of equations similar to [1] and [2], each representing a specific cycle. The length of the cycle seems to be material dependent as shown in Figure 8. Also, Zircaloy contains second phase precipitates of Zr(Cr, Fe)2 and tin as a dissolved element (see Chapter 2.07, Zirconium Alloys: Properties and Characteristics). The intermetallic precipitates are known to have a

Corrosion of Zirconium Alloys

57

1000

Weight gain (mg dm−2)

800

10 μm Figure 10 The oxide layer formed on M5™ in autoclaves at 360  C, with 10 ppm Li and 650 ppm B dissolved in the water showing the layered nature of the oxide, with periodic cracking. Bataillon, C.; Fe´ron, D.; Marchetti, L.; et al. E-DEN Monograph ‘‘Corrosion’’ Commissariat a` l’e´nergie atomique, 2008. From DEN Monographs ‘‘Corrosion and Alteration of Nuclear Materials,’’ ISBN 978-2-281-11369-3 (2010), e´ditions du Moniteur, © CEA.

600

400

200

0

280

300

320 Temperature (⬚C)

340

360

Figure 12 The effect of temperature on the oxidation kinetics of Zircaloy-4, as derived from autoclave test in water for 2500 days. Reproduced from Hillner, E.; Franklin, D. G.; Smee, J. D. J. Nucl. Mater. 2000, 278, 334.

experimentally. As shown in Figure 12, the corrosion kinetics accelerate above about 310  C. An increase of 5  C for a typical cladding temperature of 335  C results in a 26% increase in weight gain. The temperature of the metal–oxide interface (Ti) is, however, not only dependent on the temperature of the coolant, but also on the heat flux (f in W cm2):

higher oxidation resistance than the zirconium matrix.28,29 When the oxidation of the zirconium progresses, the Zr(Cr,Fe)2 precipitates are incorporated in metallic form into the oxide layer (Figure 11). However, the iron is progressively dissolved in the zirconium oxide. Tin is present in the oxide layer as nanoparticles of b-Sn, SnO, or Sn(OH)2. The slower oxidation kinetics of Zr–Nb alloys have been attributed to the absence of the second phase precipitates.7

fe l where Ts is the temperature at the water–oxide layer boundary, e the oxide layer thickness (in cm), and l the thermal conductivity of the oxide layer (W cm1 K1). Considering that zirconium oxide is a poor thermal conductor, the oxide layer will act as an insulator increasing the temperature of the metal–oxide interface. For typical values for a PWR (f ¼ 55 W cm2) and a thermal conductivity of 0.022 W cm1 K1, the interface temperature increases 1 K for an oxide layer of 4 mm.25 As a related effect, nucleate boiling can occur at the oxide–water boundary, once this boundary reaches the saturation temperature (344.5  C at 15.5 MPa in a PWR). As a result, an enrichment of Li in the liquid phase near the oxide–water boundary can occur (Figure 13), which can reach a factor of 3.25 This is not expected to increase the corrosion significantly for conditions typical for PWRs.

5.03.3.2

5.03.3.3

100 nm Figure 11 Zr(Cr,Fe)2 precipitates incorporated in metallic form into the oxide layer on Zircaloy-4. Adapted from Pecheur, D.; Lefebvre, F.; Motta, A. T.; Lemaignan, C.; Charquet, D. Oxidation of Intermetallic Precipitates in Zircaloy-4: Impact of Irradiation. In 10th International Symposium on Zirconium in the Nuclear Industry, ASTM STP 1245; Baltimore, MD, 1994; 687–70; Pecheur, D.; Lefebvre, F.; Motta, A. T.; Lemaignan, C.; Wadier, J. F. J. Nucl. Mater. 1992, 189, 2318–332.

Temperature and Heat Flux

An increase of temperature increases the oxidation kinetics, as is evident from eqn [1], and confirmed

Ti  Ts þ

Coolant Chemistry

The corrosion of Zircaloy is influenced by the chemical composition of the coolant. The PWR coolant

58

Corrosion of Zirconium Alloys

activation.) In addition, the coolant may contain small concentrations of anionic impurities that play a role in the corrosion mechanism (Figure 14). Extensive research has been performed to understand the role of lithium hydroxide and boric acid on the kinetics of the corrosion of zirconium alloys. Experiments in autoclaves have shown that the rate of oxidation of Zircaloy-4 increases significantly when boric acid is absent.25 After an initial stage where the corrosion kinetics are as expected, corrosion is accelerated in conjunction with a decrease of the thickness of the protective oxide layer,30,31 as derived from microscopic observations, especially by the ingress of Li into the oxide (Figure 15). Enhanced dissolution of the crystallite grain boundaries has been suggested as the mechanism.32 This effect was absent in the presence of boric acid, and no significant difference was observed for the oxidation kinetics for LiOH concentrations between 70 and 1.5 ppm (Figure 14). The protective effect of boric acid has been suggested to be related to the plugging of the porosity in the oxide by a borate compound.33 The coolant chemistry also influences the solubility of coolant-borne metallic impurities (e.g., iron, nickel, copper, etc. arising from corrosion release from circuit surfaces), which may deposit on fuel rod surfaces as CRUD, which is composed of metal oxides such as Fe2O3 (hematite), Fe3O4 (magnetite), FeOOH (goethite), or (Ni,Co)xFe3-xO4 (spinel).34–36 Such CRUD deposits are occurring specifically at positions with sub-cooled boiling and may have, in some cases, appeared to contribute to accelerated

contains boron and lithium. Boron, present as boric acid (1000–2000 ppm at the beginning of the cycle, depending on the cycle length, and about zero at the end of the cycle), is added to control the core reactivity through neutron absorption of 10B. The boric acid is weakly dissociated, particularly at high temperature, which could lead to a slightly acidic environment. To counteract this, small quantities of lithium hydroxide (5–10 ppm) are added in the water, to obtain a slightly alkaline pH, to avoid deposition of corrosion products on the cladding and limit the corrosion of core structures made of stainless steel or Inconel alloys. (Lithium enriched over 99% of 7Li is used, as the use of 6 Li produces the undesirable tritium through Oxide

Water

Steam bubble

Enrichment of species of low volatility

Figure 13 Schematic representation of the enrichment of species at the oxide–water boundary during nucleate boiling. Adapted from DEN Monographs ‘‘Corrosion and Alteration of Nuclear Materials,’’ ISBN 978-2-281-11369-3 (2010), e´ditions du Moniteur, © CEA. 20 70 ppm Li (B=0)

18

Oxide thickness (μm)

16 10 ppm Li 650 ppm B

14 12

3.5 ppm Li 1000 ppm B

10 70 ppm Li 650 ppm B

8

1.5 ppm Li 650 ppm B

6 4 2 0 0

100

200

300

400

500

Time (days) Figure 14 The effect of Li and B on the oxidation kinetics of Zircaloy-4. Bataillon, C.; Fe´ron, D.; Marchetti, L.; et al. E-DEN Monograph ‘‘Corrosion’’ Commissariat a` l’E´nergie Atomique, 2008. From DEN Monographs ‘‘Corrosion and Alteration of Nuclear materials,’’ ISBN 978-2-281-11369-3 (2010), e´ditions du Moniteur, © CEA.

Corrosion of Zirconium Alloys

increased significantly above the coolant’s nominal level, the increased corrosion caused by CRUD deposits is thought to be due to the role of Li, in combination with the increased metal/oxide interface temperature.37 Fluorine is produced in the coolant water by neutron capture in 18O to give 19F. Laboratory experiments have shown that the corrosion of Zircaloy-4 begins to accelerate between 19 and 190 ppm fluorine at 360  C, which is well above the coolant specification in most reactors (0.15 ppm).

16 14 12 Oxide thickness (mm)

59

10 8 6 4

5.03.3.4

2

A wealth of information exists on the in-reactor behavior of Zircaloy from worldwide fuel monitoring programs as well as from experimental research programs, from which information about the radiation effects on zirconium alloy corrosion can be deduced. These effects can be of multiple origin and include radiolysis of the coolant, changes in the metallurgical condition, displacement damage, or phase transformations. As discussed by Cox,7 no in-reactor effects are evident for an oxide layer thickness below 5–6 mm. Above that thickness, a departure from data with radiation field and for in-reactor conditions suggests an irradiation-induced acceleration of the oxide breakdown. Bataillon et al.25 suggest a factor of 2 for the oxidation rate of Zircaloy-4 between in-reactor and autoclave experiments during the first two reactor cycles. This increases to about 4 during cycles 5 and 6. This effect is shown in Figure 17, in which the thickness of the oxide layer on Zircaloy-4 as a function of exposure time is compared for three cases: (a) autoclaves without thermal gradient, (b) corrosion loop with thermal gradient generated by an electrical heating inserted in the cladding tube, and (c) in-reactor, with the effects of thermal gradient and irradiation.38

0 0

50

100

150

200

Time (days) Figure 15 The evolution of the microstructure of the oxide layer on Zircaloy-4 after oxidation in an autoclave with 70 ppm Li, without boron (360  C). The blue line shows the total oxide thickness, whereas the red line shows the thickness of the protective inner layer. Reproduced from Bataillon, C.; Fe´ron, D.; Marchetti, L.; et al. M. E-DEN Monograph ‘‘Corrosion’’ Commissariat a` l’E´nergie Atomique; 2008.

Loose deposit Adherent deposit

Oxidized cladding Figure 16 Schematic representation of the CRUD morphology.

oxidation of both BWR and PWR cladding. CRUD can have a wide variety of morphologies, from dense to porous, thus having very different thermal conductivity. The CRUD structure can generally be described an inner deposit that is tightly adherent to the oxide layer and an outer deposit that has a loose structure (Figure 16). However, the thermal conductivity of CRUD is generally better than that of zirconium oxide and therefore its added effect on deterioration of the heat flux through the corrosion layer rarely results in excessive cladding temperatures. Because the lithium concentration in the CRUD, where it is deposited as lithium borate, is

Irradiation Effects

5.03.3.4.1 Radiolysis

The ionizing radiation will interact with water molecules producing a variety of reaction products: hv H2 O ! e ; H; OH; H2 O2 ; H2

Figure 18(a) shows the results of a typical computer simulation of the speciation as a function of time.15 As one can see, numerous oxidizing species such as O2, O., HO2, and H2O2 that could accelerate the corrosion are formed. For this reason, the coolant water in PWRs is hydrogenated. This effect is

60

Corrosion of Zirconium Alloys

shown in Figure 18(b), which indicates that the presence of hydrogen significantly reduces the steady-state concentration of the oxidizing species. 5.03.3.4.2 Irradiation effects in the oxide layer

Oxide thickness (mm)

Because the growth of the oxide layer on zirconium alloys is strongly related to the diffusion of oxygen ions through the layer, as discussed above, the displacement

5.03.3.4.3 Changes in the metallurgical state of the metal

PWR, 346 ⬚C 40

Corrosion loop, 346 ⬚C Autoclave, 354 ⬚C

0 0

300

600

Time (days)

Figure 17 The thickness of the oxide layer on Zircaloy-4 as a function of exposure time for three cases: (a) autoclaves without thermal gradient, (b) corrosion loop with thermal gradient generated by an electrical heating inserted in the cladding tube, and (c) in-reactor, with the effects of thermal gradient and irradiation. Reproduced from Gilbon, D.; Bonin, B. E_DEN Monograph ‘‘Les Combustibles Nucle´aires,’’ Commissariat a` l’E´nergie Atomique; 2008.

Fast neutron irradiation can change the relative concentration of alloying elements between precipitate and matrix by a variety of mechanisms including ballistic mixing by a primarily knock-on effect.41 Fe and Cr can be dissolved from the intermetallic Zr(Cr, Fe)2 particles into the surrounding a-Zr matrix.42,43 The dissolution is linked to precipitate amorphization and the modified equilibrium between the amorphous precipitate and the matrix.44 As a result of the precipitate dissolution, the smallest particles dissolve completely, and the largest are significantly reduced in size. The ultimate location of the Fe is determined by thermal diffusion effects in the vicinity of the intermetallic particles where c-type dislocations may have formed. Postirradiation45 corrosion of such specimens shows progressive

-5

-3

-6

-4

H2

Dissolved H2 0 ppb Dissolved O2 0 ppb

-8

O2

O-2

-9

OH

-10

HO2

-11 -12 -13

HO-2

H2O2

(a)

-6 -7

O2 Dissolved O2 200 ppb Dissolved H2 500 ppb

H2O2

-8

O-2

-9

HO2

-10

OH

-11

HO-2

-12

-14 -15 -4

H2

-5 H2O2

Conc. (mol l−1) in log scale

-7 Conc. (mol l−1) in log scale

of ions from their lattice sites by fast neutron damage could lead to enhanced point defect concentrations, enhanced diffusion, and hence enhanced corrosion in-reactors.39,40 However, experimental studies have shown no clear evidence for this. Formation of electron–hole pairs and Compton electrons by b and g radiation could also theoretically lead to a significant increase in the electron conduction (electrical conductivity). The experimental evidence for this is, however, not conclusive.

-3

-2

-1

0

1

Time (s) in log scale

2

3

-13 -4

4

(b)

-3

-2

-1

0

1

2

3

4

Time (s) in log scale

Figure 18 Typical result of a computer simulation or radiolysis of water at 250  C for a dose rate of 4.5  102 Gy h1; (a) pure water; (b) with dissolved O2 and H2. Reproduced from Waterside Corrosion of Zirconium Alloys in Nuclear Power Plants; IAEA-TECDOC-996; International Atomic Energy Agency, Vienna, Austria, 1998.

Corrosion of Zirconium Alloys

degradation of the posttransition corrosion rates with increasing dose, although other studies have shown nodular corrosion improvement of irradiated material.46 Cox7 concluded that this effect is now seen as a primary contributor to enhanced corrosion in PWRs, since alloys that do not contain Fe or in which the Fe is incorporated in radiation-resistant particles, show lower in-reactor corrosion kinetics.

5.03.4 Nodular Oxidation In BWR conditions (Table 1), nonuniform, so-called nodular oxidation can also take place (Figure 19). The mechanism for nodular oxidation is yet to be fully understood, as it has proved to be challenging to study in laboratory experiments. Oxidation studies in 500  C steam demonstrated the dependence of nodular corrosion on second phase precipitate size and distribution,47 typical for various cladding batches and their metallurgical structure. However, other factors also affect nodule formation, as batches of the same cladding can behave quite differently. Such differences can be related to in-reactor phenomena such as galvanic effects, impurities, radiolytic species, and local power and flux.7,15 The question of whether nodular corrosion nucleates at intermetallic particles, between intermetallic particles, or as a collective property of a group of grains, is yet to be resolved.6 As discussed by the IAEA Expert group,6 the experimental evidence points toward the fact that nodules form away from intermetallic precipitates in

61

the alloys. In his review, Cox7 concluded that the redistribution of Fe from secondary phase particles diminishes nodular corrosion, but enhances uniform corrosion.

5.03.5 Hydrogen Embrittlement Absorption of hydrogen is a major contributor to degradation of zirconium alloys during service in nuclear systems.22 This degradation is primarily attributed to the formation of zirconium hydrides, a brittle phase that can embrittle cladding, and reduce its fracture toughness, thus enhancing the susceptibility to cracking.48 Recent studies have also shown that the addition of hydrogen can increase the creep rates in Zr–2.5Nb49 and possibly irradiation growth. The hydrogen can come from a variety of sources, some of them detailed in a 1998 report from an IAEA Expert group14: (i)

Hydrogen left over in the Zircaloy tubing after fabrication or from residual moisture due to surface preparation (the initial concentrations of hydrogen in the cladding, postfabrication but prereactor service, are on the order of 10 wt ppm). (ii) Desorption of water from incompletely dried up fuel. (iii) Hydrogen produced by (n,p) reactions in the cladding. (iv) Hydrogen ingress from the coolant water into the cladding during reactor exposure.

Uniform oxide

Nodular oxide

Zircaloy 20 μm Figure 19 Typical appearance of nodular corrosion in visual inspection and metallographic examination. Figures courtesy of Ron Adamson.

62

Corrosion of Zirconium Alloys

(v)

Absorption during the normal corrosion processes that occur in high-temperature aqueous solutions. (vi) Direct reaction of a clean (no species other than zirconium) surface with gaseous hydrogen. This hydrogen nominally could come from three sources: protons released by oxidation that form hydrogen gas, hydrogen produced by radiolysis of the water exposed to a high-energy neutron flux, and hydrogen specifically added to the cooling water to control stress corrosion cracking (see Chapter 5.02, Water Chemistry Control in LWRs and Chapter 5.08, Irradiation Assisted Stress Corrosion Cracking). (vii) Diffusion of hydrogen through a metallic bond with a dissimilar metal in which hydrogen has a higher activity. (viii) Cathodic polarization of zirconium in an electrolyte (typical for low-temperature reactors). By far, the largest source comes from the normal corrosion processes (v). 5.03.5.1 Hydrogen Production During Aqueous Corrosion of Zirconium-Base Materials The reaction of Zr with water to form zirconium oxide produces atomic hydrogen (a proton) Zr þ 2H2 O ! ZrO2 þ 2H2

½Ia

Zr þ 2H2 O ! ZrO2 þ 4H

½Ib

The proton released in the oxidation of Zr either combines with another proton to form gaseous hydrogen (eqn I(a)) or diffuses into the zirconium (eqn I(b)), where it can form zirconium hydrides. The majority of the protons formed during oxidation combine to form hydrogen gas but a fraction enters the metal. The term ‘hydrogen pickup fraction’ fH is used to relate the hydrogen absorbed to the hydrogen liberated during the corrosion reaction. fH ¼

H absorbed in cladding H generated in corrosion

Although the total amount of hydrogen absorbed is proportional to oxide thickness, the proportionality constant, the hydrogen pickup fraction, changes from alloy to alloy such that alloy design can significantly improve cladding performance. The pickup fraction

also changes with corrosion temperature and with corrosion time, but the mechanisms by which this occurs are not yet resolved. 5.03.5.2

Hydrogen Absorption

A detailed overview of the process of the absorption of hydrogen into zirconium-base materials is provided in the IAEA Technical Document ‘Waterside Corrosion of Zirconium Alloys in Nuclear Power Plants.’15 The oxide itself generally presents an effective barrier to the absorption of hydrogen such that the structure of the oxide and the electron transport mechanism can be linked to hydrogen uptake. There is some evidence that the nickel content in Zircaloy-2 increases the absorption of hydrogen (Figure 20), either by supporting direct dissociation of water or by mitigating recombination of hydrogen and oxygen.50–52 This was one reason why the nickel was removed in the formulation of Zircaloy-4. The Zr (Fe,Ni)2 intermetallic particles that exist in Zircaloy may provide a significant pathway for hydrogen uptake. Specifically, the electron current flows primarily at sites where intermetallic particles partially, or completely, short-circuit the oxide.53 Additionally, flaws have been found to exist in the oxide that are not associated with existing intermetallic particles but are at pits that may have resulted from intermetallic dissolution during pickling. These flaws allow the cathodic process to proceed. These locations are evidenced by cracks or small holes visible in the oxide.54 For some alloys, the amount of absorbed hydrogen varies as a function of oxide thickness. For example, Cox55 has shown that for Zircaloy-2, an initially high hydrogen absorption rate decreases as the oxide thickness increases, but then picks up again after the oxide reaches the transition region. Other researchers have shown an increase in H uptake just before the oxide transition.56 The additional porosity in the oxide, after transition, makes hydrogen pickup more likely. For Zircaloy-4, the pickup appears to be constant with the growth of the oxide. Oxygen additions to the water normally reduce the hydrogen uptake and hydrogen additions increase it.57 5.03.5.3

Hydride Formation

Hydrogen absorbed into the zirconium alloy cladding at levels more than the terminal solid solubility can embrittle the cladding through the formation of hydrides. The stress concentration at the ends of

Corrosion of Zirconium Alloys

. Zircaloy-2, as rolled . Zircaloy-2, β-treated . Zircaloy-4, as rolled . Zircaloy-4, β-treated

Hydrogen pickup (PPM)

60

63

~100-mil Thick specimens Zircaloy-2, β-treated

50 40

Zircaloy-2 as-rolled 30 Zircaloy-4, as-rolled or β-treated

20 10 0 0

500

1000

1500

2000

2500

3000

Hydrogen overpressure (PSI) Figure 20 Hydrogen pickup in Zircaloy-2 and Zircaloy-4 as a function of hydrogen overpressure after 14 days in 343  C water. Reproduced from Hillner, E. Hydrogen absorption in Zircaloy during aqueous corrosion, Effect of Environment, U.S Rep. WAPD-TM-411, Bettis Atomic Power Lab., W Mifflin, PA, 1964.

larger plate-type hydrides, as well as the localized deformation in the ligaments between the hydrides, leads to material weakness. As confirmed by Kerr et al.,58 performing recent in situ fracture work at the Advanced Photon Source at Argonne National Laboratory, in materials with large pregrown hydrides (100 mm), the residual stress field of the zirconium matrix governs the residual stress state of the hydride and load is shed to the notch tip hydride phase on increasing applied load. The hydrides, if formed, can be circumferential or radial (see Figure 21). The embrittlement is influenced by the orientation of hydrides relative to the stress. Hydrides that are oriented normal to the tensile load enhance embrittlement by providing an easy path for the growth of cracks through the hydrides.59 Radial hydrides are of greater concern, as they are oriented perpendicular to the hoop stress that arises during operation of the cladding tube. As one example, the stress ratio (hoop stress/axial stress ¼ sy/sz) from gas pressurization anticipated during a loss of coolant accident has an approximate value of two and most of the deformation is in the hoop direction.60 The hydrides observed in fuel cladding exposed to reactor environment are most often fcc delta hydrides ZrHx (where x  1.6). For a fixed amount of hydrogen uptake, the density and size of the formed hydrides is a strong function of the material microstructure. Initial hydride orientation has been shown to be a function of the texture of the Zircaloy-4 that develops during fabrication.61,62 As an example, as reported by

Singh and coworkers,63 Zr–2.5Nb that has been quenched and aged forms a higher density of smaller hydrides than Zr–2.5Nb that was cold-rolled and stress relieved, and the difference was attributed to the underlying grain structure, which acted as the nucleation host for the formation of hydrides. If sufficient stress is applied during the formation of the hydrides, the hydride platelets will form perpendicular to the applied stress. A specific example of deleterious hydride orientation is from ‘DHC,’ in which circumferentially oriented hydrides dissolve and reprecipitate at the crack tip, parallel to the crack orientation. Understanding DHC is of specific concern for CANDU pressure tubes and ensuring the long-term stability of spent fuel during storage and is discussed later in this section.64–69 5.03.5.4

Hydride Formation Rates

At reactor operating temperatures, the stable phases in the Zr–H phase diagram are (i) hcp-Zr with dissolved hydrogen and the delta hydride ZrHx, where x varies between 1.45 and 1.2 at high temperature. The terminal solid solubility of H in hcp-Zr is H CaZr ¼ A expðEH =T Þ

½5

where A is a constant equal to 1.2  105 wt ppm (or 0.8 mole H per cm3), and the activation energy for solid solution is 4300 K (the temperature validity for this equation is up to 865  C, the limit of the alpha phase region).

64

Corrosion of Zirconium Alloys

250 mm

(a)

100 μm

Figure 22 Hydride-rim formation in cladding on high-burnup (67 GWd/t) PWR fuel. (cladding from the H. B. Robinson plant, courtesy R. Daum ANL).

hydride precipitation), the hydrogen concentration in the cladding is essentially homogeneous. 5.03.5.5 (b)

100 μm

Figure 21 (a) Circumferential and (b) radial hydrides. Figures courtesy of Ron Adamson.

Hydrogen is very mobile in a-Zr and once it is absorbed in the cladding, it migrates easily in response to concentration, temperature, and stress gradients. The diffusion coefficient of H in Zr is H ¼ D0H expðEmH =kB T Þ DZr

EmH ¼ 3

½6

0:47eV and the preexponential factor where is 7  10 cm2 s1. This results in a high diffusion coefficient at the reactor operating temperatures (at 355  C (average cladding temperature), the diffusion coefficient is 1.1  106 cm2 s1), so that hydrogen responds quickly to changed conditions to establish a new steady state. The characteristic time to attain significant hydrogen ingress by diffusion through the thickness of the cladding at this temperature is about 12 min, which is much smaller than normal reactor exposure times. This means that at any given time, the hydrogen distribution in the cladding can be considered to be in quasi-steady state, that is, temporal variations need not be considered. Because of this, when the hydrogen is in solid solution (before

Formation of Hydride Rim

As, from eqn [5], the hydrogen solubility in Zircaloy decreases with decreasing temperature, the outer cladding arrives at the solubility limit before the inner cladding does. For an outer cladding temperature of 325  C and an inner cladding temperature of 385  C, the hydrogen solubilities are respectively 90 and 170 wt ppm. This causes hydrides to form preferentially at the outer cladding diameter. Metallographic examinations performed on cladding hydrided below the solubility limit show a more or less homogeneous hydride distribution through the thickness of the cladding. These hydrides presumably have precipitated out during the cooling from operation temperature, so that at reactor temperature, the hydrogen is in solution. As the overall hydrogen content increases as a result of increased corrosion, eventually the outer layer of the cladding reaches saturation and a hydride rim starts to form, whose thickness will increase with increasing reactor exposure. Figure 22 shows a metallograph of high-burnup cladding showing enhanced hydride formation near the outer diameter of the cladding. The hydride distribution response to stress and temperature gradients is at the root of several degradation mechanisms, such as DHC, secondary hydriding, and the degradation of cladding ductility from oxide spalling.

Corrosion of Zirconium Alloys

5.03.6 Delayed Hydride Cracking A detailed summary of the DHC phenomena is available in the report of an IAEA (International Atomic Energy Agency) Coordinated Research Project.70 An overview is presented here. The classic theory of DHC comes from the work of Dutton and Puls.71 The basis of the Dutton and Puls’ theory is sketched in Figure 23. The basis of the theory is that the crack tip hydride grows as hydrogen migrates from hydrides in the bulk of the material to the crack tip. The driving force for the diffusion of the hydrogen is the difference in the chemical potential of hydrogen between the bulk material and the crack tip hydride in response to hydrostatic stress.70,72 Dutton and Puls’ theory follows the following logic. The partial molar volume of hydrogen in the hydrides is positive. An increasing hydrostatic tensile stress reduces the chemical potential of hydrogen in the hydride relative to the bulk. This chemical potential difference causes hydrogen in solution to diffuse to the crack tip where it precipitates. In this model, the hydrides in the bulk dissolve to maintain the hydrogen concentration at the hydride interface at the solubility limit for the temperature at which the cracking is occurring. As the precipitate grows at the crack tip, it will eventually reach a critical size that is

dependent on the stress intensity factor at the crack tip and the crack will then progress an additional distance that is related to the hydride size. The relationship between crack velocity and stress intensity factor for this type of cracking is shown in Figure 24. The stable crack growth velocity is a function of temperature since that temperature affects both hydrogen solubility and hydrogen transport to the crack tip. A detailed update of the Dutton-Puls’ theory, along with a response to challenges to the theory by the group of Kim73 was recently released.74 The review by McRae supports the validity of the Dutton and Puls’ models rather than the alternatives promoted by Kim. Recently, Colas et al.75 used high-energy synchrotron X-rays at the advanced photon source, to perform in situ characterization of hydride dissolution, reprecipitation, and reorientation during thermal cycles under load. Their results, from samples initially charged with hydrogen at concentrations up to 600 ppm, indicate that the reorientation occurred above a threshold stress of 75–80 MPa. In another

s

Unstable crack growth

y

Hydride Plastic enclave

a

dg

Hydride

r=L

r = li r=l

x

Log crack velocity

Free surface

65

Vc Stable crack growth

KIC

No crack growth

rg

KH

Stress intensity factor (KI)

s Figure 23 Crack and diffusion geometry assumed in the DHC model of Puls. Reproduced from Puls, M. P. J. Nucl. Mater. 2009, 393, 350–367.

Figure 24 Schematic diagram of crack propagation by DHC in hydrided Zircaloy. Reproduced from Delayed hydride cracking in zirconium alloys in pressure tube nuclear reactors. IAEA-TECDOC-1410; Nuclear Power Technology Development Section, International Atomic Energy Agency, Vienna, Austria, 2004.

66

Corrosion of Zirconium Alloys

recent study related to hydride reorientation, Daum et al.76 found that the threshold stress is approximately 75–80 MPa for both nonirradiated and high-burnup stress-relieved Zry-4 fuel cladding cooled from 400  C. Within the uncertainty of the experiment, the irradiation was not critical to setting the threshold stress. Under ring compression at both room temperature and 150  C, Daum found that radial-hydride precipitation embrittles Zry-4. Interestingly, for nonirradiated Zircaloy-4, samples with lower hydrogen concentration (300 vs. 600 ppm) appeared to be more susceptible to radial-hydride related embrittlement. In a separate work, Daum76 also showed that failure is sensitive to hydride-rim thickness. Zircaloy-4 cladding tubes with a hydride-rim thickness >100 mm (700 wt ppm total hydrogen) exhibited brittle behavior, while those with a thickness <90 mm (600 wt ppm) remained ductile.

Many challenges remain to be addressed in understanding the corrosion and hydriding of Zr alloys in nuclear environments. Although mechanistic understanding has been developed over the past decades, still much needs to be understood, in particular, on the role of alloy design on corrosion rate, hydrogen pickup, and oxide transition. In addition, the behavior of cladding under the more challenging conditions of severe fuel duty associated with longer exposures, higher temperatures, aggressive chemistry, and incidence of boiling needs to be studied.

References 1. 2. 3.

5.03.7 Summary and Outlook The extensive research on the corrosion of zirconium alloys has resulted in an enormous flow of information during several decades. On the basis of this, a basic understanding of the processes leading to oxidation and hydriding of zirconium alloys now exists. 1. Zirconium alloy fuel cladding undergoes corrosion when subjected to the reactor environment. The corrosion is related to the oxidation of the metal by coolant water and is associated with hydrogen uptake into the cladding. The latter phenomenon is the main contributor to in-reactor degradation of cladding performance. 2. Uniform corrosion is alloy dependent and environment dependent. The higher fuel duty now imposed on nuclear fuel can lead to accelerated corrosion, which can limit fuel lifetime. 3. Corrosion is accelerated under irradiation relative to out-of-pile results. The corrosion rates can increase after a given exposure. 4. The hydrogen pickup fraction can vary from alloy to alloy and for different environments and corrosion times. The hydrogen pickup mechanisms and the influence of the alloy on the process are still under study. 5. Modern alloys such as M5 and ZIRLO provide much improved corrosion performance and show the potential for significant benefits from careful alloy design.

4. 5. 6. 7. 8.

9.

10. 11.

12. 13. 14. 15. 16.

17. 18.

Kestersson, R.; Yueh, K.; Shah, H.; et al. In 2006 International Meeting LWR Fuel Performance (Top Fuel 2006), Salamanca, Spain, 2006; pp 67–71. Bossis, P.; Pecheur, D.; Hanifi, K.; Thomazet, J.; Blat, M. J. ASTM Int. 2006, 3, paper IDJAI 12404. Sabol, G. P. ZIRLO: an alloy development success. In 14th ASTM International Symposium on Zr in the Nuclear Industry, STP 1467, Stockholm, 2005; pp 3–24. Yang, R.; Ozer, O.; Rosenbaum, H. In Light Water Reactor Fuel Performance Meeting; ANS: Park City, UT, 2000. Cox, B. J. Nucl. Mater. 1968, 28, 1. Cox, B. In Advances in Corrosion Science and Technology; Fontana, M. G., Staehle, R. W., Eds.; Plenum: New York, 1976; Vol. 5, p 173. Cox, B. J. Nucl. Mater. 2005, 336, 205. Lim, D.; Graham, N. A.; Northwood, D. O. The degradation of zirconium alloys in nuclear reactors – a review; Canadian Report, Atomic Energy Control Board, Ottawa (Now Canadian Nuclear Safety Commission) INFO-0174; Jan 1986. Lemaignan, C.; Motta, A. T. Zirconium Alloys in Nuclear Applications, Materials Science and Technology, vol. 10. Nuclear Materials Pt. 2; Frost, B. R. T., Ed.; VCH Verlagsgesellschaft mbH, Weinheim, Germany, 1994. Franklin, D. G.; Lang, P. G. In 9th International Symposium on Zirconium in the Nuclear Industry, ASTM STP 1132, 1990; pp 3–32. Bailly, H.; Me´nessier, D.; Prunier, C., Eds. The Nuclear Fuel of Pressurized Water Reactors and Fast Reactors: Design and Behaviour, Collection du Commissariat a´ L’E´nergie Atomique; English Edition: Intercept, Andover, Hants, UK, 1999. Cox, B. J. Nucl. Mater. 1969, 29, 50. Cox, B.; Pemsler, J. P. J. Nucl. Mater. 1968, 28, 73–78. Corrosion of zirconium alloys in nuclear power plants; IAEA TECDOC-684; International Atomic Energy Agency, Vienna, Austria, Jan 1993. Waterside Corrosion of Zirconium Alloys in Nuclear Power Plants; IAEA-TECDOC-996; International Atomic Energy Agency, Vienna, Austria, 1998. Garde, A. M.; Smith, G. P.; Pirek, R. C. Effects of hydride precipitate localization and neutron fluence on the ductility of irradiated Zircaloy-4. In 11th International Symposium on Zr in the Nuclear Industry, STP 1295, Garmisch-Partenkirchen, ASTM, 1996; pp 407–430. Pierron, O. N.; Koss, D. A.; Motta, A. T.; Chan, K. S. J. Nucl. Mater. 2003, 322, 21–35. Raynaud, P. R.; Motta, A.; Koss, D. A.; Chan, K. S. Fracture toughness of hydrided Zircaloy-4 sheet under

Corrosion of Zirconium Alloys

19.

20. 21. 22.

23. 24. 25. 26. 27. 28.

29. 30.

31.

32.

33. 34. 35. 36. 37. 38. 39.

through-thickness crack growth conditions. In 15th International Symposium on Zr in the Nuclear Industry, ASTM STP 1505, Sunriver, OR 2007; pp 163–177. Daum, R. S.; Majumdar, S.; Bates, D. W.; Motta, A. T.; Koss, D. A.; Billone, M. C. On the embrittlement of Zircaloy-4 under RIA-relevant conditions. In Thirteenth International Symposium on Zirconium in the Nuclear Industry, ASTM STP 1423, 2002; pp 702–718. Motta, A. T.; Gomes-da-Silva, M. J.; Yilmazbayhan, A.; Comstock, R. J.; Cai, Z.; Lai, B. J. ASTM Int. 2008, 5; paper ID# JAI10125. Kass, S. J. Nucl. Mater. 1969, 28, 315–321. Bossis, P.; Lelievre, G.; Barberis, P.; litis, X.; LeFebvre, F. Multi-scale characterisation of the metal–oxide interface of zirconium alloys. In 12th Internation Symposium on Zr in the Nuclear Industry, ASTM STP-1354, 2000; pp 918–940. Hillner, E.; Franklin, D. G.; Smee, J. D. J. Nucl. Mater. 2000, 278, 334. Bryner, J. S. J. Nucl. Mater. 1979, 82, 84. Bataillon, C.; Fe´ron, D.; Marchetti, L.; et al. E-DEN Monograph ‘‘Corrosion’’ Commissariat a` l’E´nergie Atomique; 2008. Yilmazbayhan, A. PhD. Thesis in Nuclear Engineering, Penn State University, 2004. Yilmazbayhan, A.; Motta, A. T.; Comstock, R. J.; Sabol, G. P.; Lai, B.; Cai, Z. J. Nucl. Mater. 2004, 324, 6–22. Pecheur, D.; Lefebvre, F.; Motta, A. T.; Lemaignan, C.; Charquet, D. Oxidation of intermetallic precipitates in Zircaloy-4: Impact of irradiation. In 10th International Symposium on Zirconium in the Nuclear Industry, Baltimore, MD, ASTM STP 1245; 1994; pp 687–705. Pecheur, D.; Lefebvre, F.; Motta, A. T.; Lemaignan, C.; Wadier, J. F. J. Nucl. Mater. 1992, 189, 2318–2332. Pecheur, D.; Godlewski, J.; Billot, P.; Thomazet, J. Microstructure of oxide films formed during the waterside corrosion of the Zircaloy cladding in lithiated environment. In 11th International Symposium on Zr in the Nuclear Industry, ASTM STP 1295, Garmisch-Partenkirchen, 1995; pp 94–113. Pecheur, D.; Godlewski, J.; Peybernes, J.; Fayette, L.; Noe, M.; Frichet, A.; Kerrec, O. Contribution to the understanding of the effect of water chemistry on the oxidation kinetics of Zircaloy-4 cladding. In 12th International Symposium on Zr in the Nuclear Industry; ASTM STP-1354, Toronto, 1998; pp 793–811. Cox, B.; Ungurelu, M.; Wong, Y. M.; Wu, C. In the 11th International Symposium on Zr in the Nuclear Industry, ASTM STP 1295; American Society for Testing of Materials, 1996, p 114. Britten, C. F.; Arthurs, J. V.; Wanklyn, J. N. J. Nucl. Mater. 1865, 15, 263. Roy A. Castelli, Nuclear Corrosion Modelling. The Nature of CRUD. Elsevier ISBN: 978-1-85617-802-0. Orlov, A.; Degueldre, C.; Wiese, H.; Ledergerber, G.; Valizadeh, S. J. Nucl. Mater. 2011. doi:10.1016/ j.jnucmat.2010.12.033. Janney, D. E.; Porter, D. L. J. Nucl. Mater. 2007, 362, 104–115. Adamson, R.; Garzarolli, F.; Cox, B.; Strasser, A.; Rudling, P. Corrosion Mechanism in Zirconium Alloys. ZIRAT12 Special Topic Report; ANT International, 2007. Gilbon, D.; Bonin, B. E_DEN Monograph ‘‘Les Combustible Nucle´aires’’, Commissariat a` l’E´nergie Atomique; 2008. Maguire, M. A. In the 9th International Symposium on Environmental Degradation of materials in Nuclear Power Systems, Newport Beach, CA, 1999.

40. 41.

42. 43. 44. 45.

46.

47.

48.

49. 50.

51. 52.

53. 54. 55.

56. 57.

58. 59.

60.

61.

67

Cox, B.; Alcock, A.; Derrick, F. D. J. Electrochem. Soc. 1961, 108, 129. Motta, A. T.; Lefebvre, F.; Lemaignan, C. In the 9th International Symposium on Zr in the Nuclear Industry, ASTM STP 1132; American Society for Testing of Materials, 1991; p 718. Griffiths, M.; Gilbert, R. W.; Carpenter, G. J. C. J. Nucl. Mater. 1987, 150, 53. Wang, W. J. S.; Tucker, R. P.; Cheng, B.; Adamson, R. B. J. Nucl. Mater. 1986, 138, 185. Motta, A. T.; Lemaignan, C. J. Nucl. Mater. 1992, 195, 277–285. Cheng, B. C.; Kruger, R. M.; Adamson, R. B. In the 10th International Symposium on Zr in the Nuclear Industry, ASTM STP 1245; American Society for Testing of Materials, 1994; p 400. Etoh, Y.; Kikuchi, K.; Yasuda, T.; Koizumi, S.; Oishi, M. Neutron irradiation effects on the nodular corrosion on Zircaloy-2. In International Topical Meeting on LWR Fuel Performance; American Nuclear Society: Avignon, 1991; pp 691–700. Garzarolli, F.; Stehle, H.; Steinberg, E.; Weidinger, H. Progress in the Knowledge of Nodular Corrosion, ASTM-STP-939; American Society for Testing and Materials: Philadelphia, PA, 1987; p 417. Sawatzky, A.; Ells, C. E. In Twelfth International Symposium on Zirconium in the Nuclear Industry, ASTM STP 1354; Sabol, G. P., Moan, G. D., Eds.; American Society for Testing and Materials: West Conshohoken, PA, 2000; pp 32–48. Kishore, R. J. Nucl. Mater. 2009, 385, 591–594. Hillner, E. Hydrogen absorption in Zircaloy during aqueous corrosion, Effect of Environment, U.S Rep. WAPD-TM-411; Bettis Atomic Power Lab., W Mifflin, PA, 1964. Kass, S.; Kirk, W. W. ASM Trans. Quart. 1962, 56, 77. Cox, B. The effect of some alloying additions on the oxidation of zirconium in steam, U K. Report, AERE-R4458; United Kingdom Atomic Energy Authority, AERE, Harwell. Berks, 1963. Ramasubramanian, N. J. Nucl. Mater. 1975, 55, 134–154. Cox, B.; Wong, Y. M. J. Nucl. Mater. 1999, 270, 134–146. Cox, B. Some factors which affect the rate of oxidation and hydrogen absorption of Zircaloy-2 in steam, U K. Rep. AERE-R4348; United Kingdom Atomic Energy Authority, AERE, Harwell, Berks, 1963. Harada, M.; Wakamatsu, R. J. ASTM Int. 2008, 5; Paper ID JAI101117. Kass, S. The development of the Zircaloys, Corrosion of Zirconium Alloys, ASTM STP-368; American Society for Testing and Materials, W. Conshohocken, PA, 1964; pp 3–27. Kerr, M.; Daymond, M. R.; Holt, R. A.; Almer, J. D.; Stafford, S.; Colas, K. B. Scripta Mater. 2009, 61, 939–942. Davies, P. H.; Sterns, C. P. In Fracture Toughness Testing of Zircaloy-2 Pressure Tube Material with Radial Hydrides Using Direct-Current Potential Drop; Underwood, J. H., Chait, R., Smith, C. W., Wilhem, D. P., Andrews, W. A., Newman, J. C., Eds.; ASTM STP 905, 1986, p 379. Daum, R. S.; Majumdar, S.; Tsai, H.; et al. Mechanical Property Testing of Irradiated Zircaloy Cladding Under Reactor Transient Conditions, Small Specimen Test Techniques, ASTM STP 1418; Sokolov, M. A., Landes, J. D., Lucas, G. E., Eds.; American Society for Testing and Materials: West Conshohocken, PA, 2002; Vol. 4. Vaibhaw, K.; Rao, S. V. R.; Jha, S. K.; Saibaba, N.; Jayaraj, R. N. J. Nucl. Mater. 2008, 383, 71–77.

68

Corrosion of Zirconium Alloys

62. Kiran Kumar, N. A. P.; Szpunar, J. A.; He, Z. J. Nucl. Mater. 2010, 403, 101–107. 63. Singh, R. N.; Roychowdhury, S.; Sinha, V. P.; Sinha, T. K.; De, P. K.; Banerjee, S. Mater. Sci. Eng. A 2004, 374, 342. 64. Marshall, R.; Louthan, M. Trans. ASM 1963, 56, 693–700. 65. Nagase, F.; Fuketa, T. J. Nucl. Sci. Technol. 2004, 41(12), 1211–1217. 66. Singh, R.; Kishore, R.; Singh, S.; et al. J. Nucl. Mater. 2004, 325, 26–33. 67. Choubey, R.; Pulse, M. Met. Mat. Trans. A 1994, 25A, 993–1004. 68. Barraclough, K.; Beevers, C. J. Mater. Sci. 1969, 4, 518–525. 69. Daum, R. S.; Majumdar, S.; Liu, Y.; Billone, M. C. J. Nucl. Sci. Technol. 2006, 43(9), 1054–1067. 70. Delayed hydride cracking in zirconium alloys in pressure tube nuclear reactors. IAEA-TECDOC-1410; Nuclear

71. 72. 73. 74. 75. 76.

Power Technology Development Section, International Atomic Energy Agency, Vienna, Austria, 2004. Dutton, R.; Puls, M. P. Effect of Hydrogen on Behavior of Materials, TMS-AIME: New York, 1976; pp 512–525. Eadie, R. L.; Coleman, C. E. Scripta Metall. 1989, 23, 1865–1870. Kim, Y. S. Mater. Sci. Eng. A 2010, 527, 7480–7483. Puls, M. P. J. Nucl. Mater. 2009, 393, 350–367. Colas, K. B.; Motta, A. T.; Almer, J. D.; et al. Acta Mater. 2010, 58, 6575–6583. Daum, R. S.; Majumdar, S.; Bates, D. W.; Motta, A. T.; Koss, D. A.; Billone, M. C. On the embrittlement of Zircaloy-4 under RIA-relevant conditions. In Zirconium in the Nuclear Industry: 13th International Symposium. ASTM STP 1423, 2002; pp 702–718.

5.04 Corrosion and Stress Corrosion Cracking of Ni-Base Alloys S. Fyfitch AREVA NP Inc., Lynchburg, VA, USA

ß 2012 Elsevier Ltd. All rights reserved.

5.04.1

Introduction

70

5.04.2 5.04.2.1 5.04.2.2 5.04.2.3 5.04.3 5.04.3.1 5.04.3.2 5.04.3.3 5.04.3.4 5.04.4 5.04.4.1 5.04.4.1.1 5.04.4.1.2 5.04.4.1.3 5.04.4.1.4 5.04.4.2 5.04.4.3 5.04.4.4 5.04.4.4.1 5.04.4.4.2 5.04.4.4.3 5.04.4.4.4 5.04.4.4.5 5.04.4.5 5.04.4.5.1 5.04.4.5.2 5.04.4.5.3 5.04.4.5.4 5.04.4.5.5 5.04.4.6 5.04.4.7 5.04.5 References

Ni-Base Alloy Use in PWRS/BWRS Wrought Ni–Cr–Fe Alloys Age-Hardenable Ni-Base Alloys Ni-Base Welding Alloys General Corrosion Water Chemistry Flow Rates Crevices Mitigation Stress Corrosion Cracking Environmental Conditions Temperature Water chemistry Sulfur intrusions Electrochemical potential Flow Rates Crevices Material Susceptibility Factors Heat treatment Microstructure Grain size Chemical composition Product form Stress Operating stress Residual stress Surface effects Weld geometry Stress relief annealing Irradiation Mitigation Outlook

70 70 72 73 73 75 75 75 75 75 77 77 77 80 81 81 82 82 82 83 84 84 85 85 86 86 87 87 87 88 90 90 90

Abbreviations ASME ASTM B&PV BWR CEDM CMTRs

American Society of Mechanical Engineers American Society for Testing and Materials Boiler and Pressure Vessel Boiling water reactors Control element drive mechanism Certified material test reports

CRDM ECP EPRI GMAW GTAW HAZ IASCC

Control rod drive mechanism Electrochemical potential Electric Power Research Institute Gas-metal-arc welding Gas-tungsten-arc welding Heat-affected zone Irradiation-assisted stress corrosion cracking

69

70

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

IGSCC INCO LM LWR MSE NRC PWR PWSCC RCS RUBs SAW SCC SEM SMAW TT

Intergranular stress corrosion cracking International Nickel Company Light microscopy Light water reactor Mechanical surface enhancement Nuclear Regulatory Commission Pressurized water reactors Primary water stress corrosion cracking Reactor coolant system Reverse U-bends Submerged-arc welding Stress corrosion cracking Scanning electron microscope Shielded-metal-arc welding Thermal treatment

5.04.1 Introduction Nickel–chromium–iron alloys (i.e., nickel-base alloys) are widely used in the power industry in both fossil (e.g., coal and gas) power stations and light water reactor (LWR) nuclear power stations (i.e., pressurized and boiling water reactors (PWRs and BWRs)). As a result, the service behavior of these alloys has been extensively studied,1 especially their susceptibility to corrosion and stress-induced corrosion phenomena. The power industry is concerned with the occurrence of such failure phenomena because of their effect on the safety and availability of equipment. Corrosion and, in particular, stress corrosion failures are not new. The power industry is well acquainted with stress corrosion cracking (SCC) of stainless steel in BWR piping and nickel-base alloys in PWR steam

Table 1

generators and its effect on equipment availability.2 SCC of these austenitic alloys has been known for more than 50 years.

5.04.2 Ni-Base Alloy Use in PWRS/ BWRS 5.04.2.1

Wrought Ni–Cr–Fe Alloys

The wrought nickel-base alloys that are typically used for nuclear applications are Alloy 600 and, more recently, Alloy 690, which contain approximately twice the chromium content. These materials are used primarily for their inherent resistance to general corrosion (i.e., oxidation resistance), strength at elevated temperatures, and a coefficient of thermal expansion very close to carbon and low-alloy steels. The typical chemical composition and mechanical properties of these alloys are summarized in Tables 1 and 2, respectively. Both Alloy 600 and Alloy 690 are non-agehardenable, austenitic solid-solution strengthened materials. No precipitation reaction is possible with either alloy to increase strength; however, strength can be increased by cold-working the material. They are normally used in the annealed condition; however, a low-temperature heat treatment, or ‘thermal treatment’ (TT), is also generally used with these alloys, which tends to improve the resistance to SCC in primary water chemistry conditions, which is typically known as primary water SCC (PWSCC) (see later sections of this chapter). This improvement is clearly shown to be more pronounced, at least for Alloy 600 material, in crack initiation testing.3

Chemical composition of wrought nickel-base alloys used in nuclear applications

Alloying element

Alloy 690

Alloy 600

Alloy X-750

Alloy 718

Alloy 800

Ni þ Co C Mn Fe S Si Mo Cu Cr Ti Al P Nb þ Ta Others

58.0 min. 0.04 max. 0.5 max. 7.0–11.0 0.015 max. 0.50 max. – 0.50 max. 28.0–31.0 – – – –

72.0 min. 0.15 max. 1.00 max. 6.00–10.00 0.015 max. 0.50 max. – 0.50 max. 14.0–17.0 – – – –

70.0 min. 0.08 max. 1.00 max. 5.0–9.0 0.01 max. 0.50 max. – 0.50 max. 14.0–17.0 2.25 – 2.75 0.40–1.0 – 0.70–1.20

50.0–55.0 0.08 max. 0.36 max. Bal. 0.015 max. 0.35 max. 2.8–3.3 0.30 max. 17.0–21.0 0.65–1.15 0.20–0.80 0.015 max. 4.75–5.50 B 0.006 max.

32–35 0.03 max. 0.4–1.0 Bal. – 0.30–0.70 – <0.75 20.0–23.0 <0.60 0.15–0.45 – –

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

These alloys are widely used in LWRs. In BWRs, applications include such locations as reactor vessel nozzle safe ends, core support structures, and shroud bolts. The PWR applications are typically within the reactor coolant system (RCS) such as steam generator tubing, penetrations and nozzles, control rod drive mechanism (CRDM) and control element drive mechanism (CEDM) nozzles in reactor vessel heads, and instrument nozzles in pressurizers and RCS piping, but may also be found in selected non-Class 1 components such as the Core Flood Tanks. In addition, Alloy 600 has also been used in a number of fastener applications.

Table 2

71

Figures 1–4 show typical applications of Alloy 600 material as used in the reactor coolant systems of the four major BWR and PWR vendor designs. Alloy 600 and 690 materials are available as plate, barstock, tube/pipe, or forged material. The majority of these materials were procured for the American Society for Testing and Materials (ASTM)4 or American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel (B&PV) Code5 specifications (e.g., ASTM B 166 and B 167 or ASME SB-166 and SB-167). Table 3 lists the various industry specifications that are used to procure these materials.

Typical room temperature mechanical properties of wrought nickel-base alloys used in nuclear applications

Mechanical property

Alloy 600

Alloy 690

Alloy X-750a

Alloy 718b

Alloy 800c

Yield strength, min. MPa (ksi) Ultimate tensile strength, min. MPa (ksi) Elongation, min. (%)

242 (35) 552 (80) 30

242 (35) 586 (85) 30

655 (95) 1103 (160) 20

1241 (180) 1034 (150) 10

334 (48) 572 (83) 30

a

Alloy X-750 HTH: Solution annealing at 1093  C (2000  F) and age-hardening at 704–718  C (1300–1324  F). Alloy 718: Solution annealing at 1100–1400  C (1832–2000  F) and age-hardening at 720 and 620  C (1328 and 1148  F). c Alloy 800: Solution annealing at 1038–1066  C (1900–1950  F) and age-hardening at 760  C (1400  F) for 10 h, furnace cool to 649  C (1200  F), hold for 20 h. b

Feedwater recirculation inlet/outlet welds RPV attachments/ brackets

Shroud support structure

Courtesy GE nuclear

CRD in-core housing instrumentation penetrations

Figure 1 Typical applications of Alloy 600 materials in the reactor coolant systems of a General Electric Design boiling water reactor. RPV, Reactor Pressure Vessel.

72

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Instrument and vent penetrations (both hot legs)

Decay heat line weld CRDM nozzles Pressure relief nozzle safe ends/ welds (both tanks)

PZR vent, spray, and relief line welds PZR steam and water instrument penetrations

CRDM motor housings Leak-off monitor lines PZR heater sleeves and diaphragm plates

Core flood tank instrument penetration (both tanks)

PZR surge nozzle weld

HPI/MU nozzle welds (all cold legs)

Hot legsurge nozzle weld

Core flood line welds (both lines) Core guide lugs (ID)

Piping-RC pump suction and discharge welds (all pumps)

RV bottom head instrument penetrations

Primary drain nozzles (both SGs)

Instrument nozzles and drain penetrations (all cold legs)

SG nozzle dam rings (both SGs)

Figure 2 Typical applications of Alloy 600 materials in the reactor coolant systems of a Babcock & Wilcox Design pressurized water reactor. RV, Reactor Vessel; RC, Reactor Coolant; SG, Steam Generator; PZR, Pressurizer.

5.04.2.2

Age-Hardenable Ni-Base Alloys

Alloy X-750, a high-strength precipitation-hardening alloy originally developed for gas turbines and the aerospace industry, is widely used in internal applications for both BWR and PWR designs, such as fuel assembly hold-down springs, control rod guide tube support pins, jet pump beams, and reactor internals structural bolting. This alloy is very similar in composition to Alloy 600, but contains additions of titanium and aluminum, which combine with nickel to form the g0 precipitates, Ni3Al and Ni3 (Al, Ti), for strengthening.6 Alloy 718 is another age-hardenable austenitic nickel-base alloy, originally developed for the aerospace industry, that has seen much use in the nuclear industry as a structural material due to its high strength and corrosion resistance.7 A significant increase in strength can be achieved by two precipitation reactions from solid solution involving g0 and g00 (Ni3Nb)

secondary phases within the austenitic matrix.8 The addition of niobium sets this alloy apart from other high-strength nickel-base alloys (e.g., Alloy X-750) that are strengthened by g0 alone. Both the g0 and g00 precipitates are quite small and can only be resolved with an SEM (scanning electron microscope) unless gross over-aging has taken place. The microstructures of the solution-annealed and age-hardened conditions are indistinguishable with light microscopy (LM). Alloy 718 has also been used extensively in PWR primary coolant systems, predominantly for fuel assembly hardware.9 Alloy 718 is utilized for fuel assembly hold-down springs, bolts, and spacer grids. It has been shown to possess superior SCC initiation resistance compared to Alloy X-750. Although Alloy 718 has experienced some isolated failures in PWRs due to fatigue/fretting cracking, it is considered highly resistant to intergranular SCC (IGSCC) initiation and other forms of corrosion.

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

73

PZR instrument nozzles PZR and RC pipe-surge line connections CEDM motor housing

Spray nozzlepipe weld Safety and relief valve nozzle-pipe welds and/or flanges

CEDM/ICI nozzles to RPV head welds RPV top head vent nozzle

PZR heater sleeves

RPV head leak monitor tubes (2) Surge nozzle-pipe welds Charging inlet nozzles (2 cold legs)

Instrument nozzles (all hot and cold legs) Safety injection and SDC inlet nozzle (all hot and cold legs) Let-down and drain nozzles (all hot and cold legs)

Shutdown cooling inlet nozzle (all cold legs)

Spray nozzles (2 cold legs)

Primary nozzle closure rings and welds (both SGs) Bottom channel head drain tube and welds (both SGs)

Guide lugs flow skirt ICI nozzles-ICI guide tubes (system 80 plants)

RCP suction and discharge (all cold legs)

Shutdown cooling outlet nozzle (1 hot leg)

Figure 3 Typical applications of Alloy 600 materials in the reactor coolant systems of a Combustion Engineering Design pressurized water reactor. RPV, Reactor Pressure Vessel; RC, Reactor Coolant; RCP, Reactor Coolant Pump; SDC, Shutdown Cooling; PZR, Pressurizer.

In addition, a modified Alloy 800 material, with carbide-forming elements added to limit the solid solution carbon content, has been successfully used for many years in Germany for steam generator tubing. The typical chemical compositions, mechanical properties, and industry procurement specifications are provided in Tables 1–3. 5.04.2.3

Ni-Base Welding Alloys

Welding of nickel–chromium–iron alloys is typically performed using arc-welding processes such as gastungsten-arc welding (GTAW), shielded-metal-arc welding (SMAW), and gas-metal-arc welding (GMAW).7 Submerged-arc welding (SAW) may also be used provided the welding flux is carefully selected. Alloy 82, 182, and 132 are typical filler metals used to join Alloy 600 components to carbon or low-alloy steel vessels and other component items. These weld alloys are also used as cladding in selected components within the reactor coolant system. In addition, Alloy 52 and 152 (and newly developed variants such as Alloy 52M and 152M) are filler metals that have

recently become the preferred materials used to join Alloy 690 component items to carbon or low-alloy steel vessels in the reactor coolant system. Occasionally, there is a need for welded Alloy X-750 items and the filler metals Alloy 82 or 69 (ERNiCrFe-8) were used during original fabrication. The Alloy 69 material is no longer produced and filler metal Alloy 718 is the currently recommended material for welding. The typical chemical compositions and industry procurement specifications of these alloys are summarized in Tables 3 and 4.

5.04.3 General Corrosion One of the main reasons that nickel-base alloys were chosen for LWR applications is that they have the ability to withstand a wide variety of severe operating conditions involving corrosive environments, high temperatures, high stresses, and combinations of these factors. General corrosion can be defined as uniform deterioration of a metal surface by chemical or

74

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Spray nozzlepipe weld

Safety and relief nozzle-pipe welds

Head vent pipe CRDM motor housings

CRDM nozzles

RPV head leak monitor tube

Surge nozzlepipe welds

Thermowells (all hot and cold legs)

RV nozzle pipe weld (all hot and cold legs)

SG nozzlepipe weld (all hot and cold legs)

Core support block Bottom-mounted instrument nozzles

Bottom channel head drain tube and welds (all SGs)

Primary nozzle closure rings and welds (all SGs)

Figure 4 Typical applications of Alloy 600 materials in the reactor coolant systems of a Westinghouse Design pressurized water reactor. RPV, Reactor Pressure Vessel; RV, Reactor Vessel; SG, Steam Generator.

Table 3

Typical nickel-base alloy specifications used in nuclear applications

ASME B&PV Code

ASTM Standard

Material

Product form

SB-163 SB-166 SB-167 SB-168 SB-637 SB-670 SFA 5.11 SFA 5.14

B 163 B 166 B 167 B 168 B 637 B 670 – –

Alloys 600/690/800 Alloys 600/690 Alloys 600/690 Alloys 600/690 Alloys 718/X-750 Alloy 718 Alloys 182/152 Alloys 82/52

Seamless tubing Rod and bar Seamless pipe and tube Plate, sheet, and strip Rod, bar, and forgings Plate, sheet, and strip Covered welding electrodes Bare welding rods and electrodes

electrochemical reaction with the environment. Nickel has good resistance to corrosion in the normal atmosphere, in freshwaters, and in deaerated nonoxidizing acids, and it has excellent resistance to corrosion by caustic alkalies. The high nickel content of these alloys gives them resistance to corrosion by many organic and inorganic compounds and also makes them virtually

immune to chloride-ion SCC. Chromium additions provide resistance to sulfur compounds and also provide resistance to oxidizing conditions at high temperatures or in corrosive solutions. Details of the corrosion resistance in these types of environments can be found elsewhere (i.e., see also Chapter 2.08, Nickel Alloys: Properties and Characteristics).2,10

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Table 4

75

Chemical composition of nickel-base welding alloys used in nuclear applications

Alloying element

Alloy 52a filler metal

Alloy 69 filler metal

Alloy 72 filler metal

Alloy 82 filler metal

Alloy 132 electrode

Alloy 152a electrode

Alloy 182 electrode

Ni þ Co C Mn Fe S Si Mo Cu Cr Ti Al P Nb þ Ta Al þ Ti Zr B Others

Bal. 0.04 max. 1.0 max. 7.0–11.0 0.015 max. 0.50 max. 0.50 max. 0.30 max. 28.0–31.5 1.0 max. 1.10 max. 0.030 max. 0.10 max. 1.5 max. – – 0.50 max.

70.0 min. 0.08 max. 1.0 max. 5.0–9.0 0.015 max. 0.50 max. – 0.50 max. 14.0–17.0 2.00–2.75 0.40–1.00 0.030 max. 0.70–1.20 – – – –

55.0 max. 0.05 0.1 max. 0.2 max. 0.008 max. 0.1 max. – 0.20 max. 44.0 max. 0.6 max. – – – – – – –

67.0 min. 0.10 max. 2.5–3.5 3.0 max. 0.015 max. 0.50 max. – 0.50 max. 18.0–22.0 0.75 max. – 0.030 max. 2.0–3.0 – – – 0.50 max.

62.0 min 0.08 max 3.5 max 11.0 max 0.02 max 0.75 max

Bal. 0.05 max. 5.0 max. 7.0–12.0 0.015 max. 0.75 max. 0.50 max. 0.50 max. 28.0–31.5 0.50 max. 0.50 max. 0.030 max. 1.0–2.5 – – – 0.50 max.

59.0 min. 0.10 max. 5.0–9.5 10.0 max. 0.015 max. 1.0 max. – 0.50 max. 13.0–17.0 1.0 max. – 0.030 max. 1.0–2.5 – – – 0.50 max.

0.50 max 13.0–17.0 – – 0.030 max. 1.5–4.0 – – – 0.50 max

a

Alloys 52M and 152M have controlled additions of boron and zirconium.

5.04.3.1

Water Chemistry

The water chemistry of LWRs is discussed in detail in Chapter 5.02, Water Chemistry Control in LWRs. Nickel-base alloys are essentially immune to general corrosion in LWR environments due to the formation of an adherent Cr-rich oxide on the surface. 5.04.3.2

Flow Rates

The inherent passivity of nickel-base alloys provides them with excellent resistance to flow-assisted corrosion. They are able to withstand very high flow rates, on the order of 18.3 m s1 (60 ft s1), without concern. Corrosion rates in such flowing conditions for nickelbase materials are expected to be <2.5 mm year1 (<0.1 mil year1).11 5.04.3.3

Crevices

General corrosion of nickel-base alloys in crevices is not anticipated to be of great concern in LWRs because of the passive nature of these materials. Pitting may occur occasionally in the presence of impurities, which could lead to SCC (see Section 5.04.4.3), particularly in the more oxidizing conditions of BWRs. No failures in PWRs have been directly attributed to creviced locations. 5.04.3.4

Mitigation

As general corrosion is minimal in LWR environments, mitigation is not really necessary. As noted

in this section on LWR Structural Materials, PWR environments have reducing conditions and general corrosion is not of concern. However, mitigation of corrosion concerns in the more oxidizing environment of BWRs has been through the use of hydrogen water chemistry (i.e., to make the environment more reducing, similar to a PWR) and noble metal chemical additions.12,13

5.04.4 Stress Corrosion Cracking SCC of nickel-base alloys is an important age-related phenomenon affecting LWRs. This type of failure mechanism for nickel-base alloys typically occurs intergranularly and is generally termed intergranular stress corrosion cracking (IGSCC). In PWRs, IGSCC is typically termed primary water stress corrosion cracking (PWSCC). The occurrence of SCC of nickelbase alloys has been extensively studied since the first reported observation of cracking in laboratory tests using Alloy 600 in high-purity water by Coriou et al.14 in 1959. Over the last three decades, IGSCC has been observed numerous times in LWRs and it has affected both the safe and economic operation of the reactors. In BWRs, cracking of nickel-base components such as safe ends, shroud bolts, and access hole covers has occurred; however, the predominant failures have been identified in Alloy 182 welds. In PWRs, PWSCC of Alloy 600 component items has been observed in steam generators, pressurizers, and

76

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

CRDM nozzles, and most recently in Alloy 182 and Alloy 82 welds. The mechanism of this cracking phenomenon is not completely understood, and prediction of crack initiation time has proven to be extremely difficult, if not impossible, due to the uncertainty of numerous variables (e.g., heat treatment and residual stress). In this section, emphasis will be given to the SCC of Alloy 600 materials in PWRs, given the fact that it has been the most prevalent; however, as noted above, BWR conditions are not immune to IGSCC of nickel-base alloys. It is known, however, that SCC of nickel-base materials occurs as a result of the following three factors:  susceptibility of the material  a tensile stress (including both operating and residual stress)  a corrosive environment The synergistic effect of these three factors is typically shown on a Venn-type of diagram (Figure 5). As an example, the susceptibility of Alloy 600 material to PWSCC depends on several factors, including the chemical composition, heat treatment during manufacture of the material, heat treatment during fabrication of the component, and operating parameters of the component. Chemical composition and heat treatment are interrelated in several ways. For example, one reason for annealing Alloy 600 is to solutionize the carbon in the alloy. As the material cools, chromium carbides precipitate from the solution at both intragranular and intergranular locations. If the cooldown from the

Mechanical Operational tensile stresses residual tensile stress

SCC Corrosive Susceptible material Chemical composition microstructure

Electrochemical corrosion potential temperature pH-value

Figure 5 Synergistic factors affecting stress corrosion cracking of nickel-base materials.

anneal is sufficiently slow, a greater number of carbides will precipitate at the grain boundaries (i.e., intergranularly) and the resistance to PWSCC will be improved. Well-decorated grain boundaries are an indication that an Alloy 600 material has received proper heat treatment and that sufficient carbon was available in the solution to combine with chromium. If adequate amounts of carbon and chromium exist, but the anneal is not at a high enough temperature or sufficient time is not allowed to solutionize the carbon, an adequate amount of carbon will not be available to precipitate intergranularly as chromium carbides, leading to minimal grain boundary decoration. Most precipitation occurs during cooldown following annealing; however, stress relief treatments can lead to additional precipitation. The primary goal of stress relief, however, is to allow a local realignment of highly strained regions to reduce internal stresses. Carbon and chromium concentration gradients are also reduced given the extended time at the temperature. Thus, if the anneal has not adequately solutionized carbon for chromium carbide precipitation at the grain boundaries, stress relief treatment will not reduce susceptibility to PWSCC. Tensile stresses, resulting from both residual and operating stresses, can be significant for some Alloy 600 component items. Operating stresses are produced from mechanical and thermal loading, while residual stresses are generated as a result of fabrication, installation, and welding processes. Residual stresses are more difficult to quantify than operating stresses and, in many instances, are of a higher magnitude than operating stresses. PWSCC is a thermally activated degradation mechanism, that is, as the temperature increases, the rate of PWSCC increases exponentially. Thus, the hot leg temperature of the RCS creates a more aggressive environment in which the Alloy 600 components must operate. The cracking observed in PWRs to date is typically axially oriented (although circumferentially oriented cracks have been observed) and occurs in an area, such as a weld heat-affected zone (HAZ), that has high residual tensile stresses. In a cylindrically shaped component (e.g., piping, vessel, and nozzles), the circumferential stresses are inherently higher than axial stresses. Thus, in a homogeneous material with no initial flaws, cracking would be expected to occur axially because of the higher circumferential stresses. PWSCC has been the subject of much research and analyses in recent years as a result of the many failures that have been attributed to it. However,

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

(50–70  F) temperature differences between hot and cold legs are enough to significantly influence the time to initiation and subsequent crack growth rate. Temperature is generally believed to affect the rate of SCC attack in accordance with an activation model for thermally controlled processes (Arrhenius equation), exp(Q/RT), where Q is the activation energy, R is the ideal gas constant, and T is the absolute temperature. The current consensus is that the activation energy for crack initiation falls in the range of 188–230 kJ mol1 (45–55 kcal mol1) and many predictions are based on 210 kJ mol1 (50 kcal mol1). There is also evidence that the activation energy varies with material carbon content.15

a reliable crack initiation model has yet to be developed. PWSCC of Alloy 600 components in the RCS can lead to through-wall cracking and thus leakage of primary water. (Catastrophic failure is not expected as circumferentially-oriented cracks do not occur unless very high axial stresses are generated in the component, e.g., from roll expansion methods.) 5.04.4.1

Environmental Conditions

The major environmental conditions affecting SCC of nickel-base material in LWR environments appear to be temperature, water chemistry (oxygen, hydrogen, lithium, boron, and sulfur content), and electrochemical potential (ECP). Each of these factors is evaluated as follows.

5.04.4.1.2 Water chemistry

The water chemistry of LWRs can generally be described as essentially pure water. PWRs primarily include hydrogen, boron, and lithium to produce reducing conditions. BWRs primarily operate with low levels of oxygen, but in recent times, hydrogen additions have been introduced to limit the oxidizing potential of the environment. Additional details are included in Chapter 5.02, Water Chemistry Control in LWRs.

5.04.4.1.1 Temperature

By far, temperature is the single most significant environmental factor influencing the initiation of SCC in LWR environments. This is evidenced by the fact that the vast majority of SCC of PWR steam generator roll expansion transitions have occurred on the hot leg side of the tube sheet. The 28–39  C

MA + drawn 35 % area reduction

4000

SCC initiation time (h)

77

3000

2000

1000

– – –

– – Hydrogen

200 ppm B 0.7 ppm Li hydrogen

500 ppm B 1.0 ppm Li hydrogen

1100 ppm B 2.0 ppm Li hydrogen

Water chemistry Figure 6 Stress corrosion cracking initiation times for Alloy 600 steam generator tubing at 360  C (680  F). Stress corrosion cracking initiation time as a function of primary water chemistry for as-drawn (35% area reduction) mill-annealed tubing. Reproduced from Airey, G. P. The stress corrosion cracking performance of Inconel Alloy 600 in pure and primary water environments. In Proceedings: 1983 EPRI Workshop on Primary-Side Stress Corrosion Cracking of PWR Steam Generator Tubing; EPRI NP-5498, Project S303–5, with permission from Electric Power Research Institute.

78

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

5.04.4.1.2.1 Hydrogen

The effect of dissolved hydrogen on SCC susceptibility of nickel-base alloys (e.g., Alloys 600 and X-750) has been evaluated by numerous researchers. Pathania and McIlree16 reviewed the influence of hydrogen on PWSCC in 1987. At that time, the emphasis of the work was on initiation at temperatures of 360  C (680  F) and above. The authors concluded that the susceptibility of Alloy 600 increased when the amount of dissolved hydrogen increased. Airey17 has shown that dissolved hydrogen increases the rate of PWSCC of steam generator tubing in autoclave tests at 360  C (680  F). An example of his

Temperature (⬚C) 10-5

290

325

345

365

Crack growth rate (mm s-1)

4 pts

10-6 4 pts Ave. of 4 pts

data, shown in Figure 6, shows that the SCC initiation time for pure water is decreased dramatically when hydrogen is added. However, in the primary water of PWRs, the effect of hydrogen appears to be a function of the boron and lithium content. Bandy and Van Rooyen18 have shown a similar effect with boron alone versus pure water and primary water (Figure 7). They showed that 83% of the specimens cracked in pure water with hydrogen versus only 2% in pure water without hydrogen. More recently, Norring has reported on tests to determine the effect of hydrogen overpressure in 330  C (626  F) water.19 Results of these tests, shown in Figure 8(a), suggest that the rate of PWSCC increases with increasing hydrogen overpressure. The most recent update on the influence of hydrogen on PWSCC was prepared by Cassagne et al.20 in 1997. All the data seem to indicate that the susceptibility of Alloy 600 decreases drastically for low hydrogen values (<10 kPa (1.45 psi)) regardless of temperature. For hydrogen partial pressure above 100 kPa (14.5 psi), a more progressive decrease in susceptibility seems to occur between 360 and 400  C (680 and 752  F). Data are not available in this range for lower temperatures. Between 10 and 100 kPa (1.45 and 14.5 psi), it appears that PWSCC initiation is not greatly affected for all temperatures between 400 and 310  C (752 and 590  F). At 290  C (554  F), the influence of hydrogen cannot be assessed because of a lack of data. 5.04.4.1.2.2

10-7 Flattened specs, I.E. Cold worked, pure H2O As received, pure H2O + H2 As received, pure H2O + H3BO3 As received, pure H2O As received, (0.03 % C), pure H2O Primary H2O 10-8

1.75

1.70

[

1.65 1 T(K)

1.60

1.55

] ´ 1000

Figure 7 Effect of hydrogen on Alloy 600 cracking in pure and primary water environments. Reproduced from Bandy, R.; Van Rooyan, D. Quantitative examination of stress corrosion cracking of Alloy 600 in high temperature water – Work in 1983. In Proceedings: 1983 EPRI Workshop on Primary-Side Stress Corrosion Cracking of PWR Steam Generator Tubing; EPRI NP-5498, Project S303–5, with permission from Electric Power Research Institute.

Boron and lithium

Evaluations of boron and lithium on PWSCC of Alloy 600 have been performed by numerous investigators. Norring et al.19 concluded that increasing the lithium content from 2.4 to 3.5 ppm significantly decreased the time to crack initiation (see Figure 8(b)). The most complete evaluation was performed by Ogawa et al.21. In these tests, hydrogen overpressure was kept at a constant level of 30 cm3 kg1 H2O. The results of this work are shown in Figures 9 and 10. It appears that maintaining a constant pH of 7.1–7.3 (at 285  C (545  F)) will produce a range of crack initiation times (Figure 9) and that increasing the boron content greater than 1200 ppm (at any lithium level) decreases the crack initiation times (Figure 10). Therefore, the beginning of cycle boron concentrations appears to be the worst condition for PWSCC initiation. Maintaining a pH level of 7.3 (at 285  C (545  F)) also appears to be better than a pH level of 7.1. Follow-up tests were performed at high boron concentrations with varying lithium concentrations

Temperature...........

329.9 ⬚C

329.6 ⬚C

Boron......................

1471 ppm

1441 ppm

Lithium....................

2.42 ppm

2.38 ppm

Hydrogen content...

13.1 ml kg-1

25.2 ml kg-1

7.0 kPa

13.6 kPa

7.354

7.351

Activity...

99 90

50 25 10 5 1 100 200

10 5

(a)

99 90

B&W 1700 ⬚F 7/8⬙ (6A)

25 10 5

500 1000 2000 5000 10 000 Exposure time (h)

The influence of hydrogen on the tendency to PWSCC

Temperature...........

328.9 ⬚C

329.6 ⬚C

Boron......................

1241 ppm

1441 ppm

Lithium....................

3.54 ppm

2.38 ppm

Hydrogen content...

24.7 ml kg-1

25.2 ml kg-1

13.6 kPa

13.6 kPa

7.575

7.351

Activity... pH...........................

B&W 1700 ⬚F 7/8⬙ (6)

90 50 25 10 5 1 100 200

500 1000 2000 5000 10 000 Exposure time (h)

B&W 1700 ⬚F 3/4⬙

90 50 25 10 5 1 100 200

99 Cracked specimens (%)

Cracked specimens (%)

99

Cracked specimens (%)

99

(b)

500 1000 2000 5000 10 000 Exposure time (h)

50

1 100 200

500 1000 2000 5000 10 000 Exposure time (h)

B&W 1700 ⬚F 3/4⬙

25

B&W 1700 ⬚F 7/8⬙ (6)

99 90

79

50

1 100 200

Cracked specimens (%)

Cracked specimens (%)

pH...........................

Cracked specimens (%)

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

500 1000 2000 5000 10 000 Exposure time (h) B&W 1700 ⬚F 7/8⬙ (6A)

90 50 25 10 5 1 100 200

500 1000 2000 5000 10 000 Exposure time (h)

The influence of lithium on the tendency to PWSCC

Figure 8 Effect of hydrogen and lithium on time to stress corrosion cracking of Alloy 600 steam generator tubing. Reproduced from Norring, K.; Rosborg, B., Engstrom, J., Svenson, J. Influence of LiOH and H2 on primary side IGSCC of Alloy 600 steam generator tubes, Colloque International Fontevraud II, Sept 10–14, 1990; Socie´te´ Franc¸aise d’Energie Nucle´aire, Paris, France, 1990; pp 243–249, with permission from Socie´te´ Franc¸aise d’Energie Nucle´aire.

80

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

5

Li (ppm)

4 1.7 1.6 1.5

3

1.4

1.3

1.2 1.1

2

1.0

1

0 6.3

6.4

6.5

6.6

6.7

6.8

6.9 7.0 pH 285 ⬚C

7.1

7.2

7.3

7.4

7.5

7.6

Figure 9 Isosusceptibility diagram for primary water stress corrosion cracking of Alloy 600 as a function of lithium and pH at 285  C (545  F). The numbers (1.0, 1.1, 1.2, . . .) represent the ratio of the percent intergranular fracture referenced to the response at B:280/Li:2.0 ppm. Reproduced from Ogawa, N.; et al. Nucl. Eng. Des. 1996, 165, 171–180.

5

Li (ppm)

4

3

1.7

1.6

1.5

1.4

1.3

2

1.2

1.1

1.0

1

0 2000

1500

1000 B (ppm)

500

0

Figure 10 Isosusceptibility diagram for primary water stress corrosion cracking of Alloy 600 as a function of lithium and boron. The numbers (1.0, 1.1, 1.2, . . .) represent the ratio of the percent intergranular fracture referenced to the response at B:280/Li:2.0 ppm. Reproduced from Ogawa, N.; et al. Nucl. Eng. Des. 1996, 165, 171–180.

to simulate beginning of fuel cycle conditions. Ogawa et al.22 report that there is little effect of lithium content from 2 to 10 ppm at boron concentrations greater than 1200 ppm (see Figure 11) and PWSCC susceptibility at 1600 ppm boron (2–10 ppm lithium) was higher than that at 500 or 280 ppm boron concentrations (with 2 ppm lithium). 5.04.4.1.3 Sulfur intrusions

Sulfur intrusions by themselves will not produce SCC in nickel-base material; however, sulfate will promote intergranular attack and intergranular SCC. A sensitized material microstructure is much more

susceptible (in terms of initiation time) to this type of attack, although all nickel-base materials will be attacked by sulfate. Andresen23 has shown that the time to failure decreased by 2–3 orders of magnitude in constant load tests between pure water and sulfate impurities (conductivities ranging from 5 to 55 mS cm1 due to sulfuric acid additions). Bandy et al.24 have shown that sulfates are very potent cracking agents for Alloy 600 materials. Temperature significantly accelerates the cracking and most likely decreases the threshold stress for cracking to occur.

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

81

11 (1.53) 10 1.5

9

8

(1.46)

Li (ppm)

7

6

5 1.4 4 1.3 1.2 1.1

3

1.0 1.1

2

1.2

1

0 2000

1500

1000 B (ppm)

500

0

Figure 11 Revised isosusceptibility diagram for primary water stress corrosion cracking of Alloy 600 as a function of lithium and boron. The numbers (1.0, 1.1, 1.2, . . .) represent the ratio of the % intergranular fracture referenced to the response at B:280/Li:2.0 ppm. Values in parentheses are extrapolated or interpolated from test result values. Reproduced from Ogawa, N.; Nakashiba, T.; Yamada, M..; Umehara, R.; Okamoto, S.; Tsuruta, T. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL, 1997; pp 395–401. Copyright 1997 by the American Nuclear Society, La Grange, IL, USA.

5.04.4.1.4 Electrochemical potential

100 PH2 = 0.005 MPa

IGSCC (%)

80

PH2 = 0.1 MPa

60 40 20

Alloy 600 350 ⬚C 0.01 m H3BO3 + 0.001 m LiOH soln

0 -1000

-300 0 (E ) Potential (mV) Corr

300

Figure 12 Effect of electrochemical potential on rate of primary water stress corrosion cracking in Alloy 600 material. Reproduced from Smialowka, S. Hydrogen induced IGSCC of Alloy 600 in high temperature aqueous environments. In Proceedings: 1987 EPRI Workshop on Mechanisms of Primary Water Intergranular Stress Corrosion Cracking; EPRI NP-5987SP, with permission from Electric Power Research Institute.

SCC is also significantly affected by the ECP. As shown in Figure 12 by the work of Smialowska,25 small changes in the ECP can have a large effect on SCC. Test results have demonstrated that a crack growth rate maximum, with respect to coolant hydrogen variation, is observed in proximity to a key phase transition, the nickel (Ni) to nickel oxide (NiO) phase transition, as shown bya Pourbaixdiagram (see Figures 13 and 14).26,27 The SCC hydrogen dependency is fundamentally described by the extent that the alloy’s corrosion potential deviates from the potential of the Ni/NiO phase transition. This potential difference represents the relative stability of the SCC controlling oxide films (e.g., crack tip oxides are often of a NiO structure). 5.04.4.2

Flow Rates

Flow rates in LWRs do not appear to have any effect on SCC susceptibility and no testing data are known to be available.

82

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Potentials typical of deaerated environments

Oxide film and IGSCC arrows indicate an increasing sensitivity to cracking

RHE potential

Potential

Ni/Nio equilibrium

Potentials typical of environments with dissolved hydrogen (10/30 bars)

No oxide film no IGSCC

‘Neutral’ environments

‘Caustic’ environments

Stable Cr oxides

Cr very soluble » 9/10 pH at 300/320 ⬚C

Figure 13 Effect of pH and potential on the surface films and sensitivity to stress corrosion cracking of Alloy 600 material. Reproduced from Scott, P. M.; Le Calvar, M. In Proceedings of the Sixth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; The Minerals, Metals, and Materials Society: Warrendale, PA, 1993; pp 657–667, with permission from The Minerals, Metals, and Materials Society.

5.04.4.3

Crevices

Nickel-base materials have been found to be very susceptible to IGSCC in crevice areas of BWR applications, especially at welds where both weld-induced sensitization and high residual stresses are present. Stress corrosion crack initiation and growth of nickelbase materials in BWRs have been attributed to the development of an acidic environment within crevices because of the oxygen in normal BWR water chemistry, combined with high residual and applied stresses resulting from the geometry and nearby welds. The first widely reported occurrence of SCC of Alloy 600 material in BWRs was at the Duane Arnold unit in 1978.28 This occurred in the crevice area of a recirculation inlet nozzle safe end and was initiated in the creviced area between the thermal sleeve and the safe end. However, no known failures of nickel-base materials have been directly attributed to environmental conditions within crevices in PWRs.

5.04.4.4

Material Susceptibility Factors

The PWSCC susceptibility of a particular heat of Alloy 600 material is dependent upon a variety of factors. The most important factors, based on results from a number of investigators, appear to be material microstructural features, chemical composition, and manufacturing process (or product form). Each of these has been evaluated below. 5.04.4.4.1 Heat treatment

Considerable research efforts have been made to identify the mechanism responsible for PWSCC of Alloy 600. These investigations, reviewed by Was,29 have shown that PWSCC is primarily dependent on the heat treatment received by the material. For instance, a heat treatment of mill-annealed (MA) Alloy 600 in the temperature range 650–750  C (1200–1380  F) has been found to cause a drastic improvement in the resistance to PWSCC. This has

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

83

1.50 O2

1.00

H2O

H

+

0.50

NiO2 1 ppb

V (SHE)

0.00

Acid SO4 cracking

H2 1 atm 8.2 ppm

Ni3O4

Ni++ Ni -0.50

NiO

Fe++ Fe

PWR Secondary side Primary side

Fe3O4 Fe

Caustic cracking

PWSCC

-1.00

Ni(OH)3 Caustic IGA

-1.50

-2.00

0

2

4

6

8

10

12

14

pH

Figure 14 Main domains of intergranular attack and stress corrosion cracking of Alloy 600 in aqueous solutions at 300 C (572 F) relative to the Pourbaix diagram for nickel. (Note the stability boundary for iron relative to its lowest oxidation states is also shown to indicate that iron can be oxidized under all likely PWR primary and secondary conditions whereas elemental nickel can be stable under certain PWR primary conditions depending on the corrosion potential fixed by the hydrogen partial pressure.) Reproduced from Scott, P. M.; Combrade, P. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL,1997; pp 65–73. Copyright 1997 by the American Nuclear Society, La Grange Park, IL, USA.

been termed ‘thermal treatment’ (TT). Although the exact mechanism of enhanced resistance to PWSCC is unclear and still under debate, there is general agreement that the chemical composition and structure of grain boundaries are of crucial importance. In this connection, chromium depletion, segregation of impurities to grain boundaries, intergranular carbides and their mechanical effect on stress concentrations, and grain boundary misorientation appear to be essential to the corrosion behavior of the material. Annealing temperature and time are critical to the precipitation reactions that occur. The penultimate and final anneals appear to be the most critical for PWSCC resistance. Stiller et al.,30 Briant et al.,31 Hall and Briant,32 EPRI NP-507233 all agree that a

high-temperature final anneal, sufficient to solutionize all carbon precipitates, followed by a slow enough cooling rate to precipitate copious carbides on the grain boundaries (intergranular carbides) will produce a reasonably PWSCC-resistant material. Norring34 tested Alloy 600 reverse U-bends (RUBs) in high-purity water containing hydrogen. The time to crack initiation at 320  C (608  F) increased by a factor of 8 when the annealing temperature was increased from 925  C (1697  F) to 1025  C (1877  F). 5.04.4.4.2 Microstructure

The precipitation obtained in the final material microstructure is dependent upon heat treatment and chemical composition (mainly carbon, chromium,

84

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

titanium, and nitrogen content). During the cooling that occurs after annealing, carbon combines with chromium to form chromium carbides. In addition, titanium carbides and titanium nitrides are formed. These precipitate within the grains (intragranularly) or at grain boundaries (intergranularly) depending largely upon the temperature reached during annealing and thermal treatments, the presence of prior precipitates (e.g., undissolved carbides), the time at temperature, the carbon content, and the cool down rate. A study was conducted by the Electric Power Research Institute (EPRI) in 1981 to evaluate carbide dissolution and precipitation kinetics of Alloy 600.35 This study showed that final annealing temperatures between 982 and 1010  C (1800 and 1850  F) (for carbon contents between 0.02% and 0.048%) provide a consistent thermal treatment response and an adequate precipitation of chromium carbides. The most widely accepted hypothesis for the beneficial effect of intergranular carbides on PWSCC resistance has been proposed by Bruemmer,36 who suggests that grain boundary carbides promote crack blunting due to their effectiveness as dislocation sources. Another possible explanation proposed by Smialowska25 is that Alloy 600 material passivates more readily in the presence of grain boundary carbides. Most recently, Fish et al.37 have proposed that intergranular carbides may improve passivity at the crack tip by reducing the amount of carbon segregated along the grain boundaries. Whatever the mechanism, or mechanisms, involved, there is general agreement that a microstructure with copious intergranular carbides and few intragranular carbides correlates with good resistance to PWSCC. A range of intergranular and intragranular carbide precipitation has been used by many investigators to quantify PWSCC susceptibility.15 Scott et al.38 have developed a range of material susceptibility indices, based on minimum times to failure, for Alloy 600 material used in French steam generators. These factors have also been used for CRDM nozzles.39,40 5.04.4.4.3 Grain size

Grain size can be related to yield strength and tensile strength of Alloy 600 material by the general type of Hall–Petch relationship.41 That is, a lower yield strength material will typically have a larger grain size. Data for PWSCC42 show that very small grain sizes (typically ASTM grain size numbers >9) are more prone to PWSCC than larger grain sizes (typically ASTM 4–8). Other data43 show that steam generator tubing with a grain size number larger

than ASTM 6 is less susceptible to PWSCC. However, the effect of grain size is most likely a secondorder effect, while carbide precipitation tends to be more important. 5.04.4.4.4 Chemical composition

The chemical composition of Alloy 600 has mainly been correlated to PWSCC susceptibility in terms of the carbon content. Many investigators have shown that carbon contents near the high end of the typical mill range (0.02–0.06 wt%) result in increased PWSCC susceptibility. Electricite de France (EdF) conducted a series of RUB specimen tests on steam generator tubing, using a low mill anneal temperature with carbon content ranging from 0.01 to 0.07 wt%.44 The tests were conducted in elevated temperature pure water and primary water with hydrogen overpressure. These tests show that low carbon (<0.018%) had the lowest PWSCC susceptibility. Pichon and others45 have shown that carbon contents that are relatively high (>0.063%) or very low (<0.012%) are more susceptible to PWSCC. The carbon content of Alloy 600 weld materials (Alloys 182 and 82) has also been investigated by several researchers.46,47 It appears that low carbon contents (<0.02%) are most susceptible, medium carbon contents (0.02–0.04%) are intermediate, and high carbon contents (>0.04%) have the most resistance to PWSCC. However, correlating the actual carbon content of the material may not be the best approach. This is also shown in the test data reported by Norring et al.34 A better approach may be to determine the ‘available carbon’ content. This would entail knowledge of the titanium and nitrogen content of the heat of material. If it is assumed that all the nitrogen is precipitated as titanium nitrides (TiN), then the remaining titanium can be assumed to react with the carbon in the material. With these assumptions (which are thermodynamically reasonable), the remaining carbon would be the ‘available carbon’ amount to precipitate as chromium carbides (M7C3 or M23C6 precipitates). Unfortunately, titanium and nitrogen are not normally reported in the certified material test reports (CMTRs); thus, this approach cannot be used without archive material or sampling of the heats in service. Buisine et al.48 evaluated the PWSCC resistance of nickel-based weld metals with various chromium contents. The tests clearly demonstrated that weld metals with 30% chromium were resistant to PWSCC. The threshold for PWSCC resistance appears to be between 22% and 30% chromium.

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

There have been discussions in the literature of the role of chromium depletion along the grain boundaries (‘sensitization’), as well as boron and phosphorous segregation along the grain boundaries.49 The degree of sensitization is tied to the heat treatments and there have been conflicting reports on the benefits and disadvantages of sensitization. It has been shown that a sensitized material is more susceptible to caustic attack. In addition, boron and phosphorous may be beneficial for retarding PWSCC, but the experimental evidence is weak at present. 5.04.4.4.5 Product form

Alloy 600 component items in the RCS are fabricated by press forging, hammer forging, hot rolling, and forming and machining from bar stock in accordance with ASME SB-166 or cold drawn and hot finished tubing or piping in accordance with ASME SB-167. In addition, it has been suggested that some ASME SB-167 material was actually produced from SB-166 bar stock and recertified as SB-167 material. The fabrication processes affect the material microstructure, yield strength, and hardness, which in turn affect the PWSCC resistance of the material. The PWSCC susceptibility of various product forms has been evaluated to some degree. The most susceptible product form appears to be cold worked and low temperature MA steam generator tubing, as evidenced by the vast majority of data available in the literature. However, the other product forms (i.e., forgings, piping, and bar stock) have not been assessed as extensively. One study, performed by Webb50 found differences in susceptibilities of cold-worked and hot-worked materials with similar microstructures. The cold worked and annealed tubing was more susceptible than hot worked and annealed forging material. The different behavior remains even when yield strength differences between the materials are taken into account. Another study on the relative susceptibility of high and low yield strength bar and tubing, performed for EPRI,51 concluded that PWSCC susceptibility ranked in decreasing order as (1) high-yield strength bar, (2) high-yield strength tubing, and (3) low-yield strength bar. However, no low-yield strength tubing was tested in this study. Microstructural evaluation in the study described above indicated that grain boundary carbide decoration is generally poor in bar products and better in tubing products. However, the susceptibility of Alloy 600 material also depends on surface cold work due to machining, grinding, and reaming. A material with

85

highly susceptible microstructure when subjected to a large amount of cold work on the surface (i.e., reaming or grinding) becomes very susceptible to PWSCC. Therefore, component items that were machined from bar stock and with weld root grinding should be considered highly susceptible to PWSCC. A machined surface without reaming or grinding is considered to have undergone moderate cold work and somewhat less susceptible to PWSCC. The least susceptible material is for cold drawn tubing with a high mill-anneal temperature. This type of approach has been adopted by Consumers Power Company at the Palisades Nuclear Plant.52 The most recent work was performed by Briceno et al.,53 who tested a variety of product forms. Two groups of materials appear to exist, as shown in Figure 15. The first group contains cold-worked material, thick wall tube, and steam generator tubing; whereas, the second group contains hot-worked thick wall tubes. The forged bars showed higher crack initiation times than the tubes tested. Specimens made of forged bars and hot-worked tube with similar grain boundary carbide distributions (60%) and the same grain size (ASTM 4.0–4.5) showed a significant difference in initiation times. The same tendency is seen for a low density of intergranular carbides when coldworked tube and forged bar with the same grain size were compared. 5.04.4.5

Stress

Laboratory tests and operating experience in PWRs suggest that significant PWSCC should not occur in Alloy 600 component items at stresses less than about 242 MPa (35 ksi) for temperatures up to around 324  C (615  F). As operating design stresses permitted by the ASME B&PV Code are much less than 242 MPa (35 ksi), PWSCC failure would not be expected to occur due to applied pressure and thermal loadings. This is supported by the fact that essentially all PWSCC failures have occurred at locations where (1) high residual stresses are produced during fabrication (e.g., pressurizer nozzle J-groove welds), (2) high stresses are produced as a result of strains induced during operation (e.g., steam generator-dented tube support plate intersections), or (3) high stresses are produced by geometric abnormalities (e.g., excessive steam generator tube ovality). The following paragraphs address the operating and residual stresses that may act in RCS pressure boundary component items.

86

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Specimens cracked/specimens tested

1.0

0.8

0.6 FC no cracks FB FA FX V1V2 W1 W2 no cracks GW GV

0.4

0.2

0.0 0

Bar

Thick wall tube

S.G. tube

1000

2000

3000 4000 Time (h)

5000

6000

7000

Heat

Working method

Final thermal treatment

YS (MPa) after forging

YS (MPa) after thermal treatment

%C

FC

Forged bar

745 ⬚C/2 h (a.c)

428

413

0.021

FB

Forged bar

800 ⬚C/2 h (a.c)

491

489

0.021

FA

Forged bar

800 ⬚C/2 h (a.c)

550

412

0.024

FX

Hot worked

(*)

–

291

0.03

V1

Cold rolled 1000 ⬚C/3 min (w.c)

–

301.5

0.051

V2

Cold rolled 1000 ⬚C/3 min (w.c)

–

274

0.052

W1

Hot worked

980 ⬚C

–

320.6

0.081

W2

Hot worked

980 ⬚C

–

244.7

0.067

GVW

Cold worked 927 ⬚C/3–5 min (a.c)

–

393

0.04

GV

Cold worked 927 ⬚C/3–5 min (a.c)

–

389

0.038

( * ) no heat treatment after extrusion of the tube. % C = Percentage carbon content Figure 15 Stress corrosion cracking initiation time at 330  C (626  F) for up to 10 500 h. Reproduced from Briceno, D.; Blazquez, F; Hernandez, F. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL,1997; pp 249–256. Copyright 1997 by the American Nuclear Society, La Grange Park, IL, USA.

5.04.4.5.1 Operating stress

5.04.4.5.2 Residual stress

Operating stresses, on the order of 69–138 MPa (10–20 ksi) in the hoop direction of J-groove, partial penetration welds, as well as full penetration welds, are anticipated in LWR component items. Stress levels on this order are not high enough to cause concern with SCC. Therefore, residual stresses must be taken into account.

The magnitude and sign of residual stresses in most typical RCS component items are considered to be a function of (1) surface layer hardness produced by cold working (e.g., rolling, machining, bending, or reaming) and (2) residual stresses produced by welding. Test data by Bandy and Van Rooyen (Figure 16) indicate that the initiation time for PWSCC in

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

87

Fraction of yield strength (RT)

10

AR

CW

345 ⬚C (AR)

1.0

0.1 1

100

10

1000

Failure (days) Figure 16 Correlation between stress and time to stress corrosion cracking for steam generator tubing tested in pure water at 300  C (tfailure ¼ K  Stressb, b ¼ 4). CW: cold work; AR: as-received. Reproduced from Bandy, R.; Van Rooyan D. Quantitative examination of stress corrosion cracking of Alloy 600 in high temperature water – Work in 1983. In Proceedings: 1983 EPRI Workshop on Primary-Side Stress Corrosion Cracking of PWR Steam Generator Tubing; EPRI NP-5498, Project S303–5, with permission from Electric Power Research Institute.

elevated temperature water varies inversely with the fourth power of the applied stress.18 Test data by Yonezawa et al.54 (Figure 17(a) and 17(b)) suggest that the initiation time varies inversely with the sixth or seventh power of the applied stress. In addition, Figure 18 shows data from Yonezawa that suggest that the absolute value of the stress raised to a power is not the only significant factor. Other significant factors include the amount of cold work and the ratio of the applied stress to the material yield strength. In other words, the PWSCC initiation time at a given applied stress level should increase as the material strength is increased. Conversely, for strain-controlled situations, such as at a J-groove weld, where the residual stress is a fixed percentage of the material yield strength, increasing the yield strength can actually decrease the initiation time. 5.04.4.5.3 Surface effects

The residual stress in a component item is a function of the applied mechanical forces and the heat input. Mechanical forces applied during cold work tend to produce compressive residual stresses in the surface layer. Conversely, heat input has the potential to produce tensile residual stresses. Specifically, heat input causes expansion of the surface layer yielding, if the thermal stresses exceed material yield strength and residual tensile stresses in the surface layer upon cooling. This suggests that high heat input operations such as drilling, turning,

or grinding have the potential to develop higher tensile residual stresses than low heat input operations such as reaming. Note that these high tensile residual stresses produced during machining are limited to a few mils depth. Berge et al.55 have shown, in crack initiation testing performed in France, that the time to crack initiation with a base material tensile stress of 500 MPa (72.5 ksi) is four times that of a material with a cold-worked surface stress of 1000 MPa (145 ksi). 5.04.4.5.4 Weld geometry

The welding of small penetration nozzles into larger vessel component items, by either partial penetration or full penetration welds, can produce tensile stresses in the nozzle. The magnitude of these stresses is generally on the order of the yield strength of the material. The cooling and subsequent shrinking of the weld material pull the nozzle wall toward the much stiffer component item to which it is welded. This shrinkage results in residual tensile hoop stresses in the nozzle near the weld. This weld shrinkage also results in axial bending of the nozzle. The magnitude of the bending stress, however, is typically much lower than that of the hoop stress. 5.04.4.5.5 Stress relief annealing

The primary goal of stress relief heat treatment is to allow a local realignment of highly strained regions to reduce internal stresses. As mentioned above,

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Stress measured by X-ray diffraction analysis on the top of U-bend specimens (kgf mm-2)

88

100

90 s-7

80

70

60

Crack

50 0.2% Offset proof stress at room temperature: 3.3 kgf mm-2 40

102

103

Applied stress (kgf mm-2)

(a)

104

Stress corrosion cracking testing time (h)

Crack No crack s-6 70 60 50 40 30 20

0.2% offset proof stress at 360 ⬚C

10 0 (b)

102

103

104

Stress corrosion cracking testing time (h)

Figure 17 Correlation between stress and time to stress corrosion cracking for Alloy 600 steam generator tubing. (a) Effect of stress on the stress corrosion cracking resistance of mill-annealed Alloy 600 at 360  C in high temperature water. (b) Effect of stress on the stress corrosion cracking resistance of mill-annealed Alloy 600 at 360  C in high temperature water, using constant load stress corrosion cracking test. Reproduced from Yonezawa, T.; Onimura, K.; Saito, I.; Takamatsu, H. In Materials for Nuclear Reactor Core Applications; BNES: London, 1987; pp 77–83, with permission from British Nuclear Energy Society.

improvement to the intergranular carbide precipitation is also almost always a benefit. The degree of stress relief is known to be a function of time and temperature. Available data from the International Nickel Company (INCO) are given in Figure 19.56 It can be seen that at a temperature of 482  C (900  F) for 4 h, 21% of the stress is relieved. Also, at a temperature of 870  C (1600  F) for 4 h, 88% of the stress is relieved. It should be noted that essentially all the stress relief occurs within 1 h at the temperature of interest. Residual stresses are

generally reduced from values near the room temperature yield strength to values that are near zero stress. 5.04.4.6

Irradiation

Irradiation-assisted stress corrosion cracking (IASCC) is an age-related degradation mechanism where materials exposed to neutron radiation become more susceptible to SCC with increasing fluence.57 IASCC, like PWSCC, is a distinctive subset of SCC. Despite numerous investigations and research efforts,

89

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Applied stress (kgf mm-2)

Material Heat treat ment

Stress corrosion cracking testing time (h)

Applied stress -2 Frac- (kgf mm ) ture 0.8σ0.2

Cold working

1.0sO2

—

24

8% tensile strained

45

15% tensile strained

57

´ (2.793) ´ (2.713)

IGSCC

975 ⬚C 25% 0.07 annea- tensile ling strained

72

´ (159) ´ (3.231)

IGSCC

C (%)

20% cold rolled

87

2.000 4.000 6.000 8.000 10.000 12.000 14.000

´ (1.717)

Fracture

2.000 4.000 6.000 8.000 10.000 12.000 14.000

—

9.505 ´ (3.704) ´ (5.772)

Stress corrosion cracking testing time (h)

IGSCC IGSCC

IGSCC

IGSCC IGSCC

—

45

(12.8481)

IGSCC

´ (2.374) ´ (4.191)

57

IGSCC

—

No crack ´ Crack

Figure 18 Effect of applied stress to yield strength ratio on time to stress corrosion cracking for Alloy 600 steam generator tubing. Effect of cold working on the stress corrosion cracking resistance for the same ratio of applied stress to yield strength of mill-annealed Alloy 600 at 360  C in high temperature water. Reproduced from Yonezawa, T.; Onimura, K.; Saito, I.; Takamatsu, H. In Materials for Nuclear Reactor Core Applications; BNES: London, 1987; pp 77–83, with permission from British Nuclear Energy Society.

30 482 °C (900 °F) 593 °C (1100 °F) 25 538 °C (1000 °F) Residual stress (1000 psi)

649 °C (1200 °F) 20

15 704 °C (1300 °F) 10 760 °C (1400 °F) 5 871 °C (1600 °F)

0

0

1

2 Time (h)

3

4

Figure 19 Effect of heating time and temperature on residual stress of cold-drawn, annealed Alloy 600 rod. Reproduced from Inconel Alloy 600, International Nickel Company: Huntington, WV, 1978, with permission from Special Metal Corporation.

details of the IASCC mechanism remain hypothetical.58–61 The current consensus is that IASCC results from a synergistic effect of irradiation damage to the material, water environment with possible radiolysis

effects, and a stress state. At present, interactions between these variables have not adequately been quantified and no primary IASCC controlling mechanism has been identified.

90

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

Nickel-base alloys, such as Alloy 600 and 690, are not typically utilized in areas of high fluence because of their high nickel content. Nickel will become highly radioactive when exposed to neutron radiation. No known IASCC-related studies have been reported in the open literature for these alloys. There are, however, some data available for Alloy X-750 at low neutron fluence levels. Results obtained from swelling capsule tests indicate that IASCC resistance of Alloy X-750 is approximately the same as Types 304 and 316 stainless steel.62,63 Some tensile test data of small scale Alloy X-750 HTH Condition bolts irradiated in a PWR to 1.4  1019 n cm2 (E > 1.0 MeV) performed without any failure up to 0.95% strain.64 Simulated BWR testing demonstrated an irradiation-enhanced IGSCC susceptibility of Alloy X-750 HTH Condition as a function of fluence and boron content.65 Fluence at 1019 n cm2 versus 1014–1018 n cm2 (E > 1.0 MeV) showed an increase in susceptibility. Alloy 718 has an excellent record in PWR primary water applications often at operating stresses close to or exceeding the yield point that can be 1000– 1100 MPa (145–160 ksi). Alloy 718 is used in fuel applications where very high neutron fluxes are also experienced. The few failures that have occurred have been attributed to a manufacturing defect that allowed components to enter service with preexisting intergranular defects. Alloy 718 is known to be highly resistant to crack initiation, but IGSCC will propagate rapidly in PWR primary water from preexisting defects. 5.04.4.7

Mitigation

A number of techniques have been evaluated and are available to delay or eliminate the occurrence of SCC in LWRs. Many of the available and future mitigation techniques were evaluated in a paper presented at a conference held by the US Nuclear Regulatory Commission (NRC) in 2003.66 In general, these techniques fall into three main categories: 1. Mechanical surface enhancement (MSE) 2. Environmental barriers or coatings 3. Chemical or electrochemical corrosion potential (ECP) control MSE techniques represent processes that reduce surface tensile residual stresses or induce compressive surface stresses on a component item or weld. Examples of MSE techniques include shot peening and electropolishing. Environmental barrier or coating techniques represent processes that protect the

material surface from an aggressive environment. Coating examples include nickel plating and weld deposit overlays or inlays. Chemical or ECP control techniques represent changes to the environment that alter the corrosion process or produce corrosion potentials outside the critical range for SCC. Examples of chemical or ECP control include zinc additions to the reactor coolant and modified water chemistry (e.g., hydrogen water chemistry and noble metal chemical additions to BWR coolant; variations in dissolved hydrogen levels, lithium concentrations, and boron concentrations for PWR coolant).

5.04.5 Outlook Wrought nickel-base alloys and their weld metals were originally used in LWRs due to the materials’ inherent resistance to general corrosion in a number of aggressive environments and because of a coefficient of thermal expansion that is very close to that of low alloy and carbon steel. Over the last 40 years, SCC has been observed in numerous component items and associated welds, sometimes after relatively long incubation times. The occurrence of SCC has been responsible for significant downtime and replacement power costs. Nickel-base materials will continue to be used both for repair and replacement activities in currently operating nuclear units and in the next generation of LWRs. Although the continued use of Alloy 600 material, for example, has essentially been eliminated in LWRs, acceptance and use of the higher chromium material, Alloy 690, is widely recognized by the industry. These higher chromium materials have been shown to be highly resistant to SCC in laboratory experiments and component repairs. Their use has proven to be an effective decision, since replacements have been free from cracking in operating reactors over periods up to about 20 years. Improvements in the weldability of the higher chromium weld metals (i.e., variants of Alloys 52 and 152) have been identified through significant research and the results have been put into application today.

References 1. 2.

Shah, V. N.; Macdonald, P. E. Eds. Aging and Life Extension of Major Light Water Reactor Components; Elsevier: Amsterdam, 1993. Corrosion in the nuclear power industry. Metals Handbook, Vol. 13, Corrosion, 9th ed.; ASM International: Metals Park, OH, 1987; p 927.

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys 3.

4. 5. 6. 7.

8.

9. 10. 11. 12.

13.

14.

15. 16.

17.

18.

19.

20.

Knorring, K.; Stiller, K.; Nilsson, J. In Proceedings of 5th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Monterey, CA, 1991; American Nuclear Society: La Grange, IL, 1991; pp 482–487. American Society for Testing and Materials Annual Book of ASTM Standards; ASTM: West Conshohocken, PA, 2009; Vol. 02.04. ASME Boiler and Pressure Vessel Code, Section II, Part B, Nonferrous Materials; ASME: New York, NY, 2009. INCONEL Alloy X-750 Product Handbook, 5th ed.; Inco Alloys International: Huntington, WV, 1988. Miglin, M. T.; Domian, H. A.; Baty, D. L. Microstructure and Stress Corrosion Resistance of Alloy X-750, 718, and A-286 in LWR Environments; Research Project 2181-1, NP-6392-M; Electric Power Research Institute: Palo Alto, CA, June 1989. Keiser, D. D.; Brown, H. L. A review of the physical metallurgy of Alloy 718; ANCR-1292; Aerojet Nuclear Company, Idaho Falls, ID, Feb 1976; UC-25, Materials, TID-4500, R63. Mager, T. R.; Wilson, I. L. W. Improved stress corrosion resistance of NiCrFe alloys; EPRI Research Project 2181-2, Interim Report, NP-6908-M; July 1990. Resistance to Corrosion, Product Brochure 5M 5-85 S-37, 4th ed.; Inco Alloys International: Huntington, WV, 1985. Copson, H. R.; Berry, W. E. Corrosion 1960, 16(2), 123–129. Kass, J. N.; Cowan, R. L. In Proceedings of Second International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; American Nuclear Society: La Grange Park, IL, 1985; pp 211–218. Andresen, P. L. In Proceedings of Seventh International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; National Association of Corrosion Engineers: Houston, TX, 1995; pp 563–577. Coriou, H.; Grall, L.; Pelras, M.; Vettier, S. In Third Metallurgical Symposium on Corrosion (Aqueous and Gaseous), 1959; North Holland: Amsterdam, 1960; pp 161–169. Report of Joint PWR Owners Group Alloy 600 Meeting; prepared by Dominion Engineering Inc., available from EPRI, Palo Alto, CA, Apr 27–28, 1995; Attachment #2. Pathania, R. S.; McIlree, A. R. In Proceedings of the Third International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; National Association of Corrosion Engineers: Houston, TX, 1987. Airey, G. P. The stress corrosion cracking performance of Inconel Alloy 600 in pure and primary water environments. In Proceedings: 1983 EPRI Workshop on Primary-Side Stress Corrosion Cracking of PWR Steam Generator Tubing; EPRI NP-5498, Project S303–5. Bandy, R.; Van Rooyan D. Quantitative examination of stress corrosion cracking of Alloy 600 in high temperature water – Work in 1983. In Proceedings: 1983 EPRI Workshop on Primary-Side Stress Corrosion Cracking of PWR Steam Generator Tubing; EPRI NP-5498, Project S303–5. Norring, K.; Rosborg, B.; Engstrom, J.; Svenson, J. Influence of LiOH and H2 on primary side IGSCC of Alloy 600 steam generator tubes, Colloque International Fontevraud II, Sept 10–14, 1990; Socie´te´ Franc¸aise d’Energie Nucle´aire: Paris, France, 1990; pp 243–249. Cassagne, Th; Fluery, B.; Vaillant, F.; de Bouvier, O.; Combrade, P. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL, 1997; pp 307–321.

21. 22.

23.

24. 25.

26.

27.

28. 29. 30. 31. 32. 33. 34. 35. 36.

37.

38.

39.

40. 41. 42. 43.

91

Ogawa, N.; et al. Nucl. Eng. Des. 1996, 165, 171–180. Ogawa, N.; Nakashiba, T.; Yamada, M.; Umehara, R.; Okamoto, S.; Tsuruta, T. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL, 1997; pp 395–401. Andresen, P. L. The effects of sulfate impurities in 288  C water on IGSCC of Inconel 600 in constant load and SSRT experiments. In Corrosion 84, Apr 2–6, 1984; available from National Association of Corrosion Engineers: Houston, TX; Paper Number 177. Bandy, R.; Roberge, R.; Newman, R. C. Corrosion 1983, 39(10), 391–398. Smialowka, S. Hydrogen induced IGSCC of Alloy 600 in high temperature aqueous environments. Proceedings: 1987 EPRI Workshop on Mechanisms of Primary Water Intergranular Stress Corrosion Cracking; EPRI NP5987SP. Scott, P. M.; Le Calvar, M. In Proceedings of the Sixth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; The Minerals, Metals, and Materials Society: Warrendale, PA, 1993; pp 657–667. Scott, P. M.; Combrade, P. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL, 1997; pp 65–73. Pipe Crack Study Group. Investigation and evaluation of stress-corrosion cracking in piping of light water reactor plants; NUREG-0531; NRC, Feb 1979. Was, G. S. Corrosion 1990, 46(4), 319–330. Stiller, K.; et al. Metall. Mater. Trans. 1996, 27A, 327–341. Briant, C. L.; O’Toole, C. S.; Hall, E. L. Corrosion 1986, 42(1), 15–27. Hall, E. l.; Briant, C. L. Metall. Trans. 1985, 16A, 1225–1236. Specially prepared Alloy 600 tubing; EPRI NP-5072, Project S303-17; Feb 1987. Carbide dissolution and precipitation kinetics of Inconel 600; EPRI NP-2093, Project 1708-1, Oct 1981. Bruemmer, S.; et al. Corrosion 1988, 44(11), 782–788. Fish, J. S.; Lewis, N.; Yang, W. J. S.; Perry, D. J.; Thompson, C. D. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL, 1997; pp 266–273. Scott, P.; Meyzaud, Y.; Benhamou, C. In Proceedings of International Symposium on Plant Aging and Life Predictions of Corrodible Structures, Sapporo, Japan, May 15–18, 1995; pp 1–8. Boudot, R.; Vidal, P.; Gelpi, A. A method for the evaluation of the nickel based alloys susceptibility to PWSCC. In Proceedings: 1992 EPRI Workshop on PWSCC of Alloy 600 in PWRs, EPRI TR-103345, Paper F6, Dec 1993. Cattant, F. Metallurgical investigations performed on CRDM nozzles removed from power plants. In Proceedings: 1992 EPRI Workshop on PWSCC of Alloy 600 in PWRs, Dec 1993; EPRI TR-103345, Paper B5. Park, H. B.; Kim, Y. H.; Lee, B. W.; Rheem, K. S. J. Nucl. Mater. 1996, 231(3), 204–212. Proceedings: 1985 Workshop on Primary-side SCC of PWR S/G Tubing; EPRI NP-5158, Project S303–5, EPRI. Shah, V. N.; Lowenstein, D. B.; Turner, A. P. L.; et al. Nucl. Eng. Des. 1992, 134, 199–215. Majo, D. G.; Gelpi, A.; Dallery, D.; Rouillon, Y.; Van Duysen, J. C.; Zacharie, G. Prediction of the in-service behavior of Alloy 600 tubes used in steam generators of pressurized water reactors. In Colloque International Fontevraud II Sept 10–14, 1990; Socie´te´ Franc¸aise d’Energie Nucleaire: Paris, France, 1990; pp 281–290.

92

Corrosion and Stress Corrosion Cracking of Ni-Base Alloys

44. Pichon, C.; Boudot, R.; Benhamou, C.; Gelpi, A. Residual life assessment of French PWR vessel head penetrations through metallurgical analysis. In American Society of Mechanical Engineers, Pressure Vessels and Piping Division, 1994 Pressure Vessels and Piping Conference; Service Experience and Reliability Improvement: Nuclear, Fossil, and Petrochemical Plants, 1994; Vol. 288, pp 41–47. 45. Seman, D. J.; Webb, G. L.; Parrington, R. J. Primary water stress corrosion cracking of Alloy 600 – Effects of processing parameters. In Proceedings: 1991 EPRI Workshop on PWSCC of Alloy 600 in PWRs, EPRI TR-100852, Paper E3, July 1992. 46. Mullen, J. V.; Parrington, R. J. Stress corrosion of Alloy 600 weld metal in primary water. In Proceedings: 1992 EPRI Workshop on PWSCC of Alloy 600 in PWRs, Dec 1993;EPRI TR-103345, Paper F3. 47. Buisine, D.; Vaillant, F.; Vidal, P.; Gimond, C. PWSCC resistance of nickel based weld metals with various chromium contents. In Proceedings: 1994 EPRI Workshop on PWSCC of Alloy 600 in PWRs, EPRI TR-105406, Paper D5, Aug 1995. 48. Airey, G. P. Metallography 1980, 13, 21–41. 49. Webb, G. L.; Burke, M. G. In Proceedings of the Seventh International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; National Association of Corrosion Engineers: Houston, TX, 1995. 50. Sarver, J. M.; Pathania, R.; Stuckey, K.; Fyfitch, S.; Gelpi, A.; Foucault, M. PWSCC of Alloy 600 penetrations (EPRI RP 3223). Presentation E1 in Proceedings: 1994 EPRI Workshop on PWSCC of Alloy 600 in PWRs, EPRI TR-105406, Project 3223-1, Aug 1995. 51. Hoang, P. H. Primary water stress corrosion cracking inspection ranking scheme for Alloy 600 components. In Proceedings of the Sixth symposium: Current Issues Related to Nuclear Power Plant Structure, Equipment and Piping, Dec 4–6, 1996; available from N. Carolina State University Center for Nuclear Power Plant Structures, Equipment, and Piping: Raleigh, NC. 52. Briceno, D.; Blazquez, F.; Hernandez, F. In Proceedings of the Eighth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems; American Nuclear Society: La Grange Park, IL, 1997; pp 249–256. 53. Yonezawa, T.; Onimura, K.; Saito, I.; Takamatsu, H. Materials for Nuclear Reactor Core Applications; BNES: London, 1987; pp 77–83.

54.

55. 56. 57. 58.

59.

60.

61. 62.

63. 64.

65.

Berge, P.; Buisine, D.; Gelpi, A. PWSCC effect of initial surface preparation. In Proceedings: 1997 EPRI Workshop on PWSCC of Alloy 600 in PWRs, EPRI TR-109138, paper E5, Nov 1997. Inconel Alloy 600, International Nickel Company: Huntington, WV, 1978. McNeil, M. B. Nucl. Eng. Des. 1998, 181, 55–60. Scott, P. A review of irradiation assisted stress corrosion cracking of austenitic materials for PWR core internals. Eurocorr, Framatome, France, Sept. 1996. Nelson, J. L.; Andresen, P. L. In Proceedings of Fifth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Monterey, CA, Aug, 1991; American Nuclear Society: La Grange Park, IL, 1991; pp 10–26. Andresen, P. L.; Ford, F. P.; Murphy, S. M.; Perks, J. M. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Jekyll Island, GA, Aug, 1989; National Association of Corrosion Engineers: Houston, TX, pp 1-83–1-121. Hanninen, H.; Aho-Mantilla, I. In Proceedings of 3rd International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Traverse City, MI, Aug 30–Sept 3, 1987; The Minerals, Metals, and Materials Society: Warrendale, PA, 1987; pp 77–92. Scott, P. J. Nucl. Mater. 1994, 211, 101–122. Andresen, P. L. In Stress Corrosion Cracking – Materials Performance and Evaluation; Jones, R. H., Ed.; American Society for Metals: Metals Park, OH, 1992; pp 182–210. Materials Reliability Program: Stress Corrosion Cracking of High Strength Reactor Vessel Internals Bolting in PWRs (MRP-88); EPRI Report 1003206; 2003. Bajaj, R.; Mills, W. J.; Lebo, M. R.; Hyatt, B. Z.; Burke, M. G. In Proceedings of Seventh International Symposium on Environmental Degradation of Materials in Nuclear Power Systems—Water Reactors; American Nuclear Society: La Grange Park, IL, 1997. Conference on Vessel Penetration Inspection, Crack Growth and Repair (NUREG/CP-0191). Compiled by: Mintz, T. S., Cullen, W. H., Sr.; US Nuclear Regulatory Commission, Mar 2004.

5.05 Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels U. Ehrnste´n VTT Technical Research Centre of Finland, Espoo, Finland

ß 2012 Elsevier Ltd. All rights reserved.

5.05.1

Introduction to Austenitic Stainless Steels

5.05.1.1 5.05.1.2 5.05.1.3 5.05.2 5.05.2.1 5.05.2.1.1 5.05.2.1.2 5.05.2.1.3 5.05.2.1.4 5.05.2.1.5 5.05.2.2 5.05.2.3 5.05.3 5.05.4 References

Types, Mechanical Properties, and Microstructures Welding Components Made of Stainless Steels in BWRs and PWRs Stress Corrosion Cracking IGSCC in BWR Environment Degree of sensitization Deformation Environment Stress Components at risk IGSCC in PWR Environment TGSCC in BWR and PWR Environments Pitting Corrosion Microbiologically Induced Corrosion

Abbreviations BWR CGR ECP EPR HAZ HWC IGSCC K KISCC LWR MIC NG NMC NMCA NWC PLEDGE PWR RBMK SCC TGSCC VVER

Boiling water reactor Crack growth rate Electrochemical corrosion potential Electrochemical (potentiokinetic) reactivation Heat-affected zone Hydrogen water chemistry Intergranular stress corrosion cracking Stress intensity factor Threshold stress intensity for SCC Light water reactor Microbiologically influenced corrosion Nuclear grade Nobel metal chemistry Noble metals chemistry addition Normal water chemistry Plant life extension and diagnosis by GE (General Electric) Pressurized water reactor Channel type graphite moderated reactor Stress corrosion cracking Transgranular stress corrosion cracking Water-water energetic reactor

93 94 94 94 96 96 96 98 98 100 101 101 101 102 102 103

5.05.1 Introduction to Austenitic Stainless Steels Austenitic stainless steels have rendered their vast use because of their good performance in corrosive environments, in addition to their excellent ductility, formability, toughness, and weldability. The good corrosion resistance of austenitic stainless steels is mainly due to chromium alloying, resulting in a protective, chromium-rich passive film on the material in many environments. Molybdenum, used as an alloying element in Type 316 stainless steels, further increases the corrosion resistance. Chromium and molybdenum are, however, both ferrite-forming elements, and to maintain a fully austenitic structure, a balance between austenite-stabilizing elements (C, N, Ni, Mn, and Co) and ferrite-stabilizing elements (Cr, Mo, Si, Ti, Nb, Al, V, and W) in solution must be established.1 To compensate for the molybdenum addition in Type 316 stainless steels, the amount of nickel is increased. In materials for nuclear environments, the cobalt content is kept as low as possible (<0.02%), because of its strong influence on radioactivity buildup. Stress corrosion cracking (SCC) is taken into account in the design codes for light water reactors (LWRs, 93

94

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

i.e., boiling and pressurized water reactors (BWRs and PWRs)) through a statement that SCC should not occur.2 Intergranular stress corrosion cracking (IGSCC) is still by far the largest damage mechanism for austenitic stainless steels in oxidizing BWR conditions, and work on avoiding IGSCC is still going on.3,4 Several factors affect IGSCC, one of them being sensitization. Sensitization is a result of nucleation and growth of chromium-rich carbides on grain boundaries, causing grain boundary chromium depletion. Chromiumdepleted grain boundaries are prone to corrosion and, in combination with a large enough stress and suitable environment, to IGSCC. All measures to prevent sensitization are therefore taken in all steps of component manufacturing and plant operation. Concerning the chemical composition, this can be done by reducing the carbon content to levels below 0.03%, as is done in Types 316L, 316LN, and 316NG stainless steels, where the carbon content is typically in the order of 0.02%. Since carbon is a strengthening element, nitrogen is added to these steels to still achieve good mechanical properties. Nitrogen also reduces chromium-rich carbide formation, a concept that is utilized in the French RCCM norms, which allow a carbon content of 0.035% in their nitrogen-strengthened Type 316 stainless steel with
0.08% N. The other approach to avoiding sensitization is to tie up the carbon into precipitates. This is utilized in stabilized stainless steels, which are of two main categories, titanium- and niobiumstabilized stainless steels, Type 321 and 347, respectively. To ensure that the carbon is tied into Ti(C,N) or Nb(C,N) precipitates, a high enough stabilization ratio, that is, Ti/C or Nb/C above 5 or 10, respectively, is specified. Good corrosion resistance is ensured by restricting the amount of harmful elements, especially sulfur and phosphorus, which may cause intergranular corrosion when segregated to grain boundaries. Several standardized grain boundary corrosion tests, such as the Strauss test5 and the electrochemical potential reactivation (EPR) test,6 are employed routinely as part of acceptance tests for materials and components. 5.05.1.1 Types, Mechanical Properties, and Microstructures The chemical compositions and mechanical properties of the most common stainless steels are presented in Table 1 and the main mechanical properties are summarized in Table 2.

In addition to compositional, mechanical, and corrosion resistance requirements, several other requirements are put on austenitic stainless steel materials for nuclear components. These include, for example, requirements on grain size. A small enough grain size is needed to enable reliable nondestructive inspection requirements, using ultrasonic techniques. A common requirement is that the grain size must not exceed ASTM number 4.0, which corresponds to an average grain size of 90 mm. The grain size of stabilized stainless steel components is typically smaller than this. 5.05.1.2

Welding

Stainless steels are generally welded with a slightly over-alloyed filler metal to ensure good corrosion resistance of the final joint. The weld shall contain a small amount (>3% but <10%) of d-ferrite to avoid solidification and liquation cracking.7 Welding induces residual stresses, which together with the operational stresses enhance crack initiation and growth. The aim is, naturally, always to minimize the residual stresses by a proper choice of welding parameters, by securing a good fit between the parts to be welded, etc. The use of a narrow-gap welding technique has increased remarkably during the last few decades. The narrow-gap welding method has many advances as it results, for example, in a lower level of residual stresses, a reduced weld volume, a narrower heat-affected zone (HAZ) with lower risk of sensitization, and less grain growth.8 5.05.1.3 Components Made of Stainless Steels in BWRs and PWRs Austenitic stainless steel is the main construction material in nuclear power plants (NPPs) owing to its good corrosion resistance, ease to manufacture different shapes, and good weldability. In BWRs, stainless steels are used for piping and reactor pressure vessel cladding and structures inside the pressure vessel, including the core shroud (which separates the primary water upward flow through the core from the downward flow in the annulus), the core plate (which supports the bottom of the fuel), the top guide (which aligns the top of the fuel bundles), the shroud dome, the steam separators, etc. Austenitic stainless steel is also largely used for other components such as pumps, valves, shafts, sleeves, and in auxiliary systems such as water tanks. The material choices for PWRs are essentially the same. The steam generator (SG) tubes are made of

Chemical composition of common stainless steel alloys

Type

C (max.%)

Mn (max.%)

Si (max.%)

P (max.%)

S (max.%)

Cr

304

0.08

2.00

1.00

0.045

0.030

18.0–20.0

8.00–10.5

304L

0.03

2.00

1.00

0.045

0.030

18.0–20.0

8.00–12.0

304LN

0.03

2.00

1.00

0.045

0.03

18.0–20.0

8.00–12.0

Z2CN19-10 316

0.035 0.08

2.00 2.00

1.00 1.00

0.040 0.045

0.030 0.030

18.5–20.0 16.0–18.0

9.0–10.0 10.0–14.0

2.00–3.00

316L

0.030

2.00

1.00

0.045

0.030

16.0–18.0

10.0–14.0

2.00–3.00

316LN

0.030

2.00

1.00

0.045

0.030

16.0–18.0

10.0–14.0

2.00–3.00

Z2CN18-12 321

0.038 0.08

2.00 2.00

1.00 1.00

0.040 0.045

0.030 0.030

17.0–18.2 17.0–19.0

11.5–12.5 9.00–12.0

2.25–2.75

347

0.08

2.00

1.00

0.045

0.030

17.0–19.0

9.00–13.0

Source: Wegst, C. E. Stahlschluessel Key to Steel; Stahlschluessel Wegst GmbH: Marbach, 1995. a The corresponding alloys according to the Swedish SS and the German DIN standards are also given.

Ni

Mo

N

Others

SS 2333/SS 2332 DIN 1.4301 SS 2352 DIN 1.4306 SS 2371 DIN 1.4311

0.10–0.16
0.08

SS 2343/SS 2347 DIN 1.4436/DIN 1.4401 SS 2348/SS 2353 DIN 1.4404/DIN 1.4435 SS 2375 DIN 1.4429

0.10–0.16
0.08

Corresponding alloysa

5 C min. Ti 10 C min. Nb

SS 2337 DIN 1.4541 SS 2338 DIN 1.4550

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

Table 1

95

96

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

Table 2

Minimum room temperature mechanical properties of stainless steels (hot finished and/or annealed forging)

Type

Tensile strength (min. MPa)

0.2% yield strength (min. MPa)

Elongation (min.%)

Reduction in area (min.%)

304 304L 304LN Z2CN19–10 316 316L 316LN Z2CN18–12 321 347

515 450 515 510 515 450 515 510 515 515

205 170 205 210 205 170 205 210 205 205

40 40 40 35 40 40 40 35 40 40

50 50 50 50 70 50 50

Source: Davis, J. ASM Specialty Handbook; ASM International: Materials Park, OH, 1994; ASM Standard A 240.

Ti-stabilized stainless steel in the Russian designed Water–water energetic reactors (VVERs), which are PWRs with slightly different water chemistry and horizontal instead of vertical SGs, as in Western PWRs. The SG tubes in western PWRs are made of nickel-based materials (Alloy 600, 690, or 800). Further, both the SG vessel and the pressurizer are clad with austenitic stainless steel.

5.05.2 Stress Corrosion Cracking SCC is a failure mode caused by a combination of a susceptible material, stresses, and an aggressive environment (Figure 1). There are two modes of SCC in austenitic stainless steel, namely intergranular and transgranular stress corrosion cracking (IGSCC and TGSCC). IGSCC in austenitic stainless steel is the major failure mode in BWRs, while it has not been considered as a plausible failure mode in PWR primary water under normal operation conditions. However, the number of IGSCC cases in PWRs has increased by time, showing that PWRs are not totally immune to IGSCC. Most TGSCC cases are due to chloride-induced SCC. TGSCC is rare in the primary system, and the failure cases are typically observed in auxiliary systems. 5.05.2.1

IGSCC in BWR Environment

In this section, the main factors affecting IGSCC in BWR environment are reviewed, that is, degree of sensitization, deformation, electrochemical corrosion potential (ECP), water purity, and stress. Also several

Region of potential stress corrosion cracking

Stress

Environment

Material susceptibility

Figure 1 The classic presentation of stress corrosion cracking includes the three circles: material, environment, and stress. Reproduced from General Electric Company. Alternative Alloys for BWR Pipe Application; NP-2671-LD, Final Report; San Jose, CA, 1982, with permission from BWR Owners Group.

other parameters affect IGSCC susceptibility, but a comprehensive description of these is out of the scope of this chapter. Among these parameters are temperature, hydrogen, mechanisms related to localization of deformation and to corrosion deformation interactions, such as effect of strain rate, subtle differences in chemical composition, dynamic strain aging, dynamic recovery, vacancy injection, selective dissolution, grain boundary segregation, and relaxation.9–21 5.05.2.1.1 Degree of sensitization

Sensitization is the result of nucleation and growth of chromium-rich carbides M23C6 at the grain boundaries, which results in a depletion of chromium at the grain boundaries because of faster diffusion rate

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

along the grain boundaries compared to that within the grain interior. Chromium-rich carbides form within the temperature range of 500–750  C, but continue to grow down to much lower temperatures. Sensitization can therefore occur as thermal sensitization during heat treatment and welding or as lowtemperature sensitization during long-time exposure to LWR temperatures, below the chromium-rich carbide precipitation temperature.22 In the latter case, the nucleation of carbides must have occurred previously, and the nucleated carbides grow during the long-time exposure and deplete the grain boundaries of chromium. The degree of sensitization is typically measured using the EPR test, which is sensitive to the area where the grain boundary chromium content is

below 15%, and is, thus, not a true measure of the grain boundary chromium content. The time for carbide precipitation increases as the carbon level decreases as seen in Figure 2. Nitrogen alloying delays carbide precipitation (Figure 3), while deformation accelerates diffusion and precipitation. The obvious remedy to avoid sensitization is, thus, to decrease the amount of free carbon, as explained earlier, by reducing the carbon content, or by tying carbon to Ti- or Nb-carbides and by restricting the degree of deformation. IGSCC in sensitized stainless steels occurs typically in the weld HAZ at a distance of 4–8 mm from the fusion line, at the location where a high degree of sensitization combined with high residual stresses

1000 1800 900

800 1400 316L (0.027% C)

316 (0.07% C)

Temperature (⬚F)

Temperature (⬚C)

1600 304 (0.053% C)

700

1200 600 1000 500 100

101

102

103

104

Time (min)

900 800

1832

0.069% N 0.145% N 0.247% N

1652

0.039% N

1472

700

1292

600

1112

500 0.01

0.1

1

10 Time (h)

100

1000

932

Precipitation temperature (⬚F)

Precipitation temperature (⬚C)

Figure 2 Time–temperature–precipitation diagram for stainless steels with different carbon contents. Reproduced from Shah, V. N.; MacDonald, P. E. Aging and Life Extension of Major Light Water Reactor Components; Elsevier: Amsterdam, 1993.

1000

97

Figure 3 Effect of nitrogen on precipitation of M23C6 in a 0.05C–17Cr–13Ni–5Mo stainless steel. Reproduced from Peckner, D.; Bernstein, I. M. Handbook of Stainless Steels; McGraw-Hill: New York, 1977; pp. 4-35–4-53, pp 751–757.

98

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

results in most severe conditions for IGSCC. The typical location of IGSCC is different in nonsensitized stainless steels, where IGSCC occurs very close to the fusion line, within the first few grains. 5.05.2.1.2 Deformation

Deformation increases the susceptibility of stainless steels to IGSCC, in sensitized as well as in nonsensitized stainless steels, where the role of strain is decisive. Deformation occurs as bulk cold work from rolling, bending, grinding, etc., as surface cold work from machining, grinding, etc. and as weld shrinkage from welding. Weld shrinkage can lead to up to 25% equivalent room temperature strain in the weld HAZ (Figure 4). The highest degree of deformation occurs very close to the fusion boundary, and this is also the location of observed IGSCC cracking in nonsensitized stainless steels pipes.23–27 The importance of cold deformation in IGSCC is shown in Figure 5. Deformation is estimated to be the main affecting parameter in 50% of all IGSCC cases covered in the survey (including sensitized and nonsensitized stainless steels). The effect of deformation has been studied using bulk-deformed materials,28–35 and the results show a correlation between IGSCC crack growth rate (CGR) and yield strength (Figure 6).

20

Various BWR stainless steel weld HAZs

% Strain

15

10

5

0

0

5

10

15

20

25

30

35

Distance from weld fusion line (mm) Figure 4 Deformation versus distance from the weld fusion line in various stainless steel weld HAZs. Deformation is expressed in terms of equivalent tensile strain at room temperature, and results from weld shrinkage strains during welding. Reproduced from Andresen, P. L.; et al. In Corrosion 2000, NACE 55th Annual Conference, Orlando, FL, Mar 26–31, 2000; p 12, Paper No. 00203, with permission from BWR Owners Group.

Much effort is nowadays put on deformation in terms of restrictions on bulk and surface deformation and on the development of sophisticated surface treatment procedures to remove surface cold work at critical locations.36,37 Application of narrow-gap welding results both in a decrease in the degree of deformation in the HAZ and in lower residual stresses. It should be pointed out that some components, such as bolts, can be made of intentionally cold-worked stainless steel to increase the material strength. 5.05.2.1.3 Environment

As mentioned earlier, one of the main reasons for the good behavior of austenitic stainless steels in LWR conditions is the formation of a protective passive film in high-temperature water (around 300  C). The oxide film formed in high-temperature water has a double-layered structure. The inner layer grown on the metal surface consists of a chromium spinel or magnetite and is covered by an outer layer of magnetite or Fe–Ni spinel precipitated from the aqueous phase.38,39 The double-layered oxide structure forms so that faster diffusing elements pass through the inner layer to the outer layer while the slower diffusing elements, such as chromium, remain in the inner layer and therefore the outer layer contains mainly of iron and the inner layer is enriched with chromium. Although consensus is not yet reached on the mechanistic details for corrosion and SCC in LWR environments, breakage of the passive film is generally considered to be of major importance because of the fact that if the oxide film breaks, the corrosion rate is high until passivation occurs.28,40,41 The CGR of IGSCC is highly dependent on the oxidizing power of the environment, that is, the ECP (Figure 7). The corrosion potential increases as the oxygen content increases in the high-temperature

Cold work Chemistry Material Nuclear grade Ni-base alloy Weld repair Residual stresses Sensitization

Figure 5 Cause of IGSCC in Swedish nuclear power plants. Cold work is the biggest singular parameter affecting IGSCC. Gott, K. Personal communication, April 2010.

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

99

Crack growth rate (mm s–1)

1E – 06 Very high martensite

Unsensitized 304, 304L and 316L SS and A600 288 ⬚C high purity water, 2000 ppb O2 CT tests at 27.5–30 MPa m½ Circles = High martensite SS Triangles = Alloy 600 Predicted response Two sensitized points for comparison

1E – 07

Very low or no martensite

Annealed cold worked

1E – 08 0

100

200

300 400 500 Yield strength (MPa)

600

700

800

Figure 6 Effect of yield strength (and martensite content) on stress corrosion crack growth rate of unsensitized stainless steels in oxygenated, high purity water. The predicted response is based on the PLEDGE model (plant life extension and diagnosis by GE). Reproduced from Andresen, P. L.; et al. In Proceedings of 11th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Stevenson, WA, Aug 11–14, 2003; American Nuclear Society: La Grange Park, IL, 2003; p 435, with permission from BWR Owners Group.

water, but it is not a linear relationship, and small changes in oxygen concentration can result in large changes of ECP and CGR. Important to notice is that the correlation between CGR and ECP is different for sensitized stainless steels and deformed, nonsensitized stainless steels, which show much higher CGRs compared to sensitized materials at low potentials (although lower than at high ECP). Two solutions to lower the ECP in BWRs have been developed, that is, hydrogen water chemistry (HWC) developed by General Electric and Asea Brown Boveri, and noble metal chemistry addition (NMCA, NobleChem™), developed by General Electric. The ECP in a BWR recirculation circuit during normal operation and using normal water chemistry (NWC), that is, 200 ppb oxygen, is above 100 mVSHE. The ECP is still higher in the core because of radiolytic decomposition of water forming hydrogen peroxide, H2O2. The ECP is remarkably lower in plants using either HWC, where 40–250 ppb hydrogen is added to the feed water, or noble metal chemistry (NMC™), where a small amount of platinum is added to the reactor water either at about 130  C during startup or during full power operation (OnLine NobleChem™), creating an electrocatalytic surface layer.42,43 The ECP of the buffered PWR environment is in the lower range of the ECP curve, that is, about 600 mVSHE. The trend, especially for the US BWR fleet is toward HWC and NMCA. None

of the US BWR plants operates on NWC, and 75% apply NMCA.44 The majority of the European BWRs operate on NWC. Water purity has a profound effect on both crack initiation and CGR in oxidizing environments. The main concerns are chlorides and sulfates for SCC and additionally copper for pitting corrosion (although Cu also has a synergistic effect on SCC). Sulfate and/or chloride levels already in the ppb-range increase the IGSCC susceptibility. Power plants monitor online the conductivity, which is a mirror for water purity, and analyze the amounts of impurities on regular basis from grab samples. The conductivity of BWR primary water of today has been reduced from a typical range of about 0.4 mS cm1 in the 1970s to 0.1–0.2 mS cm1 (the conductivity of theoretically pure water is 0.056 mS cm1). Dissolved oxygen is consumed inside cracks and crevices, and the local ECP is reduced to low levels, creating a potential gradient between the outer surface and the crack tip. This results in migration of anions into the lower potential area, which results in very high anion levels in the crack despite low levels in the surrounding environment.45 Further, the local environment in a crack can remain aggressive for a long time after, for example, short periods of higher impurity levels in the bulk environment. It is not only the environment during steadystate operation that needs attention, but also the

100

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

Sensitized 304 stainless steel 30 MPa m½, 288 ⬚C water 0.06–0.4 μS cm–1, 0–25 ppb SO4 SKI Round Robin Data Filled triangle = Constant load Open squares = Gentle cyclic

←200 ppb O2 ←500 ppb O2 ←2000 ppb O2

1E-05 4 dpa 304SS

10−6 316L stainless steel 25 mm CT specimen constant load 288 ⬚C water Test conditions: -2 0 ⬚C cm EPR ª27.5 MPa m½ 200 ppb O2

20% CW A600

316L (A14128, square) 304L (Grand Gulf, circle) non-sensitized SS 50%RA 140C (black) 10%RA 140C (grey)

1E-07

42.5 28.3 14.2 μmh–1

20% CW A600 GE PLEDGE predictions 30 MPa m½ sens SS

0.5 2000 ppb O2 Ann. 304SS 200 ppb O2

0.25

1E-08

Crack growth rate (mm s–1)

Crack growth rate (mm s–1)

1E-06

10−7

10−8

–1

0.1 μS cm Means from analysis of 120 L sens SS data 0.06 μS cm–1 0.06 μS cm–1

0.1

Predicted curves from PLEDGE code for typical range in ECP

GE PLEDGE predictions for unsens. SS (upper curve for 20% CW)

1E-09 –0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0.0 0.1 0.2 (a) Corrosion potential (VSHE)

10−9 0.3 0.4 (b)

10−1

100

101

Solution conductivity (μS cm–1)

Figure 7 Summary of crack growth rates of sensitized stainless steels versus corrosion potential, ECP (a) (reproduced from Andresen, P. L.; Morra, M. M. J. Nucl. Mater. 2008, 383(1–2), 97–111) and for nonsensitized stainless steels versus solution conductivity (b) (reproduced from Andresen, P. L. Corrosion 1988, 44(7), 450). The prediction curves for different water conductivity levels are according to the PLEDGE model, with permission from BWR Owners Group. RA ¼ reduction in area; CW ¼ cold work.

environment during shutdown, downtime periods, and during startup. The possible role of these will increase with plant age and amount of shutdowns and startups. 5.05.2.1.4 Stress

The stresses causing IGSCC are a combination of residual and operational stresses, although the first is considered more decisive in IGSCC failures. This is because operational stresses are kept low by design and components are usually designed to operate below 80% of their yield strength. The CGR of intergranular stress corrosion cracks increases with increasing stress intensity factor (K ) (Figure 8). The effect of stress intensity on CGR varies depending on the material and environment. Knowledge of the dependency between K and CGR is very important for structural integrity calculations, which are made to show that flaws, either postulated or detected using nondestructive inspections, are tolerable and do not pose a safety risk. Huge efforts have been put on the production of high-quality laboratory CGR data and

efforts are still going on. Approved relationships (i.e., agreement reached between national safety authority and plant operators) are called disposition lines, and examples of published lines are shown in Figure 8. Several methods to mitigate IGSCC have been applied over the years, such as last pass heat sink welding, mechanical stress improvement, and weld overlay cladding.46 All these aim at producing a compressive stress state in the HAZ. However, these methods are usually applied as temporary remedies. Measurement of residual stresses is an area of increased focus nowadays, and lack of knowledge can result in excessive under- or over-conservatism in design and in structural integrity calculations. Also other stress-related factors affect IGSCC, such as vibratory loading, thermal loads from, for example, stratification, as well as load cycles during shutdowns and startups. Much effort was earlier put on defining the KISCC; that is, the stress intensity, below which SCC would not occur. With improved laboratory testing techniques, lower and lower KISCC values have been measured and a true threshold value may not exist.

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

101

1E – 04 NRC 0313, BWR NWC MD-01, BWR NWC JSME, BWR NWC JSME, BWR HWC

–1

Crack growth rate (mm s )

1E – 05 1E – 06 1E – 07 1E – 08 1E – 09 1E – 10 1E – 11

0

10

20

30 40 50 60 Stress intensity (MPa m½)

70

80

90

Figure 8 Crack growth rate versus stress intensity according to dispositions lines for sensitized stainless steels in normal water chemistry boiling water reactors environment. Compiled by author from NRC Generic Letter GL880011. NRC Position on IGSCC in BWR Austenitic Stainless Steel Piping; January 25, 1988; http://www.nrc.org; Jansson, C.; Morin, U. In Proceedings of 8th International Symposium on Environmental Degradation of Materials in Nuclear power Systems – Water Reactors, Amelia Island, FL, Aug 10–14; American Nuclear Society: La Grange Park, IL, 1997; pp 667–674; Kobayashi, H.; Kashima, K. Int. J. Press. Vess. Pip. 2000, 77, 937–944. JSME is the Japan Society of Mechanical Engineers; NRC is the Nuclear Regulatory Commission.

5.05.2.1.5 Components at risk

The earliest incidents of SCC in BWRs occurred in stainless steel fuel cladding, before zirconium alloys were used.47 IGSCC plagued the BWRs in the 1970s and caused a clear reduction in capacity factors. Cracking was first observed in the recirculation and water cleanup systems in pipes with small diameter and later also in larger diameter pipes. The material was mainly Type 304 with a high carbon content of 0.6%. Owing to large efforts to solve the problem, including the development of Type 316NG, narrow-gap welding technique, as well as low-potential water chemistries, the number of IGSCC incidents has remarkably reduced.3 In the late 1980s, cracking in Ti-stabilized stainless steel piping was detected.48–50 Robust mitigation measures, including adoption of narrow-gap welding, change of material to Nb-stabilized stainless steel with higher stabilization ratio requirements, and reduction of the amount of welds, were applied in Germany to solve the problem. Also from Russian channel type graphite moderated reactors (RBMKs), which operate under BWR-like conditions, numerous IGSCC cases have been reported.51,52 In the 1990s, the first cases with IGSCC in nonsensitized stainless steels were reported in BWRs,23,25 first in pipings, and later numerously in core shrouds.27 Deformation (weld shrinkage in piping and surface grinding in

the core shrouds) is considered to be of major importance in these cases. 5.05.2.2

IGSCC in PWR Environment

PWRs operate at low corrosion potentials and very low oxygen levels, <30 ppb. The risk of IGSCC in austenitic stainless steels in nonoxidizing environment is, thus, much lower than in NWC BWR environment.31,53 Incidences with IGSCC under nominal PWR conditions have not been reported. Oxygen can, however, be enclosed in certain situations, such as startups, and can lead to IGSCC in austenitic stainless steels. Although the number of IGSCC cases in PWRs is still very low, the number seems to be increasing.54,55 IGSCC has been observed in pressurizer heater sleeves, canopy seals in the control rod drives, SG safe-ends, etc. Laboratory tests on cold-worked stainless steels show that IGSCC is possible also in normal PWR environment, indicating (although not generally accepted) that more IGSCC failures may occur in the operating PWR plants with time. 5.05.2.3 TGSCC in BWR and PWR Environments Austenitic stainless steels are prone to TGSCC when exposed to aggressive oxidizing water, for example,

102

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

containing chlorides, under sufficiently high stresses. As chloride levels in BWR and PWR primary systems are kept low, TGSCC is rare under normal operating conditions. The exception is TGSCC from the secondary side in VVER SG tubes, manufactured of Ti-stabilized stainless steels. Condenser leakages, frequently reported in VVERs, can result in chloride contamination of the secondary water and eventually in TGSCC in the SG tubes. Copper release from brass condensers and crevices formed by iron deposits further enhance both TGSCC and pitting corrosion on the secondary side of the SG tubes. A risk of TGSCC exists at locations where (a slow) buildup of aggressive conditions can occur. The risk of TGSCC increases with plant age, as the buildup of aggressive conditions can be very slow and can occur at unknown (uninspected) locations. Known chloride sources are old insulation and sealing materials (e.g., asbestos), leakage from cables, polymers, paints, concrete, etc. Wet insulation is the worst of these, as the insulation provides crevice conditions in addition to a chloride source. Strict regulations for expendables (grease, cleaning agents, sealing materials, etc.) allowed in NPPs are applied to reduce the risk of buildup of aggressive conditions. All bare outer surfaces of austenitic stainless steel components, where humidity may exist, can be at risk for TGSCC. TGSCC has been reported in valves,56 for example, where the source for the chlorides is assumed to be asbestos sealing used early in time and in water tanks, where chlorides probably stem from humidity and concentration buildup at the waterline. Stainless steel bellows in the BWR reactor containment are, in principle, at risk because of the high degree of cold work in the bellows. However, no SCC has been reported in these. New components can also be at risk for TGSCC, if proper measures are not taken to avoid contamination of components during transportation, storage, and installation. TGSCC can occur in oxidizing concentrated boric acid solutions although laboratory results are not fully conclusive whether chloride is also needed or not.56–58

is very often observed at same locations as TGSCC, but pitting corrosion can also occur without SCC and vice versa. The risk of pitting corrosion under normal BWR conditions is extremely low. However, pitting corrosion can occur in pressure boundary systems at locations where (slow) buildup of aggressive local conditions can occur. Such locations are, for example, areas with low water flow, dead ends, and valves with sealing. As pitting occurs only in oxidizing conditions, it is not a plausible degradation mechanism in PWR primary water under nominal environmental conditions. However, the environment may be oxidizing both locally and/or temporarily because of startups, for example. Different systems during shutdown may be filled with air, and this may cause air pockets during startup. The oxygen from air will then dissolve into the primary water and local oxidizing conditions temporarily emerge until the oxygen is consumed by the oxidation of metal surfaces. The risk of pitting corrosion (and TGSCC) is, however, highest in auxiliary systems, for example, at outer surfaces, where the temperature is low enough for condensation to occur. Thus, pitting corrosion can occur at nominally dry locations. Accumulation of aggressive local conditions is enhanced by crevices. The sources of chlorides were listed earlier. Sulfate sources have been introduced earlier, for example, in molybdenum disulfide greases, but since the harmful influence of this material was identified, it is not an allowed expendable material. Again, copper can enter the system from copper-containing structural components. Pitting corrosion is seldom considered to pose a safety problem, as the wall thicknesses of pressure boundary components are usually large enough to sustain pitting corrosion for long times without leakage. However, pitting corrosion is always an indication of a harmful environment existing at the location and is often associated with the risk of TGSCC, which can cause wall cracking in short time periods. Pitting corrosion enhances the risk of SCC as the pits increase the local stress concentration and thus act as crack initiators. Observation of pitting corrosion shall therefore not be omitted as insignificant.

5.05.3 Pitting Corrosion Pitting corrosion occurrence has several similarities to TGSCC, that is, it requires oxidizing conditions and presence of water with harmful ions, such as chlorides, fluorides, sulfates, and/or copper, but no stress is needed. The Type 304 stainless steel is more prone to pitting corrosion than Type 316 stainless steel. Pitting corrosion

5.05.4 Microbiologically Induced Corrosion A rather rare corrosion mode is microbiologically induced corrosion, or nowadays, microbiologically influenced corrosion (MIC). MIC is normal

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

electrochemical corrosion where the microorganisms either chemically or physically change the conditions on the metal surface to be favorable to corrosion.59 MIC appears as localized corrosion rather than as uniform corrosion, and in welds rather than in base materials. Pitting corrosion in the weld metal can cause preferential attack of either the austenite or the ferrite phase of the weld metal. The microorganisms of interest in MIC are mostly bacteria and fungi. The highest risk of MIC is at temperatures from 15 to 45  C and near neutral pH, that is, in the range from 6 to 8. MIC has been observed in fire-fighting systems, for example. MIC is stopped with great difficulty once it is established due to the high sustainability of the microorganisms involved. The quality of the water in all phases of the lifetime of the equipment is, thus, very important at locations with risk of MIC. Water of high quality must be used, not only during normal operation, but also during hydrotesting of the system, for example.

13. 14.

15.

16.

17. 18. 19. 20. 21. 22.

References 1. 2.

3. 4. 5.

6.

7. 8.

9. 10. 11. 12.

Peckner, D.; Bernstein, I. M. Handbook of Stainless Steels; McGraw-Hill: New York, 1977; pp 4-35–4-53, 751–757. Ford, P. F. In Proceedings of 13th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Whistler, BC, Canada, Aug 19–23, 2007; Allen, T. R., et al. Eds.; Canadian Nuclear Society: Toronto, ON, 2007. NRC. Expert Panel Report on Proactive Management of Materials Degradation; NUREG/CR-6923; US Nuclear Regulatory Commission: Washington, DC, 2007. Shah, V. N.; MacDonald, P. E. Aging and Life Extension of Major Light Water Reactor Components; Elsevier: Amsterdam, 1993. ASTM Standard A 262-02a. Standard Practices for Detecting Susceptibility to Intergranular Attack in Austenitic Stainless Steels; ASTM International: West Conshohocken, PA, 2008; p 17. ASTM Standard G 108-94. Standard Test Method for Electrochemical Reactivation (EPR) for Detecting Sensitization of AISI Type 304 and 304L Stainless Steels; ASTM International: West Conshohocken, PA, 2004. Folkhard, E. Welding Metallurgy of Stainless Steel; Springer-Verlag: Wien, 1988. Korhonen, M.; Luukas, M.; Ha¨nninen, H. In International Conference on Efficient Welding in Industrial Applications (ICEWIA), Lappeenranta, Finland, Aug 25–27, 1999; Martikainen, J., Eskelinen, H., Eds.; Lappeenranta University of Technology: Finland, 1999; pp 244–255 Aaltonen, P.; Saario, T.; Karjalainen-Roikonen, P.; et al. In Corrosion ’96; Denver, CO, Mar 24–29, 1996, NACE International: Houston TX, 1996; p 12 , Paper No. 81. Andresen, P. L.; Briant, C. L. Corrosion 1989, 45, 448–463. Arioka, K.; Yamada, T.; Terachi, T.; Miyamoto, T. Corrosion 2008, 64(9), 691–706. Birnbaum, H. K. In Hydrogen Effects on Material Behavior; Moody, N. R., Thompson, A. W., Eds.; TMS: Warrendale, PA, 1990; pp 639–660.

23.

24. 25.

26.

27. 28. 29.

30. 31.

103

Boursier, J. M.; Desjardins, D.; Vaillant, F. Corros. Sci. 1995, 37(3), 493–508. Briant, C. L.; Andresen, P. L. In Proceedings of 3rd International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Traverse City, MI, Aug 30–Sept 3, 1998; Theus, G. J., Weeks, J. R., Eds.; TMS–AIME: Warrendale, PA, 1988; pp 371–382. Couvant, T.; Vaillant, F.; Boursier, J. M. Effect of strainpath on stress corrosion cracking of AISI 304L stainless steel in PWR primary environment at 360  C. In Proceedings of Eurocorr 2004, Nice, France, Sept 12–16, 2004; p 11, Event No. 226. Ehrnste´n, U.; Ivanchenko, M.; Nevdacha, V.; Yagodzinskyy, Y.; Toivonen, A.; Ha¨nninen, H. In Proceedings of 12th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Salt Lake City, UT, Aug 14–18, 2005; Allen, T. R., King, P. J., Nelson, L., Eds.; TMS: Warrendale, PA, 2005; pp 1475–1482. Ferreira, P. J.; Robertson, I. M.; Birnbaum, H. K. Acta Mater. 1998, 46(5), 1749–1757. Hall, M. M., Jr. Corros. Sci. 2008, 50, 2902–2905. Hall, M. M., Jr. Corros. Sci. 2009, 51, 225–233. Hong, S. G.; Lee, S.-B. J. Nucl. Mater. 2004, 328, 232–242. McDonald, D. D. J. Electrochem. Soc. 1992, 139, 3434–3449. Kekkonen, T.; Aaltonen, P.; Ha¨nninen, H. Corros. Sci. 1985, 25, 821–836. Angeliu, T. M.; Andresen, P.; Hall, E.; Sutliff, J.; Sitzman, S.; Horn, R. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors. Newport Beach, CA, Aug 1–5, 1999; Ford, F. P., Bruemmer, S. M., Was, G. S., Eds.; TMS–AIME: Warrendale, PA, 1999. Angeliu, T. M.; Hall, E.; Sutliff, J.; Sitzmen, S.; Andresen, P. In Corrosion 2000; NACE International: Houston, TX, 2000; Paper No. 00186. Ehrnste´n, U.; Ha¨nninen, H.; Aaltonen, P.; Jansson, C.; Nenonen, P.; Angeliu, T. In Proceedings of 10th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Lake Tahoe, NV, Aug 5–9, 2001; Ford, F. P., Was, G., Eds.; NACE International: Houston, TX, 2001; p 10. Ooki, S.; Tanaka, Y.; Takamori, K. In Proceedings of 12th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Salt Lake City, UT, Aug 14–18, 2005; Allen, T. R., King, P. J., Nelson, L., Eds.; TMS: Warrendale, PA, 2005; pp 365–376. Yamashita, H.; Ooki, S.; Tanaka, Y.; Takamori, K.; Asano, K.; Suzuki, S. Int. J. Press. Vess. Pip. 2008, 85, 582–592. Andresen, P. L. Corrosion 1988, 44(7), 450. Andresen, P. L. In Proceedings of 10th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Lake Tahoe, NV, Aug 5–9, 2001; Ford, F. P., Was, G., Eds.; NACE International: Houston TX, 2001; p 9. Andresen, P.; Angeliu, T.; Catlin, W.; Young, L.; Horn, R. In Corrosion, 2000, NACE 55th Annual Conference, Orlando, FL, Mar 26–31, 2000; p 12, Paper No. 00203. Andresen, P. L.; Emigh, P.; Morra, M.; Horn, R. In Proceedings of 11th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Stevenson, WA, Aug 11–14, 2003; American Nuclear Society: La Grange Park, IL, 2003; pp 816–833.

104

Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels

32. Andresen, P. L.; Morra, M. M. J. Nucl. Mater. 2008, 383(1–2), 97–111. 33. Horn, R.; Gordon, G.; Ford, P.; Cowan, R. Nucl. Eng. Des. 1997, 174, 313–325. 34. Magdowski, R.; Speidel, M. O. In Corrosion ’96, NACE, 51st Annual Conference and Exposition, Denver, CO, Mar 24–29 1996; p 6, Paper No. 112. 35. Ta¨htinen, S.; Ha¨nninen, H.; Trolle, M. In Proceedings of the 6th Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, San Diego, CA, Aug 1–5, 1993; Gold, R. E., Simonen, E. P., Eds.; TMS: Warrendale, PA, 1993. 36. Prevey, P. S.; Jayaraman, N. In Proceedings of ICSP 9, Paris, Marne la Vallee, France, Sept 6–9, 2005; p 7, Paper No. 260. 37. Offer, H. P.; Morra, M.; Chan, A. In Proceedings of 13th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Whistler, BC, Canada, Aug 19–23, 2007; Allen, T. R., et al. Eds.; Canadian Nuclear Society: Toronto, ON, 2007; p 17. 38. Robertson, J. Corros. Sci. 1991, 32, 443–465. 39. Stellwag, B. Corros. Sci. 1998, 40, 337–370. 40. Ford, P.; Taylor, D.; Andresen, P.; Ballinger, R. Corrosion Assisted Cracking of Stainless Steel and Low Alloys Steel in LWR Environments; Report NP5064S; Electric Power Research Institute: Palo Alto, CA, 1987; p 124. 41. Ford, P. F.; Andresen, P. L. In Corrosion Mechanisms in Theory and Practice; Marcus, P., et al. Ed.; Marcel Dekker: New York, 1995; pp 501–546. 42. Hettiarachchi, S. In Proceedings of 11th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Stevenson, WA, Aug 11–14, 2003; American Nuclear Society: La Grange Park, IL, 2003; pp 477–487. 43. Hettiarachchi, S. In Proceedings of 10th International Conference on Environmental Degradation of Materials in Nuclear Power Systems-Water Reactors, Lake Tahoe, NV, Aug 5–9, 2007; Ford, F. P., Was, G., Eds.; NACE International: Houston, TX, 2007; p 10. 44. Andresen, P. L. Personal communication, May 2009. 45. Andresen, P. L. In Proceedings of 5th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Monterey, CA, Aug 25–29, 1995, American Nuclear Society: La Grange Park, IL, 1995; pp 209–218. 46. General Electric Company. Alternative Alloys for BWR Pipe Application; NP-2671-LD, Final Report; San Jose, CA, 1982. 47. Ha¨nninen, H.; Aho-Mantila, I. In Proceedings of 3rd International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Traverse City, MI, Aug 30–Sept 3, 1987; TMS–AIME: Warrendale, PA, 1987; pp 77–92.

48.

49.

50.

51.

52. 53.

54.

55.

56. 57.

58.

59.

Erve, M.; Wesseling, U.; Kilian, R.; et al. In 20th MPA-Seminar, Stuttgart, Oct 6–7, 1994; Staatliche Materialpru¨fungsanstalt (MPA) Universita¨t Stuttgart: Stuttgart, Germany, 1994; Vol. 2. p 21, Paper No. 29. Kilian, R. In Proceedings of 7th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Breckenridge, CO, Aug 7–10, 1995; Airey, G., Ed.; NACE International: Houston, TX, 1995; pp 529–540. Kilian, R.; Eberle, U.; Bru¨mmer, G.; et al. In Proceedings of the 9th Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Newport Beach, CA, Aug 1–5, 1999; Bruemmer, S. M., Ford, F. P., Was, G. S., Eds.; TMS–AIME: Warrendale, PA, 1999; pp 347–357. International Atomic Energy Agency. Mitigation of Intergranular Stress Corrosion Cracking in RBMK Reactors; Final Report of the Programme’s Steering Committee; IAEA-EBP-IGSCC; IAEA: Vienna, 2002. Timofeev, B.; Fedorova, V.; Buchatskii, A. Mater. Sci. 2004, 40(1), 48–59. Scott, P. M. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Newport Beach, CA, Aug 1–5, 1999; Ford, F. P., Bruemmer, S. M., Was, G. S., Eds.; TMS–AIME: Warrendale, PA, 1999. Couvant, T.; Legras, L.; Pokor, C.; et al. In Proceedings of 13th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Whistler, BC, Canada, Aug 19–23, 2007; Allen, T. R., et al. Eds.; Canadian Nuclear Society: Toronto, ON, 2007; Vol. 1. pp 499–514. Chynoweth, J.; Hyres, J. In Proceedings of 13th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Whistler, BC, Canada, Aug 19–23, 2007; Canadian Nuclear Society: Toronto, ON, 2007; Vol. 2, pp 1214–1225. Kilian, R.; Wesseling, U.; Wachter, O.; Widera, M.; Bru¨mmer, G.; Ilg, U. In Fontevraud V; Sept 23–27, 2002; SFEN: France, 2002. Berge, P.; Keroulas, F.; Gras, J.; Noe¨l, D.; Da Vunha Belo, M. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Jekyll Island, GA, Aug 6–10, 1989; Cubicciotti, D., Ed.; NACE International: Houston, TX, 1989; pp. 11-74–11-87. McDonald, D.; Cragnolino, G.; Olemacher, J.; Chen, T.; Dhawale, S. Intergranular Stress Corrosion Cracking of Austenitic Stainless Steels in PWR Acid Storage Systems; EPRI NP-2531; Electric Power Research Institute: Palo Alto, CA, 1982. Cramer, S., Covino, B., Jr., Moosbrugger, C. Eds. Handbook Volume 13A: Corrosion Fundamentals, Testing and Protection; ASM International: Materials Park, OH, 2003.

5.06 Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels H.-P. Seifert Paul Scherrer Institut, Villigen PSI, Switzerland

J. Hickling Independent Technical Consultant, Prastio-Avdimou, Cyprus

D. Lister University of New Brunswick, Fredericton, NB, Canada

ß 2012 Elsevier Ltd. All rights reserved.

5.06.1

Introduction

107

5.06.2 5.06.2.1 5.06.2.2 5.06.2.2.1 5.06.2.2.2 5.06.2.2.3 5.06.3 5.06.3.1 5.06.3.2 5.06.3.2.1 5.06.3.2.2 5.06.3.2.3 5.06.3.2.4 5.06.3.2.5 5.06.4 5.06.4.1 5.06.4.2 References

Uniform and Flow-Accelerated Corrosion Uniform Corrosion Flow-Accelerated Corrosion Controlling factors Mechanisms and models Service experience and mitigating actions Localized Corrosion and Environmentally Assisted Cracking Pitting Environmentally Assisted Cracking Basic types of EAC and major factors of influence Corrosion fatigue and strain-induced corrosion cracking Stress corrosion cracking EAC mechanisms and models Service experience and mitigation actions Conclusions Uniform and Flow-Accelerated Corrosion Localized Corrosion and Environmentally Assisted Cracking

109 109 111 111 114 118 120 120 122 122 123 128 132 136 139 139 139 140

Abbreviations AC AGR ANL ASME ASME BPV ASME III ASME XI ASTM BWR BWRVIP BWRVIP-60

Content of Cr, Mo and Cu in alloy in EPRI ‘CHECWORKS’ FAC-Code Advanced gas-cooled reactor Argonne National Laboratory, USA American Society of Mechanical Engineers ASME Boiler and Pressure Vessel Code Section III of ASME BPV Code Section XI of ASME BPV Code American Society of Testing and Materials Standards Boiling water reactor Boiling Water Reactor Vessel and Internals Program Basis document for SCC disposition lines for low-alloy steels

CANDUW

CF CRDM CS DCPD

DH DL DO DSA EAC EC ECP ECPcrit

CANada Deuterium Uranium, PHWR developed by Atomic Energy of Canada Ltd. Corrosion fatigue Control rod drive mechanism (housing) Carbon steel (Reversed) direct current potential drop crack length measurement method Dissolved hydrogen (concentration) Disposition line Dissolved oxygen (concentration) Dynamic strain ageing Environmentally assisted cracking Erosion corrosion Electrochemical corrosion potential Critical cracking potential (e.g., for SICC)

105

106

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

EPRI

Electric Power Research Institute, USA F & A model EAC model for CS & LAS developed by P. Ford and P. Andresen (GE GR) FAC Flow-accelerated corrosion FRAD Film rupture anodic dissolution EAC mechanism HAEAC Hydrogen-assisted EAC mechanism HAZ Heat-affected zone of weldment HCF High-cycle fatigue HT High temperature HWC Hydrogen water chemistry JSME Japanese Society of Mechanical Engineers LAS Low-alloy steel LCF Low-cycle fatigue LEFM Linear elastic fracture mechanics LWR Light water reactor MT Mass transfer in EPRI ‘CHECWORKS’ FAC-Code NDT Nondestructive testing NMCA Noble metal chemical addition NRC Nuclear Regulatory Commission, USA NWC Normal water chemistry PHWR Pressurized heavy water reactor PWHT Postweld heat treatment PWR Pressurized water reactor PWSCC Primary water stress corrosion cracking (in PWRs) QþT Quench and temper heat treatment RPV Reactor pressure vessel SCC Stress corrosion cracking SEM Scanning electron microscope SHE Standard hydrogen electrode SICC Strain-induced corrosion cracking SS Stainless steel SSR(T) Slow strain rate (test) SSY Small-scale yielding UTS Ultimate tensile strength VGB German Association of Large Power Plant Operators YS Yield strength

Ceq

Symbols

KI KI,i

C Cb

Concentration of Fe(II) species at the oxide–coolant interface Concentration of Fe(II) species in the bulk coolant

CODLL

d D da/dN da/dNAir da/dNCF

da/dtAir ¼ da/dNAir/ DtR da/dtCF ¼ da/dNCF/ DtR da/dtSCC da/dtSICC dCODLL/dt

de/dt de/dtcrit dKI/dt EA Fen

G h

kc kd

Thermodynamic equilibrium concentration of Fe(II) species Crack-opening displacement at load line in precracked fracture mechanics specimen Pipe diameter Diffusivity Crack advance per fatigue cycle Crack advance per fatigue cycle in air Corrosion fatigue crack advance per fatigue cycle in hightemperature water Time-based fatigue crack growth rate in air Time-based corrosion fatigue crack growth rate in hightemperature water SCC crack growth rate SICC crack growth rate Crack-opening displacement rate in slow rising load or displacement test Strain rate (sometimes locally at crack-tip) Critical strain rate (e.g., for SICC) Stress intensity factor rate in slow rising load or displacement test Arrhenius activation energy of thermally activated process Environmental correction factors, ratio of fatigue life in air at room temperature to that in water at service temperature Geometry factor in EPRI ‘CHECWORKS’ FAC-Code Mass transfer coefficient for Fe(II) species from the oxide–coolant interface to the bulk environment by convection Geometry factor in Siemens-KWU ‘WATHEC’ FAC-Code Dissolution reaction rate constant of magnetite at the oxide–coolant interface Stress intensity factor (LEFM) Stress intensity factor at the onset of SICC crack growth in slow rising load tests with precracked fracture mechanics specimens

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

KIJ

Stress intensity factor at the onset of ductile crack growth m Paris law exponent for fatigue and corrosion fatigue n Frequency exponent for corrosion fatigue pH  log[Hþ] pH value R Load ratio of minimum to maximum load Flow-accelerated corrosion rate RFAC Dissolution rate of magnetite at the Rd oxide–coolant interface Formation rate of magnetite at the Rg metal/oxide interface Mass transport rate of Fe(II) Rm species from the oxide–coolant interface to the bulk environment by convection Re ¼ dur/m Reynolds number Sc ¼ m/( rD) Schmidt number Sh ¼ hd/D Sherwood number t Time T Temperature u Coolant velocity Z Reduction of area in tensile test [X] Concentration of element/species X in water or in alloy a Steam void fraction in EPRI ‘CHECWORKS’ FAC-Code DC ¼ Ceq  Cb Undersaturation in dissolved Fe(II) species Dd Wall thinning DK ¼ KI,max  Stress intensity factor range of KI,min fatigue cycle Upper DK threshold for corrosion DKCF,H fatigue Lower DK threshold for corrosion DKCF,L fatigue DK threshold for fatigue in air DKth, Air Decline time (down-ramp) of DtD fatigue cycle Hold time at maximum load of DtH fatigue cycle Rise time (up-ramp) of fatigue cycle DtR e Mechanical strain Critical strain (e.g., for SICC) ecrit k Specific electrical conductivity m Viscosity of coolant n Loading frequency Upper critical frequency for ncrit,H corrosion fatigue

ncrit,L r s scrit t

107

Lower critical frequency for corrosion fatigue Density Mechanical stress Critical stress (e.g., for SCC) Fluid shear stress at pipe wall

5.06.1 Introduction Carbon and low-alloy steels (CS & LAS, Table 1) and their associated weld filler metals are widely used for pressure vessels and piping in both the primary and secondary coolant circuits of watercooled reactors (light water reactors (LWRs) and CANDUs – pressurized heavy water reactors (PHWRs)), as well as in service water systems.1 The main reasons for the use of CS & LAS are their combination of relatively low cost, good mechanical strength and toughness properties in thick sections (hardenability), and good weldability, as well as their good stress corrosion cracking (SCC) resistance in primary coolant environments. Compared with austenitic stainless steels and nickel-base alloys, ferritic CS & LAS exhibit only moderate corrosion and irradiation resistance. They also show a ductile-to-brittle transition in toughness properties at lower temperatures. CS & LAS components in the primary circuit of pressurized water reactors (PWRs) are clad (usually with austenitic stainless steel) and thus do not generally come into direct contact with the reactor coolant. This is also the case for the reactor pressure vessel (RPV) in boiling water reactors (BWRs), although the RPV head is sometimes left unclad and the cladding has been removed from the blend radius of many RPV feedwater nozzles. In BWRs of German and of newer General Electric designs, extensive use is also made of unclad LAS and CS in both the feedwater and steam lines, as well as in the condensate system. The primary coolant piping in conventional CANDUs is made exclusively of unclad CS. In secondary coolant systems, the steam generator pressure vessel shell is unclad, as are the feedwater, drain, and steam lines. CS & LAS pressure-boundary components, in particular in the primary circuit such as the RPV, are very critical systems with regard to plant safety and lifetime (extension). Minimizing corrosion improves plant availability and economics and is also fundamental for safe operation over extended periods of 50–60 years.

108

Typical CS & LAS piping and pressure vessel materials in Western LWRs (US designation, according to Section II of ASME BPV Code) 

Designation

Type

Product form

Cmax (%)

Mn (%)

Pmax (%)

Smax (%)

Simin (%)

Cumax (%)

Nimax (%)

Crmax (%)

Momax (%)

Vmax (%)

YS25 C (MPa)

Heat treatment

Microstructure

SA 106 Gr. B

CS C–Mn

Pipe drawn

0.30

0.29 1.06

0.035a

0.035a

0.10

0.40a

0.40b

0.40b

0.15b

0.08b

 240 300–400c

Normal.

Ferriticpearlitic

SA 333 Gr. 6

CS C–Mn

Pipe drawn

0.30

0.29 1.06

0.025a

0.025a

0.10

–

–

–

–

–

 240 300–400c

Normal.

Ferriticpearlitic

SA 516 Gr. 70

CS C–Mn

Vessel plate

0.27d

0.79 1.30

0.03a

0.035a

0.13 0.45

–

–

–

–

–

 260 300–400c

Normal.

Ferriticpearlitic

SA 533 B Cl.1

LAS Mn–Mo–Ni

(R)PV plates

0.25

1.07 1.62

0.12e (0.35)

0.15e (0.35)

0.13 0.45

0.10e

0.37 0.73

–

0.41 0.64

0.05

 345 450–550c

Q&T

Bainitic

SA 508 Gr. 3 Cl. 1

LAS Mn–Mo–Ni

(R)PV forging

0.25

1.20 1.50

0.12e (0.25)

0.15e (0.25)

0.15 0.40

0.10e

0.40 1.00

0.25

0.45 0.60

0.05

 345 450–550c

Q&T

Bainitic

SA 508 Gr.2 Cl. 1

LAS Ni–Mo–Cr

(R)PV forging

0.27

0.50 1.00

0.12e (0.25)

0.15e (0.25)

0.15 0.40

0.10e

0.50 1.00

0.25 0.45

0.55 0.70

0.05

 345 450–550c

Q&T

Bainitic

a

In modern steels, these values are less than 0.015%. Combination shall not exceed 1.0%. c Typical range. d Carbon varies with thickness up to 0.31%. e Requirement for core belt region. YS ¼ yield strength; Normal. ¼ normalized; Q & T ¼ quenched and tempered. b

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

Table 1

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

Consideration of both uniform and flowaccelerated corrosion (FAC) behavior for all unclad surfaces is important for corrosion product transport and deposition (e.g., crud formation on fuel elements) but – together with the assessment of resistance to localized corrosion phenomena such as pitting and environmentally assisted cracking (EAC) – is obviously also required for integrity reasons. In the case of EAC, however, safety considerations furthermore require that possible defects extending through the cladding be taken into account, so that the susceptibility of the RPV must be assessed as if no cladding were present. Sometimes, thick pads of Alloy 182 have been welded directly onto the RPV to act as attachment points for internal structures; the higher yield strength of Alloy 182, the thicker section and its known SCC susceptibility raise special concerns for these areas. In such cases, it is possible that SCC or thermal fatigue of the austenitic alloy will occur such that the crack tip propagates to the interface between the austenitic and ferritic alloys. Furthermore, leakage of coolant from the primary circuit in PWRs poses a special hazard for CS & LAS components, since the boric acid it contains can concentrate and lead to uniform corrosion, or ‘wastage,’ of external surfaces. This chapter covers both the uniform and localized corrosion behavior of CS & LAS pressureboundary components in the primary (BWR, PWR, and CANDU) and secondary (PWR and CANDU) coolant systems of Western reactors, whereby the discussion is focused on relevant US nuclear codes and rules together with material standards in this area. Special emphasis in Sections 5.06.2 and 5.06.3 is placed on FAC and on EAC, both of which have resulted in serious pipe ruptures (FAC) or leaks (EAC) during both nuclear and fossil service in the past. In Section 5.06.2, the uniform and boric acid corrosion behavior of CS & LAS, as well as the nature of the protective oxide film on these materials, are summarized first, followed by a condensed review of the FAC behavior of these steels. The major factors controlling FAC, the underlying mechanism and predictive models, as well as the relevant service experience and possible mitigation actions are discussed. After a brief overview of pitting in CS & LAS in the first part of Section 5.06.3, crack initiation susceptibility conditions and crack growth behavior are discussed in detail for the different types of EAC and compared with the relevant design codes and crack growth disposition curves for CS & LAS. This is followed by a review of the mechanistic understanding of EAC and of existing EAC models. LWR service

109

experience and mitigation actions with regard to EAC are then summarized and compared with this experimental and theoretical background knowledge. Finally, Section 5.06.4 summarizes the major conclusions of this review.

5.06.2 Uniform and FlowAccelerated Corrosion 5.06.2.1

Uniform Corrosion

Uniform or general corrosion does not normally cause a problem for the structural integrity of CS or LAS components in nuclear coolant systems. Corrosion rates in typical circuits are generally of the order of a micrometer per year (1 mm year1) or less – higher than those of stainless steel or nickelbased alloys, for example, but quite acceptable. Around 300  C, uniform corrosion rates of CS & LAS are minimal at a slightly alkaline pH300  C of 6–6.5 (neutral high-purity water has a pH300  C of 5.7) and intermediate dissolved oxygen levels. Under some shutdown conditions, however, LWR primary coolant can be aggressive to these materials, in particular in conjunction with increased oxygen levels (e.g., through oxygen ingress from air); below 100  C, corrosion rates may be high. Compact, defect-free oxide films grown at higher temperatures during service are kinetically quite stable at lower temperatures and usually provide sufficient protection against uniform corrosion during short shutdown periods. Nevertheless, reactor vessels and LAS piping in PWRs are clad with stainless steel, which helps reduce the build-up of crud on fuel and of radiation fields by ensuring a high degree of water purity with a low level of dissolved iron. A particular concern in PWRs arises from the leakage of borated coolant from joints such as gasketed flanges and its impingement on components such as flange studs. Up to 2001, some 140 leaks had been reported publicly.2 Solid boric acid at room temperature and dilute, deaerated boric acid solutions regardless of temperature have little effect on CS & LAS, but as the boric acid concentrates, corrosion rates up to about 1 mm year1 may be reached. Aerated solutions can be much more aggressive, with the attack increasing with acid concentration. Note that as hot coolant escapes to the environment, its boric acid content (which may be nominally 2000 ppm (1 ppm ¼ 1 mg kg1; 1 ppb ¼ 1 mg kg1) or more as elemental boron) concentrates by evaporation. At temperatures in the neighborhood of 100  C, which

110

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

are attained by surfaces impacted by coolant flashing to steam, corrosion rates can reach 250 mm year1.2 In some situations, flow effects can exacerbate the attack, as described in Section 5.06.2.2. The resistance of CS and LAS to corrosion is dependent upon the protective properties of the oxide film. Environments such as boric acid that dissolve or erode the oxide then promote corrosion. The predominant oxide on CS and LAS in coolant circuits operating above about 130  C is magnetite – Fe3O4. In deoxygenated alkaline water, the magnetite forms a double layer that has been well characterized in terms of materials performance in boiler systems at temperatures of about 300  C.3 This morphology is found on CS in CANDU primary circuits, and would be found on pressure-vessel steel exposed to PWR primary coolant in the absence of high-alloy cladding. The layers are formed by the simple oxidation of the steel by water: Fe þ 2H2 O ¼ FeðOHÞ2 þ H2

½I

The nascent hydrogen is absorbed by the metal and diffuses to the exterior. Roughly half of the ferrous species (often as the dissolved hydroxide – depending on the pH) are precipitated oxidatively at the metaloxide interface as small crystallites of magnetite, each a few tens of nanometers across, also releasing hydrogen to the coolant: 3FeðOHÞ2 ¼ Fe3 O4 þ 2H2 O þ H2

½II

The precise fraction precipitated is determined by the density of the oxide relative to that of the metal, since the inner layer occupies the volume of metal corroded.3 The remainder of the dissolved iron diffuses through the oxide to the bulk coolant and precipitates according to eqn [II] as an outer layer of magnetite crystals, each several micrometers across, again releasing hydrogen to the coolant. If metal species other than those of iron originate from alloy components elsewhere in a circulating system, they may coprecipitate and modify the locally formed magnetite. An example of double-layer formation is shown in Figure 1. The concentration of dissolved iron in the coolant governs the oxide formation. If the coolant is significantly undersaturated in iron, the outer layer cannot precipitate and the inner layer may even dissolve at the oxide–coolant interface. In nonisothermal systems, temperature gradients create solubility differences and transport iron around the circuit, modifying the oxide films accordingly (the same phenomenon transports different oxides around circuits containing other

Coolant flow Precipitation Dissolution Outer oxide

o/s interface

Inner oxide

m/o interface Corroding metal

Figure 1 Schematic of double layer oxide formation on carbon steel in high-temperature water.

materials, such as the nickel-base alloys in PWRs). Thick films may also spall and release oxide particles to be distributed by the coolant. In circuits connected to the reactor core, oxide transport may create deposits on the fuel, impeding heat transfer and leading to increased radiation fields around out-of-core components (note that the nickel-base alloys and stainless steel in PWRs can produce deposits derived from nickel ferrite, NiFe2O4; on high-burnup fuel undergoing subcooled boiling, these can harbor boron from the coolant and provoke shifts in the neutron flux, as well as affect radiation fields). Evolved hydrogen also affects magnetite solubility (by the one-third power of the concentration – as indicated by eqn [II]). Such increased solubility at the metal–oxide interface has been invoked as the reason for the lack of precipitation within pores as iron diffuses to the oxide–coolant interface.4 Magnetite films formed on steel surfaces that are pressure boundaries, where the hydrogen evolved by eqn [I] continuously effuses through the metal, tend to have a more adherent inner layer of larger crystallites than those formed on totally immersed surfaces such as test coupons, where the evolved hydrogen can only diffuse through the oxide to the bulk coolant once the metal becomes saturated.5 Under neutral oxidizing conditions, magnetite is still the predominant base oxide formed on steels.6 However, since dissolved oxygen becomes the oxidant rather than water, hydrogen generation is suppressed and the basic oxidation reactions become: 2Fe þ 2H2 O þ O2 ¼ 2FeðOHÞ2 6FeðOHÞ2 þ O2 ¼ 2Fe3 O4 þ 6H2 O

½III ½IV

The oxide layers – especially the outer one – then tend to contain the more-oxidized forms maghemite and/or hematite (both of formula Fe2O3), particularly in BWR circuits.7 Under reactor coolant conditions, corrosion rates and oxide solubilities under

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

oxidizing conditions are generally substantially lower than those under reducing conditions. At high oxygen levels, however, the risk for pitting and EAC increase significantly (see Section 5.06.3). The forms of oxide that are thermodynamically stable under various conditions in coolant circuits are indicated by Pourbaix diagrams, which plot the equilibrium potentials of the oxidizing–reducing reactions against pH; the higher the potential, the more oxidizing the environment. For dissolved species, the equilibria and therefore the lines in the diagram are dependent upon the concentration; when illustrating corrosion situations, a concentration of 106 M or less is often assumed. It should be borne in mind, therefore, that such diagrams are mainly indicative in nature and illustrate the possibilities of species formation without taking account of reaction kinetics. Figure 2, adapted from Beverskog and Puigdomenech,8 is an example for species pertinent to steel at 310  C, where a species concentration of 108 M is representative. The hydrogen line in the figure represents the equilibrium: 2H2 O þ 2e ¼ 2OH þ H2 5.06.2.2

½V

Flow-Accelerated Corrosion

5.06.2.2.1 Controlling factors

Flow-accelerated (or -assisted) corrosion (FAC), sometimes called erosion–corrosion (EC) in older literature, 0 −0.2 Fe2O3

−0.4

Fe(OH)4−

−0.6

Fe3O4

E(v)

Fe(OH)+

Fe(OH)2

−0.8

Hydr

ogen

line

Fe(OH)3−

−1 Fe

−1.2 −1.4 6.5

7

7.5

8 pH310 ⬚C

8.5

9

9.5

Figure 2 Pourbaix diagram for iron at 108 m at 310  C. Reproduced from Beverskog, B.; Puigdomenech, I. Corros. Sci. 1996; 38(12): 2121–2135.

111

is essentially the dissolution and erosion of the normally protective oxide film on CS (or LAS with a Cr-content < 0.2 wt%), exacerbated by fluid flow effects, resulting in excessive corrosion rates and substantial pipe wall thinning. Nowadays, the term EC implies the involvement of a significant mechanical component as an abrasive (e.g., by dispersed solid particles in the liquid phase) or cavitation-induced (mechanical) removal of surface material; it should therefore be differentiated from FAC, which is primarily caused by a flow-induced increase in the mass transfer of dissolving and reacting (corrosive) species at high-flow or highly turbulent locations, although fluid shear stress on the oxide film at the material surface may also make substantial contributions to the damage in some situations. FAC is a pervasive problem in most types of steam-raising system and has caused feedwater line ruptures, occasionally with fatal consequences, in both fossil and nuclear plants.9,10 In primary coolant systems also, less serious (though costly) FAC occurs chronically in the CS outlet feeders of conventional CANDUs,11 and flow effects are implicated in the corrosion of PWR pressure-vessel steel by borated coolant leaking through cracked penetrations in the RPV head.12 FAC thus occurs in the regions of high turbulence in both single and two-phase flows, but never in systems with dry steam. FAC depends on hydrodynamics (mainly steam quality, flow rate, fluid shear stress at the wall, turbulence intensity, and mass transfer coefficient), environmental factors (mainly temperature, pH, dissolved oxygen, hydrogen, and iron concentrations) and material parameters (metal composition – Mo, Cu and, in particular, Cr content).9 The critical parameter combinations for the occurrence of FAC in feedwater systems and the main parameter effects are schematically summarized in Figure 3. The conditions leading to increased FAC rates are usually related to regions with turbulent flow, to low electrochemical corrosion potentials ECP (i.e., to chemically reducing conditions), and to low iron concentrations in the water (Figures 3 and 4). Depending on the pH, the maximum FAC rates occur at about 130  C in single-phase flow, and at about 180  C in two-phase flow (in the latter, it is the condition in the liquid layer at the steel surface that controls the FAC rate, but this is difficult to measure or predict). Note that FAC can still be a problem at other temperatures, even though rates are lower. For example, feeder FAC in CANDU primary coolants occurs at 300–310  C at the core outlet, and FAC is also significant in feedwater systems at the low

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

Log FAC rate

112

pH 7

Hydrodynamics Hydrodynamics Hydrodynamics pH 9

• Shear stress at surface • Flow rate • Turbulent intensity • Mass transfer coefficient

Flow

Critical conditions for high FAC risk in feedwater

∼150 °C

Material

Environment

• [Cr] in metal < 0.2%

Log FAC rate

Log FAC rate

pH 7 • Low [Fe] • pH < 9.2 • 120 °C < T < 180 °C • [O2] < 2–40 ppb

pH 7

∼0.2% Cr

pH 9

T

∼40 ppb

Log FAC rate

pH 7

∼1−2 ppb pH 9

Log FAC rate

pH 9 [Cr] in metal

∼130 °C

pH 9

pH [O2]

Figure 3 Critical parameter combinations for flow-accelerated corrosion (derived from Uchida, S. et al. In: Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, CD-ROM. Whistler, British Columbia, Canada, 19–23 August, King, P., Allen, T., Busby, J., Eds.; Toronto, ON: The Canadian Nuclear Society, 2007) and major parameter effects on flow-accelerated corrosion under feedwater conditions.

temperature of condensate extraction. Specific geometries like elbows, bends, protruding weld roots, orifices, and valves cause local turbulence, which significantly increases FAC rates at, or immediately downstream of, the location concerned. Systems such as the moisture-separator/reheater drain lines, where steam has condensed and relatively iron-free water is flowing, are particularly susceptible. In primary coolant systems, there is the desire to keep iron concentrations low to prevent crud build-up and radiation transport problems, hence the frequent use of highalloy materials that are resistant to FAC as cladding. However, it must be recognized that a recirculating system will always tend toward equilibrium; in other words, dissolved iron concentrations on average will vary around solubility values, depending upon oxide dissolution and precipitation kinetics, temperature gradients around the circuit, and the capacity of sinks such as the purification circuit. Most studies of FAC have been performed under feedwater conditions, which generate high rates of attack that can reach several millimeters per year in some situations. Neutral chemistry, low-oxygen conditions at about 140  C, as may be found in BWR

feedtrains, can give high FAC rates, so dual-cycle PWRs or PHWRs routinely add an amine such as ammonia to raise the pH in the secondary coolant circuit. The actual pH employed depends upon the materials of construction; for all-ferrous feedtrains, a pH25  C from 9.3 to 9.6 is usually specified, but the value is kept below 9.2 to avoid excessive corrosion of copper-base alloys, if these are present. Also, to achieve a more even distribution of additive around the circuit, an amine (such as morpholine) with a coefficient of distribution between the steam and liquid phases closer to unity than that of ammonia may be used. Oxygen dissolved in the coolant is also a powerful inhibitor of FAC; it has been added routinely to feedwater systems in BWRs and certain fossil boilers for some time. Depending upon the rate of attack, levels of oxygen from a few ppb to several tens of ppb are sufficient to stifle FAC completely. Maintaining a dissolved oxygen content >30 ppb, which raises the corrosion potential ECP in the feedwater system above the Fe3O4/Fe2O3 phase boundary in the Pourbaix diagram in Figure 2, is particularly crucial in BWRs operating on hydrogen water chemistry

113

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

St37.2 (A414 Gr B-carbon steel)

Flow (kg h−1) 983

491

15Mo3 (A161 Gr T1–0.5% Mo)

907 756 605

378 302 227

15NiCuMoNb5 13CrMo44 (A213 Gr T12–1% Cr, 0.5% Mo) 10CrMo910 (A213 Gr T22–2.2% Cr, 1% Mo) 3000

4.0 pH = 9.04

pH = 7.0

2.0

Increasing flow

Loss rate (mm year−1)

3.0

1.0

Specific material wear rate (μg cm−2 h−1)

1000 300 100 30 10 3.0 1.0 0.3 0 90 (a)

100 110 120 130 140 150 160 170 Temperature (⬚C)

0.1 50 (b)

100

150

200

250

300

Temperature (⬚C)

Figure 4 Effect of temperature, flow rate (a) (data from Bignold, G. J. et al. In: Proceedings of the International Specialist’s Meeting on Erosion-Corrosion of Steels in High-Temperature Water and Wet Steam, Les Renardie´res, France, 11–12 May; EDF: France, 1982) and material (b) (data from Heitmann, H. G.; Schub, P. In: Proceedings of the Third Meeting on Water Chemistry of Nuclear Reactors, pp. 243–252, Bournemouth, UK, October; British Nuclear Engineering Society (BNES): London, UK, 1983) on single-phase flow-accelerated corrosion under different flow and chemistry conditions. Reproduced from Dooley, R. B. Power Plant Chem 2008, 10(2), 68–89.

(HWC) with high rates of hydrogen injection into the feedwater. If HWC is combined with noble metal chemical addition (NMCA), the FAC risk is reduced, since much lower hydrogen injection rates are then adequate to mitigate SCC in stainless steel recirculation piping and reactor internals. (Recombination of hydrogen and oxygen to lower the ECP requires the radiation fields present in the RPV.) Oxygen levels significantly above 50 ppb may increase the risk of strain-induced corrosion cracking and corrosion fatigue in CS & LAS feedwater piping (see Section 5.06.3). Furthermore, the deliberate addition of oxygen to feedwater systems in dual-cycle reactors may pose problems, since residual oxygen entering the steam generators can provoke SCC of the high-alloy steam-generator tubes. Nevertheless, severe FAC of

components in the feed train of advanced gas-cooled reactors (AGRs) has been successfully mitigated since the early 1980s by oxygen additions.13 Material properties have a significant impact on FAC rates, but typically the plant operator has no control over this (unless a replacement of piping is an option). Certain elements in the steel can act to retard FAC, as mentioned earlier; for example, chromium is particularly effective and a concentration of 0.1% in the metal reduces FAC in 180  C ammoniated water and water–steam mixtures at pH25  C 9 by about 70%.14 Moreover, under CANDU primary coolant conditions of 310  C and pH25  C 10.5 (adjusted with lithium), increasing the chromium content of SA-106 Grade B CS from 0.019% to 0.33% reduces FAC by about 50%.11

114

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

5.06.2.2.2 Mechanisms and models

As with uniform corrosion (discussed in Section 5.06.2.1), FAC is governed by the ability of the oxide film to protect the metal. Magnetite forms on the steel at the metal–oxide interface and is degraded at the oxide–coolant interface by fluid flow effects and by dissolution according to the general equation [VI] (which indicates the dependence of the dissolved species on pH under reducing conditions and which is equivalent to eqn [II] for b ¼ 2). The turbulence in the coolant and the solubility of the magnetite are then paramount in determining the severity of the attack. Fe3 O4 þ 3ð2  bÞHþ þ H2 ¼ 3½FeðOHÞb ð2bÞþ þ ð4  3bÞH2 O ½VI with b ¼ 0, 1, 2, or 3. Mass transfer is often assumed to control the mechanism.15 This derives from the postulate that the magnetite film attains a steady-state thickness as it dissolves at the rate Rd at its outer surface in coolant undersaturated in dissolved iron and forms continuously at the metal–oxide interface at the same rate Rg. Since the magnetite formation at the metal–oxide interface accounts for only about half of the corroded metal, the other half diffuses through the magnetite to the oxide–coolant interface, and with the iron from the magnetite dissolution is transported to the bulk coolant at the rate Rm. The FAC rate RFAC is thus twice the dissolution rate Rd of the magnetite at the oxide–coolant interface. This concept of two processes in series – dissolution Rd and mass transfer Rm– leads to the equation for the steady-state FAC rate RFAC¼ dm/dt ¼ Rm¼ 2Rd with all the variables in equivalent units of iron per unit surface and time. 1. Steady-state assumption for the serial process: Rg ¼ Rd ¼ 0:5Rm

½1

2. Dissolution rate of magnetite at the oxide–coolant interface according to eqn [VI] (assuming firstorder kinetics): Rd ¼ 0:5  dm=dt ¼ kd ðCeq  CÞ

½2

where kd is the dissolution reaction rate constant, C is the concentration of Fe(II) species at the oxide–coolant interface, and Ceq is their equilibrium concentration according to eqn [VI], which corresponds to their maximum solubility in the coolant.

3. Transport of Fe(II) species from the oxide–coolant interface to the bulk environment by turbulent mass transfer: Rm ¼ h  ðC  Cb Þ

½3

where Cb is the concentration of Fe(II) species in the bulk coolant and h is the mass transfer coefficient, which is dependent on flow conditions and geometry. From eqns [1]–[3] it follows that: RFAC ¼

h  kd  DC ð0:5  h þ kd Þ

½4

where DC ¼ ðCeq  Cb Þ is the undersaturation in iron. Models of FAC are based on the principles behind eqn [4]. We expect that kd strongly increases with temperature according to an Arrhenius law for a thermally activated process (although there are no data to confirm this over the temperature ranges of interest), whereas h only shows a moderate increase through the temperature dependence of the properties in eqn [6]. If mass transfer controls, h is small compared with kd (h kd) and the equation reverts to: RFAC ¼ hDC

½5

For a coolant of constant conditions containing little or no dissolved iron (i.e., Cb  0), the driving force DC approaches a constant value – the solubility of the oxide, Ceq – and RFAC varies as the mass transfer coefficient (which increases with increasing flow rate and turbulence). The mass-transfer model then implies that the effects of materials composition and coolant chemistry on FAC rate are brought about by their effects on oxide solubility (Figure 5). According to eqn [VI], the saturation concentration or solubility Ceq depends on temperature, pH and H2 concentration by simple chemical equilibrium thermodynamics. Accordingly, the effect of chromium in the steel can be attributed to the relative stability of mixed oxides containing chromium (iron chromite, FeCr2O4, for example, is virtually insoluble in reducing coolant and accounts for the protection afforded by stainless steel and similar alloys). As corrosion proceeds and the magnetite dissolves, chromium is not leached out in concert but continually concentrates in the film. It is interesting to note that the inhibition occurs immediately at the start of exposure and continues at about the same level, suggesting that the mechanism is the rapid formation at the metal–oxide interface of a more protective layer of oxide that is maintained throughout exposure.16 It appears that the higher the chromium content of the steel, the more protective that

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

100 NH3 (mg kg−1) pH 0.1 80

Fe (μg kg−1)

60

8.75

0.2

8.90

0.3

9.00

0.5

9.20

1.0

9.40

2.0

9.60

300

350

40

20

0 0

50

100

150

200

250

Temperature ( ⬚C)

Figure 5 Solubility of magnetite/iron as a function of temperature at various ammonia concentrations. Reproduced from Dooley, R. B. Power Plant Chem 2008, 10(2), 68–89.

oxide layer; hence, the immunity of stainless steel and certain LAS like 13CrMo44 or 10CrMo910. These observations cannot be fully explained with the DC term in the simple mass-transfer model of eqn [5], which would indicate an increasing inhibition with time as the chromium concentrates in the film. One mode of inhibition of FAC by additives in the coolant may be via incorporation in the magnetite to make it less soluble. From the experience with chromium, whatever species is added should be available to affect the metal–oxide interface consistently, presumably by being kept permanently in solution in the coolant. So far, titanium has shown promise as a coolantborne inhibitor of FAC under CANDU primary coolant conditions. Its effectiveness has been attributed to its ability to form a mixed oxide with iron – ulvo¨spinel – along with the magnetite that forms on corroding CS.17 An in-plant demonstration of titanium addition to a CANDU primary system is described in Section 5.06.2.3. The simple mass-transfer model also indicates that temperature should affect FAC partly through its influence on magnetite solubility. In ammoniated

115

water at pH25  C 9.0, there is a strong temperature dependence and the maximum FAC rate occurs between 130 and 140  C, depending on flow rate.18 The solubility of magnetite under the same conditions increases from the range 5–15 ppb at 25  C to a maximum of about 30 ppb that persists over the range 110–150  C,9,19 while mass transfer coefficients at the same mass flow should approximately double between 25 and 140  C. Similarly, in neutral water, the maximum attack for several materials occurs at about 150  C,20 while the magnetite solubility increases from about 70 ppb at 25  C to a maximum of about 140 ppb at 120–130  C. The rough correspondence between the temperatures of maximum FAC rate and of maximum magnetite solubility, as well as the effect of temperature itself on solubility, indicate the strong influence of oxide film dissolution on the FAC mechanism. It is likely that at low temperatures dissolution rates of magnetite are low enough for kd to have an effect through eqn [4] and lower the flow dependence accordingly.18 The inhibiting effect of amines and high pH at feedwater temperatures should also be realized mainly through the solubility of magnetite. Thus, in neutral water at 140  C, the solubility of magnetite is about 119 ppb, but if the pH25  C is raised with ammonia to 9.2, the value falls to the range 14–26 ppb.9,19 This would suggest that a reduction in FAC rate by a factor of 8.5–4.5 might be expected from ammoniating the coolant to pH25  C 9.2; however, experiments indicate a reduction by a factor of only about 2.16 It is also instructive to consider the mass transfer implications of the model according to eqn [5]. Mass transfer in pipe flow in aqueous systems can be described via a correlation of the mass transfer coefficient h with dimensionless numbers: h ¼ Sh

D D ¼ ARe p Sc q d d

½6

in which the Sherwood number, Sh (given by hd/D, where d is the pipe diameter and D is the diffusivity), characterizes mass transfer in terms of the flow (via the Reynolds number, Re, given by dur/m, where u is the coolant velocity, r is the density, and m is the viscosity) and physical properties (via the Schmidt number, Sc, given by m/(rD)); A, p, and q are constants. Typically, experiments on mass transfer of dissolved species yield values between about 0.6 and 0.9 for the exponent p.21,22 Recent experiments in a water loop on FAC under neutral conditions at 140  C, however, indicated that the FAC rate RFAC correlated rather weakly with Re.1.2,23 An alternative

116

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

mass transfer analysis gave an excellent correlation with fluid shear stress at the pipe wall, t: Ru ¼ Pt

½7

FAC rate ⫻ Flow rate (mm year–1⫻m s–1)

where P is a constant (see Figure 6). Thus, a steel containing 0.019% chromium gave the correlation RFACu ¼ 0.07t, while a steel containing 0.001% chromium in parallel experiments gave RFACu ¼ 0.18t, where RFAC is in units of millimeters per year, u is in meters per second, and t is in pascals.16 The predominance of mass transfer in developing such correlations depends upon the dissolution rate constant, kd in eqn [4], being large enough to make the mass transfer coefficient, h, controlling. This would seem to be valid under neutral chemistry conditions, where the solubility of magnetite is high, but under high-pH conditions, where the solubility is reduced, kd may be reduced also and its influence may become significant. However, although recent 80

indications24 are that FAC in 140  C ammoniated water at pH 9.2 is not correlated well by the simple mass-transfer model leading to eqn [6], those experiments also indicated a greater dependence on flow rate or shear stress, viz., t raised to the power 1.5–2.0. This cannot be attributed to an increasing influence of kd in eqn [4]; apparently, a different mechanism is involved. Surface texturing usually accompanies FAC. In steam–water mixtures, ‘tiger-striping’ is caused by the streaming pattern of the liquid film on the surface, while in single-phase water, ‘scalloping’ sculpts the attacked surface with grooves, flutes, or shallow depressions (Figure 7(a) and 7(b)). However, in experiments in neutral water at 140  C, in which corrosion rates of several millimeters per year were seen in tubular test sections, a low-chromium steel developed no scallops, even though it corroded at more than twice the rate of a higher chromium steel that developed distinct scallop patterns.23 The scalloping that was seen was approximately related to the pipe flow via a characteristic ‘scallop Reynolds number’:

70

Resc ¼ 1:55 104  2:6 103

60 50 40 30 20 10 0 0

200

400

600

800

1000

1200

Shear stress τ (N m−2)

Figure 6 Correlation for flow-accelerated corrosion at 140  C in neutral water: carbon steel with 0.019% Cr.

in which the characteristic dimension is the average scallop spacing. While the scallops were formed by the corrosion of the metal, it was significant that distinct oxide forms developed and were related more to scallop crests than to valleys. Those forms, shown in Figure 7(c), occurred over pearlite grains in the metal and may be described as ‘coral-like.’ They provide further confirmation of the importance of oxide dissolution in the mechanism, since they are no doubt formed by the different solubilities of the different compositions of oxide overlaying the lamellae of cementite and ferrite in the pearlite. As the magnetite

200 μm

(a)

½8

3 μm

(b)

Diameter of piping

(c)

Figure 7 Surface texturing in flow-accelerated corrosion. (a) ‘Scalloping’ or ‘horseshoe pits’ in single-phase flow. (b) ‘Tiger striping’ in two-phase flow. (c) ‘Coral-like’ oxide on carbon steel undergoing flow-accelerated corrosion in neutral water at 140  C. (a) and (b) courtesy of COMSY™ with permission of AREVA NP.

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

generally dissolves, that over the cementite lamellae is less soluble and is left standing proud. It was noted in the experiments that the ‘coral’ oxide concentrated about 50% more chromium than the surrounding oxide on the ferrite grains, possibly because the underlying pearlite contained more chromium as carbide associated with the cementite. Loop studies using tubular test sections of the material of interest16 under reactor feedwater conditions establish the basis for adding oxygen with minimal residual concentrations left at the end of piping systems. At 140  C in neutral water, about 40 ppb of dissolved oxygen are required to stifle FAC, whereas at pH25  C 9.2 with ammonia, only 1–2 ppb are required. The concentrations required for stifling are related to the measured FAC rates and it is clear that – as oxygen is progressively added to the corroding system – the cathodic reaction of water being reduced to hydrogen is replaced by oxygen reduction; at the stifling concentration, the oxygen sink disappears and with continuing addition its concentration in the loop jumps sharply. However, although there is an obvious relationship between the FAC rate at stifling and the stoichiometric flux of oxygen by mass transfer to the surface, a straightforward linear correspondence may not apply.13 While several mechanistic models of FAC in feedwater systems based mostly on the principles behind eqn [4] have been developed, empirical models have been applied extensively for some time. In the 1980s, for example, parametric studies at the laboratories of the then Siemens-KWU led to the formulation of a correlation between pipe wall thinning Dd and the system variables u (flow velocity), T (temperature), pH, O2 (oxygen concentration), M (materials composition – Cr, Mo, and Cu), and t (exposure time): Dd ¼ kc f ½u; T ; pH; O2 ; M; t 

½9

where kc is a geometry factor. The correlation was developed initially from data for single-phase water flow, but was adapted to two-phase steam-water flows, with the bulk velocity u substituted by the mean velocity of the annular film of water covering the pipe wall. The resulting computer code, ‘WATHEC,’ was restricted to steels with the content of Cr plus Mo less than 5% and exposure times greater than 200 h. The predictions of wall thinning for a large number of situations were equal to or greater than the measured values in 85% of the cases – in other words, the code was considered to be suitably conservative.25 Later, the data management tool ‘DASY’ was added to the code. The EPRI-sponsored computer code ‘CHECWORKS™’ combined an empirical equation, which

117

had some basis in mechanisms such as that leading to eqn [4], with a comprehensive data management scheme.26 The data management includes analysis of ultrasonic test data, calculation of critical wall thickness for components at risk, and organization of pertinent databases. The FAC rate RFAC is written as a function of the system variables: RFAC ¼ f ½T ; AC; MT; O2 ; pH; G; a

½10

where T is temperature; AC is alloy content of Cr, Mo, and Cu; MT is mass transfer; O2 is concentration of dissolved oxygen; G is a geometry factor; and a is the steam void fraction. The factors in eqn [10] are interrelated and the equation is nonlinear. While the absolute predictions of RFAC in CHECWORKS™ are not generally of high precision, iterations incorporating plant measurements can identify the locations of risk and can rank components in the order of vulnerability.27 The FAC of CS is most pronounced under feedwater conditions, but it also occurs at higher temperatures in the primary coolant systems of PHWRs. The phenomenon was identified in the late 1990s at the Point Lepreau CANDU in New Brunswick, Canada, where surfaces of affected outlet feeders of CS were scalloped and the wall thinning rates plotted against coolant velocity indicated a dependence on the velocity raised to the power 1.5.28 Regions of high turbulence, such as the tight-radius bends close to the reactor face, were more severely affected. It was also noted that the coolant at the core outlet was unsaturated in dissolved iron, since it entered the core at 265  C saturated after its passage through the steam generators of nickel alloy and the inlet feeders of CS; as its temperature rose in the fuel channels the solubility at the high pH rose in concert (the CANDU core contains no iron-bearing alloys, so it cannot act as a source of dissolved iron). Although the high-turbulence (and therefore highmass-transfer) regions are again the most affected in primary coolant FAC, it is unlikely that the mechanism in primary coolant is straightforward mass-transfer control based on eqn [11]. First of all, the velocity dependence is too high (the power of 1.5 rather than 0.6–0.9 as expected from correlations such as eqn [6]). Second, measurements of the dissolution rate of magnetite under chemistry conditions close to those of CANDU coolant29 have given a value of kd in eqn [4] very much lower than the mass transfer coefficient h, which would put the mechanism squarely under dissolution control with no velocity effect at all. The alternative theory proposed is that dissolution of magnetite works in synergy with fluid shear stress at the

118

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

surface to degrade the oxide. Thus, the loosening of the magnetite crystallites in the film makes them susceptible to removal by the fluid forces and as they are eroded away the film becomes less protective. A mathematical model developed on this principle was able to predict quite well the thinning of the walls of outlet feeders at an operating plant in terms of the development of the oxide film, the pattern of attack around representative bends, and the corrosion potential ECP and velocity-dependence of FAC rate in individual feeders.30 The model was adapted for predicting corrosion under conditions when the coolant is saturated in dissolved iron and gave reasonable predictions of oxide film growth and general corrosion in CANDU inlet feeders, where corrosion rates are quite low. It is probably more than a coincidence that FAC under these primary coolant conditions, when magnetite solubility is low, seems not to be controlled directly by mass transfer, while similar indications apply under feedwater conditions at high pH, when solubilities are also low. The parallel between the two situations could be clarified if measurements of kd under feedwater conditions were available and the measurements under primary coolant conditions were verified. The high rate of general corrosion of CS caused by aerated concentrated solutions of boric acid originating from leaking PWR coolant was described in Section 5.06.2.1. Some of the studies that quantified the attack were done with dynamic systems, such as evaporating sprays, and it became clear that flow has an effect.2 Of immediate concern is the corrosion of RPV steel caused by borated coolant leaking through cracked penetrations housing control rod drive mechanisms. At the Davis Besse PWR in 2002, such corrosion had threatened the integrity of the vessel. The sequence of events that can lead to cavity formation next to a nozzle was postulated12 to be in three phases: initially, slow seepage of coolant into the external annulus (crevice) in the head would be accompanied by low corrosion rates; next, when the crevice had opened enough and the crack had lengthened to give substantial leak rates, an evaporating coolant jet would accelerate the attack through flow effects; finally, leakage into a cavity would create a turbulent evaporating pool, extending the attack sideways. An extensive testing program sponsored by the Electric Power Research Institute (EPRI), Palo Alto, California, investigated the phases of boric acid attack at Davis Besse. The second phase, which experienced substantial flow effects, was simulated with laboratory experiments in which a flashing jet of borated coolant was directed onto a heated sample

of pressure-vessel steel and the damage assessed in terms of system parameters – notably, coolant chemistry and flow rate.31 Volumetric (or massive) metal loss was correlated with volumetric coolant flow and seemed to behave differently from metal penetration, which was correlated with jet velocity. FAC was in evidence through miniature scallops in the damage craters that formed around (but some distance away from) the points of jet impact. Metal loss rates attained about 3 cm3 year1 at a flow rate of 200 ml min1 with a boric acid concentration equivalent to 1500 ppm [B] and pH300  C of 6.9 adjusted with lithium; the rate depended on the volumetric flow in the jet raised to the power 0.84. Under the same chemistry conditions, the penetration rate reached 200 mm year1 at a jet velocity of 140 m s1 and the two were correlated via the velocity raised to the power 4.3. It was notable that neither pH300  C nor the boron concentration was the controlling chemistry parameter; rather, it was the ratio [B]/[Li]. 5.06.2.2.3 Service experience and mitigating actions

Many incidences of feedwater pipe thinning by FAC from two-phase coolant were reported in the 1980s. In 1985 March, a line downstream of a level control valve for a feedwater heater at the Haddam Neck PWR actually ruptured because of FAC induced by flow-impingement. However, the first major incident in a nuclear plant was the catastrophic pipe break at the Surry Unit 2 PWR in December 1986, which led to five deaths and several injuries. The 0.46 m diameter line thinned and ruptured at an elbow, 0.3 m from a 0.6 m header, as a section of the pipe wall 0.6 1.2 m was blown out. Until then, FAC by steam–water mixtures had been considered to be more serious than FAC by single-phase coolant. Six months later, excessive thinning of a feedwater line was reported at the Trojan plant and, in September 1988, Surry Unit 2 reported 20% wall loss in the suction line to a feedwater pump over a 1.2-year period. Reports of serious thinning of feedwater piping continued after the Surry incident, even though plant inspections had generally become more rigorous and chemistry control had tightened. In May 1990, the Loviisa Unit 1 WWER (Eastern type PWR) in Finland suffered a break in a 0.3 m diameter line in the turbine hall, releasing about 50 m3 of steam and water into the building, and in February 1993, a similar incident occurred in Unit 2. The latest major FAC incident in a nuclear plant was the rupture of a feedwater line at the Mihama

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

Unit 3 PWR in 2004, which led to four deaths and seven serious injuries.32 The thinning of the pipe wall from 10 to 0.6 mm by FAC caused a large section to peel back after rupture, allowing the coolant at

Figure 8 The ruptured feedwater line at Mihama-3.

140  C to flash to steam on release. The damaged pipe is shown in Figure 8. Maximum thinning rates were observed just downstream of an orifice plate, where turbulence intensity was high. It should be noted that chemistry had been maintained at high pH with ethanolamine and that hydrazine was used to scavenge oxygen. However, the location had not been inspected since the plant start-up in 1976. The possibility of adding oxygen to the feedwater is being considered and inspection procedures have been revised extensively. In 1997, at the Point Lepreau CANDU PHWR in New Brunswick, Canada, the outlet feeder pipes of CS that carry the heavy water coolant from the core to the steam generators were found to be corroding excessively. The same problem has since then been seen at all CANDUs in operation before 2000. Figure 9 shows the arrangement of pipes at a reactor face. The feeders are about 76 mm diameter and carry

3

4

1

119

2 5

9

7

6

8 10

1. Reactor outlet header 6. Calandria end shield face 2. Reactor inlet header 7. Tube spacers 3. Reactor outlet header 8. Support brackets 4. Reactor inlet header 9. Walkway 5. Feeder tube upper supports 10. End fittings

Figure 9 CANDU reactor face; cutaway view of feeder pipe arrangement, with permission from Atomic Energy of Canada Limited.

120

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

coolant at velocities between about 12 and 22 m s1. At the coolant temperature of 310  C and pH  10.6 (adjusted with lithium), the solubility of magnetite is relatively low – about 1.7 ppb – and the FAC rates accordingly attain only about 120 mm year1. The attack is only a fraction of that observed under feedwater conditions and has caused no safety issues, but it means that feeder integrity may be lifelimiting and has necessitated replacements at some reactors. Mitigating actions have been taken by reducing the primary coolant alkalinity to the bottom of the recommended pH range, where the magnetite solubility is close to the minimum, while plants in service since 2000 employ feeders with a relatively high chromium content (0.3% in contrast with the 0.02% in earlier reactors). One trial of a coolant additive was made at the Darlington Unit 3 900 MW CANDU in 2002, when a titanium dioxide slurry was added to one channel to give a concentration of ten or so microgram per kilogram at the outlet feeder33; significant reductions in FAC rate were recorded, but it was decided not to undertake the further development needed to proceed to the next stage of full-plant addition. Meanwhile, feeder replacement has become a manageable – if costly – operation at severely affected plants. The boric-acid corrosion at the Davis Besse PWR in Ohio, which in 2002 was found to have a large wastage cavity in the RPV head adjacent to a penetration housing a control rod drive mechanism, was postulated to be partly due to FAC (see Figure 10). About 2040 cm3 of the pressure-vessel steel had corroded away and about 106 cm2 of stainless steel cladding were exposed at the bottom. Substantial quantities of solid boric acid had deposited close by. Subsequent investigations12 determined that coolant had leaked from a crack in the adjacent Alloy 600 nozzle into the surrounding annulus and in time had widened it. The crack was an example of primary water stress corrosion cracking (PWSCC), of which numerous incidences have been recorded in PWRs. No simple means of mitigation have been proposed for existing plants, since the coolant boron level is fixed by reactivity considerations. Long-term prevention entails avoiding operating with coolant leaks (as is, in fact, the regulation in some jurisdictions), for example, through minimizing the possibility of PWSCC by using less-susceptible Alloy 690 material for the CRDM penetrations. In the meantime, more rigorous inspection regimes are being implemented.

Figure 10 Cavity in the reactor pressure vessel head at the Davis Besse pressurized water reactor.

5.06.3 Localized Corrosion and Environmentally Assisted Cracking 5.06.3.1

Pitting

In CS & LAS, a shallow form of pitting can occur in the complete absence of anionic water impurities as the electrochemical corrosion potential (ECP) at the steel surface is raised, for example, through oxygen and other oxidizing species. Such corrosion pits often, but not exclusively, initiate at MnS inclusions which intersect the steel surface. Figure 11, originally compiled by Hickling, shows the critical boundaries between uniform surface attack and shallow pitting in high-purity water at low flow rates as a function of temperature and dissolved oxygen level. The critical oxygen concentration for pitting drops with decreasing temperature and is further reduced by a simultaneous mechanical straining of the surface, or by increased sulfate and chloride impurity levels. Furthermore, pitting in CS & LAS is favored by high steel sulfur contents and quasi-stagnant flow conditions. In the absence of impurities, increasing the flow rate of water across the metal surface mitigates the aforementioned form of pitting corrosion. Alkalization

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

shifts this boundary to much higher values, as does the introduction of buffering and passivating species (e.g., on the secondary side of steam generators).34–37 Some degree of pitting corrosion is inevitable after long-term exposure of unclad CS & LAS surfaces to water in LWR systems and is not usually a threat to either coolant purity or to structural integrity. Shallow pitting has been observed primarily in specific piping systems with residual water because of

121

incomplete draining during nonoperational periods (shutdown corrosion). This can be avoided by adequate wet or dry preservation techniques. If pitting happens during normal plant operation at high temperatures, however, it indicates conditions under which EAC may also occur (since this is controlled by similar parameters, see Section 5.06.3.2) and can even be directly implicated in the initiation of EAC (Figure 12).

300

ck

ce

ta at

g

ce

tin

fa

pit

ng

ur

tti

ls ra ne Ge

Pi

ne

ra

ls

ur

fa

200 Ge

Temperature (⬚C)

at

ta

ck

Pitting with straining

Pitting without straining

100

SSRT

0 10

lzumiya, Tanno Videm Mizuno et al. Ford

Coupon specimens

1000

100

10 000

Dissolved oxygen content DO (ppb) Figure 11 Boundaries between uniform corrosion and pitting attack in carbon and low-alloy steel in quasi-stagnant high-temperature water. Compiled from The general and localized corrosion of carbon and low-alloy steels in oxygenated high-temperature water. EPRI-NP-2853; EPRI: Palo Alto, CA, 1983; http://my.epri.com/; Electric Power Research Institute. BWR environmental cracking margins for carbon steel piping. EPRI Report NP-2406; EPRI: Palo Alto, CA, 1982; http://my.epri.com/; Indig, M. et al. Rev Coat Corros 1982, 5, 173–225; Videm, K. In Proceedings of the 7th Scandinavian Corrosion Congress, Trondheim, Norway, May 26–28, 1975; pp 444–456.

(a) 20 mm

(b)

Acc.V Spot Magn 20.0 kV 5.9 1696x

20 mm Det WD SE 11.0 5/1;250C;HigN

Figure 12 Strain-induced corrosion cracks initiating from a corrosion pit or a (dissolved) MnS inclusion at the surface of a low-alloy and carbon steel in high-temperature water in slow strain rate experiments. Adapted from Congleton, J. et al. Corros. Sci. 1985, 25, 633–650; Atkinson, J. D. et al. In Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Whistler, British Columbia, Canada, Aug 19–23; King, P., Allen, T., Busby, J., Eds.; The Canadian Nuclear Society: Toronto, Canada, 2007; CD-ROM.

122

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

5.06.3.2 Environmentally Assisted Cracking 5.06.3.2.1 Basic types of EAC and major factors of influence

EAC is used here as a general term to cover the full spectrum of corrosion cracking ranging from stress corrosion cracking (SCC) under static load to corrosion fatigue (CF) under cyclic loading conditions (Table 2).38,39 Strain-induced corrosion cracking (SICC) involving slow, dynamic straining with localized plastic deformation of material, but where obvious cyclic loading is either absent, or is restricted to a limited number of infrequent events such plant startup and shutdown, is increasingly used as an appropriate term to describe the area of overlap between SCC and CF.38,39 Under critical parameter combinations, EAC is observed in all wrought and welded CS & LAS in high-temperature water. The EAC crack path is usually perpendicular to the direction of maximum Table 2

tensile stress and transgranular in nature, with a quasicleavage appearance showing a feathery morphology at high magnifications. The general fracture appearance is similar for SCC, SICC, and even CF (at least for strong environmental acceleration of fatigue crack growth), thus confirming that EAC is governed by the same basic process for all three loading modes. In the case of cyclic loading at frequencies 103 Hz, the fracture surface also usually contains both ductile and brittle fatigue striations, which are perpendicular to the local crack-growth direction.38 EAC initiation and growth in CS & LAS are governed by a complex interaction of environmental, material, and loading parameters, and most influencing factors are both interrelated and synergistic. The major parameters of influence, which have been identified so far, are summarized in Table 3.38,39 The effect of these parameters on EAC initiation and crack growth (including key thresholds) is

Basic types of EAC in CS & LAS and relevant nuclear codes

Mechanism

Environmentally assisted cracking (EAC) CF Corrosion fatigue

SICC Strain-induced corrosion cracking

SCC Stress corrosion cracking

Type of loading

Cyclic: low-cycle, high-cycle

Static

LWR operation condition Characterization of crack growth Characterization of crack initiation

Thermal fatigue, thermal stratification, . . . ASME XI Code Case N-643 (PWR) ASME III Fen-approach

Slow monotonically rising or very low-cycle Start-up/shut-down, thermal stratification High-sulfur line of F & A model as upper bound Susceptibility conditions: ECPcrit, de/dtcrit, ecrit

Transient-free, steady-state power operation BWRVIP-60 disposition lines scrit

Source: Seifert, H. P.; Ritter, S. Research and service experience with environmentally-assisted cracking of carbon & low-alloy steels in high-temperature water. SKI-Report 2005:60; SKI: Stockholm, Sweden, 2005; ISSN 1104-1374. http://www.stralsakerhetsmyndigheten. se/. Hickling, J. et al. PowerPlant Chem. 2005, 7, 31–42.

Table 3

Major influencing factors for EAC in C & LAS

Environmental parameters

Material parameters

Loading parameters

Corrosion potential, dissolved oxygen content Temperature

S-content, morphology, size, spatial distribution and chemical composition of MnS

2  Cl, SO2 4 , S , HS

Susceptibility to dynamic strain ageing, Concentration of interstitial C and N Hardness/yield stress if >350HV5/800 MPa

Frequency, loading or strain rate Level of load, KI, stress, strain, DK Type of loading

Flow rate

Residual stress

Source: Seifert, H. P.; Ritter, S. Research and service experience with environmentally-assisted cracking of carbon & low-alloy steels in hightemperature water. SKI-Report 2005:60; SKI: Stockholm, Sweden, 2005; ISSN 1104-1374. http://www.stralsakerhetsmyndigheten.se/. Hickling, J. et al. PowerPlant Chem. 2005, 7, 31–42.

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

discussed in detail in Seifert and Ritter38 and an interpretation of their synergism is given both there and in Section 5.06.3.2.4. 5.06.3.2.2 Corrosion fatigue and straininduced corrosion cracking 5.06.3.2.2.1 Initiation and susceptibility conditions

Slow strain rate (SSR)38,40 and low-cycle fatigue (LCF) tests40–42 with smooth specimens have clearly shown that CF and SICC can occur in CS & LAS in oxygenated, high-purity, high-temperature water if the following conjoint threshold conditions are simultaneously satisfied: Water temperature >150  C. In LCF experiments, susceptibility then increases with temperature up to 320  C. SSR tests, on the other hand, usually indicate a maximum of susceptibility between 200 and 270  C, depending on strain rate. Corrosion potential > ECPcrit ¼ 200 mVSHE or dissolved oxygen content >30 ppb. Above this threshold, EAC susceptibility then generally increases with increasing ECP/oxygen content, but saturates at very high levels. Loading which leads to (local) macroscopic strains at the water-wetted surface above the elastic limit. The susceptibility then increases strongly with the degree of plastic strain. SSR experiments with tapered specimens, and LCF tests with different waveforms, indicate a minimum critical strain of 0.15–0.2%, which is in a similar range to typical oxide film rupture strains on CS & LAS (0.05– 0.2%) in high-temperature water. Positive strain rates below 103 s1. The EAC susceptibility then increases with decreasing strain rate de/dt. In most LCF investigations, saturation of the decrease in fatigue life is observed below a strain rate of 105 s1, but SSR tests indicate a maximum of susceptibility between 105 and 107 s1, depending on ECP and temperature. EAC susceptibility increases with increasing steel sulfur content and a lower threshold is often quoted at 0.003 wt%, but experimental evidence for the latter is weak. Distinct material susceptibility to dynamic strain aging (DSA) in the critical temperature/strain-rate range, or a low yield stress, may also favor crack initiation by SICC. If one or other of these conjoint threshold conditions is not satisfied, SICC initiation is extremely unlikely and no, or only a minor, environmental reduction of

123

fatigue life is observed in high-temperature water. Furthermore, in high-purity water, a high flow rate may completely suppress SICC susceptibility and significantly retard CF crack initiation (in particular, for small strain amplitudes or slow strain rates) compared to quasi-stagnant conditions, since the risk for the formation of an aggressive, occluded water chemistry within small surface defects is significantly reduced by convection. Note, however, that high levels of chloride or sulfate may extend the range of susceptibility to less severe conditions (e.g., to lower ECP and strain levels). The range of system conditions where EAC crack growth from incipient cracks may occur is significantly extended compared to the initiation susceptibility conditions specified earlier. For example, CF crack growth has been observed in high-purity PWR water at ECPs below 500 mVSHE under certain cyclic loading conditions (102 to 10 Hz).38 Apart from local stress raisers such as welding defects, which may help overcome the strain threshold in the field, the effect of initial surface condition (surface roughness, cold work, residual stress, oxide film, and preoxidation) on SICC and CF initiation is much less pronounced than for (high-cycle) fatigue in air, or with SCC of stainless steels or Ni-base alloys. SICC cracks usually, but not exclusively, initiate at MnS inclusions or corrosion pits.38,40,43,44 Pitting, particularly if occurring actively, therefore facilitates SICC initiation (Figure 12). CF cracks, on the other hand, initiate mainly along slip bands, carbide particles, or at the ferrite–pearlite phase boundary, and less frequently at micropits or MnS inclusions.40–42 The effect of pitting and MnS inclusions on CF initiation is thus moderate, but may become more pronounced in the case of deep, high-aspect-ratio pits, mild environmental conditions, or at small strain amplitudes.40 5.06.3.2.2.2 SICC initiation and crack growth from incipient cracks

In high-purity water in the absence of any significant fatigue contribution, CS & LAS show distinct SICC susceptibility only in highly oxidizing environments. For example, it is almost impossible to initiate relevant SICC crack growth in precracked specimens in slow rising-load tests with constant load rate at KI values <70 MPa m1/2 in high-purity water at an ECP of <100 mVSHE. Even under highly oxidizing conditions (ECP  þ50 mVSHE), KI values of 25 MPa m1/2 have to be exceeded to initiate SICC in slow, risingload experiments in high-purity water. A maximum in SICC initiation susceptibility (i.e., a minimum in

124

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

(a)

101

102

60

103

SA 533 B Cl. 1, 0.018% S 288 ⬚C, 8 ppm O2’ 65 ppbSO24−

10−6

50

10−7

45 10−8

40 35 30 10−7

10−9 10−6

10−5

10−4

10−3

−1

Crack opening displacement rate dCODLL/dt (mm s )

(b)

10−8

50

40

10−9

30 Weld filler material, 0.007% S dCODLL/dt = 2–4 ⫻ 10−6 mm s−1 20 100

10−2

8 ppm O2, 65 ppp SO24−

150

200

250

300

SICC crack growth rate (ms−1)

55

100

Stress intensity factor at onset of SICC crack growth Kl,i (MPam1/2)

60

SICC crack growth rate (m s−1)

Stress intensity factor at onset of SICC crack growth Kl,i (MPa m1/2)

Stress intensity factor rate dKl/dt (MPa m½ h−1) 10−1

350

Temperature (⬚C)

Figure 13 Effect of loading rate (a) and temperature (b) on Strain-induced corrosion cracking initiation and crack growth from incipient cracks in oxygenated high-temperature water in a low-alloy steel reactor pressure vessel steel. Reproduced from Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 378, 312–326.

the KI,i value needed to initiate SICC) occurs at intermediate temperatures around 200–250  C and at slow loading rates (Figure 13). Furthermore, the SICC initiation susceptibility is affected by environmental and material parameters in a very similar way as in SSR tests with smooth specimens.38,45 Under highly oxidizing conditions (ECP  þ 50 mVSHE), SICC cracks can grow without any significant fatigue crack growth contribution and, in extreme cases, SICC growth rates during plant transients with severe thermal stratification can reach high values (up to several hundreds of micrometers per event, or several millimeters per day). Under such circumstances, SICC crack growth rates in both high-purity water and water with increased sulfate or chloride levels are comparable for all CS & LAS, independent of their sulfur content. Once growth is initiated, crack velocities rise with increasing loading rates dKI/dt (and thus crack-tip strain rates) and with increasing temperature (Figure 13), at least up to 250  C.38,45 5.06.3.2.2.3 CF initiation and crack growth from incipient cracks

Under cyclic loading, CS & LAS show a distinct susceptibility to initiation of corrosion fatigue (CF) from incipient cracks in all LWR environments. However, environmental acceleration of fatigue crack growth in these materials only occurs for certain combinations of loading and environmental parameters. The cycle-based crack growth rate da/dN then usually depends strongly on loading frequency and temperature, in contrast to fatigue in air, and can achieve 10–100 times higher values under certain conditions (although the actual effects on

component integrity in the field are much less severe than might appear at first glance from isothermal test data). The combination of material, loading, and environmental conditions, where strong environmental acceleration of fatigue crack growth (factor of 10) usually occurs, extends over a broad range in the case of highly oxidizing BWR/NWC conditions, but is restricted to a much narrower range at low ECPs (Figures 14 and 15(a)), that is, in the case of PWRs and of BWRs operating on HWC (with or without NMCA (noble metal chemical addition)). In contrast to SICC initiation, there is no threshold ECPcrit for CF in high-purity water, and material parameters usually play a less pronounced role (Figure 15(b)).46,47 The major parameters affecting CF crack growth are shown in Table 4, and the observed cracking behavior can be briefly summarized as follows.38,47 Above an upper critical frequency ncrit,H of 1–100 Hz and/or an upper DKCF,H threshold, environmental acceleration of fatigue crack growth disappears, because air fatigue crack growth rates are higher than the maximum EAC rates for CS & LAS in high-temperature water. These upper ncrit,H and DKCF,H thresholds are shifted to lower values with increasing DK and frequency, respectively, and do not depend noticeably on environmental and material parameters. Strong environmental acceleration of fatigue crack growth (factor of 10) in LWR environments occurs with all CS & LAS below these upper thresholds for the combination of temperatures 100  C, loading frequencies  ncrit,L ¼ f (ECP, DK, . . .), and DK values  DKCF,L ¼ f (ECP, n, . . .). If one or more

125

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

102 0.4 or 8 ppm O2 +50 to + 200 mVSHE 0.2 ppm O2 0 mVSHE

ASME Xl ‘air’

10−4

10−3 10−2 10−1 Frequency n (Hz)

(a)

100

101

dK/dt , n

t’

10−2

T = 288 ⬚C, R = 0.7 − 0.8, Δ K = 14 − 22 MPa m1/2 SA 533 B Cl. 1 (0.018% S), 20 MnMoNi 5 5 (0.015% S)

10−5

10−1

we

< 0.005 ppm O2 −500 mVSHE

100

Xl ‘

ASME Xl ‘Wet’

10−1

10−2

101

ME

100

SA 533 B Cl. 1,0.0018% S 250 ⬚C, kout £ 0.07 μS cm–1 0.4 ppm O2

10−3

102 (b)

AS

101

da/dNCF (μm cycle–1)

da/dNCF (μm cycle–1)

102

r’

ai

E

‘ Xl

M

AS

Kl.max = 58 MPa m1/2 0.5 MPa m1/2h−1 5 MPa m1/2h−1 35 MPa m1/2h−1 370 MPa m1/2h−1

1 10 100 Stress intensity factor amplitude Δ K (MPa m1/2)

(a)

10−6

SA 533 B Cl.1 (0.018 wt % S) T = 274 − 288 ⬚C

10−8

10−10 /d

t Air

10−12

da

NWC, + 150 mVSHE NWC, − 100 mVSHE HWC, − 500 mVSHE

10−12 10−10 10−8 10−6 Fatigue crack growth rate in air da/dtAir (m s−1)

Corrosion fatigue crack growth rate in high-temperature water da/dtCF (m s−1)

Corrosion fatigue crack growth rate in high-temperature water da/dtCF (m s−1)

Figure 14 Effect of electrochemical corrosion potential, loading frequency (a) and DK and loading rate (b) on corrosion fatigue crack growth in low-alloy steel in high-temperature water and comparison with ASME XI ‘Air’ and ‘Wet’, crack growth rates. Reproduced from Seifert, H. P.; Ritter, S. Corros. Sci. 2008, 50, 1884–1899.

(b)

10−3 Superposition-model da/dtCF = da/dtENV + da/dtAir

10−5 10−7 10−9 10−11 10−13 10−13

T = 240 − 288 ⬚C 0.4 − 8 ppm O2 <1 or 65 ppb SO42− n = 3E − 6 to 8E − 3 Hz 240 ⬚C R = 0.2–0.8 288 ⬚C 250 ⬚C ΔK = 11 − 62 MPa m1/2 r SA 533 B Cl.1 (0.018% S) i t 20 MnMoNi 5 5 (0.004% S) /d A 20 MnMoNi 5 5 (0.015% S) da

10−11

10−9

10−7

10−5

10−3

Fatigue crack growth rate in air da/dtAir (m s−1)

Figure 15 Corrosion fatigue crack growth in the time-domain as a function of electrochemical corrosion potential (ECP) for a high-sulfur low-alloy steel (a) and as a function of material and loading parameters at high ECPs >  þ 50 mVSHE (b). Reproduced from Seifert, H. P.; Ritter, S. Corros. Sci. 2008, 50, 1884–1899.

of these threshold conditions are not satisfied, the environmental effect is moderate (factor of 5), or even absent altogether. When these conjoint threshold conditions are simultaneously satisfied, the environmental acceleration of fatigue crack growth increases with increasing temperature (sometimes with a maximum around 250  C) and decreasing loading frequency and DK values, often showing maximum enhancement close to the lower thresholds. The lower thresholds ncrit,L and DKCF,L depend on material, environmental, and loading parameters (in particular, ECP and DK or frequency – see Figure 14 and Table 4). They can reach very low values of

106 Hz and of 2 MPa m1/2 (for load ratio R!1)

or 5 MPa m1/2 (R! 0) under highly oxidizing conditions (ECP  þ100 mVSHE). Below these lower thresholds, the environmental acceleration of fatigue crack growths drops to moderate values (factor of 5) and the cycle-based crack growth rate da/dN becomes independent of frequency. The DK-dependence is the same as in air and the effect of temperature is usually moderate. Such behavior is sometimes designated as true corrosion fatigue and corresponds to the low-sulfur curve of the Ford and Andresen model (see Section 5.06.3.2.4; Ford and Andresen48 and Ford49). Once conditions exceed these lower thresholds, the material behavior corresponds to the highsulfur curve of the model.

126 Table 4

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels Major parameter effects on CF crack growth with typical parameter values

Crack-tip environment

Low-sulfur, 0.1 ppm S2

High-sulfur, 1–5 ppm S2

Range da/dtCF/dat/dtAir

n < ncrit,L 2–5 6¼ f(n) m  3, that is, parallel to air curve 6¼f(n), that is, n ¼ 0 DKth, Air ¼ f(R)

ncrit,L n < ncrit,H  1 Hz 10–100 (10 000) n# ! da/dtCF/dat/dtAir" m  1–2 n ¼ 0.5–0.6 n", R", ECP", SO2 4 ", S", DSA"! DKCF,L# þ100 mVSHE: DKCF;L  2–5 MPa m1/2

500 mVSHE: DKCF,L  10–30 MPa m1/2 ECP", SO2 4 ", S" DK" ! ncrit,L# þ100 mVSHE: 1 E5 Hz to <1 E6 Hz

500 mVSHE: 1 E3 Hz to 1 E1 Hz T > 100  C: T" ! da/dtCF" EA  40–50 kJ mol1 T# ! DKCF# With minimum of DKCF at intermediate T?

da/dNCF ¼ BDKm da/dNCF ¼ Ann DKCF,L ¼ f(n, R, ECP, . . .) ncrit,L ¼ f (DK, R, ECP, . . .)

No evidence for lower threshold for low-sulfur behavior

Temperature

Moderate effect

Above 1–10 Hz, the environmental effects on crack growth disappear. Source: Seifert, H. P.; Ritter, S. Research and service experience with environmentally-assisted cracking of carbon & low-alloy steels in high-temperature water. SKI-Report 2005:60; SKI: Stockholm, Sweden, 2005; ISSN 1104-1374. http://www.stralsakerhetsmyndigheten. se/; Seifert, H. P.; Ritter, S. Corros. Sci. 2008, 50, 1884–1899.

The water flow rate has little effect on CF crack growth of long/deep cracks, in contrast to CF initiation and short-crack growth. In the latter cases, CF effects can disappear entirely if flushing out of the local crack-tip electrolyte occurs, but this is unrealistic for deep, semi-elliptical cracks. 5.06.3.2.2.4 Adequacy and conservatism of fatigue design according to Section III of ASME BPV Code in the context of environmental effects

Design against fatigue of CS & LAS primary pressureboundary components is often based on Section III of the ASME BPV Code.50 It relies on the use of fatigue curves and endurance limits, derived mainly from strain-controlled LCF tests with small, smooth specimens in air at room temperature, which do not explicitly consider the possible effects of LWR environments. The accumulated good service experience with CS & LAS primary pressure-boundary components does not indicate any generic deficiencies in the current fatigue design procedures arising from lack of consideration of environmental effects (see Section 5.06.3.2.5), even though laboratory investigations clearly reveal that the fatigue lives of CS & LAS can be substantially reduced in LWR environments (Figure 16)38,41,42,51,52: Although the microstructures and cyclic-hardening behavior of CS & LAS differ significantly, the effects of the environment on the fatigue life of all these steels are very similar. The magnitude of the reduction in this

depends on temperature, strain rate, oxygen level in the water (i.e., ECP), and sulfur content of the steel. The decrease is significant only when four conditions are satisfied simultaneously, that is, when the strain amplitude, temperature, and dissolved oxygen content in the water are above certain threshold values (0.15%, 150  C, 0.04 ppm) and the strain rate is below a key threshold value (103 s1). Although only a moderate decrease in life (by a factor of <2) is observed if any one of the aforementioned threshold conditions is not satisfied, fatigue lives of CS & LAS can be more than a factor of 20 lower in the coolant environment than in air under certain environmental and loading conditions (Figure 16). Such observations have thus raised some concern with respect to the possibility of insufficient safety margins for the fatigue design of certain CS & LAS pressure-boundary components. Conservatism in the ASME Code fatigue evaluations may arise from (a) the fatigue evaluation procedures themselves and/or (b) the fatigue design curves they use. Sources of conservatism in the procedures include the use of design transients that are significantly more severe than those experienced in service, conservative grouping of transients, and use of simplified elastic–plastic analyses that result in higher stresses/strains. The design margins of 2 and 20 on stress and cycles, respectively, in the ASME III design curve were intended to cover the effects of some variables (e.g., surface finish, material

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

127

10.0

1.0

0.1 101 (a)

A533-Gr B low-alloy steel 288 ⬚C, ea ≈ 0.4%

150−250 >250 0.05−0.2 >0.2 0.01−0.4 <0.01 ³0.006 ³0.006

Temp( ⬚C) :<150 DO (ppm) :£0.05 Rate (%/s) :³0.4 S (wt.%) :>0.006

Fatigue life (cycles)

Strain amplitude ea (%)

Carbon steel

Mean curve RT air

ASME design curve 102

103

104

104

103

Air Simulated PWR ≈0.7 ppm DO

102 105

10−5

106

Fatigue life (cycles)

(b)

10−4

10−3

10−2

10−1

100

Strain rate (%/s)

Figure 16 Comparison of fatigue initiation life of carbon steel in high-temperature water with ASME III mean and design curves in air (a) and effect of strain rate and oxygen content on fatigue initiation life in low-alloy steel. Reproduced from US NRC. Effect of LWR coolant environments on the fatigue life of reactor materials. NUREG/CR-6909; US NRC: Washington, DC, 2007, http://www.nrc.gov/. Work performed by Argonne National Laboratory, managed and operated by UChicago Argonne, LLC, for the US Department of Energy under Contract No. DE-AC02-06CH11357.

variability, load sequence, and size effects) that can influence the fatigue life of components, but were not actually investigated in the tests which provided the original data for the curves. They were not intended to cover the effects of LWR environments,38 but a recent, detailed analysis41,42 revealed that the factors of 2 on stress/strain and 20 on cycle number not only provide appropriate margins for the intended factors but may also contain excess conservatism that would partially counteract reductions in fatigue life due to EAC. Based on large research programs, methods have been developed in the United States42 and in Japan51,53 for incorporating the effects of LWR coolant environments into fatigue evaluations. Experimentally derived fatigue life correction factors Fen (defined as the ratio of life in air at room temperature to that in water at the service temperature) are used to adjust component fatigue usage values for environmental effects. Such approaches have recently been implemented in the Japanese JSME Code54 and the new US NRC Regulatory Guide 1.207,55 which is to apply to new plants. The apparent discrepancy between isothermal laboratory results (strong environmental effects) and field experience (only a few CF incidents under very specific circumstances, predominantly related to thermal transients) mainly arises from the large degree of conservatism in the fatigue evaluation procedures mentioned earlier. The significantly higher strains of design transients may fully outweigh possible environmental effects for real plant transients. Furthermore, one or more threshold conditions are often also not

satisfied for most transients. Even plant transients with strong environmental effects (e.g., at slow strain rates) are usually not very damaging with respect to fatigue damage accumulation because of their small strain amplitudes, together with the rather limited number of cycles during the whole component lifetime. Finally, as discussed earlier, the turbulent flow rate at most component surfaces significantly retards corrosion fatigue crack initiation (in particular for small strains and slow strain rates) compared to the quasi-stagnant conditions used in most lab investigations.38,41,42 5.06.3.2.2.5 Adequacy and conservatism of fatigue flaw tolerance evaluations according to Section XI of ASME BPV Code in the context of environmental effects

Fatigue flaw tolerance evaluations in combination with periodic in-service inspections are based on Section XI of the ASME BPV Code. Article A-4300 in Appendix A of Section XI contains a set of reference fatigue crack growth rate da/dN curves for CS & LAS in air (‘air curves’) or in LWR coolant environments (‘wet curves’).56 The current ASME XI wet curves (Figure 17) are based on lab data obtained prior to 1980. They depend explicitly on DK and load ratio R, but not on other variables that are known to be important in CF, such as loading frequency, ECP, or steel sulfur content. The same curves are used for different types of CS & LAS and for BWR/NWC, BWR/HWC/NMCA, and PWR primary or secondary side conditions. System conditions (or thresholds)

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

100 10−1 10−2

101

Δtrise = 14 s R = 0.95 0.018% S

100 10−1 10−2

43 s −6 0 N 20 e = s ca t ise e Δ r od .65 C 0 = R

HWC, 274/288 ⬚C

3 64 s N- 14 e = s e ca Δt ris de 95 Co = 0. R

e

rv

10−3

ME AS X I ‘W et, ME ’R XI ⱖ0 ‘W .65 et, ’R ⱕ 0.2 5

101

N-643, EAC-curve for S > 0.013 % S, R = 0.65, ΔtR = 100 s N-643, EAC-curve for S > 0.013 % S, R = 0.65, ΔtR > 1 s N-643, non-EAC curve, R = 0.65 ASME XI ‘Wet,’ R = 0.65 ASME XI ‘Air,’ R = 0.65

AS

da/dN (μm per cycle)

102

da/dNCF (μm per cycle)

128

ir’

E

XI

‘A

M

AS

Δtrise = 200–2000 s R = 0.3–0.7 0.004–0.018% S

cu

AC

-E

on

N 10−3 10−4 1 10 100 1 10 100 (a) Stress intensity factor amplitude ΔK (MPa m1/2) (b) Stress intensity factor amplitude ΔK (MPa m1/2)

Figure 17 ASME XI ‘Air’ and ‘Wet’ curves and Code Case N-643 for high and low-sulfur steels (a) and comparison of cyclic corrosion fatigue crack growth rates under hydrogen water chemistry conditions for different loading conditions with the corresponding curves (b). Reproduced from Seifert, H. P.; Ritter, S. Corros. Sci. 2008, 50, 1884–1899.

where environmental effects on fatigue crack growth can be neglected or excluded are not defined in the present ASME Section XI Code.38,47 A more specific Code Case N-64357 for fatigue crack growth in CS & LAS exposed to PWR primary environments has been developed since 2000 and may be used as an alternative to the ASME XI wet reference fatigue crack growth rate curves for this specific environment (Figure 17). Depending on system conditions, the Code Case N-643 procedure predicts either lower or higher crack growth rates than the general Section XI approach. The main advantage of this newer Code Case is that it contains criteria for the onset/cessation of EAC and that it better reflects the experimentally observed cracking behavior in PWR environments, since it considers steel sulfur and frequency–loading-rate effects to a certain extent. However, this approach has not (yet) found general acceptance, primarily because of the difficulty of defining appropriate rise times for actual plant transients and the complications involved in including these in component analyses. In a similar way to that discussed earlier for fatigue life design, conservatism in fatigue flaw tolerance evaluations may arise from (a) the fatigue flaw tolerance evaluation procedures themselves (e.g., by the use of design transients) and/or (b) the fatigue crack growth rate curves. As discussed in Seifert and Ritter,38,47 the current ASME XI wet curves conservatively cover the CF crack growth rate laboratory data under most combinations of loading, environmental, and material parameters. Even under highly oxidizing BWR/NWC conditions, they are only significantly exceeded under some very specific (but plantrelevant) circumstances (Figures 14 and 17), which

have caused some isolated CF cracking incidents in the past. The Section XI curves might therefore be regarded as an adequate, general bounding approach under most system conditions, but they do not realistically describe and reflect the experimentally observed CF crack growth behavior of CS & LAS in oxygenated high-temperature water. The curves predict crack growth rates which are either significantly too high (e.g., n 102 Hz and ECP < 200 mVSHE) or too low (e.g., n 102 Hz and ECP > 0 mVSHE). Furthermore, system conditions or thresholds (e.g., n > 1 Hz), where environmental effects on fatigue crack growth can be neglected, or even excluded, are not defined in ASME XI. 5.06.3.2.3 Stress corrosion cracking 5.06.3.2.3.1 conditions

Initiation and susceptibility

Table 5 shows an assessment scheme according to Hickling,58 based on both laboratory and field experience, for SCC initiation susceptibility and crack growth in CS & LAS at normal strength levels under BWR/NWC conditions. Initiation of propagating SCC cracks from smooth, defect-free surfaces under static load in high-purity water is only observed for the following conjoint conditions: stresses at the water-wetted surface above the high-temperature yield strength, quasi-stagnant flow conditions, and dissolved oxygen contents 0.2 ppm. Furthermore, if complete exhaustion of low-temperature creep is allowed to occur before the specimens are exposed to high-purity, high-temperature water, no SCC is observed, thus indicating ‘nonclassical’ SCC behavior and confirming the importance of slow dynamic surface straining

Assessment scheme for SCC susceptibility of CS & LAS

O2 (ppm)

Operating medium: HT water or steam condensate with T > 170  C Flow conditions

k (mS cm1)

Crack initiation by SCC?

Derivation

Crack growth by SCC?

Derivation

<0.2 <0.2 <0.2

Typical for reactor Quasi-stagnant Quasi-stagnant

Typical, that is, 0.1 0.2 Raised (e.g., by impurities)

1 1 2

Typical for reactor Quasi-stagnant

Typical, that is, 0.2 0.2

0.2–0.4

Quasi-stagnant

Raised (e.g., by impurities)

0.4

Typical for reactor

Typical, that is, 0.2

0.4

Quasi-stagnant

<1 often (0.2)

0.4

Quasi-stagnant

Raised (e.g., by impurities)

No susceptibility No susceptibility Possibility cannot be excluded, perhaps after incubation time No susceptibility Possibility cannot be excluded, perhaps after incubation time Possibility cannot be excluded, perhaps after incubation time Susceptibility is suppressed through flow Possibility cannot be excluded, perhaps after incubation time Possibility cannot be excluded, perhaps after incubation time

1 2 3

0.2–0.4 0.2–0.4

No susceptibility No susceptibility For stress levels at the water-wetted surface in the region of the HT yield pointa No susceptibility For stress levels at the water-wetted surface > HT yield pointa For stress levels at the water-wetted surface in the region of the HT yield pointa In general, no susceptibility (? at stress levelsHT yield point) For stress levels at the water-wetted surface  HT yield pointa For stress levels at the water-wetted in the region of the HT yield pointa

1 2 2 1, 3 2 2

a Possibility cannot be excluded, perhaps after long incubation time. 1 – from experiments in more aggressive environments; 2 – from appropriate autoclave experiments; 3 – no direct experimental evidence. Source: Hickling, J.; Reitzner, U. VGB Kraftwerkstech. 1992, 72, 359–367.

1 2 2 2 2 2

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

Table 5

129

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

under such conditions. Initiation of small ‘nonpropagating’ SCC surface microcracks (surface length < 1–5 grain diameter) at (MnS) inclusions or corrosion pits may occur under less severe conditions than specified earlier. Furthermore, the possibility of SCC at stresses around the high-temperature yield strength range cannot be excluded for very unfavorable parameter combinations.38,40,58 Since the primary design stresses are limited to values below the yield strength and turbulent flow conditions prevail at most component locations, SCC initiation appears to be extremely unlikely under stationary, transient-free, LWR power operation conditions, but growth of a preexisting crack could occur under certain conditions, as discussed in the next section. 5.06.3.2.3.2 SCC crack growth

Stress corrosion crack growth rate da/dtscc (m s−1)

In contrast to SICC and CF crack growth (and to the behavior of austenitic stainless steels and nickel-base alloys), CS & LAS reveal very low susceptibility to sustained SCC crack growth up to high KI levels in LWR environments at operating temperatures. Such high KI-levels would correspond to rather deep cracks in pressure-boundary components. In laboratory tests with relatively small specimens, no sustained SCC crack growth (or only very slow (<0.6 mm year1), often localized propagation) is observed up to KI values of 50–60 MPa m1/2 in a chloride-free BWR/NWC environment and up to 80–100 MPa m1/2 in PWR or BWR/HWC/NMCA environments. Above these KI thresholds, SCC growth rates increase rapidly within a small KI-interval of 10–20 MPa m1/2 and reach a plateau at high growth rates of up to several meters per year.38,59–61

(a)

10−6 T = 288 ⬚C, 0.004 − 0.018 % S O2 = 200–600 ppb, < 1– 165 ppb SO42− 2–

10−8

10−10

O2 = 8 ppm, <1–65 ppb SO4

Down-pointing arrows indicate continuous decay of SCC with subsequent crack arrest within 1000 h

BWRVIP-60 SCC DL 1

10−12 40 60 80 100 20 Stress intensity factor Kl (MPa m1/2)

The very low SCC crack growth susceptibility at KI values below 50 MPa m1/2 is also confirmed by long-term exposure of precracked specimens in actual reactors,61,62 as well as by the field experience with both unclad LAS components and with cracks in claddings or dissimilar metal welds, where the crack-tip extends to LAS at the fusion boundary (see Section 5.06.3.2.5). On the other hand, critical conditions for water chemistry (e.g., chloride content > 5 ppb þ high ECP > 0 mVSHE), for loading (e.g., ripple loading), or for the material itself (e.g., hardness > 350 HV5 or high DSA susceptibility at intermediate temperatures) can result in sustained and relatively fast (>10 mm year1) SCC for KI-values well below 60 MPa m1/2 (Figures 18 and 19). Such conditions, identified in laboratory experiments, generally appear atypical for current, optimized BWR power operation practice with modern, properly fabricated CS & LAS pressureboundary components. Changing to HWC (i.e., low ECP) always results some few hours later in a significant reduction of laboratory SCC crack growth rates (by more than one order of magnitude even under otherwise critical conditions).38,60 5.06.3.2.3.3 Adequacy and conservatism of BWRVIP-60 SCC disposition lines

With regard to structural integrity, the possible occurrence of SCC is usually more critical than CF, since fatigue was considered in the original component design, whereas even very slow SCC growth rates might result in relevant crack advance over long operational periods. Within the framework of the ‘BWRVIP program’ of the Electric Power Stress corrosion crack growth rate da/dtscc (m s−1)

130

(b)

10−6

10−8

T = 288 ⬚C, Weld HAZ, 350 HV5 + 0.022% S T = 288 ⬚C, 460 HV5 T = 288 ⬚C, KI → KIJ

10−10 BWRVIP-60 SCC DL 1

10−12

180 ⬚C < T < 270 ⬚C + high DSA Base metal Weld filler Weld HAZ

20 40 60 80 100 Stress intensity factor Kl (MPa m1/2)

Figure 18 (a) Experimental confirmation of the conservative character of the BWRVIP-60 stress corrosion cracking disposition line 1 for stationary, transient-free boiling water reactor power operation at 274–288  C. (b) Constant load test results with stress corrosion cracking growth rates above the BWRVIP-60 stress corrosion cracking disposition lines 1. Reproduced from Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 372, 114–131.

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

131

8.0 ppm DO

10–9

0.4 ppm DO

10–8

0.2 ppm DO

line’ r SCC sulfu h g ‘Hi

’ line CC S r fu sul wo L ‘

100 ppb Cl– 50 ppb Cl– 20 ppb Cl– 15 ppb Cl–, 250 ⬚C 10 ppb Cl– 5 ppb Cl–

on curv e

+200 mVSHE 8 ppm DO 0.02 wt % S

10–10

Transiti

SCC crack growth rate da/dtSCC (m s–1)

10–7

10–11

BWRVIP-60 SCC DL 2 for water chemistry transients BWRVIP-60 SCC DL 1 for stationary power operation

10–1210

20

(a)

30 40 50 60 70 80 Stress intensity factor KI (MPa m½)

90

100

Accelerated SCC crack growth (>10 mm year–1) at KI < 60 MPa m½

T = 274/288 ⬚C NWC

200

0 Continuous operation allowed Prompt shut-down

EPRI action Level limit 3

–400

EPRI action Level limit 1

–200 HWC

Corrosion potential ECP (mVSHE)

No SCC (<0.6 mm year–1) at KI < 60 MPa m½

–600 1 (b)

10 Chloride content (ppb)

100

Figure 19 (a) Effect of chloride content on stress corrosion cracking crack growth in low-alloy steel in simulated boiling water reactor/normal water chemistry environment and comparison with the BWRVIP-60 stress corrosion cracking disposition lines and the F & A model. (b) Synergistic effect of electrochemical corrosion potential and chloride content on the onset of fast stress corrosion cracking in low-alloy steel under simulated boiling water reactor conditions and comparison with the typical conditions during stationary boiling water reactor power operation and Action Levels 1 and 3 of the EPRI BWR water chemistry guidelines. Reproduced from Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 372, 114–131.

Research Institute (EPRI), SCC disposition lines for flaw tolerance evaluations (Figure 18) were developed in 1999 based on lab investigations and service experience61 and later accepted by the US NRC as an interim position. The ‘BWRVIP-60 SCC disposition line 1’ applies to crack growth in LAS under static loading and transient-free, stationary BWR/NWC or HWC power operation conditions. On the other hand, disposition line 2 (¼‘low-sulfur SCC line’ of the Ford and Andresen model63) may be used for

estimating SCC crack growth during and 100 h after transients in water chemistry (>‘EPRI action level 1 limit’ of EPRI BWR water chemistry guidelines64) or load transients not covered by fatigue evaluation procedures. These curves are currently undergoing minor revisions. The conservative character of the BWRVIP-60 disposition line 1 for SCC crack growth in CS & LAS under NWC conditions has been confirmed for temperatures in the range of 274–288  C and

132

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

both base and weld filler/heat-affected zone (HAZ) materials (Vickers hardness < 350 HV5, 0.02 wt% S) if the water chemistry is maintained within current BWR/NWC operational practice (<EPRI action level 1 limit).38,60 In the case of NWC conditions, several results clearly indicate, however, that ‘line 1’ may be slightly exceeded at intermediate temperatures (180–270  C) in CS & LAS which show distinct susceptibility to DSA. Not unexpectedly, sustained SCC with crack growth rates significantly above ‘line 1’ is observed at 288  C if excessive hardness (350 HV5) is present in the steel, for example, in bad weld HAZs. Furthermore, BWRVIP-60 disposition line 2 is significantly exceeded for the case of ripple loading (R > 0.95) or during chloride transients ( EPRI action level 1 limit), even at fairly low stress intensity values around 30 MPa m1/2 (see Figure 19). On the other hand, ‘line 2’ seems to cover even very severe sulfate transients above the EPRI action level 3 limit.38,60 At low ECP, that is, in the case of BWR/HWC or PWR conditions, line 1 appears conservatively to cover the SCC crack growth in LAS, even under the otherwise critical conditions mentioned earlier for more oxidizing, environments.38,60

FRAD-EAC-mechanism 1

2

5.06.3.2.4 EAC mechanisms and models 5.06.3.2.4.1

EAC crack growth mechanism

For the case of EAC in CS & LAS in hightemperature water, the film rupture/anodic dissolution (FRAD), also referred to as slip dissolution/ oxidation, and hydrogen-assisted EAC (HAEAC) cracking mechanisms have been proposed in the literature (see Figure 20). Either of these may be superimposed on pure mechanical fatigue. In the FRAD mechanism, the protective oxide film formed on CS & LAS in high-temperature water is ruptured by plastic straining at the crack-tip. The crack-tip then advances by anodic dissolution of the bare metal matrix. Anodic dissolution is slowed down and finally stopped by the nucleation and reformation of the oxide film (‘repassivation’). Thus, continued crack advance will depend on a further oxide rupture process due to the action of a strain rate at the cracktip. The crack propagation rate is controlled by both anodic dissolution/repassivation kinetics and the frequency of oxide film rupture at the strained crack-tip. The first part is governed by the chemical composition of the local crack-tip electrolyte and the material. The second part is determined by the fracture strain of the oxide film and the crack-tip strain rate.63

Hydrogen-assisted EAC mechanism syy

Crack surface covered by an oxide film

Bare surface of metal generated by oxide film rupture and anodic dissolution of the metal

Elastic–plastic stress distribution

Acidic oxygen free water Oxide layer 1

Void 8

8 3

2

8

9

MnS

6

MnS

Fe2+ 4

5

1. Local anodic reaction

3

Repassivation: Oxide nucleation and growth

ad

Fe → Fe2+ + 2e 2. Hydrolysis and generation of H+ Fe2+ + H2O → FeOH+ + Η+ 3. Liquid phase transport 4. Local cathodic reaction H+ + e → H MnS + 2H+ →H2S + Mn2+

7 5. Hydrogen absorption 6. Hydrogen transport in lattice 7. Hydrogen trapping on inclusions 8. Hydrogen-induced cracking 9. Linkage of microcracks to main fracture

Figure 20 Schematic illustrations of film rupture/anodic dissolution (left) and of hydrogen-assisted environmentally assisted cracking mechanisms for carbon and low-alloy steel in high-temperature water. Adapted from Ford, F. P. J. Pressure Vessel Technol. 1988, 110, 113–128; Ha¨nninen, H. et al. Corros. Sci. 1983, 23, 663–679.

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

Oxide film rupture–repassivation events at the strained crack-tip are also involved in the HAEAC model, but here, hydrogen-induced microcrack formation ahead of the crack-tip and linkage of these microcracks to the main crack are the prime sources of EAC crack growth, and result in discontinuous crack propagation. The hydrolysis of metal cations from anodic dissolution is an important source of hydrogen, but – in contrast to the FRADmechanism – anodic dissolution does not contribute significantly to crack advance. Hydrogen transport in the electrolyte and in the metal lattice is believed to be fast. Therefore, the generation of bare metal surface by film rupture and film reformation may be the rate-controlling steps and explain the strain rate dependence of EAC.65,66 The FRAD and HAEAC mechanisms may be simultaneously active and controlled by the same rate-limiting steps (e.g., oxide film rupture rate or repassivation kinetics). Both mechanisms are able to explain the observed, dominant effect of strain rate and of MnS inclusions on EAC. The EAC behavior of CS & LAS in high-temperature water can best be rationalized by a superposition/combination of the FRAD and HAEAC mechanisms. At lower temperatures (<100  C) and/or high yield strength/ hardness levels (>800 MPa/>350 HV5) and high strain rates (>103 s1), hydrogen effects are more

HCl H2SO4

NaCl

Role of MnS inclusions

Sulfur-anions as HS, S2, and SO2 may signifi4 cantly retard repassivation after oxide film rupture and therefore increase crack advance by anodic dissolution in the FRAD model.38,63 Retarded repassivation of the film-free surface and adsorbed HS, S2, or H2S also increase hydrogen absorption into the metal lattice and therefore favor HAEAC.38,65,66 Furthermore, the dissolution of MnS is a further potential source of hydrogen, and MnS inclusions in the region of maximum hydrostatic stress ahead of the crack-tip may act as strong hydrogen traps and thus HAEAC microcrack initiation sites.38,65,66 The effects of steel sulfur content are synergistic with environmental variables, such as (sulfur-) anionic impurities in the bulk environment, ECP (dissolved oxygen content), and flow rate (Figure 21).38,67 This is believed to be due to the creation of a sulfur-rich crack-tip environment responsible for EAC, which arises both from the dissolution of MnS intersected by the growing crack and by the transport of sulfuranions by migration/diffusion/convection within the crack enclave.38

SSRT, 1⫻10–6s–1 hpw, 288 ⬚C Low-flow autoclave

–100 Corrosion potential ECP (mVSHE)

Corrosion potential ECP (mVSHE)

5.06.3.2.4.2

Filled symbols TGSCC Open symbols no SCC

–100 –200

pronounced. At high temperatures (150  C) and/or lower yield strength/hardness levels and slow strain rates (<103 s1), anodic dissolution appears to dominate.38

Pure water Cl– SO42–

0

Na2SO4

–300 –400 –500

SCC –200

–300 No SCC

–400 A533-B A508 A533-B

0 (a)

133

100

200

300

400

Anion concentration (ppb)

500

A533-B

0.005 0.010 0.015 0.020 0.025

600 (b)

Sulfur content (wt.%)

Figure 21 Synergistic effect of electrochemical corrosion potential and sulfate anion concentration (a) and steel sulfur content (b) on strain-induced corrosion cracking susceptibility of low-alloy steel in high-temperature water. Data obtained from Sund, G.; Rosborg, B. The influence of impurities on the tendency to stress corrosion cracking of pressure vessel steel A533-B in water at 288  C; Personal communication; 1991.

134

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

(which affect repassivation/pH/oxide film stability) in the crack-tip environment. The onset and extent of EAC is crucially dependent on simultaneously maintaining a slow, positive crack-tip strain rate and a critical, local sulfur-anion activity of about 1–5 ppm S2. If these two conjoint requirements are not met, no SCC and SICC and only minor environmental acceleration of fatigue crack growth are generally observed. If the critical anion concentration is exceeded, the EAC crack growth rate depends primarily on the crack-tip strain rate and increases with increasing external strain rate up to an upper critical limit of around 103–102 s1 (Figure 23). Most EAC thresholds, as well as phenomena involving cessation or local retardation of crack growth, may be directly attributed to this conjoint requirement and thus to crack-tip environment chemistry and crack growth rate/crack-tip strain rate, rather than to overall loading parameters or bulk environmental conditions per se. For example, suitable combinations of different parameters may help exceed the critical crack-tip sulfur-anion concentration, provided that there is at least one source of sulfur (MnS inclusions or sulfur anions in the bulk environment).38,63,69 These two local factors are governed by a system of interrelated and synergistic corrosion parameters: In general, the crack-tip strain rate increases with increasing loading rate/frequency/level, crack growth rate, yield strength, and DSA susceptibility. The sulfur anion activity in the crack-tip environment increases with increasing steel sulfur content, concentration of sulfur anions in the bulk environment outside the crack, crack growth rate (exposure of fresh, dissolvable MnS by the growing crack), and corrosion

Under certain combinations of temperature (100– 350  C) and strain rate (108–102 s1), DSA may synergistically interact with both mechanisms to increase EAC susceptibility and provide an additional contribution to the crack growth process (Figure 22). DSA may result in the occurrence of a higher crack-tip strain/strain rate than for loading outside the DSA range, or than in a material which is not susceptible to DSA. The localization of plastic deformation and increase in planar deformation from DSA probably support mechanical rupture of the oxide film and result in a reduction of the local fracture toughness, thus favoring quasi-brittle crack extension.38,68 The concentrations of free, interstitial nitrogen and carbon (which mainly govern the DSA susceptibility in CS & LAS) might therefore be just as relevant for EAC susceptibility as the steel sulfur content, and some data suggest that DSA may even overwhelm sulfur effects under certain conditions. DSA generally results in an extension of the susceptibility region and can affect the temperature and strain-rate dependence of EAC. The most pronounced effects of DSA on EAC38 are typically observed close to crack growth thresholds (e.g., for SCC under static load,60 or for CF close to critical frequencies under cyclic load47). 5.06.3.2.4.4 Controlling factors for EAC crack growth

As shown by lab testing, EAC growth from incipient cracks is essentially governed by the crack-tip strain rate and the activity of sulfur or chloride anions

10–8

80

90 20 MnMoNi 5 5, A, 0.004 % S DO = 8 ppm, 65 ppb SO42−

–1

20 MnMoNi 5 5, A, 0.004 % S SRL tests (8 ppm DO, 65 ppb SO42−) Tensile tests in air (DIN 50125)

70

SCC crack growth da/dtSCC (m s )

80

Open symbols: no SICC detected by DCPD (Kmax ) I

75

60

70

50 65

40

Reduction of area Z (%)

Stress intensity factor at the onset of SICC crack growth KI,i (MPa m1/2)

90

85

10–9 BWR VIP 60 DL K = 60–75 MPa m1/2 I

80 10–10 75 10–11 70 10–12

Reduction of area Z (%)

5.06.3.2.4.3 Role of dynamic strain ageing

65

30 100

(a)

150

200

250

Temperature ( ⬚C)

300

10–13

60 350

(b)

0

50

100

150

200

250

300

60 350

Temperature ( ⬚C)

Figure 22 Coincidence of maximum of strain-induced corrosion cracking susceptibility in slow rising load tests (a) and maximum in stress corrosion cracking growth rates in constant load experiments (b) with maximum in DSA susceptibility in terms of temperature (and strain rate). (a) Reproduced from Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 378, 312–326; (b) reproduced from Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 372, 114–131.

SA 533 B Cl. 1, C, 0.018% S, SICC 20 MnMoNi 5 5, A, 0.004% S, no SICC 20 MnMoNi 5 5, A, 0.004% S, SICC

10−7

10−8 –6

Scatter band for dCODLL/dt =2 – 4⫻10

mm s−1

8 ppm DO, + 150 mvSHE,288 ⬚C dCODLL/dt = 2–4 ⫻ 10−6 mm s−1

10−9

10−10

135

DO = 8 ppm, <1or 65 ppb SO42–, T = 288 ⬚C, 250 ⬚C, 200 ⬚C

10−6

SCC or SICC crack growth rate da/dt scc or da/dtSICC (ms−1)

SICC crack growth rate da/dtSICC (m s−1)

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

0

100 200 300 Sulfate content (ppb)

10−6

Constant load Very low-frequency cyclic load

10−7

Slow rising load Linear fit of 95% prediction interval

10−8 10−9

10−10

400

da/dtEAC ∝ (dCODLL/dt)0.8 Model: da/dtEAC = A . de/dtn

10−11

10−9 10−8 10−7 10−6 10−5 10−4 10−3 Crack-opening displacement rate dCODLL/dt (mm s−1)

Environmental component, +200 mvSHE Mechanical component, air fatigue

10−6

Max. environm. component

e ur lin

ig

ue

-sulf

High

Shifted to higher de/dt −10 10 with ECP↓ , S↓ SO42– ↓ or Cl–↓

10−8

rf

at

10−8

10−12 −10 10

(a)

DSA↑, YS↑→ de/dt ↑ for given loading conditions

e

r

fu

ul

-s

Ai

EAC crack growth rate (ms−1)

10−4

lin

10−5

10−7

DO

DH [ppm]

21 0.79 8 0 0.4 0 HWC +150 mVSHE

DO

DH

8.4 1.26 0 0.095 0 0.150

10−4

Crack-tip strain rate de/dt

10−2

10−0

(s−1)

(b)

+

: NMCA

HWC –500 mVSHE T = 274–28 ⬚C low-alloy steel

10−9

F & A-model for NWC F & A-model for HWC PSI NWC curve da/dtAir

10−11

w Lo

10−6

Corrosion fatigue crack growth rate in high-temperature water da/dtCF (m s−1)

Figure 23 Effect of sulfate content on strain-induced corrosion cracking crack growth in a low- and high-sulfur low-alloy steel (left) and of crack mouth opening rate (as a measure for the crack-tip strain rate) on environmentally assisted cracking crack growth rate under different loading conditions (right). Evidence for the described conjoint conditions and the important role of crack-tip strain rate. Reproduced from Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 378, 312–326.

High-sulfur line Low-sulfur line

t Air

/d da

10−11

10−9

10−7

10−5

Fatigue crack growth rate in air da/dtAir (m s−1)

Figure 24 Basic environmentally assisted cracking crack growth curves of the F & A model (a) and good correlation of the observed corrosion fatigue crack growth behavior under boiling water reactor/normal water chemistry and hydrogen water chemistry or noble metal chemical addition conditions with the predictions of the F & A model (b). (a) Adapted from Ford, F. P. J. Pressure Vessel Technol. 1988, 110, 113–128 and (b) reproduced from Seifert, H. P.; Ritter, S. Corros. Sci. 2008, 50, 1884–1899.

potential/dissolved oxygen content (enhanced migration of sulfur anions toward the crack-tip).38 5.06.3.2.4.5 Ford & Andresen EAC model

The mixed mechanistic–phenomenological Ford and Andresen model (Figure 24) is the sole quantitative approach for CS & LAS and is based on the FRAD mechanism.48,49,63,69,70 In the case of corrosion fatigue, a linear superposition of fatigue and EAC crack growth is assumed. In this model, EAC crack growth is controlled by the crack-tip strain rate de/dCT and the sulfur-anion activity cCT in the crack-tip electrolyte. The EAC crack growth rate da/dtEAC increases with increasing crack-tip strain

rate de/dCT according to a power law and saturates above a critical strain rate of 103–102 s1 (continuous dissolution at a crack tip free of protective oxide). The exponent of this power law relationship is dependent on the sulfide-anion activity in the crack-tip environment. Above a sulfide concentration of 0.02 ppm (low-sulfur threshold), the formation of a new oxide layer is increasingly delayed by increasing sulfide content, which thus leads to a larger increment of crack advance by anodic dissolution per oxide-rupture event. Above a sulfide content of 2 ppm (high-sulfur threshold), the sulfide effect on repassivation saturates. Based on this experimentally

136

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

derived relationship, lower and upper limiting EAC crack growth curves for CS & LAS (Figure 24) were defined (so-called low- and high-sulfur EAC lines). A critical, high sulfur-anion content of  2 ppm S2 has to be maintained in the crack-tip electrolyte to sustain fast, high-sulfur EAC crack growth rates, otherwise the crack growth rates quickly drop down to low-sulfur crack growth rates.48,49,63,69,70 The sulfur-anion concentration in the crack-tip environment from dissolution of MnS inclusions is assumed to be proportional to the steel sulfur content and the crack growth rate itself (intersection of the crack with new, dissolvable MnS inclusions) and to increase exponentially with the potential gradient/ ECP (based on crevice experiments and modeling). Similarly, the degree to which sulfur anions are enriched in the crack-tip environment depends on how much they tend to migrate there from the bulk environment. It is thus proportional to their concentration and increases exponentially with potential gradient/ECP.48,49,63,69,70 The Ford and Andresen model correctly predicts most experimentally observed EAC data trends for CS & LAS over a wide range of conditions.38 Furthermore, it has also been successfully applied to predict LCF initiation.70 It therefore has good potential for data analysis and the definition of disposition lines. The high-sulfur line of the model conservatively covers almost all EAC crack growth rate data. Recent results indicate, however, that the transition curve between the low- and high-sulfur CF lines might be nonconservative under highly oxidizing BWR/NWC conditions (ECP > 0 mVSHE) and very low loading frequencies (<104 Hz),47 possibly as a result of DSA (which is not considered in the model). Furthermore, the model predicts much too high SCC crack growth rates at KI levels < 60 MPa m1/2 in high-purity water.38 This is related to the cracktip strain rate formulation for static loading conditions, which was originally derived for stainless steels and is generally unsatisfactory when applied at these lower KI levels.38 In addition, the model does not explain the unexpectedly strong effect of relatively low chloride levels on SCC crack growth rates (see Figure 19). 5.06.3.2.5 Service experience and mitigation actions 5.06.3.2.5.1 Service experience

The accumulated operating experience with CS & LAS pressure-boundary components in the primary and secondary coolant circuits of LWRs is very good

worldwide. Use of the current fatigue design and evaluation codes (ASME III and XI) has been generally successful in preventing fatigue cracks and failures in CS & LAS components and therefore the limited (and nonspecific) consideration of EAC in these codes would – at first sight – seem to be adequate under most operating circumstances, or even (fortuitously) conservative. However, isolated instances of EAC have occurred under specific circumstances and these always involved several unfavorable factors. EAC has occurred particularly in BWR service, most often in CS & LAS piping, less frequently in vessels, and very rarely in the clad RPV. EAC cracking has been observed in wrought, weld filler, and weld HAZ materials and has been transgranular in nature. In BWRs, steam and feedwater piping – as well as condensate systems and RPV feedwater nozzles – have been affected (Figure 25). In the secondary circuit of PWRs, cracking has been observed in feedwater piping/tanks and heat exchangers, feedwater nozzles of PWR steam generators, and steam generator girth welds. EAC damage was usually detected during in-service inspection and seldom led to through-wall penetrations with leakage.1,38,39,46,63,71–74 These EAC incidents have been clearly associated with an oxidizing environment (usually due to dissolved oxygen), severe dynamic straining (due to global and local thermal stratification/striping, or due to thermal and pressurization cycles during plant transients), and either high local stresses (around or above the high-temperature yield strength) or high secondary/residual stress (due, e.g., to welding defects, pipe bends in conjunction with inadequate pipe supports or restraints, and localized post-weld treatment). Most PWR cracking incidents involved aerated feedwater under specific start-up operating conditions (no application of nitrogen blankets in condensate storage tanks) or at special locations (e.g., locations with air present during shutdown), where the normally low contents of oxygen and low ECP during operation cannot be assumed upon startup. The fact that cracking frequently occurred in BWR lines with stagnant steam or nondegassed condensate, but not in comparable lines carrying flowing steam, is another indication that oxygen concentration and ECP are crucial parameters with respect to the occurrence of EAC, since such conditions are known to encourage the formation of a condensate film rich in dissolved oxygen. Thermal hydraulics (thermal stratification/ striping phenomena) and local stress raisers played a key role and were often more important than water

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

5 mm

(a)

(c)

(b)

Feed water pipe

Weld root

5 mm

137

5 mm

RPVnozzle

(d)

200 mm

Figure 25 (a) Circumferential strain-induced corrosion cracking crack at the feedwater pipe to reactor pressure vessel nozzle weld of a German boiling water reactor originating form a weld root notch due to severe thermal stratification during plant start-up and hot stand-by at low feedwater flow rates. Reproduced from Hickling, J.; Blind, D. Nucl. Eng. Des. 1986, 91, 305–330. (b) Crack interruptions markings on a strain-induced corrosion cracking fracture surface indicating intermittent crack growth during plant transients. Reproduced from Hickling, J.; Blind, D. Nucl. Eng. Des. 1986, 91, 305–330. (c) Formation of multiple, semielliptical strain-induced corrosion cracking cracks at a circumferential weld between a straight seamless pipe and a valve casing in a boiling water reactor. Reproduced from Hickling, J.; Blind, D. Nucl. Eng. Des. 1986, 91, 305–330. (d) Corrosion fatigue cracks in carbon steel instrumentation line at the main steam pipe of a pressurized water reactor originating from corrosion pits with intermittent corrosion during shutdown and subsequent reinitiation of cracking during plant transients. Reproduced from Roth, A. et al. In Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power System – Water Reactors, Snowbird, Salt Lake City, UT, Aug 14–18; Todd, R. A., King, P. J., Nelson, L., Eds.; The Minerals, Metals and Materials Society: Warrendale, PA, 2005; pp 795–802.

chemistry or material aspects. These cases were attributed either to SICC or CF. Cracking incidents with a major or relevant contribution of SCC to the total crack advance in properly manufactured and heat-treated CS & LAS primary pressure-boundary components have not been observed. In fact, longterm exposure and reinspection of unclad components, such as BWR feedwater nozzle radii or vessel heads, and several cracking incidents in dissimilar metal welds (e.g., in core shroud support welds, where crack-tips in the susceptible Alloy 182 weld metal arrested in the interface region to the adjacent LAS) confirm the generally excellent resistance of CS & LAS to SCC in reactor coolant.38,39,72,73 Cracks often initiated from water-wetted geometric discontinuities (e.g., welding defects) and/or in regions with stagnant flow or creviced geometry.

In several cases, EAC cracks preferentially initiated from corrosion pits (Figure 25) and corrosion during shutdown periods (pitting and sometimes corrosion products (rust) close to the water line as a possible source of anionic impurities during startup) can be a further contributing factor. In general, crack initiation has been more affected by high-frequency, high-cycle fatigue, for example, due to local thermal stratification or thermal striping loads which were limited to nearsurface regions. Crack propagation was often dominated by LCF from slower and less frequent transients, for example, due to more global thermal stratification or to operational power transients.38,39,72,73 Several incidents were related to unanticipated sources of thermal stress cycles in critical locations. In many cases, fabrication or design deficiencies (e.g., welding defects) favored local plastification of

138

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

the water-wetted material surface and sometimes resulted in an increased loading level (from residual stress) or increased EAC susceptibility of the material (e.g., excessive hardness of weld HAZs and distinct DSA susceptibility of weld metal). Reoccurrence of this kind of cracking can be avoided by improved design of the components and/or fabrication procedures, as well as by better quality control during the manufacturing process, and – in some cases – by optimized operational procedures.38,39,72–74





5.06.3.2.5.2 Critical components and operational conditions

Special attention should be paid to components which are likely to undergo significant (localized) mechanical loading, or which are exposed to increased oxygen levels in the coolant, higher conductivities, and/or quasi-stagnant flow conditions during operation. Critical components, for example, are piping carrying nearly stagnant steam, or nondegassed condensate, during normal reactor operation (e.g., in systems which are used only intermittently), feedwater nozzles and adjacent sections of horizontal piping (if thermal stratification can occur) and thin-walled piping, and pipe bends in conjunction with inadequate pipe supports or restraints. Operational transients associated with thermal stratification and thus high secondary strains (e.g., start-up, hot stand-by, or slow, but relevant pressure and temperature changes) should receive attention, in particular, if they occur with a sufficiently high frequency.38 5.06.3.2.5.3 Possible mitigation actions and countermeasures

Apart from good design, the common EAC mitigation strategy is, firstly, to exclude large preexisting defects by nondestructive examination and quality assurance measures during fabrication and installation of components and, secondly, to avoid during operation those water chemistry and stress or strain combinations which could lead to either EAC crack initiation or to accelerated crack growth. This procedure should be complemented by periodic, nondestructive, in-service inspection of critical component locations. In summary, EAC risks can be minimized by38: Selection of suitable materials, for example, lowsulfur steel (<0.003 wt% S) with low EAC and DSA susceptibility and optimized high toughness (! larger critical crack size). Selection of suitable manufacturing and fabrication practices to avoid welding defects and HAZs





with high hardness (350 HV5) and to reduce residual stress (e.g., stress relief heat treatment and narrow gap welding). Improved design to reduce regions of high local stresses (by increased wall thickness, by internal flush grinding of joints and optimization of welding technology, avoidance of discontinuities and constraints, optimized pipe supports, etc.). Reduction of the number and severity of thermal and pressurization cycles (thermal-stratification during hot stand-by, startup/shutdown), for example, by optimized operating procedures and/or by improved design of the affected component (thermal sleeves for feedwater nozzles, feedwater spargers, etc.). Avoidance of near-stagnant operating conditions (including in crevices or dead-legs) and reduction in dissolved oxygen levels (e.g., by inert gas shielding of make-up water, modification of startup procedures to allow venting of piping, improvement of drainage in horizontal lines, etc.). Maintenance of adequate water chemistry control with regard to anionic impurities (EPRI or VGB guidelines) and application of HWC/NMCA to reduce the ECP in BWRs.

5.06.3.2.5.4 Service experience vs. experimental and theoretical background knowledge

Plant-operating experience fits well to the accumulated experimental/mechanistic background knowledge on EAC and pitting behavior for CS & LAS components exposed to high-temperature water. Both show the same qualitative parameter trends, for example, for dissolved oxygen (ECP), flow rate, and strain. They confirm the high SCC resistance of CS & LAS under static loading conditions (steady-state power operation) and clearly reveal that slow, positive (tensile) straining with associated (localized) plastic yielding and sufficiently oxidizing conditions are essential for EAC initiation in high-purity, high-temperature water. Since primary design stresses are generally limited to values below the elastic limit, thermal-hydraulics (e.g., thermal stratification) and local stress raisers have therefore played a key role for SICC/CF in the field and were generally more important than material or water chemistry aspects. EAC cracks often initiated from corrosion pits and pitting of LAS is strongly favored under oxidizing conditions, especially at low and intermediate temperatures, and by slow, dynamic straining. Quasi-stagnant or low-flow conditions promote EAC initiation by the formation of an aggressive, occluded water chemistry at the base of the pit

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

(or within small surface defects), as well as by the enrichment of oxygen in steam condensate.38,39,72,73 By taking the exact boundary conditions in laboratory tests and in the field into account, no major discrepancy between operating experience and experimental trends is found. The few EAC cracking incidents in the field can be rationalized by the identified susceptibility conditions and by EAC flaw tolerance evaluations. The apparently higher cracking frequency in lab tests may be explained by the beneficial effect of a turbulent flow rate in the field (which is characteristic for most component locations) upon EAC initiation, the large degree of conservatism in the fatigue evaluation procedures, and the conjoint threshold conditions for the onset of EAC with regard to ECP (or dissolved oxygen), strain rate, strain, and temperature. In general, for most transients/component locations, one or more threshold conditions are not satisfied. The threshold strain and stress for SICC/CF and SCC initiation, in particular, are seldom exceeded in the case of good operational practice and the absence of design or fabrication deficiencies.38,39,71–73

5.06.4 Conclusions 5.06.4.1 Uniform and Flow-Accelerated Corrosion Uniform corrosion of CS and LAS poses no significant threat to the integrity of coolant systems under normal operating conditions; the typical corrosion rates of less than 1 mm year1 or so are manageable. A significant drawback of the material is the rather thick corrosion-product layer that develops in hightemperature water – double-layered magnetite in alkaline, reducing conditions (such as PWR or CANDU primary coolants) but with an admixture of oxidized species such as maghemite or hematite in neutral, oxidizing conditions (such as BWR coolant). The fouling and contamination of reactor circuits from the release of such oxides by spalling in high flows, as well as by dissolution as a result of solubility gradients, is minimized by careful chemistry control. Similarly, although the material in contact with PWR primary coolant is mainly stainless steel or nickelbased alloy, the transport of nickel-ferrite corrosion products is minimized by continuous control of alkalinity. The reactor circuit in BWRs exposes mainly stainless steel to the coolant, but – depending upon the chosen operating chemistry (NWC, HWC, HWC/NMCA) – the chief source of corrosion

139

products (CS/LAS in the feedtrain) can be controlled by the addition of metal ions to the feedwater so as to give an optimum ratio of Fe/Ni and minimize radiation field buildup. CANDUs have unclad CS in the primary circuit and operate at constant high pH; although this controls activity transport, general fouling of the circuit from corrosion products released from FAC can create thermohydraulic problems. Some CANDUs have had to clean components in the primary system in order to restore optimum operating conditions. The stainless-steel cladding in PWR primary circuits prevents corrosion of the LAS base material, especially under shutdown conditions. However, leaks at high temperature can allow the boric acid in the coolant to concentrate on unclad external surfaces and, together with oxygen from the atmosphere, can provoke highly corrosive situations. Flow effects can exacerbate such attack. Rigorous surveillance regimes are required to avoid damage to components such as flanges, valve fittings and especially, the pressure-vessel head in the vicinity of leaking penetrations. Corrosion of unclad CS & LAS in PWR secondary systems is prevented by chemistry control during HT-operation, but additional measures (e.g., nitrogen blanketing) are required during shutdown. FAC of CS in reactor systems is most damaging in feedwater circuits, where the solubility of normally protective oxide films is high, dissolved iron is low, and turbulence is intense. Chronic FAC of feeders in the CANDU primary circuit occurs through similar mechanisms, although the oxide solubility is low at the constant high pH and temperature of the coolant. Depending upon the circumstances, oxide solubility can be controlled to some extent by adjusting the chemistry with pH additives (such as ammonia) and/or adding oxygen (as is done in various feedwater circuits) or even by replacing components with higher-Cr material. Commercial software is available to pinpoint likely vulnerable areas, give some indications of possible rates of attack, and manage databases. However, rigorous surveillance of vulnerable components, for example, where high coolant turbulence is induced, is still necessary to avoid catastrophic pipe failures. 5.06.4.2 Localized Corrosion and Environmentally Assisted Cracking In CS & LAS, a shallow form of pitting can occur in the complete absence of anionic water impurities as the corrosion potential (ECP) is raised, for example,

140

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

through oxygen and other oxidizing species. Some degree of pitting corrosion is inevitable after longterm exposure of unclad CS & LAS surfaces to water in LWR systems and is not usually a threat to either coolant purity or to structural integrity. Shallow pitting has been observed primarily in specific piping systems with residual water because of incomplete draining during nonoperational periods (shutdown corrosion). This can be avoided by adequate wet or dry preservation techniques. Laboratory investigations have revealed significant effects of simulated reactor environments on fatigue crack initiation/growth, as well as the possibility of SCC crack growth under certain critical combinations of environmental, material, and loading parameters. During the last three decades, the major factors of influence and EAC susceptibility conditions have been clearly identified. The effects of most parameters on EAC initiation and growth are adequately known with acceptable reproducibility and reasonably understood by mechanistic models. Tools for incorporating environmental effects into ASME III fatigue design curves have been developed/qualified, but not yet widely adopted. The BWRVIP-60 SCC disposition lines and ASME XI reference fatigue crack growth curves are usually conservative and adequate under most PWR and BWR operational circumstances, but there is nevertheless valuable potential for the reduction of undue conservatism and the need for improvements to eliminate some specific inadequacies. The operating experience of CS & LAS primary pressure-boundary components in LWRs is very good worldwide. However, isolated instances of EAC have occurred, particularly in BWR service, most often in piping and more rarely in the RPV itself. Oxidizing conditions (usually dissolved oxygen) and relevant plastic straining were always involved. These cases were attributed either to SICC or CF and can be rationalized by the experimental/mechanistic background knowledge. Both service experience and appropriate laboratory experiments confirm the high resistance of CS & LAS to SCC under static loading conditions (representative of steady-state power operation) and clearly reveal that slow, positive (tensile) straining, with associated (localized) plastic yielding and sufficiently oxidizing conditions, are usually necessary for EAC initiation in high-temperature water. Based on the above, different remedial and mitigation actions have been qualified and successfully applied to further reduce the low EAC cracking frequency in the field.

In spite of the absence of SCC in operating reactors, several unfavorable, critical parameter combinations have been identified in the laboratory, which can lead to sustained, fast SCC, with crack growth rates well above the BWRVIP-60 SCC disposition lines. Most of them appear atypical for current BWR plant operation with properly manufactured CS & LAS components, but some further clarification is required, in particular, with regard to improved quantification of the boundaries/thresholds for the transition from very low to high SCC crack growth rates. In this context, research should be focused on the effects of chloride transients and DSA/ yield strength on the SCC crack growth behavior of CS & LAS (including weld and HAZ materials) under BWR/NWC conditions, as well as on the behavior of Alloy 182-LAS dissimilar metal weld joints. Additionally, the mitigation effect of HWC/NMCA should be confirmed under these critical conditions.

References 1. 2.

3. 4. 5. 6. 7. 8. 9. 10.

11.

12.

13.

Electric Power Research Institute. Materials handbook for nulcear plant pressure-boundary components, EPRI 1002792; EPRI: Palo Alto, CA, 2002. http://my.epri.com/. Electric Power Research Institute. Boric acid corrosion guidebook, revision 1; managing boric acid corrosion issues at PWR power stations, EPRI 1000975; EPRI: Palo Alto, CA, 2001http://my.epri.com/. Potter, E. C.; Mann, G. M. W. In Proceedings of the 1st International Congress on Metallic Corrosion, Butterworths: London, 1961; p 417. Berge, P.; et al. Corrosion 1977, 33(5), 173–178. Bloom, M. C.; et al. Corrosion 1957, 13, 297–302. Mabuchi, K.; et al. Corrosion 1991, 47(7), 500–508. Hickling, J. Private communication; EPRI: Palo Alto, CA, 2006. Beverskog, B.; Puigdomenech, I. Corros. Sci. 1996, 38(12), 2121–2135. Dooley, R. B. Power Plant Chem 2008, 10(2), 68–89. Uchida, S.; et al. In Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, CD-ROM. Whistler, British Columbia, Canada, Aug 19–23; King, P., Allen, T., Busby, J., Eds.; The Canadian Nuclear Society: Toronto, ON, 2007. Slade, J.; Gendron, T. In Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power System – Water Reactors, Snowbird, Salt Lake City, UT, Aug 14–18; Todd, R. A., King, P. J., Nelson, L., Eds.; The Minerals, Metals and Materials Society: Warrendale, PA, 2005. Marks, C.; et al. In Proceedings of the 6th International Symposium on Contribution of Materials Investigations to Improve the Safety and Performance of LWRs (Fontevraud 6), CD-ROM, Fontevraud, France, Sept 18–22. The French Nuclear Energy Society (SFEN): Paris, France, 2006. Woolsey, I.; Quirk, G. In International Conference on Flow Accelerated Corrosion (FAC2008), DVD-ROM, Lyon, France, Mar 17–20; EDF: France, 2008.

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels 14. Bouchacourt, M. EPRI Workshop on Erosion-Corrosion of Carbon Steel Piping, Washington, DC, Apr 14–15, EPRI: Palo Alto, CA, 1987. 15. Berge, P.; et al. In Proceedings of the Second International Conference on Water Chemistry of Nuclear Reactor Systems, Bournemouth, UK, October, British Nuclear Energy Society (BNES): London, UK, 1980; Vol. 1. 16. Lister, D. H.; et al. In Proceedings of the 16th Pacific Basin Nuclear Conference (16PBNC), CD-ROM, Aomori, Japan, Oct 13–18 Japanese Industrial Forum (JAIF): Japan, 2008. 17. Liu, L.; Lister, D. H. In Proceedings of the International Conference on Water Chemistry of Nuclear Reactor Systems (NPC’08), Berlin, Germany, Sept 15–18, VGB: Essen, Germany, 2008. 18. Bignold, G. J.; et al. In Proceedings of the International Specialist’s Meeting on Erosion–Corrosion of Steels in High-Temperature Water and Wet Steam, Les Renardie´res, France, May 11–12 EDF: France, 1982. 19. Tremaine, P.; LeBlanc, J. C. J. Solution Chem. 1980, 9(6), 415–442. 20. Heitmann, H. G.; Schub, P. In Proceedings of the Third Meeting on Water Chemistry of Nuclear Reactors, Bournemouth, UK, October British Nuclear Engineering Society (BNES): London, UK, 1983. pp 243–252. 21. Berger, F. P.; Hau, K. F. F. L. Int. J. Heat Mass Transf. 1977, 20, 1185. 22. Sydberger, T.; Lotz, U. J. Electrochem. Soc. 1982, 129(2), 276–283. 23. Lister, D. H.; et al. In Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, CD-ROM, Whistler, British Columbia, Canada, Aug 19–23; King, P., Allen, T., Busby, J., Eds.; The Canadian Nuclear Society: Toronto, ON, 2007. 24. Lister, D. H.; et al. Power Plant Chem. 2008, 10(11), 659–667. 25. Henzel, N.; et al. Erosion corrosion in power plants under single- and two-phase flow conditions-updated experience and proven counteractions. In Proceedings of the American Power Conference, Chicago, IL, USA, 18–20 April 1988; Vol. 50. pp 992–1000. 26. Lee, S.-H.; et al. J. Korean Nucl. Soc. 1999, 31(4), 375–384. 27. Lee, S. H.; et al. J. Korean Nucl. Soc. 1999, 31(4), 375–384. 28. Lister, D. H. In Proceedings of the 1998 JAIF International Conference on Water Chemistry in Nuclear Power Plants, Kashiwazaki, Japan, 13–16 Oct, Japan Atomic Industrial Forum (JAIF): Japan, 1988; pp 442-1–442-8. 29. Balakrishnan, P. V. Can. J. Chem. Eng. 1977, 55, 357. 30. Lister, D. H.; Lang, L. C. In Proceedings of the International Conference on Water Chemistry in Nuclear Reactor Systems (Chimie 2002), CD-ROM, Avignon, France, Apr 22–26; The French Nuclear Energy Society (SFEN): Paris, France, 2002. 31. Ponguak, J.; et al. Boric acid corrosion of reactor pressure vessel steel caused by an impinging jet of simulated PWR coolant. In Proceedings of the International Conference on Water Chemistry of Nuclear Reactor Systems, CD-ROM, Jeju Island, Korea, 23–26 Oct, 2006. 32. The Nuclear and Industrial Safety Agency. Secondary piping rupture accident at Mihama power station, unit 3 of the Kansai Electric Power Co., Inc. (Final Report), Revision 1 (translated by Japan Nuclear Energy Safety Organization (JNES)); The Nuclear and Industrial Safety Agency: Japan, 2005; http://www2.jnes.go.jp/. 33. McGrath, M. A.; et al. In Proceedings of the International Conference on Water Chemistry of Nuclear Reactor Systems (NPC’08), Berlin, Germany, Sept 15–18 VGB: Essen, Germany, 2008.

34.

35. 36. 37.

38.

39. 40.

41. 42. 43. 44.

45. 46.

47. 48.

49. 50.

51. 52. 53.

54.

141

Electric Power Research Institute. The general and localized corrosion of carbon and low-alloy steels in oxygenated high-temperature water, EPRI-NP-2853; EPRI: Palo Alto, CA, 1983; http://my.epri.com/. Electric Power Research Institute. BWR environmental cracking margins for carbon steel piping. EPRI Report NP-2406; EPRI: Palo Alto, CA, 1982; http://my.epri.com/. Indig, M.; et al. Rev. Coat. Corros. 1982, 5, 173–225. Videm, K. Corrosion of steel in high-temperature water – influence of oxygen in water of performed oxide coatings. In Proceedings of the 7th Scandinavian Corrosion Congress, Trondheim, Norway, May 26–28, 1975; pp 444–456. Seifert, H. P.; Ritter, S. Research and service experience with environmentally-assisted cracking of carbon & lowalloy steels in high-temperature water. SKI-Report 2005:60; SKI: Stockholm, Sweden, 2005; ISSN 11041374; http://www.stralsakerhetsmyndigheten.se/. Hickling, J.; et al. PowerPlant Chem. 2005, 7, 31–42. Seifert, H. P.; et al. Crack initiation due to environmentallyassisted cracking in carbon and low-alloy steels exposed to high-temperature water-Part 1: Overview of results from laboratory tests. Workshop on Detection, Avoidance, Mechanisms, Modeling, and Prediction of SCC Initiation in Water-Cooled Nuclear Reactor Plants (CD-ROM), Beaune, Burgundy, Sept 7–12, 2008. Chopra, O. K.; Shack, W. J. J. Pressure Vessel Technol. 2009, 131: 021409-1–021409-21. US NRC. Effect of LWR coolant environments on the fatigue life of reactor materials, NUREG/CR-6909; US NRC: Washington, DC, 2007; http://www.nrc.gov/. Congleton, J.; et al. Corros. Sci. 1985, 25, 633–650. Atkinson, J. D.; et al. In Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, CD-ROM, Whistler, British Columbia, Canada, Aug 19–23; King, P., Allen, T., Busby, J., Eds.; The Canadian Nuclear Society: Toronto, Canada, 2007. Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 378, 312–326. Electric Power Research Institute. Corrosion fatigue of water-touched pressure retaining components in power plants, EPRI TR-106696; EPRI: Palo Alto, CA, 1997; http://my.epri.com/. Seifert, H. P.; Ritter, S. Corros. Sci. 2008, 50, 1884–1899. Ford, F. P.; Andresen, P. L. In Proceedings of the 3rd International Atomic Energy Agency Specialist’s Meeting on Sub-critical Crack Growth, Moscow, USSR, May 14–17; Cullen, W., Ed.; US NRC: Washington, DC, 1990. Vol. 1; 105–124, NUREG/CP-0112. Ford, F. P. J. Pressure Vessel Technol. 1988, 110, 113–128. The American Society of Mechanical Engineers. The 2004 ASME Boiler and Pressure Vessel Code, Section III, Rules for Construction of Nuclear Power Plant Components; The American Society of Mechanical Engineers: New York, 2004. Higuchi, M.; Iida, K. Nucl. Eng. Des. 1991, 129, 293–306. Van Der Sluys, A. PVRC’s position on environmental effects on fatigue life in LWR applications, WRC Bulletin 487; Welding Research Council: New York, 2003. Higuchi, M. In Proceedings of the ASME 2008 Pressure Vessel and Piping Conference (PVP 2008), CD-ROM, Chicago, IL, 27–31 July; The American Society of Mechanical Engineers: New York, 2008; ISBN 0-7918-3828-5. The Japanese Society of Mechanical Engineers. Codes for nuclear power generation facilities: Environmental fatigue evaluation method for nuclear power plants, JSME S NF1–2006; The Japanese Society of Mechanical Engineers: Tokyo, Japan, 2006.

142

Corrosion and Environmentally-Assisted Cracking of Carbon and Low-Alloy Steels

55. US NRC. Guidelines for evaluating fatigue analyses incorporating the life reduction of metal components due to the effect of the light-water reactor environment for new reactors, NRC Regulatory Guide 1.207; US NRC: Washington DC, USA, 2007; http://www.nrc.gov/. 56. The American Society of Mechanical Engineers. The 2004 ASME Boiler and Pressure Vessel Code, Section XI, Rules for Inservice Inspection of Nuclear Power Plant Components; The American Society of Mechanical Engineers: New York, 2004. 57. The American Society of Mechanical Engineers. The 2004 ASME Boiler and Pressure Vessel Code, Code Cases for Nuclear Components, Code Case N-643; The American Society of Mechanical Engineers: New York, 2004. 58. Hickling, J.; Reitzner, U. VGB Kraftwerkstech. 1992, 72, 359–367. 59. Kussmaul, K.; et al. Nucl. Eng. Des. 1997, 168, 53–75. 60. Seifert, H. P.; Ritter, S. J. Nucl. Mater. 2008, 372, 114–131. 61. Ford, F. P.; et al. In Proceedings of the 9th International Conference on Environmental Degradation of Materials in Nuclear Power Systems-Water Reactors, Newport Beach, CA, Aug 1–5; Bruemmer, S., Ford, P., Was, G., Eds.; The Minerals, Metals and Materials Society: Warrendale, PA, 1999; pp 855–863. 62. Seifert, H. P.; Ritter, S. Review and assessment of SCC experiments with RPV steels in Oskarshamn 2 and 3 (ABB Report SBR 99–020), SKI Report 2005:61; The Swedish Nuclear Power Inspectorate: Stockholm, Sweden, 2005; ISSN 1104–1374; http://www. stralsakerhetsmyndigheten.se/. 63. Electric Power Research Institute. Environmentally assisted cracking of low-alloy steels, EPRI NP-7473-L; EPRI: Palo Alto, CA, 1992; http://my.epri.com/. 64. Electric Power Research Institute. BWR water chemistry guidelines – 2004 revision, BWRVIP-130, 1008192; EPRI: Palo Alto, CA, 2004; http://my.epri.com/. 65. Ha¨nninen, H.; et al. Corros. Sci. 1983, 23, 663–679. 66. Ha¨nninen, H.; et al. In Proceedings of the 2nd International Atomic Energy Agency Specialist’s Meeting on SubCritical Crack Growth, Sendai, Japan, 15–17 May;

Cullen, W. H., Ed.; US NRC: Washington, DC, 1985, Vol. 2; 73–97, NUREG/CP-0067. 67. Sund, G.; Rosborg, B. The influence of impurities on the tendency to stress corrosion cracking of pressure vessel steel A533-B in water at 288 C. Personal communication. 1991. 68. Ha¨nninen, H.; et al. In Proceedings of the 10th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, CD-ROM, Lake Tahoe, NV, Aug 6–10; The National Association of Corrosion Engineers: Houston, TX, 2001; Paper No. 47. 69. Andresen, P. L. In Proceedings of the Third International Conference on Fatigue and Fatigue Thresholds, Charlottesville, VA, June 28–July 3; Ritchie, R. O., Starke, E. A., Eds.; EMAS Publishing: Warrington, UK, 1987; 1189–1200, Vol. III. pp 1189–1200. 70. Electric Power Research Institute. Environmentally assisted fatigue crack initiation in low-alloy steels – A review of the literature and the ASME Code requirements, EPRI TR-102765; EPRI: Palo Alto, CA, 1993; http://my.epri.com/. 71. Hickling, J.; Blind, D. Nucl. Eng. Des. 1986, 91, 305–330. 72. Roth, A.; et al. Crack initiation due to environmentallyassisted cracking in carbon and low-alloy steels exposed to high-temperature water – Part 2: Overview and assessment of operating experience. Workshop on Detection, Avoidance, Mechanisms, Modeling, and Prediction of SCC Initiation in Water-Cooled Nuclear Reactor Plants, CD-ROM, Beaune, Burgundy, France, Sept 7–12, 2008. 73. Roth, A.; Hickling, J. In Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, CD-ROM, Whistler, British Columbia, Canada, Aug 19–23; The Canadian Nuclear Society: Toronto, Canada, 2007. 74. Roth, A.; et al. In Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power System – Water Reactors, Snowbird, Salt Lake City, UT, Aug 14–18; Todd, R. A., King, P. J., Nelson, L., Eds.; The Minerals, Metals and Materials Society: Warrendale, PA, 2005; pp 795–802.

5.07

Performance of Aluminum in Research Reactors

K. Farrell Formerly of Oak Ridge National Laboratory, Oak Ridge, TN, USA

ß 2012 Elsevier Ltd. All rights reserved.

5.07.1 5.07.2 5.07.2.1 5.07.3 5.07.3.1 5.07.3.2 5.07.4 5.07.5 5.07.6 5.07.6.1 5.07.6.2 5.07.6.2.1 5.07.6.2.2 5.07.6.2.3 5.07.7 5.07.7.1 5.07.7.2 5.07.7.3 5.07.7.4 5.07.8 References

Introduction Typical Applications History of Aluminum Applications in Research Reactors Properties of Aluminum Practical Characteristics Alloy Types, Temper Designations, and Tensile Properties Fuel Elements Corrosion Radiation Effects Basics Microstructures Fluence Temperature Transmutation products Property Changes Swelling Mechanical Properties Effects of Neutron Spectrum Radiation Softening, Creep, and Stress Relaxation Conclusion

Abbreviations AIME ANL ANSI ASM ASTM ATR CRC CTE EBR-II Emod ETR GR HEU HFIR HPRR IAEA INL IRV-M2

American Institute of Mining, Metallurgical, and Petroleum Engineers Argonne National Laboratory American National Standards Institute American Society for Metals American Society for Testing Materials Advanced Test Reactor Chemical Rubber Company Coefficient of thermal expansion Experimental Breeder Reactor-II Modulus of elasticity Experimental test reactor Graphite Reactor Highly enriched uranium High Flux Isotope Reactor High performance research reactor International Atomic Energy Authority Idaho National Laboratory Acronym for a recent Russian research reactor

LANL LEU MTR

OPAL ORNL ORR PIE PIREX RERTR RR SNF STP TRIGA TEM UTS VPH YS

144 144 144 145 146 147 149 153 158 158 159 160 161 161 166 166 166 169 170 173 173

Los Alamos National Laboratory Low enriched uranium Specifically, MTR is the Materials Testing Reactor at Idaho National Laboratory. Also used generically for materials test reactors Open Pool Australian Light water reactor Oak Ridge National Laboratory Oak Ridge Research Reactor Post irradiation examination Proton Irradiation Experiment facility Reduced enrichment for research and test reactors Research reactor Spent nuclear fuel Special Technical Publication Test, research, isotopes, general atomic Transmission electron microscopy Ultimate tensile stress Vickers pyramid hardness Yield stress

143

144

Performance of Aluminum in Research Reactors

5.07.1 Introduction Aluminum alloys are generally too weak or have temperature limitations that preclude their use in reactors built to produce electricity, high-temperature process heat, or marine propulsion. But in the milder conditions in most research reactors (RRs) where bulk water coolant temperatures are usually <100
C, aluminum alloys are quite comfortable and are universally employed and have greatly contributed to the success and longevity of the reactors. RRs are those whose principal function is to generate neutrons for purposes of nuclear education and training, production of medical and industrial isotopes, neutron activation analyses, neutron scattering studies, and even semiconductor doping, neutron radiography, and food preservation treatments. RRs are also employed to study basic radiation effects in materials and as test beds for evaluating candidate structural materials and fuels/assemblies for power reactors. RRs come in many shapes, sizes, and types. For descriptions of the various classes of RRs, see http://www.world-nuclear.org/ and West.1 They are generally low power, typically about a few kilowatts, thermal, but range up to about 250 MW. According to the recently updated list2 of worldwide RRs published by the IAEA, a total of 674 RRs have been built in 57 countries, of which 234 are still operational, and 7 are planned or under construction. Two new ones are OPAL, the 20 MW Open Pool Australian Light water-cooled reactor, which opened at Lucas Heights, Sydney, in April 2007, and the Russian 4 MW pooltype IRV-M2 commissioned in 2008. This chapter is a review, more a tutorial, of the behavior of aluminum alloys in water-cooled RRs. It is a somewhat personal view, based on American experience in the area. Because that experience has been adopted in many countries and is still influencing the current state of the art, this chapter should be of interest outside the borders of the United States.

5.07.2 Typical Applications Aluminum is the material of choice for construction of many components in low-temperature water-cooled-and-moderated RRs. Typical applications are the reactor tanks in open-pool reactors; containment vessels in some sealed reactors; core grids; pedestals; neutron beam tubes; cold neutron source moderator vessels; shrouds to direct and

separate water flows; shuttles (‘rabbits’) and aluminum filler powder used to convey isotope target materials and test materials rapidly in and out of the reactor via aluminum hydraulic and pneumatic tubes; sheaths and finned tubing for stationary longterm isotope target rods; cladding for control plates/ rods; cladding and liners for reflector materials; cladding and thermal conduction filler for fuel rods/ plates; and temporary plugs for closing idle irradiation facilities in and around the core. Applications outside the reactor per se are in-pool tool extension arms; transfer gates between pool sections; restraint baskets in some shipping casks; support beams for pool covers; and hot cells manipulator arms. 5.07.2.1 History of Aluminum Applications in Research Reactors Aluminum was at the forefront of the development of nuclear technology. It has the distinction of being the first nonfissile, non-neutron absorber class metal used in the world’s first continuously operating nuclear reactor, the X-10 Graphite Reactor at Oak Ridge, TN. The Graphite Reactor became critical on 4 November 1943, <1 year after Fermi’s demonstration of a self-sustaining nuclear fission chain in the graphite pile at the University of Chicago on 2 December 1942. In Fermi’s experiment, the only metals in the pile were natural uranium and the cadmium-coated control rods. The pieces of natural uranium (238U containing about 0.7 at.% 235U) and uranium oxide were bare, placed in shallow depressions carved into the upper faces of the graphite slabs, and cooled by convection of ambient air. The power level was about 2 kW. The X-10 Graphite Reactor pile3 was much bigger than the Chicago pile and was designed to operate at 1 MW thermal power, later upgraded to 4 MW. It was built to produce pilot plant quantities of plutonium isotopes. The Chicago pile had no shielding; the Graphite Reactor was shielded by a 2.2-m thickness of high-density concrete. Aluminum made its debut in the Graphite Reactor as fuel cladding to protect the highly chemically reactive uranium from contamination by air and graphite during the higher power and longer fissioning periods and to safeguard it from attack by water during subsequent radioactive decay in underwater storage. In addition, it trapped the more copious volatile radiation products resulting from the longer irradiation exposures. These aluminum–clad pieces of natural uranium, called ‘slugs,’ were the forerunners of metal–clad fuel

Performance of Aluminum in Research Reactors

elements. A slug was made by placing a solid cylinder of uranium in a thimble-shaped aluminum can 25 mm diameter  100 mm long with a 0.75 mm wall. A flat Al end cap was added, and the assembly was passed through a die to force the can walls tightly around the fuel. Surplus wall material was cut off above the cap, and the cap was welded all around its edge. These slugs were pushed end to end into the reactor via round horizontal holes through the concrete face, which were aligned with 44 mm square holes cut through the full 7.3 m width of the cubic array of graphite blocks. The square holes were oriented on edge such that the slugs occupied the lower corner, allowing cooling spaces around the slugs. Cooling was simple: two large fans at the rear of the pile sucked ambient air through the holes around the slugs and discharged it up a tall chimney. The slugs exited the pile at the rear face and were channeled into a deep water canal where they were held until shipped to hot cells for processing to extract the plutonium. Some early problems4 were encountered in the slugs, including faulty welds and blisters and formation of an intermetallic UAl3 phase by interdiffusion at the U–Al interface, especially in the high-temperature regions in the center of the reactor. The blistering was attributed to fast-growing gas bubbles in the UAl3 phase. These problems were overcome by better welding practice and the development of bonded slugs as described next. The next phase of exploitation of aluminum was in the B reactor at the Hanford site in Washington State, which went critical on 27 September 1944. The B reactor was a scaled-up production model of the Graphite Reactor designed to operate at 250 MW. At such power, forced air cooling would have been inadequate. So the horizontal holes were replaced with aluminum tubes in which aluminum–clad uranium slugs were cooled with flowing water from the Columbia River. To improve the transfer of heat from the uranium to the cladding, the spaces between them were filled with a low melting Al–12% Si eutectic alloy by melting the eutectic in situ. A bonus of this treatment was that it killed the formation of the UAl3 phase and associated blistering, presumably due to an inhibiting effect of the silicon. The successes of these upgrades established aluminum as a suitable material for use in combined conditions of intensive irradiation and a flowing aqueous environment. Aluminum became more firmly entrenched in RRs with the development of advanced fuel elements, as described in Section 5.07.4.

145

Before concluding the present subsection, another lesser-known ‘first’ for aluminum deserves mention. It has particular relevance to the nuclear power industry. It is not widely known that aluminum was involved in the earliest demonstration of electricity produced from steam made by boiling water in a nuclear reactor. Normally, the heat from nuclear fission in RRs is discarded, not used to produce electricity. However, in August 1948, two staff members at the X-10 Graphite Reactor, Mansel Ramsey and Charles Cagle, placed an aluminum can containing ten aluminum–clad uranium slugs and some water in a normally unused side channel of the reactor where it was exposed to reactor neutrons. The trapped heat generated in the slugs boiled the water. Steam from the process was piped to a small model steam engine, rated at 1/1000 hp (0.75 W), which rotated an armature mounted between the poles of a permanent magnet. Sufficient electricity was generated to light a flashlight bulb. The thermal efficiency was estimated to be 2%. The Graphite Reactor is now a National Historic Landmark and is open to the public. A commemorative plaque and a replica of the steam engine and coupled dynamo from Ramsey and Cagle’s pioneering boiling water power reactor are displayed in a small showcase in the reactor lobby. The ‘official’ first production of nuclear electricity is credited to the lighting of another bulb on December 1951 at the Experimental Breeder Reactor-I, Arco, Idaho, now the Idaho National Laboratory.

5.07.3 Properties of Aluminum Heat removal and reduced generation of heat are major considerations in the popularity of aluminum in RRs. Most of the energy released from controlled nuclear fission appears as heat. Much of the heat, >80%, arises in the fuel from nuclear fissions. However, a significant portion, 5–20%, is produced in the nonfissile materials in the core and its surroundings by bombardment with particles emanating from the fission reactions and from decay of fission products. For power reactors, the heat is essential to generate the electrical output. In the case of RRs, the heat is a nuisance product; and the goals are to minimize heat generation from the nonfissile materials in the system and to get rid of it from those materials and from the fuel as fast as possible. Hence, structural materials that create the least heat and/or conduct it away the fastest are the most

146

Performance of Aluminum in Research Reactors

favored for RRs. In this regard, aluminum is outstanding. Generally, heat production is greater with increasing material density and with decreasing specific heat. It is increased by high cross-sections for neutron absorption and scattering, which also reduce reactor efficiency by stealing neutrons from participation in fission processes. Heat removal rate is larger with higher thermal conductivity. Therefore, construction materials with low density, high specific heat, high thermal conductivity, and low nuclear cross-sections offer the best prospects for minimizing heat generation and maximizing heat removal. In Table 1, the relevant properties for aluminum are compared with those of other cladding and structural materials used in power reactors and for uranium. All values are for room temperature or 100
C. The scatter in values for a given parameter and material is due in part to sensitivity to chemical composition and heat treatment, etc. These variations do not mask the large differences between Al and the other materials. The density of Al is 1/2–1/3 of those of the other cladding materials, and only 1/7 that of U. Its specific heat capacity is twice as high as the other materials. And its thermal conductivity is 5–10 times greater than the values for the other materials. Additionally, its neutron capture and scattering cross-sections are much smaller than those of the other materials, except for nuclear-grade Zr. In that respect, it should be remembered that in the early days when Al was establishing its foothold in nuclear technology commercial Zr was contaminated with up to 3% of the strong neutron absorber Hf. It was also inordinately expensive.

Table 1

5.07.3.1

Practical Characteristics

Having attractive physical properties for reactor use is of no merit if those properties cannot be exploited in a practical manner. The commercial and economic attributes of aluminum that encourage its deployment in RRs are: It is ductile, plentiful, cheap, and light weight. It is castable, machineable, and weldable, and it can be shaped readily by conventional processes of rolling, forging, extrusion, drawing, and cupping. It has good aqueous corrosion resistance due to near-insolubility in water and formation of a passive, self-restoring surface film of hydrated aluminum oxide. It is nonmagnetic and nonsparking. Although aluminum is inherently weak, it can be strengthened by cold work, solid solution hardening, and precipitation treatments. It has an fcc crystal structure and no crystallographic phase changes. Its crystal structure is near isotropic, ensuring that it will not suffer damaging directional thermal expansion and radiation growth like those exhibited by graphite and the hexagonal metals Mg and Zr. It does not form stable embrittling hydride phase(s) as Ti and Zr do. At low temperatures, it has no ductile-to-brittle transition. On the contrary, it is somewhat special in that at cryogenic temperatures, where it gains strength, it often gains ductility too. This combination of no hydride phase, outstanding low temperature properties, and low neutron cross-sections make aluminum the prime material for building cold neutron sources. Another attractive feature is that pure aluminum has no long-lived radioisotopes. The major source of immediate radioactivity is from decay of 24Na produced via 27Al(n,a)24Na, decaying by g-emission

Relevant properties of reactor materials

Material

Aluminum Zirconium Austenitic steel Ferritic steel Uranium

Density (kg m3)

2700 6490 8000 7900 1900

Specific heat (J kg1 K1)

887–963 254–285 377–565 440–494 111–167

Thermal conductivity (W m1 K1)

160–230 8–40 11–21 17–42 11–28

Melting point (
C)

660 1852 1425 1525 1132

Emod (GPa)

70 88–98 190–201 200–210 176–208

CTE, lin. (106 K1)

23 5.7 16 12 13.9

Nuclear cross-section (barns) sabs

ss

0.23 0.19 3.0 2.5 7.6

1.5 6.4 10 11 8.9

Sources: Matos, J. E.; Snelgrove J. L. Selected Thermal Properties and Uranium Density Relations for Alloy, Aluminide, Oxide, and Silicide Fuels; IAEA- TECDOC-643, International Atomic Energy Agency, Vienna, 1992; pp 1–19, article Appendix I-1.1 in Research reactor core conversion guidebook Volume 4; Fuels (Appendices I–K). Lide, D. R., Ed. CRC Handbook of Chemistry and Physics, 86th ed.; Taylor & Francis: Boca Raton, FL, 2005–2006; Gale, W. F.; Totemeier, T. C., Eds. Smithells Metals Reference Book, 8th ed.; Elsevier and ASM International, Amsterdam and Materials Park, OH, 2004; Cverna, F., Ed. ASM Ready Reference: Thermal Properties of Metals; ASM International: Materials Park, OH, 2002.

Performance of Aluminum in Research Reactors

with a half-life of 15 h. In alloys, long-lived radioactivity arises from decay of isotopes produced from alloying elements and residual impurity elements present in the aluminum, primarily 65Zn, 51Cr, 59Fe, with half-lives of 250, 28, and 45 days respectively. So if low residual radioactivity is an objective it can be met to a large extent by avoiding alloys containing significant quantities of Zn, Cr, and Fe. Aluminum is not without its shortcomings. It has a low elastic modulus and low melting temperature. The former means that in their annealed conditions aluminum alloys have low strengths compared with annealed austenitic steels, Zr, and bcc metals. However, aluminum can be hardened by various treatments as described in Section 5.07.3.2. However, the low melting temperature of 660
C imposes operating temperature limits of 100–150
C, which are homologous temperatures of 0.4–0.45Tm where lattice vacancies are mobile and can invoke susceptibility to creep and stress relaxation. Even without imposed stresses, the strength condition of prehardened alloys can become compromised at temperatures above 150
C because of the possibility of thermal overaging as described in Section 5.07.3.2 Aluminum has poor abrasion resistance. It can be sensitive to localized galvanic and pitting corrosion. It is prone to liquid metal embrittlement, particularly Hg. Difficulties may be encountered in obtaining leak-tight fusion welded joints for hi-tech applications, mainly due to porosities resulting from solidification shrinkage (volumetric change) and dissolved gases, in particular, hydrogen.5 In addition, aluminum does not undergo a color change on heating, and during manual welding may melt abruptly without warning, allowing overheating that can cause excessive sagging and dropthrough of the weld bead. The advent of a solid-state joining process, namely friction-stir welding,6 has largely overcome those welding troubles. 5.07.3.2 Alloy Types, Temper Designations, and Tensile Properties There is no universally embraced international standard system for defining the types and conditions of aluminum alloys. The International Standards Organization does have classifications for aluminum and its alloys, but most countries adhere to their own systems. The system followed in the United States of America is ANSI H35.1-1990, instituted by the American National Standards Institute and supported by the Aluminum Association and ASM International. The ANSI system and the US alloys

147

covered by it are described in reference,7 which is an excellent source of aluminum data; it includes a short list of alloys for other nations and their national designations. The ANSI system is used herein. In its entirety, it is a morass. Here, it is outlined just to the extent that is necessary to provide an uninformed reader with enough details to understand the nomenclature and the various processing treatments and the upper service temperature limits those treatments impose for maintaining stability of the processed materials. The system has two classifications, one for wrought alloys and one for cast alloys. Only the wrought alloy classification is described here. Briefly, it is an eight-character code consisting of two groups of four characters separated by a hyphen. The first four characters are all numerals and they identify the alloy group by chemical composition. There are eight aluminum alloy groups. The first digit of the first alloy group is 1, which represents alloys with a minimum of 99.00 wt% aluminum. In this group, the major foreign elements are Fe and Si, which are really residues from the aluminum extraction process and will be found to various degrees in all aluminum alloys. The next three digits in the group identify specific alloys in the same series, and the group as a whole is denoted the 1xxx series, often vocalized as the one-thousand series. The other seven alloy series are 2xxx (major alloying element, Cu), 3xxx (Mn), 4xxx (Si), 5xxx (Mg), 6xxx (Mg þ Si), 7xxx (Zn), and 8xxx (other element). An upper case X preceding the series identifier numeral indicates an experimental alloy. The second group of four characters in the eightcharacter designation represents the temper condition, that is, the heat treatment or degree of cold work. The first character of the four-character temper condition is an upper case letter representing a type of treatment. The other three characters are digits indicating variations within the treatment. There are many temper treatments. Only the three treatments most likely to be encountered in RR materials are described here. They are ‘O’ for the fully annealed condition, ‘H’ for a strain-hardened condition, and ‘T’ for a precipitation-hardened condition. The O condition is attained by annealing the material at about 340
C then slowly cooling it. There are no specified variations of the O condition. The H temper is more complex. The first digit after the H is a 1, 2, or 3. H1 signifies strain-hardened only. H2 is strainhardened and partially annealed. H3 is strain-hardened and stabilized by a low temperature heat treatment. The second digit, that is, the one following the H1,

148

Performance of Aluminum in Research Reactors

H2, or H3 designation is a number between 1 and 8 and is the degree of reduction in thickness or crosssectional area given to the alloy in its fully annealed condition to bring it to the desired strength level. Level 8 corresponds to a maximum reduction of about 75%. Level 1 represents approximately oneeighth of 75%, 2 is two-eighths, and so on. The third digit, if used, implies a variation of the two-digit temper. Partial annealing for the H2 condition is applied to products that are strained beyond the desired final amounts and are then brought back to the needed strength level by the partial anneal. Stabilization heat treatment for the H3 condition is applied to those products that, unless stabilized, would gradually age-soften at room temperature. Partial annealing also inhibits age softening. This tendency for softening of some cold-worked aluminum alloys at room temperature is important because such recovery requires the involvement of mobile lattice vacancies and/or self-interstitial atoms that promote climb and rearrangement of the cold work dislocations. It indicates the occurrence of atomic movement at room temperature, which, as we shall see shortly, is a factor affecting the development of radiation damage in aluminum. In addition to hardening by cold work, aluminum can be strengthened by solid solution treatment and by precipitation hardening. Only two alloying elements, Mg and Li, have sufficient solubility (several %) at room temperature to provide significant solid solution strengthening. Al–Li alloys are not recommended for reactor use because natural Li contains about 7.5% 6Li, which has a large crosssection for transmutation to 3H and 4He, both of which can be highly detrimental to aluminum. So the only solid solution-hardened alloys available for reactor use are the 5xxx (Al–Mg) series. Other metallic elements, principally Cu, Si, and Zn, have little or no solubility in aluminum at room temperature but are modestly soluble at higher temperatures near the melting point. This latitude permits considerable strengthening of such alloys by quenching-andaging, also known as precipitation hardening. The ANSI designations for the precipitation-hardened T conditions comprise ten subdivisions, T1–T10. For all T treatments, the alloy is heated to a temperature of 500–540
C to dissolve segregated alloying elements, followed by a rapid quench into cold water, which gives an unstable supersaturated solid solution. Precipitation is achieved by allowing the material to sit at room temperature for periods of weeks called ‘natural aging’ (tempers T1–T4) or by ‘artificial

aging’ at temperatures of 160–190
C for times of 6–24 h (tempers T5–T10). Flattening or straightening treatments may be applied before or after the aging treatment and are indicated by numbers in the third and fourth character positions. The temper conditions for aluminum alloys most often encountered in RRs are T4, T6, and T651. A T651 condition indicates a material that has been artificially aged then subjected to a light stretching operation insufficient to change its mechanical properties from those of the T6 condition. Of the precipitation-hardened alloys, the 6xxx series hardened by precipitates of Mg2Si is by far the most popular for RR service. The 6061 alloy in its T6 and T651 conditions has been approved for service as a class 1 nuclear components material under the Boiler and Pressure Vessel Code of the American Society of Mechanical Engineers, Case N-519.8 Two types of precipitation-hardenable wrought aluminum alloys, the 2xxx series (Al–Cu) and the 7xxx series (Al–Zn), both of which can be heat treated to greater strengths than the 6xxx alloys, are not usually found in nuclear reactors. Some 2xxx alloys are prone to aqueous pitting corrosion or may release Cu ions to the coolant that could be deleterious to other materials in the reactor such as stainless steel. The 7xxx series alloys have too low ductility and are the most difficult to weld. Their high zinc contents will cause high radioactivity. Unlike the cold-worked 1xxx alloys that can undergo recovery at room temperature, the precipitation-hardened alloys are thermally stable at temperatures up to about 150
C provided they have been given appropriate natural or artificial aging treatments. However, exposure to higher temperatures will cause overaging and associated reduction in mechanical strength. This softening is illustrated in Figure 1 for 6061-T6 alloy after heating to various temperatures for various times and testing at room temperature.9 It can be seen that softening is promoted by both time and temperature. For times up to 1 h, softening commences at about 200
C and is substantial but incomplete at about 370
C. For a longer exposure of 1000 h, the softening begins around the aging temperature, indicated by the down-pointing arrow, and is essentially complete at temperatures between 260 and 300
C. The data in Figure 1 are for specimens reheated without load. If reheating occurs under loads sufficient to induce creep and stress relaxation, the softening temperatures are pushed downward.

Performance of Aluminum in Research Reactors

149

350 Softening effects of reheating temperature and time on room temperature properties of 6061-T6Al (originally aged 18 h at 160 ⬚C)

300

200

1000 h

30 min

6 min

150 50 100 Elong., 1000 h

25

Elongation

50

Elong., 6 and 30 min

0

0

100

200 Reheat temperature (⬚C)

% Elongation

0.2% yield stress (MPa)

YS 250

0 400

300

Figure 1 Softening effects of reheating temperature and time on room temperature properties of 6061-T6 aluminum (originally aged 18 h at 160
C). Data from Structural Alloys Handbook, 1989 ed., Vol. 3, Battelle Memorial Institute, Columbus, OH, 1989; p. 14.

Table 2

Example alloys and their room temperature tensile properties

Alloy

Composition (wt%)

YS (MPa)

UTS (MPa)

Elongation (%)

1100-O 4032-T6 5052-H34 6061-T651

<1 (Fe þ Si) 12.2Si, 1Ni 2.5Mg, 0.25Cr 1Mg, 0.6Si, 0.28Cu, 0.2Cr

35 320 210 280

90 380 260 310

40 9 16 17

Source: Aluminum Standards and Data, 10th ed.; The Aluminum Association: Washington, DC, 1990.

Table 2 gives typical tensile properties of various Al alloys employed in RRs. The weak 1100-O alloy is simply annealed commercial purity aluminum with no deliberate alloy additions; it is hardenable to an H condition by cold work if so desired. The 4032 alloy is a eutectic composition of Si in aluminum that has been solution-treated and aged to create finely divided precipitates of Si; this alloy is used principally as a filler wire to improve the weldability of aluminum alloys. The 5052 alloy is a solid solution alloy of 2.5% Mg with a small amount of Cr added to control grain size and strengthen the grain boundaries. The particular 5052 alloy in the table has been work hardened to a 4/8, or half-hard, condition before stabilization. The 6061-T651 alloy has been solution treated and artificially hardened by precipitates of Mg2Si phase and its precursors, then given a mild stretching treatment.

5.07.4 Fuel Elements The most crucial and demanding applications of aluminum in RRs are in the fuel elements. There it is used inside the fuel element as a thermal conduction matrix in which a dispersion of fuel particles is embedded and as a cladding material that protects the fuel from corrosive attack by the cooling water, retains fission products, and transfers heat from the fuel and matrix to the coolant. As RRs matured, considerable improvements were made in the fuel elements. Most fuel in RRs is no longer unalloyed, metallic a-phase uranium whose orthorhombic crystal structure is prone to severe radiation growth and swelling leading to distortion and cracking. It has been ousted by more stable and more isotropic uranium compounds that can also better accommodate fission gases with minimum swelling. The most

150

Performance of Aluminum in Research Reactors

common ones are U3O8, UAlx, where x can be 2, 3, or 4 but is usually considered10 to be a mixture of 3 and 4, and U3Si2. There is also a hydride fuel, U–ZrH1.6, which is used exclusively in the open-pool TRIGA (test, research, isotopes, General Atomic) types where the fuel is in the form of slugs comprised of particles of U dispersed in the ZrH1.6 phase (see Chapter 3.12, Uranium–Zirconium Hydride Fuel). Originally, the TRIGA slugs were sheathed in aluminum, which has now been replaced with stainless steel or nickel alloy. However, TRIGA reactors still contain other aluminum components. There is no outstandingly superior aluminum cladding alloy. The most common aluminum cladding alloys are 1100 and the stronger 6061. Other alloys have been investigated in neutron irradiations,11 namely 5052; X800N, where N is 1, 2, or 3 and whose compositions are Al–1Ni–1Fe; and two sintered aluminum powder alloys, M257 and M470, which were fabricated by ball milling flake powder of 1100Al in air until it contained a dispersion of 6% and 10% Al2O3, respectively, then consolidating by pressing, sintering, and hot rolling. The Mxxx alloys were deemed to be no better than 1100 and 6061 types. They are more difficult and expensive to make and harder to weld than regular melted-type alloys. In Europe, particularly in France, two preferred alloys are AlFeNi, a relative of X8001 with the composition 1Fe–1Ni–1Mg, and AG3-NET, a 5xxx-type with 2.5–3.0Mg and low residuals. The greatest concern for cladding is its corrosion behavior (see Section 5.07.5). A feature of RR fuels is that they are much more highly enriched in 235U than those in power reactors: 12–93% versus about 2.5%. Drivers for raising the 235 U levels were extended fuel cycles; the growing demands for industrial and medical isotopes, particularly 99Mo the parent of the all-important medical diagnostics tool 99mTc; and the need for higher neutron fluxes for increased production of the heavy, transuranic isotopes. The use of highly enriched uranium (HEU) meant higher heat generation and required improved means of removing the heat. The solution was the development of dispersion fuels in which particles of the enriched fuel were distributed in a matrix of thermal conductor material, all compressed together in sealed aluminum cans. The thermal conductor is aluminum powder, usually a 1xxx-type, often atomized powder of better than 99.5% purity and particle size <100 mesh (150 mm maximum, 23–48 mm mean). Atomized powder particles are denser, pour more easily than milled flake

powders, and have less low conductivity surface oxide per unit volume. The aluminum matrix may occupy more than 50 vol% of the fuel/aluminum mixture. A huge advance in fuel element morphology and heat removal efficiency took place when Eugene Wigner designed his thin, curved fuel plates for the high flux Materials Testing Reactor (MTR) built at Arco, Idaho. A thin plate has a number of advantages over cylindrical slugs. The rolling treatment used to produce the plates from a fuel slab, or from a dispersion of fuel particles in aluminum matrix powder, sandwiched between two aluminum cladding sheets gives superior mutual contact of cladding, matrix, and fuel for improved heat transfer paths to the cladding. The much larger surface-to-volume ratio of plates provides more efficient heat transfer to the coolant, thus permitting higher fuel loadings per unit volume. The benefit of a curved fuel plate is that any buckling and bowing in the plate due to irradiation will be focused in the direction of the radius of curvature. Thus, in a fuel element comprised of a stack of curved plates restrained at their edges and separated from each other by cooling channels of the same width as the thickness of the plates, any such distortions will be accommodated cooperatively in the radial direction without unacceptable narrowing of the cooling channels. An MTR fuel element contained 18 plates each about 72 mm wide and about 727 mm long bent to a curvature of 140 mm radius in the width direction. The plate thickness was 1.27 mm including a minimum cladding thickness of 0.25 mm on each face. The plate edges were brazed into sturdy side panels to seal the plate edges and impart rigidity to the assembly. The water gap was 1.27 mm. The cladding and side panels were made from 1100Al; the Al brazing alloy contained about 13% Si.12 This assembly was then enclosed in a long, rectangular aluminum box fitted with end fixtures for remote handling. The end fixtures were castings of Al–7% Si. The reactor core was built from groups of such elements assembled upright in rectangular arrays held together by aluminum grid plates. Refueling was done from the top, and any element could be replaced by a box of the same size containing a reactor experiment or materials for isotope production, or a beryllium reflector or a control rod. These MTR-type boxed fuel elements in open grid core arrangements performed very well and became very common for RRs. To satisfy demands for higher power densities and more sophisticated tailoring of local neutron fluxes, the next advancement in aluminum–clad fuel elements was the development of upright, annular

Performance of Aluminum in Research Reactors

elements using curved fuel plates in which the fuel particles may be required to be graded in concentration across the thickness and width. Beryllium reflectors surrounding the annulus direct neutrons from the fuel back to the hollow center, or ‘trap,’ of the annulus where reactor experiments and isotope targets are placed. The Be also creates additional neutrons from (n, 2n) reactions. Vertical holes bored through the reflector allow passage of cooling water and house irradiation experiments. Two highperformance beryllium-reflected reactors using annular fuel elements are the High Flux Isotope Reactor (HFIR) at Oak Ridge National Laboratory (ORNL), rated at 100 MW thermal and currently running at 85 MW, and the Advanced Test Reactor (ATR) at Idaho National Laboratory, rated at 250 MW but lately operating at 100–125 MW. The cores of these reactors are of uncommon designs and deserve comment. The ATR core13 is 1.22 m diameter and 1.22 m high. It contains a continuous serpentine-like wall of fuel elements looped around nine flux traps each about 120 mm diameter arrayed in a square 3  3 grid. In plan view, the wall forms the shape of a four-leaf clover. It fully embraces the central flux trap and the four corner ones. The other four traps lie just outside the wall; each is tucked in between the junctions of two leaves and is about half wrapped by the wall. At each corner lobe, there are four shim control cylinders just outside the wall and six shim rods at the neck of the wall inside the cloverleaf. These controls allow each of the four lobes to be run at different power levels simultaneously, as needed by the experiments in the traps. The remainder of the space in the core is occupied by blocks of Be reflector containing numerous experiment holes. The wall is built14,15 from 40 individual wedgeshaped fuel elements, each containing 19 curved fuel plates. The cross-sectional area of an element is a 45
sector of a circular annulus. Its outer arc, plate #19, has a radius of 137 mm and an arc length of 100.9 mm. Its inner arc, plate #1, has a radius of 77 mm and an arc length of 54.1 mm. The 19 fuel plates are attached by roll-swaging to 6061-T6Al side panels 64.6 mm wide  1257 mm long. Within the elements, the curved plates are concentric with the circumferences of the traps. The plates are 1.27 mm thick except for #1 and #19, which are thicker. The water gap is 1.98 mm. The ATR fuel is UAlx enriched with 235U to 93%, dispersed in a matrix of Al powder and clad with 0.38 mm thick 6061-OAl. The HFIR core16 is more compact, about the size of a small trash can, into which are packed 540 fuel

151

plates in quite a different arrangement than in the ATR. The core diameter is 435 mm and it is 791 mm tall. It has a single central flux trap, 129 mm diameter. The fuel is granules of U3O8 enriched with 235U to 93% and embedded in Al powder. The cladding is 6061Al. The core consists of two concentric annular arrays of involute-curved fuel plates, as shown in the sketch of a radial segment in Figure 2. The black region in the fuel plates is the fuel dispersed in its Al matrix; the white area is Al filler. There are 369 plates in the outer annulus and 171 in the inner annulus. The plates are 610 mm high with widths for the inner and outer annulus plates of 94 and 81 mm, respectively, before bending. The plate thickness and coolant gaps are 1.27 mm, as in the MTR-type elements. The two annuli are fabricated separately and are united when loaded into the reactor. In addition to the unique radial-like orientation of the fuel plates, the fuel particles are uniquely distributed in the plates. To minimize the radial peak-to-average power density ratio, the thickness of the compacted fuel mix is varied along the arc of the involute curve as seen in Figure 2. This shaped region is backed by filler Al containing no fuel particles. For the inner annulus, the filler powder backing the shaped fuel region contains

Al filler Al + 41 wt% U3O8 1.27 mm Coolant channel

Outer annulus sidewalls

Al + 30 wt% U3O8

Al filler + B4C poison

1.27 mm

Figure 2 Horizontal section through a small segment of the HFIR core showing fuel plate curvatures and fuel distributions in the plates. Modified from Binford, F.T.; Cramer, E. N. The High Flux Isotope Reactor; A Functional Description, Vol 1B, Illustrations; ORNL-3572 (Rev.2); Oak Ridge National Laboratory: Oak Ridge, TN, 1968.

152

Performance of Aluminum in Research Reactors

particles of B4C burnable poison. Two concentric cylindrical control plates clad in Al are located immediately surrounding the core. Outside the control plates are four concentric cylindrical Be reflectors. Because beryllium generates copious quantities of helium and tritium from neutron irradiation, it tends to swell and crack, particularly at the faces of its high neutron flux regions. To retain chips spalled from these surfaces, the reflector and any penetrations in it are clad with aluminum. Four horizontal 6061Al beam tubes and numerous vertical holes penetrate the reflector. Like most dispersion-type fuel plates, the HFIR and ATR plates are fabricated by what is called a picture frame technique. This utilizes powder metallurgy methods to disperse the fuel particles uniformly in the Al matrix and press the mixture into a hard rectangular compact. The rigid compact is placed in a window of the same size cut in an Al slab or frame, which is usually the same alloy as the cladding. Sheets of cladding material are welded to the top and bottom faces of the filled frame and the assembly is hot rolled through a large reduction in thickness to ensure that the cladding is fully bonded to the fuel charge and the frame. After verifying the location of the fuel charge, the rolled plate is cold rolled to flatten it and bring it to the specification thickness. It is then given a final anneal at 500
C to reveal any blisters and rolling defects in the cladding surfaces. After verifying the location of the fuel region, the plates are blanked to finished size in a press. Of course, it is not as simple as that. Strict quality assurance standards have to be met, and at every stage in the operation, there are numerous inspections and rigorous sizing and confirmation tests. To reproducibly obtain the graded fuel distributions in the HIFR plates, a special procedure was developed.17,18 A custom-designed contoured auxiliary die plate is mounted over the cavity of the powder press to facilitate mounding of the fuel/matrix powder mix in a semicylindrical hump. Another auxiliary die plate is added to allow filler powder to be leveled on top of the humped fuel charge. This duplex charge is withdrawn into the press cavity, the auxiliary dies are removed, the rectangular punch is inserted into the die mouth and pressure applied, and the charge is consolidated in a single cold pressing operation. The HFIR fuel plates are bent to the desired involute shape in an elastomer-faced punch and die press. They are welded into the cylindrical inner and outer sidewalls of the fuel elements. The sidewalls are

machined from extruded-type 6061 aluminum tubing in the T6511 temper. Twenty-seven equally spaced circumferential weld grooves are turned on one face of each sidewall, and slots are milled at prescribed depths and angles on the other face of the wall. The weld grooves intrude a short way into the slots. The fuel plates are slid into the slots and properly spaced with the aid of temporary Teflon separators. The plates are machine welded in place through the grooves. A 4043Al weld filler wire and an argon shield gas are used. End fixtures machined from 6061Al tubing are welded to the ends of the elements, and final machining and inspection are conducted. These multiplate fuel elements are a testimonial to designer ingenuity and superb fabrication skills, and the versatility of aluminum. Manufacturing these fuel elements is not only painstaking but also expensive. In year 2007, each HFIR element cost $1 M.19 It is replaced after its regular lifetime operating cycle of 26 days. With so much effort and cost invested in it, a rejected element is a severe financial loss. The specifications and acceptance standards are so high that the chances of producing a fuel element completely free of specification violations are very low. The first 30 000 fuel plates suffered a rejection rate of 12%, and of the first 45 fuel assemblies, only 4 inner elements passed the final inspection.20 However, the degrees of severity of the violations were all minor or were correctable. With waivers, all 45 elements were accepted and gave exemplary service. After operation of the first 60 fuel cores at the full design power level of 100 MW, 4 of them were autopsied.21 No significant faults were found. The in-reactor performance of these complex ‘aluminumbased’ fuel elements has been incredible, surpassing all expectations. Development of RR fuels and fuel plates is continuing. Concerns over the possibilities of nuclear weapons proliferation and terrorism led to establishment of the Reduced Enrichment for Research and Test Reactor (RERTR) program at Argonne National Laboratory.22 The goal of RERTR is to eliminate the use of highly enriched uranium (HEU) in RRs by converting to the use of low enriched uranium (LEU). HEU is defined as uranium that has the fraction of the fissile isotope 235U greater than 20%, LEU is less than 20%. Historically, RRs have used enrichment levels of 235U up to 93%. RERTR is intended to be achieved without impairing the safety and performance of the reactors and/or jeopardizing the production of important isotopes, and at minimum cost for changes in fuel elements.

Performance of Aluminum in Research Reactors

In some RRs with modest uranium enrichment and low power levels the RERTR LEU goal was met by diluting the fuel with natural uranium. For many of the high performance RRs (HPRRs) that must retain their 235U levels and cannot tolerate the burden of added 238U without excessive operational penalty, the RERTR dilution can be achieved by replacing the HEU fuel with LEU compounds or alloys containing higher fractions of U. To that end, the initial focus of RERTR was on the development of uranium silicide fuels, U3Si and U3Si2, dispersed in aluminum and clad with aluminum.23,24 While this move has been successful for many RRs it is not sufficient for the most demanding HPRRs. For them, attention has turned away from dispersion fuels to monolithic alloy fuels where higher U densities are attainable. The goal is to develop fuel plates built from foils of LEU alloy, 250–500 mm thick, clad with aluminum.19,25–27 In order to prevent buckling and cracking of the foil during multiple rolling and recrystallization treatments and to inhibit radiation growth and warping, there must be just enough alloying metal in the U to stabilize it in its isotropic g-phase. Several alloying metals are suitable, but the field of contenders has been reduced to the U–Mo system. A 90% LEU-10% Mo alloy currently holds the best prospects. Some serious hurdles are recognized. Interdiffusion between the cladding and the fuel foil during annealing and in-reactor exposure encourages the formation of reaction layers of uranium–aluminum compound(s) with low thermal conductivity and low resistance to growth of fission gas bubbles. Such layers threaten the integrity of the fuel/cladding interface. Development of these layers is retarded by additions of Zr or Ti to the fuel, or Si in the cladding. When Si is incorporated in the cladding, it is found to segregate at the fuel/cladding interface, acting like a diffusion barrier. Thin film diffusion barriers of Si, Zr, and ZrN applied directly to the surfaces of the fuel foil by co-rolling and thermal spraying have done well in reactor tests. The current hot roll bonding processes used for attaching cladding to dispersion fuel plates may not be fully adaptable to barrier-coated foil fuels. Other bonding methods such as hot isostatic pressing are under investigation. For HFIR plates, where the foils must be tapered in both width and length and have involute shapes, fitting and bonding diffusion films and cladding to the fuel foil on a mass production scale is a challenge. Hot roll bonding will not work because the foil and the cladding will not deform to the same extent and will result in nonuniformly

153

thick cladding, and shear deformation during rolling may damage the diffusion barrier. It is recommended19 that the tapered foil, bent to its involute shape and with an adherent diffusion barrier, should be prepared separately then sandwiched in shaped recesses in two full-length clamshells of cladding of appropriate thickness and bonded over all mating surfaces. Alternatively, if the clamshells can be made from a two-ply Al sheet, like the commercial OneSide Alclad™, the inner layer of, say, 1100Al, could contain the ingredients for a diffusion barrier. The hot isostatic pressing route may then allow bonding and barrier filming in a single operation and in batch mode. If burnable poison cannot be incorporated in the fuel foil, it may be possible to accommodate it in the inner cladding layer with the diffusion barrier components. The corrosion behavior of the Al cladding on alloy foil fuel elements will need to be explored thoroughly. A penetration of the cladding will probably be more serious than one in current dispersion fuel plates because the alloy fuel will likely be more reactive and soluble in water than the dispersant-type intermetallic and refractory fuels.

5.07.5 Corrosion Metallic corrosion, the removal of metal atoms from the metal surface by the electrochemical action of the environment, has many forms: uniform, galvanic, pitting, grain boundary, crevice, etc. Uniform corrosion and pitting are the types of most interest to RRs. The greatest worry is the aluminum fuel cladding where the environmental conditions are most aggressive and where an unexpectedly high corrosion rate might breach the cladding and allow release of highly radioactive fission products throughout the water system. Pitting corrosion is the major form of attack on the cladding of spent fuel elements during longterm storage in water basins.28 Herein, the focus is on uniform corrosion of cladding. Aluminum is a very reactive metal. In dry air, it combines with oxygen to form an adhesive, self-healing Al2O3 film that retards further oxidation at the metal surface. Such films are usually quite thin, tens of nanometers, usually described as amorphous. Films formed in moist air and water are much thicker, 1 mm or more. The water-formed reaction films developed on aluminum cladding are variously described as ‘hydrated oxides’ and ‘hydroxides,’ and ‘oxidehydrates,’ and they are generically referred to as

154

Performance of Aluminum in Research Reactors

‘oxide films.’ In HPRRs, the films grown on the fuel cladding may be 20–50 mm thick.29,30 The most common corrosion products28,30 reported on aluminum cladding are boehmite, a crystalline monohydrated aluminum oxide, Al2O3H2O, and bayerite, a crystalline trihydrated oxide, Al2O33H2O. At temperatures below about 77
C, the boehmite phase is formed preferentially but may transform to bayerite with continued immersion. At temperatures above 77
C and below 100
C, a pseudoboehmite structure grows, which may age to other hydrated oxide forms or retain its pseudoboehmite structure. Between 100 and 400
C, crystalline boehmite will form. A gelatinous boehmite is the chemical precursor of both of the crystalline hydroxides.30 The mature hydroxides are normally white color but other hues have been reported and may stem from absorption of Fe, Cr, Ni, or other metal ions leached from steels in the reactors or in the corrosion test loops. The corrosion film is both the reaction product and the medium through which the corrosion process occurs. Whether corrosion is governed by ingress of O and OH ions through the film to the metal surface or by egress of Al ions to the film/water interface, it is expected to be diffusion controlled. Thus, all else being equal, an increase in film thickness should lower the corrosion rate by increasing the diffusion length, and vice versa. Therefore, the corrosion rate should be parabolic with time and have an Arrheniustype dependence on temperature. Moreover, ideally, if all the corroded metal was retained in the corrosion film, if the chemical composition and physical structure of the film were constant throughout the thickness, and if all of the film was retained on the metal, the film thickness would be proportional to the amount of metal corroded. Alas, such ideality does not prevail. The corrosion process is confounded by a number of interacting factors, including the following: there is a one-sided heat flux on the cladding; the corrosion film is a thermal insulator compared with the Al cladding, so the temperature of the film will increase with thickness; the film may not be of uniform composition and/or structure; the film is soluble to some extent in water, and its solubility is strongly susceptible to the pH of the water, which is related to water composition; the film is subject to erosion in flowing water and to spontaneous spallation above some uncertain thickness, about 50 mm in one case.31 And to further complicate the situation, there is wide variation in the ways the corrosion tests are conducted and in the parameters that are measured.

The tests may be carried out in open cups, closed autoclaves, vented autoclaves, closed loops, bypass loops, or on used fuel plates. Evaporation or consumption of the water may require that it will need to be periodically replaced or its volume readjusted. Except in in-reactor tests and loop test systems with bypass monitoring and adjustment of the water, the chemistry of the water may change substantially during the test. Few corrosion rates for cladding materials are measured directly. They are usually derived from measurements of the thickness of the corrosion film. A thickness measurement gives the thickness of the film adhering to the substrate at the time of the measurement. It will not include film that has been dissolved and/or eroded away. On a spent fuel element, it may include film that has formed in a storage pool over time periods much longer than it experienced in-reactor, and with no forced cooling. During preparation for post irradiation examination (PIE) in a hot cell, the spent element is no longer fully immersed. It gets hot and has to be periodically sprayed with water to cool it. It has been opined21 that the resultant steaming and thermal cycling may cause more corrosion than in-reactor operation and underwater storage. There is no guarantee that the density and the composition of the film will be invariant through the film thickness. On the contrary, multilayer films are more common than not. Almost all films have a thin, monolithic base in contact with the Al surface, presumably associated with the ubiquitous air-formed Al2O3 film. On top of this base, there may be one to three distinct layers. Some films contain pores or are cracked. Only the films on irradiated fuel elements have been exposed to the effects of neutron irradiation and radiolysis of the water. The way in which the film thickness is measured may be questionable, too. At least six different methods are used, viz.: (1) Scaled measurements by optical or scanning electron microscopy of metallographically polished and etched cross-sections of the corroded test piece; (2) micrometer measurements of the thickness of the test piece before corrosion and after the corrosion product is removed by electrolytic polishing until the shiny metal is seen; (3) weight gains of coupons with film in place; (4) weight losses of coupons after removal of the film; (5) acoustic and eddy current measurements with instruments calibrated against accepted standard films; and (6) temperature increases measured with thermocouples attached to the noncorroding back surface of the test piece during the test, and related to spot film thicknesses measured metallographically after the test.

Performance of Aluminum in Research Reactors

A neglected aspect of film measurements is that almost all of the measurements have been made on specimens that, deliberately or unavoidably, were dried at room temperature or at 100
C32 before the measurement was attempted, or before the measuring instrument was calibrated. Until recently, nobody seems to have determined whether such drying treatments will shrink, spall, crack, or otherwise alter the bulk film. The gelatinous surface layer that precedes the crystalline corrosion films will almost certainly be altered during dehydration. It is not uncommon for test coupons to be dried, weighed, and placed back in the test for the next exposure period, and so on until the termination of the campaign. That was the method used in one seminal laboratory test study.33 The first periods in the full exposure sequence were the shortest ones, 1 or 2 days, and they always showed the largest weight gains, usually 60–90% of the total weight gained during the full duration of the test, which was about 22 days. Weight gains after the first period were linear with time and were relatively minor. That is not parabolic corrosion behavior. The abrupt change in weight gain indicates that something happened during the first interruption of the test that set the scene for a sudden switch from an initial rapid corrosion rate to a subsequent constant low rate. Likely, the first drying treatment irreversibly altered the structure and permeability of the hydrated film. Recent autoclave tests34 on AlFeNi alloy reinforce that suspicion. It was demonstrated that during a 34-day test, interruptions made every 7 days to remove, dry, weigh, descale, dry, reweigh, and replace the test piece in the autoclave with refreshed water for the next exposure period had serious consequences to the corrosion kinetics. Without interruptions, the inner and outer oxide layers were twice as thick, the weight gain was 26% higher, and the amount of metal removed from the substrate was 23% higher. Some efforts have been made to correlate film thicknesses with corrosion rates.31–33 Tests made under controlled conditions in a corrosion loop31 found that the thickness of the boehmite film on 1100, 6061, and X8001 alloys was about 1.4 times the depth of penetration into the aluminum regardless of changes in test parameters that changed the film thickness, as long as there was no stripping or spallation of the film. Using a literature value for the density of boehmite, it was estimated that about 70% of the corroded Al remained in the adherent film and about 30% was lost to the coolant. When spallation did occur, which was usually above a film thickness of

155

50 mm, the 1100 and 6061 alloys always showed localized attack of the aluminum under the spalled area, whereas the X8001 alloy showed only uniform attack under all conditions. This correlation was for a closed, single set of data. It should not be considered representative of all data and situations. Other data by some of the same authors,32 where the principal variables were temperature and flow rate, showed that the ratio of corrosion product retained to the weight of metal corroded ranged from a high of 0.54 at a low temperature of 170
C and flow rate of 6.1–9.5 m s1 to a low of 0.08 at 290
C and 29 at 32.6 m s1. Another source29 quotes a retention level of 50–80% of the oxide on the cladding surface, but it may be citing Griess et al.31 In general, the relationship between film thickness and corrosion rate is not well established. Film thicknesses from laboratory tests31,35–38 display power law growth with exposure time, but the time exponents, preexponential factors, and activation energies differ from one experimenter to another and may be applicable only to the particular set of data from which they were determined. Nevertheless, the laboratory tests have established that the corrosion films are sensitive to a number of interacting factors. They include the temperature and surface condition of the cladding; the heat flux density on the cladding; and the temperature, pH, flow rate, and purity of the water. In RRs, water purity is controlled by filtration and ion exchange systems; it is also linked to pH. With regard to pH, the films will dissolve if the water is strongly acidic (pH < 4.5) or strongly alkaline (pH > 8.5); films are most stable in the range 5.0–6.5, the closer to 5.0 the better. The pH of reactor water and spent fuel storage pool water tends to converge toward the desired range by carbonic and nitric acids formed from CO2 and N absorbed from air. It can be maintained close to 5.0 by controlled additions of nitric acid. The strongest increase in film growth is from increase in temperature, and the controlling temperature is that at the hydroxide/water interface.31 To lesser extents, increased heat flux density and water flow rate will raise the film growth rate. For the alloys 1100, 6061, and X8001, which all corroded alike until spallation occurred,31 the rate of oxide formation at a heat flux of 1.58 MW m2 was about half of that at 3.13–6.31 MW m2, other conditions being the same. At coolant flow rates in the range 7.6–13.7 m s1, the rate of accumulation of the corrosion product was the same for all three alloys. Corrosion rates measured on the insides of 1100Al production tubes39 were

156

Performance of Aluminum in Research Reactors

found to be unchanged by water velocities in the range 0.305–5.58 m s1. Reduced water temperatures will reduce the film growth rate. Despite differences in strengths, compositions, and microstructures, the alloys 1100, 6061, and X8001 all seem to have similar corrosion behavior under similar conditions.31,32 Spalling tends to introduce local attack in 1100 and 6061 but not in X8001.31 The AlFeNi alloy shows good performance at temperatures up to 250
C in autoclave tests35 and in-reactor exposures40 at temperatures below 120
C, but it has not been tested under high heat fluxes. AG3-NET cladding on U3Si2 dispersion fuel plates undergoing in-reactor tests failed41 at a heat flux of 5.5 MW m2. The cladding was swollen and breeched by a combination of a very thick corrosion film and subfilm intergranular corrosion. Cross-section X-ray spectroscopy analyses showed that oxygen had penetrated intergranularly all the way through the cladding to the meat. The corroded cladding was interesting in other ways. The outer oxide layer was monolithic and was exceptionally thick, 100 mm. Directly beneath it was a region about 80 mm thick containing many round 30 mm size pores. Below the porous region, the grain boundaries were enriched in Mg and oxygen. The plates were intended to reach a temperature of about 180–200
C at the exterior surface of the cladding and 220–240
C in the fuel. Temperatures estimated from the thick corrosion layers were >300
C for the water/corrosion film interface and >400
C for the fuel meat. The AG3NET alloy has a history of intergranular cracking in beam tubes and other structures in the Reacteur Haut Flux at the Institut Laue-Langevin in France. Although that cracking occurred at high fluences, the irradiation temperatures were low. Such low temperature intergranular cracking is a sign of pending weakness in the alloy and does not bode well for applications at higher temperatures as in fuel cladding. The influences of neutron flux and radiolysis of water are unclear. These parameters are omnipresent in RRs and we might imagine them to strongly affect aqueous corrosion of fuel cladding by damaging the cladding and its corrosion film and by altering the activity of the water. One researcher42 writes that reports of neutron flux effects on the hydroxide films are few and there is disagreement; he claims that the opinion of most (Russian) researchers is that neutron irradiation decreases, rather than increases, the corrosion rate. Effects of radiolysis are uncertain. According to Golosov,42 one Russian authority argues

that radiolysis may either accelerate corrosion by facilitating cathodic processes or reduce corrosion by promoting anodic passivation. Data from laboratory corrosion loop tests without radiation fields seem to be fairly compatible with data from irradiated fuel elements in terms of oxide thicknesses, compositions, and pH effects. There are no outlandish differences that would immediately draw attention to radiation effects. At least, none that has been strong enough to insist that loop tests should be repeated in irradiation fields. A similar conclusion was reached for aqueous corrosion of aluminum process tubes in production reactors.39 Therefore, irradiation effects must be modest at worst. However, there are some troubling reports that seem to indicate large effects of irradiation fields in nonreactor conditions. Sindelar et al.43 studied 6061Al coupons exposed to moist air at 150 and 200
C, with and without exposure to a 60Co g source at 1.8  106 R h1. Weight gains and film thicknesses were measured. The corrosion product was patches of loosely aggregated, randomly oriented 1 mm size boehmite crystals sitting on a thin monolithic base layer, even at 100% relative humidity where the product was permanently under a film of water. g-Irradiation seemed to double the weight gains and increase the film thicknesses by a factor of 10. There was substantial surface blistering of the base layer, attributed to hydrogen gas. The paper provided no details of the experimental conditions. Enquiries to the authors produced a lengthier publication44 with the missing details. Those details cast grave doubt on the conclusions drawn from Sindelar et al.43 In particular, the experiments with the g-field were made under radically different conditions than those without the field. Specimens for the g-irradiations were sealed in small stainless steel cans of just 78 ml and each can represented an uninterrupted test for a given exposure period of 1, 4, 8, and 12 weeks. The tests without the g-field were made in stainless steel autoclaves of volume 37 850 ml for 15 unequal exposure periods totaling about 30 weeks. At the end of each period, the specimens were removed, dried, weighed, and replaced in the autoclave with a new charge of water. In light of the effects of interruptions described in Wintergerst et al.,34 the effects of g-irradiation described in Sindelar et al.43 and Lam et al.44 are inconclusive. In the other work,45 Al coupons of undeclared composition and condition were exposed to static brackish water of pH 8–9, at undisclosed temperature for periods of up to 30 days, with and without low dose irradiations with neutrons from a 252Cf source

Performance of Aluminum in Research Reactors

(1010 n m2 s1) and, separately, g-rays from a 60Co source at 15 Sv h1 (1.6  104 R h1). Corrosion was determined from weight losses. It was not stated whether the specimens were recycled from one period to the next. The neutrons and the g-rays had the same effects and to the same degree; they promoted formation of a grayish layer on the specimen surfaces; they reduced the weight losses by 25–30%; and they almost eliminated severe pitting corrosion displayed by the unirradiated specimens. None of these three reports mentioned whether radiation heating was a factor. The laboratory loop tests have verified the expectation that the corrosion film is a thermal insulator compared with the Al cladding, and they have provided31 a thermal conductivity value of 2.25 W m1 K for boehmite, which is a factor of 70–100 less than Al. However, it is not always ascertained whether a particular film is boehmite or bayerite or a mix of both. No thermal conductivity value is available for bayerite. When insulating films build on the Al cladding of heat sources like the fuel and long-term heavy isotope targets, the temperatures of the sources and their claddings or containers will rise. This temperature rise will increase the corrosion rate and the growth rates and dissolution rates of the corrosion films. In HPRRs, a side effect of an increase in cladding temperature by the adherent corrosion product is the threat of plate buckling.31 As described earlier, the strengths of the cladding and Al fuel matrix can be decreased significantly by tens of degrees increases in temperature, and creep rates will increase. If an insulating corrosion film increases the temperature gradients between the center thickness of the fuel plate and the surface of the film, and between the fuel-loaded portions of the fuel plates and their cooler frames, the plates may distort. If the distortion is not in phase from one plate to the next, it might perturb the coolant flow and accelerate the temperature changes. Griess et al.31 envisaged that the insulation provided by the corrosion-product film might be more of a limitation on the use of aluminum–clad fuel elements in high flux reactors than is corrosion damage per se and, in the worst case, may lead to burnout of parts of the fuel plates. Fortunately, that prophecy has not been fulfilled. Serious plate distortion has not been a widespread issue. One case of plate distortion is described in Shaber and Hofman.30 Plate buckling found in some MTR elements12 was blamed on new design changes. It is recommended30 that new fuel elements should be prefilmed with a hydroxide film to reduce the rate of in-reactor buildup of the corrosion layer. Tests32 with 1100, 5154, 6061, and X8001 alloys at flow

157

rates in the range 6.1–20.4 m s1 found that preexposure of the test pieces to water at 250–300
C for 24 h in an autoclave caused a significant improvement in corrosion resistance, but not at higher flow rates. The ATR elements are pretreated30 by immersing them in water for 48 h at 180
C and pH 5.0. In the early days of HFIR operation, the new fuel elements were often stored in the reactor pool water for up to 3 months before being placed into service. This immersion resulted in the formation of a rather thick, gelatinous, corrosion product film on the element surfaces.21 In an attempt to avoid that condition, some of the elements were pretreated by boiling them in deionized water for 24 h to produce a thin, boehmite film on the surfaces of the elements before they were placed into service. When the pretreated elements were used, the coolant flow rate was found to gradually decrease and the pressure drop across the elements gradually increased during the reactor fuel cycle. No significant damage was caused. Changes in coolant flow rate and pressure drop were not observed when the reactor was operated with non-pretreated fuel elements. Metallographic examinations of cross-sections of the spent fuel plates revealed much thicker corrosion films on the pretreated plates. Pretreatment of the HFIR fuel elements was discontinued. Most RRs do not practice pretreatment of their fuel elements. It is proposed here that because of the seemingly large effects of dehydration on retarding subsequent film growth as discussed earlier, at least one in-reactor trial should be made of a prefilmed fuel plate with a dehydration step or a low temperature baking treatment added. A drying treatment might also be worthwhile for a newly spent fuel element before it enters pond storage. What we really need to learn from corrosion measurements and film thickness data is the thickness of uncorroded Al cladding remaining on the fuel element at the end of reactor service, and whether that thickness will be sufficient to continue to seal the spent fuel through further corrosion expected during cool-down storage in water basins. That is, we need reliable corrosion rates pertinent to the particular application. Corrosion product thickness data are invaluable in identifying and characterizing the major factors governing corrosion and the interplay between them, but they are meaningless to corrosion rates if a reproducible relationship between film thickness and corrosion rate is not established. We need predictability. To that end, efforts are underway to derive predictive models for film thicknesses40,46 and corrosion rates.42 These models are in their

158

Performance of Aluminum in Research Reactors

infancy. Because they lack a large body of consistent data to draw on, the authors must make many assumptions, fittings, and correlations to derive constants, correction factors, adjustment factors, and augmentation factors. With so much flexibility built into the prediction equations, it is not surprising that the authors can find good correlations with selected data from measurements made on spent fuel cladding. This is not intended as a criticism of the modelers; it is a reflection of the paucity of input data. Reliable modeling is essential. But it needs reliable input data. Data obtained from recycled test coupons should either be excluded from the models or modeled as a separate category. To be generically applicable, film thickness models and corrosion rate models should attempt to merge in a complementary manner. In low power RRs where convective flow is sufficient to take care of cooling and water quality is adequately controlled, problems from corrosion films formed on the aluminum cladding and on other aluminum components elsewhere in the reactor are uncommon. In HPRRs, the most prominent corrosion problems were those in the early days of operation that caused a milky turbidity of the coolant and a white deposit and increase in surface radioactivity on all surfaces exposed to the coolant. The turbidity was identified as a fine suspension of boehmite, and the g-radioactivity was consistent with decay of 24Na, both effects attributable to corrosion/erosion of the fuel cladding. The turbidity is created by increase in the cladding temperature due to the warming effects of the hydroxide film. In turn, the temperature of the coolant in immediate contact with the film is raised. This increases the solubility of aluminum oxide in the immediate volume of coolant. When this small volume moves on and merges with the cooler bulk coolant, the solubility falls and much of the dissolved film is released as a particulate suspension. Particles of film washed directly into the coolant by erosion of the cladding due to the high coolant flow rate contribute to the turbidity. Since turbidity ensues when the concentration of aluminum in the bulk water exceeds the solubility of the aluminum oxide, turbidity problems are brought under control by tuning demineralization treatments to remove dissolved aluminum from the bulk water and by reducing the degree of dissolution through adjustments in pH to between 5.1 and 5.4 where aluminum oxides have minimum solubility. In-reactor pitting corrosion and galvanic corrosion have not been serious problems. Pitting of Al, which is encouraged by the presence of ions of Cu, halides, and bicarbonates, is more serious in storage

pools where poorer water chemistry and nearly stagnant water conditions may exist, but diligent monitoring and control of water chemistry can mitigate these concerns. Intergranular corrosion has not been a problem in RRs, but it could become an issue at high irradiation temperatures as evidenced by the AG3-NET cladding described earlier. Overall, aluminum cladding has given very good service in water-cooled RRs and continues to do so. The major variables influencing the corrosion process (es) and corrosion products are fairly well identified except for effects of irradiation. More data from spent fuel elements are needed to guide and refine models for predicting film thicknesses and corrosion rates.

5.07.6 Radiation Effects 5.07.6.1

Basics

As in other metals, irradiation of Al with neutrons or charged particles introduces lattice vacancies, selfinterstitial atoms, and transmutation products that evolve into radiation damage microstructure, which causes swelling, radiation hardening, and loss of ductility. Radiation damage effects in aluminum differ from those in most other metals in two respects. One is that the radiation damage is affected strongly by a solid transmutation product, silicon, discussed more in Section 5.07.6.2.3. The other is that Al is much more tolerant of radiation effects than most other metals. At least, it is for irradiations conducted at ambient temperatures. Neutron irradiation of Al at temperatures between 25 and 100
C does not induce detectable radiation hardening until the fast neutron fluence exceeds about 1  1024 n m2, whereas in Fe and Zr, radiation hardening is detectable at fluences two to three orders of magnitude less than that.47 Moreover, even when Al is radiation hardened at 25–100
C, it still retains significant ductility when compared with considerably reduced ductilities in Fe and Zr. This delayed display of radiation hardening exists despite the fact that the number of atomic displacements per atom in Al are about twice as many as in other metals at the same fast fluence, which is brought about by the lower displacement threshold energy for Al. The larger part of Al’s better tolerance of radiation damage is owed to its low melting temperature, which makes its homologous temperature high compared with those for Fe and Zr. At room temperature, the homologous temperature of aluminum is 0.32Tm, versus 0.175 for austenitic steel, 0.17 for ferritic steel, and 0.26 for a-Zr if

Performance of Aluminum in Research Reactors

referred to the a ! b-transition temperature of about 860
C, or 0.14 if referred to the m.p. of b-Zr, 1852
C. In general, noticeable thermally induced movement of lattice vacancies will occur in metals at homologous temperatures above about 0.3Tm. Because of that movement at room temperature in Al, there will be a greater loss of radiation-produced vacancies and of interstitials to mutual recombination, resulting in less nucleation of the point defect clusters that are the seeds of damage microstructure, hence less radiation hardening. A feature of radiation damage in polycrystals is the absence of point defect damage microstructure at grain boundaries and other incoherent interfaces. Point defect clusters and voids do not develop on the boundaries, and damage-free zones are formed on each side of a grain boundary. This denuding may be difficult to see in many high melting point metals irradiated at temperatures below 100
C because the zones are narrow. In Al, the high Tm allows development of wide and conspicuous denuded zones. Incoherent interfaces are comprised of structural dislocation networks and high equilibrium concentrations of vacancies that make the boundaries deep sinks for absorption and recombination of freely migrating point defects. They are pulled in from the near regions of the butting grains, leaving a volume of matrix straddling the boundary that is diminished in radiation-produced point defects. Due to the greater mobility of the interstitials and the bias of dislocations for absorption of interstitials, the zone deprived of interstitials is wider than that for the vacancies. This creates an unbalanced concentration of vacancies at the rim of the denuded zone. Therefore, vacancy clusters are encouraged to form in that rim and they become more numerous and/or larger than those in the grain matrix. Impurities and transmutation products are also drawn into the grain boundary, but they are not annihilated there; they accumulate. If they are largely insoluble, as H, He, and Si are in Al, they will precipitate and grow there, the gases as bubbles and the Si as particles or films. Within the grains, the gases will stabilize embryo clusters of vacancies and facilitate their growth into voids as long as there is an excess of vacancies. Some of the Si will attach itself to the voids. Grown-in dislocations in the grain interiors are also sinks for point defects. Diffusion and binding of freely migrating point defects and solute atoms are important for understanding and analyzing radiation effects. Some useful parameters for pure Al are:48,49

159

 The threshold energy for atomic displacements is 25 eV compared with 40 eV or more for most other metals.  The self-diffusion rate, Dsd ¼ A(Qsd/kT), where A is a constant and the activation energy Qsd is the sum of the formation energy of a vacancy, Evf , and the migration energy of the vacancy, Evm .  For temperatures between 298 and 580 K, A is 1.6E  5 m2 s1 and Qsd is 1.3 eV.  For temperatures between 570 and 923 K, A is 2.0E  4 m2 s1 and Qsd is 1.48 eV.  Evf ¼ 0.6 eV.  Evm ¼ 0.7 to 0.88 eV, deduced from Evm ¼ Qsd  Evf .  Eif ¼ the formation energy of an aluminum interstitial atom, >3 eV.  Eim ¼ the migration energy of an aluminum interstitial atom, 0.1 eV.  Evb-s ¼ the binding energy between a vacancy and a solute atom and is considered to be <0.1 eV for Ag, Cu, Mg, Zn, and Si.  Eib-s ¼ the binding energy between an aluminum interstitial atom and a solute atom. No values are available, but for the solute to reduce b needs to radiation damage microstructure Eis m m b E Þ þ E . be >ðEv i vs Solutes that seem to reduce radiation damage structure most strongly at concentrations of 100 appm are Cr, Cu, Mn, Ti, V which have the largest negative lattice misfits, defined as (a-a0)/fa0, where a0 is the lattice parameter of pure Al, a is the lattice parameter of the alloy, and f is the atomic fraction of solute. No relationship is found between degree of radiation damage and thermal diffusion rates of solutes. 5.07.6.2

Microstructures

Examples of the damage microstructure in highpurity Al irradiated to a fast neutron dose of 3.5  1024 n m2 at 50
C are presented in Figure 3. The two photographs are not the same field. The one on the left shows typical radiation-produced dislocation loops. Some of the smaller spots are particles of radiation-produced Si. The other photograph is tilted to put the dislocations out of contrast and reveal the voids more clearly; they are facetted. The loops and voids are of order 30 nm diameter. In ferritic steel, austenitic stainless steel, and Zircaloy-4 alloys irradiated under similar temperature and neutron fluence conditions as in Figure 3, the radiation damage microstructure is resolvable as 1–2 nm black dots.47 The loops in Figure 3 are not faulted. Nobody has reported

160

Performance of Aluminum in Research Reactors

(a)

0.1 mm

(b)

0.1 mm

Figure 3 Dislocation loops (a) and voids (b) in high-purity aluminum after irradiation at 50
C to a fluence of 3.5  1024 n m2 (E > 0.1 MeV).

faulted loops or stacking fault tetrahedra in neutron-irradiated Al, presumably because the very high stacking fault energy (SFE) of Al,50,51 about 160–200 mJ m2, would inhibit faulting. Yet faulted loops, and multilayered loops, are formed in aluminum during electron bombardment in a highvoltage electron microscope52 and in thin foils that are water-quenched from temperatures near the melting point then aged.53 The occurrence of faulting in these cases is increased with increasing aging temperature and impurity level. It is possible that the SFE may have been reduced by contamination occurring through the foil surfaces. 5.07.6.2.1 Fluence

Experiments54 made to establish the minimum fast neutron fluence for the onset of visible radiation damage microstructure in high-purity Al foils irradiated at <60
C indicate a threshold in the vicinity of 3  1023 n m2, E > 1 MeV. The specimens were in two conditions, annealed and cold worked, and were of three thicknesses, 100 nm, 12.7 mm, and 76.2 mm. None of the 100 nm foils showed damage structure for exposures up to 5  1022 n m2. At a fluence of 3  1023 n m2, the 12.7 and 76.2 mm foils, both annealed and cold worked, showed many small loops, about 25 nm size in low concentration and spotty distribution. After a dose of 1.6  1024 n m2, the loop concentration was higher, but there was little or no change in size. At both of these doses, the annealed specimens contained loosely tangled dislocation lines at higher density than is characteristic of well-annealed high-purity Al. These dislocations were kinked, and some of them that moved while under observation in transmission electron microscopy (TEM) examination were seen to have

been pinned at the loops. In annealed specimens, the grain boundaries had well-defined denuded regions about 0.35 mm wide on each side at a dose of 1.3  1024 n m2, and in the regions next to the denuded zones many of the loops were large and there were dislocation segments among them. Some of these segments spanned the denuded zones. These denuded regions are wider than the 100 nm thick foils in which no radiation damage was found, suggesting that the radiation point defects had migrated from those foils. For cold-worked specimens, nothing was said about the cold work dislocations. It is quite possible that considerable recovery of the coldworked dislocation structure occurred before and during irradiation. At grain boundaries in the coldworked material after irradiation to 6  1023 n m2, a few loops and some dislocation segments are present within a distance of about 0.35 mm from the boundary. In the adjoining regions, loops and dislocation segments are present in high concentrations, and many of the loops are large, >0.1 mm, and they encircle smaller loops or are kinked by smaller loops. Evidently, the dislocation segments are portions of growing loops. The tangled dislocation lines in the annealed specimens at the lower doses probably arose from growth of the earliest loops. No radiation voids were seen in these experiments. It was speculated that the loops were vacancy-type, growing from collapse of vacancy clusters produced by the irradiation. French studies55,56 have corroborated and enlarged on the heterogeneous nature of evolution of early damage microstructure and the roles of dislocations in Al. With increasing fluence at constant irradiation temperature, the loops evolve into dislocation lines, and voids and Si precipitates arise, which increase in concentrations and sizes. The voids are larger and

Performance of Aluminum in Research Reactors

161

0.1 mm

Figure 4 Denuded grain boundary and associated void enhanced regions in 4–9 purity aluminum after irradiation at 50
C to a fluence of 3.4  1026 n m2 (E > 0.1 MeV).

less numerous than the Si precipitates. The voids are facetted and so are the larger particles of Si. Most, if not all, of the voids have a facetted Si particle attached to the outside of one facet of the void. The majority of the Si particles are not attached to voids. The grain boundary denuded regions are not enlarged, but the voids at their rims are exaggerated, as illustrated in Figure 4. An unpleasant potential consequence of these sheets of large voids is that in the event of an unexpected overload, they may provide paths for premature failure by a tearingalong-the-dotted-line-type separation. 5.07.6.2.2 Temperature

Raising the irradiation temperature coarsens the damage microstructure and decreases the degree of radiation hardening for a given dose. At 150
C, Figure 5, the dislocation structure almost disappears; there are fewer voids, but they are larger than at 50
C and are strongly facetted, and many of them are very much elongated.57 Particles of radiation-produced Si are attached to one face of a void, usually at the narrow end of elongated voids. Freestanding Si particles are likewise facetted and elongated, some in ribbon shapes. The denuded regions straddling grain boundaries are wider, about 1 mm each side. For annealed materials irradiated at temperatures above 150
C, certainly above 200
C, no dislocation-type radiation damage microstructure or voids are produced; coarse Si particles are seen. In cold rolled pure Al, some large cavities remained58 after irradiation at 220
C. During postirradiation annealing59 removal of damage microstructure was slowed by impurity content and by higher doses. For 1 h anneals of 1100Al, void swelling began recovery at 200
C and was almost complete at 300
C where gas

1 mm

Figure 5 Voids in high-purity aluminum irradiated at 150
C to a dose of 2  1025 n m2 (E > 0.1 MeV). Right: Preinjected with 3 appm He. Reproduced from Farrell, K.; Wolfenden, A.; King, R. T. Radiat. Eff. 1971, 8, 107–114, with permission from Taylor and Francis.

bubble swelling intervened. Void ripening preceded void elimination. Si precipitates ripened and were prevalent on grain boundaries. 5.07.6.2.3 Transmutation products

The gases helium and hydrogen produced from (n, a) and (n, p) reactions of fast neutrons with lattice atoms considerably affect the development and effects of radiation damage structure by encouraging the nucleation of voids, dislocation loops, and bubbles.57,60,61 Helium is insoluble in Al. It binds strongly with vacancies and has very limited mobility. Hydrogen is almost insoluble at ambient temperatures, becoming more soluble with increasing temperature. In the lattice, it is mobile even at room temperature. The degree of promotion of voids and loops by the gases decreases with increasing irradiation temperature, and the promotion of bubbles increases. In the righthand micrograph of Figure 5, it is evident that even for a high irradiation temperature of 150
C, the presence of just 3 appm He implanted at room temperature has boosted nucleation of voids. The string of voids in the field is not on a grain boundary. It probably marks the position of a grown-in dislocation present during implantation that has climbed away during irradiation. At higher irradiation temperatures58 or during postirradiation anneals,59 the gases form bubbles. The levels of helium and hydrogen produced in Al are not widely different from those in other metals. For the purpose of comparison we can use the production rate tables of Gabriel et al.62

162

Performance of Aluminum in Research Reactors

to estimate the levels of He and H produced in Al, Fe, and Zr during a 1-year exposure in the high flux peripheral target positions of the HFIR core. There, on the horizontal mid-plane at current operating conditions, the annual fast neutron fluence will be 2.4  1026 n m2, E > 0.11 MeV, and the thermal fluence will be 4.5  1026 n m2. In Al, this fast fluence will generate gas concentrations of 11 appm He and 63 appm H, and in Fe there will be 5 appm He and 95 appm H. In Zr, which is noted for its low cross-sections, only about 0.25 appm He and 5 appm H would be produced. The atomic displacement levels for the three metals can also be calculated. They are 36 dpa for Al, 18 dpa for Fe, and 17 dpa for Zr. The larger dpa level in Al is due primarily to its low effective displacement energy, 25 eV versus 40 eV for Fe and Zr. There are three interesting reactions with thermal neutrons that produce gases from the foreign elements Li, B, and Ni, which may be present in some Al alloys. The spatial distributions of gases from these sources are each different. Lithium has high solubility in Al and generates a uniform distribution of gases. Boron and Ni are insoluble and they produce localized concentrations of gas. Lithium is not a common impurity in Al, but there is a commercial series of Al–Li alloys developed for their lightweight highstrength properties. Laboratory-made Al–Li alloys have been used to study radiation hardening, helium embrittlement, and swelling.63,64 Natural Li contains 7.5% of the 6Li isotope, which has a capture crosssection of about 950b and decomposes to He and tritium via the reactions 6Li þ nth ! 7Li ! 4He þ 3H. The present writer65 made an Al–0.052 wt% Li alloy using 6–9 high-purity Al and Li enriched to 98% with 6Li, and irradiated it to a dose of 5.5  1025 n m2, E < 0.0253 eV and 2.2  1025 n m2, E > 0.1 MeV at 55
C. About 95–99% of the 6Li was burnt up to produce about 2200 appm each of He and tritium. The atomic displacement level was about 3 dpa, not including any displacements from the recoiled gases. The effects of these high levels of gases were striking, see Figure 6. The insert is an enlarged view of the matrix cavities. Compared with irradiated pure Al control specimens, the concentrations of matrix cavities were increased 1000-fold, and their sizes decreased tenfold; dislocation densities were increased tenfold. Most grain boundaries were crammed with large bubbles, many so interconnected that it was difficult to obtain thinned foils for TEM examination because the grain boundaries were eaten away before much thinning of the grain interiors

0.1 mm

1 mm

Figure 6 Modification of void structure by very high helium and tritium levels from burnup of 6Li. Reproduced from Farrell K.; Houston J. T. Combined Effects of Displacement Damage and High Gas Content in Aluminum, ORNL-TM-5395; Oak Ridge National Laboratory: Oak Ridge, TN, May 1976. Also available in Proceedings of International Conference on Radiation Effects and Tritium Technology for Fusion Reactors, Gatlinburg, TN, Oct. 1–3, 1975, U.S. Department of Commerce CONF-750989, Mar 1976; pp. II-209–II-233.

occurred. The grain boundary in Figure 6 is one with a low concentration of bubbles. Hardness measurements gave a Vickers pyramid hardness (VPH) of 137 MPa for the annealed, unirradiated Al and the alloy, and 382 and 902 MPa for the irradiated specimens. In bend tests made in air and liquid nitrogen (LN), the unirradiated materials and the irradiated pure Al were bent through full circles without rupture. The irradiated alloy broke with an audible crack and with no detectable plastic strain. Fracture was accompanied by release of tritium. The fracture surfaces displayed 100% intergranular failure. These are incredible hardening and embrittling effects of the gases. Electron microcopy examination of carbon replicas taken from the fracture surfaces showed huge irregular interconnected bubble cavities. Failure occurred by plastic tearing of the small areas of intact grain boundaries between the cavities. Postirradiation annealing treatments caused the appearance of a coarse distribution of large facetted matrix cavities superimposed on the small matrix cavities, and with no denuding of the surrounding small cavities. These enlarged cavities were frequently associated with large silicon particles that grew concurrently during the anneals. An anneal at 500
C showed incipient disintegration of the specimens and TEM foils could not

Performance of Aluminum in Research Reactors

be obtained. It was postulated that the large cavities grown during annealing were tritium bubbles. Al often contains trace quantities of B in the form of small B4C inclusions. Natural B contains 19.8% 10B, which has a large neutron capture cross-section of 3835b, producing Li and He via 10 B þ nth ! 11B ! 7Li þ 4He. The range of the recoiled He atoms is about 5 mm, and the He is segregated in a well-defined band in a halo around the parent inclusion. The larger Li atom has a smaller range and is soluble; it forms a diffuse halo. At low irradiation temperatures and low doses, the halos are very prominent because of heavy decoration with dislocation loops. As the dose increases, the loops grow and move off leaving dense halos of voids, especially for the He halo. The writer has seen hundreds of these halos. Most of them were circular or near circular, with an occasional cigar shape, depending on the shape of the mother particle. Most were isolated randomly, but some were in groups or were strung in chains on a grain boundary. One is illustrated in Figure 7. This particular halo is slightly squashed, following the elliptical contour of the central particle. The dark region of the outer halo is actually filled with small cavities, resolvable at higher magnification. Where a halo intercepts a grain boundary the voids seem to disappear, but during annealing they become visible as bubbles on the boundary that grow

1 mm

Figure 7 Damage halos around a suspected B4C particle in 1100-OAl irradiated to 2.9  1026 n m2 at about 55
C.

163

faster than those elsewhere in the halo. Such highly heterogeneous distributions of transmutant gas have been perceived more as a novelty than as a possible threat to the integrity of the host material. This attitude may be unwarranted. A highly localized concentration of helium in a patch on a grain boundary could be a prime site for premature helium embrittlement at stresses and temperatures below the ranges for normal helium embrittlement elsewhere in the specimen. A spongy helium halo that intercepts the surface of Al cladding may provide a potential site for initiation of local corrosion. For these reasons, it might not be a good idea to consider placing particles of B4C burnable poison in single-layer cladding on monolithic LEU fuel plates; a better location would be in the inner layer of a two-layer cladding. Some of those considerations apply to He produced from Ni in Al. It comes from the 59Ni isotope, which is not found in natural Ni. The 59Ni must first be created from 58Ni that comprises 68.1% of natural Ni. The two-step process66 to yield the He is 58 Ni þ nth ! 59Ni; 59Ni þ nth ! 56Fe þ 4He. Helium generation via this route does not scale linearly with time. It is slow to start while the 59Ni accrues, then it increases as the square of the fluence. It is favored by long-term exposures or strongly thermalized neutron spectra. Only trace quantities of Ni are found in most Al alloys except the X8001 and AlFeNi-type alloys, which contain a nominal 1 wt% Ni. These alloys were developed for cladding because early laboratory corrosion tests indicated they might have better corrosion resistance than existing cladding alloys. Trials of the X8001 have not shown superior performance. The alloying elements in X8001 are insoluble in the solid alloy and form intermetallic inclusions that are malleable and become deformed and extended into stringers during unidirectional rolling and extrusion processing. The helium atoms formed from the Ni in the stringers are recoilimplanted into the near-matrix regions surrounding the stringers where they accelerate local formation of voids and dislocations. The recoiled 56Fe atoms cause extra dpa locally.67 These He-enriched regions are not as obvious as those around B4C inclusions. They are indicated by higher concentrations of voids, and the emergence of more numerous He bubbles during postirradiation annealing. This localized damage offers an explanation of a hitherto inexplicable puzzle found in the corrosion response of X8001 alloy. A characteristic of extruded X8001 tubes undergoing aqueous corrosion in reactors is that smooth shallow troughs or discontinuous ruts lying in the direction of

164

Performance of Aluminum in Research Reactors

the tube axis are created in the corroded surfaces.68 Such troughs have not been reported in laboratory corrosion tests of X8001. It is suggested here that the troughs are stringer beds left when the Ni-rich stringers are eased out of the surface by selective corrosion/erosion of the more highly damaged He-rich matrix at the stringers. Localized enhancement of He at Ni-rich stringers is also believed to play a major role in the occurrence of axial cracking of the X8001 cladding on HFIR long-term isotope target rods.69,70 The target material is a cylindrical compact of actinide oxides in an Al powder matrix, 6.3 mm diameter  14.5 mm long, each jacketed in 1100Al. The meat contains about 10% porosity to accommodate fission gases. A target rod consists of 35 jacketed capsules stacked in a tube of X8001 alloy that is hydrostatically compressed around them to form the outer cladding. The tubes are made by extrusion and have six equally spaced longitudinal fins. Before the target slugs are loaded into the tubes, most of the fins are machined off, leaving short lengths of fins at several locations along the active length of the rod. The loaded target rod is slid into an X8001 tubular sheath with hexagonal ends, known as a hex can. The groups of remnant fins along the length of the rod act as spacers that centralize the rod in the hex can and maintain an annular water-cooling channel around the rod. A bundle of 31 sheathed target rods just fills the vertical HFIR trap. Cooling water flows inside and outside the hex cans. Cracks were found in the mid length, high flux regions of the target rods during a search for the source of a-contamination detected in the exiting coolant. Investigation showed that the cracks were intergranular and were oriented in the length of the rods at locations where the fins had been removed. Lengths were up to 66 mm. The cracks originated in the target rod cladding, but some of the larger ones had penetrated the 1100Al jackets of the target slugs. There was no evidence that corrosion was involved. The hex cans were not cracked. The exposure history of the rods is that they were first irradiated for about 1 year in the D2O environment of the C reactor at Savannah River Nuclear Laboratory at a temperature of 20
C. They were returned to ORNL and inserted in the HFIR at 46
C where, during their fifth fuel cycle, the a-leak was detected. The summed fluences were 6.9  1026 n m2, thermal, and 1.2  1026 n m2, E > 0.82 MeV. The conclusion from the investigation was that gas swelling of the target meat had imposed a hoop stress on the radiation-damaged cladding that had become too stiff

to undergo plastic flow and had cracked instead. Rupture tests on unirradiated lengths of the cladding tubes by internal pressure caused failure along the machined-off fin lines, indicating the lines were weak regions in the tubes. A possible solution to the cracking problem was to decrease the swelling-related hoop stresses on the cladding by raising the pore volume in the target meat from the then current 10% level to 20% and 25%. Trials were successful, and the cracking no longer occurs. The questions of why the cracks were located only on the fin lines and why it was intergranular were not answered, but were pursued.71 The intergranular nature of the fractures made helium embrittlement a suspect. However, calculations of helium levels from the 27Al(n, a) reaction with fast neutrons had given 7 appm, which was considered inadequate for helium embrittlement at the low irradiation temperatures experienced by the rods. But supposing it was occurring, why did it favor the fin lines? Metallographic and TEM examinations of pieces of unirradiated cladding tubing showed that the extrusion process had stretched the inclusion particles into stringers and forced sheets of them into the fins from which they extended back into the tube wall. Removal of the fins left behind an aggregation of stringers protruding into the tube wall. Ergo the weakened regions along the fin lines seen in the rupture tests of the unirradiated tubing. When the enhanced production of helium from Ni by thermal neutron capture was announced,66 a connection between aggregates of Ni-rich stringers and helium embrittlement was discerned. Irradiated pieces of high-purity Al, 1100Al, and a X8001 hex can were sent to a specialist laboratory for helium analyses. For a common thermal fluence of 1  1026 n m2, the results were 1.8, 4.8, and 9.5 appm, respectively. After a fluence of 3  1026 n m2, the corresponding values were 7.2, 18.1, and 145. A piece of hex can irradiated to 5.8  1026 n m2 yielded 220 appm. A piece of 6061Al at 13.8  1026 n m2 gave 47 appm, much less than the hex can at 5.8  1026 n m2. These results leave no doubt that the presence of Ni in Al irradiated to high thermal neutron fluences greatly boosts the helium levels. And since the He will remain close to the Ni particles, there must be very large concentrations around the stringers. Any grain boundaries overlapped by those local helium clouds will be prime candidates for helium embrittlement cracking under the influence of hoop stresses. Hence, the intergranular cracking at the fin lines. In most metals, the gaseous transmutation products play a larger role in the development and effects of

165

Performance of Aluminum in Research Reactors

radiation damage structure than do the nongaseous transmutants, one reason being that most construction metals do not produce much nongaseous transmutants. Al is different. Depending on the degree of thermalization of the neutron spectrum, Al can produce large quantities of silicon from the two-step reaction 27Al þ nth ! 28Al þ g; 28Al ! 28Si þ b. A rough guide to the quantity expected annually in the HFIR PTP spectrum can be obtained by multiplying the thermal neutron fluence by 230mb, the standard thermal neutron (0.0235 eV, 2200 m s1) absorption crosssection for Al. The result is 1.035 at.% Si (1.073 wt%). The Si is insoluble in Al at temperatures below about 350
C and is usually manifest as a precipitate of elemental Si.72 This precipitate makes a substantial contribution to radiation damage in Al, and is the dominant hardening agent at high thermal neutron fluences. There is one outstanding qualifier to that generalization. In the 5xxx-type Al–Mg solid solution series, the free Mg atoms dissolved in the Al will react with the atoms of transmutant Si to form a precipitate of Mg2Si.73 Thus, a 5xxx series alloy will be converted to a 6xxx-like alloy.73–76 Figure 8 shows the Mg2Si microstructure formed in irradiated 5052-O alloy. Because this precipitate occurred at a temperature below the usual 160
C aging temperature used to obtain the T6 tempered condition in 6061Al, the Mg2Si precipitate developed in the 5000 alloy is finer than in the 6061-T6 alloy. The microstructure of heavily irradiated 6061-T6 alloy is illustrated in Figure 9. Since there are usually larger quantities of Mg in the 5xxx alloys than in the 6xxx alloys, a greater volume of Mg2Si can be created in the former 5xxx alloys. Hence, irradiated 5xxx alloys will undergo radiation hardening and precipitation hardening simultaneously, and their overall hardening rate will be larger than in other Al alloys exposed to the same neutron fluence. Note that there are no voids in Figure 8. Sparsely distributed voids are found75 at a higher fluence of 1.8  1027 n m2. At half of that dose, the 6061-T6 alloy contains many more voids, Figure 9. The association of the transmutant Si with voids is interesting. We saw earlier that voids and particles of Si become visible in the microstructure at about the same dose. The voids are larger and fewer than the Si particles. A Si particle is usually attached to one facet of a void, and that particle is larger than its unattached brethren in the matrix. It is also facetted. As irradiation continues, a change occurs in the void-Si relationship. The voids lose their facetted shape and become rounded.77 They are completely covered with a thin coating consisting of mostly Si

100 nm

200

000

220

020

Figure 8 Precipitates of Mg2Si and excess Si in formerly 5052-OAl irradiated to 5.7  1026 n m2 at 55
C. Reproduced from Farrell, K. J. Nucl. Mater. 1981, 97, 33–43, with permission from Elsevier.

0.1 mm

Figure 9 Voids with background precipitate of radiation-produced silicon and silicon-decorated original Mg2Si precipitates in 6061-T6Al irradiated to 1027 n m2 at 55
C.

166

Performance of Aluminum in Research Reactors

0.1 mm Figure 10 Si particles and Si-coated voids on a carbon extraction replica from 1100-OAl irradiated to 1.4  1027 n m2 (E > 0.1 MeV) and 2.3  1027 n m2 (E < 0.025 eV) at 55
C. Reproduced from Farrell, K.; Bentley, J.; Braski, D. N. Scripta Metall. 1977, 11, 243–248.

with some Al. The coating is noncrystalline and flexible. The unattached Si particles in the matrix are also rounded but are crystalline. The coated voids jut out of the thinned edge of the hole in TEM foils and can be lifted from the matrix on carbon extraction replicas. Figure 10 is an example. The larger features with the dark rims are the coated voids. Four of them have partially collapsed without breaking, indicating a highly ductile coating. Many of the Si particles seem to have a layered structure. The Al–Si system is a simple eutectic; there are no compounds. It is suspected that the small amount of Al found in the void coatings may be from the Al matrix that was not completely dissolved from the voids during the electrolytic extraction process. Silicon is obviously involved in void formation and growth but its specific role is unclear.

5.07.7 Property Changes 5.07.7.1

Swelling

Radiation swelling is the increase in volume arising by accumulation of voids from excess vacancies and by formation of gas bubbles. For Al, there are also small swelling contributions from build-up of particles of transmuted Si and, in 5xxx alloys, Mg2Si, which have densities of 2329 and 1990 kg m3, respectively.78 Gas bubble swelling is not an issue for Al in RRs because the temperature is too low, except perhaps in fuel cladding where some pores

found in the cladding may have been formed by the accumulation of hydrogen. Swelling can be measured from dimensional changes. More often it is determined from changes in immersion density values. Swelling in various Al alloys is shown in Figure 11. These alloys were all irradiated in the core of the HFIR and they make the most comprehensive and consistent set of swelling data.79 For reference, the dotted line is estimated for the swelling from Si alone. It is evident that the unirradiated chemical compositions and microstructures have major effects on the degree of radiation swelling. The purest grades, sixnines and four-nines, show swelling earliest in dose and swell at the highest rates with dose. The rates decrease above a dose of about 1  1025 n m2. Swelling in the two-nines grade (1100-O) requires significantly higher doses, but the swelling rate is unchanged. The 6061-T6 alloy, with its inherent Mg2Si phase, starts swelling appreciatively later in dose than the 1100-O. This is traceable to reduced nucleation of voids, but its swelling rate is about the same as the other alloys. The greatest resistance is in the 5052-O alloy. There, the swelling is less than for the Si alone. In this alloy, much of the early swelling is not due to voids; it is caused by the silicon and the new Mg2Si phase and by the increase in the original density of the matrix, r0, as Mg is drawn from solution to create the Mg2Si. The effects of prior cold work on swelling in Al agree in general that the presence of cold work dislocation structure decreases the overall void swelling but the reduction is not massive; concurrence of dislocation recovery confuses the details.80–83 5.07.7.2

Mechanical Properties

The major consequences of radiation damage structures on the mechanical properties of Al alloys are radiation hardening and associated loss in ductility. There are too many data from too many sources to be described in detail here. A good source of compiled data, including the sparse information on fracture toughness and weldments, is Marchbanks.84 For 6061Al in particular, see Farrell.85 Strengthening and loss of ductility are demonstrated best in tensile properties. In Figure 12 we can directly compare the changes in strength and ductility of different alloys irradiated and tested under the same conditions.79 The most striking feature is the relatively rapid hardening displayed by the 5052-O alloy. As explained earlier, this is caused by the combined effects of radiation damage and in-reactor

Performance of Aluminum in Research Reactors

wt % Si 0.1

1

dpa 0.1

10

10

10

100

6-9 4-9 2-9 (1100-O) 6061-T6 5052-O

( ri )

%

Ti = 328 K (0.35 Tm) 1 6061-T6

Swelling

ro−ri

1

167

1100-O 5052-O 0.1 Pure aluminum 28

Si

0.01 1024

1025 1026 Fluence (n m-2 > 0.1 MeV)

1027

Figure 11 Radiation-induced swelling in various Al alloys as a function of fast fluence. Reproduced from Farrell, K. In Proceedings of the Conference on Dimensional Stability and Mechanical Behaviour of Irradiated Metals and Alloys, Brighton, Apr 11–13, 1983; British Nuclear Energy Society: London, 1983; Vol. 1, pp 73–76, with permission from British Nuclear Energy Society (now The Nuclear Institute).

formation of a fine precipitate of Mg2Si. In contrast, the 6061-T6 alloy, which contains Mg2Si before irradiation, begins radiation hardening at about the same fluence as the 1100-OAl, and hardens thereafter at the same rate. The 1100-O alloy contains no Mg2Si before or after irradiation. From which we deduce that preexisting Mg2Si precipitates play no role in radiation hardening. This is an interesting conclusion. It contradicts the expectation that the precipitates would diminish the degree of radiation hardening in 6061-T6Al by promoting the recombination of freely migrating vacancies and interstitials. Perhaps that expectation is wrong. But it seems satisfactory for explaining the delayed swelling in the 6061-T6 alloy, where nucleation of voids is retarded, perhaps until the transmutant gases enable achievement of critical size cavity nuclei. Alternatively, maybe the radiation-produced Si dominates the hardening process. It is a mystery. The fluence for the onset of radiation hardening in the weak 4–9Al is about one order of magnitude less than in

the other alloys, and the subsequent rate of hardening is less than for the others. Here, again, we invoke the recombination argument. This higher purity material contains less solutes and inclusions. Thus there is less trapping and annihilation of freely migrating point defects, hence more point defect clusters are formed in the early stages of irradiation. It is suspected that the reduced rate of hardening is connected with dynamic recovery of deformation during the tensile test. It was pointed out in Section 5.07.3.2 that recovery from cold work occurs readily in high-purity Al at room temperature. Loss in uniform elongation in all of the alloys is concomitant with increase in strength . . . to a point. At a fluence of about 1026 n m2 the ductility reaches a plateau of 3–5% even though the strength continues to rise. The 1100-OAl has the least ductility in the plateau region and displays an intergranular-like fracture mode that may be caused by tearing and void interconnection in the void-rich regions lying alongside the grain boundaries.

Performance of Aluminum in Research Reactors

600

Stress (MPa)

500

Tirr = 328 K Ttest = 323 K . ~ 10-4 S-1 '

168

UTS 0.2% FS

400

300

6061-T6

200 5052-O 100 1100-O 4-9AI 0

Uniform elongation (%)

40

30

20

4-9AI 5052-O 1100-O

10

6061-T6

0 1023

1024

1025

1026

1027

-2

Fluence (n m > 0.1 MeV) Figure 12 Radiation-induced changes in room temperature tensile properties of various Al alloys. Reproduced from Farrell, K. In Proceedings of the Conference on Dimensional Stability and Mechanical Behaviour of Irradiated Metals and Alloys, Brighton, Apr 11–13, 1983; British Nuclear Energy Society: London, 1983; Vol. 1, pp 73–76, with permission from British Nuclear Energy Society (now The Nuclear Institute).

Irradiated metals undergoing plastic deformation are prone to strain localization, seen as dislocation channels swept free of point defect clusters. Such channeling coincides with changes in the tensile test curve, notably introduction of a sharp yield point and a reduction in the slope of the strain-hardening portion of the curve. In severe cases, a yield point drop occurs that leads directly into prompt necking. Dislocation channeling has not been reported for neutron-irradiated Al. Nor do the tensile curves for neutron-irradiated Al display sharp yield points and pronounced decreases in work hardening rate. The reason for this apparent resistance to channeling in

Al is not known. Al is not immune to channeling. It occurs in unirradiated quenched-and-aged Al. When high-purity Al crystals are quenched from near the melting point into water at room temperature or into iced brine, then aged at room temperature for 1–4 days or at 60
C for 2 h, there is a considerable increase in yield stress (YS) and large decrease in strain hardening rate.86,87 Such quenchaging treatment produces many small vacancy clusters. Yielding occurs abruptly, and large widely spaced slip bands appear on the specimen surfaces, forming mirror images from one surface to the opposite surface. TEM examination87 reveals that

Performance of Aluminum in Research Reactors

A5 AI, ~35% CW, 77 K (0.08 Tm)

400

~1 ⫻ 1023 n m-2, E > 1 MeV

Stress (MPa)

300

~5 ⫻ 1021 n m-2

7.5 ⫻ 1021 n m-2, then held 60 h at room temperature before test at 77 K

200 Unirradiated

100

Unirradiated, tested 293 K

0 0

10

20

30

40

Elongation (%)

Figure 13 Tensile curves of pure Al irradiated and tested at cryogenic temperatures, showing recovery at room temperature. Reproduced from Bochirol, L.; Brauns, P.; Claudet, G. Prog. Refrigeration Sci. Technol. 1973, 1, 643–650, with permission from International Institute of Refrigeration, AVI Publishing Company.

the slip bands are dislocation channels prominently displayed against the background of undisturbed vacancy clusters remaining in the matrix between the channels. Another way to induce a sharp yield point and reduced strain hardening rate in Al is to irradiate and tensile test at cryogenic temperature. Figure 13 gives tensile curves for an A5 alloy (>99.95% Al) in the half-hard, H24 (35% cold worked), condition after neutron irradiation and testing at 77 K in LN.88 A curve for an unirradiated specimen tested at room temperature is included as a reference condition. Note that this unirradiated specimen has only 14% elongation, in keeping with its cold-worked condition. When tested in LN the unirradiated material is not only stronger it is much more ductile, with an elongation of 44%. This is an example of the enhancement of ductility at low temperature mentioned in Section 5.07.3.1. Irradiation to 5  1021 n m2 in LN further increases the YS, and although the elongation falls a little, it is still larger than at room temperature; moreover, the initial part of the strain hardening curve to about 8% elongation is steeper than that for the unirradiated specimen, indicating no irradiation-induced dislocation channeling. However, the reduced slope after 8% elongation and prior to onset of necking in the irradiated specimen signals possible late entry of channeling. Note that for this dose at room temperature there would be no discernible radiation damage structure and no change in tensile properties. In the LN specimen irradiated to the higher fluence

169

of 1  1023 n m2 intervention of channeling is more likely. There, very steep initial strain hardening passes though a peak at 1% elongation and plunges into prompt necking. Even greater strengthening and more exaggerated prompt necking is found for both cold-worked and annealed specimens when the irradiation and testing are done at 27 K in liquid neon.88 The deformation modes were not determined for these specimens. Identification of the deformation mode would have required TEM examinations of specimens cut from the gauge sections of the specimens and would probably not have revealed the channeling because the specimens would undergo rapid recovery from radiation hardening when they are brought up to room temperature. Such recovery is illustrated by the curve for the specimen irradiated to 1  1022 n m2 then held at room temperature for 60 h before testing in LN. Almost complete recovery has occurred, indicating elimination of the point defect clusters needed to provide background contrast for detection of dislocation channels. The good result of this recovery is that it relieves the stored energy built up by accumulation of the radiation damage microstructure at the cryogenic temperature. Advantage is taken of this recovery process in Al cold neutron sources by performing periodic in situ room temperature anneals on them. These anneals do not erase the transmutation products. Residual hardening from the gaseous products will be negligible but the accumulating Si, whose levels will be high due to the highly thermalized neutron spectrum at the cold source, will give increasing retained hardening as it forms precipitates at room temperature. For cold source vessels constructed from 5xxx alloys, the transmutation-produced Si will react with dissolved Mg at room temperature to create naturally aged fine precipitates of Mg2Si, with greater hardening effects than from Si alone. 5.07.7.3

Effects of Neutron Spectrum

When tensile data for irradiated Al from different reactors or from different regions of a single reactor are compiled in a single traditional plot of property versus fast neutron fluence, there tends to be a large scatter in the data, particularly at the higher fluences. It was noted89 that the scatter was reduced somewhat if the plot was made in terms of thermal fluence. This improvement was attributed to a spectral effect involving the production of Si precipitates by the thermal neutrons and their modification by fast neutrons. That explanation has since been taken a little

170

Performance of Aluminum in Research Reactors

wt% silicon

0.1

0.01

1

10

Tensile strength (MPa)

800 6061-T6 aluminum . grouping by fth/ft ratio

700

20

Unirradiated HFIR target HFBR surveillance HFBR V15 HFBR CRDF

600 500

21

21

21

8 5-81-3 22

5

400 5

1.3

.57 53

5

0.5

300 200 0

1023

1024

1025 1026 Thermal fluence (n m- 2)

1027

1028

Tensile strength (MPa)

800 20

700

21

600

21

21 5-8 8

500 5

22

400 5

1-3 1.3 .57 .53

5

0.5

300 200 0

1023

1024

1025

1026

1027

1028

Fast fluence, E > 0.1 MeV (n m- 2)

Figure 14 Radiation-induced strengthening of 6061-T6Al discriminated by thermal and fast fluence. Reproduced from Farrell, K. In Proceedings of Materials Research Society Symposium on Microstructure of Irradiated Materials, 1995; Vol. 373, pp 165–170, with permission from Materials Research Society.

farther.90 Figure 14 presents tensile strength data for the 6061-T6 alloy irradiated in two reactors, the HFIR and the High Flux Beam Reactor (HFBR), at 50–65
C. The same strength data are plotted against thermal fluence in the upper box and against fast fluence in the lower box. Ignoring for the moment the lines in the plots, it is evident that at fluences above 1  1025 n m2 the data points are more scattered when plotted against fast fluence. The HFIR specimens were all irradiated at the same location where the ratio of thermal-to-fast flux, ’th/ ’f, was narrow, 1.7–2.3. The HFBR was cooled and moderated by heavy water, and the HFBR specimens were taken from regions with ’th/’f values between 0.5 and 22. When the data are discriminated by ’th/ ’f as indicated by the lines labeled with the large numerals, a pattern of dependence on ’th/’f emerges. The trend is the same in both boxes but the separation is clearer in the fast fluence plot. It is evident that when irradiation is conducted in a hard spectrum, ’th/’f ¼ 0.5, there is little increase in strength, even at high fluence. As the spectrum softens, ’th/’f ! 21, the specimens harden. This response is obviously due to greater generation of silicon precipitates in the more thermalized spectra.

TEM examinations of the more heavily irradiated specimens showed that the dominant microstructural feature introduced by the irradiation in both reactors was particles of Si. They were more numerous and smaller in the softer spectrum, in agreement with the strengthening trend. After eliminating possible contributions from differences in damage rates, irradiation temperatures, and exposure times, it was concluded that the spectrum controls the ripening behavior of the Si precipitates, which determines the degree of strengthening. A hard spectrum causes more sputtering of the precipitates and more transport of the sputtered atoms, resulting in greater ripening of the precipitates and less strengthening per unit Si level or unit fluence. 5.07.7.4 Radiation Softening, Creep, and Stress Relaxation There are some highly contentious reports of radiation softening of Al alloys that demand some discussion. These are not claims of minor softening such as a decline in the strain-hardening region of a tensile curve. They are full-blown changes from an agehardened or cold-worked condition to a dead soft,

Performance of Aluminum in Research Reactors

‘annealed’ state. And they occur in specimens that are not carrying a load. The softening is usually interpreted as being due to destruction of dislocation tangles and precipitate particles through dynamic atomic displacements created by fast neutrons. It is the present writer’s opinion that they are actually unintended instances of thermal softening by radiation heating, which can happen very easily and quickly in Al if adequate precautions are not taken to ensure adequate cooling during irradiation. In each of these reports, the specimens were irradiated in sealed cans with external water cooling and no means of measuring the internal temperature. Using sealed cans is an invitation to radiation heating of the contents unless very strict actions are taken to ensure satisfactory heat removal. In the first91 of these reports, disks of 2024-O, 2024-T3, 1100, and sintered Al powder of undeclared Al2O3 content were placed in aluminum cans evacuated and back filled with helium. The cans were irradiated in the MTR for 60 days to fast neutron doses of 3  1022, 2  1023, and 8.5  1023 n m2 at estimated temperatures of 100, 119, and 138
C, respectively. Hardness tests on the 2024-T3 disks showed: unirradiated control, Hv ¼ 170; 3  1022 ¼ 178; 2  1023 ¼ 68; 8.5  1023 ¼ 60, indicating a pronounced change from the age-hardened-T3 condition to a fully softened condition. The hardnesses of the 1100 and the sintered aluminum powder alloy were unchanged by the irradiations. This softening of 2024-T3 alloy conflicts with tensile and hardness data92 for a 2024-T3 or -T4 alloy irradiated to doses between 2  1023and 1.3  1026 n m2 that show no change in properties to a dose of 1.2  1024 n m2, above which there is hardening. These latter data were for specimens that were irradiated in direct contact with flowing water to minimize nuclear heating. The writer has also tested tensile specimens of 2024-T351Al that were irradiated in contact with flowing water; there was no change in properties for fluences between 4  1021 and 1  1024 n m2, beyond which hardening was displayed.93 The second softening claim94 was made for tensile specimens of Al–3Mg in a cold-worked condition (degree of work not given), and 6061-T6 bombarded with 600–800 MeV protons in the Los Alamos Meson Physics Facility (LAMPF). Annealed specimens of the two alloys were included for reference. Assemblies of seven specimens laid side-by-side were sealed in envelopes made from 0.13 mm thick foil of 1100Al. Sixteen packages were stacked with gaps between them to allow cooling water to flow though at a rate of 1 l s1 and pressure of 17 bar (1700 kPa) to

171

ensure good contact of the foil with the specimens. The beam spot size was 50 mm, the average current was 574 mA, and the exposure period was 750 h. Displacement damage up to about 0.2 dpa was created by the direct impact of the protons and by highenergy neutrons spalled from surrounding materials and the nearby beam stop. No Si was generated. Helium and hydrogen contents were 67 and 275 appm. From the beam parameters, it can be seen that the power input was about 200 MW m2, which is about 20 times larger than the heat fluxes reached on fuel cladding. The irradiation temperature for the LAMPF specimens was estimated to be 40–50
C, certainly not exceeding 100
C. Postirradiation tests revealed that the yield strengths of the cold-worked Al–3Mg and 6061-T6 specimens were reduced by 70–80% and the ultimate strengths by 40–65%, which placed them at the same strength levels as those for unirradiated annealed specimens of the alloys. The irradiated annealed alloys did not undergo irradiation hardening; they actually softened by 10–20%. Metallographic studies and TEM examinations of the formerly cold-worked Al–3Mg alloy94,95 showed that recrystallization had occurred. At a dose of only 0.1 dpa, the tightly packed dislocation cell structure was gone, replaced by very loosely tangled dislocations, and the grains were now equiaxed. In the former 6061-T6 alloy, there were no longer any Mg2Si precipitates. In the irradiated annealed Al–3Mg specimens, there were gas bubbles on the grain boundaries. These microstructural changes are very much like those found for thermal anneals or high temperature irradiations; they require sustained, long range, high-intensity diffusion of vacancies and solutes. Usually, irradiation with energetic particles does not create or support such condition unless radiation heating occurs. Under adequate temperature control, radiation enhanced diffusion is highly local and sporadic and is rapidly quenched. Normally, radiation attrition of precipitates results in precipitate atoms being recoiled into the adjacent matrix and there forming new satellite particles of the mother phase or a new phase. A situation of total elimination, with no reappearance of the precipitate by aging, is more typical of a high-temperature solution anneal and a slow cool. Nevertheless, the authors attributed these extreme changes in microstructure to the absence of transmutation-produced Si which stabilizes radiation damage microstructure in regular neutron irradiations, and to the different nature of the damage produced by high-energy protons, this despite efforts they and others were making

172

Performance of Aluminum in Research Reactors

at the time to promote such irradiations as a substitute for slower neutron irradiations. They were unable to account for the disagreement with an earlier 600 MeV proton irradiation of a solution-treated Al–Mg–Si alloy similar to 6061 at 150
C in the Swiss PIREX facility, which induced precipitation of a high density of Mg2Si precipitates.96 Strangely, they gave no consideration to the possibility that they might have overestimated the cooling efficiency of their irradiation experiment in LAMPF. It should have been the prime consideration because at that time LAMPF was undergoing changes to make it suitable for materials irradiations. None had been performed there previously. The changes were incomplete but the experiment was made anyway. It was a pseudo atest. Since then, improvements have been made to the facility and many materials irradiations have been conducted that simulate reactor irradiations. Later irradiations of 6061-T6 in LAMPF97,98 have not encountered the softening found in the inaugural proton irradiation. The third softening claim99,100 was again for a cold-worked, nominal Al–3Mg alloy and for a precipitation hardened AlMgSi (6061-type) alloy. Tensile specimens and annealed companion specimens were irradiated to fast fluences of 8.7  1021 and 2.5  1022 n m2 in the wet channels of the Egyptian Nuclear Research Reactor. The responses of the materials to both of these fluences were almost identical and are confusing. The AlMgSi alloy in two different age-hardened conditions and its as-received cold rolled condition displayed large decreases in hardness, YS, and ultimate tensile stress (UTS). The Al–3Mg alloy in its as-received cold rolled condition showed large reduction in hardness and relatively small reductions in YS and UTS. In a partially annealed condition (1 h at 240
C) its hardness was reduced 30% but its YS and UTS were increased by 55% and 27%. In the fully annealed state (1 h at 400
C) the hardness was decreased by 5%, the YS was increased by a whopping 380%, and the UTS rose by 97%. Although no microstructural studies were made, the authors concluded that the observed softening was due to removal of cold work dislocations and dissolution of Mg2Si precipitates as seen in Lohmann et al.94 and Singh et al.95 The conclusion that Mg2Si precipitates are unstable in a neutron flux is at odds with the well-established fact that the precipitates persist to very high fluences in 6061-T6 specimens that are properly cooled during irradiation, and they are actually generated by the irradiation in Al–Mg solid solution alloys. The unusually large

hardening caused in the partially and fully annealed Al–3Mg specimens was not discussed. It was ascertained from the lead author that the specimens had been irradiated in sealed cans. It was conveyed to the author that ongoing experiments at ORNL with 6061-T6 were not finding any signs of radiation softening at fluences similar to those in his experiments. Some years later, he published a third paper101 on the AlMgSi alloy in four conditions: Cold rolled; annealed; naturally age hardened; and artificially age hardened. The specimens were wrapped in Al foil and were packed in aluminum powder in sealed Al cans. The fast neutron fluence was 3.7  1022 n m2, somewhat higher than his previous irradiations. Hardness and tensile tests revealed no softening of the cold-worked and age-hardened specimens. The annealed specimens displayed a 23% increase in Vickers hardness and a 111% increase in yield strength. This radical departure from the previous softening results was not credited to better cooling of the specimens during irradiation. Rather, the newfound nonsoftening of the prehardened materials was attributed to the longer irradiation exposure. It was explained that the prehardened alloys actually did go through a full softening process in the region of 1  1022 n m2 as before, by the mechanism of radiation-induced dissolution of the cold work dislocations and Mg2Si precipitates. This removal of these strong point defect sinks allowed irradiation damage microstructure to build up as irradiation continued. At the exposure of 3.7  1022 n m2 sufficient damage microstructure had formed to restore the strengths of the softened materials to their former values. These claims of radiation softening, controversial as they are, do serve a useful purpose. They highlight the previously discussed temperature sensitivity of cold-worked and age-hardened aluminum. Moreover, they send a message that in the event of an unplanned temperature excursion on such materials they should be examined carefully to ensure that they still meet the strength requirements for the application. Radiation creep and stress relaxation are timedependent plastic deformation processes driven by applied external stresses or by stored internal stress. They can be accelerated by point defects from atomic displacement events. The few data that exist for radiation creep in aluminum are confusing and inconclusive. Details are available in Farrell.85 Creep can lead to failure if the load is continuously applied. Strain from stress relaxation ceases once the load is relaxed. There are at least two cases81,102 of stress

Performance of Aluminum in Research Reactors

relaxation in Al reactor components. Both were found in retired reactor components that had entered service in cold-worked conditions, one 1100Al and the other precipitation-hardened 6063Al. During service they had been continuously in direct contact with the coolant water, so thermal interference was not involved. Some softening occurred due to partial recovery of the cold work structure at low doses, followed by radiation hardening at higher doses. Complete softening to an annealed level did not occur during irradiation.

5.07.8 Conclusion Aluminum and its alloys have contributed immensely to development of water-cooled RRs and to our understanding of radiation effects in metals. They continue to do so.

10.

11.

12.

13.

14.

15.

16.

Acknowledgments During preparation of this chapter, interactions with present and former Oak Ridge National Laboratory personnel, S. A. David, R. D. Godfrey, S. J. Pawel, J. D. Sease, and R. E. Stoller were sincerely appreciated. Opinions and conclusions in this chapter are solely the responsibility of the writer.

17.

18.

19.

References 1. 2. 3.

4.

5. 6.

7. 8.

9.

West, C. D. Nuclear News, Oct 1997; pp 50–56. http://www.iaea.org/worldatom/. Rosenthall, M. W. An Account of Oak Ridge National Laboratory’s Thirteen Nuclear Reactors; ORNL/ TM-2009/181; Oak Ridge National Laboratory: Oak Ridge, TN, Aug 2009, Revised Mar 2010. Williams, R. O. Terminal Report on ORNL Slug Problems: Causes and Prevention; ORNL-CF-50-7-160; Oak Ridge National Laboratory: Oak Ridge, TN, period covered Sept 1948 to July 1950. Flemings, M. C. Solidification Processing; McGraw-Hill: New York, 1974. Thomas, W. M.; Nicholas, E. D.; Needham, J. C.; Murch, M. G.; Temple-Smith, P.; Dawes, C. J. Friction-Stir Butt Welding. GB Patent No. 9125978.8. International Patent Application No. PCT/GB92/02203; Welding Research Institute: Great Britain, 1991. Aluminum Standards and Data, 10th ed. The Aluminum Association: Washington, DC, 1990. Use of 6061-T6 and 6061-T651 Aluminum for Class 1 Nuclear Components, Section III, Division 1, Cases of Boiler and Pressure Vessel Code, Supplement 10-NC, Case N-519; American Society of Mechanical Engineers: New York, 1994. Structural Alloys Handbook, 1989 ed., Vol. 3; Battelle Memorial Institute: Columbus, OH, 1989; p 14.

20. 21.

22. 23.

24. 25.

26.

173

Matos, J. E.; Snelgrove J. L. Selected Thermal Properties and Uranium Density Relations for Alloy, Aluminide, Oxide, and Silicide Fuels; IAEA-TECDOC-643, International Atomic Energy Agency, Vienna, 1992; pp 1–19, Appendix I-1.1 in Research reactor core conversion guidebook Volume 4; Fuels (Appendices I–K). Gibson, G. W.; Shupe, O. K. Annual Progress Report on Fuel Element Development for Fiscal Year 1961; IDO-167 27, TID-4500 (17th ed.), Phillips Petroleum Company: Idaho Operations Office, 1961. Bartz, M. H. In Properties of Reactor Materials, Proceedings of the 2nd United Nations International Conference on the Peaceful Uses of Atomic Energy, Geneva, Sept 1–13, 1958; Vol. 5, pp 466–474. Marshall, F. M. Advanced Test Reactor Capabilities and Future Irradiation Plans; INL/CON-06-01211; Oct 2006; www.inl.gov/technicalpublications/Documents/ 3489341.pdf. Beaver, R. J.; Adamson, G. M., Jr.; Patriarca, P. Procedures for Fabricating Aluminum-Base ATR Fuel Elements; ORNL-3632; Oak Ridge National Laboratory: Oak Ridge, TN, Jun 1964. Kim, S. S.; Pope, C.; Taylor, L. L. Criticality Analysis for Proposed Maximum Fuel Loading in a Standardized SNF Canister with Type 1a Baskets; INL/EXT-07-12326; Idaho National Laboratory: Idaho Falls, ID, Feb 2007. Binford, F. T.; Cramer, E. N. The High Flux Isotope Reactor; A Functional Description, Vol 1B, Illustrations; ORNL-3572 (Revision 2); Oak Ridge National Laboratory: Oak Ridge, TN, Jun 1968. Adamson, G. M., Jr Fabrication Procedures for the Initial High Flux Isotope Reactor Fuel Elements; ORNLTM-2196; Oak Ridge National Laboratory: Oak Ridge, TN, Feb 1969. Knight, R. W.; Morin, R. A. Fabrication Procedures for Manufacturing High Flux Isotope Reactor Fuel Elements – II; ORNL/6852; Oak Ridge National Laboratory: Oak Ridge, TN, Dec 1999. Sease, J. D.; Primm, R. T., III; Miller, J. H. Conceptual Process for the Manufacture of Low-Enriched Uranium/ Molybdenum Fuel for the High Flux Isotope Reactor; ORNL/TM-2007/39; Oak Ridge National Laboratory: Oak Ridge, TN, Sept 2007. Adamson, G. M., Jr.; Knight, R. W. HFIR Fuel Element Production and Operation; ORNL-TM-2196; Oak Ridge National Laboratory: Oak Ridge, TN, Jun 1968. Richt, A. E.; Knight, R. W.; Adamson, G. M., Jr. Postirradiation Examination of the Performance of HFIR Fuel Elements; ORNL-4714; Oak Ridge National Laboratory: Oak Ridge, TN, Dec 1971. http://www.rertr.anl.gov/. Hofman, G. L.; Rest, J.; Snelgrove, J. L. Comparison of Irradiation Behavior of Different Uranium Silicide Dispersion Fuel Element Designs. CONF-9409107-6, ANL/TD/CP-85108; Presented at the 1994 International Meeting on Reduced Enrichment for Research and Test Reactors, Williamsburg, VA, Sept 18–23, 1994. Sears, D. F.; Conlon, K. T. Development of LEU Fuel to Convert Research Reactors: NRU, MAPLE, and SLOPOKE; http://www.7ni.mfa.no/NR/rdonlyres/EC9F5.... Clark, C. R.; Knighton, G. C.; Meyer, M. K.; Hofman, G. L. Monolithic Fuel Plate Development at Argonne National Laboratory, 2003; http://www.rertr.anl/RERTR25/PDF/ Clark.pdf. Clark, C. R.; Wight, J. M.; Knighton, G. C.; Moore, G. A.; Jue, J. F. Update on Monolithic Fuel Fabrication Development; INL/CON-05-00855; Idaho National Laboratory: Idaho Falls, ID, Nov 2005.

174 27.

28. 29.

30. 31.

32.

33. 34. 35.

36.

37. 38.

39.

40. 41. 42.

43. 44.

45.

Performance of Aluminum in Research Reactors Moore, G. A.; et al. Monolithic Fuel Fabrication Process Development at the Idaho National Laboratory, 30th International RERTR Meeting, Washington, DC, Oct 5–9, 2008; INL/CON-08-14550. Corrosion of Research Reactor Spent Fuel in Water, 2003; http://www-pub.iaea.org/MTCD/publications/ PDF/TRS418_web.pdf. Griebenow, M. L.; Hanson, G. H.; Larrick, A. P. TRA Oxide Film Control and Surveillance; RE&C Report RE-A-77-059; Idaho National Laboratory: Idaho Falls, ID, Oct 1977. Shaber, E.; Hofman, G. Corrosion Minimization for Research Reactor Fuel; INL/EXT-05-00256; Idaho National Laboratory: Idaho Falls, ID, Jun 2005. Griess, J. C.; Savage, H. C.; Rainwater, J. G.; Mauny, T. H.; English, J. L. Effect of Heat Flux on the Corrosion of Aluminum by Water. Part III. Final Report on Tests Relative to the High Flux Isotope Reactor; ORNL-3230; Oak Ridge National Laboratory: Oak Ridge, TN, Dec 1961. English, J. L.; Rice, L.; Griess, J. C. The Corrosion of Aluminum Alloys in High Velocity Water at 170 to 290
C; ORNL-3063; Oak Ridge National Laboratory: Oak Ridge, TN, Jun 1961. Draley, J. E.; Ruther, W. E. Aqueous Corrosion of 2S Aluminum at Elevated Temperatures; ANL-5001; Argonne National Laboratory: Lemont, IL, Feb 1, 1953. Wintergerst, M.; Dacheux, N.; Datcharry, F.; Herms, E.; Kapusta, B. J. Nucl. Mater. 2009, 393, 369–380. Griess, J. C.; Savage, H. C.; Mauny, T. H.; English, J. L. Effect of Heat Flux on the Corrosion of Aluminum by Water. Part I. Experimental Equipment and Preliminary Test Results; ORNL-2939; Oak Ridge National Laboratory: Oak Ridge, TN, Apr 1960. Griess, J. C.; Savage, H. C.; English, J. L. Effects of Heat Flux on the Corrosion of Aluminum by Water. Part IV. Tests Relative to the Advanced Test Reactor and Correlation with Previous Results; ORNL-3541; Oak Ridge National Laboratory: Oak Ridge, TN, Feb 1, 1964. Pawel, R. E.; Yoder, G. L.; Felde, D. K.; Montgomery, B. H.; McFee, M. T. Oxid. Met. 1991, 36(1/2), 175–194. Pawel, S. J.; Felde, D. K.; Pawel, R. E. Influence of Coolant pH on Corrosion of 6061 Aluminum Under Reactor Heat Transfer Conditions; ORNL/TM-13083; Oak Ridge National Laboratory: Oak Ridge, TN, Oct 1995. Carlson, P. A.; Curtiss, D. H.; Miller, N. R.; Van Wormer, F. W. Summary Status Report: Internal Corrosion of Ribbed Aluminum Process Tubes; HW-72728; Hanford Atomic Products Operation: Richland, WA, Feb 19, 1962. Kim, Y. S.; Hofman, G. L.; Robinson, A. B.; Snelgrove, J. L.; Hanan, N. J. Nucl. Mater. 2008, 378, 220–228. Leenaers, A.; et al. J. Nucl. Mater. 2004, 327, 121–129. Golosov, O. A. Corrosion of aluminum alloys in water at temperatures up to 100
C. Paper S6-P18 in RERTR 2009 I¨ 31st International Meeting on Reduced Enrichment for Research and Test Reactors, Beijing, China, Nov 1–5, 2009. Sindelar, R. L.; Lam, P. S.; Louthan, M. R., Jr.; Iyer, N. C. Mater. Char. 1999, 43, 147–157. Lam, P. S.; Sindelar, R. L.; Peacock, H. B., Jr. Vapor Corrosion of Aluminum Cladding Alloys and Aluminum–Uranium Fuel Materials in Storage Environments (U); WSRC-TR-97-0120; Westinghouse Savannah River Company: Aiken, SC, Apr 1997. Kumar, R.; Baheti, G. L.; Chacharkar, M. P.; Khatri, P. K. J. Radioanal. Nucl. Chem. 1989, 132(1), 3–9.

46.

47.

48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.

60. 61. 62.

63.

64. 65.

66. 67. 68.

69. 70.

Kim, Y. S.; Hofman, G. L.; Hanan, N. A.; Snelgrove, J. L. Prediction model for oxide thickness on aluminum alloy cladding during irradiation. In International RERTR Meeting, Chicago, IL, Oct 5–10, 2003. Farrell, K.; Byun, T. S.; Hashimoto, N. Mapping Flow Localization Processes in Deformation of Irradiated Reactor Structural Alloys – Final Report; ORNL/TM-2003/ 63; Oak Ridge National Laboratory: Oak Ridge, TN, Sept 2003. Farrell, K.; Houston, J. T. J. Nucl. Mater. 1979, 83, 57–66. Bishop, M.; Fletcher, K. E. Int. Metall. Rev. 1972, 17, 203–225. Smallman, R. E. Modern Physical Metallurgy; Butterworths: London, 1962; p 188. Hammer, B.; Jacobsen, K. W.; Milman, V.; Payne, M. C. J. Phys. Condens. Matter 1992, 4(50), 10453–10460. Wolfenden, A. Micron 1973, 4, 295–305. Peck, R. L.; Westmacott, K. H. Met. Sci. J. 1971, 5, 155. Bierlein, T. K.; Mastel, B. J. Appl. Phys. 1962, 33–39, 2873–2875. Risbet, A.; Levy, V. Philos. Mag. 1975, 31, 975–983. Risbet, A.; Brebec, G.; Lanore, J.-M. K.; Levy, V. J. Nucl. Mater. 1975, 56, 348–354. Farrell, K.; Wolfenden, A.; King, R. T. Radiat. Eff. 1971, 8, 107–114. Risbet, A.; Levy, V. J. Nucl. Mater. 1973, 46(3), 341–352. Jostsons, A.; Long, E. J., Jr.; Stiegler, J. O.; Farrell, K.; Braski, D. N. Annealing of Voids in Aluminum; ORNL-TM3494; Oak Ridge National Laboratory: Oak Ridge, TN, Oct 1971. Ells, C. E.; Evans, W. Trans. AIME 1963, 227, 437. Farrell, K.; Chickering, R. W.; Mansur, L. K. Philos. Mag. 1986, 53, 1–26. Gabriel, T. A.; Bishop, B. L.; Wiffen, F. W. Calculated Irradiation Response of Materials Using Fission Reactor (HFIR, ORR, and EBR-II) Neutron Spectra; ORNL/ TM-6361; Oak Ridge National Laboratory: Oak Ridge, TN, Aug 1979. Shiraishi, K.; Nagasaki, R. J. Nucl. Sci. Technol. 1965, 2–12, 499–505; Shiraishi, K.; Murata, T. J. Nucl. Sci. Technol. 1966, 3–11, 466–472; Shiraishi, K. J. Nucl. Sci. Technol. 1967, 4–3, 136–142; Shiraishi, K. J. Nucl. Sci. Technol. 1971, 8–5, 250–255. Smith, I. O.; Russell, B. J. Nucl. Mater. 1970, 35, 137; 1970, 37, 96; 1971, 38, 1. Farrell, K.; Houston, J. T. Combined Effects of Displacement Damage and High Gas Content in Aluminum, ORNL-TM-5395; Oak Ridge National Laboratory: Oak Ridge, TN, May 1976. Also available in Proceedings of International Conference on Radiation Effects and Tritium Technology for Fusion Reactors, Gatlinburg, TN, Oct. 1–3, 1975, U.S. Department of Commerce CONF-750989, Mar 1976; pp. II-209–II-233. Bauer, A. A.; Kangilaski, M. J. Nucl. Mater. 1972, 42, 91–95. Greenwood, L. R.; Kneff, D. W.; Skowronski, R. P.; Mann, F. M. J. Nucl. Mater. 1984, 123, 1002–1010. Hodgson, W. H. Final Report on Program for Using X-8001 Aluminum Alloy Cladding Material for Hanford Fuel Elements PT-IP-43-A-84-MT, IP-80-A-91-FP and IP-2-I-99-FP, HW-65754; General Electric Company: Richland, WA, July 22, 1960. Cunningham, J. E. Severe Radiation Damage to Aluminum Alloys; ORNL-TM-2138; Oak Ridge National Laboratory: Oak Ridge, TN, Mar 1968. Lotts, A. L.; et al. Analysis of Failure of HFIR Target Elements Irradiated in SRL and in HFIR – An Interim Status Report; ORNL-TM-2236; Oak Ridge National Laboratory: Oak Ridge, TN, Feb 1972.

Performance of Aluminum in Research Reactors 71.

72. 73. 74. 75. 76.

77. 78. 79.

80. 81.

82. 83. 84. 85. 86. 87. 88. 89.

Farrell, K. ORNL Studies of Radiation Damage in Aluminum and Its Alloys: A Status Report (December 1973); Intra-Laboratory Correspondence UCN-430 (3 5-61); Oak Ridge National Laboratory: Oak Ridge, TN, Jan 9, 1974. Farrell, K.; Stiegler, J. O.; Gehlbach, R. E. Metallography 1970, 3, 275–284. Jostsons, A.; King, R. T. Scripta Metall. 1972, 6, 447–452. King, R. T.; Jostsons, A. Metall. Trans. A 1975, 6A, 863–868. Farrell, K. J. Nucl. Mater. 1981, 97, 33–43. Lyjbrink, E.; van Grol, H. J.; Dekker, F.; van Witzenburg, W. In Effects of Radiation on Materials: Eleventh International Symposium, Philadelphia, PA 1982; ASTM STP 782, pp 765–778. Farrell, K.; Bentley, J.; Braski, D. N. Scripta Metall. 1977, 11, 243–248. Lide, D. R. Ed. CRC Handbook of Chemistry and Physics, 86th ed.; Taylor & Francis: Boca Raton, FL, 2005–2006. Farrell, K. In Proceedings of the Conference on Dimensional Stability and Mechanical Behaviour of Irradiated Metals and Alloys, Brighton, Apr 11–13; British Nuclear Energy Society: London, Vol. 1 1983; 73–76. Risbet, A.; Levy, V. J. Nucl. Mater. 1973, 46(3), 341–352. Yoshida, H.; Kozuka, T.; Sagane, T. Reactor irradiation effects on Al 1100. In Proceedings of the 24th Japan Congress on Materials Research – Metallic Materials, Kyoto University, Japan 1981; pp 1–6. van Witzenburg, W.; Mastenbroek, A. J. Nucl. Mater. 1985, 133&134, 553–557. Munitz, A. J. Nucl. Mater. 1989, 165, 305–312. Marchbanks, M. F. Advanced Neutron Source Materials Handbook; ORNL/M-4582; Oak Ridge National Laboratory: Oak Ridge, TN, Aug 1995. Farrell, K. Materials Selection for the HFIR Cold Neutron Source; ORNL/TM-99-208; Oak Ridge National Laboratory: Oak Ridge, TN, Aug 2001. Maddin, R.; Cottrell, A. H. Philos. Mag. Ser. 7 1955, 46(38), 735–743. Mori, T.; Meshii, M. Acta Metall. 1969, 17, 167–175. Bochirol, L.; Brauns, P.; Claudet, G. Prog. Refrigeration Sci. Technol. 1973, 1, 643–650. Weeks, J. R.; Czajkowski, C. J.; Tichler, P. R. In Effects of Radiation on Materials: 14th International Symposium;

90.

91.

92.

93. 94. 95.

96. 97.

98.

99. 100. 101. 102.

175

American Society for Testing and Materials: West Conshohocken, PA, 1990; Vol. II, ASTM STP 1046, pp 441–452. Farrell, K. A spectral effect on phase evolution in neutron-irradiated aluminum? In Microstructure of Irradiated Materials, The Materials Research Society Symposium Proceedings; 1995; Vol. 373, pp 165–170. Wallack, S. The Effect of Radiation on the Physical and Mechanical Properties of Metals and Alloys; WADC Technical Report 58-605, ASTIA Document No. AD 215540; Feb 1959. Gronbeck, H. D. ETR Radiation Damage Surveillance Programs, Progress Report II; IN-1036, Radiation Effects on Materials TID-4500; Idaho Nuclear Corporation Report; Feb 1967. Farrell, K.; Mahmood, S. T. Tensile properties of neutron irradiated aluminum alloys 2024 and 7075. Paper in preparation. Lohmann, W.; Ribbens, A.; Sommer, W. E.; Singh, B. N. Radiat. Eff. 1986, 101, 283–299. Singh, B. N.; Lohmann, W.; Ribbens, A.; Sommer, W. F. In Radiation-Induced Changes in Microstructure: 13th International Symposium (Part1); American Society for Testing and Materials: Philadelphia, PA, 1987; ASTM STP 955, pp 508–519. Singh, B. N.; et al. J. Nucl. Mater. 1986, 141–143, 743–747. Sommer, W. F.; Stubbins, J. F. Los Alamos National Laboratory Accelerator Production of Tritium Project B&R GB0508302: Topical Report Materials Safety Experiment; LANL Report LA-UR-93-2850; Aug 10, 1993. Dunlap, J. A.; Borden, M. J.; Sommer, W. F.; Stubbins, J. F. In Effects of Radiation on Materials: 17th International Symposium; American Society for Testing and Materials: West Conshohocken, PA, 1996; ASTM STP 1270, pp 1047–1056. Ismail, Z. H.; Mohammed, H. G. Scripta Metall. 1989, 23, 2067–2072. Ismail, Z. H. Radiat. Eff. Defects Solids 1990, 112, 105–110. Ismail, Z. H.; Birt, B. J. Nucl. Mater. 1995, 218, 289–292. Munitz, A.; Shtechman, A.; Cotler, C.; Talianker, M.; Dahan, S. J. Nucl. Mater. 1998, 252, 79–88.

5.08

Irradiation Assisted Stress Corrosion Cracking

P. L. Andresen GE Global Research Center, Schenectady, NY, USA

G. S. Was University of Michigan, Ann Arbor, MI, USA

ß 2012 Elsevier Ltd. All rights reserved.

5.08.1

Introduction

177

5.08.2 5.08.2.1 5.08.2.2 5.08.3 5.08.3.1 5.08.3.2 5.08.4 5.08.4.1 5.08.4.2 5.08.4.2.1 5.08.4.2.2 5.08.4.2.3 5.08.4.3 5.08.5 References

Irradiation Effects on SCC: Laboratory and Plant Data Individual Effects of Radiation on IASCC Service Experience Irradiation Effects on Water Chemistry Radiolysis and Its Effect on Corrosion Potential Effects of Corrosion Potential on IASCC Irradiation Effects on Microchemistry and Microstructure Radiation-Induced Segregation Microstructure, Radiation Hardening, and Deformation Irradiated microstructure Radiation hardening Deformation mode Radiation Creep and Stress Relaxation Summary

180 180 183 187 187 189 190 190 194 194 196 198 201 202 202

Abbreviations AES AGR BWR CT CW DPA FEGSTEM FWHM HWC HWR IASCC IGSCC NWC PWR RIS RH SCC SFE SGHWR SS SSRT STEM

Auger electron spectroscopy Advanced gas cooled-reactor Boiling water reactor Compact type (specimen) Cold work Displacements per atom Field emission gun TEM Full-width – half-max (for profiles) Hydrogen water chemistry (in BWRs) Heavy water reactor Irradiation-assisted stress corrosion cracking Intergranular stress corrosion cracking Normal water chemistry (in BWRs) Pressurized water reactor Radiation-induced segregation Radiation hardening Stress corrosion cracking Stacking fault energy Steam generating heavy water reactor Stainless steel Slow strain rate test Scanning transmission electron microscopy

TEM TG

Transmission electron microscopy Transgranular

5.08.1 Introduction Nuclear power accounts for about 17% of the world’s electricity production, and the rapid expansion in nuclear power throughout the world will necessitate that they operate with high reliability and safety. Stress corrosion cracking (SCC) has occurred in all water cooled reactors, including boiling-water reactors (BWRs) and pressurized-water reactors (PWRs), with a greater incidence in unirradiated, out-of-core components, especially between 1970 and 1990. As these materials, component designs, and water chemistries have improved, an increasing percentage of cracking problems has occurred in irradiated components. While irradiation-assisted stress corrosion cracking (IASCC) has been observed since early plant operation, increasing operating time and fluence has led to an increased incidence of cracking. Setting aside zircaloy fuel cladding and pressure vessel steels, most irradiated core components consist of austenitic stainless steels and nickel-base alloys 177

178

Irradiation Assisted Stress Corrosion Cracking

exposed to environments that span oxygenated to hydrogenated water at
270–340  C. The core of a nuclear reactor is an extreme environment consisting of high-temperature water, imposed stresses and strains, and an intense radiation field that affects the water chemistry, stress, and microstructure of the core materials (Figure 1). For background, the reader is referred to Chapter 1.03, RadiationInduced Effects on Microstructure; Chapter 1.05, Radiation-Induced Effects on Material Properties of Ceramics (Mechanical and Dimensional), and Chapter 1.07, Radiation Damage Using Ion Beams for more detailed treatments of radiation effects on materials. Initially, the affected components were primarily small components (bolts, springs, etc.) or components designed for replacement (fuel rods, control blades, or instrumentation tubes). However, in the last
20 years, IASCC has been observed in structural components (e.g., PWR baffle bolts and BWR core shrouds and top guides). Extensive literature exists for SCC under unirradiated conditions, and the basic factors and dependencies are well defined and reasonably well modeled for austenitic stainless steel and nickel alloys (e.g., Alloys 600 and its weld metals).1–9 A complete consensus on the underlying mechanism of cracking has not emerged although the well-behaved continuum in crack growth rate response versus material/ composition (including from stainless steels to nickel Solution renewal rate to crack-tip Stress

Δf Anionic transport

Oxide rupture rate at crack-tip

Environment

Microstructure

g-field Crack tip f [A]–, pH Passivation rate at crack-tip Grain boundary denudation

Hardening Relaxation

N-fluence Segregation

Figure 1 Schematic of the primary engineering parameters that effect stress corrosion cracking – stress, microstructure, and environment – and the underlying crack tip processes that control stress corrosion cracking. The primary ways in which radiation affects stress corrosion cracking is also shown: segregation, hardening, relaxation, and radiolysis. Radiolysis can increase the corrosion potential, which in turn increases the potential gradient (
’) and the crack tip potential ’, anion concentration [A], and pH.

alloys), water chemistry, temperature, and radiation suggests that a common crack growth mechanism is operative.8–13 Our understanding has evolved from the view that SCC occurs under very specific and unique conditions to the view that a continuum in response exists.8–13 With steady improvement in laboratory and plant detection of SCC, it is clear that SCC occurs under a wide range of conditions and also at a wide range of growth rates. Figures 2 and 3 show examples of the effect of environment (corrosion potential and water purity), material condition (sensitized vs. cold-worked), and stress (stress intensity factor) on SCC growth rates; the solid curves are the predicted response.8,9,12,13 SCC occurs even at low corrosion potential (Figure 2), and thus the behavior in BWRs and PWRs is linked, with the primary differences being dissolved H2, temperature, and the dissolved ion chemistry (B and Li are added to PWR primary water; see Chapter 5.02, Water Chemistry Control in LWRs).8–10,14 Of these, temperature has a universal effect, variations in dissolved H2 are particularly important in nickel alloys, and B/Li has little or no effect in deaerated water.9,14 Early plant (Figure 4) and laboratory (Figure 5) observations showed that the same basic dependencies existed for unirradiated and irradiated stainless steels, and that increasing fluence produces a well-behaved increase in SCC susceptibility (Figure 6). Figure 4 shows a strong effect of water purity for both unirradiated and irradiated BWR components, and Figure 5 shows a very similar response to corrosion potential to that in Figure 2. Thus, it was proposed that radiation enhances SCC primarily in four ways: segregation, hardening, relaxation, and radiolysis (Figure 1). The neutron fluence where these processes have an effect is shown in Figure 7, along with the current end-of-life fluence for various BWR and PWR components. The primary radiation effects on materials operate in a similar range of fluences, and thus their individual contributions can be difficult to distinguish. An example of their interaction in altering SCC growth rate is shown in the prediction of cracking of a weld in a BWR core shroud (Figure 8) in which the individual effects are plotted along with the resulting crack length versus time. While many of the enhancements in SCC susceptibility from irradiation dose (neutron fluence) have been well established, it remains possible that additional factors will emerge at high fluences (e.g., >30 displacements per atom (dpa)). Intergranular (IG) SCC is promoted in austenitic stainless steels above a ‘threshold’ fluence

179

Irradiation Assisted Stress Corrosion Cracking

1.0E–05 1

25 mm CT specimen Furnace sensitized; 15 C cm–2 288 ⬚C water ; 0.1–0.3 mS cm–1 Constant load ; 25 Ksi in1/2

6

Crack growth rate (mm s–1)

42.5 μin h–1

10–7

10

8

1.0E–06

11

Crack growth rate (mm s–1)

10–6

Sensitized 304 stainless steel 30 MPa m1/2, 288 ⬚C water 0.06–0.4 μS cm–1, 0–25 ppb SO4 filled triangle = constant load open squares = ‘gentle’ cyclic

5

14

14.2 μin h–1 Theoretical curves

ααα 9

μS cm–1

0.3 0.2

0.1

10–8

2

200 ppb O2 500 ppb O2 2000 ppb O2

304 Stainless steel

3 7 4

Screened round robin data - highest quality data - corrected corr. potential - growth rates corrected to 30 MPa m1/2

42.5 28.3 14.2 –1 μin h

1.0E–07 GE pledge predictions 30 MPa m1/2

0.5 2000 ppb O2 Ann. 304SS 200 ppb O2

0.25 1.0E–08 0.1

12

0.06 μS cm–1

Hydrogen water chemistry β

Normal water chemistry (ex-core)

–600

–400 –200 0 +200 Corrosion potential (mVshe)

+400

30 MPa m1/2 1.0E–09 –0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 Corrosion potential (Vshe)

1.0E–05 Sensitized 304 stainless steel 30 MPa m1/2, 288 ⬚C water 0.06–0.4 μS cm–1, 0–25 ppb SO4 SKI round robin data filled triangle = constant load open squares = ‘gentle’ cyclic

200 ppb O2 500 ppb O2 2000 ppb O2

β

10–9

–1

0.06 μS cm Industry mean

0.2

0.3

0.4

4 dpa 304SS

Crack growth rate (mm s–1)

1.0E–06 316L (A14128, square) 304L (Grand gulf, circle) nonsensitized SS 50% RA 140 C (black) 10% RA 140 C (gray) 1.0E–07

20% CW A600 42.5 28.3 14.2 μin h–1

20% CW A600 GE pledge predictions 30 MPa m1/2 Sens SS

0.5 2000 ppb O2 Ann. 304SS 200 ppb O2

0.25

1.0E–08

–1

0.1

0.1 μS cm Means from analysis of 120 lit. sens SS data

0.06 μS cm–1

0.06 μS cm–1

GE pledge predictions for Unsens. SS (upper curve for 20% CW)

1.0E–09 –0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 Corrosion potential (Vshe)

0.2

0.3

0.4

Figure 2 Stress corrosion cracking growth rate versus corrosion potential for stainless steels tested in high-purity water at 288  C containing 2000 ppb O2 and 95–3000 ppb H2. Dissolved O2 strongly influences corrosion potential, which in turn affects crack chemistry and growth rate of sensitized stainless steels (two graphs at left) as well as cold-worked stainless steels and Alloy 600 (large rectangular symbols on right graph) and irradiated stainless steel (large triangular symbols). Cold-worked or irradiated materials have an elevated yield strength, which causes an increase in growth rate at both low and high potential. RA, Reduction in area; CW, Cold work.

180

Irradiation Assisted Stress Corrosion Cracking

10–5

Stress intensity (ksi in1/2) 6 8 10 20 30 40 60 80

4

10–3 Sens. 304 stainless steel 288 ⬚C water

10–4

10–7

NRC disposition line

10–5

* 10–8

*

Theory 15 C cm–2, –50 mVshe 0.5 ms cm–1 Theory –2 15 C cm , –50 mVshe 0.2 ms cm–1

10–9

10–6

*

Theory 15 C cm–2, –200 ® –500 mVshe 0.2 ms cm–1

10–10

4

6

Crack growth rate (in h–1)

Crack growth rate (mm s–1)

10–6

10–7

8 10 20 30 40 60 80 Stress intensity (MPa m1/2)

Figure 3 Effect of stress intensity factor on stress corrosion cracking growth rate for sensitized stainless steel exposed in various water chemistries at 288  C.

(Figures 5 and 6). This occurs in oxygenated (e.g., BWR) water above 2–5  1020 n cm–2 (E > 1 MeV), which corresponds to about 0.3–0.7 dpa, and depends on the stress, water chemistry (especially, sulfate and chloride), and other factors. Attempts to reproduce the same level of IG cracking in inert environments have been unsuccessful, confirming that it is an environmental cracking phenomenon, not simply a change in the mechanical properties and response of the material in an inert environment. Cracking in hydrogenated water (i.e., BWR hydrogen water chemistry (HWC) or PWR water) is typically observed at roughly a 4 higher fluence than in oxidizing water, with IASCC enhanced at elevated temperature (Figure 9). For both BWR and PWR conditions, the same basic dependencies exist for unirradiated and irradiated materials. It is important to distinguish the results of different kinds of SCC testing. Crack-growth testing typically uses fracture mechanics specimens, commonly a compact type (CT) specimen. It has the significant

advantage of continuous, online monitoring of crack length versus time, usually by employing an electric current, potential drop technique. It can provide a resolution of about 1 mm and can accurately characterize the inherent resistance to crack advance (i.e., beyond the ill-defined initiation stage) as well as the dependencies on corrosion potential, stress intensity factor, etc. Smooth specimen tests, whether by constant load, constant displacement, or slow strain rate (SSR), are simpler tests to perform, but represent some combination of initiation and growth. SSR tests impose failure and can overstate or understate the SCC susceptibility. Constant load or displacement tests generally require periodic interruption for examination, and ‘initiation’ can be microscopic cracks or complete failure. Work over the last 25 years has enabled many aspects of IASCC phenomenology to be explained and predicted on the basis of the experience with intergranular stress corrosion cracking (IGSCC) of nonirradiated stainless steel in high-temperature water environments. This continuum approach has successfully accounted for radiation effects on water chemistry and its influence on electrochemical corrosion potential. However, all radiation-induced microstructural and microchemical changes that promote IASCC are neither fully known nor fully reproducible in similar materials. Well-controlled data from well-characterized irradiated materials remain insufficient due to the inherent experimental difficulties and financial limitations. Many of the important metallurgical, mechanical, and environmental aspects that are believed to play a role in the cracking process are illustrated in Figure 1. Only persistent material changes are required for IASCC to occur, but in-core processes such as radiation creep and radiolysis also have an important effect on IASCC.

5.08.2 Irradiation Effects on SCC: Laboratory and Plant Data 5.08.2.1 Individual Effects of Radiation on IASCC IASCC can be categorized into radiation effects on the water chemistry (radiolysis) and on the material/stress, and the accepted definition of IASCC encompasses cases where either factor is dominant (low-fluence materials tested in water undergoing radiolysis, or preirradiated materials tested without an active radiation flux). Radiation dose rate in rads/h is often used in radiolysis, and in neutrons per square centimeter (n cm–2) or displacements per atom in materials. In light water reactors (LWRs), 1 dpa corresponds

Frequency of SCC initiation increases dramatically with increasing conductivity

0.2

Best fit

1.2 1.0 0.8 0.6

Threshold conductivity for SCC initiation increases as level of sensitization decreases Low carbon SS High carbon SS

Nonsensitized low carbon SS

0.2 0 0

0.2

(b)

1.0 0.9

Sensitized high carbon SS

0.4

0.6

Plant average conductivity (ms cm–1)

0.4

0.6

Plant average conductivity (ms cm–1)

*

0.8

Frequency of IGSCC initiation increases with plant conductivity

0.7 0.6 0.5

*

0.4

* Experienced substantial high conductivity excursions not reflected in average value

0.3 0.2

*

0.1 0

(c)

181

1.4 Upper bound

0.4

% with IGSCC/on-line months

(a)

1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0

% with IGSCC/on-line months

% with IGSCC/on-line month

Irradiation Assisted Stress Corrosion Cracking

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Plant average conductivity (ms cm–1)

Figure 4 The effects of average plant water purity shown in field correlations of the core component cracking behavior for (a) stainless steel intermediate and source range monitor dry tubes, (b) creviced stainless steel safe ends, and (c) creviced Inconel 600 shroud head bolts, which also shows the predicted response versus conductivity. Adapted from Brown, K. S.; Gordon, G. M. In Proceedings of 3rd Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): New York, NY, 1988; pp 243–248.

to
7  1020 n cm–2 when counting neutrons with E > 1 MeV, or
1.5  1021 n cm–2 for E > 0.1 MeV. The primary effects of radiation on materials8,15–22 include microcompositional effects (grain boundary chemistry) and microstructural changes (formation of dislocation loops, voids, precipitates, and the resulting changes hardening and deformation mode). In terms of their effect on IASCC, the primary effects of radiation are the following:  Radiolysis of water, in which a variety of short- and long-lived radicals and species are produced. There is no evidence that the specific species formed are important, and indeed their effect on cracking appears to be fully captured by their overall effect on the corrosion potential of the material.  Radiation-induced segregation (RIS), which produces an enrichment in some species (e.g., Ni and Si) at grain boundaries and other defect sinks, and a depletion in other species (e.g., Cr). Even though the distance over which RIS occurs is very limited (a few

nanometers), studies of unirradiated materials have shown that the narrow profiles can affect SCC.23,24  Radiation hardening (RH), which results from radiation damage and the creation of vacancy and interstitial loops, which impede dislocation motion. Once a few dislocations move along a given slip plane, they clear the ‘channel’ of most of these barriers, and subsequent dislocation occurs primarily in these channels.  Radiation creep relaxation, which reduces constant displacement stresses such as in bolts or associated with weld residual stress. During active irradiation, radiation creep can promote dynamic strain, and thereby SCC.  Swelling, which occurs to a limited extent at temperatures above
300  C, but can be sufficient to produce reloading of components such as PWR baffle former bolts. Swelling occurs differently in different materials, and is delayed in cold-worked materials. Stresses due to swelling are balanced by

182

Irradiation Assisted Stress Corrosion Cracking

Data of Jacobs( ) and Kodama( ) Postirrad. SSRT 2–3 ⫻ 10–7 s–1 288 ⬚C Comm. purity 304SS( ) and 316SS( ) 42 ppm O2-sat’d vs. 0.02 ppm O2

100

80

»2

100

40

»42

ppm O2

60 %IGSCC

% IGSCC fracture

Data shifted right by Init. grain boundary Cr enrichment

»0.2 0.02 –0.05

fC, Vshe

20 –0.4

0 1019

–0.2

0

0.2

1020 1021 Neutron fluence (n cm–2 ) (E > 1 MeV)

1022

Percentage of spot welds inspected with IGSCC

Figure 5 Dependence of irradiation-assisted stress corrosion cracking on fast neutron fluence as measured in slow strain rate tests at 3.7  107 s1 on preirradiated type 304 stainless steel in 288  C water. The effect of corrosion potential via changes in dissolved oxygen is shown at a fluence of 2  1021 n cm–2. Reproduced from Jacobs, A. J.; Hale, D. A.; Siegler, M. Unpublished Data on SCC of Irradiated SS in 288  C Water and Inert Gas; GE Nuclear Energy: San Jose, CA, 1986. SSRT, Slow strain rate test. PWR control BWR core BWR end rod failures (IASCC) component of life failures (IASCC)

100

80

1020

BWR creviced control blade sheath 60 Threshold fluence for IGSCC ≈5 ⫻ 1020 n cm–2

40

0.1

0 0

0.2

0.4

0.6

0.8

1

1.2

1.4

Neutron fluence (n cm–2 ⫻ 1021) (E > 1 MeV)

Figure 6 Dependence of irradiation-assisted stress corrosion cracking on fast neutron fluence for creviced control blade sheath in high-conductivity boiling-water reactors. Reproduced from Gordon, G. M.; Brown, K. S. In Proceedings of 4th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1990; pp 14-46–14-62.

radiation creep relaxation, but the resulting stress can be sufficient to cause IASCC.  Other microstructural changes, such as precipitation or dissolution of phases in materials. While there is no clear evidence that such changes affect IASCC response, this may only reflect the limited

PWR end of life PWR life extension

1021 1022 1023 Neutron fluence (n cm–2) (E >1 MeV) Irradiation dose (dpa) 1 10 Significant changes in grain boundary composition, alloy strength, and ductility

20

PWR baffle bolt failures (IASCC)

100 Onset of significant void swelling and possible embrittlement

Figure 7 Neutron fluence effects on irradiation-assisted stress corrosion cracking susceptibility of type 304SS in boiling-water reactor environments. Reproduced from Bruemmer, S. M.; Simonen, E. P.; Scott, P. M.; Andresen, P. L.; Was, G. S.; Nelson, J. L. J. Nucl. Mater. 1999, 274, 299–314.

characterization and IASCC studies that has been performed on high fluence materials. The individual (segregation, hardening, and creep, Figure 10) and composite (SCC, Figures 5–9) effects of radiation increase with dose in much the same manner, which makes the isolation of, and attribution to, individual contributions difficult. The dislocation loop microstructure is closely tied to radiation

Irradiation Assisted Stress Corrosion Cracking

30

20

30 25

Effect of rad segregation

20

Effect of rad hardening

15 10

Depth 10

Stress intensity (K)

5

Effect of stress relaxation 0 0

100

200 300 Time (month)

Fraction of IG cracking area

40

In PWR primary water CW316, 340 ⬚C CW316, 325 ⬚C CW316, 290 ⬚C Type 304, 325 ⬚C

0

*

* * 1.0E + 20

20 Loop line length SCC

15

10

1.0E + 21 1.0E + 22 Fluence (n cm–2) (E > 0.1 MeV)

CP-304SS 3.2 MeV protons 360 ⬚C

0 0

1

2

3 Dose (dpa)

4

5

6

Figure 10 Schematic diagram showing the dose dependence of key irradiated microstructure features (radiation-induced segregation and dislocation microstructure) and radiation hardening along with stress corrosion cracking susceptibility. Reproduced from Was, G. S. In Proceedings of 11th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; American Nuclear Society (ANS): La Grange Park, IL, 2004; pp 965–985.

irradiation. Subsequent sections focus on the possible mechanisms by which water chemistry, RIS, microstructure, hardening, deformation mode, and irradiation creep – individually or on concert – may affect IASCC.

*

* Specimen broke at the pin hole

20

Cr Cr depletion

5

100

60

25

0 500

400

Figure 8 Predicted effect of radiation segregation, radiation hardening, and radiation creep relaxation on a boiling-water reactor core shroud, where the through-wall weld residual stress profile is the primary source of stress. Less aggressive water chemistry (corrosion potential and water purity) would result in less crack advance early in life, which would give a greater opportunity for radiation creep relaxation. The leak depth is the wall thickness of the shroud. While radiation hardening continues to increase the yield strength, its effect on crack growth is reduced (see Figure 21(a)). EPR, Electrochemical potentiokinetic repassivation (a test for sensitization).

80

Hardness

Arbitrary units

Crack depth (mm)

Leak depth

38.1-mm thick 304SS, two-sided weld 0.75 Vshe, 0.15 μS cm–1 EPR0 = 10.8 C cm–2 (0.050% C) Flux = 3 ⫻ 1019 n cm–2 year

K, segregation, hardening, or relaxation

40

30

183

1.0E + 23

Figure 9 Percentage intergranular stress corrosion cracking versus fluence for cold-worked type 316 stainless steel tested at various temperatures. Despite the low potential environment of pressurized-water reactor primary water, at high fluence (especially at higher temperature) there is significant susceptibility to stress corrosion cracking.

hardening and both increase with dose until saturation occurs by
5 dpa. RIS also increases with dose and tends to saturate by
5 dpa. Although dependent on both metallurgical and environmental parameters, IASCC generally occurs at doses between 0.5 dpa (for BWRs) and 2–3 dpa (for PWRs), which encompasses the steeply rising portion of the curves in Figure 4 that describe the changes in materials properties with

5.08.2.2

Service Experience

The IASCC service experience extends over 50 years, and the early observations projected an accurate view of the important characteristics and dependencies, and pointed to the ‘assisted’ nature of radiation in enhancing SCC. As with other forms of SCC, early observations suggested that a growing incidence with time and neutron fluence should be expected. IASCC was first reported in the early 1960s7,8,15–22,25–36 and involved intergranular cracking of stainless steel fuel cladding, with the initiation of multiple cracks occurring from the water side. By contrast, mostly ductile, transgranular cracking was observed in postirradiation mechanical tests performed in inert environments at various strain rates and temperatures. Grain boundary carbide

184

Irradiation Assisted Stress Corrosion Cracking

precipitation was rarely observed although preexisting thermal sensitization was present in some cases. A correlation between time-to-failure and stress level was reported, with failure occurring first in thinwalled rods with small fuel-to-cladding gaps, where fuel swelling strains were the largest. The highest incidence of cracking occurred in peak heat flux regions, corresponding to the highest fluence and the greatest fuel–cladding interaction (highest stresses and strains.) Similar stainless steel cladding in PWR service exhibited fewer instances of intergranular failure. At that time, off-chemistry conditions or stress rupture were often considered to be the cause of PWR failures. In the last 30þ years, a growing number of other stainless steel (and nickel alloy) core components have exhibited IASCC, including neutron source holders in 1976 and control rod absorber tubes in 1978. Instrument dry tubes and control blade handles and sheaths, Figure 5, which are subject to very low stresses, are also cracked, generally in creviced locations and at higher fluences.15,18,20,37,38 These initial failures in the most susceptible components were

Table 1

followed by more numerous incidents of IASCC in the past 20 years, perhaps most notably in BWR core shrouds8,15,16,39 and PWR baffle bolts.19,40,41 A summary of reported failures of reactor internal components is listed in Table 1 and demonstrates that IASCC is not confined to a particular reactor design, material, component, or water chemistry. For example, stainless steel fuel cladding failures were reported years ago in commercial PWRs and in PWR test reactors.15,22,25,29–36,42 At the West Milton PWR test loop, intergranular failure of vacuum-annealed type 304 stainless steel fuel cladding was observed31 in 316  C ammoniated water (pH 10) when the cladding was stressed above yield. Similarly, IASCC was observed in creviced stainless steel fuel element ferrules in the Winfrith steam generating heavy water reactor (SGHWR),43,44 a 100 MWe plant in which light water is boiled in pressure tubes, where the coolant chemistry is similar to other boiling-water reactor designs. The 20% Cr–25% Ni–Nb stainless steel differs from type 304 primarily in Ni and Nb content, as well as in its lower sulfur (0.006%) and phosphorus (0.005%) contents. The ferrules were designed for

Irradiation-assisted stress corrosion cracking service experience

Component

Material

Reactor type

Possible sources of stress

Fuel cladding Fuel cladding Fuel claddinga Fuel cladding ferrules Neutron source holders Instrument dry tubes Control rod absorber tubes Fuel bundle cap screws Control rod follower rivets Control blade handle Control blade sheath Control blades Plate type control blade Various boltsb Steam separator dryer boltsb Shroud head boltsb Various bolts Guide tube support pins Jet pump beams Various springs Various springs Baffle former bolts Core shroud Top guide

304SS 304SS 20% Cr–25% Ni–Nb 20% Cr–25% Ni–Nb 304SS 304SS 304/304L/316L SS 304SS 304SS 304SS 304SS 304SS 304SS A-286 A-286 600 X-750 X-750 X-750 X-750 718 316SS cold work 304/316/347/L SS 304SS

BWR PWR AGR SGHWR BWR BWR BWR BWR BWR BWR BWR PWR BWR PWR and BWR BWR BWR BWR and PWR PWR BWR BWR and PWR PWR PWR BWR BWR

Fuel swelling Fuel swelling Fuel swelling Fabrication Welding and Be swelling Fabrication B4C swelling Fabrication Fabrication Low stress Low stress Low stress Low stress Service Service Service Service Service Service Service Service Torque, differential swelling Weld residual stress Low stress (bending)

a

Cracking in AGR fuel occurred during storage in spent fuel pond. Cracking of core internal occurs away from high neutron and gamma fluxes. AGR, Advanced gas-cooled reactor b

Irradiation Assisted Stress Corrosion Cracking

a 5-year exposure during which the peak fast neutron flux is 2–3  1013 n cm–2 s (E > 1.5 MeV), yielding a peak fluence over 5 years of 3–5  1021 n cm–2. The similarity of IASCC in BWRs and PWRs was also noted in swelling tube tests performed in the core44,45 on a variety of commercial and high-purity heats of types 304, 316, and 348 stainless steel and Alloys X-750, 718, and 625. Swelling was controlled by varying the mix of Al2O3 and B4C within the tubes; the latter swells as neutrons transmute B to He. Nominally identical strings of specimens were inserted into the core in place of fuel rods. Historically, the oxidizing potential in a PWR core is lower than in BWRs, but in the last decade, most BWRs employ an electrocatalytic technology called NobleChem™(46–48) to create a low corrosion potential. Some early investigators attributed PWR cracking to low ductility stress rupture (of course, this mechanism would apply equally to BWRs). A few laboratory studies reported small amounts of intergranular cracking of irradiated stainless steels in SSR tests in 300  C inert environments49 although in many related experiments50,51 no intergranular failure was found. Small amounts of intergranular cracking in inert tensile tests are not surprising, and since the early 1990s, the plant and laboratory IASCC data show that cracking is environmentally assisted and follows a well-behaved continuum in response over ranges in fluence, corrosion potential, temperature, stress, etc.8,15–17 Factors that distinguish PWRs from BWRs include their higher operating temperature, 10 higher maximum neutron fluence in core structural components, higher hydrogen fugacity, and borated– lithiated water chemistry (including the possibility of localized boiling and thermal concentration cells in crevices from gamma heating, which could lead to aggressive local chemistries). The possible role of RIS of Si may be especially important in accounting for the limited difference in SCC response at high potential (BWR) versus low potential (PWR) at high fluence.52,53 Brown and Gordon37,38 (Figure 4) accumulated and analyzed data for cracking in Alloy 600 shroud head bolts (first observed in 1986) as well as stainless steel safe ends (first observed in 1984) and in-core instrumentation tubes (first observed in 1984) with a focus on components that were creviced, a factor known to exacerbate cracking.7,37,38 The highest radiation exposure occurred for the intermediate range and source range monitor (IRM/SRM) dry tubes, which contain flux monitors housed in thin-walled,

185

annealed stainless steel tubes. Cracking initiated in the crevice between the spring housing tube and the guide plug at fluences between 0.5 and 1.0  1022 n cm–2 (E > 1 MeV). Wedging stresses from the thick oxide observed in the crevice were implicated, since other (applied and residual) stresses were negligible. The primary variable from plant to plant is the average coolant conductivity, which correlates strongly with cracking incidence (Figure 4(a)–4(c)). Each point in Figure 4 represents inspection results for one BWR plant, and data are normalized using reactor operating time (i.e., percentage of components with intergranular cracks divided by the online exposure time). The scatter in Figure 4(a) was attributed to variations in fluence and specific ion chemistry, as well as limitations in the resolution of underwater visual inspection. Scatter can also result from short-term excursions in conductivity, which is not adequately reflected in the average, as identified in Figure 4(c). Correlations between IASCC and conductivity were also reported for cracking in shroud head bolts (Figure 4(c)) and creviced safe ends (Figure 4(b)). The strong influence of conductivity on cracking of stainless steel has also been shown in laboratory tests and plant recirculation piping, where predictive modeling8,12,13,15,54,55 has been compared to field data on the operational time required to achieve a detectable crack (typically, 10% of the wall thickness). Preliminary prediction of the shroud head bolt cracking response8 also provides reasonable agreement with observation (Figure 4(c)). High-strength, nickel-base alloy components15,22 have also experienced IASCC (Table 1), with many incidents in lower radiation flux regions (e.g., where the end-of-life fluence is below 5  1019 n cm–2) such as cracking of Inconel X-750 jet pump beams in BWRs. Inconel X-750 cracking has also occurred extensively in PWR fuel hold-down springs, which attain an end-of-life fluence of 1–10  1021 n cm–2; it is proposed that cracking has been aggravated by vibrational stresses (corrosion fatigue). The effects of irradiation on IASCC in high-strength, precipitationhardened nickel-base alloy components as well as in stainless steels have not been characterized. BWR core shrouds8,15,16,39 and PWR baffle bolts40,41 are the two most common examples of IASCC although susceptibility clearly exists in other areas, such as control blade components, fuel components, the BWR top guide, etc. SCC in the BWR core shroud occurs almost exclusively near the welds (both circumferential and vertical), and

186

Irradiation Assisted Stress Corrosion Cracking

initiation is observed from both the inside (ID) and outside (OD) surfaces (the shroud separates the upward core flow from the downward recirculation flow that occurs in the annulus between the shroud and the pressure vessel). This large-diameter welded ‘pipe’ has little active (DP) loading, and its susceptibility arises primarily from weld residual stresses and weld shrinkage strains.56–58 Cracking is observed in both low fluence and moderate fluence areas, and the extent of the enhancement in SCC susceptibility by irradiation is limited because, while RH and RIS occur, radiation creep also relaxes the weld residual stress. SCC predictions for a BWR core shroud that account for the damaging effect of RIS and RH and the beneficial effect of radiation creep relaxation are shown in Figure 8 and illustrate the complexity of the interactions of these phenomena in the evolution of cracking. Predictions also indicate that, if SCC does not nucleate early in life (e.g., below 0.5 dpa), for example, from high coolant impurity levels or severe surface grinding, susceptibility tends to decrease with fluence in the shroud welds because of radiation-induced creep relaxation (although many shroud welds are in very low flux areas). The last decade has also seen extensive failures of PWR baffle bolts40,41 although large plant-to-plant and heat-to-heat differences are observed. Most baffle bolts are fabricated from type 316 stainless steel cold-worked to 15% to increase their yield strength. The complex baffle former structure exists in a PWR because their fuel does not have a surrounding ‘channel,’ so the baffle former structure must conform closely to the geometry of the fuel to provide well-distributed water flow. The baffle former plates are usually made from annealed material, typically type 304 stainless steel. Because of their proximity to the fuel, very high fluences can develop – up to
80 by the end of the original design life. The high gamma flux produces significant heating in the components, in some instances estimated at þ40  C, especially in designs where the PWR coolant does not have good access to the bolt shank. While the heat-to-heat variations are not understood, it is clear that plants that load-follow (and therefore undergo power level changes and thermal cycles) are much more prone to baffle bolt cracking. Another aggravant is the thermal gradient and possible boiling (resulting in altered water chemistry) in the shank area of the bolt if the coolant access is restricted. However, primary factors must be the very sizeable stress relaxation that occurs early in life (e.g., during the first

5 dpa), followed by preferential radiation swelling of the annealed baffle plates over the cold-worked baffle bolts, which will cause reloading. The dynamic equilibrium between swelling and radiation creep, which determines the ‘reloading’ stress in the bolt, is likely a complex function of many parameters, including local neutron flux, temperature, baffle plate geometry, and composition. The number of IASCC incidents continues to grow, and there can be no question that many LWR components are susceptible. Strategies to mitigate IASCC (e.g., NobleChem™(46–48)) and manage IASCC (e.g., by showing some IASCC could be tolerated, installing mechanical restraints to mitigate the impact of IASCC in BWR shrouds, or selectively inspect and replace baffle bolts) have been successful. IASCC field experience has led to the following trends and correlations:  Intergranular cracks associated with radiation effects on solution-annealed stainless steel were once thought to occur only at fluences above 0.3  1021 n cm–2. But significant intergranular cracking in BWR core shrouds (which do not have thermal sensitization) occurs over a broad range of fluences, showing that a true fluence threshold does not exist.15,16 The observations of SCC in unirradiated, unsensitized stainless steel (with or without cold work) also undermine the concept of a threshold fluence below which no SCC occurs. This also holds for thresholds in corrosion potential, water impurities, temperature, etc.8,54,59,60  SCC susceptibility is affected by fluence in a complex fashion. SCC in BWR shrouds and PWR baffle bolts does not always correlate strongly with fluence, and one important reason for this is that radiation creep produces relaxation of the stresses from welding and in bolts. The need to account for many changing factors is necessary in interpreting and predicting SCC.  Most early incidents involved high stresses or dynamic strains, but cracking has since been observed at quite low stresses at high fluences and longer operating exposure. Laboratory and field data indicate that IASCC occurs at stresses below 20% of the irradiated yield stress, and at stress intensities below 10 MPa m1/2.  Extensive laboratory and field data show that corrosion potential is a very important parameter, with its effect being consistent from low to high fluence, except in some high fluence materials and/or

Irradiation Assisted Stress Corrosion Cracking





  



under high stress intensity factor conditions. Materials prone to high radiation-induced changes in grain boundary Si level may exhibit a very limited effect of corrosion potential.52,53 There is no evidence of threshold potential, and indeed irradiated materials exhibit IASCC in deaerated water. Impurities, especially chloride and sulfate, strongly affect IASCC in BWR water (Figure 4). As noted by Brown and Gordon,37 this correlation applies equally to low and high flux regions and to stainless steels (Figure 4(a) and 4(b)) and nickel-base alloys (Figure 4(b)). Indeed, the correlation closely parallels that from out of core.8,12,13,15,38,54,55 At higher levels, the same impurities can affect SCC in PWRs. Similarly, if high corrosion potential conditions form in the PWR primary (where B and Li are present), high growth rates can result.9,14 Crevices exacerbate cracking primarily due to their ability to create a more aggressive crevice chemistry from the gradient in corrosion potential (in BWRs) or in temperature (most relevant to PWRs). Crevices can also produce stress and strain concentration. Cold-working often exacerbates cracking (especially abusive surface grinding), although it can also delay the onset of some radiation effects. Temperature has an important effect on IASCC, enhancing both crack initiation and growth rate. Preexisting grain-boundary carbides or chromium depletion is not required for susceptibility although furnace-sensitized stainless steels are clearly highly susceptible to cracking in-core. Cr depletion will develop or be magnified by irradiation, and increase IASCC susceptibility, although its effect is most pronounced in pH-shifted environments, as can develop when potential or thermal gradients exist. The role of N, S, P, and other grainboundary segregants is less clear. IASCC is enhanced at a fluence that is dependent on applied stress and strain, corrosion potential, solution conductivity, crevice geometry, etc. At sufficiently high conductivities, cracking has been observed in solution-annealed stainless steel in the field (Figure 6(a) and 6(b))37 and in the laboratory.8,12,13,15,54,55 Thus, while convenient in a practical engineering sense, the concept of a ‘threshold’ fluence (or stress, corrosion potential, etc.) is scientifically misleading8,52,59,60; cracking susceptibility and morphology are properly considered an interdependent continuum over many relevant parameters.

187

5.08.3 Irradiation Effects on Water Chemistry 5.08.3.1 Radiolysis and Its Effect on Corrosion Potential SCC susceptibility is fundamentally influenced by corrosion potential, not by the oxidant and reductant concentrations per se.8,15,61 The lower H2 concentration in BWRs distinguishes BWRs from PWRs because it permits the radiolytic formation of oxidants. Above 500 ppb (5.6 cm3 kg1) H2, radiolytic formation of oxidants is effectively suppressed, and the corrosion potential remains close to its thermodynamic minimum (which is a function of temperature, H2 fugacity, and pH). BWRs cannot achieve this H2 level in the core because H2 partitions to the steam phase, which begins to form about a quarter of the way up the fuel rods. Thus, this section is primarily relevant to BWRs. Radiolysis and the presence of oxidizing species require that many sequential and nonlinear dependencies that must be considered, for example, radiation flux produces oxidizing and reducing species, the corrosion potential is controlled by multiple reactions of these species, the crack chemistry is nonlinearly dependent on the corrosion potential and the dissolved ions present, and the SCC growth rate is a nonlinear function of the crack chemistry.8,61 The relationship between radiation flux and SCC cannot easily be determined empirically, but rather requires a fundamental understanding of each subprocess. Water is decomposed by ionizing radiation into various primary species62–65 including both radicals (e.g., e aq, H, OH, HO2) and molecules (e.g., H2O2, H2, O2), which can be oxidizing (e.g., O2, H2O2, HO2) or reducing (e.g., e aq, H, H2). The predominant species that are stable after a few seconds are H2O2 and H2, with O2 forming primarily from the decomposition of H2O2. Because H2 partitions to the steam phase and H2O2 is not volatile, 87% of the water that is recirculated in a BWR (11–14% of the core flow becomes steam) is oxidant rich. H2 is introduced in the feed water, which mixes with the recirculated water near the top of the annulus (the region of downflow between the core shroud and pressure vessel). The concentrations of radiolytic species are roughly proportional to the square root of the radiation flux in pure water. The radiation energy versus intensity spectrum influences the concentration of

Irradiation Assisted Stress Corrosion Cracking

0.4 0.3 Corrosion potential (VSHE)

each radiolytic species, which is described in terms of a yield, or G value (molecules produced per 100 eV absorbed by water). In LWRs, the G values for most species are within a factor of 3 for fast neutron versus gamma radiation. Despite this similarity, the influence of fast neutron radiation is much stronger than gamma radiation primarily because the energy deposition rate, or mean linear energy transfer (LET), is greater (40 eV nm–1 for fast neutrons versus 0.01 eV nm–1 for gamma radiation63). Also, the neutron flux in LWRs (e.g., 1.03  109 Rad h–1 core average and 1.68  109 Rad h–1 peak in a BWR4 of 51 W cm–3 power density) is also higher than the gamma flux (0.34  109 Rad h–1). Indeed, the moderate gamma levels present in the downcomer in the outside annulus of a BWR core actually promote recombination of hydrogen and various oxidants.62,65 This is a key element in the effectiveness of HWC, and BWRs that have a wider annulus and lower gamma flux near the pressure vessel respond less to a given H2 addition. The contribution of thermal neutrons and beta particles to radiolysis is small in LWRs. As in many electrochemical processes, the integrated effects of various oxidants and reductants on environmental cracking is best described via changes in corrosion potential, which controls the thermodynamics and influences the kinetics of most reactions. Since electrochemical potentials are logarithmically dependent on local oxidant, reductant, and ionic concentrations (via the Nernst relationship, ’ ¼ ’o þ RT/nF ln [products/reactants], where ’ is electrochemical potential, R is the gas constant, T is temperature in kelvin, n is the number of moles, and F is Faraday’s constant), radiation-induced increases in concentration of various species by many orders of magnitude may have comparatively small effects on the corrosion potential in hot water. Further, corrosion potentials are mixed potentials involving a balance of anodic and cathodic reactions on the metal surface. At low oxidant concentrations, the rapid drop in corrosion potential to 0.5 Vshe (Figure 11(a)) results from mass-transport-limited kinetics of oxidants to the metal surface. The relationship between dissolved oxygen and corrosion potential in hot water as a function of radiation type and flux is shown in Figure 11(a), in which the connected points represent data obtained in controlled radiation on/off experiments. The data from these latter experiments are shown in Figure 11(b) in terms of a radiation-induced shift in potential. The curves in Figure 11(a) represent the

2 points, 200 ppb H2O2 1 point, 31ppb Cu2+

0.2 0.1

0 –0.1

200 ppb H2

H2Wc BWR top core n + g Proton rad. = BWR peak –1 30–200 MGy h g

–0.2 –0.3

–1 0.3–2 MGy h g Pt, various rad levels

–0.4 –0.5

Open symbol = no radiation Lines indicate paired data (e.g., g on/off) Band of data for unirradiated conditions

–0.6 –0.7

100

(a)

101

102

103

104

Dissolved oxygen (ppb)

0.3 0.2 Dfc with radiation (VSHE)

188

0.1

200 ppb H2O2

0 –0.1

200 ppb H2 2 points

–0.2

Proton, ECP in crevice BWR top core n + g Proton rad. = BWR peak 30–200 MGy h–1g 0.3–2 MGy h–1g Pt, various rad levels

–0.3 <100 (b)

102 101 103 Dissolved oxygen (ppb)

104

Figure 11 (a) Effect of radiation on the corrosion potential of type 304 stainless steel in water at 288  C. The curves denote the range of typical values in the unirradiated corrosion potential data (reproduced from Andresen, P. L.; Ford, F. P.; Higgins, J. P.; et al. In Proceedings of ICONE-4 Conference; ASME International: New York, NY, 1996). (b) Effect of radiation on the shift in corrosion potential from the value under unirradiated conditions for type 304 stainless steel in water at 288  C. With the exception of the boiling-water reactor measurements, all data were obtained in controlled radiation on/off experiments (reproduced from Andresen, P. L.; Ford, F. P.; Higgins, J. P.; et al. In Proceedings of ICONE-4 Conference; ASME International: New York, NY, 1996). Curves in (b) show the trends in the proton-irradiated data, where the effects of radiation (on/off and over a range of fluxes) were evaluated for a variety of dissolved O2 and H2 concentrations under otherwise identical conditions. H2Wc, Hydrogen Water chemistry.

scatter band for the data obtained under unirradiated conditions. Similar scatter also exists in the irradiated corrosion potential data in Figure 11(a) and comprises contributions from both real effects and experimental error. High radiation flux experiments were

Irradiation Assisted Stress Corrosion Cracking

performed by Taylor66 and Andresen et al.67 using multiple, fundamentally different types of radiationhardened reference electrodes. In-reactor measurements have also been performed by Indig68,69 using multiple, radiation-qualified silver chloride reference electrodes. Figure 11(a) and 11(b) show that little, if any, elevation in corrosion potential results from irradiation sources that do not include neutrons or simulate their contribution (e.g., using high-energy protons67). Some studies using gamma radiation15,66 showed a significant decrease in corrosion potential, especially in the intermediate (e.g., 10–200 ppb) range of dissolved oxygen. This is consistent with enhanced recombination of oxidizing and reducing species, which occurs in the downcomer region of BWRs.65 In instances where neutrons or protons have been used, a consistent, significant elevation in corrosion potential is observed. This is more pronounced at low dissolved oxygen concentrations with no dissolved hydrogen (Figure 11(b)), where increases of over þ0.25 V occur. At higher inlet oxygen concentrations (e.g., 200 ppb), the data still show a significant shift (typically þ0.1 to 0.15 V) in corrosion potential for radiation conditions representative of peak LWR core fluxes (Figure 11(b)); less increase is observed for inlet oxygen concentrations associated, for example, with air saturation (8.8 ppm O2) or oxygen saturation (42 ppm O2 at STP). A similar elevation in corrosion potential is observed for additions of hydrogen peroxide (200 ppb H2O2, Figure 11(a)), which suggests that H2O2 may be a major factor in increasing the corrosion potential under irradiated conditions. In-core, in situ measurements in BWRs show that the corrosion potential, which is þ0.2 to þ0.25 Vshe in normal water chemistry (NWC), can be decreased by >0.5 V by sufficient additions of dissolved hydrogen in a BWR.69 This is corroborated by other measurements67 (Figure 11(b)), which show very little radiation-induced elevation in corrosion potential when the fully deaerated inlet water contains moderate dissolved hydrogen (>200 ppb H2, 0 ppb O2). At high H2 levels, the core becomes reducing, and the small concentration of 16N (transmuted from 16O) changes from soluble NO 3 to volatile NOx and NH3, causing a large increase in radiation level in the steam lines and turbine. The effect of radiation on the corrosion potential within a crack or crevice has also been of interest, with the possibility that a net oxidizing environment in the crack could be created that could elevate the

189

corrosion potential above the potential at the crack mouth. In the absence of radiation, measurements in high-temperature water in artificial crevices (e.g., tubing),70,71 at the tip of growing cracks,72 and of short crack growth behavior73 show that the corrosion potential remains low (i.e., 0.5  0.1 Vshe in pure water at 288  C) for all bulk oxygen concentrations, indicating that complete oxygen consumption occurs within the crack. Measurements of radiation effects in crevices67 show that the elevation in corrosion potential is limited to <0.05 V (’c < 0.45 Vshe) in-core; this is consistent with interpretation of available corrosion potential data on free surfaces.8,15,16 These small changes will not significantly affect the 0.75 V (þ0.25 Vshe (near mouth) minus 0.5 Vshe (in crack)) potential difference in the crack under irradiated normal BWR water chemistry conditions. The potential difference, along with other factors, controls the enhancement mechanism that can lead to an increased anion activity and altered pH at the crack tip.12,13,54,55 5.08.3.2 Effects of Corrosion Potential on IASCC Preirradiated stainless steels were evaluated in SSR tests50,74 in hot water using additions of oxygen and/or hydrogen peroxide to elevate the corrosion potential to simulate the effect of radiation. Jacobs et al.50 showed that IASCC in stainless steel irradiated to 3  1021 n cm–2 was strongly affected by dissolved oxygen (and, by inference, corrosion potential) (Figure 5(b)). Ljungberg74 also evaluated preirradiated materials in SSR tests and observed decreasing average crack growth rates with decreasing corrosion potential. Less IASCC occurs at low corrosion potential, but crack growth rate tests (discussed later) and other tests confirm that IASCC does not vanish at low corrosion potential (Figures 5 and 9). Continuously monitored fracture mechanics specimens were installed in the recirculation piping and core of the Nine Mile Point Unit 1 BWR. Furnace-sensitized type 304 stainless steel specimens in both locations showed higher growth rates in the core, where the corrosion potential was higher (Figure 12). Specimens of annealed type 304 and 304L stainless steel showed growth after different fluences but, once growth initiated, the growth rates at a given fluence were identical. The delay time (fluence) and differences in delay time were attributed to the need to transition from a transgranular fatigue precrack to IG SCC.

190

Irradiation Assisted Stress Corrosion Cracking

104

100 • Furnace sensitized 304 st. st. • 27.5 MPa m1/2

Crack length (mm)

100

Nine Mile Point 1 BWR DCB specimens for monitoring SCC

96 94

In—core

Observed Theory

Nine Mile Point 1 BWR DCB specimens for monitoring SCC C304SEN, thermally 27.5 MPa m1/2 sensitized 304 C304ST, solution treated 304

102

92 90

Crack length (mm)

98

98 96 94

C304L, low carbon 304

92 90

88 88 86

Recirculation line

86

84

84 0

(a)

500

1000 1500 2000 2500 3000 3500 4000 Time (h)

0

2000

4000

6000

8000

10 000

12 000

Time (h)

(b)

Figure 12 Data for fracture mechanics specimens of type 304 stainless steel exposed in the high flux region of the core and in the recirculation line of the Nine Mile Point Unit 1 BWR. All specimens were precracked and wedge loaded to an initial stress intensity factor of 27.5 MPa m1/2. (a) Comparison of predicted and observed crack length versus time for furnace-sensitized type 304 stainless steel specimens in the core and recirculation line. (b) Crack length versus time for one furnace-sensitized and two annealed specimens of type 304 stainless steel in the core. Adapted from Andresen, P. L.; Ford, F. P.; Murphy, S. M.; Perks, J. M. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Theus, G. J., Eds.; NACE: Houston, TX, 1990; pp 1–83; Andresen, P. L.; Ford, F. P. In Proceedings of 7th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995; pp 893–908; Was, G. S.; Andresen, P. L. Corrosion 2007, 63(1), 19–45; Andresen, P. L.; Ford, F. P. In Corrosion/89; NACE: Houston, TX, 1989; Paper no. 497; Andresen, P. L.; Ford, F. P. International Cooperative Group on Irradiation Assisted Stress Corrosion Cracking (ICG-IASCC) Minutes, Apr 1989.

Figures 2 and 13 present the first laboratory SCC growth rate testing ever performed on irradiated (4 dpa) type 304SS, which was well behaved at high corrosion potential and showed a large change in growth rate at lower corrosion potential. One evolving concern for high-fluence stainless steels is the prospect of a major loss of the effect of corrosion potential and stress intensity factor, as has been observed in unirradiated alloys possessing high Si.52,53 The detrimental effect of Si is likely associated with its ability to oxidize at all LWR-relevant potentials, coupled with its relatively high solubility in hot water (i.e., it does not form a protective oxide like Cr2O3 or Fe/Ni oxides/spinels). These data are compared with other irradiated and unirradiated data in Figure 2 based on simultaneous measurements of corrosion potential and crack growth rate in fracture mechanics specimens; the accompanying curves represent model predictions.8,13,15,16,54,55,75,76 Clearly, the in situ data compare favorably with the spectrum of unirradiated data, and data obtained on a fracture mechanics specimen of furnace-sensitized type 304 stainless steel using high-energy proton irradiation to simulate the mix of neutron and gamma radiation present in power reactors.

5.08.4 Irradiation Effects on Microchemistry and Microstructure 5.08.4.1

Radiation-Induced Segregation

RIS describes the redistribution of major alloying elements and impurity elements at point defect sinks.77–87 Because IASCC is intergranular, the sinks of greatest interest are the grain boundaries. RIS is driven by the flux of radiation-produced defects to sinks and is therefore fundamentally different from thermal segregation or elemental depletion due to grain boundary precipitation. Vacancies and interstitials are the basic defects produced by irradiation and can reach concentrations that are orders of magnitude higher than those at thermal equilibrium, resulting in large increases in diffusion rates. If the relative participation of alloying elements in the defect fluxes is not the same as their relative concentration in the alloy, then a net transport of the constituents to or from the grain boundary will occur (Figure 14). This unequal participation of solutes in the vacancy and/ or interstitial fluxes to sinks results in either enrichment or depletion of an alloying element at the grain boundary. The species that diffuse more slowly by the vacancy diffusion mechanism are enriched, and the faster diffusers become depleted.

Irradiation Assisted Stress Corrosion Cracking

191

250 12.59 2.7 × 10−8 mm s–1

12.53

150

To 0 ppb O2@1508 h

Crack length (mm)

12.55 1.3 × 106 mm s–1

12.51 12.49

100

Constant K

Dissolved oxygen (ppb)

200

12.57

50

CT#2 - irradiated 304SS, 4 dpa 21 MPa m1/2, 220 ppb O2, 288 ⬚C, 20 ppb SO4

12.47 12.45 1475

0 1495

1515

1535

1555

1575

1595

Time (h) Figure 13 Crack length versus time for Type 304 stainless steel irradiated to 4 dpa and tested in at 21 MPa m1/2 in water at 288  C. A large reduction in crack growth rate is observed as the dissolved O2 and corrosion potential are decreased. Reproduced from reproduced from Andresen, P. L.; Ford, F. P.; Higgins, J. P.; et al. In Proceedings of ICONE-4 Conference; ASME International: New York, NY, 1996.

Cv

Ci

Jv JA

Ji JB

JB

(a)

JA

(b)

CB CA0,B CA

(c) Figure 14 Schematic illustration of radiation-induced segregation in a binary 50A–50B system showing (a) the development of the vacancy concentration profile by the flow of vacancies to the grain boundary balanced by a equal and opposite flow of A and B atoms, but not necessarily in equal numbers, (b) the development of the interstitial concentration profile by the flow of interstitials to the grain boundary balanced by a equal and flow of A and B atoms migrating as interstitials, but not necessarily in equal numbers, and (c) the resulting concentration profiles for A and B. Reproduced from Was, G. S. Fundamentals of Radiation Materials Science: Metals and Alloys; Springer: Berlin, 2007.

Enrichment and depletion can also occur by association of the solute with the interstitial flux. The undersized species will enrich, and the oversized species will deplete.80 The magnitude of the buildup/depletion is dependent upon several factors such as whether a constituent migrates more rapidly by one defect mechanism or another, the binding energy between solutes and defects, the dose, dose rate, and the temperature. RIS profiles are also characterized by their narrowness, often confined to within 5–10 nm of the grain boundary, as shown in Figure 15 for an irradiated stainless steel. Segregation is a strong function of irradiation temperature, dose, and dose rate (Figure 16). Segregation peaks at intermediate temperatures since a lack of mobility suppresses the process at low temperatures, and back-diffusion of segregants minimizes segregation at high temperature (where defect concentrations approach their thermal equilibrium values). For a given dose, a lower dose rate results in a greater amount of segregation. At high dose rates, the high defect population results in increased recombination which reduces the number of defects that are able to diffuse to the grain boundary. Figure 16 shows the interplay between temperature and dose rate for an austenitic stainless steel. RIS occurs in the intermediate temperature range, and this range rises along the

192

Irradiation Assisted Stress Corrosion Cracking

18

4 JEOL 2010F 0.75 nm probe

16

3

14 12 10 8

Ni

2

Si

1

P 0

6 –20

–15 –10 –5 0 5 10 15 Distance from grain boundary (nm)

20

Figure 15 Compositional profiles across grain boundaries obtained by dedicated scanning transmission electron microscopy from a low-strain, high-purity 348 stainless steel swelling tube specimen irradiated to 3.4  1021 n cm–2 in water at 288  C in a boiling-water reactor. Composition profiles were measured using a field-emission gun scanning transmission electron microscope. Reproduced from Jacobs, A. J.; Clausing, R. E.; Miller, M. K.; Shepherd, C. M. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Ed.; NACE: Houston, TX, 1990; pp 14–21.

temperature scale with increasing dose rate to compensate for the higher recombination rate. In Fe–Cr–Ni alloys, the vacancy exchange (inverse Kirkendall) mechanism successfully explains the observed major element segregation.84,85 Studies have shown that nickel segregates to grain boundaries while chromium and iron deplete. The directions of segregation are consistent with an atomic volume effect in which the subsized solute migrates preferentially with the interstitial flux, and the oversized solute participates preferentially in the vacancy flux. The results are also consistent with the diffusivity of the solutes in Fe–Cr–Ni, in which Ni is the slow diffuser, Cr is the fast diffuser, and Fe is intermediate. In commercial austenitic stainless steels, chromium depletes at grain boundaries and nickel enriches, while iron can either deplete or enrich according to the magnitude of the diffusion coefficient relative to the other solutes.88 RIS increases with neutron dose in LWRs and saturates after several (
5) displacements per atom in the 300  C temperature range. Figure 17 shows grain boundary chromium depletion for austenitic

Homologous temperature (T/Tm)

5 Cr

Si or P concentration (wt%)

Cr or Ni concentration (wt%)

20

1105

0.8 0.7

Back diffusion of vacancies

933

0.6

761 589

0.5 0.4

Radiation-induced segregation

416

LWR 0.3 peak flux region 0.2 0.1 10–8

244

Recombination of vacancies and interstitials 10–7 10–6 10–5 10–4 –1 Radiation flux (dpa s )

72

Temperature for g stainless steel (⬚C)

22

–101 10–3

Figure 16 Dependence of radiation-induced segregation on homologous temperature and dose rate for austenitic stainless steels.

stainless steels as a function of dose.89–97 As the slowest diffusing element, nickel becomes enriched at the grain boundary. Since iron depletes in 304 and 316 stainless steels, the nickel enrichment makes up for both chromium and iron depletion and can reach very high levels up to
30 wt%. Minor alloying elements and impurities also segregate and have been implicated in the IASCC process. Mn and Mo strongly deplete at the grain boundary under irradiation,98 but neither is believed to be a significant factor in IASCC. Minor alloying or impurity elements such as Si and P also segregate under irradiation. Silicon strongly enriches at the grain boundary to as much as ten times the bulk (0.7–2.0 at.%) composition in the alloy99 and can be important in IASCC. Phosphorus is present at much lower concentrations and is only modestly enriched at the grain boundary because of irradiation.83,98 Phosphorus tends to segregate to the grain boundary following thermal treatment, which reduces the amount of additional segregation to the grain boundary during irradiation, making the contribution due to irradiation difficult to detect.98 Undersized solutes such as C, B, and N should also segregate, but there is little evidence of RIS, due in part to the difficulty of measurement. Another potential segregant is helium, produced by the transmutation of 10B. The mobility of He is low at LWR core temperatures, but the opportunity for accumulation at the grain boundary is increased by segregation of B to the boundary. Overall, the behavior of these minor elements under irradiation is not well understood.

Irradiation Assisted Stress Corrosion Cracking

Fast neutron fluence (E > 1 MeV) ⫻ 1025 n m–2 Grain boundary Cr concentration (wt%)

26

0

2

4

6

8 304 (82) 304 (13) 304 (91) 304 (92) 304 (93) 316 (82) 316 (94) 348 (91)

24 22 20

10

12

14

HP 304 (95) CP 304 (95) CP 316 (95) CP 304 (17) CP 316 (17) HP 316 (17) CP 304 Protons 96 CP 316 Protons 96

18 16 14 12 10 0

5

10 Dose (dpa)

15

20

Figure 17 Dose dependence of grain boundary chromium concentration for several 300-series austenitic stainless steels irradiated at a temperature of about 300  C. Adapted from Asano, K.; Fukuya, K.; Nakata, K.; Kodama, M. In Proceedings of 5th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Simonen, E. P., Gold, R. E. Eds.; American Nuclear Society (ANS): LaGrange, IL, 1992; p 838; Jacobs, A. In Proceedings of 7th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995; p 1021; Jacobs, A. J.; Wozadlo, G. P.; Nakata, K.; et al. In Proceedings of 6th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Gold, R. E., Simonen, E. P., Eds.; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 1993; p 597; Kenik, E. A. J. Nucl. Mater. 1992, 187, 239; Nakahigashi, S.; Kodama, M.; Fukuya, K.; et al. J. Nucl. Mater. 1992, 179–181, 1061; Jacobs, A. J.; Clausing, R. E.; Miller, M. K.; Shepherd, C. M. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Ed.; NACE: Houston, TX, 1990; pp 14–21; Walmsley, J.; Spellward, P.; Fisher, S.; Jenssen, A. In Proceedings of 7th International Symposium on Environment Degradation of Materials in Nuclear Power System—Water Reactors; NACE: Houston, TX, 1997; p 985; Was, G. S.; Busby, J. T.; Gan, J.; et al. J. Nucl. Mater. 2002, 300, 198.

Oversize solutes can affect the microchemistry or microstructure of the alloy, thereby altering the IASCC susceptibility. They are believed to affect RIS by acting as vacancy traps, thereby increasing the effective recombination of vacancies and interstitials and thus reducing RIS. Kato et al.100 conducted electron irradiations of several stainless steels at temperatures of 400–500  C up to 10 dpa. Results showed that some solutes (Zr and Hf) consistently produced a large suppression of radiation-induced chromium

193

depletion, while others resulted in less suppression or suppression at only certain temperatures. Fournier et al.101 conducted irradiation of 316 containing Hf or Pt using 3-MeV protons (400  C) and 5-MeV Ni ions (500  C). Ni irradiations showed little effect of the oversize impurity in reducing grain boundary chromium depletion (Cr depletion increased in the case of Hf), but proton irradiation showed a significant suppression of RIS of chromium at low dose (2.5 dpa) with the effect diminishing at higher (5.0 dpa) dose. Pt had a smaller effect on Cr. Ti and Nb similarly produced little change in the grain boundary chromium concentration after irradiation with 3.2-MeV protons to 5.5 dpa at 360  C. In Zr-doped 304SS, there were no consistent results of suppression of grain boundary chromium after 3.2-MeV proton irradiation to 1.0 dpa at 400  C.102 Neutron irradiation at very low dose (0.5 dpa) shows a small effect of Ti and Nb on grain boundary Cr.103 In all, the data on the effect of oversize solutes on RIS of chromium are inconsistent. RIS is understandably implicated in IASCC of stainless steels, especially in oxidizing environments, in part because of the effect of thermal sensitization in extensive data from lab and plant operational experience.15–20,37,38,104,105 As shown in Figure 17, grain boundary chromium depletion during irradiation can be severe. Figure 18(a) shows a correlation between grain boundary chromium level and IGSCC susceptibility in stainless steels where the grain boundary depletion is due to thermal sensitization.106 Much data have been accumulated to support the role of chromium depletion as an agent in IGSCC of austenitic alloys in oxidizing conditions. Numerous studies show that, as the grain boundary chromium level decreases, intergranular SCC increases. Typical chromium-depleted zone widths are of the order 100–300 nm full width at half-maximum (FWHM), providing a significant volume of depleted material adjacent to the grain boundary. Figure 18(b) shows a similar correlation between grain boundary chromium level and IASCC susceptibility as measured by the percentage IG cracking on the fracture surface during SSR experiments. A major difference between Cr depletion profiles resulting from RIS and those due to precipitation reactions is that the width of the RIS profiles can be as much as 2 orders of magnitude smaller, typically 5–10 nm. There is large scatter in the data that makes a direct correlation difficult to support, and differences in testing conditions undoubtedly contribute.

194

Irradiation Assisted Stress Corrosion Cracking

%IGSCC in slow strain rate test

100 80 60

e° = 2 ⫻ 10–7 S–1

␧° = 1 ⫻ 10–6 S–1 40 Alloy 600 SSR tests, 23 ⬚C sulfuric acid

20

Type 304SS SSR tests, 288 ⬚C 8 ppm O2 water

0 4 (a)

6 8 10 12 14 16 18 20 Minimum grain boundary chromium concentration (wt%)

100

HP 316 SS with single solute additions108 3xx SS17

80

%IG

60 40 20 0 10 (b)

12

14 16 18 20 Grain boundary Cr content (wt%)

irradiation. However, it should be noted that these alloys were not irradiated, and this difference may be important in the relevance of such experiments to IASCC. Using 1.5–5% Si stainless steels of both standard (e.g., 304L) base composition and synthetic irradiated grain boundary composition, Andresen has observed significantly increased growth rates, reduction in the benefit of lowering corrosion potential, and very little effect of stress intensity factor between 27 and 13 ksi in1/2. The data on impurity segregation effects on IASCC remain inconclusive. Extensive experiments have been conducted to isolate the effect of particular impurities such as S, P, C, N, and B in IASCC, but none have yielded unambiguous results. Sulfur has not been found to segregate under irradiation, and, while P thermally segregates to a significant extent, irradiation-induced P segregation is small in comparison. C, N, and B cannot be measured in STEM; N and B are very difficult to identify in AES; and C is a common contaminant. Overall, it has been a challenge to establish a link between impurity element segregation and IASCC in austenitic stainless steels.

22

Figure 18 Effect of grain boundary Cr content on intergranular stress corrosion cracking for (a) sensitized stainless steel and Alloy 600 and (b) irradiated stainless steels. Slow-strain-rate tests are tests in which the specimen is monotonically strained versus time. Reproduced from Bruemmer, S. M.; Was, G. S. J. Nucl. Mater. 1994, 216, 348.

Among the minor alloy elements, only Si is known to segregate to high levels, and Si segregation is correlated with IASCC. Experiments by Busby et al.107,108 on a high-purity 316 base alloy doped with 1 wt% Si showed severe IASCC in NWC and in primary water after irradiation to 5.5 dpa at 360  C. STEM measurements of grain boundary Si confirm levels up to 6 wt%. Past studies comparing Auger electron spectroscopy (AES) and scanning transmission electron microscopy (STEM) results have shown that the actual concentration of Si at the grain boundary plane may be as high as 15–20 wt%. Though the electron beam probe in STEM is very small, the measurement underpredicts the concentration at the grain boundary by a factor of 2 to 3. Yonezawa et al.109–112 and Li et al.113 have provided extensive evidence to show that increased Si in stainless steel results in increased IGSCC in alloys tailored to imitate the composition of grain boundaries under

5.08.4.2 Microstructure, Radiation Hardening, and Deformation 5.08.4.2.1 Irradiated microstructure

The microstructure of austenitic stainless steels under irradiation changes rapidly at LWR service temperatures. Point defect clusters (called ‘black dot damage’ when electron optics could not resolve the details) begin to form at very low dose, dislocation loops and network dislocation densities evolve with dose over several displacements per atom, and the possibility exists for the formation and growth of He-filled bubbles, voids, and precipitates in core components in locations exposed to higher dose and temperatures.114–120 Below 300  C, the microstructure is dominated by small clusters and dislocation loops. Near 300  C, the microstructure contains larger faulted loops plus network dislocations from loop unfaulting and cavities at higher doses. The primary defect structures in LWRs are vacancy and interstitial clusters and Frank dislocation loops. The clusters are formed during the collapse of the damage cascade associated with primary and secondary atom collisions after an interaction with a high-energy particle. The larger, faulted dislocation loops nucleate and grow as a result of the high mobility of interstitials. The loop population grows in size and number density until absorption of vacancies and

Irradiation Assisted Stress Corrosion Cracking

interstitials equalize, at which point the population has saturated. Figure 19 shows the evolution of loop density and loop size as a function of irradiation dose during LWR irradiation at 280  C. Note that saturation of loop number density occurs very quickly, by
1 dpa, while loop size continues to evolve up to
5 dpa. The specific number density and size are dependent on irradiation conditions and alloying elements, but the loop size rarely exceeds 20 nm and densities are of the order of 1  1023 m–3. The hardening process and IASCC susceptibility are influenced by small defects. The traditional view that small defect clusters are predominantly faulted interstitial loops and vacancy clusters121 may be inaccurate. Analysis of recent postirradiation annealing experiments by Busby et al.122 and Simonen et al.123 suggests that there are at least two types of defects with different annealing characteristics: vacancy and interstitial faulted loops, each with different annealing kinetics. The step change in hardness as a function of annealing time suggests that the density of vacancy loops is perhaps much higher than previously believed, and higher than the density of interstitial loops.122 Above 300  C, voids and bubbles may begin to form, aided by the increased mobility of point defects at the higher temperature. The dislocation structure will evolve into a network structure as larger Frank loops unfault. The reduction in the sink strength of the dislocation loops aids in the growth of voids and bubbles. While their size and number density increase with temperature, the dislocation microstructure continues to be the dominant microstructure component over the temperature range expected for LWR components (<350  C).

Loop density (m–3)

11 1023

Loop density

1022

Loop size

10 9 8 7

1021

1020 0

6

Type 304SS heats Type 316SS heats 1

2

3 4 5 6 Irradiation dose (dpa)

5 7

8

Average loop diameter (nm)

12

4

Figure 19 Measured change in density and size of interstitial loops as a function of dose during light water reactors irradiation of 300-series stainless steels at 275–290  C. Reproduced from Bruemmer, S. M.; Simonen, E. P.; Scott, P. M.; Andresen, P. L.; Was, G. S.; Nelson, J. L. J. Nucl. Mater. 1999, 274, 299.

195

Irradiation can also accelerate or retard the growth of second phases, modify existing phases, or produce new phases, although these processes are more pronounced above 400  C. In stainless steels, the principal second phase is chromium carbides, which are stable under irradiation. In high-strength Ni-base alloys, the second phases can undergo several types of transformations: g0 can dissolve, g00 can dissolve and reprecipitate, and Laves phase can become amorphous. A key factor in phase formation in austenitic stainless steels under LWR operating conditions is RIS, which can induce the formation of phases by exceeding the local solubility limit. Was et al.124 irradiated a high-purity stainless steel containing 1 wt% Si with 3.2-MeV protons to 5.5 dpa at 360  C. They observed the formation of g0 (Ni3Si) in the matrix but not on the grain boundary, which is puzzling since the concentration of both Si and Ni is higher in the boundary. This was also observed in a similar alloy irradiated with neutrons to 7 dpa at 300  C.125 g0 is a coherent precipitate that can significantly strengthen the matrix and has the potential to alter the deformation behavior in the unirradiated and irradiated conditions. Oversize solutes can also affect the irradiated microstructure by mechanisms similar to RIS. Proton and nickel ion irradiations show that the addition of Hf to a 316SS-base alloy increased loop density, decreased loop size, and eliminated voids.101 Platinum addition to 316SS resulted in no change in loop density and a small increase in loop size, but increased void size and density. The good agreement between proton and Ni ion irradiation results indicates that the major effect of the oversized solute is not due to the cascade (where there are large differences between proton and nickel ion irradiation), but rather is due to the post-cascade defect partitioning to the microstructure evolution. Electron irradiation experiments by Watanabe et al.126 and proton irradiation experiments by Was et al.124 showed that stainless steel with Ti additions had slightly lower dislocation loop densities and larger sizes compared to the base alloy. Nb increased only the loop size. In contrast to the base alloy, neither the Ti nor the Nb-doped alloys formed voids under the conditions tested. Zirconium addition to 304SS resulted in reduced hardness, decreased loop density, and no change in loop size in proton irradiation to 1.0 dpa at 400  C and compared to the base alloy.127 Zirconium-containing samples also had a lower void density with no change in void size as compared to the base alloy.

196

Irradiation Assisted Stress Corrosion Cracking

5.08.4.2.2 Radiation hardening

1200 1000

Yield strength (MPa)

The dislocation loop microstructure formed during irradiation hardens the alloy, which correlates with increased SCC susceptibility. Under an applied stress, dislocation network interacts elastically with the dislocation loops, producing an increase in the yield strength of the alloy, which can be detected in tensile tests or indentation hardness measurements. This is accompanied by a decrease in elongation, which is affected more than reduction in area because necking occurs relatively early in most tensile tests of irradiated stainless steel. Fracture toughness also decreases with fluence. The increase in yield strength with dose in 300 series stainless steels irradiated around 300  C is shown in Figure 20. The yield strength can reach values up to five times the unirradiated value by about 5 dpa, and its increase follows a square root dependence on dose. Both the source hardening model128 and the dispersed barrier hardening model129 provide reasonable correlations between hardening and the dislocation loop microstructure. In the dispersed barrier hardening model, the increase in hardness is proportional to (Nloop  dloop)1/2, where Nloop is the loop number density and dloop is the loop diameter. Deformation changes dramatically with radiation. Homogeneous deformation at low dose is replaced by heterogeneous deformation at higher doses as the defect microstructure begins to impede the motion of dislocations. Plasticity becomes localized to narrow channels that have been cleared of defects by preceding dislocations, providing a preferred path for subsequent dislocation motion. The channels are very narrow (<100 nm) and closely spaced (<1 mm) and typically run the full length of a grain, terminating at the grain boundaries.130 Dislocation channeling can cause localized necking and a sharp reduction in uniform elongation.131 Hardening has been cited as a key factor in IASCC susceptibility. Figures 2 and 21(a) show that increasing cold work and yield strength increases the crack growth rate in nonsensitized austenitic stainless steel tested in 288  C BWR NWC.132 This dependence carries over to irradiated material. Figure 21(b) shows a correlation between yield strength and susceptibility to IASCC in SSR tests although the correlation is complicated by other radiation-induced changes, especially RIS. The correlation is better at high values of yield strength (>800 MPa) and at very low values (<400 MPa), with more scatter at intermediate values (400–800 MPa).

800 600 400 200 0

0

2

4 6 Dose (dpa)

8

10

Figure 20 Irradiation dose effects on measured tensile yield strength for several 300-series stainless steels, irradiated and tested at a temperature of about 300  C. Adapted from Singh, B. N.; Foreman, A. J. E.; Trinkaus, H. J. Nucl. Mater. 1997, 249, 103; Seeger, A. In Proceedings of 2nd United Nations International Conference on the Peaceful Uses of Atomic Energy; United Nations: New York, NY, 1958; Vol. 6, p 250; Bruemmer, S. M.; Cole, J.; Carter, R.; Was, G. S. In Proceedings of 6th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Gold, R. E., Simonen, E. P. The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 1993; p 537; Bloom, E. E. In Radiation Damage in Metals; Peterson, N. L., Harkness, S. D., Eds.; ASM International: Materials Park, OH, 1975; p 295.

Hardening alone is not sufficient to explain IASCC. Annealing experiments tracked the change in hardness and dislocation loop microstructure versus annealing condition.133 Figure 22 shows the change in hardness and the change in the dislocation loop line length (Navg  davg) along with the change in IASCC susceptibility in SSR tests as annealing progresses.120,134,135 Except for very short annealing times, the loop line length and the hardness closely track each other, as expected if the loop structure is controlling the hardening. At very small values of (Dt)1/2, the hardening remains flat before decreased with annealing time. While both hardening and cracking are reduced with increased annealing, the behavior of the hardness does not fully explain the response. Busby et al.122 postulated that the removal of very small defects at short annealing times may account for the IASCC behavior. Short times at high temperature (500  C) may preferentially remove the small

Irradiation Assisted Stress Corrosion Cracking

10–7 10–8 10

100 Stress corrosion cracking, nonsensitized austenitic stainless steels in simulated BWR water 288 ⬚C

HP 316SS with single solute additions 108 3xxSS 17

80

304, 1.4301 347, 1.4550 321, 1.4541 316Ti, 1.4571

60 % IG

Stress corrosion cracking growth rate, Δa/Δt, (m s–1)

10–6

197

–9

40 10–10 20 10–11 Intergranular

10–12 0 (a)

200

increasing fraction of transgranular SCC

400 600 800 1000 1200 1400 1600 Yield strength, Rp0.2 (MPa) (b)

0 200

400

600

800

1000

Yield stress (MPa)

Percentage of as-irradiated feature remaining after heat treatment

Figure 21 Effect of yield strength on intergranular stress corrosion cracking. (a) Crack growth rate of cold-worked, unirradiated 300-series stainless steels tested in 288  C simulated boiling-water reactor (reproduced from Speidel, M. O.; Magdowski, R. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Ford, F. P., Bruemmer, S. M., Was, G. S., Eds.; The Minerals, Metals & Materials Society: Warrendale, PA, 1999; p 325) and (b) percentage intergranular stress corrosion cracking in slow strain rate tests on 300-series stainless steels where hardening is by irradiation (adapted from Bruemmer, S. M.; Simonen, E. P.; Scott, P. M.; Andresen, P. L.; Was, G. S.; Nelson, J. L. J. Nucl. Mater. 1999, 274, 299; Busby, J. T.; Kenik, E. A.; Was, G. S. J. Nucl. Mater. (in press)).

120 RIS

100

IASCC Hardness Loop line length RIS

80 60 40

Hardness and loop line length IASCC

20 0 0

0.0005 0.001 0.0015 0.002 0.0025 0.003 Iron diffusion distance, (DFet)1/2, (cm)

Figure 22 Removal of radiation-induced segregation and dislocation microstructure as measured by loop line length and hardness with extent of annealing as measured by (Dt)1/2 for iron, where D is the diffusivity of Fe and t is time. This relationship accounts for annealing at different times and temperatures. The effect on irradiation-assisted stress corrosion cracking in slow strain rates is also shown. Adapted from Busby, J. T.; Was, G. S.; Kenik, E. A. J. Nucl. Mater. 2002, 302, 20; Edwards, D. J.; Simonen, E. P.; Garner, F. A.; Greenwood, L. R.; Oliver, B. M.; Bruemmer, S. M. J. Nucl. Mater. 2000, 317, 32; Katsura, S.; Ishiyama, Y.; Yokota, N.; et al. In Corrosion/98; NACE: Houston, TX, 1998; Paper no. 132; Jacobs, A.; Wozadlo, G. P.; Gordon, G. M. Corrosion 1995, 51, 10, 731.

defect clusters either by annihilation or by spontaneous dissociation. The dislocation loops may absorb the free vacancies and interstitials, thus adding to their line length. The loss of the small defect clusters will be offset by the growth of Frank loops producing no net change in measured hardness or yield strength. Despite a lack of hardness change, this process may alter the deformation mode at the local level by removing the small obstacles to dislocation motion, thus changing the character of localized deformation. Hash et al.136 showed that hardening is not the sole factor in IASCC, by testing a series of samples of commercial purity 304SS with nominally the same hardness but with different contributions from cold work and proton irradiation. At the extremes were a sample that was cold-rolled to a 35% reduction in thickness and no irradiation and one with 1.67 dpa irradiation and no cold work. Three samples had varying amounts of cold work (10, 20, and 25%) and corresponding amounts of irradiation dose (0.55, 0.25, and 0.09 dpa) to give a hardness level that was within 5% over all samples. SCC susceptibility was measured by the amount of IG cracking in an SSR test in 288  C BWR NWC. IASCC susceptibility was not constant, as would be expected if hardness were the only factor, with cracking observed in only the two highest dose samples (0.55 dpa with 10% cold

198

Irradiation Assisted Stress Corrosion Cracking

work and 1.67 dpa with no cold work), irrespective of their hardness, Figure 23. The amount of cracking in the lower dose sample was higher than in a companion sample at the same dose but without cold work, indicating that cold work can enhance the IASCC susceptibility. These results also suggest that RH may promote crack initiation, as it does crack growth. Combined with the annealing results, they suggest that other factors besides the hardness level and yield strength – such as RIS and deformation mode – play a role in the IASCC process. 5.08.4.2.3 Deformation mode

Results of IASCC experiments on proton-irradiated samples over a wide range of doses and alloys have consistently showed that high Ni alloys have high IASCC resistance in SSR tests. In particular, Ni concentrations >18 wt% are highly resistant to IASCC compared to 304SS with 8 wt% Ni. Kodama et al.137 showed that there is a good, but not perfect, correlation between Ni equivalent and IASCC. Notable exceptions are alloy 800138 and a Fe–20Cr–25Ni– 1Nb alloy used in AGR that experienced IGSCC.139 Ni may affect IASCC directly through a change in composition or indirectly by changing the slip character. Higher nickel content in stainless steel increases the stacking fault energy (SFE) significantly, producing a change in slip from planar to wavy. Swan et al.140 studied the effect of SFE on slip behavior using a series of Fe–18Cr–xNi alloys where 8<x<23. He showed that for the 8% Ni alloy, the slip was entirely planar and, as Ni increases, the cross-slip increases. By 20% Ni, there was no evidence of planar slip and the deformation microstructure consisted of a web of Dose (dpa)

0

0.09

0.25

0.55

dislocation tangles, which is evidence of wavy slip in a high SFE material. Figure 24 shows that planar slip yields greater dislocation interaction with grain boundaries than does a wavy slip. SFE has been linked to SCC resistance in stainless steels by Thompson and Bernstein, who found that increasing SFE correlates well with increased reduction in area and decreased SCC susceptibility.141 The IASCC susceptibility of several irradiated stainless steels is plotted as a function of SFE in Figure 25 and shows that there is a good correlation between SFE and IASCC using Rhode’s142 and Schramm’s143 correlations for SFE. These correlations differ in the elements included and the weights given in terms of nickel equivalent (NiEq), but are unable to account for some minor elements and therefore may deviate substantially from the true SFE in some cases. There is much scatter in the Ni equivalent plots, similar to that shown in the IGSCC dependence on grain boundary Cr and yield strength. One potential source of the scatter is the inherent variability in initiationdominated phenomena. Another is that, while grain boundary Cr and yield strength are measured, NiEq and SFE are generally calculated and with considerable uncertainty. So they may be useful in identifying transitions in behavior, but are not as reliable quantitatively. Microstructure can also influence the deformation mode. Farrell et al.144 conducted neutron irradiation of 316 stainless steel at temperatures between 65 and 100  C to doses of <1 dpa, and then characterized the deformation behavior. With increasing dose, the propensity for dislocation channeling increased (Figure 26). The volume of material occupied by channels increases rapidly with dose and saturates

1.67

Extent of IG cracking

1mm

1mm Low SFE ® planar

Cold work 35%

e

a

25%

20%

10%

High SFE ® wavy slip

0%

Figure 23 Degree of irradiation-assisted stress corrosion cracking in Type 304 stainless steel samples with the same hardness but with different combinations of hardening by cold work and irradiation using 3.2-MeV protons at 360  C. Reprinted, with permission, from 21st International Symposium on Effects of Radiation on Materials, copyright ASTM International, 100 Barr Harbor Drive, West Conshohocken, PA 19428.

Grain boundary

Grain boundary

Figure 24 Micrographs and corresponding schematic illustrations of planar and wavy slip in the vicinity of a grain boundary (micrographs from Swann140).

Irradiation Assisted Stress Corrosion Cracking

at <0.5 dpa. These data show that the defect microstructure created by irradiation can induce planar deformation in the form of narrow dislocation channels. The importance of slip localization in IASCC may be in the way in which the dislocations interact with the grain boundary. In planar slip, well-defined and separated slip bands or dislocation channels (for irradiated materials) transmit dislocations during plastic deformation. These slip bands or channels terminate at grain boundaries where dislocations are fed into the grain boundary region versus forming a tangled dislocation network within the grain. In the more traditional view of dislocation infusion into grain boundaries, the pileup of dislocations in the intersecting channel creates progressively higher stresses at the grain boundary at the head of the pileup. If the stress exceeds a critical value, separation of the grain boundary could occur according to a Stroh (wedge) cracking type of mechanism.145 This cracking process could occur regardless of the environment and may in fact be the mechanism that occurs in some of the IG cracking observed in very highly irradiated steels in inert environment.131 At lower fluences, the stress at the grain boundary may promote rupture of the oxide film, leading to exposure of the metal to the solution and subsequent corrosion and IASCC. Alternatively, deformation could occur in the boundary plane, which would rupture the oxide film and promote IASCC. Alexandreanu146 showed that dislocation absorption by grain boundaries in

100

IGSCC (%)

80 60 SFE = 1.2 + 1.4% Ni + 0.6% Cr + 17.7% Mn – 44.7% Si Rhodes and Thompson142

40 20 0 0

(a)

50 100 Stacking fault energy (mJ m–2)

150

100

IGSCC (%)

80 60 SFE = –53 + 6.2% Ni + 0.7% Cr + 3.2% Mn + 9.3% Mo Schramm and Reed143

40 20 0 0

(b)

50 100 150 Stacking fault energy (mJ m–2)

200

Figure 25 Irradiation-assisted stress corrosion cracking susceptibility as measured by percentage IG cracking as a function of stacking fault energy determined using (a) Rhode’s correlation (reproduced from Rhodes, C. G.; Thompson, A. W. Met. Trans. 1977, 8A, 1901) and (b) Schramm’s correlation (reproduced from Schramm, R. E.; Reed, R. P. Met. Trans. A 1975, 6A, 1345).

0.3

1015

ND (cm–2)

0.15

1014 Channel area

0.1

ORNL NERI 316 SS Neutron irradiated

0.05 0

0.5

1 Dose (dpa)

1.5

2

0.04

0.03

0.02

0.01

Channel area (nm2 nm–2)

0.2

0.05 Strain hardening exponent (n)

0.25 Strain hardening exponent (n)

1013 0

199

0

Figure 26 Variation in dislocation channel area, dislocation loop line length, and strain hardening exponent as a function of dose for neutron-irradiated 316SS. ‘ND’ is the product of the number density and diameter of Frank dislocation loops. Reproduced from Farrell, K.; Byun, T. S.; Hashimoto, N. Mapping flow localization processes in deformation of irradiated reactor structural alloys, Report ORNL/TM-2002/66; Oak Ridge National Laboratory, July 2002.

200

Irradiation Assisted Stress Corrosion Cracking

nickel-base alloys leads to deformation in the grain boundary, usually by boundary sliding, and promotes IGSCC. He observed random grain boundaries exhibited more sliding, and boundaries where sliding occurred were four times more susceptible to IGSCC in primary water at 360  C than boundaries that did not slide. This same process could explain how dislocation injection into grain boundaries by planar slip or dislocation channeling can result in IASCC. The concept that slip planarity controls the deformation mode and IASCC is consistent with observations of IASCC in 304 and 316SS. 316SS has, in some cases83 been shown to be more resistant to IASCC than is 304SS, which is consistent with its higher SFE. Observations by Bailat et al.147 on neutron-irradiated samples and by Busby et al.148 on proton-irradiated samples show clear dislocation channel patterns on the surface of 304SS samples that cracked in SSR tests but an absence of those patterns on the 316SS samples that did not crack. In addition, SSR tests on high Ni alloys (with SFE of
40 mJ m–2) showed no IG cracking and also no dislocation channeling. Further support for the influence of slip planarity

on IASCC comes from the annealing studies cited earlier. On samples that underwent either a short, low-temperature anneal or were tested in the asirradiated condition, IG cracking was accompanied by a high density of dislocation slip bands on the surface of
65 mm2. Annealing at 500  C for 45 min resulted in the elimination of IG cracking and a reduction in the surface slip band density to
25 mm–2. In the annealing experiments described earlier, hardening by irradiation was much more effective in initiating IGSCC than was cold work, where the dislocation structure will likely be more cellular than planar. Bruemmer et al.149 showed a correlation between slip band intersection with the walls of a growing crack and the accompanying steps on the oxide on the walls. Figure 27(a) shows a TEM micrograph of SCC in a cold-worked 316 stainless steel baffle bolt. The slip bands intersect the walls of the narrow crack at 45  to the crack growth direction. The slip band–crack wall intersections are also coincident with steps in the oxide as shown schematically in Figure 27(b). This suggests that the flow of

Metal m/o interface

Oxide

100 nm Crack wall oxide growth Crack

Grain boundary deformation by dislocation

Figure 27 Micrograph of (a) deformation bands intersecting a crack in a baffle bolt (reproduced from Thomas, L. E.; Bruemmer, S. M. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Ford, F. P., Bruemmer, S. M., Was, G. S., Eds.; Metallurgical Society of the American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): Warrendale, PA, 1999; p 41). (b) A schematic of the role that the deformation bands may be playing in irradiation-assisted stress corrosion cracking.

Irradiation Assisted Stress Corrosion Cracking

250

200 Stress (MPa)

dislocations along the slip steps and into the grain boundary may have been responsible for discontinuous growth of the crack along the grain boundary. Thus, low SFE and irradiation can both lead to planar or localized deformation, which terminate at grain boundaries, which must be accommodated by the grain boundary. This results in shear strain that can rupture the oxide film and promote the initiation of intergranular cracks.

150

100

50

5.08.4.3 Radiation Creep and Stress Relaxation

0 0

5 10 Displacement level (dpa)

15

Figure 28 The effects of radiation-induced creep on load relaxation of stainless steel at 288  C. Reprinted, with permission, from 16th International Symposium on Effects of Radiation on Materials, copyright ASTM International, 100 Barr Harbor Drive, West Conshohocken, PA 19428.

1.0 0.9

Fraction of stress remaining

At LWR temperatures, radiation creep results from diffusion of the radiation-produced vacancies and interstitial atoms to dislocations, enhancing the climb-to-glide process that controls time-dependent deformation. Radiation creep can be both beneficial and detrimental. Benefits accrue from relaxation of constant displacement stresses, for example, weld residual stress and in loaded bolts and springs. However, under these conditions – and more so under constant load conditions – radiation creep also induces elevated creep rates, including grain boundary sliding, that help initiate and sustain SCC. Figures 28 and 29 show examples of load relaxation under constant displacement conditions, a process that is quite reproducible over a wide range of materials and loading modes, and generally produces sizeable (>50%) load relaxation within a few displacements per atom. Thus, for example, in areas of the BWR shroud that receive a moderate neutron flux, if SCC initiation does not occur early in life (e.g., by 1 dpa), the relaxation in residual stress should diminish the likelihood of cracking later in life. Because the effect of relaxation is significant, it tends to offset the detrimental effects of RIS and RH. Thus, it is not surprising that the incidence of SCC in BWR shroud welds, where the neutron flux can vary by 2 orders of magnitude because of the varying proximity of the fuel, does not show a strong correlation with fluence. Radiation creep relaxation also affects PWR baffle bolts, which are subject to large variations in fluence and temperature.40,41 Baffle bolts in high flux regions can accumulate more than 3 dpa year1, and thus the preload will rapidly decrease during the first several years. Thus, SCC probably initiates early in life (before significant radiation creep relaxation occurs) or later in life when reloading occurs from differential swelling in the (annealed) baffle plates relative to the (cold-worked) baffle bolts.

201

Preload 23.6 N 36.5 N

0.7

0.5

Walters -1977 370 ⬚C, EBR-II X-750 springs 0.3

0

1

2 3 Neutron dose (dpa)

4

5

Figure 29 Radiation creep relaxation of X-750 springs at 370  C. Reproduced from Walters, L. C., Ruther, W. E. J. Nucl. Mater. 1977, 68, 324.

While difficult to prove, the elevated and sustained deformation rates associated with radiation creep can only accentuate susceptibility to SCC. Estimates of crack tip deformation rates15 indicate the radiation creep is not a large contributor in actively growing cracks, but rather it is expected to promote crack initiation and to sustain crack growth (or promote crack reinitiation, if an existing crack does arrest).

202

Irradiation Assisted Stress Corrosion Cracking

It is obviously very important to factor radiation creep relaxation into initial component design and subsequent SCC analysis. Its impact is large, and it occurs in the same fluence range as RIS and RH.

5.08.5 Summary Cracking of in-core reactor components exposed to both irradiation and high-temperature water supports the significant role of irradiation in intergranular SCC. The effects of irradiation on SCC occur through changes in the water chemistry and in the alloy microstructure. It is the ‘persistent’ radiation effects on the microstructure that are responsible for the ‘threshold-like’ behavior of IASCC. The principal effect of irradiation on water chemistry is through radiolysis, which, below
500 ppb dissolved H2, can increase the corrosion potential from the formation of radiolytic species (radicals and molecules that are oxidizing and reducing). High corrosion potentials strongly correlate with severity of cracking in both laboratory tests of preirradiated specimens and in core components in operating power reactors. Irradiation causes a significant change in local composition near grain boundaries and other defect sinks. The enrichment of nickel and silicon, and the depletion of chromium, may affect the susceptibility to IASCC. Irradiation also alters the microstructure, and, under LWR conditions, faulted dislocation loops represent the primary irradiation-induced microstructure defect. The loops impede the motion of dislocations, resulting in an increase in the yield strength by factor of up to 5. RH correlates with IASCC propensity, and also induces highly localized deformation in the form of dislocation channels, which could contribute to IASCC. Irradiation also induces creep that can relax macroscopic stresses and can also enhance local dynamic deformation. Other factors, such as swelling and formation of new phases, may accentuate IASCC at high fluence. With the many effects of irradiation, which overlap spatially and temporally, more work is needed to identify the mechanism(s) of IASCC and develop a comprehensive prediction methodology.

References 1.

Staehle, R. W.; Forty, A. J.; VanRooyen, D., Eds. Fundamental Aspects of Stress Corrosion Cracking; Ohio State University, NACE: Houston, TX, 1967.

2. Staehle, R. W.; Hochmann, J.; McCright, R. D.; Slatern, J. E., Eds. Stress Corrosion Cracking and Hydrogen Embrittlement of Iron-Base Alloys, Firminy, France; NACE: Houston, TX; June 1973; p 1977. 3. Proceedings of 1st–14th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE/ANS/TMS, 1983–2009. 4. Contribution of materials investigation to the resolution of problems encountered in pressurized water reactors, In Proceedings of 1st–6th International Symposium on Fontevraud, 1986–2006; Societe Francaise d’Energie Nucleaire: Paris, France. 5. Jones, R. H., Ed. In Proceedings of Chemistry and Electrochemistry of Corrosion and SCC: A Symposium Honoring the Contributions of R.W. Staehle; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, Feb 2001. 6. Bruemmer, S. M. et al., Eds. In Proceedings of Parkins Symposium on Fundamental Aspects of Stress Corrosion Cracking; The Metallurgical Society of the American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME), 1992. 7. Cowan, R. L.; Gordon, G. M. In Stress Corrosion Cracking and Hydrogen Embrittlement of Iron-Base Alloys; Staehle, R. W., Hochmann, J., McCright, R. D., Slatern, J. E., Eds.; NACE: Houston, TX, 1977; pp 1063–1065. 8. Ford, F. P.; Andresen, P. L. In Corrosion Mechanisms; Marcus, P., Oudar, J., Eds.; Marcel Dekker: New York, NY, 1994; pp 501–546. 9. Andresen, P. L.; Was, G. S. In 17th International Corrosion Congress, Las Vegas; NACE: Houston, TX, 2008. 10. Andresen, P. L. In Corrosion/96; NACE: Houston, TX, 1996; Paper 96258. 11. Andresen, P. L. In Proceedings of Tenth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 2001. 12. Ford, F.P.; Andresen, P.L. In Proceedings of 3rd International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Theus, G.J., Weeks, J.R. Eds.; The Metallurgical Society of the American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): New York, NY; p 789. 13. Andresen, P. L.; Ford, F. P. Mater. Sci. Eng. 1988, A1103, 167. 14. Andresen, P. L.; Hickling, J.; Ahluwalia, K. S.; Wilson, J. A. Effects of PWR primary water chemistry on PWSCC of Ni alloys. In Proceedings of 13th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Canadian Nuclear Soc, Aug 2007. 15. Andresen, P. L.; Ford, F. P.; Murphy, S. M.; Perks, J. M. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Theus, G. J., Eds.; NACE: Houston, TX, 1990; pp 1–83. 16. Andresen, P. L.; Ford, F. P. In Proceedings of 7th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995; pp 893–908. 17. Bruemmer, S. M.; Simonen, E. P.; Scott, P. M.; Andresen, P. L.; Was, G. S.; Nelson, J. L. J. Nucl. Mater. 1999, 274, 299–314. 18. Was, G. S.; Andresen, P. L. Corrosion 2007, 63(1), 19–45. 19. Scott, P. M. J. Nucl. Mater. 1994, 211, 101. 20. Was, G. S.; Andresen, P. L. J. Met. 1992, 44(4), 8. 21. Jacobs, A. J.; Wozadlo, G. P. In Proceedings of International Conference on Nuclear Power Plant Aging,

Irradiation Assisted Stress Corrosion Cracking

22.

23.

24.

25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

35.

36. 37.

38.

39. 40.

41.

42.

43.

Availability Factor and Reliability Analysis; ASM International: Materials Park, OH, 1985; p 173. Hanninen, H.; Aho-Mantila, I. In Proceedings of 3rd Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): New York, NY, 1988; pp 77–92. Bruemmer, S. M.; Arey, B. W.; Charlot, L. A. In Proceedings of Sixth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 1993; p 277. Andresen, P. L. In Proceedings of Sixth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Metallurgical Society of The American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): New York, NY, 1994; pp 245–253. Cheng, C. F. Corrosion 1964, 20, 341. Duncan, R. N.; Arlt, W. H.; Williamson, H. E.; Baroch, G. S.; Hoffman, J. P.; Pashos, T. J. Nucl. Appl. 1965, 1, 413. Armijo, J. S.; Low, J. R.; Wolff, U. E. Nucl. Appl. 1965, 1, 462. Pashos, T. J.; Duncan, R. N.; Williamson, H. E.; Arlt, W. H.; Baroch, C. J.; Hoffman, J. P. Trans. Am. Nucl. Soc. 1964, 7(2), 416. Cheng, C. F. Corrosion 1964, 20, 341. Cheng, C. F. J. Nucl. Mater. 1975, 56, 11. Cheng, C. F. Reactor Technol. 1970, 13, p 310. Schaffer, L. D. Army Reactor Program Progress Report, ORNL 3231, Jan 1962. Brown, J. B., Jr.; Storhok, B. W.; Gates, J. E. Trans. Am. Nucl. Soc. 1967, 10, 668–669. Pasupathi, V.; Klingensmith, R. W. Investigation of stress corrosion in clad fuel elements and fuel performance in the Connecticut Yankee Reactor; Electric Power Research Institute (EPRI), NP-2119; Nov 1981. Multer, I. In European Operating Experience, Proceedings of Joint Topical Meeting on Commercial Nuclear Fuel Technology Today, Toronto, Apr 1975; 75-CNA/ANS100. Storre, J. T.; Locke, D. H. Nucl. Eng. Int. 1970, 2, 93–99. Gordon, G. M.; Brown, K. S. In Proceedings of 4th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1990; pp 14-46–14-62. Brown, K. S.; Gordon, G. M. In Proceedings of 3rd Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): New York, NY, 1988; pp 243–248. Horn, R. M.; Gordon, G. M.; Ford, F. P.; Cowan, R. L. Nucl. Eng. Des. 1997, 174, 313–325. Heuze, P. A.; Goltrant, O.; Cauvin, R. In Proceedings of Internationl Symposium on Fontevraud IV; French Society of Nuclear Energy: Paris, France, 1998; p 195, in French. Scott, P. M.; Meunier, M. C.; Deydier, D.; Silvestre, S.; Trenty, A. In Environmentally Assisted Cracking: Predictive Methods for Risk Assessment and Evaluation of Materials, Equipment and Structures, ASTM STP 1401; Kane, R. D., Ed.; ASTM: West Conshohocken, PA, 2000; pp 210–223. Garzarolli, F.; Rubel, H.; Steinberg, E. In Proceedings of Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1984; pp 1–24. Norris, D. I. R.; Baker, C.; Titchmarsh, J. M. In Proceedings of Radiation-Induced Sensitization of

44.

45.

46.

47.

48. 49. 50. 51.

52. 53.

54.

55.

56.

57. 58. 59.

60.

203

Stainless Steels, Berkeley Nuclear Labs; Central Electricity Generating Board: London, 1987; pp 86–98. Garzarolli, F.; Alter, D.; Dewes, P.; Nelson, J. L. In Proceedings of 3rd International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The American Institute of Mining, Metallurgy, and Petroleum Engineers (AIME): New York, NY, 1988; pp 657–664. Garzarolli, F.; Alter, D.; Dewes, P. In Proceedings of 2nd Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; American Nuclear Society (ANS): LaGrange Park, IL, 1986; pp 131–138. Andresen, P. L. In Proceedings of 7th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995; pp 563–578. Hettiarachchi, S.; Wozadlo, G. P.; Diaz, T. P.; Andresen, P. L.; Cowan, R. L. In Proceedings of 7th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995; pp 735–746. Andresen, P. L.; Diaz, T. P.; Hettiarachchi, S. In Corrosion/2004; NACE: Houston, TX, 2004; Paper no. 04668. Manahan, M. P.; Santucci, J.; Sipush, P. Nucl. Eng. Des. 1989, 113, 297–321. Jacobs, A. J.; Hale, D. A.; Siegler, M. Unpublished Data on SCC of Irradiated SS in 288  C Water and Inert Gas; GE Nuclear Energy; San Jose, CA, 1986. Onchi, T.; Dohi, K.; Soneda, N.; Navas, M.; Luisa Castano, M. In Proceedings of 11th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; American Nuclear Society (ANS): La Grange Park, IL, 2003; p 1111. Andresen, P. L.; Morra, M. M.; Catlin, W. R. In Corrosion/ 2004; NACE: Houston, TX, 2004; Paper no. 04678. Andresen, P. L.; Morra, M. M. In Proceedings of 12th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Minerals, Metals, and Materials Society (TMS): Snowbird, UT, 2005. Ford, F. P.; Andresen, P. L.; Solomon, H. D.; et al. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1990; pp 4-26–4-51. Ford, F. P.; Taylor, D. F.; Andresen, P. L.; Ballinger, R. G. Environmentally controlled cracking of stainless and low alloy steels in LWR environments; Electric Power Research Institute (EPRI), NP-5064M (RP2006–6); 1987. Angeliu, T. M.; Andresen, P. L.; Sutliff, J. A.; Horn, R. M. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The American Institute of Mining, Metallurgy, and Petroleum Engineers (AIME): Warrendale, PA, 1999. Angeliu, T. M.; Andresen, P. L.; Hall, E.; et al. In Corrosion/2000; NACE: Houston, TX, 2000, Paper no. 186. Othon, M. A.; Morra, M. M. Microsc. Microanal. 2005, 11 (Suppl. 2), 522–523; doi:10.1017/S1431927605506949. Andresen, P. L.; Angeliu, T. M.; Young, L. M. In Proceedings of Staehle Symposium on Chemistry and Electrochemistry of Corrosion and SCC; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 2001. Andresen, P. L. In Proceedings of 10th International Symposium on Environmental Degradation of Materials

204

61.

62.

63. 64. 65. 66. 67. 68.

69.

70.

71.

72.

73. 74. 75. 76. 77. 78. 79. 80. 81.

82.

Irradiation Assisted Stress Corrosion Cracking in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 2001. Andresen, P. L.; Young, L. M. In Proceedings 7th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995; pp 579–596. Lin, C. C. In Proceedings of 2nd Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; American Nuclear Society (ANS): LaGrange Park, IL, 1986; pp 160–172. Burns, W. G.; Moore, P. B. Radiat. Eff. 1976, 30, 233. Cohen, P. Water Coolant Technology of Power Reactors; Gordon and Breach Science Publishers: New York, NY, 1969. Proceedings of Water Chemistry of Nuclear Reactor Systems, Oct 23–27; British Nuclear Energy Society: London, 1989; No. 5. Taylor, D. F. In Corrosion/90; NACE: Houston, TX, 1990; Paper no. 501. Andresen, P. L.; Ford, F. P.; Higgins, J. P.; et al. In Proceedings of ICONE-4 Conference; ASME International: New York, NY, 1996. Head, R. A.; Indig, M. E.; Andresen, P. L. Measurement of in-core and recirculation system responses to hydrogen water chemistry at Nine Mile Point Unit 1 BWR; Electric Power Research Institute (EPRI) Contract RP2680–5; 1989; Final Report. Gordon, B. M. Hydrogen water chemistry for BWR, Task 27 – Materials and Environmental Monitoring within the Duane Arnold BWR; Electric Power Research Institute (EPRI) Contract RP1930–1; 1983. Taylor, D. F.; Caramihas, C. A. In Localized Crack Chemistry and Mechanics of Environmentally Assisted Cracking; The American Institute of Mining, Metallurgy, and Petroleum Engineers (AIME): New York, NY, 1983. Combrade, P.; Foucault, M.; Slama, G. In Proceedings of Atomic Energy Agency Specialists Meeting on Subcritical Crack Growth; Nuclear Regulatory Agency: Sendai, Japan, 1985; Vol. 2, pp 201–218, NUREG/CP0067. Gabetta, G.; Buzzanaca, G. In Proceedings of Atomic Energy Agency Specialists Meeting on Subcritical Crack Growth; Nuclear Regulatory Agency: Sendai, Japan, 1985; Vol. 2, pp 219–230, NUREG/CP0067. Andresen, P. L.; Vasatis, I. P.; Ford, F. P. In Corrosion/90; NACE: Houston, TX, 1990; Paper no. 495. Ljungberg, L. G.; Atom, A. B. B. Vasteras, Sweden, Personal communication, Mar 1991. Andresen, P. L.; Ford, F. P. In Corrosion/89; NACE: Houston, TX, 1989; Paper no. 497. Andresen, P. L.; Ford, F. P. International Cooperative Group on Irradiation Assisted Stress Corrosion Cracking (ICG-IASCC) Minutes, Apr 1989. Was, G. S. Fundamentals of Radiation Materials Science: Metals and Alloys; Springer: Berlin, 2007. Wiedersich, H.; Okamoto, P. R.; Lam, N. Q. J. Nucl. Mater. 1979, 83, 98. Okamoto, P. R.; Rehn, L. E. J. Nucl. Mater. 1979, 83, 2. Okamoto, P. R.; Wiedersich, H. J. Nucl. Mater. 1974, 53, 336. Lam, N. Q.; Kumar, A.; Wiedersich, H. In Proceedings of the Effects of Radiation on Materials, 11th Conference, ASTM STP 782; Brager, H. R., Perrin, J. S., Eds.; ASTM: West Conshohocken, PA, 1982; p 985. Perks, J. M.; Marwick, A. D.; English, C. A. In Proceedings of the Radiation-Induced Sensitisation of Stainless Steels; Norris, D. I., Ed.; Central Electricity Generating Board, Berkeley Nuclear Labs: Berkeley, Gloucestershire, 1987; p 15.

83. 84. 85. 86. 87. 88. 89.

90.

91.

92. 93. 94.

95.

96. 97. 98. 99.

100. 101. 102. 103.

104.

Jenssen, A.; Ljungberg, L. G.; Walmsley, J.; Fisher, S. Corrosion 1998, 54(1), 48. Allen, T.; Busby, J. T.; Was, G. S.; Kenik, E. A. J. Nucl. Mater. 1998, 255, 44. Was, G. S.; Allen, T. R.; Busby, J. T.; et al. Proc. Mater. Res. Soc. 1999, 540, 421–432. Hindmarsh, A. C. GEAR: Ordinary differential equation system solver; Lawrence Livermore Laboratory Report; UCID-30001; 1972. Simonen, E. Pacific Northwest National Laboratory, private communication, 2001. Was, G. S.; Allen, T. R. Mater. Charact. 1994, 32(4), 239. Asano, K.; Fukuya, K.; Nakata, K.; Kodama, M. In Proceedings of 5th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Simonen, E. P., Gold, R. E., Eds.; American Nuclear Society (ANS): LaGrange, IL, 1992; p 838. Jacobs, A. In Proceedings of 7th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 1995; p 1021. Jacobs, A. J.; Wozadlo, G. P.; Nakata, K.; et al. In Proceedings of 6th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Gold, R. E, Simonen, E. P. Eds.; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 1993; p 597. Kenik, E. A. J. Nucl. Mater. 1992, 187, 239. Nakahigashi, S.; Kodama, M.; Fukuya, K.; et al. J. Nucl. Mater. 1992, 179–181, 1061. Jacobs, A. J.; Clausing, R. E.; Miller, M. K.; Shepherd, C. M. In Proceedings of 4th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Ed.; NACE: Houston, TX, 1990; pp 14–21. Walmsley, J.; Spellward, P.; Fisher, S.; Jenssen, A. In Proceedings of 7th International Symposium on Environment Degradation of Materials in Nuclear Power System—Water Reactors; NACE: Houston, TX, 1997; p 985. Was, G. S.; Busby, J. T.; Gan, J.; et al. J. Nucl. Mater. 2002, 300, 198. Carter, R.; Damcott, D.; Atzmon, M.; Was, G. S.; Bruemmer, S. M.; Kenik, E. A. J. Nucl. Mater. 1994, 211, 70. Bruemmer, S. M.; Simonen, E. P.; Scott, P. M.; Andresen, P. L.; Was, G. S.; Nelson, J. L. J. Nucl. Mater. 1999, 274, 299. Bruemmer, S. M. In Proceedings of 10th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 2002. Kato, T.; Takahashi, H.; Izumiya, M. J. Nucl. Mater. 1992, 189, 167. Fournier, L.; Sencer, B. H.; Was, G. S.; Simonen, E. P.; Bruemmer, S. M. J. Nucl. Mater. 321 2003, 2–3, 192–209. Gan, J.; Simonen, E. P.; Bruemmer, S. M.; Fournier, L.; Sencer, B. H.; Was, G. S. J. Nucl. Mater. 2004, 325, 94–106. Dumbill, S.; Hanks, M. In Proceedings of 6th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Gold, R. E., Simonen, E. P., Eds.; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 1993; p 521. Nelson, J. L.; Andresen, P. L. In Proceedings of 5th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Cubicciotti, D., Simonen, E. P., Gold, R. E., Eds.;

Irradiation Assisted Stress Corrosion Cracking

105. 106. 107. 108.

109. 110. 111.

112.

113. 114. 115. 116. 117. 118. 119.

120. 121. 122. 123. 124. 125.

126. 127. 128. 129.

American Nuclear Society (ANS): La Grange Park, IL, 1992; p 10. Was, G. S.; Allen, T. R.; Busby, J. T.; et al. J. Nucl. Mater. 1999, 270, 96. Bruemmer, S. M.; Was, G. S. J. Nucl. Mater. 1994, 216, 348. Busby, J. T.; Was, G. S.; Kenik, E. A. J. Nucl. Mater. 2002, 302, 20. Busby, J. T.; Was, G. S. In Proceedings of 11th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; American Nuclear Society (ANS): La Grange Park, IL, 2003; p 995. Yonezawa, T.; Fujimoto, K.; Kanasaki, H.; et al. Corros. Eng 2000, 49, 655. Fujimoto, K.; Yonezawa, T.; Iwamura, T.; Ajiki, K.; Urata, S. Corros. Eng 2000, 49, 701. Yonezawa, T.; Fujimoto, K.; Iwamura, T.; Nishida, S. In Environmentally Assisted Cracking: Predictive Methods for Risk Assessment and Evaluation of Materials, Equipment, and Structures, ASTM STP 1401; Kane, R. D., Ed.; ASTM: West Conshohocken, PA, 2000; p 224. Yonezawa, T.; Iwamura, T.; Fujimoto, K.; Ajiki, K. In Proceedings of 9th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 1999; p 1015. Li, G. F.; Kaneshima, Y.; Shoji, T. Corrosion 2000, 56(5), 540. Maziasz, P. J.; McHargue, C. J. Int. Met. Rev. 1987, 32, 190. Maziasz, P. J. J. Nucl. Mater. 1993, 205, 118. Garner, F. A. J. Nucl. Mater. 1993, 205, 98. Zinkle, S. J.; Maziasz, P. J.; Stoller, R. E. J. Nucl. Mater. 1993, 206, 266. Edwards, D. J.; Simonen, E. P.; Bruemmer, S. M. J. Nucl. Mater. 2003, 317, 13. Edwards, D. J.; Oliver, B. A.; Garner, F. A.; Bruemmer, S. M. In Proceedings of 10th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; NACE: Houston, TX, 2002. Edwards, D. J.; Simonen, E. P.; Garner, F. A.; Greenwood, L. R.; Oliver, B. M.; Bruemmer, S. M. J. Nucl. Mater. 2000, 317, 32. Stoller, R. JOM 1996, 12, 23. Busby, J. T.; Sowa, M. M.; Was, G. S.; Simonen, E. P. Phil. Mag. 85 2005, 4–7, 609–617. Edwards, D. J.; Simonen, E. P.; Bruemmer, S. M. J. Nucl. Mater. 2003, 317, 13. Busby, J. T.; Kenik, E. A.; Was, G. S. J. Nucl. Mater. (in press). Edwards, D. J.; Simonen, E. P.; Bruemmer, S. M. Radiation-induced segregation behavior in austenitic stainless steels: Fast reactor versus light water reactor irradiations. In Proceedings of 13th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Canadian Nuclear Soc, Aug 2007. Watanabe, H.; Muroga, T.; Yoshida, N. J. Nucl. Mater. 1996, 239, 95. Hackett, M. J.; Busby, J. T.; Miller, M.; Was, G. S. J. Nucl. Mater. 2009, 265. Singh, B. N.; Foreman, A. J. E.; Trinkaus, H. J. Nucl. Mater. 1997, 249, 103. Seeger, A. In Proceedings of 2nd United Nations International Conference on the Peaceful Uses of Atomic Energy; United Nations: New York, NY, 1958; vol. 6, p 250.

130.

131. 132.

133.

134. 135. 136.

137.

138. 139. 140. 141.

142. 143. 144.

145. 146. 147. 148.

149.

205

Bruemmer, S. M.; Cole, J.; Carter, R.; Was, G. S. In Proceedings of 6th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Gold, R. E., Simonen, E. P., Eds.; The Minerals, Metals, and Materials Society (TMS): Warrendale, PA, 1993; p 537. Bloom, E. E. In Radiation Damage in Metals; Peterson, N. L., Harkness, S. D., Eds.; ASM International: Materials Park, OH, 1975; p 295. Speidel, M. O.; Magdowski, R. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Ford, F. P., Bruemmer, S. M., Was, G. S., Eds.; Metallurgical Society of the American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): Warrendale, PA, 1999; p 325. Was, G. S. In Proceedings of 11th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; American Nuclear Society (ANS): La Grange Park, IL, 2003; p 965. Katsura, S.; Ishiyama, Y.; Yokota, N.; et al. In Corrosion/ 98, NACE: Houston, TX, 1998; Paper no. 132. Jacobs, A.; Wozadlo, G. P.; Gordon, G. M. Corrosion 1995, 51, 10, 731. Hash, M. C.; Wang, L. M.; Busby, J. T.; Was, G. S. In 21st International Symposium on Effects of Radiation on Materials, ASTM STP 1447; Grossbeck, M. R., Allen, T. R., Lott, R. G., Kumar, A. S., Eds.; ASTM: West Conshohocken, PA, 2004; pp 92–104. Kodama, M.; Fukuya, K.; Kayano, H. In Proceedings16th International Symposium on Radiation Effects on Materials, ASTM STP 1175; Kumar, A. S., Gelles, D. S., Nanstad, R. K., Little, E. A., Eds.; ASTM: West Conshohocken, PA, 1994; p 889. Jenssen, A.; Ljungberg, L.; Walmsley, J.; Fisher, S. In Corrosion/96; NACE International: Houston, TX, 1996; paper no. 101. Norris, D. I. R.; Baker, C.; Taylor, C.; Titchmarsh, J. M. In Materials for Nuclear Reactor Core Applications; BNES: London, 1987; Paper no. 52. Swann, P. R. In Corrosion 19; 1963; 102t. Thompson, A. W.; Bernstein, I. M. In Advances in Corrosion Science and Technology; Staehle, R. W., Fontana, M. G., Eds.; Plenum Press: New York, NY, 1980; Vol. 7, p 53. Rhodes, C. G.; Thompson, A. W. Met. Trans. 1977, 8A, 1901. Schramm, R. E.; Reed, R. P. Met. Trans. A. 1975, 6A, 1345. Farrell, K.; Byun, T. S.; Hashimoto, N. Mapping flow localization processes in deformation of irradiated reactor structural alloys; Oak Ridge National Laboratory, Report ORNL/TM-2002/66; July 2002. Stroh, A. N. Phil. Mag. 1957, 3, 597. Alexandreanu, B.; Was, G. S. Corrosion 2003, 59(8), 705–720. Bailat, C.; Almazouzi, A.; Baluc, M.; Schaublin, R.; Groschel, F.; Victoria, M. J. Nucl. Mater. 2000, 283–287, 446. Busby, J. T.; Sowa, M. M.; Was, G. S. In 21st International Symposium on Effects of Radiation on Material; Grossbeck, M. R., Allen, T. R., Lott, R. G., Kumar, A. S., Eds.; ASTM: West Conshohocken, PA, 2004; pp 78–91. Thomas, L. E.; Bruemmer, S. M. In Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Ford; F. P., Bruemmer, S. M., Was, G. S., Eds.; Metallurgical Society of the American Institute of Mining, Metallurgical, and Petroleum Engineers (AIME): Warrendale, PA, 1999; p 41.

5.09

Material Performance in Lead and Lead-bismuth Alloy

K. Kikuchi Ibaraki University, Ibaraki, Japan

ß 2012 Elsevier Ltd. All rights reserved.

5.09.1

Recent Lead-Alloy Activity

207

5.09.2 5.09.2.1 5.09.3 5.09.4 5.09.5 5.09.6 5.09.7 5.09.8 References

Utilization of LA The Conceptual Models of ADS and MYRRHA Ferritic–Martensitic Steels Surface Treatment to F/M and Austenitic Steels Oxide Dispersion-Strengthened Steel Austenitic Stainless Steels Precipitation Formation Outlook

209 209 210 213 214 215 216 217 218

Abbreviations ADS AFM BEM DBTT EB EDX F/M steel GESA GIF ICP LA LBE LFR LINAC MA MEGAPIE MFM MYRRHA ODS OECD/NEA

SEM WDX

Accelerator-driven nuclear transmutation system Atomic force microscopy Backscattered electron microscope Ductile-to-brittle transition temperature Electron beam Energy-dispersed X-ray analyzer Ferritic–martensitic steel Gepulste Elektronenstrahlanlage Generation IV International Forum Inductive-coupled plasma atomic emission spectrometer Lead alloy Lead–bismuth eutectics Liquid-metal-cooled fast reactor Linear accelerator Minor actinides MEGA-watt Pilot Experiment Magnetic force microscopy Multipurpose hybrid research reactor for high-tech applications Oxide dispersion-strengthened steel The Organisation for Economic Co-operation and Development/ The Nuclear Energy Agency Scanning electron microscopy Wave-dispersed X-ray analyzer

5.09.1 Recent Lead-Alloy Activity A brief justification for the utilization of lead or lead bismuth for use as a coolant in nuclear energy systems was given in 2001 by Sekimoto.1 When the possibility of the utilization of nuclear energy was discovered, it was expected to be a primary energy source in the future. Fast reactors can utilize the entire energy content of natural uranium. The selection of a coolant was an important item for designing fast reactors. The neutron slowing-down caused by the coolant should be minimized. This is first made possible by decreasing the average atomic density of the coolant in the reactor core, and second by employing a nuclide with a large mass number as the coolant, whose neutron moderating power is low. A liquid metal is considered the best coolant for using the second method. Initially, liquid mercury was employed but it was not successful in either the United States or Russia. Since then, several liquid metals were considered, including lead alloys (LA), and finally, sodium was selected. However, public concern about the safety of sodium has increased following sodium leakage incidents, so the development and deployment of fast reactors on more than a prototype scale has not occurred. In the last 10 years, the study of the utilization of LA including lead–bismuth eutectics (LBE) has been ongoing for application to nuclear waste transmutation systems and lead–bismuth cooled nuclear reactors.

207

208

Material Performance in Lead and Lead-bismuth Alloy

LBE is a candidate material for a spallation target and a reactor coolant. In the accelerator-driven nuclear transmutation system (ADS), LBE is a candidate for both the subcritical-reactor coolant and the spallation neutron source target. In addition, the lead or lead– bismuth-cooled fast reactor (LFR) is one of the four reactor types investigated in Generation IV systems proposed by the Generation IV International Forum (GIF). A LBE-cooled Long-Life Safety Simple Small Portable Proliferation-Resistant Reactor has also been proposed.2 As a result of the investigations on LA, comprehensive literature has been published. The Working Group on LBE of the OECD/NEA Nuclear Science Committee3 published a handbook and review reports on LA technology. The material properties of lead and lead–bismuth are discussed in detail in Chapter 2.14, Properties of Liquid Metal Coolants. As part of the development of advanced nuclear systems, including ADS proposed for high-level radioactive waste transmutation and Generation IV reactors, heavy liquid metals such as lead or LBE were investigated as reactor core coolant and spallation targets. Heavy liquid metals were also being envisaged as target materials for high-power neutron spallation sources. The objective of the handbook is to collate and publish properties and experimental results on lead and LBE in a consistent format in order to provide designers with a single source of qualified properties and data and to guide subsequent development efforts. The handbook covers liquid lead and LBE properties, material compatibility and testing issues, key aspects of the thermal-hydraulic and system technologies, existing test facilities, and open issues and perspectives. Zhang and Li4 reviewed the studies on fundamental issues in LBE corrosion. They included phase diagrams, thermodynamics, physical properties, corrosion mechanisms, oxygen control, experimental results, and corrosion results. Some recommendations were proposed for future studies: precipitation and deposition of corrosion products; oxygen transport; oxide formation and kinetics in LA; coolant hydrodynamic effects; steel composition, microstructure, and surface effects; and corrosion models. These are key areas for future research. Fazio et al.5 characterized corrosion property for ferritic–martensitic (F/M) steels and austenitic steels in stagnant LA on the basis of the results of corrosion tests. This report briefly summarized the current status on LA activities. At a temperature below 450
C, adequate oxygen activities in the liquid

metal steels form an oxide layer that behaves as a corrosion barrier. In the temperature range above 500
C, corrosion protection because of the oxide scales seems to fail. A mixed corrosion mechanism has been observed, where both oxide scale formation and dissolution of the steel elements occurred. However, in this high-temperature range, it has been demonstrated that the corrosion resistance of structural materials can be enhanced by coating the steel with FeAl alloys. Experiments performed in flowing LA (mostly LBE) confirm that the corrosion mechanism of the steels depends on the oxygen content in LA. At relatively low oxygen concentration, the corrosion mechanism changes from oxidation to dissolution of the steel elements. The experimental activity also extends up to temperatures of 750
C for oxide dispersion-strengthened (ODS) alloys and their welded variants in Pb. The use of materials at higher temperatures will also require investigation of creep rupture. MEGAPIE was the MEGA-watt Pilot Experiment done at Paul Scherrer Institut (PSI) in 2006 for developing a LBE spallation target. The MEGAPIE project was started as an essential step toward demonstrating the feasibility of coupling a high power accelerator, a spallation target, and a subcritical core assembly. The project was expected to furnish important results regarding safe treatment of components that had come into contact with lead–bismuth.6 The design data was obtained and the operational mode was confirmed.7 Corrosion rates were estimated experimentally at 400
C for a LBE flow rate of 1 m s1 and 2.2 m s1 where the oxygen content in the LBE was <107 wt%. No protective oxide layer was produced on the steel surface. This oxygen content has been considered representative of the MEGAPIE conditions, as no oxygen control and monitoring system is anticipated to be used in the target. The estimated corrosion rates, 40–86 mm year1, indicate that in the given testing conditions, the corrosion resistance of the steel does not represent a critical issue, especially since LBE temperature is expected to be lower (320
C). The goals of the experiment were fully accomplished8: 4 months of reliable and essentially uninterrupted operation (beam trips and short beam interruptions permitted) at a power level as high as the accelerator was able to deliver (about 0.75 MW) excellent performance of the target and the dedicated ancillary systems, the proof of functionality of advanced proton beam safety devices, and, last but not least, a superb neutronic efficiency delivering about 80% more neutrons for

Material Performance in Lead and Lead-bismuth Alloy

the users compared to the previously operated leadcannelloni target. Verification of performance will be scheduled in the postirradiation experiment.

material usage in design studies. The material temperature at contact with LBE is slightly <500
C in the spallation reaction area and <550
C in the fuel core area under normal conditions. Figure 1 shows the ADS concept. A superconducting linear accelerator (LINAC) is connected with a subcritical fast reactor. A high-energy proton beam is injected into the core of the reactor. Spallation reactions produce a number of neutrons from the lead–bismuth nuclei, which are then used to transmute minor actinides (MA). The interface between the beam duct and lead bismuth is called the beam window. For example, a tank type reactor with 800 MW thermal power and LBE-coolant and spallation target was proposed.11–13 The proton beam energy was set at 1.5 GeV. The beam current varied between 10 and 20 mA according to criticality swings. In the steady-state condition, as the beam window material generates heat by spallation reactions and is cooled by flowing LBE. A temperature difference is established between the LBE, the material in contact with the LBE, and the material on the other side of the window, with the temperatures being 400, 450, and 500
C, respectively. As the MA core cladding material is gamma heated and the fuel adds to the radiation heat, temperatures reach, for example, a maximum of 500, 550, and 600
C. The maximum average velocity in the particular flow channel of LBE is 1.8 and 2.0 m s1, at the window and in the MA core region, respectively. Figure 2 shows the conceptual model of MYRRHA consisting of an inner vessel, guard vessel,

5.09.2 Utilization of LA 5.09.2.1 The Conceptual Models of ADS and MYRRHA Recent activity on materials research and development in LA, especially LBE, aims at realizing ADS, MEGAPIE, LFR, and MYRRHA (multipurpose hybrid research reactor for high-tech applications).9,10 It is valuable to know each specific environment for

Liq.He RF ADS

Injector RFQ

DTL

Superconducting LINAC

Beam duct

P Beam window

Beam window MA PbBi (Am,Cm)

n MA

Subcritical reactor

PbBi

209

Spallation reaction

Figure 1 The conceptual model of accelerator-driven nuclear transmutation system with beam window.

11

10

1. Inner vessel 2. Guard vessel 3. Cooling tubes

4

4. Cover 5. Diaphragm 6. Spallation loop 7. Subcritical core

5 8 12

9

6

11 8

7

9 10

7

11

3 13

2 1

12

8. Primary pumps 9. Primary heat exchangers 10. Emergency heat exchangers 11. In-vessel fuel transfer machine 12. In-vessel fuel storage 13. Coolant conditioning system

Figure 2 The conceptual model of subcritical reactor in multipurpose hybrid research reactor for high-tech applications. Courtesy of J Bosch. ADS Candidate Materials Compatibility with Liquid Metal in a Neutron Irradiation Environment, Doctoral Thesis, ISBN 978-90-8578-241-4, 2008; 7.

210

Material Performance in Lead and Lead-bismuth Alloy

cooling tubes, spallation loops, primary heat exchangers, and so on, but without a beam window.15 In this system, a high-energy proton beam with an energy of 600 MeV is injected directly into the free surface of the lead–bismuth in the subcritical reactor core. The MYRRHA project aims to serve as a basis for the European experimental ADS. In the first stage, the project focuses mainly on demonstrating the ADS concept, safety research of subcritical systems, and on nuclear waste transmutation studies. Subsequently, MYRRHA will be used as a fast spectrum irradiation facility dedicated to research on structural materials, nuclear fuel, liquid metal technology, and associated aspects on the one hand and as a radioisotope production facility on the other. The system consists of a proton accelerator that supplies a 600 MeV  3–4 mA proton beam to a LBE spallation target, delivering the primary neutrons, which in turn couples to a LBEcooled subcritical fast core. The structural materials for MYRRHA need to withstand temperatures ranging between 200 and 550
C (normal operating temperature between 300 and 450
C) under high spallation neutron flux and contact with liquid LBE. It is clear that the candidate materials need to fulfill challenging requirements such as high thermal conductivity, high heat resistance, low thermal expansion, low ductileto-brittle transition temperature (DBTT) shift, sufficient strength at elevated temperatures with limited loss of ductility and toughness, low swelling rate, high creep resistance, and good corrosion resistance.14 Studies of LA for developing ADS are also reported from the points of view of conceptual ideas16,17 and related facility.18

5.09.3 Ferritic–Martensitic Steels One method of using materials such as F/M and austenitic stainless steels in LA is to keep an oxide layer on the surface of the base metal in contact with LA by controlling the oxygen concentration in the LA.19–21 Too little oxygen in LA will lead to dissolution of the protective iron oxide. Excess oxygen solution in the LA will lead to the production of lead oxide that could plug the cooling tubes. Theory predicts that an adequate oxygen concentration in LA exists between, for example, 106 and 104 wt% in the temperature region of 400 and 700
C. An alternative method is to add anticorrosion elements such as Al to the surface, which leads to a protective oxide that guards base metals, as mentioned in the section on surface treatment.

The oxide scale is not a simple structure but consists of duplex layers: magnetite Fe3O4 near the LA side and spinel (FeCr)3O4 near the base metal. The original surface exits at the interface between the magnetite and spinel but not at the front surface of the magnetite near the LA. An early question was how the oxide layers on the surface of the base metal grew. Martinelli et al.22–24 reported a global study on the oxidation process of Fe–9Cr–1Mo martensitic steel (T91) in static LBE. The isotope tracer oxygen-18 was employed in the corrosion test. Also, the mass balance of Fe and Cr was investigated theoretically. They explained the Fe–Cr spinel growth rate mechanism as follows: The oxidation reaction can occur because of the presence of nano-channels. Nano-channel formation is achieved by the dissociative/perforative growth in the magnetite. The nano-channel allows a fast diffusion of oxygen to the T91/spinel interface. Oxygen cannot diffuse in the oxide lattice because its rate is insufficient for Fe–Cr spinel formation, but is instead transported via short cut diffusion paths. Even if oxygen diffusion in grain boundaries could be possible, oxygen would likely diffuse inside nano-channels. The nanochannels are, in some cases, called lead nano-channels because of the results of the LBE oxidation tests. Liquid metal does not penetrate evenly in the oxide scales; only lead penetrations are observed. Nevertheless, in pure bismuth oxidation tests, bismuth penetrations are also observed in the scales. On the other hand, the iron diffusion from T91 to the magnetite/Pb–Bi interface leads to vacancy formation at the T91/Fe–Cr spinel interface. Because of the presence of chromium atoms, these vacancies can accumulate to form nano-cavities at the T91/Fe–Cr spinel interface. This accumulation is quasi complete; very few cavities are annihilated on the T91/oxide interface. The Fe–Cr spinel grows inside the nanocavity until it is completely filled. At that moment, the oxygen can no longer reach the T91 alloy and the oxidation reaction interrupts itself. The formed Fe–Cr spinel thickness then becomes equal to the consumed T91 thickness because of this self-regulation process, as shown in Figure 3. In this process, the limiting step of the Fe–Cr spinel growth rate is thus the ‘iron diffusion’ across the oxide scale. A key issue in maintaining structural integrity is to maintain high performance of the welded materials. The corrosion properties between the base metal and the weldment were investigated.25 The materials tested were F/M steel F82H26 and the electron beam (EB) welding of F82H. The chemical composition

Material Performance in Lead and Lead-bismuth Alloy

Oxygen

Nano-channel

211

Nano-channel

LBE

LBE Oxide

Iron

Oxide

Original metal surface Nano cavity

Newly formed oxide

Figure 3 Self-regulation of the Fe–Cr spinel growth.

of F82H is 8Cr–2W–0.2VTa–bal/Fe (wt%). Oxygen concentration was controlled to (2–4)  105 mass %. Welded materials were prepared with a bead-onplate weldment with a 15 mm depth of melting. F82H steel was welded after preheating at 300
C, heat-treated at 300
C for 2 h, and then annealed at 750
C for 2 h for stress relief. Figure 4 shows optical microscope observation of cross-section for F82H specimens and an impinging-flow simulation around the specimen. It was observed that the welded metal of F82H revealed a coarse martensitic structure in comparison with the fine microstructure in the nonwelded region because of melting and resolidification in the welding process. The corrosion depth in F82H was limited near the surface of the material. A failure of the outside layer in the duplex corrosion layers was observed. The heat-affected zone showed that the martensitic structure became fine because of the rapid heating and cooling during the welding process. Regardless of the difference in microstructures, the corrosion layer showed no apparent difference. The growth of the corrosion depth, defined by the layers of magnetite and spinel, followed a parabolic law, where diffusion controls the process. The result of the flow simulation of LBE impingement indicated that the velocity varied from 0 to 1 m s1 near the specimen surface. At higher temperatures, for example, above 500
C, the internal oxide layer or diffusion zone was clearly identified. Furukawa et al.27 observed three layers, consisting of the duplex layers mentioned earlier and a diffusion zone in the base metal beneath the spinel layer in the static LBE test at 500 and 550
C under the oxygen control to 106 wt% for high Cr steel (10.54Cr–1.75W– MnMoV) with heat treatment: 1070
C, 100 min aircooled; 770
C, 440 min air-cooled. Tan and Allen tested high Cr steel material in the DELTA loop, at Los Alamos National Laboratory (LANL). The material tested was HCM12A, procured from Sumitomo Metal Industries, Ltd., with

composition provided by the supplier: 10.83Cr– 1.89W–1.02Cu–0.64Mn–0.39Ni–0.30Mo–0.27Si–0.19V– 0.11C–0.063N–0.054Nb–0.016P–0.002S–0.001Al–3.1  105 B, and balance/Fe (wt%).28 The chemical composition and heat treatment of this material are slightly different from those used in the experiment by Furukawa and Muller. HCM12A is one of the thirdgeneration 12Cr ferritic steels with tempered martensite,29 which was originally developed for heavy section components such as headers and steam pipes for use at temperatures up to 620
C and pressures up to 34 MPa30 with good resistance to thermal shocks.31 The HCM12A was received after being annealed at 1050
C and tempered at 770
C.28 They compared the oxide layer to the porous magnetite layer on the supercritical water exposed sample at 600
C, 667 h. Temperatures at both conditions were different. It was found that detachment of most of the magnetite nonprotective layer occurred on the LBE-exposed sample at 530
C–600 h earlier in time than models developed by Zhang and Li. From a technical experimental point of view, it is the issue how to detect the original surface of base metal in order to evaluate the oxide thickness. A thin yttrium coating layer will help to detect it in the LBE corrosion test. At temperatures above 600
C, the oxide layer grew thinner with increasing temperature, which suggests that around this temperature, a change occurred in the mechanism of oxidation. At 570
C, FeO-wustite is formed. Compared with magnetite, wustite has a lower standard free energy of formation, which ensures its stable existence at low oxygen potential. In fact, the layer was formed in the region between magnetite and base metal. Also in this temperature range that is beyond the point of oxidation mechanism change, dissolution attack was observed at several points, and the number of such points increased with prolongation of run duration. The observations would suggest lowering the maximum processing temperature in LBE applications from the point of view of the static LBE test.27

212

Material Performance in Lead and Lead-bismuth Alloy

Welded zone Heat-affected zone

Tip

1 mm

(a)

1 mm

(b)

Oxide layer

20 µm

Oxide layer 20 µm

(d)

(c)

Oxide layer Oxide layer

Pb–Bi (e)

(f) STAR

Specimen PROSTAR 3.10 Velocity magnitude M/S Local MX = 1.089 Local MN = 0.7602E-03 *Presentation grid* 1.089 1.012 0.9338 0.8560 0.7783 0.7005 0.6228 0.5450 0.4673 0.3895 0.3118 0.2340 0.1563 0.7851E−01 0.7603E−03

LBE flow Y

x z

(g) Figure 4 Optical microscope observation of cross-section for F82H specimens and an impinging-flow simulation around the specimen. (a) Macro structure, including welded zone and heat-affected zone, (b) macro structure at the specimen end where lead–bismuth eutectics impinges from the right hand side indicated with an arrow, (c) micro structure of welded zone tested at 450
C for 1000 h, (d) micro structure of tip region tested at 450
C for 1000 h, (e) cross-section of tip region tested at 450
C for 3000 h, (f) cross-section of tip region tested at 500
C for 1000 h, and (g) simulated flow profile of lead–bismuth eutectics around the specimen.

Material Performance in Lead and Lead-bismuth Alloy

Hosemann et al. attempted nano-scale characterization of HT-9 (11.95Cr–1Mo–0.6Mn–0.57Ni– 0.5W–0.4Si–0.33V–bal/Fe (wt%)) by using atomic force microscopy (AFM), using a function of magnetic force microscopy (MFM) and C-AFM. C-AFM is a contact mode electrical characterization technique that involves applying a voltage typically between the conductive AFM tip and the sample while monitoring variations in the local electrical properties in a range of picoamperes to microamperes. The HT-9 tube was tested at 550
C in flowing LBE under 106 wt% oxygen for 3000 h.32 It was found that the oxide consists of at least four different layers with different grain structures and therefore conductivity/magnetic properties. The outer layers seem to be Fe3O4 and have good conductivity, while the inner layer is Cr enriched and has lower conductivity or is insulating. This is in agreement with the literature where Cr additions lower the conductivity of Fe3O4. The outer layer can be divided into two distinct areas based on a change in grain structure. The inner oxide layers adopt the grain structure from the bulk steel. High pore density within these layers suggests that these are fast diffusion paths allowing Fe diffusion outward and O diffusion inward. The LBE corrosion experiment in the DELTA Loop on T91, HT-9, and EP823 conducted for 600 h at 535
C showed multilayer oxides on the tested materials. The wave-dispersed X-ray analyzer (WDX) measurements on the cross-sections revealed two Cr and Fe containing oxide layers and no Fe3O4 layer. It appears that the main difference between observed oxide layers is the Fe content and the microstructure. Nano-indentation tests across the oxide layers were performed.33 The results showed lower values of E-modulus in these oxide layers than that of the bulk steel layers and higher hardness values for the oxides than that of the bulk steel. The inner oxide layer is softer than the outer oxide layer. This might be due to the fact that the inner oxide layer has higher porosity than the outer layer. Yamaki and Kikuchi34 conducted a mechanical test of oxide scales. The beam window at the boundary of the high-energy proton beam and reactor core, as shown in Figure 1, is loaded by thermal stress and buckling load in the deep LBE of the reactor.35 The specimen was a ring made from the F/M steel pipe, HCM12A. The inner surface of the pipe had been exposed to flowing LBE during the loop operation at 400–500
C for 5500 h under an oxygen concentration in the range from 1  105 to 5  105 wt%. Apparently, the oxide layer had a duplex structure. Possibly they were outside the magnetite and inner

213

side spinel. Figure 5 shows the results of the ring compression test. The HCM12A ring was compressed by 50% and unloaded. Near position A, cracking occurred because of excess strain to the spinel layer rather than the Fe3O4 layer. This was caused by the fact that the Young’s modulus of Fe–Cr spinel layer was lower than that of Fe3O4 layer by 10% with the same hardness in both layers. Near position E, the oxide layer was spalled off from the boundary between the base metal and the spinel. F/M steels can guard the base metal by forming a spinel oxide layer. The formation mechanism is controlled by the iron diffusion rating. The magnetite oxide layer is not protected against contact with LBE. Under the tensile stresses, excess strains will spall off the oxide layer from the interface between the base metal and the spinel layer. Over 570
C, the oxide formation mechanism is changed by the formation of wustite.

5.09.4 Surface Treatment to F/M and Austenitic Steels Muller et al. demonstrated that the effect of Al-alloying into the surface of the base metal was the F/M steel OPTIFER IVc (10Cr–0.58Mn–0.56C–0.40W– 0.28V–bal/Fe (wt%)) using electron pulse treatment, GESA (Gepulste Elektronenstrahlanlage).36,37 There was no corrosion attack visible in any part of the alloyed portion after 1500 h exposure to liquid lead at 550
C with 8  106 at.% oxygen. The alumina layer that must have formed at the surface during oxidation in lead might be very thin and could not be detected. Only the unalloyed part of the surface was covered with thick oxide scales. The results also suggested that the Fe–Cr spinel layer ends at the original specimen surface. This surface treatment had a similar result when applied to an austenitic steel base metal 1.4970(16.5Cr–13.8Ni–1.91Mn– 0.81Si–MoTi–bal/Fe (wt%)). Weisenburger et al. examined T91 tubes with modified FeCrAlY coatings in LBE. These coatings are often used for turbine blade protection.38 The coating had an average thickness of 30 mm after application by a plasma spray method and was remelted using the pulsed large area GESA EB to gain a dense coating layer and to improve the bonding between the coating and the bulk material. They intended to simulate a cladding material’s environment by using a pressurized tube type specimen. The results showed

214

Material Performance in Lead and Lead-bismuth Alloy

Load Load

Delamination

Initial diameter (48 mm)

Final diameter (24 mm)

Specimen Flat plate

(a)

d width

Crack Base metal

d interval

(b) 10 mm A

B C

I

F H G

D

27.5 mm

E

(e)

Oxide scales

10 mm

(c) Base metal

A

No oxide scale

B C 100 mm

D

(f) E

(d) Figure 5 The ring compression test. (a) HCM12A ring model before compression, (b) the ring model after compression by 50%, (c) the ring after unloading, (d) simulation of maximum principal strain distribution induced at loading, (e) the cross-section near position A, and (f) the cross-section near position E.

that the tangential wall stress of about 112.5 MPa induced by an internal tube pressure of 15 MPa increased the Fe diffusion and led to enhanced magnetite scale growth. Coated specimens, however, have no magnetite layer. This is another advantage of the coating. Energy-dispersed X-ray analyzer (EDX) line scans of the cross-section of the coated T91 tube specimen after 2000 h exposure to LBE at 600
C show that the top oxide scale must consist mainly of alumina followed by a thin layer enriched in Cr. These layers protect the steel not only from LBE attack but also from oxygen diffusion into the coating and bulk material. The coating process needs some improvement to avoid coating regions with aluminum concentrations below 4 wt%; otherwise, the oxide layer will grow in the same manner as the original material. Steels with 8–15 wt% Al alloyed into the surface suffer no corrosion attack for all experimental temperatures and exposure times.39 Technical concerns about surface treatment are the effect of cyclic loading on the low cycle fatigue endurance in air and LBE. Low cycle fatigue tests were conducted in LBE containing 106 wt% dissolved oxygen with T91 steel at 550
C. T91 was

employed in two modifications, one in the as-received state and the other after alloying FeCrAlY into the surface by pulsed EB treatment (GESA process). Tests were carried out with symmetrical cycling (R ¼ 1) with a frequency of 0.5 Hz and a total elongation Det/2 between 0.3% and 2%. No fatigue effects from LBE could be detected. Results in air and LBE showed similar behavior. Additionally, no difference was observed between surface treated and nontreated T91 specimens.14 A melting process of coating materials enhances bonding between the coating and the bulk materials. Heat deposition because of a pulsed EB exposure successfully demonstrated that remelted alumina or FeCrAlY coating was effective in protecting the base metal property from LBE attack.

5.09.5 Oxide DispersionStrengthened Steel Takaya et al.40 investigated the corrosion resistance of ODS steels with 0–3.5 wt% Al and 13.7–17.3 wt% Cr, at 550 and 650
C for up to 3000 h in stagnant LBE

Material Performance in Lead and Lead-bismuth Alloy

containing 106 and 108 wt% oxygen. The ODS steels were manufactured by hot extrusion of mechanically alloyed powders at 1150
C, and consolidated bars were annealed by 60 min of heat treatment at 1150
C, followed by air cooling. Chemical compositions of ODS materials are (13.7–17.3) Cr–(1.9–3.5) Al–(0.34–0.36)Y2O3–TiSiMn–bal/Fe (wt%). Protective Al oxide scales formed on the surfaces of the ODS steels with 3.5 wt% Al and 14–17 wt% Cr, and no dissolution attack was seen in any of the cases. Addition of Al is very effective in improving the corrosion resistance of ODS steels in LBE. On the other hand, the ODS steel with 16 wt% Cr and no Al showed no corrosion resistance, except in the case of exposure to LBE with 106 wt% oxygen at 650
C. Thus, the corrosion resistance of ODS steels in LBE may not be improved solely by increasing Cr concentration. There is additional data reported by Hosemann et al. on ODS alloys in LBE. Specimens were exposed to flowing LBE in the DELTA Loop at LANL at 535
C for 200 and 600 h. The oxygen content in the LBE was about 106 wt%. The detailed manufacturing process was not disclosed. Conclusively, PM2000, which has a chemical composition of 20Cr– 5.5Al–0.5Y2O3–0.5Ti–bal/Fe (wt%), showed a very dense, thin, and protective oxide layer because of its higher Al content. The compositions of the oxide layers found on the Al alloyed materials change with depth. Elements are oxidized based on the amount of oxygen available for oxidation and the free energy of the oxide. It appears that at least 5.5 wt% Al in the alloy is necessary to form a protective Al-enriched oxide.41 The oxide scale has a duplex structure below 500
C. Over 500
C, a diffusion zone in the base metal is apparently observed. The oxide layer appears to consist of three layers, that is, the duplex layers plus the diffusion zones. The oxide scale does become unstable. The outer magnetite layer is prone to be spalled off in the flowing LA. In such an environment, the Al coating is found to be effective in enhancing corrosion resistance. Remelting processes, for example, by GESA EB exposure, make a good Al layer. The disadvantage of the coating method is the disintegration or cracking due to an uncontrolled process. This cracking could be attributed to the local reduction of Al content. The ODS alloy is developed for cladding materials. Materials development is progressing in the direction of highCr Al-ODS alloys. The recommended Al composition in ODS alloys varied from 3.5 to 5.5. An adequate amount of Al will balance the corrosion resistance and mechanical strength. Excess Al will reduce mechanical

215

strength. The reason why the Al enrichment in Febase steel improves corrosion resistance in LA will be determined in future investigations. ODS steel aims at enhancing the strength of material applicable to the cladding materials of a fast reactor. The addition of Al to ODS improves corrosion resistance in LBE at the fuel cladding temperatures. On the other hand, the excess addition of aluminum reduces the strength of materials. An optimization is needed to balance the two factors at around 5%.

5.09.6 Austenitic Stainless Steels Austenitic stainless steels are candidate materials for the spallation target window in ADS. In MEGAPIE, however, F/M steel, T91, was used for the beam window in flowing LA, and this was acceptable for a limited duration (4 months). The lifetime of the beam window of the T91 liquid Pb–Bi container in the MEGAPIE target was summarized based on the present knowledge of LBE corrosion, embrittlement, and radiation effects in the relevant condition.42 It was suggested that the lower bound of the lifetime of the T91 beam window was determined when the steel became brittle at the lowest operation temperature, 230
C, with a safety margin of 30%. Evaluation using the DBTT data and fracture toughness values of T91 specimens tested in LBE, a dose limit of about 6 dpa, corresponding to 2.4Ah proton charge to be received by the target in about 20 weeks in the normal operation condition, was set. In the ADS design, for example, the beam window material will produce about 1000 appm (3He þ 4He) a year by 1.5 GeV proton beam bombardment in the reactor core for austenitic stainless steel, Japanese Primary Candidate Alloy (JPCA), and F/M steel, F82H.43 The helium production of 1000 appm He suggests that the DBTT will increase by 400– 500
C.44 This increase will set the design temperature at the beam window at 450–500
C. The use of a F/M steel may lead to a brittle fracture, and those materials should be avoided in operations for extended times. Therefore, austenitic steel is the candidate material. The production of hydrogen and helium in JPCA was slightly larger, 3–4%, than that of F82H because of the addition of nickel and boron. JPCA, in which the chemical composition is 0.50Si–1.77Mn–0.027P–0.005S–15.60Ni–14.22 Cr–2.28Mo–0.24Ti–0.0031B–0.0039N–bal/Fe (wt%), was developed to reduce the helium embrittlement

216

Material Performance in Lead and Lead-bismuth Alloy

of austenitic steel for first wall and blanket structural components in fusion reactors.45 The optimized JPCA material is manufactured by vacuum induction melting, vacuum arc melting, and solution-annealing at 1100
C for 1 h. The TiC precipitates within the matrix and on the grain boundaries serve as trapping centers for the helium produced during neutron irradiation. However, dissolution of the MC precipitates initiates the onset of helium embrittlement as well as high swelling during high fluence neutron irradiation. The improved stability of the MC precipitates, which formed in the matrix during irradiation, prevents loss of ductility at 500
C and below. The corrosion properties of an austenitic stainless steel at low temperature demonstrated good endurance for material usage in LBE during a short time, approximately at 300 and 470
C for 3000 h for 1.4970 austenitic stainless steel,46 and at 420
C for 2000 h for 1.4970 austenitic stainless steel and 316L at an oxygen concentration of 106 wt%.47 No dissolution was seen in the aforementioned results. A thin oxide scale may protect the material from attack in LBE. As demonstrated in Figure 4, a corrosion test under impinging flow was also conducted on JPCA and its EB welded bar at an oxygen concentration of 2–4  105 wt%.48 The EB welded metal of JPCA exhibited a dendritic structure 1 mm in width, but a heat-affected zone was not visible. Scanning electron microscopy (SEM) observation showed no corrosion layer for the specimens tested at 450
C and 1000 h.

1600 1400 1200 1000 800 600 400 200 1200 1000 800 600 400 200 2000

But at 3000 h, a thin corrosion layer could be observed at 1–2 mm in depth. For the weld joint, the depth of the corrosion layer as well as corrosion morphology showed the same results as with the parent material. The results of X-ray diffraction analyses showed how the oxide layer developed at 450 and 500
C. Figure 6 shows X-ray diffraction analyses of the JPCA specimens under the conditions of 1000 h at 450
C (top, JPCA-1), 1000 h at 500
C (middle, JPCA-2), and 3000 h at 450
C (bottom, JPCA-3). Oxidation of the JPCA at 450
C progressed in the same manner as at 500
C.

5.09.7 Precipitation Formation The dissolution of Ni, Cr, and Fe from structural materials into LA was studied. Saturated solubility of Ni in LA is calculated to be a couple of wt% at 450–500
C.20 Corrosion–erosion tests have been conducted at the JLBL-1 facility of the Japan Atomic Energy Agency (JAEA). The main circulating loop was made of SS316 austenitic stainless steel, and consisted of the specimens at high and low temperatures, filters, a surge tank, a cooler, an electromagnetic flow meter, a surface-level meter, thermocouples, and a drain tank.49 The loop was operated at a maximum temperature of 450
C with a temperature difference of 50–100
C and average flow velocity of 1 m s1. The oxygen concentration was estimated to be 107 wt%, Cr0.19, Fe0.70, Ni0.11 Bi M3O4

JPCA-1

JPCA-2

JPCA-3

1500 1000 500 0 10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Figure 6 X-ray diffraction analyses of JPCA specimens under the condition of 1000 h at 450
C (top, JPCA-1), 1000 h at 500
C (middle, JPCA-2), and 3000 h at 450
C (bottom, JPCA-3).

Material Performance in Lead and Lead-bismuth Alloy

according to measurements using an oxygen probe. The testing specimen tube is a cold-drawn seamless type SS316, which was produced as a tubing form with 13.8 mm outer diameter, 2 mm thickness, and 40 cm length. The tube was solution-heat treated at 1080
C for 1.5 min and then cooled rapidly. Figure 7 shows both the EDX analyses of the low-temperature specimen after corrosion–erosion testing for the 3000 h and the SEM image of an unused specimen. The surface of unused specimen was characterized by the creviced structure. This feature resulted from the acid washing in the manufacturing process during material preparation. It was found that precipitated materials existed with the solidified LBE on the tube. The precipitation consists mainly of Fe and Cr as measured using energy dispersive X-ray analyses and apparently looks crystalline. The quantitative analyses by means of a focused-ion beam, X-rays, and Transmission Electron Microscope (TEM) showed that the weight concentration ratio was roughly Fe:Cr = 9:1, for example. Nickel was not found in the crystals or in the solidified LBE. The precipitations occur in the lead–bismuth including impurities dissolved from SS316 at the high temperature portion of the test. Zhang and Li4,50 calculated the corrosion/precipitation rate for iron using a kinetic corrosion model and the temperature profile for the JLBL-1 loop. They reported that the observed deposition zones in the JLBL-1 loop could be exactly predicted using the nonisothermal and multimodular corrosion model. The predicted corrosion rate is about 0.05–0.08 mm per 3000 h if the diffusion coefficient is selected as

SEM

BEM

Fe

Cr

217

3.9  109 m2 s1. This agrees well with the experimental results of 0.03–0.1 mm. Ni-rich precipitation was found in the JLBL-1 loop after a total operation time of 9000 h was achieved. Figure 8 shows Ni-rich precipitates in an SEM (low magnitude) and laser microscope (high magnitude) images on the surface of solidified LBE.51 The solubility of Ni is higher than that of Fe and Cr, around a couple of wt% in the temperature range of 350–450
C. For the measurement of Ni in LBE, an inductive-coupled plasma atomic emission spectrometer (ICP, ULTIMA2) was used for analyses. It was found that the Ni concentration was below 0.1 wt%. In addition, Ni-rich precipitates were found not only at the high temperature part but also at the low temperature part, and on the surface of residual LBE as well. This was not the case for Fe–Cr precipitates; they were only found at the low temperature part. The driving force for Fe–Cr precipitates was concluded to be a difference of the saturation concentration at different temperatures. It can be assumed that the Ni-rich precipitates formed on the surface of the residual LBE during a cooling period, although the precipitation rate is not known for the establishment of such a Ni-rich structure.

5.09.8 Outlook For the use of LA as coolant and spallation target, it is important that the compilations and databases of material properties are extended to include

Unused specimen surface 50 µm

20 µm 500 µm Ni

Pb

Bi

Figure 7 Energy-dispersed X-ray analyzer analyses of low-temperature specimen after corrosion–erosion test at JLBL-1 and scanning electron microscopy image of unused specimen.

Figure 8 Ni-rich precipitates of scanning electron microscopy (low magnitude) and laser microscope (high magnitude) on the surface of solidified lead–bismuth eutectics. High magnification image is taken by laser microscope. Reproduced from Kikuchi, K.; Saito, S.; Hamaguchi, D.; Tezuka, M. J. Nucl. Mater. 2010, 398, 104–108.

218

Material Performance in Lead and Lead-bismuth Alloy

the mechanical properties of structural materials in LA, such as the effect of temperature and strain rate,52 fracture toughness,25,53 weldment,54 and liquid metal embrittlement.55 They are essential to the design work for the concepts of the ADS systems. Erosion–corrosion of materials in flowing LA should be considered along with details of the flow profile. It was recognized that magnetite in the oxide layer is eliminated in the steady-state flowing LA. There is evidence of erosion detected in the expanded flow channel of the JLBL-1 test specimen. The eroded part of the specimen coincides with regions in flowing LA, such as secondary flow or vortex flow. In the last decade, research projects for ADS, LFR, MEGAPIE, and MYRRHA were launched. To design the conceptual model of the real spallation target, the performance of neutronics, reactor physics, and thermal hydraulics has been studied along with the performances of materials in lead–bismuth or lead. For the materials bombarded by high-energy particles, such as proton beams, the surface coating or surface treatment, defect formation mechanism including not only corrosion but also the synergetic effect of irradiation field, must be understood. For the materials applicable to the cladding of fuel rod, Fe–Al alloys seem to be effective in the use of LA, but optimum Al concentration must be determined.

11. 12. 13. 14. 15. 16.

17.

18. 19.

20.

21. 22.

References 1.

2. 3.

4. 5. 6. 7. 8. 9. 10.

Sekimoto, H. Introductory remarks on liquid metal cooled fast reactors. In Proceedings of the ISTC-TITech Japan Workshop on Nuclear Reactors Technologies in Russia/CIS, July 2001; p 13. Sekimoto, H.; Ryu, K.; Yoshimura, Y. Nucl. Sci. Eng. 2001, 139, 306–317. OECD/NEA Nuclear Science Committee. Working Party on Scientific Issues of the Fuel Cycle, Working Group on Lead–Bismuth Eutectic Published Handbook on Lead–Bismuth Eutectic Alloy and Lead Properties, Materials Compatibility, Thermal-hydraulics and Technologies, ISBN 978-92-64-99002-9; 2007. Zhang, J.; Li, N. J. Nucl. Mater. 2008, 373, 351–377. Fazio, C.; Alamo, A.; Almazouzi, A.; et al. J. Nucl. Mater. 2009, 392, 316–323. Bauer, G. S.; Salvatores, M.; Heusen, G. J. Nucl. Mater. 2001, 296, 17–33. Groeschel, F.; Fazio, C.; Knebel, J.; et al. J. Nucl. Mater. 2004, 335, 156–162. Wagner, W.; Gro¨schel, F.; Thomsen, K.; Heyck, H. J. Nucl. Mater. 2008, 377, 12–16. Cinotti, L.; Giraud, B.; Ait Abderrahim, H. J. Nucl. Mater. 2004, 335, 148–155. Knebel, J. U.; Aı¨t Abderrahim, H.; Benamati, G.; et al. In Proceedings of the 8th Information Exchange Meeting. Actinide and Fission Product Partitioning and Transmutation, Las Vegas, NV, Nov 9–11, 2004;

23. 24. 25. 26. 27. 28. 29.

30. 31. 32. 33. 34. 35. 36.

Organization for Economic Co-operation and Development, American Nuclear Society: La Grange Park, IL, 2005. Mukaiyama, T.; Takizuka, T.; Mizumoto, M.; et al. Progr. Nucl. Energ. 2001, 38(1–2), 107. Tsujimoto, K.; Sasa, T.; Nishihara, K.; Takizuka, T.; Takano, H. Progr. Nucl. Energ. 2000, 37(1–4), 339. Oigawa, H.; Ouchi, N.; Kikuchi, K.; et al. In Proceedings of GLOBAL 2003, New Orleans, LA, Nov 16–20, 2003; American Nuclear Society: La Grange Park, IL, 2003. Weisenburger, A.; Heinzel, A.; Fazio, C.; Mu¨ller, G.; Markow, V. G.; Kastanov, A. D. J. Nucl. Mater. 2008, 377, 261–267. Van den Bosch, J. ADS Candidate Materials Compatibility with Liquid Meal in a Neutron Irradiation Environment, Doctoral Thesis, ISBN 978-90-8578-241-4, 2008; 7. Sheffield, R. L.; Pitcher, E. J. ADS History in the USA. In Proceedings of Applications of High Intensity Proton Accelerator, October 19–21, 2009; Fermi National Accelerator Laboratory. Gohar, Y. Spallation target design for accelerator-driven systems. In Proceedings of Applications of High Intensity Proton Accelerator, October 19–21, 2009; Fermi National Accelerator Laboratory. Pitcher, E. J. J. Nucl. Mater. 2008, 377, 17–20. Gromov, B. F.; Orlov, Yu. I.; Martynov, O. N.; Ivanov, K. D.; Gulevsky, V. A. In Liquid Metal Systems; Borgstedt, H. U., Frees, G., Eds.; Plenum Press: New York, 1995; pp 339–343. Martynov, P. N.; Orlov, Yu. I.; Efanov, A. D.; Troiynov, V. M.; Rusanov, A. E.; Lavrova, O. V. Technology of lead–bismuth coolants for nuclear reactors. In Proceedings of the ISTCTITech Japan Workshop on Nuclear Reactor Technologies in Russia/CIS, July 2001; pp 80–105. OECD. Handbook on Lead–Bismuth Eutectic Alloy and Lead Properties, Materials Compatibility, Thermal-hydraulics and Technologies, 2007; 109. Martinelli, L.; Balbaud-Ce´le´rier, F.; Terlain, A.; et al. Corrosion Sci. 2008, 50, 2523–2536. Martinelli, L.; Balbaud-Ce´le´rier, F.; Terlain, A.; Bosonnet, S.; Picard, G.; Santarini, G. Corrosion Sci. 2008, 50, 2537–2548. Martinelli, L.; Balbaud-Ce´le´rier, F.; Picard, G.; Santarini, G. Corrosion Sci. 2008, 50, 2549–2559. Auger, T.; Hamouche, Z.; Medina-Almazan, L.; Gorse, D. J. Nucl. Mater. 2008, 377, 253–260. Tamura, M.; Hayakawa, H.; Tanimura, M.; Hishinuma, A.; Kondo, T. J. Nucl. Mater. 1986, 141–143, 620. Furukawa, T.; Mu¨ller, G.; Schumacher, G.; et al. J. Nucl. Sci. Technol. 2004, 41, 265. Tan, L.; Machut, M. T.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2007, 371, 161–170. Masuyama, F. New developments in steels for power generation boilers. In Advanced Heat Resistant Steels for Power Generation; Conference Proceedings, San Sebastian, Spain, Apr 27, 1998; Viswanathan, R., Nutting, J., Eds.; IOM Communications: London, 1999; pp 33–48. Viswanathan, R.; Bakker, W. J. Mater. Eng. Perform. 2001, 10, 81. Klenowicz, Z.; Darowicki, K. Corrosion Rev. 2001, 19, 467. Hosemann, P.; Hawley, M. E.; Koury, D.; et al. J. Nucl. Mater. 2008, 381, 211–215. Hosemann, P.; Swadener, J. G.; Welch, J.; Li, N. J. Nucl. Mater. 2008, 377, 201–205. Yamaki, E.; Kikuchi, K. J. Nucl. Mater. 2010, 398, 153–159. Sugawara, T.; Kikuchi, K.; Nishihara, K.; Oigawa, H. J. Nucl. Mater. 2010, 398, 246–250. Heinzel, A.; Kondo, M.; Takahashi, M. J. Nucl. Mater. 2006, 350, 264–270.

Material Performance in Lead and Lead-bismuth Alloy 37. Muller, G.; Schumacher, G.; Zimmermann, F. J. Nucl. Mater. 2000, 278, 85–95. 38. Weisenburger, A.; Heinzel, A.; Muller, G.; Muscher, H.; Rousanov, A. J. Nucl. Mater. 2008, 376, 274–281. 39. Muller, G.; Heinzel, A.; Konys, J.; et al. J. Nucl. Mater. 2004, 335, 163. 40. Takaya, S.; Furukawa, T.; Aoto, K.; et al. J. Nucl. Mater. 2009, 386–388, 507–510. 41. Hosemann, P.; Thau, H. T.; Johnson, A. L.; Maloy, S. A.; Li, N. J. Nucl. Mater. 2008, 373, 246–253. 42. Dai, Y.; Henry, J.; Auger, T.; et al. J. Nucl. Mater. 2006, 356, 308–320. 43. Nishihara, K.; Kikuchi, K. J. Nucl. Mater. 2008, 377, 298–306. 44. Dai, Y.; Wagner, W. J. Nucl. Mater. 2009, 389, 288–296. 45. Tanaka, M. P.; Hamada, S.; Hishinuma, A.; Grossbeck, M. L. J. Nucl. Mater. 1988, 155–157, 957–962. 46. Barbier, F.; Rusanov, A. J. Nucl. Mater. 2001, 296, 231.

47. 48. 49. 50. 51. 52. 53. 54. 55.

219

Muler, G.; Heinzel, A.; Koney, J.; et al. J. Nucl. Mater. 2002, 301, 40. Kikuchi, K.; Kamata, K.; Ono, M.; Kitano, T.; Hayashi, K.; Oigawa, H. J. Nucl. Mater. 2008, 377, 232–242. Kikuchi, K.; Saito, S.; Kurata, Y.; et al. JSME Int. J. B 2004, 47(2), 332–339. Zhang, J.; Li, N. J. Nucl. Mater. 2004, 326, 201–210. Kikuchi, K.; Saito, S.; Hamaguchi, D.; Tezuka, M. J. Nucl. Mater. 2010, 398, 104–108. Van den Bosch, J.; Sapundjiev, D.; Almazouzi, A. J. Nucl. Mater. 2006, 356, 237–246. Van den Bosch, J.; Coen, G.; Almazouzi, A.; Degrieck, J. J. Nucl. Mater. 2009, 385, 250–257. Van den Bosch, J.; Almazouzi, A. J. Nucl. Mater. 2009, 385, 504–509. Auger, T.; Serre, I.; Lorang, G.; Hamouche, Z.; Gorse, D.; Vogt, J. B. J. Nucl. Mater. 2008, 376, 336–340.

5.10

Material Performance in Molten Salts

V. Ignatiev and A. Surenkov National Research Centre, Kurchatov Institute, Moscow, Russian Federation

ß 2012 Elsevier Ltd. All rights reserved.

5.10.1

Introduction: Brief Review of Different Related Applications

221

5.10.2 5.10.2.1 5.10.2.2 5.10.3 5.10.3.1 5.10.3.1.1 5.10.3.1.2 5.10.3.1.3 5.10.4 5.10.5 5.10.6 References

Choice of Fuel and Coolant Salts for Different Applications Chemical Compatibility of Materials with Molten-Salt Fluorides Preparative Chemistry and Salt Purification Developments in Materials for Different Reactor Systems Molten-Salt Reactor Metallic materials for primary and secondary circuits Graphite for the core Materials for molten-salt fuel reprocessing system Advanced High-Temperature Reactor Liquid-Salt-Cooled Fast Reactor Secondary Circuit Coolants

223 226 228 229 229 230 241 242 243 246 247 249

Abbreviations AHTR

Advanced High-Temperature reactor cooled by molten salts ARE Aircraft Reactor Experiment CNRS Centre de la National Recherche´ Scientifique, France dpa Displacements per atom FLIBE Molten LiF-BeF2 salt mixture FLINABE Molten LiF-NaF-BeF2 salt mixture Hastelloy N or Ni-Mo alloy developed for MSR INOR-8 at ORNL HTR High-Temperature Reactor cooled by helium HX Heat Exchanger IGC InterGranular Cracks IHX Intermediate Heat Exchanger KI Kurchatov Institute, Russia LSFR Liquid Salt-cooled Fast Reactor LWR Light Water Reactor MA Minor Actinides MC (U,Pu)C Metal Carbide fuel form MOSART Molten Salt Actinide Recycler & Transmuter MOX (U,Pu)O2 Mixed Oxide fuel MSBR Molten Salt Breeder Reactor MSFR Molten Salt Fast Reactor MSR Molten Salt Reactor MSRE Molten Salt Reactor Experiment MWe Megawatts electrical

MWt NCL NFC NPP ODS ORNL RE REDOX RW SFR SNF TRU UOX VHTR

Megawatts thermal Natural Convection Loop Nuclear Fuel Cycle Nuclear Power Plant Oxide Dispersion-strengthened Steels Oak Ridge National Laboratory, USA Rare Earth elements Electrochemical reduction– oxidation Radioactive Wastes Sodium-cooled Fast Reactor Spent Nuclear Fuel TRans-Uranium elements UO2 Uranium Oxide fuel Very High-Temperature Reactor

5.10.1 Introduction: Brief Review of Different Related Applications In the last few years, there has been a significantly increased interest in the use of high-temperature molten salts as coolants and fuels in nuclear power and fuel cycle systems.1–5 The potential utility of a fluid-fueled reactor that can operate at a high temperature, but with a low-pressure system, has been recognized for a long time. One of the attractive

221

222

Material Performance in Molten Salts

features of the molten-salt system is the variety of reactor types that can be considered to cover a range of applications. Molten salts offer very attractive characteristics as coolants, with respect to heat transport and heat transfer properties at high temperatures. The molten-salt system has the usual benefits attributed to fluid-fuel systems. The principal advantages over solid-fuel element systems are (1) a high negative temperature coefficient of reactivity; (2) lack of radiation damage that can limit fuel burnup; (3) the possibility of continuous fission-product removal; (4) the avoidance of the expense of fabricating new fuel elements; and (5) the possibility of adding makeup fuel as needed, which precludes the need for providing excess reactivity. Indeed, fuel can be processed in an online mode or in batches in order to retrieve fission products and then reintroduced into the reactor (fuel in liquid form during the whole cycle). Molten fluoride salts were first developed for nuclear systems as a homogeneous fluid fuel. In this application, salt served as both fuel and primary coolant at temperatures
700  C. Secondary coolant salts were also developed that contained no fissile and fertile materials. In the 1970s, because power cycle temperatures were limited by the existing steam system technology, the potential for use of molten salts at extreme temperatures was not fully explored. Today, much higher temperatures (>700  C) are of interest for a number of important applications. For 60 years, nitrate salts at lower temperatures have been used as coolants on a large industrial scale in heat transport systems in the chemical industry; thus, a large experience base exists for salt-base heat transport systems.6–8 However, because these salts decompose at 600  C, highly stable salts are required at higher temperatures. Most of the research on high-temperature molten-salt coolants has focused on fluoride salts because of their chemical stability and relatively noncorrosive behavior. Chloride salts are a second option, but the technology is less well developed.9,10 As is true for most other coolants, corrosion behavior is determined primarily by the impurities in the coolant and not the coolant itself. While largescale testing has taken place, including the use of such salts in test reactors, there is only limited industrial experience. In the 1950s and 1960s, the US Oak Ridge National Laboratory (ORNL) investigated moltensalt reactors (MSRs), in which the fuel was dissolved in the fluoride coolant, for aircraft nuclear propulsion and breeder reactors.11 Two test reactors were built at ORNL: the Aircraft Reactor Experiment

(ARE)12–14 and the Molten Salt Reactor Experiment (MSRE).15 The favorable experience gained from the 8 MWt MSRE test reactor operated from 1965 to 1969 led to the design of a 1000 MWe molten-salt breeder reactor (MSBR) with a core graphite moderator, thermal spectrum, and thorium–uranium fuel cycle.16,17 In the MSBR design, fuel salt temperature at the core outlet was 704  C. The research and development effort, combined with the MSRE and a large number of natural and forced convection loop tests, provided a significant basis for demonstrating the viability of the MSR concept. Since the 1970s, with other countries, including Japan, Russia, and France, the United States placed additional emphasis on the MSR concept development.18–22 Recent MSR developments in Russia on the 1000 MWe molten-salt actinide recycler and transmuter (MOSART)1 and in France on the 1000 MWe nonmoderated thorium molten-salt reactor (MSFR)4,5 address the concept of large power units with a fast neutron spectrum in the core. Compared to the MSBR, core outlet temperature is increased to 720  C for MOSART and 750  C for the MSFR. The first concept aims to be used as efficient burners of transuranic (TRU) waste from spent UOX and MOX light water reactor (LWR) fuel without any uranium and thorium support. The second one has a breeding capability when using the thorium fuel cycle. Studies of the fast-spectrum MSFR also indicated that good breeding ratios could be obtained, but high power densities would be required to avoid excessive fissile inventories. Adequate power densities appeared difficult to achieve without novel heat removal methods. Earlier proposals for fast-spectrum MSRs used chloride salts.9 However, chloride salts have three major drawbacks: (1) a need for isotopically separated chlorine to avoid high-cross-section nuclides; (2) the activation product 36Cl, which presents significant challenges to waste management because of its mobility in the environment; and (3) the more corrosive characteristics of chloride systems relative to fluoride systems. Today, in addition to the different MSR systems, other advanced concepts that use the molten-salt technology are being studied, including the advanced high-temperature reactor (AHTR) and the liquidsalt-cooled fast reactor (LSFR). The AHTR uses clean molten salts as the coolant and the same coated particle fuel encapsulated in graphite as high-temperature gas-cooled reactors, such as the very high-temperature reactor (VHTR). The fuel cycle characteristics are essentially identical

Material Performance in Molten Salts

to those of the VHTR. This concept was originally proposed in the 1980s by the RRC-Kurchatov Institute in Russia,19 but most of the recent work is being conducted in the United States.23 The AHTR is a longer-term high-temperature reactor option with potentially superior economics due to the properties of the salt coolant. Also, better heat transport characteristics of salts compared to helium enable power levels up to 4000 MWt with passive safety systems. The AHTR can be built in larger sizes or as very compact modular reactors, it operates at lower pressure, and the equipment is smaller because of the superior heat transfer capabilities of liquid-salt coolants compared to helium. A newer concept is the LSFR, which is being investigated in the United States and France.24 Liquid salts offer three potential advantages compared to sodium: (1) molten fluoride salts are transparent and have heat transport properties similar to those of water; however, their boiling points exceed 1200  C; (2) smaller equipment size because of the higher volumetric heat capacity of the salts; and (3) no chemical reactions between the reactor, intermediate loop, and power cycle coolants. There is experience with this type of system because the ARE at ORNL used a sodiumcooled intermediate loop. The basic design of an LSFR is similar to that of a sodium-cooled fast reactor (SFR), except that a clean salt replaces the sodium and the reactor operates at higher temperatures with the potential for higher thermal efficiency. Molten-salt fluoride-based coolants allow fast-reactor coolant outlet temperatures to be increased from 500–550  C (sodium) to 700–750  C, with a corresponding increase in plant efficiency from 42% to 50%. To identify salts that produce acceptable ‘voiding’ (meaning thermal expansion) response, chlorides are also explored as salts for the LSFR, though one has to consider the 36Cl production either by neutron capture on 35Cl or (n, 2n) reaction on 37Cl. Recent MSR developments in the United States on the 2400 MWt liquid-salt-cooled, flexible-conversion-ratio reactor address the concept with a core power density of 130 kW l1 and a maximum cladding temperature of 650  C.25 Based on technical considerations, LSFRs may have significantly lower capital costs than SFRs; thus, there is an incentive to examine the feasibility of an LSFR. There are fundamental challenges to this new reactor concept, such as development of high-temperature clads that are corrosion resistant in the salt environment, can operate at high temperatures, and can withstand high neutron radiation levels.

223

There are multiple industrial uses for hightemperature heat at temperatures from 700 to 950  C.2 There is a growing interest in using hightemperature reactors to supply this heat because of the increasing prices for natural gas and concerns about greenhouse gas emissions. Such applications require high-temperature heat transport systems to move heat from high-temperature nuclear reactors (gas-cooled or liquid-salt-cooled) to the customer. There are several economic incentives to develop liquid-salt heat transport systems rather than using helium for these applications: (1) the pipe crosssections are less than one-twentieth of that of helium because of the high volumetric heat capacity of liquid salts; (2) salt systems can operate at atmospheric pressure; (3) better heat transfer characteristics of the salt reduce the size of heat exchangers; and (4) molten-salt pumps operate at much higher temperatures to provide heat in a narrow temperature interval, compared to compressors that circulate helium in a VHTR.19 For most of these applications, the transport distances will exceed a kilometer. Finally, it should be noted that fuel refining and reprocessing in systems using molten chlorides/ fluorides and liquid metals (Bi, Zn, Cd, Pb, Sn, etc.) is a promising method to solve the actinide and fission product partitioning task for advanced fuels. These approaches are considered as basic for reprocessing metal, nitride, and MSR fuels.2,4,17,19 As can be seen from the considerations above, there are several potential applications of molten salts for future nuclear power. There is great flexibility in the use of molten-salt concepts for nuclear power in liquid-fuel and solid-fuel reactors, heat transfer loops, or fuel-processing units.

5.10.2 Choice of Fuel and Coolant Salts for Different Applications Selection of salt coolant composition strongly depends on the specific design application: fluid fuel (burner or breeder), primary (LSFR or AHTR) or secondary coolant, heat transport fluid, etc. In choosing a fuel salt for a given fluid-fuel reactor design, the following criteria are applied26:  Low neutron cross-section for the solvent components  Thermal stability of the salt components  Low vapor pressure  Radiation stability

224

Material Performance in Molten Salts

 Adequate solubility of fuel (including TRU waste) and fission-product components  Adequate heat transfer and hydrodynamic properties  Chemical compatibility with container and moderator materials  Low fuel and processing costs At temperatures up to 1000  C, molten fluorides exhibit low vapor pressure (1 atm) and relatively benign chemical reactivity with air and moisture. Molten fluorides also trap most fission products (including Cs and I) as very stable fluorides, and thus act as an additional barrier to accidental fission product release. Fluorides of metals other than U, Pu, or Th are used as diluents and to keep the melting point low enough for practical use. Consideration of nuclear properties alone leads one to prefer as diluents the fluorides of Be, Bi, 7Li, Pb, Zr, Na, and Ca, in that order. Salts that contain easily reducible cations (Bi3þ and Pb2þ, see Table 1) were rejected because they would not be stable in nickel- or iron-base alloys of construction. Three basic salt systems (see Table 2)27–33 exhibit usefully low melting points (between 315 and 565  C) and also have the potential for neutronic viability and material compatibility with alloys: (1) alkali fluoride salts, (2) ZrF4-containing salts, and (3) BeF2containing salts. An inspection of the behavior of the phase diagrams for these systems reveals a considerable range of compositions in which the salt will be completely molten with concentrations of UF4 or ThF4 > 10 mol% at 500  C and >20 mol% Table 1

Thermodynamic properties of fluorides

Compound (solid state)

–DGf,1000 (kJ mol1)

Compound (solid state)

–DGf,1000 (kJ mol1)

LiF NaF KF BeF2 ThF4 UF3 ZrF4 UF4

522 468 460 447 422 397 393 389

AlF3 VF2 TiF2 CrF2 FeF2 HF NiF2 CF4

372 347 339 314 280 276 230 130

Source: Novikov, V. M.; Ignatiev, V. V.; Fedulov, V. I.; Cherednikov, V. N. Molten Salt Reactors: Perspectives and Problems; Energoatomizdat: Moscow, USSR, 1990; Ignatiev, V. V.; Novikov, V. M.; Surenkov, A. I.; Fedulov, V. I. The state of the problem on materials as applied to molten-salt reactor: Problems and ways of solution, Preprint IAE-5678/11; Institute of Atomic Energy: Moscow, USSR, 1993; Williams, D. F.; et al. Assessment of candidate molten salt coolants for the advanced high-temperature reactor, ORNL/TM-2006/12; ORNL: Oak Ridge, TN, 2006.

at 560  C.27 Trivalent plutonium and minor actinides are the only stable species in the various molten fluoride salts. Tetravalent plutonium could transiently exist if the salt redox potential is high enough. Solubility of PuF4 by analogy of ZrF4, UF4, and ThF4 should be relatively high. But for practical purposes (stability of potential container material), the salt redox potential should be low enough and correspond to the stability area of Pu (III). PuF3 solubility is maximum in pure LiF, NaF, or KF and decreases with the addition of BeF2 and ThF4.28–33 The solubility decrease is more for BeF2 addition, because PuF3 is not soluble in pure BeF2. As can be seen from Table 2 (column 1), the LiF–PuF3 system is characterized by a eutectic point with 20 mol% of PuF3 at 743  C.28 The calculated solubility of PuF3 in the matrix of LiF–NaF–KF (43.9–14.2–41.9) at T ¼ 600  C has been found to be 19.3 mol%.5 Adequate solubility of PuF3 at 600  C in burner (>2 mol%) and breeder fast-spectrum concepts (3–4 mol%) can also be achieved using 7LiF–(NaF)– BeF2 (column 3) and LiF–(BeF2)–ThF4 (column 4) systems solvent (see Table 2), respectively. The lanthanide trifluorides are also only moderately soluble in BeF2- and ThF4-containing mixtures. If more than one such trifluoride (including UF3) is present, they crystallize to form a solid, made up of all the trifluorides, on cooling of the saturated melt so that, in effect, all the LnF3 and AnF3 act essentially as a single element. If so, the total (An þ Ln) trifluorides in the end-of-life reactor might possibly exceed their combined solubility. Melts of these fluorides have satisfactory values of heat capacity, thermal conductivity, and viscosity over a temperature range of 550–1000  C and provide an efficient removal of heat when they are used as the coolant over a wide range of compositions. (See also Chapter 3.13, Molten Salt Reactor Fuel and Coolant). Transport properties of molten-salt coolants ensure highly efficient cooling with natural circulation; the salt–wall heat transfer coefficient is close to the same as that for water. The thermal diffusivity of the salt is 300 times smaller than that of sodium. Therefore, all other things being equal, the characteristic solidification time for a volume of the fluoride melt is 300 times longer than that of sodium.2 A particular disadvantage of ZrF4-containing (more than 25 mol%) melts is its condensable vapor, which is predominantly ZrF4.26 The ‘snow’ that would form could block vent lines and cause problems in pumps that circulate the fuel. Note also that the use of Zr instead of sodium in the basic solvent will lead to

Material Performance in Molten Salts

225

Table 2 Molar compositions, melting temperatures ( C),27 and solubility of plutonium trifluoride (mol%) at 600  C in different molten fluoride salts considered as candidates for the fuel and the coolant circuits in MSR concepts Alkali-metal fluorides LiF–PuF3 (80–20) 743  C28 LiF–KF (50–50) 492  C – LiF–RbF (44–56) 470  C – LiF–NaF–KF (46.5–11.5–42) 454  C 19.35 LiF–NaF–RbF (42–6–52) 435  C –

ZrF4-containing

BeF2 containing

ThF4 containing

Fluoroborates

LiF–ZrF4 (51–49) 509  C – NaF–ZrF4 (59.5–40.5) 500  C 1.831 LiF–NaF–ZrF4 (42–29–29) 460  C – LiF–NaF–ZrF4 (26–37–37) 436  C – NaF–RbF–ZrF4 (33–24–43) 420  C – NaF–KF–ZF4 (10–48–42) 385  C – KF–ZrF4 (58–42) 390  C –

LiF–BeF2 (73–27) 530  C 2.032 LiF–NaF–BeF2 (15–58–27) 479  C 2.032,33 LiF–BeF2 (66–34) 458  C 0.532,33 LiF–BeF2–ZrF4 (64.5–30.5–5) 428  C – NaF–BeF2 (57–43) 340  C 0.332 LiF–NaF–BeF2 (31–31–38) 315  C 0.432

LiF–ThF4 (78–22) 565  C 4.229 LiF–BeF2–ThF4 (75–5–20) 560  C 3.129 LiF–BeF2–ThF4 (71–16–13) 499  C 1.530 LiF–BeF2–ThF4 (64–20–16) 460  C 1.229 LiF–BeF2–ThF4 (47–51.5–1.5) 360  C –

KF–KBF4 (25–75) 460  C – RbF–RbBF4 (31–69) 442  C – NaF–NaBF4 (8–92) 384  C –

increased generation of long-lived activation products in the system. Potassium-containing salts are usually excluded from consideration as a primary coolant because of the relatively large parasitic capture crosssection of potassium. However, potassium-containing salts are commonly used in nonnuclear applications and serve as a useful frame of reference (e.g., LiF– NaF–KF). This leaves 7LiF, NaF, and BeF2 as preferred major constituents. For reasons of neutron economy at ORNL, the preferred solvents for prior Th–U MSR concepts have been LiF and BeF2, with the lithium enriched to 99.995 in the 7Li isotope. It has recently been indicated that this well-studied BeF2-containing solvent mixture needs further consideration, in view of the current knowledge on beryllium toxicity.4 Unlike the MSR, AHTR and LSFR use solid fuel and a clean liquid salt as a coolant (i.e., a coolant with no dissolved fissile materials or fission products). For the MSR, a major constraint was the requirement for

high solubility of fissile materials and fission products in the salt; a second was suitable for salt reprocessing. For AHTR and LSFR, these requirements do not exist. The requirements mainly include (1) a good coolant, (2) low coolant freezing points, and (3) applicationspecific requirements. As a result, a wider choice of fluoride salts can be considered. For a fast reactor, it is also desirable to avoid low-Z materials that can degrade the neutron spectrum. In all cases, binary or more complex fluoride salt mixtures are preferred because the melting points of fluoride salt mixtures are much lower than those for single-component salts. According to recent ORNL recommendations,26 the following two types of salts should be studied for AHTR and LSFR primary circuits in the future:  Salts that have been shown in the past to support the least corrosion (e.g., salts containing BeF2 and ZrF4 in the concentration range 25–40 mol%);

226

Material Performance in Molten Salts

 Salts that provide the opportunity for controlling corrosion by establishing a very reducing salt environment (e.g., alkali fluoride (LiF–NaF–KF) mixtures and BeF2-containing salts). Alternatively, the 2400 MWt liquid-salt-cooled, flexible-conversion-ratio reactor25 was designed, utilizing as a primary coolant the ternary chloride salt 30NaCl–20KCl–50MgCl2 (in mol%) with maximum cladding temperatures under 650  C. The selected chloride base salt has high melting points (396  C for the reference salt vs. 98  C for sodium). Claim is made that the materials used in the fuel, core, and vessel should be the same as those in the sodium and lead reactor designs but at temperatures required corrosion behavior for mentioned above materials in chloride salts is not clear yet (see details in Section 5.10.6 Secondary Circuit Coolants, Table 7). For applications that use molten salt outside a neutron field, additional salts may be considered. Candidate coolants can include salts deemed unsuitable as a primary coolant but judged as acceptable for use in a heat transfer loop. Familiar oxygen-containing salts (nitrates, sulfates, and carbonates) are excluded from consideration because they do not possess the necessary thermochemical stability at high temperatures (>600  C). These salts are also incompatible with the use of carbon materials because they decompose at high temperatures to release oxygen, which rapidly reacts with the available carbon. The screening criteria for selecting secondary salt coolants require that the elements constituting the coolant must form compounds that (1) have chemical stability at required temperatures, (2) melt at useful temperatures and are not volatile, and (3) are compatible with high-temperature alloys, graphite, and ceramics. In addition to the fluoride salts considered (see Table 2), two families of salts fulfill these three basic requirements: (a) alkali fluoroborates and (b) chloride salts. For both salt systems, there are material problems, particularly at the high end of the temperature range. The chemical stability of chloride salt mixtures seems not as good as for fluorides, though exclusion of oxygen and nitrogen is important. Sulfur from 35Cl and some fission products are potential precipitating species. Processing could be carried out, at some cost in external holdup. High-temperature processing has the potential benefits of being close-coupled, of reducing inventory, and of conserving 37Cl. Finally, a heat transport fluid is envisaged for the coupling of a reactor with a chemical plant, for

example, for hydrogen production.34 Typical salts considered are LiF–NaF–KF, KCl–MgCl2, and KF– KBF4. The ternary LiF–NaF–KF mixture provides superior heat transfer, KCl–MgCl2 has the potential to be a very low-cost salt, and KF–KBF4 may provide a useful barrier to isolate tritium from the hydrogen plant. Also, the ternary eutectic 9LiCl–63KCl– 28MgCl2 (in mol%) with melting point of 402  C appears to be the best compromise between raw material cost, performance, and melting point. As will be shown in the next sections, molten salts, first of all fluorides, are well suited for use at elevated temperatures as (a) fluid-fuel, (b) in-core coolant in a solid-fuel reactor, and (c) secondary coolant to transport nuclear heat at low pressures to a distant location. Materials are the greatest challenge for all high-temperature molten-salt nuclear applications. Current materials allow operation at 700–750  C and may be extended to higher temperatures. Operating temperatures much above 800  C will require significantly improved materials. There are strong incentives to increase the temperature to reach the full potential of the molten-salt-related systems for efficient electric and thermochemical hydrogen production. In this chapter, we review the relevant studies on materials performance in molten salts. 5.10.2.1 Chemical Compatibility of Materials with Molten-Salt Fluorides For any high-temperature application, corrosion of the metallic container alloy is the primary concern. Unlike the more conventional oxidizing media, the products of oxidation of metals by fluoride and chloride melts tend to be completely soluble in the corroding media.35–38 Owing to their solubility in the corroding media, passivation is precluded and the corrosion rate depends on other factors, including39–46 oxidants, thermal gradients, salt flow rate, and galvanic coupling. The general rule to ensure that the materials of construction are compatible (noble) with respect to the salt is that the difference in the Gibbs free energy of formation between the salt and the container material should be >80 kJ mol1 K1. The corrosion strategy is the same as that used in SFR, where the materials of construction are noble relative to metallic sodium. Many additional factors will influence the corrosion of alloys in contact with salts, but it is useful to keep in mind that the fundamental thermodynamic driving force for corrosion appears to be slightly greater in chloride systems than in fluoride systems. This treatment ignores a number of

Material Performance in Molten Salts

important salt solution effects, especially for salt mixtures that exhibit large deviations from ideal thermodynamic behavior. Additional study in the laboratory will be needed to understand whether chloride salts are fundamentally more corrosive toward alloys than fluorides, and whether corrosion control strategies can be devised that can be used with, or favor, chloride salt systems.34 As mentioned above, design of a practicable MSR system demands the selection of salt constituents that are not appreciably reduced by available structural metals and alloys whose components Mo, Ni, Nb, Fe, and Cr can be in near equilibrium with the salt (see Table 1). Equilibrium concentrations for these components will strongly depend on the solvent system. Examination of the free energies of formation for the various alloy components shows that chromium is the most active metal components. Therefore, any oxidative attachment to these alloys should be expected to show selective attack on the chromium. Stainless steels, having more chromium than Ni-base alloys developed within MSR programs, are more susceptible to corrosion by fluoride melts, but can be considered for some applications. Chemical reaction of the fluoride with moisture can form metal oxides that have much higher melting points and therefore appear as insoluble components at operating temperatures.39,40 Reactions of uranium tetrafluoride with moisture result in the formation of the insoluble oxide: UF4 þ 2H2 O $ UO2 þ 4HF

½1

The most direct method to avoid fuel oxide formation is through the addition of ZrF4, which reacts in a similar way with water vapor: ZrF4 þ 2H2 O $ ZrO2 þ 4HF

½2

The net reaction would be ZrF4 þ UO2 $ ZrO2 þ UF4

½3

Oxide films on the metal are dissolved by the following reactions: 2NiO þ ZrF4 ! 2NiF2 þ ZrO2

½4

NiO þ BeF2 ! NiF2 þ BeO

½5

2NiO þ UF4 ! NiF2 þ UO2

½6

227

Other corrosion reactions are possible with solvent components if they have not been purified well before utilization: Cr þ NiF2 ! CrF2 þ Ni

½7

Cr þ 2HF ! CrF2 þ H2

½8

These reactions will proceed essentially to completion at all temperatures within the circuit. Accordingly, such reactions can lead (if the system is poorly cleaned) to rapid initial corrosion. However, these reactions do not give a sustained corrosive attack. The impurity reactions can be minimized by maintaining low impurity concentrations in the salt and on the alloy surfaces. Reaction of UF4 with structural metals (M) may have an equilibrium constant which is strongly temperature dependent; hence, when the salt is forced to circulate through a temperature gradient, a possible mechanism exists for mass transfer and continued attack: 2UF4 þ M $ 2UF3 þ MF2

½9

This reaction is of significance mainly in the case of alloys containing relatively large amounts of chromium. Corrosion proceeds by the selective oxidation of Cr at the hotter loop surfaces, with reduction and deposition of chromium at the cooler loop surfaces. In some solvents (Li,Na,K,U/F, for example), the equilibrium constant for reaction [9] with Cr changes sufficiently as a function of temperature to cause the formation of dendritic chromium crystals in the cold zone.38 For Li,Be,U/F mixtures, the temperature dependence of the mass transfer reaction is small, and the equilibrium is satisfied at reactor temperature conditions without the formation of crystalline chromium. Of course, in the case of a coolant salt with no fuel component, reaction [9] would not be a factor. Redox processes responsible for attack by molten fluoride mixtures on the alloys result in selective oxidation of the contained chromium. This removal of chromium from the alloy occurs primarily in regions of highest temperature and results in the formation of discrete voids in the alloy.35 These voids are not, in general, confined to the grain boundaries in the metal, but are relatively uniformly distributed throughout the alloy surface in contact with the melt. The rate of corrosion has been measured and was found to be controlled by the rate at which chromium diffuses to the surfaces undergoing attack.41

228

Material Performance in Molten Salts

Graphite does not react with molten fluoride mixtures of the type to be used in the MSR concepts considered above (after carbon, borides and nitrides appear to be the most compatible nonmetallic materials). Available thermodynamic data suggest that the most likely reaction: 4UF4 þ C $ CF4 þ 4UF3

½10

should come to equilibrium at CF4 pressures <101 Pa. CF4 concentrations over graphite–salt systems maintained for long periods at elevated temperatures have been shown to be below the limit of detection (<1 ppm) of this compound by mass spectrometry. Moreover, graphite has been used as a container material for many NaF–ZrF4–UF4, LiF–BeF2–UF4, and other salt mixtures at ORNL and the RRC-Kurchatov Institute, with no evidence of chemical instability.47 In an MSR, reactions such as [11] and the later [12] were prevented by careful control of the solution redox chemistry, which was accomplished by setting the UF4/UF3 ratio at approximately (50–60)/1: UF4 þ Cr $ UF3 þ CrF2

½11

UF3 þ 2C $ UC2 þ 3UF4

½12

Additions of metallic Be to the fuel salt lead to reduction of the UF4 via 2UF4 þ Be0 $ 2UF3 þ BeF2

½13

The significance of redox control to the MOSART system with uranium-free fuel is that in some cases, where the fuel is, for example, PuF3, the Pu(III)/Pu (IV) redox couple is too oxidizing to present a satisfactory redox-buffered system. In this case, as was proposed by ORNL, redox control could be accomplished by including an HF/H2 mixture to the inert cover gas sparge, which will not only set the redox potential, but will also serve as the redox indicator if the exit HF/H2 stream is analyzed relative to inlet.48 In principle, avoiding corrosion in an MSR or in fuel-processing units with metallic components is significantly more challenging than avoiding corrosion in clean salt coolant applications (heat transport loops, AHTR and LSFR). In an MSR, the dissolved uranium and other such species in the fuel salt result in the presence of additional corrosion mechanisms that can limit the useful service temperature of an alloy. In clean salt applications, these types of corrosion mechanisms can be reduced or eliminated by (1) using purified salts that do not contain chemical species that can transport chromium and other

alloy constituents or (2) operating under chemically reducing conditions. Under chemically reducing conditions, chromium fluoride has an extremely low solubility, which limits chromium transport. The interaction of trace amounts of oxides, air, or moisture (either in the salt or cover gas) with fluoroborates often controls alloy corrosion, but these chemical interactions are complex and are not completely understood. For the secondary coolant NaF–NaBF4, corrosion is mainly determined by the selective yield of Cr from the alloy through the following reactions45: H2 O þ NaBF4 $ NaBF3 OH þ HF NaBF3 OH $ NaBF2 O þ HF 6HF þ 6NaF þ Cr $ 2Na3 CrF6 þ 3H2

½14

The hydrolysis of BF3 in the presence of any moisture in the cover gas above the salt is rapid and generates HF which is intensely corrosive to the system, especially when it is absorbed into molten salt. Some of the actual oxygen- and hydrogencontaining species that result from hydrolysis of BF3 in the salt have been identified. However, understanding of this chemistry is not complete,49 and more work is needed before preparative chemistry and online purification requirements can be defined with confidence. The behavior of hydrogen- and oxygen-containing species in fluoroborates is also important because it provides a means to sequester tritium in the salt, and thus an intermediate fluoroborate loop could serve as an effective tritium barrier. The species that is likely responsible for holding tritium in the salt was identified by Maya,50 and an engineering-scale experimental program was conducted that proved the effectiveness of sodium fluoroborate in sequestering tritium.51 5.10.2.2 Preparative Chemistry and Salt Purification Molten-salt use typically begins with the acquisition of raw components that are combined to produce a mixture that has the desired properties when melted. However, most suppliers of halide salts do not provide materials that can be used directly. The major impurities that must be removed to prevent severe corrosion of the container metal are moisture/oxide contaminants. Once removed, these salts must be kept from atmospheric contamination by handling and storage in sealed containers. During the US MSR

Material Performance in Molten Salts

program, considerable effort was devoted to salt purification by HF/H2 sparging of the molten salt, which is described in numerous reports.52–55 In addition to removing moisture/oxide impurities, the purification also removes other halide contaminants such as chloride and sulfur. Sulfur is usually present in the form of sulfate and is reduced to sulfide ion, which is swept out as H2S in the sparging operation. Methods were also developed to ensure the purity of the reagents used to purify the salts and clean the container surfaces used for corrosion testing. Another means of purification that can be performed after sparging involves simply reducing the salt with a constituent active metal such as an alkali metal, beryllium, or zirconium. While such active metals will remove oxidizing impurities such as HF, moisture, or hydroxide, they will not affect the other halide contaminants that influence sulfur removal. Therefore, it seems inevitable that the HF/H2 sparging operation, either by itself or followed by a reducing (active metal) treatment, will be a necessity. Although a great deal of effort can be devoted to purify the molten-salt mixture in the manner described above, it is primarily useful in producing materials for research purposes, without the possibility of interference from extraneous impurities. Removal of oxygen-containing impurities from chloride and fluoroborate salts is considerably more difficult because the fluoride ion more readily displaces oxygen from most compounds than does the chloride ion and because borate and hydroxyborate impurities are difficult to remove by fluorination with HF. Nearly all of the chloride salts prepared for corrosion studies have had relatively high levels of oxygencontaining impurities. The typical salt preparation for these studies involved treatment of reagent chlorides by drying the solid salt under vacuum, followed by prolonged treatment with dry HCl gas, and finishing with an inert gas purge of HCl from the salt. This treatment is not effective in removing the last portion of bound oxygen from the salt. Depending on the salt composition, oxygen contents of up to a few percent (in wt%) may remain. A more effective method for removing oxygen is needed to investigate the basic corrosion mechanism in pure chloride salts; otherwise, the effects of oxygen-containing species will dominate the apparent corrosion response. The use of carbochlorination has been recommended56 for the removal of oxygen and it has been claimed that salts with very low oxygen content (3 ppm) can be produced by this method.57

229

A similar type of purification improvement is needed for fluoroborates. Previous treatments with HF and BF3 (to avoid loss of BF3 from the melt) were not as effective as desired. There is also a need for accurate analytical methods for the determination of oxygen in melts and, in certain cases, it is necessary to identify the oxygen-containing species (oxide type, hydroxyl, etc).

5.10.3 Developments in Materials for Different Reactor Systems 5.10.3.1

Molten-Salt Reactor

When considering an MSR, the materials required fall into three main categories: (1) metallic components for primary and secondary circuits, (2) graphite in the core, and (3) materials for molten-salt fuel reprocessing systems. The metallic material used in constructing the primary circuit of an MSR will operate at temperatures up to 700–750  C. The outside of the primary circuit will be exposed to nitrogen containing sufficient air from inleakage to make it oxidizing to the metal. No metallic structural members will be located in the highest flux. The inside of the circuit, depending on design, will be exposed to salt-containing fission products and will receive maximum fast and thermal fluencies of about 1–2 1020 and 5–8 1021 neutrons cm2, respectively. The operating lifetime of a reactor will be in the range of 30–50 years, with an 80% load factor. Thus, the metal must have moderate oxidation resistance, must resist corrosion by the salt, and must not be subject to severe embrittlement by neutrons.49 The material must be fabricable into many products (plate, piping, tubing, and forgings) and capable of being formed and welded both under well-controlled shop conditions and in the field. The primary circuit involves numerous structural shapes ranging from a few centimeters thick to tubing having wall thicknesses <1 mm. These shapes must be fabricated and joined, primarily by welding, into an integral engineering structure. Thus, the activities required for development of material for the primary circuits will suffice for secondary circuits if supplemented by information on the compatibility of the material with the coolant salt. Graphite is the principal material other than salt in the core of the reference breeder reactor design with a thermal spectrum and thorium fuel cycle.16,17 In nonmoderated MSR concepts (e.g., MOSART1 and MSFR4), graphite is used only as a reflector.

230

Material Performance in Molten Salts

The graphite core and reflector structures will operate in a fuel salt environment over a range of temperatures from 500 up to 800  C. In any MSR design, graphite is, of course, subject to radiation damage. There are two overriding requirements in the graphite in MSRs, namely, that both molten salt and xenon be excluded from open pore volume. Any significant penetration of the graphite by the fuelbearing salt would generate a local spot, leading to enhanced radiation damage to the graphite and perhaps local boiling of the salt. This requires that the graphite be free of gross structural defects and that the pore structure be largely confined to diameters <106 m.49 135Xe will diffuse into graphite and affect the neutron balance. This requires graphites of very low permeability, for example, 108 cm2 s1. The requirements of purity and impermeability to salt are easily met by high-quality, finegrained graphite, and the main problems arise from the requirement of stability against radiation-induced distortion.58 Material selection for molten-salt fuel reprocessing systems depends, of course, upon the nature of the chosen process and the design of the equipment to implement the process. For MSRs,58 the key operations in fuel reprocessing are (1) removal of uranium from the fuel stream for immediate return to the reactor, (2) removal of 233Pa and fission product zirconium from the fuel for isolation and decay of 233Pa outside the neutron flux, and (3) removal of rareearth, alkali-metal, and alkaline-earth fission products from the fuel solvent before its return, along with the actinides, to the reactor. Such a processing plant will present a variety of corrosive environments. The most severe ones are (a) the presence of molten salt along with gaseous mixtures of F2 and UF6 at 500  C and that with absorbed UF6, so the average valence of uranium is near 4.5 (UF4.5) at temperatures near 550  C and (b) the presence of molten salts (either molten fluorides or molten LiCl) and molten alloys containing bismuth, lithium, thorium, and other metals at temperatures near 650  C as well as HF–H2 mixtures and molten fluorides, along with bismuth in some cases, at 550–650  C. High radiation and contamination levels will require that the processing plant be contained and have strict environmental control. If the components are constructed of reactive materials, such as molybdenum, tantalum, or graphite, the environment must be an inert gas or a vacuum to prevent deterioration of the structural material. Obviously, materials capable of long-term service under these conditions must be provided.

The main developments necessary to do this within the above-mentioned categories are described below. 5.10.3.1.1 Metallic materials for primary and secondary circuits

An extremely large body of literature exists on the corrosion of metal alloys by molten fluorides. Much of this work was done at ORNL and involved either thermal convection or forced convection flow loops. To select the alloy best suited to this application, an extensive program of corrosion tests was carried out on the available commercial nickel-base alloys and austenitic stainless steels.26,34–38 5.10.3.1.1.1 Development status of nickel-base alloys in ORNL

These tests were performed in a temperature gradient system with various fluoride media and different temperatures (maximum temperature and temperature gradient). Chromium, which is added to most alloys for high-temperature oxidation resistance, is quite soluble in molten fluoride salts. Metallurgical examination of the surveillance specimens showed corrosion to be associated with outward diffusion of Cr through the alloy. It was concluded that the chromium content should be maintained as low as reasonably possible to keep appropriate air oxidation properties. Corrosion rate is marked by initial rapid attack associated with dissolution of Cr and is largely driven by impurities in the salt.26,34–38 This is followed by a period of slower, linear corrosion rate behavior, which is controlled by a mass transfer mechanism dictated by thermal gradients and flow conditions. Minor impurities in the salt can enhance corrosion by several orders of magnitude and must be kept to a minimum. Dissolution can be mitigated by a chemical control of the redox in salts, for example, by small additions of elements such as Be. Corrosion increased dramatically as the temperature was increased and is coupled to plate-out in the relatively cooler regions of the system, particularly in situations where high flow is involved. The nuclear power aircraft application for which MSRs were originally developed required that the fuel salt operate at around 850  C. Inconel 600, out of which the Na,Zr,U/F ARE test reactor was built, was not strong enough and corroded too rapidly at the design temperature for long-term use.12–14 The existing alloys were screened for corrosion resistance at this temperature and only two were found to be satisfactory: Hastelloy B (Ni–28% Mo–5% Fe) and

Material Performance in Molten Salts

Hastelloy W (Ni–25% Mo–5% Cr–5% Fe). However, both aged at service temperature and became quite brittle due to formation of Ni–Mo intermetallic compounds.38 On the other hand, Hastelloy B, in which chromium is replaced with molybdenum, shows excellent compatibility with fluoride salts at temperatures in excess of 1000  C. Unfortunately, Hastelloy B cannot be used as a structural material in high-temperature systems because of its agehardening characteristics, poor fabrication ability, and oxidation resistance. Tests performed at 815  C especially showed Ni-base alloys to be superior to Fe-base alloys. This led to the development of a tailored Ni-base alloy, called INOR-8 or Hastelloy N (see Table 3), with a composition of Ni–16% Mo–7% Cr–5% Fe–0.05% C.35 The alloy contained 16% molybdenum for strengthening and chromium sufficient to impart moderate oxidation resistance in air, but not enough to lead to high corrosion rates in salt. Hastelloy N has excellent corrosion resistance to molten fluoride salts at temperatures considerably above those expected in MSR service; further (see Table 4), the resultant maximum corrosion rate of Hastelloy N measured in extensive Li,Be,Th,U/F loop testing at reactor operating temperatures was below 5 mm year1.42–46 Higher redox potential set in the system Li,Be,Th,U/F made the salt more oxidizing. At ORNL, the dependence of corrosion versus flow rate was tested in the range of velocities from 1 to 6 m s1. It was reported that the influence of

231

the flow rate was significant only during the first 1000–3000 h. Later, the corrosion rates, as well as their difference, decreased.43 The mechanical properties of Hastelloy N at operating temperatures are superior to those of many stainless steels and are virtually unaffected by long-time exposure to salts. The material is structurally stable in the operating temperature range, and the oxidation rate is <2 mils in 100 000 h. No difficulty is encountered in fabricating standard shapes when the commercial practices established for nickel-base alloys are used. Tubing, plates, bars, forgings, and castings of Hastelloy N have been made successfully by several major metal manufacturing companies, and some of these companies are prepared to supply it on a commercial basis. Welding procedures have been established, and a good history of reliability of welds exists. The material has been found to be easily weldable with a rod of the same composition. Inconel is, of course, an alternate choice for the primary circuit structural material, and much information is available on its compatibility with molten salts and sodium. Although probably adequate, Inconel does not have the degree of flexibility that Hastelloy N has in corrosion resistance to different salt systems, and its lower strength at reactor operating temperatures would require heavier structural components. Hastelloy N alloy was the sole structural material used in the Li,Be,Zr,U/F MSRE and contributed

Table 3

Chemical composition of the nickel–molybdenum alloys (mass %)

Element

Hastelloy N (INOR-8)

Ti-modified Hastelloy N 197258

Nb-modified Hastelloy 197658

HN80M-VI

HN80MTY (EK-50)

MONICR

Ni Cr Mo Ti Fe Mn Nb Si Al W Cu Co Ce Zr B S P C

Base 7.52 16.28 0.26 3.97 0.52 – 0.5 0.26 0.06 0.02 0.07 – – <0.01 0.004 0.007 0.05

Base 6–8 11–13 2 0.1 0.15–0.25 0–2 0.1 – – – – – – 0.001 0.01 0.01 0.05

Base 6–8 11–13 – 0.1 0.15–0.25 1–2 0.1 – – – – – – 0.001 0.01 0.01 0.05

Base 7.61 12.2 0.001 0.28 0.22 1.48 0.040 0.038 0.21 0.12 0.003 0.003 – 0.008 0.002 0.002 0.02

Base 6.81 13.2 0.93 0.15 0.013 0.01 0.040 1.12 0.072 0.020 0.003 0.003 – 0.003 0.001 0.002 0.025

Base 6.85 15.8 0.026 2.27 0.037 <0.01 0.13 0.02 0.16 0.016 0.03 <0.003 0.075 <0.003 0.003 0.003 0.014

– The elements were neither added to the melt nor determined.

232

Test loop

NCL-1255

Summary of ORNL loop corrosion tests for fuel fluoride salts Structural material

NCL-16

Hastelloy N þ 2% Nb Hastelloy N

MSRE

Hastelloy N mod. Ti
0.5 Hastelloy N

NCL-15A NCL-18

Hastelloy N Hastelloy N

NCL-21A

Hastelloy N

NCL-23

Hastelloy N, mod. 1% Nb Inconel 601

NCL-24 FCL-2b

Hastelloy N, mod. 3.4% Nb Hastelloy N Hastelloy N mod. 1% Nb

Molten salt (mol%)

Fluid test conditions

Corrosion rate (mm year1)

Circulation mode

Tmax ( C)

Natural convection

704

90

80 439

–

–

Natural convection V = 2.5 cm s1

704

170

28 000

660

1.0

65LiF–29.1BeF2–5.0– ZrF4–0.9UF4 66LiF–34BeF2 73LiF–2BeF2–5ThF4 68LiF–20BeF–11.7ThF– 0.3UF4 71.7LiF–16BeF2–12ThF4– 0.3UF4 U4þ/U3þ 104

Fuel circuit

654

22

21 800

675 700 654

0.5 0.9 8.0

Coolant circuit Natural convection Natural convection

580 677 704

35 55 170

26 100 35 400 11 600

580 677 704

no 1.5 1.2

Natural convection

704

138

10 009

704

3.5

1004

704

3.7

71.7LiF–16BeF2–12ThF4– 0.3UF4 U4þ/U3þ 40 68LiF–20BeF–11.7ThF– 0.3UF4 71.7LiF–16BeF2—12ThF4– 0.3UF4 U4þ/U3þ 100

Natural convection

704

138

721

704

34

V ¼ 1 cm s1 Natural convection

704

138

1500

704

2.5

Forced convection

704

138

4309

704

2.6

2242

704

0.4

70LiF–23BeF2–5ZrF4–1UF4– 1ThF4 66.5LiF–34BeF2–0.5UF4

Tmax ( C)

Specim. temperature ( C)

V ¼ 1 cm s1

V ¼ 2.5–5m s1

Exposure (h)

Source: Koger, J. W. Alloy compatibility with LiF–BeF2 salts containing ThF4 and UF4, ORNL-TM-4286; ORNL: Oak Ridge, TN, 1972; Keiser, J. R.; et al. Salt corrosion studies, ORNL-5078; ORNL: Oak Ridge, TN, 1975; pp 91–97; Keiser, J. R. Compatibility studies of potential molten-salt breeder reactor materials in molten fluoride salts, ORNL-TM-5783; ORNL: Oak Ridge, TN, 1977.

Material Performance in Molten Salts

Table 4

Material Performance in Molten Salts

significantly to the success of the experiment.15,16 Less severe corrosion attack (<20 mm year1) was seen for the Hastelloy N in contact with the MSRE fuel salt at temperatures up to 704  C for 3 years (26 000 h). The most striking observation is the almost complete absence of corrosion for Hastelloy N during the 3-year exposure to the MSRE coolant Li,Be/F salt (see Table 4). Two main problems of Hastelloy N requiring further study were observed during the construction and operation of the MSRE. The first was that the Hastelloy N used for the MSRE was subject to a kind of ‘radiation hardening,’ due to accumulation of helium at grain boundaries.59,60 Later, it was found that modified alloys with fine carbide precipitates within the grains would hold the helium and avoid this migration to the grain boundaries. Nevertheless, it is still desirable to design well-blanketed reactors in which the exposure of the reactor vessel wall to fast neutron radiation is limited. The second problem was the discovery of tiny cracks on the inside surface of the Hastelloy N piping for the MSRE. It was found that these cracks were caused by the fission product tellurium.61,62 Later work showed that tellurium attack could be controlled by keeping the fuel under reducing conditions.63–65 This is done by adjustment of the chemistry so that about 2% of the uranium is in the form of UF3, as opposed to UF4. This can be controlled rather easily now that good analytical methods have been developed. If the UF3 to UF4 ratio becomes too low, it can be raised by the addition of some beryllium metal, which, as it dissolves, will rob some of the fluoride ions from the uranium. When the ORNL studies were terminated in early 1973, considerable progress had been made in finding solutions to both problems.58 Since the two problems were discovered a few years apart, the research on them appears to have proceeded independently. However, the work must be brought together for the production of a single material resistant to both problems. It was found that the carbide precipitate that normally occurs in Hastelloy N could be modified to obtain resistance to embrittlement by helium. The presence of 16% molybdenum and 0.5% silicon led to the formation of coarse carbide that was of little benefit. Reduction of the molybdenum concentration to 12% and the silicon content to 0.1% and the addition of a reactive carbide former such as titanium led to the formation of a fine carbide precipitate and an alloy with good resistance to embrittlement by helium. The desired level of titanium was about

233

2%, and the phenomenon was confirmed by numerous small laboratories and commercial melts by 1972. Because the intergranular embrittlement of Hastelloy N by tellurium was noted in 1970, ORNL’s understanding of the phenomenon was not very advanced at the conclusion of the program in 1973. Numerous parts of the MSRE were examined, and all surfaces exposed to fuel salt formed shallow intergranular cracks (IGC) when strained. Some laboratory experiments had been performed in which Hastelloy N specimens were exposed to low partial pressures of tellurium metal vapor and, when strained, formed IGC very similar to those noted in parts from the MSRE. Several findings indicated that tellurium was the likely cause of the intergranular embrittlement, and the selective diffusion of tellurium along the grain boundaries of Hastelloy N was demonstrated experimentally. One in-reactor fuel capsule was operated in which the grain boundaries of Hastelloy N were embrittled and those of Inconel 601 (Ni, 22% Cr, 12% Fe) were not. These findings were in agreement with laboratory experiments in which these same metals were exposed to low partial pressures of tellurium metal vapor. Thus, at the close of the program in early 1973, tellurium had been identified as the likely cause of intergranular embrittlement, and several laboratory and in-reactor methods were devised for studying the phenomenon. Experimental results had been obtained that showed variations in sensitivity to embrittlement of various metals and offered encouragement that a structural material could be found that resisted embrittlement by tellurium. The alloy composition favored at the close of the ORNL program in 1973 is given in Table 3 with the composition of standard Hastelloy N. The reasoning at that time was that the 2% titanium addition would impart good resistance to irradiation embrittlement and the 0–2% niobium addition would impart good resistance to intergranular tellurium embrittlement. Neither of these chemical additions was expected to cause problems with respect to fabrication and welding. When the ORNL program was restarted in 1974, top priority was given to the tellurium-embrittlement problem.63–66 A small piece of Hastelloy N foil from the MSRE had been preserved for further study. Tellurium was found in abundance, and no other fission product was present in detectable quantities. This showed even more positively that tellurium was responsible for the embrittlement. Considerable effort was spent in seeking better methods of exposing test specimens to tellurium.

234

Cracking parameter (frequency (cm−1) ⫻ average depth (µm))

Material Performance in Molten Salts

175

150

Tellurium penetration (mm)

125

100

75

500 hr

200 hr

50 hr

25

100 hr

50

0 0

5

10 15 20 Square root of time (√h)

25

Figure 1 Tellurium penetration versus time for Hastelloy N exposed at 700  C to LiF–BeF2–ThF4 (72–16–12 mol%) containing Cr3Te4. Data obtained by Atomic Energy Station (AES). Reproduced from Keiser, J. R. Status of tellurium–Hastelloy N studies in molten fluoride salts, ORNL-TM-6002; ORNL: Oak Ridge, TN, 1977.

The most representative experimental system developed for exposing metal specimens to tellurium involved suspending the specimens in a stirred vessel of salt with granules of Cr3Te4 and Cr5Te6 lying at the bottom of the salt. Tellurium, at a very low partial pressure, was in equilibrium with the Cr3Te4 and Cr5Te6, and exposure of Hastelloy N specimens to this mixture resulted in crack severities similar to those noted in samples from the MSRE (see Figure 1). As a result of these studies,65,66 it was found that Hastelloy N exposed in salt-containing metal tellurides, such as LixTe and CryTex, undergoes grain boundary embrittlement similar to that observed in the MSRE. The embrittlement is a function of the chemical activity of tellurium associated with the telluride. Controlling the oxidation potential of the salt coupled with the presence of chromium ions in the salt appears to be an effective means of limiting tellurium embrittlement of Hastelloy N. The degree of embrittlement can be reduced by alloying additions to the Hastelloy N. The addition of 1–2 mass % Nb significantly reduces embrittlement, but small

900

Reducing

Oxidizing

600

300

0

10 20 40 70 100 200 400 Salt oxidation potential (U(IV)/U(III))

Figure 2 Cracking behavior of Hastelloy N exposed for 260 h at 700  C to molten-salt breeder reactor fuel salt containing Cr3Te4 and Cr5Te6. Reproduced from Mc Coy, H. E.; et al. Status of materials development for molten-salt reactors, ORNL-TM-5920; ORNL: Oak Ridge, TN, 1978.

additions of titanium or additions of up to 15 at.% Cr do not affect embrittlement. It was found that if the U(IV)/U(III) ratio in fuel salt is kept below about 60, embrittlement is essentially prevented when CrTel.266 is used as the source of tellurium (see Figure 2). However, further studies are needed to assess the effects of longer exposure times and measure the interaction parameters for chromium and tellurium under varying salt oxidation potentials. Studies of irradiation embrittlement and intergranular tellurium embrittlement have progressed to the point where suitable options are available. Postirradiation creep properties were acceptable for Hastelloy N modified with 2% Ti, 1–4% Nb, or about 1% each of Nb and Ti. The 2%-Ti-modified alloy was made into a number of products, and some problems with cracking during annealing were encountered. The other alloys were only fabricated into 1/2-in.-thick plates and 1/4-in.-diameter rods, and no problems were encountered. All alloys had excellent weldability. There are no obvious reasons why all of these alloys cannot be fabricated after development of suitable scale-up techniques. The resistance of all of these alloys to irradiation embrittlement depends upon the formation of a fine dispersion of MC-type carbide particles. These particles act as sites for trapping He and prevent it from reaching the grain boundaries where it is embrittling.

235

Material Performance in Molten Salts

8000

8926

2500 h Crack frequency (number cm–1) 3 crack depth (µm)

These alloys would be annealed after fabrication into basic structural shapes and the fine carbides would precipitate during service in the temperature range from 500 to 650  C. If the service temperature exceeds this range significantly, the carbides begin to coarsen, and the resistance to irradiation embrittlement diminishes. Although some heated specimens of the 2%-Ti-modified alloys and 3–4%-Nb-modified alloys had acceptable properties after irradiation at 760  C, it is very questionable whether these alloys can realistically be viewed for service temperatures above 650  C. One very important development related to intergranular embrittlement by tellurium was a number of experimental methods for exposing test metals to tellurium under fairly realistic conditions. The use of metal tellurides, which produce low partial pressures of tellurium at 700  C, as sources of tellurium provided experimental ease and flexibility. The inreactor fuel capsules also proved to be very effective experiments for exposing metals to tellurium and other fission products. The observation that the severity of cracking in standard Hastelloy N was influenced by the oxidation state of the salts adds the further experimental complexity that the oxidation state must be known and controllable in all experiments involving tellurium. It is unfortunate that Ti-modified alloys were developed so far because of their good resistance to irradiation embrittlement before it was learned that the titanium addition, even in conjunction with Nb, resulted in alloys that were embrittled by Te as badly as standard Hastelloy N. However, this situation was due to the time spread of almost 6 years between discoveries of the two problems and could not be prevented. The addition of 1–2% Nb to Hastelloy N resulted in alloys with improved resistance to IGC by tellurium, but that did not totally resist cracking. Samples of these alloys were exposed to Te-containing environments for more than 6500 h at 700  C with very favorable results (see Figure 3). However, cyclic tests where crack propagation is measured in the presence of Te will be required to clarify whether the Nb-modified alloys have adequate resistance to Te. The mechanism of improved cracking resistance due to the presence of Nb in the alloy is not known, but it is hypothesized that Nb forms surface reaction layers with the Te in preference to its diffusion into the metal along grain boundaries. Screening experiments with various alloys elucidated some other possibilities. Nickel-base alloys containing 23% Cr (Inconel 601) resisted cracking,

7000

6000 1000 h 5000

4000 250 h 3000

2000

1000

0

0

1

2 3 Nb content (%)

4

5

Figure 3 Variations of severity of cracking with Nb content. Samples were exposed for the indicated times to salt-containing Cr3Te4 and Cr5Te6 at 700  C. Reproduced from Mc Coy, H. E.; et al. Status of materials development for molten-salt reactors, ORNL-TM-5920; ORNL: Oak Ridge, TN, 1978.

whereas alloys containing 15% Cr (Inconel 600, Hastelloy S, and Cr-modified Hastelloy N) cracked as badly as standard Hastelloy N. However, it is questionable whether the corrosion rate of alloys containing 23% Cr would be acceptable in salt. Type 304 stainless steel and several other iron-base alloys were observed to resist intergranular embrittlement, but these alloys also have questionable corrosion resistance in fuel salts. Alloys containing appreciable quantities of chromium are attacked by molten salts, mainly by the removal of chromium from hot-leg sections through reaction with UF4, if present, and with other oxidizing impurities in the salt. The removal of chromium is accompanied by the formation of subsurface voids in the metal. The depth of void formation depends strongly on the operating temperatures of the system and on the composition of the salt mixture. If 300 series stainless steels are exposed to uranium-fueled salt under the same closed system conditions, the corrosion is manifested in surface voids of decreased Cr content to a depth of

236

Material Performance in Molten Salts

60–70 mm at 600–650  C. Data on corrosion rates obtained in experiments with molten Li,Be,Th,U/F mixtures for 304SS and 316SS at ORNL42 as well as later at the RRC-Kurchatov Institute19 for the Russianmade austenitic steels 12H18N10T (Fe–18% Cr–10% Ni–1% Ti–0.12% C) and AP-164 (Fe–15% Cr–24% Ni–1.5% Ti–4% W–0.08% C) agree well with each other. It is possible that a salt can be made adequately reducing to allow iron-base alloys to be used. This possibility must be pursued experimentally, because thermodynamic and kinetic data are not available to allow analytical determination. The discoveries that cracking severity was influenced by the oxidation state of the salt and that the salt could be made sufficiently reducing to prevent cracking in standard Hastelloy N opened many doors. Thus, alloys containing Ti could be used to take advantage of their excellent resistance to irradiation damage if they were protected from cracking by Te. Even standard Hastelloy N could be used in part of the system where the neutron flux was very low. The research toward finding a material for constructing an MSR that has adequate resistance to irradiation embrittlement and IGC by tellurium has progressed. ORNL findings suggest very strongly that an MSR could be constructed of 1–2%-Nbmodified Hastelloy N and operated very satisfactorily at 650  C. 5.10.3.1.1.2 Progress on Ni–Mo alloy development at RRC-Kurchatov Institute

In Russia, materials testing for the Th–U MSR were started at the RRC-Kurchatov Institute in 1976.19,20,47 It was substantiated by data accumulated in the ORNL MSR program on nickel-base alloys for UF4-containing salts. The Ni-base alloy HN80MT was chosen as a base. Its composition (in wt%) is Ni–6.9% Cr–0.02% C–1.6% Ti–12.2% Mo–2.6% Nb. The development and optimization of the HN80MT alloy was envisaged to be performed in two directions: improvement of alloy resistance to selective chromium corrosion and increase in alloy resistance to tellurium intergranular corrosion and cracking. About 70 differently alloyed specimens of HN80MT were tested. Among alloying elements were W, Nb, Re, V, Al, Mn, and Cu. The main finding was that alloying by aluminum with a decrease of titanium to 0.5% revealed significant improvement in both the corrosion and mechanical properties of the alloy. Chromium corrosion and intergranular

corrosion reached the minimum value at an aluminum content in the alloy of 2.5%. Irradiation effect on corrosion activity of fuels was also studied. It was shown that there was no radiation-induced corrosion at least up to a power density of 10 W/cm3 in a molten LiF–BeF2–ThF4–UF4 mixture. A subsequent radiation study of 13 alloy modifications was conducted. Specimens (in nitrogen atmosphere) were exposed to the reactor neutron field up to the fluency of 3 1020 neutrons cm2. Mechanical properties of alloys were studied at temperatures of 20, 400, and 650  C for nonirradiated and irradiated specimens. The best postirradiation properties were shown for alloys modified by Ti, Al, and V. Lastly, corrosion under the stressed condition was studied. It is known that tensile strain promotes an opening of intergranular boundaries and thus boosts intergranular corrosion and creates the prerequisites for IGC. The studies did not reveal any dependence of intergranular corrosion on the stress up to the value 240 MPa, that is, 0.8 of a tensile yield of the material and 5 times higher than typical stresses in Li,Be,Th,U/F MSR designs. The results of the combined investigation of mechanical, corrosion, and radiation properties of various alloys of HN80MT permitted the RRC-Kurchatov Institute to suggest the Ti- and Al-modified alloy as an optimum container material for the MSR design. This alloy, named HN80MTY (or EK-50), has the composition given in Table 3. In the thermal convection loop operated with the molten Li,Be,Th,U/F salt system, the HN80MTY alloy specimens have shown a maximum corrosion rate of 6 mm year1 (see Table 5) as for the HN80MT alloy it was two times lower.20,67 The corrosion was accompanied by selective leaching of chromium into the molten salt, which was evidenced by the 10-fold increase in its concentration for 500 h of exposure. Similar oxidizing conditions, characterized by the same content of Fe and Ni impurities in the salt, existed in testing a standard Hastelloy N alloy on the NCL-21A loop (see Table 4) operated with a molten Li,Be,Th,U/F salt system at ORNL.46 For the NCL-21A loop, the uniform corrosion rate of Hastelloy N specimens was about 5 mm year1. However, in the NCL-21A loop, the maximum temperature was somewhat lower (704  C) than in the RRC-Kurchatov Institute experiments (750  C), and in addition, fission products, including Te, were not added into the circuit. A comparison with corrosion data obtained at ORNL43,46 indicates that the HN80MT and

Material Performance in Molten Salts

Table 5

Summary of Russian loop corrosion tests for fluoride salts

Loop

Salt (mol%)

Specimen material

Tmax ( C)

Solaris

46.5LiF–11.5NaF–42KF

620

KI C1 KI C2 KI C3 KI F1 KI F2 KI M1 KURS-2 VNIITF

92NaBF4–8NaF

KI T1

LiF–NaF–BeF2 + Cr3Te4

12H18N10T HN80MT 12H18H10T AP-164 HN80MT HN80MT HN80MTY 12H18N10T 12H18N10T HN80MT HN80MTY MONICR HN80MT HN80MTY MONICR

71.7LiF–16BeF2– 12ThF4–0.3UF4 + Te 66LiF–34BeF2 + UF4 66LiF–34BeF2 + UF4 LiF–NaF–BeF2 + PuF3

DT ( C)

237

Duration (h)

Corrosion rate (mm year1)

20

3500

630 630 630 750 750 630 750 700

100 100 100 70 70 100 250 100

1000 1000 1000 1000 1000 500 750 1600

700

10

400

50 22 250 50 12 3.0 6.0 20 25 5 5 19 3 3 15

AP-164 alloy with a composition of Fe–22–25% Ni–14–16% Cr–4–5% W–0.5–1% Mn–1.4–1.8Ti–0.6% Si–0.08% C–0.035% P and 12H18N10T stainless steel with a composition of Fe–11–13% Ni–17–19% Cr–2% Mn–0.6–0.8% Ti–0.8% Si–0.12% C–0.035% P. Source: Novikov, V. M.; Ignatiev, V. V.; Fedulov, V. I.; Cherednikov, V. N. Molten Salt Reactors: Perspectives and Problems; Energoatomizdat: Moscow, USSR, 1990; Ignatiev, V. V.; Novikov, V. M.; Surenkov, A. I.; Fedulov, V. I. The state of the problem on materials as applied to molten-salt reactor: Problems and ways of solution, Preprint IAE-5678/11; Institute of Atomic Energy: Moscow, USSR, 1993.

HN80MTY resistance is higher than that of the standard Hastelloy N. This conclusion is confirmed by the microphotographs of HN80MT and HN80MTY alloy specimens after corrosion tests. Physical metallurgy studies were done on longitudinal metallographic sections of specimens subjected to tensile tests (see Figures 4 and 5). Under static conditions at T ¼ 600  C, there is only a slight tendency of HN80MT to IGC, and corrosion defects are observed along grain boundaries at a depth of 20–30 mm. With an increase of temperature to 750  C, the defect depth increases to 60 mm. Transition to loop tests at T ¼ 750  C show even more expressed IGC (see Figure 4). Massive defects in the material along the grain boundaries at full depth and further cracking over boundaries of the following grains were found. The defect area reached 130 mm. The alloy resistance to IGC was estimated from a parameter K, which is equal to the product of the number of cracks on a 1-cm length of a longitudinal section of specimens subjected to tensile strain multiplied by an average crack depth in micrometers. The estimated value for the parameter K in these conditions (ampoule isothermal tests at T ¼ 750  C) amounts to 1300 pc mm cm1. For the HN80MT alloy, this value is more than 5 times lower than that of a standard Hastelloy N alloy subjected to similar testing conditions.66

Therefore, the maximum operating temperature for HN80MT alloy in a reactor should be reduced at least to 700  C, and rigorous control of oxidation– reduction potential of the fuel salt is necessary. A completely different picture was observed in testing HN80MTY alloy specimens. No IGC traces were found, both in static tests under stress conditions (at 650–800  C up to 245 MPa) and in thermal convection loops up to T ¼ 750  C.20,67 The thermal convection tests show that corrosion proceeds uniformly along the entire grain volume, giving rise to a small porous layer near the material surface in contact with the fuel salt at the depth of 15–30 mm (see Figure 5). Thus, choosing effective alloying additions can solve the problem of IGC for nickel alloys in fuel salts containing fission products. The corrosion and other characteristics of the developed HN80MTY alloy makes it possible to consider it as a promising structural material for Th–U MSRs with a maximum working temperature of 750–800  C.20 The weldability of the alloy, however, needs improvement. To suppress crack formation during welding, the metal penetration regime was set up and maximum heat removal from the welded joint was ensured. These measures made it possible to increase significantly the characteristics of the welded joints. The manufacturing of a heat exchanger

238

Material Performance in Molten Salts

(a)

(b)

(c)

(d)

Figure 4 Microphotographs of the Ni–Mo alloy specimen surface layer (enlargement 100) after 500 h exposure to tellurium containing melt 71.7LiF–16BeF2–12ThF4–0.3UF4. (a) HN80MT isothermal tests, Texposure ¼ 600  C; (b) HN80MT isothermal tests, Texposure ¼ 750  C; (c) HN80MT nonisothermal tests in loop, Texposure ¼ 750  C; (d) standard Hastelloy N isothermal tests, Texposure ¼ 700  C. Reproduced from Ignatiev, V. V.; Novikov, V. M.; Surenkov, A. I.; Fedulov, V. I. The state of the problem on materials as applied to molten-salt reactor: Problems and ways of solution, Preprint IAE-5678/11; Institute of Atomic Energy: Moscow, USSR, 1993.

(a)

(b)

Figure 5 Microphotographs of HN80MTY alloy specimens surface layer (enlargement 100) after 500 h exposure to the tellurium containing melt 71.7LiF–16BeF2–12ThF4–0.3UF4. (a) Isothermal tests, Texposure ¼ 750  C and (b) nonisothermal tests in loop, Texposure ¼ 750  C. Reproduced from Ignatiev, V. V.; Novikov, V. M.; Surenkov, A. I.; Fedulov, V. I. The state of the problem on materials as applied to molten-salt reactor: Problems and ways of solution, Preprint IAE-5678/11; Institute of Atomic Energy: Moscow, USSR, 1993.

confirmed once more that the HN80MTY alloy is technologically effective both in hot and cold process stages.19 In a recent study, the central focus of the corrosion studies was the compatibility of Ni-base alloys with a molten Li,Na,Be/F salt system as applied to the primary circuit of MOSART fuelled with different compositions of actinide trifluorides from LWR spent fuel without U–Th support.68–70 Prior ORNL

examinations71 of Inconel in natural convection loops, which circulated molten 24LiF–53NaF– 23BeF2 and 34LiF–49NaF–15BeF2 (mol%) mixtures with an excess of free fluoride ion content, revealed no evidence of attack in either the hot or cold areas of the loop. However, a microscopic examination of specimens removed from the cooler coil did reveal the presence of a small amount of metallic deposit. These studies (see Table 5) included (1) compatibility tests

Material Performance in Molten Salts

Table 6

Nickel–molybdenum alloys’ mechanical properties

Alloy

HN80M-VI HN80MTY (EK-50) MONICR

239

Specimens in the delivery condition, T ¼ 20  C

Specimens after the corrosion tests, T ¼ 20  C

s0.2 (kg mm2)

sB (kg mm2)

d (%)

s0.2 (kg mm2)

sB (kg mm2)

d (%)

110.4 110.1 112.7 40.3 39.6

119 121.7 122.3 73.5 70.0

10.9 10.6 9.1 57.2 54.0

50.0 52.5 50.5

75.0 78.5 75.3

103.9 90.0 89.5 39.6 40.3 39.6 38.5 36.3 36.3

120.0 103.0 101.1 76.9 73.4 76.0 67.5 62.5 65.0

28.0 22.4 22.4 56.0 55.0 55.2 53 39 38

between Ni–Mo alloys and molten 15LiF–58NaF– 27BeF2 (mol%) salt in a natural convection loop with a measurement of redox potential; (2) the effect of PuF3 addition in molten 15LiF–58NaF–27BeF2 (mol%) salt on compatibility with Ni–Mo alloys; and (3) Te corrosion for molten 15LiF–58NaF–27BeF2 (mol%) salt and Ni–Mo alloys in stressed and unloaded conditions with measurement of the redox potential. Three Hastelloy N-type modified alloys, particularly HN80M-VI with 1.5% Nb, HN80MTY with 1% Al, and MONICR68 with about 2% Fe, were chosen for the study in the corrosion facilities (see Tables 3 and 6). Results of a 1200 h loop corrosion experiment69 with online redox potential measurement demonstrated that high-temperature operations with molten 15LiF–58NaF–27BeF2 (mol%) salt are feasible using carefully purified molten salts and loop internals. In the established interval of salt redox potential, 1.25–1.33 V relative to a Be reference electrode, corrosion is characterized by uniform loss of weight with low rate from sample surfaces. Under such exposure, the salt contained less than (in mass %): Ni – 0.004; Fe – 0.002; Cr – 0.002. Specimens of HN80M-VI and HN80MTY alloys from the hot leg of the loop exposed at temperatures from 620 to 695  C showed a uniform corrosion rate from 2 to 5 mm year1. For the MONICR alloy, this value was up to 20 mm year1 (see Figure 6). No significant change in corrosion behavior of material samples was found in the melt due to the presence of 0.5 mol% PuF3 addition in 15LiF– 58NaF–27BeF2 (mol%) salt. Specimens of HN80MVI from the loop exposed during 400 h at 650  C showed a uniform corrosion rate of about 6 mm year1. Under such exposure, the salt contained about (in mass %): Ni – 0.008; Fe –0.002; Cr – 0.002. No traces of IGC were found for any specimen after loop

54 51 53

tests, even in the melt with PuF3 addition. Data from chemical analysis of the specimen’s surface layer showed a decrease in chromium content by 10–20 mm. Tellurium IGC testing of the Ni–Mo alloys,69,70 without and under mechanical load (80 MPa), for the 15LiF–58NaF–27BeF2 (mol%) melt under dynamic and static conditions was carried out at 700  C with exposure times of 100, 250, and 400 h at 1.2 V system redox potential. Under stress exposure to tellurium in the 15LiF–58NaF–27BeF2 melt, the depth of cracks for MONICR specimens reached 220 mm (K > 10 000 pC mm cm1). For HN80M-VI specimens tested without stress, rather low IGC intensity was observed (K ¼ 690 pC mm cm1). However, under stress, the intensity of the HN80M-VI alloy cracking increased more than twice and the crack depth reached 125 mm. HN80MTY alloy is the most resistant to tellurium IGC of the Ni–Mo alloys studied. The intensity of its cracking under stress is K ¼ 880 pC mm cm1 (twice as low as that of HN80M-VI alloy). The effect on the resistance to tellurium corrosion of Nb, Al, Ti, Re, and Mn doping agents added to the HN80M-type alloy was also studied in the Li, Na,Be/F facility at the RRC-Kurchatov Institute.70 It was shown that both Re and Y additions only slightly increased the alloy’s resistance to tellurium cracking. The alloy doped with Nb alone significantly increased IGC resistance. Addition of Mn gave a significant increase in alloy resistance to tellurium IGC. Therefore, testing of alloys with various compositions of doping elements to enhance the alloy’s resistance to tellurium IGC should be continued in a thermal convection loop with long exposure times. Finally, as can be seen from the considerations above, new findings in the developments of Ni–Mo alloys for MSRs with fuel salt temperatures up to 750  C shift the emphasis from alloys modified with titanium and rare earths to those modified with

240

Material Performance in Molten Salts

9 8 7 6 HN80M-VI

5 4 3 2 1

(a) 100 mm

0 0

10

20

30

40

50

0

10

20

30

40

50

8 7 6 5 HN80M-VI

4 3 2 1

(b) 60 mm

0 7 6 5

HN80MTY

4 3 2 1

(c) 100 mm

0 10

0

20

30

40

8 7 6 5 4 MONICR

3 2 1

(d) 100 mm

0 0

10

20

30

40

50

60

Figure 6 Chromium distribution (mass %) versus depth of the surface layer (mm) of specimens after corrosion tests in the loop: (a) quenched HN80M-VI, Texposure = 690  C; (b) hot deformed HN80M-VI, Texposure = 670  C; (c) quenched HN80MTY, Texposure = 620  C; and (d) MONICR in the Scoda delivery state, Texposure = 690  C. Reproduced from Ignatiev, V. V.; et al. Nucl. Technol. 2008, 164(1), 130–142.

Material Performance in Molten Salts

niobium at ORNL58 and aluminum at the RRCKurchatov Institute.19 Subsequent steps for this type of metallic materials development must involve (1) irradiation, corrosion, tellurium exposure, mechanical property, and fabrication tests to finalize the composition for scale up; (2) procurement of large commercial heats of the reference alloy; (3) mechanical property and corrosion tests of at least 10 000 h duration; and (4) development of design methods and rules needed to design a reactor (breeder or burner) to be built of the modified alloy. 5.10.3.1.1.3 Alternative approaches

Certainly, some less mature approaches are possible and could be of interest for new MSR concepts. For example, Ni–W–Cr alloys have been recently proposed by Centre National de la Recherche Scientifique (CNRS) in France for their high potential to corrosion resistance for very high-temperature operation (>750  C).5 Temperatures >850  C would require the use of new solutions such as refractory alloys or graphite. Included in further evaluation should also be the assessment of (1) new proposed solvent systems (e.g., Li,Th/F), (2) increased fuel salt outlet temperatures >750  C, and (3) lower salt redox potentials from the point of view of establishing potentials that must be maintained to avoid IGC for Ni-base alloys. 5.10.3.1.2 Graphite for the core

Extensive prior work has demonstrated that graphite is compatible with molten fluoride salts (these are fundamental properties and are not particularly dependent on manufacturing). Much of the experience and data obtained in the gas-cooled reactor programs is directly applicable to MSRs. In particular, the limited lifetime of graphite resulted from neutron-induced damage. (See also Chapter 4.10, Radiation Effects in Graphite). By the time the MSBR program at ORNL was cancelled in early 1973, the dimensional changes of graphite during irradiation had been studied for a number of years.49,58 These changes depend largely on the degree of crystalline isotropy, but the volume changes fall into a rather consistent pattern. There is first a period of densification during which the volume decreases, and then a period of swelling in which the volume increases. The first period is of concern only because of the dimensional changes that occur, and the second period is of concern because of the dimensional changes and the formation of cracks. The formation of cracks would eventually allow salt to penetrate the graphite. The damage rate increases

241

with increasing temperature, and hence, the graphite section size should be kept small enough to prevent temperatures in the graphite from exceeding those in the salt by a wide margin. For fast neutron fluences greater than about 3 1022 neutrons cm2 (En > 50 keV), the rate of graphite expansion becomes quite rapid, and it appears that this represents an upper limit to acceptable exposure of the graphite (L Pm 200, where L is the moderator lifetime in full-power years and Pm is the maximum core power density in W cm3). For example, in the MSBR design, the maximum power density is about 70 W cm3 and the useful graphite life would be about 3–4 years at full power.16,17 It was further required that the graphite be surfacesealed to prevent penetration of xenon into the graphite. Since replacement of the graphite would require considerable downtime, there was a strong incentive to increase the fluence limit of the graphite. A considerable part of the ORNL graphite program was spent in irradiating commercial graphites and samples of special graphites with potentially improved irradiation resistance. The approach taken to sealing the graphite was surface sealing with pyrocarbon. Because of the neutronic requirements, other substances could not be introduced in sufficient quantity to seal the surface. Fission product gases, notably 135Xe, will diffuse into graphite with some effect on neutron balance (poison fraction for uncoated graphite is about 0.01–0.02). It is desirable, especially for high flux cores, to hold Xe poisoning to the lowest possible level (poison fraction of 0.005). This requires graphites of very low permeability, for example, 108 cm2 s1. The pyrolytic sealing work at ORNL was only partially successful. It was found that extreme care had to be taken to seal the material before irradiation. During irradiation, the injected pyrocarbon actually caused expansion to begin at lower fluences than those at which it would occur in the absence of the coating. Thus, the coating task was faced with a number of challenges. The most detailed creep data exist on the US and German graphites for the HTR plant designs.49 But these graphites, because of their coarse granularity and large pore size, are unsatisfactory for molten-salt applications. Fine-grained, isotropic, molded, or isostatically pressed, high-strength graphite suitable for core support structures (e.g., Carbone USA grade 2020 or Toyo Tanso grade IG-11058 and Russian-made GSP type graphite19) is available today. Past experience has also demonstrated techniques for accommodating any radiation-induced dimensional changes in the graphite reactor vessel insulation. Development of sealing

242

Material Performance in Molten Salts

techniques should continue both with the pulseimpregnation technique and isotropic pyrolytic coatings applied at somewhat higher temperatures. With relaxed requirements for breeding performance in the new wave of MSR concepts relative to the MSBR, the requirements for graphite would be diminished.58 First, the lower gas permeability requirements mean that graphite damage limits can be raised. Second, if the salt flow rate through the core is decreased from the turbulent regime down to laminar one, the salt film at the graphite surface may offer sufficient resistance to xenon diffusion so that it will not be necessary to seal the graphite. Finally, the peak neutron flux at the graphite location can be reduced to levels such that the graphite will last for the lifetime of the reactor. As noted above, the lifetime criterion adopted for the breeder was that the allowable fluence would be about 3 1022 neutrons cm2. This was estimated to be the fluence at which the structure in advanced graphites would contain sufficient cracks to be permeable to xenon. Experience has shown that, even at volume changes of about 10%, the graphite is not cracked but is uniformly dilated. For some nonbreeder devices where xenon permeability will not be of concern, the limit will be established by the formation of cracks sufficiently large for salt intrusion. It is likely that current technology graphites could be used to 3 1022 neutrons cm2 and that improved graphites with a limit of 4 1022 neutrons cm2 could be developed. Also, early efforts show promise that graphites with improved dimensional stability can be developed. Finally, for nonmoderated MSR concepts (e.g., MSFR and MOSART) with a graphite reflector, there is no strong requirement on gas permeability (108 cm2 s1), but molten salt should be excluded from the open pore volume (pore structure < 106 m). The last requirement can be met by currently available commercial graphite (See also Chapter 4.10, Radiation Effects in Graphite). 5.10.3.1.3 Materials for molten-salt fuel reprocessing system

For most established MSR concepts, processes involving (1) removal of uranium from fuel salt by fluorination and (2) selective extraction of transuranium elements and fission products from fuel salt into liquid bismuth are considered the most promising methods available. The material considerations below are oriented in these directions. Nickel or nickel-base alloys can be used for the construction of fluorinators and containment of F2,

UF6, and HF, though these metals would require protection by a frozen layer of fuel solvent over areas where contamination of the molten stream by the otherwise inevitable corrosion products would be severe. Many years of experience in fabrication and joining of such alloys have been accumulated17,49 in the construction of reactors and associated engineering hardware. The corrosion of L nickel (low-carbon nickel with: 99.36% Ni; 0.02% C; 0.26% Fe; 0.06% Cu; 0.26% Mn; 0.04% Si; 0.001% S) and its alloys in the severe environment represented by fluorination of UF6 from molten salts has been studied in some detail.72 Most of the data were obtained during operation of two plant-scale fluorinators constructed of L nickel at temperatures ranging from 540 to 730  C. A number of corrosion specimens (20 different materials) were located in the fluorinators. Several specimens, including INOR-1, had lower rates of maximum corrosive attack than L nickel.72,73 Nevertheless, L nickel, protected where necessary by frozen salt, is the preferred material for the fluorination–UF6 absorption system since the other alloys would contribute volatile fluorides of chromium and molybdenum to the gaseous UF6. Absorption of UF6 in molten salts containing UF4 is proposed as the initial step in fuel reconstitution for many Th–U MSR concepts. The resulting solution, containing a significant concentration of UF5, is quite corrosive. In principle, and perhaps in practice, the frozen salt protective layer could be used with vessels of nickel. It has been shown74,75 that gold is a satisfactory container in small-scale experiments, and plans to use this expensive, but easily fabricable, metal in engineering-scale tests have been described.76 Most of the essential separations required of the processing plant are accomplished by selectively extracting species from salt streams into bismuth– lithium alloys or vice versa. Moreover, no satisfactory alternative to the selective extraction metal transfer process for removal of rare-earth fission products has been identified (reductive extraction from moltensalt fluoride mixtures into lithium–bismuth alloys).58 These extractions pose difficult materials problems. Materials for containment of bismuth and its alloys are known, as are materials for containment of molten salts. Unfortunately, the two groups have few common members. Carbon steels are not really satisfactory long-term containers for molten fluorides.77,78 Nickel-base alloys are known17,49 to be inadequate containers for bismuth. Corrosion studies79,80 showed molybdenum to resist attack by bismuth and to have no appreciable

Material Performance in Molten Salts

mass transfer at 500–700  C for periods up to 10 000 h. Moreover, molybdenum is known to have excellent resistance to molten fluorides.17,49 The external environment could be inert gas, but the problems in fabricating molybdenum are huge. The resistance of tantalum and its alloys to molten fluorides has long been questioned, but no definitive tests had been made when previous surveys were written.17,49 Further tests are obviously necessary, but continued satisfactory operation of the Ta–16% W loop with fuel salts must be considered encouraging. Pure tantalum and some of its alloys with tungsten (in particular, T-111 alloy: 8% W, 2% Hf, balance Ta) have been shown to be usefully compatible with molten bismuth and bismuth–lithium alloys. Tantalum is easy to fabricate, but the external environment must be a high vacuum.58 Graphite, which has excellent compatibility with fuel salt, also shows promise for the containment of bismuth. Compatibility tests to date have shown no evidence of chemical interaction between graphite and bismuth containing up to 3 wt% (50 at.%) lithium. However, the largest open pores of most commercially available polycrystalline graphites are penetrated to some extent by liquid bismuth. Capsule tests81 of three commercial graphites (ATJ, AXF-5QBG, and Graphitite A) were conducted for 500 h at 700  C using both high-purity bismuth and bismuth–3 mass % lithium. Although penetration by pure bismuth was negligible, the addition of lithium to the bismuth appeared to increase the depth of permeation and presumably altered the wetting characteristics of the bismuth. Limited penetration of graphite by bismuth solutions may be tolerable. If not, several approaches have the potential for decreasing the extent to which a porous graphite is penetrated by bismuth and bismuth–lithium alloys. Two wellestablished approaches are multiple impregnations with liquid hydrocarbons, which are then carbonized and/or graphitized, and pyrocarbon coatings. Graphite can be adequately protected at the outside with an inert gas, but it is difficult to fabricate into complex shapes. As the chemistry of the processing system is engineered further through pilot plants, the precise type of hardware needed will be better defined. Significant additional research and development will necessarily be concerned with detailed tests of material compatibility and studies of welding, brazing, and other joining techniques, as well as joint design. Facilities for static testing, operation of thermal convection loop assemblies, and fabrication and operation of forced convection (pumped) loops will be required, along

243

with sophisticated equipment for welding, brazing, etc., under carefully controlled atmospheres. Such facilities have been used routinely in the past and involve little, if any, additional development.

5.10.4 Advanced High-Temperature Reactor When considering materials performance in the AHTR,82 the materials can be classified into three main categories: (1) graphite and C/C composites, (2) low-pressure reactor vessel materials, and (3) high-temperature metallic components. The graphite core, reflector and vessel insulation, and C/C composite core supports and control rods will operate in a molten-salt environment over a range of temperatures from 500 to 1100  C or higher (peak temperature being selected as a trade-off between reactor thermal inertia, thermal blanket system performance, and material properties). It is anticipated that, for the AHTR, properly designed and manufactured C/C composite structures will demonstrate similarly good properties in the presence of molten fluoride salts and better mechanical properties. The reactor vessel materials3 must be capable for operation at 500  C and may need to withstand temperature excursions to 800  C for 100 h under accident conditions. The vessel must demonstrate adequate strength and creep resistance (long-term and short-term), good thermal-aging properties, low-irradiation degradation, fabricability, good corrosion resistance, and the ability to develop and maintain a high-emissivity surface in air. As previously noted, nickel-base alloys demonstrate good corrosion resistance to molten salts. Therefore, ORNL proposed82 that stable, high-strength, hightemperature materials, such as 9Cr–1MoV, be coated with a high-nickel coat for the reactor vessel application. Should the vessel be required to withstand excessive off normal temperatures, base materials such as 304L, 316L, 347, Alloy 800H, or HT may be appropriate. In addition, monolithic materials with adequate corrosion resistance to molten fluoride salts and high-temperature strength may include Alloy 800H or HT, Hastelloy N, and Haynes 242. Performance of the suggested materials needs to be evaluated, especially at higher temperatures. Further, the ability to develop and maintain a highemissivity layer on the surface of the vessel exposed to argon or air must be demonstrated, but this is not considered a major barrier.

244

Material Performance in Molten Salts

High-temperature metallic or composite materials are needed for use up to 1000  C in the presence of molten fluoride salts on one side and an insulation system in contact with air on the other side. Piping and heat exchangers are examples for the latter conditions. Pumps and other components submerged below the primary salt pool will need to survive higher temperatures for short times or be replaceable at reasonable expense. The metallic materials used in these environments must demonstrate adequate strength (long-term and short-term), good thermal-aging properties, low-irradiation degradation, fabricability, and good corrosion resistance. Based on material maturity and the need for high nickel for fluoride corrosion resistance, stable, high-strength, high-temperature metallic materials such as Inconel 617, Haynes 230, Alloy 800H, Hastelloy X or XR, VDM 602CA, and HP modified with a coating with high-nickel content could be possible candidates for detailed evaluation.3,26 Should higher temperature alloys be required, Haynes 214, cast Ni-base superalloys (for pumps), and ODS MA 754 are possible candidates. Recent experience suggests that, should the oxidation potential of the salt be made very reducing, it may be possible to use ODS MA 956 (an iron-base alloy). These monolithic materials will require more testing and data development. For composite materials, liquid-siliconimpregnated (LSI) composites, with chemical vapor deposition carbon coatings, may be promising for use for pumps, piping, and heat exchangers.3 LSI composites have several potentially attractive features, including the ability to maintain nearly full mechanical strength to temperatures approaching 1400  C, inexpensive and commercially available fabrication materials, and the capability for simple machining and joining of carbon–carbon performs, allowing the fabrication of highly complex component geometries. As already discussed, corrosion activity of molten salts is dependent upon the major salt constituents and impurities in the salt. The coolant salt can be prepared and maintained in such a way that impurities do not control the corrosion response. It is expected that coolant salts can be used at significantly higher temperatures than were established in the MSR design because of the different corrosion characteristics of a clean salt coolant versus a molten salt-containing actinides and fission product fluorides. A wider range of material options also exists. The presence of uranium dissolved in the salt was always found to accelerate corrosion, and there exist additional methods to prevent corrosion when uranium is not present in the salt.

The equilibrium level of dissolved chromium has been measured for fuel salts, but not for coolant salts.83–85 Although information on fuel salts is not directly applicable to coolants, it is expected that fuel solvents that experience minimal corrosion would also be better coolants.26 Review of dissolved chromium levels for various fuel salts again reveals that the molten 46.5LiF–11.5NaF–42KF (in mol%) mixture stands somewhat apart from the other salts as it sustains a higher degree of corrosion. It also appears that there is some benefit in avoiding a very acidic (high ZrF4 or BeF2 content) system and that a salt mixture that has a nearly complete coordination shell (2:1 ratio of alkali halide to Zr or Be and heavier alkali salt) has the least potential for supporting corrosion based on temperature sensitivities. This approach is a significant oversimplification, as the identity of the various species is very important. For example, the saturating species that contain chromium are different for each of these salts. Although <10% of all corrosion testing was done with salts that were free of uranium, this small fraction amounts to a significant body of work because of the extensive test program carried out. The results of testing for uranium-free salts reveals that Hastelloy N (INOR-8), just as it is for fuel salts (see previous section), is a superior choice (rather than Inconel or stainless steels) for coolant salts. The corrosion is so intense and the duration so short for most Inconel tests that it is hard to make a judgment about which salt is the least susceptible to corrosion. For Hastelloy N loops at temperatures up to 700  C, the corrosion is so minor that it is hard to sort out corrosion effects due to the salt composition. Again, a molten 46.5LiF–11.5NaF–42KF (in mol%) mixture is among the worst. Some additional Inconel loop tests86,87 were conducted with special fuel salt mixtures in which the ZrF4 and BeF2 concentrations were varied in an attempt to select the best composition. However, these tests were somewhat inconclusive because of the short test duration (500 h) and impurity effects. Within the resolution of these tests, the following trends were verified: very basic (FLiNaK) and very acidic (LiF–ZrF4) salts showed the worst performance.26 Corrosion tests of Hastelloy N, Hastelloy X, Haynes-230, Inconel-617, and Incoloy-800H at a high temperature of 850  C were performed at the US University of Wisconsin-Madison in a molten 46.5LiF–11.5NaF–42KF (in mol%) mixture, with the goal of ranking alloy suitability for the AHTR

Material Performance in Molten Salts

core.88 In particular, an attempt was made to simulate material performance in the corrosion system with a primary salt coolant, metal reactor vessel, and graphite fuel materials. The isothermal tests were performed for 500 h in sealed graphite crucibles under an argon cover gas, without any redox measurement and control strategy. Certainly, graphite crucibles may accelerate the corrosion process by promoting the formation of carbide phases on the walls of the test crucibles, but they did not alter the basic corrosion mechanism. Corrosion was noted to occur predominantly by release of Cr from the alloys, an effect that was particularly pronounced at the grain boundaries of these alloys. Mass loss due to corrosion generally correlated with the initial Cr content of the alloys, and was consistent with the Cr content measured in the salts after corrosion tests. The corrosion attack was more severe for Hastelloy N (6.3% Cr), where Cr depletion up to depths of about 50 mm was observed. Hastelloy X (21.3% Cr) exhibited grain boundary attack up to depths of at least 300 mm below the surface. Inconel-617 (22.1% Cr) was uniformly depleted in Cr up to depths of about 100 mm from the surface and experienced dramatic grain boundary corrosion throughout the thickness of the sample. Similar attack was observed for Haynes-230 (22.5% Cr); however, the surface of Haynes-230 exhibited a Ni-enriched layer. For Haynes-230, W-rich precipitates were observed at the grain boundaries due to the relatively high W content of this alloy, demonstrating that W, like Mo, is resistant to attack from molten fluoride salt. The fundamental reason why Haynes230 experienced more weight loss than the other high Cr-containing alloys needs further investigation. Two Cr-free alloys, Ni-201 and Nb–1Zr, were also tested. Ni-201, a nearly pure Ni alloy with minor alloying additions, exhibited good resistance to corrosion, whereas Nb–1Zr alloy exhibited extensive corrosion attack. At various periods at ORNL, control of the oxidation–reduction state of the salt was explored as a means to minimize corrosion. However, it was not practical, because strong reductants either reduced zirconium or uranium in the salt to a metal that plated on the alloy wall or resulted in some other undesirable phase segregation. During the MSRE operation, periodic adjustment of the U(III)/U(IV) ratio was effective in limiting corrosion in the fuel circuit. Keiser89 also explored the possibility of using metallic beryllium to reduce corrosion in stainless steel containing a LiF–BeF2 salt, where the oxidation potential of the salt could be lowered by

245

buffering with metallic beryllium without concerns for disproportionation of uranium trifluoride; the corrosion rate was decreased at 650  C from 8 to 2 mm year1. This treatment was effective only as long as the metallic beryllium was immersed in the salt. There was little, if any, buffering capacity in this salt to maintain the reducing environment throughout the melt. Del Cul et al.90 have identified and tested candidate agents that could be used as redox buffers to maintain a reducing environment in the coolant circuit. None of these redox-control strategies has been developed to the extent that we can rely on them for a definite salt selection. However, some useful observations can be made in this regard. For a lower temperature system (<750  C), it appears that Hastelloy N is fully capable of serving as a containment alloy without the need for a sophisticated redox strategy. Even an alkali fluoride, such as a molten 46.5LiF–11.5NaF– 42KF (in mol%) mixture, could be suitable. For temperatures in excess of 750  C and for alloys that contain more chromium (as most higher temperature alloys do), it appears that a reducing salt will be needed to minimize corrosion. Inconel without the benefit of a reducing environment was found to be unsuitable for long-term use. Only a mildly reducing environment is possible with a ZrF4-containing salt, since a strongly reducing redox potential would reduce ZrF4 itself. Much more reducing systems can be devised with either LiF–NaF–KF- or BeF2containing salts. Some very important material compatibility issues will have to be explored in order to use a highly reducing salt at these higher temperatures because events such as carbide formation and carburization/decarburization of the alloy (not discussed in the report) become a significant threat. Should low-chromium/chromium-free alloys or suitable clad systems be devised as a container, these problems with salt selection will largely disappear. However, in the absence of this solution, ORNL has considered two strategies: (1) select a salt that should support the minimum level of corrosion in the absence of a highly reducing environment (some ZrF4 salts, BeF2-containing salts) or (2) select a salt with a large redox window that can be maintained in a highly reducing state (LiF–NaF–KF- or BeF2containing salts). Given the expense and difficulty of carrying out development work with berylliumcontaining salts, ORNL proposed to explore the most promising ZrF4 salts without strong reductants and to explore LiF–NaF–KF with strong reductants and/or redox buffers.26

246

Material Performance in Molten Salts

5.10.5 Liquid-Salt-Cooled Fast Reactor There are no metallic components in the reference MSR core. While Hastelloy N or another nickel-base alloy is suitable for the reactor vessel, heat exchangers, pumps, main circulation pipes, drain tanks, and other equipment, it may not be suitable for LSFR incore components (structure and fuel cladding), which will be subjected to higher temperatures and receive large fast neutron fluences in the core. The metal incore components are likely to be the primary technical challenge for an LSFR, given the requirements for higher temperature service, resistance to neutron radiation damage, and corrosion resistance to liquid salts. The use of binary metallic materials (either clad or coated) may be desirable for some applications (including the reactor vessel), in order to confer appropriate strength and corrosion resistance. Generally, practical metal systems are based on (1) nickel-, (2) iron-, or (3) molybdenum-base alloys.3 The nickel-base alloys for high-temperature service in molten-salt coolants (but not as in-core components) have been evaluated as part of the AHTR research and development activities (see previous section). Some of these alloys are known to have excellent chemical compatibility with molten saltcoolants; however, there is mixed experience with the irradiation performance of nickel alloys. For a UK Prototype Fast Reactor experience91,92 with PE-16 irradiation performance, a nickel alloy (17Cr, 43Ni, 3Mo, 2.5Ti, 34Fe þ Al, in mass %) was good, but at lower temperatures than required for LSFRs. At the same time, many nickel-base super alloys have poor radiation stability (grain boundary embrittlement). The potential of nickel-base alloys at the higher temperatures for use in an LSFR core spectrum is not well understood. The strength of many nickel alloys is a consequence of nickel–silicon precipitates. In irradiation fields, these precipitates can dissolve, with the silicon migrating to the grain boundaries and causing the alloy to weaken. For these alloys, it may be feasible to overcome this difficulty by the development of oxide-dispersion-system (ODS) nickel alloys. However, only very limited work has been done on these systems. The iron-base alloys have good radiation resistance. The primary LSFR concern associated with iron alloys is their long-term high-temperature corrosion resistance. Some of these alloys are known to have excellent chemical compatibility with moltensalt coolants.

For example, static corrosion tests87 were performed recently in molten 46.5LiF–11.5NaF–42KF and 66LiF–34BeF2 (in mol%) mixtures at 500 and 600  C for 1000 h. The purpose was to study the corrosion characteristics of reduced-activation ferritic steels, JLF-1 (8.92Cr–2W) in the molten salts. The concentration of HF in the melts was measured by the slurry pH titration method before and after exposure. The HF concentration determined the fluoridation potential. The corrosion was mainly caused by dissolution of iron and chromium in the melts due to fluoridation and/or electrochemical corrosion. The corrosion depth of the specimens at 600  C, which was obtained from the weight losses, was 0.637 mm in 66LiF–34BeF2 melt and 6.73 mm in 46.5LiF–11.5NaF–42KF melt. The corrosion rate of SS304 and SS316L steels in 66LiF–34BeF2 melt after 1000 h exposure at 600  C was estimated as 10.6 and 5.4 mm year1, respectively. Russian experience20 with molten-salt fluorides and AP-164 iron-base alloy (14–16 Cr, 22–25 Ni, 0.5–1 Mn, 4–5 W, 1.4–1.8 Ti, 0.08 C, in mass %) was good, but also at lower temperatures (630  C) than required for LSFRs (700–750  C). To overcome the temperature limitations on ironbase systems, there has been significant developmental work on ODS iron alloys for fast reactors.93 These alloys contain rare-earth oxides, such as yttrium oxide, that enable iron alloys to maintain strength at up to 80% of their melting point versus 50% for traditional alloys. The limited corrosion testing of iron-base alloys in molten fluoride salt coolants indicates the potential for corrosion-resistant iron-base systems. However, more corrosion testing will be required to expand upper operating temperature to 700–750  C before gaining more confidence in such an approach. Molybdenum alloys are compatible with molten salts and have good thermophysical properties.94 Molybdenum has a very high melting point (2600  C), high thermal conductivity, and moderate thermal neutron cross-section (2.65 barns). However, isotopically separated molybdenum,94 with its very low nuclear cross-section, is an option. There are significant challenges with molybdenum alloys: (1) such alloys are difficult to weld, (2) the fracture toughness is somewhat low with concerns about radiation embrittlement, and (3) high-temperature oxidizing conditions must be avoided because MoO3 has a melting temperature of 795  C. The potential oxidation should not be a significant concern for an LSFR because the molten-salt mixture (such as

Material Performance in Molten Salts

sodium) will be subjected to chemically reducing conditions. The fracture toughness is a primary concern at lower temperatures. Radiation damage is temperature dependent and is minimized by operating at higher temperatures in the range from 650 to 1000  C. Molybdenum-base alloys may ultimately allow the construction of a very high-temperature LSFR and is a class of materials where higher temperatures improve material properties.

5.10.6 Secondary Circuit Coolants In the secondary circuits of an MSR, AHTR, LSFR, or SFR, the main difference compared to the primary one for the container metal will be the absence of fission products and uranium in the coolant salt and the much lower neutron fluences. This material must have moderate oxidation resistance and must resist corrosion by salt not containing fission products or uranium. The corrosion for molten fluoride salts was discussed in detail in previous sections. Very little corrosion data are available for nuclear application of molten-salt mixtures, including nitrate, chloride, and fluoroborate salts, than for molten fluoride salts, especially for temperatures above 600  C. A nitrate mixture of 60NaN03–40KNO3 (in mass %) has been proposed for use in the intermediate circuit of SFRs and LSFRs.2 This molten-salt mixture is attractive for such applications because of its high heat capacity, its low reactivity in the event of a leak to air or steam, and the low operating pressures required for its use. However, the feasibility of such a system depends partly on the compatibility of the salt with candidate structural alloys. Alloy 800 and types 304, 304L, and 316SS were exposed in a natural convection loop filled with molten NaNO3– KNO3 salt in the temperature range 375–600  C for more than 4500 h.7 The weight change data for the alloys indicated that (1) the metal in the oxide film constituted most of the metal loss; (2) the corrosion rate, in general, increased with temperature; and (3) although the greatest metal loss corresponded to a penetration rate of 25 mm year1, the rate was <13 mm year1 in most cases. Spallation had a significant effect on metal loss at intermediate temperatures in the type 304L stainless steel loop. Metallographic examinations showed no evidence of intergranular attack. The exposure resulted in the growth of thin oxide films on significant cold-leg deposits. Weight change data further confirmed the absence of thermal gradient mass transport processes in these draw salt

247

systems. Raising the maximum temperature of the type 316SS loop from 595 to 620  C dramatically increased the corrosion rate, and it appears that 600  C may be the limiting temperature for use of such alloys in draw salt. Material corrosion resistance for the SFR intermediate circuit containing 56LiCl–44LiOH (in mol%) was studied in Russia.20 The corrosion facility was constructed according to a three-loop scheme. The first circuit was filled with sodium to a maximum temperature of 530  C. The second (intermediate) circuit was filled with a molten 56LiCl–44LiOH mixture, which was heated from sodium to a maximum temperature of 490  C. The last circuit with a steam generator was cooled down to 430  C. The loop structural material was stainless steel 10H18N10T, with the exception of the sodium-salt heat exchanger and steam generator, which were made of a perlite 10H2M steel. Specimens of 10H2M (Fe–2% Cr–1% Mo–0.1% C), 10H18N10T, H9MFB (Fe–9% Cr–1% Mo–1% V–1% Nb), 08H14MF (Fe–14% Cr–1% Mo–1% V–0.08% C), 10H14GMFB (Fe–14% Cr–1% Mn–1% Mo–1% V–1% Nb), and 10H14N5MF (Fe–14% Cr–5% Ni–1% Mo–1% V) steels for corrosion tests were inserted correspondingly in the hot and cold legs of the loops. The specimen in the molten-salt loop was held for a little over 2000 h. The highest corrosion resistance was displayed by steels 10H18N10T and 10H14GFB, and the least by 10H2M. The 10H18N10T steel uniform corrosion rate in the molten 56LiCl– 44LiOH mixture was 50 mm year1. The metallographic study also determined that 10H18N10T steel corrosion had an intergranular character (crack depth up to 60 mm year1). However, it should be noted here that according to chemical analysis data, the initial salt composition contained about 1% H2O. Also, corrosion product deposits were found in some local sections of the molten-salt loop. For the purpose of comparison, the most relevant corrosion results for chloride salts are displayed in Table 7.95 These results do not conform to any expected or predictable trends. For example, the effect of chromium content in the alloy does not seem to be an important factor, and the effect of temperature is not clear. Unexpected variability in the tests very likely reflects variation in the purity of the starting materials and the degree to which impurities were excluded from the loop during operation. The corrosion rates are rather high for these salts at a relatively low temperature (550  C). These rates are similar to those experienced with fluoride salts

248 Table 7 Loopa

Material Performance in Molten Salts Summary of Brookhaven loop corrosion tests for chloride salts Loop material

%Cr–Ni–Mo in Fe alloy

Duration (h)

Tests with LiCl–KCl eutectic salt TCL-F 347SS 17.5–1.4–0.2 5500 TCL-L1 410SS 12.4–0.2–0.1 2200 TCL-L3 2.25Cr–1Mo 2.25–0–1 697 Tests with 30NaCl–20KCl–50MgCl2 eutectic salt (mol%) TCL-L5 347SS 17.5–11.4–0.2 2467 TCL-L6 410SS 12.4–0.2–0.1 3971 FCL-M1 347SS 17.5–11.4–0.2 1034 FCL-M2 347SS 17.5–11.4–0.2 656

Tmax ( C)

DT ( C)

Corrosion rate (mm year1)

575 570 550

155 160 150

12 50 Highb

500 494 520 515

45 42 0 0

93 79 31 256

a

TCL refers to thermal convection loop, FCL refers to a forced convection loop. No specimen corrosion depth was reported, but salt analysis showed 0.11% iron. Source: Susskind, H.; et al. Corrosion studies for a fused salt-liquid metal extraction process for the liquid metal fuel reactor, BNL-585; Brookhaven National Laboratory: Brookhaven, NY, 1960 b

Table 8 Summary of Hastelloy N corrosion loops with 8NaF–92NaBF4 salt at ORNL Loop

Duration (h)

Tmax ( C)

DT ( C)

Corrosion rate (mm year1)

NCL-13A NCL-14 NCL-17 NCL-20 FCL-1 FCL-2

30.627 39.202 24.865 19.928 17.000 5.300

607 607 607 688 621 621

125 150 100 250 167 167

16 13 24 24 29 23

Source: Bamberger, C. E.; Baes, C. F. Corrosion of Hastelloy N by fluoroborate melts, ORNL-4832; ORNL: Oak Ridge, TN, 1973; pp 44–45.

in contact with stainless steels and Inconel at 800  C and are much higher than those experienced with Hastelloy N in contact with fluoride salts at temperatures as high as 815  C. The corrosion database for fluoroborates is shown in Table 8.45 Improvement in fluoroborate salt purity during the MSBR program was responsible for a steadily decreasing level of corrosion in tests. For NaF–NaBF4 secondary coolant, ORNL data45 in thermal corrosion loops containing Hastelloy N specimens lie in the interval of 5–20 mm year1 and are determined mostly by the degree of salt purification. These data are in good agreement with later RRC-Kurchatov Institute corrosion studies20 for Russian nickel-base alloy of the HN80MT type (about 10–15 mm year1 at 600  C). The ORNL experience reveals that the coolant fluoroborate salt absorbs moisture quite readily with attendant generalized corrosion. On occasions when leaks developed, the corrosion rate had increased and then decreased as the impurities

were exhausted. During these periods of high corrosion, all components of the alloy were removed uniformly from the hot leg and deposited in the cold leg. Crystals of Na3CrF6 deposited in the cold regions as its solubility was exceeded. In summing up the results of work on secondary circuit coolants, it should be emphasized that, among the presently known high-temperature energy carriers with operating temperatures ranging from 300 to 550  C, the most promising for practical utilization is nitrate–nitrite molten-salt mixtures. As for the range of higher operating temperatures >700  C, there are some alternatives with different maturity. The database exists for fluoride-containing tests in the 800–900  C temperature range with both Inconel and Hastelloy N (INOR-8) alloys. No experience exists with loop corrosion tests using chlorides or fluoroborates at temperatures approaching the levels anticipated in the loop that transports heat from the AHTR or VHTR nuclear plant to the hydrogen production plant. There is a need to demonstrate and recommend an improved method for purification of chloride and fluoroborate salts to be used in corrosion tests.26,34 This new method should become a purification standard to be used in conjunction with corrosion tests. High-temperature corrosion tests with properly purified chloride salts should be conducted to confirm the possibility of using chloride and fluoroborate salts in the loop that will transport heat from the AHTR or VHTR nuclear plant to the hydrogen production plant. These tests should include both batch exposures and loop tests and will probably also require the innovative use of redox buffers to minimize corrosion.26,34

Material Performance in Molten Salts

References 1.

2. 3.

4. 5.

6. 7. 8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

21.

22.

Ignatiev, V.; Feynberg, O.; Smirnov, V.; Tataurov, A.; Vanukova, G.; Zakirov, R. Characteristics of molten salt actinide recycler and transmuter system. In Proceedings of International Conference on Emerging Nuclear Energy Systems, Brussels, Belgium, Aug 21–26, 2005; paper ICQ064. Forsberg, C. W.; Lebrun, C.; Merlet-Lucotte, E.; Renault, C.; Ignatiev, V. Revue Generale Nucleaire 2007, 4, 63–71. Forsberg, C. W.; et al. Design options for the advanced high-temperature reactor. In Proceedings of International Congress on Advances in Nuclear Power Plants, Anaheim, CA, Jun 8–12, 2008. Delpech, S.; et al. J. Fluorine Chem. 2009, 130(1), 11–17. Renault, C.; et al. European molten salt reactor. In Proceedings of Seventh European Commission Conference on Euroatom Research and Training in Reactor Systems, Prague, Czech Republic, Jun 22–24, 2009. Chechetkin, A. V. High Temperature Coolants; Gosenergoizdat: Moscow, USSR, 1962. Tortorelli, P. F.; DeVan, J. H. Thermal convection loop study of the corrosion of Fe–Ni–Cr alloys by molten NaNO3– KNO3, ORNL/TM-8298; ORNL: Oak Ridge, TN, 1982. Mar, R. W.; et al. The use of molten nitrate salts in high temperature energy conversion systems. In Proceedings of First International Symposium on Molten Salt Chemistry and Technology, Kyoto, Japan, Apr 20–22, 1983; pp 285–288. Taube, M. Fast Reactors Using Molten Chloride Salts as Fuel; Wurenlingen: Switzerland, 1978. Migai, L. L.; Taritsina, T. A. Corrosion Resistance of Structural Materials in Halides and their Mixtures. Moscow, USSR: Metallurgy, 1988. Rosenthal, M. W.; et al. Nucl. Appl. Technol. 1970, 8(2), 107–118. Bettis, E. S.; et al. Nucl. Sci. Eng. 1957, 2(6), 804–812. Manly, W. D.; et al. ARE-metallurgical aspects, ORNL-2349; ORNL: Oak Ridge, TN, 1957. Cottrell, W. B. Disassembly and postoperative examination of the aircraft reactor experiment, ORNL-1868; ORNL: Oak Ridge, TN, 1958. Haubenreich, P. N.; Engel, J. R. Nucl. Appl. Technol. 1970, 8(2), 107–140. Bettis, E. S.; Robertson, R. C. Nucl. Appl. Technol. 1970, 8 (2), 190. Rosenthal, M. W.; Haubenreich, P. N.; Briggs, R. B. Development status of molten salt breeder reactors, ORNL-4812; ORNL: Oak Ridge, TN, 1972; p 250. Furukava, K.; et al. J. Nucl. Sci. Technol. 1990, 27, 1157–1178. Novikov, V. M.; Ignatiev, V. V.; Fedulov, V. I.; Cherednikov, V. N. Molten Salt Reactors: Perspectives and Problems; Energoatomizdat: Moscow, USSR, 1990. Ignatiev, V. V.; Novikov, V. M.; Surenkov, A. I.; Fedulov, V. I. The state of the problem on materials as applied to molten-salt reactor: Problems and ways of solution, Preprint IAE-5678/11; Institute of Atomic Energy: Moscow, USSR, 1993. Grope detravail CEA–EDF. Materiax metalliques RSF: Synfese des etudes realisees entre 1973 et 1983, Rapport EDF HT/12/77/83; dossier materiax metalliques: Paris, France, 1983. Lecarpentier, D.; Vergnes, J. Nucl. Eng. Des. 2002, 216, 43–67.

23.

249

Forsberg, C. W.; Peterson, P. F.; Pickard, P. S. Nucl. Technol. 2003, 144, 289–302. 24. Forsberg, C. W.; et al. Practical aspects of liquid-salt-cooled fast-neutron reactors. In Proceedings of ICAPP’05, Seoul, Korea, May 15–19, 2005, paper 5643. 25. Petroski, R.; Hejzlar, P.; Todreas, N. E. Nucl. Eng. Des. 2009, doi:10.1016/j.nucengdes.2009.07.012. 26. Williams, D. F.; Toth, L. M.; Clarno, K. T. Assessment of candidate molten salt coolants for the advanced high-temperature reactor, ORNL/TM-2006/12; ORNL: Oak Ridge, TN, 2006. 27. Thoma, R. E. Ed. Phase diagrams of nuclear reactor materials, ORNL-2548; ORNL: Oak Ridge, TN, 1959. 28. Barton, C. J.; Strehlow, R. A. J. Inorg. Nucl. Chem. 1961, 18, 143. 29. Atomic Energy Commission. Molten salt breeder reactor concept; Quarterly report for period ending July 21, NP-19145; Bombay, India, 1971. 30. Bamberger, C. E.; Ross, R. S.; Baes, C. F. J. Inorg. Nucl. Chem. 1971, 33, 3591–3594. 31. Ward, W. T.; Strehlow, R. A.; Grimes, W. R.; Watson, G. M. J. Chem. Eng. 1960, 5(2), 137–142. 32. Barton, C. J. J. Phys. Chem. 1960, 64, 306–309. 33. Ignatiev, V.; Golovatov, Y.; Merzlyakov, A.; Panov, A.; Subbotin, V. Atom. Energy 2006, 101(5), 364–372. 34. Williams, D. F. Assessment of candidate molten salt coolants for the NGNP/NHI heat-transfer loop, ORNL/TM-2006/69; ORNL: Oak Ridge, TN, 2006. 35. Manly, W. D.; Richardson, L. S.; Vreeland, D. E. Prog. Nucl. Energy 1960, 2(4), 164–179. 36. DeVan, J. H.; Evans, R. B., III. Corrosion behavior of reactor materials in fluoride salt mixtures, ORNL/TM-328; ORNL: Oak Ridge, TN, 1962. 37. Adamson, G. M.; Crouse, R. S.; Manly, W. D. Interim report on corrosion by zirconium-base fluorides, ORNL-2338; ORNL: Oak Ridge, TN, 1961. 38. De Van, J. H.; Evans, R. B. In Proceedings of the Conference on Corrosion of Reactor Materials, Jun 4–8, 1962; IAEA: Vienna, Austria, 1962; pp 557–579. 39. Grimes, W. R. Nucl. Appl. Technol. 1970, 8(2), 137–155. 40. Baes, C. F. J. Nucl. Mater. 1974, 51(1), 149. 41. Evans, R. B., III; DeVan, J. H.; Watson, G. M. Self-diffusion of chromium in nickel-base alloys, ORNL-2982; ORNL: Oak Ridge, TN, 1960. 42. Koger, J. W. Alloy compatibility with LiF–BeF2 salts containing ThF4 and UF4, ORNL-TM-4286; ORNL: Oak Ridge, TN, 1972. 43. Keiser, J. R.; et al. Salt corrosion studies, ORNL-5078; ORNL: Oak Ridge, TN, 1975; pp 91–97. 44. Koger, J. W. Forced-convection loop corrosion studies, Huntley WR and ORNL-4832; ORNL: Oak Ridge, TN, 1973; pp 135–137. 45. Bamberger, C. E.; Baes, C. F. Corrosion of Hastelloy N by fluoroborate melts, ORNL-4832; ORNL: Oak Ridge, TN, 1973; pp 44–45. 46. Keiser, J. R. Compatibility studies of potential molten-salt breeder reactor materials in molten fluoride salts, ORNL-TM-5783; ORNL: Oak Ridge, TN, 1977. 47. Ignatiev, V.; Fedulov, V.; Novikov, V.; Surenkov, A. Voprocy Atomnoi Nauki i Tehkniki: Atomno – Vodorodnaya Energetika i Tehnologiya 1981, 3(10), 74–76. 48. De Van, H.; Distefano, J. R.; Eartherly, W. P.; Keiser, J. R.; Klueh, R. L. Materials considerations for molten salt accelerator based plutonium conversion system. In Proceedings of the Global’93 International Conference, Las Vegas, 1993.

250

Material Performance in Molten Salts

49. McNeese, L. E. Program plan for development of molten-salt breeder reactors, ORNL-5018; ORNL: Oak Ridge, TN, 1974; pp 5–46. 50. Maya, L. J. Inorg. Chem. 1976, 15(9), 2179–2184. 51. Mays, G. T. Distribution and behavior of tritium in the coolant-salt technology facility, ORNL/TM-5759; ORNL: Oak Ridge, TN, 1977. 52. Briggs, R. B. Molten salt reactor program semiannual progress report for period ending February 28, ORNL-3282; ORNL: Oak Ridge, TN, 1962. 53. Shaffer, J. H. Preparation of MSRE fuel, coolant and flush salt, ORNL-3708; ORNL: Oak Ridge, TN, 1964; pp 288–302. 54. Briggs, R. B. Molten salt reactor program semiannual progress report for period ending February 28, ORNL-3812; ORNL: Oak Ridge, TN, 1965; pp 121–168. 55. Shaffer, J. H. Preparation and handling of salt mixtures for the molten salt reactor experiment, ORNL-4616; ORNL: Oak Ridge, TN, 1971. 56. Cherginets, V. L. Handbook of Solvents; Chemical Technology: Toronto, Canada, 2001; Chap. 10.3, pp 633–635. 57. Cherginets, V. L.; Rebrova, T. P. Electrochim. Acta 1999, 45(3), 469–476. 58. Engel, J. R.; Bauman, H. F.; Dearing, J. F.; Grimes, W. R.; McCoy, E. H.; Rhoades, W. A. Development status and potential program for development of proliferation resistance molten salt reactor, ORNL-TM-6415; ORNL: Oak Ridge, TN, 1979. 59. Harries, O. R. J. Br. Nucl. Soc. 1966, 5, 74. 60. Mc Coy, H. E.; Roche, T. K. Post irradiation creep properties of modified Hastelloy N, ORNL-5078; ORNL: Oak Ridge, TN, 1975; pp 82–84. 61. Mc Coy, H. E.; et al. Intergranular cracking of structural materials exposed to fuel salt, ORNL-4782; ORNL: Oak Ridge, TN, 1972; pp 109–144. 62. Mc Coy, H. E.; et al. Intergranular cracking of structural materials exposed to fuel salt, ORNL-4832; ORNL: Oak Ridge, TN, 1972; pp 63–76. 63. Mc Coy, H. E.; et al. Metallographic examination of samples exposed to tellurium-containing environments, ORNL-5078; ORNL: Oak Ridge, TN, 1972; pp 108–113. 64. Mc Coy, H. E.; et al. Development of modified Hastelloy N, ORNL-5132; ORNL: Oak Ridge, TN, 1976; pp 42–162. 65. Mc Coy, H. E.; et al. Status of materials development for molten salt reactors, ORNL-TM-5920; ORNL: Oak Ridge, TN, 1978. 66. Keiser, J. R. Status of tellurium–Hastelloy N studies in molten fluoride salts, ORNL-TM-6002; ORNL: Oak Ridge, TN, 1977. 67. Ignatiev, V.; Fedulov, V.; Novikov, V.; Surenkov, A. Voprocy Atomnoi Nauki i Tehkniki: Atomno – Vodorodnaya Energetika i Tehnologiya 1989, 3, 23–25. 68. Ignatiev, V. V.; et al. Experience with alloys compatibility with fuel and coolant salts and their application to molten salt actinide recycler and transmuter. In Proceedings of International Congress on Advances in Nuclear Power Plants, Reno, NV, Jun 4–8, 2006. 69. Ignatiev, V. V.; et al. Atom. Energy 2006, 101(4), 278–285. 70. Ignatiev, V. V.; et al. Nucl. Technol. 2008, 164(1), 130–142. 71. Tripton, C. R. The Reactor Handbook 1960, Vol. 3, p 464. 72. Litman, A. P.; et al. Corrosion associated with fluorination in the ORNL fluoride volatility process, ORNL-2832; ORNL: Oak Ridge, TN, 1961.

73. 74. 75. 76. 77.

78.

79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90.

91. 92. 93.

94. 95.

Hags, L.; et al. Comparative tests of L nickel, D nickel, Hastelloy B and INOR-1, ORNL-5924; ORNL: Oak Ridge: TN, 1968; pp 49–52. Cavin, O. B.; et al. Molten salt reactor program semiannual report for period ending August 31, ORNL-4728; ORNL: Oak Ridge, TN, 1971; pp 173–176. Bennett, M. R. Molten salt reactor program semiannual report for period ending February 29, ORNL-5132; ORNL: Oak Ridge, TN, 1976; p 170. Counce, R. M. Molten salt reactor program semiannual report for period ending August 31, ORNL-5078; ORNL: Oak Ridge, TN, 1975; p 157. Brown, C. H.; et al. Measurement of mass transfer coefficients in a mechanically agitated, nondispersing contactor operating with a molten mixture of LiF–BeF2– ThF4 and molten bismuth, ORNL-5143; ORNL: Oak Ridge, TN, 1976. Savage, H. C.; et al. Engineering tests of metal transfer process for extraction of rare-earth fission products from a molten salt breeder reactor fuel salt, ORNL-5176; ORNL: Oak Ridge, TN, 1977. Shimotake, H.; et al. Trans. Am. Nucl. Soc. 1967, 10, 141–142. Siefert, J. W.; et al. Corrosion 1961, 17(10), 75–78. Cavin, O. B.; et al. Molten salt reactor program semiannual report for period ending February 29, ORNL-4782; ORNL: Oak Ridge, TN, 1972; p 198. Ingersoll, D. T.; et al. Status of preconceptual design of the advanced high-temperature reactor, ORNL/TM-2004/104; ORNL: Oak Ridge, TN, 2004. Jordan, W. H.; et al. Aircraft nuclear propulsion project quarterly progress report for period ending June 10, ORNL-2106; ORNL: Oak Ridge, TN, 1956; p 95. Jordan, W. H.; et al. Aircraft nuclear propulsion project quarterly progress report for period ending September 10, ORNL-2157; ORNL: Oak Ridge, TN, 1956; p 107. Jordan, W. H.; et al. Aircraft nuclear propulsion project quarterly progress report for period ending December 31, ORNL-2221; ORNL: Oak Ridge, TN, 1956; p 125. Jordan, W. H.; et al. Aircraft nuclear propulsion project quarterly progress report for period ending September 10, ORNL-2157; ORNL: Oak Ridge, TN, 1956; p 145. Kondo, M.; et al. Fusion Eng. Des. 2009, 84(7–11), 1081–1085. Olson, L. C.; et al. J. Fluorine Chem. 2009, 130, 62–73. Keiser, J. R.; et al. The corrosion of type 316 stainless steel to Li2BeF4 ORNL/TM-5782; ORNL: Oak Ridge, TN, 1977. Del Cul, G. D.; et al. Redox potential of novel electrochemical buffers useful for corrosion prevention in molten fluorides. In Proceedings of the Thirteenth International Symposium on Molten Salts Held within the 201st Meeting of the Electrochemical Society, Philadelphia, PA, May 12–17, 2002. Brown, C.; et al. Nucl. Eng. 1994, 35(4), 122–128. Konobeev, Y.; Birzhevoi, G. Atom. Energy 2004, 96(5), 365–373. Generation IV Nuclear Energy Research Advisory Committee. Generation IV roadmap: Description of candidate liquid-metal-cooled reactor systems, GIF-017–00; Department of Energy: Washington DC, 2002. Bakker, K.; et al. Nucl. Technol. 2004, 146, 325–331. Susskind, H.; et al. Corrosion studies for a fused salt-liquid metal extraction process for the liquid metal fuel reactor, BNL-585; Brookhaven National Laboratory: Brookhaven, NY, 1960.

5.11

Material Performance in Helium-Cooled Systems

R. Wright and J. Wright Idaho National Laboratory, Idaho Falls, ID, USA

C. Cabet Commissariat a l’Energie Atomique, Gif-sur-Yvette, France

ß 2012 Elsevier Ltd. All rights reserved.

5.11.1

Introduction

252

5.11.2 5.11.3 5.11.3.1 5.11.3.2 5.11.3.3 5.11.3.4 5.11.3.5 5.11.3.6 5.11.3.7 5.11.3.8 5.11.4 5.11.4.1 5.11.4.2 5.11.4.3 5.11.4.4 5.11.4.5 5.11.5 5.11.5.1 5.11.5.2 5.11.5.3 5.11.6 5.11.7 5.11.8 5.11.9 5.11.10 References

Experience with VHTR Systems Comparison of IHX Concepts Shell-and-Tube Plate and Fin Etched Plate Microchannel Heat Exchangers Plate-Stamped Heat Exchanger Foam IHX Capillary IHX Ceramic IHX Heat Exchanger Alloys Regulatory Issues Alloy 617 (52Ni–22Cr–13Co–9Mo) Alloy 230 (57Ni–22Cr–14W–2Mo–La) Alloy 800H (42Fe–33Ni–21Cr) Alloy X (47Ni–22Cr–9Mo–18Fe) Welding Base Metal Preparation and Filler Metal Selection Preheating, Interpass Temperatures, and Postweld Heat Treatment Nontraditional Joining Methods Control Rod Materials Core Barrel Materials Environmental Effects of VHTR Atmospheres on Materials Aging Effects Summary

252 254 254 255 255 256 256 257 257 257 257 259 259 262 263 264 264 265 265 265 266 269 269 273 275 276

Abbreviations AGCNR AVR DLOC GE GMAW GTAW HE HTGRs HTR-10

Advanced gas-cooled nuclear reactor Arbeitsgemeinschaft Versuchsreaktor Depressurized loss of coolant General Electric Gas-metal-arc welding Gas-tungsten-arc welding Heat exchanger High-temperature gas reactors High-temperature reactor-10MWth in China

HTTR IHX INL NGNP ODIN ORNL PBMR PCHE PCS

High-temperature engineering test reactor in Japan Intermediate heat exchanger Idaho National Laboratory Next generation nuclear plant Online Data & Information Network Oak Ridge National Laboratory Pebble bed modular reactor Printed circuit heat exchanger, Heatrics Division Ltd. Power conversion system

251

252

Material Performance in Helium-Cooled Systems

PFHE PSHE RCS RCSS RSS SMAW THTR VHTR

Plate and fin heat exchanger Plate stamped heat exchanger Reactivity control system Reactor control and shutdown system Reserve shutdown system Shielded metal arc welding Thorium Hochtemperatur Reaktor Very high-temperature reactor

5.11.1 Introduction Over the past decade, there has been renewed interest in very high-temperature reactor (VHTR) technology. This type of reactor is of interest because of a number of unique characteristics, including passive safety, electricity production on a more modest scale compared to light water plants that might be more compatible with the electrical distribution system in developing countries, and very high outlet temperature that can be used for process heat or hydrogen production. The relative value of electricity production or process heat applications varies considerably with world economic conditions. Currently, it appears that steam for process heat and hydrogen production will drive development of this technology rather than electricity production. There are currently two operating VHTR prototypes, the high-temperature engineering test reactor (HTTR) in Japan and the high-temperature reactor (HTR-10) in China. The HTTR is a 30 MWt (megawatts thermal output) prismatic core reactor and the HTR-10 is a 10 MWt pebble bed prototype reactor. Both the operating reactors are designed to investigate electricity production with the VHTR technology; however, each program has parallel activities to develop process heat and hydrogen production as well. The challenges for high-temperature materials are not significantly different for either prismatic or pebble bed reactor designs. Interest in specific applications for VHTR technology is evolving rapidly. It appears that the most significant immediate interest is in a reactor with an outlet temperature on the order of 750
C with a steam generator for either electricity generation or process heat. This technology would use a relatively mature conventional steam generator technology and is expected to present lower technical risk. Higher outlet temperatures using a heat exchanger between the primary helium coolant and a secondary gas are viewed to be higher risk development projects that offer the opportunity for outlet temperatures from

850 to 950
C for hydrogen production by thermochemical processes or higher temperature process heat for industrial applications. The material issues associated with reactor internals are not affected significantly by the reactor outlet temperature; however, heat exchangers operating at the higher outlet temperatures represent significantly different issues compared to steam generators. The focus of this chapter is on higher outlet temperature systems because of the development challenges. The next generation nuclear plant (NGNP) being developed in the United States is one particular VHTR concept that is under very active development and is typical of the development around the world. This reactor is being developed to produce hydrogen as well as electricity. Conceptual designs call for a gascooled reactor with an outlet temperature greater than the 850
C required to efficiently operate the hydrogen generation plant, with a maximum of 950
C. While the design concepts are not yet final, it is highly probable that helium will be the primary coolant in the reactor. The primary material in the core will be graphite, and the prime candidates for high-temperature metallic components are the nickel-based alloys Alloy 617 or Alloy 230. An artist’s representation of one concept for the reactor and power conversion vessel and the associated hydrogen generation plants is shown in Figure 1. In this representation, a heat exchanger carries most of the reactor thermal output to a secondary circuit that powers a turbine for electricity generation. An additional heat exchanger takes 10% of the thermal energy of the reactor and diverts it as process heat to the hydrogen production plant. The most critical metallic component in the VHTR system is the intermediate heat exchanger (IHX). This heat exchanger will operate at a reactor outlet temperature of up to 950
C. In addition, the reactor system is intended to have a license period of 60 years. The combination of very high-temperature operation and long duration of service restricts material choices for the heat exchanger to a small number of coarse-grained solid-solution strengthened alloys that provide stability and creep resistance and have high chromium content for environmental resistance.

5.11.2 Experience with VHTR Systems Very early in the development of nuclear power for electricity generation or process heat, the concept of an inert gas-cooled, high-temperature reactor was explored.

Material Performance in Helium-Cooled Systems

253

Power conversion unit Intercooler

Generator

Turbine

Low-pressure compressor Primary heat rejection High-pressure compressor Recuperator

Commercial power

Blower Power for electrolysis

Pump

Heat exchanger Pebble-bed or prismatic reactor

Blower

Heat exchanger

Hydrogen production (electrolysis)

Hydrogen

Hydrogen production (thermochemical)

Hydrogen

Heat exchanger

Figure 1 An artist’s conception of a very high-temperature gas-cooled reactor and associated hydrogen production plants.

Table 1

Design characteristics of VHTRs that have been built and operated

Country of origin

Thermal power MWt Net electric power MWe Maximum core outlet temp (
C) Helium pressure MPa Steam temp (
C) Reactor type Vessel material Date of operation

Dragon

AVR

Peach bottom

Ft. St. Vrain

THTR-300

HTTR

OECD/Britain

Germany

USA

USA

Germany

Japan

21.5 – 750 2.0 – Sleeve Steel 1964–1975

46 13 950 1.1 505 Pebble Steel 1966

115 40 725 2.25 538 Sleeve Steel 1967

842 330 775 4.8 538 Block PCRVa 1979–1989

750 300 750 3.9 530 Pebble PCRV 1985

30 10 950 4 Prism Steel 1997

a Prestressed concrete reactor vessel. Source: Simon, R. A.; Capp, P. D. Operating experience with the dragon high temperature reactor experiment. In Proceedings of the Conference on High Temperature Reactors, Petten, NL, Apr 22–24, 2002; pp 1–6. Burnette, R. D.; Baldwin, N. L. Specialists Meeting on Coolant Chemistry, Plate-Out and Decontamination in Gas Cooled Reactors, Juelich, FRG, Dec 1980; International Atomic Energy Agency, 1980; pp 132–137. Shaw, E. N. History of the Dragon Project – Europe’s Nuclear Power Experiment; Pergamon: New York, 1983. Ba¨umer, B.; et al. AVR – Experimental High-Temperature Reactor; 21 Years of Successful Operation for a Future Energy Technology; Association of German Engineers (VDI), The Society for Energy Technologies: Du¨sseldorf, Germany, 1990. Baumer, R.; Kalinowski, I. Energy 1991, 16(1/2), 59–70. Brey, H. L. Energy 1991, 16(1/2), 47–58. Fuller, C. H. Design Requirements, Operation and Maintenance of Gas-Cooled Reactors, San Diego, CA, Sept 21–23, 1988; International Atomic Energy Agency, 1989; pp 55–61.

The Peach Bottom reactor in the United States and the European Dragon project were among the first to seriously address the technical issues associated with high-temperature environmental interaction between the cooling gas and metallic components.1–3 Proposals for a VHTR with an outlet temperature of 1000
C or above were put forward in the late 1970s.

The Arbeitsgemeinschaft Versuchsreaktor (AVR) was the first experimental pebble bed reactor. A commercial demonstration scale pebble bed, the Thorium Hochtemperatur Reaktor (THTR), was developed based on AVR experience. A summary of important design characteristics for gas-cooled VHTRs that have been operated to date are given in Table 1.1–7

254

Material Performance in Helium-Cooled Systems

The HTTR in Japan is the only one of the reactors listed in the table that is still in operation. The HTR-10 is not included in the table since there is no extensive operating experience with this reactor as yet. Operating experience with these reactors has shown that the primary helium coolant tends to contain H2O, H2, N2 as well as carbon-containing compounds CO, CO2, and CH4 at concentrations of a few parts per million. Impurities are introduced through adsorption on the fuel, leaks into the coolant, and lubricants from components like the helium circulators. In the reactors that are currently under consideration, the gas pressure is typically between 5 and 7 MPa. The coolant is circulated at high velocity, reaching velocities over 100 m s1 in some designs.

5.11.3 Comparison of IHX Concepts The IHX design for the VHTR will be influenced by a number of interrelated considerations, including the required separation distance between the reactor and the hydrogen production or other process heat plant, the heat losses from the intermediate loop piping, the operating pressure, the working fluid in the secondary loop, and the target efficiency of the hydrogen or process heat plant. The required separation distance will affect the intermediate loop piping size, the intermediate loop pumping requirements, and the piping heat losses to the environment. The intermediate loop pressure is critical; a low pressure will produce a high pressure differential between the primary and secondary sides of the IHX and high stress on the IHX. A high intermediate loop pressure will produce a high pressure differential across the intermediate loop pipe walls and within the hydrogen production or process heat equipment. Pressure drops within the IHX affect the pumping power requirements, which also depend on the intermediate loop working fluid, and the fluid temperature and pressure, and will have an effect on the overall VHTR cycle efficiency. The IHX may be arranged in parallel or in series with the VHTR power conversion system (PCS). In a serial arrangement, the total primary system flow (reactor outlet gas) passes through the IHX. The IHX receives gas of the highest possible temperature for delivery to the hydrogen production process (with slightly cooler gas going to the PCS), and must be large enough to handle the full primary flow. A parallel configuration splits the reactor outlet gas flow, with only a portion entering the IHX for the hydrogen or process heat plant, and the remainder of

primary flow going to a direct cycle power generation turbine. This results in the smallest possible IHX and the highest overall electrical power efficiency but lower process heat efficiency because of the cooler gas reaching that process. Specific IHX designs under consideration include countercurrent tube and shell, plate and fin, involute heat exchangers, microchannel heat exchangers, and the printed circuit heat exchanger (PCHE). The design has a significant influence on the required material properties. Tube-and-shell designs have the advantage of technological maturity, use heavy gauge materials, and are fabricated using conventional fusion welding methods. For the most simple tube-and-shell configuration, it has been estimated that 13 tons of high-temperature alloy is required per megawatt of heat transfer capability; helical designs can reduce this value to about 1.2 tons MW1. Compact heat exchanger designs have the potential for greater heat transfer efficiency; it is estimated that some of these designs will require only 0.2 tons of alloy per MW. The compact designs are much less technologically mature and increase the demands on material performance. Some compact designs have wall thicknesses of less that 1mm which places a premium on corrosion resistance and have significant stress concentrations that will lead to increased demand for creep resistance. In addition, several of these design concepts require diffusion bonding of multiple sheets of material or brazing in complex geometries. Neither of these joining methods has been used yet in nuclear applications, and nondestructive inspection methods have not been well developed. 5.11.3.1

Shell-and-Tube

A shell-and-tube heat exchanger is the most common type of heat exchanger. It consists of a number of tubes (often finned) placed inside a volume (shell). One of the fluids runs through the tubes while the second fluid runs across and along the tubes to be heated. In one variation of this concept, the heat transport fluid will flow on the shell side, allowing the tubes to contain the catalysts necessary for hydrogen production. In the simple configurations, the tube axis is parallel to that of the shell. The VHTR IHXproposed design features the tubes arranged in a helical configuration. This type of arrangement increases efficiency because of increased surface area and reduces the size, providing the potential to decrease the cost of materials. Tube-and-shell heat exchangers represent relatively mature technology

Material Performance in Helium-Cooled Systems

that has been widely commercialized in both nuclear and fossil energy systems. A helical design was extensively tested for the AVR reactor program and a similar system is in use in the HTTR in Japan. 5.11.3.2

Plate and Fin

The plate and fin heat exchanger (PFHE) transfers heat between two fluids by directing flow through baffles so that the fluids are separated by metal plates with very large surface areas. The fluids spread out over the plate, which facilitates the fastest possible transfer of heat. This design has a major advantage over a conventional heat exchanger because the size

of the heat exchanger is less for a comparable heat transfer capability. However, the candidate heat exchanger materials have relatively low thermal conductivities and will reduce the efficiency of a finned structure. Brazing is typically used to join the fins to the plate. Brazed plate heat exchangers are used in many industrial applications, although usually at low or even cryogenic temperatures. Although brazed products have been developed for high-temperature aerospace applications, the strength and creep properties of brazed joints in an IHX for a hightemperature reactor are of great concern. The unit cell heat exchanger is a typical modular plate-fin design that is being developed by Brayton Energy. An example is shown in Figure 2. Many of these individual unit cells would be grouped into larger heat exchanger assemblies. Integration of the modules within the vessel and with the interfacing piping is critical. Offset fin plate heat exchangers have very large heat transfer area density and effective countercurrent flow. 5.11.3.3

(a)

(b) Figure 2 Unit cell heat exchanger (a) primary side plate, (b) the unit cell showing countercurrent flow.

255

Etched Plate

Etched plate heat exchangers are diffusion-bonded, highly compact heat exchangers that can achieve a thermal effectiveness of over 98% in a single unit. Compact heat exchangers are four to six times smaller and lighter than conventional shell-and-tube heat exchangers of the equivalent heat transfer capability (Figure 3). The small size gives the compact diffusionbonded heat exchangers significant benefits over conventional heat exchangers across a range of industries. They are well established in the oil production, petrochemical, and refining industries. In addition, they are suitable for a range of corrosive and high-purity

Figure 3 The diffusion-bonded heat exchanger in the foreground undertakes the same thermal duty, at the same pressure drop, as the stack of three shell-and-tube exchangers behind.

256

Material Performance in Helium-Cooled Systems

Hot channel

Cold channel

Figure 4 Printed circuit heat exchanger configuration for the model.

streams and are particularly advantageous when space is limited and weight is critical. The most widely commercialized etched plate heat exchanger is a PCHE developed by Heatric Division of Meggitt (UK) Ltd. PCHE consists of metal plates on the surface of which millimeter-scale semicircular fluid-flow channels are photochemically milled, using a process analogous to that used for the manufacture of electronic printed circuit boards. The plates are then stacked and diffusion-bonded together to fabricate a heat exchanger core shown schematically in Figure 4. Heatric reports pressure capability in excess of 70 MPa and the ability to withstand temperatures ranging up to 900
C. Note that the channels are straight in this schematic, but in reality they have a zigzag configuration. Flow distributors can be integrated into plates or welded outside the core, depending on the design. The channel diameter, plate thickness, channel angles, and other attributes can be varied, so each PCHE is custom-built to fit a specified task. Channel dimensions are generally between 3 and 0.2 mm and the thickness of the web of material left after milling is typically less than 1 mm.8 The current fabrication limits are 1.5 m  0.6 m plates and 0.6 m stack height. The diffusion-bonded blocks made from several hundred individual sheets are modular and multiple blocks can be welded together to form larger units. 5.11.3.4

Microchannel Heat Exchangers

Microchannel heat exchangers, produced, for example, by Velocys, also feature a compact design similar to the etched plate design; however, the manufacturing process is somewhat different. They are constructed from diffusion-bonded corrugated sheets rather than

30⬚

Figure 5 Plate-stamped heat exchanger concept.

etched plates. The layers of corrugated sheet form many small-diameter channels that result in a high surface area/volume ratio and a high heat transfer coefficient. 5.11.3.5

Plate-Stamped Heat Exchanger

The plate-stamped heat exchanger (PSHE) concept consists of a set of modules, each being composed of a stacking of plates stamped with corrugated channels. The plates are stacked in such a way as to cross the channels of two consecutive plates and therefore to allow the different channels to communicate through the width of the plate as shown on the left in the figure. A general view of a plate is shown in Figure 5. Assembly of the plates into an IHX module is accomplished by welding only on the edges of the plates. No joining is performed in the active part of the plates, which gives the module relatively good flexibility. Therefore, this concept is thought to accommodate the thermal stresses better than the other concepts of plate IHXs. The location of the welded joints is also favorable to inspection, even if

Material Performance in Helium-Cooled Systems

257

this remains a difficult question. The joining processes which seem to be the most relevant are laser or electron beam welding due to the capability to perform narrow-gap joints and to avoid the overlapping of the welds of two consecutive plates. It should also be noted that the thickness of the PSHE plates is the largest among the metallic plate types IHX (1.5 mm), which means that it is the most favorable concept with respect to corrosion life. These reasons suggest that the PSHE concept may be the most promising among the plate IHXs.

reaches very high values, which increases the complexity of manufacturing, notably as assembly by narrow gap welding is required. Demonstration of the elements necessary for successful implementation of the technology is mainly based on technological feasibility tests like demonstration of individual tube to tube-plate welds by laser techniques. The results confirm the feasibility for limited thickness of the plate (a small mock-up is shown in Figure 7).

5.11.3.6

The development of IHXs made of ceramics is still at the research stage. Ceramic heat exchangers under development are either tubular or plate IHXs (mostly PFHE for the ceramic plate IHXs). Tube-and-shell heat exchangers based on SiC composite tubes have been developed for fossil energy applications for example. Joining the fiber-reinforced composite tubes to tube sheets and accommodating thermal expansion are the dominant technical challenges. Their resistance to aggressive environment is remarkable and they can operate at very high temperatures, >1000
C. Small monolithic compact designs have been developed from silicon carbide and silicon nitride through conventional ceramic forming and firing routes. In addition to technical issues, the cost of ceramic tubes of sufficient size for a VHTR IHX remains problematic. Table 2 provides a summary-level comparison of the significant attributes of the different IHX concept alternatives.

Foam IHX

The foam IHX concept is based on stacking plates separated by metallic foam. The barrier between the fluids is constituted by the separated plates and the fluids flow through the foam (see Figure 6). It is a new technology for heat exchanger application for which very high efficiency has been claimed. Several concerns have been identified regarding this type of IHX concept. The pressure losses induced by the foam are particularly high. Loss of small fragments of the foam is hardly avoidable and the geometry of the foam leads to an increased risk of clogging by graphite dust. 5.11.3.7

Capillary IHX

A concept with thread tubes between two tube-plates with external shell including bellows has been investigated. The diameter of the tubes is 2–3 mm. This kind of heat exchangers is currently being developed on an industrial scale. The small size of the tubes allows a sharp reduction in size and mass, but some difficulties arise at the same time, including the concern that the vibration risk is increased so that the supporting system needs to be very robust. The number of tubes

Figure 6 Foam heat exchanger concept.

5.11.3.8

Ceramic IHX

5.11.4 Heat Exchanger Alloys The desire for higher temperature operation resulted in the evolution of the materials under consideration, from stainless steels to iron-based high-temperature

258

Material Performance in Helium-Cooled Systems

Figure 7 Capillary heat exchanger mock-up.

Table 2

Comparison of IHX concept alternatives

PCHE PFHE PSHE Tubular IHX

Maturity

Stress behavior

Sensitivity to corrosion

Compactness

Numerous developments in conventional industry Numerous developments in conventional industry Numerous developments in conventional industry

High stress levels. 5 years lifetime seems very challenging High stress levels. 5 years lifetime seems very challenging Challenging but best stress accommodation among the plate IHXs Limit of state of the art

Sensitive

26 MW m3

Very sensitive

24 MW m3

Sensitive

35 MW m3

Better than plates but still sensitive Very sensitive (loss of fragments risk) Very sensitive

0.4 MW m3

Foam IHX

Industrial components in operation R&D

No results

Capillary IHX

Industrial developments

No results

Ceramic IHX

R&D

Difficult design because of fragile behavior

alloys to nickel-based alloys (see Chapter 2.08, Nickel Alloys: Properties and Characteristics). An extensive German program in the 1980s carried out exhaustive studies of the corrosion behavior of the iron-based Alloy 800H for control rods and nickel-based Alloy 617 for structural applications.9–12 The Japanese HTTR program extensively studied Alloy X and developed a variation known as XR with improved properties for some applications, while retaining Alloy 800H for the control rods.13 Compositions of these candidate alloys are given in Table 3.13–16 Based on creep resistance above 850
C, the leading candidate alloys for VHTRs are Alloy 617 and Alloy 230. A common characteristic of the alloys that have been put in service in high-temperature gas-cooled

Resistant

Comparable to other plate IHXs Better than classical tubular IHX Comparable to other plate IHXs

reactors is that they rely primarily on the formation of a tenacious chromia scale for long-term protection from environmental interaction with the gas-cooling environment.9,10,12,17 The alloys are also primarily solid-solution strengthened with carbides on the grain boundaries to stabilize the microstructure and enhance the creep resistance. Sustaining such a protective surface oxide requires sufficient oxygen partial pressure. The primary coolant gas of choice for VHTRs is helium. Although the helium is nominally pure and thus considered to be inert, there are inevitably impurities at the parts per million by volume (ppm) levels in the coolant in operating high-temperature reactors. Although at low levels, the impurities can significantly affect the performance of materials,

259

Material Performance in Helium-Cooled Systems

Table 3

Compositions of potential high-temperature alloys for VHTR (compositions in wt%)

Alloy

Ni

Fe

Cr

Co

Mo

Al

Alloy 617 UNS N06617 Alloy 230 UNS N06230 Alloy 800H UNS N08810 Alloy X UNS N06002

44.5

3

20–24

10–15

8–10

0.8–1.5

Bal

3

20–24

5

1–3

0.2–0.5

30–35

39.5

19–23

Bal

17–20

20.5–23

W

8–10

0.1

C

0.6

0.05–0.15 1

13–15

0.15–0.6 0.5–2.5

Ti

0.2–1

Si

Mn 1

0.05–0.15 0.25–0.75 0.3–1 0.15–0.6

0.05–0.1

0.03

0.05–0.15 <1

<1

Source: Incoloy Alloy 800H & 800HT, product sheet, Special Metals, 2004. Inconel 230, UNS N06230, product sheet, Special Metals, 2004. Inconel 617, UNS N06617, product sheet, Special Metals, 2005. Tanaka, R.; Kondo, T. Nucl. Technol. 1984, 66(1), 75–87.

depending on the chemistry of the particular alloy, the concentration of impurities, and the temperature at which the alloy can be oxidized, carburized, or decarburized. Several reviews of the behavior of metallic alloys for control rods, core internals, and heat exchangers in the reactor helium environment are available.9–12,17 Regardless of the IHX design, material selection for this component is critical. The material must be available in the appropriate product forms – both plate and sheet, weldable and suitable for use at 800
C or above. The majority of materials research and development programs in support of hightemperature gas reactors (HTGRs) were conducted from the 1960s to the early 1980s. The thrust of these programs was to develop a database on materials for application in steam-cycle and process-nuclear-heatbased HTGRs. Less work has been done on materials with emphasis on direct and/or indirect gas-turbinebased HTGRs. The available material property data were reviewed in detail, and an assessment of relevant factors was made including thermal expansion, thermal conductivity, tensile, creep, fatigue, creep– fatigue, and toughness properties for the candidate alloys. Thermal aging effects on the mechanical properties and performance of the alloys in helium containing a wide range of impurity concentrations are also considered.17 The assessment includes four primary candidate alloys for the IHX: Alloy 617, Alloy 230, Alloy 800H, and Alloy X. 5.11.4.1

Regulatory Issues

The IHX will form part of the pressure boundary for the VHTR and material selection and design will be subject to regulatory requirements. In the United States, the design will be guided by Section III of

the ASME Boiler and Pressure Vessel (B&PV) Code. This section specifies materials and design data for components in nuclear systems. Subsection NH of the Code, which specifies materials and design parameters for materials that will undergo inelastic deformation, includes only a very few materials. The temperature limits for Subsection NH Code materials, other than bolting, at 300 000 h are listed in Table 4. The maximum temperatures at which fatigue curves are provided are also listed. Note that of the materials under consideration for VHTR service, only Alloy 800H is currently contained in the appropriate section of the ASME Code and the service temperature is limited to 760
C. Any VHTR design that has an intended IHX service above this temperature will require extension of the Code to include additional materials, or at a minimum, extension of the use of Alloy 800H to higher temperatures. Other regulatory systems are in use internationally; however, it is generally true that additional materials and expanded databases are required before a new VHTR design can be finalized and licensed. 5.11.4.2

Alloy 617 (52Ni–22Cr–13Co–9Mo)

Alloy 617, also designated as Inconel 617, UNS N06617, or W. Nr. 2.4663a, was initially developed for high-temperature applications above 800
C. It is often considered for use in aircraft and land-based gas turbines, chemical manufacturing components, metallurgical processing facilities, and power generation structures. The alloy was also considered and investigated for the HTGR programs in the United States and Germany in the late 1970s and early 1980s. The high Ni and Cr contents provide the alloy with high resistance to a variety of reducing and oxidizing environments. In addition, the Al also

260

Material Performance in Helium-Cooled Systems

forms the intermetallic compound g0 -(Ni3Al) over a range of temperatures, which results in precipitation strengthening on top of the solid-solution strengthening imparted by the Co and Mo. Strengthening is also derived from M23C6, M6C, Ti(C, N), and other precipitates when in appropriate sizes, distributions, and volume fractions. Table 4 Materials specified in NH for elevated temperature service in nuclear applications NH code materials (other than bolting)

304 stainless steels (UNS S30400, S30409) 316 stainless steel (UNS S31600, S31609) Alloy 800H (UNS N08810) 2 1/4Cr 1Mo steel, annealed condition (UNS K21590) Grade 91 steel (UNS K90901)c

Maximum temperature (
C) For stress allowables S0, Smt, St, Sr up to 300 000 ha

For fatigue curves

816

704

816

704

760

760

593b

593

649

538

a

The primary stress limits are very low at 300 000 h and the maximum temperature limit. b Temperatures up to 649
C (1200
F) are allowed up to 1000 h. c The specifications for Grade 91 steel covered by subsection NH are SA-182 (forgings), SA-213 (small tube), SA-335 (small pipe), and SA-387 (plate). The forging size for SA-182 is not to exceed 4540 kg.

Table 5 Investigator

Observations and predictions of which precipitates form in Alloy 617 at given temperature ranges have not been consistent. A comprehensive review of the precipitates in Alloy 617 has been performed by Ren.18,19 Additional reviews can be found in Natesan et al.20 However, it is clear from the reviews that the kinetics of the precipitation and coarsening processes are important in determining the effects of aging on properties. The g0 intermetallic is generally too fine to be observed in optical microscopy. Other phases that have been identified include CrMo(C, N) and TiN,21 M12C and a possible Laves phase,22 and a Ni2(MoCr).23 A summary of observations is given in Table 5. The apparent trend is that in the temperature range of interest to the VHTR IHX, precipitates may form at initial exposure and the alloy may become stronger, but most of the precipitates will be dissolved after long-term exposure, and the alloy will depend on solid-solution strengthening in the long run. The most recent information on precipitation in Alloy 617 upon aging is shown in the T–T–T diagram in Figure 8. The influence of aging on the mechanical properties of the alloys under consideration for IHX applications is discussed in the section on environmental effects. The grain size also plays an important role in the strength of the alloy. For general applications, a grain size of 45 mm or coarser is typically preferred, but it has been shown that creep strength increases with increasing grain size, so microstructures of 100–200 mm grain size are often produced. A tradeoff exists, however, when fatigue is an issue, since finer grain sizes are preferred for fatigue resistance. In addition, for compact IHX, the thin sheet form

Prediction and observations of second-phase precipitates in Alloy 617 M23C6

M6C

g0

Stable Form wt < 5% persist to Thermocalc® prediction T  800
C T  780
C T ¼ 650
C (ORNL) Observation in material aged for 10 000 h and less T  1093
C Not observed Small wt% persist to Mankins21 and T ¼ 760
C Kimball24 1000
C 1000
C Not reported Kihara25 Observed Observed 550–1000
C Kirchhofer22 Observation in material aged for much longer than 10 000 h at 482–871
C Observed Observed Observed at 482, 538, & Wu23 also 593
C, not at 704
C for 43 000 h and observed longer, nor 870
C eta-MC after long time

m

Ti(C,N)

wt >10% 600  800
C

Not reported

Not observed

Not observed

Not observed Not observed

Not reported 400–1000
C

Not observed

TiN observed

Material Performance in Helium-Cooled Systems

261

1100

Temperature (⬚C)

1000

Limited γ⬘ formation

900 Ti (C, N) + M6C + M23C6 + γ⬘

800

Fine M4C + y Æ M23C6 + γ⬘ start

700 γ⬘ start extended γ⬘ end

600 500 Ti (C, N) + M4C 400 0.1

1

Ti (C, N) + M4C+ M20C4 10

100 Time (h)

1000

10 000

100 000

Ti(C, N) + Coarse M6C + M23C6 + γ⬘

Ti (C, N) + M4C4 + M4C + γ⬘5

Ti (C, N) + M4C + Coarse/Fine M6C + γ⬘

Ti (C, N) + M4C4 + M4C + γ⬘5

Ti (C, N) + M23C4 + M6 + γ⬘+ Ni2/(Mo,Cr) Figure 8 Time-temperature-transformation (T-T-T) diagram for precipitation of phases in Alloy 617 upon aging.

restricts development of large grain size. Whether the grains will significantly coarsen after the dissolution of certain grain boundary precipitates at long-term exposure is not clear. The existing mechanical property database for Alloy 617 is extensive (Table 6). This alloy has adequate creep strength at temperatures above 870
C, good cyclic oxidation and carburization resistance, and good weldability. It also has lower thermal expansion than most austenitic stainless steels and high thermal conductivity relative to the other candidates. It retains toughness after long-time exposure at elevated temperatures and does not form intermetallic or Laves phases that can cause embrittlement. Preliminary testing described later indicates that Alloy 617 has the best carburization resistance of the four alloys. During early development, Alloy 617 was systematically studied by Huntington Alloys, Inc. for applications in gas turbines, nitric acid production catalyst-grids, heat-treating baskets, Mo refinement reduction boats, etc. When Alloy 617 was considered for the HTGR, it was extensively investigated by Huntington, Oak Ridge National Laboratory (ORNL), and General Electric (GE). The Huntington data were used to develop ASME B&PV Code qualification, including the 1980s draft Code Case for the HTGR and applications covered by (nonnuclear) Section I and Section VIII Division 1. Alloy 617 is

not currently qualified for use in ASME Code Section III, although it is allowed in Section I and Section VIII, Division 1 (nonnuclear service). Efforts to gain the approval from the ASME Code committees for nuclear service were stopped when interest in VHTR technology waned in the 1990s. Both the ORNL-HTGR and GE-HTGR studies generated data from Alloy 617 that had been aged and/or tested in simulated HTGR helium. The helium impurities were the same as those considered for the VHTR system but the concentrations were different. Unfortunately, only processed data still exist; all original test curves needed for certain modeling efforts are irretrievable. Alloy 617 was also extensively investigated in Germany for its HTGR and other programs. The data generated were collected in the Online Data & Information Network (ODIN). Original test curves, if not all, are stored in the ODIN. However, the strain measurements of creep test curves were not all conducted with fine resolution and may not all be ideal for constitutive equation development. The aging effects on Alloy 617 are summarized in Ren and Swindeman.18 The development in modeling creep behavior of Alloy 617 is summarized in Swindeman et al.24 It is believed that creep–fatigue will be the most significant failure mechanism for materials in the IHX. Creep–fatigue damage results from cyclic loads superimposed on materials subjected to temperatures

262

Material Performance in Helium-Cooled Systems

Table 6

Summary of testing done on Alloy 617

Research organization

Number of heats

Number of samples

Test type

Temperature (
C)

Huntington alloys

13 þ 1 wire

ORNLa

4 plate þ 1 wire

GEa

1 plate þ 1 bar

Germany

Not specified

Honeywell aerospace Gen IV programb

Not specified Not specified

179 249 73 51 25 1 36 7 40 302 1947 29 261

Tensile Creep Tensile Creep Charpy Tensile after creep Creep Creep–fatigue Fatigue Tensile Creep Creep crack growth Low cycle fatigue Not specified Creep–fatigue

25–1093 593–1093 24–871 593–871 24 RT after 871 750–1100 950 850, 950 RT-1000 500–1000 700–1000 <500–1000

80

800, 1000

a

Some tests exposed to HTGR environment. Air, pure helium and vacuum environments.

b

5.11.4.3

Alloy 230 (57Ni–22Cr–14W–2Mo–La)

Alloy 230, also designated as Haynes 230, UNS N06230, or W. Nr. 2.4733, is a newer alloy than Alloy 617. In addition to outstanding resistance to oxidizing environments, Alloy 230 has good weldability and fabricability. It also has a lower thermal expansion coefficient than Alloy 617; it appears that thermal expansion has an inverse correlation with Ni content. Alloy 230 has a higher tensile strength than Alloy 617 up to 800
C, but above that the difference is

105

Cycles to failure (Nf)

and loads that will induce creep damage under monotonic loading. Recent creep–fatigue data for Alloy 617 and Alloy 230 are shown in Figures 9 and 10 for tests that involved fully reversed cyclic loading at total strain ranges of 0.3% and 1% with varying hold time during the tensile portion of the cycle at 800 and 1000
C. The plots show the reduction in cycles to failure with increasing tensile hold time under creep loading conditions. It can be seen that in general, increased hold time results in decreased cycles to failure. At 800
C, the two alloys have similar behavior; however, at 1000
C, Alloy 617 appears to have somewhat higher cycles to failure compared to Alloy 230. A limited number of tests have been carried out on specimens that contain weldments and it has been found that the cycles to failure in specimens containing a fusion weld is reduced. In these specimens, the cracking is typically in the weld metal and not in the heat-affected zone or at the weld–base metal interface.

Alloy 617 – 0.3% total strain Alloy 617 – 1.0% total strain Alloy 230 – 0.3% total strain Alloy 230 – 1.0% total strain Alloy 617CCA – 0.3% total strain Alloy 617CCA – 1.0% total strain

104

1000

No hold 100 10

100

1000

Hold time (s) Figure 9 Effect of hold time on cycles to failure in creep–fatigue at 800
C.

insignificant. It appears that Alloy 617 has slightly better creep properties than Alloy 230. Alloy 230 has a better thermal fatigue crack initiation resistance but a worse thermal cycling resistance compared to Alloy 617. The Ni base and high Cr content impart resistance to high-temperature corrosion in various environments, and oxidation resistance is further enhanced by the microaddition of the rare earth element La. Compared to Alloy 617, Alloy 230 has a high W concentration which replaces much of the Co in

Material Performance in Helium-Cooled Systems

105

Cycles to failure (Nf)

No hold

Alloy 617– 0.3% total strain Alloy 617–1.0% total strain Alloy 617CCA – 0.3% total strain Alloy 617CCA – 1.0% total strain Alloy 230 – 0.3% total strain Alloy 230 – 1.0% total strain

104

103

102 10

100 1000 Hold time (s)

10 000

Figure 10 Effect of hold time on cycles to failure in creep–fatigue at 1000
C.

Alloy 617. The W and Mo in conjunction with C are largely responsible for the strength of the alloy, and its relatively high B content in comparison to that in Alloy 617 can be controlled to achieve optimized creep resistance. Usually, B acts as an electron donor; it can affect the grain boundary energy and help improve ductility. In Ni-based alloys, B also segregates to grain boundaries and helps to slow grain boundary diffusion, thus reducing the creep process. On the other hand, excess boron in a neutron field could also lead to embrittlement due to transmuted He, although irradiation is not a factor for IHX applications. In the solution-annealed condition in which this alloy is typically supplied, the grain size is typically >45 mm with large carbide precipitates rich in W, presumably of the M6C type. After aging, Alloy 230 typically exhibits M6C and M23C6 precipitates. After aging for 1000 h at 850
C, very small carbide precipitates rich in Cr and M23C6 were observed along the grain boundaries. No grain coarsening was observed.25 Creep strength is believed to be brought about by solid-solution strengthening, low stacking fault energy, and precipitation of M23C6 carbides on glide dislocations.26,27 However, a negative impact of M23C6 on room temperature ductility was also reported. After aging at 871
C for 8000 h, the room temperature tensile elongation of Alloy 230 decreased from 50% to 35%, with a precipitation of M23C6 observed in microstructural examination, but an additional 8000 h of aging did not further decrease

263

ductility.26 Significant microstructural changes were also observed after thermal aging in air for 10 000 h at temperatures ranging from 750 to 1050
C. After the 750
C aging, coarser intergranular precipitation of M23C6 and coarse and blocky intra- and intergranular precipitates of M6C were observed. After the 850–1050
C aging, the M6C carbides were irregular in shape. After aging at 1050
C, the secondary intragranular M23C6 appeared to have dissolved. A decrease in toughness and ductility coincided with the appearance of the intragranular M23C6 and reached a minimum after the aging at 850
C. The toughness and ductility recovered after the aging at 1050
C.28 There is less characterization of Alloy 230 compared to Alloy 617. The major known large-scale study was tensile and creep tests by Haynes International. Creep times ranged from 15.3 to 28 391 h. Like Alloy 617, Alloy 230 is not currently qualified for use in ASME Code Section III, although it is allowed in Section VIII, Division 1 (for nonnuclear service). At present, the database for Alloy 230 is significantly smaller than that for Alloy 617 and a much larger effort is required to develop an Alloy 230 Code Case for elevated temperature application. Some recent data on environmental effects of exposure to prototypical VHTR chemistries are given in the following sections and creep–fatigue properties are included in Figures 9 and 10. 5.11.4.4

Alloy 800H (42Fe–33Ni–21Cr)

This alloy is the only iron-based alloy under consideration, although it has a solid-solution strengthened austenitic structure like the other three alloys. Upon aging, precipitates can form and somewhat reduce the tensile and creep ductility. Alloy 800H has the lowest creep rupture strength and the lowest resistance to oxidation of the four alloys. There is an additional variant of this alloy, 800HT, that has a composition similar to that of 800H, but has an additional specification for coarse grain size. The majority of material that is currently available in this alloy series is Alloy 800HT, which also meets the specification for Alloy 800H. Among the four candidate materials, Alloy 800H is the only one that is Code qualified for use in nuclear systems, but only for temperatures up to 760
C and a maximum service time of 300 000 h. Alloy 800H was the primary high-temperature alloy used in the German HTGR programs and an enormous amount of data were obtained. However, only very limited data from the German HTGR programs

264

Material Performance in Helium-Cooled Systems

are currently available on the mechanical properties of this alloy beyond 800
C, especially in impure helium environments.

5.11.5 Welding All of the solid-solution alloys that have been mentioned are readily welded using conventional fusion welding methods. Alloys 617 and 230 are described in more detail later as prototypical of these materials. Alloy 617 has excellent weldability. Alloy 617 filler metal is used for gas-tungsten-arc (GTAW) and gasmetal-arc welding (GMAW). The composition of the filler metal matches that of the base metal, and deposited weld metal is comparable to the wrought alloy in strength and corrosion resistance.29 Alloy 230 is also readily welded by GTAW and GMAW. Shielded metal-arc welding (SMAW) and resistance welding techniques can also be used. Submerged-arc welding

Stress (MPa)

Alloy X has the best oxidation resistance of the four alloys, although its carburization resistance is the worst. Above 700
C, Alloy X can form embrittling phases that result in property degradation. The creep rupture strength is not as good as Alloy 617 or 230. The limitations of this alloy will be similar to the draft code case for Alloy 617 in terms of grain size, product form, and limitations on service time. A limited database exists for Alloy X for conditions typical of a VHTR, but the high-temperature scaling in Hastelloy X has been less than optimal. As a result, a modified version, Alloy XR, has been developed in Japan; however, the United States has little access to Alloy XR material, either for evaluation or for ASME Code qualification. Japanese are currently using Alloy XR in a heat exchanger in the HTTR at temperatures of 850–950
C. The material is codified in Japan for nuclear use, which would likely accelerate code acceptance in ASME. An extensive environmental database and HTGR experience exist. However, the database may be limited to large grain material, similar to the Alloy 617 draft code case. Also, similar to the Alloy 617 draft code case, Alloy XR may have issues with weldments that need to be addressed. It is uncertain if this alloy is readily available as a commercial product. Figures 11–14 compare the creep rupture strength, oxidation behavior, carburization behavior, and allowable stress for the four alloys, respectively.

Alloy X

60

Alloy 230

50

Alloy 617

40 30 20 10 0 1

1000 100 Time to rupture (h)

10

10 000

100 000

Figure 11 Creep rupture strength at 962
C in air.

Thickness (mm)

Alloy X (47Ni–22Cr–9Mo–18Fe)

Alloy 800H

70

20 10 0 −10 −20 −30 −40 −50 −60 −70 −80 −90

Alloy 800H

Alloy 617

Oxide scale

Alloy 230

Affected zone

Alloy X Internal oxide

Figure 12 Schematic representation of isothermal oxidation behavior after 800 h exposure at 950
C in helium environment.

24 Normalized weight gain

5.11.4.5

80

Alloy X 19

Alloy 230 Alloy 617

14 9 4 −1

0

200

400

600

800

1000

1200

Exposure time (h) Figure 13 Mass change as a function of time in H2–5.5% CH4–4.5% CO2 carburizing environment at 1000
C.

is not recommended, as this process is characterized by high heat input to the base metal and slow cooling of the weld. These factors can increase weld restraint and promote cracking. The as-welded properties of these alloys are given in Table 7.30 The welds exhibit room temperature strength that matches or is slightly

Material Performance in Helium-Cooled Systems

265

better than the base metal, but a considerable decrease in ductility is observed at elevated temperatures, as shown in Table 8.

alloy, HASTELLOY S alloy, or HASTELLOY W alloy welding products may all be considered, depending upon the particular case.29,30

5.11.5.1 Base Metal Preparation and Filler Metal Selection

5.11.5.2 Preheating, Interpass Temperatures, and Postweld Heat Treatment

Prior to any welding operation, the welding surface and adjacent regions should be thoroughly cleaned with an appropriate solvent. All greases, oils, corrosion products, and other foreign matter should be completely removed. It is preferable, but not necessary, that the alloy be in the solution-annealed condition when welded.30 Alloys 617 and 230-W™ (AWS A5.14, ERNiCrWMo-1) filler wire are recommended for joining Alloy 617 and 230, respectively, by GTAW or GMAW. The filler metals are not specifically designed for nuclear application. For dissimilar metal joining of Alloy 230 to nickel-, cobalt-, or iron-based materials, 230-W filler wire, Alloy 556™

Preheat is not required, generally room temperature (typical shop conditions) is adequate. Interpass temperature should be maintained below 93
C. Auxiliary cooling methods may be used between weld passes, as needed, providing that such methods do not introduce contaminants. Postweld heat treatment is not generally required either. Table 9 shows the nominal welding parameters based on welding conditions used in the Haynes International laboratories and should serve as a guide for performing typical GTAW and GMAW operations on Alloy 230. All processes used 230-W filler wire.30 5.11.5.3

Maximum allowable stress (MPa)

250 Alloy X Alloy 230

200

Alloy 617 Alloy 800H 150 100

50

0 0

200

400

600

800

1000

Temperature ( ⬚C) Figure 14 Allowable stress for heat exchanger materials for plate, sheet, and strip forms from the ASME boiler and pressure vessel code Section VIII.

Table 7

Nontraditional Joining Methods

As noted earlier in the description of IHX designs, several of the compact heat exchanger design concepts will require the joining of sheet product to be either diffusion bonding or brazing. Diffusion bonding of these alloys is relatively well developed because of applications in aerospace systems that require this fabrication method. The etch plate compact design fabricated from austenitic stainless steel has been commercialized for petrochemical applications, and limited diffusion bonding studies have been completed using Alloy 617. Characterization of diffusion-bonded stacks of sheet indicates that mechanical properties comparable to base metal can be achieved at room temperature. The details of diffusion bonding parameters are considered proprietary by the IHX vendors, and it is not clear whether temperatures sufficiently high to cause carbide dissolution and/or grain growth is a matter of concern.

Room-temperature tensile properties of joints in as-welded condition

Alloy

Specimen

Yield strength (0.2% offset) (MPa)

Tensile strength (MPa)

Elongation (%)

Reduction of area (%)

61729

GMAWa GTAWb GMAWc

510 542 490

761 823 785

43.3 37.3 48.2

42.0 38.3

23030 a

Alloy Filler Metal 617. Average of ten tests. Alloy Filler Metal 617. Average of 17 tests. Alloy 230-W filler wire.

b c

266

Material Performance in Helium-Cooled Systems

Table 8

Tensile properties of 230 base and weld metals 23
C

GMAW deposit weld metal Cold-rolled and 1232
C solution annealed (sheet) Hot-rolled and 1232
C solution annealed (plate) Vacuum investment castings (as-cast)

538
C

871
C

UTS

YS

EL

UTS

YS

EL

UTS

YS

EL

785 838 840 615

490 422 375 325

48.2 47.2 47.7 37.8

610 699 690 450

435 303 251 230

34.8 53.7 54.6 38.2

310 308 315 285

275 234 242 185

45.4 75.0 99.5 19.0

Source: HAYNESW 230W Alloy. Haynes International, Inc. Publication H-3000H, 2004.

Table 9

Weld parameters for Alloy 230

Welding method

GMAW

Configuration (mm)

Thickness > 2.3 1.1 dia. wire

Technique

Stringer bead or slight weave 100–130a 18–21 4.3–4.8 12.7–19.1 203–356 Torch, 50 Ar-25% He

Current (A) Voltage (V) Feed rate (m min1) Stick-out (mm) Travel speed (mm min1) Gas flow (l min1) Gas

GTAW Auto

Manual

Square butt joints 1.0/1.6/3.2 thick, 1.6 electrode with 45
included shape No filler metal added 50/80/120b 8.0/8.5/9.5

V or U groove, >3.6 thick, 3.6 dia. wire, 3.6 electrode with 30
included shape Stringer bead interpass T < 100
C 120 root, 140–150 fillb 11–14

10/12/12 Shield, 14.2 backing, 4.7 Argon

102–152 Shield, 14.2–16.5 backup, 4.7 Argon

a

DCEP, torch flow CFPH. DCEN. Source: HaynesW 230W Alloy, Haynes International High-Temperature Alloys. b

Very little information on brazing these alloys is available. A general concern is that low melting point braze materials could result in poor elevated temperature properties in structures fabricated by these methods.

5.11.6 Control Rod Materials The pebble bed modular reactor (PBMR) is the most complete recent design for a VHTR. The reactor was designed to operate at about 400 MWt and primarily to produce electricity. Recent changes in the global economic climate have caused reconsideration of the design for a VHTR in South Africa; however, the analysis that went into the design and selection of materials for the control rods is illustrative of the most recent analysis of these issues. A schematic of the PBMR core is shown in Figure 15. The design outlet gas temperature for the PBMR was 900
C; the core was designed to be 11m high and 3.7 m in

diameter, and the annulus filled with about 452 000 60-mm-diameter fuel pebbles.31 The PBMR builds on the German experience of the AVR and THTR; however, it will use a direct cycle to produce power rather than a steam generator, and it will have an annular core configuration with a solid graphite central reflector. The annular core produces several advantages: it shifts the peak power radially outward, thus enabling significantly higher output; it enhances the fuel safety margin; and, by increasing the neutron flux in the outer graphite reflector, it increases the effectiveness of the control and shutdown systems.32 The reactor control and shutdown system (RCSS) has two components: the reactivity control system (RCS) and the reserve shutdown system (RSS). The RCS consists of 12 control rods and 12 shutdown rods, located in the outer reflector.33 They are evenly spaced around the core and at a radial distance of about 70mm from the inner surface of the reflector (see Figure 16).34 During normal operation, the control rods, which

Material Performance in Helium-Cooled Systems

Reactor pressure Core barrel Top Side

Cold gas riser Center Pebble bed Bottom Inlet Inl Hot gas Outlet

Figure 15 Schematic of the pebble bed modular reactor (PBMR) annular pebble bed reactor. Reproduced from Kriel, W. Material selection: High-temperature metallic materials. Slides, Sept 21–22, 2005.

Core barrel

267

penetrate a maximum distance of 1.5m into the core,34 are used for minor reactivity adjustments to keep the reactor critical, provide reactivity compensation for xenon poisoning effects during load following effects,35 and allow for some excess reactivity so that the reactor may continue operation for some time if no fuel is being loaded.36 They are also used for hot shutdown purposes.33 The reactor power is actually adjusted by regulating the mass flow rate of the gas inside the primary circuit rather than by adjusting the control rods.32,33 During scram, the additional 12 shutdown rods are lowered to the bottom of the active core. In the event of a loss of electrical power, insertion of the rods is by gravity. The first set of control rods will drop, and later the shutdown rods will drop, should the need arise. The RSS consists of eight storage containers of 10-mm-diameter small absorber spheres containing B4C that can be fed by gravity into eight channels in the central reflector. The RSS serves as both the secondary shutdown system and the cold shutdown system. It must be activated in addition to the RCSS to bring the PBMR to a cold shutdown condition (100
C). The control rod design is similar to that of previous metal control rods. A schematic is shown in Figure 17. A number of annular B4C rings are encased between two tubes of Alloy 800H, to form a section about a meter long. One unique feature is that

Reactor pressure vessel

Side reflector barrel Reactor inlet pipes Annular core

Reactor outlet pipe

Center reflector

Gas riser channels

Small absorber sphere channels Control rod

Figure 16 Top view of the PBMR core, showing the location of the control rod channels, fuel pebbles, small absorber sphere channels, and other features. Reproduced from PBMR. Data and boundary conditions to be used in VSOP, TINTE, and MCNP PBMR 400 MW ( Th) reactor models.

268

Material Performance in Helium-Cooled Systems

Control rod drive mechanism

RCS chain

Control rod segment

Control rod link

Secondary shock absorber

Figure 17 Schematic of the control rod assembly in the PBMR. Reproduced from Broom, N.; Smit, K. PBMR Design Methodology. Slides, Oak Ridge, TN, 12th April 2005.

the inner tube is much thinner than the outer tube. These sections are mechanically linked to form an articulated control rod several meters long. One difference from past designs (e.g., the AVR) is that the control rod is suspended from the drive mechanism by a chain, rather than a cable. A secondary shock absorber is in place in the channel below the control rod to protect it and the core structure in the event of a chain failure. Additional shock absorbers within the drive mechanism dampen the impact load on the control rod drives during scram.33 A control rod guide tube (not shown in Figure 16) connects the control rod drive mechanism to the core structure to guide the control rod into the core.37 During normal operation, the temperature of the control rods is estimated to be from about 650 to 700
C,38,39 and the temperature resulting from a depressurized loss of coolant (DLOC) event is estimated to be only 850
C.39 The end-of-life fast fluence is reported as 2  1022 (E > 0.1 MeV),38 and the thermal

fluence is reported as 5  1021 n cm2.39 The secondary shock absorbers have an operating temperature of 900
C; during DLOC, they can be subjected to temperatures of up to 1100
C for short periods. Under these conditions, the use of Alloy 800H was justified by the PBMR program because the high-temperature strength and creep resistance are sufficiently qualified for long-term normal operation at 700
C, and limited operation above 850
C under abnormal events can be tolerated according to available data. The response of Alloy 800H to neutron irradiation at the temperatures expected in the control rod sleeves is not well characterized. The PBMR design notes that the irradiation response has been characterized to high levels of fast fluence at lower temperatures and Alloy 800H control rods have had extensive qualification and service in previous German VHTR programs. Limited data from older VHTR programs on irradiation effects in Alloy 800H at temperatures above 600
C suggest that helium embrittlement from (n, a) reactions associated with thermal neutrons is the predominant degradation mechanism. Recent work has examined property changes associated with irradiation to 1.45 dpa at temperatures of 580 and 660
C.40,41 Significant strengthening was observed along with a sharp decrease in ductility. Material irradiated at 660
C and subsequently tensile tested at 700
C showed tensile elongation of less than 0.5%. The mechanisms of embrittlement are not yet completely clear and irradiation experiments to higher fluence at higher temperature will likely be required to examine this issue. The PBMR design includes specialized equipment to remove and replace the control rods, as well as storage for used control rods. The RCSS will be inspected every 6 years during the scheduled maintenance outage, and repaired as necessary. These outages are planned to last 30–50 days, depending on the other maintenance scheduled, with the exception of a 180-day shutdown after 24 years to replace the core reflector.34,42 Eventually, it is hoped that a VHTR similar in design to the PBMR reactor can run with an outlet temperature of 1000
C or even higher. In this case, carbon fiber-reinforced carbon composites (Cf/C) or silicon carbide fiber-reinforced silicon carbide composites (SiCf/SiC) must be considered for the more challenging temperatures of the control rods.39 Experience with irradiation of SiCf/SiC composites for nuclear fusion applications suggests that these materials have superior resistance to property degradation from neutron irradiation as well as resistance to higher temperatures and could potentially have

Material Performance in Helium-Cooled Systems

lifetimes comparable to the life of the plant. As with ceramic IHXs, application of these advanced materials in a nuclear system would require considerable further development and cooperation with appropriate standards and regulatory organizations.

5.11.7 Core Barrel Materials Another evidently important metallic internal structure shown in Figures 15 and 16 is the core barrel. The function of the core barrel is to mechanically contain the shape of the graphite blocks making up the core and to channel the flow of the primary coolant. Although schematics in Figures 15 and 16 are specific to a pebble bed design, the core barrel is essentially identical in design and function for a prismatic design as well. The temperatures, neutron fluence, and mechanical loads on the core barrel are moderate and the PBMR design, for example, proposed the use of Type 316 stainless steel for this application. Alloy 800H is also a leading candidate for the core barrel material. Although the demands on the material are modest in terms of mechanical loading and neutron irradiation, it is a very large structure (8 m high by 3.5 m diameter and 50 mm in thickness) that will need to be fabricated on site by welding in most cases. Both Type 316 stainless steel and Alloy 800H are available in the required size and are readily fabricated.

5.11.8 Environmental Effects of VHTR Atmospheres on Materials All the high-temperature reactor systems operated to date had extensive gas cleanup systems associated with the helium coolant flow. These systems are intended to keep the total impurity levels in the helium below typically 10 ppm. Particularly in the early reactors, where the fuel was either not intended to contain the fission products or was ineffective in this function, the cleanup systems were also intended to capture radionuclides.1,3 Capture of tritium that is produced (at least in part) by transmutation of lithium impurities in the graphite remains an important function of the cleanup system. In the AVR and THTR reactors, active control was maintained on the H2O and CO concentrations to reduce oxidation of the graphite reflectors, and the other impurities were routinely found to reach acceptable steady-state levels without active control.4–7,43,44 It has been noted that the

269

cleanup systems may play a secondary role in maintaining gas chemistry, with the massive amount of graphite at high temperature present in all of the reactor designs playing a dominant role.1 Molecular sieves are effective in capturing most of the gaseous impurities; however, they have difficulty capturing H2 and CO. To resolve this problem, the gas stream is passed over a bed of CuO that oxidizes the H2 to H2O and CO to CO2 upstream of the molecular sieve where these gases are effectively removed. The Peach Bottom plant attempted the use of heated Ti getters for hydrogen and tritium; however, these were not effective and oxidation of the H2 prior to removal is now accepted practice.2 In a typical plant, up to about 20% of the gas stream is diverted to the cleanup system each hour. Table 10 shows the impurity levels reported for steady-state operation for several of the VHTRs.1,2,43,44 As shown in the table, at steady state, all of the reactors for which operating data are available had similar levels of impurities. Some caution should be exercised when comparing the data for different plants, since, in some cases, there are varying values reported in different publications for the same plant. This may be associated with conversion from partial pressure of impurities (the preferred units for corrosion studies) to ppm by volume (the typical units used for comparison of one plant to another). Several plants have undergone extensive postmortem analysis of the core internals and heat exchangers.1,2 There are reports of some oxidation and at least one report of massive deposition of carbon on the internals, as discussed in more detail in the following paragraphs; however, there have been no problems with failure of components on the primary side associated with environmental effects. There have been a large number of experimental studies and modeling of the effect of VHTR helium on the high-temperature alloys listed in Table 3.10–12,16,17,45–54 Depending on the specific proposed application, different model chemistries have been developed, and the testing has focused on these. Several of the model impurity chemistries are shown in Table 11.45–52,54 Comparison of the values in Table 11 with actual operating experience suggests that the model chemistries tend to have higher impurity levels of some species than those found in operating reactors. This is notable for H2 in particular. It is not clear why these particular values were chosen; however, it can be noted that several of the proposed applications were for process heat for coal gasification, and there was concern that hydrogen would diffuse from the process plant into the primary

270

Material Performance in Helium-Cooled Systems

Table 10

Impurities reported in the helium coolant during steady-state operation of VHTRs (in ppm)

Dragon Peach bottom Fort St. Vrain AVR THTR

H2O

H2

CO

CO2

CH4

O2

N2

0.1 0.5 1 0.15 <0.01

0.1 10 7 9 0.8

0.05 0.5 3 45 0.4

0.02 <0.05 1 0.25 0.2

0.1 1.0 0.1 1 0.1

0.1 – –

0.05 0.5 – 22 0.1

Source: Simon, R. A.; Capp, P. D. Operating experience with the dragon high temperature reactor experiment. In Proceedings of the Conference on High Temperature Reactors, Petten, NL, Apr 22–24, 2002; pp 1–6. Burnette, R. D.; Baldwin, N. L. Specialists Meeting on Coolant Chemistry, Plate-Out and Decontamination in Gas Cooled Reactors, Juelich, FRG, Dec 1980; International Atomic Energy Agency, 1980; pp 132–137. Nieder, R. Specialists Meeting on Coolant Chemistry, Plate-Out and Decontamination in Gas Cooled Reactors, Juelich, FRG, Dec 1980; International Atomic Energy Agency, 1980; pp 144–152. Nieder, R.; Stroter, W. VGB Kraftwerstech. 1988, 68, 671–676.

Table 11 Model impurity chemistries used in environmental testing programs (compositions in ppm). HHT, PNP were used for German nuclear process heat projects

HHT PNP AGCNR JAERI B

H2O

H2

CO

0.75 0.75 1 0.5

250 250 200 100

20 7 20 50

CO2

CH4

0.1 1

25 10 10 2.5

O2

N2 5 <2.5 <2.5 <2.5

AGCNR was a German VHTR, and JAERI B composition was extensively studied in development of the HTTR. Source: Nieder, R. Specialists Meeting on Coolant Chemistry, Plate-Out and Decontamination in Gas Cooled Reactors, Juelich, FRG, Dec 1980; International Atomic Energy Agency, 1980; pp 144–152. Nieder, R.; Stroter, W. VGB Kraftwerstech. 1988, 68, 671–676. Bates, H. G. A. Nucl. Technol. 1984, 66(2), 415–428. Brenner, K. G. E.; Graham, L. W. Nucl. Technol. 1984, 66(2), 404–414. Christ, H. J.; et al. Mater. Sci. Eng. 1987, 87, 161–168. Christ, H. J.; et al. Oxid. Metals 1988, 30, 27–51. Christ, H. J.; et al. Oxid. Metals 1988, 30, 1–26. Fujioka, J.; et al. Nucl. Technol. 1984, 66(1), 175–185. Inouye, H. Nucl. Technol. 1984, 66, 392–403.

coolant circuit.45 Note that N2, at concentrations similar to those listed in Table 10, has never been found to contribute significantly to environmental interactions with nickel-based alloys.48,49 Interplay between the alloy surface, temperature, and gas composition determines whether corrosive oxidation, carburization, or decarburization occur. The corrosion mechanisms of particular significance to mechanical stability are carburization and decarburization. Carburization is associated with lowtemperature embrittlement, and decarburization is linked to reduced creep rupture strength. Ideally, a continuous self-healing, impermeable passivating oxide layer is needed to establish the most corrosion-resistant alloy. In the case of Alloy 617, the chromia layer (Cr2O3) is the most important barrier to the effects of corrosive reactor gases. As noted earlier, of the existing materials, Alloy 617 is the leading candidate for use in the VHTR

heat exchangers because it has the highest creep strength of the solid-solution alloys under consideration for temperatures above 850
C. Evaluation of this alloy for VHTRs began in the early 1980s, with the most comprehensive work done by Brenner and Graham,46 Christ et al.,47–49 Graham,51 and Quadakkers and Schuster.54 Alloy 230 is under consideration as an alternative to Alloy 617 because it has equivalent creep properties and may suffer from less internal oxidation. Based upon the work of Quadakkers and others, assessments of Alloy 617 stability at various gas concentrations and temperatures can be displayed graphically. Quadakkers used a diagram of the type shown in Figure 18 to display the results of the stability calculations for the nickel–chromium alloy.47–49,54 Five conditions are represented within the diagram: I – strong reduction (decarburization without a surface oxide); II – decarburization (with

Material Performance in Helium-Cooled Systems  3COþ2Cr ¼ Cr2 O3 þ3C PCO CO pressure for this equilibrium

CrnCm

log ac

III I

II

½III

The measured steady-state conditions (AGCNR gas composition) determine where the alloy sits within the modified chromium diagram; hence, the steadystate carbon activity (acss ) and the steady-state oxygen pressure (POss2 ) are tentatively calculated with the following equations (provided the kinetics of methane splitting is low compared to the water vapor dissociation on the alloy surface):

Cr2O3

IV V

271

P*CO

Cr metal

ss1=2

log Po2 Figure 18 Modified chromium stability diagram at a given temperature; Zone III is preferred for optimal chromia layer protection against corrosion.

a porous oxide); III – stable external oxide (with stable internal carbides); IV – mixed surface oxide and carbide layers (with internal carburization); and V – strong internal and external carburization. Zone III was determined to be the area of highest stability; an environment that is oxidizing and slightly carburizing. For a heat exchanger operating at temperatures of 900
C and assuming standard gas composition, called AGCNR helium and shown in Table 11, Alloy 617 would be in zone III, for example. The diagram describing alloy stability was based on the most relevant species involved in the corrosion process, namely chromium. Identifying which form – Cr2O3 chromium carbide or chromium metal – is most stable in a particular environment will determine the ultimate fate of the alloy.55 It is important to note that the gas chemistries found in operating reactors and used in the previous test programs are not in thermodynamic equilibrium. A steady-state gas composition is reached at any temperature based on kinetic considerations. As is shown later, the concentrations of H2O and CO largely determine the partial pressure of oxygen and the activity of carbon, respectively. The important features of the diagram are critical carbon activation (ac ), critical partial pressure of oxygen (PO 2 ), and the critical partial pressure of  carbon monoxide (PCO ). At a given temperature, these parameters are calculated from the following thermodynamic reactions: ac

23Cr þ 6C ¼ Cr23 C6 metal  carbide equilibrium activity

½I

PO 2

Cr2 O3 ¼ 2Cr þ 1:5O2 disassociation pressure of chromia

½II

ss =PO2 Þ acss CO ¼ Cþ0:5O2 acss ¼ K ðPCO

½IV

POss2 H2 O¼H2 þ½O POss2 ¼ðKPH2 O =PH2 Þ2 ð1ðPCH4 =PH2 O Þ1=100Þ

½V

A chromium activity of 0.75 was assumed at the Alloy 617 surface.48,49 At very high temperatures, there is a critical temperature above which the oxide layer is unstable and CO evolution will occur. This critical temperature represents the maximum application temperature for the alloy. The results of an experiment for Alloy 230 showing oxide instability and CO evolution are given in Figure 19 for a specimen that was held at a constant temperature of 900
C to establish a surface oxide before rapidly increasing the temperature to 1000
C. The chemical reaction that gives rise to CO evolution is 2Cr2 O3 þ Cr23 C6 ¼ 6CO þ 27Cr

½VI

It is also significant to note the rapid rate at which the reaction occurs, indicating that significant changes in surface condition of this alloy can occur in a few hours. The reaction [VI] (also known as the microclimate reaction) suggests that degradation of the protective oxide on the surface of the IHX alloys under consideration can be suppressed by increasing the partial pressure of CO. Figure 20 shows the upper temperature for oxide stability as a function of CO partial pressure for both Alloy 617 and Alloy 230. The onset of reaction [VI] occurs at a particular temperature, TA, when the CO concentration is no longer sufficient to drive the reaction from right to left. It results in eventual total loss of either chromium oxide or carbide, depending on concentration, then in complete carburization or decarburization of the alloy, depending on the gas composition. This degradation mechanism is considered in the stability diagrams developed using the approach of Quadakkers  , and it has been discussed extensively by as PCO Brenner and Graham,46 Christ et al.,47–49 and Graham.51

Material Performance in Helium-Cooled Systems

40

1000

Partial pressure (mbar)

35

T(°C)

800

CO

30 PCO(mbar) PCH4(mbar)

25

600

P(CO)inlet

20

400

15

Temperature (°C)

272

CH4 200

10

5

0

5⫻104

1⫻105

0

2⫻105

1.5⫻105

Time (s) Figure 19 Experimental determination of oxide instability (as evidenced by CO evolution) for Alloy 230 in He with 200 ppm H2, 21 ppm CO, 19 ppm CH4, and 0.5 ppm H2O.

1275 1250 Fit a(Cr) = 0.72 TA In K

1225 1200 1175 Alloy 230 Alloy 617 Alloy 617 after(20)

1150 1125 0

1

2

3 P(CO) ln Pa

4

5

6

Figure 20 The critical temperature for oxide instability for Alloy 617 and Alloy 230 as a function of CO partial pressure.

At the temperature TA, the reaction will go to completion, and as a result, this temperature represents a maximum use temperature for the particular alloy for a given gas composition. Experimental results for this reaction are shown for Alloys 617 and 230. A shortcoming of the stability diagram approach presented earlier is that it does not account for the fact that Cr2O3 becomes volatile at a temperature above about 950
C.56 There may be sufficient oxygen partial pressure to form the oxide as predicted from the modified stability diagram; however, it will

not be protective because of the vapor pressure of the oxide. Most laboratory studies have been at very low flow rates to more closely approach thermodynamic equilibrium for fundamental studies of corrosion mechanisms. It has been noted that these conditions may not be representative of reactor systems, where very high gas velocities are likely, for example, 75–100 m s1 at the outlet of the VHTR.48,49 With very low levels of impurities, this increases the possibility that impurities will be depleted during the experiments and may give rise

Material Performance in Helium-Cooled Systems

to anomalously low values for Cr2O3 vaporization in test systems compared to reactor operation.46–49,51,52 Micrographs of cross-sections from Alloy 617 and Alloy 230 plate material after exposure to heliumcontaining impurities that resulted in decarburizing atmosphere at 1000
C are shown in Figure 21(a) and 21(b), respectively. The microstructures are largely as anticipated. Alloy 617 shows a relatively thick chromium oxide scale with significant formation of grain boundary aluminum oxides. Alloy 230 shows less surface oxidation and notably reduced tendency for formation of grain boundary oxides. A decarburization region is also apparent for both alloys; the decarburized region is particularly notable in Alloy 230 shown in Figure 21. Internal oxidation may be of particular concern in alloy selection for compact heat exchangers where very thin material sections may be encountered. Carburization of Alloy 617 has been examined in previous work and the results shown in Figure 22 for this alloy after exposure to carburizing conditions at 1000
C are largely consistent with behavior reported in the literature. For the conditions examined, here there is little formation of a surface oxide scale. Carbon uptake in the material is manifested by increased grain boundary carbide precipitate volume fraction. The behavior of Alloy 230 is markedly different from that of Alloy 617. While there is little evidence of scale formation, there is a very large volume fraction of carbide formation in the alloy that heavily decorates both grain and twin boundaries. This large volume fraction of carbides was found through the entire thickness of the 3 mm thick coupon after 500 h at 1000
C. A micrograph from the center of an Alloy 230 coupon is shown in Figure 23. To investigate the effect of environmental interaction on the mechanical properties of the heat exchanger alloys, Alloy 617 specimens that were carburized at 900 (1000 h) and 1000
C (500 h) as well as companion specimens that were oxidized at 900
C (1000 h) were tested. Representative room temperature stress–strain curves for the materials exposed at 900
C are shown in Figure 24. It is clear from the figure that the heavily carburized specimen (IN 617-2) has higher flow strength compared to the oxidized specimen (IN 617-7), but considerably reduced ductility. Room temperature tensile results for the material carburized at 1000
C are essentially identical to those for material carburized at 900
C. Tensile testing the carburized Alloy 617 at 800
C showed increased reduction in area to about 4%; the oxidized

273

(a)

(b)

50.00 µm

Figure 21 Optical micrographs of cross-sections through (a) Alloy 617 and (b) Alloy 230 after 500 h at 1000
C under decarburizing conditions.

material had greater than 50% reduction in area in tension at 800
C. Tensile tests of Alloy 230 with extensive carburization also indicated nil ductility at room temperature and less than 1% tensile elongation at 800
C.

5.11.9 Aging Effects In addition to decreased ductility from carburization, aging will cause precipitation of carbides and other phases in the temperature range of interest for heat exchanger applications. The alloys with higher aluminum, such as Alloy 617, have a significant volume fraction of Ni3Al (g0 ) formed at some aging temperatures. This phase increases the strength and results in reduced ductility and impact properties. A T–T–T diagram for Alloy 617 illustrating the region of

274

(a)

Material Performance in Helium-Cooled Systems

50.00 mm

50.00 µm

Figure 23 Optical micrograph of carbides in the center of the Alloy 230 coupon after carburizing exposure at 1000
C.

(b)

50.00 mm

Figure 22 Optical micrographs of cross-sections through (a) Alloy 617 and (b) Alloy 230 after 1000 h at 1000
C under carburizing conditions.

stability of the g0 is shown in Figure 8. The most rapid precipitation of additional phases occurs at a temperature of 750
C, which is somewhat below the expected maximum operating temperature of the IHX. The effect of thermal aging on impact properties of Alloy 617 is shown in Figure 25 for an aging time of 1000 h. Absorption of energy from impact testing drops considerably for these aging treatments compared to the solution-annealed material. The room temperature tensile ductility of Alloy 800H has been found to be essentially unchanged by aging in air at 800
C for 30 000 h. In contrast, Alloy X showed a drop in room temperature tensile elongation from 45% to 10% for similar aging treatment. While the ductility of Alloys 617 and X decreases significantly after aging conditions, both retain substantial ductility. Addition of degradation in properties from carburization in the VHTR atmosphere could, of course, be a cause for further concern.

Stress (Mpa)

% strain vs stress (MPa) 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0

IN617-7 IN617-2

0

5

10

15 % strain

20

25

30

Figure 24 Room temperature tensile stress–strain curves for Alloy 617 that has been oxidized (IN617-7) and carburized (IN617-2) at 900
C for 1000 h.

Microstructures of Alloys X and 800H after 500 h exposure to oxidizing conditions illustrate the differences in precipitation between the two alloys upon aging. The grain size of heat-treated Alloy X is 35 mm and precipitation within the grains is heavy as shown in Figure 26. Several carbide populations coexist: fine dark carbides which form stringers at twin boundaries and round grayish carbides are within the grains; large round white carbides (Mo rich, M6C) are formed at grain boundaries. Alloy 800H microstructure shown in Figure 27 after oxidizing exposure at 750
C exhibits large and regular grains and lesser amounts of quite

Material Performance in Helium-Cooled Systems

275

Impact toughness (J) at RT

350 617 unaged 617 1000 h aged

300 250 200 150 100 50 //

0 500

600

700 800 900 Aging temperature (°C)

1000

Figure 25 Room temperature impact energy for Alloy 617 after aging at various temperatures for 1000 h.

homogeneous precipitation compared to Alloy X, except for some stringers of carbides along the rolling direction. The more pronounced precipitation in Alloy X compared to Alloy 800H and the resulting decrease in ductility for Alloy X would suggest that Alloy 800H might be the preferred alloy for use in this temperature range. Note, however, that the oxidation resistance of Alloy X is higher compared to Alloy 800X in terms of external as well as internal oxidation. For compact heat exchanger designs where thin sections are a concern, the limited internal oxidation might favor selection of Alloy X. Limited internal oxidation is a particularly attractive attribute if it can be shown that further reduction of ductility due to carburization is not a significant probability. Although aged materials retain ductility at operating temperatures, embrittlement from aging or environmental effects is an issue in some circumstances. Reduced ductility and impact properties of any of the alloys are of concern during startup from ambient temperature after an outage, for example, because stresses from thermal gradients may be large in this circumstance.

5.11.10 Summary Although designs for very high-temperature nuclear reactors are continuing, enough is known from system requirements and past experience to make reasonable material choices for the most highly challenged systems, the heat exchanger and control rod sleeves. Due to the high anticipated service temperature and the requirement for minimal change in

100 µm Figure 26 Microstructure of Alloy X after 672 h at 750
C under oxidizing conditions.

100 µm Figure 27 Microstructure of Alloy 800H after 672 h exposure at 750
C under oxidizing conditions.

material behavior for service lives up to 60 years, the material choices are limited to face center cubic solid-solution alloys. Compact heat exchanger designs are especially demanding due to small section sizes. The number of candidate alloys is further constrained by the necessity of selecting materials that are sufficiently mature to be acceptable to regulators for licensing the plant. The leading candidate alloy for IHX application at outlet temperatures in excess of 850
C is Alloy 617. Alloys 230, 800H, X, and XR are also viable candidates for some applications. Only Alloy 800H is currently in the US nuclear design code and for temperatures limited to 760
C. Additional material property characterization will be required to allow application of Alloy 800H at higher temperatures and to design with the other candidate

276

Material Performance in Helium-Cooled Systems

alloys. The need for additional data is apparent for design under any regulator around the world. Characterization of the creep and creep–fatigue properties for long times and under prototypical VHTR atmospheres will be required. For compact heat exchanger designs, joining methods that have not typically been used in nuclear service, including diffusion bonding and brazing, will need to be extensively studied and the properties of resulting structures characterized. Development of nondestructive inspection methods for the complex IHX geometries envisioned for VHTR systems is an urgent need. Environmental resistance of the candidate alloys has been extensively studied and is generally understood. Coolant chemistry has to remain in an acceptable domain which favors passive oxidation over carburization or decarburization which affects the alloy microstructure and properties. Additional work remains to validate the understanding of environmental resistance and aging behavior of the materials, particularly at the highest temperatures. The temperature and stress to which control rod sleeves will be subjected under normal operating conditions and for different accident scenarios have not yet been fully understood. The best current thinking is that control rod sleeves made of Alloy 800H will have sufficient properties to operate safely. Irradiation-induced changes in mechanical properties of this alloy are not understood, and additional testing may also be necessary to determine if the lifetime of the sleeves will be adequate. Further development and testing of composite control rod sleeves, either carbon fiber or silicon carbide fiber, is desirable to demonstrate operation at higher temperatures and for increasing the service life.

7.

8.

9.

10. 11. 12. 13. 14. 15. 16. 17. 18.

19. 20.

21. 22. 23.

References 1.

Simon, R. A.; Capp, P. D. Operating experience with the dragon high temperature reactor experiment. In Proceedings of the Conference on High Temperature Reactors, Petten, NL, Apr 22–24, 2002; pp 1–6. 2. Burnette, R. D.; Baldwin, N. L. Specialists Meeting on Coolant Chemistry, Plate-Out and Decontamination in Gas Cooled Reactors, Juelich, FRG, Dec 1980; International Atomic Energy Agency: Vienna, Austria, 1980; pp 132–137. 3. Shaw, E. N. History of the Dragon Project – Europe’s Nuclear Power Experiment; Pergamon: New York, NY, 1983. 4. Ba¨umer, R.; Barnert, H.; Baust, E. AVR: Experimental High Temperature Reactor, 21 Years of Successful Operation for a Future Energy Technology; VDI-Verlag GmbH: Du¨sseldorf, 1990. 5. Baumer, R.; Kalinowski, I. Energy 1991, 16(1/2), 59–70. 6. Brey, H. L. Energy 1991, 16(1/2), 47–58.

24.

25.

26. 27. 28. 29. 30.

Fuller, C. H. Design Requirements, Operation and Maintenance of Gas-Cooled Reactors, San Diego, CA, Sept 21–23, 1988; International Atomic Energy Agency, 1989; pp 55–61. Pua, L. M.; Rumbold, S. O. Presented at the First International Conference on Microchannels and Minichannels; Apr 24–25, 2003. Rochester, New York, USA. Bodmann, E.; et al. High Temperature Metallic Materials for Gas-Cooled Reactors, Cracow, Poland, June 20–23, 1988; International Atomic Energy Agency, 1988; pp 14–26. Nickel, H.; Schubert, F. Nucl. Technol. 1984, 66(3), 649–660. Nickel, H.; Schubert, F.; Breitling, H.; Bodmann, E. Nucl. Eng. Des. 1990, 121(2), 183–192. Nickel, H.; Schubert, F.; Schuster, H. Nucl. Eng. Des. 1984, 78, 251–265. Incoloy Alloy 800H & 800HT, product sheet, Special Metals, 2004. Inconel 230, UNS N06230, product sheet, Special Metals, 2004. Inconel 617, UNS N06617, product sheet, Special Metals, 2005. Tanaka, R.; Kondo, T. Nucl. Technol. 1984, 66(1), 75–87. Natesan, K.; et al. Materials Behavior in HTGR Environments; ANL-02/37 NUREG/CR-6824; Argonne National Laboratory, Feb 2003. Ren, W.; Swindeman, R. W. A review of aging effects in Alloy 617 for Gen IV nuclear reactor applications. In Proceedings of the 2006 ASME Pressure Vessels and Piping Division Conference, Vancouver, BC, July 23–27, 2006. Ren, W.; Swindeman, M. J. Development of a Controlled Material Specification for Alloy 617 for Nuclear Applications; ORNL/TM-2005/504; ORNL, May 30, 2005. Natesan, K.; Moisseytsev, A.; Majumdar, S.; Shankar, P. S. Preliminary Issues Associated with the Next Generation Nuclear Plant Intermediate Heat Exchanger Design; ANL/EXT-06-46; Argonne National Laboratory, Sept 2006. Mankins, W. L.; Lamb, S. ASM Handbook (formerly Metals Handbook), 10th ed.; ASM International: Materials Park, OH, 1990. Kirchhofer, H.; Schubert, F.; Nickel, H. Nucl. Technol. 1984, 66, 139–148. Wu, Q.; Vasudevan, V. K. Characterization of Boiler Materials for Ultracritical Coal Power Plants; Annual Progress Report for Period August 1, 2002 to July 30, 2003; Under UT-Battelle Sub Contract Number 4000017043; Jan 28, 2004. Swindeman, R. W.; Swinderman, M. J.; Ren, W. A brief review of models representing creep of Alloy 617. In Proceedings of the 2005 ASME Pressure Vessels and Piping Conference, Denver, CO, July 17–21, 2005. Se´ran, J. L.; Cabet, C. J.; et al. Metallic and graphite materials for out-of-core and in-core components of the VHTR: First results of the CEA R&D Program. Beijing, China, Sept 22–24, 2004. Klarstrom, D. L. In Materials Design Approaches and Experiences; Zhao, J. C., et al., Eds.; TMS: Warrendale, PA, 2001; pp 297–307. Tawancy, H. M. J. Mater. Sci. 1992, 27, 6481–6489. Jordan, C. E.; et al. In Long Term Stability of High Temperature Materials; Fuchs, G. E., et al., Eds.; TMS: Warrendale, PA, 1999; pp 55–67. High Temp Metals Inc. Inconel 617 Technical Data. Haynes International High-Temperature Alloys HaynesW 230W Alloy.

Material Performance in Helium-Cooled Systems 31. Ion, S.; Powlowski, J. P.; et al. Pebble Bed Modular Reactor the First Generation IV Reactor to Be Constructed. 32. IAEA (International Atomic Energy Agency) Current Status and Future Development of Modular High Temperature Gas Cooled Reactor Technology; IAEA-TECDOC-1198; February, 2001. 33. Matzner, D. In Proceedings of the Conference on High Temperature Reactors HTR-2004, Sept 22–24, 2004; International Atomic Energy Agency, 2004; pp 1–26. 34. Gougar, H. Personal communication with J. K. Wright, 2006. 35. Matzner, D. Proceedings of the 2nd International Topical Meeting on High Temperature Reactor Technology, Beijing, China, Sept 22–24, 2004; International Atomic Energy Agency, 2004; pp 1–13. 36. PBMR. PBMR Control Rod Design and Requirements. Nuclear Engineering 1–19. 37. IST Nuclear. The Pebble Bed Modular Reactor – A Solution for the 21st Century. 38. Kriel, W. Material selection: High temperature metallic materials; Slides, Sept 21–22, 2005. 39. Fazluddin, S.; et al. The use of advanced materials in VHTR’s. In 2nd International Topical Meeting on High Temperature Reactor Technology, Beijing, China, Sept 22–24, 2004. 40. Gan, J.; Cole, J. I.; Allen, T. R.; Shutthanandan, S.; Thevuthasan, S. J. Nucl. Mater. 2006, 351(1–3), 223–227. 41. Nanstad, R. K.; McClintock, D. A.; Hoelzer, D. T.; Tan, L.; Allen, T. R. J. Nucl. Mater. 2009, 392(2), 331–340. 42. Slabber, J. PBMR (Pty) Ltd., Technical Description of the PBMR Power Plant; 016956; Feb 14 2006.

43.

44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

55. 56. 57. 58.

277

Nieder, R. Specialists Meeting on Coolant Chemistry, Plate-Out and Decontamination in Gas Cooled Reactors, Juelich, FRG, Dec 1980; International Atomic Energy Agency, 1980; pp 144–152. Nieder, R.; Stroter, W. VGB Kraftwerstech. 1988, 68, 671–676. Bates, H. G. A. Nucl. Technol. 1984, 66(2), 415–428. Brenner, K. G. E.; Graham, L. W. Nucl. Technol. 1984, 66(2), 404–414. Christ, H. J.; Kunecke, U.; Meyer, K.; Sockel, H. G. Mater. Sci. Eng. 1987, 87, 161–168. Christ, H. J.; Kunecke, U.; Meyer, K.; Sockel, H. G. Oxid. Metals 1988, 30, 27–51. Christ, H. J.; Schwanke, D.; Uihlein, T.; Sockel, H. G. Oxid. Metals 1988, 30, 1–26. Fujioka, Junzo; Fukasako, Norio; Murase, Hirokazu; Nishiyama, Yukio Nucl. Technol. 1984, 66(1), 175–185. Graham, L. W. J. Nucl. Mater. 1977, 171, 155–178. Inouye, H. Nucl. Technol. 1984, 66, 392–403. Nickel, H.; Bodmann, E.; Seehafer, H. J. Energy 1991, 16(1/2), 221–242. Quadakkers, W. J.; Schuster, H. High Temperature Metallic Materials for Gas-Cooled Reactors, Cracow, Poland, June 20–23, 1988; International Atomic Energy Agency, 1988; pp 73–78. Roine, A. Hsc Chemistry for Windows, Version 5.1. Grabke, H. J.; Horz, G. Annu. Rev. Mater. Sci. 1977, 7, 155–178. PBMR. Data and boundary conditions to be used in VSOP, TINTE and MCNP PBMR 400 MW (Th) reactor models. Broom, N.; Smit, K. PBMR Design Methodology. Slides, Oak Ridge, TN, April 12, 2005.

5.12

Material Performance in Supercritical Water

T. R. Allen, Y. Chen, X. Ren, K. Sridharan, and L. Tan University of Wisconsin, Madison, WI, USA

G. S. Was and E. West University of Michigan, Ann Arbor, MI, USA

D. Guzonas Atomic Energy of Canada Limited (AECL), Deep River, ON, Canada

ß 2012 Elsevier Ltd. All rights reserved.

5.12.1

Introduction

280

5.12.1.1 5.12.2 5.12.2.1 5.12.2.1.1 5.12.2.1.2 5.12.2.1.3 5.12.2.1.4 5.12.2.1.5 5.12.2.2 5.12.2.2.1 5.12.2.2.2 5.12.2.2.3 5.12.2.3 5.12.2.3.1 5.12.2.3.2 5.12.2.4 5.12.3 5.12.3.1 5.12.3.1.1 5.12.3.1.2 5.12.3.1.3 5.12.3.1.4 5.12.3.1.5 5.12.3.2 5.12.3.2.1 5.12.3.2.2 5.12.3.2.3 5.12.3.2.4 5.12.3.3 5.12.3.4 5.12.3.5 References

Experimental Issues Corrosion in SCW Effect of Alloy Class Ferritic–martensitic steels Austenitic stainless steels Ni-based alloys Zr-based alloys Ti-based alloys Oxide Structures Ferritic–martensitic steels Austenitic steels Ni-based alloys Surface Composition Modification Effect of grain size refinement Effect of grain boundary structure optimization Summary of Corrosion in SCW Stress Corrosion Cracking in SCW Austenitic Stainless Steels Temperature Microstructure Water chemistry Pressure/dielectric constant Irradiation Nickel-Based Alloys Alloy Temperature Water chemistry Irradiation Ferritic–Martensitic Alloys Other Alloys Summary of SCC Behavior in SCW

283 286 286 288 292 294 295 296 296 296 296 298 300 301 302 305 305 309 309 311 311 313 313 315 315 317 318 319 320 322 322 322

Abbreviations AECL AOA AR ASTM

Atomic Energy of Canada Limited Axial offset anomaly As received condition American Society for Testing and Materials

BWR CANDU CERT CGR CIPS CLSBs

Boiling water reactor Canada Deuterium Uranium Constant extension rate tensile Crack growth rate Crud-induced power shifts Coincidence site lattice boundaries

279

280

Material Performance in Supercritical Water

CLT DCPD EBSD ECAP ECP EDX or EDS F/M FFTF GBE GenIV IGSCC JPCA LWR ODS OT PHWR PWR RBMK RHAB SAM SCC SCW SCWO SCWR SEM TEM TGSCC XPS XRD

Constant load tensile DC potential drop Electron backscatter diffraction Equal channel angular processing Electrochemical corrosion potential Energy dispersive X-ray spectroscopy Ferritic-martensitic steels Fast Flux Test Facility Grain boundary engineering Generation IV Intergranular stress corrosion cracking Japanese Prime Candidate Alloy Light water reactor Oxide dispersion strengthened Oxygenated treatment Pressurized heavy water reactor Pressurized water reactor Reactor Bolshoy Moshchnosty Kanalny Random high angle boundaries Scanning auger microscopy Stress corrosion cracking Supercritical water Supercritical water oxidation Supercritical water-cooled reactor Scanning electron microscopy Transmission electron microscopy Transgranular stress corrosion cracking X-ray photoelectron spectroscopy X-ray diffraction

5.12.1 Introduction The Generation IV (Gen IV) advanced reactor initiative is pushing nuclear reactor technology to completely new performance requirements, as a result of





higher temperatures and process efficiency longer lifetimes and higher neutron fluxes challenges in materials technologies novel technologies introduced for reactor fuel and fuel cycles

One promising Gen IV concept is the supercritical water-cooled reactor (SCWR). The idea of using a supercritical water (SCW) coolant in a water-cooled reactor dates back to the 1960s,1,2 although no such reactor was ever built. More recently, two types of

SCWR concept have evolved from existing light water reactor (LWR) and pressurized heavy water reactor (PHWR) designs: (1) a number of designs3–5 consisting of a large reactor pressure vessel containing the reactor core (fueled) heat source, analogous to conventional pressurized water reactor (PWR) and boiling water reactor (BWR) designs, and (2) designs with distributed pressure tubes or channels containing fuel bundles, analogous to conventional CANDU (CANDU®, Canada Deuterium Uranium, is a registered trademark of Atomic Energy of Canada Limited (AECL)) and RBMK (Reactor Bolshoy Moshchnosty Kanalny) nuclear reactors.6 The balance-of-plant is typically a direct-cycle design, and the out-of-core portions of both concepts are similar to existing fossilfired generators. The SCWR will have core outlet temperatures well above the thermodynamic critical point of water (374  C, 22.1 MPa); the reference design for the SCWR7,8 calls for an operating pressure of 25 MPa and an outlet water temperature of up to 620  C. Peak fuel cladding temperatures could be as high as 850  C in some designs (e.g., Chow and Khartabil9). Figure 1 illustrates the proposed operating conditions of an SCWR, as well as the operating ranges of existing reactor designs, fossil-fired SCW power plants, and supercritical water oxidation (SCWO) processes. Operation in the SCW regime gives the SCWR many advantages compared to the state-of-the-art LWRs and PHWRs, including the use of a single-phase coolant with high enthalpy, a direct, once-through steam cycle that enables the elimination of components, such as steam generators and steam separators and dryers, a low coolant mass inventory resulting in smaller components, a much higher efficiency (45% vs. 33% in current LWRs), and nearly 50 years of industrial experience from thermal-power stations with a SCW cycle. Besides the design concept itself, the most important technical issues are likely to be the identification of materials for in-core and out-of-core components and the identification of appropriate coolant chemistry. As noted, there is significant industry experience with the use of SCW in nonnuclear power generation,10,11 with about 268 944 MWe (462 units) of installed capacity in coal-fired SCW power plants worldwide11 as of 2004. However, a nuclear reactor core is significantly different from a fossil-fired boiler; Figure 2 illustrates the scale differences between core components in an SCWR and the corresponding components in a fossilfired SCW power plant. A fossil-fired boiler contains a large number of relatively thick-walled (6–12 mm thickness) fire tubes that circulate water on the inside.

Material Performance in Supercritical Water

281

70 60 Supercritical

Pressure (MPa)

50 Liquid

40

SCWO region SCFP region

30 SCWR core

20 PWR core

10 0 100

Critical point

CANDU core BWR core

200

300

Possible peak cladding temperature

Steam

400 500 600 Temperature (⬚C)

700

800

900

Figure 1 The temperature–pressure phase diagram of water showing the critical point and the supercritical regime. The operational regions of present boiling water reactor and pressurized water reactor plants, as well as supercritical fossil-fired boilers, are presented along with the design area of supercritical water-cooled reactors. The large area covered by supercritical water oxidation processes is fully in the supercritical water region. Adapted from Heikinheimo, L.; Guzonas, D.; Fazio, C. GENIV materials and chemistry research – Common issues with the SCWR concept. 4th International Symposium on Supercritical Water-Cooled Reactors, Heidelberg, Germany, Mar 8–11, 2009.

The wall thickness for the fuel cladding in the reference SCWR design is 0.63 mm (but may be as low as 0.4 mm12) and the wall thickness for the water rods is 0.40 mm. These very thin-walled components provide little margin for corrosion in an SCWR core, where the consequences of failure are significant. Oxide films of several hundred micrometer thickness are not unusual for fossil-fired plant boiler tubes, but are unacceptable for an SCWR water rod or fuel cladding. In addition to oxide films formed on fuel cladding surfaces by corrosion of the base metal, the deposition of corrosion products released from corrosion of outof-core components onto fuel cladding surfaces could result in (1) overheating of the cladding surface or underdeposit corrosion, leading to fuel failures, (2) changes in reactivity in the core (crud-induced power shifts (CIPS), also known as axial offset anomaly (AOA)), and (3) increased radiation fields on out-ofcore piping. It is important to note that, with respect to corrosion product deposition, corrosion rates of out-of-core materials that may be acceptable for fossil-fired SCW plants may not be acceptable for an SCWR.13 In addition to the operating experience and research in support of fossil-fired SCW plants, a large amount of data on materials degradation in SCW was acquired in the development of SCWO processes. While the chemistry conditions in these

tests are generally not of direct relevance to an SCWR, typically being acidic with high concentrations of aggressive species such as chloride, these data do provide some insights into the key parameters affecting corrosion phenomena in SCW.14 However, in spite of information available from current reactor designs, modern boiler technologies, and research in support of SCWO, significant gaps still exist in our understanding of the corrosion properties of materials under proposed SCWR operating conditions. There is a strong interplay between coolant chemistry and materials selection in any water-cooled nuclear power plant system. As noted, the SCW coolant in both the pressure vessel and pressure tube concepts lies in both the liquid and supercritical fluid areas of the T–P phase diagram (Figure 1). The coolant passes through the critical point at some location in the reactor core. The corrosivity of SCW varies widely depending upon the values of properties, such as density, ion product, and dielectric constant, as well as on the nature of any solutes present (impurities, dissolved oxygen) and their concentrations.14 At the low density (100 kg/m3) expected at the core outlet of an SCWR, SCW is a nonpolar solvent capable of dissolving gases such as oxygen to complete miscibility. While the solubility of ionic species is expected to be extremely low under these conditions, the formation of neutral complexes increases with temperature

282

Material Performance in Supercritical Water

Column 2 Column 5

Column 9

Column 13 Row 13 Coolant Fuel rod/clad Water duct

(5 6) Water rod Assembly duct

Row 1 (a)

Central instrumentation rod

04-GA50011-02

Fuel rod outer diameter = 12 mm Clad wall thickness = 0.4 – 0.6 mm

Fire tube outer diameter = 50 mm Tube wall thickness =10 mm (b)

Water rod outer diameter = 40 mm (square) Water rod wall thickness = 0.4 mm

Figure 2 (a) A 1/8 assembly model of the 21  21 supercritical water-cooled reactor fuel assembly and (b) comparison of typical fossil boiler tube dimensions to fuel rod and water rod dimensions in the reference supercritical water-cooled reactor design. Part (a) reproduced from Feasibility Study of Supercritical Light Water Cooled Reactors for Electric Power Production; Final Report, DE-FG07-02SF22533, INEEL/EXT-04-02539, Jan 2005.

and can become important under near-critical and supercritical conditions. It has been suggested that the most important temperature region is from 300 to 450  C; over this temperature range, the properties of water change markedly, and solvent compressibility effects exert a huge influence on solvation. With the exception of a few recent studies (see Wesolowski et al.15), the thermochemistry of neutral hydrolyzed metal species is poorly understood, even at temperatures well below the critical point. In current water-cooled reactor designs, degradation of system components is minimized by selecting and then controlling a set of chemistry parameters that together reduce the aggressiveness of the coolant

to the specific alloys used in the system. Chemistry performance requirements are set by the sometimes conflicting desires to minimize corrosion (general and localized), minimize corrosion product and activity transport, optimize thermal performance, and maximize system lifetime. The primary requirement of chemistry control is to reduce degradation rates such that design lifetimes are achievable for the entire system. Although fossil-fired SCW plants have used a variety of water chemistries,16,17 to date, most experimental work on SCWR materials has been carried out using a limited subset of these water chemistries, namely, low-conductivity, neutral-pH water, low-conductivity, neutral-pH water with added

Material Performance in Supercritical Water

oxygen (50–8000 ppb) (this encompasses the oxygenated treatment (OT) commonly used in fossil-fired SCW plants, in which a low concentration of oxygen (50–150 ppm) is added to the boiler feedwater to reduce corrosion of mainly carbon steel components), and hydrogen water chemistry (H2 concentration 30 cm3 kg1 water). Additional testing under a wider range of water chemistries, developed based on a deeper understanding of SCW chemistry, may prove beneficial. In addition to the water chemistry issues discussed earlier, which, with the exception of activity transport, are common to both nuclear and fossil-fired SCW power plants, reactor in-core components must also contend with the effects of irradiation on water chemistry and alloy microstructure. Water radiolysis can increase the concentrations of oxygen and other oxidizing species (e.g., OH, H2O2, and HO2/O 2 ), raising the corrosion potential and increasing susceptibility to processes such as stress corrosion cracking (SCC). While current PWRs and PHWRs limit the formation of oxidizing species in-core by ensuring the presence of excess hydrogen at concentrations sufficient to chemically lower the net production of oxidizing species by radiolysis, the existing data are insufficient to determine whether this strategy would be effective in an SCWR. Coupled with the high solubility of oxygen in SCW, uncontrolled radiolysis could lead to very oxidizing coolant conditions in-core and immediately downstream of the core. Water chemistry experiments at very high temperatures and pressures, especially beyond the critical point of water, are difficult to perform, and advances in this area will require a combination of computer simulations and experiments. Water in a reactor core is subject not only to extreme conditions of high temperature and pressure, but also to an intense flux of ionizing radiations (energetic neutrons, g-rays, recoil protons, and heavy ions), making its radiolysis difficult to determine experimentally. Currently, the limited experimental data available on SCW radiolysis (see, e.g., Bartels et al.,18 Meesungnoen et al.,19 Pommeret,20 Katsumura21) have shown that virtually no free radical reaction rates follow an Arrhenius law. As a result, rate constants for key reactions cannot be extrapolated from data measured at subcritical temperatures and must be measured. Once sufficient data exist to predict the corrosion potential, laboratory experiments can be carried out to more accurately simulate environmental conditions expected for the core of an SCWR. Ultimately, model predictions and out-of-core test data will need to be validated against experimental data obtained

283

from in-reactor test loops.22 As direct measurement of the electrochemical corrosion potential (ECP) of a test specimen in an in-core SCW loop is not possible with existing technologies, such testing will also require development of a reference electrode that can withstand the SCW environment. For in-core components, perhaps the most challenging problem is the role of irradiation on microstructure and how these changes affect SCC. Irradiation-assisted SCC has been a generic problem in LWRs of all types and covering many austenitic and nickel-based alloys.23,24 This chapter reviews the current understanding of the response of candidate materials for SCWR systems, focusing on the corrosion and SCC response, and highlights the design trade-offs associated with certain alloy systems. With the exception of the effect of irradiation on SCC, the important issues of radiation response and radiolysis are not addressed in this review. For an overview of corrosion and SCC in light water-cooled reactors, see Chapter 5.03, Corrosion of Zirconium Alloys; Chapter 5.04, Corrosion and Stress Corrosion Cracking of Ni-Base Alloys; Chapter 5.05, Corrosion and Stress Corrosion Cracking of Austenitic Stainless Steels; and Chapter 5.08, Irradiation Assisted Stress Corrosion Cracking. 5.12.1.1

Experimental Issues

Performing corrosion experiments at elevated temperatures and pressures below the critical point is challenging, and becomes even more so under supercritical conditions. For example, material selection for autoclaves and loop piping is complicated by the fact that corrosion rates of these materials under SCW conditions are often similar to those of the materials under investigation. Electrochemical methods, often employed for corrosion studies at subcritical temperatures, become difficult or impossible at the low ionic conductivities of low-density SCW.25 Tests using irradiated materials or simulated radiolysis conditions, required for the development of an SCWR, significantly increase the experimental challenges. Key experimental variables affecting corrosion in SCW identified to date are temperature, water density (pressure), dissolved oxygen concentration, water conductivity, and surface preparation. The effect of solution pH has not been studied for the SCWR, although some fossil-fired SCW plant water chemistries are slightly alkaline. Tests have been performed at temperatures ranging from below the critical

Material Performance in Supercritical Water

temperature up to 650  C. Test durations have ranged from 24 to 3000 h. After exposure to SCW, test specimens have been characterized using techniques ranging from weight change measurements to various surface analytical methods such as scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX or EDS), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, and scanning Auger microscopy (SAM). Corrosion experiments in SCW can be performed in static autoclaves and capsules, and refreshed autoclaves and loops, complicating the comparison of data from various research groups. Loops can be either recirculating, in which a major amount of the loop water is circulated through the loop circuit without purification (a side-stream purification system is often used to reduce impurity concentrations) or ‘oncethrough,’ which includes loops that cool and purify all of the water to remove ionic impurities prior to its reintroduction to the test section. Static autoclaves are relatively simple to operate, but significant, and possibly unrepresentative, concentrations of dissolved corrosion products can build up in the autoclave as a result of the corrosion of both the test specimens and

the autoclave body in SCW. Figure 3 shows the results of experiments carried out in Hastelloy C and Alloy 625 autoclaves, in which the autoclave body itself acted as the test specimen.26 The concentrations of various metals in the solutions before and after exposure to SCW conditions were measured. The data show that significant amounts of Ni, Mo, and W were released into solution. These dissolved species can deposit onto other system surfaces, such as test specimens, by precipitation or direct incorporation into the growing corrosion film. Daigo et al.27 have shown that chromium released from the autoclave body by corrosion can migrate to the surfaces of test specimens in the autoclave, leading to improved corrosion resistance of the test alloy. (Tests were carried out at 400  C and 30 MPa in 0.01 M H2SO4 and 0.025 M oxygen (800 ppm) for the development of SCWO processes.) It was proposed that the observed reduction in corrosion rate was due to (1) changes in the solubility of Cr2O3 due to temperature gradients in the loop and (2) the possible formation of soluble Cr6+ species under the highly oxidizing conditions of these tests. The buildup of corrosion products can be minimized using a refreshed autoclave or a once-through loop. Once-through loops maintain constant, well-controlled

Elemental concentration (ng mL-1)

3500

45 40

3000

35 2500 30 2000

25

1500

20 15

1000 10 500

Percent of element in alloy

284

5

0 Cr

Mn

Fe

Ni

Zr Nb + Ta Mo Element

Blank concentration Alloy %

0 Sn

W

Water sample after 280 h exposure in hastelloy C autoclave

Figure 3 Concentrations of various metals in solution after 280-h exposure of a Hastelloy C autoclave containing deionized water at 450  C. The concentrations of various elements in the alloy and the concentrations in the water before the test are also shown. Reproduced from Guzonas, D.; Tremaine, P.; Brosseau, F. Predicting activity transport in a supercritical water cooled pressure tube reactor. 4th International Symposium on Supercritical Water-Cooled Reactors, Heidelberg, Germany, Mar 8–11, 2009.

Material Performance in Supercritical Water

chemistry conditions ideal for studying mechanistic effects, but these conditions may also be unrepresentative of those expected in a real plant, because the coolant in an operating SCWR will likely have some measurable concentration of impurities, especially metal ions, which could alter the dissolution of surface films and therefore change the corrosion rate. A recirculating loop with sidestream purification is a good compromise, allowing both good chemistry control and controlled corrosion product transport. It is important that the experimenter be aware of both the advantages and potential shortcomings of a particular test facility when planning an experiment, and ensure that all the required test parameters are identified and controlled. In all types of test facilities, characterization of water samples during and after the tests, and/or measurement of the solution conductivity before and after the test specimens, can provide some insight into the release of metal ions into solution. Under SCW conditions, the solvent density can be varied by changing either the temperature at constant pressure or the pressure at constant temperature. The former case is more relevant to an SCWR. At the relatively low pressures of interest for an SCWR, the widest variation in density can be achieved at temperatures just above the critical temperature. Figure 4 illustrates the variation in density that can be achieved as a function of temperature at a pressure of 25 MPa. It should be noted that at this pressure, the variation in density above 450  C is small, so that experiments carried out above this temperature can be considered to be at constant density (about 100 kg m3).

As in any corrosion testing, care must be taken to avoid galvanic coupling between corrosion coupons and the test section or autoclave; the coupon holder should be constructed of an insulating material. This can be a challenge in SCW, as many ceramics, including ZrO2, can dissolve at very high temperatures. The concerns about galvanic effects are mitigated to some extent at higher temperatures because of the low ionic conductivity of SCW under typical test conditions. The surface treatment of corrosion coupons prior to exposure to SCW has received only limited attention in recent investigations, although the effect has been known for decades. For example, Ruther et al.28 reported that surface preparations that left a strainfree surface resulted in much higher corrosion in superheated steam than surface preparations that severely work the surface. This effect was attributed to higher chromium diffusivity for mechanically worked surfaces. Thus, care must be taken to assess the surface preparation methods used when comparing results from different research groups. Weight gain is commonly used as a measure of the extent of corrosion, but the results are generally only semiquantitative because (1) some of the metal released from the base metal by corrosion can enter the test solution rather than remain in the oxide, (2) corrosion products from other sources in the test system can deposit on the surface, and (3) spalling of the oxide can occur. Measuring the weight loss obtained by removing the oxide film from the test coupon surface by descaling (e.g., ASTM G 1-03)

800 Tcritical

700

Density (kg m-3)

600 500 400 300 200 100 0 300

350

400

285

450 500 Temperature (⬚C)

550

600

650

Figure 4 Dependence of the density of supercritical water on temperature at a constant pressure of 25 MPa.

286

Material Performance in Supercritical Water

provides a better measure of the corrosion rate, but finding a suitable descaling method for an alloy can be challenging.

5.12.2 Corrosion in SCW The worldwide programs to study corrosion (see Chapter 5.01, Corrosion and Compatibility for an overview of corrosion in nuclear systems) in SCW have examined many material classes, including ferritic–martensitic (F/M) steels, austenitic steels, Ni-based alloys, Zr-based alloys, and Ti-based alloys29–93 (see Chapter 2.07, Zirconium Alloys: Properties and Characteristics; Chapter 2.08, Nickel Alloys: Properties and Characteristics; and Chapter 2.09, Properties of Austenitic Steels for Nuclear Reactor Applications for descriptions of each alloy class). When data from the fossil-fired SCW industry and programs to develop SCWO systems are considered, some corrosion data relevant to the development of an SCWR exist for at least 90 alloys, although there are few alloys for which a good understanding of the corrosion mechanism exists. Table 1 lists the types of materials tested and the test conditions from programs to measure corrosion in SCW; of the five classes listed, most of the work has focused on the first three. Test temperatures have ranged from subcritical up to 650  C, with dissolved oxygen concentrations ranging from <10 ppb to 8000 ppb. Tests have been performed with durations Table 1

ranging from 24 to 3000 h. It should be noted that in spite of the experimental issues raised in Section 5.12.1, in most cases, there is reasonable agreement in the corrosion data for the same alloy obtained under the same test conditions in different laboratories within the estimated experimental error. Some specific alloys have multiple designations. For example, HCM12A and T122 have the same composition as do NF616 and T92. In the figures that follow, the alloy designation used is the same as that used in the original reference. Table 1 can be used to crossreference alloys with multiple designations. 5.12.2.1

Effect of Alloy Class

The corrosion response in SCW is a strong function of alloy class, as illustrated in Figures 5 and 6, which show data from corrosion tests performed for up to 3000 h at 500  C in SCW, with a dissolved oxygen concentration of 25 ppb at the test section inlet. Of the three classes of materials included in the figures, oxidation occurs most rapidly for F/M steels and is slowest for the nickel-based alloys. Within each alloy class, there are certain alloys with better oxidation resistance. From a strictly general corrosion mitigation viewpoint, improvements in oxidation resistance are most critical in the F/M steels, as a high oxidation rate would remove too much of the base metal in a thinnedwalled reactor component like the fuel cladding. Generally, temperature was found to have the greatest influence on oxidation/corrosion of the

Summary of experiments on corrosion in pure supercritical water

Alloy class

Alloy

Temp. ( C)

Water chemistry

Exposure time (h)

Austenitic stainless steel

304, 304L, 316, 316L, 316 + Zr, 310, 310S, 310 + Zr, 347H, Sanicro28, D9, 800H, AL6XN, Carpenter 20C B3, Nitronic-50, PNC1520, alloy 1.4970 600, 625, 690, 718, 825, C22, B2, C276, MAT21, MC

290–650

Deaerated (<10 ppb) to 8000 ppb dissolved oxygen

100–3000

290–600

100–3000

T91, T91a, T91b, HCM12A (T122), HCM12, HT-9 (12Cr–1Mo–1WVNb), NF616 (T92), MA956, 2.25Cr–1Mo (T11), P2 9Cr, 12Cr, F/M, 316, Inconel, Hastelloy G-30, 19Cr, 14Cr–4Al, 16Cr–4Al, 19Cr–4Al, 22Cr–4Al Zr, Zr–Nb, Zr–Fe–Cr, Zr–Cr–Fe, Zr–Cu–Mo, Zr-2, Zr-4 Ti–3Al–2.5V, Ti–6Al–4V, Ti– 15Mo–5Zr–3Al, Ti–15V–3Al–3Sn–3Cr

290–650

Deaerated (<10 ppb) to 8000 ppb dissolved oxygen, <0.1 mS cm1 Deaerated (<10 ppb) to 8000 ppb dissolved oxygen, <0.1 mS cm1

360–600

25 ppb

200–3000

400–500

Deaerated (<10 ppb dissolved oxygen), <0.1 mS cm1 8000 ppb dissolved oxygen, 0.1 mS cm1

<2880

Nickel-based Ferritic– martensitic Oxide dispersion strengthened Zirconium-based Titanium-based

290–550

100–3000

500

Material Performance in Supercritical Water

287

8 3000 h 500 °C, 25 ppb 7

Weight gain (mg cm-2)

6

Ferritic– martensitic

5 4 3 2 Ni-based Austenitic 1 0 625

NF709

800H

316

D9

ODS

NF616 HCM12A T91

Alloy Figure 5 Comparison of weight gain for various alloys exposed to supercritical water at 500  C for 3000 h. The oxide dispersion strengthened alloy is the 9Cr oxide dispersion strengthened alloy listed in Table 1.

5 500 °C, 25 ppb NF616

Weight gain (mg cm-2)

4

3

2

316

1

0 625

-1

0

500

1000

1500 2000 Time (h)

2500

3000

3500

Figure 6 Comparison of weight gain for various alloys exposed to supercritical water at 500  C for up to 3000 h. NF616 represents a typical ferritic–martensitic steel, 316 stainless steel represents a typical austenitic steel, and Alloy 625 represents a typical nickel-based alloy.

materials (Figure 7). With an increase of exposure temperature, the weight gain associated with oxidation typically increases significantly. While corrosion studies performed in support of the development of SCWO systems suggested a maximum in the

corrosion rate at temperatures around the critical point, where the water density decreases significantly,12 most measurements of the corrosion rate under SCWR conditions do not clearly show this effect. As noted in Section 5.12.1, between about

288

Material Performance in Supercritical Water

10 20–25 ppb dissolved oxygen 1026 h

HCM12A

Weight gain (mg cm-2)

8

6

4

D9

2

625

0 Pseudo critical point -2 350

400

450

500 550 Temperature (°C)

600

650

Figure 7 Comparison of weight gain for various alloys exposed to supercritical water at temperatures from 360 to 600  C for 1026 h. HCM12A represents a typical ferritic–martensitic steel, D9 stainless steel represents a typical austenitic steel, and Alloy 625 represents a typical nickel-based alloy.

300 and 450  C, the properties of water change dramatically. At higher water densities (and higher dielectric constants), an electrochemical mechanism based on the formation and dissolution of a surface film is expected to dominate.94,95 Below some threshold water density, it has been proposed that the dominant mechanism changes to one based on the molecular interactions of the metal with oxygen and/or water and the transport of cation and anion defects in the oxide. For example, Betova et al.96 concluded that at temperatures of up to 500  C, the oxidation of stainless steels proceeds in a manner analogous to that in high-temperature subcritical water, whereas at higher temperatures, the oxidation kinetics seems to be closer to that in water vapor. Yi et al.85 concluded that corrosion in SCW was similar to that in gaseous conditions, where oxide formation occurs without metal dissolution. 5.12.2.1.1 Ferritic–martensitic steels

As seen in Figure 8, the bulk chromium concentration in F/M alloys has a significant effect on oxidation response, with higher bulk chromium concentrations leading to reduced weight gain. T91, a 9 wt% Cr alloy, has the worst oxidation response as reflected by the largest weight gain. The newer alloys, HCM12A and NF616, perform slightly better and have similar oxidation responses. The 9Cr oxide dispersion strengthened (ODS) alloy has the best oxidation response of the F/M

alloys studied, even though it has a lower bulk chromium concentration. The 9Cr ODS alloy has two significant microstructural differences compared to conventional F/M steels. It contains distributed nanosized Y–Ti–O particles91 and is composed of fine equiaxed ferritic grains about 1 mm in size.50 The typical effects of time and temperature on oxide growth in F/M steels are shown in Figure 9, using 9Cr ODS, NF616, and HCM12A as examples. The time exponents, n, obtained by fitting the weight gain data using the generalized equation △W = kptn, which is usually employed to evaluate high-temperature oxidation kinetics, are also summarized in Figure 9. In the equation, △W is the weight change of the steel (mg cm2), kp is a rate constant, and t is the exposure time (h). The oxide growth kinetics vary between parabolic and cubic rate laws, and the oxide growth rate increases significantly with temperature and tends to follow a parabolic growth rate at higher temperatures. Figure 10 shows a TEM image of the morphology of grains and grain boundaries in the internal oxidation layer of the 9Cr ODS alloy after the corrosion tests, as well as EDS analyses, indicating the presence of yttrium in selected areas.43 Yttrium was enriched in oxide ribbons that formed along the oxide–metal grain boundary regions in the internal oxidation layer. Formation of these ribbons reduced the cation flux in these regions, resulting in depletion of iron cations in the (FeCr)3O4 and the gradual decrease of scale density with the increase of exposure time.

Material Performance in Supercritical Water

289

6 500 °C, 25 ppb T91

5

Weight gain (mg cm-2)

HCM12A NF616

4

9Cr ODS 3

2

1 HT9 0 0

500

1000

1500

2000 Time (h)

2500

3000

3500

4000

Figure 8 Comparison of weight gain for various ferritic–martensitic alloys exposed to supercritical water at 500  C for up to 3000 h. HT9 and HCM12A are 12 wt% Cr alloys. T91 and NF616 are 9 wt% Cr alloys.

14 600 ⬚C 500 ⬚C 360 ⬚C HCM12A HCM12A HCM12A NF616 NF616 NF616 9Cr ODS 9Cr ODS 9Cr ODS

Weight change (mg cm-2)

12

n = 0.42

10

n = 0.37 n = 0.39

8 6 4

n = 0.39 n = 0.38 n = 0.25

2 0 -2 0

200

400

600 800 Exposure time (h)

1000

1200

Figure 9 Weight change as a function of exposure time for ferritic–martensitic steels 9Cr oxide dispersion strengthened, HCM12A, and NF616 in 500 and 600  C supercritical water and 360  C subcritical water with an inlet-dissolved oxygen content of 25 ppb. Reproduced from Chen, Y.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2007, 371, 118–128.

The strong effect of temperature on the general corrosion of F/M alloys is seen in the weight gain data for HCM12A in Figure 7. Weight gain from oxidation increases with increasing temperature from 360 to 600  C. Yi et al.85 reported a weight loss

for T91 exposed to 370  C water for 200 h at 25 MPa with <10 ppb dissolved oxygen, and Hwang et al.73 reported weight losses for T91, T92, and T122 in deaerated SCW at 370  C and 25 MPa. As noted above, there may be a change in corrosion mechanism

290

Material Performance in Supercritical Water

100 nm (a) Cr-K

Counts

2200

O-K Fe-K

Counts

0

(b)

0.000

20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0.000

keV

20.480 Y-K

2.000

4.000

6.000

8.000

10.000

12.000

14.000

16.000

18.000

keV

Figure 10 (a) Transmission electron microscopy image shows the morphology of grains and grain boundaries in the internal oxidation layer formed in the oxidized 9Cr oxide dispersion strengthened alloy and (b) X-ray spectra showing the presence of yttrium in ribbon along grain boundary. Reproduced from Chen, Y.; Sridharan, K.; Allen, T. R.; Ukai, S. J. Nucl. Mater. 2006, 359(1–2), 50–58.

Material Performance in Supercritical Water

291

0.22 HCM12A 500 °C 168–236 h

Weight gain (mg cm-2 day-1)

0.20

0.18

0.16

0.14

0.12 10

1000

100

104

Oxygen (ppb) Figure 11 Weight gain as a function of temperature for HCM12A exposed to low oxygen concentration supercritical water for 1026 h. Data are compiled from two different institutions. Reproduced from Was, G. S.; Ampornrat, P.; Gupta, G.; et al. J. Nucl. Mater. 2007, 371, 176–201.

as the density decreases above the critical point. The observed weight losses suggest that most of the metal released by corrosion dissolves into the solution phase. As the SCWR coolant will pass through the critical point in the SCWR core, additional studies of corrosion near the critical point would be helpful in determining whether a corrosion maximum exists in this region. Regardless of the uncertainties in corrosion behavior near the critical point, all studies of F/M alloys show very high corrosion rates at high temperatures relative to the austenitic and Ni-based alloys. Dissolved oxygen concentration influences the oxidation behavior, as shown in Figure 11, which presents data from test systems with flowing refreshed water. For HCM12A, the weight gain at 500  C decreases with increasing dissolved oxygen concentration between 10 and 300 ppb. At much higher dissolved oxygen concentrations (2000 ppb), the weight gain increases considerably above that experienced between 10 and 300 ppb. This dependence of oxidation on dissolved oxygen concentration is exploited in the fossil-fired energy industry. Combined water chemistry control in fossil-fired power plants97,98 involves addition of small amounts of oxygen to the feedwater to enhance the formation of hematite crystals between the magnetite grains, thus reducing the oxidation rate, perhaps by reducing the diffusion of oxygen through the multiphase film. In studies examining F/M materials

for use in SCWRs, significant hematite crystals have not been found between the magnetite grains at 10–300 ppb oxygen, where oxidation is the slowest. Figure 12 compares weight gain results at 25 and 8000 ppb for T91. Starting around 500 h, the weight gain for 25 ppb exposure appears to be greater than that from 8000 ppb exposure. This trend is the opposite of that seen for HCM12A, where high oxygen concentrations led to large oxide growth. However, in Figure 12, the data come from two different research groups using different experimental methods; the 8000 ppb data are from a static autoclave and the 25 ppb data from a flowing system. The corrosion rate in the static autoclave may become lower at longer exposure times because the test solution becomes saturated with corrosion products, inhibiting their further release into the solution. Figure 13 demonstrates the effects of both temperature and time on the corrosion of HCM12A. Similar to the data in Figure 7, weight gain for HCM12A in deaerated water is seen to be a strong function of temperature. This figure combines the measurements from two different institutions, showing consistency for nominally similar conditions. Figure 14 shows similar trends for T92 (also known as NF616) in nondeaerated (8000 ppb) SCW. Oxidation of T92 is also a strong function of temperature. The weight gain data for 500  C and 8000 ppb

292

Material Performance in Supercritical Water

6 T91 500 °C 5

Weight gain (mg cm-2)

25 ppb oxygen 4

3

2 Nondeaerated (8000 ppb)

1

0

–1 –500

0

500

1500 2000 Time (h)

1000

2500

3000

3500

Figure 12 Weight gain on T91 exposed to supercritical water at 25 and 8000 ppb at 500  C. The 8000 ppb data are from a static autoclave and the 25 ppb data from a flowing system. Nondeaerated data displayed in green are from KAERI. Reproduced from Jang, J.; Han, C. H.; Lee, B. H.; Yi, Y. S.; Hwang, S. S. Corrosion behavior of 9Cr F/M steels in supercritical water. Proceeding of ICAPP’05, Seoul, Korea, 2005; Paper no. 5136. Two symbols are shown as KAERI tested two different heats of alloy T91. Test results displayed in red are from the University of Wisconsin. Copyright 2005 by the American Nuclear Society, La Grange Park, Illinois. 10

HCM12A <10 to 25 ppb 600 °C

Weight gain (mg cm-2)

8

6

4

500 °C 2

500 °C

0

400 °C –2 –500

0

360 °C 500

1000

1500 2000 Time (h)

2500

3000

3500

Figure 13 Weight gain as a function of time for HCM12A exposed to deaerated water at 360, 400, 500, and 600  C. Test results displayed in red are from the University of Wisconsin and test results displayed in blue are from the University of Michigan. Data from Ampornrat, P.; Bahn, C. B.; Was, G. S. Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Minerals, Materials and Metals Society, 2005; p 1387.

is compared with data for 500  C and 25 ppb. In contrast to HCM12A (Figure 10) and similar to T91, T92 undergoes a larger weight gain at lower dissolved oxygen concentration.

5.12.2.1.2 Austenitic stainless steels

Figure 15 demonstrates that variability in oxidation response also occurs among austenitic stainless steels of different composition. Of the three alloys shown,

Material Performance in Supercritical Water

293

4 T92 (NF616)

3.5 550 °C

Weight gain (mg cm-2)

3 2.5 2

500 °C 25 ppb

1.5 500 °C 8000 ppb

1

400 °C 350 °C

0.5 0 0

100

200

300

400

500

600

700

800

Time (h) Figure 14 Weight gain as a function of time for T92 (also known as NF616) exposed to 8000 ppb water at temperatures ranging from 350 to 550  C (8000 ppb data from KAERI displayed as green points are adapted from Jang et al.62 and compared to data taken at 500  C and 25 ppb dissolved oxygen from the University of Wisconsin.

6 600 °C 20–25 ppb dissolved oxygen

5

Weight gain (mg cm-2)

316 4

D9

3

2

1 800H 0

0

200

400

600

800

1000

1200

Time (h) Figure 15 Weight gain as a function of time for austenitic stainless steels, D9, 316, and Alloy 800H, exposed to low oxygen concentration supercritical water at 600  C.

weight gain is greatest for 316 stainless steel. The weight gain for Alloy 800H is less than that for the other alloys, but erratic. This erratic weight gain is associated with spallation of the oxide. Guzonas et al.26 reported weight change data for the austenitic alloys Carpenter 20CB3, Nitronic-50, and AL6XN after exposure to SCW

at 450  C for 483 h at 23 MPa. Carpenter 20CB3 lost weight, while the other two alloys gained weight. As with the F/M alloys, temperature has a strong effect on oxide growth for austenitic steels. This is demonstrated in Figure 16, which shows significantly greater weight gains in 316 stainless steel with

294

Material Performance in Supercritical Water

316 stainless steel exposed at 500  C shows a slightly greater weight gain from oxidation at higher dissolved oxygen concentrations (2000 ppb vs. 25 ppb). At 550  C, the samples exposed to deaerated water (<10 ppb dissolved oxygen) show significantly greater weight gain from oxidation than samples exposed to 8000 ppb.

increasing temperature. The measured activation energies are around 210 kJ mol1 for the stainless steels (210 kJ mol1 for 304 stainless steel and 214 kJ mol1 for 316L stainless steel), corresponding to cation diffusion through the oxide as the rate-limiting step.81 The effect of dissolved oxygen concentration on the oxidation of austenitic steels does not demonstrate a clear trend as evidenced by Figure 17.

5.12.2.1.3 Ni-based alloys

As shown in Figure 5, Ni-based materials show relatively little oxidation compared to F/M and austenitic steels. As an example, Figure 18 shows that the weight gain for Alloy 625 is small and does not change significantly with time for material exposed at 500 and 600  C. The Ni-based alloys were the only class of alloys to show increasing weight gain (retained oxide) at temperatures below the critical point (360  C), where the density of the water is greater than just above the critical point. As with other alloys, weight gain due to oxidation increases with temperature, as demonstrated in Figure 19. The measured activation energies are 140 kJ mol1 for the Ni-based alloys (134 kJ mol1 for Alloy 690 and 142 kJ mol1 for Alloy 625), corresponding to cation diffusion through the oxide as the rate-limiting step.81 At low temperatures (400  C), Alloy 625 exhibits a weight loss due to pitting. In Ni-based materials, the weight change for shorter durations is determined

1.4 485 h Weight gain (mg cm-2)

1.2

316 stainless steel <10 ppb oxygen

1.0 0.8 0.6 0.4

419 h

0.2 575 h 0

400

578 h 450 500 Temperature (°C)

550

Figure 16 Effect of temperature on the oxide growth of 316 stainless steel in deaerated water for roughly 500 h. Reproduced from Was, G. S.; Teysseyre, S.; Jiao, Z. Corrosion 2006, 62(11), 989–1005.

0.07 316L < 10 ppb

Weight gain (mg cm-2 day-1)

0.06 0.05 316 and 316L 10–8000 ppb dissolved oxygen

0.04 0.03

316 2000 ppb

0.02 0.01 316L 8000 ppb

316 25 ppb 0 250

300

350

400

450

500

550

600

Temperature (°C) Figure 17 Weight gain rate as a function of temperature for 316L exposed in <10 ppb dissolved oxygen (blue circles), 316L exposed in 8000 ppb dissolved oxygen (green diamonds), 316 exposed in 25 ppb dissolved oxygen (red triangles), and 316 exposed in 2000 ppb dissolved oxygen (black squares). Data for exposure in 2000 ppb and less from Was et al.37 Data at 8000 ppb from Kasahara.60

Material Performance in Supercritical Water

by both the weight loss from pitting and the weight gain from general oxidation. As a result, the actual weight gain may be underestimated. Nevertheless, experimental evaluation of layer thickness showed that this class of alloys exhibits the lowest oxide layer thickness.

0.3

Weight gain (mg cm-2)

360 °C

625 20–25 ppb dissolved oxygen

0.25 0.2 0.15 0.1

600 °C 0.05 0

-0.05

500 °C 0

200

400

600 Time (h)

800

1000

1200

Figure 18 Weight gain as a function of time for Alloy 625 exposed to low oxygen concentration supercritical water at 360, 500, and 600  C. Adapted from Was, G. S.; Allen, T. R. Time, temperature, and dissolved oxygen dependence of oxidation of austenitic and ferritic–martensitic alloys in supercritical water. Proceedings of ICAPP’05; Paper no. 5690; Ren, X.; Sridharan, K.; Allen, T. R. Corrosion 2007, 63, 603.

295

5.12.2.1.4 Zr-based alloys

Zirconium-based alloys have been considered for incore SCWR applications (fuel cladding in all designs, pressure tubes in a pressure tube design) because of their good neutron economy, although current commercial alloys lack the high-temperature strength for load-bearing application at SCW temperatures. The CANDU SCWR design retains the use of zirconium alloy pressure tubes by using an insulated fuel channel design9; in this design, the pressure tube, which forms the in-core pressure boundary, operates at about the moderator temperature (80  C), where the mechanical strength is higher and the corrosion rate is lower. The zirconium alloy Excel,99 a Canadian zirconium-base alloy with nominal composition of Zr–3.5Sn–0.8Nb–0.8Mo, has been chosen for this application. Alloys investigated for use as an SCWR fuel cladding in international programs include Zircaloy-2, Zircaloy-4, zirconium, and model alloys including Zr–Nb, Zr–Fe–Cr, and Zr–Cu–Mo.34,52,69,77,92 Initial studies of Zircaloy-2 and Zircaloy-4 showed extremely high oxidation, suggesting that zirconiumbased alloys would be unacceptable for use at the temperatures required in an SCWR core. Further studies of model alloys52,92 have indicated that optimized compositions can reduce the oxidation rate to levels less than that of F/M steels, although still higher than that of austenitic steels. Oxidation results

0.03 0.025

485 h

Alloy 625 <10 ppb oxygen

Weight gain (mg cm-2)

0.02 0.015 0.01

578 h

419 h

0.005 575 h 0 -0.005 -0.01

400

450

500

550

Temperature (°C) Figure 19 Weight gain as a function of temperature for Alloy 625. Reproduced from Was, G. S.; Teysseyre, S.; Jiao, Z. Corrosion 2006, 62(11), 989–1005.

296

Material Performance in Supercritical Water

2.5 500 °C, <10 ppb O2 Processed at 580 °C

Weight gain (mg cm-2)

2

1.5 Processed at 720 °C

1

0.5

0

0

500

1000

1500

2000

2500

3000

3500

4000

Time (h) Figure 20 Weight gain results for Zr–0.4Fe–0.2Cr alloys exposed to supercritical water at 500  C. The two datasets represent materials processed at two different temperatures before exposure. Graphic produced from the data in Motta et al.92

for one set of the model Zr–Fe–Cr alloys are shown in Figure 20. 5.12.2.1.5 Ti-based alloys

Japanese programs have studied the corrosion of Ti-based alloys in SCW. Alloys examined include Ti–15Mo–5Zr–3Al, Ti–3Al–2.5V, Ti–6Al–4V, and Ti–15V–3Al–3Sn–3Cr.65 These Ti-based alloys were exposed in static autoclaves with 8000 ppb dissolved oxygen for 500 h at temperatures of 290, 380, and 550  C. Little difference was noted in the weight gain of samples exposed at 290 and 380  C. Weight gain of the samples exposed at 550  C was significantly higher than that of samples exposed at lower temperatures. At 550  C, the Ti–15V–3Al–3Sn–3Cr and Ti–15Mo–5Zr–3Al alloys had roughly a factor of three lower weight gain than the Ti–3Al–2.5V and Ti–6Al–4V alloys. The weight gain at 550  C for Ti–15V–3Al–3Sn–3Cr is similar to the weight gain experienced by 304 stainless steel exposed at the same temperature and <10 ppb dissolved oxygen. 5.12.2.2

Oxide Structures

5.12.2.2.1 Ferritic–martensitic steels

The oxide formed on a F/M steel exposed under low (ppb) dissolved oxygen concentration consists of two layers plus an internal oxidation layer.48 Examples

of oxide layer growth on F/M steels are shown in Figure 21. A diagram of the process is shown in Figure 22.100 All F/M steels form thick, but stable, surface oxide layers. The scale consists of a layered structure with an outer Fe3O4 layer, an inner (FeCr)3O4 layer with the spinel structure, and an innermost internal oxidation layer. (Spinel refers to a 2 class of compounds of general formulation A2+B3+ 2 O4 that crystallize in the cubic (isometric) crystal system. The oxygen anions are arranged in a cubic closepacked lattice and the cations A and B occupy some or all of the octahedral and tetrahedral sites in the lattice. Magnetite and nickel ferrite are examples of iron oxides having the (inverse) spinel structure.) The outer Fe3O4 layer predominantly consists of coarse columnar grains and the inner (FeCr)3O4 layer is composed of very fine equiaxed grains. Depending on the steel, the thickness of the internal oxidation zone was observed to vary significantly. None of the F/M steels tested showed any indication of oxide scale spallation at exposures of up to 3000 h for all temperatures tested (290–650  C, see Table 1). 5.12.2.2.2 Austenitic steels

Most examinations of oxidation layers formed on austenitic alloys exposed to SCW reveal that the oxide scale has a layered structure consisting of an outer magnetite or mixed magnetite–hematite layer

Material Performance in Supercritical Water

297

9Cr ODS

15 kV ⫻ 3000

8 mm

15 kV ⫻ 3000

8 mm

15 kV ⫻ 3000

8 mm

15 kV ⫻ 3000

8 mm

HCM12A

Figure 21 Top figure is a cross-section of the oxide formed on NF616 exposed at 500  C for 1026 h. Bottom figure is a comparison of HCM12A and 9Cr oxide dispersion strengthened alloy exposed at 500  C in 25 ppb dissolved oxygen for 1026 h. 9Cr oxide dispersion strengthened alloy has a thinner and less porous oxide.

2

1

Atomic concentration (at.%)

100

3

2 mm

O Cr Fe

80

60

40

20

0 0

5

Outward growth

10

15 20 Distance (mm)

25

30

35

Inward growth Fe2/3+, eCr3+, e-

Diffusion

Diffusion zone

Bulk Fe–Cr alloy

O2-

SCW

Fe3O4

Fe–Cr spinel

Fe Figure 22 Typical cross-section for the oxide formed on a ferritic–martensitic steel exposed to 25 ppb supercritical water. The upper left image is a scanning electron micrograph of the oxide cross-section. Between points 1 and 2 is the outer magnetite layer. Between points 2 and 3 is the inner spinel layer. To the right of point 3 is the diffusion layer and the bulk metal.

298

Material Performance in Supercritical Water

5 mm Figure 23 Cross-section morphology of the D9 stainless steel after exposure to supercritical water with 25 ppb of inlet oxygen content at 500  C for 505 h.

scale, in which the oxide grains display preferential orientations imposed by the original austenite grain orientations.101 5.12.2.2.3 Ni-based alloys

20 µm

20 µm

20 µm

Figure 24 Cross-sectional morphologies of the oxide layers formed after exposure to supercritical water with an inlet-dissolved oxygen content of 25 ppb at 600  C: (a) 316 stainless steel for 1000 h, (b) D9 stainless steel for 1000 h, and (c) NF709 for 667 h.

and an inner (Fe, Cr, Ni, M)3O4 (M: other minor elements) layer with the spinel structure. In some cases, particularly at higher temperatures, bulk microstructure rather than bulk composition was observed to promote changes in the multilayered structures and significantly alter the corrosion behavior of the alloys. Figure 23 shows that the oxide layer formed on the 15Cr–15Ni austenitic D9 stainless steel exhibits a nodular structure with an uneven surface after exposure at 500  C.101 The spacing of these nodules is in the range of 5–10 mm, which is comparable to the size of the austenitic grains in the D9 stainless steel. At 600  C, a dual-layered structure was observed on 316 stainless steel, as shown in Figure 24(a). For the 15Cr–15Ni D9 stainless steel, an additional continuous Cr2O3 layer was observed to form between the spinel layer and the base alloy.47 Thereafter, D9 stainless steel showed much lower weight gain than even the more highly alloyed NF709 after exposure to 600  C SCW, implying good corrosion resistance to the SCW environment. For the 20Cr–25Ni alloy, NF709, the near-surface austenite grains were readily oxidized intragranularly to form the inner oxide

The oxides found on Ni-based alloys are thin compared to those found on the other alloy classes; Guzonas26 reported oxide thicknesses of about 50 nm on Alloy 625 and Alloy 690 after 483-h exposure to SCW at 450  C and 23 MPa. The surface morphologies of the Ni-based alloys, Alloys 625 and 718, after exposure to SCW at 500  C for about 500 h are shown in Figure 25. For Alloy 625, the oxide layer consisted of oxide particles with a typical size of about 1 mm on a uniform oxide layer composed of very fine oxide particles in the size range of tens of nanometers. This particle size was verified by high magnification imaging, which is not shown here. Pits with varying diameters were also observed on the sample surface. Similarly, Alloy 718 exhibited oxidation, consisting of oxide particles on a uniform oxide layer, which leads to weight gain, and local pitting, which leads to weight loss during SCW exposure. For Alloy 625, increasing the SCW test temperature to 600  C resulted in the formation of a uniform oxide surface with fine particulates, while the grain boundaries were outlined by a topographically elevated oxide, as shown in Figure 26(a). Gray-scale EDS mapping images for oxygen, chromium, and nickel shown in Figure 26(b), 26(c), and 26(d), respectively, indicate that grain boundaries on the oxide surface were enriched in oxygen and chromium, whereas nickel was depleted. These results indicate that grain boundaries are the primary paths for cation migration in these corrosion regimes.47 There were no identifiable X-ray diffraction (XRD) peaks corresponding to oxide from Alloy 625 at 500  C for 505 h because the oxide layer formed was too thin. However, at 600  C for 1026 h, three types of oxide products were identified on Alloy 625. The major phase was a spinel phase with a stoichiometry of Ni(Fe,Cr)2O4, and the other two oxides were identified as Cr1.3Fe0.7O3 and NiO.47

Material Performance in Supercritical Water

(a)

299

(b) Pitting in Alloy 625

Oxidation in Alloy 625

(c)

10 mm

(d) Oxidation in Alloy 718

Pitting in Alloy 718

Figure 25 Morphological features of surface corrosion for Alloy 625 and 718 samples after exposure to supercritical water with an inlet-dissolved oxygen content of 25 ppb at 500  C for 500 h. Reproduced from Ren, X.; Sridharan, K.; Allen, T. R. Corrosion 2007, 63, 603.

15

25 mm

(a) Cr K

(c)

65535

28

133

25 mm

OK

6

25 mm

(b) Ni K

(d)

75

20

130

25 mm

Figure 26 Intergranular corrosion observed in Alloy 625 after exposure to supercritical water at 600  C for 1026 h. Scanning electron microscope image is shown in (a) and energy dispersive X-ray spectroscopy mapping images of Cr, Ni, and O are shown in (b), (c), and (d). Reproduced from Ren, X.; Sridharan, K.; Allen, T. R. Corrosion 2007, 63, 603.

Guzonas et al.26 reported Raman spectra from the oxides formed on Alloy 625 and Alloy 690 after exposure to SCW at 450  C for 483 h. The Raman spectra suggested that the oxide on these alloys was

mainly NiFe2O4; the spectrum obtained from the Alloy 690 surface showed an additional, unidentified band. Auger electron spectroscopy of the surface of Alloy 625 shows a duplex oxide structure after

300

Material Performance in Supercritical Water

80

In625 at 500 ⬚C for 500 h

14

Oxygen

60

Nickel

50 40 Chromium 30 20

600 ⬚C 600 ⬚C

10 8 6 4

Molybdenum 2

10

0

0 0

3

6

80

12 9 Sputter time (min)

15

18

In625 at 600 ⬚C for 1026 h

70 Atomic concentration (%)

NF616 AR 500 ⬚C 500 ⬚C Y

12 Weight gain (mg cm−2)

Atomic concentration (%)

70

60

Nickel

Oxygen

50 40 Chromium

30 20

Molybdenum

10

0

200

400

600 Time (h)

800

1000

Figure 28 Weight change data as a function of exposure time for as-received (AR) and Y-coated ferritic steel Alloy NF616 after exposure to supercritical water with an inlet-dissolved oxygen concentration of 25 ppb at 500 and 600  C for up to 1000 h. Other ferritic–martensitic steel samples exhibit comparable trends before/after coating with a 250 nm yttrium layer. Adapted from Tan, L.; Machut, M. T.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2007, 371, 171–175; Chen, Y.; Sridharan, K.; Allen, T. R. In 13th Environmental Degradation of Materials in Nuclear Power Systems 2007; Allen, T. R., Busby, J. T., King, P. J., Eds.; Canadian Nuclear Society, 2007.

Iron

0 0

10

20 30 Sputter time (min)

40

Figure 27 Chemical composition versus depth profiles as determined by Auger electron spectroscopy for Alloy 625 after exposure to supercritical water with an inlet-dissolved oxygen content of 25 ppb at (a) 500  C for about 500 h and (b) at 600  C for 1026 h. Reproduced from Ren, X.; Sridharan, K.; Allen, T. R. Corrosion 2007, 63, 603.

exposure to SCW at both 500 and 600  C.47 The oxide consisted of a nickel/iron-rich outer layer and a chromium-rich inner spinel layer, as indicated in Figure 27. Moreover, a diffusion layer existed between the oxide layer and the base alloy, where the chemical compositions of all elements gradually changed from that of the oxide to the bulk alloy concentration. 5.12.2.3

Surface Composition Modification

In an attempt to mitigate the high oxidation rates seen in the F/M alloys, modification of the surface composition was investigated. Shown in Figure 28 as an example, coating a thin yttrium film on the surface of the F/M steel samples significantly mitigated the

corrosion rate in SCW environments.49,51 Figure 29 shows the cross-sectional view and EDS mapping of the Y-coated 9Cr ODS alloy after exposure to SCW at 500  C. In comparison with the oxide structure for uncoated 9Cr ODS alloy (Figure 21), a yttrium-rich layer formed on the Y-surface-treated 9Cr ODS alloy. This layer was observed to divide the magnetite layer into two parts.49,51 The yttrium, shown in the elemental mapping (Figure 29(c)), is mostly concentrated in one central layer, indicating that this layer formed in the initial stages of exposure because of the high thermodynamic stability of Y-based oxide, and was subsequently driven outward because of the growth of the fine Fe3O4 grains underneath. XRD analyses have indicated that yttrium is predominantly present in the form of a YFeO3 phase.51 The yttrium layer converted to a discrete YFeO3 particulate layer, and this layer bisected the magnetite layer into two regions with distinct morphologies. The outer magnetite layer was porous, while the inner magnetite layer was fine grained. The YFeO3 particulates acted as effective local diffusion barriers for cation diffusion; however, the interparticulate regions provided adequate diffusion paths for cations to promote the formation of a fully developed oxide layer.

Material Performance in Supercritical Water

4 mm

(a) 100

Atomic concentration (at. %)

80 60

O Cr Fe

40 20 0 5

Y W

4 3 2 1 0 0

5

10

(b)

15

20

25

30

Distance (mm)

O

Cr

Fe

Y

301

5.12.2.3.1 Effect of grain size refinement

Grain boundaries are fast diffusion pathways. Since diffusion is critical to oxide growth, one possible way to influence oxidation rates is to change the grain size. This has two potential offsetting effects. Smaller grains allow oxide-forming elements, like chromium, to quickly diffuse to the surface and form protective oxides. On the contrary, smaller grains can allow oxygen to more rapidly enter the bulk metal and form an oxide. This is not desired. Typically, the amounts of protective oxide-forming elements, like chromium, in the bulk metal determine whether smaller grains are beneficial to mitigating oxidation. A smaller grain size was obtained on T91, using equal channel angular processing (ECAP) and shot peening. Before the grain refinement treatments, the grain size was 5 mm. Following treatment, as shown in Figure 30, the ECAP material had an average grain size of 1.25 mm. In the affected near-surface region, the shot-peened material had an average grain size of 0.93 mm. Figure 31 shows that weight gain is reduced by both ECAP and shot-peening treatments. The reduction in weight gain is roughly 30%. As demonstrated in Figure 32 for the shot-peened material, the spinel layer is thinnest in those regions where the bulk metal grains are smallest, providing indications that smaller grains in the bulk metal can lead to a more protective oxide. Figure 33 provides an additional example of improvement in oxidation resistance through grain refinement. The data in Figure 33

0.12

(c) Figure 29 (a) Cross-sectional scanning electron microscopy image of the Y-coated 9Cr oxide dispersion strengthened ferritic steel after exposure to supercritical water at 500  C with an inlet-dissolved oxygen concentration of 25 ppb for 667 h, (b) corresponding energy dispersive X-ray spectroscopy elemental composition profile across the oxide thickness, and (c) energy dispersive X-ray spectroscopy mapping showing the elemental distribution of O, Cr, Fe, and Y. Reproduced from Chen, Y.; Sridharan, K.; Ukai, S.; Allen, T. Effect of yttrium thin film surface modification on corrosion layers formed on C rods steel in supercritical water. In 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Vancouver, Canada, August 2007.

Area fraction

0.10 0.08 0.06 0.04

ECAPed avg = 1.25

0.02 Shot-peened avg = 0.93 0.00 0

1

3 2 Grain size (µm)

4

5

Figure 30 Comparison of grain size for T91 produced by equal channel angular processing and shot-peening processes. Both treatments lead to substantial grain refinement. Ren, X.; Sridharan, K.; Allen, T. R. Mater. Corros. 2010, 61(9), 748–755. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

302

Material Performance in Supercritical Water

6 TP1 500 °C, 25 ppb

As-received

Weight gain (mg cm–2)

5

Nanograined

4

Shot peened 3

2

1

0 0

500

1000

1500

2000

2500

3000

3500

4000

Time (h)

Figure 31 Comparison of weight gain for T91 in three different metallurgical conditions exposed to supercritical water at 500  C for times up to 3000 h.

near-surface region significantly reduces the oxidation rate by 60% in Alloy 800H. 5.12.2.3.2 Effect of grain boundary structure optimization

15 µm Figure 32 Example of changing oxide thickness with local grain size in T91 exposed to supercritical water at 500  C. The spinel layer (black) is thinnest in regions where the bulk metal grains (right side of figure) are the smallest. Ren, X.; Sridharan, K.; Allen, T. R. Mater. Corros. 2010, 61(9), 748–755. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

are for an austenitic steel, Alloy 800H. After shot peening, the surface grain size of the Alloy 800H was around 15 nm. During exposure to SCW, the grain size grew to 180 nm. As with T91, shot peening and the subsequent grain size reduction in the

Given that grain size significantly affects the oxidation response in both F/M and austenitic alloys, studies were performed to determine whether changing the character of the grain boundary distributions is important. Grain boundary engineering (GBE) was applied to Alloy 800H by means of one cycle of thermomechanical processing, followed by annealing to improve the spallation resistance of the protective oxide layer.38,53,102,103 Weight change measurement and SEM evaluations after SCW corrosion tests showed that extensive oxide spallation occurred in the as-received samples, but not in the GBE-treated samples (Figures 34 and 35). To explain this distinct difference, cross-sections of the samples were examined using electron backscatter diffraction (EBSD). Figure 36 shows the EBSD maps illustrating the distribution of phases and strain in the as-received sample (Figure 36(a) and 36(b)) and the GBE-treated sample (Figure 36(c) and 36(d)) after exposure to SCW at 500  C for 505 h and 690 h. As shown in these figures, austenite (fcc structure), magnetite, Fe–Cr spinel, and hematite were identified for both types of samples. Magnetite and Fe–Cr spinel are not distinguishable by EBSD because of their identical crystal structure, but EDS analysis showed

Material Performance in Supercritical Water

303

0.8 Alloy 800H 500 °C, 25 ppb 0.7

As received

Weight gain (mg cm–2)

0.6 0.5 0.4 0.3 0.2 0.1 0

Shot peened

0

500

1000

1500 2000 Time (h)

2500

3000

3500

Figure 33 Comparison of weight gain for Alloy 800H in large- and small-grained metallurgical conditions exposed to supercritical water at 500  C for times up to 3000 h.

0.5 800 H 0.4 Weight gain (mg cm–2)

TM processed Annealed (1 mm) Annealed (600 grit)

500 °C 25 ppb

0.3

0.2

0.1

0.0

–0.1 0

200

400

600

800

1000

1200

Time (h) Figure 34 Weight gain versus time for Alloy 800H exposed at 500  C. TM refers to thermomechanically processed condition. Annealed samples were polished to 1 mm and 600 grit finish prior to exposure. TM samples were polished to 1 mm finish prior to exposure.

that the inner layer is an Fe–Cr spinel phase and the outer layer is magnetite. Comparing Figure 36(a) with 36(c), it can be seen that the GBE-treated sample showed a higher fraction of hematite mixed with a smaller amount of magnetite in the outer layer, and an inner layer of a spinel phase mixed with phases identified as austenite.

The strain distribution shown in Figure 36(b) is presented by the local average misorientation between each EBSD data point measurement and its neighbors, excluding any higher angle boundaries (>5 ). This figure indicates that for the as-received sample, there is a strain accumulation close to the Fe–Cr spinel–magnetite interface (the interface

304

Material Performance in Supercritical Water

(a) AR: 500 °C/3-week

(b) GBE: 500 °C/4-week

Exfoliated

Compact and continuous 2000 µm

Retained oxide

2000 µm

(c) AR: 600 °C/6-week

(d) GBE: 600 °C/6-week

Exfoliated

Compact and continuous

Hematite 200 µm

200 µm

Figure 35 Plan view images indicate that spallation of the oxide was significant in as-received Alloy 800H, but not in the grain boundary-engineered Alloy 800H. Reproduced from Tan, L.; Sridharan, K.; Allen, T. R.; Nanstad, R. K.; McClintock, D. A. J. Nucl. Mater. 2008, 374, 270–280.

fcc

(c) GBE (phase)

(b) AR (strain)

S

M

fcc

H

(d) GBE (strain)

fcc + S

(a) AR (phase)

M H

2 mm

2 mm Inner

Outer

2 mm

2 mm Inner

Outer

Inner

Outer

Inner

Outer

Figure 36 Electron backscatter diffraction maps of cross-sections of oxide layer formed on Alloy 800H samples, demonstrating the phase and strain (0 5 average misorientation) distribution: (a, b) as-received (AR) samples exposed to supercritical water with an inlet-dissolved oxygen content of 25 ppb at 500  C for 3 weeks and (c, d) grain boundary engineering-treated samples exposed to supercritical water at 500  C for 4 weeks. The labels fcc, S, M, and H denote phases with face-centered cubic structure, such as the substrate austenite, spinel [(Fe,Cr)3O4], magnetite [Fe3O4], and hematite [Fe2O3], respectively. The specimen layout during electron backscatter diffraction analysis is schematically shown beside the electron backscatter diffraction maps, with the direction of TD [010] and RD parallel to oxide growth and oxide surface, respectively. Adapted from Tan, L.; Sridharan, K.; Allen, T. R.; Nanstad, R. K.; McClintock, D. A. J. Nucl. Mater. 2008, 374, 270–280.

between the inner and the outer layers). The strain distribution shown in Figure 36(d) for the GBE-treated sample is relatively uniform in the oxide scale.

By integrating the strain intensity along the direction parallel to the Fe–Cr spinel–magnetite interface, relative strain intensity as a function of the location across the oxide scale was obtained and

Material Performance in Supercritical Water

305

1.0

Strain intensity (normalized)

+ 0.8

AR GBE

0.6

0.4

0.2 Metal Inner oxide Outer oxide 0

1

2

3 4 Distance (mm)

5

6

Figure 37 Normalized strain intensity across the oxide scale on the as-received and the grain boundary engineeringtreated Alloy 800H samples. The two lines are aligned at the inner–outer oxide interface. Reproduced from Tan, L.; Sridharan, K.; Allen, T. R.; Nanstad, R. K.; McClintock, D. A. J. Nucl. Mater. 2008, 374, 270–280.

is shown in Figure 37. These data clearly show that there is a sharper strain change at the Fe–Cr spinel– magnetite interface in the as-received sample compared to the GBE-treated sample. The strain change between the inner or outer oxide layers was significantly reduced after GBE treatment. This is believed to be attributable to the increased population of low energy coincidence site lattice boundaries (CLSBs) for the GBE-treated samples,104,105 and results in lower levels of oxide spallation for the GBE-treated alloy. 5.12.2.4

Summary of Corrosion in SCW

1. All F/M steels exhibited a dual-layered oxide structure after exposure to SCW, composed of an outer Fe3O4 magnetite layer and inner Fe–Cr–O spinel layer. 2. Oxide growth, as measured by weight gain of the samples, increased dramatically with temperature. 3. The 12Cr F/M steels exhibited a lower weight gain compared to those containing 9% Cr, with the exception of 9Cr ODS alloy, which showed a lower weight gain than all the F/M steels tested. TEM examination of the oxidized layer in the 9Cr ODS alloy indicated segregation of yttrium to the grain boundary area in the internal oxidation layer, which may retard outward cation diffusion. 4. For austenitic alloys, the weight gain and oxide thickness were substantially lower than for F/M steels. Weight change measurements and SEM

cross-sectional examination showed that the oxide layer in these alloys is susceptible to spallation. In some of these alloys, an innermost Cr2O3 layer was also observed. 5. In terms of weight gain, the Ni-based alloys showed the best performance, but these alloys also exhibited some pitting that may numerically offset some of the weight gain from oxidation. 6. Surface modification approaches, such as yttrium surface treatment, can reduce the oxidation kinetics and hold promise, particularly for ferritic steels. 7. Grain refinement techniques, such as ECAP and shot peening, lowered the oxide growth rate in both F/M and austenitic alloys. GBE reduced the tendency for oxide spallation in austenitic alloys, and strain measurements across the oxide layer showed a more gradual change in strain distribution for the GBE-treated samples, which may be in part responsible for the reduced spallation.

5.12.3 Stress Corrosion Cracking in SCW This section reviews the current understanding of SCC behavior of candidate materials for SCWR systems. With the exception of the effect of irradiation on SCC, the important issues of radiation response and radiolysis are not addressed in this review. In general, the data reviewed here are distinct from the body of literature covered by SCWO experiments;

306

Material Performance in Supercritical Water

however, most of the data reported below on the effects of water chemistry and of density or dielectric constant are from the SCWO literature. In support of the SCWR program, corrosion and SCC have been studied in pure SCW in austenitic stainless steels, Ni-based alloys, F/M steels, and Tibased alloys. Test temperatures have ranged from 288 to 732  C. Dissolved oxygen concentration has ranged from <10 ppb to 8000 ppb. The effects of chemical additions have been examined in the context of the development of SCWO systems, specifically the effects of addition of H2SO4, HCl, H2O2, and NaCl. Additionally, the effects of system pressure and dielectric constant on SCC resistance have been studied. SCC experiments in SCW include constant extension rate tensile (CERT) tests, constant load tensile (CLT) tests, and crack growth rate (CGR) tests. A total of four alloy systems have been investigated for their SCC response in varying levels of detail: austenitic stainless steels (nine alloys), nickel-based alloys (eight alloys), F/M steels (six alloys), and one titanium and one F/M ODS alloy. Intergranular stress corrosion cracking (IGSCC) has been found in all but the titanium alloy system, which exhibited transgranular stress corrosion cracking (TGSCC). The alloy systems, specific alloys and range of test conditions are summarized in Table 2. The existing database is characterized by a large number of alloys and parameters, yet experiments were conducted by relatively few investigators and laboratories. The selections of the reference alloy condition (sensitized vs. unsensitized) and the reference water chemistry (deaerated vs. 8000 ppb dissolved oxygen) tend to be specific to the laboratory in which the experiments were conducted. The result is that there exist few instances in which data can be compared between laboratories because of the inherent differences in experimental conditions. As such, the following results and their interpretation are compromised somewhat by the lack of a systematic approach across the various laboratories and investigators conducting the experiments. An issue of particular importance in the determination of SCC susceptibility is the way in which this susceptibility is measured. Two different measures are presented in the literature. One is based on the amount of intergranular cracking (%IG) of the fracture surface following failure, and the other is based on the quantification of cracks on the sample surface (density, length, length/unit area, depth). Tsuchiya et al.,61,106,109 Fujisawa et al.,110 and Saito107 use the %IG on the fracture surface as a measure of SCC

susceptibility. Was and coworkers84,111,122 use crack depth and crack density on the gauge surface as an indicator of the IGSCC susceptibility. Unfortunately, these measures are not always in agreement. As described later, results often show that extensive IG cracking occurred in multiple locations in the gauge section, but very little was observable on the fracture surfaces. This is likely due to two principal factors: obscuring of the fracture mode by severe oxidation following fracture, and the poor statistical probability of the fracture surface capturing the extent of IG cracking. It has been shown in many studies that ductile alloys that fail by IG fracture in service (e.g., Alloy 600 in primary water) often exhibit only small amounts of IG cracking in CERT tests.127 Because CERT tests are typically conducted at a high strain rate relative to the growth rate of a stress corrosion crack, the plastic strain tends to ‘outrun’ the IG cracks that have initiated, resulting in a largely ductile fracture mode. For example, a rough measure of the CGR can be obtained by dividing the maximum crack depth by the time of the test. For 304 stainless steel tested in 400  C deaerated SCW,111 this value falls between 1.0  108 mm s1 (400  C) and 3.2  107 mm s1 (550  C). So, even the fastest CGR is still an order of magnitude below the extension rate (7  106 mm s1) imposed by the test. As an aside, these CGRs are similar to those experienced by 304 stainless steel in BWRs under low-potential conditions.128 In these cases, the fracture surfaces are largely ductile, but the samples display significant IG cracking on the surfaces, but with little depth penetration. As such, a better measure of cracking is obtained through analysis of the gauge section and the cross-section of the tensile bars. Cracking severity, as measured by the crack length per unit area (crack density  crack length on the gauge surface), incorporates both the density and length of the cracks. Further, crack length measured on the gauge surface correlates well with crack depth measured in cross-section. This agreement is supported by extensive work by Santarini,129 who showed through successive crack sectioning that there is a functional relationship between crack length and crack depth in CERT samples. While the significance of cracking on the gauge surface is not yet understood, the degree of IG cracking is known to depend on the strain rate in CERT tests, and at lower strain rates, the amount of IG cracking could increase substantially.130,131 However, a systematic analysis of the effect of strain rate on IGSCC in SCW has not yet been conducted. While it

Table 2

Summary of alloys and conditions used in stress corrosion cracking experiments in supercritical water Alloy

Temp. ( C)

Water chemistry

Loading mode

Results

References

Austenitic stainless steel

304, 316L (sensitized) 310S (fine grain) 316L

290–550

8000 ppb DO, 0.06 mS cm1, 25 MPa

slow strain rate test 4  107 s1

IG 304 below 400  C

290, 550 400 400–550

slow strain rate test 4  107 s1 slow strain rate test 2.78  106 s1 slow strain rate test 3  107 s1

No IGSCC IGSCC

304L, 316L

IG at all temperatures

[111]

316 (sensitized)

360, 400

[112]

500, 650

slow strain rate test 2.78  106 s1 slow strain rate test 3  107 s1

IG, pressure dependent

347H, 316NG, 1.4970, BGA4 316

8000 ppb DO, 25 Mpa 8000 ppb DO, 25 MPa, up to 0.01 mol l1 HCl, up to 3.6 MPa H2 Deaerated (<10 ppb DO), nondeaerated (8000 ppb DO), <0.1 mS cm1, 25 MPa 8000 ppb DO, 25–60 MPa, 2.4–13.4 dielectric constant 100–150 ppb, <0.1 mS cm1, 25 MPa

[61, 106, 107, 108] [109] [110]

IGSCC 316NG and BGA4

[113]

288–500

CGR

CGR temperature dependent

[114]

316L, 316LGBE

500

Nondeaerated (2000 ppb DO), deaerated (<10 ppb DO), 0.1– 0.5 mS cm1 10.3–24.8 Mpa <10 ppb DO, <0.1 mS cm1, 25 MPa

slow strain rate test 3  107 s1

[115]

316 (sensitized)

360, 400

316L (SA)

360, 400

S31266

400–500

8000 ppb, 25–60 MPa, 2.4–13.4 dielectric constant 8000–800 000 ppb, up to 0.01 M H2SO4, 30–60 MPa 5–25 MPa, 10 wt% H2O2, up to 1.6 g l1 NaCl, up to 1 g l1 HCl

316, 347

732

Deaerated, 34.5 MPa

TG crack in 316

[90]

316L, D9 (0, 7 dpa) 316L (0, 7 dpa) 316L (SA and CW, 27–44 dpa)

400, 500

<10 ppb DO, <0.1 mS cm1, 25 MPa

slow strain rate test 2.78  106 s1 slow strain rate test 2.78  106 s1 slow strain rate test 103 s1, 5  107 s1 Constant load – 87% sy, 100% sy Constant load 103, 83 MPa for 168 h slow strain rate test 3  107 s1

IG reduced in GBE condition IG, strong dependence on pressure IG, strong dependence on pressure, H2SO4 IG crack under certain condition

IG at both temperatures

[117]

400, 500 400, 500

<10 ppb DO, <0.1 mS cm1, 25 MPa <10 ppb DO, hydrogenated, <0.1 mS cm1, 24–27.6 MPa

slow strain rate test 3  107 s1 slow strain rate test 3  107 s1

[118] [119]

800H

500

400–600 ppb DO, <0.2 mS cm1, 25 MPa

800H (0, 7 dpa)

400, 500

<10 ppb DO, <0.1 mS cm1, 25 MPa

slow strain rate test 0.8  107 s1 or 1.5  107 s1 (conflicting rates given) slow strain rate test 3  107 s1

IG at both temperatures IGSCC correlates w/hardening and hydrogen No SCC IG at both temperatures

[117, 121]

[66] [66, 68] [116]

[120]

307

Continued

Material Performance in Supercritical Water

Alloy class

308

Continued

Alloy class

Alloy

Temp. ( C)

Water chemistry

Loading mode

Results

References

Nickelbased

600 625, C276, MC

290–550 400

[106] [110]

290–550 400–550

slow strain rate test 4  107 s1 slow strain rate test 2.78  106 s1 slow strain rate test 4  107 s1 slow strain rate test 3  107 s1

No IGSCC IGSCC for 625, C276

600, 690, 625 625, 690

IGSCC 625 at 550  C IG at all temperatures

[107] [122]

690, 690GBE

500

8000 ppb DO, 0.06 mS cm1, 25 MPa 8000 ppb DO, 25 MPa, up to 0.01 mol l1 HCl, up to 3.6 MPa H2 8000 ppb DO, <0.1 mS cm1, 25 MPa Deaerated (<10 ppb DO), <0.1 mS cm1, 25 MPa <10 ppb DO, <0.1 mS cm1, 25 MPa

slow strain rate test 3  107 s1

[115]

690, 718 N06625, N06030

400 390–500

Aerated SCW, 25 Mpa 22.5–25 MPa, 10 wt% H2O2, up to 1.6 g l1 NaCl

690 (0, 7 dpa) 690 (0, 7 dpa) 12-Cr-1Mo1WVNb T91, T92

400, 500 400, 500 290, 550

<10 ppb DO, <0.1 mS cm1, 25 Mpa <10 ppb DO, <0.1 mS cm1, 25 Mpa 8000 ppb DO, <0.1 mS cm1, 25 Mpa

slow strain rate test 1  106 s1 slow strain rate test 103 s1, 5  107 s1 Constant load – 100% sy, 140% sy slow strain rate test 3  107 s1 slow strain rate test 3  107 s1 slow strain rate test 4  107 s1

IG reduced in GBE condition IGSCC in 718 IGSCC in 625

IG at both temperatures IG at both temperatures No IGSCC

[117, 121] [118] [107]

370–600

slow strain rate test 0.8– 3  107 s1, fatigue CGR slow strain rate test 3  107 s1

No SCC

400

<10–600 ppb, <0.1–171.1 mS cm1, 25 MPa <10 ppb DO, <0.1 mS cm1

[73, 83, 120] [124]

400, 500, 600

10, 100, 300 ppb DO, 25 MPa

slow strain rate test 3  107 s1

370, 500

<10 ppb DO, <0.1 mS cm1 inlet, 25 MPa

Corrosion fatigue

400, 500

<10 ppb DO, <0.1 mS cm1 inlet, 25 MPa 8000 ppb DO, <0.1 mS cm1, 25 MPa 100–150 ppb, <0.1 mS cm1, 25 MPa 8000 ppb DO, 7.8 MPa

slow strain rate test 3  107 s1

IG in HT-9 only, increase with irradiation, temp, DO Fatigue CGR increased due to fatigue-oxidation interaction IG cracking in all conditions TGSCC No SCC No SCC

Ferritic– martensitic

T91, T91 CSL HCM12A, HT-9 (0–10 dpa) T91, T91 CSL, HCM12A, HT-9 (0, 7 dpa) T91

Ti Alloy ODS

HT-9, HT-9 CSL (0, 7 dpa) Ti-15Mo-5Zr-3Al PM2000 19Cr, 19Cr-4.5Al

290, 550 500, 650 288

DO, dissolved oxygen; SSRT, slow strain rate test; IG, intergranular; CGR, crack growth rate test.

slow strain rate test 4  107 s1 slow strain rate test 3  107 s1 slow strain rate test 104 to 3  107 s1

IG in HT-9 only

[123] [116]

[59, 125] [85] [126] [107] [113] [64]

Material Performance in Supercritical Water

Table 2

Material Performance in Supercritical Water

1.2

Maximum stress (MPa)

600 500 400

IGSCC fraction water density (mg m–3)

304 Was et al.(84,111) 304(61,106,108) 316L(84,106,111,122) 316L(108) 316(61,112) 310(109) 347H(113) 1.497(113) BGA4(113)

300 200

400

450

500 550 600 Temperature (°C)

650

700

Figure 38 Maximum stress versus temperature in constant extension rate tensile tests on austenitic stainless steel in supercritical water. Solid symbols denote solutionannealed samples in deaerated supercritical water, and open symbols denote sensitized samples in 8000 ppb dissolved oxygen.

50

Failure strain (%)

25

0.8

20 Water density

30

20

304(84,111) 304(61,108) 316L(84,111,122) 316L(108) 316(61,112) 347H(113) 1.497(113) BGA4(113)

10

400

450

500

550

600

650

700

Temperature (⬚C) Figure 39 Strain to failure versus temperature in constant extension rate tensile tests on austenitic stainless steel in supercritical water. Solid symbols denote solution-annealed samples in deaerated supercritical water, and open symbols denote sensitized samples in 8000 ppb dissolved oxygen.

has been argued that cracking on the gauge section is a more accurate measure of SCC susceptibility, both measures of IG cracking propensity, %IG on the fracture surface and crack density on the gauge surface, are presented.

Crack density

0.6

15 10

0.4 IGSCC fraction

5

0.2

250

300

350 400 450 500 Temperature (°C)

550

0 600

Figure 40 Intergranular stress corrosion cracking fraction, crack density and water density of 304 stainless steel strained at 4  107 s1 in supercritical water containing 8000 ppb oxygen at a pressure of 25 MPa. Reproduced from Tsuchiya, Y.; Kano, F.; Sato, N.; Shioiri, A.; Moriya, K.; Kasahara, S. SCC properties of metals under supercritical-water cooled power reactor conditions. Corrosion 2004, NACE International, Houston, TX, 2004; Paper No. 04485.

5.12.3.1

40

0 350

1

0 200

100 0 350

30 Sensitized 304

Crack density (cracks mm–2)

700

309

Austenitic Stainless Steels

Austenitic stainless steels that have been tested in pure SCW include 304, 310, 316, 316L, 347H, D9, 1.4970, BGA4, UNS S31266, and 800H. All of these alloys have experienced some degree of SCC, most of which is intergranular. Figures 38 and 39 show the maximum stress and failure strain, respectively, for several stainless alloys tested in CERT mode in SCW. Note that there is a general decrease in the maximum stress with temperature, but the failure strain does not show a clear trend. These datasets consist of essentially two groups, those for CERT tests conducted on sensitized alloys in 8000 ppb dissolved oxygen and a set consisting of solution-annealed samples tested in deaerated (<10 ppb O2). 5.12.3.1.1 Temperature

Plots of the extent of IG cracking versus temperature are shown in Figures 40 and 41. Tsuchiya et al.61 noted that IGSCC susceptibility measured as %IG on the fracture surface of sensitized 304 stainless steel in 8000 ppb O2 dropped from 100% at 290  C to 0% above 400  C (Figure 40). However, they also noted that over the same temperature regime, there was an increase in the density of cracks on the gauge surface. Teysseyre and Was111 noted that in annealed 304 and 316 L stainless steel in deaerated SCW,

310

Material Performance in Supercritical Water

100

350

80 2500

304 60

2000 40

316L 1500

20

Maximum crack depth (µm)

304 Crack density (cracks mm–2)

Crack length per unit area (µm mm–2)

3000

400

(a)

450 500 550 Temperature (°C)

304 316L

250 200 150 100 50

316L 1000 350

300

0 350

0 600 (b)

400

450 500 550 Temperature (°C)

600

Figure 41 (a) Crack length/unit area and crack density versus temperature, and (b) crack depth versus temperature for 304 and 316L stainless steel strained at 3  107 s1 in deaerated supercritical water at 25 MPa. Reproduced from Teysseyre, S.; Was, G. S. Corrosion 2006, 62(12), 1100–1116.

Crack growth rate (mm s–1)

10-6 304 316L 625 690

10-7

Q = 105 kJ mol–1 Q = 85 kJ mol–1 Q = 84 kJ mol–1 Q = 87 kJ mol–1

10-8

10-9 1.2

1.25

1.3

1.35

1.4

1.45

1.5

1/T (10-3 K-1)

Figure 42 Arrhenius behavior of crack growth rates and activation energy for cracking for austenitic stainless steels and nickel-based alloys tested in pure, deaerated (<10 ppb O2) supercritical water. Rates were determined from crack depth and test time in constant extension rate tensile experiments. Reproduced from Teysseyre, S.; Was, G. S. Corrosion 2006, 62(12), 1100–1116.

the crack density and crack length per unit area decreased with temperature, but the crack depth showed a significant increase with temperature (Figure 41). The IG amounted to only a few percent. Crack depth measurements permitted determination of the activation energy for crack growth, which was found to be in the range 85–105 kJ mol1 for the stainless steel alloy, as shown in Figure 42. The cracking mechanism is discussed in more detail in the section on nickel-based alloys. Examples of

cracks in 304 and 316L stainless steel in 550  C SCW are shown in Figure 43. Tests conducted in 500  C SCW containing 100–150 ppb oxygen on 316NG, 347H, and 1.4970 stainless steel showed no clear evidence of SCC on the fracture surface, but cracks were evident on the side surface of all alloys.113 Both TG and IG cracks were identified on the 316NG stainless steel sample. Austenitic alloy BGA4 showed IGSCC on both fracture and side surfaces. Tests in 650  C SCW showed IGSCC and TGSCC on the fracture surface, but no cracks were observed on the side surface. A single study was conducted on the CGR of 316L stainless steel in SCW. Experiments were conducted on 20% cold-worked 316L stainless steel in deaerated SCW from 288 to 500  C at a constant K value of 25 MPa m1/2.114 CGR increased with temperature in the subcritical regime between 288 and 360  C. In the supercritical regime, the rate decreased with temperature from 400 to 450  C and then to 500  C, as shown in Figure 44. Analysis of the fracture surface using SEM revealed that SCC in both subcritical and SCW was intergranular. Comparison of the crack length from SEM and DC potential drop (DCPD) showed that DCPD measurements of crack length are within 30% of that measured from the fracture surface. Crack blunting by rapid oxidation in the supercritical regime during a CGR test, as opposed to film rupture by the higher strain rate of a CERT test, may explain the apparent contradiction of the temperature dependence of cracking in CGR versus CERT tests.

Material Performance in Supercritical Water

311

(b)

(a) 304

316

100 mm

Det SE T304-550

100 mm

Magn Det 250x SE T316-550

(c)

(d) 690

625

Magn Det WD 250x SE 9.7 T625-550

100 mm

Acc.V Spot Magn 15.0 kV 5.0 2000x

Det WD BSE 10.0

10 mm

Figure 43 Micrographs of crack morphologies on cross-sections of samples tested in pure, deaerated supercritical water at 550  C, (a) 304 stainless steel, (b) 316L stainless steel, (c) Alloy 625 and 500  C (d) Alloy 690. Reproduced from Teysseyre, S.; Was, G. S. Corrosion 2006, 62(12), 1100–1116.

Very limited tests have been conducted on Alloy 800H. Alloy 800H was tested in CERT mode in 500  C SCW containing 500 ppb dissolved oxygen.120 The sample failed at about 38% strain and exhibited evidence of brittle-type fracture over a portion of the fracture surface. 5.12.3.1.2 Microstructure

Some work has been conducted to understand the role of grain boundary character on cracking behavior.115 Both 316L stainless steel and Alloy 690 were given thermomechanical treatments to increase the fraction of special boundaries, and then were tested in CERT mode in 500  C water. Results showed that the fraction of cracked grain boundary length in the specimens with higher fractions of special boundaries is reduced for 316L stainless steel and Alloy 690 by factors of 9 and 5 at 15% strain, and 3 and 2 at 25% strain, respectively. This reduction is due to the special boundaries which, at 35% strain, have a

probability of cracking that is 9–18 times lower than that for a random high angle boundary (RHAB). Figure 45 shows the probability of crack occurrence at special boundaries and RHABs following straining to 15% and 25% in 500  C SCW. Note the very large difference in probability at both strains, regardless of the special boundary fraction of the sample. Fine grain 310 stainless steel, tested at 290 and 550  C in SCW containing 8000 ppb oxygen, showed no cracking on either the fracture surface or the side surface.109 However, the authors do not report the grain size. 5.12.3.1.3 Water chemistry

Several studies were conducted on the effect of chemical additions to SCW during the development of SCWO processes; chemicals added include H2SO4, HCl, NaCl, and H2O2. The addition of HCl to 400  C SCW containing 8000 ppb dissolved oxygen resulted in a decreased strain to failure and increased IGSCC susceptibility in 316 stainless

312

Material Performance in Supercritical Water

10-6

1

Crack growth rate (mm s–1)

0.8 10-7 0.6

0.4

CGR deaerated CGR in 2 ppm O2 water

-8

10

0.2

10-9 250

300

350

400

450

500

Oxide weight gain rate (mg dm–2 day–1) water density (g cm–3)

Water density Oxidation rate

0 550

Temperature (⬚C) Figure 44 Crack growth rate versus temperature across the subcritical–supercritical line for a 0.5T CT specimen of unsensitized type 316L stainless steel in pure water. Redrawn from Peng, Q. J.; Teysseyre, S.; Andresen, P. L.; Was, G. S. Corrosion 2007, 63, 1033–1041. Water density is adapted from: W. Wagner, A. Prub, J. Phys. Chem. Ref. Data 31 (2002), p. 327, and oxide weight gain rate over a period of 500 hr is from: G. S. Was, S. Teysseyre, Z. Jiao, Corrosion 62 (2006), p. 989.

0.12 15% strain

Cracking probability

0.1

25% strain

Special RHAB

0.08

0.06

0.04

690GBE

690

316LGBE

316L

690GBE

690

316LGBE

0

316L

0.02

Figure 45 Probability of crack occurrence at special boundaries and random high angle boundaries following straining to 15% and 25% in 500  C supercritical water. Reproduced from West, E. A.; Was, G. S. J. Nucl. Mater. 2009, 392, 264–271.

steel over the range 0.001–0.01 mol l1.110 Addition of 0.01 M H2SO4 to SCW resulted in a sharp reduction in strain to failure in solution-annealed 316L stainless steel and severe intergranular cracking.66,68 Two other austenitic stainless steels, ‘superaustenitic stainless steel’ UNS S 31266 and 347H, were tested in SCW with addition of oxidizers. UNS S 31266 was tested at 400 or 450  C in SCW with additions of 10 wt% H2O2, 1 g l1 HCl, or 10 wt% H2O2 þ 1.6 g l1 NaCl in both CERT mode (5  107 s1) and CLT (100% of the elastic limit).116 The CERT test in 400  C SCW þ 10 wt% H2O2 resulted in SCC, but the CLT test in the same environment at 450  C showed no evidence of cracking. However, CLT tests in 10 wt% H2O2 þ 1.6 g l1 NaCl at 450  C and in 1 g l1 HCl at 400  C resulted in cracks ranging from 5 to 130 mm in length 316 and 347H stainless steel were exposed to 732  C, 34.5 MPa SCW and stressed to 90% of the stress required to cause rupture in 1000 h (103 and 83 MPa, respectively).90 After 168 h of exposure, the 316 stainless steel sample had failed and exhibited several small cracks in addition to the crack that caused failure.

10

0.35

8

0.3

6

0.25 0.2 20

4

30

40

50

60

2 70

Pressure (MPa) Figure 46 Failure strain and dielectric constant versus pressure for 316 stainless steel strained at 2.8  106 s1 in supercritical water containing 8000 ppb oxygen. Reproduced from Watanabe, Y.; Abe, H.; Daigo, Y. Environmentally assisted cracking of sensitized stainless steel in supercritical water: Effects of physical property of water. Proceedings of GENES4/ANP2003, Kyoto, Japan, Sept 15–19, 2003; Paper No. 1183.

5.12.3.1.4 Pressure/dielectric constant

Watanabe and coworkers66,68,112 conducted a systematic study of the effect of SCW pressure on the cracking propensity of sensitized 316 stainless steel in pure water containing 8000 ppb dissolved oxygen. Varying the water pressure from 25 to 60 MPa at 400  C, they observed a monotonic decrease in strain to failure and maximum stress and an increasing amount of IG fracture with pressure, or dielectric constant, as shown in Figure 46. By 60 MPa, the failure was almost totally intergranular. Watanabe suggested that at high pressure, the higher dielectric constant has several effects; it favors the formation of metal ions, tends to increase the solubility of metal oxides, and increases the electrical conductivity of the water, resulting in enhanced anodic dissolution and greater cracking propensity. The reverse is true under the low-density, gas-like conditions of SCW at low pressure. 5.12.3.1.5 Irradiation

Irradiation of 304 and 316L stainless steel significantly increased the IG cracking propensity at both 400 and 500  C. Samples irradiated with 2 MeV protons showed significant increases in the amount of IG cracking with

8000

120 Crack length per unit area

6000 80

5000 %IG

4000

60

3000

40 Crack depth

0

2000

400 °C SCW 316L 7 dpa

20

0

1

(a)

(b)

7000

100

120

2

3 4 5 Dose (dpa)

6

7

1000 8

0

8000

Crack length per unit area

7000

100

6000

%IG 80

5000 4000

60 Crack depth

3000

40

2000 20

Crack length per unit area (µm mm–2)

0.4

313

316L

1000

7 dpa

Crack length per unit area (µm mm–2)

12 Dielectric constant of water

Failure strain (%)

0.45

14 Sensitized 316 400 °C

Crack depth (µm) + %IG on fracture surface

0.5

Crack depth (µm) + %IG on fracture surface

Material Performance in Supercritical Water

0 0 380 400 420 440 460 480 500 520 SCW temperature (°C)

Figure 47 Crack depth, crack length per unit area and %IG on the fracture surface (a) versus dose, and (b) versus temperature for 316L stainless steel strained at 3  107 s1 in deaerated supercritical water at 25 MPa. Reproduced from Zhou, R.; West, E. A.; Jiao, Z.; Was, G. S. J. Nucl. Mater. 2009, 395, 11–22.

dose, as measured by either crack length per unit area, crack depth, or %IG on the fracture surface, as shown in Figure 47.117,118 The crack length per unit area and the %IG cracking were greater at 400  C than at 500  C, but the behavior of the crack depth was reversed. The degree of cracking increased with hardness at 400  C (Figure 48(a)), and there appears to be a correlation with the grain boundary chromium concentration (Figure 48(b)). Austenitic D9 stainless steel and Alloy 800H also showed increased susceptibility to IGSCC following irradiation.121 The intergranular crack length per unit area normalized to strain showed that irradiation has a very similar effect on D9 and

314

Material Performance in Supercritical Water

7500

Crack length per unit area (µm mm–2)

316L, 400 °C

316L, 400 °C

7000

6500

6000

5500

5000 150

200

(a)

250 Hardness (Hv)

300

350 10 (b)

12 14 16 18 Grain boundary Cr concentration (wt%)

20

Figure 48 Dependence of intergranular cracking on (a) hardness and (b) grain boundary chromium content for unirradiated and irradiated (2–7 dpa at 400  C) 316L stainless steel and strained at 3  107 s1 in 400  C deaerated supercritical water at 25 MPa. Reproduced from Zhou, R.; West, E. A.; Jiao, Z.; Was, G. S. J. Nucl. Mater. 2009, 395, 11–22.

Crack length per unit area normalized to strain to failure (mm mm–2)

300 7 dpa Unirradiated 250

D9

316L

200

150

100

50

690

800H

0 Figure 49 Crack length per unit area measurements on unirradiated and 7 dpa proton irradiated alloys D9, 316L, 690, and 800H strained in 400  C supercritical water. Reproduced from West, E. A.; Teysseyre, S.; Jiao, Z.; Was, G. S. In Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Canadian Nuclear Society; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, ON, 2007.

316L stainless steel (Figure 49). Cracking susceptibility in Alloy 800H increased with irradiation, but the magnitude of the cracking is considerably lower than that in 316L and D9 stainless steel.

Japanese Prime Candidate Alloy ( JPCA) was tested in SCW after having undergone neutron irradiation to doses between 26.9 and 43.9 dpa at temperatures ranging from 390 to 520  C in the Fast Flux Test Facility (FFTF).119,132 This alloy is similar to a Tistabilized 316 stainless steel. Overall, samples were highly susceptible to IG cracking. Figure 50 shows the fracture surface of a cold-worked sample irradiated to 26.9 dpa at 390  C after straining to failure in 400  C SCW. The sample failed at <5% total strain and exhibited considerable IG cracking. Figure 51 shows a stress–strain plot of samples irradiated to 33.2 dpa at 520  C and tested in either Ar gas or deaerated SCW at 400  C. The sample tested in SCW exhibits considerably lower strain to failure and is characterized by significant IGSCC, which is absent in the sample tested in Ar. As such, the observed IG cracking is attributed to IGSCC. Results also showed that whether hardening is due to cold-work or irradiation, cracking was higher for the cases with higher hardness (Figure 52). Experiments were also conducted on the effect of water density and temperature on cracking. Samples tested in either 385  C SCW at 27.6 MPa or 400  C SCW at 23.4 MPa cracked more at the higher temperature/lower pressure condition, regardless of the dose and temperature at which they were irradiated, as shown in Figure 53. These results indicate that temperature is a much more important factor than water density in irradiated samples. Finally, a set of

Material Performance in Supercritical Water

X50 500 mm

315

X250 100 mm

Figure 50 Fracture surface of alloy Japanese Prime Candidate Alloy in the cold-worked condition and following irradiation to 26.9 dpa at 390  C in fast flux test facility, and strained at a rate of 3  107 s1 in 400  C deaerated supercritical water at 25 MPa. Adapted from Teysseyre, S.; Was, G. S. Stress corrosion cracking of neutron-irradiated steel in supercritical water. In 13th International Conference on Degradation of Materials in Nuclear Power Systems – Water Reactors; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, 2007; Was, G. S.; Teysseyre, S. Stress corrosion cracking and corrosion of candidate alloys for the supercritical water reactor; Final Report to U. S. Department of Energy, Idaho National Laboratory, project #0003852, Sept 30, 2008.

1000

Stress (MPa)

800 SCW

Argon

600

400

200

0 0

2

4 6 Strain (%)

8

10

Figure 51 Stress–strain behavior of Japanese Prime Candidate Alloy irradiated to 33.2 dpa in fast flux test facility at 520  C, and strained in either argon or deaerated supercritical water at 400  C and at a strain rate of 3  107 s1 in deaerated supercritical water at 25 MPa. The sample strained in supercritical water failed in the pinhole by severe intergranular stress corrosion cracking. Adapted from Teysseyre, S.; Was, G. S. Stress corrosion cracking of neutron irradiated steel in supercritical water. In 13th International Conference on Degradation of Materials in Nuclear Power Systems – Water Reactors; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, 2007; Was, G. S.; Teysseyre, S. Stress corrosion cracking and corrosion of candidate alloys for the supercritical water reactor; Final Report to U. S. Department of Energy, Idaho National Laboratory, project #0003852, Sept 30, 2008.

experiments was conducted to assess the effect of hydrogen addition to the water. It was found that addition of 500 ppb H to 400  C SCW at 23.4 MPa was very effective in suppressing the degree of IGSCC (Figure 54).

In general, austenitic alloys were observed to be susceptible to IGSCC in SCW, where the extent of susceptibility varied with both alloy and water chemistry conditions. Higher temperatures resulted in greater cracking on the gauge section, but a reduction in the %IG on the fracture surface. Overall, 304 stainless steel exhibits greater susceptibility than does 316 and additions of small amounts of acid to the solution increase the susceptibility. Increasing SCW pressure favors IG cracking in unsensitized 316 stainless steel in pure water þ 8000 ppb oxygen. Irradiation greatly increases susceptibility to IGSCC in SCW at both 400 and 500  C. 5.12.3.2

Nickel-Based Alloys

A total of eight nickel-based alloys have been tested for SCC susceptibility in SCW (see Table 2). The tests were performed at temperatures ranging from 390 to 550  C and pressures ranging from 22.5 to 25.5 MPa. The dissolved oxygen content was maintained either in the deaerated condition or at 500 or 8000 ppb. The water chemistry was altered in some tests by the addition of HCl, H2O2, NaCl, or H2SO4. Of the eight alloys tested, only the MC Alloy110 and the UNS N06030 alloy116 did not show evidence of SCC. 5.12.3.2.1 Alloy

Alloy 718 proved to be extremely susceptible to SCC.123 A CERT test conducted on a sample of Alloy 718 in aerated (8000 ppb) SCW at 400  C failed at a strain of only 9% and a maximum stress of

316

Material Performance in Supercritical Water

25

16 427 °C–43.9 dpa specimen

14

SA: 296 Hv

12

CW: 334 Hv

385 ⬚C:27.6 MPa 400 ⬚C:23.4 MPa

20

15 % IG

%IG

10 8

10

6 4

5 2 0

400 ⬚C deaerated

(a)

385 ⬚C deaerated

400 ⬚C hydrogenated

50 368 Hv

40

%IG

30

20

346 Hv 334 Hv

10

SA:407 ⬚C:41.1 dpa MPa

SA:427 ⬚C:43.9 dpa

CW:427 ⬚C:43.9 dpa

Alloy

Tested environments 400 °C deaerated SCW

0

296 Hv

Figure 53 Effect of water density and temperature on cracking of Japanese Prime Candidate Alloy in 385  C:27.6 MPa supercritical water and 400  C:23.4 MPa. Adapted from Teysseyre, S.; Was, G. S. Stress corrosion cracking of neutron-irradiated steel in supercritical water. In 13th International Conference on Degradation of Materials in Nuclear Power Systems – Water Reactors; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, 2007; Was, G. S.; Teysseyre, S. Stress corrosion cracking and corrosion of candidate alloys for the supercritical water reactor; Final Report to U. S. Department of Energy, Idaho National Laboratory, project #0003852, Sept 30, 2008.

0 SA:407 °C:41.1 dpa SA:427 °C:43.9 dpa

(b)

CW:390 °C:26.9 dpa CW:427 °C:43.9 dpa

Alloy

Figure 52 Influence of hardening on cracking in Japanese Prime Candidate Alloy. Cracking increases with hardening whether hardening is due to (a) cold-work on specimen irradiated at 427  C to 43.9 dpa and tested in different environments, or (b) irradiation condition. Adapted from Teysseyre, S.; Was, G. S. Stress corrosion cracking of neutron-irradiated steel in supercritical water. In 13th International Conference on Degradation of Materials in Nuclear Power Systems – Water Reactors; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, 2007; Was, G. S.; Teysseyre, S. Stress corrosion cracking and corrosion of candidate alloys for the supercritical water reactor; Final Report to U. S. Department of Energy, Idaho National Laboratory, project #0003852, Sept 30, 2008.

1300 MPa, and the fracture surface was completely intergranular fracture. The authors attributed this cracking behavior to the oxidation and swelling of the primary niobium carbides that cause cracks to initiate at these carbides.

CERT tests on Alloy 690 have produced mixed results. Fournier et al.123 tested Alloy 690 in 400  C, 25 MPa SCW under aerated conditions and found that Alloy 690 failed by completely ductile rupture and showed a significant amount of necking. Companion tests in air revealed behavior for each alloy that was similar to that in SCW at 400  C. Was and coworkers84,111,122 found IGSCC in Alloy 690 at 400, 450, 500, and 550  C, with the extent of cracking increasing with temperature (see Figure 41). Alloy 600 was tested at 290, 380, and 550  C in water containing 8000 ppb dissolved oxygen.108 The two lower temperature experiments failed by ductile fracture with no sign of SCC. The sample at 550  C exhibited a crack density of about 26 cracks mm2. MC Alloy was tested in pure SCW at 400  C and 8000 ppb and was also tested with the addition of 0.001 mol l1 HCl and 0.01 mol l1 HCl. No cracks were observed on the specimen and there was no reduction in the strain to failure when the HCl additions were made.110

317

14

H-SCW Deaerated-SCW

12

%IG

10 8 6 4 2

10 000

600 625

8000

500 400

6000

625 300

4000 200 2000

690 690

0 350

400

(a)

450

500

550

100

Crack density (cracks mm–2)

16

Crack length per unit area (mm mm–2)

Material Performance in Supercritical Water

0 600

Temperature (⬚C) 100

CW SA 427 ⬚C–43.9 dpa sample’s condition Figure 54 Influence of 500 ppb hydrogen addition on cracking of Japanese Prime Candidate Alloy condition SA:427  C:43.9 dpa in 400  C supercritical water. Adapted from Teysseyre, S.; Was, G. S. Stress corrosion cracking of neutron-irradiated steel in supercritical water. In 13th International Conference on Degradation of Materials in Nuclear Power Systems – Water Reactors; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, 2007; Was, G. S.; Teysseyre, S. Stress corrosion cracking and corrosion of candidate alloys for the supercritical water reactor; Final Report to U. S. Department of Energy, Idaho National Laboratory, project #0003852, Sept 30, 2008.

5.12.3.2.2 Temperature

Figure 55 shows the temperature dependence of IG cracking in nickel-based Alloy 625 and Alloy 690 compared to that for the stainless steels. Both alloy systems show a strong dependence on temperature, with the extent of cracking rising exponentially with temperature. Figure 43 shows that the maximum IG crack depth obeys an Arrhenius behavior, with an activation energy between 84 and 87 kJ mol1. These values are within the range of activation energies for cracking of nickel-based Alloy 600 in lowpotential primary water in PWRs, for which the activation energy for CGR in the temperature range of 310–420  C is between 80 and 220 kJ mol1.129 The low value of the activation energy for crack growth compared to oxide growth (200 kJ mol1)133 may indicate either a role of aggressive species in the water or a short-circuit growth path, for example, grain boundary oxidation ahead of the growing crack. Besides a slip-oxidation mechanism, the latter mode could also occur by selective internal oxidation (SIO) that has been observed in nickel-based alloys,

Maximum crack depth (mm)

0

80 60 40 20 0 350

(b)

690 625

400

450 500 550 Temperature (⬚C)

600

Figure 55 (a) Crack length/unit area and crack density versus T, and (b) crack depth versus T for Alloy 690 and Alloy 625 strained at 3  107 s1 in deaerated supercritical water at 25 MPa. Reproduced from Teysseyre, S.; Was, G. S. Corrosion 2006, 62(12), 1100–1116.

as described by Scott,134 at the upper end of the temperature range used in this study. In fact, the measured CGRs are consistent with the diffusion of oxygen in nickel by Bricknell and Woodford.135 These results show that both slip-oxidation and SIO are possible mechanisms for IGSCC in SCW. Alloy 690 was also tested in both the thermally treated and solution-annealed conditions.136 Samples were tested at 385  C and 27.6 MPa and at 400  C and 25.4 MPa. The water density at 385  C was 465 kg m3 versus 175 kg m3 at 400  C. Results of CERT tests showed that the crack depth was consistently worse, regardless of sample condition, in the higher temperature, lower-density environment (Figure 56(a)). These results indicate that temperature is a more important factor than water density for Alloy 690. It should also be noted that cracking was mainly observed along TiC nodules in the 385  C, 26.7 MPa tests. Alloy 600 in the mill-annealed condition was also tested along with the

318

Material Performance in Supercritical Water

1400 400 ⬚C, 25.4 MPa 175 kg m–3 385 ⬚C, 27.6 MPa 465 kg m–3

8

6

4

2

690(84,111,122) 690(107,123) 625(107,111,122) 625(116) 625(108,110) 600(107,108) 718(123) MC alloy(110) UNS N 10276(110) 800H(120)

1200 Maximum stress (MPa)

Max crack depth observed (µm)

10

1000

800

600

400 0 UM-690 (a)

UM-TT690

EPRI-TT690

200 350

Alloy

400

400 ⬚C, 25.4 MPa 175 kg m–3 385 ⬚C, 27.6 MPa 465 kg m–3 150

600

60

100

50 50

0 EPRI-600 (b)

550

Figure 57 Maximum stress versus temperature in constant extension rate tensile tests on nickel-based alloys in supercritical water.

UM-690 UM-TT690 EPRI-TT690 Alloy

Figure 56 Maximum crack depth observed after failure in 400  C:25.4 MPa and 385  C:27.6 MPa deaerated supercritical water for (a) Alloy 690, and (b) for all alloys including mill-annealed Alloy 600. Reproduced from Materials Reliability Program: Constant Extension Rate Stress Corrosion Cracking Testing of Alloys 600 and 690 in Supercritical Water (MRP-233), Technical Report 1016154, EPRI, Palo Alto, CA, 2007.

Alloy 690 samples136 and was found to be highly susceptible to SCC in SCW, and intergranular cracking was clearly identified (Figure 56(b)). The temperature dependence of the maximum stress and strain to failure for all of the data on nickel-based alloys is shown in Figures 57 and 58. Perhaps because of the wide range of conditions, there is little observable dependence on temperature for either parameter. However, as shown in the preceding paragraph, samples tested within a single program tend to show a cleaner correlation.

Strain to failure (%)

Max crack depth observed (µm)

200

450 500 Temperature (⬚C)

40

30 690(84,111,122) 690(123) 625(107,111,122) 625(108,110) 600(106,107,108) 718(123) MC alloy(110) UNS N 10276(110) 800H(120)

20

10

0 350

400

450 500 Temperature (⬚C)

550

600

Figure 58 Strain to failure versus temperature in constant extension rate tensile tests on nickel-based alloys in supercritical water.

5.12.3.2.3 Water chemistry

As part of a study of materials for use in the upgrading of low-quality hydrocarbons using SCW, Fujisawa et al.110 studied SCC of Alloy C-276, Alloy 625, and MC Alloy in 400  C SCW containing 8000 ppb dissolved oxygen. No IG cracks were observed in Alloy C-276 in pure water, but the addition of small

40

20

0 –20 15

20

25

30 35 Cr content

40

45

50

Figure 59 Effect of the chromium content of various nickel-based alloys on the percentage intergranular stress corrosion cracking in 400  C supercritical water containing 8000 ppb dissolved oxygen and additions of HCl. Reproduced from Fujisawa, R.; Nishimura, K.; Kishida, T.; Sakaihara, M.; Kurata, Y.; Watanabe, Y. Corrosion 2005; NACE International: Houston, TX, 2005; Paper no. 05395.

amounts (0.001 mol l1) of HCl resulted in severe IG cracking. Alloy 625 failed by IGSCC at 0.01 mol l1 HCl, but not at lower concentrations or in deionized water. Bosch and Delafosse116 also found that Alloy 625 cracked intergranularly at both 390 and 450  C in constant load tests in water containing 10 wt% H2O2 þ 1.6 g l1 NaCl. CERT tests in SCW þ 10 wt% H2O2 revealed IG cracking at both 400 and 500  C, with considerably greater cracking at 500  C. No cracks were found in the MC Alloy after CERT tests in 0.01 mol l1 HCl at 400  C and 8000 ppb dissolved oxygen. Alloy UNS N 06030 was tested in constant load mode (80% or 100% of yield strength) in SCW containing either 10 wt% H2O2 or 10 wt% H2O2 þ 1.6 g l1 NaCl at 400–450  C.116 No cracking was observed in the solution containing only hydrogen peroxide and only localized corrosion, but no cracking, was observed in the latter solution. Fujisawa showed that IGSCC susceptibility generally follows the inverse of the alloy chromium content in SCW containing HCl additions, with the austenitic stainless steel alloys exhibiting the greatest amount of IGSCC and the high nickel, MC Alloy the least (Figure 59). However, Teysseyre and Was111 observed that Alloy 625 exhibited the worst IGSCC resistance among 304 and 316L stainless steel, Alloy 625 and Alloy 690 tested in pure, deaerated SCW.

10 000 690 400 ⬚C SCW

25

8000

20 15 4000 10 %IG

2000

5 Crack length per unit area

0 1

2

3

(a)

(b)

6000

Crack depth

4 5 Dose (dpa)

6

7

0 8

10 000

120 100

Crack length per unit area (µm mm–2)

%IGSCC

60

0.001 mol/1 HCl 0.01 mol/1 HCl

30

319

690 7 dpa

8000 Crack length per unit area

80

6000

60 %IG

Crack depth

4000

40 20

2000

Crack length per unit area (µm mm–2)

400 ⬚C 8 ppm DO 25 MPa

Crack depth (µm) + %IG on fracture surface

80

Crack depth (µm) + %IG on fracture surface

Material Performance in Supercritical Water

0 0 380 400 420 440 460 480 500 520 SCW temperature (⬚C)

Figure 60 Crack length per unit area, crack depth and %IG versus (a) dose, and (b) temperature for Alloy 690 irradiated to 7 dpa at 400  C and tested in pure supercritical water at a strain rate of 3  107 s1. Reproduced from Zhou, R.; West, E. A.; Jiao, Z.; Was, G. S. J. Nucl. Mater. 2009, 395, 11–22.

5.12.3.2.4 Irradiation

Irradiation effects on nickel-based alloys are confined to Alloy 690 and both 400 and 500  C with 2 MeV protons to doses of 2, 4, and 7 dpa.117,118 As shown in Figures 49 and 60, the crack length per unit area, crack depth, and %IG fracture, all increase with dose for CERT tests conducted in 400  C SCW. The effect of irradiation on cracking is greatest at 500  C, where the crack length per unit area is 40 times greater on the irradiated side of the sample than on the unirradiated side. The nickel-based alloys exhibit much the same dependence of cracking on key experimental parameters

320

10

0

0

400 ⬚C 500 ⬚C Ar 500 ⬚C

500 ⬚C

600 ⬚C

+300 ppb O2

as do the austenitic stainless steels. However, cracking is generally more severe in some of the alloys such as Alloy 625. Chromium content appears to be a significant factor in cracking in SCW containing low concentrations of HCl. Also, irradiation has a greater effect on cracking in Alloy 690 than it does on 316L stainless steel. 5.12.3.3

Ferritic–Martensitic Alloys

Six F/M alloys have been tested in SCW to date: T91, T91a, T92, T92a, HCM12A, and HT-9. Of these alloys, the only one that showed susceptibility to SCC was HT-9.59,73,83,108,120,125 The crack densities for HT-9 (2–18 mm2) were generally much lower than the crack densities for the austenitic stainless steels and Ni-based alloys. Tests on F/M alloys were conducted in pure water with dissolved oxygen contents ranging from the deaerated condition (<10 ppb) to 500 ppb and temperatures ranging from 400 to 600  C. The strains to failure of the F/M alloys were much lower than those of the austenitic alloys, ranging from 10.9% to 24.4%. One CERT test was performed on a sample of T91 in argon at 500  C and the strain to failure was still only 15%. T91a was also tested as U-bends at 500 and 550  C and showed no evidence of cracking. HT-9 shows increasing susceptibility to cracking with temperature and dissolved oxygen content. Figure 61 shows that both the crack density and

500 ppb118

<10 ppb73,118

<10 ppb73,118 <10 ppb59,137

300 ppb59,137

10

5

0 HCM12A

HT-9

T91

T92

Figure 62 Strain to failure versus oxygen content in ferritic–martensitic alloys tested in constant extension rate tensile mode in supercritical water at 500  C.

600 HCM12A(59) HT-9(35,124) T91(59) T91(120) T92(120) PM2000(113)

500 Maximum stress (MPa)

Figure 61 Effect of temperature and dissolved oxygen content on crack density and crack depth in HT-9 tested in constant extension rate tensile mode in nominally deaerated supercritical water or in 500  C Ar. Reproduced from Ampornrat, P.; Gupta, G.; Was, G. S. J. Nucl. Mater. 2009, 395, 30–36.

300 ppb59,137

10

15

<10 ppb59,137

20

20

20

300 ppb59,137

30

30

25 <10 ppb59,137

40

Strain to failure (%)

40

500 ⬚C 25 MPa

100 ppb59,137

30

50 Crack density Maximum crack depth

Maximum crack depth (µm)

Crack density (cracks mm–2)

50

100 ppb73,118

Material Performance in Supercritical Water

400

300

200

100

0 350

400

450

500 550 600 Temperature (⬚C)

650

700

Figure 63 Maximum stress versus temperature for ferritic–martensitic alloys. Data points at 500  C are slightly shifted horizontally so that all symbols could be discerned.

crack depth increase with temperature between 400 and 600  C, and with dissolved oxygen concentrations between 10 and 300 ppb.125 Cracking is aggravated by both high temperature and dissolved oxygen in SCW. Although cracks were not observed on any of the T91, T91a, T92, T92a, or HCM12A samples, it is important to note that they did show a substantially lower strain to failure with increasing dissolved

Material Performance in Supercritical Water

14

50 HCM12A(59) HT-9(35,124) T91(59) T91(120) T92(120) PM2000(113)

12 Maximum crack depth (mm)

Strain to failure (%)

40

30

20

10

500 ⬚C 8 6 4

0

400

450

500

550

600

650

700

Temperature (⬚C) Figure 64 Strain to failure versus temperature for ferritic– martensitic alloys. Data points at 500  C are slightly shifted horizontally so that all symbols could be discerned.

10–6 Fatigue crack growth rate, da/dN (m per cycle)

400 ⬚C

10

2

0

350

321

Air Water

10–7

300

400 500 Temperature (⬚C)

600

700

Figure 65 Fatigue crack growth rates of T91 in air and water at DK  24 MPa m1/2 as a function of temperature. Reproduced from Yi, Y.; Lee, B.; Kim, S.; Jang, J. Mater. Sci. Eng. A 2006, 429(1–2), 161–168.

oxygen concentration in SCW.73,83,124 Figure 62 shows the effect of oxygen concentration on strain to failure at 500  C. The maximum stress and strain to failure are plotted as a function of temperature in Figures 63 and 64, respectively. Note that the maximum stress decreases and the strain to failure increases with

Unir

Irr

Unir

Irr

Figure 66 Effect of irradiation on intergranular crack depth in HT-9 tested in 500  C, deaerated supercritical water. The temperatures given in the figure are the irradiation temperatures. Data from Gupta, G.; Was, G. S. Effect of proton irradiation and grain boundary engineering on stress corrosion cracking of ferritic–martensitic alloys in supercritical water. Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Minerals, Materials and Metals Society: Warrendale, PA, 2005; pp 1359–1367.

temperature. Such a temperature dependence would not be observed for alloys that exhibit increasing susceptibility to SCC. Hence, the observed dependence of mechanical properties on temperature is not related to the SCC propensity. Yi et al.85 conducted fatigue crack growth tests on 1T CTsamples of T91 in deaerated water at 370 and 500  C and in air. Tests were conducted in air at a frequency of 1 Hz, a R-ratio of 0.1, and a △K of 23–24 MPa m1/2, and in water at a frequency of 0.0042 Hz, a R-ratio of 0.16, and a △K of 24–25 MPa m1/2. The CGR in water at 500  C was about 1.4 times that in air and the enhancement in cracking was attributed to rapid oxidation (Figure 65). Irradiation was also found to enhance the amount of IG cracking in SCW. HT-9 is known to be susceptible to cracking following irradiation and testing in air.138 Samples of HT-9 were irradiated with protons at either 400 or 500  C to a dose of 7 dpa, and tested in CERT mode in 500  C deaerated SCW. Results showed that crack density and crack depth increased because of irradiation and the increase was greatest at 400  C (Figure 66). It should be noted that irradiation hardening at 400  C is quite significant, but

322

Material Performance in Supercritical Water

30 Crack density Max. crack depth

25

25

20

20

15

15

10

10

5

5

0

Maximum crack depth (mm)

Crack density (cracks mm–2)

30

0 AR-irr

GBE-irr

AR-unirr

GBE-unirr

Alloy condition Figure 67 Crack density and maximum crack depth for HT-9 samples in the as-received (AR) condition and the grain boundary-engineered condition, and the effect of irradiation on cracking propensity for each condition. Irradiation and constant extension rate tensile testing were both done at 500  C. Reproduced from Gaurav, G.; Ampornrat, P.; Ren, X.; Sridharan, K.; Allen, T. R.; Was, G. S. J. Nucl. Mater. 2007, 361(2–3), 160–173.

is only minimal at 500  C.126 Hardening may be controlling the increased susceptibility to IGSCC in HT-9. One additional set of experiments was performed on grain boundary engineered HT-9, in which the fraction of 1 grain boundaries was increased by about 30% compared to the as-received (tempered) condition by a deformation and heat treatment process.126 Figure 67 shows that both the crack depth and crack density in the grain boundary-engineered samples are less than that in the as-received samples irradiated at 500  C and tested in CERT mode in pure, deaerated SCW at 500  C. 5.12.3.4

Other Alloys

Very little data on Ti alloys in SCW exist, though there is a significant database in subcritical water. The only experiment conducted in SCW was on Ti–15Mo–5Zr–3Al in pure SCW at 550  C, 25 MPa, and 8000 ppb dissolved oxygen. The maximum stress was 249 MPa, and the strain to failure was 38%, resulting in a crack density of 26 cracks mm2.107,108 While several experiments have been reported on ODS alloys in subcritical water,63,64 only two experiments have been conducted to date in SCW. ODS steel PM2000 was strained in 500 and 650  C SCW containing 100–150 ppb oxygen at an initial rate of

3  108 s1, and then at 3  107 s1 after about 10% strain had been accumulated.113 The alloy showed no susceptibility to SCC at either temperature. An ODS alloy was also strained in 600  C in deaerated SCW to 4%, at which point the sample broke in the threads. There was no evidence of cracking on the gauge surface. 5.12.3.5

Summary of SCC Behavior in SCW

The following trends were observed in the SCC behavior in SCW. 1. Austenitic stainless steels and nickel-based alloys exhibit susceptibility to IGSCC in pure SCW over the temperature range of 400–650  C. 2. IGSCC susceptibility, as measured by %IG on the fracture surface, decreases with temperature, but as measured by cracking on the gauge section, it generally increases with temperature. 3. The CGR of cold-worked 316L stainless steel increases with temperature up to 400  C and decreases after that because of crack blunting by rapid oxidation. 4. Small additions of HCl or H2SO4 increase the susceptibility to IGSCC in the austenitic alloys. 5. Increasing system pressure caused an increase in the severity of IGSCC in sensitized 316L stainless steel in pure water with 8000 ppb dissolved oxygen. 6. Nickel-based alloys are very susceptible to IGSCC, comparable to austenitic stainless steels. 7. There appears to be an effect of alloy chromium content on SCC cracking propensity in nickelbased alloys in dilute HCl solutions, with higher Cr alloys showing greater resistance to IGSCC. 8. With the exception of HT-9, F/M alloys are resistant to IGSCC in pure SCW up to 600  C. 9. Irradiation strongly enhances IGSCC in pure SCW, increasing both the crack density on the gauge surface and the crack depth, at temperatures between 400 and 500  C.

References 1. Dollezhal’, N. A.; Emel’yanov, I. Ya.; Aleshchenkov, P. I. Development of Power Reactors such as the Reactor at the Beliy Yar Nuclear Power Plant with NuclearSuperheated Steam; Report No. 309 at the 3rd Geneva Convention, 1964. 2. Wright, J. H.; Paterson, J. F. Proc. Am. Power Conf. 1966, XXVIII, 139. 3. Oka, Y.; Koshizuka, S. Nucl. Technol. 1993, 103, 295.

Material Performance in Supercritical Water 4.

5. 6. 7. 8. 9. 10. 11.

12. 13.

14. 15.

16.

17.

18.

19.

20.

21.

Schulenberg, T.; Starflinger, J. European research project on the high performance light water reactor; 4th International Symposium on Supercritical Water-Cooled Reactors, Heidelberg, Germany, Mar 8–11, 2009. Bae, Y. Y.; Jang, J.; Kim, H. Y.; Yoon, H. Y.; Kang, H. O.; Bae, K. M. Nucl. Eng. Technol. 2007, 39, 273. Torgerson, D. F.; Shalby, B. A.; Pang, P. Nucl. Eng. Des. 2006, 236, 1565. A Technology Roadmap for Generation IV Nuclear Energy Systems; Report No. GIF002-00, Dec 1, 2002; http://nuclear.gov. Feasibility Study of Supercritical Light Water Cooled Reactors for Electric Power Production; Final Report, DE-FG07-02SF22533, INEEL/EXT-04-02539, Jan 2005. Chow, C. K.; Khartabil, H. F. Nucl. Eng. Technol. 2007, 40, 139. Buongiorno, J. Massachusetts Institute of Technology, Personal communication, June 2005. Viswanathan, R.; Purgert, R.; Rao, U. Materials for ultrasupercritical coal-fired power plant boilers. Challenges for High Temperature Alloys for Aerospace, Land-Based Gas Turbines, Power & Transportation Symposium, Session 2, High Temperature Alloys in Power Generation, ASM Materials Solutions Conference, Columbus, OH, Oct 18–21, 2004. Erlich, K.; Konys, J.; Heikinheimo, L. J. Nucl. Mater. 2004, 327, 140. Burrill, K. A. Water chemistries and corrosion product transport in supercritical water in reactor heat transport systems. 8th BNES Conference on Water Chemistry of Nuclear Reactor Systems, Bournemouth, UK, Oct 22–26 2000. Kritzer, P. J. Supercritical Fluids 2004, 29, 1. Wesolowski, D. J.; Ziemniak, S. E.; Anovitz, L. M.; Machesky, M. L.; Benezeth, P.; Palmer, D. A. In Aqueous Systems at Elevated Temperatures and Pressures: Physical Chemistry in Water, Steam, and Hydrothermal Solutions; Palmer, D. A., et al. Eds.; Elsevier: Amsterdam, 2004; Chapter 14. Guzonas, D.; Tremaine, P.; Brosseau, F. Predicting activity transport in a supercritical water cooled pressure tube reactor. 4th International Symposium on Supercritical Water-Cooled Reactors, Heidelberg, Germany, Mar 8–11, 2009. Kysela, J.; Sˇvarc, V.; Ru˚zˇicˇkova´, M. Water chemistry specifications for supercritical water cooled reactors – Possible options. 4th International Symposium on Supercritical Water-Cooled Reactors, Heidelberg, Germany, Mar 8–11, 2009. Bartels, D. M.; Anderson, M.; Wilson, P.; Allen, T.; Sridharan, K. Supercritical water radiolysis chemistry. Supercritical water corrosion. Available from http:// nuclear.inel.gov/deliverables/docs/ uwnd_scw_level_ii_sep_2006_v3.pdf. Meesungnoen, J.; Jay-Gerin, J. P.; Guzonas, D. Montecarlo simulations of the radiolysis of water at elevated temperature and in the supercritical regime. International Workshop on Supercritical Water Coolant Radiolysis, June 14–15, 2007; Nuclear Research Institute: Rez, Czech Republic, 2007. Pommeret, S. Radiolysis of supercritical water: An experimental and theoretical challenge. International Workshop on Supercritical Water Coolant Radiolysis, June 14–15, 2007; Nuclear Research Institute: Rez, Czech Republic, 2007. Katsumura, Y. In Charged Particle and Photon Interactions with Matter; Mozumder, A., Hatano, Y., Eds.; Marcel Dekker: New York, NY, 2004; p 697.

22. 23.

24. 25.

26.

27. 28. 29.

30. 31.

32. 33.

34.

35. 36. 37.

38. 39. 40. 41.

42. 43.

323

Ru˚zˇicˇkova´, M.; Ha´jek, P.; Sˇmida, S.; Vsˇolcˇa´k, R.; Petr, J.; Kysela, J. Nucl. Eng. Technol. 2007, 40(2). Special issue on the 3rd International Symposium on SCWR. Was, G. S.; Busby, J. T.; Andresen, P. L. In Corrosion in the Nuclear Industry; Corrosion: Environments and Industries, ASM Handbook; ASM International: Houston, TX, 2006; Vol. 13c, pp 386–414. Was, G. S.; Andresen, P. JOM 1992, 44(4), 8–13. Kriksunov, L. B.; Macdonald, D. D. Corrosion testing and prediction in SCWO environments. Proceedings of the ASME Heat Transfer Division, 1995; p 281, HTD-Vol 317-2. Guzonas, D. A.; Wills, J.; Do, T.; Michel, J. Corrosion of candidate materials for use in a supercritical water CANDUW reactor. 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, 2007; pp 1250–1261. Daigo, Y.; Watanabe, Y.; Sue, K. Corrosion 2007, 63, 277. Ruther, W. E.; Schlueter, R. R.; Lee, R. H.; Hart, R. K. Corrosion 1966, 22, 147. Allen, T. R.; Tan, L.; Chen, Y.; et al. Corrosion and radiation response of advanced ferritic–martensitic steels for Generation IV application. Proceedings of Global, 2005; Paper IL001. Allen, T. R.; Tan, L.; Chen, Y.; et al. Corrosion of ferritic– martensitic alloys in supercritical water for Gen IV application. Proceedings of Global, 2005; Paper 419. Allen, T. R.; Chen, Y.; Tan, L.; Ren, X.; Sridharan, K. Corrosion of candidate materials for supercritical water-cooled reactors. In 12th Environmental Degradation Conference of Materials in Nuclear Power Systems – Water Reactors; Allen, T. R., King, P. J., Nelson, L., Eds.; The Minerals, Metals and Materials Society: Warrendale, PA, 2006; p 1397. Allen, T. R.; Was, G. S. Corrosion 2007, Nashville, TN; NACE International: Houston, TX, 2007; Paper 07RTS9. Allen, T. R.; Sridharan, K.; Chen, Y.; Tan, L.; Ren, X.; Kruizenga, A. Materials development and selection for corrosion resistance: A critical issue in supercritical water reactors. 16th Pacific Basin Nuclear Conference (16PBNC), Aomori, Japan, Oct 13–18, 2008; Paper P16P1032. Sridharan, K.; Zillmer, A.; Licht, J. R.; Allen, T. R.; Anderson, M. H.; Tan, L. Corrosion behavior of candidate alloys for supercritical water reactors. Proceedings of ICAPP 04, Pittsburgh, PA, 2004; p 537; Paper 4136. Tan, L.; Allen, T. R. EPD Congress 2005; Schlesinger, M. E., Ed.; The Minerals, Metals and Materials Society: Warrendale, PA, 2005; p 61. Chen, Y.; Sridharan, K.; Allen, T. R. Corrosion 2005; NACE International: Houston, TX, 2005; Paper 05391. Was, G. S.; Allen, T. R. Time, temperature, and dissolved oxygen dependence of oxidation of austenitic and ferritic–martensitic alloys in supercritical water. Proceedings of ICAPP’05; Paper 5690. Tan, L.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2006, 348, 263. Chen, Y.; Sridharan, K.; Allen, T. Corrosion Sci. 2006, 48(9), 2843. Tan, L.; Yang, Y.; Allen, T. R. Corrosion Sci. 2006, 48, 3123. Sridharan, K.; Tan, L.; Chen, Y.; Ren, X.; Kruizenga, A.; Allen, T. R. In Proceedings of the Materials Science and Technology Annual Conference, Cincinnati, OH; Rebak, R. B., Baker, B. A., Pint, B. A., Eds.; 2006; pp 363–374. Ren, X.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2006, 358, 227. Chen, Y.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2006, 359, 50.

324 44. 45. 46.

47. 48. 49. 50. 51.

52.

53. 54.

55. 56. 57. 58. 59.

60.

61.

62. 63. 64.

65.

Material Performance in Supercritical Water Tan, L.; Yang, Y.; Allen, T. R. Corrosion Sci. 2006, 48, 4234–4242. Chen, Y.; Sridharan, K.; Allen, T. R. Corrosion 2007, Nashville, TN; NACE International: Houston, TX, 2007; Paper 07408. Chen, Y.; Kruizenga, A.; Ren, X.; et al. Progress in understanding corrosion in supercritical water systems. 3rd International Symposium on SCWR-Design and Technology, Shanghai, China, Mar 12–15, 2007; Paper SCR2007-P002. Ren, X.; Sridharan, K.; Allen, T. R. Corrosion 2007, 63, 603. Was, G. S.; Ampornrat, P.; Gupta, G.; et al. J. Nucl. Mater. 2007, 371, 176–201. Tan, L.; Machut, M. T.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2007, 371, 171–175. Chen, Y.; Sridharan, K.; Allen, T. R. J. Nucl. Mater. 2007, 371, 118–128. Chen, Y.; Sridharan, K.; Allen, T. R. In 13th Environmental Degradation of Materials in Nuclear Power Systems 2007; Allen, T. R., Busby, J. T., King, P. J., Eds.; Canadian Nuclear Society: Toronto, ON, 2007. Motta, A. T.; Siwy, A. D.; Kunkle, J. M.; et al. In 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems 2007; Allen, T. R., Busby, J. T., King, P. J., Eds.; Canadian Nuclear Society: Toronto, ON, 2007; pp 1501–1513. Tan, L.; Sridharan, K.; Allen, T. R.; Nanstad, R. K.; McClintock, D. A. J. Nucl. Mater. 2008, 374, 270. Ren, X.; Sridharan, K.; Allen, T. R. Effect of shot peening on the oxidation of ferritic–martensitic steels in supercritical water. In Corrosion 2008; Paper 08425. Nanstad, R. K.; McClintock, D. A.; Hoelzer, D. T.; Tan, L.; Allen, T. Am. Nucl. Soc. Trans. 2008, 98, 1161–1162. Tan, L.; Sridharan, K.; Allen, T. R. Mater. Sci. Forum 2008, 595–598, 409–418. Tan, L.; Ren, X.; Sridharan, K.; Allen, T. R. Corrosion Sci. 2008, 50, 2040–2046. Tan, L.; Ren, X.; Sridharan, K.; Allen, T. R. Corrosion Sci. 2008, 50, 3056. Ampornrat, P.; Bahn, C. B.; Was, G. S. Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Minerals, Metals and Materials Society: Warrendale, PA, 2005; p 1387. Kasahara, S. General corrosion of iron, nickel and titanium alloys as candidate materials for the fuel cladding of supercritical-water cooled power reactor, GENES4/ ANP2003, Kyoto, Japan, Sept 2003; Paper 1132. Tsuchiya, Y.; Kano, F.; Saito, N.; et al. SCC and irradiation properties of metals under supercritical-water cooled power reactor conditions, GENES4/ANP2003, Kyoto, Japan, Sept 2003; Paper 1096. Jang, J.; Han, C. H.; Lee, B. H.; Yi, Y. S.; Hwang, S. S. Corrosion behavior of 9Cr F/M steels in supercritical water. Proceeding of ICAPP’05, Seoul, Korea, 2005; Paper 5136. Cho, H. S.; Kimura, A.; Ukai, S.; Fujiwara, M. J. Nucl. Mater. 2004, 329–333, 387–391. Kimura, A.; Cho, H.; Toda, N.; et al. Fuel cladding materials R and D for high burn-up operation of advanced water-cooling nuclear energy systems. Proceedings of the American Nuclear Society – International Congress on Advances in Nuclear Power Plants 2005, ICAPP’05, Seoul, Korea, May 2005; pp 1737–1745 Paper 5338. Kaneda, J.; Kasahara, S.; Kuniya, J.; et al. General corrosion properties of titanium based alloys for the fuel claddings in the supercritical water-cooled reactor. Proceedings of the 12th International Conference on

66.

67. 68.

69.

70.

71.

72.

73. 74. 75. 76. 77. 78. 79.

80. 81. 82. 83.

84.

Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, 2006; pp 1409–1419. Abe, H.; Watanabe, Y.; Sue, K. Stress corrosion cracking mechanisms of 316 stainless steels in sub-critical liquid water and supercritical water from a ‘‘dielectic constant of water’’ point of view. Proceedings of ICAPP’04, Pittsburgh, PA, June 2004; Paper 4226. Fujisawa, R.; Nishimura, K.; Nishida, T.; Sakaihara, M.; Kurata, Y.; Watanabe, Y. Corrosion 2006, 62(3), 270–274. Watanabe, Y.; Abe, H.; Diago, Y.; Nishida, T. Corrosion resistance and cracking susceptibility of 316L stainless steel in sulfuric acid-containing supercritical water, Corrosion 2004; NACE International: New Orleans, TX, 2004; Paper 04493. Jeong, Y. H.; Park, J. Y.; Kim, H. G.; et al. Corrosion of zirconium-based fuel cladding alloys in supercritical water. Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, 2006; pp 1369–1378. Guzonas, D. A.; Wills, J.; McRae, G. A.; et al. Corrosionresistant coatings for use in a supercritical water CANDUW reactor. Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, 2006; pp 1379–1386. Kaneda, J.; Kasahara, S.; Kuniya, J.; Kano, F.; Takahashi, H.; Matsui, H. Material properties of stainless steels modified with addition of zirconium for supercritical water-cooled reactor. International Congress on Advances in Nuclear Power Plants – ICAPP 2007, 2008; pp 2795–2805. Kaneda, J.; Kasahara, S.; Kuniya, J.; et al. Corrosion film properties of the candidate materials for the fuel claddings of the supercritical-water cooled power reactor. International Congress on Advances in Nuclear Power Plants 2005; ICAPP’05, 2005; pp 2963–2972. Hwang, S. S.; Lee, B. H.; Kim, J. G.; Jang, J. J. Nucl. Mater. 2008, 372(2–3), 177–181. Filippov, G. A.; Grishanin, E. I.; Lebedev, Yu. E.; Trubachev, V. M.; Fal’kovskii, L. N.; Fonarev, B. I. Atom. Energ. 2007, 102(3), 204–210. Daigo, Y.; Watanabe, Y.; Sue, K. Corrosion 2007, 63(12), 1085–1093. Li, Q.; Zhou, B.; Yao, M.; Liu, W.; Chu, Y. Rare Metal Mater. Eng. 2007, 36(8), 1358–1361. Peng, Q.; Gartner, E.; Busby, J. T.; Motta, A. T.; Was, G. S. Corrosion 2007, 63(6), 577–590. Barringer, E.; Faiztompkins, Z.; Feinroth, H.; et al. J. Am. Ceram. Soc. 2007, 90(1), 315–318. Watanabe, Y.; Daigo, Y. Mater. Sci. Forum 2006, 522–523, 213–220. High-Temperature Oxidation and Corrosion 2005 – Proceedings of the International Symposium on High-Temperature Oxidation and Corrosion 2005. Nagae, M.; Yoshio, T.; Oda, K. Corrosion behavior of structural ceramics in supercritical water. 11th International Ceramics Congress, 2006; p 5. Was, G. S.; Teysseyre, S.; Jiao, Z. Corrosion 2006, 62(11), 989–1005. Daigo, Y.; Watanabe, Y.; Sugahara, K.; Isobe, T. Corrosion 2006, 62(2), 174–181. Jang, J.; Hwang, S. S.; Han, C. H.; Lee, B. H. Corrosion and stress corrosion cracking of high Cr ferritic– martensitic steels in supercritical water. Proceedings of the 2006 International Congress on Advances in Nuclear Power Plants, ICAPP’06, Reno, NV, June 4–8, 2006; pp 2187–2192, Paper No. 6235. Was, G. S.; Teysseyre, S. Challenges and recent progress in corrosion and stress corrosion of austenitic alloys for supercritical water reactor core components.

Material Performance in Supercritical Water

85. 86.

87. 88. 89.

90. 91. 92. 93.

94. 95. 96.

97.

98. 99. 100.

101. 102. 103. 104. 105. 106.

Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, The Minerals, Materials and Metals Society: Warrendale, PA, 2006; pp 1343–1357. Yi, Y.; Lee, B.; Kim, S.; Jang, J. Mater. Sci. Eng. A 2006, 429(1–2), 161–168. Bojinov, M.; Heikinheimo, L.; Saario, T.; Tuurna, S. Characterisation of corrosion films on steels after longterm exposure to simulated supercritical water conditions. Proceedings of the American Nuclear Society – International Congress on Advances in Nuclear Power Plants 2005, ICAPP’05, 2005; pp 1799–1807. Yoon, J. H.; Son, K. S.; Kim, H. S.; et al. Mater. Sci. Forum 2005, 475–479(Pt. 5), 4207–4210. Jang, J.; Lee, Y.; Han, C. H.; Yi, Y. S.; Hwang, S. S.; Sik, S. Mater. Sci. Forum 2005, 475–479, 1483–1486. McKinley, J.; Teysseyre, S.; Was, G. S. Corrosion and stress corrosion cracking of austenitic alloys in supercritical water. In Proceedings of Global 2003, pp 2363–2364. Boyd, W. K.; Pray, H. A. Corrosion 1957, 13(6), 33–42. Yamashita, S.; Ohtsuka, S.; Akasaka, N.; Ukai, S.; Ohnuk, S. Philos. Mag. Lett. 2004, 84(8), 525. Motta, A. T.; Yilmazbayhan, A.; Gomes da Silva, M. J.; et al. J. Nucl. Mater. 2007, 371(1–3); Nucl. Fuels Struct. Mater. 2007, 1, 61–75. Penttila¨, S.; Toivonen, A.; Heikinheimo, L.; Novotny, R. Corrosion studies of candidate materials for European HPLWR. Proceedings of ICAPP’08, Anaheim, CA, June 8–12, 2008; Paper 8164. Macdonald, D. D. J. Supercrit. Fluids 2004, 30, 375. Kriksunov, L. B.; Macdonald, D. D. J. Electrochem. Soc. 1995, 142, 4069. Betova, I.; Bojinov, M.; Kinnunen, P.; Lehtovuori, V.; Penttila, S.; Saario, T. Surface film electrochemistry of AISI316 stainless steel and its constituents in supercritical water. Proceedings of ICAPP 2007, Nice, France, May 13–18, 2007. Dooley, B.; Larkin, B.; Webb, L.; Pocock, F.; Bursik, A. Oxygenated treatment for fossil plants. Proceeding of 53rd International Water Conference, IWC-92-16, Pittsburgh, PA, Oct 1992; pp 154–161. Dooley, B.; Mathews, J.; Pate, R.; Taylor, J. Ultrapure Water July/August 1995, 48–55. Chow, C. K.; Khartabil, H. F.; Bushby, S. J. A fuel channel design for CANDU-SCWR. 14th International Conference on Nuclear Engineering (ICONE-14), 2006. Graham, M. J.; Hussey, R. J. In Oxidation of Metals and Associated Mass Transport; Dayananda, M. A., Rothman, S. J., King, W. E., Eds.; The Metallurgical Society: Warrendale, PA, 1987; p 85. Chen, Y.; Sridharan, K.; Kruizenga, A.; Tan, L.; Allen, T. R. Corrosion behavior of austenitic alloys in supercritical water, submitted to Corrosion Science, 2008. Flewitt, P. E. J.; Wild, R. K. Grain Boundaries – Their Microstructure and Chemistry; Wiley: England, 2001. INCOLOY Alloy 800H & 800HT, Technical Bulletin, Special Metals Publication SMC-047, Mar 2004, http://www.specialmetals.com. Tan, L.; Allen, T. R. Metall. Mater. Trans. A 2005, 36(7), 1921–1925. Randle, V. The Role of the Coincidence Site Lattice in Grain Boundary Engineering; The Institute of Materials: London, 1996. Tsuchiya, Y.; Kano, F.; Sato, N.; Shioiri, A.; Moriya, K.; Kasahara, S. Corrosion 2004; NACE International: Houston, TX, 2004; Paper no. 04485.

107.

108. 109. 110. 111. 112.

113.

114. 115. 116. 117. 118. 119.

120.

121.

122. 123.

124.

125. 126. 127.

325

Saito, N.; Tsuchiya, Y.; Kano, F.; et al. SCC properties of candidate alloys for SCWR core components. Proceedings of ICAPP’06, Reno, NV, June 4–8, 2006; Paper no. 6137. Kasahara, S.; Matsui, H.; Kano, F.; Saito, N.; Ookawa, M.; Kaneda, J. Proceedings of Workshop on Fueled Loop Tests for SCWR R&D; OECD/NEA: Paris, 2006. Tsuchiya, T.; Kano, F.; Saito, N.; Ookawa, M.; Kaneda, J.; Hara, N. Corrosion 2007; NACE International: Houston, TX, 2007; Paper no. 07415. Fujisawa, R.; Nishimura, K.; Kishida, T.; Sakaihara, M.; Kurata, Y.; Watanabe, Y. Corrosion 2005; NACE International: Houston, TX, 2005; Paper 05395. Teysseyre, S.; Was, G. S. Corrosion 2006, 62(12), 1100–1116. Watanabe, Y.; Abe, H.; Daigo, Y. Environmentally assisted cracking of sensitized stainless steel in supercritical water: Effects of physical property of water. Proceedings of GENES4/ANP2003, Kyoto, Japan, Sept 15–19, 2003; Paper no. 1183. Penttila, S.; Toivonen, A.; Heikinheimo, L.; Novotny, R. SCC properties and oxidation behavior of austenitic alloys at supercritical water conditions. 4th International Symposium on Supercritical Water-Cooled Reactors, Heidelberg, Germany, Mar 8–11, 2009; Paper no. 60. Peng, Q. J.; Teysseyre, S.; Andresen, P. L.; Was, G. S. Corrosion 2007, 63, 1033–1041. West, E. A.; Was, G. S. J. Nucl. Mater. 2009, 392, 264–271. Bosch, C.; Delafosse, D. Corrosion 2005; NACE International: Houston, TX, 2005; Paper 05396. Zhou, R.; West, E. A.; Jiao, Z.; Was, G. S. J. Nucl. Mater. 2009, 395, 11–22. Teysseyre, S.; Jiao, Z.; West, E.; Was, G. S. J. Nucl. Mater. 2007, 371, 107–117. Teysseyre, S.; Was, G. S. In 13th International Conference on Degradation of Materials in Nuclear Power Systems – Water Reactors; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, ON, 2007. Jang, J. Recent results on corrosion/SCC/CGR of F/M steels & high Ni alloys in SCW at KAERI. Presented at the Project Management Board Meeting for the SCWR Materials and Water Chemistry, Ann Arbor, MI, Mar 2006. West, E. A.; Teysseyre, S.; Jiao, Z.; Was, G. S. In Proceedings of the 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Canadian Nuclear Society; Allen, T. R., Busby, J., King, P. J., Eds.; Canadian Nuclear Society: Toronto, ON, 2007. Was, G. S.; Teysseyre, S.; McKinley, J. Corrosion 2004; NACE International: Houston, TX, 2004; Paper 04492. Fournier, L.; Delafosse, D.; Bosch, C.; Magnin, Th. NACE International Annual Conference, Corrosion 2001; NACE International: Houston, TX, 2001; pp 1–11, Paper 01361. Gupta, G.; Was, G. S. Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; The Minerals, Materials and Metals Society: Warrendale, PA, 2005; pp 1359–1367. Ampornrat, P.; Gupta, G.; Was, G. S. J. Nucl. Mater. 2009, 395, 30–36. Gaurav, G.; Ampornrat, P.; Ren, X.; Sridharan, K.; Allen, T. R.; Was, G. S. J. Nucl. Mater. 2007, 361(2–3), 160–173. Begley, J. A. Strain-Rate Damage Model for Alloy 600 in Primary Water; Report No. NP-7008s; Electric Power Research Institute, 1990.

326

Material Performance in Supercritical Water

128. Andresen, P. L. In Stress Corrosion Cracking; Jones, R. H., Ed.; ASM International: Materials Park, OH, 1992; p 183. 129. Santarini, G. Corrosion 1989, 45, 389. 130. Was, G. S.; Sung, J. K.; Angeliu, T. M. Metall. Trans. A 1992, 23A, 3343–3359. 131. Congleton, J.; Parkkins, R. N. Nucl. Eng. Des. 1987, 103, 301–311. 132. Was, G. S.; Teysseyre, S. Stress corrosion cracking and corrosion of candidate alloys for the supercritical water reactor, FinalReport to U. S. Department of Energy, Idaho National Laboratory, project #0003852, Sept 30, 2008. 133. Was, G. S.; Teysseyre, S.; McKinley, J.; Jiao, Z. Corrosion 2005; NACE International: Houston, TX, 2005; Paper no. 5397. 134. Scott, P. M. In Ninth International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors; Bruemmer, S., Ford, P.,

135. 136.

137.

138.

Was, G. S., Eds.; The Minerals, Metals and Materials Society: Warrendale, PA, 1999; p 3. Bricknell, R. H.; Woodford, D. A. Scr. Metall. 1982, 16(6), 761–763. Materials Reliability Program: Constant Extension Rate Stress Corrosion Cracking Testing of Alloys 600 and 690 in Supercritical Water (MRP-233), Technical Report 1016154, EPRI, Palo Alto, CA, 2007. Heikinheimo, L.; Guzonas, D.; Fazio, C. GENIV materials and chemistry research – Common issues with the SCWR concept. 4th International Symposium on Supercritical Water-Cooled Reactors, Heidelberg, Germany, Mar 8–11, 2009. Alamo, A.; Averty, X. Status of post-irradiation examinations performed on RA and conventional martensitic steels irradiated at 325 C in osiris reactor. Progress Report UT-SM&C-LAM2, CEA report, NT SRMA 99-2316, Apr 1999.

5.13

Material Performance in Sodium

T. Furukawa and E. Yoshida Japan Atomic Energy Agency, O-arai, Ibaraki, Japan

ß 2012 Elsevier Ltd. All rights reserved.

5.13.1

Introduction

327

5.13.2 5.13.3 5.13.3.1 5.13.3.2 5.13.4 5.13.4.1 5.13.4.2 5.13.4.3 5.13.4.4 5.13.4.5 5.13.4.6 5.13.4.7 5.13.5 5.13.5.1 5.13.5.2 5.13.5.3 5.13.6 5.13.6.1 5.13.6.2 5.13.7 5.13.7.1 5.13.7.2 References

Material Selection in the Consideration of Application in Sodium Corrosion Mechanism of Materials in Sodium Corrosion Produced by the Dissolution of Alloy Elements to Sodium Corrosion Produced Through Chemical Reaction with the Impurities in Sodium Corrosion Behavior and Factors Affecting Steel Immersion Time Temperature Dissolved Oxygen Sodium Velocity Alloy Elements Carburization and Decarburization Corrosion Estimation of FBR Materials Effect of Sodium on the Mechanical Strength of Steels Austenitic Stainless Steel Ferritic Steels Others (Ni Base Alloys, ODS) Damage to Steels with Sodium Compounds Sodium–Water Reaction Sodium Leak Tribology Self-Welding Frictional Wear

328 328 328 330 331 331 332 332 333 333 333 334 336 336 338 338 338 338 339 339 339 339 340

Abbreviations EBR-II FBRs FFTF FMS JAEA Monju ODS PFR PNC

Experimental Breeder Reactor No. 2 (USA) Fast breeder reactors Fast Flux Test Facility (USA) Ferritic–martensitic stainless steel Japan Atomic Energy Agency Japanese Prototype Fast Breeder Reactor Oxide dispersion-strengthened steel Prototype Fast Reactor (UK) Power Reactor and Nuclear Fuel Development Corporation (present JAEA)

5.13.1 Introduction Sodium is one of the elements that exhibit the characteristics demanded of coolants for fast breeder reactors (FBRs). The physical properties of sodium

are shown in Table 1 and see Chapter 2.14, Properties of Liquid Metal Coolants. Sodium is a solid at room temperature, and the melting and boiling temperatures are 97.82 and 881.4
C, respectively. Therefore, sodium is in the liquid phase at FBR operating temperatures without pressurization. For this reason, it is not necessary to adopt the proof-pressure design employed in light water reactors for sodium-cooled FBRs. Moreover, the thermal conductivity and specific heat of sodium at 550
C are 0.648 W cm1
C1 and 1.256 J g1
C1, respectively, and sodium can transfer the heat of the reactor core to the power generation system efficiently. Furthermore, its insignificant neutronmoderating capability is suitable for the coolant for the FBR, in which fast neutrons play a major role in the nuclear reaction. On the other hand, the weakness of sodium as a coolant is its reactivity with oxygen and/or water.

327

328 Table 1

Material Performance in Sodium Physical properties of sodium

Atomic number Atomic weight Melting point (
C/K) Boiling point (
C/K) Volume increase on melting (%) Density (g cm3) Thermal conductivity (W cm1 K1) Specific heat (J mol1 K1)

11 22.9898 97.8/371.0 881.5/1154.7 2.71 0.856 at 400
C 0.821 at 550
C 0.727 at 400
C 0.657 at 550
C 29.40 at 400
C 28.88 at 550
C

Table 2 elements1

Element Solubility equation Cu Ag Au Mg Zn

When high-temperature sodium is leaked into the atmosphere, it reacts with the oxygen and moisture in the atmosphere, and the by-products of this reaction are known to cause structural damage to the reactor. Moreover, in the event of a steam generator tube failure, high-temperature steam blows off into the sodium, and wastage of the adjoining tubes occurs.

5.13.2 Material Selection in the Consideration of Application in Sodium In the process of material selection, it is necessary to take the environment into consideration by estimating the mechanical properties and thermal characteristics. Environments specific to the FBR components include: (a) contact with the coolant (liquid metallic sodium), (b) high temperature at which the creep effects must be taken into account, and (c) neutron irradiation. In this chapter, an outline of the compatibility of materials with sodium is provided. The oxides formed on the surface of the material are easily reduced in high-temperature sodium, resulting in direct contact between the material and sodium. Under this condition, the dissolution of elements contained in the material, such as iron, chromium, and nickel, in sodium and the reverse phenomenon (deposition) occur on the material surface due to the difference in chemical potential. The behavior is fundamentally controlled by the solubility of the material elements in sodium and by the diffusion rate of the materials. The solubility equation of the various elements in sodium is shown in Table 2. Among the elements in austenitic stainless steel, the solubility of nickel is greatest. Therefore, the phase transformation of austenite to ferrite through nickel dissolution is observed on the surface in longterm immersion in sodium. The compatibility of various metals with sodium reported by Borgstedt and Mathews1 is shown in

Solubility equation of metallic and amphoteric

Cd Al Ga In U Pu Sn Pb Bi Cr Mo Mn Fe Co Ni

log Swwpm ¼ 5.450  3055/T (K) log Swwpm ¼ 7.22  1479/T (K) Swt% ¼ 11 þ 0.52  (T(K)  273.15)  6  104  (T(K)  273.15)2 Swt% ¼ 0.1414 þ 2.08  106  (T(K)  273.15) þ 1.248  103  (T(K)  273.15)2 Swppm ¼ 1.4 þ 0.057  (T(K)  273.15) log Swt% ¼ 3.67  1209/T (K) log Swwpm ¼ 1.4 þ 0.057/T (K) log Swt% ¼ 1.349  1010/T (K) log Swt% ¼ 4.48  1552/T (K) log Swwpm ¼ 4.36  6010.7/T (K) log Swwpm ¼ 8.398  10.950/T (K) log Swt% ¼ 5.113  2299/T (K) log Swt% ¼ 6.1097  2636/T (K) log Swt% ¼ 2.15  2103/T (K) log Swt% ¼ 5.67  4038/T (K) log Swwpm ¼ 9.35  9010/T (K) log Swwpm ¼ 2.738  2200/T (K) log Swwpm ¼ 3.640  2601/T (K) log Swwpm ¼ 4.720  4116/T (K) log Swwpm ¼ 0.010  1493/T (K) log Swwpm = 2.07  1570/T (K)

Temperature range (K) 623–773 377–806 373–873

373–573 373–600 423–773 375–573 373–573 560–970 560–970 473–673 393–523 398–563 563–923 948–1198 500–720 550–811 658–973 673–973 673–973

Table 3. The data can be used to create an index for the selection of the material used in sodium.

5.13.3 Corrosion Mechanism of Materials in Sodium There are two known mechanisms of sodium corrosion. One is corrosion produced by the dissolution of alloy elements to sodium, and the other is corrosion produced through chemical reaction with the impurities in sodium. These two corrosion mechanisms are described in Sections 5.13.3.1 and 5.13.3.2. 5.13.3.1 Corrosion Produced by the Dissolution of Alloy Elements to Sodium In this case, corrosion is dependent on the solubility in sodium of the elemental composition in the material, temperature, and the rate of solution. The solution rate Rc is given by the following formula: Rc ¼ K ðCs  Ci Þ

½1

Material Performance in Sodium

Table 3

329

Compatibility of materials with alkali metals1 Compatible with alkali metal up to (
C)

Material

Mg alloys Al alloys Cu alloys Ag and its alloys Au and its alloys Zn coatings Pb and its alloys Sn and its alloys Fe Low-alloy steels Ferritic steels High-Cr steels Austenitic steels Ni alloys Mo alloys W alloys Ti alloys Zr alloys V alloys Nb alloys Ta alloys Sintered A12O3 Stab. ZrO2/CaO Stab. ThO2/Y2O3 Glass UO2 UC

Factors influencing compatibility

Li

Na

K

Rb and Cs

n.c. n.c. 300 n.c. n.c. n.c. n.c. n.c. 500 500 500 500 450 400 1000 1000 700 700 700 700 700 350 350 400 n.c.

n.c 350 400 n.c. n.c. n.c. n.c. n.c. 700 700 700 700 750 600 1000 1000 700 700 700 700 700 500 350 550 250 750 750

300 400 400 n.c. n.c. n.c. n.c. n.c. 700 700 700 700 750 600 1000 1000 700 700 700 700 700 500 350 550 250

300 450 400 n.c. n.c. n.c. n.c. n.c. 700 700 700 700 750 600 1000 1000 700 700 700 700 700 500 350 550 250

Metal solubility, oxygen exchange Metal solubility Metal solubility High metal solubility High metal solubility High metal solubility Very high metal solubility Very high metal solubility Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Flow velocity Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Nonmetallic impurities Thermomechanical action Intergranular corrosion Intergranular corrosion Chemical reaction Excess of oxygen Nonmetallic impurities

n.c. ¼ not compatible.

1.0E + 02

1.0E + 00 Ni Solubility (ppm)

Mn 1.0E − 02

Fe

C Mo 1.0E − 04 Si 1.0E − 06 Cr 1.0E − 08 1

1.2

1.4 1000/T (K)

1.6

Figure 1 Solubility of alloy elements in sodium.

1.8

where K is the solution rate constant, Cs is the solubility limit in sodium, and Ci is the actual concentration in sodium. The solution rate constant K is controlled by diffusion. The solubility of the alloy elements of the steel is shown in Figure 1.1–5 Included in Figure 1 are the major elements of austenitic stainless steels (type 304SS and 316SS) and the Cr–Mo steels (2.25Cr–1Mo and Mod. 9Cr–1Mo) which are used as the structural materials of FBRs. The solubility of each of the elements in sodium at 550
C is less than a few parts per million. This means that the compatibility of the steels with sodium is fundamentally excellent. In the isothermal sodium condition, the corrosion of the steels stops when the dissolved elements reach saturation concentration at the temperature of sodium. However, in the nonisothermal sodium condition, corrosion resulting from the difference in activity between sodium and the material surface occurs continually. This corrosion behavior is called thermal gradient mass transfer. In the cooling system, the elements in the materials in the high-temperature section

330

Material Performance in Sodium

ΔW (mg cm-2)

Temperature (⬚C)

dissolve as a result of the temperature dependency of the solubility of the elements in sodium, and the dissolved elements are deposited on the steel surface in the low-temperature section by the same mechanism. The results of thermal gradient mass transfer using a sodium loop made of SUS316, which is equivalent to AISI type 316SS, is shown in Figure 2. The weight loss caused by the dissolution of the elements in the steel is measured in the high-temperature section, and the weight gain caused by the deposit of the dissolved elements in sodium is observed in the low-temperature section. Figure 3 shows the microstructure of the inner surface of the sodium pipe taken from the flowing sodium loop operated for 82 000 h. Dissolution of the elements in the material is observed in the hightemperature section, and precipitation is observed in the low-temperature section located in the lower stream. Generally, selective corrosion occurs at the initial stage as a result of the dissolution of the elements

600 Sodium flow→ 550 500 500 400 +1 390

600 520 470 410

0 –1

-1

Sodium velocity :3 m s Oxygen level :9 ppm Immersion time 5512 h 7141 h 10644 h

Weight loss

-2 -3 0

5

10 15 20 Distance from electro magnetic pump of sodium loop (m)

25

29

Figure 2 Weight change in SUS316 after corrosion test in flowing sodium. Reproduced from Maruyama, A.; Nomura, S.; Kawai, M.; Takani, S.; Ohta, Y.; Atsumo, H. J. Atomic Energy Soc. Jpn. 1984, 26, 59.

in the steel, and then, the behavior moves to general corrosion with the progress of time. In the hightemperature section, the dissolution of nickel, chromium, manganese, and silicon to sodium occurs easily, and molybdenum and iron remain in the material. In the low-temperature section, chromium deposits easily with decreasing temperature. Carbon transfer in the steel affects the mechanical properties of the FBR structural materials. Decarburization occurring in the high-temperature section, in particular, has the potential of degrading creep strength. The effect of sodium on the mechanical strength is described in Section 5.13.5. 5.13.3.2 Corrosion Produced Through Chemical Reaction with the Impurities in Sodium The most important element in the impurities in sodium is oxygen. Sodium is the reducing agent, and its affinity to oxygen is very strong. The temperature dependence of the oxygen saturated in sodium is shown in Figure 4. The solubility of oxygen in sodium is significantly higher than in water. However, the control of impurities in sodium can be achieved by using the cold trap technique6 based on the theory of the deposit of dissolved impurities in sodium. The introduction of oxygen into sodium may occur during nuclear plant construction, refueling, the supplementing of the reactor cover gas, the opening of the coolant boundary for maintenance operations, etc. These are the paths for contamination through oxides adhering to the components and the impurities in the gas. The relationship between the standard free energy of the formation of iron oxides and the temperature is shown in Figure 5. The thermodynamically stable oxide in sodium is sodium oxide, Na2O. Sodium flow

(a) 525 ⬚C, 82 000 h (BD-1)

(b) 575 ⬚C, 82 000 h (BD-2)

(c)

(d) 625 ⬚C, 82 000 h (BD-3)

420 ⬚C, 82 000 h (BD-4)

10 μm

Figure 3 Corrosion of the inner surface of sodium loops made of type 304SS operated for 82 000 h. Reproduced from Yoshida, E.; Kato, S.; Wada, Y. Liquid Metal Systems; Plenum press: New York, 1995.

Material Performance in Sodium

Dissolved oxygen (ppm)

1.0E + 04

oxidize iron thermodynamically (i.e., iron oxide cannot be formed). The iron is oxidized by the formation of complex Na–Fe oxides.7 In addition to oxygen, impurities in sodium include elements in the steel, hydrogen, and nitrogen. Hydrogen and nitrogen induce changes in the microstructure that lead to the potential degradation of mechanical properties.

1.0E + 03 1.0E + 02 1.0E + 01

5.13.4 Corrosion Behavior and Factors Affecting Steel

1.0E + 00 1.0E - 01 1

1.5

2

2.5

3

1000/T (K) Figure 4 Temperature dependence of saturated oxygen concentration in sodium.

-300 O2 Na 2

2/3

ΔG0f (KJ mol-1 O2)

-400

Fe 2O 3 O

2Fe

The basic mechanism of the steel corrosion in sodium is described in Section 5.13.3. In this section, the corrosion behavior and its effects are described. The major data are obtained for austenitic stainless steels. The important factors which influence the corrosion of the steels in sodium are: (1) immersion time, (2) temperature, (3) dissolved oxygen, (4) sodium velocity, (5) alloy elements, and (6) carburization and decarburization. The data on each factor are described in the following sections. 5.13.4.1

-500

1/2

e 3O 4

a 2O

F

2N O3

2/3

-600 O4

1/2

Fe Na 3

O3

e a 4F

N 2/3

e a 5F

N

-700 O2

Fe Na 2

-800 200

331

400

600 800 Temperature (K)

1000

1200

Figure 5 Relationship between
G0f and temperature in oxides consisting of iron and sodium.

Iron, the main element in the steel, is corroded by the following reaction: Fe½s;l þ 3Na2 O½s;l ! Na4 FeO3 ½s þ 2Na½s;l 0 0 ¼ þ43:8 KJ mol1 ;
Gr298 ¼ þ26:1KJ mol1
Hr298

where [g], [l], and [s] stand for gas, liquid, and solid. This is an endothermic reaction. However, Gibbs energy of reaction (
Gr0 ) becomes negative at 380
C, and the reaction progresses spontaneously above that temperature. In addition, Na2O cannot directly

Immersion Time

As mentioned in Section 5.13.3.1, nickel, chromium, and manganese dissolve easily in sodium. Therefore, when austenitic stainless steel is immersed in sodium, the selective corrosion caused by the dissolution of these elements progresses in the initial stage. This process is based on the solid diffusion of the elements in the steel, and is described by Fick’s second law. Diffusion in sodium is given by the following two formulas: dM dc ¼ D dt dx

½2

2 pffiffiffiffiffi Mðt Þ ¼ pffiffiffi Dt ðC0  Cs Þ p

½3

where M is the diffusion amount, D is the diffusion coefficient, C0 is the initial concentration, and Cs is the surface concentration at time (t). Formula [2] is Fick’s first law and its integration is given as the total diffusion. Formula [3] is the function given as the approximate formula. From these formulas, it is understood that the corrosion behavior at the initial stage observes the parabolic law proportional to the root of time. At the initial process of the corrosion, the parabolic type behavior caused by the selective corrosion is dominant. The corrosion behavior is called start-up corrosion.

332

Material Performance in Sodium

100

400

Weeks et al.9

Steady-state corrosion

Metal loss

Start-up

Temperature (⬚C) 600 500

700

Immersed time, t Selective loss (Ls) ∝ t1/2 Bulk loss (Lb) ∝ t Total loss (Ls + Lb)

Corrosion rate (µm year–1)

Bagnall and Jocobs12

10

Menken30

Zebroski10

Kolster11

Thorley and Tyzack8 JAEA

1

Figure 6 Schematic representation of corrosion data for austenitic stainless steel.

The corrosion behavior of iron, which is the main element of stainless steel, is general corrosion and the corrosion progresses as a linear function of time. Therefore, the corrosion of the dominant factor changes from start-up corrosion to general corrosion with the progress of time (Figure 6). The corrosion behavior that dominates general corrosion is called steady-state corrosion. The time required for a change from start-up corrosion to steady-state corrosion is 2000–5000 h, although it is dependent on conditions such as sodium volume, temperature, velocity, and dissolved oxygen. 5.13.4.2

Temperature

Temperature dependence of the corrosion rate in the sodium of austenitic stainless steels is shown in Figure 7. The corrosion rate CR is described by the following Arrhenius function:   E ½4 CR / exp  RT where R is the gas coefficient (8.3145 J mol1 K1), E is apparent activation energy, and T is temperature. The apparent activation energy E is a function of temperature. The values are reported as 100–120 kJ mol1 by numerous researchers (Table 4). The energy is lower than that of the activation energy of the diffusion of iron, chromium, and nickel, in

Austenitic stainless steel Oxygen content: 10 ppm Na velocity: 3.8 m s–1 0.1 0.9

1.1

1.3

1.5

1000/T (K) Figure 7 Corrosion rate of austenitic stainless steels in flowing sodium.

stainless steel (250–300 kJ mol1), and it agrees with the activation energy of solubility in sodium (Figure 1, Fe: 82.5 kJ mol1 and Cr: 104 kJ mol1). In fact, it is understood that the corrosion of stainless steel is dominated by the dissolution process of the major elements (iron, chromium, and nickel) of the steel. The effect of dissolved oxygen on sodium, which influences the dissolution reaction, is described in the following section. 5.13.4.3

Dissolved Oxygen

The effect of dissolved oxygen on the corrosion rate is described by the following formula because the corrosion process is dominated by the reaction process of oxides. CR / ½O2 n

½5

where CR is the corrosion rate, [O2] is the oxygen concentration, and n is a constant. The constant n, reported by the researchers, is listed in Table 5. This result suggests the possibility

Material Performance in Sodium Table 4 Comparison of apparent activation energy on corrosion rate Sodium temperature (
C)

Activation energy (kJ mol1)

Reference

Thorley Weeks Zebroski Kolster Bagnall Maruyama

450–725 538–705 500–700 650–700 593–723 500–650

73.5 108.8 110.5 114.2 167.4 92–109

[8] [9] [10] [11] [12] [13]

Table 5 Comparison of oxidation coefficient on corrosion rate Bibliography

O2 content (ppm)

Coefficient (n)

Reference

Thorley Zebroski Roy Kolster

5–100 12, 50 5–30 1–8 8–40 2.5–9

1.5 1, 1.56 1.2 0.91 >1 0.8

[8] [10] [14] [11]

Maruyama

[13]

that the control of dissolved oxygen may significantly influence the corrosion behavior. 5.13.4.4

Sodium Velocity

It has been observed that the corrosion rate in sodium increases as sodium velocity increases. However, it is known that the increase ends when the velocity reaches a certain limit. It is believed that the limit is a function of oxygen concentration in sodium and/ or the structure of the sodium loop. According to Thorley and Tyzack,8 Roy,14 and Kolster,11 the limit is 3.8, 6–7, and 3 m s1, respectively. It is believed that the effect of such a sodium velocity is based on the thickness of the boundary layer between the material surface and the flowing sodium. In fact, the thickness of the boundary layer decreases as sodium velocity increases, and the diffusion of alloy elements via the boundary layer increases. 5.13.4.5

Alloy Elements

The effect of the alloy elements on corrosion is examined because long-term corrosion occurs as a result of thermal gradient mass transfer. The effect of the chromium and nickel concentration on the steels is particularly significant. Figure 8 shows the effect

10 O2 cont.: 1 ppm Velocity: 1.48 m s–1 Immersion time: 2930–5254 h

Corrosion rate (µm year–1)

Bibliography

333

650 ⬚C

1.0 600 ⬚C

0.1

0

20 30 10 Nickel concentration in steel (mass %)

40

Figure 8 Effect of nickel content in stainless steel on the corrosion rate.

of the nickel content of stainless steel on the corrosion rate. On the other hand, the dependence of the corrosion rate on elements in the steels is hardly observed in austenitic stainless steels, such as types 304SS, 316SS, and 321SS, because of the slight difference in chemical composition (Figure 9). 5.13.4.6

Carburization and Decarburization

In monometallic sodium loops, the difference of the carbon activity, which is the driving force of the carbon transfer of the material, increases as temperature increases. Therefore, decarburization occurs in the high-temperature section and carburization occurs in the low-temperature section. On the other hand, in bimetallic sodium loops which consist of austenitic stainless steel and ferritic steel, it is easy for decarburization to occur in the ferritic steel, which has a high carbon activity due to the difference in carbon activity between different materials, whereas carburization in austenitic stainless steel, which has low carbon activity, easily occurs at elevated temperature. Carbon is an important element in maintaining the superior mechanical strength of steel. Therefore, the carburization/decarburization behavior of the steels via sodium is important from the perspective of mechanical properties.

334

Material Performance in Sodium

750

650

Temperature (⬚C) 550 450 304SS 316SS

Corrosion rate (μm year–1)

10

Fuel cladding tube (FCT)

321SS

~9 ppm O2 ~2.5 ppm O2 ~9 ppm O2 ~2.5 ppm O2 ~9 ppm O2 ~2.5 ppm O2

1 20 ppm O2 10 ppm O2 5 ppm O2

Sodium velocity 2–4 m s-1 (FCT: 2–4.8 ms-1) 0.1

0.9

2.5 ppm O2

Immersion time 1000–7200 h 1.1

1 ppm O2 1.3 1000/T (K)

1.5

Figure 9 Comparison of corrosion rates of austenitic stainless steels.

Carbon concentration in sodium (ppm)

100 7 5 3 2 10–1 7 5 3 2

Decarburization Type316 (0.06 wt% C)

FFTF (566 ⬚C) FFTF (474 ⬚C)

Type304 L (0.025 wt% C) Data band of the carbon concentration in EBR-II primary cooling system

EBR-II (470 ⬚C)

10–2 7 5 3 2 10–3 400

Carburization

500

600 700 Temperature (⬚C)

Figure 10 shows the boundary between carburization and decarburization in monometallic sodium loops consisting of austenitic stainless steel (single alloy).15 At a carbon concentration of 0.2 ppm in sodium, the temperature boundary is 650
C, with decarburization occurring over that temperature and carburization occurring below that temperature. Although the boundary is influenced by the carbon concentration in sodium (carbon activity), it is necessary to take decarburization and carburization into consideration to apply austenitic stainless steel when the temperature is above 550
C, such as fuel cladding tubes. On the other hand, in bimetallic sodium loops that consist of ferritic steel and austenitic stainless steel (two alloys), decarburization and carburization also occur in the temperature range of the structural materials <550
C. The thickness of decarburization of the component (
WD ) can be estimated by the following formula: pffiffi
WD ¼ K t ½6 where K is the decarburization coefficient (g cm2 s1/2) and t is time (s). The temperature dependence of the decarburization of 2.25Cr–1Mo steel in bimetallic sodium loops consisting of type 304SS and 2.25Cr–1Mo steel, which have the same structure as those in FBR Monju, is shown in Figure 11. In recent times, the application of high chromium ferritic steel, which has excellent mechanical strength at elevated temperatures and superior corrosion resistance with the coolant, has been used as the structural material for advanced FBRs.16 The relationship between decarburization and chromium concentration in the steel is shown in Figure 12. It is known that the decarburization/carburization behavior of the ferritic steels is dependent on chromium concentration. The precipitation of chromium carbide increases as the chromium concentration increases. In this case, decarburization of the steel is inhibited because the carbon activity in steel decreases.

800

Figure 10 Boundary between carburization and decarburization in monometallic sodium loops consisting of austenitic stainless steel. Reproduced from Snyder, R. B. An analysis of carbon transport in the EBR-II and FFTF primary sodium systems. In Proceedings of the International Conference on Liquid Metal Technology in Energy Production, Champion, 1976.

5.13.4.7 Corrosion Estimation of FBR Materials The corrosion rate in the sodium of austenitic stainless steel and ferritic steel is very slight (a few microns per year). Furthermore, since the estimated corrosion thickness during the operation period is

335

Material Performance in Sodium

550

Temperature (⬚C) 500 450

400

∗The depth in microns from the surface is indicated beside each plot by numerals in brackets.

350 0.25

Carbon concentration near the surface∗ (%)

’s nju Mo or

f ine dl de en n sig de

(15)

(16)

0.15 (10)

(29)

(11) (16)

0.10

(14)

(11) (17)

Tube/ pipe

Bar Coupon

2517 5071

0 0

10180 17768 1.1

1.2

1.4 1.3 1000/T (K)

1.5

1.6

Figure 11 Temperature dependency of the decarburization coefficient of 2.25Cr–1Mo steel in sodium.

small compared with the thickness of the structures (permanent components) of FBRs, it is not a critical point in the reactor design. For the structural materials of Monju, therefore, the following corrosion formula was applied for all materials for the design 103 log10 R ¼ 0:85 þ 1:5log10 Co  3:9 T þ 273

½7

where R is the corrosion rate (mm year1), Co is the dissolved oxygen (ppm, 5  Co  25), and T is the temperature (
C, 400  T  650). This equation is valid for materials types 304SS, 316SS, 321SS, and 2.25Cr–1Mo steel. After the construction of Monju, the data for corrosion in sodium for the structural materials for advanced FBRs were obtained by the Japan Atomic Energy Agency (JAEA),16,17 and it was confirmed that the corrosion rate of the materials is also described by the aforementioned formula. The corrosion formula is expected to be applied to advanced steels as well.

1

(9) 9Cr–1Mo→

0.05

Specimen

7Cr–1Mo→

Time (h)

10-10 1.0

(17) (21)

(6)

5Cr–1Mo→

10

-9

(14)

0.20

2.25Cr–1Mo→

Decarburization coefficient (G cm-2 s-1/2)

mm co Re

10-8

Testing conditions 1000 h, 8 ppm O2 1000 h, 50 ppm O2 3000 h, 9 ppm O2

3Cr–1Mo→

10-7

700 650 600

2 3 4 5 6 7 8 9 Chromium concentration in steel (mass %)

10

Figure 12 Relationship between decarburization and chromium concentration in high chromium ferritic steel. Reproduced from Matsumoto, K.; Ohta, Y.; Kataoka, T. Nucl. Technol. 1976, 28(3), 452–470.

On the other hand, since the thickness of the fuel cladding materials is thin, the ratio of the corrosion thickness to the original thickness is larger than that of the structural materials. Therefore, it is necessary that corrosion estimation be performed with greater precision. The corrosion formula is proposed for each core material by JAEA.18–20 Monju core materials (PNC316 [16Cr–14Ni–B– P–Ti–Nb]), type 316SS) X ti CRi ½8 CNa ¼ 8760 i   104 5 CRi ¼ 4:92710 exp 1:647 Oxi ½9 Ti þ 273 where CNa is the corrosion thickness (mm), ti is the operation time (h) at temperature Ti with dissolved oxygen Oxi, CRi is the corrosion rate (mm year1) at sodium temperature Ti with dissolved oxygen Oxi, Oxi: dissolved oxygen (ppm, Oxi  5), and Ti is the sodium temperature (
C 400  T  700) For PNC1520 [15Cr–20Ni–B–P–Ti–Nb] CNa ¼ ðCR1 þ CR2  ti ÞOxi

½10

336

Material Performance in Sodium

log10 CR1 ¼ 7:6036  6:6021

103 Ti þ 273

 log10 CR2 ¼ 1:5172  108 exp 2:4275

½11

 104 Ti þ 273

½12

where CNa is the corrosion thickness (mm), CR1 is the initial corrosion (mm) at temperature Ti with dissolved oxygen Oxi, CR2 is the steady corrosion rate (mm h1) at temperature Ti with dissolved oxygen Oxi, ti is the operation time at temperature Ti with dissolved oxygen Oxi, Oxi is the dissolved oxygen (ppm, Oxi  5), and Ti is the sodium temperature (
C, 400  T  650) For PNC-FMS [11Cr–2W–Mo–Nb–V] CNa ¼ ðCR1 þ CR2  ti ÞOxi

½13

103 Ti þ 273

½14

log10 CR1 ¼ 9:078  8:251

 log10 CR2 ¼ 4:1666  104 exp 1:7580

104 Ti þ 273



½15

where CNa is the corrosion thickness (mm), CR1 is the initial corrosion (mm) at temperature Ti with dissolved oxygen Oxi, CR2 is the steady corrosion rate (mm h1) at temperature Ti with dissolved oxygen Oxi, ti is the operation time at temperature Ti with dissolved oxygen Oxi, Oxi is the dissolved oxygen (ppm, Oxi  5), and Ti is the sodium temperature (
C, 400  T  650).

5.13.5 Effect of Sodium on the Mechanical Strength of Steels The effect of sodium on the mechanical strength of steel is determined by three factors: (a) corrosion and mass transfer, (b) decarburization and carburization, and (c) nonoxidation (reducing atmosphere) by sodium. For creep strength whose dominant factors are time and temperature, the behavior of the carbon and minor elements in the steels are important factors. For the fuel cladding materials, which are thin structures, it is necessary to consider the effect of metal loss by corrosion. On the other hand, fatigue strength in sodium is enhanced in a nonoxidizing environment such as vacuum. Thus, the effect of sodium on mechanical strength appears as the synthesis of both the corrosion and mass transfer of the steel elements. In this section, the important points concerning the effects of sodium on

the mechanical strength of core materials, such as fuel cladding and wrapper tubes, and structural materials, such as reactor vessels, main components, and pipes, are described. To estimate the sodium environmental effect of the core materials, the following three points must be taken into consideration: (a) the thin structure (about 0.5 mm thickness for fuel cladding tubes), (b) the high temperature (maximum 650
C), and (c) the exchangeable components. Regarding (a), since the ratio of the sodium-effect layer (corrosion layer) to the thickness of the fuel cladding tubes becomes large, the stress increment affecting the base metal is larger than that of the structural materials. Regarding (b), since carbon transfer (decarburization/ carburization) becomes active, the high-temperature strength of the core materials changes easily. Regarding (c), on the other hand, it can be advantageous to allow a design that takes into account the degree of material deterioration. The structural materials are thick compared with the core materials, and the operation temperature is 550
C or less. Therefore, the effect of sodium is lower than it is on the core materials. However, since the structural materials make up the permanent structure, the long-term (30 years of more) structural integrity of the materials has to be estimated appropriately. The FBR structure materials are used at high temperatures under which creep phenomenon will occur. Therefore, the plant design requires that creep effects as well as tensile and fatigue strength properties be taken into account. In particular, in the case of the fuel cladding tubes, internal pressure creep resulting from the generation of fission product gas is one of the main failure modes. In the case of the structural materials, since cyclic stress occurs during plant start-up/shut-down and creep stress occurs during operation, creep fatigue, which is determined by both the creep and fatigue effects, is likely to be one of the main failure modes. 5.13.5.1

Austenitic Stainless Steel

The creep rupture test results in sodium for the austenitic stainless steel type 304SS are shown in Figure 13. The creep strength in sodium is equivalent to that in air at 650
C or less, and no effect of sodium on creep strength is observed. The same behavior is observed in the creep rupture test of material preimmersed in sodium for 20 000 h and the material cut from the sodium loops operated for 100 000 h.21

337

Material Performance in Sodium

5.0 Type 304SS : In sodium : In air

3.0 2.0 1.0

30

500 ⬚C 550 ⬚C

0.5

600 ⬚C 10 5 101

650 ⬚C 102

103 Time to rupture (h)

104

105

Figure 13 Creep strength of type 304SS in sodium. Reproduced from Yoshida, E.; Aoki, M.; Kato, S.; Wada, Y. 28th Symposium on High Temperature Strength, The Society of Materials Science: Japan, 1990.

Transgranular

450 ⬚C 3.0 2.0 1.0 0.5 550 ⬚C 3.0 2.0 1.0 0.5 650 ⬚C

Stress (MPa)

500

2⫻102 650 ⬚C

200

700 ⬚C

20 102

Intergranular

103 104 105 Number of cycle to failure (cycles)

5⫻105

Figure 15 Low-cycle fatigue properties of type 304SS in sodium. Reproduced from Wada, Y. Genshiryoku-kogyo 1990, 36(3); in Japanese.

100 50

Type of surface crack

Material: SUS304 In air In sodium

Total strain range (%)

Stress (kg mm-2)

100

In air In sodium 103 104 Time to rupture (h)

105

Figure 14 Creep strength of PNC316 in sodium. Reproduced from Yoshida, E.; Wada, Y. Creep rupture properties of austenitic stainless steel in elevated temperature sodium. High Temperature Strength Committee of the Society of Materials Science, Japan, 1992.

Furthermore, no effect on creep strength in the carburization environment of a bimetallic sodium system was observed. However, a slight degradation of creep ductility was observed in the carburization environment. In this case, micro cracks caused by carburization were also observed in the ternary creep region.22 On the other hand, the strength reduction caused by the dissolution of the alloy elements, including carbon, in sodium was observed at temperatures greater than 650
C (Figure 14). The results of the low-cycle fatigue test in sodium are shown in Figure 15. In the temperature range of 550
C or less used for the structural materials of FBRs, the strength in sodium is greater than in air. This can be explained by the distribution behavior of surface cracks. Many surface microcracks were observed on the surface of the specimen tested in air due to the fact that oxides formed by

high-temperature oxidation in air cause the appearance of fatigue cracks. On the other hand, hardly any microcracks were observed on the in-sodium specimens because of the effect of nonoxidation (reducing atmosphere) by sodium. This is the equivalent of fatigue behavior in a vacuum, and showed excellent strength in the low-strain range in particular. However, the superiority of the fatigue strength in sodium decreases as temperature increases, and the fatigue strength in sodium at 650
C is equivalent to that in air. It is thought that the strength of the grain boundary decreases due to the fact that the selective dissolution of the material elements to sodium via the grain boundary increases as the temperature increases.23 The major failure mode of the structural materials in the FBR is creep-fatigue loading. In this case, creep damage becomes the dominant factor, and the nucleation and growth of the creep cavity exacerbates the failure. Therefore, it is thought that the effect of sodium on the failure mode is negligible. Moreover, it has been reported that the creep fatigue strength of type 304SS, which produced accelerative carburization in sodium, is comparable to its strength in air.24 From these results, it is concluded that the effect of sodium on mechanical strength is negligible for austenitic stainless steels.

338

Material Performance in Sodium

5.13.5.2

Ferritic Steels

The corrosion behavior in the sodium of ferritic steel is fundamentally the same as that of austenitic stainless steel. However, the carbon content in ferritic steel is high compared with austenitic stainless steel, and the steel tends to undergo strength reduction by decarburization. Figure 16 shows the creep strength of PNC-FMS steel, which is planned as the core material for the advanced Japanese FBRs. At 650
C, the strength in sodium is lower than that in air, and this tendency becomes more evident when kept for longer periods of time. Such strength reduction is associated with the carbon concentration in steel. Therefore, it is necessary to give the strength reduction coefficient for the design of the FBRs and to estimate the strength reduction in sodium conservatively. 5.13.5.3

Others (Ni Base Alloys, ODS)

Nickel base alloys, which have excellent mechanical strength at high temperature, are used as the liner for protection from high-cycle fatigue failure caused by thermal striping near the coolant surface. The corrosion thickness of the nickel base alloy in sodium is equivalent to that of austenitic stainless steels. However, its tensile and fatigue strength in sodium at elevated temperatures are lower than in air because of the nickel’s dissolution to sodium and the formation of an intermetallic compound by thermal aging in high temperature.25 Oxide dispersion-strengthened (ODS) ferritic steel is a promising candidate for the fuel cladding tubes of advanced FBRs. It was reported that the

ODS steel had excellent creep strength in sodium as well as in air due to the effect of nano-oxide particles dispersed in the matrix. On the other hand, it was reported that the nickel, which is dissolved from the nickel-containing steels in sodium, penetrated to ODS steel in flowing sodium at elevated temperatures (Figure 17). To apply the steel simultaneously with the nickel-containing steels, such as austenitic stainless steel, it is necessary to carefully estimate the microstructure change by nickel penetration and its strength reduction.

5.13.6 Damage to Steels with Sodium Compounds 5.13.6.1

Sodium–Water Reaction

One of the weak points of the present FBRs is the steam generator, the main component of the power generation system. When the steam generator tube is damaged, high-temperature pressurized steam is blown into sodium, and the following chemical reactions occur: Na½s;l þ H2 O½g ! NaOH½s;l þ 1=2H2 ½g ½16 0 0
Hr298 ¼ 183:8KJmol1 ;
Gr298 ¼ 150:9KJmol1

2Na½s;l þ H2 O½g ! Na2 O½s;l þ H2 ½g

½17

0 0 ¼ 172:8KJmol1 ;
Gr298 ¼ 147:2KJmol1
Hr298

where [g], [l], and [s] stand for gas, liquid, and solid. The environment around the failed tube is heated to high temperature in the aforementioned chemical

Stress (MPa)

600 ⬚C 650 ⬚C

100 Material: PNC-FMS In sodium In air 50 101

102

103

104

Time (h) Figure 16 Creep strength of PNC-FMS in sodium. Reproduced from Ito, T.; Yoshida, E.; Kobayashi, T.; Kimura, S.; Wada, Y. Materials Design Base Standard Supplement (Tentative) of High Strength Ferritic/Martensitic Steel (PNC-FMS) Core Components for LMFBR; Research report of the Japan Atomic Energy Agency, PNC TN9410 93–045; 1993.

Ni and Cr concentration (mass %)

500 20.0 16.0

Immersed condition 4.5 m s–1, 2604 h

Material: ODS Temperature: 973 K

12.0 Cr 8.0 4.0 Ni 0.0 0

50 100 150 200 250 Distance from sodium exposed surface (μm)

Figure 17 Typical result of Ni diffusion for the ODS steel in sodium. Reproduced from Yoshida, E.; Kato, S. J. Nucl. Mater. 2004, 329–333, pp 1393–1397.

Material Performance in Sodium

reactions. Depending on the condition of failure, the temperature may exceed 1200
C. Furthermore, the combination of corrosion by the aforementioned chemical compounds and erosion by the jet blast causes damage to the steam generator tubes at a significant rate. This damage behavior is called wastage. Figure 18 shows an example of wastage test results from Monju safety analysis. The wastage rate of the steam generator tubes increases as sodium leak rate increases. Furthermore, the progression of wastage may cause an unstable fracture called a high-temperature rupture. The phenomenon was observed on the superheater tubes in the Dounreay’s Prototype Fast Reactor (PFR) in the UK. 5.13.6.2

2Na½s;l þ 1=2 O2 ½g ! Na2 O½s;l

½Fe1:75 ¼ kt

½18

0 0
Hr298 ¼ 414:6KJmol1 ;
Gr298 ¼ 147:2KJmol1

2Na½s;l þ O2 ½g ! Na2 O2 ½s;l

½19

0 0
Hr298 ¼ 510:9KJmol1 ;
Gr298 ¼ 447:5KJmol1

Wastage rate (mm s–1)

100

10-1 2.25 Cr–1Mo steel

10-2

10-3

Although sodium peroxide is a thermodynamically stable oxide in air (P(O2) = 0.21 atm), sodium oxide usually exists stably due to the reduction effect of nonburned sodium. The corrosion reaction occurring in this case is the same as that caused by chemical reaction with the oxygen dissolved in sodium. The corrosion is expressed by the formula shown subsequently. Note that the formula shows the maximum corrosion in the state where the Na2O that is necessary for the progress of the corrosion is always supplied. In air, where an excess of moisture is supplied to the sodium combustion, the corrosion reaction known as the molten salt type may occur. The corrosion was caused by the peroxide ion in the molten salt pool composed of NaOH-containing Na2O and Na2O2.26

Sodium Leak

When high-temperature sodium is leaked into the atmosphere, it reacts with oxygen, and sodium oxides are formed by the following chemical reactions:

Austenitic stainless steel

10−1

100 Leak rate (g s–1)

339

½20

k ¼ 2:01101 expð17 100=RT Þ where [Fe] is the number of reaction moles (mol cm2), t is the time (h), R is the gas constant, and T is the temperature (K).

5.13.7 Tribology 5.13.7.1

Self-Welding

Parts of the components in FBRs that come into contact and rub are the pad part of the fuel assembly, the contact part between the fuel cladding tubes and the wrapping wire, the axle hole of the mechanical pump, the control rod actuator, and the steam generator tube support. In sodium, the oxides of the component surface are reduced with sodium. Therefore, it is easy for self-welding and frictional wear to occur.27 Although frictional wear is also observed in air, self-welding is peculiar to sodium. Therefore, hard-facing material, such as Co–Cr base alloy (e.g., Stellite®) and Ni–Cr base alloy (e.g., Colmonoy), is used for such parts to protect against the tribological phenomena. The self-welding phenomenon is based on the diffusion of the metallic elements that occurs between the contact surfaces of the material. Therefore, the dominant factors in self-welding are temperature, stress, and time.

101

Figure 18 Relationship between leak rate and wastage rate. Reproduced from Tanabe, H.; Himeno, Y. Genshiryoku-kogyo 1988, 34(1), 69–76.

5.13.7.2

Frictional Wear

It is important to maintain sodium at a high level of purity from the viewpoint of the inhibition of

340

Material Performance in Sodium

Dynamic friction coefficient

1.0

Material: Chromium-carbide (LC-1H) Temperature: 600 ⬚C

0.8

3. 4. 5. 6.

0.6 7.

0.4

0.2

8. 9.

0 100

220 140 180 Cold trap temperature (⬚C)

260

Figure 19 Relationship between oxygen concentration in sodium and friction characteristics. Reproduced from Yoshida, E.; Hirakawa, Y. In-Sodium tribological study on cobalt-free hard-facing materials for contact and sliding parts of FBR components. In Fourth International Conference on Liquid Metal Engineering and Technology, Avignon, France, 1988, Vol. 2.

material corrosion. On the other hand, since the oxides formed on the surface are reduced, frictional wear is promoted. When the oxygen concentration of sodium increases, complex sodium oxides and material elements form on the surface. The oxides function as a lubricant, and frictional wear is reduced. For example, the relationship between the oxygen concentration in sodium and friction characteristics is shown in Figure 19. On the other hand, it is known that there is an intimate relation between frictional wear and material hardness. Under the same operating conditions, frictional wear decreases as material hardness increases. Therefore, to control frictional wear, generally, surface-hardening materials are used for the components of nuclear plants. Co–Cr base alloy, Ni–Cr base alloy, and chromium carbide are used as the surface-hardening materials for FBRs.

References 1. 2.

Borgstedt, H. U.; Mathews, C. K. Applied Chemistry of the Alkali Metals; Plenum Press: New York, 1987; ISBN 0–306–42326-X. Stanaway, W. P.; Thompson, R. Solubility of metals, iron and manganese in sodium. In Proceedings of the 2nd

10. 11.

12. 13. 14. 15.

16. 17.

18.

19.

20.

21. 22. 23. 24.

International Conference on Liquid Metal Engineering and Technology, Harwell, Oxford, 1980; p 1854. Claar, T. D. Reactor Technol. 1970, 13(2), 124. Stanaway, W. P.; Tompson, R. Liquid Metal Systems; Prenum Press: New York, 1982; p 421. Iizawa, K. Nucl. Eng. 1987, 33(11), 62. Eichelberger, R. L. The Solubility of Oxygen in Liquid Sodium: A Recommended Expression; Atomics International: Canoga Park, CA, 1968; AI-AEC-12685. Furukawa, T.; Yoshida, E.; Aoto, K.; Nagae, Y. The hightemperature chemical reaction between sodium oxide and carbon steel. In Proceedings of the Symposium on High Temperature Corrosion and Material Chemistry, The Electrochemical Society, 98–9, San Diego, CA, 1998; pp 312–323. Thorley, A. W.; Tyzack, C. Liquid Alkali Metals; BNES: London, 1973. Weeks, J. R.; Klamut, C. J.; Gurinsky, D. H. Alkali Metal Coolants; H M Stationery Office: London, 1967; E C 1, pp 3–24. Zebroski, E. L. Liquid Alkali Metals; BNES: London, 1973; pp 195–211. Kolster, R. H. Corrosion transport and deposition of stainless steel in liquid sodium. In Proceedings of the International Conference on Liquid Metal Technology in Energy Production, Champion, PA, 1976; p 368. Bagnall, C.; Jocobs, D. C. Relationship for corrosion of Type 316 stainless steel in liquid sodium, 1975; WARDNA-3045–23. Maruyama, A.; Nomura, S.; Kawai, M.; Takani, S.; Ohta, Y.; Atsumo, H. J. Atomic Energy Soc. Jpn. 1984, 26, 59. Roy, P. ASME Publ., 74-PWR-19, 1975. Snyder, R. B. An analysis of carbon transport in the EBR-II and FFTF primary sodium systems. In Proceedings of the International Conference on Liquid Metal Technology in Energy Production, Champion, 1976. Ito, T.; Kato, S.; Aoki, M.; Yoshida, E.; Kobayashi, T.; Wada, Y. J. Nucl. Sci. Technol. 1992, 29(4), 367–377. Wada, Y.; Yoshida, E.; Kobayashi, T.; Aoto, K. Development of new materials for LMFBR components – evaluation on mechanical properties of 316FR steel. International Conference on Fast Reactors and Related Fuel Cycles, I, Kyoto, Japan, p 7.2. ‘The Tentative Materials Strength Standard of Advanced Austenitic Stainless Steel (PNC1520) for FBR’s Core Components’ Research report of the Japan Atomic Energy Agency; PNC TN9410 90–051; 1990. Ito, T.; Yoshida, E.; Kobayashi, T.; Kimura, S.; Wada, Y. Materials Design Base Standard Supplement (Tentative) of High Strength Ferritic/Martensitic Steel (PNC-FMS) Core Components for LMFBR, Decrease in Thickness by Sodium Corrosion, Strength Reduction Factors by Decarburization and Fatigue Curves; Research report of the Japan Atomic Energy Agency, PNC ZN9410 93–045; 1993. Kaito, T.; Mizuta, S.; Uwaba, T.; Ohtsuka, S.; Ukai, S. The Tentative Materials Strength Standard of ODS Ferritic Steel Cladding; Research report of the Japan Atomic Energy Agency, PNC TN9400 2005–015; 2005. Yoshida, E.; Kato, S.; Wada, Y. Liquid Metal Systems; Plenum press: New York, 1995. Yoshida, E.; Aoki, M.; Kato, S.; Wada, Y. In 28th Symposium on High Temperature Strength; The Society of Materials Science: Japan, 1990. Wada, Y. Genshiryoku-kogyo; 1990, 36(3); in Japanese. Asayama, T.; Kagawa, H.; Komine, R.; Wada, Y. J. Mech. Behav. Mater. VI 1991, 2.

Material Performance in Sodium 25. Yoshida, E.; Komine, R.; Ueno, F.; Wada, Y. Liquid Metal Systems; Plenum press: New York, 1995. 26. Aoto, K. In Proceedings of the Symposium on High Temperature Corrosion and Materials Chemistry, 98–9, San Diego, CA, 1998; pp 287–298. 27. Yoshida, E.; Hirakawa, Y. In-Sodium tribological study on cobalt-free hard facing materials for contact and sliding parts of FBR components. In Fourth International Conference on Liquid Metal Engineering and Technology, Avignon, France, 1988, Vol. 2.

28.

29. 30.

341

Materials-oriented Little Thermodynamic Database for Personal Computers (MALT-2). The Japan Society of Calorimetry and Thermal Analysis, 1992. Weeks, J. R. In Proceedings of Symposium on Chemical Aspects of Corrosion and Mass Transfer in Liquid Sodium, 1971; pp 207–222. Menken, G. In 2nd International Conference on Liquid Metal Technology in Energy Production, XIII-2–3, Richland, WA, 1980.

5.14

Spent Fuel Dissolution and Reprocessing Processes

J.-P. Glatz European Commission, Joint Research Centre, Institute for Transuranium Elements, Karlsruhe, Germany

ß 2012 Elsevier Ltd. All rights reserved.

5.14.1 5.14.2 5.14.3 5.14.3.1 5.14.3.2 5.14.3.2.1 5.14.3.2.2 5.14.3.2.3 5.14.3.2.4 5.14.3.2.5 5.14.3.2.6 5.14.3.2.7 5.14.3.3 5.14.4 5.14.4.1 5.14.4.1.1 5.14.4.1.2 5.14.4.1.3 5.14.4.1.4 5.14.4.2 5.14.4.2.1 5.14.4.2.2 5.14.4.2.3 5.14.4.2.4 5.14.4.3 5.14.4.3.1 5.14.4.3.2 5.14.4.3.3 5.14.4.3.4 5.14.4.3.5 5.14.4.3.6 5.14.4.3.7 5.14.4.3.8 5.14.4.3.9 5.14.4.4 5.14.5 References

Introduction Fuel Cycle Industrial Reprocessing The Irradiated Fuel The Process Scheme Shearing/dissolution/off-gas treatment Dissolver product liquor conditioning Hulls and fines handling Solvent extraction Product finishing Reprocessing waste management High-level waste Safeguarding and Criticality of the Reprocessing Advanced Reprocessing Advanced Aqueous Reprocessing Uranium extraction Coextraction of actinides Direct extraction Purex adapted for Np recovery Extended PUREX Process for MA Recovery Fundamental studies Extraction mechanisms Separation of trivalent actinides from lanthanides Process development Pyro-reprocessing IFR pyroprocess European pyrochemistry projects Basic data acquisition Core processes Electrorefining on solid aluminum cathode in molten chloride media Exhaustive electrolysis Liquid–liquid reductive extraction in molten fluoride/liquid aluminum Technical uncertainties of the pyro-reprocessing Head-end conversion processes The Direct Use of Pressurized Water Reactor Spent Fuel in CANDU Process Outlook

Abbreviations ADS AEA AFCI

Accelerator-driven system Global Consulting Firm based in the UK Advanced Fuel Cycle Initiative

AREVA ASTRID

345 345 346 347 348 348 348 348 349 349 349 349 350 350 352 352 352 353 353 353 353 354 355 355 356 357 358 359 359 359 361 362 363 363 365 365 366

International Group and World leader in the energy sector Advanced Sodium Technological Reactor for Industrial Demonstration

343

344

Spent Fuel Dissolution and Reprocessing Processes

ATALANTE

Major Nuclear Cycle R&D facility in Marcoule (France) BNFL British Nuclear Fuel BPP Bismuth phosphate process BTP Bis-triazine-pyridine BTBP Bis-triazine-bis-pyridine BUTEX b,b0 -dibutyoxydiethyl ether. A process-based on a solvation extraction CANDU CANada Deuterium Uranium Reactor CEA Commissariat a` l’e´nergie atomique et aux e´nergies alternatives CMPO n-octylphenyl-N,N-diisobutylcarbamoylmethylphosphine oxide. COEX Coextraction of actinides CRIEPI Central Research Institute of Electric Power Industry DIAMEX Diamide extraction DIDPA Disodecylphosphoric acid DIREX Direct extraction DMDBTDMA Dimethyldibutyltetradecylmalon amide DMDCHMA Dimethyldicylohexanomalonamide DMDOHEMA Dimethylsioctylhexylethoxymalon amide DMDPhMA Dimethyldiphenylmalonamide DTPA Diethylentriaminepentacetic acid DUPIC Direct use of pressurized water reactor spent fuel in CANDU EDX Energy-dispersive X-ray spectroscopy analysis ENEA Italian National Agency for New Technologies, Energy, and Sustainable Economic Development EBR-II Experimental Breeder Reactor-II EURATOM European Atomic Energy Community FP Fission products FZ Ju¨lich Forschungszentrum Ju¨lich GENIV Generation IV GIF Generation IV International Forum GNEP Global Nuclear Energy Partnership HDEHP Diethylhexylphosphoric acid HEDTA Hydroxyethyl ethylenediamine triacetic acid HLLW High-level liquid waste HLW High-level waste ILW Intermediate-level waste IFR Integral fast reactor

INL ITU JAEA JNC JRC KAERI LLW LWR MA MELOX METAPHIX MOX NAS NMR NOx NPT OMEGA ORNL PHENIX PUREX PREFRE P&T PWR QSAR R&D REDOX RIAR SANEX SEM SETFICS

SPIN TBP THORP TOPO TPTZ TRPO TRU TRUEX UREX

Idaho National Laboratory Institute for Transuranium Elements Japan Atomic Energy Agency Japanese Nuclear Cycle Development Institute Joint Research Centre Korea Atomic Energy Research Institute Low-level waste Light water reactor Minor actinides Plant design (MOX fuel manufacturing) Metallic fuel irradiation ad PHENIX Mixed oxide National Academy of Sciences Nuclear magnetic resonance Nitrogen oxides Nuclear Nonproliferation Treaty Options for Making Extra Gains from Actinides Oak Ridge National Laboratory French Fast Reactor Plutonium and uranium extraction Fuel Reprocessing Plant (Tarapur, India) Partitioning and transmutation Pressurized water reactor Quantitative structure–activity relationship Research and Development Reduction–oxidation reaction Research Institute of Atomic Reactors (Dimitrovgrad, Russia) Selective actinide extraction progress Scanning electron microscope Solvent Extraction for Trivalent f-Elements Intra-group Separation in CMPO-Complexant System SeParation–Incineration Tributyl phosphate Thermal oxide reprocessing plant Trioctylphosphinoxide Tripyridyltriazine Trialkyl phosphine oxide Transuranium elements Transuranium extraction process Uranium extraction process

Spent Fuel Dissolution and Reprocessing Processes

5.14.1 Introduction The first large-scale nuclear reactors were designed for the production of weapon grade plutonium during the Second World War. It is obvious that the reprocessing technology was focused on the extraction of plutonium from the irradiated fuel. The bismuth phosphate process (BPP) was the first process to be developed and tested in the early 1940s at the Oak Ridge National Laboratory (ORNL) and scaled up to the kilogram scale in 1944 at the Hanford site. This precipitation process had already been used in 1942 by Glenn Seaborg to separate microgram quantities of Pu. However, the recovery of uranium is not possible. In the BPP process, the irradiated fuel is dissolved in nitric acid and the Pu precipitated with the fission products (FPs) using sodium phosphate and bismuth nitrate as Pu3(PO4)4 after adjustment of the valence with sodium nitrite to Pu(IV). To separate Pu from the FPs, the precipitate is redissolved in nitric acid, Pu is oxidized to Pu (VI), and the FPs are reprecipitated. Several cycles are necessary to achieve sufficient decontamination. The first solvent extraction process used in reprocessing is the reduction–oxidation reaction (REDOX) process, a continuous process where both uranium and plutonium are recovered at high yields and with high decontamination factors from FPs. Both uranyl and plutonyl nitrates are selectively extracted from dissolved fuel. After development at the Argonne National Laboratory and testing at the pilot scale at the ORNL from 1948 to 1949, a REDOX plant was built in Hanford in 1951. The b,b0 -dibutyoxydiethyl ether (BUTEX) process utilizes a dibutyloxydiethyl ether solvent and nitric acid. This process was developed in the late 1940s at the Chalk River Laboratory and operated at an industrial scale at the Windscale plant in the UK until 1976. Again at ORNL in 1949, a successful solvent extraction process for the recovery of pure uranium and plutonium was developed, initially to separate 239-Pu for military purposes. The plutonium and

Table 1

345

uranium extraction (PUREX) process was invented by Herbert H. Anderson and Larned B. Asprey at the Metallurgical Laboratory at the University of Chicago, as part of the Manhattan Project.1 The so-called PUREX process is still the standard method of extraction for the reprocessing of commercial nuclear fuels. The first industrial reprocessing plant for commercial fuels was the UP1 facility at Marcoule in France. During the 1960s and 1970s, reprocessing activities were launched in Belgium, France, Germany, India, Japan, the Russian Federation, the United Kingdom, and the United States. For various reasons, however, only some of these plants are still in operation (see Table 1), namely, at the International Group and World leader in the energy sector (AREVA) NC La Hague site in France, the THermal Oxide Reprocessing Plant (THORP) operated by the British Nuclear Group Sellafield (BNGSL) in Sellafield in the United Kingdom, the RT-1 plant in Mayk in Russia, the PREFRE facility in Tarapur, India, and, since 2010, the Rokkasho plant operated by JNFL in Japan. The RT-1 facility in Mayak is the only plant where fast reactor fuel, from the BN 600 reactor, is reprocessed on a large scale. The total amount of used fuel cumulatively generated worldwide by the beginning of 2010 was approximately 300 000 tons HM. Between now and 2030, some 400 000 tons of used fuel is expected to be generated worldwide, including 60 000 tons in North America and 69 000 tons in Europe. Worldwide, the used fuel generated in 2010 was in the order of 11 000 tons HM. About one-third of the fuel inventory is reprocessed at present; the rest is placed into interim storage facilities, mostly at the reactor sites.

5.14.2 Fuel Cycle The various activities associated with the production of electricity from nuclear reactions are referred to

Major commercial reprocessing plants in operation today

Plant

Country

Site

In operation since

Capacity (tons/year)

UP2 THORP RT-1 PREFRE RRP

France United Kingdom Russia India Japan

La dHague Sellafield Mayak Tarapur Rokkasho-Mura

1990 1994 1976 1982 2009

800 1000 400 150 800

346

Spent Fuel Dissolution and Reprocessing Processes

Reprocessing plant High level waste

U mining

Uranium storage Depleted uranium

Natural uranium Enrichment

Processing

Fissile and fertile

Repository

Spent nuclear fuel

Fuel fabrication

SNF storage

Spent fuel storage

Reactor

Nuclear reactor

Figure 1 The nuclear fuel cycle.

collectively as the nuclear fuel cycle (see Figure 1). The nuclear fuel cycle starts with the mining of uranium and ends with the disposal of nuclear waste. When using reprocessing of used fuel as an option for nuclear energy, the different stages form a true cycle. Nuclear energy systems of the future, as they were defined by the Generation IV International Forum (GIF), are supposed to provide a sustainable energy generation for the future (http://www.ne.doe.gov/ genIV/neGenIV1.html). The corresponding fuel cycles will play a central role in the achievement of this goal. The major benefits of used fuel recycling are the conservation of natural uranium resources, reduced dependence on foreign fossil fuel, and reduction of the nuclear waste radiotoxicity and the heat load of repositories. Major challenges to the implementation are significant costs, safety, and increasing proliferation concerns, also affecting the public acceptance of this technology. The present reactors use less than 1% of the uranium available in nature. With such a low efficiency, the uranium resources identified worldwide will be sufficient for only about 100 years with the currently installed nuclear power infrastructure. Depending on the growth rate in the use of nuclear systems in the future, this time span could be significantly lower. New energy systems using a technology based on the combination of fast neutron reactors with advanced multirecycling of the fuel would improve the usage of natural uranium resources by at least a factor of 50. The new reactor concepts under development will be able to recycle not only most of the fertile and fissile uranium and plutonium but also the other long-lived actinides produced in the nuclear fuel.

The consequence is that on one hand the fuel refabrication will be more complex and difficult, but on the other, the long-term waste radiotoxicity can be considerably reduced. All this should be achieved while maintaining or even improving the safety and the economic competitiveness, and minimizing the risks of proliferation. It is obvious that this change toward an enhanced sustainability is a progressive process, which has already started. Indeed, the current industrially operated fuel recycling technologies are being constantly improved and optimized in view of natural resource utilization and economic competitiveness.

5.14.3 Industrial Reprocessing The reprocessing of used commercial fuel is done exclusively by the PUREX extraction process. In a reprocessing facility, the used fuel is separated into three fractions: uranium, plutonium, and waste, which contains FPs and minor actinides (MAs). Reprocessing enables recycling of the uranium and plutonium into fresh fuel. Since 2004, commercial reprocessing is used by the nuclear industry in several countries to separate and reuse plutonium in a mixture with uranium as mixed oxide (MOX) fuel in electricity producing reactors. The first irradiation of MOX was done in 1960 in the BR3 reactor in Belgium. Today, the world’s largest MOX fabrication facility called MELOX, with a capacity of 1500 HM/year, is operated by AREVA in Marcoule in the South of France.

347

Spent Fuel Dissolution and Reprocessing Processes

In some countries, reprocessed uranium is also reused after enrichment as nuclear fuel. The uranium from reprocessing, which typically contains a slightly higher concentration of U-235 than that occurring in nature, can be reused as fuel after conversion and enrichment. However, reprocessed uranium also contains U-236, typically 0.5%, which increases at higher burn-up. This isotope is a neutron absorber; therefore, only reprocessed uranium from low-burn-up fuel is reused in light water reactors (LWRs), while that from high burn-up fuel is best used for blending or MOX fuel fabrication. 5.14.3.1

The Irradiated Fuel

Generation II reactors were typically designed to achieve a burn-up of about 40 GWd/MTU. With the improved fuel technology, these same reactors are now capable of achieving up to 60 GWd/MTU, and research and development (R&D) efforts are ongoing to further increase this burn-up value. The incentive is the achievement of a better economy of the energy production process: To produce a given amount of energy, a smaller number of fresh nuclear fuel elements are required and a lesser amount of used nuclear fuel elements are generated; furthermore, as a consequence of this, the downtime for refueling is reduced. At some stage, however, the build-up of FP neutron poisons achieves values that necessitate the reactors being shut down and refueled. Used fuel is a highly radioactive and very complex material, and at an average burn-up of 45 GWd tons1, it contains about 94% U-238, approximately 1% U-235 that has not fissioned, almost 1% plutonium, and 4.5% FPs with the following approximate composition: Rare earths, Y: 24% Ru, Tc, Rh, Pd: 16% Kr, Xe: 15% Zr, Nb: 14%

Mo: 13% Cs, Rb, I, Te: 11% Ba, Sr: 7% Depending on their thermophysical behavior during irradiation, the FPs exhibit a totally different behavior. A detailed classification of FPs was published by Kleykamp in 1985.2  Dissolved in the matrix: Rb, Sr, Y, Zr, Nb, Te, Cs, Ba, La, Ce, Pr, Nd, Pm, Sm, Eu  Partly precipitated at grain boundaries (oxides): Rb, Sr, Zr, Nb, Mo, Se, Te, Cs, Ba  Metallic precipitates: Mo, Tc, Ru, Rh, Pd, Ag, Cd, In, Sn, Sb, Se, Te  Volatiles: Br, Kr, Rb, I, Xe, Cs, Te Especially at the beginning of the irradiation process when the fission event density is the highest, leading to the highest linear power, a significant relocation of FPs takes place, depending on their volatility. In fact, in an oxide fuel, temperature gradients of at least 500  C between the fuel periphery (500  C) and the fuel center (>1000  C) lead to significant migration and diffusion processes. The grain structure of the fuel initially produced by pressing UO2 powder, induces under irradiation precipitation of some of the FPs at the grain boundaries; noble elements partially form metallic precipitates. The most volatile elements can migrate outside of the fuel pellets where they are deposited or potentially form compounds, with the cladding material as well. Parts of the volatiles are found in the fuel rod plenum. The above-mentioned burn-up also has a considerable impact on the content of transuranic elements which are formed by neutron capture of U-238. Table 2 shows the composition (major transuranium (TRU) elements and some FPs) of LWR fuels at various burn-ups in comparison to MOX fuel. Especially for Cm, the content is increased by almost a factor of 10 if the burn-up is increased from 33 to 60 GWd tons1. A similar increase is

Table 2 Composition (major transuranium elements and some fission products) of LWR fuels at various burnups in comparison to MOX fuel Fuel type

LWR 1

Average burn-up (GWd t ) Constituent

1

Pu (g tU ) Np (g tU1) Am (g tU1) Cm (g tU1) Zr (g tU1) Tc (g tU1) Ru (g tU1)

MOX

33

45

60

45

9.740 433 325 23 3.580 814 2.165

11.370 611 521 92 4.740 1.085 3.068

12.990 887 765 213 6.280 1.403 4.156

48.850 161 4.480 810 3.440 977 3.924

348

Spent Fuel Dissolution and Reprocessing Processes

observed for MOX and LWR fuels at the same burnup of 45 GWd tons1. New generation fast reactors are using MOX fuel with Pu content before irradiation of about 20% instead of 5% in LWRs and because they are less sensitive to increasing amounts of FPs, burn-ups up to 200 GWd tons1 are possible. It is obvious that all this will have a major impact on the reprocessing process. 5.14.3.2

The Process Scheme

The well-proven hydrometallurgical PUREX process used by the commercial reprocessing plants involves the dissolution of the fuel elements in 5–6 M nitric acid, the extraction of uranium and plutonium by the tributyl phosphate (TBP) solvent, the chemical separation of uranium, and a conditioning of the products (see Figure 2). The raffinate of the extraction process is a high active waste (HAW) solution, which contains the major part of the FPs and the MAs. Uranium and plutonium can be returned to the fuel cycle – the uranium to the conversion plant prior to re-enrichment and the plutonium to MOX fuel fabrication. 5.14.3.2.1 Shearing/dissolution/off-gas treatment

The fuel elements are transferred to the dissolver equipment, where the shearing equipment cuts the fuel pins into segments of a few centimeters to ensure effective fuel dissolution. Dissolver systems with a critically safe geometry can be operated in a continuous or in a batch mode. For high throughputs in large-scale reprocessing, continuous rotary dissolvers are preferred. The sheared fuel falls into the dissolver basket where it is immersed in hot nitric acid, contained in the dissolver. Similar reactions can be written for the direct dissolution of the uranium oxide fuel pellets (not showing the dissolution of the remaining actinides and FPs):

Spent fuel

HNO3

TBP solvent

Shear

Spent fuel dissolver

Extraction

Off-gas

Hulls storage

Vitrified HAW storage

Figure 2 Simplified PUREX process scheme.

3UO2 þ 8HNO3 ¼ 3UO2 ðNO3 Þ2 þ 2NO þ 4H2 O

½1

The basket retains the bulk insolubles contained in the fuel and the cladding material, also called hulls, allowing them to be removed from the vessel after the dissolution process is complete. Finer insoluble solids, not retained in the basket, are removed with the product liquor and separated subsequently by settling or centrifugation, according to their size. Insolubles are washed before being removed. Further, the off-gas containing mainly nitrogen oxides, iodine, ruthenium, carbon 14, fuel dust, and aerosols is treated in a dedicated off-gas treatment plant before being either recycled (NOx) or discharged to the atmosphere. 5.14.3.2.2 Dissolver product liquor conditioning

Following its removal from the dissolver, the product liquor containing the dissolved uranium, plutonium, MAs, and FPs, clarified from any solid material, together with recovered washings is accurately measured for adherent radioactive material, before further conditioning. Therefore, accountancy measurement tanks are fitted with highly efficient mixing systems, multilevel sampling, high accuracy level determination and density instrumentation, and very precise tank weighing systems. After accountancy determination, the liquor is transferred to conditioning tanks for further adjustments, necessary for the solvent extraction process. 5.14.3.2.3 Hulls and fines handling

The hulls are checked to be free of residual fuel and product liquor using gamma spectrometry and neutron measurement techniques (active and passive). In the unusual case of a high residual fuel content, the hulls are returned to the dissolver for further treatment; otherwise, they are either compacted or encapsulated in a cement matrix. The insoluble residues removed from the product liquor are added to the calcined high-level waste (HLW) for vitrification.

Uranyl nitrate

U evaporator

UO2 conversion

UO2 storage

Pu evaporator

MOX conversion

MOX storage

U, Pu separation

Pu nitrate

Spent Fuel Dissolution and Reprocessing Processes

5.14.3.2.4 Solvent extraction

The central part of the reprocessing is of course the solvent extraction based on the well-proven PUREX process (see Figure 2). The solvent is TBP diluted with odorless kerosene. The extraction happens through formation of an uranylnitrato complex with two TBP molecules in the organic phase according to the following equation:  UO2þ 2 þ 2NO3 þ 2TBP ¼ UO2 ðNO3 Þ2  2TBP ½2

For the primary separation cycle to remove FPs and to separate uranium and plutonium, a series of pulsed columns are used. The aqueous, highly active raffinate containing the FPs from the primary separation cycle is treated by a water steam strip to remove residual solvent. After storage, the solution is concentrated and immobilized by vitrification in view of a final disposal. This vitrification process shows high flexibility because insoluble residues (see Section 5.14.3.2.3) and alkaline effluents from the solvent regeneration can also be incorporated in the glass matrix. Uranium and plutonium in the solvent phase are separated by adding uranium IV which acts as a plutonium reductant. The reduced plutonium is back extracted into an aqueous phase which is routed to the plutonium purification and finishing lines. Where possible, equipment is designed to operate without routine maintenance during the life of the plant. Equipment in contact with radioactivity can be remotely cleaned and dismantled. In cases where contaminated equipment must be maintained, it may be remotely dismantled and rebuilt, or in other cases, it is routed to special decontamination plant systems to allow contamination to be removed and also to allow ‘hands on’ maintenance. Because of the time involved in this type of activity, duplicate spares are generally provided for units requiring routine removal for decontamination and maintenance. Radioactively contaminated components are consigned for disposal or waste treatment. Appropriate materials have to be selected according to the requirements of each item of equipment. In addition, the integrity of all process equipment in contact with active materials has to be ensured by quality control during manufacturing, installation, inspection, and testing, in order to minimize maintenance requirements and plant downtime. Stainless steel is the standard material used in the construction of the majority of the process systems, with special materials such as titanium or zirconium utilized for particularly demanding applications.

349

All materials to be used in hot cells are subject to checks for reliability in a radiation environment. Radiation-sensitive items are either located outside the hot cells or locally shielded to minimize radiation effects. Significant progress has been achieved in the development of suitable materials. However, even more reliable materials are needed and R&D efforts are continuing with a view to enhancing the qualities of materials used in modern plants.3 5.14.3.2.5 Product finishing

After purification, the plutonium is precipitated by addition of oxalic acid. The plutonium oxide product, which is produced by calcination of the oxalate, is packaged in stainless steel containers. These containers are arranged in a way to provide a criticality safe geometry for storage. The solvent loaded with uranium from the primary separation cycle passes to purification and the resulting uranyl nitrate solution is evaporated and converted to uranium trioxide by thermal denitration. The uranium trioxide product is packaged in drums for interim storage in an engineered storage. Both the uranium and the plutonium products are produced to internationally agreed specifications and in a form suitable for recycling. 5.14.3.2.6 Reprocessing waste management

A number of categories of radioactive waste are defined, each of them requiring a specific management approach. HLW is defined as the category of waste where the heat generated by radioactive decay significantly affects the design of the waste management route. Solid low-level waste (LLW) is defined as the solidwaste with radioactivity levels less than the authorized limits for the shallow land disposal. Intermediatelevel wastes (ILW) are those wastes between HLW and LLW. In addition, very low-level liquid and aerial effluents are produced, which are discharged into the environment, provided their monitoring shows compliance with discharge authorization values. 5.14.3.2.7 High-level waste

The major waste fraction from the radioactivity point of view is the HLW. The general management strategy internationally adopted for this type of waste is the storage of the liquor for radioactive decay in storage tanks. The aqueous solution of FPs and MAs is concentrated up to about a factor of 15 before it is vitrified at 1150  C using a borosilicate glass matrix (see Chapter 5.18, Waste Glass). In France, a cold crucible vitrification process is currently

350

Spent Fuel Dissolution and Reprocessing Processes

proposed as a replacement for the conventional system, aiming at a simplified single-step process. Commercial vitrification plants in Europe produce about 1000 tons per year of such vitrified waste (2500 canisters) and some have been operating for more than 20 years. The glass properties must be guaranteed to ensure the satisfactory long-term performance of the waste package. The alteration behavior of the glass is therefore assessed against the performance criteria required for interim storage or disposal purposes. 5.14.3.3 Safeguarding and Criticality of the Reprocessing The goal to foster the peaceful uses of nuclear energy based on the Treaty on the Non-Proliferation of Nuclear Weapons (NPT) is achieved through the implementation of a highly efficient safeguarding process at reprocessing plants. The particular interest in bulk-handling facilities like reprocessing plants where large quantities of plutonium are handled is obvious. Nuclear material flows (in or out) are monitored at key measurement points, such as storage areas (tanks, containers, used fuel ponds), the headend fuel treatment, shearing and dissolution area, and product storage area (plutonium, uranium). The National Academy of Sciences (NAS) has declared that the large and growing stocks of plutonium from weapons dismantlement in the United States and the former Soviet Union are a ‘clear and present danger’ to peace and security. Moreover, experts consider that plutonium of any isotopic blend is a proliferation threat; this means of course that plutonium produced in the civilian fuel cycle is itself a proliferation threat. Assuring that separated plutonium, from dismantled warheads as well as from civilian power programs, is under effective control has (again) become a high priority worldwide. If plutonium is considered as an energy resource, it is mandatory to safeguard it against diversion, putting it into active use in the civilian power program. The ultimate choice cannot be separated from the long-term strategy for use of peaceful nuclear power. However, continued use of a once-through fuel cycle will also lead to an ever-increasing quantity of excess plutonium, requiring safeguarding as well. Alternatively, recycling the world’s stocks of plutonium in fast reactors will cap the world supply of plutonium and hold it in working inventories for generating power. Transition from the current-generation LWRs to a future fast-reactor-based nuclear energy supply under

international safeguards would limit world plutonium inventories to the amount necessary and useful for power generation, with no further excess production. A concept like the integral fast reactor (IFR) in the United States foresees complete recycle of plutonium, and indeed, of all transuranics, with essentially no transuranics sent to waste, so the need for perpetual safeguards of IFR waste is eliminated. The pyrorecycle process is more proliferation resistant than the current PUREX process because at every step of the IFR recycle process the materials meet the ‘usedfuel standard.’ The scale of IFR recycle equipment is compatible with colocation of power reactors and their recycle facility, eliminating off-site transportation and storage of plutonium-bearing materials. Self-protecting radiation levels are unavoidable at all steps of the IFR cycle, and the resulting limitation of access contributes to making covert diversion of material from an IFR very difficult to accomplish and easy to detect. Another key issue for any reprocessing activity is the criticality. As already mentioned several times in the process description section above, the criticality control in the PUREX process is mandatory throughout the process scheme and in this respect plutonium is a key element, especially in view of increasing burn-up, the usage of MOX fuel, and in the long term the implementation of fast reactor systems. The factors that mainly affect criticality safety are  the fissile nuclides (235U, 238Pu, and to a lesser extent 233U);  the fraction of fertile nuclide diluting fissile nuclides (238U and 240Pu);  the mass and concentration of fissile nuclides;  the geometries and volumes of fissile materials in the facility; and  the neutron moderators, reflectors, and absorbers.

5.14.4 Advanced Reprocessing A sustainable energy generation for the future with the major objectives of effective fuel utilization and waste minimization through recycling of all actinides can only be achieved with substantial modification of the corresponding fuel cycles. The waste minimization goal is in fact based on a waste management strategy, its main motivation being the reduction in the long-term radiotoxicity. In this partitioning and transmutation (P&T) scenario studied for many decades already, long-lived radionuclides are recovered

Spent Fuel Dissolution and Reprocessing Processes

(partitioning) and converted into shorter-lived or stable isotopes by irradiation (transmutation). The transmutation efficiency should be especially high in dedicated reactors such as accelerator-driven systems (ADS), where a subcritical reactor is connected to a cyclotron or linear accelerator. Numerous research activities carried out in P&T have shown that efficient P&T scenarios can shorten the time needed for isolation of nuclear waste from >100 000 years down to about 500 years. From the viewpoint of radiotoxicity reduction of the actual waste, P&T must first concern the actinides, particularly plutonium and the MAs (mainly Am, Cm), which make up more than 99% of the radiotoxicity already after a few hundred years of storage.4 Advanced reactor systems of the IVth generation, especially those using a fast neutron spectrum, offer excellent transmutation features. Therefore, an inherent P&T scheme can be used to reduce the long-term waste radiotoxicity. On the partitioning side, one can rely on the considerable scientific and technical progress made through domestic and international projects such as SeParation–Incineration (SPIN) (France),5 Options for Making Extra Gains from Actinides (OMEGA) ( Japan),6 Global Nuclear Energy Partnership (GNEP)/Advanced Fuel Cycle Initiative (AFCI) (USA) (http://www.gnep.gov/), as well as bilateral cooperations and European Atomic Energy Community (EURATOM) Framework Programs7–11 over the last couple of decades. The most long-lived radionuclides contained in used nuclear LWR fuel are listed in Table 3. Two types of processes can be applied to the separation of long-lived radionuclides: hydrochemical (wet) and pyrochemical (dry) processes. Both have advantages and disadvantages and should be Table 3

351

applied in a complementary way. If a so-called double-strata concept, for example, as proposed in the above-mentioned OMEGA project is adopted, the well-established industrial reprocessing of commercial LWR fuel with recycling of U and Pu based on PUREX extraction should be logically combined in the first stratum with an advanced aqueous partitioning scheme, also based on liquid–liquid extraction to separate the long-lived radionuclides. In the second stratum, new generation reactor systems should preferably be combined with pyro-reprocessing, because most of the fuels under investigation for advanced reactor systems are more soluble in molten salts; shorter fuel cycles are possible because of a higher radiation resistance, and a higher proliferation resistance is due to reduced product purity. Therefore, the decision on the partitioning process to be applied should depend on the boundary conditions, such as the type of fuel material to be treated, but aqueous- and pyropartitioning are not to be seen as competitive options to achieve the partitioning of long-lived MAs and FPs from used nuclear fuel. In any case, an efficient and selective recovery of the key elements from the spent nuclear waste is absolutely essential for a successful and sustainable fuel cycle concept. This necessitates the selective separation of Am and Cm from lanthanide FPs, certainly the most difficult and challenging task in advanced reprocessing of used nuclear fuel because of the very similar chemical behavior of the trivalent elements. There are three major reasons to separate actinides from lanthanides:  Neutron poisoning: lanthanides (esp. Sm, Gd, Eu) have very high neutron capture cross sections, for example, >250 000 barn for Gd-157.

Long-lived radionuclides in used nuclear fuel

Category

Element

Isotope

Period (years)

Mass (g t1)

Isotope content (%)

Minor actinides

Np Am

237 241 243 243 244 245 79 93 99 107 126 129 135

2 140 000 432 7380 28.5 18.1 8530 65 000 1 500 000 210 000 6 500 000 100 000 15 700 000 2 300 000

430 220 100 0.3 24 1 4.7 710 810 200 20 170 360

100 67 31 1 94 5 9 20 100 16 40 81 10

Cm Fission products

Se Zr Tc Pd Sn I Cs

352

Spent Fuel Dissolution and Reprocessing Processes

 Material burden: in used LWR fuels, the lanthanide content is up to 50 times that of Am/Cm.  Segregation during fuel fabrication: upon fabrication, lanthanides tend to form separate phases, which grow under thermal treatment; Am/Cm would also concentrate in these phases. Further, the lanthanide – actinide separation can be derived from aqueous or pyrochemical partitioning processes of MAs. 5.14.4.1

Advanced Aqueous Reprocessing

The actual PUREX process is the industrial hydrochemical reprocessing technique to separate pure U and Pu fractions from used fuel. For the advanced fuel cycles mentioned above, world-wide efforts are made to use modified versions of the present PUREX process with the goal to cope with sustainability goals and to improve the economy and the proliferation resistance. 5.14.4.1.1 Uranium extraction

The US Department of Energy proposes the uranium extraction (UREX)þ process in the frame of their advanced fuel cycle development programs, where only uranium is recovered and recycled. The central feature of this concept is the increased proliferation resistance by leaving the plutonium with other transuranics for a grouped recycling in fast reactors. Several variations of the UREXþ process have been developed, with different options on how the plutonium is combined with various MAs, lanthanide, and nonlanthanide FPs. A major challenge is the fuel fabrication mainly because of the americium volatility and the fact that curium is a neutron emitter. Remote fuel fabrication facilities would be required, leading to high fuel fabrication costs and significant technological development.

Spent fuel

Shear

Off-gas

5.14.4.1.2 Coextraction of actinides

AREVA and Commissariat a` l’e´nergie atomique et aux e´nergies alternatives (CEA) have developed the COEX (coextraction of actinides) process on the basis of extensive French experience with PUREX (see Figure 3). The COEX process is based on coextraction and coprecipitation of uranium and plutonium (and usually neptunium), as well as a pure uranium stream, but without separation of a pure plutonium fraction. This process allows the production of a high-quality MOX for both light water and fast reactors. An industrial deployment for LWR-MOX is foreseen for the near term. The sodium fast reactor prototype Advanced Sodium Technological Reactor for Industrial Demonstration (ASTRID) planned for deployment in the early 1920s could also be based on the COEX process. In the longer term, the goal is to have a technology validated for industrial deployment of generation IV (GENIV) fast reactors around 2050; at this stage, the present La Hague plant will also be due for replacement around 2050. The long-term goal is to make a large capacity of spent fuel reprocessing (in the range 2000–3000 tons year1) available with a potential to further reduce the reprocessing costs and to address the potentially increasing spent fuel reprocessing needs. Another objective is to enhance the flexibility in material management with a design adapted to the treatment of a wide spectrum of fuel types, that is, legacy fuel stored for decades, newly discharged fuel for reprocessing, and fuels with high fissile isotopes content (MOX fuel, very high burn-up fuels). The goal is also to have the spent fuel reprocessing and fresh fuel refabrication on same site (limited fuel transports and storage needs). Also the implementation of MA reprocessing would be facilitated.

HNO3

TBP solvent

Spent fuel dissolution

Extraction

Hulls storage

Vitrified HAW storage

Figure 3 COEX: a simplified PUREX process scheme.

Depleted U

Coconversion

Fuel pellet manufacturing

Spent Fuel Dissolution and Reprocessing Processes

5.14.4.2 Extended PUREX Process for MA Recovery

5.14.4.1.3 Direct extraction

Another alternative reprocessing technology being developed by Mitsubishi and Japanese R&D establishments is Super-DIREX (supercritical fluid direct extraction). This technology is designed to cope with uranium and MOX fuels from light water and fast reactors. The fuel is dissolved in a mixture of nitric acid, TBP, and supercritical CO2, resulting in complexation and extraction of uranium, plutonium, and MAs with TBP.

For the separation of MAs, the PUREX process has to be modified/extended using also hydrochemical extraction techniques.13 Extensive R&D is carried out worldwide to synthesize special extractants and to develop the corresponding process schemes required for a selective separation of MAs (mainly Am and Cm) from high-level liquid waste (HLLW). The process development requires a good basic understanding on the extraction mechanisms.

5.14.4.1.4 Purex adapted for Np recovery

In the standard PUREX process, Np is partially extracted by TBP; this part follows the U stream, is separated in the second U purification cycle, and then added to the HLW and vitrified. In the fuel solutions feeded to the first decontamination cycle, Np is present as a mixture of Np (IV), Np (V), and Np (VI), but only Np(IV) is extracted. Therefore, in the PUREX process adapted for Np recovery,12 Np is completely oxidized to the oxidation state VI and then coextracted with U and Pu in the first decontamination cycle where it again follows the U stream. Finally it is, as in the standard process, recovered through a reducing scrub in the second U cycle. After separation, the Np nitrate, contaminated by b–g emitters, may be purified by solvent extraction with TBP and finally transformed to oxide by calcination of the oxalate.

5.14.4.2.1 Fundamental studies

As aqueous partitioning is based on liquid–liquid extraction from an acidic solution into an organic phase, it is crucial to understand extraction selectivity, thermodynamics, mechanisms, and kinetics. In aqueous MA partitioning schemes, two main routes are possible (see Figure 4). The optimal strategy would be of course a process, where MAs are directly extracted from the PUREX raffinate, HLLW. However, till date, no extractant capable of selective and efficient separation of the MAs at high acidities (>2 M HNO3) in a highly radioactive solution containing all FPs, among them lanthanide elements in a mass excess of 20 times compared to MAs, has been found. Partitioning of MAs involving coextraction of lanthanide (Ln) elements and a subsequent

LWR fuel Dissolved fuel PUREX

U, Pu, (Np)

HLLW

FP Selective extraction

MA extraction (org. complexant)

Coextraction of MA, Ln MA /Ln

Selective stripping MA stripping (aq. complexant)

Ln

Selective

Ln

extraction

High acid MA extraction

MA Am/Cm sep. Am

353

Cm

Transmutation Figure 4 Strategies for the separation of the minor actinides from HLLW.

Developed Future ?

FP (Ln)

354

Spent Fuel Dissolution and Reprocessing Processes

separation of the two element groups is therefore the only viable option at present. 5.14.4.2.2 Extraction mechanisms

One of the major concerns to be addressed with respect to the extraction of lanthanides (III) and actinides (III) from aqueous nitrate solutions requires the knowledge of the nature of the extracted species. A dual mechanism of extraction would be based on the formation of solvates having the general formula M(NO3)3Ln according to the following equation: M3þ þ HNO3 þ nL ¼ MðNO3 Þ3 Ln þ 3Hþ with M(III) ¼ Ln(III) or An(III) and L ¼ organic extractant. In European research programs, the reference organic extractant is based on the diamide molecule with the general formula (R(R0 )NCO)2CHR00 (where R, R0 , and R00 are alkyl or oxyalkyl groups, e.g., N,N0 dimethyl-N,N0 -dibutyltetradecyl-1,3-malonamide (DMDBTDMA); see Figure 5). For concentrated aqueous nitric acid solutions, as encountered when extracting U(VI) or actinide (IV) from nitric acid media by monoamide extractants ion-pairs, of formula [LHþ]n, [M(NO3)3þn]n3n. Several experiments, involving UV-visible and 13 C NMR (nuclear magnetic resonance) spectroscopies and solvent extraction, have been conducted to answer this question. From the data obtained so-far, one can conclude that even if a dual extraction mechanism exists, the second mechanism does not seem to be an ion-pair mechanism involving a protonated diamide. It can therefore be concluded that the occurrence of an ion-pair mechanism is unlikely. A comparison of diamides with different R0 groups (butyl, phenyl, and chlorophenyl) as regards their ability to extract An(III) or Ln(III) from aqueous nitrate media shows that a less basic malonamide has better extraction properties for the M(III) nitrate. If in the central R00 position the alkyl group is replaced by a dioctylhexylethoxy group (see Figure 6), the diamide dimethyl-dioctyl-hexylethoxy malonamide (DMDOHEMA) exhibits better affinities for M(III) nitrates.

Arrhenius activation energies close to 40 kJ mol1 for all M(III) studied indicate that the extraction is chemically limited at the aqueous–organic interphase. For a diffusion limited kinetic regime, this energy is generally found close to be 20 kJ mol1. The extraction kinetics of M(III) nitrates by DMDBTDMA were found to be much slower than for the extraction of U(VI) or Pu(IV) nitrates by TBP (extractant of the PUREX process). Crystal structures were determined by X-ray absorption spectroscopy and using synchrotron light for a large number of lanthanide – and actinide – diamide complexes. Molecular modeling studies have been conducted to compare calculated structures and X-ray determined crystal structures and to propose structural explanations for experimental differences observed during extraction of M(III) metallic nitrates by several malonamides. Using the Quanta/CHARM code, the lowest conformation calculated for dimethyldiphenylmalonamide (DMDPhMA), dimethyldicylohexanomalonamide (DMDCHMA), and BDMDPhMA structures were found to be similar to the experimentally determined crystal structures. The differences between the structures of DMDPhMA and BDMDPhMA, and of DMDCHMA were also confirmed by calculations. The differences in M(III) extraction efficiency between cyclohexano (DMDCHMA) and phenylsubstituted (DMDPhMA and BDMDPhMA) malonamides can be correlated with the difference of the preferred conformations of the malonamide extractants. Using the Gaussian 94 program, protonation of cyclohexano (DMDCHMA) and phenyl-substituted (DMDPhMA) malonamides was studied. Results are equivalent for both malonamides and show that monoprotonated malonamide contains an intramolecular hydrogen bond, while the di-protonated malonamide does not. A quantitative structure–activity relationships (QSAR) study related to the extraction of Nd(III)

O C8H17

O C4H9

N CH3

O

N

N CH3

N

C4H9

C14H29 CH3

Figure 5 N,N0 -dimethyl-N,N0 -dibutyltetradecyl-1, 3-malonamide (DMDBTDMA).

O C8H17

C2H4 CH3 O C6H13

Figure 6 N,N0 -dimethyl-N,N0 -dioctylhexylethoxymalonamide (DMDOHEMA).

Spent Fuel Dissolution and Reprocessing Processes

nitrate by a set of 17 malonamides supported the above mentioned improved M(III) nitrate extracting properties in the presence of an oxygen ether atom in the R00 substituent. 5.14.4.2.3 Separation of trivalent actinides from lanthanides

To explain the great affinity of actinides for nitrogenbearing molecules, numerous fundamental studies were carried out using a wide range of experimental methods, including spectroscopy. For Ln(III) and An (III) ions, the formula, stability, and structure of the complexes were determined both in aqueous solution and in various solvent media. It has been demonstrated that bonds between the nitrogen atoms of these ligands and Ln(III) and An(III) ions include some definite covalence. The covalence observed in bonds with the electron-donor nitrogen atoms of ligands seems higher for An(III) ions than for Ln (III) ions, and could be an indication of the greater affinity of these ligands for An(III); however, the difference is too small to really explain the sometimes very high differences in the distribution factor. Theoretical studies in the fields of quantum chemistry and molecular dynamics have provided greater insight into certain crucial aspects of reactions between these metal ions and nitrogen-bearing ligands. In particular, the synergetic extraction mechanism of Ln(III) ions using a mixture of a nitrogen-bearing ligand and a carboxylic acid has been identified by computer calculations. The calculated synergetic complex seems consistent with the experimental results. 5.14.4.2.4 Process development

Three alternative approaches are proposed. The first is based on coextraction of trivalent MAs and lanthanides (Lns) and separation of MA and Ln fractions in a second step.13 For the first part, the following are the most important processes:  The TALKSPEAK process (the Unites States)14 and disodecylphosphoric acid (DIDPA) process ( Japan)15 use acidic organophosphorus extractants.  The TRansUranium Extraction (TRUEX) process (the Unites States)16 and Solvent Extraction for Trivalent f-elements Intra-group Separation in CMPO-complexant System (SETFICS) ( Japan)17 are based on the use of CMPO (n-octyl-phenyldiisobutyl-carbomoylmethyl-phosphine-oxide).  The Trialkyl phosphine oxide (TRPO) process (China) uses a trialkyl phosphine oxide. The hot

355

demonstration of this process using genuine HLLW has been done at the Institute for Transuranium Elements (ITU) (Karlsruhe).18  The DIAMEX (diamide extraction) process using malonamides as extractant19 has been developed at CEA (France) and is also the reference process under investigation in the European partitioning projects. For an efficient recycling scheme, losses of the relevant elements should be as low as possible (0.2% or less), and a compromise between extraction and back extraction has to be made. The MA/Ln separation can be achieved by the socalled selective actinide extraction process (SANEX). The major options are as follows:  The BTP (bis-triazine-pyridine) developed at FZK-INE Germany20 or BTBP (bis-triazinebis-pyridine), which is capable of achieving the selective extraction of MAs at high nitric acid concentration (2 M).  The TPTZ (tripyridyltriazine) developed at CEA, France to be used at much lower nitric acid concentrations.21  Variants of the dithiophosphinic acids (ClPh) 2PSSH mixed with trioctylphosphinoxide (TOPO) at Forschungszentrum Ju¨lich (FZ Ju¨lich), Germany.22 Promising results have been obtained on simulated as well as on genuine solutions at lab scale. Among many extractants tested worldwide, the combination of DIAMEX and BTP (see Figure 7)23 is shown to be the best combination for an efficient recovery of MAs from HLLW or transmutation targets. Diamides do not require feed adjustment, can easily be recycled to the process, and do not leave any residue upon incineration. With regard to the separation of MAs from Ln, BTP has been shown to be the most efficient extractant, giving at the same time the highest separation factor with no feed acidity adjustment required. Separation factors between MAs and lanthanides up to 80 are reached in a single-stage extraction. These values are considerably improved in a continuous multistage process,

N N

N

N N

N

N

Figure 7 2,6-Bis-(5,6-di-isopropyl-1,2,4-triazine-3-yl)pyridine (iPr-BTP).

356

Spent Fuel Dissolution and Reprocessing Processes

and an Am/Cm product containing less than 1% of Ln is obtained. Unfortunately, an industrial application of the BTP molecule requires further investigation because it is highly sensitive to hydrolysis and radiolysis. The second alternative under investigation aims at a direct selective extraction of MAs from the PUREX raffinate in a single operation leaving all the lanthanides in the HLLW. A third option is the COEX and lanthanides with DMDOHEMA, as in the extraction step of the DIAMEX process, followed by selective stripping of the trivalent actinides from the loaded diamide solvent using a mixture of hydroxyethyl ethylenediamine triacetic acid (HEDTA) (actinide-selective polyaminocarboxylate complexing agent) and citric acid.24 The scientific feasibility of this process has been demonstrated by the CEA in the Major Nuclear Cycle R&D (ATALANTE) facility in Marcoule, France. An MA recovery of 99.9% with less than 0.3 wt% Ln in the MA fraction was achieved with a flow sheet, where the DIAMEX solvent was supplemented by an acidic extractant, diethylhexylphosphoric acid (HDEHP), to ensure effective extraction at pH > 2. In Japan the Japan Atomic Energy Agency (JAEA) has studied an advanced aqueous process combined with a U crystallization process. The main features compared with the conventional PUREX are as follows:  The purification steps of U and Pu in the conventional PUREX are eliminated, resulting in coextraction of U/Pu/Np, and the simplification of the system. A compact-sized centrifugal type equipment is used to reduce the size of the reprocessing facility.  Crystallization method is used to separate excess U before extraction of U/Pu/Np.  A combination of the SETFICS process, developed by Japanese Nuclear Cycle Development Institute ( JNC), and the TRUEX process is

applied for the recovery of Am and Cm. A recovery ratio of U/TRU has been estimated to be 99.7%, and the decontamination factor of the reprocessed product is higher than 102. Another process developed by JAEA is known as the ‘Four-Group Separation Process’; it includes the following features:  Extraction of all TRU elements including Np (V) with DIDPA at 0.5 M nitric acid.  Separation of Tc and platinum group metals by precipitation through denitration.  Separation of Sr and Cs by adsorption with inorganic ion exchangers.  Selective back extraction of Am and Cm by 0.05 M dietylentriaminepentacetic acid (DTPA). In Table 4, the separation efficiency and estimated recovery values obtained in the various processes mentioned above are compared to target values for the recovery of TRU elements and some key FPs in advanced reprocessing. The separation efficiency and the estimated recovery of TRU elements are quite high and almost fulfill the target recovery. The recoveries of Tc and platinum group metals are around 90–95% which is lower than the target recovery. This lower recovery is less important because of a lower potential radiotoxicity contribution of HLW. 5.14.4.3

Pyro-reprocessing

Pyrochemical processes rely on refining techniques at high temperature (500–900  C) depending on the molten salt eutectic used. Typically chloride systems operate at lower temperature compared to fluoride systems. In nuclear technology, the processes are mainly based on electrorefining or on extraction from the molten salt phase into liquid metal. For more than 50 years, pyrometallurgy has been studied as an alternative strategy in the reprocessing

Table 4 Target recovery, experimentally obtained separation efficiency, and estimated recovery of elements in the fourgroup partitioning process Elements

Target recovery (%)

Separation efficiency (%)

Estimated recovery (%)

Np Pu Am Cm Tc Sr, Cs

99.5 99.9 99.99 99.9 99 99

>99.95 >99.99 >99.99 >99.99 98 >99.9

99.85 99.85 99.97 99.97 95 >99.9

Spent Fuel Dissolution and Reprocessing Processes

of used fuel. Until now, only two processes have been developed up to the pilot scale, both in chloride media; the first one developed by Research Institute of Atomic Reactors (RIAR) in Dimitrovgrad (Russia) is for oxide fuels25 and the second one is using metallic fuel and is being developed in the United States as part of the so-called IFR. The RIAR process can be operated in an air atmosphere, whereas the metallic process require a more or less pure Ar atmosphere, However, only the metallic fuel process allows also the treatment of TRU elements and is therefore discussed in more detail in the following paragraph. 5.14.4.3.1 IFR pyroprocess

The electrometallurgical process was applied for the first time as a part of the IFR system in the pyrochemical separation processes for the recovery of uranium and, to some extent, of plutonium. These processes have been investigated for decades26,27 and remain the core process in the present Experimental Breeder Reactor-II (EBR-II) Spent Fuel Treatment Program. Many of the pyroprocessing systems presently proposed for development are spin-offs of this process, shown in Figure 8.

The fuel is recycled using an electrochemical process based on molten chloride salts and liquid metals. The molten salt medium for electrorefining is a solution of a certain amount of UCl3 dissolved in a LiCl– KCl eutectic. At an operating temperature of about 500  C, chopped used fuel is loaded into the electrorefiner using specially designed stainless steel baskets. The fuel is electrochemically dissolved using an appropriate potential between the basket used as anodes and a stainless steel electrode in the salt phase being used as cathode. Once the fuel starts to dissolve, uranium and a small part of the TRU elements are collected on the cathodes. Once the fuel is dissolved and most of the uranium is deposited on the solid steel, this cathode is replaced by a liquid cadmium cathode, and the remaining TRUs can be codeposited with the remaining uranium. A liquid cadmium cathode is a ceramic crucible containing molten cadmium that can be lowered into the salt bath. The cadmium in the crucible is put at cathodic potential.27 Because of the chemical activities of the TRU elements in cadmium, they can be more easily deposited with uranium in liquid cadmium cathodes than on solid cathodes. The cathode products from electrorefining operations are further processed to Refabrication for recycle Casting furnace

Electrorefiner Cathode processor Oxide reduction Metal

Uranium, transuranics, salt

Oxide Spent fuel

Metal

Salt Zeolite + FPs

Cladding + noble metal + FPs

Legend Product line Cleanup and waste

Furnace

Salt Zeolite + FPs

Metal casting furnace

Zeolite columns Highlevel waste Metal waste form

Figure 8 Metal and oxide fuel pyroprocess flow sheet.

357

Glass powder

Ceramic waste form

358

Spent Fuel Dissolution and Reprocessing Processes

distill adhering salt and cadmium and to consolidate the recovered actinides. Those are remotely fabricated into new fuel for recycling. The alkali, alkaline earth, rare earth, and halide FPs remain primarily dissolved in the salt phase. These elements can be separated from the salt phase (e.g., by extraction or precipitation processes) and are eventually conditioned in a ceramic HLW before being disposed. More than 90% of the noble metal FPs and fuel alloy material are retained in the chopped fuel cladding segments in the anode baskets. This residue can be stabilized into a metal HLW to prepare it also for disposal. Adaptations of this technology exist for the treatment of both oxide and nitride fuels. The flow sheet for the treatment of nitride fuels is similar to that of the metal fuel. The nitride fuels are also fed directly into the electrorefiner; the actinides are dissolved from the fuel cladding and collected all together electrochemically in liquid cadmium or bismuth cathodes. A specificity of this process is the evolution of nitrogen gas. If the formation of 14C from 14N is to be avoided during the fuel irradiation, the initial nitride fuel should be enriched in 15N. Depending on an economic assessment, it should be decided where and when nitrogen should be recycled. This process and the fuel refabrication are of course not very easy. After distillation of the cadmium, the recovered nitrides are separated and then fabricated into new fuels using a vibro-packing step. This process is being developed in Japan.28 5.14.4.3.2 European pyrochemistry projects

On the basis of these past studies, pyrometallurgy based on the US process has been considered not only as the reference route for the molten salt reactor fuel treatment, but also as an alternative technology that could be applied to some types of fuels envisaged for Gen IV systems or ADSs, that is, in case they turn out to be incompatible with current hydrometallurgical processes. The European pyro-reprocessing projects have the following main objectives:  to obtain basic data to allow conceptual design and assessment of reprocessing processes suitable for many different types of fuel and targets;  to assess the feasibility of separating uranium, plutonium, and MAs from FPs using pyrometallurgy in a molten chloride or fluoride systems;  to identify and characterize solid matrices for the conditioning of the wastes issuing from the pyroprocesses;

 to carry out system studies for comparing selected reprocessing of used fuels of advanced nuclear reactors including the ADS;  to revive and consolidate European expertise in pyroprocessing. As an underpinning support for the pyroprocess developments, basic properties of An and some FPs in molten salts (chlorides and fluorides) and in liquid metal solvents have been studied. A very important work was done in the thermodynamic data acquisition in molten chloride media, with a comprehensive study of actinides, lanthanides, and some other important FPs. In comparison to molten chloride salts, studies in molten fluoride are much less developed. Even though a lot of experiments were carried out on various salts, it seems in this case to be more difficult to get relevant thermodynamic data, mainly because of the lack of a reliable reference electrode. Especially for Cm, the data available are very scarce. Two efficient processes for the separation of An from Ln have been selected as promising core processes: (i) electrorefining process on a solid reactive cathode in molten chloride and (ii) liquid–liquid reductive extraction in liquid metal–molten fluoride. As a result of the data collected for a variety of liquid metals, aluminum was the clear choice for both the cathode material for the electrochemical process in molten chlorides and the extractant for the reductive extraction process in molten fluorides. Several reference flow sheets have been assessed. These results were used to optimize the two reference core processes. Moreover, several new experimental installations for process tests have been designed and constructed. In the United Kingdom, Nexia Solutions has built a new facility in an alphaactive glove-box and in Italy Italian National Agency for New Technologies, Energy and Sustainable Economic Development (ENEA) has commissioned the Pyrel II facility for process scale-up and modeling. It has become clear that the construction of a largescale electrolyzer for studies in molten salts is a complex and laborious task requiring a lot of additional efforts to be successful. Another key issue is similar to the aqueous technology, specifically the waste issue. A successful recycling should have similar targets for dry and aqueous reprocessing regarding the loss of fissile materials and the long-lived radionuclides to be recovered. A realistic value is below 0.1% for all actinides. Furthermore, the pyroprocess should also produce

Spent Fuel Dissolution and Reprocessing Processes

the lowest achievable amounts of waste, and the waste produced must be converted into a convenient form for storage or disposal. Here, real progress has been made in the decontamination of used chloride salts resulting from electrorefining, and the complementary techniques of zeolite ion-exchange filtration and phosphate precipitation have been selected for their potential to clean up used salt efficiently. A number of specific matrix materials for salt confinement have been identified (sodalite, pollucite); however, a lot of work is still to be done in this field. The system studies which were performed in the course of the European Research Programs included (i) double-strata concept (ADS), (ii) IFR, and (iii) molten salt reactor. In a first step, the general principles for the assessment of pyrochemical separation processes were defined and a common methodology for technical and economic comparisons and the selected flowsheets was determined. During the second step, the work was focused on detailed flowsheet studies and mass balance calculations. The major interest of these studies is the validation of the ‘process approach,’ a very useful tool to identify key issues and eventually reorient R&D programs. Nevertheless, as the flowsheets address different scenarios and fuels, it is very difficult to make a direct intercomparison in terms of advantages and drawbacks. 5.14.4.3.3 Basic data acquisition

As mentioned in the previous paragraph, a large variety of basic properties of An and some FPs in molten salts (chlorides and fluorides) and in liquid metal solvents have been studied.29–31 Concentrated efforts were made in basic data acquisition for molten chloride media, mainly at ITU, with a comprehensive study of actinides (U, Pu, Np, Am, Cm), lanthanides, and some other important FPs. Thermochemical properties are derived from the electrochemical measurements and from basic thermodynamic data, for instance, in the case of Np of NpCl3 and NpCl4 in the crystal state.32,33 It could be demonstrated, that the NpCl3 has a strong nonideal behavior in molten LiCl–KCl eutectic. For these experiments, a double glove box has been constructed, where the outer glove box is operated under nitrogen and the inner box under a purified argon atmosphere at overpressure conditions. This allows keeping a very pure Ar atmosphere and thereby excellent conditions for a precise determination of the required data. Auxiliary equipment is devoted to chlorination, material processing, and electrochemistry in room temperature ionic liquids,

359

a potential alternative to the high-temperature molten salt systems.34 5.14.4.3.4 Core processes

Initially, three potential chemical routes were identified as candidates for core process development activities. The first one was based on selective precipitation; it was also investigated by RIAR in Russia as a possible option in the selective separation of the TRU elements. However, the success of this process is not very encouraging; the decontamination factors that can be obtained are always very low. The second route is the electrochemical one, which includes electrolysis or electrorefining techniques, in either chloride or fluoride molten salts. The third one is based on the liquid–liquid reductive extraction between a molten salt and a liquid metal phase. Therefore, only the processes based on electrorefining on solid aluminum cathodes in molten chloride and the one based on liquid–liquid reductive extraction in molten fluoride/liquid aluminum were extensively studied in the European programs. In parallel, some studies were carried out on electrolysis in molten fluoride or liquid–liquid reductive extraction in molten chloride but with a much lower priority. 5.14.4.3.5 Electrorefining on solid aluminum cathode in molten chloride media

To comply with the sustainability goals defined for innovative reactor systems, a major objective is the development of a grouped actinide recycling process based on molten salt electrorefining. Special emphasis is given to a selective electrodeposition of actinides with an efficient separation from lanthanide FPs. In contrast to the IFR concept, where U is deposited on a solid stainless steel cathode and TRU actinides on a liquid Cd cathode,35 the electrorefining processes rely on codeposition of all actinides on a solid Al cathode material. In fact, the choice of the cathode material onto which the actinides are deposited in the electrolysis is essential in this context.36 In contrast to stainless steel or tungsten, aluminum is a reactive electrode material, that is, it forms stable alloys with the actinides, thereby avoiding the redissolution of trivalent actinides. Also the redox potentials on solid cathodes show a much larger difference in the reduction potential between actinides and lanthanides. Figure 9 shows the reduction potentials for U3þ, Pu3þ, Am3þ, La3þ, and Nd3þ determined by transient electrochemical techniques (mainly cyclic voltammetry and chronopotentiometry) on different cathode materials. On Bi

360

Spent Fuel Dissolution and Reprocessing Processes

and Cd, the selectivity of the MA recovery seems to be limited because of the small difference in reduction potentials between actinides and lanthanides. Solid Al has therefore been selected essentially because of two reasons:

experiments in which the cathodic potential was maintained at a suitable level for separation of An from Ln. With an increase in the charge passed, that is, with the buildup of a surface layer of An–Al alloy, the applied current is gradually reduced in order to stay above the cathodic potential limit. On the basis of a large set of data obtained for the electrodeposition on aluminum cathodes, the process scheme is being proposed as shown in Figure 10. The electrorefining process as presented here is operated in a batch mode. After multiple use of the eutectic salt bath, an exhaustive An electrolysis is required to avoid losses >0.1% to the waste, before the cleaning of the salt bath takes place. It is evident that the electrodeposited An–Al alloy in the exhaustive electrolysis contains more Ln than in the runs where metallic fuel is deposited and must eventually be recycled. For the cathode processing, three options are possible, chlorination, back extraction, and electrorefining. Among these, chlorination is the most promising. This step is needed to recycle the actinides to the fuel fabrication. Laboratory experiments have shown that 3.72 g of actinides were deposited in 4.17 g Al, corresponding to 44.6 wt% An in Al or 68 wt% of the maximal loading, considering that AnAl4 alloys are formed.37 A successful demonstration of the Am/Nd separation was carried out using a mixture of 255 mg Am, 281 mg Pu, and 140 mg Nd. Am and Pu were codeposited in two steps on two Al cathodes of 0.8 g each. The cathodes used were made of Al foam to increase the reaction surface area. The Nd content in the deposit of only about 0.5% proves the feasibility of a selective actinide separation by electrolysis onto Al electrodes. The results were confirmed in a multiple run experiment inducing an accumulation of lanthanides

1. Stable actinide deposits (alloys) are formed and are consequently very adherent to the cathode; at the same time, redissolution of the trivalent An by comproportionation with the trivalent actinides in the salt to form divalent Ans can be avoided (cf. equation: Am(III) þ Am(0) ¼ 3 Am(II). 2. The difference in the reduction potentials compared to lanthanides is sufficiently high to avoid their codeposition. In these electrolytic processes, the rate of the alloy formation depends on the diffusion of the involved elements in and through the solid alloy phase. Therefore, the maximum amount of actinides that can be collected on a single Al electrode has been investigated in constant current electrorefining

Potential (V vs. Ag/AgCl)

-1 -1.2 -1.4

Pu Am La

U Pu Am

Pu Am La

Nd La

U

-1.6

Pu

-1.8

Am

-2

Nd La

-2.2 Liquid Bi

Liquid Cd Solid W

Solid Al

Figure 9 Reduction potentials of some actinides and lanthanides on different cathodic materials.

Used salt with high content of FP

Salt + remaining An’s + Ln’s Metallic An–Ln fuel Electrorefining on AI cathode

Exhaustive electrolysis Cathode processing An–Al alloys

An AI

Three identified ways

Chlorination Back-extraction Electrorefining

Figure 10 Process scheme for the electrorefining of metallic fuels.

Salt cleaning Salt + waste and/or storage

Spent Fuel Dissolution and Reprocessing Processes

in the salt. The fuels used for these experiments had already been developed in the frame of the IFR concept (see previous paragraph) in the mid-1980s in the United States. These fuels contain about 15% of Zr in the metallic alloy to stabilize the fuel during reactor irradiation. The same type of fuel, used for transmutation studies initiated by Central Research Institute of Electric Power Industry (CRIEPI), Japan, in collaboration with ITU, was irradiated in the metallic fuel irradiation ad PHENIX (METAPHIX) experiment in the PHENIX reactor in France.38 This fuel containing 2% of Am and lanthanides (U61Pu22Zr10Am2Ln5) was fabricated at ITU and the remnants of the fuel fabrication campaign were used for separation studies. In the pyro-reprocessing, the metallic alloy is anodically dissolved in a LiCl–KCl eutectic39 and the actinides are collected together onto Al cathodes as alloys, leaving lanthanides in the salt phase. It is very likely that a large-scale pyroprocessing by molten salt electrorefining will be operated as a batch process similar to the industrial Al fabrication process. In view of a large-scale development of the process, an experiment with 25 successive runs was carried out to demonstrate the feasibility of a grouped actinide recovery from larger amounts of fuel without changing the salt bath.36 A total amount of more than 5 g of U61Pu22Zr10Am2Ln5 fuel was treated in this experiment and various process parameters were studied. Figure 11 shows

361

the cyclovoltamogram of the alloy on Al and W electrodes. The goal of this 25-run test was to find optimal conditions for the recovery of Am. The recovery rate of actinides was difficult to evaluate because new fuel was added in each run. Nevertheless, a stable recovery rate [mAn/(mln þ mLn)], nearly 99.9%, was achieved throughout the whole experiment. Uranium, the main constituent of the fuel with a less electronegative electrodeposition potential is preferentially deposited in the earlier runs. At the same time, the relative Am content in the actinide deposit and the separation from lanthanides (mAm/mLn) increase despite an increasing content of lanthanides not electrodeposited in the salt. This means that the target of 99.9% recovery can be reached for this process. The results of this 25-run electrorefining experiment for which genuine fuel materials were used and for which the salt bath was not changed are very promising in view of a large-scale development of pyro-reprocessing in advanced nuclear fuel cycles. 5.14.4.3.6 Exhaustive electrolysis

When a salt bath is being used for the electrorefining of large amounts of fuel, the FPs are accumulated in the salt bath and their concentration becomes too high and thereby prevent a selective deposition of actinides on the cathode. An exhaustive electrolysis is proposed for the first purification step, a complete

150 W electrode Al electrode

Al => Al3+ 100 U3+ => U Current (mA)

50

Cl− => Cl2

Np3+ => Np Pu3+ => Pu

0 U3+ => UAl4

−50 −100

Pu3+ => PuAl4 Ln3+ => Ln Li+ => Li

−150 −3.00 −2.50

Cut-off potential (−1.25 V)

−2.00 −1.50

−1.00 −0.50

0.00

0.50

1.00

1.50

Potential (V vs. Ag/AgCl) Figure 11 Cyclic voltammogram of U61Pu22Zr10Am2Ln5 on W and Al wires. Reference electrode: Ag/AgCl – 1 wt%, v ¼ 100 mV s1, T ¼ 450  C. Salt composition in wt%: U – 0.29, Np – 0.12, Pu – 0.28, Am – 0.06, Zr < 0.07, and Ln – 1.0.

362

Spent Fuel Dissolution and Reprocessing Processes

i -

+ Cl2 e-

Cl2 (g) producing anode

e-

FPn+

Selective reduction

An–Al alloy

Ann+

An–Al alloy

AI

Cl-

Molten LiCl-KCI (450 °C) Figure 12 Principle of the exhaustive electrolysis process.

grouped recovery of the remaining actinides without further fuel dissolution on a solid aluminum cathode (see Figure 10). The anode basket is therefore replaced by a chlorine electrode. Partial oxidation of the chloride salt to chlorine gas allows the actinide reduction on the cathode side. A scheme of the process is shown in Figure 12. In order to prove feasibility of the method, two galvanostatic electrolyses were carried out using a mixture of UCl3 and NdCl3.40 The potentials of both electrodes were constantly followed and a decrease of the uranium concentration from 1.7 to 0.1 wt% with no codeposition of neodymium was observed. Although the maximum applicable current densities were relatively low, the results are promising and showing high current efficiency and selectivity of the proposed method. 5.14.4.3.7 Liquid–liquid reductive extraction in molten fluoride/liquid aluminum

The alternative process to electrorefining in molten chloride salts is the liquid metal/molten salt process. This option was extensively studied by CEA in several European Research Programs.41–43 An experimental device and a process scheme have been developed to study the distribution of actinides and lanthanides between molten fluoride salt and liquid metal media. The results obtained with plutonium, americium, cerium, and samarium in the (LiF–AlF3)/(Al–Cu) medium revealed the excellent potential of the system for separating actinides from lanthanides. With a salt composition corresponding to the basic eutectic (LiF–AlF3, 85–15 mol%), up to 99% of Pu

and Am could be recovered in a single stage, with cerium and samarium separation factors exceeding 1000. The effect of the AlF3 concentration in the salt has been investigated. The distribution coefficients logically go down as the initial AlF3 concentration increases. A thermodynamic model to describe the extraction as a function of the fluoroacidity has been developed on the basis of the experimental results for cerium and samarium. The model clearly reveals a difference in solvation between divalent and trivalent lanthanides in fluoride media. The results obtained for each element were confirmed by demonstration experiments under more realistic conditions, at a lab scale. Two runs were carried out at 830  C using LiF–AlF3 (85–15 mol%) as a salt phase: one with an Al–Cu alloy (78–22 mol%) as metallic phase, the other with pure Al, to check the influence of Cu on the extraction, both in terms of separation performance and in terms of process implementation (phase separation). The metal phase was treated with salt with the following composition (wt%): PuF3 (11), AmF3 (0.2), CeF3 (2.5), SmF3 (0.5), EuF3 (0.5), and LaF3 (0.5). The results show that the distribution ratios of Pu and Am are in the same order of magnitude, similar to the ones previously measured at low concentration without lanthanides. The results obtained with Al–Cu and Al are very similar. The distribution coefficients of the lanthanides are low and thus the separation from actinides is very efficient. In a test with Al without Cu, the distribution coefficient of Cm (trace concentration in Am starting material) has been measured for the first time; it is very close to the values obtained for the other actinides (U, Np, Pu, Am). The tests without Cu addition to the metallic phase show that a satisfactory phase separation can be achieved; therefore, Cu addition is not mandatory for the process implementation. In Table 5, the main test results are summarized. The results show that the distribution ratios of Pu and Am have similar high values independent from the presence of Cu in the metallic phase and that in all cases high separation efficiency from lanthanides can be achieved. The actinide back extraction from the Al is of course an important step in view of fuel refabrication. In a bibliographic study three possible routes were identified44:  Electrorefining: Main drawback is the complexity of the process which requires three steps.  Volatilization of the Al matrix by a chlorinating reagent: It is a simple and efficient method.

Spent Fuel Dissolution and Reprocessing Processes

363

Table 5 Mass distribution coefficients and separation factors of actinides and lanthanides with and without Cu in the metallic phase Al–Cu (78–22 mol%)

Al

Metal

Distribution actor

Separation factor Am/metal

Metal

Distribution factor

Separation factor Am/metal

Pu Am Ce Sm Eu La

197  30 144  20 0.142  0.01 0.062  0.006 <0.013 <0.06

0.73  0.21 1 1014  213 2323  488 >11 000 >2400

Pu Am Cm Ce Sm Eu La

273  126 213  30 185  31 0.162  0.02 0.044  0.004 <0.03 0.03

0.78  0.47 1 1.15  0.35 1315  289 4954  1139 >7100 7100

Nevertheless, high volumes of chlorination gas have to be managed and an additional step is necessary to convert the AlCl3 to Al metal.  Oxidizing liquid–liquid extraction in molten chloride. An experimental study is necessary to select the most efficient option. 5.14.4.3.8 Technical uncertainties of the pyro-reprocessing

In the Spent Fuel Treatment Program at Idaho National Laboratory (INL), many parts of the pyroprocess fuel cycle could be demonstrated up to the 100 kg scale. Nevertheless, there are key aspects that have yet to be demonstrated, particularly the recovery of transuranics. Large-scale equipment designed and constructed was never tested beyond the laboratory scale, because of the termination of the IFR program. The remote fabrication of IFR fuel was not part of the Spent Fuel Treatment Program, but this technology was used to fabricate cold fuel for EBR-II and a demonstration of another pyroprocess (melt refining) for recycling EBR-II in the 1960s employed remote fabrication for 34 500 fuel elements.23 Another key challenge for a pyroprocessing system is the selection of appropriate construction materials for the high-temperature processes. Material improvements are needed in order to reduce the formation of dross streams and to increase the material recovery and throughput. The quantity of waste generated requiring geological disposal from pyroprocessing seems to be quite similar to that in present modern commercial aqueous processes. Advancements are being pursued to further reduce the disposal volumes using specially adapted zeolite ion-exchange technology, which has at present not yet been demonstrated beyond the laboratory scale.

Most of the radioactive work performed to date has been on the pyroprocessing cycle for metal fuel. Laboratory work has been performed on the headend operations for oxide reduction and on the nitride fuel cycle. Demonstrations of these technologies with actual used fuel have started at a laboratory scale. Additionally for nitride fuels, a demonstration of the above-mentioned recycling of nitrogen (15N) is essential for the economic considerations. 5.14.4.3.9 Head-end conversion processes

Today, all commercial reactors are operated with oxide fuels, and advanced reactor systems selected in the GENIV roadmap also rely on oxides as one of the major fuel options. As mentioned above, the pyrometallurgical process based on oxides developed in RIAR, Dimitrovgrad (Russia) does not allow the recycling of MAs. Pyro-reprocessing where all actinides are recycled is based on metallic materials; therefore, a head-end reduction step for oxide fuels is needed to convert oxides into metals. This conversion can be performed chemically, for example, by reaction with lithium dissolved in LiCl at 650  C. The recovered metal can directly be subjected to electrorefining and the Li2O is converted back to lithium metal by electrowinning. A more elegant method is the so-called direct electroreduction.43 In this case, the heat generating FPs are removed and the fissile materials are recovered as an alloy, which can be again directly reprocessed by electrorefining. Numerous experiments are carried out today to study this conversion process. The lithium reduction process using lithium metal as a reductant is carried out in molten lithium chloride. The reduction of UO245 and simulated used LWR fuel46 was studied mainly by CRIEPI in Japan in collaboration with AEA Technology in the United Kingdom. The optimized thermodynamic conditions for the reduction of TRU

364

Spent Fuel Dissolution and Reprocessing Processes

elements47 and the behavior of major FP elements46 were determined. Li is converted into Li2O and constantly removed during the process from the molten salt bath to prevent the reoxidation of the reduced fuel material. Li is recovered by electrochemical decomposition of the Li2O and recycled to the process. A simulated used oxide fuel in a sintered pellet form, containing the actinides U, Pu, Am, Np, and Cm, and the FPs Ce, Nd, Sm, Ba, Zr, Mo, and Pd, was reduced with Li metal in a molten LiCl bath at 923 K. The pellet remained in its original shape; it became porous, and a shiny metallic color was observed throughout the pellet. The Pu/U ratio did not change during the reduction process. The reduction yield of U and Pu determined by measuring the H2 formed on reaction of the reduction product with HBr and using a gas burette was more than 90%. A small fraction of Pu has formed an alloy with Pd. The RE elements are found in the gap of the porous U–Pu alloy. As expected from the oxygen potential of Ce, Nd, Sm, and Li, they remained in an oxide form. Small fractions of the actinides and lanthanides are leached from the pellet into the molten LiCl bath or found as precipitates on the crucible bottom. A large part of Am is found in the RE oxide phase rather than in the reduced U–Pu alloy. This represents of course a major problem for a grouped actinide recovery. In addition, the handling of highly reactive Li and problems in developing the corresponding equipment, especially for the lithium recovery, are major drawbacks of this process. The electrochemical reduction process is clearly the more reliable technique to convert oxides into metal. The difficult handling of Li metal and recycling through reconversion from Li2O can be avoided. The oxide ion produced at the cathode is simultaneously consumed at the anode and thus the concentration of oxide ions in the bath can be maintained at a low level. A more complete reduction of the actinide elements can be achieved and the subsequent electrorefining to separate actinides as described in the previous paragraph can be carried out in the same device.47 An electrochemical process is being developed, mainly in the United States at INL in Idaho and also in Japan at CRIEPI in Tokyo, in collaboration with the EC, DGJRC/ITU in Karlsruhe, Germany. Both unirradiated and irradiated fuel materials were treated with slightly different concepts. The oxide fuel is loaded into a permeable stainless steel basket as crushed powder.48 The basket immersed into a molten LiCl–1 wt% Li2O electrolyte at 650  C is used as the cathode and a platinum wire is

used as anode. The reduced fuel is retained in the basket. The oxygen ions liberated at the cathode diffuse to the Pt anode, where they are oxidized to oxygen gas. The corresponding reactions are as follows: Cathode: MxOy þ 2ye ¼ xM þ yO2 Anode: yO2 ¼ y=2O2 ðgÞ þ 2ye where M ¼ metal fuel constituent. The Li2O present in the salt is reduced to Li together with U and reduces chemically the fuel oxide. Consequently, the INL process is a combined chemical–electrochemical process. The molten salt can be either LiCl or CaCl2. In CaCl2, the higher temperature of 1123 K in comparison to 923 K for LiCl induces a faster diffusion of oxygen ions to the anode. At the same time, an increased initial reaction rate leads to the formation of a thin dense metal layer at the fuel surface hampering the diffusion of oxygen ions into the salt. For the CRIEPI/ITU process, the anode is made of carbon, and the fuel is not crushed but loaded as fuel element segments in a cathode basket that is made of Ta.49 The corresponding cathodic and anodic reactions are as follows: Cathode: MxOy þ 2ye ¼ xM þ yO2 Anode: yO2 þ y=2C ¼ CO2 ðgÞ þ 2ye or yO2 þ yC ¼ COðgÞ þ 2ye The INL process scheme was successfully demonstrated using irradiated used LWR oxide fuel in a hot cell. More than 98% of the U was reduced. Cesium, Ba, and Sr were dissolved in the salt phase, as expected. The rare earth and noble metal FPs remained with U and transuranics Pu and Np were reduced together with U; however, about 20% of the Am remained as oxide. The CRIEPI/ITU process was tested on various MOX (Pu content 5–45%) fuels which were reduced. It could be shown that U and Pu are efficiently coreduced, but because of the problems mentioned above, the complete reduction requires very long reaction times. The reduction of irradiated FR fuel particles at ITU was considerably faster and a complete reduction of all fuel constituents including FPs and MAS was achieved. Figure 13 shows the reduced fuel particles in the cathode basket. The analyses of the salt bath used for these experiments, the examination of the reduced product by scanning electron microscope (SEM)/energy-dispersive

Spent Fuel Dissolution and Reprocessing Processes

Figure 13 Schematic layout of an electroreduction process developed by CRIEPI/ITU.

X-ray spectroscopy analysis (EDX), and the analysis of the reduced fuel after dissolution allow for establishing a mass balance of the electroreduction process. The results show that the fuel is completely reduced; that is, all actinides are in the reduced product, the light FPs Rb, Mo, Cs, Ba, Se are dissolved in the salt, and the lanthanide FPs are divided between the reduced fuel and an oxide precipitate found at the bottom of the salt crucible. A first experiment has shown that the reduced fuel can be treated similar to the metallic fuels described above and using the same equipment and the same type of salt bath as the one used for the electrorefining tests. 5.14.4.4 The Direct Use of Pressurized Water Reactor Spent Fuel in CANDU Process Another approach to used nuclear fuel recycling which could be employed by some countries is the Direct Use of Pressurized Water Reactor Spent Fuel in CANDU (DUPIC) process,49 which enables direct recycling of used pressurized water reactor (PWR) fuel in CANada Deuterium Uranium (CANDU) reactors. CANDU reactors use natural uranium fuel without enrichment and could therefore be fuelled with uranium and plutonium from used LWR fuel. In the DUPIC process, the used fuel assemblies from LWRs are dismantled and refabricated into fuel assemblies for CANDU reactors. This process could involve simple cutting of used LWR fuel rods to be adapted as CANDU fuel elements (about 50 cm), resealing, and reengineering them into cylindrical bundles suitable for CANDU geometry. The more likely alternative is a dry reprocessing treatment, where the volatile FPs are removed from the used LWR fuel. No materials are separated during the refabrication process. After removal of the

365

cladding, the used LWR fuel is converted into powder by a thermal–mechanical process and fresh natural uranium is added before CANDU pellets are sintered and pressed. However, as noted above, used nuclear fuel is highly active and generates heat. The high radioactivity of the materials to be handled in the DUPIC process requires heavy shielding and remote operation. The restricted diversion of fissile materials and hence increased proliferation resistance go together to make a much more complex manufacturing process. Canada, where the CANDU reactor line has been developed, and South Korea, which hosts four CANDU units as well as many PWRs, have initiated a bilateral joint research program to develop the DUPIC process, and the Korean Atomic Energy Research Institute (KAERI) has been implementing a comprehensive development program since 1992 to demonstrate the DUPIC fuel cycle concept. Challenges that remain include the development of a technology to produce fuel pellets of the correct density, the development of remote fabrication equipment, and the handling of the used PWR fuel. However, KAERI successfully manufactured small DUPIC fuel elements for irradiation tests inside the HANARO research reactor in April 2000 and fabricated full-size DUPIC elements in February 2001. Research is also underway on the reactor physics of DUPIC fuel and the impacts on safety systems. A trial period of the technology has started in 2010 with irradiation of used LWR fuel in the Qinshan reactor in China.

5.14.5 Outlook Industrial reprocessing as it is in operation today mainly in France, United Kingdom, and Japan will certainly for several decades continue operation; new capacities will be installed or extended in China, Russia, and India in the near future and France and Japan consider installation of new or additional capacities in a few decades from now. If the sustainability goal strongly promoted in the GENIV initiative and also in INPRO coordinated by IAEA or the European SNE-TP platform is to be inherent to new generation reactor systems, the waste minimization will require recycling of long-lived waste constituents including MA. As a consequence, extended and modified reprocessing technologies will have to be implemented on a large scale. As a first step, the actual PUREX process will be adapted to these needs. If advanced fuel materials such as composites, metals,

366

Spent Fuel Dissolution and Reprocessing Processes

nitrides, or carbides are selected for the new reactor systems, adapted reprocessing technologies based on pyroprocesses might be well suited to reprocess these fuels. Significant efforts are being made in South Korea, India, Japan, and United States to develop these processes to an industrial scale. A possible strategy for the second half of this century could be based on a double-strata concept with an advanced aqueous reprocessing of LWR fuel in the first stratum combined with a fast reactor–pyroprocess combination in the second stratum to reach the sustainability goal.

20. 21. 22. 23. 24. 25. 26. 27.

References 1. 2. 3. 4. 5.

6. 7. 8.

9. 10. 11. 12. 13.

14. 15. 16. 17. 18.

19.

US Patent 2924506. Kleykamp, H. J. Nucl. Mater. 1985, 131(2–3), 221. Roberts, T. Mater. World 1994, 2(12), 628. Magill, J.; Berthou, V.; Haas, D.; et al. Nucl. Energy 2003, 42(5), 263. Salvatores, M.; Zaetta, A.; Girard, C.; Delpech, M.; Slessarev, I.; Tommasi, J. Proceedings of the International Conference on GLOBAL’93, Seattle, Sept 12–17 1993; p 548. Mukaiyama, T.; Takano, H.; Ogawa, T.; Takizuka, T.; Mizumoto, M. Prog. Nucl. Energy 2002, 40(3), 403–413(11). Madic, C.; Lecomte, M.; Testard, F.; et al. Proceedings of the International Conference GLOBAL 2001, Paris, France, Sept 9–13, 2001. Madic, C.; Hudson, M. J. In Proceedings OECD-NEA: 8th IEM on Actinide and Fission Product Partitioning and Transmutation, Las Vegas, Nevada, USA, Nov 9–11, 2004. Adnet, J. M.; et al.; Proceedings of GLOBAL 2005, Tsukuba, Japan, Oct 9–13 2005; Paper No. 119. Madic, C.; Hudson, M. J.; Liljenzin, J. O.; et al. New Partitioning Techniques for Minor Actinides; European Report, EUR 19149; 2000. Madic, C.; Testard, F.; Hudson, M. J.; et al. PARTNEW – New Solvent Extraction Processes for Minor Actinides – Final Report; CEA-Report 6066; 2004. Takanashi, M.; Homma, S.; Koga, J.; Matsumoto, S. J. Alloys Compd. 1998, 271–273, 689. Serrano-Purroy, D.; Baron, P.; Christiansen, B.; Malmbeck, R.; Sorel, C.; Glatz, J. P. Radiochim. Acta 2005, 93, 351; Malmbeck, R.; et al. In Proceedings of the Euradwaste’99 – Radioactive Waste Management Strategies and Issues, Luxembourg, Nov 15–18, 1999. Nash, K. L. Solvent Extr. Ion Exch. 1993, 11(4), 729–768. Morita, Y.; Glatz, J. P.; Kubota, M.; Koch, L. Solvent Extr. Ion Exch. 1996, 14(3), 385. Arai, K.; et al. J. Nucl. Sci. Technol. 1997, 34(5), 521–526. Koma, Y.; Watanabe, M.; Nemoto, S.; Tanaka, Y. Solvent Extr. Ion Exch. 1998, 16(6), 1357–1367. Glatz, J. P.; Song, C.; He, X.; Bokelund, H.; Koch, L. In Proceedings of the Special Symposium on Emerging Technologies in Hazardous Waste Management, Atlanta, Georgia, Sept 27–29, 1993; Tedder, D. W., Ed.; ACS: Washington, DC, 1994. Nigond, L.; Musikas, C.; Cuillerdier, C. Solvent Extr. Ion Exch. 1994, 12(2), 261.

28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

42. 43. 44.

45. 46. 47. 48. 49.

Kolarik, Z.; Mu¨llich, U.; Gassner, F. Solvent Extr. Ion Exch. 1999, 17(1), 23. Hill, C.; He´re`s, X.; Calor, J. N.; et al., In Proceedings of Global´99, Jackson Hole, WY, USA, Aug 29 to Sept 3, 1999. Modolo, G.; Odoj, R. Solvent Extr. Ion Exch. 1999, 17(1), 33–53. Geist, A.; Hill, C.; Modolo, G.; et al. Solvent Extr. Ion Exch. 2006, 24(4), 463–483. Miguirditchian, M.; Chareyre, L.; He´re`s, X.; Hill, C.; Baron, P.; Masson, M. In Proceedings of GLOBAL 2007, Boise, ID, Sept 9–13, 2007; Paper No. 81. Vavilov, S.; Kobayashi, T. H. I.; Myochin, M. J. Nucl. Sci. Technol. 2004, 41(10), 1018. Laidler, J. J.; Battles, J. E.; Miller, W. E.; Ackerman, J. P.; Carls, E. L. Prog. Nucl. Energy 1997, 31(1/2), 131–140. McPheeters, C.; Pierce, R. D.; Mulcahey, T. P. Prog. Nucl. Energy 1997, 31(1/2), 175–186. Takano, H.; Akie, H.; Osugi, T.; Ogawa, T. Prog. Nucl. Energy 1998, 32(3–4), 373–380. Serp, J.; Konings, R. J. M.; Malmbeck, R.; Rebizant, J.; Scheppler, C.; Glatz, J. P. J. Electroanal. Chem. 2004, 561, 143–148. Caravaca, C.; Co´rdoba, G.; de Toma´s, M. J.; Rosado, M. J. Nucl. Mater. 2007, 360, 25–31. Masset, P.; et al. J. Electrochem. Soc. 2005, 152(6), 1109–1115. Konings, R. J. M.; Serp, J.; Malmbeck, R. J. Nucl. Sci. Tech. 2001, 12(3), 906–909. Masset, P.; Konings, R.; Malmbeck, R.; Serp, J.; Glatz, J. P. J. Nucl. Mater. 2005, 344, 173–179. Plechkova, N. V.; Seddon, K. R. Chem. Soc. Rev. 2008, 37, 123–150. Kato, T.; Inoue, T.; Iwai, T.; Arai, Y. J. Nucl. Mater. 2006, 357, 105–114. Cassayre, L.; Malmbeck, R.; Masset, P.; et al. J. Nucl. Mater. 2006, 360, 49–57. Serp, J.; Konings, R. J. M.; Malmbeck, R.; Rebizant, J.; Scheppler, C.; Glatz, J. P. J. Electroanal. Chem. 2004, 561, 143. Ohta, H.; Yokoo, T.; Inoue, T.; et al. Nucl. Technol. 2009, 165, 96–110. Iizuka, M.; Kinoshita, K.; Koyama, T. J. Phys. Chem. Solids 2005, 66, 427–432. Soucˇek, P.; Malmbeck, R.; Mendes, E.; Nourry, C.; Glatz, J. P. J. Radioanal. Nucl. Chem., DOI: 10.1007/ s10967-010-0739-6, published online, 2010. Lacquement, J.; Bourg, S.; Boussier, H. In Progress of the R&D Program on Pyrochemistry at CEA Proceedings of GLOBAL 2005, Tsukuba, Japan, Oct 9–13 2005; Paper No. 153. Conocar, O.; Douyere, N.; Glatz, J.-P.; Lacquement, J.; Malmbeck, R.; Serp, J. Nucl. Sci. Eng. 2006, 153, 253–261. Conocar, O.; Douyere, N.; Lacquement, J. J. Nucl. Mater. 2005, 344(1–3), 136–141. Soucˇek, P.; Malmbeck, R.; Mendes, E.; Nourry, C.; Jardin, R.; Glatz, J. P. In Separation of Uranium from Uranium-Aluminium Alloys by Chlorination in Global 2009, Paris, France, Sept 6–11, 2009; p 1214. Sakamura, Y.; Kurata, M.; Inoue, T. J. Electrochem. Soc. 2006, 153, D31–D39. Sakamura, Y.; Omori, T.; Inoue, T. Nucl. Technol. 2008, 162, 169–178. Iizuka, M.; Sakamura, Y.; Inoue, T. J. Nucl. Mater. 2006, 359, 102–113. Herrmann, S.; Li, S.; Simpson, M. J. Nucl. Sci. Technol. 2007, 44(3), 361–367. Kurata, M.; Inoue, T.; Serp, J.; Ougier, M.; Glatz, J. P. J. Nucl. Mater. 2004, 328, 97–102.

5.15

Degradation Issues in Aqueous Reprocessing Systems

B. J. Mincher Idaho National Laboratory, Idaho Falls, ID, USA

Published by Elsevier Ltd.

5.15.1

Introduction

368

5.15.2 5.15.2.1 5.15.3 5.15.3.1 5.15.3.2 5.15.3.3 5.15.4 5.15.4.1 5.15.4.2 5.15.5 5.15.5.1 5.15.5.2 5.15.6 5.15.6.1 5.15.6.2 5.15.6.3 5.15.7 References

Chemistry of Irradiated Aqueous and Organic Solutions Produced Transient Reactive Species PUREX Process Tributylphosphate Degradation Alkane Diluent Degradation Radiolysis Versus Hydrolysis Fission Product Extraction Processes HCCD–PEG Process Radiolysis FPEX Process Radiolysis Minor Actinide and Lanthanide Extraction Processes TRUEX Solvent Degradation Diamide-Based Processes Processes for the Separation of Minor Actinides from Lanthanides TALSPEAK Process Bis(triazinyl)pyridine-Based Processes Dithiophosphinic Acid-Based Processes Conclusions

369 369 371 371 373 374 375 375 376 378 378 380 382 382 383 384 385 386

Abbreviations ALINA

BATP BOBCalixC6 BTP (ClPh)2PSSH CMPO Cs-7SB DCH18C6 DHOA DIAMEX DMDOHEMA DtBuCH18C6 DTPA FPEX FS-13

ActinideIII-Lanthanide INtergroup separation from Acidic medium Annulated bis(triazinyl)pyridine Calix[4]arene-bis-(t-octylbenzo)crown-6 Bis(triazinyl)pyridine Bischlorophenyldithiophosphinic acid Octylphenyldiisobutylcarbamoylmethyl phosphine oxide 1-(2,2,3,3,-tetrafuoropropoxy)3-(4-sec-butylphenoxy)-2-propanol Dicyclohexano 18-crown-6 Dihexyloctanamide Diamide extraction Dimethyl dioctyl hexylethoxymalonamide Di-t-butylcyclohexano-18-crown-6 Diethylenetriamine pentaacetic acid Fission product extraction Phenyltrifluoromethylsulfone

H2MBP H2MEHP HCCD–PEG HDBP LET MDOHEMA NMR NPH Ph2PSSH PUREX R2POOH R2PSOH R2PSSH SANEX TALSPEAK

TAP TBP

Monobutylphosphoric acid Monoethylhexyl phosphoric acid Cobalt dicarbollide–polyethylene glycol Dibutylphosphoric acid Linear energy transfer Methyldioctyl 2-hexyloxyethyl malonamide Nuclear magnetic resonance spectroscopy Normal paraffinic hydrocarbon Bisphenyldithiophosphinic acid Plutonium and uranium refining by extraction Phosphinic acid Monothiophosphinic acid Dithiophosphinic acid Selective actinide extraction Trivalent actinide lanthanide separation by phosphorousreagent extraction from aqueous komplexes Triamylphosphate Tributyl phosphate

367

368

Degradation Issues in Aqueous Reprocessing Systems

TCE TODGA TOPO TRUEX UREX

Trichloroethylene Tetraoctyldiglycolamide Trioctylphosphine oxide Transuranium extraction Uranium extraction

5.15.1 Introduction The first use of nuclear energy to produce electricity occurred in 1951 at the Experimental Breeder Reactor 1 in Idaho, USA.1 For several decades thereafter, nuclear energy promised an inexpensive and inexhaustible supply of electricity. However, concerns about reactor safety, nuclear weapons proliferation, and nuclear waste disposal resulted in nuclear power becoming politically, if not technically, untenable. During that period, many countries abandoned electricity generation by nuclear reactors. Today, in the early twenty-first century, renewed interest in nuclear power is being expressed around the globe. New power plants are under construction in China, Russia, and India, and orders for new constructions are anticipated in the United States. This rebirth of interest in nuclear technology has been brought about by concerns about the contribution of fossil fuel burning to climate change and the unreliable sources of those fossil fuels. Currently, about 437 reactors generate 17% of the world’s electricity. To create a nuclear power industry that can supply an increased portion of world energy demand, it is likely that the reprocessing of spent fuel will be necessary. Spent fuel contains unfissioned 235U and other fissionable actinides generated from uranium by neutron capture reactions during reactor operations. It has been estimated that the fuel available for recycling could extend natural uranium supplies by 30%.2 However, in the United States, a once-through fuel cycle has been adopted wherein used fuel is buried, rather than recycled. Japan, France, and the United Kingdom, however, continue to reprocess spent nuclear fuel using the PUREX (plutonium uranium refining by extraction) solvent-extraction process. Future nuclear fuel recycling operations for the oxide fuels used in the current generation of reactors will have goals in addition to the recovery of uranium and/or plutonium for fuel. For nuclear power to be environmentally sustainable, recycling must also minimize the amount of long-lived radioactive waste produced and create proliferation-resistant process streams. Therefore, a global collaboration is currently

investigating modified solvent-extraction processes to meet these goals. Based on decades of success with the conventional PUREX process,3 new extractions will recover neptunium and plutonium in a proliferationresistant form, selected fission products of high environmental consequence, and the minor actinides for incorporation in fast reactor fuels. Such a process will both recover energy-containing elements for energy production and minimize the amount of hazardous, long-lived, radioactive waste that must be disposed in a geological repository. Aqueous reprocessing begins with the chopping and dissolution of the used fuel in nitric acid. In the currently used PUREX process, uranium, neptunium, and/or plutonium are recovered by the extraction of the acidic phase with 30% tributyl phosphate (TBP) in an alkane diluent. Uranium is always hexavalent and extracted, while neptunium and plutonium can be extracted or rejected depending on their oxidation state adjustments prior to the contact. Proposed fuel cycles will contain additional extraction steps designed to recover additional elements of interest. Although these new extractions are still undergoing research and development, some have been successfully tested on a pilot scale. In the United States, a series of extractions collectively known as UREX (uranium extraction) has been developed with five main steps. In the first, uranium is extracted using TBP from an acidic nuclear fuel dissolution containing a reducing agent to prevent neptunium and plutonium extraction. The raffinate is next extracted with a combination of crown ethers selective for strontium and cesium,4 the main heatgenerating fission products. After treatment to remove the reducing agent, neptunium and plutonium are then removed together with another TBP contact. The raffinate still contains the minor trivalent actinides, long-lived a-emitters that are undesirable to bury from an environmental perspective but are valued as fuel for fast reactors. In the United States, they are proposed to be extracted using octylphenyldiisobutylcarbamoylmethyl phosphine oxide (CMPO) in the TRUEX (transuranium extraction) process.5 A complicating factor is that the fission-product lanthanides are also trivalent and thus have chemistry similar to that of the minor actinides. These lanthanides are very radioactive although relatively short-lived, and some are also neutron poisons. They must be removed from the minor actinide product prior to the fabrication of new fuel. The extraction proposed for this final step is the TALSPEAK (trivalent actinide lanthanide separation by phosphorous-reagent extraction from aqueous

Degradation Issues in Aqueous Reprocessing Systems

komplexes) process in the United States.6 The aforementioned sequence may change as improvements are discovered by researchers working in the collaborative countries. Competitive technologies are also under investigation. In European and Japanese work, for example, alkylamides and diamides are being investigated as replacement compounds for the phosphorouscontaining TBP and CMPO. The absence of phosphorous in the amides may be an important advantage in that molecules containing only C, H, O, and N atoms are incinerable, simplifying the disposal of spent solvents. Other benefits may include easier actinide back-extraction and solvent regeneration, and the production of relatively benign radiolysis products. Mainly in French work, a process known as DIAMEX (diamide extraction) would rely on the diamide dimethyl dioctyl hexylethoxymalonamide (DMDOHEMA) for trivalent actinide and lanthanide extraction.7 Tributylphosphate would be replaced by dihexyloctanamide (DHOA) in Indian work,8 or tetraoctyldiglycolamide (TODGA) in Japanese9 and European10 proposals. A process analogous to TALSPEAK employing a combination of malonamides and bistriazinebipyridines is called DIAMEX/ SANEX (diamide extraction/selective actinide extraction) and is being studied for the separation of the minor actinides from lanthanides. Dithiophosphinc acids are also under investigation as softdonor ligands for this difficult separation.11 Regardless of which complexing agents are adopted, they will need to be effective in a harsh environment. Dissolved nuclear fuel is very acidic and highly radioactive. This subjects complicated organic molecules that are designed to selectively complex specific metal ions to a severe hydrolysis and radiolysis environment. The adverse effects of these reactions could include a loss in solventextraction efficiency due to ligand degradation and concentration decrease, a decrease in separation factors between desired and undesired metals due to the creation of degradation products that are new complexing agents, and changes in viscosity or phase disengagement time due to the generation of high molecular weight decomposition products.12 For a process to be successful, the ligands and their diluents must be relatively stable in such an environment. Short contact times, such as those achievable with centrifugal contactors, are of benefit in minimizing the effects of hydrolysis and radiolysis on organic solvent-extraction formulations. This chapter reviews the radiation and acid hydrolysis stability of the main ligands and diluents currently being studied for application in the fuel cycle of the future.

369

5.15.2 Chemistry of Irradiated Aqueous and Organic Solutions 5.15.2.1 Produced Transient Reactive Species Most of the radiolytic damage done to ligands in an irradiated solvent-extraction solution occurs due to indirect radiolysis. Indirect radiolysis is caused by the reactions of the ligands with reactive transient species created by direct radiolysis of the diluent. Since ligands are normally present in millimolar concentrations, their diluent absorbs most of the radiation energy. In nuclear aqueous solvent extraction, these diluents are usually alkanes in the organic phase and aqueous nitric acid in the aqueous phase. Direct radiolysis ionizes and excites diluent molecules to produce short-lived but highly reactive ionic and free-radical species. These produced reactive species may undergo recombination to create molecular species, or they may diffuse away from their point of origin to react with ligands. For water irradiated by low linear energy transfer (LET) particles, the species in eqn [1] are produced with their yields (G-values in mmol J
1) shown in brackets13: H2 O


!½0:28 OH þ ½0:27e
aq þ ½0:06H þ ½0:07H2 O2 þ ½0:27H3 Oþ þ ½0:05H2

½1

The most reactive species produced are the oxidizing hydroxyl radical (OH) and hydrogen peroxide (H2O2), and the reducing aqueous electron (e
aq ) and hydrogen atom (H). Massive, highly charged particles such as a-particles (He2þ) generated mainly by actinide decay have short ranges and high LET (156 eV nm
1 for 5 MeV He2þ)14 and therefore deposit energy in areas of high localized concentrations of reactive species. These may undergo recombination, and thus higher yields of molecular species and lower yields of radicals are found. Oxidizing and reducing species have identical initial yields but in aerated nitric acid, the environment is predominantly oxidizing due to fast scavenger reactions. The aqueous electrons and hydrogen atoms are scavenged according to þ  e
aq þ H ! H

k ¼ 2:3  1010 M
1 s
1ð13Þ

½2


e
aq þ O2 ! O2

k ¼ 1:9  109 M
1 s
1ð13Þ

½3


2
 e
k ¼ 9:7  109 M
1 s
1ð15Þ aq þ NO3 ! NO3 

H þ O2 !HO2 k ¼ 2:1  1010ð16Þ

½4 ½5

370

Degradation Issues in Aqueous Reprocessing Systems 

ð17Þ HO2 $ Hþ þO
2 pka ¼ 4:8

½6

HNO3


This leaves H2O2 and the strongly oxidizing OH radical as the most reactive species in solution. The  OH radical reacts with organic solutes by hydrogen abstraction, addition to unsaturated carbon bonds or with organic and inorganic solutes by electron transfer reactions. Rate constants for OH radical reactions have been tabulated.13,18 Additional reactive species are created by nitric acid radiolysis. A thorough review of the radiation chemistry of nitric acid is given by Katsumura.19 Among the most important species, the oxidizing  NO3 radical is produced indirectly by the reaction of the OH radical with undissociated nitric acid and directly by radiolysis of the nitrate anion dissociation product of nitric acid20:

NO
3




OH þ HNO3 !NO3 þ H2 O k ¼ 5:3  107 M
1 s
1 NO
3


 ! e
aq þ NO3

½7 ½8

The reactions of the NO3 radical are similar to those of the OH radical, including hydrogen atom abstraction and addition and electron transfer reactions, although with decreased electrophilic character and lower reaction rates.21,22 Rate constants for its reaction with numerous species have been compiled.23 In acidic solution, the NO2
 radical product of 3 eqn [4] protonates24:  $ HNO
 NO2
 3 3 $ H2 NO3 pK a1 ¼ 4:8; pK a2 ¼ 7:5

½9

The product H2 NO3 decays to produce the NO2 radical25: H2 NO3 !NO2 þ H2 O

k ¼ 7  105 s
1

½10



This species is less reactive than the NO3 radical, but has been shown to add to unsaturated carbon bonds, or to carbon-centered radicals to produce nitrated derivatives of the original compound.26–28 Perhaps more importantly, the addition product of these two nitrogen-centered radicals decays to produce nitronium ion, a powerful nitrating species: 

NO2 þNO3 ! N2 O5 k ¼ 1:7  109 M
1 s
1ð20Þ
ð29Þ

N2 O5 ! NOþ 2 þ NO3

½11 ½12

Another important product of nitric acid radiolysis is nitrous acid. It is produced by direct and indirect effects:

ð30Þ

!O þ HNO2

þ NO
2 þ H $ HNO2 


ð31Þ

!O þ NO2

pK a ¼ 3:2ð32Þ

NO2 þ  NO2 ! N2 O4 kf ¼ 4:5  108 M
1 s
1ð33Þ kr ¼ 6:0  103 s
1ð33Þ

N2 O4 þ H2 O ! HNO2 þ HNO3 k ¼ 18M
1 s
1ð25Þ

½13 ½14 ½15 ½16f  ½16r ½17

Nitrous acid is also an important nitrating agent that catalyzes nitration reactions via production of the nitrosonium ion: HNO2 þ Hþ ! NOþ þ H2 O34

½18

Nitrosonium ion may then react with the organic compound by addition, shown for an aromatic species in eqns [19]–[21],35 or by electron transfer as shown in eqns [22] and [23]34: NOþ þ ArH ! ArHNOþ

½19

ArHNOþ ! ArNO þ Hþ

½20

ArNO þ HNO3 ! ArNO2 þ HNO2

½21

ArH þ NOþ ! ArHþ þ NO

½22

ArHþ þ  NO2 ! ArNO2 þ Hþ

½23

Nitrous acid also affects actinide oxidation states, especially for neptunium, in irradiated aqueous nitric acid solution. The valence state of neptunium is set by the concentration of nitrous acid according to eqn [24]36: þ
NpOþ 2 þ 3=2H þ 1=2NO3

$ NpO2þ 2 þ 1=2HNO2 þ H2 O

½24

The maximum concentration of radiolytically produced nitrous acid is limited by radiolytically produced H2O2, as shown37,38: H2 O2 þ HNO2 ! Hþ þ NO
3 þ H2 O

½25

It should be noted that irradiation is not required to produce nitrous acid in nitric acid solution. The nitric acid oxidation of solution constituents is also accompanied by the production of nitrous acid, an example of which was shown in eqn [21]. Equations [18]–[21] show that nitrous acid nitration is catalytic,

Degradation Issues in Aqueous Reprocessing Systems

and once produced, only small concentrations are necessary to cause the nitration of organic compounds. Irradiation enhances the production of nitrous acid in nitric acid, enhancing yields of nitrated products. Radiolysis also occurs in the organic phase. Normal and branched-chain alkanes are the typical organic diluents for the ligands used in nuclear solvent extraction. The radiolysis of alkanes is represented by39: CH3 ðCH2 Þn CH3


þ ! e
sol þ CH3 ðCH2 Þn CH3

þ CH3 ðCH2 Þn CH2 þCH3 þ H þ H2

½26

Specific yields for the products depend on the specific alkane irradiated. Branch-chain alkanes have higher product yields, in the range of 0.2–0.6 mmol J
1 for molecular hydrogen and 0.005–0.1 mmol J
1 for methane. Bishop and Firestone40 reported a yield of H atom of 0.07 mmol J
1 for C6–C10 hydrocarbons. The carbon-centered radical products may also be produced indirectly by hydrogen abstraction reactions with OH or NO3 radicals. Regardless of their origin, these carbon-centered radicals react with solutes by hydrogen atom abstraction or they may undergo radical–radical addition to create higher molecular weight products: CH3 ðCH2 Þn CH2 þ CH3 ðCH2 Þn CH2 ! CH3 ðCH2 Þ2nþ2 CH3

½27

As these higher molecular weight products accumulate, they change the physical characteristics of the solvent, including the phase disengagement time, density, and viscosity. These physical changes may in turn affect process performance. Alkane radicals may also undergo disproportionation, to produce unsaturated products41: CH3 ðCH2 Þn CH2 þ CH3 ðCH2 Þn CH2 ! CH3 ðCH2 Þn CH3 þ CH3 ðCH2 Þn
1 CH ¼ CH2

½28

Unsaturated products are susceptible to addition reactions by OH radical or N-centered radical species. Among the most important of carbon-centered radical reactions is oxygen addition to produce peroxyl radicals42: R þ O2 ! ROO

½29

Peroxyl radicals will undergo addition reactions to form tetroxides which then decompose to produce aldehydes, ketones, and alcohols from the original compound.43 They are therefore important intermediates in the oxidative mineralization of organic compounds by radiolysis.

371

5.15.3 PUREX Process 5.15.3.1

Tributylphosphate Degradation

The PUREX process for the extraction of the major actinides, especially uranium, consists of 30% TBP in an alkane diluent. It has been in use worldwide for decades for the recovery of fissionable materials from used nuclear fuel.3 The metal-loaded solvent is stripped with a mildly acidic aqueous phase for recovery of the actinide metal ions, and the solvent is recycled. However, its recycle potential is limited by the radiolytic degradation of TBP and its diluents. It has long been recognized that the major products of TBP radiolysis are hydrogen, methane, and dibutylphosphoric acid (HDBP), with monobutylphosphoric acid (H2MBP) and phosphoric acid produced in lesser amounts. The radiation chemistry of TBP was recently reviewed and the following discussion is abbreviated from that source.44 The accumulation of radiolytic degradation products in the PUREX solvent results in decreased extraction performance with regard to separation factors and physical parameters.45–47 The acidic radiolysis products of TBP degradation are complexing agents that interfere with uranium and plutonium stripping and fission-product separation factors.48–50 Interfacial crud formation and poor phase separation have been attributed to the formation of precipitable complexes of zirconium with H2MBP and phosphoric acid.47,51–55 The adverse effects of the buildup of these acidic phosphate products in the organic phase are mitigated during process extractions by solvent washing with aqueous Na2CO3.56,57 However, with continued recycling, washing becomes less effective and the washed solvent shows increased retention of undesirable Pu, Zr, and Ru and increased solution viscosity.58 This has been attributed to the accumulation of radiolysis products of higher molecular weight, resulting from radical addition reactions. These species have high organic phase solubility.47,59 The result is a permanently degraded and radioactively contaminated solvent that is expensive to dispose. Several mechanisms have been proposed to explain the production of HDBP in irradiated TBP solutions. Zaitsev and Khaikin60 reported that dissociative electron capture resulted in the production of the butyl radical and HDBP in irradiated neat TBP: 
e
sol þ ðC4 H9 OÞ3 PO ! C4 H9 þ ðC4 H9 OÞ2 OPO ½30

Jin et al.61 attributed the formation of HDBP under these conditions to a combination of dissociative

372

Degradation Issues in Aqueous Reprocessing Systems

electron capture and decay of excited TBP molecules, while Haase et al.62 reported that electron attachment could also result in free hydrogen atoms and a TBP carbon-centered radical. However, the electron-initiated reactions are unlikely to be of consequence in solvent extraction due to the scavenging of electrons by nitrate and hydronium ions in the acidic phase (eqns [2]–[4]). Additionally, the direct excitation of TBP becomes less important when TBP is dissolved in a diluent. Burr63 proposed that HDBP was formed by the decay of the TBP carbon-centered radical: ðC4 H9 OÞ2 ð C4 H8 OÞPO !



C4 Hþ 8

þ ðC4 H9 OÞ2 OPO

ðC4 H9 OÞ2 PðOÞ
O
ðC4 H8 Þþ þ H2 O ! ðC4 H9 OÞ2 PðOÞ
O
þ C3 H7 CHO þ 2Hþ

½36 68

Wilkinson and Williams proposed yet another mechanism for HDBP formation based upon direct TBP radiolysis: ðC4 H9 OÞ3 PO


þ ! e
sol þ ðC4 H9 OÞ3 PO

ðC4 H9 OÞ3 POþ ! ðC4 H9 OÞ2 PðOHÞOHþ þ CH2 ¼ CHCH CH3




½31

This TBP radical could be produced by hydrogen atom abstraction63 by reaction with either the radiolytically produced H atom or OH radical: ð OHÞH þ TBP ! ðC4 H9 OÞ2 ð C4 H8 OÞPO þ ðH2 OÞH2 kOH ¼ 5:0  109 M
1 s
1ð64Þ kH ¼ 1:8  108 M
1 s
1ð64Þ

½32a ½32b

Besides the direct decay to HDBP shown earlier, the TBP radical could also undergo hydrolysis to again produce HDBP65: ðC4 H9 OÞ2 ð C4 H8 OÞPO þ H2 O ! C4 H8 OH þ ðC4 H9 OÞ2 POO
þ Hþ ½33 Khaikin66 suggested that HDBP was also the stable product of dissolved oxygen addition to the TBP radical, which gives the TBP peroxyl radical: ðC4 H9 OÞ2 ð C4 H8 OÞPO þ O2 ! ðC4 H9 OÞ2 PðOÞ
O
ðC4 H8 ÞOO ½34 Superoxide elimination followed by hydrolysis produces HDBP and butyraldehyde, also a measured product.67 ðC4 H9 OÞ2 PðOÞ
O
ðC4 H8 ÞOO þ ! O
2 þ ðC4 H9 OÞ2 PðOÞ
O
ðC4 H8 Þ

½35

½37

½38

ðC4 H9 OÞ2 PðOHÞOHþ ! ðC4 H9 OÞ2 POOH þ Hþ

½39

Intramolecular hydrogen bonding between the phosporyl oxygen and a butoxy hydrogen atom of the radical cation formed in eqn [37] forms a ring structure, which decays as shown in eqns [38] and [39]. However, based on a relative rates analysis, Mincher et al.64 calculated that the literature G-values of 0.05– 0.5 mmol J
1 for TBP decomposition could be explained solely by radical reactions, without a need to invoke direct radiolysis. However, even with a contribution due to direct radiolysis, the reactions described earlier would all lead to the production of HDBP, and occur competitively. Continued irradiation produces H2MBP and phosphoric acid from HDBP via analogous reactions. It should be noted that although the G-value is commonly used to report the degradation rates of ligands in solvent formulations, it is not an ideal figure of merit for this purpose. Unlike the radical yields for a neat solution presented in eqn [1], the degradation yield of a solute irradiated in a solvent is highly dependent on the experimental conditions used. These include the presence or absence of an aqueous phase, the dissolved oxygen content of the solution, and even the ligand concentration in the irradiated solvent. Thus, they should be interpreted with caution when comparing various studies. Further, not all studies report results in terms of G-values. With these considerations in mind, G-values for some of the ligands discussed in this chapter are presented in Table 1. Among the less abundant but still important TBP radiolysis products are the acid phosphates of higher molecular weight. These species with varying alkane chain lengths suggest that radical addition reactions

Degradation Issues in Aqueous Reprocessing Systems

373

Table 1 Radiolytic degradation yields (G-value; mmol J
1) for selected ligands discussed in the text. It should be noted that the magnitude of the G-value depends on experimental conditions and ligand concentration Ligand

Solution

G

References

TBP Crown ether CMPO DMDOHEMA TODGA

Alkane solution in contact with aqueous HNO3 Alkane solution in contact with aqueous HNO3 TRUEX dodecane in contact with aqueous HNO3 0.65 M in alkane in contact with aqueous HNO3 Neat

0.05–0.5 0.1 0.01 0.4–0.5 0.85

44, 64 69 70 7, 71, 72 73, 74

occur between TBP, HDBP, and alkane solvent radicals, including the production of TBP dimers.52,75 These compounds of higher molecular weight are among those species that are not adequately removed from irradiated solvent by aqueous carbonate washing. Additional products of TBP radiolysis in the presence of nitric acid are nitrated phosphates, which also impede stripping efficiency. He et al.76 proposed that  NO3 reacts with TBP by hydrogen atom abstraction, producing the TBP radical: 

NO3 þ TBP ! ðC4 H9 OÞ2 ð C4 H8 OÞPO þ HNO3 k ¼ 4:3  106 M
1 s
1ð64Þ

½40

The TBP radical was then postulated to undergo reaction with additional NO3 to produce nitrated TBP77: 

NO3 þðC4 H9 OÞ2 ð C4 H8 OÞPO ! ðC4 H9 OÞ2 ðOC4 H8 NO3 ÞPO

½41



The NO2 radical might be expected to add in the same way: 

NO2 þðC4 H9 OÞ2 ð C4 H8 OÞPO ! ðC4 H9 OÞ2 ðOC4 H8 NO2 ÞPO

½42

Methylated, hydroxylated, and nitrated phosphates, resulting from the radical addition reactions of methyl radical, hydroxyl radical, and the nitro-radicals shown earlier, have been identified in postirradiation TBP solutions by numerous investigators.77–80 5.15.3.2

Alkane Diluent Degradation

The alkane diluent, usually kerosene, dodecane, or a mixture of normal paraffinic hydrocarbons (NPHs), also undergoes radiolytic degradation to generate metal complexing agents that do not wash out in solvent treatment with alkaline solutions.45 The decomposition of diluents in TBP solvent extraction

was reviewed by Tahraoui and Morris.81 Alkanes undergo radiolytic nitration of their carbon-centered radicals, in analogy with the aforementioned eqns [41] and [42]. Stieglitz and Becker47 and Tripathi et al.58 identified both nitro-, and nitrosoalkanes (RNO2 and RONO2) in alkanes irradiated in the presence of nitric acid. These nitroparaffins and their hydroxamic acid reaction products have been implicated in fission-product complexation in the PUREX process. Nitroparaffins are thought to be converted to the complexing enol form by contact with the alkaline scrub intended to remove the acidic products of TBP decomposition56: RCH2 N ¼ OðOÞ $ RCH ¼ NOðOÞ


½43

45,54

are metal complexing and Hydroxamic acids reducing agents formed from nitroparaffins by the Victor Meyer reaction: RCH2 NO2 ! RCONHOH

½44

Although hydroxamic acids rapidly hydrolyze in acidic media, small but steady-state organic phase concentrations have been identified in irradiated solvent.82–84 Additional diluent radiolysis products identified in irradiated TBP–dodecane–nitric acid include alkane oligomers, aliphatic ketones, and acids.85 Thus, stripping difficulties and poor separation factors result from a combination of higher molecular weight acidic phosphates from TBP radiolysis and from compounds produced by diluent nitration. As has been previously noted, nitration may also occur in the absence of irradiation when solvents are exposed to nitric acid. For example, when Blake et al.56 boiled Amsco 125-82 (a commercially available branch-chain aliphatic hydrocarbon solvent in the C12–C14 range) in contact with nitric acid under reflux, infrared analysis of the treated solvent showed absorption bands attributable to nitroparaffins and alkane oxidation products. Intense absorbance in the UV spectrum at 200 nm was also attributed to nitroparaffins, and was also obtained upon g-irradiation of

Degradation Issues in Aqueous Reprocessing Systems

the solvent. An absorbed dose of 450 kGy produced the same effect as an 11 h reflux in the presence of 2 M nitric acid. Less severe effects were noted for purified normal alkane diluents such as dodecane. According to Moore,86 the exclusion of nitrous acid from hot nitric acid solutions by the use of scavengers such as urea or sulfamic acid prevented degradation of alkane diluents. The products of reaction with nitrous acid were nitrite esters, nitroso compounds, and oxidation products, and the degree of reaction was found to be dependent on the nitric acid concentration. This resulted in enhanced fissionproduct retention by the solvent and increases in phase disengagement times, similar to the effects produced by the radiolytic nitration of the solvent. These reactions may be more important than those of  NO3 and NO2 radicals in irradiated acid.87 Finally, it should be noted that the use of more stable TBP diluents results in higher yields of HDBP due to increased TBP degradation. This has been attributed to the diluent ionization potential.81 Direct diluent radiolysis produces the diluent radical cation, as was shown in eqn [26]. A common radical cation stabilization route would be that of charge transfer by reaction with a solute such as TBP: ½CH3 ðCH2 ÞCH3 þ þ TBP ! CH3 ðCH2 Þn CH3 þ TBPþ

½45

Thus, a diluent with a higher ionization potential than the ligand is not necessarily desirable. For this reason, it has often been noted that the use of aromatic diluents, with their low ionization potentials, leads to more radiation-stable solvent-extraction systems. Examples of this will be discussed later. 5.15.3.3

Radiolysis Versus Hydrolysis

The nitration reactions in irradiated and unirradiated nitric acid produce similar products. It is also often reported that the products of radiolysis and hydrolysis are the same, although sometimes with differing yields. For example, Brodda and Heinen88 reported hydrolytically produced HDBP in 30% TBP/n-paraffin solutions that had been preequilibrated with nitric acid at 23  C. They found a linear relationship between the amount of HDBP produced and the acid concentration in the organic phase, although hydrolysis did not continue after the phases were separated. The product H2MBP was not detected, since it would have partitioned to the aqueous phase and that phase was not analyzed.

Hydrolysis presumably occurs according to the following reaction, in the presence of acidity: ðC4 H9 OÞ3 PO þ H2 O ½46

! C4 H9 OH þ ðC4 H9 OÞ2 POOH

As with radiolysis, continued dealkylation would result in the products H2MBP and H3PO4. Stieglitz und Becker47 measured the production of HDBP, H2MBP, and phosphoric acid from hydrolyzed 0.001 M TBP in various concentrations of aqueous nitric acid. Hydrolysis was represented according to eqn [47]: k1

k2

k3

TBP ! HDBP ! H2 MBP ! H3 PO4

½47

They noted a linear dependence of the rate constants k1–k3 on the concentration of nitric acid, as shown in Figure 1. The half-life for the loss in TBP concentration in 3 M HNO3 at a temperature of 50  C was 2310 h, and at 90  C, it was 30 h. Using mixed-phase experiments of 30% TBP in alkane diluent in contact with 3 M HNO3, it was found that hydrolysis rates were lower by a factor of 40–60. For mixed-phase experiments in the presence of uranium or zirconium, it was found that the hydrolysis rate increased and the dependence on nitric acid concentration was no longer evident. The authors attributed this to preferred hydrolysis of the metal–TBP complex. 0.015 k1 First-order rate constant, k (h-1)

374

0.01

0.005

k2 k3

0 -1

0

1 2 3 4 Nitric acid concentration (M)

5

6

Figure 1 The first-order rate constants (h
1) for the hydrolytic production of dibutylphosphoric acid (HDBP) from aqueous tributyl phosphate (k1), monobutylphosphoric acid (H2MBP) from HDBP (k2), and phosphoric acid from H2MBP (k3) as a function of nitric acid concentration at 75  C. Modified from data of Stieglitz, L.; Becker, R. Kerntechnik 1985, 46, 76–80.

Degradation Issues in Aqueous Reprocessing Systems

The related extractant triamylphosphate (TAP) was investigated for changes in Pu retention ability by Venkatesan et al.89 following both hydrolysis and radiolysis. Chemical degradation was studied by stirring the 1.1 M TAP/NPH solution with 4 M HNO3 at room temperature or at 75  C for periods up to 200 h. Increased Pu retention following hydrolysis was mitigated by sodium carbonate washing, and thus, it was attributed to the mono- and dibasic acids of TAP, in analogy with the hydrolysis and radiolysis of TBP.

5.15.4 Fission Product Extraction Processes In some proposals, the major heat-generating but short-lived fission products will be extracted for separate disposal following UREX. Separate disposal of these short-lived activities would permit more efficient utilization of the long-term geological repository. Two solvent-extraction processes have been considered for the separation of the main heatemitting radionuclides 137Cs and 90Sr from fuel dissolution. They are the cobalt dicarbollide– polyethylene glycol (HCCD–PEG) and fission product extraction (FPEX) processes, discussed subsequently. 5.15.4.1

HCCD–PEG Process Radiolysis

The HCCD–PEG process consists of cobalt dicarbollide (HCCD) for cesium extraction and PEG for strontium extraction, dissolved in a phenyltrifluoromethylsulfone diluent called FS-13 (C6H5SO2CF3).90 The structure of the sulfone is shown in Figure 2. The dependence of the cesium and strontium extraction distribution ratios for g-irradiated 0.12 M HCCD and 0.027 M PEG-400 in FS-13, in the presence of 1.5 M HNO3, is shown in Figure 3.91 There was no deleterious effect observed for cesium extraction efficiency even at an absorbed dose CF3 O

S

O

Figure 2 The structure of the sulfone diluent phenyltrifluoromethylsulfone (FS-13) used in the cobalt dicarbollide–polyethylene glycol fission product extraction process. The direct ionization of this compound initiates reactions that interfere with strontium extraction.

375

of 432 kGy. In contrast, strontium extraction distribution ratios decreased with absorbed dose in an exponential fashion, suggesting that PEG is susceptible to radiolytic damage. The apparent decrease in PEG concentration was proposed to be initiated by reaction with the OH radical product of water radiolysis, and the NO3 radical product of nitric acid radiolysis. The hydrogen atom abstraction reactions of these species and their associated rate constants with ethylene glycol are shown in eqns [48] and [49]92,93: HOCH2 CH2 OH þ OH ! H2 O þ CHOHCH2 OH k ¼ 2:4  109 L mol
1 s
1

½48

HOCH2 CH2 OH þ NO3 ! HNO3 þ CHOHCH2 OH k ¼ 1:6  106 L mol
1 s
1

½49

In either case, a carbon-centered PEG radical will be produced, which initiates oxidative decomposition of PEG by oxygen addition to produce the corresponding peroxyl radical (eqn [29]). However, as can be seen from Figure 3, the strontium extraction distribution ratios decreased exponentially, albeit at a lower rate, even when irradiated in the absence of an aqueous phase. This was attributed to reaction with the products of direct sulfone diluent radiolysis. The ionization of sulfone by irradiation is shown in eqn [50]: C6 H5 SO2 CF3 ! C6 H5 Sþ O2 CF3 þ e


½50

The decay of the produced radical cation results in the generation of the carbon trifluoride radical: C6 H5 Sþ O2 CF3 ! C6 H5 Sþ O2 þ CF3

½51



The addition product of the CF3 radical, hexafluorethane, is a measured product of sulfone irradiation. Hydrolysis of the sulfone radical cation product of eqn [50] with trace water in the solvent produces benzenesulfonic acid, also a measured product. C6 H5 Sþ O2 þ H2 O ! C6 H5 SO2 OH þ Hþ

½52

However, if the sulfur-centered cation product shown in eqn [51] undergoes an analogous reaction with PEG, benzenesulfonic acid also results, and the functionality of PEG as a strontium extractant would be decreased: C6 H5 Sþ O2 þ
CHðOHÞCH2 CHðOHÞ
! C6 H5 SO2 OH þ
CHðOHÞCH2 CHþ


½53

It can be seen from Figure 3 that HCCD–PEG solvent irradiated to high absorbed doses, and in the

376

Degradation Issues in Aqueous Reprocessing Systems

700

Strontium distribution ratio

600

500 Neat organic 400

300

200 Aqueous phase present 100

0 -100

0

100 200 300 Absorbed dose (kGy)

400

500

Figure 3 The decrease in strontium distribution ratios versus absorbed g-dose for cobalt dicarbollide–polyethylene glycol solvent formulation irradiated in the presence and the absence of the acidic aqueous phase.

presence of nitric acid, still provided adequate performance from which to implement a process. However, the HCCD–PEG process is unusual in that it relies on a heavy-phase organic solvent.90 This limits its compatibility with the other advanced reprocessing extractions, and it has been supplanted in current process designs by the FPEX process. 5.15.4.2

FPEX Process Radiolysis

The FPEX process was developed using the calixarene crown ether calix[4]arene-bis(t-octylbenzo)-crown-6 (BOBCalixC6) and the conventional crown ether: di-t-butylcyclohexano-18-crown-6 (DtBuCH18C6) in combination to facilitate the simultaneous extraction of cesium and strontium from acidic dissolved nuclear fuel.4 This formulation also contains 1-(2,2,3,3,-tetrafluoropropoxy)-3-(4-sec-butylphenoxy)2-propanol (Cs7SB) as a solvent modifier to help solubilize the BOBCalixC6 and the calixarene and crown ether metal complexes in an alkane diluent. The structures of the ligands and Cs7SB are shown in Figure 4. Steady-state 60Co g-ray irradiation experiments using 0.007 M BOBCalixC6, 0.075 M DtBuCH18C6, and 0.75 M Cs7SB, in Isopar L® have been reported.94

When irradiated in contact with 1.5 M HNO3, the organic phase changed from colorless to yellow, turning to yellow-orange with increasing absorbed dose. In an unpublished work by the same authors, the same color changes were noted in unirradiated samples after long exposures to the nitric acid phase. Investigation by gas chromatography with electron capture detection (ECD) revealed products that appeared only in the presence of the acidic aqueous phase. This combination of color change and new peaks with high ECD response was attributed to nitration reactions. Since the decrease in metal distribution ratios after irradiation was the same for both strontium and cesium, the effect of irradiation was proposed to be on the Cs7SB modifier, the only solvent constituent in common to both metal complexes. However, cesium and strontium extraction was only modestly affected to an absorbed dose of 200 kGy. A plot of distribution ratio versus absorbed dose revealed only a slight negative slope for both extracted metals, as shown in Figure 5. Thus, the FPEX process also appears to be robust enough to support a process. The radiation chemistry of the calixarenes and crown ethers was recently reviewed.95 The crown ethers exhibit high radiation stability, with low degradation G-values of about 0.1 mmol J
1 in organic

Degradation Issues in Aqueous Reprocessing Systems

377

OCH2CF2CF2H O

OH

O

O

O O

O O

O O

O

O

O O

O

O

O

O

O

O

Figure 4 The structures of 1-(2,2,3,3,-tetrafluoropropoxy)-3-(4-sec-butylphenoxy)-2-propanol (Cs7SB) (top); calix[4]arenebis-(tert-octylbenzo-crown-6) (BOBCalixC6) for Cs extraction (left); and 4,40 ,(50 )-di-(t-butyldicyclohexano)-18-crown-6 (DtBuCH18C6) for Sr extraction (right), used in the FPEX fission product extraction process.

12 11 Strontium Distribution ratio

10 9 8 7 6 Cesium 5 4 -50

0

50 100 150 Absorbed dose (kGy)

200

250

Figure 5 The analogous decrease in strontium and cesium distribution ratios versus absorbed g-dose for the fission product extraction solvent formulation irradiated in the presence of the acidic aqueous phase. Error bars shown are 10%.

solution irradiated in contact with aqueous nitric acid.69 This value is about half that for TBP degradation under the same conditions.44 Noncyclic polyether products result. For dicyclohexano-18-crown-6 (DCH18C6) irradiated in 1 M aqueous nitric acid to a very high absorbed dose of 3290 kGy, seven noncyclic products containing 3, 5, or 6 oxygen atoms were reported.69,96 They represent a range in product ether fragments from as small as ethane and up to fully half the macrocycle and suggest simultaneous splitting of two C–O bonds. When these water-soluble products were synthesized and millimolar amounts of each was

added to a surrogate spent-fuel raffinate solution, it was found that the extraction of uranium and strontium by DCH18C6 was not affected.97 However, plutonium extraction was suppressed by about 20% and this metal was precipitated by the most abundant radiolysis product (G ¼ 0.04 mmol J
1). The same products were produced by hydrolysis and radiolysis, although the abundance of the lower molecular weight products was favored in hydrolysis.98 For neat crown ether irradiation, recyclization to produce lighter molecular weight ethers has often been reported.95 However, recyclization in solution is not expected due to the abundance of reactive species in irradiated diluents available to react with crown ether radicals. The initiation of crown ether degradation to noncyclic ethers is thought to be due to hydrogen abstraction reactions99 to produce carbon-centered radicals. This initiates additional reactions resulting in C–O bond rupture. When Horwitz et al.100 irradiated DtBuCH18C6 in octanol, they reported that DSr was apparently stable until about 200 kGy, after which it dropped until it was one-third the original value at 300 kGy. Stripping distribution ratios climbed gradually with the absorbed dose, but were still 0.1 at 300 kGy. This decrease in performance was at least partially attributable to diluent radiolysis; however, solvent performance was acceptable, given that the absorbed doses employed were in excess of those expected in a process application. Decreases in DCs and DSr of 20–35% and an increase in DZr of more than 30% were reported by Shuler et al.101 for irradiated solutions of DCH18C6-derivative crown ethers. The macrocycles were dissolved in a TBP–kerosene

378

Degradation Issues in Aqueous Reprocessing Systems

O2N O

O O

O

O O

O O

O O O

O

Figure 6 The structure of a nitrated calix[4]arene, showing the presence of the nitro group which interferes with cesium complexation.

solution and irradiated in contact with the acidic aqueous phase over the dose range 20–100 kGy. The increase in DZr was probably due to TBP radiolysis products. The nitration of calixarene phenyl rings resulted in decreased cesium extraction efficiency as reported for the irradiation of these macrocycles in the presence of nitric acid.102,103 Higher levels of nitration, as determined by mass spectrometry, resulted in lower DCs, suggesting that nitro substitution occurred on the calixarene phenyl rings to sterically hinder cesium binding. This is shown in Figure 6. However, as noted earlier, no significant reduction in cesium extraction was found for the FPEX irradiations.94 This may be because BOBCalixC6 has alkylated phenyl groups activated toward nitration that are remote from the phenyl rings adjacent to the crown moiety, which is responsible for cesium complexation (Figure 4). These remote phenyl rings may scavenge nitrogen-centered radicals or nitrosonium ions that might otherwise react at the cesium binding site. Thus, the structure of BOBCalixC6 appears to provide a radiation stability superior to that of many other calix[4]arene-crown-6 ethers. While the result was fortuitous, it represents an example of the possibility of designing ligand molecules with high radiation resistance for use in nuclear solvent extraction.

5.15.5 Minor Actinide and Lanthanide Extraction Processes Following the extraction of major actinides and heat-emitting fission products in the steps described earlier, the aqueous raffinate contains the minor actinides americium and curium, as well as the lanthanides. The solution will also contain plutonium and neptunium depending on whether their valence states were previously adjusted to remove them in the initial

separation. The minor actinides, with Z 95, are predominantly trivalent in acidic aqueous solution and therefore have separations chemistry very similar to that of the lanthanides. Because they are long-lived a-emitters and are fissionable, it is desirable to partition them for inclusion in fast reactor fuels, rather than bury them as long-lived, high-level waste. This would allow of their use in energy production and their transmutation to short-lived fission products. However, the lanthanides are neutron poisons, and being relatively short-lived, they are intended for geological disposal. Thus, the separation of the trivalent 4f and 5f elements is necessary, yet this is a challenging separation even at the analytical scale. The separation is proposed to be performed in two steps. In the first, the actinides and lanthanides are extracted together using a conventional oxygendonor ligand. These are then stripped to an aqueous phase of proper acidity and the f-element separation is affected using soft-donor ligands. The radiation chemistry of these actinide–lanthanide extraction processes is discussed subsequently. 5.15.5.1

TRUEX Solvent Degradation

The ligand for the group separation extraction in US proposals is octyl(phenyl)-N,N-diisobutylcarbamoylmethyl phosphine oxide (CMPO), developed at the Argonne National Lab by Horwitz et al.5 The structure of this compound is shown in Figure 7. The TRUEX solvent is composed of 0.1–0.2 M CMPO, 1–2 M TBP, and a NPH diluent. The role of CMPO is to increase the extraction efficiency of the trivalent actinides and lanthanides over that of TBP alone, due to its more basic phosphoryl group.104 The g-irradiation of CMPO has been studied with respect to its effects on the solvent-extraction efficiency of americium. Irradiation has adverse affects on both forward extraction and stripping. For example, the effect of radiolysis on extraction by 0.2 M CMPO/1.4 M TBP in NPH was studied by Simozadeh et al.105 The DAm decreased from 30 to 10 at an absorbed dose of 848 kGy, for samples irradiated in contact with nitric acid concentrations between 1.6 and 6.0 M. Similar results were obtained by Logunov et al.106 for 0.2 M CMPO dissolved in m-nitrobenzotrifluoride. The DAm was decreased from 6 to about 3 after an absorbed dose of 1650 kGy, when in contact with 1 M HNO3. The irradiated solvent was reported to develop a dark reddishblack color, some of which was extracted by sodium

Degradation Issues in Aqueous Reprocessing Systems

O

O

O C4H9

P

O

H17C8

C4H9

P

N

OH

CMPO (0.12)

+

HN

H17C8

C4H9

379

C4H9

Octylphenylphosphinyl acetic acid (0.17)

O +

P

CO2

CH3

O H17C8 P

Octylphenylphosphine oxide OH

H17C8 Octylphenylphosphinic acid (0.22) Figure 7 The radiolytic decomposition of octylphenyldiisobutylcarbamoylmethyl phosphine oxide (CMPO) used in the transuranium extraction (TRUEX) process. The initial reaction is C–N bond cleavage to produce an acid and an amine. Decarboxylation of the acid produces a phosphine oxide which maintains the forward extraction of americium. Oxidation of the phosphine oxide to an organic soluble phosphinic acid interferes with stripping. The G-values (mmol J
1) given in brackets are for 0.2 M CMPO/1.2 M TBP–dodecane TRUEX formulation in contact with 5 M HNO3, as measured in Nash et al.70

carbonate washing, implying the creation of acidic, nitrated products. Americium stripping was also affected, for samples irradiated in contact with the nitric acid concentrations of 2.5 or 6 M, the DAm using 0.01 M HNO3 actually exceeded those of the forward extractions at the highest absorbed dose of about 600 kGy. For samples irradiationed in contact with 0.25 M nitric acid, the strip values were 1.107 The acid-induced thermal hydrolysis of CMPO solutions using the same nitric acid concentrations at elevated temperatures affected DAm similarly. The inability to strip after radiation exposure was attributed to the buildup of the acidic degradation products of CMPO and/or TBP radiolysis.107 Similarly, Chiarizia and Horwitz107 reported decreasing DAm for forward extractions and increasing DAm for stripping extractions for the irradiation of CMPO solutions in contact with 5 M HNO3. The changes in extraction efficiency were linear with an absorbed dose for 0.25 M CMPO in decalin. Americium distribution ratios from 2 M HNO3 decreased from 80 to 26 at an absorbed dose of 195 kGy, while stripping distribution ratios increased to 3000 over the same absorbed dose range. Carbonate washing decreased CMPO–decalin stripping distribution

ratios, although not restoring them to their initial values. This suggested the production of an organophilic acidic degradation product. The addition of TBP to CMPO solutions in the TRUEX formulation decreased stripping distribution ratios, perhaps due to TBP hydrogen bonding to CMPO degradation products.107 However, TBP would also scavenge radicals that might otherwise have reacted with CMPO. While the kinetic constants for the reactions of CMPO with the main radicals produced in irradiated aqueous nitric acid are unknown, TBP reacts at high rates with OH and  NO3.64 Thus, the presence of TBP as a TRUEX constituent not only improves CMPO solubility and third phase and metal loading characteristics,108 but also contributes to the radiation stability of CMPO. The radiolytic decomposition of CMPO was found to be first order versus absorbed dose, with
GCMPO ¼ 0.012 mmol J
1 for TRUEX–dodecane, 0.045 mmol J
1 for TRUEX–trichloroethylene (TCE), 0.055 mmol J
1 for CMPO–TCE, and 1.64 mmol J
1 for CMPO–CCl4.70 This dramatically illustrates the role of diluents in the radiolytic degradation of ligands. The chlorinated solvents resulted in lower radiation stability due to the production of chlorine radicals and HCl upon irradiation. These results also

380

Degradation Issues in Aqueous Reprocessing Systems

show the ability of TBP to protect CMPO from degradation. The formation G-values for the main radiolysis products of CMPO when irradiated in contact with 5 M HNO3 were measured by Nash et al.70 and are given in Figure 7. It can be seen that those for product appearance are actually higher than those for CMPO degradation. This inability to close the mass balance probably represents the difficulty of quantifying these product concentrations, or of making accurate absorbed dose measurements. These compounds, and other related species, were also identified by Mathur et al.109 using GC and GC–MS techniques. The products responsible for degraded stripping performance probably include organophilic acids; octylphenylphosphinic acid, octylphenylphosphorylacetic acid, and octylphenylphosphinylacetic acid from CMPO degradation, and HDBP from TBP degradation. The hydrolysis and radiolysis products of CMPO were found to be identical; however, they were produced in different yields.110 A mechanism was proposed by Chiarizia and Horwitz107 wherein CMPO degradation was initiated by C–N bond scission, resulting in the creation of a carboxylic acid and an amide. As shown in Figure 7, decarboxylation of the carboxylic acid results in methyloctylphenylphosphine oxide, a neutral compound that would complex americium to maintain high DAm values even as CMPO is degraded. Oxidation of this phosphine oxide would produce octylphenylphosphinic acid. The addition of these compounds to unirradiated CMPO–CCl4 solutions was used to qualitatively reproduce the changes in americium extraction efficiency observed in irradiated solutions.107 5.15.5.2

Diamide-Based Processes

In most of the European and Japanese work, tetraalkyldiamides have been proposed as replacement compounds for CMPO.111 The radiation chemistry of these compounds was recently reviewed, and the following discussion is adapted from that review.112 The potential advantages and uses of amides as actinide extractants have been reviewed by Gasparini and Grossi.113 They concluded that the products of tetraalkyldiamide radiolysis were water soluble and thus easily washed from the solvent. These authors also reported that the major hydrolysis products of alkylamides were due to C–N bond rupture to generate carboxylic acids and secondary amines, in analogy with the mechanism shown for CMPO in Figure 7.

The pentaalkylpropane diamides (malonamides) have the general structure (RR0 NCO)2CHR00 . Cuillerdier et al.114 compared the hydrolysis and radiolysis of diamide derivatives and concluded that stability as a function of R00 increased in the order H < C2H5 < C2H4OC6H13 < C2H4OC2H4OC6H13. Long oxyalky chains appear to protect extraction efficiency. The radiolysis products were not identified, although hydrolysis yielded octanol for the diamide with R00 ¼ C2H4OC8H17, suggesting cleavage at the ether linkage. It appears that the inclusion of a sacrificial ether linkage in the molecule may result in the generation of relatively harmless radiolysis and hydrolysis products while maintaining DIAMEX capability. This provides an example of intelligent ligand design, with respect to degradation effects. This ether cleavage is reminiscent of C–O bond cleavage in crown ether radiolysis, or C–N cleavage in amides, and may also be initiated by hydrogen abstraction reactions for irradiated solution. Other monoamides and diamide products were also reported. For the diamide with R00 ¼ C2H4OC2H4OC6H13, actinide extraction performance was reported to remain satisfactory after 700 kGy absorbed dose.114 The DIAMEX solventextraction process for the recovery of the minor actinides was developed using the malonamide DMDOHEMA shown in Figure 8. The hydrolysis and g-radiolysis of this compound have been thoroughly investigated by Berthon et al.7,71 The radiolysis of DMDOHEMA and two other malonamides in dodecane or TPH solutions in contact with nitric acid was examined using a suite of analytical techniques, including gas chromatography, potentiometry, and mass spectrometry. Higher acid concentrations led to greater rates of malonamide decomposition, with a
GDMDOHEMA of 0.37 mmol J
1 for 3 M HNO3. For 0.65 M DMDOHEMA in C6H13 O CH3

CH3

N

N C8H17

H17C8 O

O

Figure 8 The structure of the malonamide: dimethyl dioctyl hexylethoxymalonamide (DMDOHEMA) used in the DIAMEX actinide extraction process. The radiolytic decomposition of this compound is analogous to that for CMPO in Figure 7.

Degradation Issues in Aqueous Reprocessing Systems

dodecane also containing a dialkylphosphoric acid ligand, a
GDMDOHEMA of 0.55 mmol J
1 was obtained for irradiation in the presence of either a pH 3 or a 0.5 M nitric acid aqueous phase.71 Bisel et al.72 reported
GDMDOHEMA of 0.49 mmol J
1 for a 0.5 M alkane diluent solution also containing diethylhexylphosphoric acid (HDEHP), preequilibrated with 3 M HNO3 prior to irradiation. These G-values are somewhat higher than those usually reported for TBP.44 Organic acids were major radiolysis and hydrolysis products of diamides, and alcohols were created by a rupture of the ether linkages in the R00 group. The hydrolysis products of DMDOHEMA included seven species detectable by capillary zone electrophoresis with UV detection, including C4 through C7 carboxylic acids, three unidentified species, and an acidamide.115 The loss of one amide function by C–N bond rupture produced an amine and an acidic amide as products, when irradiated in the presence of the acidic aqueous phase. Products detected by Berthon et al.71 for DMDOHEMA radiolysis included the monoamide methyloctylhexyloxybutanamide, the acid amide methoxyoctylcarbamoyl 4-hexyloxybutanoic acid, the malonamides dimethyloctyl 2-hexyloxyethyl malonamide (DMOHEMA) and methyldioctyl 2-hexyloxyethyl malonamide (MDOHEMA), and the amine methyloctylamine. Thermal decomposition of the acidic amide resulted in the monoamide. The monoamide was a significant product in hydrolyzed solutions, but is rapidly degraded in irradiated solution. The mechanisms are similar to those proposed for the aforementioned CMPO radiolysis. In a diamide molecule which has suffered both the rupture of the C–N and C–O bonds, an amide lactone was produced by an intermolecular reaction between the acid and alcohol functional groups. For DMDOHEMA, with its eight carbon R0 group, all these species are soluble in the organic phase. A dose of 690 kGy to 1 M DMDOHEMA in dodecane in contact with 4 M nitric acid resulted in a final malonamide concentration of 0.59 M.116 The decrease in malonamide concentration was accompanied by decreases in DAm, DEu, and DNd. When solutions containing varying amounts of the major degradation products were prepared, all products were found to interfere with Am and Nd extraction, the amine being the most harmful. An acidic solvent wash quantitatively removed the amine degradation product, while an alkaline wash removed about 80% of the acidic amide. The combination of an acid wash,

381

followed by water scrubbing and then an alkaline wash restored the solvent’s extraction capabilities to that expected for the decreased malonamide concentration.71,72 In testing using a DMDOHEMA– HEDHP–TPH solvent irradiated and hydrolyzed in a continuous flow system that included alkaline solvent washing, Bisel et al.72 reported that most diamide degradation products were removed from the organic phase, with only the monoamide, the acidic amide, and MDOHEMA accumulating. The products due to an exposure equivalent to 1–2 years of process operation (800 kGy) did not affect metal distribution ratios, phase disengagement, surface tension, refractive index, or solvent density. Acceptable increases in viscosity were also measured. A related compound, the diglycolamide extractant N,N,N 0,N 0 -tetraoctyl-3-oxapentane-1,5-diamide (TODGA), shown in Figure 9, has also been examined as an actinide extractant. Exponential decreases in TODGA concentration with absorbed dose have been measured, with a
GTODGA for the neat compound of 0.85 mmol J
1. This is about twice the yield of decomposition as for TBP.44 Sugo et al.73,74 measured even higher rates of TODGA decomposition in dodecane and octane than for neat TODGA, with increasing rates with higher diluent content. The addition of benzene to dodecane solutions9,73 or the use of aromatic diluents9 decreased TODGA decomposition due to their low ionization potentials, as has been found for other ligands.117 Glycolamide derivatives containing aromatic functional groups also exhibited higher radiation stability, possibly due to intramolecular energy transfer to the benzyl ring, providing another example of attempts to design radiation-stable ligands.73 The products of TODGA radiolysis included dioctylamine, dioctylacetamide, dioctylglycolamide, and dioctylformamide.9,74 The presence of nitric acid did not enhance radiolytic degradation yields but did favor cleavage of the C–N bond to form the amine and acidic products over C–O bond cleavage to form acetamide and glycolamide, in analogy with the results for DMDOHEMA reported earlier. C8H17

C8H17 N

N O

H17C8 O

C8H17 O

Figure 9 The structure of diglycolamide extractant N,N, N0 ,N0 -tetraoctyl-3-oxapentane-1,5-diamide (TODGA).

382

Degradation Issues in Aqueous Reprocessing Systems

Modolo et al.10 investigated the postirradiation solvent-extraction performance of 0.2 M TODGA in alkane diluent. The DAm gradually decreased with the absorbed dose for TODGA and TODGA– TBP mixtures irradiated in the presence and absence of nitric acid. The DEu remained unchanged over the absorbed dose range 0–1000 kGy. No degradation products were detected by NMR. In an unpublished work, Modolo reported nine products of irradiated 0.1 M TODGA in TPH in the absence of the aqueous phase, as determined by HPLC–mass spectrometry. These included products created by the rupture of C–N bonds to eliminate one or more octyl groups from the amide nitrogen and C–N bond rupture to generate secondary amines. Carboxylic acids were not detected, however, the glycolamide and acetamide products of C–O bond cleavage were, in agreement with Sugo et al.9 for irradiation in the absence of the acidic aqueous phase. In general, the foregoing discussions demonstrate that the diamides are about twice as susceptible to degradation by radiolysis and hydrolysis as is CMPO. However, diamide degradation appears to have less severe affects on solvent extraction performance. The inclusion of the sacrificial ether linkage in the malonamides mitigates the production of harmful acidic degradation products by favoring the production of innocuous water-soluble alcohols. Solvent washing steps appear to be necessary for both processes, however.

5.15.6 Processes for the Separation of Minor Actinides from Lanthanides 5.15.6.1

TALSPEAK Process

To effect a separation between the trivalent actinides and lanthanides, soft donors such as nitrogen- or sulfur-containing complexants must be exploited to selectively complex the minor actinides. The process which has received most study in the United States is TALSPEAK, based upon the competition between HDEHP in the organic phase and lactate-buffered diethylenetriamine pentaacetic acid (DTPA) in the aqueous phase.6,118 The preferential complexation of the actinides by DTPA holds them in the aqueous phase while the lanthanides are extracted. Tachimori and Nakamura119 investigated the effects of HDEHP irradiation on solvent extraction as a component of the TALSPEAK formulation. When neat HDEHP was irradiated with g-rays, and then used to prepare a 0.5 M solution in NPH

diluent, both DAm and DNd for extractions from 0.05 M DTPA–1 M NaNO3 solutions at pH 3 gradually increased in a fashion analogous to the case of the maximum absorbed dose of 2000 kGy. This was attributed to an increase in the concentration of monoethylhexylphosphoric acid (H2MEHP), which also acts as a complexing agent. The separation factor was, however, unchanged. These authors also irradiated the aqueous phase as either 0.05 M DTPA in 1 M NaNO3, or 0.05 M DTPA in 1 M lactic acid.119 When these pH 3, irradiated aqueous solutions were then spiked with the lanthanide and actinide metals, they were next extracted with unirradiated and appropriately preequilibrated 0.5 M HDEHP solution. For the nitrate solution, there was a rapid increase in D for both metals, probably due to a decrease in concentration of the DTPA. The increase was more severe for DAm, resulting in a decrease in aAm/Nd. In contrast, the lactate solution showed a comparatively modest increase in D for both metals, but with a constant separation factor. The loss in DTPA concentration was also less for irradiated lactate solution, suggesting that lactate, in addition to serving the role of a pH buffer, was acting as a radical scavenger to protect DTPA. When the organic and aqueous phases were mixed during irradiation, simulating an actual process, lactate not only mitigated the adverse effects on the extraction but also minimized the radiolytic yield of H2MEHP and inorganic phosphates from HDEHP decomposition. Martin et al.120 measured the rate constant for the reaction of the OH radical with lactate ion and lactic acid over a range of temperatures. They found room temperature values of 7.8  108 (pH 6) and 5.2  108 M
1 s
1(pH 1), respectively. When compared with the literature value for the DTPA reaction of 5.2  109 M
1 s
1,121 it was determined via a relative rates analysis that lactic acid was incapable of completely protecting DTPA from OH radical reactions. Only 33% of OH reactions occur with the DTPA under these conditions. Increased protection for DTPA could be had by increasing the lactic acid concentration (2 M lactic acid would reduce DTPA reactions to 20%), or by substituting an organic acid with a faster OH radical reaction rate constant. Numerous acids have been used in TALSPEAKlike formulations,122 with a-hydroxybutyric and butyric acids having high OH radical reaction rate constants. These, and the associated aEu/Am for those systems, are given in Table 2. Since the OH radical reaction is likely to be one of H atom abstraction,120

Degradation Issues in Aqueous Reprocessing Systems

it can be predicted that organic acids with large, branched alkane groups will provide greater protection to DTPA. Unfortunately, none of the alternatives investigated to date provides both a good actinide– lanthanide separation factor and a high rate constant. 5.15.6.2 Bis(triazinyl)pyridine-Based Processes Among compounds that have shown promise for minor actinide–lanthanide separations are the tridentate bis(triazinyl)pyridines (BTPs). Their exceptional americium solvent-extraction efficiency and high separation factors for americium from europium aAm/Eu have prompted a great deal of investigation of these molecules for the purposes of minor actinide recovery from nuclear fuel dissolution. The general structure of the BTPs is shown in Figure 10. The key functional groups are the tridentate arrangement of soft-donor nitrogen atoms and the large alkyl Table 2 The OH radical reaction rate constants
1
1 (M s ) and Eu/Am separation factors for organic acids that have been used in TALSPEAK formulations Acid

aEu/Am

kOH

Formic Acetic Propionic Butyric Glycolic Lactic Citric Malonic a-Hydroxyisobutyric Glycine–HNO3

19 24 29 10 84 91 105 57 62 16

1.3  108 1.7  107 6.2  108 2.2  109 6.0  108 5.2  108 5.0  107 2.0  107 1.0  109 1.7  107

Rate constants are those of Buxton et al.,12 at pH 1–2, except for lactic acid which is from Martin et al.120 The aEu/Am are those of Nilsson and Nash.118

R2

383

substituents on the triazinyl rings that minimize the solubility of the protonated form of the extractant in the aqueous phase.123 Unfortunately, the BTPs have limited hydrolytic and radiolytic stability. When octanol solutions of the BTP 2,6-bis(5,6-n-propyl-1,2,4-3-yl)pyridine (nPr-BTP), were investigated for hydrolytic stability, it was found that DAm decreased by 80% only after 2 days of exposure to 1 M HNO3.124 Even greater losses in extraction efficiency were encountered in the presence of nitrous acid. A concentration of 0.01 M nitrous acid in 1 M nitric acid caused a decrease in DAm of 50% in only 2 h. Hydrolysis appeared to be initiated by attack on the a-CH2 group of the n-propyl chain of the triazinyl ring, and substitution with branched alkane chains greatly decreased its rate. Hydrolysis and radiolysis products of nPr-BTP included BTP derivatives that were hydroxy, keto, or nitro substituted at the a-CH2 position of one or more propyl chains, and dealkylation products.125 The isopropyl derivative (iPr-BTP) is more resistant to hydrolysis than nPr-BTP.126 In radiolysis experiments, 1.8  10
3 M 2,6-di (5,6-diethyl-1,2,4-triazin-3-yl)pyridine (tE-BTP) was irradiated by Nilsson et al.127 in 1-hexanol solution in the absence of an aqueous phase. Americium distribution ratios decreased with increasing absorbed dose, falling below unity at <5 kGy for extraction from 0.01 M HClO4. The authors proposed a mechanism for BTP degradation initiated by direct radiolysis of the alcoholic diluent: RCH2 OH


þ ! e
sol þ ½RCH2 OH

½54

Dissociative electron capture would then dealkylate the BTP, increasing its water solubility and decreasing distribution ratios for americium. Nilsson et al.127 found that the addition of nitrobenzene to hexanol solutions of tE-BTP inhibited the decrease in DAm and thus presumably the decrease in BTP concentration with absorbed dose. This was attributed to nitrobenzene scavenging of solvated electrons as shown in eqn [55]128: 
C6 H5 NO2 þ e
aq ! C6 H5 NO2

R1

N

N

R1

N N R1

N

N N

R1

Figure 10 General structure of the bis(triazinyl)pyridines (BTPs), soft-donor ligands used in the separation of the minor actinides from the lanthanides.

k ¼ 3:8  1010 M
1 s
1

½55

While these reactions adequately explain the situation in a pure organic phase, in the presence of an aqueous phase, the reactions of the OH radical with BTP must also be considered. This species, which probably reacts by H atom abstraction, also reacts quickly with nitrobenzene (3.9  109 M
1 s
1),12 and

384

Degradation Issues in Aqueous Reprocessing Systems

thus, nitrobenzene should provide the same protection under those circumstances. Kolarik129 and Ekberg et al.123 have reviewed the history of the development of these and other heterocyclic nitrogen donor ligands with regard to attempts to improve radiolytic and hydrolytic stability. The alkyl groups attached to the triazinyl rings were replaced with cyclohexyl rings, and the new molecules were designated BATPs, or annulated BTPs.130 These rings were also substituted with alkyl groups to eliminate a-benzylic hydrogen atoms, thus mitigating H atom abstraction reactions from initiating BTP decomposition at that vulnerable position. The resulting compound, shown in Figure 11, is 2,6-bis (5,5,8,8-tetramethyl-5,6,7,8-tetrahydrobenzo[1,2,4] triazine-3-yl)pyridine, or BATP-1, or CyMe4-BTP. This molecule provided both good hydrolytic stability and a very high aAm/Eu  5600 for extractions from 0.5 M HNO3, using malonamide-modified n-octanol diluent. The malonamide DMDOHEMA was added to the solvent formulation to improve actinide extraction kinetics, which can be slow with BATPs alone. Unfortunately, an absorbed dose of 100 kGy resulted in 80% decomposition of BATP-1 in n-octanol.130 Similarly, Hill et al.126 reported that DAm was reduced by 50% for 0.01 M solution of BATP-1 in malonamide-modified n-octanol in contact with 1 M HNO3, after an absorbed g dose of 100 kGy. Once again, the radical-scavenger nitrobenzene reduced the decomposition of CyMe4-BTP, with only 15% loss at the same absorbed dose. The very high distribution ratios for americium using the BATPs are actually a disadvantage due to difficulties in stripping. Therefore, a second pyridine ring was introduced into these compounds to moderate this high extraction efficiency.131 The resulting derivative of CyMe4-BTP is 6,60 -bis (5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-benzo[1,2,4] triazin-3-yl)[2,20 ]bipyridine, or CyMe4-BTBP, shown in Figure 12.

A solvent formulation containing 0.01 M CyMe4BTBP/0.25 M DMDOHEMA in n-octanol was found to provide a DAm  10 and aAm/Eu  120, for extractions from 1 M HNO3. These parameters were unchanged after a 2-month contact of the solvent with 1 M HNO3, indicating adequate hydrolytic stability. Radiolytic stability studies of CyMe4-BTBP as a ligand for use in the SANEX process were performed by Magnusson et al.132 The solvent formulation was 0.015 M CyMe4-BTBP/0.25 M DMDOHEMA in n-octanol, in contact with an equal volume of americium-spiked 1 M HNO3, using either a- or g-sources. These authors reported that g-irradiation effects were more severe, resulting in a loss in about 35% of extraction efficiency within 100 kGy, in agreement with Hill et al.126 and 70% within 1200 kGy. The loss in extraction efficiency was 40% greater due to g-irradiation. 5.15.6.3 Dithiophosphinic Acid-Based Processes Comparatively little work has been done on the radiation stability of the dithiophosphinic acids, which are also proposed for minor actinide separations. These compounds tend toward oxidation upon storage, even in the absence of hydrolysis and radiolysis. The products of the oxidation are the corresponding phosphinic acids, resulting from the replacement of S with O atoms.133 Modolo and Odoj134 studied bis(2,4,4-trimethylpentyl)dithiophosphinic acid (R2PSSH or Cyanex 301) radiolysis by irradiating both neat, purified R2PSSH and 0.5 M R2PSSH in n-dodecane. Using NMR, they found that for pure, neat R2PSSH, 80% was decomposed at 1000 kGy, with the production of 6% R2PSOH and 5% R2POOH. Numerous

N

N

N

N

N N

N N N

N

N

N N N

N

Figure 11 The structure of 2,6-bis(5,5,8,8-tetramethyl5,6,7,8-tetrahydrobenzo[1,2,4]triazine-3-yl)pyridine (BATP-1 or CyMe4-BTP) used in the separation of the minor actinides from the lanthanides.

Figure 12 The structure of 6,60 -bis(5,5,8,8-tetramethyl5,6,7,8-tetrahydro-benzo[1,2,4]triazin-3-yl)[2,20 ]bipyridine, or CyMe4-BTBP used in the separation of the minor actinides from the lanthanides.

Degradation Issues in Aqueous Reprocessing Systems

unidentified products were also detected. For 0.5 M R2PSSH in n-dodecane, irradiation over the range 0–700 kGy produced a steady increase in DEu, due to the production of monothiophosphinic and phosphinic acids. The pH of the aqueous phase was controlled in these experiments, thus the effects on distribution ratios cannot be accounted for by changing acidity. Following an absorbed dose of 700 kGy, the DEu was nearly as high as DAm. The aAm/Eu dropped from 1000 to 10 at pH 3.3. Substitution of dithiophosphinic acid alkane chains with aromatic groups apparently increases radiation stability, but also necessitates the use of aromatic diluents and neutral organophosphorus compound synergists.135 The aromatic derivatives bisphenyldithiophosphinic acid (Ph2PSSH) and bischlorophenyldithiophosphinic acid ((ClPh)2PSSH) were investigated for radiation stability as 0.5 M solutions in toluene with and without a 0.25 M TBP phase modifier by Modolo and Odoj.135 No aqueous phase was present during these irradiations. Over the range 100–700 kGy, the DAm decreased slightly, while DEu increased greatly resulting in a decreased separation factor. This interesting result was attributed to the radiolysis products of TBP. When TBP was added to irradiated, TBP-free dithiophosphinic acid solutions, the decrease in aEu/Am was not as dramatic. The difference in the effect of irradiation for the extraction of the two metals suggests that the actinides and lanthanides form different complexes in this mixed solvent system. Slope analysis was used to confirm this finding. In continued work, Modolo and Seekamp11 investigated the performance of (ClPh)2PSSH (Figure 13) for both radiolytic and hydrolytic stability. This

Cl

S P SH

Cl Figure 13 The structure of bis(chlorophenyl) dithiophosphinic acid, used in the separation of the minor actinides from the lanthanides.

385

compound has now been proposed for use in the ALINA (ActinideIII-Lanthanide INtergroup separation from Acidic medium) solvent system. Hydrolysis with 2 M HNO3 oxidized (ClPh)2PSSH to produce (ClPh)2PO(SCH3), and (ClPh)2PS(OCH3), (ClPh)2PO(OCH3), after 1-day exposure. These are the same products created during radiolysis, except that the abundances of the individual products differed. Hydrolytic degradation did not occur with exposure to HCl, and was insignificant with exposure to H2SO4. Further, the addition of nitrous acid scavengers such as hydrazine and urea mitigated the hydrolysis of (ClPh)2PSSH exposed to 2 M HNO3. When (ClPh)2PSSH was irradiated as a 0.5 M solution in toluene, about 40% of the starting material was destroyed after the high absorbed dose of 2000 kGy, with only slightly more decomposed when irradiated in the presence of 0.5 M HNO3. This probably illustrates the radical scavenging protective function of the aromatic diluent. When 0.5 M (ClPh)2PSSH was irradiated as a component of the ALINA solvent, also containing a 0.15 M trioctylphosphine oxide (TOPO) phase modifier in toluene, in the presence of 0.5 M HNO3, the distribution ratios for the lanthanides remained nearly constant while DAm decreased due to loss of the dithiophosphinic acid. This resulted in a rapidly decreasing separation factor. However, the absorbed doses studied were as high and the aAm/Eu was still >10 at an absorbed dose of 600 kGy.

5.15.7 Conclusions The conditions under which the organic molecules used in nuclear solvent extraction will be employed are demanding. They will be subjected to both acid hydrolysis and to reaction with the transient reactive species produced by direct diluent radiolysis. However, in the presence of the aqueous phase, the reducing transient species are scavenged by acidity and dissolved oxygen to produce secondary, less reactive species, and the system can thus be thought of as predominantly oxidizing with respect to organic compounds. The OH and NO3 radicals are especially important transient species, with fast reaction kinetics. Both are capable of electron transfer and hydrogen atom abstraction reactions, and addition to unsaturated compounds or carbon-centered radicals. Hydrogen atom abstraction or electron transfer reactions may result in dealkylation by creating carbon-centered radicals, initiating the reactions

386

Degradation Issues in Aqueous Reprocessing Systems

that cause the rupture of C–C bonds. Dealkylation may produce new complexing agents such as HDBP in the example of TBP radiolysis, which may be more water soluble as in the example of BTP radiolysis. C–N and C–O bonds are more susceptible, and therefore the rupture of these bonds results in the generation of amides from diamides and phosphinic acids, and lighter phosphine oxides from CMPO. This susceptibility has been used to design diamides containing both linkages that allow the compound to undergo radiolysis in a predictable way to produce nondeleterious products. Nitration of diluents and ligands must be expected even in the absence of irradiation in the presence of nitric acid. Nitration of aromatic functional groups and alkane diluents occur due to reaction with the nitrosonium ion when nitrous acid occurs in nitric acid. Nitrous acid is the product of nitric acid oxidation reactions, and is also produced by radiolysis. Nitration reactions also occur due to the addition of nitrogen-centered radicals such as NO3 and NO2 with carbon-centered radicals, although direct radical addition reactions are probably not common in the condensed acidic phase. Nitration of the diluent results in the production of undesirable nitroparaffinic complexing agents, while the nitration of ligands such as calixarenes may reduce extraction efficiency. The use of very stable diluents, however, may actually increase radiolytic decomposition of the ligands. The addition of scavengers, especially aromatic compounds and/or functional groups, may be used to reduce ligand and diluent decomposition. The radiolytic and hydrolytic degradation of solvent formulations is unavoidable in nuclear solvent extraction. However, an understanding of the reaction mechanisms involved can allow of intelligent ligand design or scavenger choices that improve the efficiency and lifetime of solvent formulations. The elucidation of these mechanisms requires the ability to perform both steady-state and pulse-radiolysis experiments.

References 1. 2. 3.

Hore-Lacy, I. Nuclear Energy in the 21st Century; World Nuclear University Press: London, 2006. Paviet-Hartmann, P.; Lineberry, M.; Benedict, R. In Nuclear Engineering Handbook; Kok, K. D., Ed.; CRC Press: Boca Raton, FL, 2009; pp 315–366. Schulz, W. W.; Navratil, J. D. Eds. Science and Technology of Tributyl Phosphate; CRC Press: Boca Raton, FL, 1984.

4. Riddle, C. L.; Baker, J. D.; Law, J. D.; et al. Solv. Extract. Ion Exchange 2005, 23, 449–461. 5. Horwitz, E. P.; Kalina, D. C.; Diamond, H.; Vandegrift, G. F.; Schulz, W. W. Solv. Extract. Ion Exchange 1985, 3, 75–109. 6. Weaver, B.; Kappelmann, F. A. J. Inorg. Nucl. Chem. 1968, 30, 263–272. 7. Berthon, L.; Morel, J. M.; Zorz, N.; Nicol, C.; Virelizier, H.; Madic, C. Separ. Sci. Technol. 2001, 36, 709–728. 8. Parikh, K. J.; Pathak, P. N.; Misra, S. K.; Tripathi, S. C.; Dakshinamoorthy, A.; Manchanda, V. K. Solv. Extract. Ion Exchange 2009, 27, 244–257. 9. Sugo, Y.; Saski, Y.; Tachimori, S. Radiochim. Acta 2002, 90, 161–165. 10. Modolo, G.; Asp, H.; Schreinemachers, C.; Vijgen, H. Solv. Extract. Ion Exchange 2007, 25, 703–721. 11. Modolo, G.; Seekamp, S. Solv. Extract. Ion Exchange 2002, 20, 195–210. 12. Pikaev, A. K.; Kabakchi, S. A.; Egorov, G. F. Radiat. Phys. Chem. 1988, 31, 789–803. 13. Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data 1988, 17, 513–886. 14. Pastina, B.; LaVerne, J. A. J. Phys. Chem. A 1999, 103, 1592–1597. 15. Elliot, A. J. Radiat. Phys. Chem. 1989, 34, 753–758. 16. Gordon, S.; Hart, E. J.; Thomas, J. K. J. Phys. Chem. 1964, 68, 1262–1264. 17. Bielski, B. H. J.; Cabelli, D. E.; Arudi, R. L.; Ross, A. B. J. Phys. Chem. Ref. Data 1985, 14, 1041–1100. 18. Dorfman, L. M.; Adams, G. E. Reactivity of the hydroxyl radical in aqueous solution; National Standard Reference Data Series, Report of the United States National Bureau of Standards 46; National Bureau of Standards: Washington DC, 1973. 19. Katsumura, Y. In N-Centered Radicals; Alfassi, Z. B., Ed.; Wiley: New York, 1998; pp 393–412. 20. Katsumura, Y.; Jiang, P. Y.; Nagaishi, R.; Oishi, T.; Ishigure, K.; Yoshida, Y. J. J. Phys. Chem. 1991, 95, 4435–4439. 21. Neta, P.; Huie, R. E. J. J. Phys. Chem. 1986, 90, 4644–4648. 22. Shastri, L. V.; Huie, R. E. Int. J. Chem. Kinet. 1990, 22, 505–512. 23. Neta, P.; Huie, R. E.; Ross, A. B. J. J. Phys. Chem. Ref. Data 1988, 17, 1027–1284. 24. Gra¨tzel, M.; Henglein, A.; Taniguchi, S. Ber. Bunsen-Ges. Phys. Chem. 1970, 74, 292–298. 25. Løgager, T.; Sehested, K. J. Phys. Chem. 1993, 97, 6664–6669. 26. Olah, G. A.; Lin, H. C.; Olah, J. A.; Narang, S. C. Proc. Natl. Acad. Sci. USA 1978, 75, 1045–1049. 27. Dzengel, J.; Theurich, J.; Bahnemann, D. W. Environ. Sci. Technol. 1999, 33, 294–300. 28. Ershov, B. G.; Gordeev, A. V.; Bykov, G. L.; Zubkov, A. A.; Kosareva, I. M. Mendeleev Commun. 2007, 17, 289–290. 29. Sworski, T. J.; Matthews, R. W.; Mahlman, H. A. Adv. Chem. Ser. 1968, 82, 164–181. 30. Nagaishi, R.; Jiang, P. Y.; Katsumura, Y.; Ishigure, K. J. Chem. Soc. Faraday Trans. 1994, 90, 591–595. 31. Daniels, M. Adv. Chem. Ser. 1968, 82, 153–163. 32. Lammel, G.; Perner, D.; Warnek, P. J. Phys. Chem. 1990, 94, 6141–6144. 33. Gra¨tzel, M.; Henglein, A.; Lilie, J.; Beck, G. Ber. Bunsen-Ges. Phys. Chem. 1969, 73, 646–653. 34. Ridd, J. H. Acta Chem. Scand. 1998, 52, 11–22. 35. Turney, T. A.; Wright, G. A. Chem. Rev. 1959, 59, 497–513. 36. Siddall, T. H.; Dukes, E. K. J. Am. Chem. Soc. 1959, 81, 790–794.

Degradation Issues in Aqueous Reprocessing Systems 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.

60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70.

Vione, D.; Maurino, V.; Minero, C.; Borghesi, D.; Lucchiari, M.; Pelizzetti, E. Environ. Sci. Technol. 2003, 37, 4635–4641. Bhattacharyya, P. K.; Veeraraghavan, R. Int. J. Chem. Kinet. 1977, 9, 629–640. Spinks, J. W. T.; Woods, R. J. An Introduction to Radiation Chemistry, 3rd edn.; Wiley: New York, 1990. Bishop, W. P.; Firestone, R. F. J. Phys. Chem. 1970, 74, 2274–2284. Dewhurst, H. A. J. Chem. Soc. 1958, 80, 5607–5610. Alfassi, Z. B. Ed. Peroxyl Radicals; Wiley: New York, 1997. von Sonntag, C.; Schuchmann, H. P. In Peroxyl radicals; Alfazzi, Z. B., Ed.; Wiley: New York, 1997; pp 173–234. Mincher, B. J.; Modolo, G.; Mezyk, S. P. Solv. Extract. Ion Exchange 2009, 27, 1–25. Lane, E. S. Nucl. Sci. Eng. 1963, 17, 620–625. Neace, J. C. Separ. Sci. Technol. 1983, 18, 1581–1594. Stieglitz, L.; Becker, R. Kerntechnik 1985, 46, 76–80. Tripathi, S. C.; Ramanujam, A. Separ. Sci. Technol. 2003, 38, 2307–2326. Davis, W., Jr. In Science and Technology of Tributyl Phosphate; Schulz, W. W., Navratil, J. D., Eds.; CRC Press: Boca Raton, FL, 1984; pp 221–266. Tripathi, S. C.; Sumathi, S.; Ramanujam, A. Separ. Sci. Technol. 1999, 34, 2887–2903. Egorov, G. F.; Tkhorzhnitskii, G. P.; Zilberman, B. Ya.; Schmidt, O. V.; Goletskii, N. D. Radiochemistry 2005, 47, 359–363. Rochon´, A. M. Radiochem. Radioanalyt. Lett. 1980, 44, 277–286. Sugai, H.; Munakata, K. Nucl. Technol. 1992, 99, 235–241. Egorov, G. F.; Afanas’ev, O. P.; Zilberman, B. Ya.; Makarychev-Mikhailov, M. N. Radiochemistry 2002, 44, 151–156. Miyake, C.; Hirose, M.; Yoshimura, T.; Ikeda, M.; Imoto, S. J. Nucl. Sci. Technol. 1990, 27, 157–166. Blake, C. A., Jr; Davis, W., Jr; Schmitt, J. M. Nucl. Sci. Eng. 1963, 17, 626–637. Reif, D. J. Separ. Sci. Technol. 1988, 23, 1285–1295. Tripathi, S. C.; Bindu, P.; Ramanujam, A. Separ. Sci. Technol. 2001, 36, 1463–1478. Wagner, R. M.; Towle, L. H. Radiation Stability of Organic Liquids; Semi Annual Report # 3 of the Stanford Research Institute AECU-4053; Oak Ridge National Laboratory: Oak Ridge, TN, 1958. Zaitsev, V. D.; Khaikin, G. I. High Energ. Chem. 1994, 28, 269–272. Jin, H.; Wu, J.; Zhang, X.; et al. Radiat. Phys. Chem. 1999, 54, 245–251. Haase, K. D.; Schulte-Frohlinde, D.; Kourˇı´m, P.; Vacek, K. Int. J. Radiat. Phys. Chem. 1973, 5, 351–360. Burr, J. G. Radiat. Res. 1958, 8, 214–221. Mincher, B. J.; Mezyk, S. P.; Martin, L. R. J. Phys. Chem 2008, 112, 6275–6280. von Sonntag, C.; Ansorge, G.; Sigimori, A.; Omori, T.; Koltzenburg, G.; Schulte-Frohlinde, D. Z. Naturforsch. Teil B 1972, 27, 471–472. Khaikin, G. I. High Energ. Chem. 1998, 32, 287–289. Clay, P. G.; Witort, M. Radiochem. Radioanalyt. Lett. 1974, 19, 101–107. Wilkinson, R. W.; Williams, T. F. J. Chem. Soc. 1961, 4098–4107. Draye, M.; Favre-Re´guillon, A.; Faure, R.; Lemaire, M. Radiat. Phys. Chem. 2008, 77, 581–584. Nash, K. L.; Rickert, P. G.; Horwitz, E. P. Solv. Extract. Ion Exchange 1989, 7, 655–675.

71.

387

Berthon, L.; Journet, S.; Lalia, V.; Morel, J. M.; Zorz, N.; Berthon, C.; Amekraz, B. In Proceedings of Atalante2004, Advances for Nuclear Fuel Cycles, Nıˆmes, France, June 21–25, 2004. European Commission Joint Research Center, Institute for Transuranium Elements, Karlsuhe, Germany, 2004; CD-ROM, Paper 2-08. 72. Bisel, I.; Came`s, B.; Faucon, M.; Rudloff, D.; Saucerotte, B. In Proceedings of Global-2007, Advanced Nuclear Fuel Cycles and Systems, Boise, ID, Sept 9–13, 2007; pp 1870–1873, CD-ROM. 73. Sugo, Y.; Sasaki, Y.; Kimura, T.; Sekine, T. In Proceedings of Global 2007, Advanced Nuclear Fuel Cycles and Systems, Boise, ID, Sept 9–13, 2007; American Nuclear Society, 2007; pp 1870–1873, CD-ROM. 74. Sugo, Y.; Izumi, Y.; Yoshida, Y.; et al. Radiat. Phys. Chem. 2007, 76, 794–800. 75. Adamov, V. M.; Andreev, V. I.; Belyaev, B. N.; et al. Kerntechnik 1990, 55, 133–137. 76. He, H.; Lin, M.; Muroya, Y.; Kudo, H.; Katsumura, Y. Phys. Chem. Chem. Phys. 2004, 6, 1264–1268. 77. Tripathi, S. C.; Ramanujam, A.; Gupta, K. K.; Bindu, P. Separ. Sci. Technol. 2001, 36, 2863–2883. 78. Adamov, V. M.; Andreev, V. I.; Belyaev, B. N.; et al. Soviet Radiochem. 1988, 29, 775–781. 79. Nowak, Z. Nukleonika 1977, 22, 155–172. 80. Lasage, D.; Virelizier, H.; Jankowski, C. K.; Tabet, J. C. Spectroscopy 1997, 13, 275–290. 81. Tahraoui, A.; Morris, J. H. Separ. Sci. Technol. 1995, 30, 2603–2630. 82. Zaitsev, V. D.; Karasev, A. L.; Egorov, G. F. Soviet Atomic Energy 1987, 63, 524–527. 83. Hao-xin, H.; Guo-hui, Z.; Shu-bin, H. Radiochim. Acta 1988, 46, 159–162. 84. Ohwada, K. J. Nucl. Sci. Technol. 1968, 5, 163–167. 85. Becker, R.; Stieglitz, L.; Bautz, H. Untersuchung der strahlenchemischen TBP-Zersetzung unter den Bedingungen des PUREX-Prozesses; KfK3639; Report of the Kernforsschungzentrum Kahrlsruhe, Kahrlsruhe: Germany, 1983. 86. Moore, R. H. Chemical stability of PUREX and uranium recovery process solvent; Report of the Atomic Energy Commission HW 34501; General Electric Company: Schenectady, NY, 1955. 87. Mincher, B. J.; Elias, G.; Martin, L. R.; Mezyk, S. P. J. Radioanalyt. Nucl. Chem. 2009; DOI 10.1007/s 10967009-0156-x. 88. Brodda, B. G.; Heinen, D. Nucl. Technol. 1977, 34, 428–437. 89. Venkatesan, K. A.; Robertselvan, B.; Antony, M. P.; Srinivasan, T. G.; Vasudeva Rao, P. R. Solv. Extract. Ion Exchange 2006, 24, 747–763. 90. Herbst, R. S.; Law, J. D.; Todd, T. A.; et al. Separ. Sci. Technol. 2002, 37, 1807–1831. 91. Mincher, B. J.; Herbst, R. S.; Tillotson, R. D.; Mezyk, S. P. Solv. Extract. Ion Exchange 2007, 25, 747–755. 92. Matheson, M. S.; Mamau, A.; Silverman, J.; Rabani, J. J. Phys. Chem. 1973, 77, 2420–2424. 93. Pikaev, A. K.; Sibirskaya, G. K.; Shirshov, E. M.; Glazunov, P. Ya.; Spitsyn, V. I. Doklady Proc. Acad. Sci. USSR 1974, 215, 328–331. 94. Mincher, B. J.; Mezyk, S. P.; Bauer, W. F.; Elias, G.; Riddle, C.; Peterman, D. R. Solv. Extract. Ion Exchange 2007, 25, 593–601. 95. Mincher, B. J.; Modolo, G.; Mezyk, S. P. Solv. Extract. Ion Exchange 2009, 27, 331–353. 96. Draye, M.; Chomel, R.; Doutreluingne, P.; Guy, A.; Foos, J.; Lemaire, M. J. Radioanalyt. Nucl. Chem. Lett. 1993, 175, 55–62.

388 97. 98. 99. 100. 101. 102. 103. 104. 105.

106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116.

117.

Degradation Issues in Aqueous Reprocessing Systems Draye, M.; Favre-Re´guillon, A.; Foos, J.; Guy, A. Radiochim. Acta 1997, 78, 105–109. Draye, M.; Favre’-Re´guillon, A.; Chomel, R.; et al. Bull. Chem. Soc. France 1996, 133, 183–197. Netserov, S. V.; Panov, O. Yu. J. Radioanalyt. Nucl. Chem. 1998, 231, 87–93. Horwitz, E. P.; Dietz, M. L.; Fisher, D. E. Solv. Extract. Ion Exchange 1991, 9, 1–25. Schuler, R. G.; Bowers, C. B., Jr; Smith, J. E., Jr; Van Brunt, V.; Davis, M. W., Jr. Solv. Extract. Ion Exchange 1985, 3, 567–604. Jankowski, C. K.; Van Calsteren, M. R.; Aychet, N.; Dozol, J. F.; Lamouroux, C. Can. J. Chem. 2005, 83, 1098–1113. Jankowski, C. K.; Allain, F.; Dozol, J. F.; et al. Rapid Commun. Mass Spectrom. 2003, 17, 1247–1255. Kalina, D. G.; Horwitz, E. P.; Kaplan, L.; Muscatello, A. C. Separ. Sci. Technol. 1981, 16, 1127–1145. Simonzadeh, N.; Crabtree, A. M.; Trevorrow, L. E.; Vandegrift, G. F. Radiolysis and hydrolysis of TRUEX-NPH solvent; Report of the Argonne National Lab ANL-90/14; Argonne National Laboratory: Argonne, IL, 1992. Logunov, M. V.; Voroshilov, Yu. A.; Starovoitov, N. P.; et al. Radiochemistry 2006, 48, 55–61. Chiarizia, R.; Horwitz, E. P. Solv. Extract. Ion Exchange 1986, 4, 677–723. Schulz, W. W.; Horwitz, E. P. Separ. Sci. Technol. 1988, 23, 1191–1210. Mathur, J. N.; Murali, M. S.; Ruikar, P. B.; et al. Separ. Sci. Technol. 1998, 33, 2179–2196. Nash, K. L.; Gatrone, R. C.; Clark, G. A.; Rickert, P. G.; Horwitz, E. P. Separ. Sci. Technol. 1988, 23, 1355–1372. Thiollet, G.; Musikas, C. Solv. Extract. Ion Exchange 1989, 7, 813–827. Mincher, B. J.; Modolo, G.; Mezyk, S. P. Solv. Extract. Ion Exchange 2009, 27, 579–606. Gasparini, G. M.; Grossi, G. Solv. Extract. Ion Exchange 1986, 4, 1233–1271. Cuillerdier, C.; Musikas, C.; Hoel, P.; Nigond, L.; Vitart, X. Separ. Sci. Technol. 1991, 26, 1229–1244. Rivasseau, C.; Blanc, P. J. Chromatogr. A 2001, 920, 345–358. Nicol, C.; Cames, B.; Margot, L.; Ramain, L. In Proceedings of Atalante-2000, Scientific Research on the Back-End of the Fuel Cycle for the 21st Century, Avignon, France, Oct 24–26, 2000; European Commission Joint Research Center, Institute for Transuranium Elements, Karlsuhe, Germany, 2000; CD-ROM, Paper 3–22. Bellido, A. V.; Rubenich, M. N. Radiochim. Acta 1984, 36, 61–64.

118. 119. 120. 121. 122.

123. 124.

125.

126.

127. 128. 129. 130. 131. 132. 133. 134. 135.

Nilsson, M.; Nash, K. L. Solv. Extract. Ion Exchange 2007, 25, 665–701. Tachimori, S.; Nakamura, H. J. Radioanalyt. Nucl. Chem. 1979, 52, 343–354. Martin, L. R.; Mezyk, S. P.; Mincher, B. J. J. Phys. Chem. 2009, 113, 141–145. Cabelli, D. E.; Rush, J. D.; Thomas, M. J.; Bielski, B. H. J. J. Phys. Chem. 1989, 93, 3579–3586. Weaver, B.; Kappelmann, F. A. TALSPEAK: A new method of separating americium and curium from the lanthanides by extraction from an aqueous solution of an aminopolyacetic acid complex with a monoacidic organophosphate or phosphonate; Report of Oak Ridge National Laboratory ORNL-3559; Oak Ridge National Laboratory: Oak Ridge, TN, 1964. Ekberg, C.; Fermvik, A.; Retegan, T.; et al. Radiochim. Acta 2008, 96, 225–333. Hill, C.; Guillaneux, D.; Berthon, L. SANEX-BTP process development studies. In Proceedings of the International Solvent extraction Conference, Cape Town, RSA, Mar 17–21, 2002; Chris van Rensburg, 2002; pp 1205–1209, ISEC 2002. Hill, C.; Berthon, L.; Bros, P.; Dancausse, J. P.; Guillaneux, D. SANEX-BTP process development studies. In Proceedings of the 7th International Exchange Meeting on Actinide and Fission Product Partitioning, Jeju, Republic of Korea, Oct 14–16, 2002; pp 453–461, OECA/NEA. Hill, C.; Berthon, L.; Madic, C. Study of the stability of BTP extractants under radiolysis. In Proceedings of Global 2005, Nuclear Energy Systems for Future Generation and Global Sustainability, Tsukuba, Japan, Oct 9–13, 2005; Japan Atomic Energy Agency, 2005; CD-ROM, Paper 283. Nilsson, M.; Andersson, S.; Ekberg, C.; et al. Radiochim. Acta 2006, 94, 103–106. Peiris, S. A.; Freeman, G. R. Can. J. Chem. 1991, 69, 884–892. Kolarik, Z. Chem. Rev. 2008, 108, 4208–4252. Hudson, M. J.; Boucher, C. E.; Braekers, D.; et al. New J. Chem. 2006, 30, 1171–1183. Geist, A.; Hill, C.; Modolo, G.; Foreman, M. R. S. t. J. Solv. Extract. Ion Exchange 2006, 24, 463–483. Magnusson, D.; Christiansen, B.; Malmbeck, R.; Glatz, J. P. Radiochim. Acta 2009, 97, 497–502. Higgins, W. A.; Vogel, P. W.; Craig, E. G. J. Am. Chem. Soc. 1955, 77, 1864–1866. Modolo, G.; Odoj, R. J. Radioanalyt. Nucl. Chem. 1998, 228, 83–88. Modolo, G.; Odoj, R. Solv. Extract. Ion Exchange 1999, 17, 33–53.

5.16

Spent Fuel as Waste Material

P. Carbol, D. H. Wegen, and T. Wiss European Commission, Joint Research Centre, Institute for Transuranium Elements, Karlsruhe, Germany

P. Fors Vattenfall Power Consultant AB, Go¨teborg, Sweden

ß 2012 Elsevier Ltd. All rights reserved.

5.16.1 5.16.1.1 5.16.2 5.16.2.1 5.16.2.1.1 5.16.2.1.2 5.16.2.1.3 5.16.3 5.16.3.1 5.16.3.2 5.16.3.2.1 5.16.3.2.2 5.16.3.2.3 5.16.3.2.4 5.16.4 5.16.4.1 5.16.4.2 5.16.4.2.1 5.16.4.2.2 5.16.4.3 5.16.4.3.1 5.16.4.3.2 5.16.4.3.3 5.16.4.3.4 5.16.4.3.5 5.16.4.3.6 5.16.4.3.7 5.16.5 References

Introduction Waste Handling Strategies Fuel Evolution During Storage Concentration of Radionuclides/Elements in Fuel UO2 fuel MOX fuel Comparison of UO2 and MOX Mechanical Stability of Spent Nuclear Fuel After Irradiation Evaluation of the Temperature and Thermal Conductivity Structural Stability of the Fuel Radiation damage Helium built-up Oxidation of the fuel Fuel cladding interaction/mechanical properties Chemical Stability of Spent Nuclear Fuel After Irradiation Distribution and Chemistry of the Fission Products Storage Before Final Disposal Wet storage of fuel in cooling ponds Dry storage of fuel Repository Storage Groundwater composition Corrosion of copper Iron canister corrosion Radiolysis Corrosion of spent fuel under reducing conditions Corrosion of fuel under oxidizing conditions Instant release Summary

Abbreviations ANDRA

CANDU CASTOR dpa EBS FGR FIAP

Agence Nationale Pour la Gestion des De´chets Radioactifs (The French National Radioactive Waste Management Agency) Canadian deuterium uranium reactor Cask for storage and transport of radioactive material Dislocations per atom Engineered barrier system Fission gas release Fraction of inventory in aqueous phase

GW GWd/tHM HBS HLW IRF KBS-3

LWR MBq MOX

390 390 392 393 393 395 397 399 399 400 400 401 402 402 402 402 404 405 405 405 405 408 408 408 409 414 415 417 417

groundwater Giga Watt day per tonne heavy metal High burn-up structure High-level waste Instant release fraction Ka¨rnavfallets Behandling och Slutfo¨rvaring-3 (nuclear waste treatment and disposal, report number 3) Light water reactor Mega Becquerel Mixed oxide fuel ((U,Pu)O2)

389

390

Spent Fuel as Waste Material

NAGRA

PWR SHE SKB

SNF STP tHM XRD Zircaloy

Nationale Genossenschaft fu¨r die Lagerung Radioaktiver Abfa¨lle (Swiss National Cooperative for the Disposal of Radioactive Waste) Pressurized water reactor Standard hydrogen electrode (a reference) Svensk Ka¨rnbra¨nslehantering AB (Swedish Nuclear Fuel and Waste Management Co.) Spent nuclear fuel Standard temperature and pressure (273.15 K, 100 kPa) tones of heavy metal X-ray diffraction Zirconium alloy, used as fuel pellet cladding

5.16.1 Introduction The moment a nuclear fuel assembly is taken out of a commercial reactor, it becomes spent nuclear fuel (SNF). At this point, two main options exist for the future handling of the fuel. The decision on the preferred option varies among the countries and is based on whether the state considers the SNF as waste or a resource. The once-through cycle encompasses final storage of the SNF in, for example, a deep underground repository. The emplacement takes place after around 40 years of storage in cooling ponds or after 1 year of water cooling at the nuclear power plant followed by decades of dry storage in, for example, Castor® containers. In the closed fuel cycle, the fuel is reprocessed, that is, the uranium and plutonium (95 wt%) is separated from the rest, leaving fission products such as Tc, Mo, I, Cs, and lanthanides and heavy actinides such as Np, Am, and Cm as waste.1 The U and Pu are reused in new fuel assemblies, whereas the waste is vitrified in glass and intermediately stored at the power plant while waiting for a final storage solution. At the moment, most spent fuel is kept in storage at nuclear power plants, at centralized storage sites, and at reprocessing facilities. This fuel has been successfully and safely stored in wet and dry conditions for several decades. As of the end of 2004, a total of 276 000 tonnes of heavy metal (tHM), that is, U and Pu, had been discharged from nuclear power reactors in the world, of which 90 000 tHM had been reprocessed2

(see Chapter 5.14, Spent Fuel Dissolution and Reprocessing Processes). The progress in finalizing spent fuel repositories has been slow, forcing countries to enlarge their intermediate storage capacity for spent fuel. Without decisions on more permanent solutions, there could be the prospect of continued storage for times of up to and beyond one hundred years. Looking to the future, the projection of the total amount of spent fuel to be discharged from reactors would reach 445 000 tHM in 2020.2 Consequently, the management of SNF will be a key issue for strategic, economic, and safety reasons for the future of nuclear power, and is an issue that many states yet have to decide upon.3 This chapter focuses on SNF as waste material and aims to highlight the changes of the most important fuel properties during storage. Many types of SNF exist; however, as the vast majority of the SNF kept in storage today is UO2 and MOX (mixed U and Pu oxide) fuels, these fuel types are given most focus. The main emphasis in this chapter, when it comes to the type of repository, is laid on a deep granitic repository since this type is closest to a stepwise implementation (Sweden handled in their application to build the repository, to the regulatory body, in 2011). 5.16.1.1

Waste Handling Strategies

The nuclear power-producing countries have undertaken different strategies for the final storage of their SNF. Most countries have developed concepts in which the fuel is emplaced in subsurface repositories below the groundwater table. The selected geological formation varies among these countries, but the main options are crystalline bedrock, clay, or salt. These repository types are all expected to become free of oxidants once they are closed. The United States, on the other hand, have developed a dry storage concept above the groundwater table. In this concept, the fuel containers will be surrounded by air. Table 1 summarizes the key features of some different concepts. Depending on the type of repository, different engineered barrier systems (EBS) have been developed. The following paragraphs aim to give some details on the EBS in each repository concept. The strategy in Belgium is to dispose the spent fuel in a deep geological repository excavated in the Boom clay formation. The main feature of EBS is the, socalled, supercontainer. In the supercontainer, containment is achieved by placing the canisters of spent fuel assemblies in a carbon steel overpack and surrounding

Table 1

Characteristics of selected disposal systems for spent fuel repositories (main information from Bennett and Gens4)

Country

Waste support/inner container

Overpack

Buffer material

Buffer pore water pH

Host rock

Groundwater

Peak temperature at outer surface of overpack

Peak g-dose rate at outer surface of overpack

Belgium

3 cm carbon steel

Boom clay

Reducing NaHCO3 waters, pH
8.5 Repository site search from 2009


95  C


25 Gy h1

5 cm copper

Portland cement Bentonite

12.5–13.5

Canadaa

Cast iron and sand Cast iron

Finland

Cast iron

5 cm copper

Bentonite

7–8


70–90  C


0.33 Gy h1

France

Cast iron

5.5 cm carbon steel

None

–

Callo-Oxfordian clay


90  C

<10 Gy h1

Sweden

Cast iron

5 cm copper

Bentonite

7–8

Fractured crystalline basement

Switzerland

Carbon steel

15 cm carbon steel

Bentonite

7–8

Opalinus clay

United Kingdom United Statesc,d

Cast iron stainless steel, 316

5 cm copper Nickel–chromium alloy, Alloy-22

Bentonite Titanium drip shields

7–8 5.5–12.0 depend on evaporation

Not determined Tuff, Yucca Mountain (reconsidering)

Brackish Na–Cl to saline Na–Ca–Cl waters. Eh –300 mV, pH 7.5–8 Reducing Ca–Na–CO3 waters, near-neutral pH Dilute Na–HCO3 waters, brackish Na–Ca–Cl waters and saline Ca–Na–Cl waters; reducing Eh Reducing, nearneutral, Na–Cl waters Assumed reducing Oxidizing


90  C

140–160  Cb


0.035 Gy h1

Assumed <100  C
100–200  C

Suggested repository concept.5 See Johnson and King.6 c The Yucca Mountain Project came to an official halt in March 2010. d See refs. EPW,7 Carter and Pigford,8 Mon and Hua,9 Long and Ewing,10 and Rosenberg et al.11 In order to reduce the number of radionuclides, only the radionuclides that were present with substantial activity after 1 year, have a deviation of more than  40%, and have a half-life longer than 1 year, are listed. Positive values indicate that radionuclides are produced in larger proportion in MOX than in UO2 fuel. b

Spent Fuel as Waste Material

a

Canadian shield granite, Ordovician sedimentary rock basins Fractured crystalline basement

391

392

Spent Fuel as Waste Material

the overpack with a Portland cement concrete buffer and an outer stainless steel envelope. Both Finland and Sweden are moving toward geological disposal of SNF using the, so-called, KBS-3 (Ka¨rnavfallets Behandling och Slutfo¨rvaring-3 (nuclear waste treatment and disposal, report number 3)) concept. In this concept, the spent fuel will be disposed of at a depth of about 500 m in crystalline bedrock. The main barrier to radionuclide release is a cylindrical copper-lined cast-iron canister.4 The canister is surrounded by a bentonite clay buffer, which is designed to provide mechanical protection and to limit the access of groundwater, living microbes, and corrosive substances to their surfaces (see Chapter 5.17, Waste Containers). Reducing conditions are expected to prevail inside the canister due to the anoxic corrosion, both before and after water intrusion, of the cast-iron inserts. A number of concepts have been developed, by Nuclear Waste Management Organization, in Canada. The present is based on the storage of the SNF in a deep rock formation and shows great similarities with the KBS-3 concept. The EBS is a copper-lined castiron container surrounded by a bentonite clay buffer. However, the decision taken at a governmental level is that, although a technically safe disposal could be achieved, the level of public acceptance is insufficient to allow the proponents to progress to a site selection stage (2008). The French National Radioactive Waste Management Agency, ANDRA (Agence Nationale Pour la Gestion des De´chets Radioactifs), is investigating reversible and irreversible radioactive waste storage/ geological disposal in deep granite and clay formations. The French concepts for spent fuel disposal foresees carbon steel as the first choice material for the overpack because under the relevant geochemical conditions, it is less prone to localized corrosion than materials that are passivated (e.g., stainless steels, nickel-based alloys).4 The overpack will be surrounded by a bentonite clay buffer. The standpoint in France, in 2010, was that all French SNF will be reprocessed. In Switzerland, concept studies (by NAGRA, Nationale Genossenschaft fu¨r die Lagerung Radioaktiver Abfa¨lle (Swiss National Cooperative for the Disposal of Radioactive Waste)) have been performed for both crystalline12 and clay options.13,14 An Opalinus clay formation in northern Switzerland has been chosen as the first priority option. The reference design concept for spent fuel canisters involves a cast carbon steel body, with a machined central square channel fitted with cross-plates to permit emplacement of

different types of fuel assemblies. Carbon steel was selected as a canister material because there is long industrial experience with its fabrication, it has high strength, and it has a relatively low and predictable corrosion rate in anoxic environments. When emplaced in the repository, the SNF containers would be surrounded by a bentonite clay backfill.4 In United Kingdom, the responsibility for radioactive waste disposal is with the Nuclear Decommissioning Authority. Until now, 2011, the working model is the ‘UK reference high-level waste, HLW/ SNF geological repository concept’ an adapted KBS3 concept in terms of canister length, diameter, and structure of the insert to handle HLW and spent fuel from the UK’s advanced gas cooled reactors and pressurized water reactor (PWR). The United States has started to construct a repository in the Yucca Mountain. This repository is built within a mountain ridge of volcanic tuff in the Nevada Desert. The rock in the ridge is unsaturated, that is, not all fractures and pores are filled with water. As a result, air can move freely through interconnected, dry fractures. This keeps the conditions in the repository oxidized. The spent fuel containers consist of an inner Type 316 stainless steel vessel and an outer highly corrosion-resistant cylindrical barrier composed of an nickel based-alloy (Alloy 22). Titanium drip shields have been constructed to keep the containers dry from dripping water. An option discussed is to pack the waste containers narrow to increase the surrounding rock temperature above 100  C to vaporize any water in the vicinity of the repository and make it migrate outwards.8,9 The Yucca Mountain project was, due to a political decision, halted in March 2010. However, the termination of the project has been questioned by a large number of congress members and a legal action regarding the withdrawal of its repository license application is to be resolved by the U.S. Court of Appeals and the Nuclear Regulatory Commission.

5.16.2 Fuel Evolution During Storage This section discusses the physical changes of SNF in terms of radioactive decay and temperature. The timeframe of these changes stretches beyond 10 000 years into the far future. The long-term behavior of the SNF will be dependent on its irradiation history. However, this chapter discusses the changes as they evolve for a general fuel, that is, a fuel without a history of extreme power excursions,

Spent Fuel as Waste Material

393

106 years. 239Pu is created by neutron capture reactions in 238U followed by two b-decays. The 239Pu itself is susceptible to further neutron capture that leads to the formation of 240Pu. Due to the ingrowth of 239Pu and 240 Pu during irradiation, plutonium is the element with the second highest concentration in the SNF. The concentration of elements such as Pb, Bi, Th, Pa, and Np increases with time since they are longlived, or stable, daughters in the decay chains of more short-lived heavier actinides. The helium concentration also increases with time as He is formed through all a-decays. Americium is, on this timescale, an intermediate element whose concentration during the initial 100 years increases due to ingrowth from the b-decaying 241Pu and after this period decreases slowly due to decay of 243Am (t1/2 7380 to 245Cm (t1/2 8502 years) but also due to the presence of smaller amounts of 243Am (t1/2 7380 years) in the spent fuel. The absence of (with cutoff at 1 mg/tHM) Po, At, Rn, Fr, and Ac is a consequence of the short half-life of the isotopes of these elements. Figure 2 shows the concentration of fission products in the same UO2 fuel as exemplified above, but on a shorter timescale.

ramping (described in Chapter 2.19, Fuel Performance of Light Water Reactors (Uranium Oxide and MOX)), or extreme burn-up etc. 5.16.2.1 Concentration of Radionuclides/Elements in Fuel In performance assessments of nuclear repositories, the boundary conditions are governed by the radiotoxicological impact of individually released radionuclides rather than by the amount of released elements. It is therefore important to understand the amounts, radioactivity, and spatial distribution of the radionuclides in order to make long-term predictions of the SNF behavior in a deep underground repository. 5.16.2.1.1 UO2 fuel

After irradiation, the UO2-fuel still contains around 95 wt% UO2. The remaining 5 wt% are mainly composed of heavier actinides (Np, Pu, Am, and Cm) and fission products. In Figure 1, the concentration of the selected elements in a typical spent UO2 fuel is shown for five different times after the end of irradiation. Due to the long half-life of 238U (4.699  109 years), the uranium concentration remains constant even after 1 000 000 100 000

Concentration (g per tHM)

10 000 1000 100 10 1 0.1 0.01 0.001 He

Tl

Pb Bi

Po

At

Discharge

Rn

Fr Ra Ac Th Elements 1

100

10 000

Pa

U

Np Pu Am Cm

1 000 000

Figure 1 Elemental composition of actinides and decay daughters in a UO2 fuel with an initial enrichment of 4 wt% 235 U irradiated for five cycles in a PWR to a burn-up of 40 GWd/tHM as a function of time (years) after discharge from reactor.

394

Spent Fuel as Waste Material

10 000 Discharge 1000

10 1000

Concentration (g per tHM)

100

10

1

0.1

I

Xe Cs Ba La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er

0.001

H He Li Be C Ge As Se Br Kr Rb Sr Y Zr Nb Mo Tc Ru Rh Pd Ag Cd In Sn Sb Te

0.01

Elements

Figure 2 Elemental composition of fission products in a UO2 fuel with a burn-up of 40 GWd/tHM at different cooling times (discharge from reactor, 10, and 1000 years).

The two ‘humps’ of the two fission fragments created after the initial 235U and later 239Pu fissions, in the element intervals Rb–Rh and Xe–Nd, are easily recognizable. The majority of the fission products decay by b-emission. The majority of fission products have, in comparison to actinides, relatively short half-lives (<10 years) that lead to smaller change in their concentration on the timescale 10– 1000 years. Nevertheless, some long-lived fission products (t1/2 > 104 years) can be seen, for example, 79 Se (3.27  105 years), 93Zr (1.53  106 years), 94Nb (2.03  104 years), 99Tc (2.11  105 years), 107Pd (6.5  106 years), 126Sn (2.3  105 years), 129I (1.57  107 years), and 135Cs (2.3  106 years). The drop in hydrogen concentration is related to the decay of 3H (tritium, t1/2 12.346 years). The decrease of strontium is due to the significant amount of short-lived 90 Sr (t1/2 28.503 years). Niobium has only one stable isotope, 93Nb, and this is shielded by the long-lived radionuclide 93Zr (t1/2 1.53  106 years). A shielded radionuclide is a radionuclide that is, on the isobar, preceded by a long-lived mother. As a result, the concentration of Nb, which initially is dominated by its short-lived isotope 95Nb (t1/2 35.2 days), decreases four orders of magnitude during the initial 10 years. As 93Zr decays, the concentration of 93Nb increases, which leads to a total increase of the niobium concentration. The decrease in the cesium concentration and increase

of barium are due to 137Cs (t1/2 30.104 years) decay into Ba, which within 153 s decays to stable 137Ba. The stepwise decrease in promethium concentration is a result of the single, relatively long-lived, Pm radionuclide 147Pm (t1/2 2.62 years). Based on Figures 1 and 2, it can be concluded that (i) after an initial period of 1000 years, only minor changes in the elemental concentrations occur in the spent UO2 fuel, and (ii) the changes occur at a concentration range, which is a fraction of the total mass of U. Accordingly, the chemical composition of the fuel will only change within a narrow concentration range on repository timescales. Despite that, the radioactivity emitted from the fuel changes orders of magnitudes over the repository timescale. Figure 3 shows the total number of a- and b-decays in the same fuel. a-particles can only be emitted by quantum tunneling out of the nucleus. Since the probabilities for tunneling are extremely small and decrease with the energy of the a-particles, low-energy a-emitters are generally much more long-lived than b(g)-emitters.15 For the SNF, this results in a gradual change in the radiation field. Fresh SNF emits a very intense b(g)-field due to the decay of short-lived radionuclides, such as 85 Kr, 89Sr, 90Sr–90Y, 91Y, 95Zr–95Nb, 106Ru–106Rh, 125 Sb, 134Cs, 137Cs–137mBa, 144Ce–144Pr, 147Pm, and 154,155 Eu. However, already after 300–500 years in the 137m

Spent Fuel as Waste Material

395

1.0E + 19 1.0E + 18

βtot

Radioactivity (Bq per tHM)

1.0E + 17 1.0E + 16

αtot 1.0E + 15 1.0E + 14 1.0E + 13 1.0E + 12

αcharge 1.0E + 11 1.0E + 10 0.001

0.01

0.1

1

10

100

1000

10 000

100 000 1 000 000

Time (years)

Figure 3 Radioactivity, divided in total number of a- and b-emissions, in UO2 fuel with a burn-up of 40 GWd/tHM as a function of cooling time. The lower line indicates the a-activity in the UO2 fuel before irradiation in the pressurized water reactor.

repository, most of the b(g)-emitters have decayed and a-radiation will dominate the energy deposition to the surrounding material. The tail of the total b-activity at cooling times longer than 1000 years is due to the long-lived fission products: 79Se, 93Zr, 94 Nb, 99Tc, 107Pd, 126Sn, 129I, 135Cs, and b-decay in the decay chains of 233,236,238U. The initial a-activity mainly originates from the decay of 242,244Cm, 238Pu, and 241Am, whereas the late decrease is caused by the decay of long-lived a-emitters such as 239,240Pu and 243 Am (see Figure 4). During the first years after irradiation, the decay of curium completely dominates the a-field. The sharp ingrowth of 241Am during the first 100 years originates in decay of 241Pu (t1/2 15.16 years) which is a daughter of the long-lived 245Cm (t1/2 8.5  103 years). In the time period between 2000 and 200 000 years, the a-field is exclusively dominated by the decay of 239Pu and 240Pu. The activity of the most intensive b-decaying radionuclides is given in Figures 5 and 6. The left part of the graph shows the radionuclides that mainly contribute to the b-activity at discharge from the reactor such as (in declining order) 140La–140Ba, 141Ce, 103Ru, 95 Zr–95Nb, 143Pr, 103mRh, 144Pr–144Ce, 91Y, 89Sr, and 106 Ru–106Rh. Many of these radionuclides have short half-lives and decay within the time period 1–5 years. The b-activity from decay of 137Cs–137mBa, 90Sr–90Y, 147 Pm, 134Cs, 106Ru–106Rh, 85Kr, 144Ce–144Pr, 154Eu, and 125 Sb dominates the fuel-cooling period, 5–40 years,

in a wet or dry cask storage. During the period 40–400 years, mainly the radionuclides 137Cs–137mBa, 90 Sr–90Y, 85Kr, 151Sm, and 154Eu contribute to the b-field. At longer time periods, >400 years, the main b-activity comes from the long-lived radionuclides 99Tc, 135Cs, 107Pd, 126Sb, 129I, and 79Se. Observing the intermediate- and the long-lived radionuclides in Figures 5 and 6 explains the shape of the total b-decay curve in Figure 3. 5.16.2.1.2 MOX fuel

MOX fuel is a blend of around 93 wt% natural, reprocessed, or depleted UO2 and PuO2, and is an alternative to low-enriched UO2 for use in light water reactors. It can be fabricated through different processes yielding slightly different fuel types and properties. The fabrication process and fuel properties, such as the microstructure and fission gas release (FGR), are given by Fisher et al.16 and also discussed in Chapter 2.15, Uranium Oxide and MOX Production and Chapter 2.20, Fission Product Chemistry in Oxide Fuels. Depending on the production technique, MOX fuel often obtains a duplex structure with the plutonium concentrated in Pu-rich agglomerates. This is, for example, the case for the optimized comilling process (OCOM) where Pu agglomerates up to 200 mm in size are dispersed in a matrix of natural UO2.17 The Pu content in the agglomerates often ranges up to 30 wt%. During irradiation to a normal burn-up,

396

Spent Fuel as Waste Material

1.0E + 16 233U 234U

1.0E + 15

236U

a-activity (Bq per tHM)

238U 237Np

1.0E + 14

238Pu 239Pu 240Pu

1.0E + 13

241Pu 242Pu 241Am

1.0E + 12

243Am 242Cm 243Cm

1.0E + 11

244Cm

U start

1.0E + 10 0.001

0.01

0.1

1

10

100

1000

10 000

100 000 1 000 000

Time (year)

Figure 4 The decay of actinides in UO2 fuel with a burn-up of 40 GWd/tHM as a function of cooling time. The dashed line, U start, indicates the a-activity in the UO2 before irradiation.

1.0E + 17

79Se 85Kr 87Rb

1.0E + 16

89Sr 90Sr

b-activity (Bq per tHM)

90Y 91Y

1.0E + 15

95Zr 95Nb 99Tc

1.0E + 14

103Ru 106Ru 103mRh

1.0E + 13

106Rh 107Pd 110Ag

1.0E + 12

110mAg 126Sn 125Sb

1.0E + 11 0.001

126Sb

0.01

0.1

1

10

100

Time (years)

1000

10 000

100 000 1 000 000

126mSb

Figure 5 The decay of lighter fission products (mass number 79–126, the heavier mass numbers are shown in Figure 6) in UO2 fuel with a burn-up of 40 GWd/tHM as a function of cooling time. In order to reduce the number of radionuclides from 613 to 45, only these radionuclides that were present at the end on the irradiation in a concentration >1  1016 Bq/tHM and still existed after 1 year are shown.

around 75% of the fissions occur in the Pu-rich agglomerates (calculation based on data from Walker et al.18). This creates an uneven fission product distribution in the fuel. The high local burn-up in the

agglomerates, up to 270 GWd/tHM,18 results in a microstructure of fine grains and large pores with a diameter of several microns,19 which is similar to the high burn-up structure (HBS) formed at the

Spent Fuel as Waste Material

397

1.0E + 17 125mTe 127Te 127mTe 129Te

1.0E + 16

129mTe 129I

b-activity (Bq per tHM)

134Cs

1.0E + 15

135Cs 137Cs 137mBa 140Ba

1.0E + 14

140La 141Ce 144Ce 143Pr

1.0E + 13

144Pr 147Pm 148Pm

1.0E + 12

148mPm 151Sm 154Eu

1.0E + 11 0.001

155Eu

0.01

0.1

1

10 100 Time (years)

1000

10 000

100 000 1 000 000

Figure 6 The decay of heavier fission products (mass number 125–155, complementary to Figure 5) in UO2 fuel with a burn-up of 40 GWd/tHM as a function of cooling time. In order to reduce the number of radionuclides from 613 to 43, only these radionuclides that were present at the end on the irradiation in a concentration >1  1016 Bq/tHM and still existed after 1 year are shown.

periphery (rim) of UO2 pellets with local burn-up above 60–80 GWd/tHM.17 The plutonium concentration in the MOX fuel is typically reduced by 25% during irradiation, as an example a 5.00 wt% Pu OCOM MOX fuel has been measured to contain 3.68 wt% Pu after irradiation.17 This heterogeneity is maintained during the irradiation and gives a spent fuel with different characteristics such as local intensive radiating agglomerates in a relatively low-radiating UO2 matrix, a subgrain disintegration (<1 mm), and different chemical composition (less uranium, larger fraction of fission products, plutonium, and higher actinides) compared with the UO2 matrix. 5.16.2.1.3 Comparison of UO2 and MOX

A simulation of MOX-fuel irradiation in a PWR using the same irradiation parameters as for the UO2 fuel in Section 5.16.2.1.1 shows that burned MOX fuel contains significantly more long-lived radionuclides such as 107Pd, 126Sn–126Sb, and 135Cs (see Table 2). On the other hand, significantly less long-lived 87Rb is formed in MOX than in UO2 fuel. This is a consequence of the slightly shifted fission

yield curve toward higher mass for fission of 239Pu in comparison to 235U. On shorter timescales, <1000 years, the dose contribution due to decay of 106Ru–106Rh, 125Sb, 151Sm, and 154,155Eu is more prominent in MOX fuel than in UO2 fuel, whereas 85Kr and 90Sr–90Y are found in greater quantities in the UO2. Based on Table 2, it can be concluded that recycled MOX fuel gives a moderately different composition of both long- and short-lived radionuclides, and that the difference reflects the mass shift in fission yield of the two fissile radionuclides. An increased a-activity after irradiation of MOX fuel in comparison to UO2 is clearly observable if Figure 7 is compared with Figure 4. The main reason is the build-up of larger amounts of heavier actinides (mainly Am and Cm) due to the initial presence of Pu in the MOX fuel. In general, the same a-emitting radioisotopes are created but at an elevated concentration in the MOX fuel. The consequence of the larger actinide content is a more intensive a-radiation field, which leads to a higher fuel temperature and for repository conditions, a more intense a-radiolysis

398

Spent Fuel as Waste Material

(see Section 5.16.4.3.4) upon contact with groundwater. A comparison between the two kinds of fuels, fresh and after irradiation at different cooling times, up to 100 000 years, shows that the total a-activity in the Table 2 Comparison of amount radioactivity produced in a MOX fuel compared to a UO2 fuel under the same irradiation conditions (PWR, 5.0 wt% recycled Pu, 40 GWd/ tHM, and five annual cycles) after 1 year of cooling time Radionuclides

(MOX–UO2)/UO2 (%)

85

41 46 50 70 128 82 88 58 112 60 130

Kr Rb 90 Sr–90Y 106 Ru–106Rh 107 Pd 125 Sb 126 Sn–126Sb 135 Cs 151 Sm 154 Eu 155 Eu 87

In order to reduce the number of radionuclides, only these radionuclides that were present with substantial activity after 1 year disclose a deviation of more than 40% and have a half-life longer than 1 year. Positive values show radionuclides produced in larger proportion in MOX than in UO2 fuel.

MOX is nearly 7000 times larger than the a-activity in the UO2 before being loaded into the reactor (Table 3). On the other hand, once the fuels are discharged from the reactor, the difference is only a factor of 6 having decreased by a factor of 1000. The difference in power output between the two fuel types remains almost constant with time. Ingrowth of long-lived radionuclides such as 233U, 236 U, and 237Np in the irradiated MOX fuel restrains the power output to decrease to levels comparable to those in unirradiated UO2, even on geological timescales. The higher power output leads to higher temperatures in the repository and as a consequence higher reaction kinetics of all repository-relevant geochemical reactions of minerals. Reactions that under low temperatures are kinetically hindered might become dominating under ‘hot’ repository conditions as temperatures in the range of 70–100  C are expected to prevail in most repositories over substantial timescales (Table 1). It can be concluded that the fission product composition after irradiation of MOX and UO2 fuels is similar. The plutonium content in the MOX will be nearly 30% lower after irradiation than initially introduced in the reactor. However, the spent MOX fuel

1.0E + 17 1.0E + 16 233U

1.0E + 15

234U 236U

α-activity (Bq per tHM)

1.0E + 14

238U 237Np 238Pu

1.0E + 13

239Pu 240Pu

1.0E + 12

241Pu 242Pu

1.0E + 11

241Am 243Am

1.0E + 10

242Cm 243Cm

1.0E + 09

244Cm

MOX start

1.0E + 08 1.0E + 07 0.001

0.01

0.1

1

10

100

1000

10 000

100 000 1 000 000

Time (years) Figure 7 The decay of actinides in MOX fuel with a burn-up of 40 GWd/tHM (PWR, five irradiation cycles) as a function of cooling time. The dashed line indicates the a-activity in unirradiated MOX fuel.

Spent Fuel as Waste Material

399

Table 3 Comparison of total a-activity in fresh UO2 and MOX fuel and after irradiation (PWR, five irradiation cycles) to a burn-up of 40 GWd/tHM and after different cooling times Time

a-activity in UO2 (Bq/tHM)

a-activity in MOX (Bq/tHM)

MOX/UO2

Charge into reactor Discharge from reactor Cooling time of 1000 years Cooling time of 10 000 years Cooling time of 100 000 years

8.6  1010 2.7  1015 7.0  1013 1.7  1013 4.0  1012

5.8  1014 1.7  1016 3.1  1014 6.6  1013 1.4  1012

6823 6.3 4.5 3.8 2.9

will contain considerably larger amounts of heavier actinides than a corresponding spent UO2 fuel. This will result in a hotter fuel, which needs longer time in the final repository to decay to background level. It should be noticed that when radionuclides decay other elements are created. As most of the fission products decay by b(g)-decay, they form an element with one proton number higher. For example, 137Cs decays to 137mBa which within 153 s decays to 137Ba, a stable nuclide. With this transformation of one element into another, the chemical composition changes which affects the chemical stability of the fuel. In the example given above, cesium is more soluble in a granitic groundwater than barium, which can form insoluble BaCO3. This situation is more accentuated when long-lived daughters are part of a decay chain, for example, the chain: 244Cm (t1/2 18.1 years)–240Pu (t1/2 6.537  103 years)–236U (t1/2 2.34  107 years)–232Th (t1/2 1.41  1010 years). This issue is further discussed in Section 5.16.4.

5.16.3 Mechanical Stability of Spent Nuclear Fuel After Irradiation This section discusses the changes of inherent properties affecting the mechanical stability of the SNF. Examples of such properties are He gas buildup, dislocations, grain disintegration, hardness, heat capacity, swelling, and microstructural changes of the spent fuel matrix. The changes are discussed with focus on the SNF alone, that is, without interaction with the surrounding environment. Additional information can be found in Chapter 1.03, RadiationInduced Effects on Microstructure; Chapter 1.05, Radiation-Induced Effects on Material Properties of Ceramics (Mechanical and Dimensional); Chapter 1.06, The Effects of Helium in Irradiated Structural Alloys; Chapter 2.17, Thermal Properties of Irradiated UO2 and MOX; Chapter 2.18, Radiation Effects in UO2; and Chapter 2.20, Fission Product Chemistry in Oxide Fuels.

5.16.3.1 Evaluation of the Temperature and Thermal Conductivity The temperature in a granitic spent fuel repository will be considerably above the temperature in undisturbed bedrock (15  C at 500 m depth).20 The increased bedrock temperature is a consequence of radioactive decay of radionuclides in the fuel (Figure 8) and heat transfer through container walls. In the figure, it can be observed that both fuel types release about the same amount of thermal power right after removal from the reactor since they contain roughly the same amount of b-activity and as the b-activity dominates over a-activity by three orders of magnitude (Figure 3), despite the fact that an average a-decay emits more or less one order of magnitude more energy per decay than b-decay. With time, MOX fuel gives of three to four times more thermal power due to its higher content of actinides, hence a-decaying elements (Table 3). The spatial distribution of the canisters in a granitic repository is laid out so that the temperature in the outer bentonite layer, surrounding a container, is lower than
90  C (Table 1). The thermal conductivity of oxide fuels deteriorates very rapidly at the beginning of irradiation, mostly due to formation of radiation damage, and then with increased burn-up by the formation of a solid solution of UO2 with the oxides of Zr, Sr, the rare earths, and other fission products in lower concentrations. This effect is counterbalanced, but not quenched, by the production of metallic fission products inclusions precipitated within the UO2 and grain boundaries.21 At discharge of a LWR fuel with a burn-up of 40 GWd/tHM and a temperature of 500 K, the thermal conductivity has decreased by
50%, from 5.8 W m1 K1 as in fresh fuel to 3.0 W m1 K1.22 The lower the temperature of the fuel, the higher the degradation of the thermal conductivity. However, the total thermal power decreases with time, and in wet storage conditions, the fuel temperature will not

400

Spent Fuel as Waste Material

1.0E + 07 MOX MOX before irradiation

1.0E + 06

UO2 fuel UO2 before irradiation

Thermal power, Ac + FP (W per tHM)

1.0E + 05

1.0E + 04

1.0E + 03

1.0E + 02

1.0E + 01

1.0E + 00 1.0E - 01 1.0E - 02 0.001

0.01

0.1

1

10

100

1000

10 000

100 000 1 000 000

Time (years) Figure 8 The thermal power production in irradiated UO2 and MOX fuel with a similar burn-up of 40 GWd/tHM (PWR, five irradiation cycles) as a function of cooling time. The points indicate the thermal power in unirradiated MOX and UO2.

exceed 200  C (which would also be the limit in dry conditions after a few decades23). 5.16.3.2

Structural Stability of the Fuel

The stability of the fuel in the repository, as long it is not exposed to water, is mainly governed by its microstructural and chemical evolution due to radiation damage, helium build-up, and oxygen potential variation. In the event of exposure of SNF to groundwater in a final nuclear fuel repository, the possible preferential dissolution of grain boundaries rather than matrix – a process driven by the penetration rate of water – could cause a rapid increase of the fuel wet surface area, and therefore increase the fraction of inventory becoming available for fast release. While this has a direct impact on the mobilization of radionuclides, the water penetration, additionally, weakens the cohesion between UO2 grains which leads to degradation of the overall mechanical stability of the fuel pellet. Furthermore, an increase of the wetted surface area increases the amount of oxidants produced through a-radiolysis.24

5.16.3.2.1 Radiation damage

Defects created by displacement cascades are extensively recombined in UO2 and PuO2; hence, no matrix amorphization could be detected in those materials.25,26 Nevertheless, lattice parameter increase caused by a-radiation damage that displaces atoms in the lattice, and leading to microscopic swelling was measured. In polycrystalline UO2, the lattice parameter reaches saturation (
0.4%, which corresponds to 1% volume increase) for a dose of
1 dpa (displacements per atom) because of recombination of defects created by the displacement cascades.27 One of the more spectacular evolutions during the irradiation in reactor of the light water reactor, LWR oxide fuels is the eventual formation of the high burn-up structure, HBS. In the periphery region of the pellet, plutonium builds up more than in the central region because of the neutron resonant capture cross-section of 238U; as a consequence, more plutonium burns up in the ‘rim’ region. Finally, the local burn-up at the rim is a few times higher than at the center of the pellet. Therefore, the ‘rim’ structure is associated with this higher burn-up. This restructuring is characterized by the existence of highly

Spent Fuel as Waste Material

dense small subgrains whose size is
200 nm and the accumulation of small pores with an average size of around 1 mm.28 This restructuring could influence the fuel performance including, for example, fission gas release (FGR), fuel temperature, hardness, and swelling during irradiation, and moreover, would be the first exposed layer in case of water contacting. The specificity of the HBS region has therefore to be taken into account in any failure scenario. For more detailed information on the HBS region, see Chapter 1.03, Radiation-Induced Effects on Microstructure and Chapter 1.05, Radiation-Induced Effects on Material Properties of Ceramics (Mechanical and Dimensional). 5.16.3.2.2 Helium built-up

Helium is continuously produced in the spent fuel due to a-decay of actinides. An example of the accumulated amount of He produced in a UO2 and MOX fuel rod during storage is shown in Figure 9. The initial content of 3–5 wt% Pu in MOX fuel gives a higher He accumulation in the fuel rod. In storage condition, it is of prime importance to assess whether the fuel can retain the radiogenic helium in order to predict its long-term mechanical stability (see also Chapter 1.06, The Effects of Helium in Irradiated Structural Alloys). Indeed, accumulation of gas which ultimately creates helium bubble and of a-damage defects, evidenced by lattice

parameter increase, might cause swelling of the fuel and therefore its early failure. Interactions between the radiogenic helium formation and the defects induced by a-particles and recoil nucleus need to be understood to predict the evolution of the system. The helium behavior becomes even more relevant for the MOX fuel because of the higher production of this rare gas due to the higher actinide content. Two major parameters have to be considered as regarding the helium behavior: its solubility and diffusion coefficient. These two parameters need to be evaluated from a thermodynamical point of view and extrapolated to their ‘apparent’ values (considering the fuel as a nonperfect crystal, that is, containing defects, hence trapping sites for the gas). In an early work, Rufeh et al.29 have determined a He solubility in UO2 of 6.7  104 cm3 (STP) gUO2 1 atm1. The solubility in a UO2 monocrystal is approximately one order of magnitude lower (2.4  105 cm3 (STP) gUO2 1 atm1).30 Recent work have been carried out to determine the solubility in conditions of high pressure confirming that the Henry law applies for the helium solubility in UO2. The reported value for the solubility is 5.1  104  104  105 cm3 gUO2 1 for samples infused at a pressure of 1 kbar and a temperature of 1570 K.31 Work is ongoing to assess the solubility in nonstoichiometric UO2.

12 000 UO2 MOX Helium production (cm3 STP/rod)

10 000

8000

6000

4000

2000

0 1

10

100

401

1000 Time (years)

10 000

100 000

1 000 000

Figure 9 Accumulated volume of helium produced through a-decay of actinides in UO2 and MOX, inside a PWR fuel rod, with a burn-up of 40 GWd/tHM (PWR, five irradiation cycles) as a function of cooling time.

402

Spent Fuel as Waste Material

The higher dissolution of He in polycrystalline UO2 in comparison to single crystal is attributed to the larger amount of defects, particularly vacancies, able to accumulate He. Through thermal and nonthermal diffusion, He accumulates in gas bubbles inside the matrix and at grain boundaries. A part of the He is released to the free volume of the fuel rod with the consequence of increased internal rod pressure and thereby risk for cladding creep failure. The accumulation of He bubbles, especially at grain boundaries, can eventually result in loss of cohesion of the grains.32 5.16.3.2.3 Oxidation of the fuel

The chemical composition change due to decay has to be accounted for, for its possible implication on the oxygen potential change of the fuel. The oxygen content in a UO2 fuel after discharge from reactor is, in general, more or less stoichiometric, the O/U ratio is very close to 2. This ratio has a major impact on the radionuclide mobility in the fuel as any deviation from stoichiometry increases the diffusion coefficient of cation vacancies (O/U > 2) or anion vacancies (O/U < 2) in the matrix. Additionally, the thermal conductivity decreases when the stoichiometry changes from 2.00. Once the fuel is removed from the reactor, it exhibits intensive b-decay during the initial 300 years (Figure 3) dominated by the b-decay of 137Cs ! 137mBa, 90 Sr ! 90Zr, and 241Pu ! 241Am. The b-decay results in formation of elements of different valence state than the parent element and could eventually change the oxidation state of the spent fuel by modifying the overall oxygen ratio. Nevertheless, measurements of high burn-up UO2 fuel33 showed that the oxygen potential in the fuel appears to be controlled by molybdenum (Mo/ MoO2), an abundant fission product in spent fuel (see Figure 2). Even though the composition of the spent fuel and the temperature change during the initial 300 years of storage, the oxygen potential seems to be stabilized at approximately 450 kJ mol1, independently of burn-up, and is controlled by the oxidation of Mo.34 5.16.3.2.4 Fuel cladding interaction/ mechanical properties

Once the fuel is out-of-pile and the temperature has decreased below 800 K, the fuel-cladding (Zircaloy) interaction of fission products (Cs, Te, and I) is minimal.21 Due to the a-damage, the mechanical properties of the fuel might evolve depending on

the temperature and recovery of the defect or their evolution toward extended defects. The associated swelling due to the damage and gas production could also cause an increase in the stresses in the spent fuel. The evolution of the hardness of UO2 due to a-damage has been studied by monitoring the Vickers hardness evolution as function of the damage of UO2 samples doped with intensive a-emitters (238Pu). Two types of UO2 samples have been used, doped either with 8.8 wt% 238Pu ((U0.9Pu0.1)O2) or with
0.088 wt% 238Pu (U0.999Pu0.001)O2). Figure 10 shows the results of hardness measurements as a function of time for the two a-doped materials. Each data point represents the average of five indentations. The values measured for (U0.9Pu0.1)O2 increased quite rapidly, and after
2 months appeared to have reached a saturation value, corresponding to an increase of
20% in hardness. Even after 2 years of storage the values for (U0.999Pu0.001)O2 show a clear augmenting trend, with an overall increase of
13%, or
1.6  104 kg mm2 d1 in hardness. A comparison with the initial rate of increase of the hardness for (U0.9Pu0.1)O2, which can be roughly estimated as
1.2  102 kg mm2 d1, is consistent with the 100-fold difference in activity between the two a-doped materials.35 The a-damage evidently diminishes the yield strength (elastic limit), which in case of strong swelling could also contribute to early failure of the fuel. Additional information is given in Chapter 3.16, Ceramic Fuel–Cladding Interaction.

5.16.4 Chemical Stability of Spent Nuclear Fuel After Irradiation Knowledge of the relative amount, distribution, and nature of each element created in the fuel during reactor operation is important for the understanding of the long-term behavior of the fuel. The distribution and chemical form are of special interest when studying the corrosion behavior of the fuel. 5.16.4.1 Distribution and Chemistry of the Fission Products Figure 11 shows a schematic of an irradiated fuel pellet. The different regions marked in the figure indicate locations with potential fission product accumulation.

Spent Fuel as Waste Material

950

403

(U0.9,238Pu0.1)O2 (U0.999,238Pu0.001)O2

Hardness (kg mm-2)

900

850

800

750 Measurement time ~ 3000 d

700 10-4

10-3

10-2

10-1

100

Displacements per atom (dpa) Figure 10 The evolution of the hardness is directly related to the a-damage rather than to the composition in the case of the a-doped material. Courtesy of Thierry Wiss.

Gap

Crack

3. oxide precipitates: Rb, Cs, Sr, Ba, Zr, Nb, Mo, and Te; 4. dissolved as oxides in the UO2 matrix: Sr, Ra, Zr, Nb, Y, Te, Cs, Ba, La, Ce, Pr, Nd, Pm, Sm, Eu, and the actinides.

Grain

Grain boundary Cladding

Pellet gap

Figure 11 Schematic illustration of an irradiated fuel pellet with cladding. The fuel pellet cracks during irradiation due to temperature gradients. Volatile fission products accumulate in the cracks and in the gap between the fuel and the cladding. Metallic and oxide precipitates are formed at the grain boundaries and in the grains of the fuel. Courtesy of Patrik Fors.

Kleykamp36 divided the fission products and actinides into four classes depending on their chemical state in the fuel and the list was complemented with more elements by Ferry et al.34 1. volatile and gaseous fission products: He, Kr, Xe, Br, Rb, Cs, Te, and I; 2. metallic precipitates: Mo, Tc, Ru, Rh, Pd, Ag, Cd, In, Sn, Sb, and Te;

The abundance of the major fission products for two 45 GWd/tHM fuels, one UO2 and one MOX, is listed in Table 4. The elements are divided into the four groups given by Kleykamp.36 The first group will be found distributed between: (i) fission-induced solution within the oxide lattice, (ii) closed intragranular and intergranular bubbles within the fuel,37 and (iii) in the open porosity of the fuel and the gap between the fuel and cladding.21 This group of elements is expected to dissolve quickly in case of fuel contact with groundwater. The approximate rare gas distribution in a UO2 fuel rod is shown in Table 5. The second group forms metallic precipitates in the fuel, usually referred to as e-phase particles. In these particles, alloys are formed whose composition is dominated by molybdenum and ruthenium.38 The e-particles are generally accumulated at the grain boundaries. The e-particles are, in performance assessments, assumed to become dissolved in groundwater. However, as pointed out by Shoesmith39 this assumption is conservative as most of the grain

404

Spent Fuel as Waste Material

235 Table 4 U and 239Pu fission product creation rates in 45 GWd/tHM burn-up PWR UO2 and MOX19

Table 5 Distribution of the rare gases (Kr, Xe) in UO2 fuel rod with a burn-up of 60 GWd/tHM34

Typea

Element

PWR UO2 (wt%)

PWR MOX (wt%)

Gas fraction in free volume

Rare gases and volatiles

Xe Kr I

12.7 1.1 0.6

13.3 0.6 0.8

Ru Pd Rh Tc Ag Cd Sn

6.9 3.6 1.2 2.3 0.3 0.2 0.2

8.9 7.3 2.2 2.4 0.6 0.5 0.2

Mo Te

9.6 1.4

9.0 1.6

Cs Ba Rb

11.0 4.4 1.0

11.4 4.2 0.5

Zr

10.4

7.4

Y Ce Nd La Pr Sr Pm Sm Eu Gd

1.4 7.7 11.1 3.5 3.2 2.6 0.4 2.0 0.5 0.2

0.7 6.9 9.8 3.4 3.0 1.3 0.5 2.4 0.7 0.3

Metallic precipitates

Oxide precipitates

In solution in the oxide

The initial 235U enrichment in the UO2 was 4.25 wt%, the Pu content in the MOX was 7.8 wt%.19 a The grouping of elements is made according to Kleykamp36 and not as originally given by Bailly et al.19

boundaries are not immediately accessible to water, and the dissolution, therefore, to a significant degree could be controlled by the corrosion/dissolution of the fuel matrix. A part of the elements in group 3 are separated from the fuel matrix into cubic perovskitestructured precipitates, sometimes referred to as gray phases.21,40 This precipitate is a solid solution with the general composition ABO3 with Ba, Sr, and Cs in the A sites and Zr, Mo, U, Pu, and rare earths in the B sites.21,41 The gray phase is found to be distributed in the entire fuel, but accumulates in the grain boundaries of the colder parts of the fuel.21,42 The relative partition of molybdenum between metallic and oxide precipitates depends on the oxygen to metal ratio, and the chemical potential of oxygen, in the fuel.40,43 As the oxygen potential increases with the fuel burn-up, the amount of oxidized molybdenum

Intergranular gas fraction

0.03 0.019–0.042 in central part (r < 0.5r0) 0.014–0.048 intermediated part (0.5r0 < r < 0.975r0) 0.045–0.062 in the rim pores (0.975r0 < r)

Total intergranular gas fraction


0.08–0.15

Fraction in the grains


0.80–0.90

increases and, as a result, the molybdenum fraction in the metallic precipitates decreases.40 Table 6 shows an example of a metallic precipitate composition in a UO2 fuel.44 The elements in group 4 form solid solution with the UO2 matrix, and can only be released by dissolution of the UO2 grains themselves. Some elements such as Cs, Sr, Ba, Ra, Zr, Nb, and Mo can exist in more than one phase. The UO2 matrix encompasses the major part of the elements in solid solution, but it also contains almost all gaseous and volatile fission products contained in gas bubbles.37 The metallic precipitates exist in the UO2 matrix, as inclusions, as well as at the grain boundaries.34 The oxide precipitates are present either as individual inclusions or, to a minor degree, as soluble oxides in the UO2 matrix. It is estimated that about 30% of the fission products form solid solutions with the matrix.45 However, due to the limited migration of the nonsoluble elements (groups 1–3), about 95% of all fission products are present in the oxide lattice46 and can be released only by dissolution of the latter. Consequently, the dissolution of UO2 will be the major factor in determining the release rate of radionuclides from the repository, in case of a canister failure. 5.16.4.2

Storage Before Final Disposal

The chemical behavior of the fuel in the repository depends to some degree on the intermediate storage conditions during the time the fuel is cooled and stored in the final repository. Nowadays, two main intermediate storage strategies are employed, wet and dry storage. The release rate of fission products from the spent fuel in the repository depends to some degree on the previous handling of the fuel. Of primary importance is

Spent Fuel as Waste Material

Table 6

405

The relative composition of elements in e-particles together with the theoretical fission yields44

Measurement (fraction) Fission yield (fraction)

Mo

Ru

Tc

Pd

Rh

Te

0.327 0.429

0.405 0.260

0.070 0.101

0.117 0.109

0.042 0.059

0.038 0.043

Data from a UO2 fuel with a burn-up of 23 GWd/tHM (irradiated at a linear power effect of 20 kW m1).

to avoid oxidation of the fuel before it is encapsulated. This requires that the spent fuel is kept in an airtight container in the Zircaloy cladding during the complete intermediate cooling period (see also Chapter 5.17, Waste Containers). 5.16.4.2.1 Wet storage of fuel in cooling ponds

The experience of more than 40 years of wet storage of several million light water reactor rods shows that SNF with zirconium alloy cladding has had excellent durability.47 Investigations made on CANDU fuel rods as well as on PWR Zircaloy-cladded rods showed no detectable changes in the cladding characteristics after 20–30 years of wet storage. These investigations were supported with destructive and nondestructive examinations of fuel rods, visual evidence, and coupon studies all showing resistance to aqueous corrosion.47 Additionally, there have been no reports of FGRs, indicative of cladding failure in wet storage. Given the fact that the Zircaloy corrosion rates in wet storage are very low, indicative techniques were applied and yielded a suggested corrosion rate of <0.007 mm year1.47 5.16.4.2.2 Dry storage of fuel

SNF with zirconium alloy cladding has been placed into dry storage in approximately a dozen countries.47 The resistance of Zircaloy cladding to corrosion in water and in water vapor, the absence of a hydrated zirconium oxide, and the low oxidation rate of Zircaloy cladding in air mean that there are few chemical problems in drying Zircaloy clad fuel, or indeed in storing incompletely dried unfailed fuel. For power reactor fuel, the decay heat promotes drying when the fuel is removed from water. Due to the high decay heat of the SNF, the cladding temperatures in dry storage needed to be limited by design. Acceptable cladding temperature limits tend to be based on mechanical properties such as creep, but cladding strength can be degraded by internal fission product attack, especially at burn-ups above 45 GWd/tHM. The technical basis for satisfactory dry storage of fuel clad with zirconium alloys includes hot cell tests

on single rods, whole assembly tests, demonstrations using casks loaded with irradiated fuel assemblies, and theoretical analysis. The principal method of monitoring cladding behavior has been to periodically sample cover gases for evidence of 85Kr since its presence offers an indication that cladding has been breached. The technique is used in the assembly and in cask storage tests and demonstrations. The fact that only a few rod failures have been detected provides evidence that cladding integrity is satisfactory in dry storage. The CASTOR V/21 dry storage demonstration referred to above involved visual, nondestructive, and destructive examination of PWR Zircaloy clad fuel rods after 14 years in dry storage. There was no indication of degradation of the cladding or fuel and little or no FGR from fuel pellets during the period of dry storage. The cladding retained significant creep ductility after dry storage.47 5.16.4.3

Repository Storage

Most concepts for SNF repositories are designed to obtain oxygen-free (more commonly expressed as ‘reducing’) conditions within a few years after closure. The oxygen-free conditions will be achieved since any oxidants remaining in the repository at the time of its closure will be consumed by microbiological processes and reducing minerals.48,49 This section focuses on the mechanisms governing the spent fuel behavior under oxygen-free conditions and takes the Swedish KBS-3 concept as an example (see Figure 12). The KBS-3 concept has been chosen to illustrate the chemistry of the repository since it is well examined and close to implementation. Generally, the same reactions will be applicable in other repository types with oxygen-free conditions; however, differences in groundwater composition will be seen if the repository is placed in a clay or salt formation. 5.16.4.3.1 Groundwater composition

The water that will be in contact with the container will govern the corrosion processes of the

406

Spent Fuel as Waste Material

Groundwater in bedrock is oxygen-free

Oxygen-free H2O reacts with canister iron to produce hydrogen 3Fe + 4H2O ® Fe3O4 + 4H2

a-radiation splits the groundwater in a process known as radiolysis Iron canisters containing the spent fuel are emplaced in the repository boreholes

a

H2O ® H2O2, H2, and O2

Figure 12 Granitic repository for spent nuclear fuel.50 Geologic repositories contain spent nuclear fuel enclosed in massive copper/iron or steel/cast-iron containers 500–1000 m deep in the bedrock. The containers prevent water from interacting with the spent fuel for at least 1000 years. By the time spent fuel may become exposed to groundwater, b(g)-emitting nuclides in the fuel will have decayed to a negligible level.51 Therefore, only long-lived a-emitters will be responsible for the formation of oxidants at the fuel–water interface through radiolysis (for simplicity, the radiolytic reaction presented in the figure is unbalanced; a complete set of reactions can be found in Pastina and LaVerne52). Anoxic corrosion of iron will form large amounts of hydrogen, a reductant.51 The graphics to the left is used with permission from SKB, illustrator: Jan Rojmar.

container itself, during the initial time period, as well as the cladding and spent fuel at the time the canister fails. The composition of the natural groundwater moving through the repository will change as it passes the different barriers since the barriers contain foreign material, in respect to mass and composition, in comparison to the bedrock. An example of different groundwater compositions in respect to repository type is given in Table 7. It should be remembered that the natural groundwater composition will change over the long period of an ‘active’ repository, especially due to the processes occurring during glaciations and deglaciations.57 The groundwater compositions shown in Table 7 are selected as ‘representative’ groundwater at repository depth. This means that there exists a variation in groundwater composition, at each repository site, that depends on the fracture orientations and fracture pathways. The chemical composition of ‘old’ (>5000 years, small exchange/mixing with surface waters) groundwater, shown by its 18O and 3H content, depends on the composition of the fracture filling minerals it has exchanged ions with. Geochemical modeling of groundwater composition changes as a function of fracture filling minerals composition increases our understanding of groundwater behavior and can be extrapolated to changes in future

groundwater compositions due to water mixing as a result of changed groundwater flow paths, for example, during glaciations and deglaciations.54 Groundwater compositions from two suggested repository sites located in Sweden, Forsmark (selected as first choice) and A¨spo¨ which are
400 km apart can be compared in Table 7. Although the difference in composition of cations such as Na, K, Ca, and Mg are varying by factor of 6, these differences mirror different fracture filling mineral compositions. Since the solubility of the different ionic complexes will determine when saturation is reached, the rate of spent fuel dissolution will depend on the composition and concentration of ions in the groundwater and their respective ability to form ionic complexes. Alkali metal cations, in millimolar range, together with high concentration (in micrometer range) of dissolved U(VI) can form precipitates such as sodium diuranate, Na2O (UO3)26H2O. Alkaline earth metals (Ca, Mg, and Sr) together with Si and U(VI) form U(VI) silicates such as uranophane, Ca(UO2)2(SiO3OH)25H2O. Uranophane and soddyite ((UO2)2(SiO4)(H2O)2) are uranium (VI)-containing silicate minerals which have a relatively low solubility and in presence of silicates can keep the U(VI) concentration below 106 M.58 Coffinite, USiO4(cr), a uranium(IV) silicate mineral

Spent Fuel as Waste Material

407

Table 7 Representative groundwater compositions in crystalline bedrock (Forsmark, borehole KFM02A, 509–516 m depth), A¨spo¨ crystalline bedrock groundwater (borehole KAS02, 530-m depth), A¨spo¨ crystalline bedrock groundwater after contact with bentonite (MX-80), Opalinus clay pore water, and salt (Gorleben inclusion brine) Components

Granitic groundwater at Forsmark, Sweden53 (mol l1)

Granitic groundwater at A¨spo¨, Sweden54 (mol kg1)

Granitic groundwater from A¨spo¨ equilibrated with bentonite55 (mol l1)

Opalinus clay pore water, Switzerland13 (mol l1)

Brines at Gorleben, Germany56 (mol per kg H2O)

[Naþ]tot [Kþ]tot [Ca2þ]tot [Mg2þ]tot [Sr2þ]tot [Fe]tot [Mn]tot [Si]tot [F]tot [Cl]tot [Br]tot [HCO 3] [SO2 4 ] [S2] [S]tot T( C) log pCO2 pH Eh against SHE (V)

8.9  102 8.8  104 2.3  102 9.3  103

9.1  102 2.1  104 9.6  103 1.7  103 4.0  104 4.4  106 5.3  106 1.5  104 7.9  105 1.8  101 5.0  104 1.6  104 5.8  103

6.4  101 3.1  103 1.0  102 4.9  103 2.5  104

1.7  101 5.7  103 1.1  102 7.5  103 3.0  104 4.3  105 2.4  105 1.8  104 1.7  104 1.6  101 2.4  104 2.7  103b 2.4  102 1.4  1011

15

20 3.5 8.4 Oxidizing

0.060 0.016 0.051 5.70 0.0014 0.0206 0.0003 n.g. 0.0163 11.35 n.g. n.g. 0.0011 n.g. 0.0011 n.g. n.g. 7.0 n.g.

3.3  105 1.8  104 1.5  101 3.0  104 1.8  103 6.8  103 7.0 0.143

8.1 0.258

3.2  104 6.4  101 2.5  103a 1.6  102

25 2.2 7.24 0.167

n.g.: Not given in the reference. a [C]tot. b [CO3]tot.

is mainly found in association with UO2(cr) and SiO2(cr). Its solubility, at high silica concentrations, is considered to be even lower than for UO2(cr).59 Iron and manganese are normally present in their reduced form and control the redox potential, Eh, in the groundwater. Of the anions, Cl and Br form relatively weak complexes with the cations and U(VI) in comparison to hydroxide (OH) and carbonate   (CO2 3 ). On the other hand, at higher Cl and Br concentrations (>1 mM), they take part in radiolysis reaction scheme through OH scavenging, resulting in decreased production of H and thereby creating a more oxidizing condition.60 Calcite (CaCO3) and carbonate-containing minerals in the bedrock control the pH of the groundwater. The change of the water composition when contacted with bentonite (MX-80) is also shown in Table 7. The main change occurs due to the dissolution of impurities of calcite present in bentonite, thereby increasing the carbonate concentration and decreasing pH.61 Presence of minor amounts of CaSO4 in the bentonite gives an almost tenfold increase of the sulfate concentration.

Due to its marine origin, the pore water of the Opalinus clay is relatively saline and sodium chloride dominated, while the pH is close to neutral (Table 7). Redox conditions are reducing as evidenced by the large amounts of unoxidized pyrite and siderite.13 A comparison of the granitic groundwater from A¨spo¨ and the pore water from the Opalinus clay shows significant differences mainly in Na, K, Fe, Mn, and carbonate. While Na is twice as high in Opalinus clay pore water compared to A¨spo¨ GW, K is almost 30 higher because of the high cation exchange capacity of the layered clay. Comparing the redox potential in A¨spo¨ groundwater with Opalinus clay pore water shows a higher redox potential despite the higher concentration of total Fe and Mn in Opalinus clay pore water, exemplifying the importance of determining the oxidation state of dissolved Fe, Mn, and S. For the case where the SNF repository is situated in Opalinus Clay, as planned in Switzerland (Table 1), the canisters will be exposed to bentonite pore water that is expected to have a composition similar to that of the Opalinus clay pore water, with minor differences in composition due to ion-exchange processes with the bentonite clay.6

408

Spent Fuel as Waste Material

Granitic or argillaceous host rocks contain formation water in fractures or in pores of the rock matrix. In contrast to these groundwater bearing rock formations, salt rock formations are free of fresh flowing water and relatively impermeable. The Gorleben salt dome in Germany has been studied as a potential site for the disposal of SNF. Brine reservoirs found in the Gorleben complex having so far occurred in the explored parts of the salt dome had a limited volume and do not have any contact with the groundwater. Reservoir tests resulted in brine occurrences in the order of magnitude of maximum some thousand cubic meters.62–64 The composition of Gorleben inclusion brine, which was characterized and used for corrosion experiments with SNF,56 is given in Table 7. Considering the reducing conditions, it can be concluded that the groundwater compositions in Tables 1 and 7 show groundwaters at repository depth with a redox potential, Eh, below zero (SHE). 5.16.4.3.2 Corrosion of copper

The slow corrosion of the copper lining, surrounding the iron insert in the repository, plays a crucial role giving the intensively radioactive fission products time to decay (see Figure 3). Corrosion experiments on copper coupons in contact with bentonite have been performed under reducing conditions in the A¨spo¨ hard rock laboratory. The result showed an average corrosion rate of <0.5–3 mm year1.65 No pit corrosion could be recognized. The corrosion product identified was cuprite, Cu2O, and paratacamite, Cu2(OH)3Cl. The low corrosion rate was verified by real-time corrosion experiments performed at the same location, using copper electrodes inserted into bentonite. The results of these experiments showed corrosion rates in the range of 0.1–2.4 mm year1. It was concluded that general copper corrosion is not a problem either from a structural integrity point of view or from any risk of impairing the properties of the bentonite.65 Assuming an average copper corrosion rate of 3 mm year1 and a 50-mm-thick copper wall, a copper lining penetration will occur after 17 000 years. At this time, mainly long-lived, a-emitting actinides will dominate the radiation field (see Figure 3). A more detailed discussion on this subject can be found in Chapter 5.17, Waste Containers. 5.16.4.3.3 Iron canister corrosion

The oxygen trapped in the repository after its closure will have had been consumed by bacteria and

reducing minerals, a process already finished within a few years of closure48,49 and leading to an anoxic deep groundwater. If the copper lining is penetrated due to corrosion, after >10 000 years, and water comes into contact with the cast-iron canister, anaerobic corrosion of the iron will start: Fe þ 2H2 O ! FeðOHÞ2 þ H2 ðgÞ

½1

3Fe þ 4H2 O ! Fe3 O4 þ 4H2 ðgÞ

½2

These processes will produce large amounts of Fe2þ and H2 since the reverse reactions occur first at H2 pressures in the order of 100 MPa.66 Experiments have shown that the reactions are not limited by dissolved ferrous ions.67 The anoxic corrosion rate of the iron in the canister is very low, starting in the order of 10–30 mm year1, but decreases to 0.1 mm year1, as soon as an oxide film is formed on the iron surface.67 The amount of hydrogen present in the damaged canister will initially be determined by the rate of hydrogen production and the transport of dissolved H2 away from the production site. Since the transport of dissolved gas is limited through the capillaries in the bentonite clay buffer, H2 will accumulate in the canister. After a few years, the pressure of H2 becomes higher than the hydrostatic pressure at the depth of the repository (i.e., 5 MPa at 500-m depth) and a gas phase will start to form. The pressure inside the gas phase will continue to increase until it is higher than the hydrostatic pressure plus the swell pressure of the clay (in a repository at 500-m depth, this is expected to be around 10–15 MPa). Once the swell pressure is exceeded, a breakthrough of gas will occur that brings the pressure in the gas phase back to the hydrostatic pressure of the repository. The bentonite will recover totally after this breakthrough, and the pressure increase will start again.68 5.16.4.3.4 Radiolysis

Although the redox conditions in deep granitic waters are reducing (200 to 350 mV(SHE) at pH 7–8),65 oxidizing conditions might be created near the fuel surface as a result of the fuel’s radioactive decay (maximal range of a 5 MeV a-particle in pure water is 37 mm). The energy released when the radioactive particles interact with the groundwater leads to splitting, or radiolysis, of the water molecules (see Figure 13). Generally, equimolar amounts of oxidizing and reducing species are created. However, due to the much higher reactivity of the oxidizing species, an oxidative environment is expected to result from the

Spent Fuel as Waste Material

409

Ionizing radiation (α, β, γ) H2O H2O+

e–

H2O

H2O H 2O

H+ H2O

•

OH

OH– H•

OH• H2

H2O2 + e– + H+ HO2•

H2O

HO2•

O2

Figure 13 Simplified radiolysis scheme. Radiation interacts with the electron structure in the water molecule and deposits enough energy to ionize one of the atoms. This results in the release of an electron. The positively charged water thereby created decomposes into a hydrogen ion and a hydroxide radical; the free electron continues to interact with water molecules until its kinetic energy is low enough for the electron to be absorbed by one of the water molecules. The resulting negatively charged water decomposes into a hydroxide ion and a hydrogen radical. Hence, the sum reaction is that a water molecule is split into one hydrogen radical and one hydroxide radical. The reactive radicals, the water, and the ions react further with each other in multiple steps according to a scheme of nearly 80 reactions.52,69 Some of the most important are shown. The overall result of the radiolysis is the formation of new molecular species: H2O2, H2, and O2 (produced through the decomposition of H2O2 on UO2 surface with a yield of 0.267). Courtesy of Patrik Fors.

radiolysis.51 The radiolytic yield of the interaction process varies with the different types of emission. b(g)-radiation leads to sparse ionizations and, therefore, low probability for reaction between different radicals and a radical-rich yield. a-radiation, on the other hand, due to the a-particles’ larger size, leads to a dense track of ionizations, in which combination reactions between radicals are likely. a-radiolysis, therefore, mainly results in the formation of the molecular species H2O2 and H2.70 A more detailed description of the radiolysis process is given in Chapter 5.02, Water Chemistry Control in LWRs. The radiolysis processes can be simulated computer programs, such as CHEMSIMUL71 or MACKSIMACHEMIST72, which calculates the yield of radiolytic species as a function of time. The necessary rate constants can be obtained from the literature, as for example from Kelm et al.60 5.16.4.3.5 Corrosion of spent fuel under reducing conditions

As mentioned above, the inventory of different elements in spent fuel is divided between the gap, cracks,

grain boundaries, and the fuel grains. The elements in the gap, cracks, and the grain boundaries will be the first to come into contact with water if the canister is breached. Some of them will be quickly dissolved in the water and are referred to as the instant release fraction (IRF), see separate discussion in Section 5.16.4.3.7. Even though the IRF will be quickly released, the main problem from activity spreading point of view is dissolution of the UO2 matrix, since it contains more than 95% of the radionuclides (see Section 5.16.4.1). Two different types of UO2 matrix dissolution can take place; (i) chemical dissolution governed by the solubility of UO2 in the groundwater; (ii) oxidative corrosion of the uranium leading to dissolution of uranium in its much more soluble hexavalent oxidation state. 5.16.4.3.5.1

UO2 dissolution

The chemical dissolution of UO2 will be governed by its solubility in groundwater. The solubility in turn will be dependent on factors such as pH, temperature, ionic strength, complexing ions, but also on the crystallinity of UO2.

410

Spent Fuel as Waste Material

In crystalline UO2(cr), the uranium and oxygen atoms build up a fluorite lattice. This perfect structure is hard to dissolve as many bonds have to be broken to bring an atom into solution. However, surface U(IV) atoms are strongly hydrated in contact with water.73 As the coordination of water molecules takes place already on the UO2 surface before detachment, it helps break the bonds in the fluorite lattice. Once the U(IV) is dissolved, mononuclear hydroxo complexes will be formed, due to the strong hydrolysis of U(IV).74 In pure water with pH above 5, the hydroxo complexes consist of a U(IV) atom with four coordinated OH ions in tetrahedral positions.75 The dissolution can be described by the reaction: UO2 ðcrÞ þ 2H2 OðlÞ ! UðOHÞ4 ðaqÞ

½3

When the U(OH)4(aq) precipitates from a supersaturated solution, a solid is created by condensation reactions between the OH-groups. However, this process is not perfect and results in an amorphous solid (Figure 14). Due to the imperfect structure, the amorphous UO2 is easier to dissolve and equilibrates at higher solution concentrations than the crystalline.76 A theoretical solubility of crystalline and amorphous UO2 can be calculated as a function of pH from thermodynamic hydrolysis and formation constants.77 The resulting plot, for the temperature of 25  C, is shown in Figure 15. As shown in the figure, a solution with pH between 7 and 9 (i.e., in the range relevant for groundwater in a repository) in contact with UO2(am) will equilibrate at a dissolved uranium concentration of 108.5 M. The lower curve shows the theoretical solubility of crystalline UO2 in water at 25  C. Owing to the low UO2 solubility, the dissolution rate of the spent fuel matrix will be limited by the diffusive or advective transport of uranium away from the fuel. As those transports are slow, the practical fuel dissolution under reducing conditions becomes negligible.39

(a)

(b)

However, the reducing condition is easily disrupted by the presence of oxidants near the fuel surface since uranium, under oxidizing conditions, can be oxidized to U(VI) and the solubility of hexavalent uranium, for example, as UO2þ 2 , is many orders of magnitude higher than that of U(IV) from UO2.75 The redox potential, Eh, of the groundwater determines whether the uranium will become oxidized or not. If the Eh is more positive than the equilibrium potential for fuel dissolution, that is, the potential for the UO2/UO2þ 2 couple, the potential difference will govern the corrosion rate of the fuel. The highest Eh at which UO2 still should be resistive to oxidation could be thermodynamically calculated to 50 mV (SHE) at pH 8.5 (based on data from Grenthe et al.,80 Paquette et al.,81 and Lemire and Tremaine82 according to calculations by Shoesmith39). Experiments have, for example, shown the presence of an a-activity threshold in UO2 pellet below which the oxidants created by a-radiolysis are too few to corrode the fuel surface.83 Today, this a-activity threshold is suggested to be as high as 18 MBq/g (UO2), which corresponds to
10 000-year-old fuel (with a burn-up of 47 GWd/tHM).84 5.16.4.3.5.2 Radiolysis-driven dissolution of fuel under reducing conditions

As mentioned above, a-radiolysis affects the redox conditions at the fuel surface. The otherwise reducing environment in the groundwater (300 mV (SHE)) inevitably becomes partly oxidizing near the fuel surface. On the other hand, the corrosion of cast-iron canister gives rise to high concentration of Fe2þ (in the range of 1  105 to 1  106 M85 and H2 concentration of 1–30 mM, corresponding 1–50 bar H2 pressure) in the groundwater in contact with the fuel. In order to test the influence of Fe and/or H2 on fuel corrosion, ‘young’ fuels such as irradiated UO2 and MOX fuels cooled for 5–30 years have been corroded in presence and absence of Fe/H2.

(c)

Figure 14 Amorphous UO2. (a) Two U(OH)4 approach each other, which (b) leads to a condensation reaction between the OH-groups. (c) The hydrous, or amorphous, solid-phase generated after multiple condensation reactions constitutes an irregular network. Courtesy of Patrik Fors.

Spent Fuel as Waste Material

411

2 UO2 100C; Parks and Pohl74

0

UO2; Ollila et al.79 –2

UO2; Rai et al.78 233UO

log [U(IV)]

–4

2;

Carbol et al.50

–6 UO2(am)

–8 –10 –12

UO2(cr)

–14 –16 0

2

4

6

8

10

12

14

pH Figure 15 Concentration of U(IV) in pure water in contact with UO2 as a function of pH. The input data for the calculations of the UO2(am) and UO2(cr) were taken from Neck and Kim.77 The total U concentrations measured in different experiments performed under anoxic/reducing conditions, and found in the literature, are inserted (detailed information on the experimental conditions can be found in Carbol et al.,50 Parks and Pohl,74 Rai et al.,78 and Ollila et al.79). © European Atomic Energy Community, 2011.

A summary of these experiments is given in Table 8. It can be deduced that corrosion of irradiated fuels independently of UO2 or MOX fuel give three to four orders of magnitude lower final uranium concentration when metallic Fe/H2 is present in the leachant in comparison with the cases when H2 is absent, for example, under oxidizing condition in air. A number of possible explanations exist in the literature for the observed low U concentration measured during corrosion of fuels, such as:  reaction of Fe2þ and radiolytic oxidants;  a reaction between H2 and radiolytic oxidants;  catalytic dissociation of H2 on e-particles (Ru, Rh, and Pd);  catalytic dissociation of H2 on UO2. The different processes are discussed in the following sections. 5.16.4.3.5.3 Fuel corrosion in presence of iron

Dissolved Fe2þ ions have been shown to be able to react with H2O2 according to the so-called Fenton reaction95,96: Fe2þ þ H2 O2 ! Fe3þ þ OH þ OH Fe2þ þ OH ! Fe3þ þ OH

½4 ½5

These reactions lead to the reduction of H2O2 in the aqueous phase and precipitation of oxidized Fe3þ. Simulations have shown that the Fenton reaction may reduce the rate of fuel dissolution with as much as a factor of 50.97 But, even though the Fe2þ leads to a decreased H2O2 concentration at the fuel surface, the Fenton reaction will not prevent the fuel surface from corroding. Loida et al.98 showed this experimentally by corroding fuel in presence of magnetite (Fe3O4, a mixed Fe2þ and Fe3þ oxides). In the experiment, very little effect of dissolved Fe2þ was seen and the atmosphere in the autoclave became oxidizing. On the other hand, a reduction of oxidants and U(VI) has been observed if the magnetite is replaced by metallic iron.69,79,89,99 In a fuel corrosion experiment in presence of metallic iron, Loida et al.89 measured a partial pressure of 2.7 bar H2 after 4 years (due to the reactions with iron given above). Additionally, an analysis of iron samples from this experiment showed formation of reduced iron corrosion products (green rust and magnetite). Moreover, it was found that negligible amounts of U(IV) had precipitated on the iron. Since it is well known that U(VI) is quickly reduced and precipitates on Fe(s),100–103 the observations by Loida et al.89 clearly show that no larger amounts

412

Spent Fuel as Waste Material

Table 8 Steady-state uranium concentrations in leachate together with experimental parameters for reported corrosion experiments on irradiated UO2 fuel and spent MOX fuel Burn-up (GWd/tHM)

H2 (MPa)

Temperature ( C)

pH

Corrosion time (d)

Concentration U (M)

Solution

UO2 fuel with H2 Spahiu et al.86 43

5.0

70


8.5

312

1  1010

Spahiu et al.86

43

5.0

25


8.5


50

5  109

Albinsson et al.87 Spahiu et al.88

41

1.0

25

8.1

21

5  109

10 mM NaCl/ 2mM NaHCO3 10 mM NaCl/2 mM NaHCO3 Mod. Allard þ Fe strip

43

0.5

70


8.5

376

2  1010

Loida et al.89 Loida et al.89

50 50

0.32 0.28

25 25

7.8 9.5

1095 1619

1  108 1  108

Fors et al.90

67

4.1

23

8.1

502

2  1010

MOX fuel with H2 Carbol et al.91 47.5

5.3

23

8.1

2100

7  1010

10 mM NaCl/ 2mM NaHCO3

UO2 fuel in air Forsyth et al.92 Je´gou et al.93

42 60

0 0

25 25

8.2 6.3

1083 15

1  105 1  108

Synthetic groundwater Pure water

MOX fuel in air Je´gou et al.94

49

0

25

6.2

14

3  107

Pure water

of oxidized uranium can have reached the iron at any stage of the experiment. Since Fe2þ alone is not able to keep the uranium reduced, the results observed by Loida et al.89 must be related to the ingrowth of H2. A similar system was studied by Albinsson et al.87 in which a clearly more efficient reduction of oxidized species was obtained by simultaneous presence of H2 and Fe2þ in comparison with H2 alone. A synergistic effect of H2 and Fe2þ could be expected in the bulk solution since the dissolved H2 will react quickly with the OH radicals104,105 created by the Fe2þ in the Fenton reaction (reactions [4] and [5]). As the radical reaction: H2 þ OH ! H þ H2 O

½6

converts an oxidizing radical (E0 ¼ 1.8 V in basic solution) into a reducing (E0 ¼ 2.3 V),70 the combined effect of Fe2þ and H2 will be much stronger than the effect of Fe2þ alone (compare reactions with Fe2þ [4] and [5]). Interestingly, Albinsson et al.87 did not report any red Fe3þ precipitates in their test. This indicates that probably a back reduction of the oxidized Fe3þ by the hydrogen in solution took place.

10 mM NaCl/ 2mM NaHCO3 5.6 mol NaCl (kg H2O) 1 5.6 mol NaCl (kg H2O) 1 þ Fe powder 10 mM NaCl/ 2mM NaHCO3

Although Albinsson et al.87 only saw a minor effect in presence of pure hydrogen, Loida et al.89 did see an equally efficient reduction of uranium in presence of 3.2 bar H2 as they did in the system with presence of both Fe2þ and H2. The reductive effect of pure hydrogen is clearly contradictory in these two experiments. Dissolved H2 has, thermodynamically, the capacity to reduce both radiolytic oxidants and dissolved U(VI). However, it is kinetically hindered at low temperatures and should be inert.46 Nevertheless, there exists today, except from the results reported by Loida et al.,89 substantial experimental evidence that hydrogen becomes activated on spent fuel surfaces. 5.16.4.3.5.4

Fuel corrosion in presence of H2

Early pulse radiolysis studies showed that an electron beam from a linear accelerator induces an OH scavenging reaction with H2 in homogeneous solutions according to radical reaction [6].104,106,107 The radical yield from the accelerated electrons is comparable with the one obtained from b- and g-radiation emitted from spent fuel. The production of oxidants by the radical-rich, low-linear energy transfer b- and g-radiation emitted by spent fuel will be considerably lower in hydrogen containing solutions because of

Spent Fuel as Waste Material

this relatively fast reaction. However, modeling has shown that this reaction will not protect the fuel surface from oxidation.108,109 More recent experiments on 4He ions accelerated to 5 MeV (comparable with a-radiation emitted from fuel) in a tandem Van de Graaff accelerator show that dissolved H2 has almost no influence on the production of radiolytic oxidants.52 Since the a-radiolysis will be dominating in the repository after 300–500 years of storage (Figure 3), any OH scavenging effect induced in the bulk solution by b(g)-radiation will be negligible. Consequently, a-radiolysis has been predicted to give oxidizing conditions in the near surroundings of the fuel surface inside a failed canister in a future repository, even in presence of H2. Even though oxidation is expected, many experiments have clearly shown that dissolved hydrogen does block oxidative corrosion of UO2 on (i) spent fuel surfaces,86,89–91,110,111 (ii) a-doped UO2,50,84,111 and (iii) solid UO2 and g-radiation112 or a-radiation.113 It is clear from the accelerator studies mentioned above that bulk reaction between H2 and radiolytic radicals cannot explain these results. Since H2, as mentioned above, is kinetically hindered at room temperature (thermal activation excluded), and experiments on dissolved UO2þ 2 in contact with H2 (without any UO2 surface) showed no reaction between the species,114 the hydrogen must become activated on UO2 and fuel surfaces. Ample results exist to support a catalytic reaction between H2 adsorbed on e-particles on the fuel surface and oxidants reacting with the surrounding uranium: (i) electrochemical studies report a clear suppression of the fuel corrosion potential in presence of dissolved H2,112,115 (ii) corrosion studies show a clear influence of noble metal doping and e-particles,44,116–119 and (iii) modeling of e-particles, galvanically coupled to the UO2 matrix, has been shown to explain the absence of corrosion.97,115,120 Despite the notable amount of results supporting an e-particle-mediated reduction of oxidants by H2, it is clear that other surfaces can support this reaction. Already in the late 1950s, Bunji and Zogovic121 reported reduction of UO2þ in presence of H2 2 and UO2 surfaces only. Later, experiments without e-particles present, but with iron in the system, have shown to become reduced due to the Fe2þ and H2 created in the system,79 and electrochemical experiments on pure UO2 surfaces have indicated a surface reaction between oxidants and dissolved H2 in presence of g-radiation.122 Additionally, the experiment with dissolved UO2þ in presence of H2 mentioned 2

413

above showed immediate signs of H2 activation after introduction of pure UO2 surfaces.114 A comprehensive summary of corrosion experiments in presence of H2 is given by Cui et al.123 5.16.4.3.5.5

Influence of fuel heterogeneity

Elemental heterogeneity in fuels could be found in irradiated MOX (see Section 5.16.2.1.2). Such heterogeneity will also exist in HBS of UO2 fuels. These structures are formed at the periphery (rim) of irradiated UO2 pellets with an average burn-up exceeding 45 GWd/tHM and are a result of the high local burn-up at the periphery (about twice the average burn-up).37 Since these heterogeneous structures give rise to high a-dose rates, and therefore high local concentrations of radiolytic oxidants, the question is whether these surfaces will be protected by the hydrogen. Irradiated MOX fuels have previously been shown to have a higher dissolution rate under aerated conditions than corresponding UO2 fuels.124,125 This indicates that the intense a-field from the Pu agglomerates accelerates the dissolution and that the agglomerates themselves are relatively readily dissolved. Nevertheless, the experimental evidence presented by Carbol et al.91 shows that heterogeneous MOX fuel surfaces become completely protected by dissolved hydrogen, in the same way as irradiated UO2 surfaces. Furthermore, Carbol and coworkers saw an indication of lower dissolution of the Pu-rich agglomerates, in comparison to surrounding UO2 matrix dissolution, which implies that the Pu-rich agglomerates are harder to dissolve than the matrix. Similar indications were found by de Pablo et al.,126,127 in a comparison of the dissolution behavior of UO2 fuel periphery (containing rim) and core. Also in this case, the rim seemed to be stabilized against corrosion. A possible explanation for the lower dissolution of the rim and Pu agglomerates, in comparison to the surrounding matrix, might be found in the stabilization of the UO2 matrix caused by solid solution of fission products (mainly lanthanides, Ln) and actinides (Pu, Am, and Cm).128,129 Both MOX and UO2 fuels become restructured (disintegrated from grain size of 510 mm to 0.2– 0.3 mm), in Pu agglomerates or at the pellet rim, once the local burn-up exceeds 45 GWd/tHM. Due to this restructuring, a part of the fission products are expected to be found at the grain surfaces instead of the interior of the grains. As a result, corrosion of HBS fuel is assumed to release significantly more fission products than nonrestructured fuel. Additionally, the

414

Spent Fuel as Waste Material

increased amount of fission products at the fuel–water interface might affect the surface’s ability to interact with dissolved hydrogen. A corrosion experiment of HBS containing UO2 fragments, presented by Fors et al.,90 showed a significant release of cesium (3.5 wt% of total Cs), as expected. However, once the cesium available at the fuel surface had been dissolved, the cesium release stopped. In the experiment, it was found that the activation of dissolved hydrogen takes place also on restructured grain surfaces. The uranium concentration of 1.5  1010 M obtained after long-term stable redox conditions was lower than previously found in a similar system with a lower burn-up UO2 fuel, 5  109 M.86 Consequently, all three studies on heterogeneous fuel structures show that these do not increase the corrosion rate in comparison to normal (40 GWd/tHM) UO2 fuels.

the 1-mM dissolved H2 concentration in the Carbol experiment is probably close to the actual threshold. It should be mentioned that studies have been performed on UO2–H2O–H2 systems without e-particles,50 clearly indicating that UO2 does not corrode despite presence of a-radiolysis (the UO2 was doped with 10 wt% 233U). These results indicate that another H2 activation mechanism, so far unknown, must take place in these systems. The threshold pressure of dissolved H2, necessary for keeping the uranium reduced in the experiment performed by Carbol et al.,91 was found to be in the range of 105 M (corresponding to 0.01 bar partial pressure of H2). Interestingly, this value is also well correlated with the model predicted by Jonsson et al.97 (if the 10 wt% 233U-doped UO2 is assumed to be a 3000year old ‘fuel’), despite the lack of e-particles.

5.16.4.3.5.6 Threshold pressure of H2

5.16.4.3.6 Corrosion of fuel under oxidizing conditions

Another issue concerns the threshold H2 concentration at which H2 starts to inhibit corrosion of fuel. Carbol et al.91 in their paper on corrosion of MOX have made an attempt to determine the lowest amount of dissolved H2 necessary for keeping the uranium reduced. The results indicate that a stable uranium concentration level of 7  1010 M could be kept at the lowest tested amount of 1 mM dissolved H2. The amount of 1 mM dissolved H2 corresponds to a gas pressure of
1 bar. This amount of hydrogen has also been shown to reduce uranium dissolution rates in flow-through experiments of SNF by four orders of magnitude (as compared to oxidizing conditions).110 Based on these results, SNF is expected to be protected by dissolved hydrogen already at concentrations around 1 mM. Jonsson et al.97 concluded, based on experimental evidence and modeling, that e-particle-catalyzed solid-phase reduction by H2 will efficiently block any oxidative dissolution of fuel 100 years or older already at a partial pressure of 0.1 bar H2. They did not extrapolate their discussion to younger fuels. However, as a nearly linear relation is seen between the logarithm of the fuel age and the logarithm of the dissolution rate in their results, it is possible to get an approximate value of the theoretically required hydrogen pressure for the nearly 20-year-old MOX fuel studied by Carbol et al.91 If this extrapolation is made, around 0.5 bar H2 is found as the required pressure. Considering the difference in fuel and burn-up, MOX, 44 GWd/tHM, in comparison to UO2, 38 GWd/tHM, in the paper of Jonsson et al.,97

Corrosion of SNF in oxidizing conditions is motivated, for example, to assess the impact of possible accident scenarios during the cooling of the fuel in storage pools. Three sets of experiments were set up to study corrosion of spent UO2 fuel fragments (60 GWd/tHM) after having them washed free from the inventory in gap and grain boundaries.93 The first was carried out in presence of an external g-radiation field of 650 Gy h1, representing a fuel bundle, the second in presence of 1.2  104 M H2O2 in the start solution, as oxidative a-radiolysis product, and the third with only fuel fragments and water. The leachant was pure water (except when H2O2 was added) at 25  C, with pH 6.3. The highly oxidizing conditions motivated a relatively short experimental time of 14 days. The results showed that the hydrogen peroxide concentration was almost equally high (1  104 M) for the external irradiated fuel and in the experiment with added H2O2, while it was 1  107 M in the aerated solution containing only fuel fragments. The U concentration, at the end of the experiment, reached for the externally irradiated fuel 1  106 M, for the one with addition of H2O2 1  107 M, and for the aerated 1  108 M, which shows that under continuous g-irradiation, the U concentration continues to increase until the solubility of studtite (UO2(O2) 4H2O) is reached (
5  106 M). Presence of studtite on the fuel surface, as needlelike and rodlet crystals, was confirmed by XRD both on the fuel surfaces in the experiment with g-irradiated fuel and on the surfaces of the H2O2-exposed fragments. No studtite was observed on the fuel surface

Spent Fuel as Waste Material

in the experiment with only fuel fragments and water. A mass balance showed that 70–90% of the dissolved uranium has precipitated on the fuel surfaces. Postexperimental leaching of the fuel surfaces with aerated carbonated water resulted in dissolution of UO2þx but no studtite, while change of the leachant to Ar/4 vol% H2 purged carbonated water gave a disintegration of studtite, as could be expected due to the simultaneous presence of the reactive peroxide and H2. Spent fuel corrosion in simulated groundwater under oxidizing conditions was also studied by Ekeroth et al.130 The authors observed an almost linear increase in corrosion, measured as fraction of inventory in aqueous phase (FIAP), of fuels with increased burn-up, up to 40–45 GWd/tHM, whereas spent fuels with a higher burn-up than 50 GWd/tHM showed a decrease in corrosion. The authors offered a possible explanation in that the UO2 matrix may become chemically stabilized against oxidation by the increased fraction of lanthanides present in the fuel, either as fission products or as added dopant (Gd). Similar results were obtained by Kline et al.131 when they compared the corrosion of UO2 with Gd-doped UO2, under oxidizing conditions. It can be concluded that doping of the UO2 matrix with lanthanides seems to stabilize the matrix against oxidation and dissolution. 5.16.4.3.7 Instant release

When the container breaches, groundwater can come in contact with the waste. For repository safety assessments, the fission gas release, FGR has been used as an indicator for rapid release of segregated radionuclides such as 14C (t1/2 5.73  103 years), 36 Cl (t1/2 3.01  105 years), 79Se (t1/2 3.27  106 years), 99 Tc (t1/2 2.11  105 years), 107Pd (t1/2 6.5  106 years), 126 Sn (t1/2 2.3  105 years), 129I (t1/2 1.57  107 years), and 135Cs (t1/2 2.3  106 years).132 The principle is that the above-mentioned elements can form volatile compounds, in-pile, at a temperature range of 300– 1200  C, which behaves similarly as the fission gases. An initial release of radionuclides is expected to occur from all easily accessible sites. At first, the fuel assembly structural materials and the fuel cladding (e.g., steel, Zircaloy, etc.) will be affected. Here long-lived activation products of structural materials and impurities like 14C, 36Cl, 59Ni, and 63Ni are considered, which are homogeneously distributed in the material or trapped in, for example, a porous oxide layer. In case of damaged fuel cladding, water can penetrate into the pin, and a radionuclide release

415

can occur initially from the gap filling material and the outer periphery of the fuel pellets. This is followed by a release from the gap region and the fuel grain boundaries. At these sites, volatiles (129I, 137Cs, 135 Cs, 79Se, and 126Sn) and fission gases have accumulated during irradiation and fission gases are also present. It is expected that release from the gap region and the grain boundaries will be faster than the release from the fuel matrix where the fission products are incorporated in solid solution or trapped in bubbles and other structural defects. The fraction of radionuclides released by these rapid initial processes is generally defined as instant release fraction, IRF. Up-to-now, there is no consensus on the exact definition of IRF. Table 9 gives an overview of different components of the spent fuel assembly system, key radionuclides associated with them, and characteristics affecting the release mode of these radionuclides, indicating also those possibly belonging to the IRF. Experimental determination of IRF of SNF is very difficult. As a result, experimental IRF data based on leaching of spent fuel are limited. Some measured IRF values of volatile fission products like 129I, 135Cs, and the activation product 36Cl show a good correlation with the FGR. Therefore, FGR is often used to estimate the IRF of those nuclides.134 Cesium, for instance, has a low solubility in UO2 and a high amount is therefore removed from the UO2 fuel matrix during reactor operation. Similar release of cesium and xenon from fuel grains was found above 1200  C indicating that both Cs and Xe follow the same release paths in the fuel135 and accumulate at colder outer parts of the pellet. Due to the transport processes determining the relocation of volatile fission products, the IRF depends on the irradiation history of the fuel. Fuels that are irradiated with high linear power like commercial CANDU fuel present high pellet centerline temperatures and show a strong correlation between FGR and linear power. Commercial PWR fuels are irradiated at lower and more constant linear power, but up to much higher burn-up. In this case, the effect of burn-up on FGR is stronger than that of linear power.27 Table 10 summarizes calculated IRF estimates of PWR UO2 fuel for burn-ups between 41 and 75 GWd/tU.136 Despite the experimental difficulties, a small number of experiments have been performed to determine the gap and grain boundary fraction of 129I and 135Cs. Recently, Roudil et al.137 have determined the IRF of Cs from UO2 fuels with different burn-ups, in the range

416

Spent Fuel as Waste Material

Table 9

Expected distributions of radionuclides in fuel assemblies and possible modeling approaches133

Components

Key radionuclides

Characteristics and possible modeling approach

Fuel assembly structural materials 14 C (organic?) Zirconia (corroded Zircaloy)

Oxide film typically about 40–80 mm thick is formed in reactor (about 10% of cladding thickness). The oxide has a low solubility; the outer part is porous and may incorporate nuclides present in Zircaloy as the film grows Limited data on oxidized Zircaloy indicating preferential release; consider 14C as part of IRF. No data on steels cladding

14

Very low general corrosion rate Release of all nuclides plus remaining 14 C congruent with the slow corrosion rate

Fission gases, volatiles (129I, 137Cs, 135 Cs, 36Cl, 79Se, 126Sn(?)). Also 14C (nonvolatile but partially segregated)

Reliable data available for some nuclides. Assessment through fission gas release measurements and correlation with leaching experiments Part of IRF

Rim porosity

Fission gases, volatiles (129I, 137Cs, 135 Cs, 36Cl, 79Se, 126Sn(?))

Rim width is a function of burn-up; good data available. Large proportion of nuclides in rim region segregated into pores and secondary phases during in-pile restructuring. No experimental data indicating release Could be part of the IRF in a very conservative assumption

Rim grains

Actinides, fission products

Release through dissolution when water arrives. FP may also diffuse to rim pores by a-self-irradiation enhanced diffusion (a-SIED) FP inventory may thus be part of IRF or matrix release

Grain boundaries

Fission gases, volatiles (129I, 137Cs, 135 Cs, 36Cl, 79Se, 126Sn(?)). Segregated metals (99Tc, 107Pd)

Limited data As for the rim pores, it could be considered part of IRF in a very conservative assumption

Grains

Most of the actinides, fission products, and activation products inventory

Belongs to matrix release

Zircaloy, Inconel, and steel

Uranium oxide Gap

C (organic?), 36Cl, 59Ni, 63Ni

of 22–60 GWd/tHM, and one MOX fuel (47 GWd/ tHM). The linear powers of the investigated fuels were low, about 20 kW m1, and the FGR were 0.14, 0.23, 0.41, and 2.8% for the UO2 fuel burn-ups of 22, 37, 47, and 60 GWd/tHM, respectively. The gap inventories were determined by static leaching experiments using 20-mm-long clad segments and 1 mM HCO 3 leachant. The tests were performed in air atmosphere and lasted for 62 days. Determination of the Cs concentration in the leachates after 3, 10, 24, and 62 days made it possible to calculate the FIAP value of Cs. The results showed that, for the UO2 fuels, the FIAP stabilized after day 10. However, due

Table 10 IRF estimates (% of total inventory) for various radionuclides for PWR UO2 fuel Radionuclide

Fission gases (Kr, Xe) C 36 Cl 90 Sr 99 Tc, 107Pd 129 I 135 Cs, 137Cs 14

Burn-up (GWd/tHM) 41

48

60

75

1 (2) 10 5 1 (2) 0.1 (1) 1 (3) 1 (2)

2 (4) 10 10 1 (3) 0.1 (3) 2 (4) 2 (4)

4 (8) 10 16 1 (5) 0.1 (5) 4 (8) 4 (8)

8 (16) 10 26 1 (9) 0.1 (9) 8 (16) 8 (16)

Best estimate values, with conservative estimate values in brackets.136

Spent Fuel as Waste Material

to the difficulty in determining the breakpoint between the inventory in gap and the grain boundaries, the conservative values for the 62 days of leaching were reported. The experimentally determined inventory of cesium in the gap increased from 0.3% to 2.2% with increasing UO2 fuel burn-up up from 22 to 47 GWd/tHM. At the highest burn-up (60 GWd/ tHM), the Cs release was only 1%. This could possibly be explained by a closed fuel-cladding gap at this burn-up. A higher Cs release, about 3%, was observed for the 47 GWd/tHM MOX fuel. However, the actual release might be even higher as the leachate concentration in this experiment had not fully reached a stable FIAP plateau after 62 days. In another experiment,137 a pseudodynamic leach test was carried out on UO2 fuel powder with a burn-up of 60 GWd/tHM. The powder had been prepared from fuel fragments sampled from the center of a 35-mm-long clad segment previously leached in 1 mM HCO 3 leachant to eliminate the gap inventory. The powder samples contained particles with a size fraction of 20–50 mm. The number of grains in each particle was estimated at about 40. Thirty sequential leach cycles were carried out, initially of short duration (1–2 h) to avoid any precipitation resulting from leaching of the oxidized UO2þx layer, then longer (24–48 h) to determine the grain boundary inventory. The FIAPCs/FIAPU ratio decreased toward 1 after the sixth leaching cycle which showed that further Cs releases originate from the UO2 matrix. After subtracting the uranium release fraction, the Cs inventory at the grain boundaries was determined to be 0.38 wt% for UO2 fuel with a burn-up of 60 GWd/tHM. The measurement performed by Roudil et al.137 show that, for a 60 GWd/tHM UO2 fuel, the labile Cs inventory (gap þ grain boundaries) is about 1.4%, of which 1% is located in the gap and 0.4% in the grain boundaries. More detailed descriptions of IRF can be found elsewhere.27,34,83,133,138

5.16.5 Summary It can be concluded that the segregation of the elements in the spent fuel will mainly depend on the in-pile linear power (>25 kW m1) and thereafter on the fuel burn-up. The higher linear power the fuel has experienced, the larger fraction of the in-UO2 matrix nonsoluble elements (volatile and gaseous

417

fission products: He, Kr, Xe, Br, Rb, Cs, Te, I) will have diffused to the grain boundary and the gap (including cracks). The length and the time period at which the increased linear power has been applied will influence the fraction of elements moved to grain boundaries and gap. The longer and yet later in the fuel cycle the linear power has been applied, the larger fraction of the elements will exist in the grain boundaries and the gap. This segregation of elements in the fuel will govern the release rate of these elements in the deep underground nuclear fuel repository. The burn-up affects the same process by increasing the amount of fission products in the UO2 matrix which might lead to the fact that they reach their solubility in UO2 and consequently move to the grain boundary and the gap. Even though reducing conditions prevail in deep underground repositories (Eh around 0.3 V(SHE)), a spent fuel younger than 10 000 years will create partly oxidative conditions near the fuel surface upon contact with groundwater due to a-radiolysis of the water. Since the chemical environment is, in general, oxygen-free, a number of reactions will come into action that can consume the radiolytically produced oxidants, mainly OH, H2O2, and O2. Examples of such oxidants scavenging reactions are oxidation of canister iron, Zircaloy cladding and rock minerals, and oxidation of the fuel matrix. The competition between the reducing and oxidizing species at the fuel surface will determine the oxidation state of the UO2 and thereby its dissolution rate. It is univocally important that hydrogen is present in the repository since its catalytic dissociation reaction on the e-particles leads to the formation of hydrogen radicals. This radical is strongly reducing (2.6 V) and will reduce radiolytically oxidized UO2þx to UO2 and thereby hinder the dissolution of the UO2 matrix and consequently the release of fission products, present in the matrix, to the groundwater. A number of experiments with a-doped UO2 show that the e-particles are not alone responsible for the corrosion-inhibiting effect seen on fuel in presence of H2, and the interactions in the UO2–H2O–H2 system clearly show that, despite presence of a-radiolysis, the UO2 oxidation is stopped.

References 1. Ewing, R. C. Elements 2006, 2, 331–334. 2. International Atomic Energy Agency. IAEA Technical Report IAEA-TECDOC-1558, released in September

418

3.

4. 5.

6. 7.

8. 9.

10. 11. 12.

13.

14.

15. 16. 17. 18. 19. 20.

21. 22.

Spent Fuel as Waste Material 2007; International Atomic Energy Agency: Vienna, Austria. IAEA/NEA. In Management of Spent Fuel from Nuclear Power Reactors. Proceedings of an International Conference Organised by the International Atomic Energy Agency in cooperation with the OECD Nuclear Energy Agency, Vienna, Jun 19–22, 2006. Bennett, D. G.; Gens, R. J. Nucl. Mater. 2008, 379, 1–8. NWMO. Moving Forward Together: Designing the Process for Selecting a Site. Nuclear Waste Management Organisation (NWMO): Canada, 2009; (URL: http://nwmo.ca/uploads_managed/MediaFiles/ 470_InvitationtoReviewaProposedProcessfor SelectingaSite.pdf). Johnson, L.; King, F. J. Nuclear Mater. 2008, 379, 9–15. U.S. Senate Commission on Environment & Public Works. Report to the Chairman Senator James M. Inhofe U.S. Senate Committee on Environment and Public Works; March, 2006; URL: http://epw.senate.gov/ repwhitepapers/YuccaMountainEPWReport.pdf. Carter, L. J.; Pigford, T. H. Science 2005, 310, 447–448. Mon, K. G.; Hua, F. In Materials Degradation Issues in the U.S. High-Level Nuclear Waste Repository. Proceedings of the Twelfth International Conference on Environmental Degradation of Materials in Nuclear Power System – Water Reactors; Allen, T. R., King, P. J., Nelson, L., Eds. TMS (The Minerals, Metals & Materials Society), 2005; 1439–1456. Long, J. C. S.; Ewing, R. C. Ann. Rev. Earth Planet. Sci. 2004, 32, 363–401. Rosenberg, N. D.; Gdowski, G. E.; Knauss, K. G. Appl. Geochem. 2001, 16, 1231–1240. Grisak, G. E.; Pickens, J. F.; Belanger, D. W.; Avis, J. D. Hydrogeologic Testing of Crystalline Rocks During the Nagra Deep Drilling Program, NAGRA NTB 85–08; Nationale Genossenschaft fu¨r die Lagerung radioaktiver Abfa¨lle: Wettingen, Switzerland, 1985. NAGRA. Project Opalinus Clay – Safety Report: Demonstration of Disposal feasibility for spent fuel, vitrified high-level waste and long-lived intermediate level waste (Entsorgungsnachweis), NAGRA NTB 02–05; Nationale Genossenschaft fu¨r die Lagerung radioaktiver Abfa¨lle: Wettingen, Switzerland, 2005. NAGRA. Project Opalinus Clay: Models, Codes and Data for Safety Assessment: Demonstration of Disposal Feasibility for Spent Fuel, Vitrified High-Level Waste and Long-Lived Intermediate-Level Waste (Entsorgungsnachweis), NAGRA NTB 02–06; Nationale Genossenschaft fu¨r die Lagerung radioaktiver Abfa¨lle: Wettingen, Switzerland, 2002. Krane, K. S. Introductory Nuclear Physics; Wiley: New York, 1988. Fisher, S. B.; White, R. J.; Cook, P. M. A.; et al. J. Nucl. Mater. 2002, 306, 153–172. Walker, C. T.; Goll, W.; Matsumura, T. J. Nucl. Mater. 1997, 245, 169–178. Walker, C. T.; Goll, W.; Matsumura, T. J. Nucl. Mater. 1996, 228, 8–17. Bailly, H. Me´nessier, D.; Prunier, C.; Eds. The Nuclear Fuel of Pressurized Water Reactor and Fast Neutron Reactors; Lavoisier: Paris, 1999; pp 123–124. Duro, L.; Grive´, M.; Cera, E.; Gaona, X.; Dome`nech, C.; Bruno, J. Determination and Assessment of the Concentration Limits to be Used in SR-Can, SKB TR-06–32; Swedish Nuclear Fuel and Waste management Co.: Stockholm, Sweden, 2006. Kleykamp, H. J. Nucl. Mater. 1985, 131, 221–246. Ronchi, C.; Sheindlin, M.; Staicu, D.; Kinoshita, M. J. Nucl. Mater. 2004, 327, 58–76.

23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

40. 41. 42. 43. 44. 45. 46. 47.

48.

49.

50. 51. 52.

Ferry, C.; Poinssot, C.; Cappelaere, C.; et al. J. Nucl. Mater. 2006, 352, 246–253. Marchetti, I.; Belloni, F.; Himbert, J.; Carbol, P.; Fangha¨nel, Th. J. Nucl. Mater. 2010, DOI: 10.1016/j. jnucmat.2010.10.065. Matzke, H. Radiat. Effects 1982, 64, 3–33. Van Brutzel, L.; Rarivomanantsoa, M.; Ghaleb, D. J. Nucl. Mater. 2006, 354, 28–35. Johnson, L.; Ferry, C.; Poinssot, C.; Lovera, P. J. Nucl. Mater. 2005, 346(1), 56–65. Rondinella, V. V.; Wiss, T. Mat. Today 2010, 13(12), 24–32. Rufeh, F.; Olander, D. R.; Pigford, T. H. Nucl. Sci. Eng. 1965, 23, 335–338. Sung, P. PhD Thesis, U.M.I. Dissertation Services, University of Washington, 1967. Maugeri, E.; Wiss, T.; Hiernaut, J.-P.; et al. J. Nucl. Mater. 2009, 385, 461–466. Ronchi, C.; Hiernaut, J.-P. J. Nucl. Mater. 2004, 325, 1–12. Matzke, H. J. Nucl. Mater. 1995, 223, 1–5. Ferry, C.; Poinssot, C.; Broudic, V.; et al. Commissariat a` L´E´nergie Atomique; Report CEA-R-6084; Saclay, France, 2005. Rondinella, V. V.; Cobos, J.; Matzke, H.; Wiss, T.; Carbol, P.; Solatie, D. Mater. Res. Soc. Symp. Proc. 2001, 663, 391–398. Kleykamp, H. Nuclear Technol. 1988, 80, 412–422. Walker, C. T.; Bremier, S.; Portier, S.; Hasnaou, R.; Goll, W. J. Nucl. Mater. 2009, 393, 212–223. Kleykamp, H. J. Nucl. Mater. 1989, 167, 49–63. Shoesmith, D. W. Used Fuel and Uranium Dioxide Dissolution Studies: A Review, NWMO TR-2007-03: 1–69; Nuclear Waste Management Organization: Toronto, Canada, 2007. Kleykamp, H.; Paschoal, J. O.; Pejsa, R.; Thu¨mmler, F. J. Nucl. Mater. 1985, 130, 426–433. He, H.; Keech, P. G.; Broczkowski, M. E.; Noe¨l, J. J.; Shoesmith, D. W. Can. J. Chem. 2007, 85(10), 702–713. Sari, C.; Walker, C. T.; Schumacher, G. J. Nucl. Mater. 1979, 79(1), 255–259. Johnson, I.; Johnson, C. E.; Crouthamel, C. E.; Seils, C. A. J. Nucl. Mater. 1973, 48(1), 21–34. Cui, D.; Low, J.; Sjo¨stedt, C. J.; Spahiu, K. Radiochim. Acta 2004, 92, 551–555. Bruno, J. Bosbach, D. Kulik, D.; Navrotsky, A.; Eds. Chemical Thermodynamics, OECD: Paris, France, 2007; Vol. 10. Johnson, L. H.; Shoesmith, D. W.; Eds. Radioactive Waste Forms for the Future; North-Holland Elsevier: Amsterdam, The Nederlands, 1988. IAEA. Understanding and Managing Ageing of Materials in Spent Fuel Storage Facilities; International Atomic Energy Agency, IAEA Technical Reports Series No. 44; Vienna, Austria, 2006. Wersin, P.; Spahiu, K.; Bruno, J. Time Evolution of Dissolved Oxygen and Redox Conditions in a HLW Repository, SKB TR-94–02; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 1994. Puigdomenech, I.; Ambrosi, J.-P.; Eisenlohr, L.; et al. O2 Depletion in Granitic Media, SKB TR-01-05; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2001. Carbol, P.; Fors, P.; Gouder, T.; Spahiu, K. Geochim. Cosmochim, Acta 2009, 73, 4366–4375. Shoesmith, D. W. J. Nucl. Mater. 2000, 282(1), 1–31. Pastina, B.; LaVerne, J. A. J. Phys. Chem. A 2001, 105(40), 9316–9322.

Spent Fuel as Waste Material 53.

54. 55.

56.

57.

58. 59.

60. 61. 62.

63. 64.

65. 66. 67.

68. 69. 70. 71.

72.

Guimera`, J.; Duro, L.; Delos, A. Changes in Groundwater Composition as a Consequence of Deglaciation, SKB R-06-105; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2006. Metz, V.; Kienzler, B.; Schu¨bler, W. J. Contaminant Hydrol. 2003, 61, 265–279. Hunter, F.; Bate, F.; Heath, T.; Hoch, A.; Assurance, S. Geochemical Investigation of Iron Transport into Bentonite as Steel Corrodes, SKB TR-07-09; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2007. Kienzler, B.; Schu¨ßler, W.; Metz, V. Erstellung eines integrierten Nahfeldmodells von Gebinden hochaktiver Abfa¨lle im Salzstock Gorleben: geochemisch fundierter Quellterm fu¨r HAW-Glas abgebrannte Brennelemente und Zement. Quellterme fu¨r HAW-Glas, abgebrannten Kernbrennstoff und zementierte Abfa¨lle, FZK-INE 008/99; Forschungszentrum Karlsruhe: Karlsruhe, Germany, 2000. Bjo¨rck, S.; Svensson, N.-O. Climatic Changes and Uplift Patterns – Past, Present and Future. SKB TR 92-38; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 1992. Gorman-Lewis, D.; Mazeina, L.; Fein, J. B.; Szymanowski, J. E. S.; Burns, P. C.; Navrotsky, A. J. Chem. Thermodyn. 2007, 39, 568–575. Grenthe, I.; Fuger, J.; Konings, R. J. M.; et al., Eds. Chemical Thermodynamics of Uranium, Chemical thermodynamics; North-Holland Elsevier: Amsterdam, The Netherlands, 2004. Kelm, M.; Metz, V.; Bohnert, E.; Janata, E.; Bube, C. Radiat. Phys. Chem. 2010; DOI:10.1016/j. radphyschem.2010.10.011. Arcos, D.; Bruno, J.; Karnland, O. Appl. Clay Sci. 2003, 23, 219–228. Bornemann, O.; Behlau, J.; Fischbeck, R.; et al. Standortbeschreibung Gorleben, Teil 3, Ergebnisse der u¨ber- und unterta¨gigen Erkundung des Salinars. Geologisches Jahrbuch Reihe C, Band C, 2008, 73. Bornemann, O.; Schramm, M.; Tomassi-Morawiec, H.; et al. Geologos 2008, 14, 73–90. Bundesamt fu¨r Strahlenschutz. Brine Reservoirs in the Gorleben Salt Dome. Bundesamt fu¨r Strahlenschutz/ Federal Office for Radiation Protection, Salzgitter, Germany, 2010; URL: http://www.bfs.de/en/endlager/ gorleben/loesungen_salzstock.html. Rosborg, B.; Werme, L. J. Nucl. Mater. 2008, 379(1–3), 142–153. Garrels, R. M.; Christ, C. L. Solutions, Minerals, and Equilibria; Harper & Row: New York, 1965. Smart, N. R.; Blackwood, D. J.; Werme, L. The Anaerobic Corrosion of Carbon Steel and Cast Iron in Artificial Groundwaters, SKB TR-01–22; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2001. Bonin, B.; Colin, M.; Dutfoy, A. J. Nucl. Mater. 2000, 281(1), 1–14. Roth, O.; Jonsson, M. Central Eur. J. Chem. 2008, 6(1), 1–14. Choppin, G. R.; Liljenzin, J.-O.; Rydberg, J. Radiochemistry and Nuclear Chemistry, Butterworth-Heinemann: Woburn, MA, 2002; 3rd edn. Kirkegaard, P.; Bjergbakke, E. CHEMSIMUL, a Program Package for Numerical Simulation of Chemical Reaction Systems; Riso¨ National Laboratory: Roskilde, Denmark, 1998. Carver, M. B.; Hanley, D. V.; Chaplin, K. R. MACKSIMA-CHEMIST, A Program for Mass Action Kinetics Simulation by Automatic Chemical Equation Manipulation and Integration Using Stiff Techniques, AECL 6413; AECL: Canada, 1979.

73. 74. 75. 76. 77. 78. 79.

80. 81. 82. 83.

84. 85.

86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101.

419

Langmuir, D. Geochim. Cosmochim. Acta 1978, 42(6, Part 1), 547–569. Parks, G. A.; Pohl, D. C. Geochim. Cosmochim. Acta 1988, 52(4), 863–875. Guillaumont, R. Fangha¨nel, T.; Fuger, J.; et al., Eds. Chemical Thermodynamics; North-Holland Elsevier: Amsterdam, The Nederlands, 2003; Vol. 5. Langmuir, D.; Ed. Aqueous Environmental Geochemistry; Prentice Hall: New Jersey, 1997. Neck, V.; Kim, J. I. Radiochim. Acta 2001, 89, 1–16. Rai, D.; Yui, M.; Moore, D. A. J Sol. Chem. 2003, 32(1), 1–17. Ollila, K.; Albinsson, Y.; Oversby, V. M.; Cowper, M. Dissolution Rates of Unirradiated UO2, UO2 Doped with 233 U, and Spent Fuel Under Normal Atmospheric Conditions and Under Reducing Conditions Using an Isotope Dilution Method, SKB TR-03-13; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2003. Grenthe, I. Fuger, J.; Konings, R. J. M.; et al., Eds. Chemical Thermodynamics; North-Holland Elsevier: Amsterdam, The Nederlands, 1992. Paquette, J. L.; Lemire, R. J. Nucl. Sci. Eng. 1981, 79, 26–48. Lemire, R. J.; Tremaine, P. R J. Chem. Eng. Data 1980, 25(4), 361–370. Poinssot, C.; Ferry, C.; Kelm, M.; et al. Final Report of the European Project Spent Fuel Stability under Repository Conditions, CEA CEA-R-6093; Commissariat a´ l´E´nergie Atomique: Saclay, France, 2005. Muzeau, B.; Je´gou, C.; Delaunay, F.; et al. J. Alloys Comp. 2009, 467(1–2), 578–589. Hunter, F.; Bate, F.; Heath, T.; Hoch, A.; Assurance, S. Geochemical Investigation of Iron Transport into Bentonite as Steel Corrodes, SKB TR-07-09; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2007. Spahiu, K.; Werme, L.; Eklund, U.-B. Radiochim. Acta 2000, 88(9–11), 507–511. Albinsson, Y.; O¨degaard-Jensen, A.; Oversby, V. M.; Werme, L. O. Mater. Res. Soc. Symp. Proc. 2003, 757, 407–413. Spahiu, K.; Cu, D.; Lundstro¨m, M. Radiochim. Acta 2004, 92, 625–629. Loida, A.; Metz, V.; Kienzler, B.; Geckeis, H. J. Nucl. Mater. 2005, 346(1), 24–31. Fors, P.; Carbol, P.; Van, S.; Spahiu, K. J. Nucl. Mater. 2009, 394, 1–8. Carbol, P.; Fors, P.; Van Winckel, S.; Spahiu, K. J. Nucl. Mater. 2009, 392, 45–54. Forsyth, R. S.; Werme, L. O.; Bruno, J. J. Nucl. Mater. 1986, 138, 1–15. Je´gou, C.; Muzeau, B.; Broudic, V.; Roudil, D.; Deschanels, X. Radiochim. Acta 2007, 95, 513–522. Je´gou, C.; Caraballo, R.; De Bonfils, J.; et al. J. Nucl. Mater. 2010, 399, 68–80. Fenton, H. J. H. J. Chem. Soc. Transact. 1894, 65, 899–910. Sutton, H. C.; Winterbourn, C. C. Free Radic. Biol. Med. 1989, 6(1), 53–60. Jonsson, M.; Nielsen, F.; Roth, O.; Ekeroth, E.; Nilsson, S.; Hossain, M. M. Environ. Sci. Technol. 2007, 41, 7087–7093. Loida, A.; Kienzler, B.; Geckeis, H. Mater. Res. Soc. Symp. Proc. 2003, 757, 433–439. Cui, D.; Spahiu, K. Radiochim. Acta 2002, 90, 623–628. Grambow, B.; Smailos, E.; Geckeis, H.; Muller, R.; Hentschel, H. Radiochim. Acta 1996, 74, 149. Fiedor, J. N.; Bostick, W. D.; Jarabek, R. J.; Farrell, J. Environ. Sci. Technol. 1998, 32(10), 1466–1473.

420

Spent Fuel as Waste Material

102. Gu, B.; Liang, L.; Dicke, M. J.; Yin, X.; Dai, S. Environ. Sci. Technol. 1998, 32(21), 3366–3373. 103. Farrell, J.; Bostick, W. D.; Jarabek, R. J.; Fiedor, J. N. Ground Water 1999, 37(4), 618–624. 104. Schmidt, K. H. J. Phys. Chem. 1977, 81(13), 1257–1263. 105. Christensen, H.; Sehested, K. J. Phys. Chem. 1983, 87(1), 118–120. 106. Schwarz, H. A. J. Phys. Chem. 1962, 66(2), 255–262. 107. Dorfman, L. M. Science 1963, 141, 493–498. 108. Merino, J.; Cera, E.; Bruno, J.; et al. J. Nucl. Mater. 2005, 346(1), 40–47. 109. Cera, E.; Bruno, J.; Duro, L.; Eriksen, T. Experimental Determination and Chemical Modelling of Radiolytic Processes at the Spent Fuel/Water Interface. Long Contact Time Experiments, SKB TR-06-07; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2006. 110. Ro¨llin, S.; Spahiu, K.; Eklund, U. B. J. Nucl. Mater. 2001, 297(3), 231–243. 111. Carbol, P.; Cobos-Sabate´, J.; Glatz, J.-P.; et al. The Effect of Dissolved Hydrogen on the Dissolution of 233 U Doped UO2(s), High Burn-up Spent Fuel and MOX Fuel, SKB TR-05–09; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2005. 112. King, F.; Shoesmith, D. W. Electrochemical Studies of the Effect of H2 on UO2 Dissolution, SKB TR-04-20; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2004. 113. Sunder, S.; Boyer, G. D.; Miller, N. H. J. Nucl. Mater. 1990, 175, 163–169. 114. Spahiu, K.; Devoy, J.; Cui, D.; Lundstro¨m, M. Radiochim. Acta 2004, 92(9–11), 597–601. 115. Broczkowski, M. E.; Noel, J. J.; Shoesmith, D. W. J. Nucl. Mater. 2005, 346(1), 16–23. 116. Nilsson, S.; Jonsson, M. J. Nucl. Mater. 2008, 374(1–2), 290–292. 117. Nilsson, S.; Jonsson, M. J. Nucl. Mater. 2008, 372(2–3), 160–163. 118. Trummer, M.; Nilsson, S.; Jonsson, M. J. Nucl. Mater. 2008, 378(1), 55–59. 119. Trummer, M.; Roth, O.; Jonsson, M. J. Nucl. Mater. 2009, 383(3), 226–230. 120. Eriksen, T. E.; Jonsson, M.; Merino, J. J. Nucl. Mater. 2008, 375(3), 331–339. 121. Bunji, B.; Zogovic, B. Proc. 2nd U.N. Int. Conf. Peaceful Uses Atomic Energy 1958, 3, 350–355. 122. King, F.; Quinn, M. J.; Miller, N. H. The Effect of Hydrogen and Gamma Radiation on the Oxidation of UO2 in 0.1 moldm3 NaCl Solution, SKB TR-99-27; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 1999.

123. 124. 125. 126.

127. 128. 129. 130. 131. 132.

133.

134.

135. 136. 137. 138.

Cui, D.; Ekeroth, E.; Fors, P.; Spahiu, K. Mater. Res. Soc. Symp. Proc. 2008, 1104, 1104-NN03-05. Loida, A.; Grambow, B.; Karsten, G.; Dressler, P. Mater. Res. Soc. Symp. Proc. 1998, 506, 923–924. Je´gou, C.; Peuget, S.; Broudic, V.; Roudil, D.; Deschanels, X.; Bart, J. M. J. Nucl. Mater. 2004, 326(2–3), 144–155. De Pablo, J.; Casas, I.; Gime´nez, J.; et al. In Contribution to Leaching Studies of High-Burnup Spent Nuclear Fuel. 12th International High-Level Radioactive Waste Management Conference (IHLRWM), Las Vegas, Nevada, Sept 7–11, 2008; pp 412–417. De Pablo, J.; Serrano-Purroy, D.; Gonzalez-Robles, E.; et al. Mater. Res. Soc. Symp. Proc. 2009, 1193, 613–620. Thomas, L. E.; Einziger, R. E.; Buchanan, H. C. J. Nucl. Mater. 1993, 201, 310–319. McEachern, R. J.; Taylor, P. J. Nucl. Mater. 1998, 254(2–3), 87–121. Ekeroth, E.; Low, J.; Zwicky, H.-U.; Spahiu, K. Mater. Res. Soc. Symp. Proc.; 2009, 1124, 1124-Q02–07. Kline, A. J; Hanson, B. D.; Miller, W. H. Transac. Am. Nucl. Soc. 2007, 97(1), 73–74. Werme, L.; Johnson, L. H.; Oversby, V. M.; et al. Spent fuel performance under repository conditions: A model for use in SR-Can, SKB TR-04–19; Swedish Nuclear Fuel and Waste Management Co.: Stockholm, Sweden, 2004. Johnson, L.; Poinssot, C.; et al. Estimates of the Instant Release Fraction for UO2, NAGRA NTB 04–08; Nationale Genossenschaft fu¨r die Lagerung radioaktiver Abfa¨lle: Wettingen, Switzerland, 2004. SKI. Spent Fuel Dissolution and Source Term Modelling in Safety Assessment, SKI Report 2007:17, Swedish Nuclear Power Inspectorate (SKI): Stockholm, Sweden, 2007. Walker, C. T.; Bagger, C.; Mogensen, M. J. Nucl. Mater. 1996, 240(1), 32–42. Ferry, C.; Piron, J.-P.; Poutesquen, A.; Poinssot, C. Mater. Res. Soc. Symp. Proc. 2008, 1107, 447–454. Roudil, D.; Je´gou, C.; Broudic, V.; Muzeau, B.; Peuget, S.; Deschanels, X. J. Nucl. Mater. 2007, 362 (2–3), 411–415. Grambow, B.; Lemmens, K.; Minet, Y.; et al. Final Synthesis Report RTD Component 1: Dissolution and release from the waste matrix. NFPRO Understanding and Physical and Numerical Modelling of the Key Processes in the Near-Field and their Coupling for Different Host Rocks and Repository Strategies. Deliverable D1.6.3. Integrated Project NF-PRO. Thematic Priority FP6-EURATOM-NUWASTE-20033.2.1.1–3. Contract Number: FI6W-CT-2003-02389, 2008.

5.17

Waste Containers

F. King Integrity Corrosion Consulting Ltd, Nanaimo, BC, Canada

ß 2012 Elsevier Ltd. All rights reserved.

5.17.1

Introduction

422

5.17.2 5.17.2.1 5.17.2.1.1 5.17.2.1.2 5.17.2.1.3 5.17.2.1.4 5.17.2.1.5 5.17.2.1.6 5.17.2.1.7 5.17.2.1.8 5.17.2.1.9 5.17.2.1.10 5.17.2.1.11 5.17.2.2 5.17.3 5.17.3.1 5.17.3.2 5.17.3.3 5.17.4 5.17.4.1 5.17.4.2 5.17.4.3 5.17.4.4 5.17.4.5 5.17.5 5.17.6 References

Repository Environments Environmental Factors Temperature Chloride concentration Groundwater and pore-water chemistry pH Sulfur species Redox conditions Radiolysis and irradiation Degree of saturation Mass-transport conditions Microbial activity External loads and residual stress Evolution of the Environment Material Selection Target Lifetime Active Versus Passive Materials Environmental Considerations Corrosion Behavior of Candidate Canister Materials Carbon Steel Stainless Steels Copper Titanium Alloys Nickel Alloys Canister Lifetime Predictions Conclusions

423 423 423 423 424 424 425 425 426 426 426 426 426 426 427 428 428 428 429 429 433 437 441 443 446 447 448

Abbreviations AISI C-steel EVA HFIC HIC HLW IGA ILW LILW MIC PREN SCC SCE SF

American Iron and Steel Institute Carbon steel Extreme-value analysis High-field ion-conduction (mechanism) Hydrogen induced cracking High-level waste Intergranular attack Intermediate-level waste Low- and intermediate-level waste Microbiologically influenced corrosion Pitting resistance equivalent number Stress corrosion cracking Standard calomel electrode Spent fuel

SHE SKB TBD UNS

Standard hydrogen electrode Svensk Ka¨rnbra¨nslehantering AB To be determined Unified Numbering System

Symbols D EB ECORR EFB EOC Ercrev ERP

Diffusivity or diffusion coefficient Pitting or breakdown potential Corrosion potential Flat band potential Open-circuit potential Crevice repassivation potential Repassivation potential

421

422 HC d DE

Waste Containers

Critical absorbed hydrogen concentration Diffusion layer thickness Potential difference

5.17.1 Introduction Methods for the permanent disposal of nuclear waste are being investigated internationally.1–9 Without exception, these methods involve a series of natural and engineered barriers in an underground geological repository, with one of the engineered barriers being a

metallic canister (Figure 1). The canister is unique among these various barriers since it is the only one that is absolute, the others being permeable to a greater or lesser degree. Canisters are required for both high-activity wastes, typically spent fuel (SF) or high-level waste (HLW) from reprocessing operations immobilized in some sort of insoluble matrix, and certain low and intermediate-level wastes (LILWs). Different types of geological setting are being considered in different countries, but here the emphasis is on the long-term performance of the canister material. The canisters may ultimately fail as a result of either mechanical overload or

1

2

4 3 5 7

Legend

6

1. Surface facilities 2. Main shaft complex 3. Placement rooms 4. Backfill

5. Container and buffer 6. Container 7. Used nuclear fuel

Figure 1 Illustration of a typical multibarrier system for the disposal of nuclear waste comprising various natural and engineered barriers. Reproduced with permission from Nuclear Waste Management Organization, Toronto, ON, Canada.

Waste Containers

penetration by corrosion. Mechanical overload should be avoidable through proper engineering design, although the time-dependent effects of creep, even at near-ambient temperature, can be more difficult to predict. Nevertheless, the most likely cause of canister failure, and that addressed here, is corrosion. Selection of an appropriate canister material depends on a number of factors, including
target canister lifetime,
nature of the repository environment and how it changes over time,
effect of the canister material on other barriers,
robustness of lifetime predictions,
ease of canister fabrication, sealing, and inspection,
resource availability, and
cost. Various classes of alloy have been considered for canisters for SF/HLW and LILW.7–17 For SF/HLW, for which absolute containment is a prerequisite and for which longer canister lifetimes are often desirable, the main classes of materials that have been considered are oxygen-free copper,5,6,10,11 carbon or mild steel (C-steel),2,3,13,14 titanium alloys,1,2,7,8,12 and certain nickel-based alloys.1,7–9 For LILW, for which absolute containment is not always required (indeed, some LILW canister designs incorporate a vent to allow gases produced by degradation of the waste form to escape), the main classes of materials considered are C-steel and various grades of stainless steel.8,15–17 An important aspect of the selection of nuclear waste canister materials is the ability to make justifiable long-term predictions of the corrosion behavior. Various methods have been developed for predicting either the rate of, or the susceptibility of a given material to, different types of corrosion, including general corrosion, localized corrosion in the form of pitting or crevice corrosion, stress corrosion cracking (SCC), microbiologically influenced corrosion (MIC), and the effects of absorbed hydrogen. This chapter discusses the major aspects of the corrosion of canister materials for the disposal of nuclear wastes, namely material selection, the corrosion behavior under repository conditions, and lifetime prediction. The nature of the disposal environments is also briefly described in order to provide necessary background for the corrosion discussion. Specifically excluded from the current discussion are corrosion issues associated with interim storage, the long-term behavior of nonmetallic canister materials such as ceramics, and any discussion of canister

423

design, fabrication, and sealing except inasmuch as it might impact the corrosion behavior.

5.17.2 Repository Environments 5.17.2.1

Environmental Factors

Table 1 provides a brief summary of selected national nuclear waste programs. Detailed descriptions of the environmental conditions in different repositories in different host rocks can be found elsewhere. The most important of these conditions have been discussed. 5.17.2.1.1 Temperature

High-activity wastes emit heat, resulting in elevated canister surface temperatures. A maximum canister surface temperature of 100  C is specified in many repository designs in order to minimize adverse effects on other barriers and to minimize the effects of the initial thermal transient on the near-field environment.2,3,5,6 In contrast, the Yucca Mountain Repository in the United States was designed to operate at an elevated temperature (maximum canister surface temperature of 200–250  C) for a period of several hundred years, partly to preclude the possibility of aqueous corrosion processes until the initial activity of the waste had decayed significantly.1 Low- and intermediate-level wastes (ILWs) do not produce significant decay heat and containers for these wastes will be subject to ambient repository temperatures, except for a period of elevated temperature due to the curing of cementitious backfilling materials, if used.17 5.17.2.1.2 Chloride concentration

The Cl concentration depends largely on the nature of the host-rock environment and can vary from a few tens to many thousands of milligrams per liter. Sedimentary clay host rocks in Belgium, for example, exhibit Cl concentrations in aquifers close to the proposed Boom Clay host formation of 18 mg L1.22 In contrast, salt formations contain brine inclusions with Cl concentrations of up to 200 g L1.8 In some host-rock formations, such as fractured granitic formations in Finland, Sweden, and Canada, the salinity of the groundwater increases with increasing depth (Figure 2).5,6,21 Higher salinity groundwaters tend to be older groundwaters that have concentrated due to prolonged water–rock interactions and/or waters that were originally formed from ancient seawaters. In either case, the fact that there is a distinct gradation of salinities indicates that the deeper waters have had little interaction with the shallower fresher waters,

424 Table 1

Waste Containers Main characteristics of various national nuclear waste disposal programs Waste type Host rock

Repository locationa

Backfillb

UK

ILW

Generic

Saturated

Cementitious Ambient

UK Germany Spain

HLW (SF) HLW/SF SF

Saturated Saturated Saturated

TBD 100? Crushed salt 200 Bentonite 100

Belgium France

HLW (SF) HLW/SF

Generic Salt Granite, clay Boom clay Clay

Saturated Saturated Saturated

Cementitious <100 None – HLW <100 Bentonite-SF Bentonite 130

Saturated Saturated Saturated Saturated

Bentonite Bentonite Bentonite Bentonite

Switzerland HLW/SF Sweden Finland Japan Canada

SF SF HLW SF

USA

SF/HLW

Opalinus clay Granite Granite Generic Granite Shale/ limestone Tuff

Unsaturated None

Maximum canister Groundwater Notes References temperature ( C) [Cl] (mg L1) To be determined (TBD) TBD 300 000 6550

1

17

2 3 4

18 8 8

27 2000

4,8,19 20

3450

3

<100 <100 100 100

6900 16 000 TBD 34 300 200 000

5 6

5 6 2 14,21

200–220

7

7

1

a Refers to whether the proposed repository location is below the water table in the saturated zone or above the water table in the unsaturated, or vadose, zone. b Refers to the material in contact with the canister. The same or other materials may be used as backfill elsewhere in the repository. 1 – No specific location has been identified to host an ILW repository in the United Kingdom. 2 – No specific location has been identified to host a HLW (and SF if it is deemed to be a waste) repository in the United Kingdom. Three generic types of host rock have been defined (a lower-strength sedimentary rock, a higher-strength rock, and an evaporite) and indicative groundwater compositions defined. 3 – After a period of some R&D activity, progress on the German program has slowed. 4 – Spain has considered both clay and granite host-rock formations. The groundwater [Cl] given is that for granite host rock. 5 – The Japanese program has considered various generic host rock types and associated groundwaters and awaits a volunteer host community. 6 – Canada has investigated both granitic and sedimentary host rocks. The Cl concentrations given are the maximum for each type of host rock. 7 – At the time of writing, the US Department of Energy has applied to withdraw the License Application for the Yucca Mountain Repository. The [Cl] concentration refers to that of J-13 well water, but the seepage water dripping on to the hot drip shields and waste packages will be evaporated to dryness, potentially resulting in highly concentrated electrolytes.

an obvious advantage for deep geological disposal of nuclear wastes. 5.17.2.1.3 Groundwater and pore-water chemistry

Chemical species other than Cl can also influence the corrosion behavior of the canister. Sulfur species are discussed in more detail below. Carbonates can induce passivity of active materials, such as copper or C-steel, and can inhibit the localized corrosion of passive materials, such as stainless steel and Ni-based alloys.7 Cations can also influence the corrosion behavior; for example, the hydrolysis of Mg2þ leads to acidification of brine inclusions, and species such as Ca2þ and Mg2þ can become incorporated into corrosion product layers.7,8 It is important to recognize that the composition of the aqueous phase contacting the canister may differ from that of the groundwater.6 In many disposal systems, the canister is to be surrounded by a bentonite clay buffer material.2,3,5,6 Bentonite is an ion-exchanger

and will modify the cationic composition of the groundwater depending on the nature of the counterion (typically either Naþ or Ca2þ).3,6 Naturally occurring bentonite clays also contain impurity minerals, such as gypsum, anhydrite, calcite, pyrite, and halite, the dissolution of which will further modify the composition of the solution contacting the canister. 5.17.2.1.4 pH

Natural groundwaters tend to have neutral or slightly alkaline pH values. The presence of calcite in the bentonite will buffer the pore water at pH 8.3,6 In some disposal systems, a cementitious material is used to control the interfacial pH in the near field, either to promote passivation of the canister material as in the case of the Belgian Supercontainer or to control the solubility of radionuclides.4,8,17,19,22 The pH of a cementitious backfill will evolve as the pore fluids are flushed by incoming groundwater, from an initial value of pH 13 due to the presence of alkali metal hydroxide phases, to a period of pH 12

Waste Containers

KR 1 KR 3 Sea

Dilute Na–Cl–HCO3 brackish

Oxid. of org. matter Silicate + Calc. dissolution

Brack. Na–Cl–SO4

425

KR 23

KR 4 SO4 red. + DOC oxid. Calcite + pyrite ppt.

100 m HZ19 Carb. reduction Brack. Na–Cl

Methanogenesis

SO4 red. + CH4. oxid. Calcite + pyrite ppt.

300 m

Saline Na–Ca–Cl

HZ20

400 m

Carb. reduction Saline Ca–Na–Cl

Methanogenesis HZ21 900 m

Figure 2 Hydrogeochemical site model of the baseline groundwater conditions and main water–rock and microbial processes at Olkiluoto. KR1, KR3, KR4, and KR23 denote different boreholes and HZ19, HZ20, and HZ21 denote major reaction zones between compositionally different ground waters. Reproduced from Pastina, B.; Hella¨, P. Expected Evolution of a Spent Fuel Repository at Olkiluoto; Posiva Oy: Olkiluoto, Finland, 2006; POSIVA 2006-05.

controlled by Ca(OH)2 phases, and eventually to the background pH of the groundwater.23 5.17.2.1.5 Sulfur species

Pyrite and other sulfide minerals are found in many potential host-rock formations, as well as being present as an impurity in some of the clay materials that may be used to backfill and seal the repository.3,5,6,22 Oxidation of these minerals by atmospheric O2 during the operational and early postclosure phases will not only consume O2 but will also produce acid and a range of oxidized sulfur species, including thiosulfate 8,24 Sulfide and thiosulfate are particularly (S2O2 3 ). aggressive toward a number of candidate canister materials.7,8,10,11 In addition to sulfur-containing minerals in the host rock and backfill materials, the groundwater itself can be a source of sulfide. Most groundwaters contain sulfate which can be reduced by sulfate-reducing

bacteria to form sulfide.25 A number of deep groundwaters also contain natural background levels of sulfide ions.6,8 5.17.2.1.6 Redox conditions

There are two distinct types of redox behavior, determined by the location of the repository with respect to the water table. Because deep groundwaters are anoxic, for the majority of repository designs that are located below the water table, the redox conditions will become anaerobic after a brief period of aerobic conditions during which initially trapped atmospheric O2 is consumed.3,6,7,13,21,26 The transition from aerobic to anoxic conditions will have a significant impact on the corrosion behavior of the canisters. (Here, the term ‘anoxic’ is used to denote the absence of O2, whereas the term ‘anaerobic’ is used in a relative sense to describe a system that is less oxidizing.)

426

Waste Containers

For repositories located above the water table, the Yucca Mountain repository being the most obvious example, the system is essentially continuously aerated.1 This has a significant impact on material selection.27 5.17.2.1.7 Radiolysis and irradiation

High-activity wastes (SF and HLW) produce a significant g-radiation field that can penetrate the wall of the canister and produce both oxidizing and reducing species by the radiolysis of the surrounding pore or groundwater. However, canister designs generally comprise a sufficiently large wall thickness so that the external radiation absorbed dose rate is minimal (1 Gy h1).28 Radiolysis effects are generally insignificant for the corrosion of LILW canisters. 5.17.2.1.8 Degree of saturation

Repositories located above the water table are, by definition, permanently unsaturated. This has the advantage that the rate of transport of dissolved radionuclides is low, but does mean that the rate of transport of atmospheric O2 to the canister surface is rapid, resulting in permanently aerobic conditions and the need for highly corrosion-resistant materials.1,27 In addition, groundwater can evaporate on the canister surface resulting in highly concentrated aqueous phases.1 Even for repositories located below the water table, there will be a period of unsaturated conditions until the excavated workings have resaturated and/or until the thermal pulse from heat-generating wastes has subsided.29 5.17.2.1.9 Mass-transport conditions

Besides being important for the transport of radionuclides following canister failure, the rate of transport of oxidants to and of reactants away from the canister surface also affects the corrosion behavior.30 In bentonite-backfilled systems, mass transport will be by diffusion only. The steady-state mass-transfer coefficient (D/d, where D is the diffusivity and d is the diffusion layer thickness) in compacted bentonite is five to seven orders of magnitude lower than in bulk solution, with obvious implications for film formation and the supply of oxidants.31 Conversely, in nonbackfilled repository designs, the rate of mass transport can be rapid. 5.17.2.1.10 Microbial activity

In general, nuclear waste repositories are inhospitable environments for microbes.25 The combined effects of high temperature, radiation fields, saline

groundwaters and pore fluids, mechanical forces from swelling smectite clays, redox conditions, the lack of organic nutrients and terminal electron acceptors, and low water activity contribute to this lack of microbial activity. Although microbes can be identified that can survive extremes of each of these conditions, the combined effects of a number of environmental stressors will limit both the activity and diversity of the microbial population in the repository. 5.17.2.1.11 External loads and residual stress

Nuclear waste canisters will be subject to a number of sources of stress.5,6,13,26 Residual stress from HLW/SF canister fabrication and sealing is difficult to relieve by heat treatment because of thermal limits for the waste forms, although various nonthermal stress-relief methods are available.1 For bentonitebackfilled repository designs, external forces due to buffer swelling and hydrostatic pressures will exert loads of 5–10 MPa, depending on bentonite density and the depth of the repository.5,6,13 An additional 20–30 MPa hydrostatic load may occur during periods of glaciation. Lithostatic loads will result from the creep of the rock in the case on nonself-supporting host formations, such as salt.8 Canisters may also be exposed to shear loading due to seismic activity or, in nonbackfilled repository designs, rock impact from degradation of the excavated tunnels and rooms.1 5.17.2.2

Evolution of the Environment

Regardless of the repository design, the environmental conditions to which the canisters are exposed will evolve over time.5,6,13,21 For heat-generating wastes in repositories located beneath the water table, this evolution can be generally described as a shift from an initially warm and aerobic phase to a period of longterm cool, anoxic conditions (Figure 3).9,21 Even for permanently aerobic repositories in the unsaturated zone, environmental conditions will become more benign as the heat from the canisters dissipates throughout the host rock.1 Other aspects of the repository environment will also change over time, such as the degree of saturation or the composition of the backfill pore fluids (Figure 4).29 This evolution in the repository environment has profound implications for the corrosion behavior of the canisters.7 Just as the environmental conditions will evolve, so too will the corrosion mechanisms to which the canisters are subjected. In general terms, relatively rapid forms of localized corrosion or SCC are most likely during the initial warm, aerobic phase.

Waste Containers

Vault redox conditions

Oxidizing

Container temperature (⬚C)

100

Warm, oxidizing period

427

Cool, nonoxidizing period

Temperature

80

60

40

20

[Oxygen]

Nonoxidizing 10

102 103 Time since emplacement (year)

104

Figure 3 Schematic illustration of the evolution of the repository environment from an initial warm, oxidizing phase to an eventual cool, nonoxidizing period.

Dry-out phase Redistribution of initial moisture content Poor thermal conduction No swelling of sealing materials Dissolution and reprecipitation of minerals No deliquescence No corrosion No gas generation No O2 consumption No microbial activity

Transition phase

Saturated phase

Gradual saturation of bentonite Improving thermal conduction Development of swelling pressure Dissolution of minerals Deliquescence of precipitated salts Corrosion due to formation of thin surface water film Gas generation possible for nonnoble canister materials once O2 depleted O2 consumption No microbial activity

Complete saturation of bentonite Good thermal conduction Swelling pressure fully developed Gradual equilibration of pore water with groundwater Corrosion in fully saturated bentonite Gas generation possible for nonnoble canister materials once O2 depleted and for Cu/steel container once outer shell breached O2 completely consumed Onset of anoxic conditions No microbial activity

Time Figure 4 Summary of some of the processes accompanying the evolution of the environment in a bentonite-backfilled repository during and after the thermal transient.

As the environmental conditions become more benign, so too will the forms of corrosion, with relatively slow general corrosion replacing the rapid localized attack. This is an important consideration since if it is possible to design a canister that will remain intact for the duration of the relatively short early transient period, then there is a possibility of achieving long-term containment. Although most forms of corrosion become more benign as the environment evolves, it can be argued

that various forms of MIC and hydrogen-related degradation become more likely.25,32

5.17.3 Material Selection A number of factors come into the decision about which material(s) should be considered for the fabrication of nuclear waste canisters, some of which are discussed briefly below. Table 2 summarizes the

428

Waste Containers

Table 2

Reference and alternative candidate canister materials in various national nuclear waste programs Waste type

Reference canister material(s)

Alternative canister material(s) considered

References

UK UK Germany

ILW HLW (SF) HLW/SF HLW/SF

Belgium France Switzerland Sweden

HLW (SF) HLW/SF HLW/SF SF

BS4360 C-steel C-steel, copper, Ti alloys, Ni–Cr–Fe alloys Inconel 625, Incoloy 825, Cu10Ni, Cu30Ni, cast steel, cast iron AISI 316L, Hastelloy C-22, oxygen-free copper AISI 316L hMo Undefined passive alloy Copper Ti alloy

17 18 8

Spain

AISI 304L/316L – TStE 355 C-steel, Hastelloy C-4, Ti 99.8-Pd, pure copper TStE 355 C-steel

Finland

SF

–

6,11

Japan Canada USA

HLW SF SF/HLW

Ti alloys, copper Ti alloys C-steel, various Ni–Cr–Mo alloys

2 9,14,35–38 1,7,27,39–46

C-steel C-steel C-steel Copper, cast iron structural support Copper, cast iron structural support C-steel Copper, C-steel Alloy 22, Ti-7 drip shield

reference and alternative canister materials in a number of national nuclear waste programs. 5.17.3.1

Target Lifetime

A primary consideration when selecting a canister material is the desired period of containment.9 A number of materials can be expected to provide containment for a period of 300 years, the length of time generally associated with the period of maximum radiological hazard (since it corresponds to ten half-lives of 137Cs). However, if ‘indefinite’ containment is required (generally taken to be a period of 105–106 years) to ensure the safety of the entire disposal system, then only materials that may become thermodynamically stable (e.g., copper) or which provide excellent corrosion resistance (e.g., certain Ti and Ni–Cr–Fe alloys) are suitable candidates. 5.17.3.2

Active Versus Passive Materials

Materials can be generally classified as being either passive or active depending on whether they form a highly protective passive film or whether they dissolve actively under repository conditions.7 Examples of active materials are pure copper and C-steel in chloride-dominated environments at near-neutral pH. The range of passive materials that have been considered as canister materials include stainless steels, a-titanium alloys, certain Ni–Cr–Fe alloys, and, in the presence of a cementitious backfill, C-steel. Active materials will tend to corrode uniformly and are less subject to localized forms of corrosion. Passive materials, on the other hand,

8 4,8,19 20 13 5,10,26,33,34

exhibit slow or negligible rates of general corrosion but can be subject to rapid forms of localized attack under certain conditions. An associated consideration is the robustness of any prediction of the future corrosion performance of the canister. In this regard, there has been a preference to select active materials since there is more confidence in making long-term predictions of materials that undergo general corrosion. Although significant advances have been made in recent years in our mechanistic understanding of, and the development of predictive models for, localized corrosion and of environmental-assisted cracking, there is still a reluctance to accept that we can make justifiable longterm predictions for materials that are susceptible to localized attack or cracking. 5.17.3.3

Environmental Considerations

The environmental conditions to which the canisters will be exposed clearly influence the choice of material. Stainless steels are clearly unsuitable in nearneutral pH chloride-rich environments because of the risk of localized corrosion and SCC.47,48 The thermodynamic immunity enjoyed by copper in anoxic environments is compromised by sulfide ions, in the presence of which copper will corrode with the evolution of hydrogen.10,11,49 In permanently aerobic environments, highly corrosion-resistant materials are required, such as Ni–Cr–Fe and crevicecorrosion-resistant Ti alloys.1,27,42,50 However, it is always necessary to consider the nature of the interfacial environmental conditions

Waste Containers

when selecting a canister material. For example, copper may be perfectly suitable for use in sulfidecontaining groundwaters if the rate of supply of sulfide to the canister surface is limited by the presence of highly compacted bentonite.5,6,26 Similarly, stainless steels may be used in chloride environments without threat from localized corrosion or SCC provided the pH is alkaline and/or the environment is anoxic.15,16,51 It is possible, therefore, to modify the near-field environment to improve the canister corrosion performance through appropriate repository design and engineering.

5.17.4 Corrosion Behavior of Candidate Canister Materials In a given environment, a given canister material will be subject to one or more corrosion mechanisms. The corrosion mechanisms investigated in the various national programs are summarized in Table 3 and discussed in more detail below for each (class of ) canister material. 5.17.4.1

Carbon Steel

Carbon steels are alloys of iron and carbon with a C content of <2 wt% and typically contain Mn (<1.65 wt%), Si (<0.60 wt%), Cu (<0.60 wt%), and small amounts of other alloying elements, such as Cr, Ni, Mo, W, V, and Zr.68 The alloys proposed as candidate canister materials fall into the subset of alloys referred to as low-carbon (0.05–0.15 wt% C) or mild (0.16–0.29 wt% C) steels. In some cases, cast steels or cast iron (an iron–carbon alloy containing Table 3

UK Germany Spain Belgium France Switzerland Sweden Finland Japan Canada USA a

429

>2 wt% C and 1–3 wt% Si) have been proposed either as canister materials or, in the latter case, as structural inserts inside a primary corrosion barrier.8,26 Carbon steels have been proposed for use in a number of countries (Table 2). Although specific grades have been defined in some programs, the majority of countries are still at a relatively early stage of investigation, and refinement of the chemical, microstructural, and physical properties of the alloy is still required. Although it was at one time considered as a corrosion barrier for the permanently aerobic Yucca Mountain repository,39 C-steel is best suited as a canister material in repositories located in the saturated zone which will become anoxic. Not only can the rate of corrosion of C-steel be excessive in the presence of O2 (tens to hundreds of micrometers per year depending upon salinity and the rate of mass transport), but localized corrosion and certain forms of SCC are also more likely under aerobic rather than anaerobic conditions.14 Although there will always be a period of aerobic conditions even in repositories located beneath the water table, the extent of the oxidizing phase will be limited and the resultant corrosion minimal. Once the initially trapped O2 has been consumed, the rate of general corrosion will be lower and the probability (and potential severity) of localized corrosion and SCC much diminished. However, anaerobic corrosion will produce hydrogen and the consequences of gas production for both the material itself and the performance of the other engineered and natural barriers need to be considered. Figure 5 shows the potential-pH (Pourbaix) diagram for the Fe–H2O system at 25  C.69 Iron, and by

Corrosion processes for reference canister materials in various national nuclear waste programs Reference canister material(s)

Corrosion processes considereda

References

AISI 304L/316L TStE 355 C-steel TStE 355 C-steel C-steel C-steel C-steel Copper Copper C-steel Copper C-steel Alloy 22 Ti-7 drip shield

Atmospheric corrosion, pitting, SCC, general corrosion General corrosion, pitting, SCC, intergranular attack (IGA) General corrosion, localized corrosion, SCC, IGA General corrosion Oxidation, general corrosion General corrosion, localized corrosion, SCC, HIC General corrosion, localized corrosion, SCC, MIC General corrosion, localized corrosion, SCC, MIC General corrosion, localized corrosion, SCC, HIC General corrosion, localized corrosion, SCC, MIC General corrosion, localized corrosion, SCC, HIC, MIC General corrosion, crevice corrosion, SCC, MIC General corrosion, HIC, SCC

15–17,51–54 8 8 4,8,19,55 20,56 3,13,32,57,58 5,10,26,59 6,11 2,60–62 25,29,35–37,63–65 14,25,66 1,7,25,27,40–45,67 1,12,17

Corrosion processes considered to be most important are given in italic.

430

Waste Containers

−2 2.2

−1

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

2

2

1.8

1.8

9⬘

1.6

1.6

20

1.4 1.2

b

1.4

0 −2 −4 −6

Fe3+

2⬘

FeO42−

10⬘

1.2

1 0.8

1 FeOH2+ 4⬘

0.4 E(V)

0.8

3⬘

5⬘

0.6

11⬘

0.6

Fe(OH)+2

28

0.4

6⬘

Fe2O3

0.2 0

a

Fe2+

0.2

7⬘

−0.2 −0.4

16 2.2

0

8⬘

−0.2

26 23

−0.6

0 −2 −4 −6

−0.4

17

−0.6

Fe3O4 13

−0.8

27 −6 −0.8 HFeO2 −1 24 −6 −1.2

−1 1⬘

Fe

−1.2 −1.4

−1.4

−1.6

−1.6

−1.8 −2

−1

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

−1.8 16

pH Figure 5 Potential-pH (Pourbaix) diagram of the iron–water system at 25  C considering Fe2O3 and Fe3O4 as the stable Fe (III)- and Fe(II)-containing solid phases. Lines a and b represent the equilibria between water and hydrogen and water and oxygen, respectively, for unit fugacity of the gaseous species. The 0, 2, 4, and 6 indicate the equilibrium conditions between the various solid and dissolved phases for activities of 0, 102, 104, and 106, respectively. Reproduced from Pourbaix, M. Atlas of Electrochemical Equilibria in Aqueous Solutions, 2nd ed.; NACE International: Houston, TX, 1974.

inference C-steel, corrodes at all pH and redox conditions associated with the various repository environments. Furthermore, solid corrosion products are predicted to be stable over a wide range of conditions, with Fe(III) corrosion products stable under the relatively oxidizing conditions associated with the presence of O2 (represented by Fe2O3 in Figure 5) and Fe(II)-containing corrosion products (here represented by Fe3O4) stable under anaerobic conditions. Magnetite, in particular, is thermodynamically stable over a wide range of pH in anaerobic environments, extending from the near-neutral pH conditions associated with bulk groundwater or bentonite pore water to the alkaline range representative of cement pore waters.

The thermodynamic tendency for C-steel to corrode under anaerobic conditions has significant implications since it involves the formation of H2.2,13,32,57,70 The most likely reaction scheme for the formation of Fe3O4 involves the initial formation of ferrous hydroxide followed by the transformation of Fe(OH)2 via the Schikkor reaction Fe þ 2H2 O ! FeðOHÞ2 þ H2

½I

3FeðOHÞ2 ! Fe3 O4 þ 2H2 O þ H2

½II

A large fraction of the hydrogen that forms from reactions [I] and [II] will be evolved as gaseous H2. The generation of gaseous H2 in a repository that is,

Waste Containers

431

Corrosion rate (mm year–1)

30

20

10

0

0

1

2

3

4

5

Time (years) Figure 6 Time dependence of the anaerobic corrosion rate of carbon steel in compacted bentonite saturated with synthetic seawater or groundwater. Reproduced from Taniguchi, N.; Kawasaki, M.; Kawakami, S.; Kubota, M. Corrosion behaviour of carbon steel in contact with bentonite under anaerobic condition. In Prediction of Long Term Corrosion in Nuclear Waste Systems, Proceedings of the 2nd International Workshop, Nice, France, Sept 2004; pp 24–34, European Federation of Corrosion and Andra.

by design, meant to be sealed can be problematic.3 In nonbackfilled repositories, typically those for the disposal of LILW, the concern is that the large volumes of metallic and organic wastes (and metallic container materials) may result in the formation of excessive gas pressures and either the enhanced release of gaseous radionuclides or structural damage to the host rock.71 The high organic content of LILW results in significant microbial activity and the conversion of H2 and CO2 to methane by methanogens. In bentonite-backfilled HLW/SF repositories, even small amounts of anaerobic corrosion can result in the formation of a H2-gas phase because the permeability of the compacted bentonite is so low. Hydrogen is likely to be transported through the bentonite by two-phase flow involving both dissolved and gaseous H2. A similar transport mechanism is likely to occur in low-permeability host rocks, such as the sedimentary Opalinus Clay. The rate of general corrosion determines not only the canister lifetime (and, hence, the required wall thickness for a given target lifetime), but also the rate of H2 generation (under anaerobic conditions). As noted above, the rate of aerobic corrosion of C-steel in saline environments can be of the order of tens to hundreds of micrometers per year.14 If the amount of available O2 is limited, as in a repository located beneath the water table, the rate of aerobic corrosion is not particularly important since the total amount of corrosion is determined by the initial oxygen inventory and is typically limited to a few hundred micrometers.

Of greater interest is the rate of anaerobic corrosion, since it is this rate that primarily determines the lifetime of the canister (in the absence of localized corrosion or environmentally assisted cracking). Figure 6 shows the measured time dependence of the anaerobic corrosion rate of C-steel in compacted bentonite saturated with either synthetic seawater or simulated groundwater.72 The corrosion rate decreases with increasing exposure time due to the formation of a protective corrosion product layer. The long-term corrosion rate is of the order of a few micrometers per year, although steady state has not been achieved even after 4-year exposure. In bulk groundwater solution at near-neutral pH, the corrosion rate reaches an apparent steady state after a period of 4–6 months exposure and is of the order of 0.1–1.0 mm year1.5,7,70 In simulated cement pore-water solutions, the anaerobic corrosion rate of C-steel is even lower, being of the order of 0.01–0.1 mm year1, due to the stability of a passive Fe3O4 film.8,16 The nature of the corrosion products on C-steel vary depending upon the redox conditions and the composition of the aqueous phase (Figure 7). Regardless of the environmental conditions, the initial phase formed is precipitated ferrous hydroxide, Fe(OH)2. Under aerobic conditions various forms of ferrioxyhydroxide (a-FeOOH, goethite; g-FeOOH, lepidocrocite; and, more rarely, b-FeOOH, akaganeite) develop, although in the presence of Cl, sulfate, and carbonate ions, various forms of Green Rust

432

Waste Containers

Fe

103

+HCO3–

Fe(OH)2 +O2 GR1, GR2

α-, (β-), γ-FeOOH

FeCO3, (Fe,Ca)CO3

Anaerobic (Schikkor reaction)

+Fe(II)

47 kinds of soils 1–18 years 1350 data

102

101

Fe3O4 Long-term slow oxidation

α-, γ-Fe2O3

Pitting factor

Data in various soils (Romanoff, 1989)

γ-Fe2O3

Figure 7 Summary of the formation and transformation of corrosion products on carbon steel in aerobic, anaerobic, and carbonate-containing environments. GR1 and GR2 represent Green Rust 1 and Green Rust 2, respectively.

are the more-likely corrosion products. Under anaerobic conditions, the Fe(OH)2 is transformed to Fe3O4 via the Schikkor reaction [II]. Magnetite can also form from the transformation of ferrioxyhydroxide in the presence of Fe(II), a process that will occur during the transition from aerobic to anaerobic conditions during the evolution of the repository environment. In the presence of compacted bentonite, the pore water of which will contain bicarbonate ions due to the dissolution of calcite, or in CO2-saturated systems (as may occur in a LILW repository), the primary anaerobic corrosion product will be FeCO3 (with possible incorporation of other cations such as Ca2þ). The surface of C-steel canisters will be subject to a certain degree of localized attack.2,13,32 In nearneutral pH environments, such as in compacted bentonite or bulk groundwater environments, the surface is protected by a layer of precipitated corrosion products, but the surface film is not sufficiently protective as to be deemed passive in the classical sense as is the case for stainless steel or titanium alloys. Instead, the precipitated corrosion product layers tend to be spatially heterogeneous or ‘patchy,’ resulting in localized dissolution and the formation of low aspect ratio ‘pits,’ that is, local asperities with a width that greatly exceeds the depth of corrosion. Any localized penetration tends to degenerate into uneven surface roughening as corrosion progresses. This evolution toward more uniform corrosion with increasing depth (or exposure time) can be seen in Figure 8 in which the pitting factor (the depth of the maximum penetration divided by the average depth of corrosion) is plotted against the average depth.2 This form

100 –5 10

10–4

10–3 10–2 10–1 100 101 Average corrosion depth (mm)

102

Data in various soils (surface area, 182–364 cm2) Results in carbonate/chloride solution (surface area, 182 cm2) Results in carbonate/chloride solution (surface area, 364 cm2) Figure 8 Depth dependence of the pitting factor for carbon steel exposed to soil and simulated repository environments. Reproduced from JNC. Repository Design and Engineering Technology, Supporting Report 2; Japan Nuclear Cycle Development Institute, 2000.

of surface roughening is a consequence of different degrees of protection of the surface as a result of the spatially inhomogeneous corrosion product and may continue even under anaerobic conditions. Localized corrosion in the classical sense, that is, spatially separated but electrochemically coupled anodic and cathodic reactions, can only occur during the aerobic phase in the presence of O2 or, possibly, in the presence of Fe(III). Permanent separation of anodic and cathodic reactions is only likely to occur if the surface film is both passive and also semiconducting so that it can support the cathodic reaction. These conditions are most likely to be met under more-alkaline conditions when the surface is protected by a passive Fe3O4 film. However, the pH of fresh cement pore waters is too alkaline and permanent film breakdown is unlikely even in highly saline waters.16 Although detailed analyses have not been performed, it is likely that the initially trapped O2 will be consumed before the pore-water pH decreases due to the flushing of the alkaline cement minerals. Such an analysis requires knowledge of the critical [Cl]:[OH] ratio above which film breakdown is possible. Carbon steels are known to be susceptible to SCC in a range of environments, although the possibility of

Waste Containers

SCC in repository environments seems remote.14 Many of the environments in which C-steel has been shown to be susceptible to SCC, such as phosphate and nitrate solutions, high-temperature water (>250  C), and environments containing CO–CO2–CH4, are irrelevant for nuclear waste repositories. Still others, such as carbonate/bicarbonate and caustic environments, are unlikely to cause cracking either because of the absence of cyclic loading which appears to be necessary in the former case or, in the latter case, because the caustic concentration in cement pore waters is less than those in which cracking has been observed. However, prudence dictates that a thorough assessment of the possibility of SCC should be carried out in these latter cases. Certainly, engineering solutions, such as selecting an SCC-resistant grade (if such exists) or minimizing tensile residual stress through a suitable nonthermal stress-relief process, should also be sought. Hydrogen will be absorbed by the canister during anaerobic corrosion.2,58,66 Hydrogen can cause a number of failure mechanisms in steels but these generally occur either in aggressive environments (e.g., sour systems containing high concentrations of H2S and CO2) or in high-strength materials. The most likely forms of hydrogen degradation for C-steel in repository environments are blister formation and hydrogen-induced cracking (HIC). However, even these forms of degradation seem unlikely because58,66:
repository environments are relatively benign, resulting in low diffusible H concentrations in the steel,
the materials proposed for use as nuclear waste canisters are generally low strength, with maximum yield strengths of the order of 500–600 MPa,
the maximum applied stress (30–40 MPa) is relatively low,
the surface residual stress can be mitigated by a suitable nonthermal stress-relief treatment,
surface defects that may act to concentrate stress, and hence H, can be machined following the final closure weld, and
the canister material and the design of the final closure weld can be selected to minimize H effects. One location on the canister for which it is not possible to mitigate the residual stress or notches or crack-like defects is the inner surface of the final closure weld.58 Over time, the H2 partial pressure in the canister will achieve the same level as that on the outside of the canister (perhaps 6–8 MPa). It is

433

possible, therefore, that cracking could initiate from the inside of the canister. This possibility needs to be taken into consideration when designing the final closure weld and when selecting the canister material. Carbon steels offer a number of advantages and disadvantages as canister materials for the disposal of nuclear waste. The advantages include18:
good corrosion characteristics, corroding actively in compacted bentonite or under passive conditions in a cementitious backfill with minimal impacts of localized corrosion, SCC, or H-related degradation;
sufficient structural strength that simple singleshell canister designs are possible;
extensive experience in fabricating and sealing large C-steel structures, with a corresponding large number of suppliers;
canister lifetimes of thousands to tens of thousands of years are achievable, with robust predictive models supported by archaeological artifacts;
suitable for use in a number of repository designs. The major disadvantage of the use of C-steel is that the products of anaerobic corrosion, H2 and Fe(II), can adversely affect the properties and performance of other barriers. The effects of H2 have been discussed above. Ferrous species can cause alteration of bentonite to a nonswelling, and hence nonsealing, illitic clay structure.73 5.17.4.2

Stainless Steels

Stainless steels are Fe–Cr alloys containing a minimum of 11–13 wt% Cr to impart ‘stainless’ properties.47 A wide range of alloying additions are possible resulting in an equally wide range of properties. Stainless steels are generally classified according to their method of manufacture (wrought, cast, or precipitation hardened), their microstructure (hence, ferritic, austenitic, martensitic, and duplex ferritic–austenitic alloys), and, within these categories, according to their overall corrosion resistance (hence austenitic and superaustenitic or duplex and super-duplex alloys). Figure 9 shows the relationship between the different classes of stainless steel.47 Stainless steels have not been widely considered as a canister material for HLW/SF because of the poor performance of many austenitic alloys in warm chloride environments. Chloride is ubiquitous in deep groundwaters and, even if it is present in only small concentrations (a few mg g1), evaporative concentration on the warm canister surface can lead to localized

434

Waste Containers

Superferritic stainless steels

Ni–Cr–Fe alloys

Add Cr, Mo

Add Ni for corrosion resistance in high temperature environments

430

No Ni, ferritic

347

Add Nb + Ta to reduce sensitization

Add Cr and Ni for strength and oxidation resistance

316L 317L

Superaustenitic stainless steels

304 Fe–19Cr–10Ni Add Mo for pitting resistance

304L

Add S or Se for machinability

Duplex stainless steels

309, 310, 314, 330

Add Ti to reduce sensitization

321

303, 303 Se

Lower C to reduce sensitization

Increase Cr, lower Ni for higher strength Precipitation hardening stainless steels

Add Cu, Ti, Al, lower Ni for precipitation hardening

Add Mn and N, lower Ni for higher strength 201, 202

316 Add more Mo for pitting resistance

Add Ni, Mo, N for corrosion resistance

317

No Ni addition lower Cr, martensitic

403, 410, 420

Figure 9 Compositional relationship between different classes of stainless steels. Reproduced from Sedriks, A. J. Corrosion of Stainless Steels, 2nd ed.; Wiley: New York, NY, 1996.

corrosion (crevice corrosion and/or pitting) and SCC.47,48 Type 316L austenitic stainless steel was considered for the Boom Clay repository in Belgium (Table 2) with a pore-water Cl concentration of 27 mg g1 (Table 1), but was rejected because of localized corrosion caused by thiosulfate ions (S2O2 3 ) formed from the oxidation of pyrite.8,24 Where stainless steel has found acceptance is for the disposal of nonheat-producing LILW in cement-backfilled environments (Table 2).17 Overall, however, it is probably accurate to say that insufficient consideration has been given to the entire range of stainless steels, particularly the more corrosion-resistant and duplex alloys, primarily because of the perceived (and real) problems with the common austenitic alloys. Stainless steels are ‘stainless’ because they are protected by a Cr(III)-containing passive film (Figure 10). The passive film is stable over a wide range of pH representative of those in natural waters or cement pore water and over a range of reducing and oxidizing potentials. However, the film is unstable in acidic solutions (as may form in pits or crevices), dissolving

as Cr3þ, and at more-oxidizing potentials, dissolving as Cr(VI). The range of passivity is reduced in the presence of Cl, other halides, and S2O2 3 and by increased temperature.24,47,48 The susceptibility to localized corrosion and SCC is diminished by sulfate, carbonate, and nitrate ions. Because of the passive nature of stainless steels, the rates of general corrosion are typically low.8,15–17,51 Rates tend to be higher in aerobic environments than in the absence of O2 and increase with decreasing pH, increasing [Cl], and increasing temperature. Even under aerobic conditions in near-neutral pH environments, however, the rate of general corrosion is of the order of 1 mm year1 or less. In simulated cement pore water, the corrosion rate can be up to a factor of 100 times lower.18 Localized corrosion of stainless steels takes the form of either crevice corrosion of physically occluded regions or the pitting of exposed surfaces.48 Crevice or pit initiation is accompanied by film breakdown and the transpassive dissolution of the film as Cr(VI). Passivity breakdown can be studied

Waste Containers

2 1.8 log C = 0 1.6

H2CrO4

1.4

1

0

–1

–2 2.2

7⬘

2

4

3

7

6

5

8

9

10 11 12 13 14 15 16

–2 –4 0 –6

Cr2O27 –

6⬘

16⬘

1.6

11⬘

9⬘

HCrO4–

1.8

1.4

b

CrO42–

1.2

1

1 53⬘ 52

0.8 0.6 Cr3+

E (V)

0.8

?

18⬘ 17⬘

0.4

0.6 19⬘

0.4

54

0 –2 –4 –6

27⬘

0.2

0.2

26⬘

a

0

34

1⬘

0

–0.2

–0.2

Cr (OH)3

–0.4

28⬘

12⬘

−6

–0.6

–1

2.2 2

–2 –4 –6

1.2

–0.8

435

Cr2+ 0 –2 –4 –6

38

47

–0.6 –0.8

39

15⬘

–1.2

CrO23– 30

–1.4

5⬘

–1

CrO–3

42

0

1

2

3

4

–1.2 –1.4

Cr

–1.6 –1.8 –2 –1

–0.4

5

6

7 8 pH

9

–1.6

–1.8 10 11 12 13 14 15 16

Figure 10 Potential-pH (Pourbaix) diagram of the chromium–water system at 25  C considering Cr(OH)3 as the stable Cr(III)-containing solid phase. Lines a and b represent the equilibria between water and hydrogen and water and oxygen, respectively, for unit fugacity of the gaseous species. The 0, 2, 4, and 6 indicate the equilibrium conditions between the various solid and dissolved phases for activities of 0, 102, 104, and 106, respectively. Reproduced from Pourbaix, M. Atlas of Electrochemical Equilibria in Aqueous Solutions, 2nd ed.; NACE International: Houston, TX, 1974.

electrochemically by polarizing the sample to increasingly positive (more-oxidizing) potentials. Figure 11 shows a series of cyclic voltammograms in which the potential of the electrode is scanned first in the positive direction and then in the reverse direction toward the starting value.1 Passivity breakdown is characterized by a rapid increase in current (a measure of the rate of dissolution), with the potential at which it occurs referred to as the pitting or breakdown potential (EB).48 Upon reversing the potential scan, the measured current is higher because the film is no longer protecting the surface. The potential at which the current is the same on the forward and reverse scans is referred to as the repassivation potential (ERP) and represents the potential at which a pit or crevice no longer propagates. Breakdown and repassivation potentials can be defined for both crevice corrosion and pitting by using either occluded or planar samples, respectively.

The resistance to localized corrosion can be increased by increasing the Cr, Mo, and N contents.47 The resistance of a particular alloy can be estimated based on the pitting resistance equivalent number (PREN) which is given by Sedriks47 PREN ¼ %Cr þ 3:3ð%MoÞ þ 16ð%NÞ

½III

A critical temperature exists above which an alloy is susceptible to crevice corrosion or pitting for a given environment. Figure 12 shows a so-called corrosion map showing the dependence of the critical temperature on Cl concentration for a range of alloys, including: the austenitic Type 316L (UNS S31603, with a typical PREN of 27), duplex grade 2205 (S31803, PREN ¼ 34), and two higher austenitic alloys 904L (N08904, PREN ¼ 36) and 254 SMO (S31254, PREN ¼ 44).74 Theoretically, a given alloy is immune to localized corrosion if the combination of temperature and Cl concentration lies to the left of the

436

Waste Containers

1500 316L stainless steel

Greater than 1500 mV without breakdown

Titanium grade 7 Alloy 22

Ercrev 500 ΔE Alloy 22

0

Ecorr

00264DC_LA_0780c.ai

Potential (mV vs. Ag/AgCl)

1000

Ecorr –500

Ecorr

10–10

10–8

10–6 Current density

10–4

10–2

(A cm–2)

Figure 11 Potentiodynamic scans for Type 316L stainless steel, titanium grade-7, and Alloy 22 in concentrated chloride solution. Reproduced from DOE. Yucca Mountain Reposiotry license application; US Department of Energy, DOE/RW-0573; 2008.

the critical temperature decreases with increasing Cl concentration as would be expected and that the critical pitting temperature is higher than the critical crevice temperature. Furthermore, the localizedcorrosion susceptibility is seen to follow the trend predicted by the PREN. An alternative method for assessing susceptibility to localized corrosion is to compare the value of the corrosion potential (ECORR) with EB or ERP.48 The condition for film breakdown under naturally corroding conditions is that

80

Temperature (⬚C)

70 60 50 40 30 20 100

10 000 1000 Chloride concentration (µg g–1) 316L (S31603) crevice 316L (S31603) pitting 2205 (S31803) crevice 2205 (S31803) pitting

1 00 000

904L (N08904) crevice 904L (N08904) pitting 254 SMO (S31254) crevice 254 SMO (S31254) pitting

Figure 12 Critical temperature for the crevice corrosion and pitting of various grades of stainless steel as a function of chloride concentration. Reproduced from ASM. ASM Handbook, Volume 13B, Corrosion: Materials; American Society for Metals International: Metals Park, OH, 2005.

respective line, but immunity is also a function of the nature and duration of the test used to determine the critical temperature. The figure shows that

ECORR > EB

½1

To maintain conservatism, the criterion for localized corrosion is often based on the repassivation potential ECORR > ERP

½2

Clearly, as the repository becomes more anaerobic, the probability of localized corrosion diminishes as ECORR shifts to more-negative values. The susceptibility of stainless steels to localized corrosion is also a function of the [Cl]:[OH] ratio.51 The film breakdown potential is observed to increase significantly above pH 10–11.47 Pitting of Type 304L stainless steel was only observed at a

Waste Containers

300

SCC

Temperature (⬚C)

250

SAF 2507 No cracking

200

SAF 2205 SAF 2304

150

100 AISI 304/304L 50

AISI 316/316L No SCC

0

0.0001 0.001

0.01

0.1

1

10

Chloride concentration (%) Figure 13 Corrosion map for the susceptibility of stainless steels to stress corrosion cracking in aerated chloride environments as a function of temperature. Reproduced from Sedriks, A. J. Corrosion of Stainless Steels, 2nd ed.; Wiley: New York, NY, 1996.

temperature of 60  C in simulated cement pore water for Cl concentrations >50 000 mg g1, which represents a [Cl]:[OH] ratio of 29 at pH 13.8 The Mo-containing Type 316L alloy was even more resistant. Localized corrosion of stainless steel in contact with cement backfill is only likely, therefore, once the pore-water pH has decreased due to flushing of the alkaline phases from the cement, but then only if there is sufficient O2 present that the criterion represented by eqns [1] or [2] is met. The susceptibility of stainless steels to SCC is closely related to their localized-corrosion resistance because pits and crevices act as locations for crack initiation.47,48 Figure 13 shows an SCC susceptibility map for various austenitic and duplex stainless steels.47 The austenitic 304 and 316 alloys are clearly not suitable for use as HLW/SF canisters because of the elevated temperatures (on the assumption that the possibility of canister failure by SCC is unacceptable). The ferritic–austenitic duplex alloys 2304, 2205, and 2507 are significantly more resistant to SCC (and to localized corrosion, Figure 12), although they have not been widely considered for this application. Stainless steels possess a number of positive attributes as candidate canister materials, but also exhibit a number of potential weaknesses.18 The strengths of stainless steels include:

437


low rates of general corrosion, especially under anaerobic or alkaline conditions, permitting the use of thin-walled canisters;
a wide range of alloying options and corresponding mechanical and chemical (corrosion) properties;
extensive experience in fabricating, sealing, and inspecting stainless steels vessels with a correspondingly large number of possible suppliers; and
minimal impact on other barriers in the repository. However, these attributes are countered by the perceived susceptibility of stainless steels to localized corrosion and SCC in repository environments. Certainly, the common austenitic grades are unsuitable as canister materials for HLW/SF, although they might be acceptable when used in conjunction with a cementitious backfill. Furthermore, study of the use of duplex alloys seems warranted. 5.17.4.3

Copper

Copper was the first material to be proposed for use as a SF canister material34 and its suitability is demonstrated by the fact that it continues to be the reference material in Sweden, Finland, and (for crystalline rock) in Canada (Table 2).5,6,10,11,38 Copper has also been investigated in the Spanish and Japanese programs and even for the salt repository program in Germany.2,8 The main characteristic of copper that makes it suitable as a canister material for the disposal of HLW/SF is that, in the absence of sulfide, it is thermodynamically stable in deep groundwaters or bentonite pore water (Figure 14). As a consequence, once the initially trapped O2 and any Cu(II) species that are produced by the oxidation of Cu(I) by O2, have been consumed then corrosion will theoretically cease.10,11 Although there are a large number of copper alloys, the main focus of research activities has been oxygen-free copper. In Sweden, Svensk Ka¨rnbra¨nslehantering AB (SKB) have developed specific requirements and propose an alloy referred to as oxygen-free copper with added phosphorus (OFP copper), where the P is added to improve the creep rupture strain of the material.5,26 Because of the high ductility of copper, canisters must include an internal structural support to withstand the external load, which is typically either cast iron or C-steel.5,6 Apart from its thermodynamic stability in O2-free water, the other major attribute of copper as a candidate canister material for HLW/SF is the fact that it will tend to corrode uniformly in the repository environment with little tendency for localized corrosion

438

Waste Containers

1.2

−2 −4

0 (b)

−6

−6 −4 Cu2O3 Hydr.

CuCl+ 0.8

CuCl2 •3Cu (OH)2 −6 0

0.4

HCuO2- CuO22

CuCl E (VSHE)

−4

CuO

CuCl2-

0

−2 −4

Cu2O

−6 −0.4

Cu

−0.8

−1.2

0

2

4

(a)

8

6

10

12

14

pH Figure 14 Potential-pH diagram for the Cu–H2O–Cl system at 25  C with 1 mol dm3 Cl. The shaded box represents the range of corrosion potentials observed in laboratory experiments in O2-containing solutions.

or SCC.10,11 Copper dissolves actively in Cl environments to form soluble complexes, such as CuCl 2 and CuCl2 3 . The Pourbaix diagram in Figure 14 shows that the solubility of these complexes exceeds 102 mol dm3 in acidic 1 mol dm3 Cl solution with the formation of solid CuCl at higher concentrations. At near-neutral pH, Cu2O is the thermodynamically stable solid phase and will precipitate at increasingly lower dissolved copper concentrations as the pH increases. However, these reactions do not occur spontaneously in the absence of O2. Oxygen will be present during the initial period in the evolution of the repository environment.10,11 Figure 15 shows the various electrochemical, chemical, precipitation/dissolution, adsorption/desorption, redox, and mass-transport processes involved in the corrosion of copper in compacted bentonite in the presence of O2 (and sulfide, see below). In sulfide-free environments, the anodic dissolution of copper as CuCl 2 is coupled to the cathodic reduction of O2 on the canister surface. The relative rates of these reactions establish the value of ECORR, the range of measured values of which in aerobic and anaerobic solution are shown by the shaded box in Figure 14.31 Three of these measurements are shown in Figure 16, where the atmosphere above the solution has been gradually changed to make the environment less aerobic.10,11 These changes in the partial pressure of O2 in the experiments (up to

point ‘D’ on curves (a), (b), and (c)) simulate the evolution of the repository environment as the initial aerobic phase transitions to the long-term anoxic period. Even in deaerated solution (point ‘D’), the value of ECORR lies above the H2O/H2 equilibrium line in Figure 14, indicating that H2O is not reduced with the evolution of H2. (Note the difference in potential scales in Figures 14 and 16. The most-negative steady-state potential in Figure 16 of 0.4 VSCE is equivalent to a potential of 0.16 VSHE on the scale used in Figure 14. The pH of the tests in Figure 16 was pH 7.) The other processes shown in Figure 15 have been discussed in detail elsewhere.10,11 Some researchers have questioned the conventional wisdom that copper is thermodynamically stable in O2-free water.75–77 On the basis of a series of experiments, it has been claimed that not only does copper corrode in O2-free water, but that the rate of corrosion exceeds that observed in the presence of O2. Evidence is presented for both evolved H2 gas and large amounts of H apparently absorbed into the copper. It is proposed that a previously unrecognized copper hydroxide species is sufficiently stable that the potential of the copper surface is at or below that required for the reduction of H2O,76 although the only potentials reported by the authors do not support this position. Furthermore, ab initio modeling studies which have been performed in support of the claimed mechanism involved a hydroxide radical

Waste Containers

CuCl2•3Cu(OH)2 k3

kAb

kBf

CuClADS

kS2

CuCl2−

kBb

Cu(HS)ADS

k2

kS3

JCu(II)

k6

k1

kS1

Cu

k3

k −4

k −3

Cu2+ Fe(II)

kD kAf

k4

k −3

CuCl2•3Cu(OH)2

Cu2+

Cu2+

Cu

Cu(II)ADS

Fe(II)

k1

k6

Fe(III) JCuCl−

k −2

Fe(III)

CuCl2−

FeS2

2

(R0e-αPt + R1)

Cu2O Cu2S

HS− + Fe(II)

k7

Fe(II)

FeS k7

HS− kE

HS−

SRB

JHs−

H2 + S2− Fe(III) + SO42−

OH−

439

kC

O2

SO2− 4 Aerobic microbes k9

FeS2

+O2 k5

k10

O2(aq)

JO2

k8

O2(g) Figure 15 Reaction mechanism for the corrosion of copper in compacted bentonite in the presence of O2 or sulfide and saline groundwater.

−0.2

C

D

E

A B

(b) C

−0.4

D

A B

ECORR (VSCE )

C

−0.6

2 h for curve (a) only

−0.8

−1.0

E

(c)

(a) 0

50

100

150

200 Time (h)

250

300

350

400

Figure 16 Time dependence of the corrosion potential (ECORR) of copper electrodes in 1 mol dm3 NaCl solution at 25  C in the presence of dissolved oxygen or sulfide. Curve (a) is for a bare copper electrode and curves (b) and (c) are for a copper electrode covered by a 1-mm-thick layer of compacted Na-bentonite. At points A, B, and C, the original air purge was changed for gas mixtures of 2 vol% O2/N2, 0.2% O2/N2, and pure N2, respectively. At points D and E, Na2S was added to the solution to produce a dissolved HS concentration of 10 and 100 mg dm3, respectively.

species, rather than a hydroxide anion.77 The evidence presented to date is far from compelling, but the question of whether water does corrode in pure water is being actively studied.

Figure 16 shows the effect of sulfide on the corrosion behavior of copper. At points ‘D’ and ‘E,’ various concentrations of HS were added to the solution to simulate the effect of sulfide in the

440

Waste Containers

groundwater. In bulk solution (curve (a)), the value of ECORR fell precipitously to approximately 0.9 VSCE, while a similar effect was observed for the claycovered electrode (curve (c)) after a delay period during which the HS diffused through the clay to the copper surface. Such a potential lies below the H2O/H2 equilibrium line (Figure 14) and indicates that copper is corroding with the evolution of H2.49 This reaction mechanism is shown schematically in Figure 15. This observation has important implications for copper canisters since it demonstrates that copper will continue to corrode following the consumption of the initially trapped O2 if sulfide is present and can reach the canister surface.5,6,26 In an overall sense, curves (a) and (c) in Figure 16 simulate the expected evolution of the corrosion behavior of copper canisters in the repository as the initial aerobic environment gives way to one that is sulfide dominated. Copper may be susceptible to some degree of localized corrosion in the repository environment, although the treatment of localized effects has changed over the years as we have learnt more about the corrosion behavior of copper in disposal environments.10,11 The earliest analyses of the corrosion behavior of copper canisters were theoretically, rather than experimentally, based.33,34 Because of the wellknown phenomenon of the pitting of copper potable water pipes, it was deemed appropriate to make an allowance for pitting. In the absence of specific experimental studies in repository environments, literature data from a long-term underground corrosion study were used to estimate a pitting factor, the ratio of the maximum to the average penetration. Some of the data from this study showed a pitting factor as high as 25.34 In subsequent analyses using additional sources of data, a pitting factor of between 3 and 5 was proposed.33 In other approaches, extreme-value statistics was applied to the same dataset to predict the depth of the deepest pit on a canister over a period of 106 years, resulting in a maximum pit depth of 6 mm.36 Both approaches, however, incongruously predict pit depths of many mm for a material that is known to be dissolving actively. It is now recognized that copper is not susceptible to localized corrosion under repository conditions but instead corrodes unevenly resulting in a roughened surface.10,11 The evidence for this comes from experiments in which copper samples have been corroded in contact with compacted bentonite saturated with an O2-containing Cl solution. At the end of these tests, the average depth of corrosion was of the

order of 40 mm, with local penetrations of an additional 10–15 mm. There was no evidence for the permanent spatial separation of anodic and cathodic reactions, as would be required for classical localized-corrosion mechanisms. Oxygen-free copper is known to be susceptible to SCC in a limited number of environments that include ammonia, nitrite, acetate, and, possibly, sulfide.10,11,59,62,63 Although the canister will be largely under compressive loading (due to the hydrostatic load and bentonite swelling pressure), it is difficult to argue convincingly that there will not be some part of the canister that is not under tensile load at some stage. Therefore, most studies to date have focused on whether the appropriate environment can form. In the case of ammonia, nitrite, and acetate, it is apparent that cracking only occurs for specific combinations of potential and pH and that the presence of an oxidant (O2 or, more likely, Cu(II)) is required.63 Since such oxidizing conditions will only be present in the repository for a relatively short period of time and since the concentration of these SCC agents in the repository is so small, it is argued that cracking due to the presence of ammonia, acetate, or nitrite is highly unlikely.59 The experimental evidence for the SCC of copper in sulfide solutions suggests a threshold [HS] of 0.001 mol dm3.62 Such a high concentration is unlikely in the repository (due to precipitation with Fe(II) as FeS) making cracking equally unlikely, especially since the interfacial sulfide concentration will be minimal because of transport control through the compacted bentonite.59 The possibility of SCC is closely tied to the issue of MIC because the most likely source of ammonia, acetate, and nitrite in the repository is microbial activity.64 All of the candidate canister materials considered here, with the possible exception of Ti-based alloys, are susceptible to MIC of one form or another.78 From the viewpoint of the impacts on the canister, the approach for the different materials is similar; to demonstrate that microbial activity will either not occur in the repository or will be extremely limited. Therefore, the discussion here for copper canisters is applicable to the other classes of canister material. In general, the repository environment is inhospitable for microbes.25 The combined effects of high temperature, radiation fields, saline groundwaters and pore fluids, alkaline cementitious pore water, mechanical forces from swelling smectite clays, redox conditions, the lack of organic nutrients and terminal electron acceptors, and low water activity

Waste Containers

will suppress microbial activity. Although microbes can be identified that can survive extremes of each of these conditions, the combined effects of a number of environmental stressors will limit both the activity and diversity of the microbial population in the repository. For bentonite-backfilled repositories, there is much evidence to indicate that the combined effects of high salinity and high clay density suppress microbial activity due to the low water activity.79 Although some microbial activity has been reported in compacted bentonite, it has only been reported in the presence of added nutrients.80 For cementbackfilled repositories, the alkaline pore fluids will suppress the activity of the vast majority of microbes, and taken in conjunction with other environmental stressors, they will also minimize any impact of microbes.81 In contrast, microbial activity is expected to be dominant in LILW repository designs with no or limited use of cementitious materials.71 Oxygen-free copper would appear to be a suitable canister material for the disposal of HLW/SF, a conclusion that has been reinforced by the results of extensive studies performed since the idea was first proposed in the late 1970s. The primary advantages of the use of copper are its thermodynamic immunity in the absence of O2 (and sulfide) and the tendency toward general, as opposed to localized, corrosion in the chloride-rich environments encountered in most deep groundwaters. If the conventional thermodynamic wisdom is proven correct, copper canister lifetimes in excess of 105 years are entirely possible, and supported by numerous natural and archaeological artifacts.82 5.17.4.4

Titanium Alloys

Titanium alloys have been considered as candidate canister materials for the disposal of HLW or SF in a number of countries (Table 2). Early work focused on the properties of commercially pure grades, such as Ti-1 and Ti-2, and the dilute Ni, Mo alloy Ti-12.2,7,12 The key to the use of Ti alloys, however, is to prevent crevice corrosion and most recent efforts have been directed toward the potential use of the Pd-containing alloys Ti-7 (0.08–0.17 wt% Pd) or Ti-16 (0.04–0.08 wt% Pd).7,12 In the US Yucca Mountain Project, Ti alloys were proposed for use as a drip shield to be placed over the waste packages to protect them from seepage drips and rockfall.1 To provide structural strength, crevice-corrosionresistant Ti-24 (0.04–0.08 wt% Pd) or the Ru-containing Ti-29 were proposed, both alloys being based on the basic 6Al–4V Ti-5 composition.

441

The passive film on Ti exhibits one of the broadest ranges of thermodynamic stability of any oxide for any common structural material (Figure 17). Titanium dioxide is stable at all potentials corresponding to the stability of water and only becomes soluble in acidic or highly alkaline solution. The absence of multiple oxidation states in the oxide makes the oxide electrically insulating and may contribute to the apparent immunity to MIC.25 Titanium(III) forms under cathodic polarization. As would be expected for such a passive film, the rate of general corrosion of Ti alloys in repositorytype environments is very low.12 The average corrosion rate of Ti-7 in a range of aerated solutions with salinities up to several moles per cubic decimeter is 20 nm year1 and is independent of temperature over the range 60–90  C.12 Even lower rates (of the order of 1 nm year1) have been reported in bentonite environments.83,84 Localized corrosion of Ti alloys is important not only because it could lead to more rapid wall penetration but also because it can lead to rapid absorption of hydrogen.2,7 The pitting potential of Ti alloys in natural environments is >3 V at temperatures 100  C and pitting does not therefore occur under freely corroding conditions.7,12 Some Ti alloys, however, are susceptible to crevice corrosion in Cl environments resulting in acidification of the crevice region and sustained creviced propagation supported by a combination of O2 reduction on exposed surfaces outside of the occluded region and the reduction of Hþ inside the crevice. For Ti-2 and, especially, Ti-12 the internal reduction of Hþ can account for the majority of the crevice propagation.7,9,12 However, continued external O2 reduction is required to sustain the acidified crevice, and without it the crevice would stifle due to the loss of the critical crevice chemistry. In that regard, even Ti alloys that are susceptible to crevice corrosion may not fail in a sealed repository in which the amount of available O2 is limited. The more prudent approach to crevice corrosion, however, is to select a crevice-corrosion-resistant grade.1,7,12 Additions of Pd suppress crevice corrosion by promoting passivity by catalyzing Hþ reduction, thus reducing the local acidity and increasing ECORR. Figure 11 demonstrates the resistance of Ti-7 to crevice corrosion and the extensive range of passivity of this particular alloy that extends over more than 1500 mV.1 Another benefit of using a crevice-corrosionresistant Ti alloy is that the amount of absorbed H is greatly diminished. Titanium forms hydrides and exhibits slow crack growth or fast fracture above a

442

Waste Containers −2 2.2

−1

0

1

TiO22+

2 ? 1.8

6⬘

2

3

4

5

6

7

8

9

10

11

?

12

13

14

0

?

15

16 2.2

?

2

TiO3•2H2O

22 ?

1.8

1.6 0 1.4

? −2

−4

b

?

7⬘

?

− ? HTiO4 TiO42−

1.6 1.4

−6 ?

1.2

1.2

15 1 0.8

?

TiO2+

−6

1 0.8

−4

1⬘

0.6

0.6 ?

0.4

0.4 5⬘

0.2 E (V)

TiO2•H2O

a

0

0

20

−0.2

Ti3+

−0.4

2⬘

−0.6

−2 −4 −6

−0.8

−0.8

11 ?

18

−1

Ti2+

21

Ti(OH)3

−1.2 −1.4

−1.8

−0.4

3⬘ 0

−1.4

4⬘

17

TiO

−1.6

9

−1.8

−2

−2

8

Ti

−2.2 −2.4 −2

−1 −1.2

13 0 −2 −4 −6

−0.2

HTiO3−

14

−0.6

−1.6

0.2

16

−2.2 −1

0

1

2

3

4

5

6

7 pH

8

9

10

11

12

13

14

15

−2.4 16

Figure 17 Potential-pH (Pourbaix) diagram of the titanium–water system at 25  C considering TiO2
2H2O as the stable Ti(IV) containing solid phase. Lines a and b represent the equilibria between water and hydrogen and water and oxygen, respectively, for unit fugacity of the gaseous species. The 0, 2, 4, and 6 indicate the equilibrium conditions between the various solid and dissolved phases for activities of 0, 102, 104, and 106, respectively. Reproduced from Pourbaix, M. Atlas of Electrochemical Equilibria in Aqueous Solutions, 2nd ed.; NACE International: Houston, TX, 1974.

critical absorbed hydrogen concentration7,12 or a critical hydride layer thickness.85 The most common routes for the absorption of H are cathodic polarization or galvanic coupling to a more-active material (neither of which are of concern for nuclear waste canisters) or via an acidified crevice or pit. Increased acidity not only promotes H generation but also destabilizes the normally passive TiO2 film.

In the absence of crevice corrosion, however, hydrogen must be absorbed through the passive oxide which greatly reduces the fraction of H absorbed (Figure 18).12 The rate of H absorption is a function of temperature, potential, and the presence of intermetallic particles. Increasing temperature and decreasing potential promote the rate of Hþ discharge and the fraction of H absorbed by the oxide

Waste Containers

Crevice corrosion

Development of acidity/ oxide dissolution

443

Film breakdown/recrystallization/ rapid thickening ≥70 ⬚C

Active corrosion under occluded conditions

Increasing temperature

Ti3+ dissolution in acidic solution Rutile formation

TiIV → TiIII (oxide)

TiH2 (−1.0 V ) TiOOH

H2 evolution H absorption In metal

Hydrogen-induced cracking

(−0.65 V) (−0.54 V)

Threshold for H absorption In oxide

EOc

H2O absorption TiO2

Pitting

(8–9 V)

Amorphous/crystalline Anatase formation (TiIIII, OIIv ) Film growth (HFIC) ND decreases

Film breakdown/ recrystallization/ rapid thickening

EFB (acidic solution)

Figure 18 Schematic illustration of the properties of the passive film on titanium as a function of potential. Reproduced from Shoesmith, D. W. Corrosion 2006, 62, 703–722.

and subsequently by the Ti substrate. Increasing temperature increases the diffusivity of H in a-Ti, with a temperature of 80  C considered a threshold for the onset of H problems in practical applications.86 Although intermetallic phases, such as Ti2Ni in Ti-12 or Pd particles in Ti-7 and Ti-16, can suppress crevice corrosion, they may also promote H absorption, both by catalyzing the reduction of Hþ and by acting as preferential sites for H absorption. There is wide variation in the reported H absorption efficiency on passive Ti, with values as high as 2%.12 The solubility of H in a-Ti is of the order of 20–150 wppm.12 The critical absorbed H concentration (HC) for HIC is alloy-dependent, varying from 500–800 wppm for Ti-2, 400–600 wppm for Ti-12, and 1000–2000 wppm for Ti-16. Although Ti alloys have not been widely adopted as candidate canister materials, they offer a number of advantages for the disposal of HLW/SF.18 Chief among these advantages are the prospect of very long canister lifetimes for crevice-corrosion-resistant grades, the minimal impact on other barriers, and design flexibility for use in backfilled or nonbackfilled repositories. However, on the assumption that Ti canisters would be constructed with relatively thin walls, an internal structural support would be required. Perhaps the greatest hurdle to the adoption of Ti as a canister material, however, is the perceived difficulty of predicting and, more

importantly, justifying long-term predictions of passivity and H absorption. 5.17.4.5

Nickel Alloys

Of the various classes of Ni-based alloys, those considered as candidate canister materials fall into the Ni–Cr–Mo or Ni–Fe–Cr–Mo families. Within the Ni–Cr–Mo family of alloys, Inconel 625 was considered for the salt repository program in Germany, along with the more-corrosionresistant Hastelloy C-4 (Table 2).8 Hastelloy C-22 (Alloy 22) has been considered in the Spanish program8 and was the reference canister material in the US Yucca Mountain Project.1 The Ni–Fe–Cr–Mo alloy Incoloy 825 was also considered in Germany (Table 2).8 As for stainless steels, Ni–Cr–Mo and Ni–Fe– Cr–Mo alloys rely on the formation of a passive Cr(III) film for their general corrosion resistance. Alloying additions of Mo and W (Alloy 22) or Co (Hastelloy C-4) improve the resistance to localized corrosion.86 Addition of Si to Inconel 625 and Incoloy 825 improves corrosion resistance under oxidizing conditions. One of the main advantages of Ni alloys is the high solubility of many alloying elements, resulting in both a wide range of material properties and the general absence of adverse effects due to secondary phases.

444

Waste Containers

1.00

0.50

0.25

Low uncertainty Medium uncertainty High uncertainty Measured data

0.00 0

5

20 10 15 General corrosion rate (nm year−1)

25

00264DC_LA_0468.ai

Cumulative probability

0.75

30

Figure 19 Cumulative distribution function of the rate of general corrosion of Alloy 22 in saline simulated evaporated waters at 60  C. Reproduced from DOE. Yucca Mountain Repository license application; US Department of Energy, DOE/RW-0573; 2008.

Because of their superior passivity, Ni alloys exhibit low rates of general corrosion in a wide range of environments. Figure 19 shows a cumulative distribution function of corrosion rates of Alloy 22 exposed to various aerated synthetic saline solutions at a temperature of 60  C for a period of 5 years.1 The mean corrosion rate is 5–10 nm year1, implying that a 6.35-mm-thick (0.25 in.) canister would have a lifetime of 106 years. What has not been adequately addressed in making these sorts of predictions is how the corrosion rate will vary over the very long term as the oxide thickens and the underlying substrate becomes depleted in Cr, Mo, W, and other alloying elements. Corrosion rates are higher in aggressive acidic ‘Q’ brine solutions representative of those that may be present in salt formations, with rates of up to 0.9 mm year1 at 200  C.8 As for stainless steels, Ni alloys are susceptible to localized corrosion.1,27,42 Although pitting of exposed surfaces does occur, most attention has been focused on the possibility of crevice corrosion as it is difficult to guarantee that occluded regions will not be present somewhere on the canister surface and because crevice corrosion initiates under less-aggressive conditions than pitting. Figure 11 illustrates the relative susceptibility to crevice corrosion of Alloy 22, Type 316L

stainless steel, and the Pd-containing Ti-7. Although not as resistant as Ti-7, there is a significant difference between the corrosion potential and either the breakdown or repassivation potential for Alloy 22. (Note: the value of ECORR is not strictly given by the point of zero net current during a potentiodynamic scan, as indicated in Figure 11. Nevertheless, the difference of 900 mV between ‘ECORR’ and Ercrev is an indication of the significant resistance of Alloy 22 to crevice corrosion.) The crevice corrosion behavior of Alloy 22 was extensively investigated as part of the Yucca Mountain Project.1 The focus of these studies was the initiation of localized attack, with less attention paid to crevice propagation. The susceptibility to crevice corrosion was assessed based on the critical potential criterion in eqn [2].1 A series of measurements were made of ERP which were then fitted to a polynomial expression involving T, [Cl], and [NO 3 ]. The crevice repassivation potential shifts to more-negative values with increasing T and [Cl], but nitrate ions inhibit localized corrosion. A similar series of longterm ECORR measurements were made and similarly fitted to a polynomial involving the same variables. Increasing temperature and nitrate concentration ennoble ECORR whereas increasing [Cl] shifts

Waste Containers

445

Repassivation potential (mVSCE)

1200 1000 Alloy 825

800

Alloy 625

No crevice Crevice corrosion corrosion Autoclave Glass cell

600 400 Alloy 22 200 0

316L

-200 -400 10-4

10-3

10-2 10-1 100 101 Chloride concentration (mol dm−3)

102

Figure 20 Chloride concentration dependence of the crevice repassivation potential for various Ni-based alloys. Reproduced from Brossia, S.; Browning, L.; Dunn, D. S.; Moghissi, O. C.; Pensado, O.; Yang, L. Effect of environment on the corrosion of waste package and drip shield materials; Report No: CNWRA 2001-03; Center for Nuclear Waste Regulatory Analysis, 2001.

1000

Cracking No cracking 100 Time to failure (h)

ECORR in the active direction.1 The net effect of these dependencies is that the driving force for localizedcorrosion initiation, that is, the difference DE between ECORR and ERP, increases with increasing T and [Cl], but decreases with increasing [NO 3 ]. Figure 20 shows the dependence of ERP on Cl concentration for the crevice corrosion of various Nibase alloys and Type 316L stainless steel at 95  C.40 The resistance to crevice corrosion increases in the order 316L (least resistant) < Incoloy 825 < Inconel 625 < Alloy 22 (most resistant) (on the assumption that the respective values of ECORR are not significantly different). Possibly because of the continuously aerated environment in the Yucca Mountain repository, it was generally assumed that once initiated a crevice would propagate rapidly to failure. However, there is convincing evidence that this is not the case and that the material naturally stifles, without any assistance from the depletion of O2 that would occur in a saturated repository.44 Stifling is clearly a result of the loss of the critical chemistry necessary to sustain crevice propagation, but quite why this should occur is not fully understood. It may well be due to the inherent resistance of the material (e.g., catalysis of Hþ reduction on Mo- or W-enriched surface layers in the crevice or the stability of Mo- and/or W-containing films in the low-pH crevice solution), or it could be due to mass-transport or iR restrictions. Regardless, there is reason to believe that the extent of crevice propagation would be limited. An advantage of Ni alloys over stainless steels is that they are relatively immune to SCC in warm

10 Minimum time to cracking

1 0

60 20 40 Nickel content (wt%)

80

Figure 21 Effect of Ni content on the SCC susceptibility of stainless steels and nickel alloys in boiling MgCl2 solution. Reproduced from Sedriks, A. J. Corrosion of Stainless Steels, 2nd ed.; Wiley: New York, NY, 1996.

chloride environments.47 Figure 21 shows the dependence of the time-to-failure of tensile samples on the Ni content of the alloy for a range of stainless steels and various Ni alloys in boiling MgCl2 solution

446

Waste Containers

at 154  C. The Ni–Cr–Mo alloys with a minimum Ni content of 50 wt% are clearly more resistant to SCC than the 300-series stainless steels with Ni contents in the range 8–15 wt%. This is borne out by corrosion testing in which it has proven impossible to crack Alloy 22 at ECORR in simulated Yucca Mountain environments.1 Only anodic polarization or the addition of Pb(II) were found to induce SCC. The passivity of Ni alloys is known to be lost because of the accumulation of S at the Ni/oxide interface.87 The source of S is typically the alloy itself with S accumulating at the interface as a consequence of the slow dissolution of the alloy. This is one mechanism by which the alloy could periodically lose passivity and the oxide spall off. However, following oxidation and dissolution of the accumulated S, it would seem likely that the alloy would simply repassivate. Nickel alloys are suitable canister materials for a wide range of environments (both permanently oxidizing and eventually anoxic) and have minimal impact on other barriers in the overall system. There is good experience with the design and fabrication of Ni-alloy vessels in various industries. The corrosion

Table 4

resistance of Ni alloys is excellent, given appropriate alloy selection. However, Ni alloys suffer from the same negative perception as Ti alloys regarding the ability to predict the long-term corrosion behavior of a material that depends on a thin passive layer. This perception is unfortunate since passive alloys could provide very long-term containment of nuclear wastes.

5.17.5 Canister Lifetime Predictions Ultimately, it is necessary to make some form of lifetime prediction for the canister. Table 4 summarizes the general approaches that have been taken toward predicting different forms of corrosion for the different classes of candidate canister materials. The most common method for predicting the extent of general corrosion is simply to extrapolate empirically measured corrosion rates, typically determined using mass-loss measurements. Where appropriate, as in the general corrosion of Alloy 22 waste packages in the Yucca Mountain repository, the temperature (and, hence, time) dependence of

Methods used for the long-term prediction of the corrosion behavior of canister materials General corrosion

Localized corrosion

Environmentally assisted cracking

MIC

C-steel

Mass-balance, electrochemical modeling, and empirical rates for aerobic corrosion Empirical rates for anaerobic corrosion

Pitting factor Maximum penetration based on extremevalue analysis (EVA)

Reasoned argument for no SCC No effects of H because of low H concentration and use of low-strength steel

Stainless steel

Empirical corrosion rates

Suppression of SCC by use of cement backfill

Copper

Mass-balance or detailed reactivetransport modeling for aerobic phase Mass-transport limited corrosion due to sulfide Empirical corrosion rates

Suppression of localized corrosion by use of cement backfill Pitting factor or EVA maximum pit depth Allowance for surface roughening

Mass balance based on either organic C or sulfate limitation Reasoned argument based on lack of microbial activity –

Reasoned argument for no SCC

Reasoned argument based on lack of microbial activity

Limited propagation argument for Ti-2, Ti-12 or use of resistant Ti-7, Ti-16, Ti-29 alloys Threshold potential (ERP) for initiation, followed by rapid propagation

HIC based on either critical absorbed H concentration or critical hydride layer thickness

–

Slip dissolution model

Enhancement factor for general corrosion

Ti alloys

Ni alloys

Empirical corrosion rates

Waste Containers

the corrosion rate may also be taken into account.1 To account for the time-dependent decrease in corrosion rate due to the formation of a protective or passive film, empirical rates are generally determined from relatively long-term tests (up to 5–10 year duration).1,70,72 For repositories in which there is a limited inventory of O2, the empirical extrapolation of measured rates is typically used for the anaerobic period, with a mass-balance approach based on the available O2 used for the aerobic period.5,6,13,32 In the case of copper canisters in compacted bentonite, a reactive-transport model, coupled with a mixed-potential model to predict ECORR, has been developed.37 The results of this detailed model are consistent with those based on mass-balance and/or mass-transport calculations. There have been many different methods used to predict the localized-corrosion behavior of canister materials. For materials that are generally active, that is, C-steel and copper, the most common approach has been to use empirical data to either derive a pitting factor or a time-dependent maximum pit depth.2,13,32–34,36 For passive materials, the approaches taken include:
demonstrating that localized corrosion will not initiate in the repository environment, in the case of stainless steels with a cementitious backfill;51
arguing that the extent of propagation will be limited, for example, for Ti-2 and Ti-12 in a bentonitebackfilled repository with limited O29; and
determining whether crevice corrosion will initiate based on a threshold potential criterion and assuming rapid propagation for those canisters for which initiation does occur, the approach adopted for the localized corrosion of Alloy 22 waste packages.1,27,42 Typically, proponents of the use of a given material have developed reasoned arguments that the material is not susceptible to SCC.2,13,14,59,63 This is generally the preferred approach as crack growth rates are fast compared with repository timescales. If a material is deemed to be susceptible to SCC, then this would be sufficient reason to exclude this material from further consideration. However, even though the material seems to be immune to SCC under repository conditions, a slip dissolution was developed for Alloy 22 waste packages in the Yucca Mountain Project.1 Hydrogen-related degradation mechanisms are issues only for C-steel and Ti alloys. In the case of C-steel, it is argued that the grades of material

447

under consideration (i.e., low-strength steels) are not overly susceptible and that repository environments will not produce a sufficiently large absorbed H concentration.2,66 Titanium alloys, on the other hand, will inevitably form hydrides once sufficient H has been absorbed. In this case, therefore, the issues are how fast H is absorbed and the failure criterion that is to be used for determining failure.7,12 For crevice-corrosion-resistant Ti alloys, H absorption can only occur under passive conditions, which is a slow and not entirely understood process. The two failure criteria that have been developed are based on either a critical absorbed H concentration7,12 or a threshold hydride layer thickness.85 Various attempts have been made to predict the extent of MIC. Mass-balance arguments based on the availability of organic C88 or sulfate2 have been made, and an enhancement factor based on the rate of general corrosion has been proposed.1,27 Alternatively, it is possible to argue that the extent of microbial activity in the repository will be restricted by the adverse effects of various environmental stressors.25 A reactive-transport model to predict the rate of microbial activity has been developed based on such an approach.64

5.17.6 Conclusions The five classes of alloys that have been proposed as canister materials for the disposal of nuclear waste are: C-steels, stainless steels, copper and copper alloys, Ti alloys, and Ni alloys. Each of these materials has advantages and disadvantages for this application and the selection of the appropriate material should be based on consideration of the target canister lifetimes, the nature of the disposal environment and how it evolves over time, and the ability to make robust and justifiable lifetime predictions. The corrosion behavior of candidate canister materials has been the subject of extensive research over the past 30 years and a number of trends have emerged during that time:
For repositories that are located below the water table and for which the amount of oxidant is limited, the preferred approach appears to be based on the use of either oxygen-free copper (structurally supported by a steel or iron insert) or C-steel, both in combination with a bentonite backfill. Copper is the preferred material for crystalline host rocks for which less credit can be taken for the geosphere as a barrier, whereas C-steel is

448













Waste Containers

preferred in sedimentary host rocks which themselves provide significant containment. There are exceptions to this general rule, the most obvious being the C-steel/cement overpack in the Belgium program and the nonbackfilled disposal boreholes being considered for the disposal of HLW in France. Passive materials are primarily considered for highly aggressive environments, such as those in salt formations or during the early thermal transients in unsaturated repository designs. The apparent emphasis of active over passive canister materials is partly a consequence of the perceived difficulty in making long-term predictions of passive behavior and of localized corrosion. This perception is considered to be unfortunate, since we have learnt much about predicting the longterm behavior of passive materials over the past 30 years and the use of thin-walled canisters made from passive alloys offers a number of advantages. At the time of writing, there is no active program for repositories located in the unsaturated zone, the Yucca Mountain Program apparently being on the brink of cancellation. Although prediction of the nature of the environment in unsaturated repositories is challenging, location of a nuclear waste repository above the water table does offer certain advantages and the issues related to predicting long-term performance of the canisters are not considered to be insurmountable.

References 1. 2. 3. 4.

5.

6. 7. 8.

DOE. Yucca Mountain Reposiotry license application; US Department of Energy, DOE/RW-0573; 2008. JNC. Repository Design and Engineering Technology, Supporting Report 2; Japan Nuclear Cycle Development Institute, 2000. Nagra. Project Opalinus Clay, Safety Report; Nagra Technical Report NTB 02-05; National Cooperative for the Disposal of Radioactive Waste, 2002. ONDRAF-NIRAS. A review of corrosion and material selection issues pertinent to underground disposal of highly active nuclear waste in Belgium; ONDRAF Report Ref: NIROND 2004-02; 2004. SKB. Long-term safety for KBS-3 repositories at Forsmark and Laxemar – A first evaluation; Main Report of the SR-Can project; Technical Report TR-06-09; Swedish Nuclear Fuel and Waste Management Co., 2006. Pastina, B.; Hella¨, P. Expected Evolution of a Spent Fuel Repository at Olkiluoto; Posiva Oy: Olkiluoto, Finland, 2006; POSIVA 2006-05. Shoesmith, D. W. Corrosion 2006, 62, 703–722. Kursten, B.; Smailos, E.; Azkarate, I.; Werme, L.; Smart, N. R.; Santarini, G. COBECOMA, State-of-the-Art Document on

the Corrosion Behaviour of Container Materials; Contract No. FIKW-CT-20014-20138, Final Report;European Commission, 2004. 9. Johnson, L. H.; Tait, J. C.; Shoesmith, D. W.; Crosthwaite, J. L.; Gray, M. N. The disposal of Canada’s nuclear fuel waste: Engineered barriers alternatives; AECL-10718, COG-93-8; Atomic Energy of Canada Limited, 1994. 10. King, F.; Ahonen, L.; Taxe´n, C.; Vuorinen, U.; Werme, L. Copper corrosion under expected conditions in a deep geologic repository; Report No: SKB TR 01-23; Swedish Nuclear Fuel and Waste Management Company, 2001. 11. King, F.; Ahonen, L.; Taxe´n, C.; Vuorinen, U.; Werme, L. Copper corrosion under expected conditions in a deep geologic repository; Posiva Oy Report POSIVA 2002-01; 2002. 12. Hua, F.; Mon, K.; Pasupathi, P.; Gordon, G.; Shoesmith, D. Corrosion 2005, 61, 987–1003. 13. Johnson, L. H.; King, F. Canister options for the disposal of spent fuel; Technical Report 02-11; Nagra, 2003. 14. King, F. Overview of a carbon steel container corrosion model for a deep geological repository in sedimentary rock; Report No: NWMO TR-2007-01; Nuclear Waste Management Organization, 2007. 15. Smart, N. R.; Wood, P. Corrosion resistance of stainless steel radioactive waste packages; Report N/110; UK Nirex Limited, Nirex, 2004. 16. Smart, N. R.; Blackwood, D. J.; Marsh, G. P.; et al. The anaerobic corrosion of carbon and stainless steels in simulated repository environments: A summary review of nirex research; AEAT/ERRA-0313; 2004. 17. Smart, N. R.; Naish, C. C.; Pritchard, A. M. Corrosion principles for the assessment of stainless steel radioactive waste containers; Report No: SA/EIG/14921/C010; Serco Assurance Report to Nirex, 2006. 18. King, F.; Padovani, C. Corrosion Eng. Sci. Technol. 2011, in press. 19. Bel, J. J. P.; Wickham, S. M.; Gens, R. M. F. In Scientific Basis for Nuclear Waste Management XXIX, Materials Research Society Symposium Proceedings; Van Iseghem, P., Ed.; Materials Research Society: Warrendale, PA, 2006; Vol. 932, pp 23–32. 20. Fe´ron, D.; Crusset, D.; Gras, J. M. Corrosion 2009, 65, 213–223. 21. McMurry, J.; Dixon, D. A.; Garroni, J. D.; et al. Evolution of a Canadian deep geologic repository: Base scenario; Report No: 06819-REP-01200-10092-R00; Ontario Power Generation Nuclear Waste Management Division, 2003. 22. SAFIR 2. Safety Assessment and Feasibility Interim Report 2; ONDRAF/NIRAS; 2001. 23. Berner, U. R. Waste Manag. 1992, 12, 201–219. 24. Kursten, B.; Druyts, F. In Corrosion/2000; NACE International: Houston, TX, 2000; paper no. 00201. 25. King, F. Corrosion 2009, 65, 233–251. 26. SKB. Fuel and canister process report for the safety assessment of SR-Can; Report No: SKB TR 06-22; Swedish Nuclear Fuel and Waste Management Company, 2006. 27. Farmer, J.; McCright, D.; Gdowski, G.; et al. In Proceedings of Transportation, Storage, and Disposal of Radioactive Materials – 2000; American Society of Mechanical Engineers: New York, NY, 2000; PVP-Vol. 408, pp 53–69. 28. Shoesmith, D. W.; King, F. The effects of gamma radiation on the corrosion of candidate materials for the fabrication of nuclear waste packages; Report No: AECL-11999; Atomic Energy of Canada Limited, 1999. 29. King, F. Review and gap analysis of the corrosion of copper containers under unsaturated conditions; Report

Waste Containers

30. 31. 32. 33. 34.

35.

36. 37. 38.

39. 40.

41. 42.

43. 44.

45. 46. 47. 48. 49. 50. 51.

No: 06819-REP-01300-10124-R00; Ontario Power Generation, Nuclear Waste Management Division, 2006. King, F.; Kolar, M.; Shoesmith, D. W. In Corrosion/96; NACE International: Houston, TX, 1996; paper no. 380. King, F.; Litke, C. D.; Quinn, M. J.; LeNeveu, D. M. Corrosion Sci. 1995, 37, 833–851. Johnson, L. H.; King, F. J. Nucl. Mater. 2008, 379, 9–15. SKB. Final storage of spent nuclear fuel – KBS-3; Report No: KBS-3; Swedish Nuclear Fuel Supply Company, 1983; Vol. I–IV. Swedish Corrosion Institute. Copper as canister material for unreprocessed nuclear waste – Evaluation with respect to corrosion; Report No: KBS-TR-90; Swedish Nuclear Fuel Supply Company, 1978. King, F.; Kolar, M. The copper container corrosion model used in AECL’s second case study; Report No: 06819-REP-01200-10041-R00; Ontario Power Generation, Nuclear Waste Management Division: Toronto, ON, 2000. King, F.; LeNeveu, D. In Proceedings of Conference on Nuclear Waste Packaging, FOCUS ’91; American Nuclear Society: La Grange Park, IL, 1992; pp 253–261. King, F.; Kolar, M.; Maak, P. J. Nucl. Mater. 2008, 379, 133–141. Maak, P. The selection of a corrosion-barrier primary material for used-fuel disposal containers; Report No: 06819-REP-01200-10020-R00; Ontario Power Generation, Nuclear Waste Management Division, 1999. Brossia, C. S.; Cragnolino, G. A. Corrosion 2000, 56, 505–514. Brossia, S.; Browning, L.; Dunn, D. S.; Moghissi, O. C.; Pensado, O.; Yang, L. Effect of environment on the corrosion of waste package and drip shield materials; Report No: CNWRA 2001-03; Center for Nuclear Waste Regulatory Analysis, 2001. BSC (Bechtel SAIC Company). Aqueous corrosion rates for waste package materials; Prepared for US DOE; ANL-DSD-MD-000001; Oct 2004. Cragnolino, G.; Dunn, D. S.; Brossia, C. S.; Jain, V.; Chan, K. S. Assessment of performance issues related to alternate engineered barrier system materials and design options; Report No: CNWRA 1999-003; Center for Nuclear Waste Regulatory Analysis, 1999. Dunn, D. S.; Cragnolino, G. A.; Sridhar, N. In Corrosion/96; NACE International: Houston, TX, 1996; paper no. 97. Mon, K. G.; Gordon, G. M.; Rebak, R. B. In Proceedings of the12th International Conference on Environmental Degradation of Materials in Nuclear Power System – Water Reactors; Allen, T. R., King, P. J., Nelson, L., Eds.; The Minerals, Metals & Materials Society: Warrendale, PA, 2005; pp 1431–1438. Rebak, R. B.; Koon, N. E.; Dillman, J. R.; Crook, P.; Summers, T. S. E. In Corrosion/2000; NACE International: Houston, TX, 2000; paper no. 00181. Sridhar, N.; Cragnolino, G. A. Corrosion 1993, 49, 967–976. Sedriks, A. J. Corrosion of Stainless Steels, 2nd ed.; Wiley: New York, NY, 1996. Szklarska-Smialowska, Z. Pitting and Crevice Corrosion; NACE International: Houston, TX, 2005. Smith, J.; Qin, Z.; King, F.; Werme, L.; Shoesmith, D. W. Corrosion 2007, 63, 135–144. King, F.; Kolar, M.; Kessler, J. H.; Apted, M. J. Nucl. Mater. 2008, 379, 59–67. Smart, N. R. Review of effect of chloride in cementitious environments on corrosion of stainless steels; Serco Assurance Report for UK Nirex Limited; SA/SIS/14921/ R001; 2002.

52. 53.

54.

55. 56. 57.

58.

59. 60.

61.

62. 63.

64.

65.

66. 67.

68. 69. 70.

449

Smart, N. R. Atmospheric corrosion of stainless steel waste containers during storage; Report No: AEA-TPD-262, issue C; AEA Technology, 2000. Smart, N. R. Atmospheric pitting corrosion of stainless steel radioactive waste containers; Serco Assurance Report to Nirex; Report No: SA/EIG/14921/ C0050; 2005. Blackwood, D. J.; Gould, L. J.; Naish, C. C.; et al. The localised corrosion of carbon steel and stainless steel in simulated repository environments; AEAT/ERRA 0318; Dec 2002. Kursten, B.; Druyts, F. J. Nucl. Mater. 2008, 379, 91–96. Andra. Materials Baseline; Andra Report C.RP. AMAT.01.060; 2001, Vol. 4. King, F. Corrosion of carbon steel under anaerobic conditions in a repository for SF and HLW in Opalinus Clay; Nagra Technical Report 08-12; Nagra: Wettingen, Switzerland, 2008. Turnbull, A. A review of the possible effects of hydrogen on lifetime of carbon steel nuclear waste containers; National Cooperative for the Disposal of Radioactive Waste; Nagra Technical Report NTB 09-04; 2009. King, F.; Newman, R. C. Stress corrosion cracking of copper canisters; Swedish Nuclear Fuel and Waste Management Company Report; SKB TR-2010-04; 2010. Asano, H.; Wakamatsu, H.; Akashi, M. In Proceedings of High Level Radioactive Waste Management Conference, Las Vegas, NV, Apr 12–16, 1992; American Nuclear Society/American Society of Civil Engineers: La Grange Park, IL/New York, NY, 1992; pp 1658–1669. Akashi, M.; Fukuda, T.; Yoneyama, H. In Scientific Basis for Nuclear Waste Management XIII, Materials Research Society Symposium Proceedings; Oversby, V. M., Brown, P. W., Eds.; Materials Research Society: Pittsburgh, PA, 1990; Vol. 176, pp 525–532. Taniguchi, N.; Kawasaki, M. J. Nucl. Mater. 2008, 379, 154–161. King, F.; Kolar, M. Preliminary assessment of the stress corrosion cracking of used fuel disposal containers using the CCM-SCC.0 model; Report No: 06819-REP-01300-10103-R00; Ontario Power Generation, Nuclear Waste Management Division, 2005. King, F.; Kolar, M. Consequences of microbial activity for corrosion of copper used fuel containers – Analyses using the CCM-MIC.0.1 code; Report No: 06819-REP-01300-00120-R00; Ontario Power Generation, Nuclear Waste Management Division, 2006. King, F.; Stroes-Gascoyne, S. In Microbial Degradation Processes in Radioactive Waste Repository and in Nuclear Fuel Storage Areas; Wolfram, J. H., Ed.; Kluwer: Dordrecht, Netherlands, 1997; pp 149–162. King, F. Hydrogen effects on carbon steel used fuel containers; Report No: NWMO TR-2009-29; Nuclear Waste Management Organization, 2009. Lloyd, A. C.; Schuler, R. J.; Noe¨l, J. J.; Shoesmith, D. W.; King, F. In Scientific Basis for Nuclear Waste Management XXVIII, Materials Research Society Symposium Proceedings; Hanchar, J. M., Stroes-Gascoyne, S., Browning, L., Eds.; Materials Research Society: Warrendale, PA, 2004; Vol. 824, pp 3–9. ASM. Metals Handbook, Ninth Edition, Volume 13, Corrosion; American Society for Metals International: Metals Park, OH, 1987. Pourbaix, M. Atlas of Electrochemical Equilibria in Aqueous Solutions; 2nd ed.; NACE International: Houston, TX, 1974. Smart, N. R.; Blackwood, D. J.; Werme, L. O. The anaerobic corrosion of carbon steel and cast iron in artificial groundwaters; SKB Technical Report TR-01-22; 2001.

450

Waste Containers

71. Suckling, P.; Calder, N.; Humphreys, P.; King, F.; Leung, H. In Proceedings of the 12th International Conference on Environmental Remediation and Radioactive Waste Management, ICEM’09/DECOM’09, Liverpool, UK, Oct 11–15, 2009; American Society of Mechanical Engineers: New York, NY, 2009; paper ICEM2009-16291. 72. Taniguchi, N.; Kawasaki, M.; Kawakami, S.; Kubota, M. Corrosion behaviour of carbon steel in contact with bentonite under anaerobic condition. In Prediction of Long Term Corrosion in Nuclear Waste Systems, Proceedings of the 2nd International Workshop, Nice, France, Sept 2004; pp 24–34, European Federation of Corrosion and Andra. 73. Carlson, L.; Karnland, O.; Olsson, S.; Rance, A.; Smart, N. Experimental studies on the interactions between anaerobically corroding iron and bentonite; Posiva Working Report 2006-60; 2006. 74. ASM. ASM Handbook, Volume 13B, Corrosion: Materials; American Society for Metals International: Metals Park, OH, 2005. 75. Hultquist, G. Corrosion Sci. 1986, 26, 173–177. 76. Szaka´los, P.; Hultquist, G.; Wikmark, G. Electrochem. Solid State Lett. 2007, 10, C63–C67. 77. Hultquist, G.; Szaka´los, P.; Graham, M. J.; et al. Catal. Lett. 2009, DOI: 10.1007/s10562-009-0113-x, published online: July 28, 2009. 78. Little, B.; Wagner, P.; Mansfeld, F. Int. Mater. Rev. 1991, 36, 253–272.

79.

Stroes-Gascoyne, S.; Hamon, C. J.; Dixon, D. A.; Kohle, C.; Maak, P. In Scientific Basis for Nuclear Waste Management XXX; Dunn, D., Poinssot, C., Begg, B., Eds.; Materials Research Society: Pittsburgh, PA, 2007; paper 0985-NN13-02. 80. Masurat, P.; Eriksson, S.; Pedersen, K. Appl. Clay Sci. 2009, DOI:10.1016/j.clay.2009.01.004. 81. Aerts, S. Effect of geochemical conditions on bacterial activity; Report prepared by SCK-CEN for ONDRAF/ NIRAS, SCK(CEN-ER-75); 2009. 82. Crossland, I. In ICEM’05: 10th International Conference on Environmental Remediation and Radioactive Waste Management, Sept 4–8, 2005; American Society of Mechanical Engineers: New York, NY, 2005; paper ICEM05-1272. 83. Mattsson, H.; Olefjord, I. Werkst. Korros. 1990, 41, 383–390. 84. Mattsson, H.; Li, C.; Olefjord, I. Werkst. Korros. 1990, 41, 578–584. 85. Nakayama, G.; Sakakibara, Y.; Taniyama, Y.; et al. J. Nucl. Mater. 2008, 379, 174–180. 86. ASM. ASM Handbook, Volume 13A, Corrosion: Fundamentals, Testing, and Protection; American Society for Metals International: Metals Park, OH, 2003. 87. Marcus, P. In Corrosion Mechanisms in Theory and Practice; Marcus, P., Oudar, J., Eds.; Marcel Dekker: New York, NY, 1995; pp 239–263. 88. Marsh, G. P.; Taylor, K. J. Corrosion Sci. 1988, 28, 289–320.

5.18

Waste Glass

E. Vernaz and S. Gin Commissariat a` l’Energie Atomique et aux Energies Alternatives, Bagnols sur Ce`ze, France

C. Veyer Veyer-Consultant, St Waast la Valle´e, France

ß 2012 Elsevier Ltd. All rights reserved.

5.18.1

Introduction

452

5.18.2 5.18.2.1 5.18.2.2 5.18.2.3 5.18.2.4 5.18.2.5 5.18.2.6 5.18.3 5.18.3.1 5.18.3.1.1 5.18.3.1.2 5.18.3.1.3 5.18.3.2 5.18.3.2.1 5.18.3.2.2 5.18.3.2.3 5.18.3.2.4 5.18.3.3 5.18.3.3.1 5.18.3.3.2 5.18.3.3.3 5.18.3.3.4 5.18.4 5.18.4.1 5.18.4.2 5.18.4.2.1 5.18.4.2.2 5.18.4.2.3 5.18.4.3 5.18.4.3.1 5.18.4.3.2 5.18.4.3.3 5.18.4.3.4 5.18.4.3.5 5.18.4.3.6 5.18.4.3.7 5.18.4.3.8 5.18.4.3.9 5.18.4.4 5.18.5 5.18.5.1 5.18.5.1.1 5.18.5.1.2

Glass and Vitreous State Phenomenological Approach of Glass Glass Transition Temperature Glass Structure at Atomic Scale Polymerization and Depolymerization of the Glass Network Structure of R7T7-Type Containment Glass Crystallization Mechanisms Waste Glass Definition and Characterization The Waste Streams to Vitrify Nature and composition of HLW solutions Other kinds of nuclear waste Hazardous waste Glass Formulation Establishment of glass formation diagrams Optimization of glass formulation Validation of the reference formulation Sensitivity to chemical composition Glass Characteristics of Interest Microstructural homogeneity Physical properties Thermal stability and crystallization potential Chemical durability Long-Term Behavior of Nuclear Waste Glasses Glass Crystallization and Long-Term Thermal Stability Glass Resistance to Self-Irradiation Investigations of glasses doped with a short half-life actinide Atomistic modeling of glass self-irradiation External irradiation of glasses Nuclear Glass Alteration by Water Basic mechanisms of glass alteration Initial rate of glass dissolution Alteration rate in saturated conditions and final rate of glass dissolution Essential role of the ‘passivating reactive interphase’ Influence of glass composition Influence of groundwater and environmental materials Influence of glass fracturing Modeling glass long-term behavior Natural an archaeological analogues Conclusions on Glass Long-Term Behavior Vitrification Processes Existing Processes for Radioactive Waste Vitrification The French two-step continuous vitrification process Liquid-fed ceramic melters

453 453 454 454 456 456 456 457 457 457 458 458 459 459 461 461 461 462 462 463 464 464 465 466 467 467 468 469 470 470 470 470 470 471 471 471 472 473 473 474 475 475 477 451

452

Waste Glass

5.18.5.2 5.18.5.2.1 5.18.5.2.2 5.18.6 References

Emerging Processes for Radioactive Waste Vitrification Cold-crucible induction melters Incineration–vitrification processes Conclusions and Outlook on Waste Glasses

Abbreviations AVM

Atelier de Vitrification de Marcoule (the first industrial vitrification plant in France) BO Bridging oxygen CEA Commissariat a` l’Energie Atomique et aux e´nergies alternative (The French Atomic and Alternative Energy Commission) EXAFS Extended X-ray absorption fine structure FP Fission product GCR Gas-cooled reactor GRAAL Glass reactivity with allowance for the (model) alteration layer HLW High-level waste ICP-AES Inductively coupled plasma – atomic emission spectroscopy ICP-MS Inductively coupled plasma – mass spectroscopy LWR Light water reactor MAs Minor actinides MOX (fuel) Mixed oxide (fuel) NBO Nonbridging oxygen NMR Nuclear magnetic resonance

PUREX SEM SIMS STEM TEM UP1

479 479 480 481 482

Plutonium and uranium refining by extraction Scanning electron microscope Secondary ion mass spectrometry Scanning transmission electron microscope Transmission electron microscope The first French reprocessing plant located in Marcoule

5.18.1 Introduction Fission products (FPs) and minor actinides (MAs) produced during fuel irradiation in a nuclear reactor represent only about 5% of the weight of used nuclear fuel, but about 98% of its radioactivity. When the fuel is reprocessed, these FPs and MAs end up in concentrated solutions (called high-level waste (HLW) solutions) that are stored in tanks fitted with stirring systems and cooling facilities to evacuate the heat resulting from radioactive decay. Figure 1 presents an example of such a storage tank, made of stainless steel, in a commercial reprocessing plant.

Figure 1 Example of a high-level waste storage tank for concentrated fission product solutions.

Waste Glass

Such a storage principle can be safe for several decades. However, it requires active monitoring and maintenance, and cannot reasonably be extended for the durations required for complete decay of the activity (thousands of years). As early as the mid-1950s, the major western countries started designing plans for their nuclear waste, and work on FP immobilization was initiated at Oak Ridge (USA), Harwell (UK), Chalk River (Canada), and Saclay (France). Several materials were considered at first, with a rapid convergence on glass or glass-ceramics compositions. For instance, the first attempts at the Commissariat a` l’Energie Atomique et aux e´nergies alternative (CEA) in 1957 targeted crystals of mica-phlogopite (M2Mg6(AlSi3)2O20F4, M being an alkali or an alkaline earth metal), but this was soon abandoned because of the impossibility of incorporating all the elements of the concentrated solution within one specific mineral. During these first tests, a glassy component was frequently observed at the bottom of the crucible, with often better durability than that of the targeted mineral. As, at the same time, some favorable results had been obtained at Chalk River, where a confining glass had been obtained by melting natural aluminosilicates impregnated with FP solutions at 1350
C, glass was selected for further investigations in France.1 A new application for glass was born: glass for the containment of radioactivity. At least, the idea was born, but a long path remained to be covered from the idea to industrial deployment, to optimize glass compositions adapted for each type of FP solution, and to develop processes operable in highly radioactive environments. Similar exploratory work was performed in the other countries, which ended up 20 years later in the quasi-unanimous selection of borosilicate glass as the preferable matrix.2

5.18.2 Glass and Vitreous State Glass is one of the oldest materials known to man. During prehistoric times, man used natural glasses (volcanic) to make knives or arrowheads. The first glass actually melted by man could date back to 4500 BC. Although, in common language, the term ‘glass’ often refers to a fragile and transparent material, the scientific approach regarding vitreous state is both much wider (for instance, nuclear glasses are not transparent) and more difficult to define. This

453

chapter aims at providing the basis to understand vitreous state, by considering the aspects of the formation of a glassy structure, the major glass properties (viscosity, durability, and thermal stability), and the fine atomic scale structure of glass.3 5.18.2.1 Phenomenological Approach of Glass Why are most of natural rocks crystalline, while only a small number of them display amorphous structures (absence of diffraction peaks as evidenced by XRD)? Most mineral compounds, when in the molten state, form liquids with a low viscosity (some centipoises). (The Poise (P) is a viscosity unit commonly used in the glass industry; 1 P ¼ 0.1 Pa s1, 1 cP ¼ 103 Pa s1.) Upon cooling, these liquids easily crystallize when they reach their melting temperature. Some of these liquids, however, are very viscous in the range of their melting temperature (typically 105–107 P). Such liquids, if they are kept below their liquidus temperature (in this case, they are supercooled liquids), will tend to crystallize very slowly. If the cooling rate is faster than the crystallization rate, crystallization will not occur. During cooling, the viscosity of the supercooled liquid increases progressively until the material rigidifies: the liquid ‘vitrifies’ or transitions from supercooled liquids to the ‘vitreous state.’ A phenomenological definition of ‘glass’ could then be ‘glass is a rigidified supercooled liquid.’ This definition is nevertheless too restrictive, because a glass can be obtained by other routes, (sol–gel for instance). Several alternative approaches can be proposed:  Structural approach: absence of order in the distribution of elementary structural units at scales larger than 10–30 A˚,  Thermodynamic approach: glass is in a metastable state. It is nevertheless not unstable because the energy gap that must be overcome to bring it to its more stable crystallized state is generally significant due to the high viscosity,  Physical approach: glass is a nonporous, impermeable, isotropic, noncleavable, elastic solid with a fragile rupture behavior (absence of plastic deformation before failure),  Kinetic approach: glass is a material which transitions continuously and reversibly from liquid to solid state with temperature (Figure 2). Figure 2 provides an example of a typical evolution of viscosity between 250 and 1500
C. One can

454

Waste Glass

h = 10x P A

Solid glass (elastic)

B x = 15 Tg

Blowing (hollow glass)

x = 10

C Working range

x=5

Transformation domain (plastic glass)

Devitrification zone

D E

Devitrification domain 250

500

Liquid glass (viscous)

750

1000

1250

1500 ⴗC

Figure 2 Typical evolution of glass viscosity with temperature.

uid

Volume d

liq

ole

co er

p

Su

A

id iqu

L

B E

s

Glas

l

Crysta

C

D

Temperature Tg

Tf

Figure 3 Evolution of the specific volume V of a glass or a crystal during cooling.

distinguish an elastic solid domain below 500
C, a plastic domain between 500 and 1000
C, where the glass can be worked (blown, made to fibers, moulded, etc.), and a liquid domain above 1000
C. 5.18.2.2

Glass Transition Temperature

Figure 3 compares the evolutions of the specific volumes (inverse of densities) for a glass and a crystal with temperature. During cooling of a liquid, a sharp step is observed in the evolution of density when the material crystallizes at melting temperature. If the material does not crystallize, the specific volume continues to decrease smoothly below this melting temperature until a change of slope is observed at a

temperature Tg. This temperature is called the glass transition temperature. At this temperature, the material transitions from a supercooled liquid to a solid whose expansion coefficient (the slope of the curve) is roughly one-third of that of the liquid. Similar glass transition patterns are observed for other thermodynamic parameters such as specific heat. The glass transition temperature corresponds to a glass viscosity of 1013 P. One can then propose another definition: ‘‘glass is an amorphous solid that displays the glass transition behavior.’’ 5.18.2.3

Glass Structure at Atomic Scale

Numerous compounds can be stabilized in the vitreous structure: oxide (silicates, borates, phosphates, etc.), chalcogenide (sulfides, selenides, tellurides, etc.), ionic compounds (BeF2, ZrF4–BaF2, AlF3, ZnCl2, etc.), specific metallic alloys subjected to overhardening (Pd4Si, FeB, ZrCo, CaMg, etc.), and also organic compounds such as glycerine, polyethylene, or glucose. For instance, the shiny caramel of a tiered cake is obtained by cooling molten sugar fast enough to obtain a vitreous state; candy floss is an organic glass fiber whose Tg is around 55
C!4 The most common glass compositions are silicabased oxide glasses. Even if pure silica can vitrify by itself (application to optic fibers), all-days glasses are enriched with other oxide-based components. They include, for instance, window glass (alumina– sodium–calcium–silicate), Pyrex® (a high silica

455

Waste Glass

borosilicate), crystal-glass (silicate glass with a high content of lead oxide), optical glass (such as ‘flint’ glass with high barium oxide), nuclear glass (alumina borosilicate), and so on. The constitutive oxides of an oxide glass can be categorized into three families:  Glass network forming oxides (network formers): these oxides are able to form glass by themselves; SiO2, B2O3, GeO2, and P2O5 are the most common. In silica glass, the basic structural units are silica tetrahedra [SiO4] sharing corners. Their linking creates a continuous network with some degree of disorder (groupings in cycles of five or six tetrahedra for instance), while, in a crystalline form such as quartz, the tetrahedra are perfectly ordered.  Network modifiers: these oxides cannot give a glass by themselves. When they are coupled with network formers they are inserted into the vitreous structure and modify the properties of the material. Typically, these oxides are alkali or alkaline earth oxides (Na2O, Li2O, Cs2O, CaO, BaO, etc.). As an example, the introduction of sodium oxide in a silicate network induces the breakdown of strongly covalent Si–O bonds and their replacement by Si–O. . .Naþ bonds with a more ionic character. O |

O |

O

O |

Na+ −

O − Si − O − Si − O + Na2O → O − Si − O | | | O O O

Na+

| − O − Si − O | O

The oxygen atoms bonding two silicon atoms are said to be ‘bridging oxygen’ (BO) atoms. The introduction of two sodium atoms in the structure creates two ‘nonbridging oxygen’ (NBO) atoms. The same mechanisms may be observed with CaO, but here only one Ca2þ is needed to compensate the negative charge of the two NBO atoms. The presence of NBO atoms weakens the vitreous structure and allows decreasing melting temperature (the glass becomes less refractory) and viscosity (the glass can be poured more easily). This explains why alkali oxides are often used as fluxes. NBO atoms also loosen the network, and then help incorporating more elements in the structure. A counterpart is a decrease in chemical durability. Finally, it should be recalled that the amount of modifiers must remain limited if one wants to obtain a glass from a molten liquid. If the number of NBO atoms is too high, the liquid becomes very fluid and tends to crystallize easily upon cooling.

 Intermediate oxides: these oxides cannot give a glass by themselves. However, when mixed with network modifiers, they behave like network formers within the vitreous structure. These oxides are typically Al2O3, Fe2O3, ZnO, ZrO2, PbO, TiO2, and so on. Alumina is specifically important for the glass industry, as it improves chemical durability (for container glass, particularly). In order to behave like a network former and to form a tetrahedron similar to that of silica, the [AlO4/2] ion needs a positive charge to maintain local electroneutrality. This will be achieved by fixing an alkali ion (or one alkaline earth ion for two [AlO4/2]).This results in the following structure: O Na+ | O − Si − O− | Na+ O

O | O− − Si − O + Al2O3 → 2 O − | O

O |

O |

| O

| O

Na Si − O − Al − O

Introducing alumina into an alkali silicate glass thus ‘hardens’ the glass, as NBO atoms are eliminated according to the stoichiometry (1Naþ for 1Al or 1Ca2þ for two Al). This will allow keeping a structure slightly looser than that of a pure silica glass (as the Al–O bonds are slightly weaker than the Si–O bonds), while leaving only a small number of NBO atoms. The introduction of calcium and some alumina into the formulation allowed the transition from the poorly durable alkali glass used for medieval stained glass to the window glass used at present.  The specific role of boron: the behavior of boron oxide in the presence of alkalis differs from the general pattern described for silica. Adding alkalis to pure B2O3 or to a mixture of SiO2 and B2O3 leads at first to the formation of BO atoms, as a result of the formation of BO4 tetrahedra from the BO3 triangles initially present in the B2O3 glass (in this case, the alkali cation is used to compensate the charge of the [BO4/2] structural unit). (The elementary structural units of glass are often written [SiO4], [BO4], [AlO4] to indicate the tetrahedral environment of Si, B, or Al; however, for more detailed structural descriptions, the more rigorous notation [SiO4/2], [BO4/2], [AlO4/2], is preferred. In addition to indicating that the base atom is surrounded by four oxygen atoms, this notation stresses the fact that the oxygen atoms are shared between two tetrahedra and that, for B and Al, a negative charge exists, which will need compensation by a neighboring cation.)

456

Waste Glass

It is only for higher alkali contents that NBO atoms are formed, in the environment of boron or silicon atoms. Consequently, the addition of alkalis to a borate or borosilicate glass initially induces an increase in viscosity. It is only for higher additions that viscosity starts to decrease. This deviation in the behavior of boron when compared to silica is often termed ‘boron anomaly.’ Boron, in right amount, plays a key role in nuclear borosilicates for several reasons: (1) it helps decrease the glass melting temperature without drop of the durability as [BO4] units decrease the number of NBO by fixing an alkali ion; (2) it helps digest a number of chemical elements that would be only sparingly soluble in pure silica; (3) it prevents glass crystallization; (4) it contributes to the long-term good glass behavior by allowing the formation of a very fine gel structure (without boron) and by decreasing the final pH. 5.18.2.4 Polymerization and Depolymerization of the Glass Network On the basis of the above classification, it is observed that the progressive addition of network modifiers to silica (SiO2) leads to network depolymerization by the formation of NBO atoms. At a certain limit, the liquid is strongly depolymerized, its viscosity becomes very low, and it is not possible to obtain a glass by quenching any more. On the contrary, adding boron or intermediate oxides favors network polymerization by fixing alkalis or alkaline earths as charge compensators. The very good performances of nuclear borosilicate (moderate melting temperature coupled with good durability) come from an adequate balance between boron and intermediate elements (Al, Fe, Zr, etc.) on one side and the alkali elements on the other side, allowing a high polymerization rate as most of the alkalis are found as charge compensators rather than forming NBOs. 5.18.2.5 Structure of R7T7-Type Containment Glass The atomic structure of nuclear waste glasses has been studied using spectroscopic techniques such as nuclear magnetic resonance (NMR) or extended X-ray absorption fine structure (EXAFS), at first on simplified glass compositions (SiO2–B2O3–Na2O– Al2O3), and then on compositions with increasing complexity.

These studies show that all the intermediate cations in the R7T7 glass composition are in positions of network formers. Their substitution on Si4þ sites induces charge deficits which are compensated by a network modifier cation in order to ensure charge neutrality. The sum of negative charges cre2   ated by the (AlO 4/2), (ZrO6/2), (FeO4/2), and (BO4/2) groups in R7T7 amounts to 5.19 mol per 100 mol of elements (computed from elemental mol%). On the other hand, the sum of positive charges created by alkalis and alkaline earths is 9.44 mol%. This results in the following situation:  All the intermediate cations behave as network formers,  A somewhat limited number of network modifiers remain available to create NBOs. These considerations show that the network of this glass composition is homogeneous and very well copolymerized, owing to the good incorporation of intermediate cations in the network. The network is most probably composed of numerous mixed Si–O– M bonds (with M ¼ Al, B, Zr, Zn, and even some rare earths). This strong polymerization, as well as all the structural data, is confirmed by molecular dynamics modeling. 5.18.2.6

Crystallization Mechanisms

Crystallization in a supercooled liquid or a glass occurs via a nucleation-growth mechanism which starts by the formation of small ordered germs. The germination can be enhanced by the presence of rough interfaces in the melt. Figure 4 gives an example of nucleation and growth curves for a glass. A r b i t r a r y u n i t s

(n), (g)

(n) is the nucleation rate in number of seeds per time unit

Tg 600 ⬚C

700 ⬚C

(g) is the seeds growth rate in microns per time unit

800 ⬚C

Temperature

Figure 4 Nucleation and growth of a crystalline phase within a glass.

Waste Glass

Several crystalline phases can be formed during cooling of a supercooled liquid, according to its composition. The chemical composition of the formed crystals can be very different from those of the liquid or of the glass. For a given phase, graph (n) plots the number of supercritical germs per unit volume and time as a function of temperature. Graph (g) plots the growth rate for this phase as a function of temperature. If those two curves do not overlap, the given crystal cannot form spontaneously during cooling. In this case, when the glass is in the growth zone, there are no germs liable to grow, and when it is in the nucleation zone, the temperature is too low to allow the formed nuclei to grow. Crystallization will then be possible only after a two-stage heat treatment, one stage for nucleation and the second for growth. This is a favorable situation for the production of glassceramics with homogeneous and well controlled crystallization.5,6 On the other hand, if the two curves overlap, a risk of uncontrolled crystallization appears in the overlap region, where germs can be formed and can grow. The relative position of the nucleation and growth curves in the temperature field will then be an important parameter determining the sensitivity of the glass to crystallization; it must be studied for each of the phases liable to be formed in a given glass composition.

5.18.3 Waste Glass Definition and Characterization Vitrification is not a simple encapsulation process (as immobilization in bitumen for instance) but consists of making a new material in which the waste components are contained at the atomic scale within the matrix and can only be released by destruction of the network bonds. One major requirement is that the selected matrix should be able to incorporate all of the waste stream components in its structure. By using the flexibility brought about by the disordered and relatively loose structure of a glass, it is possible to design glass compositions able to integrate a very wide range of elements within their structure, and which are tolerant to compositional variations in the waste stream. This approach constitutes waste vitrification, where waste components are usually mixed with suitable additives and molten to give a glass wasteform. A recent development of vitrification has been the design of glass-ceramics that combine the flexibility

457

of glass formulation to digest most of the waste components with the possibility of targeting well defined crystalline phase(s) for specific waste components that may not be soluble in large amounts in glass, as molybdenum for instance. Vitrification usually involves a small number of processing steps, with a robust design, compatible with operation in a highly radioactive or hazardous environment. 5.18.3.1

The Waste Streams to Vitrify

A large variety of nuclear waste compositions have been considered for vitrification since the first attempts in the late 1950s, including not only highlevel, but also intermediate or low-level effluents. More recently, this approach has been extended to other types of inorganic hazardous waste for which other types of immobilization were not considered suitable. 5.18.3.1.1 Nature and composition of HLW solutions

FPs and MAs produced in the reactor by fission or neutron capture display a wide range of atomic numbers. One can find for instance alkalis (rubidium and cesium), alkaline earths (strontium and barium), a wide range of transition elements (zirconium, molybdenum, etc.), noble metals (ruthenium, rhodium, and palladium), chalcogenides (Se and Te), nonmetals (As, Sb, etc.), lanthanides, and actinides. The amount and composition spectrum of the FPs and MAs varies with fuel initial composition, enrichment, and burnup. The separation process used for light water reactor (LWR) oxide fuel at the commercial reprocessing plants of La Hague (France), Rokkasho Mura ( Japan), or Sellafield (UK) is a hydrometallurgical process based on plutonium and uranium refining by extraction (PUREX), where, after nitric dissolution of the fuel and a series of solvent extraction steps used for U and Pu recycling, most FPs and MAs from the fuel end up in a concentrated nitric solution (HLW solution), which constitutes the major target feed for vitrification. In addition to the isotopes extracted from the fuel, the HLW stream holds some chemicals added during reprocessing. The current PUREX-based process for LWR oxide fuel has been designed to minimize these additions, by using essentially chemicals that will not add to the waste load. However it is not possible to avoid the addition of some selected chemicals, mainly sodium, during

458

Waste Glass

ancillary operations such as solvent purification or equipment cleaning operations. Some impurities resulting from a slight corrosion of the piping (namely Fe, Cr, and Ni) or solvent degradation (phosphate) also end up in the HLW solution. In addition to commercial LWR fuel, other fuels have been reprocessed worldwide in the past, and other processes have been used or are considered for separating uranium and/or plutonium from the fuel. Fuel alloy metals (Al and Mo) or dissolved cladding material (Al, Mg, and stainless steel) may follow the HLW stream. The chemicals used to perform the dissolutions or separations can also be very diverse (mercury, fluorine, ferrous sulfamate, bismuth phosphate, etc.). In several instances, the HLW solution (which is, initially, acidic) has been neutralized by adding massive amounts of caustic to prevent corrosion of the tanks, thereby precipitating most of the FPs and MAs as hydroxides. This is for instance the case in the United States, at Hanford or Savannah River; The HLW solutions are thus quite complex and not unique, with a large number of constituents. Typically, in France, concentrated HLW solutions are nitric solutions (1–2 N) with high bg activities (several tens of TBq per liter) including suspended solids such as colloids (zirconium phosphate and cesium phosphomolybdate) and some metallic fines (insoluble residue, cladding fines generated during the fuel shearing operation). Table 1 gives examples of solutions derived from reprocessing various types of fuel from past or present reactors in France. 5.18.3.1.2 Other kinds of nuclear waste

In addition to HLW, vitrification is increasingly considered for waste of lower radioactivity content, although other, less costly, immobilization methods, such as grouting, are more common. Despite the cost, vitrification has the advantage of providing durable matrices, with significant volume reduction. For liquid effluents, a vitrification plant is for instance being designed and built at the Hanford site, USA, as part of the Waste Treatment Plant (WTP) to immobilize the low activity fraction of the waste from 177 underground tanks. This low activity fraction consists mainly of concentrated sodium nitrate and sodium hydroxide solutions, with some aluminum, chrome, sulfur, and some minor metals. In Russia, at the Radon facility near Moscow, borated low activity power plant effluents are also vitrified. Vitrification is also applied to low activity solid combustible waste, in conjunction with incineration: the combustible fraction is incinerated and the

Table 1 France

Example of HLW compositions to vitrify in

Reactor type

PWR

Gas-cooled, graphite moderated, natural uranium reactor

Fuel type

UO2

SiCrAl

UMo–MoSnAl

Burnup (MWd t1) HLW solution (l tU1) Oxide contents (g l1) Free acidity (N) Oxide composition (g l1) Fe Al Cr Ni Na Mg Zn Mo Sn P F PF oxides (g l1) Actinide oxides

33 000

4000–6000

1500–3500

660

100–120

75–100

90–100

240

0.95

1.0

0.8

13 0–2 2.3 1.9 20 – 0–1 – – 1.26 – 52.23

7–14 19–38 0–1.5 0–1.3 7–11 3–7 – – – 1–2 2–8 45.1

3.14 2.76 0.65 0.31 14 1.6 – 163.2 0.67 41.16 – 11.46

3.83

3.65

3.62

inorganic fraction, under the form of ashes, is vitrified. This is for instance the case at Ulchin, S. Korea, where a facility based on French technology has been recently commissioned to simultaneously incinerate and vitrify dry active waste and resins from nuclear power plants. At the Zwilag waste management facility, in Switzerland, a plasma-torch powered facility has been built to incinerate, melt, and vitrify low activity waste conditioned in drums. In the end, vitrification has also sometimes been considered for actinide-bearing waste or residues. In the United States, for instance, a significant program has been launched in the 1990s in the frame of the START-II program, to study the possibility of vitrifying plutonium-rich residues from the weapon industry that could not be easily made into mixed oxide (MOX) fuel. 5.18.3.1.3 Hazardous waste

Vitrification, alone or combined with incineration, provides a convenient method for immobilizing or rendering hazardous waste innocuous. For instance,

Waste Glass

in the United States, vitrification has been considered as Best Demonstrated Available Technology (BDAT) by the US-EPA (US Environmental Protection Agency) for waste containing toxic metals such as arsenic. In situ vitrification of soils has been applied in some contaminated sites in the United States. Owing to the potential for immobilizing inorganic, nonvolatile, toxic metals into a stable and durable matrix, vitrification is used in several facilities worldwide, including France and Japan, to immobilize the slag and ashes from municipal waste incineration or waste-to-energy conversion facilities. The product is completely inert and can be disposed of in a conventional disposal facility. Efforts are under way in several institutions to improve the product for reuse, mostly in the building industry (as road base or tiles for instance). Another significant application of vitrification is the destruction of asbestos: once vitrified, asbestos becomes a compact, harmless substance. Such a facility for asbestos vitrification has been operating in southwestern France since 2003. 5.18.3.2

Glass Formulation

Glass formulation is aimed at defining a compositional domain within which the matrix will display a number of required characteristics related to technological feasibility, durability, containment properties, and all the properties required for the intended use or destination of the product. For HLW glass, for instance, the product will have to ensure long-term safety in the proposed geological disposal environment. Glass formulation then consists in reaching the best compromise between a large number of

constraints: glass formation domain (solubility of the various waste components, waste loading, and homogeneous glass), technological feasibility (melting temperature limits, viscosity allowing to pour the product, minimum volatilization, and minimum corrosion of the melter), as well as stability and containment properties (thermal stability, resistance to self-irradiation, chemical durability, mechanical properties, etc.) (Figure 5). For radioactive waste immobilization, borosilicate systems represent the best compromise between the various constraints, and they have been selected as the reference matrix compositions for HLW in most countries (France, USA, UK, Japan, Germany, Belgium, etc.). Other matrices for HLW include phosphate glasses in Russia. Glass formulation involves several successive (and often iterative) steps illustrated below by the methodology used at the French CEA for HLW borosilicate glass formulation: in this instance, it was necessary to design a matrix with very good containment and long-term properties, to immobilize a highly radioactive, heat-generating, waste stream, whose composition was expected to be very stable throughout the years. In other countries, or for other waste streams, the practical organization of formulation studies may differ, but these steps are always necessary to ensure consistent and reliable waste form properties. 5.18.3.2.1 Establishment of glass formation diagrams

The glass-forming domain is determined by establishing quaternary phase diagrams with the expected four major components of the glass, to identify the

Ability to accommodate the waste Solubility (Cr, Ru, Rh, Pd, Ce, Pu, SO4, Cl) Phase separation (Mo, SO4, Cl, P) Devitrification (Mo, P, F, Mg, ...) Maximize the waste loading

Process/technology Ease of processing Melting temperature Viscosity, reactivity, residence time, electrical condition, thermal condition Additives needed

Figure 5 Waste glass formulation is a compromise.

459

Glass performance Properties for storage/disposal Thermal stability Chemical durability Resistance to self-irradiation Mechanical properties

460

Waste Glass

glass-forming regions. For waste solutions dominated by FP oxides, the four major components can be summarized as silica SiO2, boron oxide B2O3, sodium oxide Na2O, and FP oxides (considered collectively, with a chemical spectrum corresponding to the expected spectrum for the solution). This leads to ‘SON’-type glass formulations such as those selected for the LWR FP solutions at La Hague. When the waste solution is dominated by aluminum, the quaternary diagram considers SiO2, Al2O3, B2O3, and Na2O. This leads to ‘SAN’-type glass formulations, such as those selected for the gas-cooled reactor (GCR) FP solutions processed at the Atelier de Vitrification de Marcoule (AVM) facility at Marcoule (Figure 6). The glass-forming domain is established by melting a large number of glass compositions in small laboratory crucibles and performing visual or

microscopic observations. As can be seen in Figure 7, boron helps to dissolve the entire FP spectrum into the glass, prevent crystallization, and lower viscosity. The chosen boron content is however limited to the minimum needed to have a sufficient domain of homogeneous glass while keeping the best durability (for instance 18 wt% B2O3 is sufficient in the quaternary systems shown on Figure 7). Once this glass-forming region has been determined, additional constraints are used to further limit this domain, for instance, limitation of the melting temperature, which will result in limits on the refractory elements such as silica or alumina, limitation of waste loading by thermal considerations (for HLW), and limitation of network modifiers to keep an acceptable chemical durability. Each of these constraints establishes one side of the diamond in which the glass composition will be chosen. B2O3

B2O3

24% B2O3 24% B2O3 10% B2O3 10% B2O3 SiO2

Na2O

SiO2

Na2O

Ox PF

Al2O3

Figure 6 Basic quaternary glasses for gas-cooled reactor on left (SAN-type glasses) and light water reactor on right (SON-type glasses).

SiO2 SiO2 730

Ox. PF +

110 40

315 75

80

310 230

Ox.

115

Na2O 10

PF + Act.

Act.

24% B2O3

18% B2O3 Nonglassy state Heterogeneous glasses Homogeneous glasses Viscosity (P [= dPa.s] )

Figure 7 Effect of boron oxide content on the vitrification domain in quaternary glasses. Domains of glass formation into the quaternary diagram SiO2–Na2O–OxPF–B2O3 for two different B2O3 contents.

Waste Glass

5.18.3.2.2 Optimization of glass formulation

Optimization consists in adding or substituting various additives (such as Li2O, CaO, ZnO, ZrO2, etc.) to improve the matrix properties. Among all properties of interest, the waste load (considering the final glass homogeneity), the viscosity of the molten glass, and the chemical durability have to be specially mentioned. Depending on the choice for industrial process, other properties may be targeted in addition. As an example, small amounts of lithium may be substituted for sodium to decrease melt viscosity, without affecting the chemical durability. Aluminum and calcium oxide may be substituted for a small amount of silica to improve durability. It has to be pointed out that those optimizations result from sharp compromises and high-level expertises in mixing the appropriate oxides in a synergic way (for instance too high aluminum content would enhance short term durability but would decrease long term durability; it would also increase the risk of crystallization of undesirable phases; controlled etc.). At the end of this phase, a reference composition is proposed for a given effluent composition and a given melting technology. Dozens of compositions are published in the literature, for a very wide range of waste compositions, in many countries. Table 2 gives the composition of the French reference glass for the HLW solutions derived form processing 33 GWdt1 LWR fuel at La Hague: the so-called ‘R7T7’ composition, named after the vitrification facilities R7 and T7 at La Hague plant. 5.18.3.2.3 Validation of the reference formulation

Once a reference composition has been defined at lab-scale for a specific simulated waste stream, it needs to be validated for actual industrial application. (During the development stage of reference matrix, as the main consideration is on the basis of chemical incorporation, radioisotopes issued from FPs and MAs are replaced by their inactive representatives.) To this effect, several additional programs are performed:  complete characterization of the reference composition and determination of all its important properties,  validation of technological feasibility in industrial conditions by performing long-duration demonstrations on an inactive industrial-scale pilot facility, and confirmation of the maintenance of

461

Table 2 Chemical composition range of the French R7T7 glass Chemical composition range of R7T7 glasses produced in the AREVA – La Hague plant workshops Oxides

SiO2 B2O3 Al2O3 Na2O CaO Fe2O3 NiO Cr2O3 P2O5 Li2O ZnO Oxides (FP þ Zr þ actinides) fines suspension Actinide oxides SiO2 þ B2O3 þ Al2O3

Specified interval for the industry (wt%) Min

Max

42.4 12.4 3.6 8.1 3.5

51.7 16.5 6.6 11.0 4.8 <4.5 <0.5 <0.6 <1.0 2.4 2.8 18.5

1.6 2.2 7.5

>60

Average composition of industrial glasses (wt%)

45.6 14.1 4.7 9.9 4.0 1.1 0.1 0.1 0.2 2.0 2.5 17.0 0.6 64.4

the characteristics defined at lab-scale on the product made during scale one demonstrations, and  validation of the representativeness of the inactive glasses made in the laboratory by making a glass of the same composition with actual radioactive waste in a hot laboratory and determining its characteristics. The fabrication of active samples is also necessary to study the effect of self-irradiation and the specific behavior of some radioactive elements. 5.18.3.2.4 Sensitivity to chemical composition

In industrial situations, one must expect day-to-day variations in the composition of the solution to be vitrified, and also process upsets or variability which may affect the product. It is then necessary to determine what is the flexibility of the formulation towards these variations, in order to provide a wide enough operational domain, while keeping acceptable glass properties. For this, a sensitivity study is launched, to study systematically the effect of selected composition variations on glass properties, and establish acceptable limits for these components.

462

Waste Glass

For instance, in order to be pourable, the glass must display a viscosity lower than about 100 P at the melting temperature. This will limit the possible increase in refractory elements such as SiO2 or Al2O3 or the decrease in fluxing components such as Na2O or Li2O. On the other hand, in order to keep an acceptable chemical durability, the increase in alkali oxides will be limited. In such a multidimensional space, with more than ten constituents of interest in a formulation, sensitivity studies can become very cumbersome and involve a large number of crucible melts and the associated characterization measurements. In order to minimize the number of experiments while obtaining a reliably acceptable composition domain, statistical tools, property–composition models, and experiment design are implemented. These allow identifying acceptability limits and modeling matrix properties within these limits. For illustration, Table 2 displays the range of acceptable composition for the R7T7 glass produced at La Hague. One can see that the acceptable variations may be small for some components, and much larger for other components. 5.18.3.3

(Figure 8). The black color, analogous to the color of some volcanic glasses, is the result of the wide diversity of chemical elements included in the glass and, more particularly, transition metals and rare earths, which absorb light over a wide range of wavelengths. By visual observation, some small bubbles and some cracking associated with stress relaxation during cooling can be observed on industrial-size blocks. Conservative durability modeling accounts for the impact of this physical heterogeneity (see Section 5.18.4.3.7). Under the microscope (optical or scanning electron microscope, SEM), a small volume of chemical heterogeneities can be observed, essentially noble metal inclusions (Pd, Rh, or Pd–Te alloy), ruthenium oxide, nickel, zinc, and iron chromites (Figure 9). This type of inclusion represents a small volume of the matrix, and has been shown to have no effect on the long-term behavior of R7T7 glass. For other waste glass compositions, it is necessary to make sure that the precipitated phases do not alter the matrix properties, such as chemical durability. For instance, for Fe- and Cr-rich glasses, spinels have been observed, which do not alter glass chemical durability. On the other hand, for Al-rich glasses, it is necessary to avoid the formation of nepheline

Glass Characteristics of Interest

Glass characterization is the determination of all the physical and chemical properties of the product. Generally, and as observed in many laboratories, the products made with simulated waste by substituting the inactive counterparts for the radioactive components of the waste display properties that are very representative of those of the actual radioactive product. Glass characterization is thus generally performed on a simulated, nonradioactive glass for convenience, and validated in the end on radioactive glasses (see Section 5.18.3.2).

Figure 8 Obsidian glass and nuclear waste glass (cannot be distinguished with the naked eye).

20 μm

5.18.3.3.1 Microstructural homogeneity

As it is generally more difficult to characterize and demonstrate a favorable long-term behavior for a chemically heterogeneous material, the objective is usually to produce a homogenous glass, or a glass that contains, after cooling of industrial-size blocks, only a small amount of harmless precipitates or insoluble phases. Nevertheless, the notion of homogeneity is relative and depends on the resolution of the measuring instrument used. Upon visual observation, HLW glasses such as the R7T7 composition are homogenous, black, and shiny

Palladiumtellure

Ruthénium oxide Chromites

Figure 9 Scanning electron microscope observation on a zone of platinoI¨de concentration into the glass.

Waste Glass

(aluminosilicate) crystals on cooling, as these tend to be detrimental to resistance to aqueous leaching. In the end, the presence of precipitates or insoluble alloys in the molten glass may promote settling in the melter and have detrimental effects on the process (pouring difficulties and electrical short-circuiting in some types of melters). For the La Hague process, stirring allows processing melts with significant amounts of insoluble noble metals. Another form of heterogeneity is phase separation, whereby the glass separates (during melting or during cooling) into two or more different liquid or glassy phases. This type of phenomenon has to be carefully controlled to avoid the formation of a phase that is ‘weaker’ than the other, inducing some loss of general durability. Nevertheless, in some specific instances, it may be acceptable to formulate microcrystalline or phaseseparated glasses, if it can be demonstrated that this does not alter chemical durability. Such an approach has been retained in France to formulate glasses to immobilize molybdenum-rich solutions resulting from the processing of U–Mo fuel. Lower resolutions of observations (such as the one reachable by transmission electron microscope, TEM, as well as structural techniques) are not relevant to link the heterogeneities at that scale to a significant effect on durability, but are of great use for understanding. 5.18.3.3.2 Physical properties

The physical properties of nuclear waste glasses are usually quite comparable to those of classical industrial glasses,7 as illustrated on Table 3. Most of them have to range in between a minimum and a maximum value to suit industrial processes. The density of waste glasses is slightly higher than those for the industrial glasses, owing to the presence of heavy metals. The viscosity at 1100
C is much lower. This is because nuclear glasses are

Table 3

463

formulated to be poured at 1100
C while the melting temperature of industrial glasses is significantly higher. It should also be noticed that the presence of noble metals or other heterogeneity in the nuclear glass can significantly modify its rheological behavior. Glass transition temperature occurs in similar ranges, although it is slightly lower for nuclear waste glass. The thermal expansion coefficient of nuclear waste glasses is similar to that of window glass, and significantly higher than that of Pyrex: Pyrex is formulated specifically to resist thermal shocks. Thermal conductivity is similar for all three types of glass. It should be noted that this low value is significant in nuclear waste glasses which hold heat-generating FPs: temperature at the centre of the glass block will need to be controlled. Young’s modulus, which characterizes rigidity, is slightly higher for nuclear waste glass, while fracture toughness is similar. Electrical resistivity, not included in the table is another significant parameter if the glass is heated by direct electromagnetic induction or within electrodes (in the case of Joule Melter technology, this property can be neglected). Electrical resistivity decreases when temperature increases and when the alkali content of the glass increases. Electrical resistivity depends mainly on ionic diffusion in the material. It decreases from around 1.5  104 at 500
C to a few O cm at 1200
C. Other parameters not included are the thermal conductivity and the redox behavior. Thermal conductivity impacts directly the energetic efficiency in the melter. If too high, the thermal losses out of the melter make the process energetically (and economically) inefficient; if too low, the energy transmitted to the glass is nonhomogenous. Regarding redox behavior, the presence of multivalent elements (Ce oxide for instance) in the glass composition may induce foaming at high temperature due to their reduction. The monitoring of oxygen partial pressure in the

Comparison of physical properties of selected nuclear and industrial glass compositions

3

Density (kg m ) Viscosity at 1100
C (P) Tg (
C) Expansion coefficient (106 K1) Thermal conductivity (W m1 K1) Young’s modulus (1010 Pa) Fracture toughness KIc (106 Pa m1/2)

Typical HLW glass compositions

Pyrex glass

Window glass

2.50–2.75 50–150 510 8.3–9.9 1.0 8.4–8.6 0.75–0.95

2.28 80 000 565 3.2 1.09 6.1 0.85

2.46 4000 527–547 9.3 1.05 7.3 0.70–0.80

464

Waste Glass

molten glass, and if necessary the use of redox buffer (FeII/FeIII) as well as physical means can overcome this inconvenience. 5.18.3.3.3 Thermal stability and crystallization potential

Devitrification is the process by which the glass loses part or all of its glassy nature through crystallization. It depends on the composition of the glass and its thermal history. For instance, increasing the levels of FPs, noble metals, molybdenum, phosphorus, chrome, nickel, iron, or magnesium can favor crystallization in a nuclear waste glass. Furthermore, the time needed to reach the glass transition temperature (Tg) from melting temperature is also important (this depends mainly on glass thermal conductivity and specific heat, canister geometry, and process parameters such as pouring rate). However, it is considered that once Tg is reached in cooling conditions, the devitrification process is kinetically frozen (cf. Section 5.18.4.1). Devitrification studies are on the basis of subjecting glass samples to short-duration heat treatments (around 15 h) at stabilized temperatures and observing the heat-treated samples under the microscope to detect, observe, and quantify the crystals formed. Several indicators are determined:  starting crystallization temperature: the temperature below which crystalline phases can be observed in the bulk of the sample after about 10 h of isothermal heat treatment,  crystallization temperature range: range in which these crystals are observed,  maximal crystal growth rate (generally expressed in mm mn1),  crystallization potential: characterizes the ability of a composition to devitrify. It is the maximum percentage of crystals that can form after a heat treatment. This can vary from 0% (pure borate glass) to 100% (lithium disilicate glass). It is about 4% for R7T7 glass. XRD and X-ray microprobe are used to identify the crystalline phases formed within a glass while image analysis and quantitative XRD are used to evaluate the percentage of phases formed. The amount of crystals that form in an actual large size industrial glass block is different from the maximal values found at lab-scale: indeed, the thermal profile in the glass canister involves a continuous decrease of temperature, and is different in the various parts of the canister (close to the canister wall,

cooling is faster than in the centre of the glass block). Several approaches can be used to bracket an estimation of the amount of crystals in the industrial glass block. In France, for instance, the maximum possible amount of crystallization is determined on laboratory samples, by subjecting them to a heat treatment designed to promote crystallization (5 h at 610
C – nucleation temperature – and 100 h at 780
C – maximum growth temperature – this cannot happen in a real glass block) and it is postulated that the amount of crystals in the glass block cannot be higher than the fraction determined in this way (which is in fact quite small for the R7T7 glass). In the United States, where the glass blocks are quite large, crystallization studies are performed by establishing systematic TTT (time–temperature– transformation) diagrams and by considering the cooling profile at the centre of the canister (CCC, canister centerline cooling curve), which is the slowest cooling part of the canister. Whatever be the approach, the important aspect is the fact that crystallization must not be detrimental to glass durability. It is thus necessary to avoid depleting the glass matrix of elements that are favorable to durability, such as silica or alumina. 5.18.3.3.4 Chemical durability

Water is the major cause of glass alteration and radionuclide (or hazardous metal) dispersion during the life of the product, and more particularly during geologic disposal. The resistance of glass to aqueous alteration is generally called chemical durability. It is the essential property required for a containment matrix. It is also essential for classical industrial glass compositions (container glass, window glass, etc.) and as such is also considered during their formulation. The process by which glass constituents are washed into water is called leaching, and it is the combination of a variety of mechanisms described in Section 5.18.4. Chemical durability is assessed by various ‘leach tests’ during which glass samples are contacted with aqueous solutions, under a large range of experimental conditions. The leaching behavior of glasses varies according to glass composition, test conditions, and time. There is no unique evaluation of chemical durability, and the performance of different glass compositions can only be compared under the conditions of a given test. In order to fully evaluate the behavior and performance of a given glass, it is advisable to understand the various mechanisms that are involved during its interaction with water and the

Waste Glass

environment, and to design testing to obtain the parameters that allow modeling this performance. Several types of leach tests can be listed:  Static tests, during which glass samples, either monolithic or crushed to powders, are exposed to a solution and left standing there for the duration of the test. During these tests, the components dissolved from the glass progressively accumulate into the solution and are left to interact between one another and with the sample. Two of these tests have been normalized in the United States: the so-called ‘MCC-1’ test on monolithic samples (also known as ASTM-C1220) and the ‘PCT’ test on crushed glass samples (also known as ASTM-C1285). Both these tests are available with several variants (temperature, volume of solution, type of solution, etc.).  Flow-through tests, during which the sample is exposed to a continuous flow of fresh leachant to prevent the accumulation of reaction products into the solution, thereby testing the ‘initial’ alteration. Among those tests, one can list the ‘Soxhlet’ test (also known as ISO 16797:2004), by which a monolithic sample is exposed to a continuous flow of condensed water at 100
C recirculated in a distillation apparatus, or the ‘flow-through’ test (also known as ASTM C1662-07) by which the sample is inserted in a column or a container and subjected to a continuous flow of fresh solution.  ‘Integral tests’ or ‘service conditions tests’ or ‘tests including environmental materials’ are tests designed to account for the overall service environment expected during the life of the product. Glass alteration may be measured in three different ways:  Elemental analysis (ICP-MS, ICP-AES, etc.) of the leachate (solution after contacting the sample), in order to obtain information on the leaching kinetics for each element, as a function of time.  Weighing of the sample, to determine the overall weight loss.  Analysis of the alteration rind on the sample (SEM, secondary ion mass spectrometry, SIMS, scanning transmission electron microscope, STEM, etc.) in order to study the thickness of the altered layer and to understand the fate of the various elements released from the glass and retained in the alteration layer. The quantitative expression describing the rate of release of an element in water is the ‘leach rate,’ usually expressed in g m2 d1, and normalized with

465

respect to the mass fraction of the given element in the glass. LRðiÞ ¼ mi =xi St where mi is the weight (or activity) of element i released into solution, xi is the mass fraction (or specific activity) of this element in the glass, S is the area of the glass surface exposed to the leachant, and t is time. During a static test the ‘normalized mass loss’ is often used as an indicator: NLðiÞ ¼ mi =xi S;

expressed in g m2

The evolution of NL(i ) ¼ f (t) describes the overall kinetics of the process for the given experimental conditions. This curve is the basis for all the glass behavior studies. If an element (i) is a good alteration tracer, that is, congruently released with glass dissolution and not trapped back into the alteration products, then the equivalent thickness of altered glass Eth can be calculated by dividing the normalized mass loss by the glass density r: Eth ðiÞ ¼ NLðiÞ=r Boron is most often a good tracer of glass alteration. Sodium, lithium, and molybdenum are good tracers in dilute media, but they may be integrated in alteration products in more concentrated media. More information on glass alteration mechanisms is provided in Section 5.18.4. Exhaustive leaching characterizations are performed on inactive but chemically representative samples; some tests are performed on fully active material to check the impact of radioactive environment on leaching behavior.

5.18.4 Long-Term Behavior of Nuclear Waste Glasses The main phenomena that could alter glass containment properties over the long term are heat (for HLW only), radiation damage, and alteration by water. Their occurrence is not expected at the same time scale (Figure 10). For instance, the risk of crystallization is principally limited to the thermal phase, that is, in interim storage over the few decades during which the maximum glass temperature will decay from about 400 to <100
C. For thousands of years, the glass matrix is expected to remain dry and the major potential

466

Waste Glass

glass alteration mechanism is self-radiation damage. Can it change the glass containment properties when, after a few thousands of years, canister and over-pack breach, and glass alteration by water start? Many studies on glass self-radiation damage address this question. Eventually, in the very long term, the rate of radio nuclides released into the near field will be controlled by the rate of glass alteration by water. Worldwide, for over 30 years, large research efforts have been conducted to understand all the mechanisms of glass alteration by water and to develop comprehensive models, and to adapt them for the evaluation of repository performance. (A repository is the final destination of HLW glass, a disposal site selected and engineered to definitively isolate the radioactivity contained within the waste from the biosphere and man.) 5.18.4.1 Glass Crystallization and Long-Term Thermal Stability Thermal stability constitutes one of the prime criteria for glass selection. It underlies the preservation of a homogeneous glass over time. Theoretically, glass could naturally evolve to a crystalline state that is thermodynamically more stable. But such thermally induced transformation gets dramatically slow (or even stops), when the glass is maintained at temperatures lower than the glass transition temperature (Tg). Predicting thermal stability at low temperature and in the long term therefore requires experiments performed in supercooled liquids, as well as modeling. For nuclear glasses, the main work in this field was performed by Orlhac,8 which helped confirm the thermal stability of R7T7 glass in the very long term. Devitrification experiments conducted on this glass9 made it possible to identify three major

100

crystalline phases (CaMoO4, CeO2, and ZnCr2O4) and two minor phases (albite NaAlSi3O8 and silicophosphate) between 630 and 1200
C. Yet, their crystallization remains limited (a maximum 4.24 wt%), as these phases consist of minor glass constituents. Even after a heat treatment designed for a maximum crystallization (100 h at 780
C), no change can be observed in the main properties of the nuclear glass (chemical durability and mechanical properties). Plotting the nucleation and growth curves of these phases highlighted several essential points:  Nucleation sharply emerges during the first few hours of the treatment, and then stops beyond this period of time. Nucleation is heterogeneous, inducing crystallization on the preexisting active sites. Moreover, nucleation curves are strongly amplified and shift to lower temperatures in the presence of insoluble noble metal particles;  Seed crystal growth is very low, and, after a few dozen hours, displays a saturation phenomenon;  Strong nucleation coupled with slow growth globally leads to a material which can hardly be devitrified (Figure 11). The stability of high-level R7T7-type nuclear glass at low temperature and in the long term was then investigated by modeling. The mathematical model selected is on the basis of the KJMA theory (KJMA, for Kolgomorov, Johnson, Mehl, and Avrami), and describes the transformation kinetics as a function of time and temperature.10 Atomic diffusion is the main factor which limits crystallization, as demonstrated by measuring the diffusion activation energy. Viscosity is therefore the key parameter which determines the nucleationgrowth kinetics in glass: it is this very parameter which conditions diffusive atomic transport in the silicate-based liquid. Consistently, nucleation-growth kinetic processes can be determined by means of

1000

10 000 Time (years)

Figure 10 The sequence of alteration of a vitrified waste package.

Waste Glass

467

Maximum theoretical crystallized fraction (wt%) 0.6%

ZnCr2O4 zincochromite

1%

2.44%

CeO2 cerianite

CaMoO4 powellite

Total 4.24% NaAISi3O8 albite Silicophosphate

600

700

800

900

1000

1100 1200 Temperature (⬚C)

Figure 11 Temperature range for the nucleation and growth of the main crystalline phases likely to be formed after a devitrifying thermal treatment of a borosilicate nuclear glass, which confines high-level effluents arising from Uranium Oxide (UOX) fuel treatment. From CEA DEN Monographs ‘‘Nuclear Waste Conditioning’’. Etienne Vernaz, E´ditions du Moniteur: Paris, 2009; ISBN 978-2-281-11380-8.

independent viscosity measurements in a broad range of temperatures. The model validation was achieved under isothermal conditions on a simplified barium disilicate glass, known for its homogeneous, fast crystallization. Applying this model to the R7T7 glass shows that periods of several millions of years are required for the three main phases to be completely crystallized at any temperature below 600
C. Clearly, if during the slow cooling of large industrial block there is no, or minute, crystallization in the temperature range 900–600
C, then there will be neither any other crystallization, nor crystal growth in the long term, for kinetics reasons. These results confirm the thermal stability of actual HLW confining glasses. 5.18.4.2 Glass Resistance to Self-Irradiation The main sources of irradiation in nuclear glasses are a-decays from actinides, b-decays from FPs, and g transitions accompanying a- and b-decays.11 Alpha disintegrations are characterized by the production of a heavy recoil nucleus and the emission of a light a-particle, yielding a helium atom at the end of the path. Recoil nuclei (RN), shedding large amounts

of energy over a short distance result in atom displacement cascades, thus breaking a large number of chemical bonds. Alpha decays are thus the main cause of atomic displacement. On the other hand, b disintegrations produced by FPs and g transitions lead mainly to electronic interactions (electron excitation and ionization) with the glass network atoms but not to atomic displacements. The effects of self-irradiation have been studied by investigation of glasses doped with short half-life actinides, by atomistic modeling, and by external irradiations. 5.18.4.2.1 Investigations of glasses doped with a short half-life actinide

This investigation method is the most representative of nuclear glasses aging: isotropic irradiation is produced in the whole of the glass volume preserving the electrical neutrality (unlike external irradiations), a particle and the recoil nucleus are produced allowing electronic interactions and atomic displacements, and eventually helium builds up in the glass, exactly as it will in the real case. The effects of a disintegrations were mainly investigated through studying 244Cm-doped-glass glasses that can integrate within a few years doses

468

Waste Glass

equivalent to those to be delivered to the nuclear glass during thousands of years under disposal conditions.12 Figure 12 shows the ‘DHA’ Atalante laboratory. It is an example of a shielded line at Marcoule (France) devoted to HLW studies, where actinidedoped glass samples are fabricated to carry out ‘accelerated’ studies of the effects of self-irradiation. The inset in the figure shows a 238Pu-doped glass block manufactured in the Vulcain laboratory in 1975. Results produced all over the world, in France,13 UK,14 US,15 or Japan16 are quite consistent. Because of the effect of a-decay, the glass density decreases

slightly (Figure 13) and its mechanical properties appreciably improve, especially fracture toughness that characterizes glass resistance to cracking. The variations in these properties reach a saturation level and stabilize beyond 21018 a g1. Up to a dose of 1019 a g1, no helium accumulation (He bubble) is observed. No significant change in glass durability is observed either. Furthermore, there is no dose rate effect, as variations can be reproduced among the various glasses under study that exhibit quite different integration rates, spreading over four orders of magnitude. 5.18.4.2.2 Atomistic modeling of glass self-irradiation

The second approach to understand radiation damages focuses on atomistic modeling. In particular, molecular dynamics can provide insight into the ballistic effects induced by the deceleration of RN emitted at the end of a decay. Numerous studies conducted on simplified glasses representative of the basic matrix nuclear glass (SiO2, B2O3, Al2O3, Na2O, and ZrO2) demonstrated the remarkable capacity of this type of glass to restore its structure following the passage of a recoil nucleus. The following conclusions could be drawn from the whole of the calculations performed in relation to individual cascades in glasses:

Figure 12 The Atalante laboratory for high-level waste ‘DHA.’

a-decay dose (α g−1) 1018

2.1018

3.1018

4.1018

5.1018

Density variation (%)

0.1 0

244Cm-doped

glass (0.04%)

244Cm-doped

glass (0.4%)

−0.1

244Cm-doped

glass (1.2%)

244Cm-doped

glass (3.25%)

Fitting by an exponential law

−0.2 −0.3 −0.4 −0.5 −0.6 −0.7

Figure 13 Evolution of the density of curium-doped glasses with the a-decay dose. From CEA DEN Monographs ‘‘Nuclear Waste Conditioning’’. Etienne Vernaz, E´ditions du Moniteur: Paris, 2009; ISBN 978-2-281-11380-8.

Waste Glass

Figure 14 Evolution of displacement cascade from the initial glass to the reconstruction of the glass structure after the ballistic phase. From CEA DEN Monographs ‘‘Nuclear Waste Conditioning’’. Etienne Vernaz, E´ditions du Moniteur: Paris, 2009; ISBN 978-2-281-11380-8.

 displacement cascades take place in two separate steps (Figure 14): – The ballistic stage during which collisions between atoms take place as a whole. This phase coincides with the strong heating of the matrix and a depolymerization of the structure by interatomic bond breaking. In parallel, a decrease in atom density can be observed within the cascades; – The relaxation stage during which glass structure reconstruction takes place. The glass structure then experiences significant reconstruction to a state close to its initial state, but still with a slight structural depolymerization and a slight swelling on the whole. In Figure 14, four stages of the evolution of displacement cascade can be seen. In the top left corner, the initial glass containing the uranium atom (light blue atom) to be accelerated with 800 eV energy (t ¼ 0 ps). In the top right corner, the start of the ballistic phase induced by the uranium projectile (t ¼ 0.013 ps). In the bottom left corner, the final step of the ballistic phase when the maximum number of broken bonds can be observed (t ¼ 0.038 ps). In the bottom right corner, reconstruction of the glass structure after the ballistic phase (t ¼ 0.25 ps). A model of cumulative local quenching was built from these data in order to help explain the origin of

469

the small evolution observed under irradiation, as well as the origin of their stabilization under high doses.17 As displacement cascades accumulate, the glass structure is fully destabilized by nuclear interactions. Then the material can be quickly reconstructed without any external energy and its structure is close to that of glass frozen at high temperature, which results in the observed evolutions. When the whole of the glass volume has been damaged once by the displacement cascades, any new a disintegration produced will again temporarily destabilize the structure, but the latter will be rebuilt in the same way as after the first damage. So the glass no longer undergoes significant change, which could explain why its properties are stabilized beyond a given dose. It is worthwhile mentioning that the saturation dose experimentally observed in relation to macroscopic property evolutions (as seen on Figure 13) coincides with that required for full glass damaging by displacement cascades, which corroborates with the proposed model. As a conclusion, the insignificant evolution of nuclear glasses under a self-irradiation with respect to crystallized minerals could be explained by the self repairing properties of the glass structure. 5.18.4.2.3 External irradiation of glasses

This complementary type of investigation is based upon the use of nonradioactive glasses in which irradiation stress is simulated by external irradiation techniques (neutrons, heavy ions, electrons, and g). The major disadvantages of this experimentation type consist of the upsetting effects of injected high dose rates in low irradiated volumes. Today, accurate knowledge of these offsets allows relevant effects to be sorted out from experimental artifacts and the results obtained are quite comparable to those obtained with other investigation methods. This type of study was undertaken as early as the 1980s to evaluate the effects of b disintegrations. Glasses with chemical compositions representative of the industrially produced nuclear glasses were thus irradiated by electrons during 1 year up to doses equivalent to those received in about 1000 years of disposal, that is, 70% of the total bg dose received in 1 My. These irradiations have not entailed detectable modifications of the macroscopic properties (density and mechanical properties). In addition, glass still displays a homogeneous microstructure after irradiation.

470

Waste Glass

5.18.4.3

Nuclear Glass Alteration by Water

The mechanisms which control nuclear glass leaching kinetics have been investigated worldwide for more than three decades. This large accumulated knowledge allows building computational models likely to be used for performance assessment of a geological repository. These models have to be applicable to all the vitrified waste packages industrially produced, taking into account many different environmental conditions.

With an activation energy of about 75 kJ mol1, r0 ranges over seven orders of magnitude between the temperatures of 4 and 300
C. This is why natural glasses exhibit very low alteration at room temperature even after millions of years but high alteration under hydrothermal condition (‘hyaloclastites’). This value explains also the inherent difficulty of measuring r0 at room temperature, (alteration of about 1 nm per day hindered by interdiffusion) while a ‘Soxhlet’ measurement operating at 100
C will allow a prompt measurement, representative of the sample in its mass.

5.18.4.3.1 Basic mechanisms of glass alteration

5.18.4.3.3 Alteration rate in saturated conditions and final rate of glass dissolution

In contact with water, the main alteration mechanisms of borosilicate glass are the following18,19:

In a closed system or under conditions in which the water renewal rate is very slow, such as in the case of a geological repository, apparent silica saturation is observed in the leachate and a strong ‘rate drop’ is systematically evidenced for borosilicate glass. (‘Leachate’ refers to the leaching solution after contact with the glass sample.) This rate drop has been related both to affinity effects (a decrease in the hydrolysis rate coupled with an increase in the concentrations in solution), and to the formation of an alteration gel standing as a diffusive barrier between glass and solution.20 Once these ‘saturation conditions’ are established, a steady rate of glass alteration is generally observed. This ‘residual rate’ seems to be related to the phenomenon of gel dissolution and to the rate of secondary phase precipitation. For most nuclear borosilicate glass composition this final rate is very slow (about 5 nm year1 at 50
C for R7T7-type glass). The origin of such a small residual rate of glass alteration observed in saturation conditions seems to be related to the persistence of a weak dissolution rate of the gel probably itself resulting from the slow precipitation of secondary phase (mainly phyllosilicates if pH does not exceed 10).15 In some specific cases ‘alteration resumption’ can be observed, once a residual rate regime has been established. This so-called ‘renewed alteration’ is observed only for very alkaline conditions (equilibrium pH higher than 10) and specific glass composition (with high Al/Si ratio). This can be related to a high precipitation rate of some specific secondary phases (zeolites).21

 Exchange and hydrolysis reactions involving the mobile glass constituents (alkalis, boron, etc.) rapidly occur during the initial stages.  Slower hydrolysis, especially of silicon, drives the initial glass dissolution rate.  The in situ condensation of many hydrolyzed species (Si, Zr, Al, Ca, RE, etc.) results in the creation of an amorphous gel layer at the glass/solution interface regardless of the alteration conditions. This layer is more or less reorganized by hydrolysis and condensation mechanisms according to the environmental conditions.  This amorphous layer can soon constitute a barrier against the transport of water toward the glass and of solvated glass ions into solution. The existence of this transport-inhibiting effect rapidly causes this layer to control glass alteration when water renewal becomes very low.  Some glass constituents released from the glass during process can precipitate as crystallized secondary phases. The precipitation of these crystallized phases within or on top of this amorphous layer or in solution, can sustain glass alteration by consuming the elements that form the protective barrier. 5.18.4.3.2 Initial rate of glass dissolution

The initial dissolution rate r0 is the hydrolysis rate obtained in pure water when no diffusion barrier slows down the kinetics of alteration. It is an intrinsic property of the material that characterizes its chemical durability in water. This rate depends mainly on the glass composition, the temperature, and the pH. It is the maximum rate of glass alteration for a given temperature and pH.

5.18.4.3.4 Essential role of the ‘passivating reactive interphase’

Historically, this apparent saturation state described above was expressed in equations as if equilibrium with the fresh glass could be achieved. Today it is

Waste Glass

considered that a saturation state can only be achieved with respect to a hydrated layer. As saturation is approached in solution, the rate of condensation of many gel forming elements (Si, Al, Zr, Ca, etc.) increases, allowing the formation of a thin amorphous layer. Frugier et al.19 proposed the term ‘passivating reactive interphase’ (PRI) to take into account the fact that not all of the gel layer becomes passivating but only a thin inner layer in which a high condensation rate has led to closed porosity.22 Diffusion coefficients in the PRI are consistent with diffusion in solids with values of the order of magnitude of 1020 m2 s1. Furthermore, Monte Carlo modeling of the gel layer formation by hydrolysis and condensation mechanisms allows describing the conditions for which porosity closure is reached, in good agreement with experimental data.23 It will be noticed that phosphate glass that can in some cases24 display quite low initial alteration rates is not expected to form any PRI, and therefore, no large rate drop can be expected in saturation condition. 5.18.4.3.5 Influence of glass composition

Within a given domain, glass dissolution kinetics strongly depends on its composition.25 Some nonlinear effects have been evidenced on the basis of semiempirical statistical methods26 or, in a few cases, fully explained using experimental approaches and Monte Carlo numerical model. A given element usually modifies specifically each kinetic regime. As an example, Ca has no effect on the forward rate but strongly favors the rate drop as it is incorporated into the PRI in contrast to Mg that is incorporated in secondary clay minerals acting as a silicon sink and promoting higher dissolution rate. From an operational point of view, rates of actual glass or rates corresponding to the best or the worst glass within a given domain are calculated from an empirical equation built from a limited number of tested glasses. Considering the R7T7 domain, forward rate ranges from 1.6 to 4.1 g m2 d1 at 100
C and residual rate from 2.7105 to 3.2104 g m2 d1 at 90
C in initially pure water. 5.18.4.3.6 Influence of groundwater and environmental materials

In a geological repository, water chemistry initially in equilibrium with the host rock will be disturbed by engineered materials (stainless steel canister, overpack, liner, concrete, etc.) and also by the heat produced by the glass canister itself. As a consequence, it is expected that glass dissolves in an open medium in which temperature, water flow, and composition

471

evolve with time and space, at least during the first ten thousand years (this time depends on the host rock and the disposal design). Chemical elements brought by water, like Ca, Mg, organic matter, etc., may influence glass dissolution mechanisms, for example, by promoting the PRI condensation,27 increasing the hydrolysis rate of the silicate network,28 or the allowing the precipitation of secondary crystalline phases.29 In a fractured rock environment, such as a granitic disposal site, the water renewal rate will be the main environmental parameter. In a clay environment, with no flow rate (or too low compared to diffusion), the predominant effects of environmental materials will be silica sorption onto likely oxide and hydroxide minerals, and low precipitation of silicate minerals that act as silica sinks.30 In any environment, these phenomena can also be expected with iron corrosion products. This kind of reaction will maintain a high glass dissolution rate until the close environment of the glass is saturated. Beyond this transient regime that can be investigated by reactive-transport codes, the final rate regime will control long-term glass dissolution. Predicting such final rate is a challenge that requires specific integrated mock-ups, in situ tests, simulation by reactive-transport codes, and validation by natural or archaeological analogues. Finally, in salt rock, lower rates than in pure water are expected, especially if the water is weakly renewed. Numerous studies have investigated effects of ionic strength and chemical composition of the brine.31–34 Because most of these studies are quite ancient, and also because the prediction of the water availability near the glass and the migration of soluble species in salt is more complicated than in other host rocks, their conclusions, namely in terms of performance assessment should be revisited as mechanistic knowledge has been improved. 5.18.4.3.7 Influence of glass fracturing

As glass is fractured after cooling, the reactive surface is greater than the geometrical one. Neither the surface of an actual glass block nor the evolution of the reactive surface in geological disposal condition has been precisely determined or estimated. Several experimental techniques have been employed to investigate glass blocks cracking networks and determine their impact on glass lifetime.35 Considering an inactive R7T7 glass block, the largest cracks are estimated to increase the geometric surface by a factor around 5 and smallest ones by a factor around 40. Up to now, all countries have calculated glass

472

Waste Glass

package lifetimes considering a constant cracking factor. Mechanistic modeling under development will help address this issue in the near future. However, most studies on natural or archaeological analogues have shown that inner cracks have generally a minor contribution to the overall glass alteration (cf. Section 5.18.4.3.9). 5.18.4.3.8 Modeling glass long-term behavior

Predicting long-term behavior of glass requires a multiscale approach as space and time scales related to key phenomena are too large to be simulated by a single mechanistic model. As a consequence, discrete modeling approaches have been developed from ab initio calculation at atomistic level36 up to performance assessment model at macroscopic level (also called operational models).37 In between, Monte Carlo model allows bridging the gap between atomistic level and measured dissolution kinetics.38,39 One key point is the glass dissolution rate law. Many mechanistic models based on rate equations have been proposed.40–42 The most advanced one is probably the ‘glass reactivity with allowance for the alteration layer, GRAAL,’ model proposed by Frugier et al.19 In this model, the glass-related parameters are the solubility limit of the PRI, the water, and solvated ions interdiffusion coefficient in this interphase, the PRI dissolution rate. The other model parameters are relative to secondary phases likely to precipitate, depending on the chemical elements supplied by the glass or by the surrounding medium: phase solubility limits and precipitation kinetics. For R7T7

glass, a very good agreement is observed between simulation and experimental data for a very large set of experimental conditions.43 The operational model that is proposed to assess the R7T7 glass long-term behavior in the proposed setting for the French geological repository, is the socalled ‘r0!rf ’ model.44 In this model, the rate of glass alteration is supposed to keep its initial r0 value until all conditions are obtained to get the final rate (i.e., full saturation of the media, with all silica sorption sites saturated). Then the final rate rf is applied. For R7T7 glass, the model parameters (alteration rates r0 and rr, and glass surface area accessible toward water) were determined as a function of temperature, pH, and glass composition throughout the whole R7T7 composition range. Uncertainties on the parameters values were also determined. The model can be used to calculate the glass block lifetime depending on the time–temperature profile, the pH of the medium, the date of water ingress in contact with the glass, and the quantity of accessible silicon sorption sites on the metal canister alteration products. A typical calculated glass lifetime plot is given on Figure 15. Two assumptions concerning the quantity of unsaturated sorption sites during the initial rate phase are proposed and water ingress in contact with the glass after 4000 years45 is assumed. Should the environment be saturated, the total package lifetime of a R7T7 glass will exceed 300 000 years, even in a pessimistic scenario where a large amount of iron corrosion product acts as a silica sink.

60

100.0%

50 10.0%

Numerical uncertainties (1s)

40

30 1.0%

20

Temperature (⬚C)

Total fraction of altered glass (%)

Mass of corrosion products = 2730 kg Mass of corrosion products = 100 kg

Two environmental conditions

10

0.1% 1000

10 000

100 000 Time (years)

1000 000

0 10 000 000

Figure 15 Operational modeling; example of calculation of R7T7 glass lifetime in two different environmental conditions.

Waste Glass

5.18.4.3.9 Natural an archaeological analogues

Validating predictive models is one of the major difficulties of investigating the long-term behavior of containment materials because the relevant time scales largely exceed what is accessible to laboratory experimentation. Whenever possible, therefore, natural or archaeological analogues are examined for this purpose. They enable to check that no long-term mechanism is forgotten. They give us some very long-term integrated experiments, against which predictive models can be qualitatively validated. For instance, archaeological glass blocks from a shipwreck discovered near the French Mediterranean island of Les Embiez have been examined because of their morphological analogy with nuclear glasses and their known, stable environment. Like nuclear glasses, these blocks were fractured after production; they were then leached for 1800 years in seawater.46 In that specific case, a quantitative agreement has been achieved between geochemical simulation and measurements on the archaeological artifact showing that bridging the gap between short-term laboratory data and long-term natural system is possible via a rigorous methodology47 (Figure 16). The same methodology could be applied to much older basaltic glasses for which the environment can be characterized. These glasses not only exhibit the same alteration mechanisms and kinetics as nuclear glasses in laboratory experiments, but their alteration

473

products also reveal strong similarities, especially between the palagonite on basaltic glasses and the gel on nuclear containment glasses, which can constitute a diffusion barrier. These studies can contribute to a finer definition of the chemical model of nuclear glasses and to the long-term validation of the gel protective properties.48 5.18.4.4 Conclusions on Glass Long-Term Behavior For more than 30 years, a very significant research effort on nuclear glass alteration mechanisms has been carried out worldwide, and a large database has been produced. Academic researches on longterm crystallization, radiation damage, and alteration by water, enable nowadays a good mechanistic understanding of the key phenomena that can alter nuclear waste glass properties in the long term. A sound methodology was established to use the best of academic knowledge of alteration mechanisms for performance assessment of glass package in complex environments. This methodology includes the following:  Assessing the evolution of the boundary conditions including normal and incidental scenario of evolution.  Understanding elementary alteration processes at a mechanistic level.

100 Embiez glass 80

Total

% of altered glass

Sext. (r0) Sint. (D) 60

Total (measured) Sext. (measured)

40

Sint. (measured)

20

0 102

103

104

105

Time (years) Figure 16 Predicted percentage of alteration for Embiez glass (curves) and measured alteration of both kinds of surfaces (stars). Geochemical modeling has been achieved using the Hytec code including the glass reactivity with allowance for the alteration layer model for glass dissolution, water diffusion within smallest cracks, advection within the largest ones, and specific boundary conditions related to 56-m deep seawater.

474

Waste Glass

 Assessing the couplings between the different mechanisms which simultaneously occur within a given scenario. Such couplings may indeed modify significantly the global evolution. Finally, the models describing the different processes have to be integrated in a global predictive model which often requires to be simplified by selecting the most significant processes and parameters. Operational models also include a conservative approach to overcome the lack of knowledge and wrap the general trend. The use of this kind of operational model demonstrates that waste glass lifetime can be over millions of years if the glass composition is optimized and disposal conditions are appropriate. Furthermore, through this long-term research on waste glass a new ‘science of long-term behavior’ has been developed. This science and methodology is now applied to numerous other matrixes (cement, bitumen, spent fuel, etc.).

5.18.5 Vitrification Processes As nuclear energy is very concentrated, the overall volume of nuclear waste is small. This is especially true for HLW that will be concentrated into a small volume of glass. (For example, the amount of HLW glass produced each year in France, related to the reprocessing of the spent fuel of about 50 reactors, is in the range of 100 m3.) Consequently, the scale of radioactive waste vitrification facilities is usually much smaller than that found in traditional glassmaking. In addition, and especially when processing HLW, the very high levels of radiation preclude direct contact with the equipment. Any waste resulting from exchange of failed equipment, for instance, becomes radioactive waste and must be managed as such. Consequently, HLW vitrification facilities must be designed to be remotely operated, and to minimize maintenance as well as secondary waste generation. Off-gas treatment systems must be very efficient to remove any volatilized or entrained radionuclide. As most of waste streams are nitrate-rich, NOx fumes are produced and must be abated. The whole vitrification process must be contained efficiently in order to prevent release of radionuclides to the environment. Another significant difference between traditional glassmaking and waste vitrification is that, most often, the waste is in a liquid form while, in glassmaking, the batch materials are dry solids. For waste vitrification

it is then necessary to evaporate the liquid and calcine the salts prior to reacting them with the glassformers. This operation requires large amounts of additional energy provided directly in the melter or in a specific pretreatment step. In the end, the glass product must be disposed of, usually in metallic canisters. For that purpose, most of the time, the glass product must be poured into these canisters. This requires, first, that glass viscosity be around 100 P or lower at the time of pouring and, second, that the vitrification equipment be designed with a pouring function. According to the nature of the waste to be vitrified, and the context, a number of processes have been studied, among which several have been deployed industrially. The first attempts were extrapolations of the crucible work performed in the laboratories. The process was performed batch wise in a single crucible, where all the operations of evaporation, calcination, vitrification, and evacuation of the product were performed successively. The melting crucible could be the canister itself (the process was then a ‘lost-crucible’ process) or a melter from which the glass product was poured into the canister. The first French industrial facility, PIVER (Figure 17), for instance, was of this type. The metallic melter was heated from the outside by a stack of inductors. The facility was used to process actual high-level radioactive waste into 100 kg glass blocks. Similar facilities, operated with lost crucibles or not, were designed or built in various other countries (UK, Italy, etc.). Very soon, however, it was concluded that batch processes did not allow throughputs compatible with commercial operation. The PIVER throughput, for instance, was around 5 kg h1 of glass. Most countries, then, decided to abandon batch processes and design continuous vitrification processes, with two major options for feeding the waste:  One-step processes, where liquid waste is fed directly to the melter and all the steps of evaporation, calcinations, and vitrification are performed in it. This is the case for instance of the Defense Waste Processing Facility at Savannah River, USA.  Two-step processes where liquid waste is first fed to a calciner before entering the melter. This is the case for instance in France, at the AVM facility at Marcoule or at the R7 and T7 facilities at La Hague. In the following sections we will describe the major existing facilities and the emerging new processes

Waste Glass

HLW solutions

475

Glass frit

Inductors

Feeding evaporation

Calcination

Melting refining

Pouring

Figure 17 The PIVER process in France (1969–1980).

being designed to further improve the capabilities and efficiency of these processes. 5.18.5.1 Existing Processes for Radioactive Waste Vitrification

FP solution Recycling

Additive

Calciner

Gas

5.18.5.1.1 The French two-step continuous vitrification process

Following the PIVER experience, the French CEA started to develop a two-step process, in order to separate the functions of evaporation-calcination and vitrification, as illustrated in Figure 18. This allows keeping a melter of relatively small size, as most of the energy is provided at the level of the calciner. Another major decision was to select a vitrification method by which power is supplied to the glass from the outside, without direct contact of the glass with the power source. A metallic melter heated by induction provided by an external stack of inductors was selected, following the good results obtained with PIVER. This disposition allows protecting the power source from contamination. On the other hand, the size of the melter is limited by the ability to transmit heat to the core of the molten bath. The first industrial facility for vitrifying HLW in France was the AVM which was the first industrial vitrification facility in the world, commissioned in 1978. This facility has vitrified the HLW solutions from the UP1 reprocessing plant and is now used to vitrify the effluents resulting from the decommissioning and decontamination of the same UP1 plant. This mission is nearing completion, and the AVM facility is now facing decommissioning after more than 30 years of successful operation. The experience

Glass frit

Scrubber

Melter

Canister

Figure 18 The French two-step continuous vitrification process.

gained from the operation of AVM has been later incorporated into the design of the larger facilities R7 and T7 at La Hague, with three vitrification lines each, which started operation in 1989 and 1992 respectively. The same technology has been selected for the WVP (Waste Vitrification Plant) at Sellafield in the United Kingdom. In the French continuous process used at La Hague, the concentrated HLW solutions are received and stored in cooled and stirred tanks. After sampling and analysis, they are fed at a metered rate to

476

Waste Glass

the calciner. In the calciner, they are heated progressively up to about 400
C to evaporate the liquid and transform them into a finely divided powder called the calcine. The calcine falls into the melter, together with glass-formers which are fed under the form of a prefabricated glass called frit. The mixture is heated at the surface of the molten glass bath and undergoes the final vitrification reactions (from temperatures of about 700
C) and finally become digested into the homogeneous molten glass at about 1100
C. The molten glass is then poured batchwise into metallic canisters which are then weld-sealed and evacuated. The calciner is a tilted rotating tube inserted into a furnace heated by resistors. In the calciner, the solution is evaporated and most of the nitrate salts (with the exception of alkalis) are converted to oxides by decomposition of the nitrates. A calcination additive, which decomposes under the action of temperature and reacts with the nitrates, is added to ease the fragmentation of the calcined mass and to limit the volatility of some radionuclides. The calcined mass then falls into the melter together with the glass frit. The melter is a metallic crucible made of nickel-base alloy heated by induction. In order to promote heat transfer and enhance melt homogeneity, the melter is equipped with stirring and gas sparging devices. The melter fills progressively during continuous feeding. When the higher operating level is reached in the melter, a batch of 200 kg of the molten glass is poured into a stainless steel canister through a pouring nozzle situated below the melter. Pouring is activated by heating the nozzle with a specific inductor. The melter then continues to process the next batch. Each glass canister holds two batches of 200 kg of glass. After filling and cooling, the glass canisters are closed tightly by welding a cover on top of their mouth. The sealed canisters are decontaminated by shot-blasting and checked for absence of residual contamination. They are then transferred to a storage facility (Figure 19) where they are stacked in pits cooled by a forced flow of air to evacuate the residual heat produced by radioactive decay of the FPs. At the time of production, the heating power of each individual canister can be higher than 2 kW. After several years of cooling in a forced ventilation storage facility, the residual power decreases sufficiently to allow transferring the canisters to a facility cooled by natural convection. The off-gas from the calciner and the melter is composed of water vapor, nitrogen compounds, and entrained material. It is extracted at the top of the

Figure 19 The French La Hague high-level waste glass storage facility. # AREVA.

calciner and goes through a dust-scrubber, to remove most of the large particulate material and aerosols for recycling to the calciner, a condenser, washing columns, and filters to decontaminate the gas prior to release to the stack in compliance with radioactive and chemical release standards. The liquid effluent from off-gas treatment is collected and treated in specific effluent treatment facilities to concentrate the activity and recycle most of it to vitrification. This technology is now used in the French industrial facilities of R7 and T7 at La Hague and has proven its efficiency and operability. By the end of 2009 more than 18 300 glass canisters have been produced (of which about 3400 have been produced at the AVM facility and about 14 900 at La Hague). This amounts to 7240t of glass and more than 2.5108 TBq of activity safely immobilized. The small size and modular design of the technology make it easily operable and maintainable. The following are the major limitations of this process:  The life expectancy of the metallic crucible which is in direct contact with hot (1100
C) and corrosive molten glass. Through continuous developments, it is now possible to replace the metallic melters about once a year in the current La Hague facility. On the other hand, this operation is made easy by the small size and design of the facility; melter exchange is performed within a week, and the resulting waste can be size-reduced and processed together with the other metallic waste of the reprocessing facilities.  Capacity : As the glass is heated from the outside, the size of the melter, and thus the capacity, is limited. In the current facilities, the capacity of a metallic melter is about 25 kg h1 of glass, and several lines are needed if a higher throughput

Waste Glass

is required. On the other hand, this small size is also an advantage for maintenance and waste generation, as seen above.  Limitation in melt temperature: In order to preserve the integrity of the metallic crucible, operation temperature is limited to about 1100–1150
C. This, in turn, is a limit for throughput (as throughput theoretically varies like T4). Moreover, this temperature limit also reduces the range of glass compositions that can be processed in such a facility to those which have melting temperature below or around 1150
C. 5.18.5.1.2 Liquid-fed ceramic melters

Not all the liquid waste is amenable to separate calcination: when the waste holds large amounts of alkalis, the corresponding nitrate salts tend to form molten phases in the calciner and prevent adequate calcination, generating numerous sticking and caking problems. In such situations, direct liquid feeding of the melter has been implemented. This is the case in the United States (Defense Waste Processing Facility, DWPF; West-Valley Demonstration Plant, WVDP; Hanford WTP), where the acidic high-level liquid waste has been neutralized by caustic prior to storage. Direct liquid feeding can also be selected with the intent of keeping only one processing step such as in Belgium (PAMELA), Germany (VEK), Russia, or Japan (Tokai Vitrification Plant (TVF) and K-Plant in Rokkasho Mura). In these situations, the liquid waste and glass frit (or separate glass-formers) are fed continuously to the top of the melter, and evaporation, calcination, and vitrification reactions are performed in the ‘coldcap,’ a colder layer that sits on top of the molten bath and progressively dissolves into the melt. In such a configuration, it is necessary to supply all the heat to perform those transformations through the surface of the glass bath and the cold-cap. The power requirements are such that, in order to obtain an adequate throughput, the surface area of the melt must be extended (throughput theoretically varies proportionally to the melt surface area). Even if boosting can be provided by implementing radiative heaters in the melter plenum above the cold-cap, this results in much larger melters, which cannot be heated from the outside any more. A technology directly inspired from traditional glassmaking melters has been selected. The melters are lined with layers of refractory bricks in order to protect the cooled metallic walls, and heating is performed by directly applying current to the conductive melt

477

through metallic electrodes, usually made of Inconel 690 and sometimes cooled by an internal circulation of air. The current heats the melt by Joule effect. The melt, in its turn, transfers the heat to the cold-cap. As, owing to the larger surface area, these melters hold large volumes of glass, pouring is usually performed by overflow, air-lift, or vacuum siphon and can be continuous. Batch bottom pouring is nevertheless implemented for some specific applications described below. Several facilities have been operated or are still in operation worldwide. The first industrial facility to have been operated with a ceramic melter has been the PAMELA facility in Belgium, commissioned in 1985, which has processed 490 MT (metric tonne) of glass in about 6 years. The melter had a flat bottom and pouring was performed essentially by overflow. Two melters were used, with a melter life of around 3 years. The melter was 2.6  2  2 m in size, weighed about 20 MT and held about 300 l (750 kg) of glass. This facility experienced difficulties owing to the settling of glassinsoluble noble metals from the waste at the bottom. This conductive settled layer tended to disturb current distribution in the bulk of the melt and led to loss of capacity and, ultimately melter failure. In order to prevent such occurrences when dealing with noble-metal-rich feeds, the bottom of the cavity can be designed with a ‘dead zone’ below the level of the electrodes, to collect the sludge in a manner that should prevent any interactions with power distribution to the melt. This solution was implemented at WVDP in the United States for the vitrification of a backlog of HLW solutions from a pilot commercial reprocessing plant. This melter started operation in 1996 and produced 275 canisters of glass before being stopped and emptied in 2002. The melter, with a design throughput of about 45 kg h1 of glass, was of large dimensions (33.23.3 m), with a glass hold-up of about 860 l (1150 kg) and a melt surface area of 2.2 m2. The melter was equipped with two pouring chambers and a bottom electrode to promote evacuation of the noble metals. Despite these dispositions, at the end of its life, the melter showed signs of declining capacity and power distribution upsets, probably attributable to the slow accumulation of conductive noble metals at the bottom. Another solution to deal with this issue is to provide sloped walls above a bottom pouring device, to promote the evacuation of the settled noble metals. This solution has been implemented in the Tokai Vitrification Facility (TVF) (Figure 20) and

478

Waste Glass

HLLW feed line Glass feed line

Off-gas line

Start-up heater

Main electrode Auxiliary electrode Discharge nozzle Figure 20 The Tokai Vitrification Facility ceramic melter (Japan). Reproduced from Aoshima, A.; Kozaka, T.; Tanaka, K. Glass melter replacement and melter technology development in the Tokai vitrification facility. In Proceedings of ICONE12 International Conference on Nuclear Engineering, 2004; with permission from ASME.

K-plant in Japan, (but did not completely suppress the issue) and in VEK in Germany, which has been recently commissioned. TVF has been commissioned in 1995 and the first melter processed 130 canisters before its replacement in 2002. Melter #2, with an improved bottom configuration, is now implemented and operating. The melter section is 0.80.83 m, with a glass hold-up of 350 l, and a design throughput of 9 kg h1. The sloped walls at the bottom make an angle of 45
with respect to the vertical direction. Pouring is activated by heating the pouring nozzle with an inductor. Pouring is stopped by blowing cold air to freeze the glass inside the nozzle. The electrodes are cooled with air. In Rokkasho (Japan), a much larger plant, with a design throughput of 80 kg h1 and a similar conception, is in the process of active start-up. VEK in Germany has been designed to process a backlog of HLW produced by a pilot reprocessing plant. The melter has a conical bottom, with slopes of around 60
with respect to the vertical direction. The melter is designed to process 10 l h1 of feed, or produce 7 kg h1 of glass. It is a cylindrical melter with an outside diameter of 1.5 m and a height of

1.7 m, with a glass hold-up of 150 l (375 kg), a melt surface area of 0.44 m2, and an overall weight of 8 MT. The electrodes are cooled with air. Very large capacity ceramic melters have also been commissioned or are in the process of commissioning for Defence High Level Waste in the United States. These wastes are not as radioactive as the waste from commercial reprocessing, and they do not contain significant amounts of noble metals. For these applications, throughput is the major concern, as the volumes of waste to be vitrified are quite impressive. The first vitrification facility commissioned in the United States was the DWPF, at Savannah River, in the United States. The liquid-fed ceramic melter has been designed to process a thick slurry retrieved from the Savannah River Site (SRS) tank farms at a design rate of up to 100 kg h1 of glass (or 200 l h1 of feed). The frit is introduced as a powder and mixed with the feed suspension. This facility has been commissioned in 1996 and has produced more than 5300 MT of glass. The melter holds a melt volume of 2500 l (6.5 MT), with a melt surface area of 2.6 m2. Pouring is continuous except at the time of canister changeout, and performed via a siphon. The overall weight of the melter (including the glass and cooling water) is around 80 MT. Melter #1 was decommissioned in October 2002 and Melter #2 started operation 5 months later, in March 2003 (Figure 21). After completing its useful life, Melter #1, together with its supporting rack, was inserted into a box, evacuated on a trailer, and entombed for longterm storage in a specific underground cavity on site. For Hanford WTP, two vitrification facilities are being built. For HLW, the facility will host two ceramic melters with surface areas of 3.75 m2 for a design throughput of 125 kg h1 per melter. For low activity waste, it is intended to implement two elongated melters with melt surface areas or 10 m2, for a throughput of 625 kg h1 per melter. For these melters, capacity is critical. In order to improve capacity, and enhance heat transfer from the melt to the coldcap, extensive air sparging is being implemented. Liquid-fed ceramic melters are also used in Russia to process HLW into a phosphate-based matrix. The selection of the phosphate-based matrix allowed reducing melting temperature to below 1000
C, but this matrix is still very corrosive and detrimental to the melter lifetime. The interest in liquid feeding is because one single piece of equipment is needed to perform all the necessary reactions. Ceramic melters are also quite

Waste Glass

Feed tubes

479

Lifting bail

T.V. camera/borescope assembly

Off-gas film cooler

Riser heater

Dome heaters

Vapor thermowell/ conductivity probe

Pour spout heater

Melt thermowell Electrodes Drain valve

Figure 21 The Defense Waste Processing Facility melter (USA). Reproduced from Norton, M. R.; Shah, H. B.; Stone, M. E.; Johnson, L. E.; O’Driscoll, R. Overview – defense waste processing facility operating experience. In WM’02 Conference, Tucson, AZ, Feb 24–28, 2002; Westinghouse Savannah River Company: Aiken, SC, 2002; Copyright WM Symposia, Inc., 2002, reprinted by permission.

stable in operation. The major limitations of these melters are the following:  They are large pieces of equipment which, once used, become a large quantity of waste. Melter exchange can require several months.  With some exceptions, once started, they must be maintained hot until their decommissioning, in order to avoid deterioration of the ceramics.  Their tolerance to glass-insoluble elements, such as noble metals, or crystal forming elements, is limited. As the molten glass is in direct contact with the refractories and electrodes, operating temperature is

limited, usually around 1150
C, sometimes a little higher, in order to preserve the integrity of the refractories and electrode material. This is a limiting factor for capacity and for the type of glass that can be processed, as for the hot metallic melters described above. 5.18.5.2 Emerging Processes for Radioactive Waste Vitrification 5.18.5.2.1 Cold-crucible induction melters

As seen above, the two major current processes for radioactive waste vitrification, although they have proven their sturdiness and adaptation to very

480

Waste Glass

demanding environments, have reached their limits, in terms of throughput, tolerance to some elements in the glass, or glass composition. Other solutions have then been sought in the last two decades, with the emergence of the cold-crucible induction melter (CCIM) technology both in Russia and France. For this technology, the crucible itself is composed of a water-cooled metallic structure. The heating mode is direct induction in the molten bath, a technique that allows transmitting power directly at the heart of the glass, using a water-cooled inductor (Figure 22). In order to allow the penetration of the magnetic field provided by the inductor past the metallic crucible (to avoid the Faraday cage effect), the crucible is made of several sectors separated by a thin layer of isolating material. The power (typically around 600 kW) is directly induced in the glass by the high-frequency magnetic field (typically 300 kHz in France). As cold glass is not conductive, melting is started by introducing a conductive material (for instance graphite) on the cold glass in order to generate the first molten zone. This molten zone then progressively extends to fill the whole volume of the melter. A thin layer of molten material freezes at the zone of contact with the cold metallic wall. This frozen layer, several mm thick, forms a ‘cold crucible’ which prevents further contact between the molten mass and the metallic wall. This technology offers a number of advantages over the previous processes:  As the molten glass does not contact the melter wall, temperature is not limited any more, offering a whole new range of possibilities for glass or even ceramics formulations.  The cold layer also protects the equipment from corrosion, and thus allows processing effluents that

Cold cap

CCIM Cold glass layer Inductor

Molten glass

Figure 22 The cold-crucible induction melter (CCIM).

   

were deemed too corrosive for the previous techniques (such as effluents containing sulfur, chlorine, molten salts, etc.). The molten glass is not polluted by any material from the crucible. Significant increase in capacity associated with the possibility of raising the operating temperature. As the crucible is protected from both heat and corrosion, its lifetime is extended. The cold crucible layer can be easily removed at the end of operation, thus allowing easier decontamination of the crucible.

The use of CCIMs in industrial facilities has started. The cold-crucible technology is used in Russia to vitrify low activity effluents from nuclear power plants. In S. Korea, an incineration–vitrification facility based on the French CCIM technology has been recently commissioned to process dry active waste and resins from nuclear power plants. And in France, for high-level radioactive waste, a first CCIM has been retrofitted in one of the lines of the R7 facility to process corrosive decontamination effluents.49 Hot start-up is in progress. It is the first cold-crucible melter (CCM) implemented in a high activity cell. The retrofitting has been performed fully remotely. This deployment is the apex of more than 20 years of R&D work and will shift the status of this technology to that of a fully mature technology. 5.18.5.2.2 Incineration–vitrification processes

Alternative processes of incineration–vitrification coupling CCM and plasma torches are developed with the aim of drastically reducing intermediatelevel waste (ILW) volume while confining it in a more durable glassy matrix. The proposed technology relies on the great skill developed by the CEA’s vitrification teams on the CCM and also on the

Waste Glass

oxygen plasma transferred mode plasma torches that have been tested for more than 10 years in Marcoule. The Systeme Hybride d’Incineration Vitrification Avance´ (SHIVA) process (Figure 23) has been developed to study the feasibility of these technologies for the future.50 It allows performing in the same vessel the incineration of the burnable wastes, the vitrification of their mineral charge, and the combustion of off-gases. Significant advantages can be obtained by supplying the waste directly into oxygen arc plasma located above a glass bath heated by direct induction in a cold crucible. The temperature is very high and so is the efficiency of the combustion in the excited free oxygen rich atmosphere that also promotes a thorough oxidation of the glass. The treatments of different kind of wastes have been investigated. Burnable wastes such as ionic exchange resins, bituminous wastes, sulfate slurries, and graphitic sludge have been successfully incinerated with a good incorporation of their mineral charge in the glass.51 In addition, mineral wastes such as sludge issued from nuclear treatment have also been treated. The current studies concern the incineration–vitrification of chlorinated organic wastes. In this case, the main difficulty is to manage the volatile metallic chlorides in the process. Some studies are also in progress on the treatment of mixed wastes containing organics and metals.

Waste + Glass precursor Oxygen Oxygen

Cathode

Anode

Burned gases exhaust Metallic cooled walls

Plasma Molten glass

Inductor HF current Figure 23 The Systeme Hybride d’Incineration Vitrification Avance´ process principle.

481

5.18.6 Conclusions and Outlook on Waste Glasses Today vitrification is the world reference solution to the containment of HLW. Since the first trials of vitrification of FPs solutions in Saclay (France) in the 1950s, until the recent development of a new generation of CCM in the 2010s, vitrification is a success story that allowed having waste glasses that meet all industrial requirements while providing an excellent long-term behavior. Thus, in all countries where reprocessing of nuclear fuel is practised, vitrification plays a vital role in our ability to safely manage the HLW from nuclear energy. Given the importance of this matrix for HLW management, a huge number of research activities devoted to nuclear glasses have been supported over the last 30 years either by the nuclear industry or by governmental organizations. These studies have focused on both optimizing complex glass properties (solubility of components, structure, durability, etc.) understanding the containment properties and the long-term behavior, and on the continuous improvement of vitrification processes operating in hostile environments. Through these studies, general knowledge of complex glasses has considerably progressed, particularly in the area of understanding the mechanisms of glass alteration by water. To date, there are probably many more publications on the alteration of the R7T7 glass than on the alteration of window glass. Significant progress was achieved on radiation damage of glass too. If 30 years ago fear existed that the glasses could disintegrate under the effect of selfirradiation, today it is known that glass is a self repairing material whose macroscopic properties are not affected by long-term autoirradiation. Great progress has also been made in atomistic modeling of complex glasses that allows checking that our models are on the basis of an atomistic understanding of basic phenomena. This success of vitrification is expected to further increase in the next 30 years, for at least two reasons. First, in a context of global nuclear renaissance with the need to conserve resources, more and more countries will choose nuclear fuel recycling and vitrification will prevail for waste treatment. On the other hand, with the increasing desire to protect the environment, a large number of matrices used today to confine intermediate nuclear waste and even hazardous waste, will be probably replaced by glass, as incineration–vitrification allows both a drastic waste

482

Waste Glass

volume reduction and a final containment with improved performance. With the proliferation of waste types treated, we should expect a proliferation of glass compositions too. One should have to be very careful about the quality of glass products to never degrade the image of this excellent containment matrix. In fact, the glass is a wonderful material that can pass almost continuously from a soluble borax glass to almost eternal obsidian. Care should be taken that the great glass flexibility is not used at the expense of its quality. The development of good containment glass requires mastering glassy materials and process aspects, all together with the science of their long-term behavior.

References 1.

2. 3. 4. 5. 6. 7.

8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Bonniaud, R.; Cohen, Y.; Sombret, C. Essais d’incorporation des solutions concentre´es de produits de fission dans les verres et les micas. Deuxie`me Confe´rence Internationale des Nations Unies sur l’utilisation de l’Energie Atomique a` des fins pacifı`ques, Gene`ve, Sept 1, 1958. Mendel, J. L.; Ross, W. A.; Roberts, F. P. Annual Report on the Characteristics of High-Level Waste Glass BNWL-2252; Jun 1977. Zarzycki, J. Les verres et l’e´tat vitreux; Masson: Paris, 1982. Jo, P. Mate´riaux non cristallins et science du de´sordre; Presses polytechniques et universitaires Romandes, 2001. Re´aumur, M. Me´m. Acad. Sci. 1739, 370. MacDowell, J. F. Ind. Eng. Chem. 1966, 58(3), 38–45. Vernaz, E. Physico-chemical properties and log-term behaviour of French R7T7 nuclear waste glass. In Xth National Scientific Technical Conference with International Participation on Glass and Fine Ceramics, Varna, Bulgaria, Oct 18–20, 1990. Orlhac, X. E´tude de la stabilite´ thermique du verre nucle´aire. Mode´lisation de son e´volution a´ long terme. Ph.D. Thesis; Montpellier II University; 2000; CEA Report CEA-R-5895. CEA DEN Monographs ‘‘Nuclear Waste Conditioning’’. Etienne Vernaz, E´ditions du Moniteur: Paris, 2009; ISBN 978-2-281-11380-8. Orlhac, X.; Fillet, C.; Sempere, R.; Phalippou, J. J. Non Cryst. Solids 2001, 291, 1–13. Vernaz, E.; Jacquet Francillon, N.; Bonniaud, R. Science et Recherche (Echos du groupe CEA) 1982, 3, 38–43. Matzke, H. J.; Vernaz, E. J. Nucl. Mater. 1993, 201, 295–309. Peuget, S.; Cachia, J. N.; Je´gou, C.; et al. J. Nucl. Mater. 2006, 354, 1. Marples, J. A. C. Nucl. Instrum. Methods Phys. Res. B 1988, 32, 480. Weber, W. J.; Ewing, R. C.; Angell, C. A.; et al. J. Mater. Res. 1997, 12, 1946. Inagali, Y.; Furuya, H.; Idemitsu, K.; Banba, T.; Matsumoto, S.; Murakoa, S. In Materials Research Society Symposium Proceedings; 1992; Vol. 257, p 199. Delaye, J. M.; Ghaleb, D. Phys. Rev. B 2005, 71, 224204. Vernaz, E.; Gin, S.; Jegou, C.; Ribet, I. J. Nucl. Mater. 2001, 298(1–2), 27–36.

19.

Frugier, P.; Gin, S.; Minet, Y.; et al. J. Nucl. Mater. 2008, 380(1–3), 8–21. 20. Vernaz, E.; Dussossoy, J. L. Appl. Geochem. 1992, (Suppl. 1), 13–22. 21. Ribet, S.; Gin, S. J. Nucl. Mater. 2004, 324, 152. 22. Jollivet, P.; Angeli, F.; Cailleteau, C.; Devreux, F.; Frugier, P.; Gin, S. J. Non Cryst. Solids 2008, 354(45–46), 4952–4958. 23. Cailleteau, C.; Angeli, F.; Devreux, F.; et al. Nat. Mater. 2008, 7, 978–983. 24. Sales, B. C.; Boatner, L. A. Mater. Lett. 1984, 2(4, Pt 2), 301–304. 25. Frugier, P.; Martin, C.; Ribet, I.; Advocat, T.; Gin, S. J. Nucl. Mater. 2005, 346, 194–207. 26. Ribet, S.; Muller, I.; Pegg, I.; Gin, S.; Frugier, P. In Materials Research Society Symposium Proceedings; 2004; Vol. 824, pp 309–314. 27. Chave, T.; Frugier, P.; Gin, S.; Ayral, A. Submitted to Geochim. Cosmochim. Acta. 2010. 28. Gin, S.; Godon, N.; Mestre, J. P.; Vernaz, E.; Beaufort, D. Appl. Geochem. 1994, 9, 255–269. 29. Gin, S.; Frugier, P.; Godon, N.; Jollivet, P.; Verney-Carron, A. Submitted to the XIIIth Water Rock Interaction Conference; 2010; http://www.nextag.com/ Water-Rock-Interaction-XIII-1231825593/specs-html. 30. Pozo, C.; Bildstein, O.; Raynal, J.; Jullien, M.; Valke, E. J. Nucl. Mater. 2007, 35, 258–267. 31. Pederson, L. R.; et al. Mater. Res. Soc. Symp. Proc. 1984, 26, 417–426. 32. Grambow, B.; Muller, R. Mater. Res. Soc. Symp. Proc. 1990, 176, 229–240. 33. Lutze, W.; Muller, R.; Montserrat, W. Mater. Res. Soc. Symp. Proc. 1987, 112, 572–584. 34. Feng, X.; Pegg, I. Phys. Chem. Glasses 1994, 35, 98–103. 35. Minet, Y.; Godon, N. In Environmental Issues and Waste Management Technologies in the Ceramic and Industries VIII; Book Series: Ceramic Transactions; 2003; Vol. 143, pp 275–282. 36. Geneste, G.; Bouyer, F.; Gin, S. J. Non Cryst. Solids 2006, 352, 3147–3152. 37. Ribet, I.; Betremieux, S.; Gin, S.; Angeli, F.; Je´gou, C. In Global Conference Proceedings, Paris, France, Sept 6–11, 2009. 38. Devreux, F.; Barboux, P.; Filoche, M.; Sapoval, B. J. Mater. Sci. 2001, 36, 1331–1341. 39. Ledieu, A.; Devreux, F.; Barboux, P.; Minet, Y. Nucl. Sci. Eng. 2006, 153, 285–300. 40. Grambow, B.; Muller, R. J. Nucl. Mater. 2001, 298, 112–124. 41. Bourcier, W. L.; Pfeiffer, D. W.; Knauss, K. G.; McKeegan, K. D.; Smith, D. K. Mater. Res. Soc. Symp. Proc. 1990, 176, 209–216. 42. Oelkers, E. H. Geochim. Cosmochim. Acta 2001, 65, 3703–3719. 43. Frugier, P.; Chave, T.; Gin, S.; Lartigue, J. E. J. Nucl. Mater. 2009, 392, 552–567. 44. Ribet, I.; Betremieux, S.; Gin, S.; Jegou, C. Proceedings of Global 2009, Paris, France, Sept 6–11, 2009. 45. Vernaz, E.; Poinssot, C. Overview of the CEA French research program on nuclear waste. In Scientific Basis for Nuclear Waste Management XXXI, MRS Symposium Proceeding; 2008, Vol. 1107, pp 21–32. 46. Verney-Carron, A.; Gin, S.; Libourel, G. Geochim. Cosmochim. Acta 2008, 72, 5372–5385. 47. Verney-Carron, A.; Gin, S. Geochim. Cosmochim. Acta 2010, 74, 2291–2315. 48. Techer, I.; Lancelot, J.; Clauer, N.; Liotard, J. M.; Advocat, T. Geochim. Cosmochim. Acta 2001, 65(7), 1071–1086.

Waste Glass 49. Naline, S.; Girold, C.; Gouyaud, F.; Robineau, V.; Carpentier, B. Vitrification 2010: A challenging French vitrification project to retrofit a cold crucible inductive melter at the la Hague plant. Waste Management Symposium (WM – 2010), Phoenix, AZ, Mar 7, 2010–Mar 11, 2010. 50. Lemont, F.; Charvin, P.; Russello, A.; Poizot, K. An innovative hybrid process involving plasma in a cold crucible melter devoted to the future intermediate level

51.

483

waste treatment: The SHIVA technology. In 12th International Conference on Modern Materials and Technologies (CIMTEC – 2010), Montecatini Terme, Italy, Jun 6, 2010–Jun 11, 2010. Girold, C.; Lemort, F.; Pinet, O. The vitrification as pathway for longlife organic waste treatment. Waste Management Conference, Tucson, AZ, Feb 26, 2006–Feb 28, 2006; Etats-unis (Waste Management Symposium 2006 Proceedings).

5.19

Ceramic Waste Forms

E. R. Vance Institute of Materials Engineering, Australian Nuclear and Technology Organisation, Menai, NSW, Australia

Crown Copyright ß 2012 Published by Elsevier Ltd. All rights reserved.

5.19.1

Introduction

485

5.19.2 5.19.3 5.19.4 5.19.5 5.19.6 5.19.7 5.19.8 5.19.9 5.19.10 5.19.10.1 5.19.10.2 5.19.10.3 5.19.11 5.19.12 References

Desirable Performance Characteristics of High-Level Nuclear Waste Forms Design of Waste Form Ceramics Historical Evolution of Candidate Ceramic Waste Forms for HLW Titanate Ceramics Glass–Ceramics Aqueous Dissolution Radiation Damage Thermodynamic Stability of Multication Oxides Processing of Ceramics and Glass–Ceramics Hot Uniaxial Pressing Hot Isostatic Pressing Melting Cements and Geopolymers Conclusions

488 488 490 492 494 495 495 496 496 497 497 498 499 500 501

Abbreviations An ANSTO

Actinide Australian Nuclear Science and Technology Organisation AWE Atomic Weapons Establishment BET Brunauer–Emmett–Teller DWPF Defense Waste Processing Facility EA Environmental Assessment FP Fission product FUETAP Formed under extreme temperatures and pressures HIP Hot isostatic press HLW High-level waste HUP Hot Uniaxial Press ICPP Idaho Chemical Processing Plant ILW Intermediate-level waste LLNL Lawrence Livermore National Laboratory LLW Low-level waste MCC Materials Characterization Center MOX Mixed oxide MPP Magnesium potassium phosphate NIMBY Not in my Backyard NZP Sodium zirconium phosphate OPC ordinary Portland cement PCT Product consistency test PNNL Pacific Northwest National Laboratory RE Rare earth

SI SRL Synroc USDOE

International System of Units Savannah River National Laboratory Synthetic rock United States Department of Energy

5.19.1 Introduction Worldwide inventories of high-level waste (HLW) from nuclear power and weapons production, other than spent fuel, constitute hundreds of thousands and perhaps millions of tonnes.1 Figures 1 and 2 show the abundances and the time dependences of the radioactivity of HLW from nuclear fuel reprocessing, respectively. Disposal of HLW other than spent fuel itself basically involves adding specific material to the waste and processing it to transform it to a dense refractory solid – the waste form – that can withstand prolonged immersion in groundwater. In a large majority of cases, the solid would be an oxide, but metallic components may also be present, depending on the waste and the processing conditions. The solid, which would be a glass, ceramic, glass–ceramic, or cementitious material, depending on the nature of the waste to be dealt with, would then be transferred in a metal disposal canister to a geological repository (Figure 3), in which the radioactive and toxic 485

486

Ceramic Waste Forms

10

14 Mev 103

1 Fission product total activity

Fission yield (%)

102 0.1

10 Actinides total activity

1 0.01 10−1

Thermal 0.001

0.0001 70

10−2

10−3 90

110 130 Mass number

150

Figure 1 Relative atomic abundances of fission products elements in a commercial power plant nuclear fuel.

components would be immobilized for long periods, up to millions of years if necessary, depending on the half-lives of the dangerous species. Glass is currently seen as a baseline option for HLW. In this article, the focus of the immobilization is on the waste form, but of course the repository would also be expected to play a significant role in preventing the transport of radionuclides and toxic ions to the biosphere. Because it seems unlikely that spent fuel in the form of irradiated UO2, a ceramic, will be chemically processed, but rather simply allowed to cool as the short-lived fission products decay, and then containerized for eventual repository disposal, we will not discuss spent fuel in this article, but it will be the subject of Chapter 5.16, Spent Fuel as Waste Material. Tables 1 and 2 show approximate compositions of Purex-type reprocessing and tank wastes in the United States. Since the activity of the waste falls with increasing time, it is technically advantageous to store the waste as long as possible, although this should not be seen as an excuse to delay immobilization unduly. Moreover, the method of storage is critical and needs frequent attention. For instance, a strong initial driver of HLW cleanup in the United States in the early 1990s was that at the Hanford reservation in the state of Washington, the stainless steel tanks containing the old military wastes from nuclear weapons production were leaking into the

10−4

10−5

10−6

1

10

102 103 104 105 106 107 Time (year)

Figure 2 Time dependence of radioactivity of reprocessing waste from commercial power plant nuclear fuel.

Earth’s surface

Rock Backfill, plus cement

1–3 km

Canisterized HLW Figure 3 Schematic diagram of a deep geological repository.

surrounding environment. Also, the tanks in which the water had largely evaporated over the years of storage, because of radiogenic heating, gave periodic gas evolution in the form of large hydrogen bubbles,

Ceramic Waste Forms

which had safety implications via potential radionuclide removal from the tank into the atmosphere and ground area adjacent to the tanks, as well as ignition and explosion. Historically, there has been explicit and implicit competition between ceramics and borosilicate glasses for immobilization of HLW (e.g., the ‘Atlanta shootout’) for Savannah River Laboratory (SRL) waste in 19812,3 and the US Department of Energy (USDOE) decision on impure US/Russian surplus Pu in 1998.4,5 However, as there are now so many types of HLW or intermediate-level wastes (ILWs) having a large variety of chemical compositions remaining to be immobilized around the world, it is becoming clear that glass, glass–ceramics, ceramics, and on occasion, cementitious materials have their place for different HLWs and ILWs. In this article, the focus is on ceramics and glass– ceramics targeted mainly to immobilize HLW.

Table 1 Approximate compositions (wt%) and halflivesa of main fission product (FP) and actinide oxides in Purex fuel reprocessing HLW that has been stored for >10 years FP oxideb

wt%

Half-life (year)

Cs2O SrO BaO RE2O3 ZrO2 MoO3 TcO2 AnO2 RuO2 PdO Rh2O3

6 3 4 15 15 15 6 6 10 6 2

30 30 – 100a – – 210 000 >10 000 – – –

Water excluded; RE: rare earth; An: actinide. a Group half-lives are very approximate as they range from short to long times for different components. Absence of half-life value ¼ stable elements. b Contains additional stainless steel corrosion products.

487

The specific activity of reprocessing HLW is in the order of tens of terabecquerels per liter, while the US tank wastes from Pu production for atomic weapons production have specific activities that are perhaps 1000 times lower. Indeed, in other countries, some of these latter wastes would be considered as ILW. We note that hot isostatic pressing (HIPing) in waste form production has been validated recently at INL, ID, USA (Idaho National Laboratory) by US regulators, and a de facto validation of crystalline waste forms may be taken from the 1998 decision,4 albeit overturned in 2001 in the sense of not proceeding with the immobilization option, to use ceramics for surplus Pu disposition in the United States. Ceramic materials for the incorporation and the immobilization of nuclear waste range from refractory, dense, fine-grained ceramic, or glass–ceramic candidates for HLW to cementitious product options for low-level waste (LLW). Here, we discuss the design of mainly ceramics and glass–ceramics for HLW, but we also discuss cementitious materials for ILW. How such materials meet the regulatory criteria for properties such as retention of radionuclides and toxic ions when the materials are exposed to water, fire resistance, and mechanical strength, together with other properties such as waste loading, and simplicity or otherwise of processing methods, will be outlined. The influence of self-damage due to decay of the radioactive species is also an important feature and is briefly mentioned, noting that detailed discussions of this phenomenon appear in Chapter 5.22, Minerals and Natural Analogues. The final destination of HLW is generally agreed to be a deep (0.5–3 km) geological repository (Figure 3), so the principal object of fabricating a waste form for HLW is, using the simplest and cheapest possible processing methods, to incorporate the waste in a solid that is minimally porous and has high leaching resistance when exposed to water, adequate strength, and a high fraction (>20 wt%) of waste per

Table 2 Approximate oxide compositions and half-lives of An, FP, and process chemical oxides in US tank wastes and surplus Pu-bearing waste (ignoring minor cations and assuming anions such as nitrates, nitrites, hydroxides, and carbonates are removed upon calcination) US tank waste (wt%)

Half-life (year)

Surplus Pu-bearing waste (wt%)

Half-life (year)

Na2O (40) K2O (5) Al2O3 (50) Fe2O3 (4) FP (<1) AnO2 (<1)

– – – – various >10 000

PuO2 (20) UO2 (20) Fe2O3 (20) Al2O3 (20) CaF2 (20)

24 000 4  109 – – –

An: actinide.

488

Ceramic Waste Forms

unit volume. A limiting feature for HLW is the radiogenic heat output (see, e.g., Sizgek6) as it is desirable to keep the repository temperature to 100  C or less to minimize the reactivity of the waste form with groundwater. After presenting some desirable characteristics of waste forms and features impacting on the design of multiphase ceramics for wastes containing many radionuclides and process chemicals, we outline the historical development of ceramic waste form research. Alternative means of preparing appropriate candidate waste form ceramics is put forward. Then, we conclude that the best way to treat a given nuclear waste depends on the nature of the waste itself and that there is no single optimum method of treating all nuclear wastes. We further note that disagreement among waste form proponents as to the best way to dispose of a given waste does not suggest that the ‘nuclear waste problem’ is not basically solved. Given this, it would follow that nuclear power is sustainable from that point of view. In the design of an appropriate ceramic, it is advantageous if the component phases are similar to those of naturally occurring multication oxide minerals such as oxyapatite (Ca2La8(SiO4)6O2) or zirconolite (CaZrTi2O7) that are known to survive in hot, wet conditions for millions of years. This feature and the relative thermodynamic stability of crystalline material versus amorphous glass were the principal stimuli in the 1970s for the candidacy of ceramic waste forms for HLW.

5.19.2 Desirable Performance Characteristics of High-Level Nuclear Waste Forms Again, it is imperative that waste forms have very high chemical durabilities in terms of resistance to leaching by groundwater. The durability of the waste form can be subjected to laboratory study and then optimized. In this context, normalized leach rates of <1 g m2 day1 at temperatures below 100  C are considered as baseline. One gram per square meter per day (1.16  105 g m2 s1) corresponds to around 0.2–0.5 mm day1 in terms of thickness. (While g m2 day1 is not an SI unit, it is almost universally used within the nuclear waste community.) The normalized leach rate is defined as the gross release rate of the ion in question divided by the concentration of the ion, and the aim of using normalization was to defy the ‘dilute and disperse’

strategy in which sufficient dilution could reduce gross leach rates to apparently acceptable values. The latter approach is objectionable because it creates very large volumes of (weakly) radioactive waste, which take up space and inflate cleanup costs. The effect of self-irradiation needs to be taken into account in these determinations. The higher the proportion of waste that can be incorporated per unit volume of the waste form, the less repository space will be needed and so the costs will be minimized. The waste form needs to be easily and reliably processed in a remote environment, and minimization of secondary wastes such as radioactive off-gases (needing recycle) during the production of the waste form is important. Given the need for aqueous durability, it is obvious that open porosity would be a very bad feature because entry of water into the interior of the waste form solid greatly increases the potential for leaching. This is a key factor in fabrication of waste forms and is discussed later in regard to cementitious materials.

5.19.3 Design of Waste Form Ceramics When waste ions, assumed to be present as oxides or as compounds that form oxides upon melting in air, are incorporated in borosilicate glass, they are normally incorporated as network formers or network modifiers, and the waste oxides can simply be added to the precursor glass chemicals, which are usually in the form of glass frit produced by pouring the molten glass into water. However, the situation is somewhat different when waste oxides are added to targeted ceramic precursors with the object of forming nearly water-insoluble multication mineral oxide phases, insofar as the waste ions in the mineral phases enter by a substitution mechanism, not by simple addition. The partitioning of the waste ions will depend on the cation site in which they substitute so that unequal partitioning of the waste ions will result in the formation of extra phase(s) if the waste ions are just added to the mineral phase precursor. Therefore, substitution of waste ions in a multication mineral phase requires prior adjustment of the overall stoichiometry and detailed knowledge of the preferred sites of the waste ions in the mineral phase. Moreover, if the valences of the guest and host ions are different, charge compensation is necessary to preserve electroneutrality of the phase. If the valence of the guest ion is less than that of the host ion, charge compensation can be

Ceramic Waste Forms

maintained by oxygen vacancies or by the addition of cations with higher valences than that of the host. So, for example, if it is desired to substitute two monovalent alkali ions into two divalent host ion sites, we can maintain phase electroneutrality by 2M2þ þ 2O2 $ 2Mþ þ O2 þ hO

½I

□O is a vacant oxygen site. However, we can also introduce a trivalent charge compensating cation for a monovalent ion substituted in divalent host sites via 2M2þ $ Mþ þ M3þ

½II

Similar considerations arise when it is desired to incorporate higher valence guest ions on to host sites, except that the charge compensators will be cation vacancies or substitutional cations with valences less than those of the host ions. Ceramics containing several phases are necessary when a full range of fission products, minor actinides, and process chemicals require immobilization, and the same principles of chemical accounting apply to multiphase ceramics made up of mutually compatible mineral phases. A first guess as to the way that the waste ions will incorporate themselves in the mineral phases will derive from a similarity of the approximately known7 ionic sizes between the host sites and the waste ions. The ionic sizes depend also on the valence of the ions so that the valences of the waste (and the host) ions need to be known. Also, the more similar the valences of the guest and host ions, generally the larger the solid solubility of the guest ions in the host phase. The valence depends on the crystal chemistry of the host site as well as on the prevailing oxygen fugacity (note that some elements such as the Pd group will likely form metals under some processing conditions, even oxidizing environments). A good example of crystal-chemical valence stabilization is the fact that monazite, CePO4, in which Ce exists as Ce3þ, can be fabricated by firing in air at 1200  C, whereas CeO2 (Ce4þ) can also be similarly formed in air, and analogous results are obtained for PuPO(48) and PuO2. Table 3 is a partial ‘library’ of candidate ceramic phases for radionuclide immobilization. Experimental studies to define the valences of variable-valence ions fortunately do not require the use of radioactive ions if stable ions of the corresponding elements exist and can be made from X-ray near-edge absorption spectroscopy, electronic optical spectroscopy, electron paramagnetic resonance,

Table 3

489

Partial ‘library’ of ceramic host phases

Phase

Radionuclide

Pollucite, CsAlSi2O6 Hollandite, (Cs,Sr,Ba,Rb)1.14 (Al,Ti3þ,Fe)2.28TiO16 Feldspar, CaAl2Si2O8 Apatite, Ca10([P,Si]O4)6(OH,F,Cl)2 NZP (Na,Ca0.5)(Zr,Ti)2(PO4)3 Monazite, REPO4 Garnet, Ca1.5GdTh0.5FeFe3SiO12 Zircon, ZrSiO4 Xenotime, YPO4 Zirconolite, CaZrTi2O7 Perovskite, CaTiO3 Fluorite, (RE,An)O2 Pyrochlore, RE2Ti2O7 Titanite, CaTiSiO5 Rutile, TiO2 Sodalite, Na4Al3Si3O12I

Cs Cs, Rb, Sr, Ba Sr, Ba RE, An, Sr, Ba Many RE, An RE, An RE, An RE, An RE, An Sr, RE, Tc, An RE, An RE, Zr, An RE, An, Sr Tc I

RE: rare earth; An: actinide; NZP: sodium zirconium phosphate.

electronic energy-loss spectroscopy, and Mossbauer experiments, to name but a few. Scanning/transmission electron microscopy can be employed to derive the stoichiometry of the various phases. For actinides and Tc, the actual radioactive isotopes have to be employed, although popular inactive simulated substitutes are Ru for Tc and Ce for the trivalent and tetravalent actinides. Facilities for examination of significantly radioactive samples are restricted to a few national laboratories in various countries. Having established the relevant ionic valences and the desired set of candidate mineral analog phases in the multiphase ceramics, the next requirement is to establish that the different phases are compatible (can coexist at elevated temperatures). Although attempts have been made to construct single phases to immobilize single or a range of waste ions, this strategy is difficult (see Section 5.19.4). Also, for a given waste form for a given HLW chemical composition, it is important that the waste form properties are flexible and not unduly compromised because of mismatches of waste/additives ratios, and variations of waste form chemistry, noting that HLWs are very frequently inhomogeneous mixtures of solutions and sludges, calcines, etc. Flexibility derives from the use of multiple phases and chemical buffering via the presence of a phase(s) that does not include radionuclides – then variations of chemical composition just result in a change of the proportions of the phases present, not the identities of the phases themselves.

490

Ceramic Waste Forms

5.19.4 Historical Evolution of Candidate Ceramic Waste Forms for HLW Although borosilicate glass worldwide had been the main candidate waste form for Purex-type HLW (see Chapter 5.18, Waste Glass) up to the mid-1970s, Pennsylvania State University workers noted9 that glasses were fundamentally unstable from a thermodynamic point of view, and they devised ceramic waste forms for HLW derived from nuclear fuel reprocessing, based on the known natural longevity of crystalline silicate, phosphate, and molybdate minerals. These so-called supercalcine ceramics10 were sintered in air at 1100  C and had very high loadings of fission products, typically 70 wt% of fission product oxides, and the chemistry of the different phases was driven by the fission products as majority components. Typical phases were pollucite (CsAlSi2O6), powellite (CaMoO4), and rare earth apatites (Ca2RE8(PO4)6O2) and phosphates (e.g., monazite, REPO4) (RE is trivalent rare earth). All of these had mineral analogs that were known to be very durable in the hot, wet conditions likely to characterize a deep geological repository for the waste. Following work at Sandia Laboratories in the United States on phase assemblages occurring on heating sol-gel titania particles on which simulated HLW fission products and actinides were sorbed,11 Ringwood and his coworkers in Australia in the late 1970s devised multiphase titanate-based ceramics in which nearly all the fission products and actinides in HLW from nuclear fuel reprocessing were incorporated substitutionally in the various mineral analog phases.12–14 Typical waste loadings were 20 wt% of HLW oxides and the production technology was slurry mixing of the waste and precursor oxides, calcination of the waste/precursor mixture in a reducing atmosphere, followed by hot uniaxial pressing (HUP) at 1100  C to make a dense ceramic. These ceramics are discussed in a little more detail in following sections. Primarily, these ceramics were focused on Purex-type HLW, but some work was done to immobilize SRL tank wastes from weapon production. At about the same time, and perhaps in part driven by the appearance of the synroc-type ceramics, there was a worldwide surge of interest in the immobilization of HLW. However, in the United States, a key decision was made during 1981–1982 to use borosilicate glass to immobilize HLW at SRL,2,3 and there was a substantial decrease in US funding for HLW waste form

research from then on. Nevertheless, a variety of alternative waste form development work continued around the world, and the book by Ewing and Lutze15 gives an excellent survey of research up to nearly the end of the 1980s. Candidate materials included glasses, ceramics, glass–ceramics, cermets, coated materials, and cements. However, in the course of time, it has been widely (but not universally) agreed that the only real remaining candidate types of material for HLW immobilization are glasses, ceramics, and glass–ceramics. These can be produced by Joule or cold-crucible melters, sintering or hot pressing, and particularly HIPing. However, cementitious materials may yet be useful for less active HLW such as US defense wastes from weapons production. Waste form development for HLW is still continuing in some shape or form in different nuclear countries, although Japan chose borosilicate glass in the mid-1990s and therefore ceased work on alternatives except in some niche areas, such as immobilization of 129I. France instituted the ‘law of 1991,’ which placed a moratorium on waste disposal until 2006, giving them 15 years of research to make a decision on the best choices of waste forms for their particular HLWs. In the 1990s, much work was done in France on apatites as candidate ceramics, and collaborations with the ANSTO group on titanate ceramics were initiated. A Th double phosphate Th4(PO4)4P2O7 ceramic for immobilization of tetravalent actinides has also been under development in France for several years. The material is claimed to have good aqueous chemical durability even when amorphized by swift heavy ions.16 Although numerous scientific papers on candidate ceramic waste forms still emanate from France, it is as yet not totally clear whether France will ultimately continue only with vitrification for their HLW or whether ceramics may play some role. Work has continued in Australia on titanate ceramics (see Section 5.19.5) although since around 2000, the work has evolved from a narrow focus on titanate ceramics to the use of HIPing technology to fabricate candidate glass–ceramic and ceramic waste forms to immobilize a wide range of HLW/ILWs that are generally perceived to be not very suitable for vitrification. Such waste forms have been directed at Pu-bearing wastes, Magnox sludges (UK), K-basin sludges (Hanford, WA, USA), Tc-rich waste,17 I-rich waste,18 Cs/Sr/Ba/Rb partitioned wastes,19 pyroprocessing wastes,20 and U-rich 99Mo production wastes.21 Collaborative work has also been done with UK workers on titanate ceramics, and UK workers have also

Ceramic Waste Forms

worked on other ceramic formulations. The AWE plc (Aldermaston, UK) has had an ongoing 10-year program to immobilize ILW derived from pyrochemical processing of Pu metal: they are using chlorapatiteand spodiosite (Ca2(PO4)Cl)-based ceramics.22 At ANSTO, work is being carried out under the Europart effort on pyroprocessing to try to immobilize waste pyroprocessing chloride salt in apatite-based ceramics.20 Russian workers are investigating murataite and garnet-based ceramics, together with perovskites.23,24 Also in Russia, investigations of the crystal chemistry of sodium zirconium phosphate (NZP)structured and other phosphates having a wide range of ionic substitutions have been carried out by Orlova’s group for many years.25 The effects on the NZP structure of a-damage from 238Pu and 239Pu substitutions have been observed.26 Other phosphates, notably monazites, have also been studied in regard to a-damage effects from 238Pu doping.27 It is appropriate at this stage to reiterate the diverse nature of HLW, depending in part on whether it derives from commercial Purex-type reprocessing or military Pu production. Generally speaking, the Purex wastes are highly active and are basically solutions of fission products and minor actinides in 3 mol l1 nitric acid solutions. The Pu production tank wastes, however, consist mainly of process chemicals, a relatively small amount of fission products in an alkaline solution derived from the need to neutralize nitric acid that would in time attack the stainless steel tanks. Thus, these wastes consist of a concentrated solution of Na salts plus hydroxide-rich sludges and are very inhomogeneous (and largely uncharacterized) even in single tanks. Hence, there is a possible need to actually separate individual wastes into solution and sludge fractions, and a definite need to design waste forms that can cope with diversity and compositional uncertainty. More importantly, longer-term (periods of years) tests need to be carried out to gain a mechanistic chemical understanding of the leaching behavior as distinct from the raw numbers in the prescribed leach tests (see Section 5.19.7). An important facet of leaching and long-term aqueous durability is the existence or otherwise of natural analogs of the phases making up HLW waste forms, because if the natural mineral can be found to exist in a wet environment, knowledge of the local geology can give information on the time of exposure to water, and measurements of trace quantities of natural radionuclides in the mineral (such as U, Th, K, Rb) and their daughter products can allow the age of the mineral to be determined.

491

In favorable circumstances, it can be determined that mineral analog phases can last up to millions of years in wet environments, just what is needed for the manmade phases for the sequestration of HLW. Thus, again, there is a powerful argument to use waste forms based on natural analog minerals that have demonstrated their survival over geological time frames. The well-known Oklo phenomenon in the African country of Gabon is worthy of mention in this context. Roughly 2000 Ma, enough U had aggregated in a geological formation there to form an intermittently self-sustaining nuclear reactor in the presence of water to moderate the neutron flux. At that time, the natural 235 U content of U was around 6%, as distinct from 0.71% today. The principal evidence for nuclear activity is a low abundance of 235U in the residual U-bearing material. It is reassuring that the movement of the fission products having half-lives long enough to be still active has been shown to be only a few meters at most away from the residual ore body. A recent description of the disposition of fission products at Oklo has been given by Hidaka.28 Single-phase ceramics have been widely advocated for both single radioactive elements formed by partitioning of reprocessing wastes or even for the entire complement of waste elements. NZP structures have been widely studied and/or advocated for the full range of fission products and actinides.29–31 Monazite, apatite, and zircon have been studied to immobilize actinides, while pollucite10 and CaAlSi5O1232,33 have been investigated for Cs immobilization. However ‘single phase waste forms’ lack chemical flexibility. An exact match of waste and precursor stoichiometries in single-phase multication hosts, such as those mentioned earlier, is industrially unrealistic when dealing with radioactive waste materials. What is needed is an ‘extra’ minor durable phase whose abundance may vary as the waste/precursor ratio varies while still maintaining the same qualitative phase assemblage-as in the synroc-type ceramics. As mentioned previously, the sintered supercalcines10 consisted of apatite and monazite phosphates, powellites, feldspar, pollucite, etc. But there are difficulties of diluting it with materials such as alumina, silicate, or phosphate to deal with radiogenic heat production, apart from the very inelegant approach of using cold fission products as diluents.34 Also, volatility losses during sintering would be severe, although these could be minimized by heating in neutral or reducing atmospheres, and using HIPing. The Rockwell Science Center (CA, USA) realized this latter factor as early as 1981, and they put

492

Ceramic Waste Forms

forward HIPing as the preferred consolidation method for their ceramics35 directed at the tank waste-type HLWs at the Savannah River Laboratory. The ceramics based on alumina tailoring contained magnetoplumbite [Ca(Al,Fe)12O19], UO2, spinel [Mg (Al,Fe)2O4], nepheline (NaAlSiO4), and corundum, with the former phase being seen as a near-universal solvent for fission products other than gaseous species. A titanate-based ceramic was broadly similar to the synroc-D ceramic (the following section) and contained zirconolite, nepheline, spinel, magnetoplumbite, perovskite, murataite (a complex fluorite-based phase), and glass. The waste loadings were around 60 wt%, and HIPing was carried out at 1040  C/60 MPa. There were no problems in these materials with radiogenic heat because the waste was quite dilute in fission products.

5.19.5 Titanate Ceramics Ringwood et al.,12–14 devised ceramics containing phases based on durable natural titanate minerals. These ceramics were called ‘synroc’ (synthetic rock). To deal with Purex-type waste, these theoretically dense materials are made by first mixing inactive precursors of Al, Ba, Ca, Ti, and Zr oxides with liquid (simulated) HLW, drying, and calcining in a H2/N2 atmosphere for 1 h at 750  C. The calcine was then mixed with 2 wt% of powdered Ti metal for redox control and then subjected to uniaxial graphite die hot-pressing or HIPing at 1100  C. The precursor composition and the titanate phases in the early synroc-C titanate ceramic designed for reprocessed commercial power reactor wastes are given in Table 4. Since 1984, rather than using oxides, a slurry mixture of Ba and Ca hydroxides and

transesterified Al, Ti, and Zr alkoxides has been used as the precursor. This provides better solid-state reactivity than the corresponding powdered metal oxides and hydroxides. The principal advantage of this synroc-C ceramic was that the waste ions were dilutely incorporated in durable titanate mineral phases that were considerably more insoluble in water than the silicates and phosphates, and the like used in supercalcine. The waste loading could be varied between zero and 35 wt% using the same inert additive chemistry without substantially changing the basic zirconolite þ perovskite þ hollandite þ rutile phase assemblage, although of course the percentages of the different phases varied somewhat.36 This flexibility is seen as a large advantage. There were minor alumina-rich phases in the more dilute formulations. The grain size is on the order of 1 mm (Figure 4) to optimize mechanical properties and prevent subsequent radiation-induced microcracking (see Section 5.19.8). For comparison, the grain size of a synroc-C sample prepared by sintering at 1300  C is somewhat larger (Figure 5). The alloy phases derive from ions that form metals under the reducing conditions prevailing during hot pressing. These ions are Mo, Tc, Pd, Rh, and Pd, plus any corrosion products from stainless steel. The leach rates at 90  C in water from synroc-C of the most soluble elements, alkalis and alkaline earths, are typically <0.1 g m2 day1 for the first few days, and they decrease asymptotically to values of 105 g m2 day1 after 2000 days (Ringwood et al.14 and see Figure 6). Leach rates of other elements are much

Table 4 Composition and mineralogy of synroc-C (20 wt% reprocessing waste loading) Phase

wt%

Radionuclides in lattice

Hollandite, BaAl2Ti5O14 Zirconolite, CaZrTi2O7 Perovskite, CaTiO3 Ti oxides, mostly TiO2 Alloy phases

30

Cs, Rb

30

RE, Zr, An

20 15

Sr, RE, An None

5

RE: rare earth; An: actinides.

Tc, Pd, Ru, Rh, Mo, Ag, Cd, Se, Te

6864

1.5 kV

X1000

10 µm WD37

Figure 4 Backscattered electron micrograph of Synroc-C. Large feature at bottom right is a partly oxidized lump of Ti metal that was added for redox control. Bright spots are metallic alloys and the remaining micron-sized features are the ceramic phases, rutile, hollandite, perovskite, and zirconolite.

Ceramic Waste Forms

493

Table 5 Seven-day MCC-1 leach results for different elements in synroc-C

P

R

Element

Leach rate (g m2 day1)

Element

Leach rate (g m2 day1)

Mo Cs Tc Ru Sr Ca

0.4 0.1 0.05 0.03 0.02 0.02

Al Zr Ti RE An

<0.4 8  104 2  104 104 – 103 2  105 – 5  104

RE: rare earth; An: actinides.

9636

1.5 kV

X3000

10 µm

WD 8

Figure 5 Synroc-C prepared by pressureless sintering at 1300  C. The black phase is rutile (R). The dark-gray phase is perovskite (P), the white spots are the metallic alloys, and the matrix is a mixture of Ba-hollandite and zirconolite of similar contrast.

Differential leach rate (g m–2 day–1)

1.0E+00

1.0E−01

1.0E−02 Cs 1.0E−03

Ba

Sr

Mass 1.0E−04 0

10

20

30

40 50 60 Time (days)

70

80

90

Figure 6 Leaching of synroc-C in seven-day Materials Characterization Center-1 test.

lower (see Table 5). Leach rates of 105 g m2 day1 correspond to a corrosion rate of ≲1 nm day1. In parallel, a HIPed synroc-D formulation having comparable performance to synroc-C was put forward in conjunction with the Lawrence Livermore National Laboratory (LLNL) in the United States13,37 to deal with the SRL defense waste, and this synroc derivative was based on zirconolite, perovskite, spinel, and nepheline. In the 1980s, the inactive Synroc production process was scaled-up via the Synroc Demonstration

Plant at ANSTO (then the Australian Atomic Energy Commission) to produce 50 kg monoliths containing 20 wt% of simulated Purex HLW (synroc-C), with leaching and microstructural properties as good as those of gram-sized laboratory samples. In the early 1990s, the synroc ceramics were tailored toward the study of zirconolite-rich materials for immobilization of actinide-rich wastes such as Pu or partitioned transuranic elements. The initial work during 1991–1994 was directed at the latter application in conjunction with the Japanese Atomic Energy Research Institute. There was a strong focus on radiation damage via the incorporation38 of the a-emitter 244 Cm (18 year half-life), as had been done with synroc-C and a Na-doped variant thereof.39 Perovskite was also studied for comparison. The work on surplus Pu immobilization, with LLNL as the lead laboratory for the USDOE, moved from zirconolite- to pyrochlore-rich ceramics during 1994–1997. This was because of solid solution limits in the first instance when the target of the work changed from immobilization of 10 wt% Pu (impure) alone to the additional inclusion of 20 wt% U. The estimated time for amorphization of these ceramics to be complete is on the order of 1000 years and the resultant volume expansion would be around 6%.40 This expansion in polycrystalline samples doped with 238Pu, which became X-ray amorphous after around 2 years’ storage at ambient temperatures, did not produce microcracking, and no significant radiation-enhanced aqueous dissolution rates were observed with the crystalline ! amorphous transition. In addition, these ceramics incorporated an atom each of neutron-absorbing Gd and Hf for each atom of Pu to deal with potential criticality in the sample. Nearfield aggregation of Pu due to leaching was shown to be not a problem from the criticality aspect either, because the measured leach rates of Pu were spanned by those of the neutron absorbers5: hence, any leached

494

Ceramic Waste Forms

1 µm 10 wt% HLW, cooled 20 ⬚C min–1 Figure 7 Pellets of Pu-bearing ‘hockey pucks’ prepared by sintering.

Pu would be accompanied by neutron absorbers, which in turn would inhibit criticality problems. The final baseline (no impurities) version4 of the pyrochlore-rich ceramics chosen by the US DOE in 1998 contained 95 wt% of a pyrochlore-structured Ca0.89Gd0.23Hf0.23U0.44Pu0.22Ti2O7 phase plus 5 wt% of rutile-structured Ti0.9Hf0.1O2. The form of the ceramic was to be 76-mm diameter pellets weighing 500 g (Figure 7). Five hundred and sixty such pellets were to be enclosed in a US standard canister of Savannah River DWPF glass to provide a radioactive barrier (gamma field) to prevent diversion. This product was the first crystalline material to be validated in the United States. However, in early 2002, it was decided to remove the disposal option for US/Russian surplus Pu and to proceed only with a MOX fuel option for utilization. This latter option however has not been realized. While it has been realized that substituting Zr for Ti in these ceramics would severely limit the amount of radiation damage sustained by the pyrochlore lattice,41,42 it has to be stressed that this would both increase the fabrication temperature by roughly 300  C and severely restrict the entry of impurities in the target impure Pu into the pyrochlore structure.43 Other synroc derivatives have been devised for immobilization of Tc44,45 and Cs/Sr/Rb/Ba19 formed when reprocessed waste is subjected to partitioning into chemically similar groups.

5.19.6 Glass–Ceramics Glass–ceramics in principle combine the advantages of glasses and ceramics. They can be made by melting, cooling, and reheating at 1000  C to induce

Figure 8 Phase separation shown by scanning electron microscopy in sphene glass–ceramic body after furnace cooling from melt. The continuous phase is Si-rich and the discontinuous phase is a (Ca, Ti, Si)-rich phase that forms sphene on reheating. Etched for 30 s in 1% HF solution.

crystallization, or processed at subsolidus temperatures and slowly cooled to ambient temperatures. Careful design can produce crystalline mineral analog phases chosen for their immobilization qualities together with a durable glass that can provide further tolerance for variations in the waste/precursor ratios and variations of the waste feedstock. Sphene glass–ceramics were developed for 6 years in Canada for HLW arising from a reprocessing option, until it was eventually decided in 1984 to follow the United States and concentrate on spent fuel immobilization. The Canadian glass–ceramics consisted of sphene, CaTiSiO5, in a durable aluminosilicate matrix. The overall composition of the wastefree precursor in mol% was: Na2O (6.6); Al2O3 (5.1); CaO (16.5); TiO2 (14.8); SiO2 (57.0) and considerable variations in this composition were allowable without compromise of the essential properties. The material was produced by melting at 1350  C, cooling to ambient conditions, then reheating for 1 h or so at 950–1050  C to allow the sphene to crystallize within the durable glass phase.46,47 Considerable phase separation occurred during the postcooling step (Figure 8). Loadings of Purex-type HLW were feasible, although additional perovskite and other phases were observed at waste loadings of >10 wt% fission product oxides. Workers at the Hahn-Meitner Institute in Germany studied the properties of borosilicate glasses containing Purex-type HLW and the glasses were devitrified. Different formulations yielded celsian (BaAl2Si2O8), fresnoite (Ba2TiSi2O8), diopside (CaMgSi2O6), or perovskite as major crystalline phases.48,49 The best versions were the materials yielding celsian and these were also studied in the United States. In the United States,

Ceramic Waste Forms

different groups studied the glass–ceramics derived from melting mixtures of natural basalt powder and HLW calcines.50,51 Hanford (WA, US) tank wastes are rich in alkali nitrates and transition metal hydroxides, and a range of glass–ceramics was designed for these.52,53 The 4400 m3 of calcines stored at the INL are rich in alumina, zirconia, and CaF2. While only about 30 wt% of these calcines can be incorporated in glass,54 glass–ceramics studied in the late 1980s and early 1990s and produced by HIPing to immobilize the calcines had waste loadings of around 70 wt%.55 These utilized SiO2-rich frit additives. Subsequently, ANSTO workers in unpublished reports have recently developed separate glass–ceramics for immobilization of the alumina-rich and the zirconia-rich ICPP (Idaho Chemical Processing Plant) calcines. Actinides in various HLWs have been preferentially partitioned toward synroc phases, principally zirconolite, in boroaluminosilicate glass matrices (unpublished work at ANSTO, Loiseau et al., 56 Mahmoudysepehr and Marghussian57). These glass–ceramics have waste loadings of 30–80 wt% and leach rates are often 10–100 times lower than those for standard US EA glass, the baseline glass to pass the product consistency test (PCT) leach test (see next section). These glass– ceramics were prepared by melting, apart from the ANSTO work in which the HIP method described in Section 5.19.10.2 was used.

5.19.7 Aqueous Dissolution Reactivity with water of solids in the first instance depends on the state of aggregation of the solid and clearly the dissolution rate of a solid body will be less than that of a fine powder. The dissolution rate itself can be quantified by measuring the concentration of dissolved species in the water in relation to the original inventory in the solid before the onset of leaching. Dissolution rates can then be expressed as elemental loss by mass per unit surface area (expressed as geometrical or Brunauer–Emmett–Teller (BET) values) per unit time (see Section 5.19.5). These rates, however, can depend critically on (surface area/liquid volume), pH, temperature, presence of salts in the water, etc. Dissolution rates can be further complicated by the presence of colloids and adherence of primarily dissolved species to vessel walls. However, separation of the solids from the liquid, followed by acidification of the liquid, can dissolve species attached to the leach container walls as well as colloids.

495

Colloids themselves can be detected by light scattering measurements. Ultimately, the dissolution rate of a solid in water is controlled by thermodynamics and various software programs are available to describe the dissolution process, although they are usually limited by lack of basic data for some of the ions in the solid. Many laboratories now use apparatus in which the liquid of interest flows over the solid at a given rate, and the rate is not too high to prevent a measurable concentration of dissolved species to accumulate and not too low as to allow the buildup of high concentrations of dissolved materials and consequent complications by solution saturation. In most repositories, the limiting temperature would be designed as 100  C, so measurements on candidate waste forms are experimentally relatively simple. Key regulatory leaching tests are the Materials Characterization Center (MCC)-158 and PCT-B59 protocols (see Chapter 5.18, Waste Glass), which employ polished flat samples and powders, respectively, that are exposed to hot water. The required time of immobilization for real HLW is variously targeted as 104–106 years, and it is worthy of inquiry as to how real-time measurements can be accelerated. This is very difficult as attempts to accelerate leaching by using higher temperatures or more aggressive solutions are easily compromised because the thermodynamics of the solid–liquid interaction can be grossly affected by such means. In practice, short-term (a few days) leach rates of 1 g m2 day1, normalized to take account of the fractional elemental extraction, rather than the absolute quantities, are considered as satisfactory and these rates correspond to  0.1 mm year1, further noting that the leach rates of solids tend to decrease with increasing leaching time even at high degrees of dilution. This is generally attributed to the presence of ‘active surface sites’ on the cut or polished prepared surface of a candidate solid.

5.19.8 Radiation Damage The radionuclides to be immobilized in an HLW waste form include a, b, and g emitters. The most serious damage to the waste form derives from a-decay, in which the a-particle displaces around 100 atoms in the solid during its 20 mm traverse and more importantly, the heavy a-recoil atom which displaces 1500 atoms over its 20 nm trajectory. b- and g-processes produce ionization damage but very few atomic displacements. These effects

496

Ceramic Waste Forms

have been amply demonstrated in natural minerals that contain small amounts of U and Th and have ages of many millions of years, and they have been essentially reproduced in experiments on synthetic materials doped with a few percent of short-lived actinides (238Pu or 244Cm, which have half lives of 87 and 18 year, respectively). Thus, only radiation damage processes in waste forms hosting significant amounts of actinides, especially Pu and other transuranics, need serious consideration.60 The variety of radiation effects include a crystalline ! amorphous transformation after hundreds or thousands of years, with an associated lattice expansion and an associated decrease of several percent in density (e.g., 16% in zircon, ZrSiO4), the production of lattice defects in solids that do not undergo amorphism, formation of gas bubbles, potential for enhanced leaching, and radiolytic effects in which radiolysis of the water leads to the production of species such as H2O2 that may responsible for enhanced leaching. Careful work at Pacific Northwest National Laboratory (PNNL) by Strachan et al.40 has shown that there is no significant leachability increase in pyrochlore- and zirconolite-based ceramics (see Section 5.19.5) as radiation damage progressively builds up. Although it has also long been argued that these kinds of radiation effects on glasses are relatively trivial, it needs to be remembered that the baseline leachabilities of glasses tend to be some orders of magnitude higher than those of crystalline waste forms. Another effect is ‘transmutation damage,’ due to the ionic size and valence changes that may accompany a- or b-emission. Particular examples are Csþ ! Ba2þ and Sr2þ ! Y3þ ! Zr4þ , where the ionic size decreases are 20 and 30%, respectively, for the full decay schemes. These effects have not been studied in much detail because of the intense radioactivity associated with waste forms containing several percentages of the parent isotopes, but sympathetic valence changes in the matrix ions, for example, Csþ þ Ti4þ ! Ba2þ þ Ti3þ and/or the production of hole centers can help to mitigate the charge changes in these decay series.61 Although for ions with half-lives of several years the radioactivity is extremely intense (terabecquerels per gram) and so severely limits available experimental data for these effects, transmutation effects are amenable to study using current atomistic modeling codes and predictions may soon be made by these techniques.

In polycrystalline ceramics, there are two significant potential effects that can lead to microcracking.62 Microcracking can lead to greatly increased leach rates because of the increase in surface area available for water to contact. First, if the actinide-bearing phase is anisotropic, stresses are set up because of unequal lattice dilatations from radiation damage, and secondly, different actinide concentrations in the different phases lead to unequal lattice dilatations between the different phases. However, these effects can be minimized by minimizing the grain size, and this will be discussed in the processing section later.

5.19.9 Thermodynamic Stability of Multication Oxides The intrinsic thermodynamic stability of crystalline waste forms over glasses has long been discussed. For multication oxides, one question is whether the multication oxide has a lower free energy than the component oxides. If not, the multication oxide is unstable with respect to the component oxides. Another question is whether a multication oxide is stable with respect to a simpler multication oxide. The fundamental experimental method is to measure the heat of dissolution of the oxides in question in molten salts such as Pb borates or molybdates at temperatures of around 700  C, and studies of many candidate ceramic waste form phases have been made by Navrotsky’s group in the United States.63 For pyrochlore-type phases, it has been found that several are unstable with respect to forming perovskite phases, but stable with respect to decomposition into simple oxides, whereas both Zr- and Hf-zirconolites are stable with respect to both decomposition modes.64

5.19.10 Processing of Ceramics and Glass–Ceramics The principal options for the production of refractory ceramic or glass–ceramic materials are sintering, uniaxial hot pressing, HIPing, and melting. As mentioned earlier, it is clearly advantageous if these materials are fully dense or at least devoid of open porosity to prevent ingress of water into the interior of the material. All wastes are calcined to remove organics, nitrates, water, carbonates, hydroxides, etc. If relatively small amounts of material are to be dealt with

Ceramic Waste Forms

(say a few tens of tons only), it might be advantageous to mix the waste with precursors before calcination. Though sintering is the baseline method in industry for making (inactive) ceramics, making dense ceramics via sintering is not a trivial task. Metal oxide phases, preferably with multiple cation sites to allow substitution of a variety of fission products, actinides, and process chemicals form the main constituents of ceramic waste forms, and it is necessary to achieve good mixing at an early stage of processing. Best mixing would utilize water-soluble liquid precursors such as metal nitrates to achieve atomic-scale mixing for highquality homogeneous ceramics, but this is easily seen to probably constitute overkill because the footprint of such a plant would be much larger, there are more process steps and nitric acid–based gases evolved which need to be dealt with as a separate (low-level) waste materials. Moreover, if Pu-bearing or enriched U-bearing materials are being immobilized, the use of water presents a criticality risk. So, it is not surprising that the synroc derivative (see above) for surplus US/Russian impure Pu utilized dry feeds that were attrition milled to achieve mixing and reactivity with the dry precursors. MOX fuel is also made by dry powder milling and sintering. These dry operations are most useful if radioactive volatile losses upon sintering are very small, as in the two cases just mentioned, but if volatile losses are potentially significant, HIPing of waste form ceramics has advantages (see Section 5.19.10.2).

5.19.10.2

497

Hot Isostatic Pressing

In HIPing of ceramics or glass–ceramics, the reactive calcined waste form (waste þ additives) material is first packed by vibratory means inside a relatively thin-walled metal can. This is then evacuated and heated to 300–600  C for several hours to remove adsorbed gases, sealed, and then consolidated to full density by compressing it with several tens or even hundreds of MPa of argon gas during the heating cycle. The use of a suitable metal container, which may be stainless or mild steel, nickel, or other metals, can help to achieve the desired redox conditions, minimize any potentially deleterious reaction between the waste form and the container, and of course prevents offgas escape. So the entire process produces offgas only in the calcination stage where temperatures are much lower than those in the final consolidation (roughly the same as those used for vitrification, that is, 1000– 1400  C). Figure 9 shows the steps involved in HIPing. Figure 10 shows stainless steel HIP cans before and after consolidation, and Figure 11 shows a diametrically cut section. The process inherently has a batch approach but cans containing more than 100 kg are feasible, with a processing time of 10 h. Work to shorten this time

Radioactive wastes

5.19.10.1 Hot Uniaxial Pressing Hot Uniaxial Pressing (HUP) is a batch process. An inductively heated graphite die can be used and this imposes reducing conditions on the sample, even if the sample is contained in a collapsible metal container or can within the bore of the graphite die. Only 20 MPa of pressure can be exerted with graphite, but very high temperatures (>2000  C) are possible. However, for waste forms, it is usual to keep the temperature to a maximum of around 1400  C to minimize volatile losses. The use of alumina dies allows the use of more pressure and oxidizing conditions, but it is more difficult to extricate the sample, especially when it is radioactive. Of course, in principle, the hot-pressed sample may be left in the graphite or alumina die, which then would constitute an expensive transport container insert. However, HIPing has basically superseded HUP as a processing option for oxide-based ceramics.

Additives

Mixer/drier

Calcination (if required)

Off gas treatment

Fill and seal HIP canister

Preheat

HIP

Canister disposal Figure 9 Flowsheet for hot isostatic pressing.

498

Ceramic Waste Forms

Figure 10 Stainless steel cans before and after hot isostatic pressing.

submarines and has been validated at INL as a credible (and advantageous) method of consolidating radioactive ceramic waste forms. Moreover, the method is widely used in industry for preparing inactive ceramics. A large advantage is also the relatively small footprint of HIP equipment, arising in the first instance because of the absence of off-gas in the hotconsolidation step. Moreover, the main part of the HIP equipment can be located outside the hot cell so that the HIP does not experience significant radioactive contamination and therefore require disposal at the end of the waste treatment campaign. Also, the HIP process can be used for encapsulation in metal for some wastes. Such examples that have been demonstrated inactively are Sn encapsulation of 129I sorbed on zeolites18 and unpublished ANSTO work on Cu encapsulation of spent fuel pellets and zircalloy liners. For radioactive ceramic waste forms, a prime advantage is to achieve theoretical density with minimum temperature and therefore minimum grain size, thereby adding to the overall strength and reducing the potential of microcracking via radiation damage when the waste form contains a substantial amount of a-emitting waste actinides. In addition, it has been shown for several types of ceramic waste forms that HIP can/ceramic interactions are not deleterious.65–67 5.19.10.3

Figure 11 Diametrically cut can that was hot isostatically pressed.

using hot calcined powders instead of allowing them to cool to ambient temperatures is under way, and throughputs of tonnes per day are targeted. Figure 10 shows that the consolidated can has a basic cylindrical shape, that is part of the can design allowing it to occupy a minimum of space in a cylindrical US transport container. While the relationship of the shapes of the can before and after HIPing is quite complex, the basic variable is the ratio of the densities of the final ceramic and the calcined powder. The dumbbell shape of the can gives quite a deal of flexibility, but it is advantageous to maximize the density of the calcined powder to avoid undue rippling and substantial deviations from cylindrical geometry of the HIPed can. The HIP process, invented by the Battelle company in the United States in the 1950s, has been used since the 1960s in preparing nuclear fuel for

Melting

Joule melters employ large refractory ceramic baths (several square meters in area and 1 m or so deep) containing ceramic electrodes to directly heat mixtures of glass frit and nuclear waste to a molten state. The melt is then poured into steel transport canisters in which it cools slowly to yield a borosilicate glass. Such melters are only viable in the longer term if temperatures do not exceed 1150  C. For higher temperatures, cold crucible melters are necessary. Cold crucible melting utilizes inductive coupling between a water-cooled high-frequency coil and conductive waste form.68 This coupling is inclined to be low when heating is commenced, so metal or graphite often needs to be added to the ceramic charge. This oxidizes and the metal oxide combines with the charge when high temperatures are reached, while the graphite is lost as CO2 under these conditions. While some ceramics such as synroc-C can readily be melted at temperatures below 1500  C,69,70 there are serious questions of fission product volatility (captured and recycled), especially if reducing conditions are not maintained. Moreover, on cooling from the melt

Ceramic Waste Forms

at rather low cooling rates governed by the large size of canisters into which the melts would be poured (pouring would not be easy if melts were not produced in an air atmosphere), crystallization of the waste form would allow the formation of quite large grains (Figure 12). This factor would adversely affect the mechanical strength and the response to a-radiation if the wastes contained significant actinide inventories.

attendant fission product volatility can be avoided. The composition of ordinary Portland cement (OPC) is given approximately in Table 6. The curing of cement after mixing it with water and aggregate is complex and takes place over months and years; the phases in the dry cement clinker gradually transform to hydrated phases, with a CaO–SiO2–H2O (C–S–H) tobermorite phase of variable composition as the main contributor to the eventual strength of the material. OPC can be diluted with flyash from coal-fired power plants or ground blast-furnace slags (approximate compositions in Table 7). In addition to utilizing waste materials, the well-known alkali-aggregate deleterious reaction can be minimized. The behavior of such cements in immobilizing LLWs and ILWs has been extensively reviewed.72–74 Work on cements processed under steam pressure conditions at 100–400  C (FUETAP) at the end of the 1970s and early 1980s75 indicated that good compressive strengths and thermal conductivities were achievable, and the derived leach rate was in the zone of 1 g m2 day1 or better for most species, especially if Cs had been sorbed on to zeolite and then incorporated in the cement, apparently showing that cements were intrinsically leach resistant, nearly on a par with borosilicate glasses. However, cements in which the Cs is not presorbed on to zeolites fail the standard leach tests described earlier by factors in the range of 10–100 for ions that do not form insoluble hydroxides – such as Cs (especially in MCC-1 tests in which the sample geometrical surface area to leach liquid volume is low) – and they are not these days seen as serious candidates for HLW immobilization. Also, the presence of water even in a bound state is problematical for radiolytic hydrogen buildup. However, cementitious material is currently seen generally as still having strong potential for LLW and ILW immobilization.

5.19.11 Cements and Geopolymers Cement is at first sight a particularly attractive means of HLW consolidation since high temperatures and

15277

23 µm Figure 12 Microstructure of synroc formulation designed71 for (Al,U)-rich high-level waste and processed by cold crucible melting. Dark areas: spinel, MgAl2O4; Light gray: pyrochlore-structured (Ca,U) titanate phase; darker gray areas: hollandite and rutile.

Table 6

499

Approximate cement and geopolymer compositions (excluding hydrous material and carbonate)

Composition (wt%)

CaO

SiO2

Al2O3

Fe2O3

Na2O

MgO

SO3

OPC Geopolymer

60 5

20 50

5 25

5 5

2 15

5

3

Table 7

Flyash Slag

Approximate blast-furnace slag and flyash compositions CaO

MgO

Al2O3

SiO2

Fe2O3

C

Na2O

K2O

5 40

1 10

25 10

50 40

10

5

3

2

500

Ceramic Waste Forms

Geopolymers are a class of cementitious materials that can be described as alkali-activated cements.76–79 They are made by reacting at ambient temperatures aluminosilicates such as metakaolin, fly ash, or ground blast-furnace slags with alkaline solutions, usually strong NaOH solutions. The baseline stoichiometries (Table 6) for the materials participating in the reaction are typically Na/Al ¼ 1 and Si/Al ¼ 2 on a molar basis, maximizing the strength80 and a minimum amount of water (H2O/Na  7 on a molar basis) is used to assure approximately a maximum amount of reaction. The properties are relatively insensitive to variations in the molar ratios at the level of 10 or 20%. The aluminosilicates partly dissolve in the solutions and polymerize and solidify. Curing is carried out at 40–90  C. Extensive studies by solid-state nuclear magnetic resonance have been carried out over the years together with porosity studies, so it is now accepted81 that geopolymers consist of nanoporous aluminosilicate networks, with water in the pores, although micro- and macroporosity is also present. Further evidence for nanoporosity has been gleaned from transmission electron microscopy and mercury porosimetry82 and positron annihilation lifetime measurements.83 Samples that pass the PCT aqueous dissolution test84 (but which have high MCC-1 leach rates of alkalis) can be fabricated by this technique. Systematic studies of the aqueous leaching behavior of geopolymers have not yet been carried out, but measurements of time dependence (1–90 days) and surface area/volume ratios suggest that at ordinary temperatures (90  C) the principal leaching mechanism derives from exchange of the pore water with the leaching solution, rather than attack of the aluminosilicate framework. As expected from the strength measurements,80 metakaolinite-based geopolymers having Na/Al  1 and Si/Al  2 molar ratios have maximum aqueous durability.85,86 Further measurements in progress at ANSTO are looking at the effects of temperature over the range of 20–90  C, pH in the range of 2–12, and the effect of bicarbonate and chloride ions in the leaching solution. Geopolymers have better fire and acid resistance than standard OPC. Moreover, geopolymers are serious candidates for ILW, especially as they can be dewatered by heating to 300–400  C without significant effects on their mechanical and chemical properties87 as long as the thermal ramp rate is kept fairly low to minimize structural disruption from the egress of the water. This is to be contrasted with OPC in which the C–S–H strength-building phase is decomposed at

temperatures above 200  C, thereby severely impacting the mechanical properties. Geopolymers have been used in Slovenia and Kazhakstan to immobilize large amounts of ILWs.88 Magnesium potassium phosphate (MPP) ceramics based on MgKPO46H2O have been developed at the Argonne National Laboratory since the early 1990s.89,90 Inactive versions of these materials have also been used in fertilizers for agricultural purposes.91 These materials can in some sense be regarded as cementitious or low-temperature ceramics. The distinction is perhaps academic as phosphate cements are well known. MPPs are prepared by mixing calcined MgO with strong solutions of KH2PO4 and allowing the following reaction to take place: MgO þ KH2 PO4 þ 5H2 O ! MgKPO4  6H2 O While it has recently been suggested that MPPs are appropriate for HLW immobilization,92 the leach rates in that work were based on BET surface areas. Perhaps more importantly, the hydrous nature of MPPs gives rise to a radiolytic H2 hazard, and unpublished experimental work at ANSTO has shown that MPPs become very weak structurally when heated above 400  C for the purposes of dewatering and/ or attempted conversion to anhydrous ceramics.

5.19.12 Conclusions In spite of more than 40 years of work, the disposition of high-level nuclear fuel wastes around the world in future is still subject to many uncertainties, especially with Yucca Mountain as an HLW repository in the United States being currently abandoned. Apart from political and NIMBY arguments, much of the scientific debate surrounds the question of how to validate physical models that lead to the calculated maximum radiation dose to persons living close to the repository, and more particularly how to convince a lay audience that the very complex calculations, including the uncertainties, are meaningful. However, the waste form is a key containment barrier because it can be subjected to rigorous experimental study, and optimization of its behavior can be studied directly at least over a few years. In this respect, more analog studies of natural minerals are needed, where although the water/thermal history of the analog mineral itself may be hard to derive, the history may well be derivable from the surrounding minerals in the rock formation. It seems clear that

Ceramic Waste Forms Table 8 Radioactive HLW particularly amenable to ceramification over vitrification Waste

Difficulty for vitrification

Contaminated metalsa High-Al and high-Zr wastes (ICPP, for example) Tc and Cs Actinides

Incompatible with glass Low solubility in silicate glass Volatile losses Low solid solubility of lower actinide valence states

a

Can be encapsulated in ceramics using HIPing.

5.

6. 7. 8. 9.

waste form development for the large spectrum of chemically distinct HLWs already in existence for many years, plus those yet to be generated by ongoing and future nuclear power programs, will continue. The twin foci of this continuance are simply (a) increased waste loading, especially in cases where radiogenic heating is not serious, to ease the amount of space required to contain the waste forms in repositories and (b) cost savings via development of technologies that maximize waste form throughputs and minimize plant footprints and radioactive offgas emissions. It is certain that HIPing will play an important role in achieving these improvements, particularly for wastes which are problematic for vitrification (Table 8) and for which there will be a strong preference for ceramics and glass–ceramics. Cements and particularly geopolymers remain as potentially viable for ILW. (See also Chapter 5.22, Minerals and Natural Analogues; Chapter 5.18, Waste Glass and Chapter 1.05, Radiation-Induced Effects on Material Properties of Ceramics (Mechanical and Dimensional)).

10. 11.

12. 13.

14.

15. 16. 17.

18. 19.

Acknowledgements

20.

The author wishes to acknowledge numerous colleagues at ANSTO and around the world for many discussions and contributions over many years.

21.

References

23.

1. 2.

3. 4.

Ewing, R. C. MRS Bull. 2008, 33, 338–341. USDOE. The Evaluation and Selection of Candidate High-Level Waste Forms; US Department of Energy Report DOE/TIC 11611; Savannah River Operations Office: Aiken, SC, 1982. Hench, L. L.; Clark, D. E.; Campbell, J. Nucl. Chem. Waste Manag. 1984, 5, 149–173. Cochran, S. G.; Dunlop, W. H.; Edmunds, T. A.; MacLean, L. M.; Gould, T. H. Fissile Material Disposition

22.

24. 25.

501

Program-Final Immobilization Form Assessment and Recommendation, UCRL-ID-128705; Record of Decision for the Surplus Plutonium Disposition. Final Environmental Impact Statement. Jan 4, 2000, US Department of Energy, 1997. Vance, E. R.; Jostsons, A.; Moricca, S.; et al. Synroc derivatives for excess weapons plutonium. In Ceramics Transactions (Environmental Issues and Waste Management Technologies IV); Marra, J. C., Chandler, G. T., Eds.; American Ceramic Society: Westerville, OH, 1999; Vol. 93, pp 323–329. Sizgek, G. D. Nucl. Eng. Des. 2005, 235, 101–109. Shannon, R. D. Acta Crystallogr. A 1976, 32, 51–67. Bjorklund, C. J. J. Am. Chem. Soc. 1958, 79, 6347–6350. McCarthy, G. J.; White, W. B.; Pfoertsch, D. E. Mater. Res. Bull. 1978, 12, 1239–1245. McCarthy, G. J. Nucl. Technol. 1977, 32, 92–104. Dosch, R. G.; Lynch, A. W.; Headley, T. J.; Hlava, P. Titanate waste forms for high level waste – an evaluation of materials and processes. In Scientific Basis for Nuclear Waste Management; Moore, J. G., Ed.; Plenum: New York, NY, 1980; Vol. 3, pp 123–130. Ringwood, A. E.; Kesson, S. E.; Ware, N. G.; Hibberson, W.; Major, A. Nature (London) 1979, 278, 219–223. Ringwood, A. E.; Kesson, S. E.; Ware, N. G. Immobilization of U.S. defense nuclear wastes using the Synroc process. In Scientific Basis for Nuclear Waste Management; Northrup, C. J. M., Ed.; Plenum Press: New York, NY, 1980; Vol. 2, pp 265–273. Ringwood, A. E.; Kesson, S. E.; Reeve, K. D.; Levins, D. M.; Ramm, E. J. Synroc. In Radioactive Waste Forms for the Future; Lutze, W., Ewing, R. C., Eds.; North-Holland: Amsterdam, 1988; pp 233–334. Lutze, W.; Ewing, R. C., Eds. Radioactive Waste Forms for the Future; North-Holland: Amsterdam, 1988. Tamain, C.; Ozgumus, A.; Dacheux, N.; et al. J. Radioanal. Nucl. Chem. 2007, 273, 597–601. Carter, M. L.; Stewart, M. W. A.; Vance, E. R.; Begg, B. D.; Moricca, S.; Tripp, J. HIPed tailored ceramic waste forms for the immobilization of Cs, Sr and Tc. Global 2007: Advanced Nuclear Fuel Cycles and Systems, Boise, ID, Sept 9–13, 2007, pp 1022–1028. Vance, E. R.; Perera, D. S.; Moricca, S.; Aly, Z.; Begg, B. D. J. Nucl. Mater. 2005, 341, 93–99. Carter, M. L.; Gillen, A. L.; Olufson, K.; Vance, E. R. J. Am. Ceram. Soc. 2009, 92, 1112–1117. Vance, E. R. Disposition of Salt Waste from Pyrochemical Nuclear Fuel Processing. Global 2007: Advanced Nuclear Fuel Cycles and Systems, Boise, ID, 2007; pp 1412–1415. Carter, M. L.; Li, H.; Zhang, Y.; Vance, E. R.; Mitchell, D. R. G. J. Nucl. Mater. 2009, 348, 322–326. Metcalfe, B. L.; Donald, I. W.; Fong, S. K.; Gerrard, L. A.; Strachan, D. M.; Scheele, R. D. J. Nucl. Mater. 2009, 385, 485–488. Weber, W. J.; Navrotsky, A.; Stefanovsky, S.; Vance, E. R.; Vernaz, E. MRS Bull. 2009, 34, 46–53. Stefanovsky, S. V.; Yunditsev, S. Y.; Perevalov, S. A.; Startseva, I. V.; Varlakova, G. A. J. Alloys Compd. 2007, 444–445, 618–620. Orlova, A. I.; Kitaev, D. B.; Orlova, M. P.; et al. Synthesis and investigations of the phosphates of Th, U, Np, Pu, Am and lanthanides with monazite, zircon, kosnarite and langbenite mineral-like structures. In Special Publication No. 305(Recent Advances in Actinide Science); Alvarez, R., Bryan, N. D., May, I., Eds.; RSC Publishing: Cambridge, UK, 2006; pp 454–456.

502

Ceramic Waste Forms

26. Orlova, A. I.; Volkov, Y. F.; Melkaya, R. F.; Masterova, L. Y.; Kulikov, I. A.; Alferov, V. A. Radiochemistry 1994, 36, 322–325. 27. Burakov, B. E.; Yagovkina, M. A.; Garbuzov, V. M.; Kitsay, A. A.; Zirlin, V. A. Self-irradiation of monazite ceramics: Contrasting behaviour of PuPO4 and (La,Pu) PO4 doped with Pu-238. In Scientific Basis for Nuclear Waste Management XXVIII; Hanchar, J. M., StroesGascoyne, S., Browning, L., Eds.; Materials Research Society: Warrendale, PA, 2004; pp 219–224. 28. Hidaka, H. J. Nucl. Radiochem. Sci. 2007, 8, 99–103. 29. Vance, E. R.; Ahmad, F. J. Dissolution and reaction studies of cesium zirconium phosphate in aqueous media. In Scientific Basis for Nuclear Waste Management; Brookins, D. G., Ed.; Elsevier: New York and London, 1983a; Vol. 6, pp 105–112. 30. Roy, R.; Vance, E. R.; Alamo, J. Mater. Res. Bull. 1982, 17, 585–589. 31. Roy, R.; Yang, L. J.; Alamo, J.; Vance, E. R. A single phase {[NZP]} ceramic radioactive waste form. In Scientific Basis for Nuclear Waste Management VI; Brookins, D. G., Ed.; North-Holland: Amsterdam, 1983; pp 15–21. 32. Vance, E. R.; Ahmad, F. J. Nucl. Chem. Waste Manag. 1983b, 4, 171–176. 33. Adl, T.; Vance, E. R. J. Mater. Sci. 1982, 17, 849–855. 34. Vance, E. R. Earth Miner. Sci. 1980, 50, 19–21. 35. Morgan, P. D. E.; Clarke, D. R.; Jantzen, C. M.; Harker, A. B. J. Am. Ceram. Soc. 1981, 64, 249–258. 36. Vance, E. R. J. Australas. Ceram. Soc. 2002, 38, 48–52. 37. Hoenig, C.; Otto, R.; Campbell, J. Densification Studies of Synroc D for High-Level Defense Waste, 1983, UCRL-53392. 38. Mitamura, H.; Matsumoto, S.; Stewart, M. W. A.; et al. J. Am. Ceram. Soc. 1994, 77, 2255–2264. 39. Mitamura, H.; Matsumoto, S.; Hart, K. P.; et al. J. Am. Ceram. Soc. 1992, 75, 392–400. 40. Strachan, D. M.; Scheele, R. D.; Buck, E. C.; et al. J. Nucl. Mater. 2005, 345, 109–135. 41. Sickafus, K. E.; Minervini, L.; Grimes, R. W.; et al. Science 2000, 289, 748–750. 42. Begg, B. D.; Hess, N. J.; McCready, D. E.; Thevuthsan, S.; Weber, W. J. J. Nucl. Mater. 2001, 289, 188–193. 43. Stewart, M. W. A.; Begg, B. D.; Vance, E. R.; et al. The replacement of titanium by zirconium in ceramics for plutonium immobilization. In Scientific Basis for Nuclear Waste Management XXV; McGrail, B. P., Cragnolino, G. A., Eds.; Materials Research Society: Warrendale, PA, 2002; pp 311–318. 44. Hart, K. P.; Vance, E. R.; Day, R. A.; Begg, B. D.; Angel, P. J.; Jostsons, A. Immobilization of separated Tc and Cs/Sr in Synroc. In Scientific Basis for Nuclear Waste Management XIX; Murphy, W. M., Knecht, D. A., Eds.; Materials Research Society: Warrendale, PA, 1996; pp 281–288. 45. Vance, E. R.; Hart, K. P.; Carter, M. L.; Hambley, M. J.; Day, R. A.; Begg, B. D. Further studies of Synroc immobilisation of HLW sludges and Tc for Hanford tank waste remediation. In Scientific Basis for Nuclear Waste Management XXI; McKinley, I. G., McCombie, C., Eds.; Materials Research Society: Pittsburgh, PA, 1998; pp 289–296. 46. Hayward, P. J. Glass–ceramics. In Radioactive Waste Forms for the Future; Lutze, W., Ewing, R. C., Eds.; North-Holland: Amsterdam, 1988; pp 427–493. 47. Hayward, P. J.; Vance, E. R.; Cann, C. D.; Doern, D. C. J. Am. Ceram. Soc. 1989, 72, 579–586. 48. Lutze, W.; Borchard, J.; De, A. K. Characterization of glass and glass–ceramic waste forms. In Scientific Basis for Nuclear Waste Management; McCarthy, G. J., Ed.; Plenum Press: New York, NY, 1979; Vol. 1, pp 69–81.

49.

De, A. K.; Luckscheiter, B.; Lutze, W.; Malow, G.; Schiewer, E.; Tymochowicz, S. Fixation of fission products in glass–ceramics. In Proceedings of Symposium on Management of Radioactive Wastes and the Nuclear Fuel Cycle, Vienna, Austria, 1976; Vol. 2, pp 63–73. 50. Welch, J. M.; Schuman, R. P.; Sill, C. W.; et al. Iron enriched basalt for containment of nuclear waste. In Scientific Basis for Nuclear Waste Management; Topp, S. V., Ed.; North-Holland: Amsterdam, 1982; Vol. 6, pp 23–30. 51. Kelsey, P. V.; Schuman, R. P.; Welch, J. M.; Owen, D. E.; Flinn, J. E. Iron enriched basalt and its applications to Three-Mile Island radioactive waste disposal. In Scientific Basis for Nuclear Waste Management; Topp, S. V., Ed.; North-Holland: Amsterdam, 1982; Vol. 6, pp 533–540. 52. Vance, E. R.; Day, R. A.; Carter, M. L.; Jostsons, A. A melting route to Synroc for Hanford HLW immobilization. In Scientific Basis for Nuclear Waste Management XIX; Murphy, W. M., Knecht, D. A., Eds.; Materials Research Society: Warrendale, PA, 1996; pp 281–288. 53. Vance, E. R.; Carter, M. L.; Day, R. A.; Begg, B. D.; Hart, K. P.; Jostsons, A. Synroc and synroc-glass composite waste forms for Hanford HLW immobilization. In Proceedings of the International Topical Meeting on Nuclear and Hazardous Waste Management (Spectrum-’96); American Nuclear Society: La Grange Park, IL, 1996; pp 2027–2031. 54. Bolon, W. D.; Herzog, J. D.; Olson, A. L. Status Report: The Glass–Ceramic Processing Flowsheet for ICPP High-Level Waste; WINCO-1091; Westinghouse Idaho Nuclear Company: Idaho Falls, 1991. 55. Vinjamuri, K. Glass–ceramic waste forms for immobilization of the fluorinel-sodium, alumina, and zirconia calcines stored at the Idaho Chemical Processing Plant. In Environmental and Waste Management Issues in the Ceramic Industry II; Bickford, D., Bates, S., Jain, V., Smith, G., Eds.; American Ceramic Society: Westerville, OH, 1994; pp 3–12. 56. Loiseau, P.; Caurant, D.; Baffier, N.; Mazerolles, L.; Fillet, C. J. Nucl. Mater. 2004, 335, 14–32. 57. Mahmoudysepehr, M.; Marghussian, V. K. J. Am. Ceram. Soc. 2009, 92, 1540–1546. 58. USDOE. MCC-1 Static Test, Nuclear Waste Materials Handbook, US Report No. DOE/TIC-11400; Material Characterization Centre: Hanford, 1984. 59. ASTM Designation: C1285–97. Determining Chemical Durability of Nuclear Hazardous and Mixed Waste Glasses: The Product Consistency Test (PCT), 1997. 60. Weber, W. J.; Ewing, R. C.; Catlow, C. R. A.; et al. J. Mater. Res. 1998, 13, 1434–1484. 61. Vance, E. R.; Roy, R.; Pepin, J. G.; Agrawal, D. K. J. Mater. Sci. 1982, 17, 947–952. 62. Vance, E. R. Radiation effects in solids designed for immobilisation of high level nuclear waste. In Ceramic Interfaces: Properties and Applications V; Uematsu, K., Ed.; Trans. Tech Publications Ltd.: Switzerland, 2003; pp 267–277. 63. Navrotsky, A. Phys. Chem. Miner. 1997, 24, 222–241. 64. Putnam, R. L.; Gallegos, U. F.; Navrotsky, A.; et al. Formation energetics of ceramic phases related to surplus plutonium disposition. In Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries VI; Spearing, D. R., Smith, G. L., Putnam, R. L., Eds.; American Ceramic Society: Westerville, OH, 2001; pp 147–151. 65. Li, H.; Zhang, Y.; McGlinn, P. J.; Moricca, S.; Begg, B. D.; Vance, E. R. J. Nucl. Mater. 2006, 355, 136–141. 66. Zhang, Y.; Li, H.; Moricca, S. J. Nucl. Mater. 2008, 377, 470–475. 67. Zhang, Y.; Li, H.; McGlinn, P. J.; Yang, B.; Begg, B. D. J. Nucl. Mater. 2008, 375, 315–322.

Ceramic Waste Forms 68. Gombert, D.; Richardson, J. R. Nucl. Technol. 2003, 141, 301–308. 69. Sobolev, I. A.; Stefanovsky, S. V.; Ioudintsev, B. I.; Nikonov, B. S.; Mokhov, A. V. Study of melted Synroc doped with simulated high-level waste. In Scientific Basis for Nuclear Waste Management XX; Gray, W. J., Triay, I. R., Eds.; Materials Research Society: Warrendale, PA, 1997; pp 363–370. 70. Sobolev, I. A.; Stefanovsky, S. V.; Omelianenko, B. I.; Ioudintsev, S.; Vance, E. R.; Jostsons, A. Comparative study of Synroc-C ceramics produced by hot-pressing and inductive melting. In Scientific Basis for Nuclear Waste Management XX; Gray, W. J., Triay, I. R., Eds.; Materials Research Society: Warrendale, PA, 1997; pp 371–378. 71. Vance, E. R.; La Robina, M.; Li, H.; Davis, J. Synroc-D type ceramics produced by hot isostatic pressing and cold crucible melting for immobilisation of (Al,U) rich nuclear waste. In Scientific Basis for Nuclear Waste Management XXX; Dunn, D., Poinssot, C., Begg, B., Eds.; Materials Research Society: Warrendale, PA, 2007; pp 181–186. 72. Glasser, F. P. Cement Concr. Res. 1992, 22, 201–216. 73. Quillin, K. C.; Duerden, S. L.; Majumdar, A. J. Accelerated ageing of blended OPC cements. In Scientific Basis for Nuclear Waste Management XVII; Barkatt, A., Van Konynenburg, R. A., Eds.; Materials Research Society: Pittsburgh, PA, 1994; pp 341–348. 74. Milestone, N. B. Adv. Appl. Ceram. 2006, 105, 13–20. 75. McDaniel, E. W.; Delzer, D. B. Fuetap concrete. In Radioactive Waste Forms for the Future; Lutze, W., Ewing, R. C., Eds.; North-Holland: Amsterdam, 1988; pp 565–588. 76. Davidovits, J. Mineral Polymers and Methods of Making Them. U.S. Patent 4,349,386, 1982. 77. Davidovits, J. J. Therm. Anal. 1991, 37, 1633–1656. 78. Shi, C.; Krivenko, P. V.; Roy, D. M. Alkali-Activated Cements and Concretes; Taylor and Francis: London and New York, 2006; pp 313–319. 79. Provis, J. L.; van Deventer, J. S. J., Eds. Geopolymers: Structure, Processing, Properties and Industrial Applications; Woodhead: Abington Hall, Granta Park, Cambridge, UK, 2009. 80. Rowles, M. A.; O’Connor, B. H. J. Mater. Chem. 2003, 13, 1161–1165. 81. Barbosa, V. V. F.; Mackenzie, K. J. D.; Thaumaturgo, C. Int. J. Inorg. Mater. 2000, 2, 309–317.

82. 83. 84.

85. 86. 87.

88.

89.

90.

91. 92.

503

Bell, J. L.; Gordon, M.; Kriven, W. M. Microsc. Microanal. 2006, 12(Suppl. 2), 552–553. Vance, E. R.; Hadley, J. H.; Hsu, F. H.; Drabarek, E. J. Am. Ceram. Soc. 2008, 91, 664–666. Perera, D. S.; Vance, E. R.; Aly, Z.; Davis, J.; Nicholson, C. L. Immobilization of Cs and Sr in geopolymers with Si/Al molar ratios of 2. In Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries XI; Herman, C. C., Marra, S. L., Spearing, D., Vance, L., Vienna, J., Eds.; American Ceramic Society: Columbus, OH, 2006; pp 91–96. Ly, L.; Vance, E. R.; Perera, D. S.; Aly, Z.; Olufson, K. Adv. Technol. Mater. Mater. Process. J. 2006, 8, 236–247. Aly, Z.; Vance, E. R.; Perera, D. S.; et al. J. Nucl. Mater. 2008, 378, 172–179. Perera, D. S.; Vance, E. R.; Finnie, K. S.; et al. Disposition of water in metakaolinite-based geopolymers. In Advances in Ceramic Matrix Composites XI, Ceramic Transactions; Bansal, N. P., Singh, J. P., Kriven, W. M., Eds.; American Ceramic Society: Westerville, OH, 2006; Vol. 175, pp 225–236. Majersy, D.; Sekely, S.; Breza, M. In Verified Possibilities of Specific and Historical Waste Solidification; IAEA RCM on Behaviour of Cementitious Materials in Long Term Storage and Disposal of Radioactive Waste Meeting, Moscow, Russia, Sept 10–14, 2007; CD ROM, IAEA: Vienna, Austria. Jeong, S.; Wagh, A. S.; Singh, D. Chemically bonded phosphate ceramics for stabilizing low-level radioactive wastes. In Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries II; Jain, V., Peeler, D., Eds.; American Ceramic Society: Westerville, OH, 1996; pp 179–188. Singh, D.; Wagh, A. S.; Cunnane, J.; Saturia, M.; Kurokawa, S.; Mayberry, J. Phosphate-bonded ceramics as candidate final waste form materials. In Environmental Issues and Waste Management Issues in the Ceramic Industry II; Bickford, D., Bates, S., Jain, V., Smith, G., Eds.; American Ceramic Society: Westerville, OH, 1994; pp 165–174. Salutsky, M. C.; Steiger, R. P. J. Agric. Food Chem. 1964, 12, 486–491. Vinokurov, S. E.; Kulyako, Yu, M.; Slyuntchev, O. M.; Rovny, S. I.; Myasoedov, B. F. J. Nucl. Mater. 2009, 385, 189–192.

5.20

Metallic Waste Forms

W. L. Ebert Argonne National Laboratory, Argonne, IL, USA

Published by Elsevier Ltd.

5.20.1

Introduction

506

5.20.2 5.20.2.1 5.20.2.2 5.20.2.3 5.20.2.4 5.20.3 5.20.4 5.20.4.1 5.20.4.2 5.20.5 5.20.5.1 5.20.5.2 5.20.5.3 5.20.6 5.20.6.1 5.20.6.2 5.20.6.3 5.20.7 5.20.7.1 5.20.7.2 5.20.7.3 5.20.8 5.20.8.1 5.20.8.2 5.20.8.3 5.20.8.4 5.20.8.5 5.20.8.6 5.20.8.7 5.20.9 References

Waste Streams Electrometallurgical Wastes Aqueous Processing Wastes for UREX + Flowsheets Hulls and Hardware Processing Additives Alloy Compositions Processing Methods Alloy Production Waste Conditioning Testing Objectives Matrix Degradation and Radionuclide Release Waste Form Consistency Through Process Control Waste Form Acceptance and Regulatory Requirements Modeling Matrix Corrosion and Radionuclide Release Mechanisms Performance in Disposal System Processing Control Test Methods Electrochemical Test Methods Corrosion Test Methods Service Condition Test Methods Tests with INL MWF Formulation and Phase Compositions Radionuclide Distribution Electrochemical Tests Corrosion Tests Corrosion Models Repository Model Metallic Waste Form Product Consistency Summary

509 509 510 512 512 513 514 514 515 515 515 516 517 517 518 520 520 521 522 523 524 525 525 526 528 529 531 533 534 535 536

Abbreviations DOE EBR-II EBS FCR&D FPEX GTCC HLW MWF

US Department of Energy Experimental breeder reactor Engineered barrier system Fuel Cycle Research and Development Fission product extraction Greater-than-Class C High-level radioactive waste Metallic waste form

NIST

US National Institute of Science and Technology NL(i ) Normalized mass loss (based on element i) RCRA Resource conservation and recovery act SEM Scanning electron microscopy TALSPEAK Trivalent actinide–lanthanide separations by phosphorous reagent extraction from aqueous complexes

505

506

Metallic Waste Forms

TEM TRUEX UDS UREX VHT

Transmission electron microscopy Transuranic element extraction Undissolved solids Uranium extraction Vapor hydration test

5.20.1 Introduction This chapter provides an overview of current strategies and approaches for formulating, processing, testing, and developing performance models for the metallic waste forms that are being designed to immobilize high-level radioactive wastes. Some aspects have been completed and demonstrated, while others are planned approaches based on previous experience that have not yet been fully implemented. Issues that must be addressed when designing a waste form are conveniently grouped as performance, processing, and waste acceptance issues, although these are interrelated. Waste acceptance issues include regulatory requirements that the waste form and disposal system as a whole must meet, identification of waste form properties that can be measured to demonstrate that the waste form is acceptable, and specifications for performing those measurements. Performance issues include determining the capacity of the waste form to restrict the release of radionuclides to rates at which the engineered repository can meet regulatory requirements (this is the basis for waste form acceptability), and, importantly, the ability to predict the release rates over very long durations under the anticipated range of environmental conditions with a mathematical model. Processing issues include the ability to produce waste forms from the anticipated range of waste stream compositions and additives that have consistent, predictable, and acceptable chemical, radiological, and physical properties by maintaining a consistent phase assemblage and distribution of radionuclides. Although the primary role of a waste form is to retard the release of the radionuclides it contains to the surrounding environment, this should occur in a predictable manner that can be related to both long-term performance and waste form production controls. The link between production and performance will provide confidence to regulators that the performance of any waste form produced within the anticipated range of waste stream compositions is adequately represented in the performance assessments used to license the disposal facility. The radionuclide release rates and the degradation rate of the host waste form matrix are used as

source term models in performance assessment calculations to predict the capacity of the disposal system to meet regulatory limits, such as groundwater dose limits or contamination levels at or beyond the system boundary. Assessments must address very long time periods and both anticipated environmental conditions and possible extreme conditions, including highly unlikely scenarios. The unprecedented challenge of predicting material performance for waste management over a geologic time scale has led to highly conservative bounds being used in risk and safety assessments. For example, the assessment calculations for the proposed Yucca Mountain repository include igneous intrusion (a magma plume from volcanic activity) and eruption scenarios. The inclusion of catastrophic events that have extremely low probabilities of occurring in performance assessments may be necessary for identifying conditions under which a repository may fail to meet regulatory requirements. Although those conditions will likely be far outside the anticipated range of service conditions, the possible impacts on the performance of the waste form should be considered when developing testing programs. In the case of metallic waste forms, this may call for an understanding of the impacts of long-term heating on the phase composition and retention of radionuclides of the host phases. The term ‘phase composition’ is used to denote the assemblage of component phases in the waste form, including solid solution and intermetallic phases, the chemical composition of each phase, and the distribution of radionuclides. The matrix material used to immobilize radioactive waste must be compatible with both the waste stream and the disposal environment. Although the specific disposal site may not be identified at the time waste forms are being designed, general site characteristics may be known, for example, if the site will be located in clay, salt, tuff, granite, or other medium, if it will be in a reducing or oxidizing environment, in a hydrologically saturated or unsaturated horizon, etc. If not even the generic disposal conditions are known, then testing programs must consider waste form compatibility in all potential environments. In most cases, the general compatibility of a waste stream and potential host matrix can be anticipated from known materials characteristics, but it must be evaluated experimentally to determine details of the phase composition and microstructure. The radionuclide(s) may be incorporated into the host matrix or sequestered in separate phases that are encapsulated by the matrix. For example, radionuclide-bearing oxide phases may be encapsulated within a metal

Metallic Waste Forms

matrix. In that case, compatibility may simply mean that the waste form can be processed to encapsulate the radionuclide-bearing phases within the matrix. Some waste forms may simply serve as a diffusion barrier to the ingress of water or the diffusive release of the radionuclides, but degradation of metallic waste forms will likely be required prior to radionuclide release. Note that ‘degradation’ is used to refer to alteration that results in the release of radionuclides, whereas ‘corrosion’ is used to refer to alteration that does not necessarily result in release. Regardless of the mode, the radionuclide release rate from each waste package must be predictable over the service life of the disposal system as a source term in performance assessment calculations. This usually means that waste form products must have consistent composition, degradation behaviors, radionuclide inventories, and size (exposed surface area). Consistent does not mean identical in the present context. Rather, it means within the range deemed to ensure acceptable performance. The anticipated range of waste stream compositions, which may vary due to variations in the fuels being treated and efficiencies of separations processes, and the capacity to blend waste streams and adjust additives, will affect the composition range of the waste forms. The impact of the anticipated range of waste form compositions on radionuclide release must be determined and taken into account in the source term model and preliminary performance assessment calculations. The impact on the performance of the disposal system can then be used to establish the acceptable composition range. This may require reformulation of the waste form, such as lowering the waste loading or changing the additives, to accommodate the range of waste stream compositions. Some radioactive waste streams are better immobilized in a metallic matrix than in glass or other matrices because the waste materials are metallic, because the radionuclides in the waste can be more efficiently retained during waste form processing if they are in the metallic state, or because the radionuclides are more effectively retained in the waste form if they are in the metallic state. Technetium is an important example of a radionuclide that is best processed and immobilized as a metal. The predominant species in most aqueous waste streams is the pertechnetate ion TcO
4 , but Tc(VII) is neither readily processed due to volatility of species such as CsTcO4 nor strongly incorporated into the structure of a glass or other matrix. The most stable form in an oxidizing disposal environment, TcO2, is sparingly

507

soluble in water, but sublimes at 900  C and Tc(IV) disproportionates to metallic Tc and gaseous Tc2O7 at about 1100  C.1 In contrast, metallic Tc can be processed at high temperatures under reducing conditions and alloyed in a metallic waste form without the loss of volatile species. Waste streams with high concentrations of Tc and other transition metals can also be immobilized in a metallic waste form at higher waste loadings than in a glass or other waste form matrix. Specific waste streams that are expected to be amenable to metallic waste forms are discussed in Section 5.20.2. In a metallic waste form, most radionuclides are dissolved or alloyed in a host metal with other waste constituents. Formulations of alloy waste forms must identify host metals and mixtures that can be processed at practicable temperatures and form durable phases to incorporate all radioactive components. Many metallic components in waste streams have very high-melting points and must be reactively dissolved into molten metals before they can be alloyed. Available binary and ternary phase diagrams provide insights into likely melting temperatures and phase compositions for processing complex mixtures of wastes and added metals. For example, significant amounts of Zircaloy scraps from cladding hulls in a waste stream can be dissolved into molten copper or iron below 1600  C, even though Zircaloy itself (Zircaloy-2 and Zircaloy-4) melts at about 1850  C. Other metals present in spent nuclear fuel and expected to be present in metallic waste steams either as pure metals or in alloys include Mo (2623  C), Ru (2334  C), Tc (2204  C), Rh (1963  C), and Pd (1555  C). The five-metal alloy formed from these components during irradiation of oxide fuels is expected to have a melting point near 2000  C (see Kaye et al.2). These high-melting wastes can be processed into a metal waste form at temperatures lower than their melting points by utilizing eutectic mixtures. The binary phase diagrams give insights into the solubilities of waste components in iron and identify intermetallic phases likely to form in the two-component systems. Elemental substitutions can also be predicted based on metallurgical data. For example, each component in the five-metal alloy is expected to be dissolved by molten iron at a temperature well below 1600  C, and so is the five-metal alloy. The ability to produce a waste form is an obvious initial requirement in waste form design. Formulation of metal alloy waste forms is discussed in Section 5.20.3. Contaminants will be present in the waste stream (e.g., oxides and sulfates) and absorbed by the molten

508

Metallic Waste Forms

metal from the crucible and furnace atmosphere (trace O2, CO, etc.) during processing. Slags (and dross) typically form on the surfaces of molten metals from impurities that are not incorporated into the alloys. Silica, calcium, and other oxidizable materials are commonly added to molten metals in industrial processes to help sequester impurities in slag layers that can be removed from the surface of the molten metal and discarded. The addition of silica usually results in the formation of a glassy phase in the slag that encapsulates other phases of metal oxide, carbide, sulfate, etc., and may itself be quite durable. Whereas slags are removed and discarded as waste from metal and alloy products, they can serve to immobilize contaminants excluded from the alloy phases of a metal waste form. The use of slag phases as a component of metallic waste forms is being studied.3 Processing methods are discussed in Section 5.20.4. The durability of a metallic waste form will depend on the durabilities of the component phases and their interactions with other phases in the waste form and with engineering materials, primarily through galvanic couples. The phases that are present and potential couples will affect the capacity of the waste form to retain radionuclides, and long-term performance models must take both into account. Clearly, a constant phase assemblage and a constant distribution of radionuclides between those phases will simplify performance models. Although a mechanistic understanding of waste form corrosion and radionuclide release adds confidence to predictions of long-term waste form performance, a fully mechanistic model for waste form degradation may be impractical due to budget limitations or unachievable due to an incomplete understanding. Many aspects can be adequately captured with empirical models, such as pH and temperatures effects, but confidence in empirical models based solely on experimental observations is usually limited to the duration of the longest experiments. A mechanistic understanding of processes that establish the time-dependent aspects of corrosion should be the top priority of a testing program. This could include understanding how passivating layers form, mechanisms for their breakdown, and the kinetics of their healing. The term ‘passivating’ is used in the very general sense of slowing the corrosion, regardless of the controlling factor. Key testing objectives are summarized in Section 5.20.4, including performance modeling, process control, and waste acceptance. Modeling is discussed in Section 5.20.5 and testing to develop a mechanistic degradation model is discussed in Section 5.20.6.

Consistency is another important requirement for metallic waste forms related to the need to predict waste form performance (particularly radionuclide release) over a very long disposal system service life. The performance of a metallic waste form in a given disposal system will, in general, depend on the phase assemblage that comprises the waste form and the radionuclide inventory. The preferred approach for producing consistent waste forms is control of the feed (waste streams and additives) and the processing conditions. Controlling the feed controls the gross composition of the waste form, and controlling the composition and processing condition controls assemblage of phases that form, and presumably, maintains a consistent distribution of radionuclides among the component phases. A key objective of waste form design and testing is to determine the relationships between (1) the process conditions and the phase composition and (2) the phase composition and waste form performance. These relationships link the radionuclide release rates calculated using the waste form degradation model with the controlled production of the waste form. These two sets of relationships should take precedence in waste form development and their importance should not be underestimated. Whereas both high waste loadings and high chemical durabilities are key objectives of waste form design, the acceptance by regulators for disposal will likely require evidence that the waste forms will perform to the level modeled in performance assessment calculations. Because it is not practical to subject samples of each waste form product to the suit of tests needed to characterize and model performance in the disposal system, the performance of each waste form product that is made can instead be related to the performance measured for the representative material based on the phase composition and microstructure. That relationship must be established through the waste form composition and processing conditions as a part of waste form development. Therefore, it is important to determine and experimentally demonstrate the ranges of waste form compositions and processing conditions that are represented by the model. That range of compositions is the basis for waste form acceptance. For borosilicate waste glasses, the relationship between performance and process control is demonstrated using the Product Consistency Test (ASTM standard test method C 1285), which is a partial dissolution test conducted under specified test conditions.4 An extensive database was generated to show the relationship between glass composition and test response,5 and the test response

Metallic Waste Forms

can be related to the modeled glass degradation and radionuclide release behavior.6 An analogous test method and database must be developed to establish the link between production control and performance for metallic waste forms. A metallic waste form was designed and developed to immobilize metallic wastes generated during the electrometallurgical treatment of spent sodiumbonded nuclear fuel from the Experimental Breeder Reactor-II (EBR-II).7
9 That waste stream consists of Type 316 stainless steel anode baskets, steel cladding hulls, and residual metals from the fuel that were not oxidized during the electrorefining procedure (primarily Zr). The primary radionuclide in the metallic waste from the fuel is 99Tc, but the waste stream also includes a small amount of entrained salt that contains actinides (primarily U and Pu). The work done to formulate this waste form, measure the corrosion behavior and radionuclide release, develop a performance model, and develop an approach for tracking waste form consistency provides a valuable example of the general approach for developing a metallic waste form and generating the database to support acceptance for disposal. To distinguish the specific metallic waste form alloy developed for electrometallurgically treated EBR-II fuel from other metallic waste forms, it will be referred to herein as the EBR-II metallic waste form (MWF). The work done to develop, test, and model the EBR-II MWF is summarized in Section 5.20.8. All testing, modeling, and production activities related to waste form development and qualification will likely be governed by Quality Assurance requirements. Most of the work done to formulate compositions, determine degradation mechanisms, develop models, and evaluate processing effects can probably be performed at less stringently controlled research and development levels. Other aspects, such as parameterizing the degradation models and establishing processing and consistency test limits, must likely meet higher Quality Assurance standards. These typically involve the use of documented and approved testing and analytical methods, certified and calibrated standards and equipment, fully documented and technically reviewed data and analyses, etc. Additional requirements apply to processing equipment and the actual waste form products.

5.20.2 Waste Streams Waste streams amenable to a metallic waste form either include a significant amount of metal waste

509

components or include radionuclides that are best immobilized in the metallic state. Waste streams that are currently being immobilized in a metal waste form10 or being evaluated for immobilization in a metallic waste form3,11 are discussed in the following sections. 5.20.2.1

Electrometallurgical Wastes

An electrometallurgical process has been developed under the auspices of the US DOE Office of Nuclear Energy to treat the spent sodium-bonded fuel inventory from the Argonne Experimental Breeder Rector-II (EBR-II) at the Idaho National Laboratory.9,12 The inventory includes 1.1 MT of what are referred to as ‘driver fuel rods,’ which contain the majority of the fission products, and 22 MT of blanket fuel rods.10 The driver rods contain an enriched U
10Zr fuel alloy (63% 235U) and the blanket rods contain depleted U. Both fuel types are fixed in the rods with metallic sodium, which provides a thermal bond between the fuel and cladding. Treatment is necessary to separate waste radionuclides from the metallic sodium prior to their disposal. The electrometallurgical process recovers the uranium from the fuel and generates two waste streams: sodium and most fission products dissolve in the waste salt, whereas steel cladding hulls and metals not oxidized under the refining conditions remain as metallic waste. A ceramic waste form has been developed to immobilize the salt wastes and a metallic waste form has been developed to immobilize the metallic wastes. The electrometallurgical process may be used to treat other DOE-owned spent sodium-bonded fuel in the future, for example, 34 MT of sodium-bonded blanket fuel from the Fermi 1 reactor,10 but that decision remains to be made. Electrometallurgical treatment of other spent fuels that would generate metallic waste streams is being considered. Scoping work was conducted to evaluate the use of electrorefining on oxide fuels that have been reduced as an alternative to aqueous reprocessing of oxide fuels, but that option is not currently being pursued. The more likely use of electrorefining will be treatment of the used metallic fuels being developed as part of a closed fuel cycle within the DOE Fuel Cycle Research and Development program (FCR&D). New metallic waste forms will be needed for these wastes due to the planned used of Zircaloy cladding for the new fuels. The alloy developed for the steel-clad sodium-bonded EBR-II fuel is not appropriate for Zircaloy-clad fuel. However, it should be noted that Zircaloy cladding could

510

Metallic Waste Forms

be recycled in the future and the waste stream compositions might not be dominated by Zircaloy cladding hulls. About 90% of the EBR-II inventory is steel-clad fuel and the rest is Zircaloy-clad fuel. The blanket fuel cladding is Type 304L stainless steels and the driver fuel claddings include Type 316, Type D9, and Type HT9 stainless steel. The waste form made with steel cladding was the primary focus during waste form development research, and only few scoping studies were carried out to evaluate potential waste form compositions for waste streams dominated by Zircaloy cladding.8,13
17 The composition and the microstructure of the waste form are affected primarily by the relative amounts of stainless steel fuel cladding and Zr from the driver or blanket fuel in the mixture used to make the alloy. The cladding supplies Fe, Cr, Ni, Mo, Mn, Co, Cu, V, and Si to the metal waste stream, plus trace amounts of Sn, C, and S. The chemical composition of the waste stream is dominated by Fe and Zr. The low-melting eutectic composition Fe
15Zr was selected as the target composition of the waste form to allow processing at about 1600  C. This requires adding Zr to the mixture. The predominant phases that are formed are those predicted by the Fe
Zr binary phase diagram, namely, similar amounts of an iron solid solution and an Fe2Zr intermetallic phase. Radionuclides and other components in the steel (e.g., Ni and Cr) are distributed between these phases and minor amounts of other phases (e.g., Fe23Zr6). This is discussed in detail in Section 5.20.8.1. The primary radionuclide is 99Tc, but the waste contains a small amount of carry-over salt with actinides that are retained as contaminants in the waste form. A small amount of depleted uranium is added to some waste mixtures to down-blend the 235U content in the waste form to below 20 mass%. Most waste forms are expected to contain about 1 mass% U, but some could contain as much as 11 mass% U. Table 1 Nuclide 14

C Co 93m Nb 94 Nb 59 Ni 63 Ni 60

The radionuclides expected to be immobilized that provide the highest activities are listed in Table 1, expressed as the activity in a disposal canister containing two waste form ingots (see Appendix A in Ebert.)18 The 18 radionuclides listed in Table 1 provide 99.6% of the total 3  1012 Bq expected in a canister. For comparison, the total radionuclide activity in an average canister of high-level radioactive waste (HLW) glass is estimated to be more than 2.8  1015 Bq. 5.20.2.2 Aqueous Processing Wastes for UREX + Flowsheets The US Department of Energy is evaluating methods to reprocess commercial spent fuel from light water reactors through the FCR&D program. The technologies developed to partition fuel elements and recycle actinides are intended to expand the nuclear power capacity to meet the global future energy needs while minimizing the amount of high-level radioactive waste requiring geologic disposal. The near-term goal is to develop the technologies needed for a proliferation-resistant nuclear fuel cycle. To this end, various operational flowsheets are being developed and tested to recover uranium and actinide elements from spent oxide fuel for recycle and separate long-lived fission products from other fuel wastes for separate immobilization and disposal. An important potential benefit of the processing operation is minimizing the heat load of wastes destined for geologic disposal. The partitioning of short-lived heatgenerating radionuclides, such as 137Cs and 90Sr, from short-lived (e.g., lanthanides) and long-lived radionuclides that do not generate significant heat loads, such as the 129I, 99Tc, and 79Se, provides the opportunity to tailor waste forms to the properties of particular waste streams and dispose these waste forms according to risk.3,11 Although current US regulations classify all waste streams from reprocessed fuel

Estimated radionuclide activities in EBR-II metallic waste form Bq/canister 10

2.63  10 1.01  1011 3.24  1010 1.68  1010 6.59  1010 1.93  1012

Nuclide 234m

Pa Pu 125 Sb 126 Sb 126m Sb 126 Sn 239

Bq/canister 8

5.99  10 2.04  109 3.30  109 2.42  109 1.73  1010 6.22  109

Nuclide

Bq/canister

99

7.84  1011 8.07  108 5.99  108 4.74  109 1.55  108 5.99  108

Tc Te 234 Th 234 U 235 U 238 U 125m

Data from Ebert, W. L. Testing to Evaluate the Suitability of Waste Forms Developed for Electrometallurgically Treated Spent SodiumBonded Nuclear Fuel for Disposal in the Yucca Mountain Repository, Argonne National Laboratory Report ANL-05/43; Argonne National Laboratory: Argonne, IL, 2005.

Metallic Waste Forms

as high-level radioactive waste, safety analyses might indicate some waste streams/waste forms as benign and the risks low enough that they could be regulated as Class C or Greater-Than-Class C (GTCC)waste rather than high-level waste. The FCR&D program is currently studying the application of aqueous separation technologies to spent oxide fuels. Prior to dissolution, the cut fuel is heated in a voloxidation step intended to remove gaseous and volatile fuel components, including 3H, 14 C, and 129I. These are volatilized at increasing temperatures and are captured (separately) from the offgas for immobilization and disposal. Dissolution of these fuels (typically in hot HNO3 with HF) is incomplete, with residues including five-metal particles and Zr from the fuel and minute scraps of Zircaloy cladding generated when the cladding was cut. (Additional gas release occurs during dissolution.) The undissolved solids (UDS) are removed from the dissolved fuel solution by filtration. The UDS waste stream is suitable for immobilization in a metallic waste form. The dissolved fuel solution is subjected to a uranium extraction (UREX) operation that separates the U and Tc in the solvent phase and the other fuel components in the raffinate. The Tc (as pertechnetate ion) is removed from the U
Tc solution using an anion exchange change column. The U is recovered for recycle. The pertechnetate is later eluted from the column and can be recovered as either TcO2 or tetra n-butyl pertechnetate using one of the two methods that are being developed. The TcO2 can be reduced to Tc metal by heating in a reducing atmosphere (e.g., CO or Ar/H2 gas) and the tetra n-butyl pertechnetate can be reduced to Tc metal by steam reformation in a separate operation. The recovered metallic Tc is suitable for immobilization in a metallic waste form. The UREX raffinate can be treated to separate Rb, Sr, Cs, and Ba from the rest of the dissolved fuel with the fission product extraction (FPEX) operation. The 137 Cs and the 90Sr are the major heat-generating constituents of the fuel waste. Both have half-lives of about 30 years, decaying to stable 137Ba and 90Zr, respectively. The Cs/Sr/Ba/Rb stream will likely be immobilized in a glass waste form. The transuranic elements and lanthanides are next separated from the fuel solution by a transuranic element extraction (TRUEX) operation. The TRUEX raffinate contains dissolved transition metals from the fuel, but is dominated by ferrous sulfamate added to improve the efficiency of the TRUEX separation for Np and Pu. The sulfur must be removed from the TRUEX raffinate before it can be immobilized in either a glass or a metallic

511

waste form. The dissolved transition metal fission products can be recovered by increasing the pH to coprecipitate them with iron as hydroxides and then reduced for incorporation into a metallic waste form at higher waste loadings. Finally, the lanthanides are separated from the actinides using TALSPEAK (trivalent actinide–lanthanide separations by phosphorous reagent extraction from aqueous complexes). The actinides are recycled (as fuel) and the lanthanides immobilized in a glass waste form with the Cs/Sr/Ba/Rb stream. For a typical light water reactor fuel, this scheme would result in about 37% of the waste immobilized in a metallic waste form, about 46% immobilized in glass, and about 17% immobilized as captured offgas in another waste form. The compositions of the waste streams will depend on the dissolution conditions, separation efficiencies, fuel burn-up (weakly), and fuel aging prior to reprocessing. The compositions of waste stream suitable for immobilization in metallic waste forms 51 GWd/ MTHM (gigawatt days per metric ton metal) fuel stored 20 years before reprocessing are given in Table 2.19

Table 2 Estimated metallic waste stream compositions for 20-year-old 51 GWd/MTHM fuel

S Fe Se Rb Zr Mo Tc Ru Rh Pd Ag Cd Sn Sb Te Total

UDS (kg/ MTHM)

Recovered TRUEX Recombined Tc with (kg/MTHM) ferrous sulfamate (kg/MTHM) kg/ Mass MTHM %



0.00463 0.114 1.86 4.75 0.0275 1.64 0.0238 0.0684 0.0517 0.0513

0.166 9.876







0.875







0.875

13.5 23.5 0.0394
3.78 0.357
1.85 0.373 1.67 0.0631 0.146 0.141 0.0279 0.589 32.536


23.5 0.0857 0.114 5.64 5.11 1.15 3.49 0.611 2.35 0.115 0.197 0.141 0.0279 0.755 43.288


54.29 0.20 0.26 13.03 11.80 2.66 8.06 1.41 5.44 0.27 0.46 0.33 0.06 1.74

Data from Ebert, W. L. Immobilizing GNEP Wastes in Pyrochemical Process Waste Forms, US Department of Energy Report GNEP-WAST-PMO-MI-DV-2008-000150; Idaho National Laboratory: Idaho Falls, ID, 2008.

512

Metallic Waste Forms

Current waste form research under the FCR&D program is focused on two options for these waste streams: combining the UDS, recovered Tc, and pretreated TRUEX raffinate wastes in a single alloyed waste form or combining the UDS and recovered Tc in a metallic waste form and immobilizing the TRUEX raffinate wastes separately or in glass. The composition of the three combined streams is dominated by Fe from the TRUEX raffinate, which accounts for 54% of the total mass. The next most abundant elements are Zr, Mo, and Ru. Those elements dominate the combined UDS and recovered Tc streams. The amount of ferrous sulfamate assumed in Table 2 is conservatively high, but Fe is expected to still dominate the waste stream composition after the amount of ferrous sulfamate is optimized. Research is in progress to replace ferrous sulfamate with an alternative reductant. However, iron could be added to the TRUEX raffinate to recover the dissolved metals by coprecipitation. The gross composition of the recombined waste streams is similar to the composition of the EBR-II SS
15Zr metallic waste form considering that Mo, Ru, and other components will dissolve in iron to produce a solid solution similar to the steel phase in the SS
15Zr alloy. 5.20.2.3

Hulls and Hardware

The metallic wastes from the electrometallurgical treatment of EBR-II spent sodium-bonded fuel are dominated by the cladding hulls, and the metallic fuel wastes (primarily Zr) are only a small component of the waste stream. In the aqueous reprocessing flow sheets for oxide fuels, the fuel can be mechanically separated from the cladding hulls after voloxidation (if that operation is used) or after fuel dissolution. The hulls can be acid washed to remove adhering TRU contamination and disposed of separately; the hulls will likely retain enough residual contamination and activation products in the bulk metal to require disposal as GTCC waste. Some of the Zircaloy hulls could be used as a source of Zr added to alloy other waste streams. The remaining hulls could be alloyed separately, combined with the separations waste streams, or compacted with the assembly hardware for separate disposal. Preliminary studies17 indicate that Zircaloy hulls can be alloyed with waste components if significant amounts of metal additives are added to lower the processing temperature to practicable levels. However, the amount of added metal needed to lower the processing temperature is by itself adequate to alloy the wastes (i.e., without the hulls). Because there will be about 10 times

more cladding hulls than metallic fuel wastes, compacting the hulls as a separate waste form will probably be more economical than alloying them with the wastes. Neutron activation products generated by components in the various hardware materials used in fuel assemblies will include 14C, 55Fe, 59Ni, 60Co, 63Ni, 93 Zr, 94Nb, and 99Tc. The activity contribution of each can be calculated from the elemental concentration in the metal, the neutron flux, the abundance and thermal neutron capture cross-section of the reacting isotope, and the specific activity of the product radionuclide. The last three are constants and their product provides a measure of the efficiency of each neutron activation reaction; that product is referred to herein as the ‘efficiency factor’ and provides insight into the likely contribution of that radionuclide to the overall radioactivity. Activation of 98Mo leads to the production of 99Tc with an efficiency factor 2  109 Bq barns g
1. Small amounts of Mo (2
3 mass%) are present in Type 316 and other stainless steels, Inconel-718, and in trace amounts in the Zircaloys. Activation of nitrogen, carbon, and oxygen in the fuel and hardware produces 14 C with efficiency 1.27  1012, 1.70  108, and 1.48  108 Bq barns g
1, respectively. Activation of 92Zr in Zircaloy (only) produces 93Zr, which has a very long half-life, but the efficiency factor for 73Zr is only 5.55  108 Bq barns g
1. Activation of Ni and Nb produces 59Ni, 63Ni, and 94Nb with efficiency factors 8.99  1011, 9.16  1013, and 7.96  1011 Bq barns g
1, respectively. All hardware materials contain Ni, but 94Nb will only be produced in assemblies with Inconel components. Generation of 63Ni will dominate the initial radioactivity, but it will decay faster than 59Ni and 94Nb. Both 55Fe and 60Co are short lived and neither will be a significant long-term dose contributor. The efficiency factors are 1.09  1015 and 8.14  1015 Bq barns g
1, respectively. 5.20.2.4

Processing Additives

Metals can be added to the waste streams to tailor the phases comprising the waste form or serve a role in the processing operation. The steel anode baskets used in the electrorefiner are a source of Fe, Cr, and Ni to the waste form. Zirconium is added to provide a neareutectic composition to facilitate processing and generate a consistent phase composition. Some of the metal components used in waste separations and containment can be selected based on how they would impact the waste form. One method for recovering 99 Tc that is being investigated within the FCR&D

Metallic Waste Forms

program is reductive deposition on steel wool. The steel wool and the canister in which it is housed could provide the Fe and other components used to make the waste form. Many components of steels are added to improve the durability. Some of the waste components could serve the same role in the steel-like phases formed in Fe-based alloy waste form. The waste components might also increase the durability of intermetallic phases that form. The approach taken when formulating metallic waste form compositions is to minimize the amounts of additives (and thereby minimizing the mass and volume of waste) while optimizing the processing conditions, waste form durability, and consistency of the phase composition and distribution of radionuclides. The addition of metal to the waste streams allows for better control of the waste form phase composition. Variances in the amounts or compositions of the waste streams from different processed fuels can be compensated for by adjusting the amounts of additives to result in a consistent waste form. The need for a consistent phase composition (because that results in consistent performance) will likely take precedence over the desire to minimize the amount of waste forms produced.

5.20.3 Alloy Compositions Metallic waste forms are formulated to incorporate high concentrations of radionuclides into durable phases at low processing temperatures. The very high-melting temperatures of many metallic waste components require reaction with a lower-melting metal to be dissolved. An example of this is the Fe
15Zr composition used for the EBR-II MWF (see Section 5.20.8.1). The melting temperature of Zr is 1855  C, but reaction with molten Fe allows Zr to be dissolved at temperatures as low as 1337  C at the 84.9Fe
15.1Zr wt% eutectic and at 928  C at the 16.2Fe
83.8Zr wt% eutectic. Eutectics permit waste processing at practical temperatures. Compositions slightly off the eutectic can be used to promote a particular phase assemblage. In the case of the EBR-II MWF, the fact that the actinide elements report to the intermetallic phase demands that an adequate amount of that phase be present in the alloy to avoid forming another phase, which would probably change the waste form performance. Using an alloy composition that is slightly Zr-rich relative to the eutectic will generate a sufficient amount of the intermetallic phase to sequester all the actinides.

513

On the other hand, a sufficient amount of a solid solution phase may be needed to encapsulate the intermetallic phases in a physically robust monolith. Binary and ternary phase diagrams are useful for estimating the solubilities of waste components in a host metal, the intermetallic phases that may form, and the required processing temperature. In complex waste mixtures, competition between solutes (e.g., for interstitial sites in intermetallic phases) will usually lower the solubilities and could affect the phase assemblage and waste form performance. This is most readily determined experimentally. Multiphase alloy waste forms having a simple phase composition are preferred over more complex compositions due to the importance of waste form product consistency. The need to take into account only a few phases will benefit performance calculations, waste form analyses, and waste form qualification. Host matrices that can accommodate a wide range of waste elements within a few phases balance the desire to minimize the amount of waste form produced with the benefits of a simple phase composition that provides enough flexibility to accommodate the anticipated range of waste stream compositions. The austenite and the ferrite solid solution phases form in the SS
15Zr alloy developed for EBR-II wastes provide compositional flexibility for the formation of intermetallic phases having a narrower range of stoichiometries by accommodating the excesses. In a multiphase alloy waste form, the corrosion potential of one phase will necessarily be lower than those of the other phases, and that phase will corrode preferentially and mitigate corrosion of the other phases. The dissolving phase serves as the anode in the oxidation
reduction reactions of the other phases, which act as cathodes. Such a phase is often engineered into systems to serve as a sacrificial anode to protect other metals. The waste form should be designed so that the phases containing radionuclides are not the least durable phases. Although a sacrificial phase could be added as a surface coating similar to zinc-coated iron, it could instead be distributed throughout the bulk of the waste form. This would have the advantage of not being scratched off the waste form by abrasion, but the disadvantage that corrosion of that phase would leave a porous waste form with a high specific surface area. Pits and crevices usually provide microcells that promote corrosion and should be avoided. Steel is galvanized by dipping into a bath of molten zinc at about 455  C. Although steel melts at about 1510  C, an oxide-free steel surface will react

514

Metallic Waste Forms

with molten zinc to form various Fe
Zn intermetallic phases (referred to collectively as the ‘alloy layer’) between the steel and the pure zinc outer surface within minutes in a batch process. It may not be necessary to coat the entire surface, since steel corrosion commonly occurs at point defects at the surface and where microcells develop.20

5.20.4 Processing Methods Requirements for processing facilities will depend on the processing temperature and atmospheric conditions needed to melt the wastes and added metals into an assemblage of durable phases in a physically robust waste form. Small prototype alloys have been made using laboratory-scale furnaces under vacuum and under flowing inert gases. Reducing gases such as Ar/H2 and CO could also be used to control the furnace atmosphere. It may be necessary to contain the furnace in a radioactive hot cell. 5.20.4.1

Alloy Production

Alloys can be produced in resistance or induction furnaces using single-batch or continuous casting methods. Batch melts provide a convenient means of controlling the composition through the relative amounts of wastes and additives placed in the crucible and may be better suited for the required process controls. Continuous casting commonly provides energy and total cost savings, but can result in products having microstructures significantly different than those made with a batch method and a different assemblage of phases. Continuous casting usually generates coarser and more elongated phases, although the microstructure can be modified by annealing. The need to contain the furnace, such as in a hot cell, may dictate which process is used. The EBR-II MWF to be made at INL will be processed using an induction furnace and cast from batch melts as ingots in a crucible. The planned method is summarized here as an example of designing a batch processing method for a specific waste stream. The EBR-II metallic waste stream will be composed of 0.025-m (1-in.) sections of chopped cladding hulls with adherent metallic fuel wastes, primarily Zr, and a small amount of entrained salt. These are removed from the steel anode basket and loaded into a crucible with appropriate amounts of added Zr metal and depleted U. A 2-piece crucible assembly is used to facilitate loading the hulls and

accommodating the 60% volume reduction that occurs upon melting. The crucible assembly includes a crucible funnel that drains into a casting crucible. Both are made of graphite coated with an yttria refractory. The crucible funnel is also lined with steel to protect the refractory-coated graphite funnel during loading. The wastes are loaded into the funnel, which channels the metal into the receiving crucible as it melts. The steel liner melts during processing and drains into the casting crucible with the waste and becomes incorporated into the waste form. Typical processing conditions ramp to a maximum processing temperature of 1600  C over about 2.5 h and hold the mixture at that temperature for about 2 h. A vacuum of about 133 Pa (1 torr) is applied while the temperature increases to about 1300  C to distill the entrained salt. The distilled salt is condensed and recovered outside the furnace; it can be recycled back to the electrorefiner or added to the salt waste stream. The furnace is then backfilled with argon gas to atmospheric pressure (or slightly higher) before the furnace temperature is further increased. Melting begins when the temperature exceeds about 1350  C, but processing at 1600  C provides a 250  C overtemperature that helps dissolve waste metals into the molten iron and form an intimate mixture of iron solid solutions and intermetallic phases as the mixture is allowed to cool. Laboratory tests have shown the importance of locating the lower-melting Fe waste on top to flow over the higher-melting Zr waste. Cooling rates for prototype waste forms have been about 10  C min
1, but fullsized waste forms may cool more slowly. Low cooling rates facilitate solid-state diffusion impacting the formation of the thermodynamically preferred phases, but phase changes predicted to occur at below about 1000  C are not expected to be complete because of slow diffusion. Subsequent heat treatments of prototype waste forms have shown that modifications to the phase composition do occur, such as increasing amounts of intermetallics.21 Ferritic and, to a lesser extent, austenitic steels are susceptible to accelerated intergranular corrosion, primarily due to depletion of Cr near the interface, and this may be exacerbated by heat treatments. The concentrations of some waste components may have a similar effect on the severity of intergranular corrosion in metallic waste forms. The aggressive environment created by molten metal dictates the use of durable materials. Yttrialined graphite crucibles have been used to successfully cast alloys during the development and

Metallic Waste Forms

demonstration phases of electrometallurgical treatment. These are sufficiently durable in the furnace environment, but fragile. Developing better crucible materials is an on-going effort. Waste forms can be cast directly in these crucibles, but it may not be practical to pour or drain the molten metal into separate molds. A steel rod can be inserted into radioactive cast alloys to simplify both removal of the waste form from the furnace and loading it into a disposal canister. 5.20.4.2

Waste Conditioning

Some waste streams contain nonmetallic components that cannot be immobilized in an alloy. Small amounts of these contaminants are expected to be sequestered within slag phases that form on or near the surface of the waste form, but large amounts will probably need to be removed prior to alloying. As described earlier, contaminant chloride salts are removed from EBR-II metals by distillation within the furnace. Other waste streams to be alloyed will likely contain contaminants from the fuel, from separations operations, from storage containers and transfer lines, etc. It may be possible to remove some contaminants within the processing furnace, as is done with entrained EBR-II salt, whereas other contaminants will need to be removed beforehand. It may not be possible or worthwhile to remove low levels of contaminants from the feed. The presence of some nonmetallic materials may even be desirable to help immobilize other contaminants, and waste forms could be formulated to utilize phases to sequester contaminants. Small amounts of slag were formed in surface layers 1
10-mm thick on all SS
15Zr materials made for testing during the development of the EBR-II MWF. These were commonly enriched in C, O, and, in some cases, N. For example, examination of the slag that formed during the production of a SS
15Zr material in a graphite casting mold with scanning electron microscopy (SEM) revealed the presence of ZrC and ZrO2 inclusions within the alloy. The alloy associated with the slag was also enriched in Zr relative to the underlying intermetallic phases. Slags were not detected on Zr
8SS materials. This may be due to the higher solubilities of C, N, and O in Zr(b) than in Fe(g), which are the metal phases that form initially when the Zr
8SS and SS
15Zr materials are produced. These convert to Zr(a) and Fe(a) as the alloy cools. Contaminants are less soluble in these phases, but cannot diffuse out during the phase transition.

515

Slags were formed intentionally on some test materials by oxidizing the steel (in air) before melting. These were used in tests to measure the durability of the slag, which was found to be similar to the durability of the alloy. Whereas slags can simply be removed from industrial metal products, they must remain a part of the waste form, and their role in immobilizing radionuclides must be taken into account in performance models. The presence of a slag layer will not necessarily be deleterious to the waste form and could be used to an advantage. Although metallic radioactive waste forms are likely to be produced with induction or resistance melters, blast furnace slags provide insight useful for tailoring slag phases in waste form alloys. Blast furnace slag consists of nonmetallic contaminants from the ore and from additives (e.g., coke and limestone) that are not incorporated in molten metals but combine to form slag. The slag is less dense and floats on the surface of the molten iron where it can be physically separated from the metal product. Materials are often added to the mixture to facilitate the formation and removal of slag. Slag is readily skimmed from the surface of the molten metal or the metal can be drained from beneath the slag, but slag can adhere strongly to the cooled metal and be very difficult to remove. It may be possible to control the composition of the slag phase to influence both its physical nature and its capacity to retain radionuclides. The role and benefits of a slag layer remain to be studied and evaluated.

5.20.5 Testing Objectives The laboratory tests conducted with radioactive waste forms can be grouped according to three primary objectives to be addressed by testing: (1) waste form performance, (2) waste form consistency, and (3) regulatory acceptance for disposal. These objectives are not independent, since both performance and consistency must be deemed adequate for disposal, etc., but listing information needs in each group can help guide the planning and interpretation of laboratory and field tests conducted. 5.20.5.1 Matrix Degradation and Radionuclide Release The key measure of waste form performance is the rate at which radionuclides are released under the anticipated range of environmental conditions. Tests

516

Metallic Waste Forms

are conducted to determine the modes and rates of waste form degradation and radionuclide release, to develop a mechanistic model that takes into account the effects of key variables in the disposal environment, to quantify the effects of environmental variables on the waste form degradation and radionuclide release kinetics, and to parameterize the model. Several test methods are needed to highlight particular processes. The initial tests are conducted to determine and understand the processes that control the release of radionuclides from the host phase and the degradation of the host phase. The test methods and testing conditions are selected based on conceptual models for degradation and release. For metallic phases, oxidation of the metal is usually the first step followed by dissolution of the metal oxide. The oxidation of some radionuclides, such as Tc, may result in the formation of a species with diffusion-controlled release, such as TcO
4 . Therefore, the intent of the initial tests should be to determine whether there is an effect of the electrochemical potential (Eh) or an imposed potential on the generation and release of soluble species. A separate set of tests is needed to determine the effects of environmental variables on the degradation and release rates. These could include temperature, solution pH, solution or atmospheric Eh, dissolved halide concentrations, groundwater composition, radiolysis products, etc. The sensitivity of the rates to these variables should be measured under conditions spanning the anticipated values of the disposal system to ensure that the model bounds the modeled environment. How the dissolution step is coupled with the oxidation step, if this is indeed the mechanism, should also be determined. Although a fully mechanistic model is desired, it is usually not possible to fully describe the waste form degradation and radionuclide release based on atomistic and theoretical understanding. Instead, empirical models are usually required to describe some aspects of the observed corrosion and degradation behavior. For example, although the effect of Cl
on steel corrosion is well known, quantifying the effect of the Cl
concentration on the measured release of Tc will probably require an empirical model supported by a series of laboratory tests. To the extent possible, the tests used to measure the dependence should be based on an understanding of the effect so as to highlight its impact on the test response. Some degree of coupling in the measured parameter values is unavoidable, such as temperature and pH. Separating the effects of temperature and pH neglects the possible dependence of the activation energy on the pH.

After the model has been developed, it should be validated using tests other than those used to develop and parameterize the model. Validation is intended to show that the model adequately represents the process determined to control waste form degradation and radionuclide release under relevant conditions. Separate confirmation tests may be needed to demonstrate that the mechanism that is being modeled is appropriate for the integrated disposal system. The computational requirements for atomistic and detailed mechanistic models usually exceed the capabilities of repository assessment calculations. Simplified (abstracted) models that represent the effects of environmental conditions and the degradation and release rates are used instead. Confidence in the predictions using simplified models is based on the underlying mechanistic models and, usually, conservative aspects built into the simplified model. Performance assessment calculations are usually based on reactive transport models in which radionuclide release due to waste form degradation is treated as a source term that is coupled with expressions for groundwater flow, diffusion, sorption, and other relevant transport processes (vapor transport, diffusion through thin films of water, etc.) in equations expressing conservation of mass and conservation of energy. These are used to calculate contaminant transport over time under a range of conditions to determine dose to individuals under various scenarios such that the impact of the radionuclide release on the calculated dose depends on many other factors. Depending on the approach taken for risk assessment, a specific performance requirement might not be assigned for radionuclide release or waste form degradation rates. Specific performance requirements were not assigned to waste forms for the Yucca Mountain total system performance assessment, but are specified for lowactivity waste glasses in the performance assessments for the Hanford disposal system, which includes glass and grouted waste forms. 5.20.5.2 Waste Form Consistency Through Process Control Consistent waste form behavior is of primary importance for waste form acceptance. Performance assessment calculations will relate regulatory dose limits to radionuclide release rates, which will be related to waste form composition (and phase composition) by the source term model(s). Confidence in the performance assessment relies on confidence in the waste form composition, which must be controlled during

Metallic Waste Forms

production. Important aspects of a metallic waste form composition might include the domain sizes of the component phases, the degree of mixing, the fraction of slag phases, etc. Process controls might include the amounts and chemical compositions of the feed materials, both the wastes and additives, and particular processing conditions, including the mixing and loading of the feed materials, the heating schedule (temperature, time at temperature, cooling rate, heat treatment, etc.), furnace atmosphere, and crucible material. The operating range of each processing variable that results in consistent waste products must be determined based on its direct impact on the component phases that form and the distributions of radionuclides (and the indirect impact on performance). The relationships that must be established by testing and modeling are summarized in eqn [1]. Production control $ Phase composition control $ Matrix degradation rate control $ Radionuclide release control $ Acceptable performance

½1

Ultimately, the processing limits must be related to the performance of the disposal system, usually through the release rates of radionuclides contributing to the regulated dose. In the performance assessment calculations, transport through natural and engineered barriers can decrease the dose contributions of some radionuclides at the point of interest so that the radionuclide that is released the fastest or to the greatest extent may not have the greatest impact. For example, a low solubility limit and strong adsorption to mineral phases retard the transport of plutonium isotopes and their impact on total dose at a regulated point. The relative impact of Tc may be greater regardless of the relative amounts of Pu and Tc released from the waste form due to the high solubility and negligible transport retardation of the pertechnetate ion in an oxidizing environment. 5.20.5.3 Waste Form Acceptance and Regulatory Requirements In the United States, regulatory requirements include compliance with Nuclear Waste Policy Act, as amended, and dose limits specified by the US Environmental Protection Agency for disposal systems. Disposed waste is also subject to hazardous waste regulations, including the Resource Conservation

517

and Recovery Act (RCRA). Additional safeguard and security requirements may be added in the future. Although waste form acceptance ultimately occurs through a US Nuclear Regulatory Commission license for a facility to receive and dispose waste, separate acceptance criteria may be established between various DOE agencies responsible for producing, characterizing, packaging, licensing, and disposing the waste forms. In general, the acceptance criteria are intended to provide assurance that the chemical, physical, and radiological performances of the waste form are such that the disposal system will meet regulatory requirements over its service life as predicted by performance assessment calculations. These are obviously closely related to the degradation behavior of the waste form (and radionuclide release), the radionuclide inventory, and the disposal environment. For example, the DOE Office of Civilian Radioactive Waste Management has provided the Waste Acceptance System-Requirements Document22 to address the acceptance of commercial spent nuclear fuel, DOE-owned spent nuclear fuel, naval spent nuclear fuel, and high-level radioactive waste for disposal in the planned Yucca Mountain repository. Requirements are described in that document for the waste form materials, such as phase composition, radionuclide content, product consistency, and chemical durability; for the canistered waste forms, such as criticality and thermal outputs; and for the canisters themselves, such as material, dimensions, maximum weight, labeling, handling fixtures, required levels of surface decontamination, etc. Although most of these requirements are not regulatory requirements, they provide insight into additional issues that should be considered when designing a testing program.

5.20.6 Modeling Modeling activities should be integrated with testing activities at the earliest stages of developing test plans. The initial model will probably be a conceptual model based on the identity of the waste stream and matrix material, the expected material alteration modes based on literature surveys, anticipated environmental conditions for the disposal system, and insights from the behaviors of available analog materials. Testing and modeling should evolve iteratively as test results are collected and the model is modified. When the model progresses to the point of predicting degradation behavior, validation tests should be conducted for direct comparisons with model

518

Metallic Waste Forms

predictions. Other tests are needed to confirm that the modeled process is operational under disposal conditions. Eventually, the validated and confirmed model is adapted for use in performance assessment calculations. An important aspect of the performance model is correlation of the phase composition to performance. For homogeneous glass waste forms, the chemical composition can be directly related to performance through the glass dissolution rate. For multiphase alloyed waste forms, performance will probably depend on the assemblage of alloyed phases that form and the distribution of radionuclides. The more complicated corrosion mechanism of alloys is an added challenge to modeling and to demonstrating process control of waste form performance. 5.20.6.1 Matrix Corrosion and Radionuclide Release Mechanisms A mathematical model is required to represent the release of radionuclides from the waste form to become available for transport through the disposal system as a function of time. Release may occur by diffusion of radionuclides out of the waste form or by degradation of the waste form matrix and phase(s) that contain the radionuclide. A mechanistic model that takes into account the key processes in waste form degradation and radionuclide release will provide maximum confidence to long-term predictions and performance assessments. Rate equations have been developed for various release models, including constant dissolution rate models, concentration-dependent models, reaction affinity models, and diffusion-controlled release models. Empirical rate expressions have been developed based on detailed measurements. An approach for developing a source term model for waste forms is provided in ASTM Standard C 1174 Prediction of the Long-Term Behavior of Materials, Including Waste Forms, used in Engineered Barrier Systems (EBS) for Geological Disposal of High-Level Radioactive Waste.4 This standard provides a philosophy and logic that can be followed to develop a performance model by integrating testing and modeling activities. It stresses the importance of a clear definition of the problem, characterizing the disposal environment, maintaining interfaces between testing and modeling to evolve the model based on test results and design tests based on model predictions, using analog materials and systems, and both validating the fidelity of the model to the mechanism and confirming the appropriateness of the model for the disposal system.

A variety of test methods is needed to understand the degradation mechanism of a waste form material and how environmental and engineering factors will affect the releases of important radionuclides, and to develop and parameterize a performance model. Many of the information needs related to waste form processing and waste form acceptance require an understanding of the corrosion mechanism. It is usually wise to address aspects of the mechanistic, environmental, and engineering system early in a testing program to gain confidence that all the important factors are being taken into account in the model. As the mechanistic understanding develops, other aspects of waste form behavior that must be investigated usually become apparent. The general information needs for developing a performance model are summarized as follows. Identify the radionuclide release mechanism: The objective is to determine whether each radionuclide is released stoichiometrically as the matrix degrades or whether separate models are required. In most cases, these tests will serve to confirm the release mode based on an understanding of the matrix material and how the radionuclide is incorporated. For the various waste forms being developed for reprocessing wastes, the release of radionuclides may be controlled by diffusion (leaching), congruent dissolution of the matrix, or degradation of the matrix to expose the sequestering phase that contains the radionuclide, which could then dissolve, be leached of radionuclides, be released as a colloid, etc. Release may require prior oxidation of the sequestering phase and/or the radionuclide itself, as is expected for Tc in a metallic waste form. Characterize the matrix degradation mechanism: It is expected that the release of radionuclides will be affected physically or chemically by the degradation of the waste form matrix. Degradation of the matrix may be required before a radionuclide can be released; it may simply need to be physically or chemically altered rather than dissolved. For some of the multiphase waste forms considered for reprocessing wastes, dissolution of an encapsulating phase may be required before water can react with the phase bearing the radionuclide. Radionuclide release will be affected by both the radionuclide-bearing phase and the host matrix. These could degrade by different mechanisms, for example, in the case of an oxide phase encapsulated in a metal matrix. Characterize the effects of environmental variables: The release rates of radionuclides and the degradation rates of matrix materials will probably be affected

Metallic Waste Forms

directly by environmental variables such as temperature and the groundwater chemistry (pH, Eh, dissolved component concentrations, etc.) and indirectly by sorption onto colloids, minerals, and engineering materials in the disposal system. The groundwater chemistry may change significantly due to interactions with engineering materials and waste form degradation. Tests generally need to be conducted under conditions exceeding the expected range of values for environmental variables to initially determine and model the dependency and then to verify that the mechanism remains operative over the anticipated range of disposal conditions. A variable that affects the degradation of a waste form but is not tracked in performance assessment calculations can be captured using another variable that is tracked or through the uncertainty ranges of coefficients in the rate equation. Develop a degradation model that accounts for environmental variables: If possible, the dependence of the degradation rate or radionuclide release on each environmental variable that affects the rate should be modeled mathematically according to the mechanism. It may be beneficial (or necessary) to combine the effects of two or more variables in a bounding semiempirical model rather than modeling each explicitly. This may be necessary if factors cannot be distinguished in tests or practically differentiated by modeling. Residuals from fitting the test data using separate dependencies will include the effects of cross terms, and the interdependence will be taken into account through both the regressed values and the uncertainties. Consider waste package interactions: The effects of the waste package materials on the groundwater chemistry, transport of released radionuclides, etc. must be considered in performance assessment calculations. Although these are taken into account separately from waste form degradation in performance assessments, the interactions that affect the groundwater prior to contacting the waste form should be considered when defining the range of environmental variables to be represented in tests. Standardized test methods routinely call for control tests and/or blank tests in which interactions of test solutions, dissolved components, and test specimens with vessels and supports are measured and taken into account in analyses of the test data. Interactions between metallic waste forms, waste packages, and engineering materials need to be evaluated, such as Galvanic coupling, and taken into account in models. Alloyed waste forms should not serve as sacrificial anodes for other metallic components in the disposal system.

519

Measure model coefficients: The same test methods used to characterize the effects of environmental variables often provide data that can be used to determine the values of model coefficients. To measure model coefficient values, the effects of variables other than the one being quantified must be held constant or reliably taken into account when evaluating the results. Typically, this involves conducting a series of tests in which the values of all variables are held constant, except the variable of interest. Such tests can be conducted at constant temperature, in pH-buffered solutions, with spiked leachant solutions, in controlled atmospheres, etc. It is important that the measured coefficients are appropriate over the anticipated range of conditions in the disposal system, and it is possible that different coefficient values could be required for specific ranges if the degradation mechanism changes. Model validation: The intent of model validation is to demonstrate that the mathematical model adequately represents the particular process of interest, be it oxidation, diffusion, radionuclide release, or matrix degradation. A model is considered validated if it predicts responses consistent with the results of separate tests that were not used for model development or parameterization with an acceptable range. The validation test should focus on the specific process being modeled and the measured response should be sensitive to that process. For example, if diffusion-controlled release from a solid is being modeled, the validation test should be conducted under conditions that avoid saturation in the contacting fluid. The effects of other processes are considered in model confirmation tests. Model confirmation: The intent of model confirmation is to demonstrate that the process represented in the degradation model controls the waste form behavior under the range of conditions and other interactions expected in the disposal system. Confirmation tests generally couple the modeled process with other system processes that could affect waste form behavior. Model confirmation cannot be completed until the disposal system is identified and characterized. However, aspects of potential disposal systems likely to affect waste form behavior should be taken into account when developing the model, including the groundwater chemistry and flow characteristics, interactions with engineering materials and the host geology, radiation fields, etc. In most cases, the laboratory tests used to determine waste form corrosion behavior must subject the test specimen to more aggressive conditions than those

520

Metallic Waste Forms

expected in a disposal system to generate a measurable response. For example, the dissolution of a waste form due to contact by groundwater may have been modeled using the results of static immersion tests, whereas the disposal system is hydrologically unsaturated. A possible model confirmation test would be to expose the waste form to intermittent contact by dripping water to see how well the model represents degradation under these conditions. Other test methods that could be used include vapor hydration tests to simulate contact by vapor and unsaturated flow tests to simulate intermittent contact with flowing water. The responses of model confirmation tests will usually be too complicated for use in determining mechanisms for specific processes. This is because they will be affected by several processes acting simultaneously. They provide data that are responses to several interacting processes that must be taken into account in the model either directly through model variables or the ranges of model coefficient values. Nevertheless, such tests can provide confidence in the testing process that the correct process is being modeled early for a particular environment. 5.20.6.2

Performance in Disposal System

As with the waste form degradation and radionuclide release model, performance assessment models based on a mechanistic understanding of processes affecting the spread of radionuclides through engineered and natural barriers would provide the highest confidence. This is usually not possible because of the complex pathways and interactions affecting transport. Instead, transport models use effective distribution coefficients to model complex interactions with various minerals in the soil, including reversible and irreversible sorption to fixed and mobile substrates, effective flow rates for advection through fractures and along grain boundaries, diffusion and dispersion, etc. Small variations in temperature and water content spatially and temporally are neglected, but large variations can be modeled by separating the regions of interest into several cells that can be assigned different characteristic values.23 The solute transport models that have been developed for environmental impact analyses generally include equations to account for water infiltration rates through the system, source term release rates for important nuclides, and solute transport rates through the geologic media. Based on the conservation of mass for a transported radionuclide, the

dependence of the change in concentration C with time on these factors is written as @C @C @2C X Qn þ r ¼
n þ D 2 þ @t @x @x n

½2

Reactive transport models usually only consider transport in the direction of flow (x in this example), except for dispersive effects.23,24 The first term on the right-hand side represents advective transport, where n is the groundwater velocity, and the second term represents dispersive transport. The dispersion coefficient, D, is sometimes written as the sum of an effective diffusion coefficient for the geologic medium and Pthe coefficient of mechanical dispersion. The term n Qn represents the sum of the effects from each of the n processes that affect the solution concentration C, which may include waste form degradation, leading to radionuclide release (the source term), sorption, colloidal and secondary phase formation, and several other processes depending on the species of interest and the disposal system. The value of C is usually constrained in models by a speciesspecific solubility limit. The fourth term in eqn [2], r, takes into account the in-growth and decay of the radionuclide of interest. Most transport calculations for contamination assessments include transport through hydrologically unsaturated (vadose) zones. A dimensionless volumetric moisture content y can be included to represent the fraction of the total pore volume available in the medium that is occupied by water, and C can be replaced in eqn [2] by the product Cy. This simply scales the solution volume by the moisture content. The radionuclide release model provides the source term Q used in the transport model. Depending on the form of the rate equation for radionuclide release from a metallic waste form, its integration into a reactive transport expression may require additional terms to represent electrochemical processes. 5.20.6.3

Processing Control

The ability to maintain acceptable waste form performance (the retention of radionuclides) using processing controls must be demonstrated by establishing a correlation between the assemblage and compositions of alloy phases in the waste form and waste form performance measured in laboratory tests. The correlation should have a mechanistic basis, and, ideally, the dependence of the performance on the phase composition should be quantified and

Metallic Waste Forms

parameterized in the degradation model. This may not be possible because waste form performance cannot be measured with a single test method. Instead, a test method having a response that is correlated with the performance of the anticipated range of waste form compositions can be used to represent the relative performance. In the case of high-level waste glasses, the Product Consistency Test (ASTM method C 1285)4 is used to correlate performance with composition; the response in that test is used by DOE to identify acceptable waste glasses.22 The use of an analogous approach for metallic waste forms will require the identification of a test method and development of a correlation model that can be related to the performance model. These will then allow processing limits to be established for producing acceptable waste forms, such as waste loadings, additives, and processing temperature.

5.20.7 Test Methods A variety of test methods is required to understand the corrosion behavior of metallic waste forms, measure the kinetics of various processes, and accelerate corrosion to simulate aspects of long-term degradation. The degradation rates of well-designed waste forms should be difficult to measure under anticipated service conditions if they effectively retain radionuclides. Test conditions that differ significantly from the service conditions are often required to produce a measurable test response or measure the effect of a variable. Relating the corrosion occurring under extreme test conditions to service conditions usually requires an understanding of the corrosion mechanism, even though an understanding of the mechanism is often the objective of the tests. Testing and modeling to determine the corrosion mechanism should be coordinated and iterative activities. The selection of initial test methods should be based on a conceptual model and hypothesized mechanism, both of which should be developed from literature reviews and considerations of possible analog materials. In the case of metal waste forms, experience with steel corrosion provides a logical starting point. The initial tests should be conducted to evaluate the hypothesis and conceptual model, and both the conceptual model and the suite of tests should evolve as new data are collected. The degradation of a metal waste form and subsequent release of radionuclides is hypothesized to involve oxidation reactions to convert the metal

521

components to oxides and hydrolysis reactions to dissolve the oxides. Diffusion of oxygen through the evolving oxide layer is another important process required to continue the corrosion. It is likely that the oxidation, dissolution, and diffusion processes may each control the radionuclide release rate at some point and under some reaction conditions. Components might be released as quickly as the metal surface is oxidized early in the reaction but slow as an oxide layer develops and slowly dissolves. The oxide layer will likely serve as a diffusion barrier to oxygen, such that diffusion rate controls the continued corrosion rate. Passivation of the surface (of each component phase) may occur to prevent further corrosion, or the layer could become thick enough that the mismatch between the metal surface and the crystalline oxide layer causes the layer to spall off the metal surface and corrosion resumes. To characterize this hypothetical corrosion mechanism, separate test methods would be needed to highlight the oxidation reactions, the dissolution reactions, and the long-term stability of a passivating layer. The majority of test methods and techniques developed by industry for metal corrosion address shortterm behavior (surface oxidation, pitting, passivation, etc.) related to the service objectives, whereas the release of radionuclides to the solution over very long corrosion times is of primary interest for alloys used as waste forms. The oxidation of metal components will have no effect on waste form performance until transportable radionuclides are released to solution, either by dissolution or colloid formation. Many of the tests used to study the degradation of glass, minerals, and ceramics address the effects of the solution chemistry on the dissolution reaction. These materials dissolve by an affinity-controlled mechanism that is sensitive to the accumulation of components in the solution. The effect of the solution chemistry on the degradation of metals is not expected to be as significant because the most abundant components (e.g., Fe and Zr) are sparingly soluble in most solutions and will have a very limited dissolved concentration range. Nevertheless, similar test methods are needed to study dissolution of the oxide layers resulting in radionuclide release and diffusion (e.g., of oxygen) through the layers. Passivity is usually investigated by studying the nature of the films that form and the kinetics of the electrochemical processes. Corrosion tests in which specimens are immersed in solution or contacted by steam can be used to measure the release of components to solution and generate oxide layers for detailed examination. Electrochemical tests can be used to

522

Metallic Waste Forms

measure the kinetics of processes taking place on clean surfaces. Several useful electrochemical and corrosion test methods are discussed in the following sections. 5.20.7.1

Electrochemical Test Methods

Metal corrosion by oxidation reactions is conveniently and efficiently measured with traditional electrochemical techniques.25,26 These include construction of potentiostatic curves (anodic and cathodic charge curves), studying potential shift curves and double-layer capacity, electrochemical impedance spectroscopy, etc. Measurements can be conducted at a range of temperatures and in various solutions to determine and quantify the effects of many environmental parameters on the oxidation rate, such as pH and the presence of various solute species, and combined effects such as the Cl
concentration and critical pitting temperature for various alloy formulations. Electrochemical techniques used to determine corrosion rates utilize the corrosion current to calculate the corrosion current density, corrosion intensity, and corrosion penetration rate. The current that is passed upon application of an external potential can be related to the mass of material reacted using Faraday’s law and then expressed as a rate using the reaction time and specimen area. This gives the reaction rate in terms of the current density. The polarization resistance can be determined from a plot of the applied potential against the current and can be related to the current density by the anodic and cathodic Tafel constants for the material. The corrosion rate of the material can be calculated from the current density and the density and atomic weight of the metal (or the equivalent weight of an alloy). This gives the rate for uniform corrosion (oxidation) of a bare metal or alloy surface. If the corrosion is localized rather than uniform, the penetration rates at particular sites can be orders of magnitude higher than the average value. Some of the reacted specimens should be examined to look for areas of localized corrosion (pitting). Because the data analysis techniques require extrapolating measured curves, the accuracy of the experimental measurements is crucial to the calculated rates and their sensitivity to the compositions of the alloy and the solution. Standardized procedures should be followed whenever possible. The ASTM procedure Standard Test Method for Conducting Potentiodynamic Polarization Resistance Measurements (G 59) provides a standardized method for measuring corrosion potentials and potentiodynamic polarization resistance for determining the

general corrosion rates of metals.27 A small potential scan near the corrosion potential of the metal is applied to a sample and the resulting currents are measured. Tests conditions can be varied to study the effects of temperature, pH, dissolved O2, solution composition (e.g., oxidizers), etc. Potentiodynamic measurements (in general) provide a convenient comparison of the oxidation behaviors of various metals and alloys, and a rapid and economical approach for studying the oxidation step of the corrosion mechanism. Interpretation of the measurements requires assumptions regarding Tafel slopes and equivalent weights of alloys. A galvanic cell is formed when two (or more) metals or alloys having different polarization resistances are coupled and in contact with an electrolyte solution, including the separate phases in a multiphase alloy. In such a system, the least stable material will act as the anode material and corrode in preference to the other materials, which will act as cathodes. Galvanic couples can also form between waste form phases and engineering materials such as waste canisters. Obviously, radionuclide-containing alloy phases should not serve as the anode of a galvanic couple; rather, such phases should be protected by phases without hazardous components. Galvanic couples with sacrificial anode materials are used extensively to protect other metals, such as steels galvanized with zinc and iron anodes in residential water heaters. It may be possible to add such a phase as part of the waste form to protect radionuclidebearing phases, but it is crucial that radionuclidebearing phases do not act on the anode. The ASTM procedure Standard Guide for Conducting and Evaluating Galvanic Corrosion Tests in Electrolytes (G 71) addresses conducting and evaluating galvanic corrosion tests.27 The standard addresses selection of materials and preparation of test specimens, selection of test conditions (electrolyte, materials coupling, atmosphere, duration, etc.), and the evaluation of test results. Galvanic corrosion measurements provide insights into likely materials interactions, including interactions between different phases within an alloy and between an alloyed waste form and the waste canister material. Microscopic examination of the specimen before and after analysis can provide confirmatory evidence of the relative durabilities of the components phases, although test results are sensitive to specimen preparation and electrical contact. Several other standardized test methods, practices, and guides are available for conducting and

Metallic Waste Forms

evaluating electrochemical measurements, and consensus standards should be followed when available and appropriate. Besides providing the benefits of experience, standardized methods often provide reference test results for comparisons and quantify testing bias. For example, the ASTM standards G 1, G 3, G 5, G 15, G 50, G 60, and G 102 provide helpful insights into preparing specimens, conducting electrochemical tests and measurements, and interpreting test results.27 5.20.7.2

Corrosion Test Methods

Corrosion tests are used to study the release of waste form components into solution during waste form degradation. Many corrosion test methods have been developed to study the dissolution of oxide materials, such as glasses, that do not require a preceding oxidation step. These methods can be applied to a metal waste form, but the results may be complicated by the coupled oxidation reactions and depend on whether the metal test specimen is grounded or insulated. The ASTM procedure Standard Practice for Laboratory Immersion Corrosion Testing of Metals (G 31) provides a standardized method for directly measuring mass loss in laboratory immersion tests to determine the general corrosion rates of metals.27 Conditions that may be controlled include test duration, temperature, pH, oxygen concentration, solution flow, solution composition, and solution concentration. This procedure addresses factors important to sample preparation, test conditions, methods of cleaning specimens, and cautions regarding the interpretation of test results that are relevant to any test method. The relative contributions of generalized corrosion cannot be distinguished from localized or intergranular corrosion based on the mass loss alone. Although the mass can increase due to the formation of oxide surface layers or decrease as the layers dissolve, the corrosion products are removed from reacted specimens prior to weighing in the ASTM G 31 procedure. Simply tracking the mass change in test specimens will not be sufficient to understand the corrosion mechanism, but other analyses can be conducted to gain insight into the corrosion behavior, such as microscopic examinations of reacted specimens to detect pitting and identify corrosion products. The ASTM Test Method for Static Leaching of Monolithic Waste Forms for Disposal of Radioactive Waste (C 1220) provides a standardized method for measuring the mass of a monolithic test specimen

523

that has dissolved into solution under static conditions.4 The test results are sensitive to the corrosion behavior of the waste form with little feedback from dissolved components occurring in short-term tests. The geometric surface area of the specimen can be measured to allow accurate calculation of the specific release rate of soluble components. The test is easy to run, can be conducted under a wide range of conditions, provides sufficient solution volumes for analysis, and is economical. Only small volumes of waste solution are generated. Short-term tests can be used to measure the effects of temperature, pH, and components in the leachant on the release rates. Responses may become affected by the chemical affinity of the solution in long-term tests. Aspects of ASTM G 31 regarding specimen preparation and cleaning should be followed when conducting ASTM C 1220 tests with metallic specimens. Two important variations of ASTM C 1220 involve interrupting the test intermittently to recover solution for analysis: the partial replacement test and the solution exchange test. In the partial replacement test, a small portion of the solution is removed for analysis and replaced with fresh leachant solution. This allows the extent of reaction to be tracked in an otherwise static test, but dilutes the remaining solution slightly. The impact of diluting the solution is usually negligible in tests with metallic specimens because dissolved components do not affect the continued corrosion to a significant degree. The entire volume of solution is replaced with fresh leachant in a solution exchange test. Solution exchange can have a significant effect on metal corrosion due to opening the vessel to refresh the air and oxygen content. Replacing the entire solution volume will also replace all the dissolved oxygen (assuming that the added solution is air saturated), whereas partial placement will only replace a small fraction of the dissolved oxygen. This will be important for oxidation reactions forming oxide layers. In addition, solution exchange will re-establish the initial pH, but partial replacement will allow the pH to drift as the specimen corrodes. ASTM Standard Test Method for Diffusive Releases from Solidified Waste and a Computer Program to Model Diffusive, Fractional Leaching from Cylindrical Waste Forms (C 1308) provides a standardized method for solution replacement tests that should be consulted prior to running a modified ASTM C 1220 test.4 The test conditions specified in the standard are better suited to rapid release from grouted waste forms than slow release expected from

524

Metallic Waste Forms

metallic waste forms, although long-term tests at elevated temperatures could lead to measurable test responses. The vapor hydration test (VHT) is a static test in which a monolithic specimen is suspended in a sealed vessel with a small amount of water.28 When heated (typically in the range 125
200  C), the vapor phase becomes saturated and a thin film of water condenses on the specimen. The specimen temperature rises more slowly than that of the test vessel and provides the coolest surface when the test is initiated. As the specimen corrodes, the condensed solution usually becomes hygroscopic, so the thin film of water remains on the test specimen even after thermal equilibrium. In the standard VHT, the amount of water added to the vessel is carefully controlled so that a condensed layer forms on the specimen but no liquid water remains in the vessel to establish a reflux cycle. Alteration of the reacted sample can be analyzed and thickness of the altered surface layer measured on a cross-sectioned specimen. This method has been used to promote the formation of oxide layers on metallic specimens for microscopic investigations. Neither the actual acceleration factor nor the chemistry and the volume of the solution contacting the specimen during reaction are known in VHTs. The test response is sensitive to the volume of water that condenses on the specimen, which cannot be controlled accurately or measured during the test. The extent of corrosion can be estimated based on the amounts of alteration phases that formed. However, the VHT is not well suited for quantifying corrosion rates, because the precision of the test (i.e., the amounts of alteration phases formed) is usually poor. The consistency of the corrosion mechanisms at the relatively high temperatures at which most VHTs are performed and at temperatures relevant to disposal environments must be established to extrapolate VHT results to service conditions. In a modification of the VHT, enough water is added to promote refluxing during the test to flush components released from the specimen to the bottom of the vessel. The test is interrupted intermittently to recover and analyze the solution and track the release of waste form constituents. This modification is similar to a Soxhlet test,29 except that water vapor condenses on the sample itself rather than in a separate condenser and maintains an adhering layer of water on the sample rather than filling a sample boat. The modified VHT method has been used to

measure the release of U from metallic specimens (see Section V.C.3 in Ebert).18 5.20.7.3

Service Condition Test Methods

The electrochemical and corrosion tests discussed earlier are usually conducted to understand the corrosion mechanism and to develop and parameterize a degradation model. The test conditions that are used are typically very different than the environment that will exist in a disposal system (referred to as the ‘service conditions’). Most laboratory tests are conducted by contacting a small metal surface area with a large volume of water to highlight alteration processes or at elevated temperatures to promote corrosion and facilitate measurements. The service conditions of most disposal sites will be restricted to small volumes of water and thin films contacting small waste form surface areas. Metallic waste forms will likely not be thermally hot due to the low activities expected in amenable waste streams (e.g., those given in Table 2). The effects of limited groundwater volumes on waste form degradation have been studied in laboratory tests in which groundwater is periodically dripped onto crushed material or a monolithic specimen and in tests where a mixture of water and air is passed through crushed material. The unsaturated test was developed to simulate small volumes of groundwater dripping through a breached waste container onto spent fuel or HLW glass.30 The water collects on and reacts with the specimen, and occasionally drips off the specimen into the bottom of the test vessel. The accumulated solution is collected and analyzed periodically to track the release of radionuclides. The pressurized unsaturated flow test was developed to simulate disposal in a hydrologically unsaturated environment.31,32 A mixture of water and air is pumped through a column of crushed materials that may include soil, rock, waste form, and backfill materials to simulate a disposal system. Water interacts with these materials as it flows through the column, and the effluent chemistry is tracked during the test. Changes in the solids due to dissolution and precipitation of secondary phases can be measured at the end of the test. Column tests are commonly used to simulate groundwater interactions with soils and minerals to measure dissolution rates, distribution coefficients of contaminants, transport parameters, etc.33,34 Column tests can be conducted in the laboratory to control

Metallic Waste Forms

test conditions, such as temperature, water content, and flow rates, and to simplify sample collections. Column tests are also conducted in the field using lysimeters.35
38 Lysimeters are simply columns placed in the ground and filled with soil (sand, rock, etc.) of interest. Containers and probes can be placed at the bottom and various intermediate locations to analyze and collect groundwater that passes through the column. The top of the lysimeter can be left open to allow meteoric water (rain and snow) to enter and flow through the lysimeter or intermittently covered to regulate the amount of water entering the lysimeter. Waste form corrosion can be coupled with transport by emplacing one or more waste form specimens within the soil. The waste form specimens and adjacent soil can be recovered for analysis after testing. Finally, materials can be simply buried in the soil and allowed to corrode naturally with no attempt to control, collect, or analyze the groundwater. Examples of this are underground corrosion studies conducted by US National Institute of Standards and Technology (NIST), formerly the National Bureau of Standards, such as the corrosion of stainless steels buried in various soils in the early 1970s.20 These studies were focused on the corrosion effects on the specimens, such as etching, blistering, cracking, and pitting, which are important for steels used as canisters (see Sullivan39), but less so for steels used to immobilize radionuclides that are components of the alloy or solid solution.

5.20.8 Tests with INL MWF Electrometallurgical treatment is being used to condition used sodium-bonded fuel from the EBR-II reactor for disposal as high-level radioactive waste.9,12,40 Waste forms have been developed for the salt and metallic wastes resulting from the electrorefining process. The development of the waste form for metallic wastes is summarized in the following sections as an example of formulating and testing an alloy to address performance and processing issues, and development of a degradation model. A few key results are provided here; a more extensive summary of the test results and full bibliography are provided in Appendix B of Ebert.18 Detailed analyses of the alloy structure and phase composition provide a crucial link between the consistency of the metallic waste form maintained by process controls and the performance of the waste form in the disposal site as

525

modeled in assessment calculations and utilized for waste acceptance. Of course, the waste form performance reflects the combined behavior and interactions between all the component phases, which are not modeled individually for the MWF. 5.20.8.1 Formulation and Phase Compositions Two alloy compositions were identified as potential waste forms to immobilize waste streams that were dominated by stainless steel or Zircaloy cladding hulls. These target compositions were selected to maintain consistency in the phase compositions for the fairly narrow range of waste stream compositions and simplify the waste form qualification process. The binary Fe
Zr phase diagram41 was used as a guide for waste forms made by alloying stainless steel and Zircaloy claddings with the metallic wastes.17 An eutectic exists at about 84.9 mass% Fe and 15.1 mass% Zr with a melting temperature of 1337  C, whereat Zr6Fe23 and g-Fe are predicted to form. A maximum processing temperature of 1600  C was desired to allow use resistance or inductively heated furnaces. Alloys having various proportions of steel and Zr were made for analysis, and the microstructures, likely distribution of radionuclides, and durabilities of these materials were studied. Steel-dominated compositions from SS
5Zr to SS
42Zr and zirconiumdominated compositions from Zr
50SS to Zr
8SS were studied. The analog of the Fe
15Zr alloy was selected as target composition for waste streams dominated by steel cladding hulls based on phase composition, chemical and physical durability, and processing considerations. The analog of the Zr
8Fe alloy was identified as a possible target composition for waste streams dominated by Zircaloy cladding hulls. Both mixtures result in alloys having similar amounts of metallic and intermetallic phases but with one important difference: the metallic waste components in the fuel (Zr, Mo, Ru, Rh, Pd, Tc, etc.) and other metals in the cladding (e.g., Cr and Ni) form solid solutions in iron but are essentially insoluble in zirconium. Whereas the Fe-dominated alloys include an iron solid solution phase that serves to encapsulate the intermetallic phases in a physically durable monolithic waste form, the Zr-dominated alloys comprise loosely aggregated intermetallic phases without a binding metal phase. Because the initial focus was on steel-clad fuels, only the SS
15Zr mixed alloy

526

Metallic Waste Forms

was developed as a waste form for EBR-II wastes. However, these initial results suggest that a Zrdominated mixed alloy that is not well consolidated will not be useful as a waste form. 5.20.8.2

Steel

Radionuclide Distribution

The likely distributions of many elements in the component phases of the mixed alloy can be inferred from well-established binary phase diagrams, but the distribution of the radionuclides from the fuel in these phases must be measured because they are less studied and more important to performance than other waste components. Several SS
15Zr materials were made to study the dispositions of Tc and other radionuclides in the component alloy phases. The behavior of these radionuclides during corrosion was also measured for some of these materials. The results of tests and analyses with the materials made during development of EBR-II MWF alloys (both with and without added radionuclides) provide valuable insight into potential waste forms for the other waste streams that could be generated in the aqueous and pyrochemical processes. Samples of SS
15Zr were prepared with various amounts of added U, Np, Pd, Ru, and Tc. The capacity of the MWF to incorporate U is of interest, whether U is present as a contaminant or is added intentionally to down-blend 235U to below 20 mass %42 to meet criticality limits. The backscattered electron SEM image of a SS
15Zr
11U mixed alloy in Figure 1 shows U to be concentrated nonuniformly in the intermetallic phase.43 The darkest phase is the iron solid solution, the gray phase is the intermetallic phase without U, and the white phase is the intermetallic with U. Note the intimate mixing of the iron and intermetallic phases on a 100-mm scale. The polytypes detected in MWF materials during analyses are noted later, although the effects on performance are not known. Intermetallic phases with the general formula AB2 are commonly referred to as ‘Laves phases’; the A and B sites are usually occupied by several different elements in the Laves phases formed in the MWF. For example, the Fe sites in ZrFe2 are occupied by various amounts of Fe, Cr, and Ni in the MWF materials that have been studied by McDeavitt and coworkers.17 Some of the intermetallic phases can exist in cubic, hexagonal, and dihexagonal structures depending on the packing arrangement of identical (or very similar) layers. These related phases are referred to as polytypes.

U-rich intermetallic

U-poor intermetallic

200 µm

Figure 1 Backscattered electron scanning electron microscopy image of SS
15Zr
11U showing iron solution phase (black), U-poor intermetallic phase (gray), and U-rich intermetallic phase (white). Modified from Figure 4 of Keiser, D. D. Jr.; Abraham, D. P.; Sinkler, W.; Richardson, J. W. Jr.; McDeavitt, S. M. J. Nucl. Mater. 2000, 279, 234
244, used with permission. License number 2287090615691.

The compositions of the (mostly) ferrite and intermetallic phases of three MWF materials are summarized in Table 3.43 The value reported for each element has an absolute uncertainty of 2 at.%. Actinides were only detected in the ZrFe2-type intermetallic phase of SS
15Zr. The U-rich and U-poor designations indicate analyses of the brighter and the darker regions of the intermetallic phases, respectively. (The U-rich areas are brighter due to the higher electron density of U.) Differences in the relative concentrations of Zr and U in the U-rich and U-poor zones and the fairly similar Fe concentrations indicate that U substitutes at Zr sites rather than Fe sites in the intermetallic phase. Comparison of the SS
15Zr
5U alloy with the SS
15Zr alloy (see Table 4 in Keiser et al.18) showed relative increases of 5, 5, and 15 vol.% in the amounts of ferrite, ZrFe2 (cubic), and Zr6Fe23, and decreases of 4 and 20 vol.% in the amounts of austenite and ZrFe2 (dihexagonal). Although Zr6Fe23 could not be distinguished from ZrFe2 in the SEM analyses of most Ucontaining MWF materials, neutron diffraction revealed its presence in trace amounts and indicated that the lattice was contracted about 2%.43 Analysis of Zr6Fe23 that could be distinguished in the SS
15Zr
11U
0.1Pd
0.6Ru
0.3Tc material indicated that phase accommodates less U than the ZrFe2 phase. This is consistent with the preferred accommodation of U in cubic structures.

Metallic Waste Forms

Table 3

527

Phase compositions in SS
15Zr MWF materials

Additions to SS
15Zr

Composition (at.%)a

Phase

5U 2U
1Nb
1Pd
1Rh
1Ru
1Tc 11U
0.1Pd
0.6Ru
0.3Tc

Ferrite U-rich Laves U-poor Laves Ferrite U-rich Laves U-poor Laves Ferrite U-rich Laves U-poor Laves Zr6Fe23

Fe

Cr

Ni

Zr

U

67.5 44.9 49.1 67.6 43.2 46.5 65.9 49.3 53.3 58.1

23.2 3.3 6 22.3 3.3 4.1 26.8 3.1 6.5 11

5 25.7 18 3.5 22.5 17.9 3.1 18 12.1 9.5

Negligible 7.6 20.6 Negligible 10.9 19.1 Negligible 8 21.9 17.2

Negligible 17.2 1.5 Negligible 12.2 3 Negligible 19.3 2.7 1.7

a Absolute uncertainty of each value estimated to be 2 at.%. Data from Keiser, D. D. Jr.; Abraham, D. P.; Sinkler, W.; Richardson, J. W. Jr.; McDeavitt, S. M. J. Nucl. Mater. 2000, 279, 234
244, used with permission. License number 2287090615691.

Table 4

Phase compositions for actinide-bearing SS
15Zr alloys

Alloy

SS
15Zr
2U
2Pu SS
15Zr
6Pu SS
15Zr
10Pu SS
15Zr
2Np SS
15Zr
6Pu
2Np

Phase

Ferrite U-rich Laves U-poor Laves Ferrite U-rich Laves U-poor Laves Ferrite U-rich Laves U-poor Laves Ferrite U-rich Laves U-poor Laves Ferrite U-rich Laves U-poor Laves

Composition (at.%)a Fe

Cr

Ni

Zr

U

Pu

Np

74 49 57 71 38 55 70 35 55 71 42 58 70 46 60

21 3 7 24 3 8 25 2 7 23 3 7 24 3 8

3 22 13 4 34 14 3 34 14 5 34 21 5 30 17

1 21 22 1 8 22 1 7 22 0.4 6 13 0.5 6 13

0.2 9 0.5













0.2 5 0.5 0.2 17 1 0.5 20 2


0.4 10 1










0.5 15 0.5 0.1 5 0.5

a Absolute uncertainty of each value estimated to be 2 at.%. Data from Keiser, D. D. Jr.; Abraham, D. P.; Sinkler, W.; Richardson, J. W. Jr.; McDeavitt, S. M. J. Nucl. Mater. 2000, 279, 234
244, used with permission. License number 2287090615691.

Table 4 shows the compositions of phases in SS
15Zr materials made with added U, Np, and Pu.43 Like U, both Np and Pu are contained almost entirely (and in the same vicinity) within the intermetallic. Selected area electron diffraction and EDS in an analytical transmission electron microscopy study indicated that the U was accommodated primarily in the cubic structure and segregated to the edges of the intermetallic adjacent to ferrite domains. The higher magnification transmission electron microscopy image in Figure 2 of a SS
15Zr
5U

material shows the actinides to be segregated at grain boundaries of the intermetallic phase.44 The distribution of U and other radionuclides within the intermetallic phase is not uniform on a millimeter scale.45 The measured concentrations of contaminants in the phases formed in the SS
15Zr
2Tc materials are summarized in Table 5.46 Similar amounts of cubic and dihexagonal ZrFe2 were detected in neutron diffraction analysis (19 vol.% dihexagonal and 22 vol.% cubic), but these were not distinguishable in

528

Metallic Waste Forms

SEM analysis. Compared with SS
15Zr, the SS
15Zr
2Tc material had more cubic ZrFe2, less dihexagonal ZrFe2, and more Zr6Fe23. Slightly lower lattice parameter values were measured for both ZrFe2 structures in the SS
15Zr
2Tc material than in SS
15Zr, which suggests that Tc was substituted at the Zr sites in that lattice. Small increases were seen in both the ferrite (0.104%) and the austenite (0.056%) lattices. Small amounts of Cr, Mn, Ni, and Si were detected in all phases, and Mo was detected in all but perhaps the Zr6Fe23 intermetallic phase. The assessed (theoretical) Tc
Fe binary phase diagram47 indicates that Tc is soluble in Fe(d) and Fe(g) (up to about 30%), but is essentially insoluble

Intermetallic grains

Steel

in Fe(a), which is the thermodynamically preferred phase below about 912  C. Discrete Tc-bearing phases were not detected in any of the samples that were examined. However, similar amounts of Tc were detected in the ferrite, austenite, and ZrFe2-type and Zr6Fe23-type intermetallic phases formed in a SS
15Zr
2Tc material.43 The Tc was probably dissolved in the Fe(d) formed initially at high temperatures and not excluded when it transformed to Fe(g) and then partially to Fe(a) as the material cooled. That is, the Fe(a) is probably supersaturated with Tc. Although austenite could be converted to ferrite by aggressive heat treatment, austenite was usually seen to remain as a metastable phase in as-cast MWFs made with austenitic steel.43 The authors suggested that ‘the solubility of technetium in the austenite and ferrite may be due to the presence of chromium and nickel in these iron solid solution phases.’ The alloy that was studied only contained about 1.7 wt% Tc. The capacity of the iron solid solution in the MWF for Tc is not known, but the assessed Fe
Tc phase diagram suggests that up to 66 wt% Tc can be accommodated in an FeTc1 + x intermetallic phase.

Actinide-enriched

5.20.8.3

10 µm Figure 2 Scanning electron microscopy backscattered electron images of SS
15Zr
5U alloy. The dark gray areas are steel and the lighter areas are intermetallic phases. Several small white areas seen between some of the intermetallic grains (located by the arrows) are enriched in actinides. Modified from Figure 2b of Janney, D. E. J. Nucl. Mater. 2003, 323, 81
92, used with permission. License number 2287090487208.

Table 5 Host phase

Ferrite Austenite ZrFe2b Zr6Fe23 a

Electrochemical Tests

The initial corrosion rates (of bare metal surfaces) were calculated from linear polarization resistance measurements of several alloy compositions in various solutions. The results of tests in simulated tuff groundwater solutions and groundwaters spiked with NaCl at 20  C are summarized in Table 6 (data from Snyder et al.48). The simulated tuff groundwater solution SJ-13 had a pH of 9 and contains 109 kg m
3
3
3 HCO
Si, 18 kg m
3 3 , 51 kg m , Na, 34 kg m 2

3
SO4 , and 4.3 kg m Cl , and the concentrated
3 SJ-13 contained 12 700 kg m
3 HCO
3 , 5300 kg m
3
3 2
Na, 30 kg m Si, 22 kg m SO4 , and 727 kg m
3 Cl
and had a pH of 8.2. In general, the MWF alloys

Elemental compositions of phases formed in SS
15Zr
2Tc alloy Elemental concentration (at.%)a Cr

Fe

Mn

Mo

Ni

Si

Tc

Zr

23.5 17.8 5.7 8.4

66.2 66.4 45.4 51.7

1.6 1.9 1.6 1.0

1.2 0.7 0.7 Negligible

5.3 11.3 23.9 19.1

0.4 0.6 1.4 1.3

1.6 1.3 1.0 0.9

Negligible Negligible 20.2 17.6

Absolute uncertainty of each value estimated to be 2 at.%. Includes cubic and dihexagonal polytypes. Data from Keiser, D. D. Jr.; Abraham, D. P.; Richardson, J. W. Jr. J. Nucl. Mater. 2000, 277, 333
338, used with permission. License number 2287090753236. b

Metallic Waste Forms

529

Table 6 Corrosion rates measured in a simulated tuff groundwater solution SJ-13, SJ-13 spiked with NaCl, and concentrated SJ-13 at 20  C, in micrometer per year Material

SJ-13

SJ-13 þ 1000 kg m
3 Cl


SJ-13 þ 10 000 kg m
3 Cl


Concentrated SJ-13

SS
5Zr
2Nb
1Ru
1Pd SS
15Zr
1Nb
1Ru
1Pd
1Rh SS
20Zr
2Nb
1Ru
1Pd Type 316 stainless steel Alloy C22 AISI 1018

0.11  0.05 0.20  0.02 0.19  0.03 0.42  0.15 0.17  0.09 16.9  6.60

0.70  0.56 0.52  0.17 0.99  0.59 1.70  0.65 0.56  0.04 105  24

0.75  0.86 1.53  1.89 2.12  1.62 2.31  1.41 0.81  0.56 176  14

1.25  1.00 2.18  2.02 1.80  0.91 2.18  1.40 0.88  0.38 2.20  0.19

Data from Snyder, C. T.; Barnes, L. A.; Fink, J. K. Metal Waste Form Corrosion Release Data from Immersion Tests, Argonne National Laboratory Report ANL-04/15; Argonne National Laboratory: Argonne, IL, 2004.

are more durable than Type 316 stainless steel and became more reactive as the dissolved Cl
concentration was increased in both the spiked SJ-13 and the concentrated SJ-13 solutions. However, the corrosion rates increased by less than a factor of 12. The results in Table 6 are based on at least three linear polarization measurements, and the standard deviations indicate large uncertainties for some values. The effect of pH on the corrosion rates was likewise small, less than a factor of 20 for rates measured at pH 2, 4, 7, and 10.17,49 These test responses reflect the oxidation rate; the effects of the solution chemistry on the dissolution step may be different. A galvanic cell is formed when alloys having different corrosion potentials are coupled by direct contact or through an electrolyte solution. Possible galvanic couplings between the phases in the EBR-11 MWF and Alloy C22 were studied. Alloy C22 was expected to be the most stable metal in the waste package for the proposed Yucca Mountain disposal system and to act as the cathode in all galvanic couples. Tests were conducted to measure the coupling of Alloy C22 with SS
15Zr and SS
15Zr
1Nb
1Pd
1Rh
1Ru waste form compositions.49 The SJ-13 solution (pH 9) and the SJ-13 solution adjusted to pH 2 were used as electrolytes. Both materials were found to be electrochemically noble when coupled with Alloy C22. Companion tests coupling AISI 1018 steel with Alloy C22 showed the steel to be electrochemically active and preferentially oxidized. The galvanic potentials were about
200 and 62 mV for the SS
15Zr and SS
15Zr
1Nb
1Pd
1Rh
1Ru specimens and
600 mV for AISI 1018 steel. Similar results were obtained in the pH 2 solution, except that all potentials were shifted to higher values and the currents were slightly higher. The currents measured in tests with each MWF specimen were very small (but nonzero) and positive, indicating that both alloys are noble

relative to Alloy C22. Enhanced corrosion of the EBR-II MWF is not expected due to galvanic coupling with Alloy C22 or any other material (e.g., carbon steel) that is part of the waste package. 5.20.8.4

Corrosion Tests

Various test methods have been used to study the degradation behavior of alloys and the release of radionuclides. Initial tests indicated that degradation of the alloy occurs through a two-step mechanism of oxidative dissolution in which the metal atoms at the surface are first oxidized to form an oxide layer, and then the oxide layer dissolves. The oxidation step was studied with electrochemical tests and the combined oxidative dissolution processes were studied with static immersion tests, solution-exchange tests, and vapor hydration tests. The release of waste components as the steel and intermetallic phases corrode has been measured in immersion tests under a range of conditions. The metal waste form was subjected to test methods commonly applied to glass waste forms for direct comparison. Composition differences prevented direct comparisons of particular elements, but the releases of all components from the EBR-II MWF were well below the releases of elements from glasses.50 Other tests were conducted with saw and drill shavings of EBR-II MWF material for comparison with tests conducted with crushed glass to assess the possible use of ASTM C 1285 to track metallic waste form consistency.51 Particles reacted under those conditions for 2 years were examined with transmission electron microscopy (TEM) and found to generate oxide layers similar to those formed on alloys reacted at 200  C for shorter durations.52 This suggests that temperature can be used to accelerate corrosion progress.

Metallic Waste Forms

0.5 U Tc

0.4 NL(i) (g m–2)

In most tests with EBR-II MWF materials, the solutions in which the test specimens were immersed were either partially or completely replaced periodically to detect and minimize any effects that released components accumulating in the solution might have on the release rate. For example, the concentration of dissolved Zr may affect the dissolution of the intermetallic phase. Dissolution of EBR-II MWF materials was not stoichiometric, and U was found to be released to the greatest extent under most test conditions. The fractional release of U from SS
15Zrbased materials was typically about 10 times greater than the fractional release of Tc (i.e., when normalized to the amounts of U and Tc in the materials). The results of tests conducted with EBR-II MWF materials containing different amounts of U and Tc are shown in Figure 3.53 The normalized mass loss, NL(i ), is calculated as the mass of an element i that is released into solution divided by the mass fraction of element i in the corroding solid and by the surface area of the solid. This allows the releases of different elements in the same and different tests to be compared directly. The release rates of both U and Tc are high initially (i.e., during the first 50 days) but decrease and then become nearly constant over time. The initial decrease in the release rate is interpreted to correspond to the formation of oxide layers over the steel and intermetallic phases, which limit the continued corrosion (oxidation) of those metal phases. The negligible release at longer sampling times is probably due to the very slow dissolution of the oxide layers. To characterize the oxide layers, relatively thick layers were generated on the surfaces of exposed steel and intermetallic phases by reacting a SS
15Zr material in steam at 200  C for 91 days.54,55 Oxide material at the solution interface appears less compacted and has a higher oxygen content than oxide at the metal interface. The difference is probably due to dissolution of oxide at the outer surface, while corrosion of the metal generates oxide at the oxide layer/ metal interface. Oxygen must diffuse through the oxide layer to oxidize the metals and grow the inner layer, while the outer region of the oxide layer dissolves slowly into solution. As seen in the photomicrographs in Figure 4, both oxide layers adhere to the underlying metallic phase.56 Solids analyses indicate that both Tc and U are retained in the oxide layers.56 This implies that Tc is oxidized to insoluble TcO2 that is immobilized within the oxide layer and that further oxidation is required to release the soluble species TcO
4 . Immersion tests with various Tc-doped EBR-II MWF materials indicate that Tc is

0.3

0.2

0.1

0

0

50

(a)

100 150 200 250 300 350 400 Time (days)

0.4 U Tc

0.3 NL(i) (g m–2)

530

0.2

0.1

0 0 (b)

200

400

600 800 Time (days)

1000

1200

Figure 3 Releases of U and Tc measured in a long-term solution exchange tests with (a) SS
15Zr
4NM
2U
1Tc and (b) SS
15Zr
11U
0.6Ru
0.3Tc. Data from Johnson, S. G.; Noy, M.; DiSanto, T.; Keiser, D. D. Jr. In Proceedings of the DOE Spent Nuclear Fuel and Fissile Materials Management Meeting, Charleston, SC, Sept 17
20, 2002; American Nuclear Society: La Grange Park, IL, 2002; Waste Form Testing session.

released initially (presumably as TcO
4 ) when the surface is still active, but that the release decreases with time as the oxide layers form on the underlying steel and intermetallic phases. This is probably due to the slow dissolution of the oxide layers and their roles as diffusion barriers to oxygen infiltration. Additional evidence is needed to demonstrate that the passivating effect of the oxide layers will be maintained over geological time scales. A series of immersion tests was conducted to characterize the effects of temperature, pH, and Cl
concentration on the release of U from a SS
15Zr
10U alloy.57 A modification of the ASTM C 1220 method was used wherein monolithic test specimens were immersed in a pH buffer solution spiked with 0, 1000, or 10 000 kg m
3 NaCl at a specimen surface

Metallic Waste Forms

531

1

Oxide layers

1

2

2

3

Steel

Oxide layers

100 nm

(a)

(b)

Intermetallic

100 nm

Figure 4 Transmission electron microscopy photomicrographs of oxide layers formed over (a) steel phase and (b) intermetallic phase of MWF reacted at 200  C for 91 days. Modified from Dietz, N. L. Transmission Electron Microscopy Analysis of Corroded EBR-II Metallic Waste Forms; Argonne National Laboratory Report ANL-05/09; Argonne National Laboratory: Argonne, IL, 2005.

area-to-solution volume ratio of 10 m
1 at 50, 70, and 90  C. The entire solution was removed for analysis and replaced with fresh buffer solution after 14, 28, and 70 days. Although the test conditions are not representative of disposal conditions, the tests provide a convenient means to measure the effects of temperature, pH, and Cl
concentration on the release of U as the oxide layer forms. These dependencies can be incorporated into a more realistic model that more accurately represents the disposal environment. Figure 5(a) and 5(b) shows the results of tests conducted at 50  C in pH 4 and pH 8 solutions spiked with 0, 1000, and 10 000 kg m
3 NaCl, where the normalized mass losses based on the cumulative mass of U released to solution are plotted against the reaction time.57 Note that each test under a particular set of pH/Cl/temperature conditions was conducted with a single specimen to eliminate the variance in U concentration at the surface. These results show significant effects of pH, Cl
concentration, and time on the amount of U released. The release slows with time such that release during the initial 14 days accounts for about half the cumulative release over 70 days. The release of U is of primary interest because U is the most rapidly released radionuclide and is used to provide a conservative bound for the release rates of all radionuclides in performance assessments. The average release rates based on the cumulative values over 70 days are plotted against the final pH in Figure 5(c) for all tests at 50  C, where NR(U) ¼ NL (U)/70 days.57 The variance in the results of replicate tests may reflect the nonuniform distribution of U in the alloy and between test specimens. The release of U is faster than most other elements in most tests. Figure 5(c) shows that the release rates do not have a simple pH dependence.

5.20.8.5

Corrosion Models

The corrosion, degradation, and release of radionuclides from the EBR-II MWF is hypothesized to occur through an oxidation
dissolution mechanism in which metallic components exposed at the surface are first oxidized and form an oxide layer and then the outer layer dissolves. Formation of oxide layers on the exposed surfaces of the steel and intermetallic phases slows the release of all components to solution. The release rates of all radionuclides were modeled to equal that of the most efficiently released constituent in each particular test regardless of whether the component was released from the steel or intermetallic phase. In this regard, the waste form was modeled to be a homogeneous single phase that dissolved stoichiometrically. An empirical model for EBR-II MWF degradation was developed based on the results of electrochemical and dissolution tests with several materials.48,58,59 That model incorporated the dependencies of the release rate on temperature, pH, and Cl
concentration that were measured in test environments ranging from pH 2 to 12, 25 to 90  C, and about 0 to 10 000 kg m
3 Cl
. The model is based on the general concept that the initially bare alloy surface becomes covered with an oxide layer that slows the releases of radionuclides and matrix components to solution. The layer is modeled to protect the EBR-II MWF surface from continued corrosion, but credit is only taken for the length of time this was observed in the tests that were used to measure the dependencies. At the end of that period, the passivating effect of the layer is modeled to vanish and then redevelop at the same rate during the next period. That is, the layer is modeled to periodically spall from the underlying metal and

532

Metallic Waste Forms

pH 4

10

1

0.1

0

20 40 60 Test duration (days)

(a)

pH 8

1

NL(U) (g m–2)

NL(U) (g m–2)

100

0.1

0.01

0.001 0

80

20 40 60 Test duration (days)

(b)

80

NR(U) (g m–2 days–1)

0.1

0.01

0.001

0.0001 2

4

6

(c)

8 pH

10

12

14

▪

Figure 5 Results of static tests at 50  C in solutions with (●) 10 000 kg m
3 Cl
, ( ) 1000 kg m
3 Cl
, and (⧫) without added NaCl: NL(U) at (a) pH 4 and (b) pH 8, and (c) average release rates NR(U) through 70 days for tests at various pH values in solutions with 1000 kg m
3 Cl
. Data from Ebert, W. L.; Lewis, M. A.; Barber, T. L.; DiSanto, T.; Johnson, S. G. Static Leach Tests with the EBR II Metallic Waste Form; Argonne National Laboratory Report ANL-03/29; Argonne National Laboratory: Argonne, IL, 2003.

then reform. This moderation of the passivating effect is due to the absence of direct evidence regarding the long-term stability of the oxide layers. The model presumes a common time dependence of the oxidation and dissolution reactions for releasing constituents to solution and growing the oxide layer. If the corrosion and release rates decrease exponentially due to growth of the oxide layers, then the cumulative release of all constituents should follow the logarithmic form: Cumulative constitute release ¼ a  lnð1 þ bt Þ ½3 where a and b are fitting parameters and t is time. For sparingly soluble oxides, the thickness of the layers should increase following the same equation. The fitting parameters a and b have the following significance59: the product a  b gives the initial release rate prior to formation of the layer, 1/b gives the

characteristic time required to passivate the surface (i.e., until the release rate becomes negligible compared to experimental uncertainty), and a represents the extent of corrosion necessary for the layer to significantly slow the release to solution. The metallic waste form does not dissolve stoichiometrically in laboratory tests because of differences in the oxidation rates of individual elements, their solubility limits, and their sequestration in alteration phases. Although there is no experimental evidence that the slowing effect of the oxide layers will not persist over long times, for example due to the layers sloughing off, application of the model to the EBR-II MWF was conservatively limited to 1-year periods, which is the longest duration over which most test series were conducted. The average release rate over the time interval Te that the oxide layer remains an effective barrier is:

Metallic Waste Forms

Release rate ¼

amax  lnð1 þ bt Þ Te

½4

The term amax is used to indicate that the element released the fastest under particular conditions was used to fit the model. Values of the model parameters amax and b capture the dependencies on temperature ( C), pH, and the Cl
concentration (kg m
3). Expressions were determined by assuming a simple linear or quadratic dependence on these variables, using a least-squares fit of the experimental data, and minimizing the number of free parameter combinations. The dependencies, which were determined by fitting experimental results,48,53 are given in eqns [5] and [6] (see Bauer et al.58,59): ln ðb  amax Þ ¼
0:10105 þ ð0:015112 þ 5:8201  10
6  ½Cl
Þ  T
0:69848  pH ½5 and ln amax ¼ 7:9812 þ ð2:3938  10
4  ½Cl
Þ
1:2273  pH

½6

The fitted dependencies given in eqns [5] and [6] are not unique relationships, but they provide physically sensible fits for conditions relevant to a geological disposal site. The modeled rate increases with increasing temperature and Cl
concentration, and decreases with increasing pH. Equation [4] can be rewritten in terms of the fitted dependencies as   amax  ln 1 þ b amaxamax t ½7 Bounding release rate ¼ Te where the cumulative release over the interval Te can be used to provide a conservative time-independent rate that can be compared with the rates calculated for other waste forms. Note that the rate equation given in eqn [7] represents release from a metallic waste form having the specific phase composition and chemical composition for treated EBR-II spent fuel. Rate equations for other compositions would require conducting separate suites of tests to determine parameter values for each composition. 5.20.8.6

Repository Model

The EBR-II waste forms were not included in the performance assessment calculations conducted in support of the Yucca Mountain repository license application for construction. To evaluate the likely

533

acceptability of the EBR-II MWF, the source term for metal waste form degradation using eqn [7] was compared to that for HLW glass.60 The radionuclide release rates used in performance assessment calculations are calculated as the product of the release rate from the waste form degradation model, the radionuclide inventory, and the reacted surface area.61 The products of the degradation rate, inventory, and surface area for each waste form can be compared to estimate the impact of replacing some of the HLW glass with EBR-II MWF on repository performance. For both the EBR-II MWF and the HLW glass, the release rates of all radionuclides are modeled to be equal to the waste form degradation rate. Solubility and transport limits are based on individual radionuclides and independent of the source. The effects of temperature and pH on the degradation rate of the EBR-II metal waste form can be compared to the temperature and pH dependence of the defense HLW glass degradation model developed for use in the Yucca Mountain performance assessment calculations. Because the glass model does not include a term for the effects of Cl
(HLW glass dissolution is not affected by dissolved Cl
), the metal waste form model rates were calculated using the highest anticipated Cl
concentration for comparison with the glass model. The Cl
concentration is treated as a constant rather than a variable for the purpose of comparison with the glass model. The concept for disposing EBR-II wastes is to codispose ceramic waste forms with metal waste forms in the same canister to distribute the weight of the metal waste forms among several canisters and waste packages. Degradation of the ceramic waste form would provide Cl
to groundwater contacting the metal waste form. It was estimated that a maximum of 620 kg m
3 Cl
could be dissolved in the water inside a breached canister based on the volume and surface area of ceramic waste form available to infiltrating water. The model predictions can also be compared with the rates measured in modified ASTM C 1220 tests conducted with leachants spiked with NaCl (these were discussed in Section 5.20.8.4). The dissolution rates of a SS
15Zr
10U alloy were measured in tests conducted at 50, 70, and 90  C over a range of pH values using buffer solutions spiked with NaCl to attain 1000 kg m
3 Cl
. The cumulative amount of U released over 70 days was used to calculate the average dissolution rate for comparison with the HLW glass and EBR-II metal waste form models.60 The rates from tests conducted at 50 and 90  C are

534

Metallic Waste Forms

plotted in Figure 6 along with the lines showing the maximum rates from the defense HLW glass models at these temperatures over the full pH range. The dashed lines in Figure 6 show the rates calculated with the empirical metal waste form model given in eqn [6] calculated at 50 and 90  C with 620 kg m
3 Cl
and a time interval Te ¼ 1 year; the rates calculated with 1000 kg m
3 Cl
are only slightly higher. The EBR-II MWF model is representative of the measured rates in acidic and neutral solutions, but underestimates the rates measured in alkaline solutions by more than an order of magnitude. The poor fit in alkaline solutions calls into question the simple pH dependence that is used in the EBR-II MWF model. The key finding demonstrated in Figure 6 is that the rates calculated with the defense HLW glass model bound both the rates calculated with the EBR-II MWF model and the rates measured in the modified ASTM C 1220 tests (solution exchange tests) over the entire pH range, including the rates in alkaline solutions. One exception is the rate measured in the test at 50  C and pH 9, which was slightly higher than the glass model. It is important to note that the MWF model pessimistically ignores the likely long-term stability of the oxide layers that will probably protect the MWF surface throughout the service life of the repository. The oxide layers

3

log rate g (m–2 days–1)

2

MWF immersion 90 ⬚C MWF immersion 50 ⬚C

HLW glass model at 90 ⬚C

1

HLW glass model at 50 ⬚C

0 –1

5.20.8.7 Metallic Waste Form Product Consistency

EBR-II MWF

–2 model at 90 ⬚C –3 –4 0

EBR-II MWF model at 50 ⬚C

2

4

likewise have a minor effect on the rates determined from the short-term modified ASTM C 1220 tests. In effect, the EBR-II MWF model assumes that the oxide layers disappear and reform on an annual basis, such that the average degradation rate from the model becomes increasingly conservative over time. The surface area of HLW glasses available for degradation in performance assessment calculations is based on the dimensions of the glass pour canisters and a cracking factor. The exposed surface area of an average HLW glass log is taken to be 30 m2 for comparison with the EBR-II MWF, which will be cast as ingots having a right cylinder geometry 0.35
0.41 m (14
16 in.) in diameter and 0.05
0.13 m (2
5 in.) thick. One or two ingots will likely be codisposed with two ceramic waste form monoliths in a disposal canister. The ingots are not expected to fracture due to cooling or impact, so the geometric surface area represents the maximum surface area that can be exposed to water. The surface area of a representative ingot 0.4 m (16 in.) in diameter and 0.1 m (4 in.) thick is about 0.38 m2, so the surface area of two EBR-II MWF ingots in a breached canister is about 0.76 m2. This is about 2.5% of the glass surface area. The total inventory of radionuclides in the EBR-II MWF is only about 0.1% of the total inventory of HLW glass, on a per canister basis. The predominant radionuclides are 1.92  1012 Bq 63Ni, 7.77  1011 Bq 99 Tc, 9.99  1010 Bq 60Co, and 6.66  1010 Bq 59Ni. (The total inventory of the EBR-II ceramic waste form is about 16% of the HLW glass inventory, on a per canister basis.) Based on comparison of the combined release rate, inventory, and surface area, the performance of the EBR-II MWF is conservatively bounded by the performance of HLW glass in total system performance assessments.

6

8

10

12

14

pH Figure 6 Comparison of measured MWF degradation rates (data points), rates from empirical MWF model (dashed lines), and rates from HLW glass degradation model (solid lines) at 50 and 90  C. Data from Ebert, W. L. Testing to Evaluate the Suitability of Waste Forms Developed for Electrometallurgically Treated Spent SodiumBonded Nuclear Fuel for Disposal in the Yucca Mountain Repository; Argonne National Laboratory Report ANL-05/ 43; Argonne National Laboratory: Argonne, IL, 2005.

The objective of tracking the consistency of waste form products is to link the controls applied to waste form production with predictable performance in a disposal system (see Section 5.20.5.2). The narrow range of waste stream compositions anticipated from the EBR-II spent fuel inventory simplified the task of tracking waste form product consistency. The control limits placed on the feed compositions are 1
11 mass% total U and 5
20 mass% Zr with targets of 10 mass% total U and 15 mass% Zr. The composition Fe
15Zr results in nearly equal amounts of the steel and Fe2Zr intermetallic phases

Metallic Waste Forms

being formed, and variations in the amount of Zr in the alloy will change the relative amounts of each phase. Acceptable performance requires that the waste form contains at least the minimum amount of the Fe2Zr intermetallic phase required to sequester the actinides present in the waste. It was determined that 5 mass% Zr in the alloy is adequate to produce enough intermetallic phase to accommodate the maximum amount of actinides that could be present in the waste. Most waste streams will contain <5 mass % Zr, and additional Zr must be added to generate an adequate volume of intermetallic phase. The addition of Zr is a critical step in waste processing. Because Zr is not soluble in the steel phase and the actinides substitute for Zr in the intermetallic,44 the Zr content of the waste form provides a reliable measure of the amount of intermetallic phase that will form. It is expected that waste streams from several electrorefining runs will be blended to control the U enrichment in the EBR-II MWF. The Zr content will be tracked, but not optimized. In practice, 5 mass% Zr can be added to any waste mixture to ensure that the total Zr will exceed the minimum content of 5 mass% by the amount of Zr present in the fuel wastes stream. Measurement of the Zr content can be used to track waste form consistency by verifying enough Zr is present.42 Based on the Fe
Zr binary phase diagram and scoping tests, about 30 mass% Zr can be accommodated in the phase structure of the waste form, although the temperature required to reactively dissolve the Zr will increase with the Zr content of the mixture. However, scoping tests showed that alloys with that much Zr were brittle, so an upper limit of 20 mass% Zr is imposed to provide enough of the steel phase to maintain the physical integrity of the alloy. As part of this approach, it must be demonstrated that the composition analysis of small samples taken from waste forms having the fine-grained microstructure of the SS
15Zr alloy will provide a sufficiently accurate measure of the Zr content. This can be done by comparing the analytical results of various sample sizes to bulk compositions based on known formulations or analysis of larger aliquants.

5.20.9 Summary Metallic waste forms are appropriate for waste streams that include significant amounts of metallic components or contain radionuclides that are most effectively processed and immobilized as metals. The current design of metallic waste forms is based on the

535

compatibility of the waste stream with the targeted assemblage of phases that form from the waste and added metals. Other metals can be added to facilitate melting the waste metals or forming phases that effectively immobilize the waste components in a consistent assemblage of phases. Binary phase diagrams provide useful insights regarding the likely melting temperatures, eutectic melting compositions, likely phase compositions and assemblages, and solubilities for various mixtures of wastes and additives. The degradation behaviors of radionuclidebearing alloy phases must be modeled to predict radionuclide release and migration over very long times in performance assessment calculations. These calculations are used to demonstrate that regulatory requirements will be met over the service life of the disposal system as the waste forms degrade. This requires testing to understand and quantify radionuclide release rates under the range of conditions anticipated in a disposal facility and those possible under extreme conditions. Long-term predictions based on mechanistic models are more reliable than those based on empirical models, but mechanistic models are very difficult to develop. An empirical model with mechanistic underpinnings has been developed for the specific metal waste form composition formulated and designed to immobilize metallic wastes from the electrometallurgical treatment of spent EBR-II fuel. Tests and analyses of the EBR-II MWF provide valuable insight regarding the disposition of radionuclides in component phases and their releases as those phases degrade. They also show the difficulties of quantifying a complicated and coupled corrosion process. Development of a mechanistic model for the degradation of alloyed waste forms is probably the most important research need, especially a term to represent the effect of the phase composition. Qualification of metallic waste forms for disposal will likely require evidence that acceptable waste form performance (i.e., controlling the radionuclide release) can be expected based on how the waste form is produced. That is, controlling the gross composition of waste streams and additives and controlling the processing conditions (processing temperature, furnace atmosphere, etc.) will result in a waste form with predictable phase assemblage and radionuclide disposition that can be related to waste form performance. Because a very large number of waste forms will be made over long times with varying waste stream compositions, controlling waste form performance through processing controls is a very

536

Metallic Waste Forms

important aspect of waste form design. This is necessary to provide confidence that the performances of all waste forms produced are bounded by the radionuclide release rates used in the performance assessment calculations. A variety of test methods is required to provide the range of data needed to determine the radionuclide release mechanism(s), the influence of environmental variables and waste form composition on the release rate (both the chemical and phase composition), develop and parameterize a model for calculating radionuclide release over very long times, and establish process control limits for making consistent products. The electrochemical and immersion test methods used to characterize the EBR-II MWF have provided an extensive database to support its qualification for disposal in the proposed Yucca Mountain repository. Lacking a mechanistic model for radionuclide release, the approach taken was to show that the impact of disposing EBR-II waste forms would be dwarfed by the impact of commercial fuel and HLW glass and bounded by performance assessment calculations. A mechanistic model of radionuclide release is an important goal for advancing the development of metal waste forms. From the conceptual model of sequential oxidation and dissolution steps, new test methods coupling electrochemical techniques with test methods that accelerate corrosion are probably needed to characterize radionuclide release under a range of conditions. The oxidation step may also complicate coupling waste form degradation with the transport models used in performance assessments. For example, the need to take into account galvanic couples within the waste from and between the waste form and other metals may require additional terms in transport models. Methods to accelerate metal corrosion processes are also needed to help understand long-term performance, support waste form modeling, and lead to acceptance for disposal. Testing and modeling completed to date indicates that metallic waste forms provide a preferable option for several existing and anticipated waste streams from fuel treatment and recycling processes. Several challenges remain in development of metallic waste forms for high-level and low-activity radioactive waste streams in the areas of performance and consistency testing, developing waste form degradation and radionuclide release models, and integrating those into performance assessments. Innovative research in electrochemical measurements and modeling is needed to support continued

model development and waste form formulations, and replace the current empirical approaches with sound mechanistic models. (See also Chapter 3.01, Metal Fuel; Chapter 5.14, Spent Fuel Dissolution and Reprocessing Processes; Chapter 5.16, Spent Fuel as Waste Material; Chapter 5.18, Waste Glass and Chapter 5.19, Ceramic Waste Forms).

Acknowledgments The submitted manuscript has been created by UChicago Argonne, LLC, Operator of Argonne National Laboratory (‘‘Argonne’’). Argonne, a U.S. Department of Energy Office of Science laboratory, is operated under Contract No. DE-AC02-06CH11357. The U.S. Government retains for itself, and others acting on its behalf, a paid-up nonexclusive, irrevocable worldwide license in said article to reproduce, prepare derivative works, distribute copies to the public, and perform publicly and display publicly, by or on behalf of the Government.

References 1.

Rard, J. A. Critical Review of the Chemistry and Thermodynamics of Technetium and Some of Its Inorganic Compounds and Aqueous Species; Lawrence Livermore National Laboratory Report UCRL-53440; Lawrence Livermore National Laboratory: Livermore, CA, 1983. 2. Kaye, M. H.; Lewis, B. J.; Thompson, W. T. J. Nucl. Mater. 2007, 366, 8–27. 3. Gombert, D. Global Nuclear Energy Partnership Integrated Waste Management Strategy; US Department of Energy Report GNEP-WAST-AI-RT-2008-000214; Idaho National Laboratory: Idaho Falls, ID, 2008. 4. ASTM. Annual Book of ASTM Standards; ASTMInternational: West Conshohocken, PA, 2009; Vol. 12.01. 5. Jantzen, C. M.; Pickett, J. B.; Beam, D. C. Process/ Product Models for the Defense Waste Processing Facility (DWPF): Part I. Predicting Glass Durability from Composition Using a Thermodynamic Hydration Energy Reaction Model (THERMO); Westinghouse Savannah River Report WSRC-TR-93-672; Westinghouse Savannah River Company: Aiken, SC, 1994. 6. Ebert, W. L.; Cunnane, J. C.; Thornton, T. A. In Proceedings of the 9th International High-Level Radioactive Waste Management Conference, Las Vegas, NV, Apr 29
May 3, 2001; American Nuclear Society: La Grange Park, IL, 2001; CD-ROM Section I-10. 7. McDeavitt, S. M.; Abraham, D. P.; Keiser, D. D.; Park, J. Y. In Proceedings of the Spectrum’96 Meeting, Nuclear and Hazardous Waste Management International Topical Meeting, Seattle, WA, Aug 18
23, 1996; American Nuclear Society: La Grange Park, IL, 1996; pp 2477–2484. 8. McDeavitt, S. M.; Abraham, D. P.; Park, J. Y.; Keiser, D. D. JOM 1997, 49(7), 29–32.

Metallic Waste Forms 9.

10.

11.

12.

13. 14.

15. 16. 17. 18.

19.

20.

21.

22.

23.

24.

Goff, K. M.; Teske, G. M.; Howden, K. L.; Johnson, T. A. Electrometallurgical treatment of EBR-II spent fuel. In Waste Management’04 Conference, Tuscon, AZ, Feb 29
Mar 4, 2004. Goff, K. M.; Benedict, R. W.; Teske, G. M.; Johnson, T. J. In Proceedings of the Fifth Topical Meeting on DOE Spent Nuclear Fuel and Fissile Materials Management, Charleston, SC, Sept 17
20, 2002; American Nuclear Society: La Grange Park, IL, 2002; (DVD) Electrometallurgical Processes section. Gombert, D. Global Nuclear Energy Partnership Integrated Waste Management Strategy Waste Treatment Baseline Study; US Department of Energy Report GNEP-WAST-AIRT-2007-000324; Idaho National Laboratory: Idaho Falls, ID, 2007. Vaden, D.; Li, S. X.; Johnson, T. A. In Proceedings of the Fifth Topical Meeting on DOE Spent Nuclear Fuel and Fissile Materials Management, Charleston, SC, Sept 17
20, 2002; American Nuclear Society: La Grange, IL, 2002; (DVD) Electrometallurgical Processes Section. Abraham, D. P.; McDeavitt, S. M.; Park, J. Y. Metall. Mater. Trans. 1996, 27A, 2151–2159. Abraham, D. P.; McDeavitt, S. M.; Park, J. Y. In Proceedings of the Embedded Topical Meeting on DOE Spent Nuclear Fuel and Fissile Material Management, Reno, Nevada, June 16
20, 1996; American Nuclear Society: La Grange Park, IL, 1996; pp 123–128. Abraham, D. P.; Richardson, J. W., Jr.; McDeavitt, S. M. Scr. Mater. 1997, 37, 239–244. Abraham, D. P.; Richardson, J. W., Jr.; McDeavitt, S. M. Mater. Sci. Eng. 1997, A239
240, 658–664. McDeavitt, S. M.; Abraham, D. P.; Park, J. Y. J. Nucl. Mater. 1998, 257, 21–34. Ebert, W. L. Testing to Evaluate the Suitability of Waste Forms Developed for Electrometallurgically Treated Spent Sodium-Bonded Nuclear Fuel for Disposal in the Yucca Mountain Repository; Argonne National Laboratory Report ANL-05/43; Argonne National Laboratory: Argonne, IL, 2005. Ebert, W. L. Immobilizing GNEP Wastes in Pyrochemical Process Waste Forms; US Department of Energy Report GNEP-WAST-PMO-MI-DV-2008-000150; Idaho National Laboratory: Idaho Falls, ID, 2008. Gerhold, W. F.; Escalante, E.; Sanderson, B. T. The Corrosion Behavior of Selected Stainless Steels in Soil Environments; National Bureau of Standards Report NBSIR 81-2228; National Bureau of Standards: Washington, DC, 1981. Abraham, D. P.; Richardson, J. W. Phase stability of laves intermetallics in a stainless steel–zirconium alloy. In Proceedings of the Long Term Stability of High Temperature Materials Conference, Warrendale, PA; Fuchs, G. E., Dannemann, K. A., Deragon, T. C., Eds.; 1999; pp 169–179. DOE US Department of Energy Office of Civilian Radioactive Waste Management. Waste Acceptance System Requirements Document, Rev. 5; US Department of Energy Office of Civilian Radioactive Waste Management Report DOE/RW-0351 Rev. 05, May 2007; US Department of Energy: Washington, DC, 2007. Bacon, D. H.; White, M. D.; McGrail, B. P. Subsurface Transport over Reactive Multiphases (STORM): A Parallel, Coupled, Nonisothermal Multiphase Flow, Reactive Transport, and Porous Medium Alteration Simulator, Version 3.0. User’s Guide; Pacific Northwest National Laboratory Report PNNL-14783; Pacific Northwest National Laboratory: Richland, WA, 2004. Sullivan, T. M.; Suen, C. J. Low-Level Waste Shallow Land Disposal Source Term Model: Data Input Guides; Nuclear

25. 26. 27. 28. 29.

30. 31.

32. 33. 34.

35.

36.

37.

38. 39.

40.

41. 42.

43.

537

Regulatory Commission Report NUREG/CR-5387; Nuclear Regulatory Commission: Washington, DC, 1989. Jones, D. A. Principles and Prevention of Corrosion; Prentice Hall: Englewood Cliffs, NJ, 1992. Stansbury, E. E.; Buchanan, R. A. Fundamentals of Electrochemical Corrosion; ASM International: Materials Park, OH, 2000. ASTM. Annual Book of ASTM Standards; ASTMInternational: West Conshohocken, PA, 2009; Vol. 03.02. Ebert, W. L.; Bates, J. K.; Bourcier, W. L. Waste Manag. 1991, 11, 205–221. ISO. Nuclear energy—Soxhlet-mode chemical durability test—Application to vitrified matrixes for high-level radioactive waste; International Standard ISO 16797:2004; International Standard Organization: Geneva, Switzerland, 2008. Wronkiewicz, D. J.; Bates, J. K.; Gerding, T. J.; Veleckis, E.; Tani, B. S. J. Nucl. Mater. 1992, 190, 107–127. McGrail, B. P.; Martin, P. F.; Lindenmeier, C. W. In Scientific Basis for Nuclear Waste Management XX, Proceedings of the Materials Research Society Symposium, Boston, MA, Dec 2
6, 1996; Gray, W. J., Triay, I. R., Eds.; Materials Research Society: Pittsburgh, PA, 1997; Vol. 465, pp 253–260. Pierce, E. M.; McGrail, B. P.; Valenta, M. M.; Strachan, D. M. Nucl. Technol. 2005, 155(2), 149–165. White, A. F.; Brantley, S. L. Chem. Geol. 2003, 202, 479–506. Fuhrmann, M.; Schoonen, M. Leaching of Slag from Steel Recycling: Radionuclides and Stable Elements; Brookhaven National Laboratory Report BNL-71445; Brookhaven National Laboratory: Upton, NY, 1997. Arnold, G.; Colombo, P.; Doty, R. M.; et al. Lysimeter Study of Commercial Reactor Waste Forms: Waste Form Acquisition Characterization and Full-Scale Leaching; Brookhaven National Laboratory Report BNL-51613; Brookhaven National Laboratory: Upton, NY, 1983. Oblath, S. G.; Grant, M. W. Special Wasteform Lysimeters Initial Three-Year Monitoring Report; Savannah River Laboratory Report SRL-DP-1712; Westinghouse Savannah River Company: Aiken, SC, 1985. Last, G. V.; Serne, R. J.; LeGore, V. L. Field Lysimeter Studies for Performance Evaluation of Grouted Hanford Defense Wastes; Pacific Northwest National Laboratory Report PNL-10166; Pacific Northwest National Laboratory: Richland, WA, 1995. Jantzen, C. M.; Kaplan, D. I.; Bibler, N. E.; Peeler, D. K.; Plodinec, M. J. J. Nucl. Mater. 2008, 378, 244–256. Sullivan, T. Waste Container and Waste Package Performance Modeling to Support Safety Assessment of Low and Intermediate-Level Radioactive Waste Disposal; Brookhaven National Laboratory Report BNL-74700-2005IR; Brookhaven National Laboratory: Upton, NY, 2004. Federal Register. In Record of Decision for the Treatment and Management of Sodium-Bonded Spent Nuclear Fuel, Sept 19, 2000; Federal Register: Washington, DC 2000; Vol. 65(182), pp 56565–56570. Arias, D.; Granovsky, M. S.; Abriata, J. P. In Phase Diagrams of Binary Iron Alloys; Okamoto, H., Ed.; ASM International: Materials Park, OH, 1993. Keiser, D. D., Jr.; Johnson, S. G.; Ebert, W. L. In Proceedings of the DOE Spent Nuclear Fuel and Fissile Materials Management Meeting, Charleston, SC, Sept 17
20, 2002; American Nuclear Society: La Grange Park, IL, 2002; Poster Session. Keiser, D. D., Jr.; Abraham, D. P.; Sinkler, W.; Richardson, J. W., Jr.; McDeavitt, S. M. J. Nucl. Mater. 2000, 279, 234–244.

538

Metallic Waste Forms

44. Janney, D. E. J. Nucl. Mater. 2003, 323, 81–92. 45. Janney, D. E.; Keiser, D. D., Jr. JOM 2003, 55(9), 59–60. 46. Keiser, D. D., Jr.; Abraham, D. P.; Richardson, J. W., Jr. J. Nucl. Mater. 2000, 277, 333–338. 47. Okamoto, H. In Binary Alloy Phase Diagrams; Massalski, T. B., Ed.; ASM International: Materials Park, OH, 1990. 48. Snyder, C. T.; Barnes, L. A.; Fink, J. K. Metal Waste Form Corrosion Release Data from Immersion Tests; Argonne National Laboratory Report ANL-04/15; Argonne National Laboratory: Argonne, IL, 2004. 49. Abraham, D. P.; Peterson, J. J.; Katyal, N. K.; Keiser, D. D.; Hilton, B. A. In Proceedings of the Corrosion 2000 Conference, Orlando, FL, Mar 26
31, 2000; NACE International: Houston, TX, 2000; Paper No. 00205. 50. Johnson, S. G.; Noy, M.; DiSanto, T.; Barber, T. L. In Scientific Basis for Nuclear Waste Management XXV. Proceedings of the Materials Research Society Symposium, Boston, MA, Nov 26
29, 2001; McGrail, B. P., Cragnolino, G. A., Eds.; Materials Research Society: Warrendale, PA, 2001; Vol. 713, pp 705–711. 51. Johnson, S. G.; Keiser, D. D.; Frank, S. M.; DiSanto, T.; Warren, A. R.; Noy, M. In Scientific Basis for Nuclear Waste Management XXIII, Proceedings of the Materials Research Society Symposium, Boston, MA, Nov 29
Dec 2, 1999; Smith, R. W., Shoesmith, D. W., Eds.; Materials Research Society: Warrendale, PA, 2000; Vol. 608, pp 589–594. 52. Luo, J. S.; Abraham, D. P. In Scientific Basis for Nuclear Waste Management XXIII, Proceedings of the Materials Research Society Symposium, Boston, MA, Nov 20
Dec 2, 1999; Smith, R. W., Shoesmith, D. W., Eds.; Materials Research Society: Warrendale, PA, 2000; Vol. 608, pp 583–588. 53. Johnson, S. G.; Noy, M.; DiSanto, T.; Keiser, D. D., Jr. In Proceedings of the DOE Spent Nuclear Fuel and Fissile Materials Management Meeting, Charleston, SC, Sept

54. 55. 56.

57.

58.

59.

60.

61.

17
20, 2002; American Nuclear Society: La Grange Park, IL, 2002; Waste Form Testing session. Abraham, D. P.; Dietz, N. L.; Finnegan, N. In Proceedings of the Corrosion 2001 Conference; NACE International: Houston, TX, 2001; Paper No. 01139. Abraham, D. P.; Dietz, N. L. Mater. Sci. Eng. 2002, A329
331, 610–615. Dietz, N. L. Transmission Electron Microscopy Analysis of Corroded EBR-II Metallic Waste Forms; Argonne National Laboratory Report ANL-05/09. Argonne National Laboratory: Argonne, IL, 2005. Ebert, W. L.; Lewis, M. A.; Barber, T. L.; DiSanto, T.; Johnson, S. G. Static Leach Tests with the EBR II Metallic Waste Form; Argonne National Laboratory Report ANL-03/29. Argonne National Laboratory: Argonne, IL, 2003. Bauer, T. H.; Abraham, D. P.; Fink, J. K.; Johnson, I.; Johnson, S. G.; Wigeland, R. A. In Proceedings of the International High-Level Radioactive Waste Management Conference, Las Vegas, NV, Apr 29
May 3, 2001; American Nuclear Society: La Grange Park, IL, 2001; CD-ROM. Session E-3 Source Term-I: General Modeling Topics. Bauer, T. H.; Johnson, S. G.; Snyder, C. T. In Proceedings of the DOE Spent Nuclear Fuel and Fissile Materials Management Meeting, Charleston, SC, Sept 17
20, 2002; American Nuclear Society: La Grange Park, IL, 2002. Ebert, W. L.; Lewis, M. A.; Barber, T. L.; Johnson, S. G. In Scientific Basis for Nuclear Waste Management XXVI, Symposium, Boston, MA, Dec 2
5, 2002; Finch, R. J., Bullen, D. B., Eds.; Materials Research Society: Warrendale, PA, 2003; Vol. 757, pp 71–80. BSC. Defense HLW Glass Degradation Model; Bechtel SAIC Company Report ANL-EBS-MD-000016, Rev. 02; Bechtel SAIC: Las Vegas, NV, 2004.

5.21

Graphite

J. Fachinger Furnaces Nuclear Applications Grenoble ZU Hanau Research and Development, Hanau, Germany

ß 2012 Elsevier Ltd. All rights reserved.

5.21.1

Introduction

540

5.21.2 5.21.2.1 5.21.2.2 5.21.2.3 5.21.2.4 5.21.2.5 5.21.3 5.21.4 5.21.4.1 5.21.4.1.1 5.21.4.2 5.21.4.2.1 5.21.4.2.2 5.21.5 5.21.5.1 5.21.5.2 5.21.5.3 5.21.5.4 5.21.6 5.21.6.1 5.21.6.2 5.21.6.3 5.21.6.4 5.21.7 5.21.8 References Appendix 1

Amounts of i-Graphite and Its Origin Russia United Kingdom France United States Others Retrieval of i-Graphite Graphite Properties Physical Properties Wigner energy Chemical Properties Oxidation in gaseous phases Graphite reactions with liquids Graphite Radioactivity Formation of 3H Formation of 14C Formation of 36Cl Diffusion of Radionuclides in Graphite Graphite Treatment for Disposal or Recycling Waste Packages and Encapsulation Thermal Treatment The Russian ‘Self-Propagating High-Temperature Synthesis SHS’ Recycling of i-Graphite Final Disposal Summary

540 540 541 541 541 542 542 542 542 543 543 543 545 545 546 546 547 549 550 550 551 552 553 553 556 557 558

Amounts of Irradiated Graphite in Different Countries

Abbreviations AGR AM-1 AMB AVR b BEPO CP1 EDF EL2 FRJ-1

Advanced gas-cooled reactor Prototype of RBMK Aтoм Mиpный Бoльшoй Allgemeiner Versuchsreaktor (a small HTR prototype reactor in Germany) Barn (10
24 cm
2) British experimental pile ‘0’ Chicago pile-1 E´lectricite´ de France SA Graphite moderated test pile in France Research Reactor Ju¨lich 1

GLEEP

Graphite low energy experimental pile HTGR High-temperature gas-cooled reactor HTR High-temperature reactor IAEA International Atomic Energy Agency MAGNOX Magnesium alloy graphite moderated gas cooled uranium oxide reactor RBMK Reaktor Bolschoi Moschtschnosti Kanalny THTR Thorium high-temperature reactor UNGG Uranium naturel graphite gaz reactor WAGR Windscale’s advanced gas-cooled reactor

539

540

Graphite

5.21.1 Introduction Graphite has been used in nuclear technology since the birth of this technology. The first artificial nuclear reactor, the Chicago Pile-1, consisted of a pile of uranium and graphite. It was the fundament for future developments in the different graphite-moderated nuclear power reactors such as the Uranium Naturel Graphite Gaz reactors (UNGG) in France, Magnox and advanced gas-cooled reactors (AGR) in the United Kingdom, or RBMK in Russia. The culmination of this development was the high-temperature reactor, for example, the Fort Saint Vrain reactor in the United States or the Thorium-Hochtemperaturreaktor (THTR) in Germany. Worldwide, more than 230 000 tons of irradiated graphite (i-graphite) exist, which will eventually need to be managed as radioactive waste.1 The major part of i-graphite is still in operational or shutdown nuclear power reactors. Actually, the reactor cores have been removed from Fort Saint Vrain, GLEEP, and BEPO. The removal of the core from the Allgemeiner Versuchsreaktor (AVR) is proposed for the next few years. Smaller quantities of i-graphite result additionally from operation, in the form of graphite sleeves, which have to be replaced during operation, and from different kinds of research reactors. The total amount of i-graphite is assumed to be in the range of 220 000–250 000 tons. A more detailed overview of the amount of i-graphite is given in Section 5.21.2. Graphite changes its properties during irradiation in a nuclear reactor. Most important for the treatment and disposal of i-graphite is the possibility of storing energy in the form of structural defects, the so-called Wigner energy. The entire Wigner energy could be released rapidly after an initial local release. Furthermore, the graphite will be contaminated by radionuclides. They result from the activation of 13C and impurities in the graphite matrix as well as from the depletion of fission products released from the fuel elements. This is described in Section 5.21.5. Various treatment methods have been developed or are under development to decontaminate i-graphite or to optimize the disposal volume and behavior, respectively. Section 5.21.6 gives a short overview of all these developments. A major issue is the establishment of a close graphite cycle, which is essential for the future development of graphite-moderated nuclear reactors.

The last section is dedicated to the final disposal options for i-graphite and the behavior of i-graphite under different disposal conditions.

5.21.2 Amounts of i-Graphite and Its Origin Generally, one can distinguish between four different types of nuclear reactors that utilize graphite as neutron moderator and reflector.  Aircooled graphite piles with a low power density as test facilities, prototypes, and first-generation plutonium production reactors.  Carbon-dioxide-cooled reactors (Magnox and UNGG) for electricity supply and/or plutonium production.  Helium-cooled high-temperature reactors for electricity generation and process heat generation.  Graphite-moderated water-cooled reactors for plutonium production and/or electricity generation. The last feature, electricity generation, has been optimized especially in the Russian RBMK reactors.

5.21.2.1

Russia

The main sources of i-graphite are the RBMK nuclear power plants as well as high-capacity plutonium production reactors. Five RBMK power plants with 11 reactors are still in operation in Russia.2,3 The original license foresaw a lifetime of 30 years. However, lifetime extensions are already licensed or envisaged. Therefore, the first shutdowns are expected in 2013, with a replacement program starting in 2015.3 The amount of graphite from these reactors is given in Table A.1. An important fact about this graphite-moderated water-cooled reactor type is that a helium–nitrogen mixture gives the graphite moderator a protective atmosphere, which will have an important impact on the generation of 14C.4 Four more graphite-moderated power reactors, Bilibino-1–4, are in operation. These dual purpose reactors for electricity and heat contain 133 tons of graphite each.3 The AMB-1 and -2 in Beloyarsk, and the AM-1, prototypes of the RMBK reactors, were shutdown in 1981, 1989, and 2004 respectively. The fuel has been removed from the reactor core and stored in cooling basins. The reactor units with the graphite core are under safe storage conditions of IAEA Stage I, under surveillance; further dismantling is planned.

Graphite

Besides these graphite-moderated power plants, 13 high-capacity plutonium power reactors had been in operation in Russia. All of them were shutdown between 1987 and 2008. The reactor units with the graphite core are under safe storage conditions of IAEA Stage II. The first dismantling concepts proposed the transformation of the reactor shafts into a final radwaste repository. However, this disposal concept is not in accordance with the new Russian waste disposal regulations.5 Therefore, a further 21 000 tons of graphite internals of the reactor units, as well as 8000 tons of graphite rings at on-site storage facilities, have to be managed as radioactive waste in Russia. An additional source of i-graphite are research reactors and other critical assemblies. The determination of the exact number of these reactors and assemblies is complicated. However, it can be assumed that more than 110 such facilities exist in Russia. Furthermore, it was not possible to figure out which moderator had been used in these facilities and therefore the amount of i-graphite could not be evaluated. 5.21.2.2

United Kingdom

The United Kingdom has the largest amount of i-graphite that has to be managed as radioactive waste1,6 because most of the British nuclear power plants are gas-cooled graphite-moderated reactors as opposed to those in other countries, which utilize water-moderated reactors as an alternative. The Magnox reactor type was utilized after test reactor and prototype development in the late 1950s. The name was derived from the fuel cladding made from a magnesium–aluminum alloy. The last Magnox reactor was commissioned in 1971. The next generation of gas-cooled reactors were the AGR, commissioned between 1976 and 1989. Both reactor types were graphite-moderated and cooled with CO2. Totally about 80 000 tons of graphite have to be handled as radioactive waste now or in the next two decades after the shutdown of the still operational AGR reactors. An overview of the amount of graphite in the different UK reactors is given in Table A.2 in the Appendix. 5.21.2.3

France

The first graphite-moderated reactor in France was the pile EL2 built at Saclay in 1952.7,8 It was an experimental reactor like the EL3 built in 1957. The total mass of graphite in these two reactors was 109 tons. The operation of these reactors ended in 1965 and 1979, respectively.

541

The first UNGG were built and operated by CEA in Marcoule as plutonium production reactors for the French nuclear deterrent forces. The main characteristics of these reactors are given in Table A.3. While the G1 was still air cooled, all other UNGG used CO2 as cooling gas. The graphite bricks, used as moderator as well as shielding for the internal walls of the reactor, were mounted to a horizontal reactor core. The decommissioning has achieved IAEA level 2. Six more UNGG reactors have been built and operated by EDF for electricity production. The main characteristics of these reactors are also given in Table A.3. All of them are under decommissioning. With the exception of Chinon A1 the design was changed from a horizontal to a vertical shaft for improved fuel handling. However, the fuel cartridges had to be protected by graphite sleeves to withstand the mechanical forces of the upper fuel cartridges. These graphite sleeves are temporarily stored in silos except those from Buggy, which have been disposed at the final disposal sites at Manche and Aube. 5.21.2.4

United States

The development of nuclear reactors started in the United States with the graphite-moderated pile 1 (CP1) in Chicago. The largest amounts of i-graphite in the United States are from the plutonium production reactors at the Hanford site. The Hanford B-Reactor was the first large-scale plutonium production reactor in the world. The reactor was graphite moderated and water cooled. It consisted of an 8.5  11 m horizontal tube and contained 1100 tons of graphite.9 All reactors at the Hanford site were intended for plutonium production. In all, nine plutonium production reactors were operated at the Hanford site. The reactors were shutdown between 1964 and 1971 after an average life span, except the lastbuilt N Reactor (1963), a dual purpose facility for civil electricity generation (shutdown in 1987). Most of the reactors have been entombed after defueling and act as interim storage for the graphite moderator and structural materials to allow the decay of radioactive material until dismantling is possible with a low dose risk. Reactors exclusively for civil application were the HTGR at Peach Bottom and Fort Saint Vrain with a graphite block as moderator. Both were shutdown after a relatively short operational time of 7 and 13 years, respectively. An overview of the amount of graphite in the different reactors is given in Table A.4 in the Appendix.

542

Graphite

5.21.2.5

Others

Table A.5 gives an overview of graphite-moderated reactors in other countries. Most of them are RBMK reactors in countries of the former Soviet Union or Magnox reactors. High-temperature reactors have been constructed in China, Germany, and Japan. Two graphite-moderated high-temperature reactors, the AVR and the THTR, were operated in Germany. About 1000 metric tons of i-graphite and irradiated carbon bricks have to be managed as radioactive waste. Furthermore, about 1 million irradiated fuel pebbles were produced during the operational time of these reactors. They consist mainly of a graphite matrix that contains about 10 000 so-called coated particles (see Chapter 3.06, TRISO Fuel Production and Chapter 3.07, TRISO-Coated Particle Fuel Performance). These particles safely enclose the nuclear fuel and the major part of the fission products. The AVR is under dismantling to the green field. Therefore the reactor vessel with the graphite core will be pulled out as one piece and transferred to an interim storage facility. It will stay there for about 80 years before further treatment. The development in Japan is based on a graphite block core similar to that in the United States. The Chinese HTR-10 is a pebble HTR like the one in Germany.

whole reactor core was flooded before cutting the graphite internals. This procedure provides two advantages. First, the water acts as shielding, which minimizes the dose rate of the employees, and second, the water prevents the formation of graphite dust. However, water purification requires additional effort. A third approach is being used for dismantling the AVR in Ju¨lich, Germany. The whole reactor vessel, including the graphite internals, built as one part, has been lifted out of the reactor and transferred to an interim storage. Before lifting, the reactor core was filled with light concrete to consolidate the internals and to reduce impacts, in case the vessel falls down during the lifting procedure. Further dismantling will be performed after the decay of the main part of g-emitters, especially 60Co. This method represents a way between complete dismantling and long-term safe storage and enables fast cleanup of a site. The choice of the best applicable retrieval method depends on several site-specific facts and therefore there is no ‘best procedure.’ They include the mechanical and physical properties of the graphite, the dose rate of the graphite, and the surrounding structures, as well as the specific side constructions, which determine the space available for the installation of equipment. A detailed overview of retrieval and available procedures and tools is under compilation by an expert team of the European Carbowaste project.10

5.21.3 Retrieval of i-Graphite 5.21.4 Graphite Properties The retrieval of i-graphite is based on two main principles, dry and wet retrieval.10 Dry retrieval has been chosen for the decommissioning of WAGR and GLEEP. GLEEP was a low-energy and low-radiation test reactor. The resulting total activity was so low that the graphite could be removed manually without shielding. Only protective overalls and gloves were required to avoid incorporation. The graphite should be treated in an industrial incinerator that is licensed for the discharge of small amounts of radioactivity. It was noted that the graphite blocks showed only small effects after treatment at 1400 K for 3 h in air. Less then 2% of the graphite was lost during the process. However, about 87% of 3H and 63% of 14C were released from the graphite. The activity and dose rate were so high that a manual retrieval of the graphite was not acceptable for the WAGR. A remote removal system was developed for the retrieval of the graphite stack. Wet retrieval was utilized for dismantling the hightemperature reactor at Fort Saint Vrain. Therefore, the

5.21.4.1

Physical Properties

The properties of graphite are related to its polycrystalline structure (see Chapter 2.10, Graphite: Properties and Characteristics). Graphite crystallites are built by graphite planes, which are loosely bound by van der Waal’s forces. The single planes consist of carbon six rings with a sp2 electron configuration of the carbon atoms and a dislocated p-system on both sides of the plane. Therefore the properties of the crystallites are anisotropic with respect to the orientation parallel or vertical to the planes. Irradiation induces damages in these graphite crystallites, which lead to anisotropic effects in the crystallites. A good example is a radiation-induced expansion in one direction and shrinkage in the other direction. Therefore the macroscopic changes can be anisotropic, if the crystallites have an overall preferred orientation direction, or isotropic, if the crystallites are randomly distributed. This depends on the shape of the crystallites as well as on the production process. For example,

Graphite

extruded graphite shrinks parallel to the extrusion direction and expands perpendicular to the extrusion direction at temperatures below 300  C and shrinks in both directions at higher temperatures. More isotropic molded graphite initially exhibits shrinkage in all directions under all irradiation conditions. The irradiation-induced shrinkage proceeds to a point of maximal density. Further irradiation causes an expansion to the original density and beyond. Besides being important for reactor operation, this effect is also a key issue for waste management. It affects the mechanical stability, which has a large influence on the retrieval of graphite piles from the reactor for decommissioning. Furthermore, the density and porosity may influence the radionuclide release in intermediate storage and especially in final disposal. Another important parameter for disposal is the reduction of the thermal conductivity. Small amounts of fast neutrons will reduce thermal conductivity and can be decreased further by 2 orders of magnitude to 2 W m
1 K
1, depending on neutron dose and irradiation temperature. Another very important property for disposal is porosity, which allows the penetration of aqueous phases into the graphite matrix and therefore an undisturbed transport of radionuclides through the graphite matrix. 5.21.4.1.1 Wigner energy

The Wigner effect is named after its discoverer Eugene Paul Wigner. This effect describes the displacement of atoms in a solid caused by neutron irradiation, which can occur in any solid. However, it has a special importance for solid moderator materials such as graphite. An atom can be moved from its position in a crystal lattice by collision with neutrons, if they have energies above 25 eV. Therefore, high-energy neutron, for example, 1 MeV, causes cascades of damages with about 900 displacements in a graphite moderator. Not all of the displacements lead to lattice defects because the displaced atoms could also fill lattice vacancies. Atoms that cannot be placed in lattice vacancies remain as interstitial atoms between the lattice planes and therefore they are associated with a higher energy.11 When such an atom has sufficient thermal energy, it is able to move to normal lattice position and release excess energy if the position energy is higher than the energy required for the return to a lattice place. If such a process has been initiated, the hole-stored Wigner energy can be released immediately and heat up a graphite pile. This was the cause of the Windscale fire accident.12

543

Wigner energy can cause the following problems related to the management of i-graphite: 1. Initiation of an uncontrolled release of Wigner energy during retrieval. 2. Sawing or cutting of the graphite core can lead to a local heat increase, which may lead to an uncontrolled release of Wigner energy. Therefore such operations should be performed with sufficient cooling. 3. Release of the Wigner energy during final disposal. 4. The temperature of the final disposal sites for lowand medium-level wastes is normally strictly limited. These limits depend on corrosion processes or microbial degradation, and higher temperatures may disturb the integrity of the disposal or increase the reaction rates of release mechanisms. Therefore, the Wigner energy should be dissipated before storage by annealing the graphite at temperatures above 250  C or it has to be demonstrated that the disposal site will not be affected by such an energy release. This has been tried by NIREX but they concluded that Wigner energy is not adequately understood to guarantee that a release of Wigner energy cannot affect the safety of a disposal site.13 Despite this potential risk, only low amounts of i-graphite have considerable amounts of Wigner energy. They are related to reactors operated at low temperatures, for example, reflectors of material test reactors. High reactor operation temperatures, for example, achieved in an AGR would directly cause the annealing of the graphite.13 5.21.4.2

Chemical Properties

‘Burning of radioactive graphite’ has been in public discussion since the accident at Chernobyl. But graphite has an extremely low chemical reactivity, which explains its geochemical stability, proved by the presence of natural graphite ores in the earth’s crust. Graphite needs extremely powerful oxidation agents to convert it into the thermodynamic-favored carbon dioxide. This also allows the utilization of graphite under extreme conditions in industrial processes, for example, as electrode in arc melting at temperatures up to 3000  C or its use as fire extinguisher. 5.21.4.2.1 Oxidation in gaseous phases

Graphite can react with gases such as O2, CO2, H2O, and H2 at elevated temperatures and the temperature depends on the perfection of the graphite’s crystal structure14 and therefore on the amount of impurity.

544

Graphite

Generally, heterogeneous reactions involving a porous solid and a gas can be controlled by one or more of three idealized steps: 1. Mass transport of the reacting gas from gas stream to the exterior graphite surface. 2. Mass transport of the reacting gas from the exterior surface to an active site and mass transport of the products in the opposite direction. 3. Chemical reaction at the active sites. The variation of the reaction rate with temperature for gas–carbon reactions can be divided into three main zones (Figure 1). In the low-temperature zone (zone I), the reaction is controlled by the chemical reactivity of the solid (step 3). There will be almost no concentration gradient of reacting gases throughout the whole volume of the sample because of low reaction rate, and this provides uniform access to the interior surface of porous materials. For graphite – oxygen reaction, the upper limit for temperature will be 500  C, and for graphite–steam system, it will be 850  C.15 In the intermediate-temperature zone (zone II), step 2 becomes important. The diffusion of reactants in pores will influence the oxidation rate of material. At higher temperatures, the concentration gradient of the reacting gas becomes steeper within graphite and the gas concentration becomes zero at a distance R nearer the surface. The activation energy Ea in this zone amounts to half of the true activation energy Et. For graphite–steam reaction, this temperature region is characterized by a temperature range of 850–1350  C and graphite oxygen reaction, by 500–900  C. In the high-temperature zone (>900  C for graphite oxygen and >1250–1400  C for graphite–water steam) – zone III – the concentration of the reacting gas is low at the exterior of the solid and the rate is

controlled by step 1. As bulk gas-transfer processes have low activation energies, the apparent activation energy for gas–carbon reactions in zone III is low. The reactions occurring in the gas–graphite system are Reaction with oxygen ð
r H : standard enthalpy of reaction at 25 CÞ ½1

½2

Ln (reaction rate)

II

a


r H ¼
283:0 kJmol
1

COðgÞ þ 1=2O2 ðgÞ!CO2 ðgÞ

½3

Reaction with carbon dioxide Boudouard reaction : CðsÞ þ CO2 ðgÞ ! 2CO


r H ¼ þ 172:5 kJ mol
1 ½4

The equilibrium can be shifted with increasing CO pressure16,17 or in the presence of a catalyst. Reaction with water CðsÞþH2OðgÞ!COðgÞþH2ðgÞ


r H ¼þ131:3 kJ mol
1 ½5

COðgÞ þ H2 OðgÞ ! CO2 ðgÞ þ H2 ðgÞ
r H ¼
213:7 kJ mol
1

½6

The hydrogen and CO2 produced can then react with carbon

C þ H2ðgÞ ! CH4 ðgÞ

b


r H ¼
110:5 kJmol
1

CðsÞ þ 1=2O2 ðgÞ!COðgÞ

C þ CO2ðgÞ ! 2COðgÞ

III


r H ¼
394:5 kJmol
1

CðsÞ þ O2 ðgÞ ! CO2 ðgÞ


r H ¼ þ172:5 kJ mol
1

½4


r H ¼
71:81 kJ mol
1 ½7

The presence of hydrogen can shift reaction [5] left [4]. Reaction with hydrogen

I 1/ T

Figure 1 Ideal reaction zones in graphite: I – reaction rate is controlled by chemical reactivity of the sample; II – reaction rate is controlled by diffusion in pores; III – reaction rate is controlled by gas transport to the exterior surface of the sample; a and b are transition zones.

C þ H2ðgÞ ! CH4 ðgÞ
r H ¼
71:81 kJ mol

½7

The mechanism and kinetics of these reactions are described by Walker15. The approximate relative rates of gas–carbon reactions at 800  C and 0.1 atm. are given in Table 1.

Graphite Table 1 Approximate relative rates of gas–carbon reactions at 800  C and 0.1 atm. pressure Reaction

Relative rate

C–O2 C–H2O C–CO2 C–H2

1  105 3 1 3  10
3

In the literature, there are a number of investigations of nuclear graphite reactivity in different oxidation conditions. Results of oxidation of HTR-10 nuclear graphite IG-1118 exhibited three regimes: 400–600  C with an activation energy of 158.56 kJ mol
1, 600–800  C, at which the activation energy was 72.05 kJ mol
1, and the ‘third-zone’, over 800  C regime with a very low oxidation energy. The comparison of reactivity of the two types of graphite used in HTR in oxygen and air at 650–900  C (regime II) leads to the conclusion that there is no difference in the behavior of matrix graphite (A3-27) and standard graphite V483T during oxidation.19 At the same time, at a lower temperature (400  C, regime I) matrix graphite is more reactive with respect to air. For the temperature range 350–520  C, the activation energy Ea for A3-3 graphite matrix amounts to 110 kJ mol
1.20 The oxidation in air and moisture has to be considered for dismantling and interim storage, whereas the reaction with humidity and aquatic phases is important for final disposal. Several investigations into virgin and irradiated graphite have been carried out, mostly in air at ambient pressure. A comprehensive review was made by Stairmand,21 who concluded that significant graphite oxidation can be excluded at temperatures below 350  C. However, graphite oxidation in air can occur in high irradiation fields. Duwe et al.22 showed the consumption of oxygen and the production of carbon dioxide during the interim storage of HTR fuel pebbles in sealed cans. But the dose of the irradiation field from freshly irradiated fuel pebbles is normally not relevant to the interim storage or final disposal of i-graphite. 5.21.4.2.2 Graphite reactions with liquids

Graphite does not react with alkaline and acidic solutions if no oxidizing agent is present. Dissolved oxidants such as nitric acid, ozone, hypochlorides, and hydrogen peroxide attack graphite to different degrees.23–26 An important factor is the surface area, which depends mainly on pore volume and pore size.

545

The reaction with oxidizing agents, for example, concentrated nitric acid, leads finally to the evolution of carbon dioxide: C þ 4HNO3 ! 2H2 O þ 4NO2 þ CO2 However, different stable intermediate reaction products can be formed: graphitic oxide (C7H2O4), mellitic acid (C6(CO2H)6), and hydrocyanic acid (HCN). The yield of these products and carbon dioxide depends on the reaction conditions and the nature of the graphite material. Contact of i-graphite with aqueous phases during interim storage or final disposal cannot be excluded with an absolute certainty. In such a case, the oxidation of i-graphite depends mainly on the irradiationinduced production of highly reactive species by radiolysis of the aqueous phase and the accessible graphite surface. Corrosion experiments with A3-3 graphite show that the corrosion rate of graphite is increased in final repository relevant aqueous phases by external g-irradiation with a dose rate of 2 kGy h
1.27 The obtained corrosion rates are in the range from 10
5 down to 10
7 g m
2 day. High chloride concentrations accelerate the graphite corrosion probably by the formation of hypochlorides. This clearly indicates that irradiation-induced corrosion processes are relevant to the final disposal of graphite. However, this is an extremely high dose rate not relevant to the disposal of i-graphite. The first attempt to determine the relation between the dose rate and the corrosion rate was made in the framework of the European RAPHAEL project.28 However, the low number of performed measurements and the scattering of the obtained data did not allow the derivation of a validated data set for such a correlation.

5.21.5 Graphite Radioactivity The utilization of graphite in a reactor leads to two different types of radioactive contamination in the graphite material, the contaminants being  Activated impurities in the bulk graphite material  Radioactive isotopes occurring in the reactor circuit The activation products are more or less homogenously distributed in the graphite, depending on the original location of the impurities, as well as on the possibility of their migrating in the graphite by thermal gradients induced by the reactor conditions and repulse effects during the activation process itself. The radioactive isotopes from the reactor

546

Graphite

circuit are located (adsorbed) primarily at the graphite surface and migration into the bulk material requires a transport force, which could be a thermal gradient. The depleted isotopes have different origins:  Activation products of the coolant.  Impurities in the coolant.  Corrosion products from steel constructions of the reactor distributed in the coolant and activated in the reactor core.  Release of fission products from fuel elements with a cladding failure. These different sources of the radioactive contamination indicate that the activities of i-graphite depend on the reactor type, the type of the utilized virgin graphite material, and the operational conditions of the reactor. Therefore, even i-graphite of similar reactor types can show different contamination levels and different isotope ratios and a detailed characterization of i-graphite is required before retrieval from a specific reactor in addition to calculated radionuclide inventories. A good example of such an approach was the compilation of the so-called ‘Aktivita¨tsatlas des AVR’29 which was calculated and validated on the basis of radiochemical analysis of i-graphite from different locations in the reactor core. However, a detailed consideration of the different i-graphite materials from different reactor types and different graphite types will not be helpful. Furthermore, the amount of detailed information would definitely be out of the scope of this review and not indicate the significant general problems of the waste management of i-graphite. The dose rate, one of the key parameters for the retrieval and interim storage of i-graphite, depends mainly on the 60Co activity. 60Co has a half-life of 5.3 years. The main source of 60Co in i-graphite is the abrasion of fine metal parts from the pebble loop system, which has been built up in the pipes by neutron activation. Therefore, waiting for some decay periods can be helpful to reduce the dose per person for the workers at dismantling. Another important parameter for retrieval is the release of radionuclides into air. This could occur in the form of contaminated dust, which can be handled by adequate exhausting methods. More problematic is the release of tritium as gaseous component. However, it must be ensured that information specific to the reactor is retained during this period. For final disposal, 14C and 36Cl have been identified as key nuclides with respect to the long-term

safety, due to their long half-life, mobility, and biocompatibility. 5.21.5.1

Formation of 3H

The radionuclide 3H, tritium has a half-life of 12.3 years. The contribution of radioactivity in nuclear graphite resulting from tritium is significant.29–31 It is produced by the following reactions:  Fission reactions of uranium impurities in the graphite and fuel cladding failure, such as 235 U(n,f) 3H reactions.  Lithium impurities in the graphite via 6Li(n,a) 3 H reactions.  3He (n,p) 3H in HTR reactors, which utilize helium as coolant.  10B (n,2a) 3H reactions in absorber rods (negligible for designs without core rods). The chemical properties of tritium are essentially the same as those of ordinary hydrogen. Tritium generated from lithium impurities is produced mostly in graphite bulk. The release of tritium is controlled by its diffusion out of the grain boundaries and into the pore system. 5.21.5.2

Formation of 14C

Three routes, shown in Table 2, have to be considered for the formation of 14C. In the reactor core materials, nitrogen is present only as an impurity, whereas carbon and oxygen are in some cases major constituent elements of the coolant, moderator, or fuel. In spite of this fact, the 14N activation reaction is usually more important for 14C production due to its large cross-section. Therefore, the location and the chemical form of nitrogen are important for the location of the formed 14C. Nitrogen levels vary widely from 10 to 100 ppm in different reactor graphite types30 and sometimes they are not known very precisely. A comprehensive study of 14C has been carried out by Marsden et al.31 Calculations showed that about 70% of the 14C originates from nitrogen impurities with an assumed amount of 50 ppm by weight.

Table 2

Activation reactions generation 14C

Reaction

Abundance of isotope in natural element (%)

Capture crosssection (barn)

14

99.63 1.07 0.04

1.81 0.0009 0.235

N(n,p)14C C(n,g)14C 17 O(n,a)14C 13

Graphite

Another source of 14C is the oxygen pathway from the coolant. A birth ratio of 14C has been calculated for an AGR from 17O:14N:13C as 25:21:1, assuming 50 ppm nitrogen. Besides the level, the location of nitrogen impurity in reactor core materials is an important parameter. The nitrogen content of graphite is reduced during manufacture by several high-temperature treatment steps. However, the different heating and cooling processes cause the formation of cracks and closed pores, which could be filled with air. Therefore, the absorption of nitrogen on graphite surfaces as well as the nitrogen diffusion in the graphite matrix is one of the major parameters for the local distribution of 14C in i-graphite. Takahashi et al. reported that the nitrogen content in graphite depends on the surface area of the graphite and decreases from the surface to a depth of about 30 nm32 and that 14C produced from nitrogen will remain at its original position. This is in agreement with the 14C distribution of HTR fuel pebbles from the German AVR33 (Figure 2) and with the observation of a preferential release of 14C by surface oxidation of i-graphite from the German high-temperature reactor AVR in a nitrogen/steam atmosphere.36 Takahashi et al.32 reported that the kinetic energy of the formed 14C atom is about 470 kJ mol
1, which is in the range of covalent carbon bonding energies and therefore the 14C atom will be attached to nodes of the carbon lattice. They suggested that 14C will not be released by radiolytic oxidation of the graphite.

µCi g–1(C)

6

547

However, this is also a surface-related reaction and a similar release should be assumed as observed for thermal surface oxidation. This is in contradiction to Finn reporting a backscattering energy of 40 keV (4  106 kJ mol
1) for 14C formation.37 This would be significantly above any chemical bonding energy and would lead to movements in the lattice and the formation of new species. This high backscattering energy as well as the large number of displacements of carbon atoms during irradiation should lead to a more homogenous distribution of 14C. However, the displacements are in the range of 1–2 mm so that 500 displacements in one direction would be required for a transport of 1 mm. So generally, it can be assumed that 14C produced by activation of 13C is more or less homogenously distributed as opposed to 14C generated from 14N which is located in near-surface areas. However, it cannot be concluded in general that the 14C in i-graphite at the end of the reactor life is generated mainly by nitrogen activation. Surface oxidation of i-graphite irradiated in carbon dioxide during reactor operation could reduce surface-bound 14 C. This reaction, as well as low amounts of nitrogen impurities, could result in the remaining 14C inventory being generated mainly by activation of 13C. Activation calculation for the Bugey 1 reactor shows that the 13C activation leads to 96% of the measured inventory 14C in i-graphite and only 4% of the inventory must be addressed to nitrogen activation.38 This would also be in agreement with the results obtained by Pichon,39 which show a fast release of about 0.1% of the 14C inventory followed by a negligible leaching phase (Figure 3). A possible explanation could be a covalent bonding of 14C resulting from 13C activation in the graphite matrix and leachable 14C fraction from nitrogen activation loosely absorbed at the surface.

4

5.21.5.3

2

10

20

30

mm

Figure 2 Distribution of 14C in an high-temperature reactor (HTR) pebble fuel element. Adapted from Schmidt, P. Alternativen zur Verminderung der C-14-Emission bei der Wiederaufarbeitung von HTR-Brennelementen; Forschungszentrum Ju¨lich: Ju¨lich, 1979.

Formation of 36Cl

The dominant formation of 36Cl is by the neutron activation of 39K (2.2 barn), the main stable natural chlorine isotope with an occurrence of about 75%. Chlorine itself is used for the removal of metal impurities in graphite by the formation of volatile chlorides. However, residual amounts of chlorine remain in the graphite. Therefore, new cleaning methods for nuclear graphite avoid the utilization of chlorine. Furthermore 36Cl can be built by n,a-reaction of 39 K (4.3 mbarn) or from 34S via an n,g-reaction to 35 S (2.3 barn) followed by a b-decay to 35Cl. But these reactions have no significant relevance.

548

Graphite

Cumulative released fraction of 14C (%)

0.08

N ⬚8 – Lime water

0.07

N ⬚10 – Lime water

N ⬚9 – Pure water

0.06 0.05 0.04 0.03 0.02 0.01 0 0

50

100

150

200

250 300 Time (days)

350

400

450

500

Figure 3 Leaching behavior of 14C from French G2. Adapted from Pichon, C.; Guy, C.; Comte, J. Cl-36 and C-14 behaviour in UNGG graphite during leaching experiments, 2008.

Fraction of data less than concentration

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 Measured by NAA Inferred from 36Cl

0.2 0.1 0 0

0.2

0.4

0.6

0.8

1

1.2

Initial chlorine concentration (ppm) Figure 4 Initial chlorine concentration in Oldbury moderator graphite as measured by NAA and as inferred from 36 Cl activation. Adapted from Brown, F.; Palmer, J.; Wood, P. Derivation of a radionuclide inventory for irradiated graphite-chlorine-36 inventory determination. In IAEA Technical Committee Meeting on Nuclear Graphite Waste Management, Manchester, UK, 1999.

An investigation of core graphite from the Oldbury reactor shows the correlation of the initial chlorine impurities and the 36Cl inventory (Figure 4).40 Furthermore, the investigation shows chlorine loss during irradiation. This is explained by the release of chlorine from open pores and an activation of chlorine in the closed pores. However, radiolytic oxidation during operation will open the closed pores by graphite oxidation, which results in an additional release path for 36Cl (see Figure 5). Leaching experiments with French i-graphite from G2 showed that a major amount of 80–85%

will be leached from the graphite in 1 month. A further 5–10% will be leached in a period of about 1½ years (Figure 6).39 This is in agreement with the proposed chloride form of 36Cl located at water-accessible surfaces and its high solubility. A small part of 5–10% 36Cl remained in the graphite. This could be explained by 36Cl located in graphite areas that are not in contact with the leaching media, for example, closed pores, or by covalent bonds between the chloride and the carbon atoms of the graphite lattice. Further investigations are required to clarify the nature of the nonleachable 36Cl.

Graphite

Closed porosity

35Cl

Open porosity

Oxidation

Activation

549

Primary circuit

Release

35Cl

35Cl

Activation Oxidation

36Cl

Release

36Cl

36Cl

Figure 5 Schematic of activation and release of chlorine in graphite.

Cumulative released fraction of 36Cl (%)

100 90 80 70 N ⬚5 – Lime water N ⬚10 – Lime water N ⬚8 – Lime water

60 50

N ⬚6 – Pure water N ⬚9 – Pure water N ⬚2 – Pure water

40 0

100

200

300

400

500

Time (days) Figure 6 Leaching behavior of 36Cl from French G2. Adapted from Pichon, C.; Guy, C.; Comte, J. Cl-36 and C-14 behaviour in UNGG graphite during leaching experiments, 2008.

5.21.5.4 Diffusion of Radionuclides in Graphite Diffusion in polycrystalline graphite is a complex topic strongly related to the structure of the graphite. The three general diffusion types, listed in the order of increasing diffusion velocity, are:  Volume diffusion by movements of atoms due to the presence of lattice defects or exchange of lattice positions.41  Diffusion along grain boundaries.  Pore diffusion. All the three different diffusion types can occur in graphite: Volume diffusion in the graphite crystallites, grain boundary diffusion at the micropores between the crystallites, and pore diffusion in the pores between the graphite particles. Self-diffusion of carbon in graphite occurs at temperatures about 2000  C42 and may be important for central zones of HTR fuel elements.

Diffusion of fission products in graphite has been studied intensively with respect to radionuclide release from HTR fuel elements. All these processes become effective at higher temperatures and can be neglected at temperature ranges relevant to retrieval, interim storage, and final disposal. However, they might be interesting for decontamination processes, especially for tritium. Table 3 shows some diffusion coefficients measured for A3-3 graphite from HTR fuel pebbles, pitch coke (AS1-500), and petrol coke (AL2-500) after irradiation.43,44 The diffusion and release processes of radionuclides in i-graphite depend strongly on the nature of the graphite and especially on the anisotropy of the graphite.45,46 Tritium can be released from graphite more or less completely by thermal treatment under inert atmosphere at temperatures in the order of 1300  C.36

550

Graphite

Table 3

Diffusion coefficients of tritium nuclear graphite 

Type of graphite

Temperature ( C)

Diffusion coefficient, D0 (s
1)

A3a A3a A3a A3b AS1-500b AL2-500b

800 850 900 1000 1050 1025

1.72  10
9 9.09  10
9 6.89  10
8 8.18  10
9 9.83  10
11 1.83  10
10

a

Irradiation at temperatures < 100  C. Irradiation at temperatures 500  C.

b

5.21.6 Graphite Treatment for Disposal or Recycling 5.21.6.1 Waste Packages and Encapsulation Containers or drums are used as a packaging option for i-graphite, mainly for safe handling in the operational phase of waste management and not as a barrier for long-term safety aspects. No special designs of containers or drums have been made for i-graphite and standards from common waste management are applied. Therefore, this aspect is not covered in this chapter. However, it must be mentioned that graphite can act as a noble metal and accelerate the galvanic corrosion of stainless steel containers and measures should be implemented to isolate the graphite from the container or container materials with a guaranteed lifetime until final disposal should be used. Another problem may arise while filling the waste container, especially if more or less rectangular graphite bricks are filled into drums. The disposal costs depend on the classification and volume of the radioactive waste and not on the weight. Therefore different methods have been developed to achieve a high packing density but all methods will generate secondary wastes in the form of graphite dust. Bach et al. compared grinding, plasma cutting, jet cutting, wire sawing, and hydraulic breaking of graphite, especially with respect to the related release of graphite dust. The encapsulation aims at a safe enclosure of the waste by retardation or immobilization of radionuclides to avoid a release into the environment or at least to reduce the release to an unobjectionable level. Generally two options exist for encapsulation. Embedding in a matrix material. Impregnation of the graphite to fill the open pores.

The reference waste package concept for graphite waste envisages the embedding of i-graphite in cement pastes. The cement will establish an alkaline environment in the pore water, which will reduce the solubility of many key nuclides. Especially 14C will form insoluble carbonates if it is oxidized to CO2 by radiolysis processes. Further, the different cement phases combined with a large pore surface area will be able to absorb radionuclides or build insoluble secondary phases. On the other hand, the porous structure will not hinder the contact between aquatic phases and the waste and therefore a radionuclide release cannot be excluded, especially for 36Cl, which shows no significant retardation by cement. Alternative embedding materials could be glass, polymers or resins, bitumen, low-melting metals, or ceramics. The organic materials would all result in a dense waste package that protects the graphite from leaching. However, the production process and the handling are related to a potential fire hazard. Furthermore, the long-term stability could be less than the half-life of the key nuclides due to radiolysis or ageing processes and therefore water exclusion cannot be guaranteed for the final disposal time scale. Low-melting metals may have sufficient corrosion stability, which has not been sufficiently determined for disposal conditions. However, their own toxicity may create a problem. For example, the license for the German low-level waste underground disposal site Konrad allows only the disposal of 72 Mg tin, 539 Mg zinc and 33 400 Mg lead due to the water protection law of Lower Saxony. The vitrification of graphite will result in a wellknown product similar to vitrified high-level waste. Besides the known problems of the final disposal of high-level waste such as fracturing, the graphite may react with the glass melt and form dispersed gas pebbles (bubbles?), which is known from the embedding of coated particles from HTR fuel. The closing of the open pore system of graphite has been successfully tested by impregnation with bitumen, epoxy resins, and tar. Therefore, the graphite has to be evacuated and then immersed in bitumen or resin under high pressure at elevated temperatures to obtain a sufficient fluidity. Leaching tests with such an impregnated material have proved the high immobilization potential of this procedure. However, this would lead to problems similar to those already described for materials used as embedding. A new approach to fill the pore system of i-graphite is a process that can be classified between embedding and impregnation. It foresees the granulation of the

Graphite

i-graphite and a high-temperature compaction after mixing with glass in an amount equal to the pore volume. First attempts show a density of about 2.2 cm3 g
1, which is close to the theoretical density, and that the glass does not increase the total volume. Furthermore, this method would lead to volume-optimized waste packages because the produced block dimensions can be adjusted to the waste container dimensions. However, the proposed good leaching resistance and mechanical properties are yet to be demonstrated. 5.21.6.2

551

Centre, Ju¨lich (former KFA Ju¨lich). This development was related to the reprocessing of HTR fuel pebbles. Another process, based on inductive heating, has been developed by Westinghouse for graphite fuel compacts. However, the incineration of graphite would result in a total release of 14C as CO2 together with the bulk 12CO2, which may causes local increases of the 14C activity in the surrounding area of the incineration plant. Therefore no public acceptance could be achieved for such a graphite treatment option, even if the released 14C activity would be negligible in comparison with the natural 14C amounts. The trapping of CO2 is no alternative. Solidification of the CO2 as insoluble calcium carbonate from 1.2 tons of graphite (0.7 m3) would produce 10 tons CaCO3 (3.7 m3). However, such a process has the advantage that the 14 C has been transferred into a defined species and will have a more or less homogenous distribution. An advanced thermal treatment method has been developed first at the Research Centre, Ju¨lich. It was shown that the majority of the carbon 14C inventory could be removed from the AVR reflector and fuel graphite and graphite from the thermal column of the research reactor FRJ-1 by partial oxidation.34 The AVR Graphite was irradiated at a high temperature in an inert helium atmosphere and the other graphite at room temperature in an air atmosphere. The thermal treatment process for 14C separation was performed in nitrogen or argon plus 2% oxygen or humidified nitrogen or argon. First examinations by Podruzhina showed a 14C release of about 70% with a total graphite oxidation in the range of 20 to 30%.34 This results were

Thermal Treatment

The most effective volume reduction would be the complete oxidation of the i-graphite with a small ash residue which contains the nonvolatile radionuclides. Volatile radionuclides like tritium, 36Cl, or 137Cs may cause some problems but could be trapped from the off-gas and solidified for final disposal. (Tritium in the form of HTO could be used for the production of cement paste used as embedding material for radioactive waste.) Another problem is the incineration of nuclear graphite due to its chemical purity. The high thermal conductivity will conduct the heat from the surface into the bulk material, which inhibits incineration. The poor combustibility of graphite was shown by the first attempt of CEA, utilizing a coal stove. Therefore, the material must be crushed before incineration. Milling can be performed technically without dust release but requires great effort. The burning itself could be performed in furnaces or in fluidized bed reactors. The burning of crushed graphite has been demonstrated at the Research 100

Release rate (%)

80

60

14

C release rate; nitrogen + steam Total carbon release rate; nitrogen + steam C release rate; nitrogen + 2% oxygen Total carbon release rate; nitrogen + 2% oxygen 14

40

20

0 0

Figure 7

14

50

100

150 200 Time (min)

250

300

350

C release and total carbon oxidation by thermal treatment of Allgemeiner Versuchsreaktor graphite at 1230  C.

552

Graphite

confirmed by Jansen.35 Higher release rates were obtained by Florjan.36 Up to 60% of 14C will be released within the first 60 min followed by a slower release of 20–30% in the next 2–7 h (Figure 7). The best 14C release rates have been obtained at temperatures of about 1200  C, whereas the separation of 12C and 14C is better at lower temperatures (Tables 4 and 5). But the release rates of Florjan could not be repeated until now. Furthermore, these results show the different 14C release behavior of the different graphite types under similar treatment conditions. The best results have been obtained with AVR graphite. This is explained by the inhomogenous distribution of 14C with higher 14C concentrations on the surface and the existence of more reactive 14C containing species. This indicates that the irradiation conditions have an important influence on this process and that further investigations will show whether this process can be applied to CO2-cooled reactors or RMBK reactors. However, this process could be an alternative waste treatment option only when 5% of the graphite materials have to be oxidized and captured as CaCO3, if sufficient decontamination factors can be achieved with these graphite types. The removal of 14C from graphite has been considered the main problem in decontaminating graphite. However, the separation of radionuclides other than 14C has to be managed, which can be performed by different methods. Table 4

5.21.6.3 The Russian ‘Self-Propagating High-Temperature Synthesis SHS’ Graphite is homogenously mixed aluminum and titanium dioxide. The amounts are related to the following reaction: 3C þ 4Al þ 3 TiO2 ! 2 Al2 O3 þ 3TiC The exothermic reaction is self-propagating and only an initial start is required. The formed stable

14

C release by thermal treatment in N2 + 2% O2

Sample

R7

Origin

AVR

Treatment time (h) Temperature ( C) Total carbon release (%) 14 C release (%) Separation factor

Table 5

An option would be the complete incineration of the residual graphite, which would result in two more waste streams. The residual 14C including the CO2 stream could be released in the environment if sufficient decontamination rates can be achieved or fixated as CaCO3, which can be sent to a surface disposal site. The second waste stream will be a very small quantity of high active ashes and filter dust which must be disposed as high-level waste after an appropriate fixation. A typical volume reduction would be in the range of 160 for an incineration process for nuclear graphite.1 A second option would be direct disposal in a nearsurface disposal site. However, this would require a sufficient reduction of the 36Cl inventory (see Chapter 1.06, The Effects of Helium in Irradiated Structural Alloys), which has not been investigated yet.

R8

K8

K9

M3

M4

FRJ2

Reflector graphite

Fuel graphite

3 900 2.85 61.0 21

3 900 2.69 62.2 23

3 1230 2.94 78.8 27

3 900 1.87 43.2 23

3 1230 2.32 64.4 28

K5

K6

M2

MS2

3 1230 4.04 79.8 20

Thermal column

14

C release by thermal treatment in N2 + steam

Sample

R6

Origin of graphite

AVR

Treatment time (h) Temperature ( C) Total carbon release (%) 14 C release (%) Separation factor

R10

FRJ2

Reflector graphite

Fuel graphite

7 900 0.85 41.0 48

7 900 1.48 70.0 47

7 900 1.55 45.0 29

Thermal column 7 1280 5.40 92.6 17

7 900 4.12 69.8 17

7 900 0.02 49.0 2250

Graphite

titanium carbide contains 14C and the other radionuclides incorporated into the corundum and titanium carbide lattice. Additional confining additives can be added to the reaction mixture, for example, zirconium, which build even more stable crystalline phases with selected radionuclides such as uranium and plutonium. Furthermore, additives are used to improve the final product quality and to minimize the volatilization of radionuclides. Therefore, this process is also suitable for graphite contaminated with actinides from the Russian production reactors. The process requires a carefully controlled regime to minimize the radionuclide release. 5.21.6.4

Recycling of i-Graphite

The reuse of i-graphite may open a waste management route that has the potential to reduce the amount of i-graphite for disposal. The easiest attempt would be the direct use of i-graphite without any treatment for the production of new materials for the nuclear industry. Generally, it is known that used graphite can be recycled as additive material for graphite production. However, this cannot be done with i-graphite in the existing graphite production facilities. Even lowcontaminated graphite must be handled in a closed manufacturing unit to avoid the release of contaminated graphite dust. Furthermore, the amount of i-graphite suitable for direct reuse is small in comparison with the total amount of i-graphite. Therefore reuse will be associated with decontamination of i-graphite. The success of the decontamination depends on the achievable decontamination factors, especially of 14C. In principle, two methods could be proposed for decontamination: The wet method of leaching the graphite with suitable decontamination agents such as mineral acids or alkaline solutions. Decontamination by thermal treatment in steam or diluted oxygen atmosphere. Both options are under investigation in the European Carbowaste project.47 At the moment, there are not enough results from the leaching process to evaluate the feasibility of this method, whereas the thermal treatment mentioned in Section 5.21.6.2 has already proved its potential to remove 14C from the i-graphite matrix. Figure 8 shows potential routes to obtain feedstock material for graphite recycling. Two options can be considered after thermal

553

14

C decontamination. The first option is the total oxidation of the residual graphite to find out if the remaining 14C amount in the graphite would allow a free release of the off-gas into the atmosphere. The expected residues are in the range of a few kilograms per Mg graphite. The alternative is treatment with graphite cleaning methods known from the graphite industry to remove the residual nonvolatile radionuclides to a level that can be handled in further production steps. Potential products could be silicon carbide, waste additives, and feedstock material for new graphite materials in the nuclear industry. The production of materials will definitely be cheaper if fresh feedstock materials are used, but the benefit will be the reduced waste volume (see next section). Another interesting aspect of the separation of the 14 C is the option to replace 14C production as tracer material for scientific purposes by irradiation of nitrates.

5.21.7 Final Disposal Figure 9 shows the general routes for radioactive waste classification in different countries. Among the European Union states, the Belgian and French schemes are very similar and are closely related to the EU classification scheme, which is based on the general IAEA recommendations. These schemes formally recognize the lifetimes of the predominant radionuclides within waste packages, and segregate low- and intermediate-level waste into short-lived and long-lived categories, on the basis of whether the half-lives of these nuclides are less than or greater than 30 years respectively. Generally i-graphite can be assumed to be low- or medium-level radioactive waste by these regulations, whereby the classification for final disposal of i-graphite is determined mainly by the inventory of long-living radionuclides 14C and 36Cl. The high bio-compatibility and the good solubility of 36Cl if it occurs as chloride and therefore its high mobility require larger efforts to provide a safe enclosure from the environment. The French surface disposal site Centre de l’Aube has a total limit for the disposal of 0.4 TBq for 36Cl. Figure 10 shows a calculation of the 36Cl inventory of the stack of Bugey 1 plus some measured data. The graphite core of the reactor Bugey 1 would have a 36Cl inventory of about 0.1 TBq, assuming an average level of 50 Bq 36Cl g
1, which is probably too low as shown in Figure 10. Measured values for Bugey 1 reveal an

554

Graphite

i-graphite Disposal

Solidification

14

C depleted i-graphite

Partial oxidation in steam or diluted oxygen

14 C enriched off-gas in the form of CO or CO2

14

C separation

14

C products

Option 1 Total oxidation

Option 2

Free release of the off-gas (depending on the residual 14C inventory) Reconversion of CO and CO2 off-gas to carbon

Solidification and disposal

Further graphite cleaning by high temperature treatment may be in presence of halogens as decontamination agent

Other carbon-based products (e.g., SiC, lamp black, waste additives) Additive for new graphite products Figure 8 Process scheme of a potential process for thermal treatment of i-graphite.

average of about 200 Bq g
1. Furthermore 36Cl is easily leached, which has been discussed in Section 5.21.5.3. Therefore, a near-surface disposal of graphite is not an acceptable waste management option in France. The situation in the United Kingdom is similar. An estimate of the total 36Cl inventory is given by David Lever.48 It is in the range of 23 TBq for the British i-graphite. This 36Cl inventory will not allow the near-surface disposal at the Drigg site if the release could not be excluded over geological time scales.

A further aspect of UK reactors is the release of C if it is in the form of methane. Figure 11 shows a risk analysis of such a release if all the 14C is released as methane. The assumption will not be true for i-graphite; however, no quantitative results that give a clear figure about the relation between 14CO2, 14CO, and 14CH4 in the long-term release of 14C under disposal conditions are available. However, there is some evidence that organic 14C compounds cannot be neglected. Leaching of HTR fuel spheres shows a 14

Belgium

France VLLW < 100 Bq g-1

Cal A – low concentrations short half-lives (Criteria X and Y)

LLW Short-lived halflives < 30 years; activity between 100 and 105 Bq g-1 ILW Short-lived halflives < 30 years; activity between 105 and 108 Bq g-1

European Union

Transition waste

IAEA

United Kingdom

EW – Exempt waste

VLLW – less than 400 kBq of b/g activity per 0.1 m3 material

LILW-SL

LILW-SL

Short-lived half-lives < 30 years

Short-lived half-lives < 30 years

LLW Long-lived halflives > 30 years; Cal B – medium or activity between 100 LILW-LL long half-lives in and 105 Bq g-1 relatively high Long-lived half-lives ILW concentrations > 30 years Long-lived halfpower < 20 W m-3 lives > 30 years;

LILW-LL Long-lived half-lives > 30 years

activity between 105 and 108 Bq g-1 Cal C – substantial HLW; amounts of b- and activity between 108 a-emitters and 1010 Bq g-1 power > 20 W m-3

HLW

HLW

Figure 9 Comparison of radioactive waste classification schemes.

Spent nuclear fuel Waste with negligible heat generation

LLW – < 4 GBq t-1 of a and <12 GBq t-1 of b/g activity

Landfill/free disposal Surface disposal Geological disposal

High-level waste (HLW) similar to European definitions; arises mainly from manufacture of nuclear weapons Transuranic waste (TRU): radioactive waste containing more than 3.7 ´ 103 Bq g–1 (100 nCi g−1) of a-emitting transuranic isotopes with half-lives > 20 years nuclear weapons

ILW >4 GBq t-1 of a or >12 GBq t-1 of b/g activity, no heating consideration in storage HLW as ILW and with cooling in storage facilities

United States

Uranium mill tailings Naturally occurring radioactive material

Heat-generating waste

Low-level radioactive waste (LLW): by definition: everything else

Graphite

Generic disposal routes

Germany

555

556

Graphite 1.00E + 03

Bq g de 36Cl

1.00E + 02

Measures C2J6 C1K7 B319 C4H2 B8J0 C2L4 = C2G6 ABlB C3F9 D9Jl = A619 Risque = 50.00% Risque = 2.50% Risque = 0.10% Initial centre Initial + 2 sygma

1.00E + 01

1.00E + 00

36Cl measurements

Activation calculation results for each channel where the samples come from

1.00E – 01

11

13 15 17 19 21 Height (m) in the graphite stack (cooling gas flow direction)

23

Figure 10 Activation calculation results on BU1 stack with data assimilation method: 36Cl.

1E – 02 1E – 03 1E – 04

Total Magnox spheres

1E – 05 1E – 06 1E – 07

Uranium spheres Stainless steel spheres Carbon steel spheres

1E – 08

Organic degradation

1E – 09 1E – 10

Radiolysis organics Release from graphite Risk target

1E – 11 1E – 12 1E – 13 1E + 00 1E + 01 1E + 02 1E + 03 1E + 04 1E + 05 1E + 06 Years postclosure (postclosure starts calendar year 2150)

Figure 11 Radiological risk versus time for 14CH4 by contributory sources. Adapted from Lever, D. Graphite Wastes: Disposal Issues; Manchester, UK, 2006.

higher release of organic 14C than 14CO2, but as dissolved organic species and not as gaseous species.49 In Germany, radioactive waste is divided into two classes: waste with and without significant heat development. Deep underground disposal is planned for both types. The former Konrad iron mine has been designated as the disposal site for the nonheat developing waste and is proposed to be ready for waste disposal in 2013. The graphite from the reactor core of the AVR and THTR clearly belongs to the category of non heat developing waste and therefore could be disposed of at this site. However, the 14C inventory of about 2.9  1014 Bq for the ceramic core interior from the AVR will claim a major share of the licensed

C inventory (4  1014 Bq) of this disposal site and will limit the disposal of other radioactive waste. Furthermore, the actual interim storage stage will extend beyond the proposed operational time of this disposal site and therefore alternatives are required. 14

5.21.8 Summary A general solution for the management of i-graphite has not been established yet. Only France has an ambitious final disposal plan for its i-graphite, with the proposal of the near-surface underground disposal site at a depth of about 200 m. Other countries like

Graphite

the United Kingdom have not made a final decision for a reference waste management route until now. Three main challenges have been identified for the waste management of i-graphite. The first is the retrieval of the major amounts of i-graphite from the reactor cores. Some experience is available from the decommissioning of the BEPO, GLEEP, and Fort Saint Vrain reactors. However, no general methodology can be recommended because the retrieval depends on many site-specific factors. A major concern is the need for more data on i-graphite. This includes data on property changes of the structure and mechanical properties due to irradiation and the radionuclide inventory, as well as fundamental data concerning the behavior of i-graphite during treatment procedures, and disposal behavior. Future research should focus also on the speciation of the chemical form of radionuclides because the chemical form determines the long-term release behavior under final disposal conditions. Furthermore, the localization of the key nuclides 14C and 36Cl at a nano-scale is a major challenge because a near-surface distribution and a homogenous distribution in the bulk would lead to completely different release characteristics. Another challenge is the development of a safe waste management route. Generally, two principal methodologies could be utilized: the decontamination of i-graphite by chemical or thermal treatment to obtain a carbonaceous material for further use in nuclear technology or the final disposal of i-graphite. The most advanced plan for a final disposal has been achieved by France, which is planning an underground disposal site for low-level radioactive waste containing longliving radionuclides, especially 36Cl, 14C, and radium. Therefore, i-graphite will be grouted in drums or containers and disposed of afterward. This could be assumed as the actual reference concept. Other conditioning methods, which ensure the safe long-term enclosure of 36Cl and 14C as the Russian RSH method or a long-term stable sealing of the graphite pore system, for example, with glass,50 may be alternatives to enable a near-surface disposal of i-graphite from reactor cores.

4. 5.

6.

7. 8.

9. 10. 11. 12. 13.

14. 15.

16. 17. 18. 19. 20.

21. 22. 23.

References 1. 2. 3.

Marsden, B.; Wickham, A. Characterization, Treatment and Conditioning of Radioactive Graphite from Decommissioning of Nuclear Reactors; IAEA: Vieanna, 2006. Energoatom Concern OJSC > Nuclear Power Plants. Labyntseva, M. Russian Federation Country Report; Marseilles, France, 2008.

24. 25. 26. 27.

557

RBMK – Wikipedia. Russian Standards Catalog. GOST/R/50996/96/// Collection/storage/treatment/and/burial/of/radioactive/ waste/Terms/and/definitions/GOST/R/50996/96///Sbor/ khranenie/pererabotka/i/zakhoronenie/radioaktivnykh/ otkhodov/Terminy/i/opredeleniia/. Holt, G. The decommissioning of commercial magnox gas cooled reactor power stations in the United Kingdom (Magnox Electric plc. Berkeley Centre, Berkeley (United Kingdom)); IAEA-TECDOC–1043, pp 71–83. Mary Byrd Davis: Natural uranium graphite gas reactors (UNGG), Nuclear France: Materials and Sites, WISE-Paris, 2002. Bastien, D. (CEA Centre d’Etudes Nucleaires de Saclay, 91 - Gif-sur-Yvette (France) )HTGR Knowledge Base: Twenty-nine years of French experience in operating gas-cooled reactors; http://www.iaea.org/inisnkm/nkm/ aws/htgr/abstracts/abst_iwggcr19_16.html Hanford Cultural and Historic Resources Program. Hanford Site Historic District: History of the Plutonium Production Facilities, 1943–1990; Battelle: Columbus, OH, 2003. EC Project, Carbowaste, Contract FP7-211333, Workpackage 1 & 2. Wigner effect – Wikipedia, the free encyclopedia; http:// en.wikipedia.org/wiki/Wigner_energy, accessed on Mar 1, 2010. Windscale fire – Wikipedia, the free encyclopedia; http:// en.wikipedia.org/wiki/Windscale_fire, accessed on Mar 1, 2010. Wise, M. Management of UKAEA Graphite Liabilities. In Proceedings of the IAEA Technical Committee Meeting on Nuclear Graphite Waste Management; http://www.iaea. org/inisnkm/nkm/aws/htgr/abstracts/abst_manchester. html, 1999. Laine, N.; Vastola, F.; Walker, P. J. Phys. Chem. 1963, 67, 2030. Delle, W.; Koizlik, K.; Nickel, H. Graphitische Werkstoffe fu¨r den Einsatz in Kernreaktoren. Teil 2: Polykristaliner Graphit und Brennelementmatrix; Thiemig AG: Mu¨nchen, 1983. Hedden, K.; Lo¨we, A. Carbon 1967, 5, 339. Hedden, K.; Wicke, E. About some influences on the reactivity of carbon, In Third Biennial Carbon Conference 1957, S249–256. Xiaowei, L.; Jean-Charles, R.; Suyuan, Y. Nucl. Eng. Des. 2004, 227, S281. Moormann, R.; Hinssen, H. K.; Ku¨hn, K. Nucl. Eng. Des. 2004, 227, 281. Katscher, W.; Moormann, R.; Hinssen, H. K.; Stauch, B. ersuche zur Graphitkorrosion in Luft bei Temperaturen unterhalb 873 K, Vorschlag zur Modifizierung des Verbrennungs-Head-Ends fu¨r HTR-Brennelemente, Ju¨lich, 1984. Stairmand, J. Graphite Oxidation – A Literature Survey; AEA Technology Report, AEA-FUS-83, 1990. Duwe, R.; Bru¨cher, H.; Fachinger, J. R&D on Intermediate Storage and Final Disposal of Spent HTR Fuel; IAEA: Vienna, 1997. Gomez-Serrano, V.; Acedo-Ramos, M. A.; Lopez-Peinado, A. J. Thermochim. Acta 1995, 254, 249. Donnet, J. B.; Schutz, A. Carbon 1967, 5(2), 113–125. Donnet, J. B.; Ehrburger, P. Carbon 1973, 11(4), 309–316. Gmelin Handbuch der Annorganischen Chemie, Verlag Chemie GmbH, 1966. Fachinger, J.; den Exter, M.; Grambow, B.; Holgerson, S.; Landesmann, C.; Titov, M.; Podruhzina, T. Behaviour of

558

28. 29. 30. 31. 32.

33.

34.

35.

36. 37.

Graphite spent HTR fuel elements in aquatic phases of repository host rock formations. In Proceedings of the 2nd International Topical Meeting on High Temperature Reactor Technology, Bejing, China, 2004. EC Projekt ‘‘RAPHAEL’’; EC Contract 516508 (FI6O); Internal report D-B3.5. Aktivita¨tsatlas des AVR, Ju¨lich: WTI. Bush, R.; Smith, G.; White, I. F. Carbon-14 Waste Management; Office for Official Publications of the European Communities: Luxemburg, 1984. Marsden, B.; Wickham, A.; Hopkinson, K. The chemical form of Carbon-14 within graphite, Serco Assurance, 2002. Takahashi, R.; Toyahara, M.; Maruki, S.; Ueda, H.; Yamamoto, T. Investigation of morphology and impurity of nuclear-grade graphite, and leaching mechanism of carbon-14. In Proceedings of IAEA Technical Meeting, Manchester, UK, on Nuclear Graphite Waste Management, published on CD-ROM IAEA-NGWM/CD 01-00120, October, 1999. Schmidt, P. Alternativen zur Verminderung der C-14-Emission bei der Wiederaufarbeitung von HTR-Brennelementen; Forschungszentrum Ju¨lich: Ju¨lich, 1979. Podruzhina, T. Graphite as radioactive waste: Corrosion behaviour under final repository conditions and thermal treatment, Forschungszentrum Ju¨lich (Diss. RWTH Aachen 2004), Berichte des Forschungszentrum Ju¨lich 4166, 2004. R. Jansen, Thermische Dekontamination von aktiviertem Reaktorgraphit – Planung, Konstruktion und Test eines Ofensystems. Forschungszentrum Ju¨lich (Dipl. RWTH Aachen 2005), 2005. Florjan, M. Dekontamination von Nuekleargraphit durch thermische Behandlung (im Druck), RWTH Aachen 2009. Finn, R.; Ache, J.; Wolf, A. J. Phys. Chem. 1969, 73(11), S3928–S3933.

38. 39. 40.

41. 42. 43. 44. 45. 46. 47. 48.

49.

50.

Poncet, B. Personal communication. Pichon, C.; Guy, C.; Comte, J. Personal communication. Brown, F.; Palmer, J.; Wood, P. Derivation of a radionuclide inventory for irradiated graphite-chlorine-36 inventory determination. In IAEA Technical Committee Meeting on Nuclear Graphite Waste Management, Manchester, UK 1999. Mrowec, S. Acta Cryst. 1981, A37, 141–142. Sach, R.; Williams, W. Carbon 1974, S425. Fischer, P. G. Verhalten von Tritium in Reaktorgraphiten; Forschungszentrum Ju¨lich GmbH: Ju¨lich, 1975. Malka, V. J.; Raitz von Frentz, R. Adsorption und Desorption von Tritium an Graphit; Forschungszentrum Ju¨lich GmbH: Ju¨lich, 1978. Ashida, K.; Watanabe, K. J. Nucl. Mater. 1991, 183, 89–95. Saeki, M. J. Nucl. Mater. 1985, 131, 32. EC Project, Carbowaste, Contract FP7–211333, Workpackage 4. Lever, D. Graphite Wastes: Disposal Issues, Presentation at the University Manchester, UK, 2006; http://www.hse. gov.uk/aboutus/meetings/iacs/nusac/031006/ presentation1.pdf. Fachinger, J.; Zhang, Z. X.; Brodda, B. G. Graphite corrosion and hydrogen release from HTR fuel elements in Q-brine. In Proceedings of the International Conference on Radioactive Waste Management and Environmental Remediation, 5th, Berlin, Sept 3–7, 1995; Vol. 1, pp 637–640. Fachinger, J.; Hrovat, M.; Grosse, K.; Seemann, R. Impermeable graphite: A new development for embedding radioactive waste and an alternative option of managing irradiated graphite. In Proceedings of the Waste Management Symposium 2010, 10027, Phoenix, 2010.

Appendix 1 Amounts of Irradiated Graphite in Different Countries

Table A.1

Graphite-moderated reactors in Russia

Reactor

Graphite amount (Mg)

Scheduled shutdown

Kursk 1 Kursk 2 Kursk 3 Kursk 4 Leningrad 1 Leningrad 2 Leningrad 3 Leningrad 4 Smolensk 1 Smolensk 2 Smolensk 3 Total

2000 2000 2000 2000 2638 1798 1798 1798 2158 1798 1798 19 988

2021 2024 2013 2015 2018 2020 2009 + 20 years 2011 + 20 years 2013 2020

Graphite

Table A.2

559

Graphite-moderated reactors in United Kingdom

Location

Reactor

Type

Graphite in reactor (tons)

Graphite total (tons)

Dungeness Dungeness Hartlepool Hartlepool Heysham Heysham Heysham Heysham Hunterston Hunterston Hinkley Point Hinkley Point Torness Torness Bradwell Bradwell Calder Hall Calder Hall Calder Hall Calder Hall Chapelcross Chapelcross Chapelcross Chapelcross Dungeness Dungeness Hinkley Point Hinkley Point Oldbury Oldbury Sizewell Sizewell Wylfa Wylfa Berkeley Berkeley Hunterston Hunterston Trawsfynydd Trawsfynydd Windscale Winfrith Windscale Windscale Harwell Harwell Total

B-1 B-2 1 2 Unit I-1 Unit I-2 Unit II-1 Unit II-2 B1 B2 B1 B2 1 2 Unit 1 Unit 2 Unit 1 Unit 2 Unit 3 Unit 4 Unit 1 Unit 2 Unit 3 Unit 4 A1 A2 A1 A2 Unit 1 Unit 2 A1 A2 A1 A2 Unit 1 Unit 2 A1 A2 Unit 1 Unit 2 WAGR Dragon Pile 1 Pile 2 BEPO GLEEP

AGR AGR AGR AGR AGR AGR AGR AGR AGR AGR AGR AGR AGR AGR Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox Magnox AGR HTR Air cooled Air cooled Air cooled Air cooled

850 850 1360 1360 1520 1520 1520 1520 970 970 970 970 1520 1520 1810 1810 1164 1164 1164 1164 1164 1164 1164 1164 2150 2150 2210 3310 2061 2061 2237 2237 3470 3470 1938 1938 1780 1780 1900 1900 285 40 <2000 2000 766 505 65 080

ND ND ND ND ND ND ND ND ND ND ND ND ND ND

1931 1931 1630 1630 1630 1630 1630 1630 1630 1630 2237 2237 2457 2457 2090 2090 2240 2240 3740 3740 1650 1650 2150 2150 1980 1980 285 40 <2000 2000 766 505 77 000

560

Graphite

Table A.3

Graphite-moderated reactors in France

Location

Reactor

Type

Graphite in reactor (tons)

Graphite total (tons)

Marcoule Marcoule Marcoule Loyettes Avoine Avoine Avoine Orleans Orleans Total

G1 G2 G3 Bugey 1 Chinon A1 Chinon A2 Chinon A3 St. Laurent A1 St. Laurent A2

Air cooled UNGG UNGG UNGG UNGG UNGG UNGG UNGG UNGG

1200 1207 1207 2039 1050 2200 2530 2572 2440 16 445

1200 1207 1207 3600 1060 2500 4000 4240 4100 23 114

Table A.4

Graphite-moderated reactors in United States

Location

Reactor

Type

Graphite in reactor (tons)

Graphite total (tons)

Platteville, CO Peach Bottom, PA Hanford Hanford Hanford Hanford Hanford Hanford Hanford Hanford Hanford Savannah River Savannah River Oak Ridge Brookhaven Chicago Pacific North West Labs Pacific North West Labs Argonne National Laboratory

Fort Saint Vrain Peach Bottom B D F DR H C KW KE N SR-305 SP 8 GR (X-10) BGRR CP1 HTLTR HTR CP2

HTGR HTGR LWGR LWGR LWGR LWGR LWGR LWGR LWGR LWGR LWGR Test

ND ND 1080 1080 1080 1080 1080 1080 1080 1080 1080 ND ND ND 440 ND ND ND ND

ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND

Air cooled Test Test Air cooled United States Graphite pile

Graphite

Table A.5

Germany Belgium Italy Japan North Korea Lithuania Spain Ukraine

China

561

Graphite-moderated reactors in other countries Location

Reactor

Type

Graphite in reactor (tons)

Graphite total (tons)

Shutdown date

Juelich Uentrop Mol Borgo Sabotino Tokai Oarai Nyongbyon Visaginas Visaginas Hospitalet de l’Infant Chernobyl Chernobyl Chernobyl Chernobyl INET

AVR THTR 300 BR-1 Latina Tokai 1 HTTR Nyongbyon 1 Ignalina 1 Ignalina 2 Vandellos

HTGR HTGR Air cooled Magnox Magnox HTTR Magnox LWGR LWGR Magnox

225 300 472 2065 920 ND ND 1700 1700 2440

ND 300 472 ND 1600 ND ND 2000 2000 ND

1988 1989 S

Unit 1 Unit 2 Unit 3 Unit 4 Tsinghua HTR-10 ND ND ND

LWGR LWGR LWGR LWGR HTR

1700 1700 1700 <1700 111

2000 2000 2000 <2000 2000

1996 S 1991 S 2000 S 1986 S

Air cooled LWGR LWGR

ND ND ND

ND ND ND

Baotou Jiuquian Guangyuan

1987 S 1998 S

1990 S

5.22

Minerals and Natural Analogues

G. R. Lumpkin Australian Nuclear Science and Technology Organisation, Kirrawee, NSW, Australia

T. Geisler-Wierwille Universita¨t Bonn, Bonn, Germany

Crown Copyright ß 2012 Published by Elsevier Ltd. All rights reserved.

5.22.1 5.22.2 5.22.3 5.22.3.1 5.22.3.2 5.22.3.3 5.22.3.4 5.22.3.5 5.22.3.6 5.22.3.7 5.22.3.8 5.22.4 5.22.4.1 5.22.4.2 5.22.4.3 5.22.4.4 5.22.5 5.22.5.1 5.22.5.2 5.22.5.3 5.22.5.4 5.22.6 5.22.6.1 5.22.6.2 5.22.6.3 5.22.6.4 5.22.7 References Appendix:

Introduction a-Decay Damage in Minerals Oxides Pyrochlore Group Zirconolite Brannerite Perovskite Baddeleyite Crichtonite ABO4 and AB2O6 Minerals (B ¼ Nb, Ta, and Ti) Hollandite Silicates Zircon Thorite Titanite (Sphene) Allanite Phosphates Monazite Apatite Group Kosnarite and Related NZP Materials Xenotime Ore Deposits: Analogs for Spent Fuel Preamble General Aspects of Uraninite Alteration in Natural Systems Natural Fission Reactors in Gabon Uranium Migration in the Koongarra Ore Deposit Conclusions List of Mineral Names and Compositions Oxides Silicates Phosphates, Arsenates, Vanadates Carbonates, Fluorocarbonates, Fluorides Sulfides Uranyl Minerals (Section 5.22.6)

Abbreviations apfu AES AFM

Atoms per formula unit Auger electron spectroscopy Atomic force microscopy

BSE

DTA

564 565 565 565 570 574 575 578 578 579 580 581 581 585 585 586 587 587 588 590 590 591 591 591 592 593 594 595 599 599 599 599 599 600 600

Backscattered electron, an imaging mode in SEM, sensitive to mean atomic number Differential thermal analysis

563

564

Minerals and Natural Analogues

EDX EELS EPMA

EXAFS HLW ICP-OES IMF IR

Ln M–M M–O M–O–M MOX NZP RDF SEM Synroc Synroc-C Synroc-F TEM TOF-SIMS Urf6 vpfu XANES XPS XRD

Energy dispersive X-ray analysis Electron energy loss spectroscopy Electron probe microanalysis, quantitative X-ray analysis with crystal spectrometers Extended X-ray absorption fine structure High level waste Inductively coupled plasma-optical emission spectrometry Inert matrix fuel Infrared, a type of vibrational spectroscopy complementary to Raman spectroscopy Lanthanide series elements, La–Lu Metal–metal distance in a crystal or amorphous material Metal–oxygen distance in a crystal or amorphous material Metal–oxygen–metal angle Mixed oxide fuel, composed of uranium and plutonium oxide Sodium zirconium phosphate Radial distribution function, a way of describing M–O and M–M distances Scanning electron microscopy Synthetic rock Hollandite + perovskite + rutile + zirconolite-based material for HLW Pyrochlore + uraninite ceramic for partially reprocessed fuel Transmission electron microscopy Time-of-flight-secondary ion mass spectrometry A uranyl group with six equatorial oxygen atoms Vacancies per formula unit X-ray absorption near edge structure X-ray photoelectron spectroscopy X-ray diffraction

5.22.1 Introduction The immobilization and long-term disposal of nuclear waste is one of the greatest challenges that modern society faces today. Various types of high level waste (HLW) have been generated from nuclear operations around the world; for example, spent fuel from commercial nuclear power stations, liquid waste from the reprocessing of spent fuel, and waste from the

production of nuclear weapons and weapons grade plutonium resulting from nuclear disarmament treaties between the United States and Russia. Some plutonium, particularly in France, has been used in mixed oxide fuel (MOX, composed of uranium and plutonium oxide) in place of the standard uranium oxide fuel. Previous US policy adopted the strategy of a once-through fuel cycle followed by direct disposal of the spent fuel; however, recent changes have seen a move toward more effective use of uranium-based fuels in programs that combine reprocessing, transmutation, and separations technology in advanced fuel cycles (e.g., Generation IV nuclear power systems). Borosilicate glass is the currently accepted waste form of choice for many countries that reprocess their commercial spent fuel (see Chapter 5.18, Waste Glass), but there exists a significant fraction of ‘legacy’ waste and other nuclear materials that are very complex in physical form and chemical composition (e.g., the Na-, Al-, and Zr-rich waste stored in tanks at sites in the United States). These complex waste materials, together with impure plutonium, and the separated fission products and actinides generated from the various partitioning strategies may be better suited for existing or new types of high-performance crystalline waste forms or glass-ceramics (see Chapter 5.19, Ceramic Waste Forms). Some of these materials, for example, inert matrix fuels (IMFs), are being designed for recycling of reactor-grade plutonium and minor actinides in commercial power stations, followed by geological disposal, an attractive option that does not generate new plutonium.1 As envisaged by G.J. McCarthy, A.E. Ringwood, and others in the 1970s, there exist alternative crystalline waste forms that may be capable of providing a much higher level of chemical durability than borosilicate glass or directly disposed spent fuel. Many of these materials have been extensively developed over the previous 20–25 years, while others are relatively new. Materials such as tailored ceramics,2 the synthetic rock (Synroc) polyphase titanate waste forms,3,4 and related special purpose waste forms are reasonably well developed and have been the subject of extensive leach testing and radiation damage studies. Pyrochlore is the major component of Synroc-F, a polyphase ceramic designed for partially reprocessed nuclear fuel5 and later appeared as the principal host phase for excess weapons Pu and U in a crystalline titanate ceramic form. Zirconolite has also been proposed as an ideal host phase for actinides due to a combination of crystal chemical flexibility and very high durability in aqueous fluids,6 and hollandite

Minerals and Natural Analogues

may provide an excellent host material for separated long-lived radioactive Cs for similar reasons.7 Additional special purpose waste forms for actinides include zircon,8 monazite,9 and zirconium-based materials having the fluorite, defect fluorite, or pyrochlore structures.10,11 Except for zircon, none of these materials has been studied to the same extent as the titanate waste forms. Nevertheless, monazite and Zr-based materials are promising in view of their resistance to amorphization and excellent chemical durability. Nuclear waste form materials must meet several requirements in order to reach final consideration for use in a repository, including a high level of durability in aqueous fluids, crystal chemical flexibility allowing the material to cope with variations in the composition of the waste stream, reasonably high waste loadings, volume reduction, and reliable and cost-effective processing technologies. Information derived from minerals can be used to assess all but the latter criterion. Furthermore, studies of U ore deposits are useful in the assessment of the performance of spent fuel, including transport of U away from the repository. The purpose of this chapter is to summarize the performance of minerals in terms of their response to a-decay damage and their interactions with natural aqueous fluids in geological environments. For comparison, we also discuss some of the relevant literature results involving accelerated radiation damage of synthetic compounds doped with short-lived actinides, controlled laboratory experiments on the dissolution of synthetic materials with and without short-lived actinides, and dissolution of radiation-damaged natural samples. Following a brief introduction to uraninite and its alteration products in natural systems, we conclude with an overview that uranium ore deposits as analogs for spent fuel under repository conditions, including aspects of the natural fission reactors in Gabon and uranium migration around the Koongarra ore body, Northern Territory, Australia.

5.22.2 a-Decay Damage in Minerals In this section, we briefly summarize the effects of a-decay damage on the structures of some of the more important natural analogs. Here, it is important to point out that a-decay involves two concurrent processes, the release of a high energy (4–5 MeV) a-particle together with a low energy (70–100 keV) recoil atom. The process can be expressed in the following general way: A nþ ZP

ðn2Þþ 4 !A4 þ2 He2þ Z2 R

½I

565

In this expression, P is the parent isotope, R is the recoiling daughter isotope, and the charged He atom is the emitted a-particle. The symbols A, Z, and n represent the mass, atomic number, and nuclear charge, respectively. The massive recoil nucleus has a range of 20–25 nm and typically displaces on the order of 1000 atoms primarily by nuclear stopping processes; whereas, the a-particle has a range of about 10–15 mm and loses most of its energy through electronic interactions before displacing on the order of 100 atoms near the end of its track. The valence states of the a-particle and recoil atom are rapidly reconfigured in the solid to produce 4He and to return the recoil atom to a more stable state. This charge reconfiguration process may be complex in the case of U that can exist as the U4+, U5+, and U6+ ions in solids or in general if there are other elements present with variable valence states such as the transition metals (see Section 5.22.4.3). The important parent isotopes in minerals are 238 U, 235U, and 232Th. These isotopes decay to the stable isotopes 206Pb, 207Pb, and 208Pb through their respective decay series. Based on each decay series, the total a-decay dose D can be calculated using the following equation: D ¼ 8N238 ðel238 t  1Þ þ 7N235 ðel235 t  1Þ þ 6N232 ðel232 t  1Þ

½1

In this equation, t is the geological age, N represents the present-day concentration of the parent isotope, and l is the decay constant. This equation is strictly applicable to samples wherein the isotopic composition of the U has been determined. In situations where only the Th and U elemental concentrations have been determined, one may assume that the U isotopic composition consists of 99.28% 238U and 0.72% 235U. Alternatively, the second term in the equation may be ignored without major consequence, as the associated error is usually smaller than the uncertainty in the geological age. In this chapter, we give all dose values in units of 1016 a-decays per milligram (this is because 1  1016 a-decays per milligram is approximately equivalent to one displacement per atom in minerals).

5.22.3 Oxides 5.22.3.1

Pyrochlore Group

Pyrochlore is an anion-deficient derivative of the fluorite structure type with a doubled a cell

566

Minerals and Natural Analogues

parameter and change in space group from Fm3¯ m to Fd3¯ m.12–14 Minerals of the pyrochlore group conform to the general formula A2mB2X6wY1npH2O, where A represents cations in eightfold coordination, B represents cations in sixfold coordination, and X and Y are anion sites. The basic structural element of pyrochlore is the framework of corner-sharing octahedra. Within this framework, continuous tunnels exist parallel to the h110i directions. Both the A-site cations and Y-site anions are located in these tunnels. In synthetic systems, some A-site cation exchange capacity has been demonstrated in defect pyrochlores, in which the values of m in the general formula can be quite large. Most natural pyrochlores form under magmatic conditions in granitic pegmatites, nepheline syenite pegmatites, and carbonatites, or late-stage veins associated with these rock types. The composition of common pyrochlore usually approaches the stoichiometric form (Na, Ca, Ln, U)2(Nb, Ta, Ti)2O6(F, OH, O), but the structure type is extremely flexible in terms of the sheer number of elements that can be incorporated and is particularly amenable to the incorporation of actinides. Natural samples are known to contain up to 30 wt% UO2, 9 wt% ThO2, and 16 wt% Ln2O3, an important consideration for the issue of nuclear criticality. However, as shown by the general formula, the crystal chemistry of pyrochlore is complicated by the potential for vacancies at the A-, X-, and Y-sites (m ¼ 0.0–1.7, w ¼ 0.0–0.7, and n ¼ 0.0–1.0) and the incorporation of water molecules (p ¼ 0–2) in the vacant tunnel sites. The total water content of the natural defect pyrochlores may be as high as 10–15 wt% H2O (with speciation as both water molecules and OH groups). In a little known but classic paper, Krivokoneva and Sidorenko15 examined a suite of Russian pyrochlores using X-ray diffraction (XRD) methods. An analysis of the line broadening showed that strain increased from 0.0009 to 0.0035 as crystallite dimensions decreased from 100–120 nm down to 35–40 nm in the initial stages of damage, a decrease to 15 nm was observed in the latter stages of damage. These authors also carried out an analysis of the radial distribution function (RDF) of an amorphous sample and showed that there was no long-range order present beyond the second coordination sphere. However, peaks in the RDF representing the major M–O and M–M distances showed that the fundamental structural units (e.g., the coordination polyhedra) still existed in the amorphous state. Lumpkin and Ewing16 also used XRD to determine both the

beginning (Di) of the crystalline–amorphous transformation and the critical amorphization dose (Dc) for a large suite of pyrochlores from different localities. They showed that the transformation zone increased in dose as a function of the geological age of the samples. Both dose curves are well-described by an equation of the form: Di;c ¼ D0 etK

½2

In this expression, D0 is the intercept dose for Di or Dc and K is a rate constant. Analysis of the dose-age data gives a value of D0 ¼ 1.4  1016 a per milligram for the amorphization dose curve and K ¼ 1.7  109 year1.17 The Bragg peak intensities of a subset of these samples were fitted to an equation of the form: I =I0 ¼ eBD

½3

Here, I/I0 represents the total intensity of all observable Bragg peaks divided by the total intensity obtained from an undamaged sample of similar composition and B is a constant related to the amount of material damaged by each a-decay event. Equation [3] gives an excellent fit to the data with B ¼ 2.6  1016 mg per a-particle, corresponding to an average cascade radius of 2.3 nm in which a maximum of 2600 atoms are displaced. An analysis of line broadening in these samples showed that crystallite dimensions decreased from about 500 to 15 nm with increasing dose. Strain initially increased with dose and reached a maximum of 0.003 before falling to values below 0.0005 at higher dose levels, consistent with a description of the crystalline–amorphous transformation as a type of ‘percolation’ transition.18 With increasing a-decay dose, transmission electron microscopy (TEM) images reveal mottled image contrast due to strain, followed by the appearance of local amorphous domains that increase in volume and begin to overlap to produce larger amorphous areas until they are connected throughout the material. This is the first percolation transition. With further increases in dose, the crystalline areas diminish in volume until they become isolated, giving way to a microstructure dominated by amorphous pyrochlore.16 This is the second percolation transition. During the 1980s, Greegor and coworkers19–22 carried out several studies of the local structure and bonding around Ti, Nb, Ta, and U atoms in pyrochlore using EXAFS–XANES (extended X-ray absorption fine structure–X-ray absorption near edge structure). Results of these studies demonstrated that the M–O coordination polyhedra of amorphous pyrochlore exhibit reduced bond

567

Minerals and Natural Analogues

Zircon Apatite Pyrochlore Zirconolite

14 12 DV/ V0 (%)

distances, reduced coordination number, and increased distortion relative to the undamaged crystalline structure. Furthermore, there was no periodicity in evidence beyond the second coordination sphere, with some disruption of the M–M distances. From these studies, it was realized that only a slight increase in the mean M–M distance was required in order to explain the overall increase in volume caused by a-decay damage and that this could be facilitated by increased M–O–M angles. The thermal behavior of radiation-damaged natural pyrochlore was investigated using differential thermal analysis (DTA) and XRD.23,24 Results of this study indicated that the samples recrystallized in the range of 400–700  C, depending upon the composition and degree of crystallinity. Measured values of the recrystallization energy are 125–200 J g1 and are inversely correlated with the level of crystallinity. In the early to mid-1980s, Clinard and his colleagues conducted an extensive set of experiments on ‘cubic zirconolite’ CaPuTi2O7 in which the Zr is completely replaced by 238Pu (t1/2 ¼ 87.7 years). This material is actually a pyrochlore compound similar to synthetic CaUTi2O7. In their first publication, Clinard et al.25 reported that CaPuTi2O7 has a total volume expansion of 4.7%, an ‘apparent’ lattice volume expansion of 2.2%, and a critical amorphization dose of 0.3  1016 a per milligram based on XRD analysis. Further analysis of the data showed that CaPuTi2O7 exhibits a bulk volume expansion of 5.4% at ambient temperature and becomes amorphous at a dose of 0.5  1016 a per milligram based on bulk swelling curves. For samples held at 302  C, the bulk swelling saturates at 4.3%, and the material becomes amorphous at a dose of 1  1016 a per milligram as estimated from the swelling data. This is a very significant result, as it indicates that the critical dose for an experiment lasting 3 years at 302  C is roughly equivalent to what nature produces in 107–109 years. When stored at a temperature of 602  C, CaPuTi2O7 did not become amorphous; however, the material showed a bulk expansion of 0.4% consistent with accumulation of lattice point defects.26 In retrospect, this is a stunning result and represents the first and only realistic ‘bracket’ for the critical temperature for amorphization of a nuclear waste form material. Also during the 1980s, Weber et al.27 investigated synthetic Gd2Ti2O7 doped with 3 wt% 244 Cm (t1/2 ¼ 18.1 years) and determined the amorphization dose of 0.4  1016 a per milligram with B ¼ 4.4  1016 mg per a-particle, a total volume expansion of 5.1% at saturation (Figure 1), and

10 8 6 4 2

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Dose (1016 a per milligram) Figure 1 Plot showing total volume expansion of synthetic pyrochlore Gd2Ti2O7 doped with 244Cm, zirconolite CaZrTi2O7 doped with 244Cm, apatite (e.g., britholite) CaNd4(SiO4)3O doped with 244Cm, and zircon (ZrSiO4 doped with 238Pu) as a function of increasing a-decay dose.

an increase in fracture toughness together with a decrease in hardness and elastic modulus. Changes in microstructure with increasing dose mimic the results for natural pyrochlores described earlier. Based on the results of DTA experiments, an activation energy of Ea ¼ 3.8 eV was determined for recrystallization of this pyrochlore. A recrystallization temperature of 700–800  C was determined by isochronal annealing. The authors also performed leach tests on single-phase Cm-doped Gd2Ti2O7 pyrochlore samples. In this work, the leach tests were limited to annealed, fully crystalline and fully amorphous samples, and were exercised at 90  C in pure water for 14 days. The experiments revealed weight losses of 0.02% and 0.05% for the crystalline and amorphous pyrochlore samples, respectively. The results of this study also indicated that the leach rate of Cm increased by a factor of 17 as a consequence of amorphization. More recently, Strachan and coworkers28 investigated the effect of 238Pu on the structure of four synthetic pyrochlore samples with variable amounts of Al, Gd, Hf, and U. The results of detailed XRD and bulk swelling measurements indicate that the critical dose for amorphization is (0.2–0.4)  1016 a per milligram and is associated with a total volume expansion of <6%. Based on changes in the cubic lattice parameter with time, it appears that the unit

568

Minerals and Natural Analogues

cell expansion of the pyrochlore phase is on the order of 2.9–4.7%. Using purpose-built equipment for flow through dissolution tests, Strachan et al.28 examined the behavior of 238Pu-doped pyrochlore at pH ¼ 2–12 and 85–90  C. They found very low release rates based on Pu and to a lesser extent U, possibly due to solubility controls on these elements. Experiments carried out on amorphous and recrystallized samples demonstrated very similar release rates for Gd at pH ¼ 2 and 85  C. This is a very important conclusion of this major research project and, together with the observation that the materials did not develop cracks with increasing dose, lends substantial credence to the use of these ceramics as nuclear waste forms. Following a careful analysis of the effects of flow rate and specimen surface area, forward dissolution rates of (0.7–1.3)  103 g m2 day1 were obtained for two different samples. Other experiments have been conducted in order to determine the kinetics of U release from pyrochlore, (Ca,Gd,Ce, Hf,U)2Ti2O7, but without the complicating effects of short-lived actinides.29,30 Both studies report that the pH dependence follows a shallow v-shaped pattern with a minimum near neutral pH. The release rates for U, converted from the limiting rate constants given by Zhang et al.,30 range from 6  107 to 7  105 g m2 day1 for all experimental conditions (e.g., T ¼ 25–75  C and pH ¼ 2–12). Geisler et al.31–33 performed hydrothermal experiments with a natural, crystalline Ta-based pyrochlore (microlite) from Lueshe near Lake Kivu of the Democratic Republic of Congo in pure water and acidic solutions (pH ¼ 0) at 175 and 200  C. The hydrothermal treatment in the acidic solutions causes the partial replacement of the microlite by a new defect pyrochlore that is characterized by a larger unit cell volume, a large number of vacancies at the A-site (A ¼ Ca, Na) and anion vacancies, by molecular water, and possibly, OH groups. Analyses of the experimental fluid further revealed that U was lost to the solution. TEM investigations of the interface between the new defect pyrochlore and the unreacted microlite revealed a topotactic relationship between both pyrochlore phases. Furthermore, the interface between both phases was found to be sharp on the nanoscale with a sharp, step-like decrease of the Ca and Na content at the interface toward the defect pyrochlore. Time-of-flight-secondary ion mass spectrometry (TOF-SIMS) and confocal micro-Raman mapping of the defect pyrochlore produced in an acidic solution that was enriched with 18O (47.5 at.%) revealed that the defect pyrochlore is strongly

enriched with 18O with a sharp 18O gradient to relict unreacted areas. The authors suggested that the replacement of microlite by a defect pyrochlore occurs by a pseudomorphic reaction that involves the dissolution of the pyrochlore parent accompanied by the simultaneous reprecipitation of a defect pyrochlore at a moving dissolution–reprecipitation front; a process that has been named interface-coupled dissolution–reprecipitation process. It is noteworthy that the treatment in pure water for 14 days at 175  C did not produce reaction zones detectable by backscattered electron (BSE) imaging. However, significant spectral changes in powder infrared (IR) spectra of the reaction product and the detection of Na and Ca in the experimental solution indicated that the microlite has also reacted in pure water. The experimental, chemical, and textural alteration features bear a remarkable resemblance to those seen in naturally altered microlite samples (see Figure 2 and subsequent discussion). Later, Po¨ml et al.34 studied the hydrothermal alteration of a natural, heavily radiation-damaged pyrochlore (betafite) from a rare earth pegmatite from Lindvigskollen near Kragerø, South Norway and a synthetic titanate-based pyrochlore ceramic [(Ca0.76Ce0.75Gd0.23Hf0.21)Ti2O7] produced at the Lawrence Livermore National Laboratory, USA. The authors treated cuboids of both samples with edge lengths of 3.3 mm in a 1 M HCl solution containing 43.5 at.% 18O at 250  C for 72 h. During the experiments, both samples were transformed mainly into rutile with subordinate anatase. The degree of transformation was significantly higher for the natural radiation-damaged pyrochlore; for example, 18O was highly enriched with the reaction products of both samples with a sharp gradient (on a micrometer scale) toward the unreacted pyrochlore and no apparent diffusion profile. The replacement reaction retained even fine-scale morphological features typical for pseudomorphs. Based on these observations, the authors suggested that the dissolution of pyrochlore is spatially and temporally coupled with the precipitation of stable (metastable) TiO2 phases at an inwardly moving reaction front, a mechanism that is essentially the same as proposed for the experimental replacement of crystalline microlite by a defect pyrochlore as discussed in the previous paragraph. The authors pointed out that their results produced under relatively extreme batch-experimental conditions show similarities with nature as well as with results derived from experiments conducted under moderate conditions rather expected in a nuclear repository.

Minerals and Natural Analogues

(a)

(b)

20 mm Na+

BSE

(c)

(d)

Ca+

18O-/O-

0 Reaction rim / new pyrochlore phase Unaltered core / untreated microlite Altered natural microlite Unaltered natural microlite

A 0.00

0.25

1.00

0.25

0.50

0.75

0.00 1.00

A+

0.75

0.50

(e)

0.25

1.00

O–5 (f)

0.00

0.3

1.00

0.75

0.25

1.00 0.00 A+2

0.00

0.2

y

0.50

0.50

0.75

0.1 +

0.25

0.75

0.50

x

569

0.25

0.50

0.75

0.00 1.00

F

Figure 2 Experimental hydrothermal aqueous alteration of a natural pyrochlore (e.g., Ta-rich variety known as microlite) in highly acidic solution at 200  C. Backscattered electron image (a) indicates increased average atomic number in altered (brighter) areas, due to selective loss of the light cations Na (b) and Ca (c), and enrichment in 18O (d). Triangular diagrams also indicate that changes in cation (e) and anion (f) compositions are similar to those observed in nature.

Numerous investigations have demonstrated that natural pyrochlores are susceptible to alteration via reaction with aqueous fluids over a range of conditions involving pressure, temperature, and fluid composition. At higher temperatures (300–650  C, <400 MPa) in highly evolved late-stage magmatic fluids, Ca enrichment is commonly observed; whereas, the main effect of alteration at moderate temperatures under hydrothermal conditions (200–350  C, <200 MPa) is the loss of Na and F, often combined with cation exchange for Sr, Ba, REE, and Fe. Further removal of Na, F, Ca, and O may occur in low temperature hydrothermal or weathering environments, resulting in the maximum numbers of A-site, Y-site, and X-site vacancies, maximum hydration levels, and more limited exchange large cations such as K, Sr, Cs, Ba, Ce, and Pb in certain environments.35–45 Ti-rich pyrochlore (betafite) from hydrothermal veins (Figure 3) in the contact metamorphic zone adjacent to the Adamello igneous massif in northern Italy contain 29–34 wt% UO2 and are chemically the closest natural analogs presently known for nuclear waste forms. Electron microscopy and microanalytical work have revealed that these pyrochlore samples

Figure 3 Photograph showing dark-colored Ti-rich hydrothermal veins in marble from the contact metamorphic zone of the Adamello igneous massif in northern Italy. These veins have transported U, Th, and lanthanide elements as demonstrated by the presence of pyrochlore (e.g., Ti-rich type known as betafite) and zirconolite. Width of image ¼ 10 cm.

have only suffered a minor late-stage hydration event as evidenced by lower backscattered electron image contrast around the rims of the grains (Figure 4).

570

Minerals and Natural Analogues

of the U-rich rim. During this alteration, which is the result of exposure to tropical conditions, Na, Ca, and F were removed from the pyrochlore, thus leading to increased A-site vacancies (up to about 1.8 A-site vpfu). The alteration also led to localized redistribution of radiogenic Pb and to hydration, but U remained immobile. Although U loss has been documented in certain geological environments, it appears that U is highly stable on the A-site of the pyrochlore structure. As natural pyrochlore can accommodate U6+ in significant amounts, it is possible that the geometry of the A-site, which is similar to the Urf6 topology in uranyl minerals with two short bonds and six long equatorial bonds,49 plays a role in the geochemistry of pyrochlore. Figure 4 Backscattered electron image of pyrochlore (betafite) and zirconolite from Ti-rich veins, Adamello, Italy. Pyrochlore occurs as overgrowths on highly zoned zirconolite crystals. Note the slightly darker rim on the pyrochlore, due to minor loss of Na, F, and increased hydration. Zirconolite was unaffected by this alteration event. Width of image ¼ 60 mm.

Results of this study demonstrate quantitative retention of U and Th for time periods of 40 Ma, even though the crystals experienced cumulative total a-decay doses of 3–4  1016 a per milligram.46 In two samples from Bancroft, Ontario, Canada, Lumpkin and Ewing39 had previously concluded that the major result of alteration was hydration, with only minor changes in elemental composition, apart from the precipitation of galena due to mobility of radiogenic Pb. In contrast to these examples, the Ti- and U-rich pyrochlores from granitic pegmatites in Madagascar exhibit a range of alteration effects, including relatively hightemperature, postmagmatic hydrothermal processes, and lower temperature alteration.39,47 If the Ca content falls below 0.2–0.3 apfu, these Ti-rich pyrochlores may exhibit various levels of recrystallization to liandratite þ rutile (or anatase). In the most severe cases documented, this may be accompanied by a loss of up to 20–30% of the original amount of U and local redistribution of the radiogenic Pb. A recent study of a 440 Ma pyrochlore from Mozambique provides qualitative information on the effect of radiation damage on the alteration of pyrochlore.48 These pyrochlore crystals exhibit a distinct growth zoning, characterized by a U-free core and a U-rich rim (up to 17 wt% UO2). Following uplift and cooling, groundwater penetrated these fractured crystals and led to the deposition of clay minerals along both fractures and cleavage planes. This low-temperature process also led to chemical alteration of the pyrochlore but only within the zone

5.22.3.2

Zirconolite

The structure of zirconolite14 is also considered to be an anion-deficient derivative of the fluorite structure type and can be viewed as a volumetrically condensed, layered version of pyrochlore with reduced symmetry and several polytypic forms (monoclinic 2 M or 4 M, both with space group C2/c; orthorhombic 3O with space group Acam, hexagonal 3T with space group P3121). The chemical composition of zirconolite 2 M corresponds to CaZrTi2O7, but in nature, it commonly deviates from this end-member composition due to extensive substitution of Y, Ln, Th, and U for Ca and Nb, Fe, and Mg for Ti.50,51 In natural samples, Zr is subject to only limited substitution by other elements (e.g., minor amounts of Y, Ln, U, and Ti). Experimental work has shown that extensive substitution of REE, Th, and U generally results in a polytypic phase transformation from monoclinic 2 M to trigonal 3T or from monoclinic 2 M to monoclinic 4 M. Zirconolite is an important, highly durable host phase for actinides and fission products, with the ability to incorporate up to 24 wt% UO2, 22 wt% ThO2, and 32 wt% Ln2O3 in natural systems. Lumpkin et al.23,24 discussed the results of an extensive study of amorphous and annealed zirconolite from Sri Lanka using a variety of methods, including XRD, EXAFS–XANES, TEM, and DTA. Electron diffraction and high-resolution TEM studies suggested that amorphous zirconolite lacked periodicity beyond the second coordination sphere, consistent with a random network model of the amorphous state. EXAFS–XANES results provided more detailed information for the Ti- and Ca-sites and indicated that amorphous zirconolite lacked periodicity beyond the first coordination sphere, with reduced M–O

Minerals and Natural Analogues

bond lengths, reduced coordination number, and increased distortion of the Ti–O polyhedra (determined from a prominent pre-edge feature in the XANES results). Farges et al.52 provided additional results for the Zr-, Th-, and U-sites and reported nearly identical coordination numbers and bond lengths for the amorphous and annealed samples. However, a significant increase in the range of Zr–O and Th–O distances was observed, leading the authors to conclude that slight variation of the M–O–M angles can have a profound effect on long-range periodicity and medium-range order. Later work also confirmed the reduced coordination of Ti in the amorphous samples, pointing specifically to a fivefold coordination environment.53 Annealing studies of the amorphous zirconolites from Sri Lanka were performed using DTA and showed exotherms at 780  C due to recrystallization.23,24 For two different samples, the energy release associated with recrystallization was 43 and 48 J g1. TEM investigation of samples annealed at 1100  C showed that the structure recovered to monoclinic zirconolite2 M, but the material was highly twinned and contained some stacking faults and intergrowths of other polytypes on (001). A few polycrystalline grains were also observed, which gave electron diffraction patterns consistent with the fluorite structure type. From these observations, it was suggested that amorphous zirconolite initially recrystallizes to a disordered, defect fluorite structure. A detailed study of highly zoned zirconolite samples from the contact metamorphic zones of the Bergell intrusion (30 Ma) on the Swiss–Italian border and the Adamello igneous complex (40 Ma) in northern Italy provided the first detailed results on the crystalline–amorphous transformation in natural zirconolites.54 The crystals from both localities contain a wide range of ThO2 and UO2 concentrations up to a combined maximum of over 20 wt%, thus ensuring a substantial range of a-decay dose (see Figure 4). Due to the small grain size, the two suites of samples were characterized by analytical electron microscopy. In particular, a series of TEM dark field images revealed the percolation-like behavior with increasing dose: (1) the appearance of mottled diffraction contrast (0.08  1016 a per milligram), (2) extensive development of amorphous domains in a crystalline matrix (0.3–0.5  1016 a per milligram), and (3) overlap of collision cascades to produce larger amorphous areas that eventually dominate the structure as crystalline domains diminish in size to <10 nm (0.7–0.9  1016 a per milligram).

571

A study of seven suites of zirconolite samples ranging in age from 16 to 2060 Ma and with a-decay doses of 0.008  1016 to 24  1016 a per milligram has also been reported.55,56 For each suite of samples, the beginning of the crystalline–amorphous transformation (onset dose) was defined as the first appearance of mottled diffraction contrast in bright field images. The end of the transformation (critical amorphization dose) was defined by the complete disappearance of Bragg diffraction spots, leaving only diffuse rings at 3.0 and 1.8 A˚ in diffraction patterns. These dose ‘brackets’ were used to construct a plot of dose versus age, revealing a pattern of upward curvature with increasing age for both the onset dose and critical dose. As in the previous work on pyrochlore, this upward curvature was interpreted as evidence for long-term annealing of isolated a-recoil collision cascades back to the crystalline structure. The data were fitted using eqn [2] in order to determine the intercept dose and annealing rate constant. Curve fits gave values of D0 ¼ 0.11  1016 a per milligram and K ¼ 1.0  109 year1 for the onset dose curve; and D0 ¼ 0.94  1016 a per milligram and K ¼ 0.98  109 year1 for the critical dose curve. Due to potential dose rate effects and other factors, it is instructive to compare the work summarized earlier with actinide doping experiments. Weber et al.27 published a detailed body of work on radiation damage in CaZrTi2O7 doped with 3 wt% 244Cm. Synthetic zirconolite becomes amorphous at a dose of 0.5  1016 a per milligram with a total volume expansion of 6.0% at saturation (Figure 1) and B ¼ 4  1016 mg per a-particle for the amount of material damaged in the collision cascade. The crystalline–amorphous transformation in this material also leads to an increase in fracture toughness and a decrease in hardness and elastic modulus. DTA studies show that the fully amorphous material exhibits exothermic reactions at 500–530 and 680–700  C and releases about 13 and 114 J g1 of stored energy, respectively, in the two exotherms. Isochronal annealing indicated that the main phase of recrystallization occurs at 500–700  C; however, the zirconolite initially forms as a pseudocubic (rhombohedral) structure before transformation to monoclinic zirconolite at about 900  C. As shown in Figure 5, this material exhibits anisotropic lattice expansion with increasing dose (data from Wald and Offermann57). The a lattice parameter increases by 0.3% and exhibits saturation behavior at higher doses, the b parameter increases marginally by 0.1% at low dose and then remains nearly constant, whereas the c cell dimension

572

Minerals and Natural Analogues

1.25 0.729

1.248

b (nm)

a (nm)

1.249

0.728

1.247 0.727

1.246

(a)

(b)

1.035

1.155

1.145

Vc (nm3)

c (nm)

1.03 1.15

1.02

1.14

(c)

1.025

1.015

(d)

0.5

1

1.5

2

2.5

Dose (1015 a per milligram) Figure 5 Plots showing anisotropic lattice expansion in zirconolite (CaZrTi2O7) doped with 244Cm. As shown in (a) and (b), the a and b parameters exhibit normal ‘saturation’ behavior and increase by 0.32% and 0.14%, respectively. The c cell parameter does not show saturation and increases by 1.6% up to the maximum dose of the experiment. The total unit cell volume expansion (d) is 2.0%.

continues to expand with dose and reaches a maximum value of 1.5% relative to the undamaged zirconolite. In a concurrent study of synthetic zirconolite doped with 238Pu, Clinard et al.58 reported a total volume expansion of 5.5% and a critical amorphization dose of 0.5  1016 a per milligram. In a recent study of radiation damage via actinide doping experiments, Strachan and coworkers59 investigated the effect of 238Pu on the structure of three synthetic zirconolites that also contained variable amounts of Al, Gd, Hf, and U but no Zr. Although these are not pure single-phase specimens, detailed XRD and bulk swelling measurements indicate that the critical dose for amorphization is (0.3–0.5)  1016 a per milligram and is associated with a total volume expansion of 5%. By employing similar procedures to those used in their previous study of 238Pu-doped pyrochlore (Section 5.22.3.1), Strachan et al.59 showed that the forward dissolution rate of radiation-damaged zirconolite is 1.7  103 g m2 day1 at pH ¼ 2 and 90  C. The dissolution tests exhibited very little dependence on pH, were not dependent on

the level of radiation damage, and no cracking was observed in the zirconolite specimens. The dissolution of synthetic zirconolite without short-lived actinides has been determined as a function of pH using pure water in single pass flow through tests at temperatures of 75  C and lower.29,30 These authors have independently studied a Ce–Gd–Hf zirconolite containing about 16 wt% UO2 and the results of the two studies are similar. Release rates determined by Zhang et al.30 for Ti and U indicate that zirconolite dissolves congruently after about 20 days following an initial period where U is released at a somewhat faster rate than Ti. The limiting rate constants are equivalent to U release rates of 6.4  107 to 1.3  105 g m2 day2 for zirconolite over the entire pH range of 2–12 and temperature range of 25–75  C. The dissolution rate of zirconolite is characterized by a shallow v-shaped pattern with a minimum near pH ¼ 8, similar to the results obtained for pyrochlore. Isotopic age dating work by Oversby and Ringwood60 and electron microscopy studies by Ewing et al.61 have shown that natural zirconolite exhibits closed

Minerals and Natural Analogues

system behavior for U, Th, and Pb for up to 650 Ma with little, if any, evidence for geochemical alteration. More recently, Rasmussen and Fletcher62 proposed that zirconolite may become the principal mineral for age dating in mafic igneous rocks due to its ability to retain radiogenic Pb. Their analysis of 1200 Ma dolerite intrusive rocks from Western Australia demonstrated that zirconolite returned the same 207Pb/206Pb age as zircon and baddeleyite, but the zirconolite age was much more precise (by factors of 3.3 and 13, respectively). Lumpkin et al.17 and Hart et al.63 have described the alteration of amorphous zirconolite from the 2060 Ma carbonatite complex of Phalaborwa, South Africa in somewhat greater detail. Electron microprobe analyses, element mapping, and backscattered electron images demonstrate that the alteration is localized along cracks and resulted in the incorporation of Si and loss of Ti, Ca, and Fe. However, in these samples, the Ln, Y, Th, and U contents remained relatively constant across the alteration zones. Radiogenic Pb appears to have been mobile and precipitated mainly within the altered areas as galena. In carbonatites, zirconolite may be replaced along cracks and within micron-sized domains by an unidentified Ba–Ti–Zr–Nb–ACT silicate phase, suggesting that zirconolite may not be stable in the presence of relatively low temperature hydrothermal fluids enriched with aqueous silicate species.64,65 At higher temperature and pressure in magmatic, hydrothermal, or metamorphic systems, zirconolite may be altered by dissolution–reprecipitation or replacement mechanisms. Giere´ and Williams66 have described zoned zirconolites from Adamello, Italy, which exhibit corrosion and replacement by a new generation of zirconolite together with loss of Th and U to a hydrothermal fluid. Importantly, a thermodynamic analysis of the mineral assemblages was performed in this work and indicated that the zirconolite crystallization and corrosion occurred at 500–600  C in a reducing hydrothermal fluid rich in H2S, HCl, HF, and P and relatively poor in CO2. Pan67 has also described the breakdown of zirconolite to a new mineral assemblage consisting of zircon, titanite, and rutile in metamorphosed ferromagnesian silicate rocks at Manitouwadge, Canada. This reaction can be expressed as follows (modified slightly from Pan67): CaZrTi2 O7 þ 2SiO2 ¼ ZrSiO4 þ CaTiSiO5 þ TiO2 ½II This reaction illustrates the potential instability of zirconolite at high temperature and pressure in silica-rich

573

systems; however, the phase relations of zirconolite and other minerals in the system CaO–SiO2–TiO2–ZrO2– H2O–CO2 remain elusive, especially at low temperature and pressure. Malmstro¨m68 investigated the performance of several zirconolite compositions under hydrothermal conditions (150–700  C, 50–200 MPa) in fluids containing different concentrations of HCl, NaOH, H3PO4, silicate, or carbonate, in addition to pure water. Starting materials consisted of near end-member CaZrTi2O7, together with single-phase samples doped with Nd, Al, U, Ce, Gd, and Hf. The results of these experiments demonstrate that zirconolite is most highly reactive in the NaOH-bearing fluids, but temperatures in excess of 500  C are required to produce a continuous alteration layer consisting of perovskite þ calzirtite at 50 MPa or perovskite þ baddeleyite at 200 MPa. In the case of HCl, similar temperatures are required to produce an alteration layer consisting of rutile and anatase. Somewhat surprisingly, scanning electron microscopy (SEM) observations revealed that the silicate and carbonate fluids had no visible effect on the zirconolite surface after experimental runs at 550  C and 50 MPa. Only limited reaction was observed in pure water or H3PO4 fluids at the same temperature and pressure, with rutile and monazite appearing as products on the surface. Recently Po¨ml,69 experimentally investigated the hydrothermal alteration of a crystalline 239Pu-doped and an X-ray amorphous 238Pu-doped (D  7  1018 a-decay per gram) zirconolite ceramics with the composition Ca0.87Pu0.13ZrTi1.74Al0.26O7. A disk of each ceramic sample was treated in a Teflon# vessel with 2 ml of 1 M HCl at 200  C for 3 days under autogeneous pressure. The analyses of the experimental fluids by inductively coupled plasma-optical emission spectrometry (ICP–OES) revealed that significantly higher Ca, Al, and Pu concentrations were released into solution from the 238Pu-doped than from the crystalline 239Pu-doped sample. Optical and SEM investigations of the 239Pu-doped sample after the experiment revealed no signs of alteration, while the X-ray amorphous 238Pu-doped sample showed strong alteration features even under the optical microscope. Parts of the disc were covered by a ‘carpet’ of TiO2 crystals. Energy dispersive X-ray (EDX) analyses further showed that the uncovered areas lost Ca, Pu, and Al and have a composition close to ZrTiO4. Surface areas yielding the original composition could not be found. Such an observation indicates a diffusion-controlled leaching process from the X-ray amorphous 238Pu-doped zirconolite. However, further research is necessary before any

Minerals and Natural Analogues

conclusion about the alteration mechanism can be made. Regardless of what details constitute the alteration process, the comparative experiments clearly demonstrated that self-irradiation damage has a strong effect on the aqueous stability of zirconolite. In addition, further hydrothermal experiments were conducted with powders of different Ce-doped zirconolite ceramics (XCe ¼ 00.225 apfu) at temperatures between 100 and 300  C, with different surface-area-to-fluid volume ratios, and different solution compositions (1 M HCl, 2 M NaCl, 1 M NaOH, 35% H2O2, 1 M NH3, pure H2O). Experiments at temperatures >200  C were carried out in silver and nickel reactors, while for those at lower temperatures, Teflon vessels were used. The results of the different experimental series can be summarized as follows: (1) The alteration rate was insignificant for all solutions other than the 1 M HCl solution. A 1 M HCl solution was therefore used for all other experimental series. (2) Rutile (and anatase) and baddeleyite replaced the zirconolite grains to varying degrees in the 1M HCl solution, that is, zirconolite dissolution was found to be incongruent. (3) No clear correlation between the Ce-doping level and the degree of alteration could be observed. (4) The degree of alteration increased only slightly with increasing temperature. (5) The alteration rate was found to be independent on the surface-to-volume ratio. (6) Ag dissolved from the silver reactors dramatically increased the reaction rate, while Ni from the Ni reactors reduced the solubility of Ti and Zr in the HCl solution, indicating that background cations have a strong effect on the alteration rate. In summary, considering that only the 1 M HCl solution caused any significant alteration at temperatures between 100 and 300  C, crystalline zirconolite proved to be extremely stable in aqueous solutions. 5.22.3.3

Brannerite

The crystal structure of brannerite, ideally UTi2O6, is based on a distorted array of hexagonal closepacked oxygens. The structure is monoclinic, space group C2/m, and consists of layers of Ti octahedra connected by columns of U octahedra.70,71 Natural and synthetic brannerites can incorporate substantial amounts of Ca, Ln, Th, and other elements. In both cases, the incorporation of lower valence elements on the A-site may be charge-balanced by partial oxidation of U4+ to U5+ and/or U6+ ions.72,73 Lumpkin et al.74 have examined a small suite of brannerites from different localities by SEM–EDX, showing

Calibration B4 brannerite B10 brannerite B12 brannerite

0.7

M5/(M4 + M5)

574

0.68

0.66

0.64

4

4.5

5

5.5

6

Average U valence Figure 6 This diagram shows the electron energy loss spectroscopy branching ratios M5/(M4 + M5) obtained from the U M4,5 spectra of three natural brannerite samples. Calibration data indicate average valence states between about 4.3 and 4.8, for example, significant amounts of U5+ and/or U6+ are present.

that unaltered areas of the samples contain up to 7 wt% CaO, 8 wt% Ln2O3, 7 wt% PbO, and 15 wt% ThO2, together with minor amounts of Al, Si, Mn, Fe, Ni, and Nb. Using electron energy loss spectroscopy,73 measured the U–M4,5 spectra of three of these samples and reported average U valence states of 4.4–4.8 based on the individual M5/(M4 þ M5) branching ratios (Figure 6). Amorphous brannerite may contain nanocrystal inclusions of uraninite. All natural brannerite samples with ages greater than 20 Ma appear to be amorphous due to a-decay damage. Electron diffraction patterns of relatively unaltered areas of the brannerite samples typically consist of two broad, diffuse rings characteristic of amorphous materials. Unaltered natural brannerites have average a-decay doses of (2–170)  1016 a per milligram in part due to the very high U content.74 A partially crystalline brannerite from Binntal, Switzerland,75 allows the critical dose to be estimated at 1  1016 a per milligram, a value similar to geologically young pyrochlore and zirconolite samples. Geochemical alteration of brannerite is common and appears to increase in severity with geological age, although the P–T–X conditions are poorly understood.76 Alteration usually occurs around the rim of the brannerite or along cracks and may involve the formation secondary such as anatase, galena, and thorite. As reported for pyrochlore and zirconolite,

Minerals and Natural Analogues

the galena may precipitate due to the combined effects of radiogenic Pb migration and the presence of S species in the aqueous fluid. Altered brannerite typically loses U, and the concentration may fall to 1 wt% UO2 in the most heavily altered areas. The loss of U may be compensated in part by the incorporation of up to 18 wt% SiO2 and 16 wt% FeO, together with significant amounts of P2O5, As2O5, and Al2O3 in certain examples. In some cases, the associated rock or mineral matrix surrounding the brannerite may be highly fractured, providing pathways for the migration of U, evidenced by the precipitation of secondary U minerals.74 Several important experimental studies of synthetic and natural brannerite have been conducted and are summarized as follows: For temperatures of 20–50  C and pH values in the range of 2–12, Zhang et al.30 reported limiting rate constants for U equivalent to dissolution rates of about 103 to 105 g m2 day1. The lowest dissolution rate was obtained at 20  C and pH ¼ 5.6. However, the pH dependence was examined in detail at 70  C, and the results show a shallow v-shaped pattern similar to that of pyrochlore and zirconolite, although the release rates of U from these two phases are 1–2 orders of magnitude less than that of brannerite. In further detailed work, Zhang et al.77 demonstrated that the U release rate is strongly dependent on bicarbonate concentration and that bicarbonate does not interact strongly with titanium, either on the solid surface or in solution. The dissolution of brannerite is incongruent and a preferential release 100

U/ Ti atomic

pH = 2 pH = 11

10

1

0.1

5

10

15 20 Time (days)

25

Figure 7 Experimental aqueous dissolution of synthetic brannerite showing how the U/Ti atomic ratio in solution changes with time in acidic (pH ¼ 2) and alkaline (pH ¼ 11) solutions. The experiment at pH ¼ 2 mimics the U loss observed in natural brannerite samples.

575

of uranium occurs at pH ¼ 2 (Figure 7). TEM examination after the experiment identified a relatively small amount of secondary TiO2 (anatase brookite) containing variable amounts of U and trace amounts of other elements. Overall, the dissolution of the brannerite at pH ¼ 11 is nearly congruent, and TEM examination after the experiment shows that the sample develops large areas of an amorphous secondary phase. X-ray photoelectron spectroscopy (XPS) analyses indicate the existence of oxidized U5+ and U6+ species on specimens both before and after leaching, and U6+ was the dominant component on the specimen leached in the pH 11 solution. Zhang et al.78 also conducted thermal annealing studies of amorphous brannerite in Ar at 500, 700, 900, and 1100  C followed by batch dissolution experiments in solution at pH ¼ 4 at 30  C. XRD analysis indicates that the major recrystallization of the sample occurs at 900–1100  C, but detailed TEM examination revealed that partial recrystallization started at 700  C. Analysis of the starting material showed that it contained small amounts of Al and Si due to alteration and nanocrystals of UO2+x within the amorphous matrix. Interestingly, the authors found that the dissolution rate increases with annealing temperature, and this is attributed to the growth of UO2+x and to the formation of an aluminosilicate glass phase at the highest temperature. Relative to synthetic brannerite,30 the dissolution rate of the amorphous brannerite is about one order of magnitude higher. 5.22.3.4

Perovskite

Perovskite is an ABX3 structure type based around a framework of corner sharing, octahedral B-site cations with B ¼ Ti, Fe, and Nb, together with minor amounts of Mg, Al, Zr, and Ta, depending on bulk rock chemistry and mineralogy. The large A-site cations occupy the center of a large cavity formed by 8 B-site octahedra and are coordinated to 12 oxygens in the ideal cubic structure. Most perovskites are distorted via octahedral tilting and generally have lower symmetry. In nature, only near end-member SrTiO3 is cubic, most other compositions are orthorhombic.79 The major A-site cations in natural perovskites are Na, Ca, Sr, and REE, with minor amounts of K, Ba, and U. The concentration of Th is usually low, but may reach levels as high as 18 wt% ThO2 in certain alkaline rocks.80,81 Only a limited amount of information is available on radiation damage in natural perovskite. In one study, TEM methods were used to examine the

Minerals and Natural Analogues

microstructure and electron scattering properties of a large suite of samples having a range of geological ages.55,56 Perovskite grains containing 1.4–3.1 wt% ThO2 indicate that the beginning of the transformation is at a dose of 0.3–0.6  1016 a per milligram for t ¼ 295–520 Ma. A preliminary analysis of the data using eqn [2] gave D0 ¼ 0.19  1016 a per milligram and K ¼ 2.3  109 year1. Perovskite from Bratthagen, Norway, contains up to 6.0 wt% ThO2, equivalent to a-decay doses of (0.8–1.2)  1016 a per milligram. These grains are partially damaged based on the appearance of a weak diffuse ring in electron diffraction patterns. During this study, it was discovered that the perovskite crystals contain a hydrothermal alteration phase known as lucasite. Although lucasite has a slightly higher Th content relative to the associated perovskite, it is completely electron diffraction amorphous at a dose of (1.5–1.7)  1016 a per milligram, suggesting a lower radiation tolerance than the host perovskite. Th-rich perovskites from the Khibina alkaline complex, Russia, have been studied by electron probe microanalysis (EPMA) and XRD.80,81 The crystals are zoned and contain 2.3–18.5 wt% ThO2, increasing from core to rim. XRD results indicate that the cores (2.3–7.4 wt% ThO2) retain some crystallinity, but the rims (8.7–18.5 wt% ThO2) are completely amorphous. These data are consistent with a critical dose of 2  1016 a per milligram based on the maximum and minimum Th contents of the cores and rims, respectively. There have also been very few laboratory studies of radiation damage in perovskite by doping with short-lived actinides. Mosley82 studied CmAlO3, in which about 95% of the actinide content is 244Cm, and determined that the material is amorphous by XRD after 8 days of the experiment, from which we estimate a dose of 0.15–0.2  1016 a per milligram (equivalent to 0.2–0.3 dpa). Although the critical dose may be slightly higher than the X-ray value, it is still substantially lower than the value estimated for natural perovskite. Using Mosley’s lattice parameter versus time data, we have plotted the unit cell volume expansion, DVc/Vc0, as a function of the estimated dose (Figure 8). This plot indicates that the lattice volume expansion may be as high as 8.8% at saturation based on the curve fit; however, the last data point indicates a lattice volume expansion of 6% prior to amorphization. The latter value is close to the total volume swelling observed in pyrochlore and zirconolite doped with 238Pu or 244Cm. Perovskite commonly releases Ca in aqueous fluids even at low temperature, breaking down to

0.06 0.05 DVc / V0

576

0.04 0.03 0.02 0.01 0.05 Dose (10

0.1 16

a per milligram)

Figure 8 Unit cell volume expansion of CmAlO3 plotted as a function of a-decay dose. The total expansion at saturation is 8.8%; however, the highest measured expansion is 6.0%. These data suggest that the total macroscopic swelling could be in excess of 10 vol.%.

one or more polymorphs of TiO2 (generally anatase brookite or TiO2–B). This is well illustrated by the alteration of perovskite to anatase, cerianite, monazite, and crandallite group minerals during severe weathering of carbonatites in Brazil.83 Using electron microscopy, Banfield and Veblen84 have proposed that the perovskite–anatase reaction mechanism involves topotactic inheritance of layers of the perovskite Ti–O framework. In hydrothermal systems, Mitchell and Chakhmouradian80,81 have described the alteration of perovskite to kassite, anatase, titanite, calcite, and ilmenite in the presence of a CO2- and SiO2-rich fluid phase at temperatures of 250–600  C in alkaline rocks. These observations are consistent with the thermodynamic properties of perovskite and related minerals. Nesbitt et al.85 have shown that perovskite is unstable with respect to titanite, rutile, calcite, and quartz in many hydrothermal fluids and groundwaters at 25–300  C. Lumpkin et al.55,56 and Chakhmouradian et al.86 described the hydrothermal alteration of Na-bearing perovskite (ideally Na0.5Ln0.5TiO3) in alkaline igneous rocks. The primary result of this alteration is removal of Na from the original perovskite, producing lucasite, LnTi2O6x(OH,F)xH2O. Lucasite is isostructural with kassite, CaTi2O4(OH)2H2O, another alteration product of perovskite (CaTiO3). The simplified reaction relationships for perovskite, kassite, and lucasite are shown subsequently: 2Hþ þ H2 O þ 2CaTiO3 ¼ CaTi2 O4 ðOHÞ2  H2 O þ Ca2þ

½III

Minerals and Natural Analogues

Hþ þ H2 O þ 2Na0:5 Ln0:5 TiO3 ¼ LnTi2 O5 ðOHÞ H2 O þ Naþ

½IV

In both cases, the replacement product is usually reported as having a distinctly fibrous or prismatic morphology, a feature that can be observed by SEM (Figure 9). In the following paragraphs, we will discuss some of the important experimental data on perovskite dissolution. Thermodynamic calculations and data for natural groundwaters and hydrothermal fluids (up to 300  C) revealed that perovskite is generally unstable with respect to titanite, titanite þ quartz, rutile, or rutile þ calcite.85 Measurements of the dissolution rates of two natural perovskites and synthetic SrTiO3 and BaTiO3 samples were obtained in pure water at 25–300  C, indicating that elemental release rates are 101 to 103 g m2 day1 for Ca, Sr, and for Ba. Kamizono et al.87 have also examined Ce, Nd, and Sr-doped CaZrO3 in acidic (HCl, pH ¼ 1) and near neutral (deionized water, pH ¼ 5.6) solutions at 90  C. In the acidic solution, the dissolution rates of the impurity elements were near 0.1 g m2 day1; whereas Ca and Zr were released at rates on the order of 103 and 103 g m2 day1, respectively. Leach rates were about two orders of magnitude lower in the experiment using deionized water. 6729

577

Surface analytical studies of synthetic perovskite after leaching at 150–250  C in silica-saturated aqueous fluids were reported by Myhra et al.88 These authors, using a combination of Auger electron spectroscopy (AES), SEM–EDX, and XPS techniques, identified the presence of a surface reaction layer ranging in thickness from a few monolayers to several hundred nanometers, depending on the leaching conditions. The alteration layer was composed mainly of crystalline TiO2, a thin siliceous layer, and possible calcium carbonate or hydroxide species. Following this work, Myhra et al.89 reported additional results for CaTiO3 and BaTiO3 leached at 300  C in pure water. They determined release rates after 14 days of 1.4  102 g m2 day1 for Ca and 3.3  102 g m2 day1 for Ba from the two samples, respectively. Surface analytical work showed that Ca and Ba were depleted to a depth of about 200 nm and that oxygen was enriched near the surface, consistent with the release of Ca and Ba to solution and the formation of a crystalline TiO2 layer. The studies noted earlier were extended to lower temperatures of 20–100  C by Pham and coworkers,90,91 who showed that near surface decreases in the Ca/Ti ratio determined by XPS are accompanied by the formation of an amorphous Ti-rich layer up to 10-nm thick, as observed by TEM. The authors proposed a base catalyzed hydrolysis and ion exchange model to account for their observations, whereby surface Ca2+ is released to solution via exchange with H+ and Ti–O–Ti surface species are converted to Ti–OH species via reaction with OH and H2O. The overall reaction can be written as follows90: CaTiO3 þ ð6  xÞHþ ¼ Ca2þ þ TiðOHÞxð4xÞþ þ ð3  xÞH2 O

50 mm Figure 9 Example of hydrothermal alteration of natural perovksite in alkaline rocks from Bratthagen, Norway. This backscattered electron image shows partial replacement of a large perovskite (e.g., Na-rich variety known as loparite) single crystal by a brighter, fibrous secondary phase, lucasite.

½V

This reaction indicates that the aqueous Ca2+/H+ ratio and the presence of Ti–OH surface species control the dissolution of perovskite. The presence and buildup of the amorphous Ti–OH surface film at low temperatures may be due to kinetic factors, as crystalline anatase or rutile are thermodynamically favored at low temperatures.90 McGlinn et al.92 have investigated the pH dependence of the release of Ca from two perovskite samples: endmember CaTiO3 and Ca0.78Sr0.04Nd0.18Ti0.82Al0.18O3. Results of this study, performed at 90  C with pH ranging from 2.1 to 12.9, demonstrated that the Ca release rates generally decrease with increasing pH. After 43 days of leaching, the Ca release rate for the end-member perovskite decreased from

578

Minerals and Natural Analogues

8.9  102 g m2 day1 at pH ¼ 2.1 to 2.2  103 g m2 day1 at pH ¼ 12.9. Similar results were obtained for the doped perovskite. SEM imaging clearly showed the development of ‘agglomerated, submicron, titanaceous particles’ on the perovskite surface after the dissolution experiments performed in acidic aqueous solutions. In a more recent study by Zhang et al.,93 thermally annealed perovskite (CaTiO3) surfaces were characterized by SEM, TEM, XPS, and atomic force microscopy (AFM) techniques before and after aqueous dissolution testing in deionized water at room temperature, 90  C, and 150  C. Results of this work demonstrated that, although mechanical damage caused higher Ca release initially, it did not affect the long-term Ca dissolution rate. However, the removal of surface damage by annealing did lead to the subsequent spatial ordering of the alteration product, which was identified as anatase (TiO2) by both X-ray and electron diffraction, on CaTiO3 surfaces after dissolution testing at 150  C. The effect of Ca2+ in the leachant on the dissolution reaction of perovskite at 150  C was also investigated, and the results suggest that under repository conditions, the release of Ca from perovskite is likely to be significantly slower if Ca2+ is present in groundwater. 5.22.3.5

Baddeleyite

The monoclinic form of ZrO2, baddeleyite is the only natural analog for the proposed cubic zirconia waste forms. A recent literature review by Lumpkin94 shows that, even though baddeleyite is rather widespread as a trace phase in natural systems, it has a limited composition range of 87–99 wt% ZrO2, with most of the remainder comprised of FeO, TiO2, and HfO2. Natural baddeleyite has a distorted fluorite structure due to the low concentrations of other large cations such as Ca and REEs. The total concentration of Th and U is usually <2000 ppm by weight; nevertheless, the cumulative a-decay dose is in the range of (0.1–1.1)  1016 a per milligram for samples with ages of 1100–1200 Ma. Natural samples of baddeleyite appear to remain crystalline even at the highest observed a-decay dose levels. A case study of baddeleyite from the Jacupiranga carbonatite complex of southern Brazil shows that the mineral can incorporate up to 4.1 wt% Nb2O5 and 1.2 wt% Ta2O5.94 Incorporation of Nb5+ and Ta5+ is partially compensated by the incorporation of up to 0.4 wt% MgO and 0.3 wt% FeO in a charge-balanced substitution of the form 3Zr ¼ 2(Nb, Ta) þ (Mg, Fe).

These samples are also resistant to hydrothermal alteration which affected associated pyrochlore crystals, even though the baddeleyite crystals received a-particle doses as high as (3–4.5)  1016 a per milligram along common grain boundaries with a U-rich pyrochlore phase.94 These observations are consistent with results from hydrothermal experiments on zirconolite ceramics, wherein zirconolite is often replaced by baddeleyite.68,69 5.22.3.6

Crichtonite

Minerals of the crichtonite group conform to the general formula A1xM21O38 and are generally found in mafic–ultramafic and granitic igneous rocks. In these minerals, A ¼ Na, Ca, Sr, Pb, and the larger Ln elements (e.g., La, Ce, Pr, and Nd) and M ¼ Ti with variable amounts of Mg, Al, V, Cr, Mn, and Zr. They are of some interest here due to the ability to incorporate substantial amounts of U in the mineral davidite. Based on a review of the literature, Gong et al.95 proposed that this structure type may be an effective host phase for a range of fission products and actinides in high-level nuclear waste. One member of the group, loveringite, commonly occurs in titanate-based nuclear waste forms as a minor phase. Very little is known about the radiation effects of this mineral; however, we report here some previously unpublished data from our laboratory for samples (five davidite, one crichtonite). SEM–EDX analyses demonstrate that the chemical composition varies considerably from sample to sample, but individual samples are relatively uniform in composition with only limited evidence for zoning. The U content ranges from 0.2 to 9.5 wt% UO2 (0.02–0.65 apfu). The Th content is much lower, ranging from <0.1 to 1.3 wt% ThO2 (<0.01–0.09 apfu). Maximum amounts of other notable cations include 3.7 wt% V2O3, 4.1 wt% Cr2O3, 2.5 wt% Y2O3, 5.6 wt% La2O3, 6.0 wt% Ce2O3, 4.0 wt% MnO, 2.4 wt% ZnO, 2.7 wt% SrO, and 4.9 wt% PbO. Estimates of the geological age are available for these samples, giving a dose range of (0.3–42)  1016 a per milligram using eqn [1]. At present, the critical amorphization dose is poorly constrained, but is probably somewhere between 0.5  1016 and 2  1016 a per milligram based on our TEM observations. This is consistent with recent data for a 270 Ma davidite sample from Bektau-Ata, Kazakhstan, which is heavily damaged at a dose of 1.5  1016 a per milligram in a recent study by Malczewski et al.96 Two of our samples showed evidence for alteration in the form of replacement by an assemblage of rutile þ ilmenite or rutile þ titanite.

Minerals and Natural Analogues

579

5.22.3.7 ABO4 and AB2O6 Minerals (B ¼ Nb, Ta, and Ti) A number of other oxide minerals are known to contain substantial amounts of Th and U and are interesting from the point of view of the effect of structure type and composition on response to a-decay damage. Minerals in this category are principally fergusonite, aeschynite, euxenite, samarskite, and to a lesser extent, columbite–tantalite. Fergusonite, ideally YNbO4, may contain substantial amounts of heavy Ln, Th, and U on the Y-site together with Ti and Ta on the Nb site.97 This rare element oxide mineral is commonly reported to be amorphous as a result of a-decay damage. Giere´ et al.97 have given a detailed description of the chemistry and radiation damage of fergusonite occurring in a 40 Ma granitic pegmatite from Adamello, Italy. These authors used a combination of quantitative EPMA, TEM–EDX, and microRaman analyses to provide an upper limit on the critical amorphization dose of 1  1016 a per milligram for a sample characterized by TEM–EDX. This is consistent with previous work on amorphous fergusonite from Norway, containing 4.9 wt% UO2,98 and having an estimated a-decay dose of 4  1016 a per milligram based on an assumed age of 320 Ma. By way of comparison, amorphous fergusonite from the Rutherford pegmatite, Amelia, Virginia, has a well-defined age of 289 Ma, and contains 2.2–4.7 wt% ThO2 and 1.5–7.4 wt% UO2, giving a dose range of (2–7)  1016 a per milligram.99 Fergusonite from this locality is commonly altered along grain boundaries and cracks, accompanied by loss of most of the Y and Ln elements and uptake of Ta, Ca, Fe, and some Th (Figure 10). The orthorhombic AB2O6 oxides aeschynite, euxenite, and polycrase (with idealized compositions CeNbTiO6 or YNbTiO6) typically occur in granitic pegmatites where they incorporate Y and a range Ln series elements, together with Ta, Th, and U, and small amounts of Ca and Fe. Ewing100 investigated a suite of AB2O6 oxide minerals and reported all of them to be amorphous or ‘metamict’ due to a-decay damage. Based on Ewing’s carefully measured compositions, we have estimated that these samples received a-decay doses of (2–21)  1016 a per milligram wherein the lowest dose provides a good upper limit for amorphization of samples with ages of 250 Ma. According to Ewing, the AB2O6 oxides are commonly altered by hydrothermal fluids whereupon they show a consistent increase in the Ca content together with OH and H2O, generally at the

Figure 10 Backscattered electron image showing hydrothermal alteration of natural fergusonite (left) and monazite (right) from the Rutherford #2 granitic pegmatite, Amelia County, Virginia. Fergusonite is heavily altered to calciotantite + fersmite, thereby losing Y and heavy lanthanides to the fluid phase. The adjacent monazite crystal has also been partially altered during this process, but in this case, the alteration involves minor chemical exchange with the fluid phase. Width of image ¼ 0.2 mm.

expense of Y, Ln, Th, and U. At lower temperatures, for example, during weathering processes, A-site cations tend to be globally depleted. Due to the complications imposed by alteration, radiation damage, and potential phase transformations between AB2O6 minerals, thermal annealing is generally complex. The observations of Ewing and Ehlmann101 show that, in the simplest case, the aeschynite structure type is the first phase to form beginning at 400  C, followed by transformation to the higher temperature euxenite form at 700–750  C. However, pyrochlore and rutile are commonly formed in many samples. Lumpkin et al.23,24 examined a small group of amorphous Nb–Ta–Ti oxide minerals with stoichiometries of ABO4 and AB2O6 and reported recrystallization energies in the range of 40–85 J g1. Rare examples of columbite–tantalite, ideally (Mn,Fe)(Nb,Ta)2O6, were reported to contain enough U to induce amorphization (see Lumpkin102 and references to previous work). These 1800 Ma samples are zoned in both the TiO2 (2.2–4.8 wt%) and UO2 (0.2–2.6 wt%) contents and exhibit a critical amorphization dose of 8  1016 a per milligram for specimens examined by electron microscopy. This unusually high dose was attributed to long-term annealing of a-recoil collision cascades back to the original structure. We have recently found zoned columbite–tantalite crystals from the High Peak mine,

580

Minerals and Natural Analogues

in the Elk Mountain district of northern New Mexico. These zoned crystals have U-rich cores with up to 2.5 wt% UO2 and U-poor rims. The rims, however, contain inclusions of amorphous AB2O6 oxides resulting in cracking due to differential swelling (Figure 11). This example provides dramatic evidence for mechanical failure and subsequent geochemical alteration in course-grained materials with inclusions rich in U and Th. Columbite–tantalite is commonly altered to a secondary assemblage consisting of pyrochlore (NaCaTa2O6F), fersmite (CaNb2O6), or calciotantite (CaTa4O11) during postmagmatic hydrothermal activity, depending upon the activities of Naþ, Ca2þ, Fe2þ, Hþ, and HF in the fluid medium (see Lumpkin and Ewing38 and references therein). EXAFS and XANES studies generally indicate that the fully amorphous structures of these radiation-damaged minerals do not possess atomic periodicity beyond the second coordination environment (M–M distances), and in some cases, there is evidence for ‘disruption’ of the second coordination sphere. The first coordination sphere (M–O distances) remains intact albeit with minor changes in the bond lengths and degree of distortion of the M–O polyhedra. In the case of fergusonite, the Nb–O bond lengths of two longer bonds in the crystalline phase appear to decrease slightly in the amorphous material, consistent with a reduced preedge feature in the Nb–K edge.103 Thus, the distorted 4 þ 2 coordination geometry of fergusonite is somewhat ‘homogenized’

Figure 11 Backscattered electron image showing cracking and geochemical alteration of U-rich amorphous AB2O6 mineral inclusions in U-poor columbite–tantalite from the High Peak pegmatite, Elk Mountain, New Mexico. The cracking is clearly induced by differential swelling of the inclusions and the host phase, allowing aqueous fluid to migrate through the cracks. Width of image ¼ 0.5 mm.

by a-decay damage. Apart from the loss of and considerable disruption of the second coordination sphere, Nakai et al.103 found little difference in the mean bond length and distortion of the Nb site in euxenite. This result differs somewhat from the analysis of the Ti-site geometry in amorphous and annealed aeschynite and euxenite presented by Greegor et al.104 Although the disruption of the second coordination sphere is a common feature of both studies, it was found that amorphization produced a slight reduction in both the mean Ti–O bond length and coordination number of these minerals, primarily due to displacement of the longer Ti–O bonds.104 5.22.3.8

Hollandite

Although they do not contain actinides in natural or synthetic systems, minerals of the hollandite group are extremely important for the encapsulation of the relatively short-lived heat-generating radionuclides 90 Sr and 137Cs and long-lived 135Cs. The crystal structure of hollandite, A1.1–1.7B8O16, is similar to that of rutile and consists of edge-sharing chains of octahedra connected via corner sharing to form a three-dimensional framework. Hollandite, however, has two octahedral chains connected by edge sharing instead of the single chain found in rutile, resulting in a rather large 2  2 tunnel capable of accommodating large A-site cations like K, Rb, Cs, and Ba.105–107 These cations exhibit various ordering sequences over the available tunnel sites, commonly resulting in superlattice peaks in X-ray or electron diffraction patterns. The space group is typically I4/m or C2/m depending upon the A/B cation radius ratio. Numerous synthetic samples have been produced with Ti, Mn, Mo, noble metals, or Sn as the most common major elements and with Mg, Al, V, Fe, Co, Ni, Zn, and Sb, among others, as minor elements on the B-site. The composition of hollandite in titanate-based waste forms is generally 4þ given as ðBax Csy Þ½ðTi; AlÞ3þ 2xþy Ti82xy O16 in which charge compensation for Ba and Cs is usually provided by Al or Ti3þ.108 In natural samples, typical B-site cations are Ti, V, Cr, Fe, Mn2þ, and especially tetravalent Mn4þ. It is of interest to mention here one geological occurrence in particular, the Mn-rich metamorphic rocks Le Coreaux, Belgium, where deep purple and violet metasedimentary layers are host to quartz veins that contain Ba–Sr hollandite and other Mn minerals.109 These stable hollandites formed at a pressure of 1–2 kbar and temperature below 360–380  C, they

Minerals and Natural Analogues

contain 19–8 wt% BaO and 0.3–8 wt% SrO, and Sr is the dominant cation in the tunnels at concentrations above 6 wt% SrO. This suggests that hollandite may also be considered as a potential host phase for radioactive Sr in nuclear waste forms. The closest natural analog for synthetic titanium hollandite is the mineral priderite that occurs in Western Australia and elsewhere and has a composition of approximately (Ba,K)1.2–1.6[Ti,Fe,Mg]8O16.110,111 Very little information exists on the geochemical behavior of hollandite in natural systems, therefore, in this section, we will briefly outline some of the relevant experimental studies. Pham et al.112 carried out experimental work on synthetic Ba-hollandite doped with Cs and containing Al on the B-site for charge balance. These authors suggested that, following the initial release of Cs and Ba from reactive surface sites, the first few monolayers of the structure rapidly dissolved due to the release of Al and consequent precipitation of Al–OH species, driving solution pH to lower values. However, the alteration process was mediated via the formation of a continuous Al- and Ti-rich surface layer. Further evidence for selective removal of Ba and enrichment of Al and Ti on the surface of hollandite leached at 250–300  C was presented by Myhra et al.108 These conclusions were largely based on the incongruent release of Ba (0.113 g m2 day1), Al (6.6  103 g m2 day1), and Ti (<8  104 g m2 day1) after 14 days of leaching, combined with XPS analyses of the altered surfaces. In a study combining dissolution experiments and detailed characterization by electron microscopy, Carter et al.7 demonstrated that the release rates of Ba and Cs from hollandite are nearly identical, whereas Al and Ti are below the detection limits at 90  C and only Al was detected at 150  C (however, only by a factor of 2–3 above the detection limit). This is similar to previous observations; however, SEM work revealed the presence of nodular secondary phases on the surface of the hollandite at both temperatures. This was confirmed by TEM, which identified both Ti-rich and Al-rich nodules in a ratio of about 10 to 1, respectively. Furthermore, XPS analysis of the hollandite surfaces after the dissolution experiments indicated the presence of an Al-rich layer but only for the samples used in the experiments at 150  C. These authors also examined the pH dependence of the elemental release rates at 90  C, finding that the release of Ba decreases linearly from about 2  103 g m2 day1 at pH ¼ 2.5 down to 4  104 g m2 day1 at pH ¼ 12.9.

581

5.22.4 Silicates 5.22.4.1

Zircon

Zircon is classified as an ABO4 type orthosilicate (space group I41/amd) due to the presence of isolated SiO4 tetrahedra, which constitute the B-site, but the structure actually consists of a framework of edge-sharing silicate tetrahedra and eight coordinated A-sites.113 In nature, the composition of zircon often approaches ideal ZrSiO4 and is a common accessory mineral found in a variety of geological environments. Natural zircon may contain trace amounts of Ca, Ln, Hf, Th, and U on the A-site and P on the B-site. Some or all of these elements may be enriched significantly in zircon specimens from highly fractionated granitic rocks and especially in granitic pegmatites. Maximum concentrations of 26 wt% Ln2O3, 10 wt% ThO2, 10 wt% UO2, and 5 wt% CaO have been reported in zircon, but these values are exceptional. Natural zircons may also contain both OH and H2O species, some of which may be incorporated as a means of providing local charge balance at radiation damage sites.114 It is of interest to note here that U-rich zircon, containing up to 12.9 wt% U (0.1 U per formula unit) occurs in the silicate melt that formed during the accident at the Chernobyl nuclear plant in 1986.115 This melt, often referred to as ‘Chernobyl lava,’ resulted from partial melting of the nuclear fuel, related structural materials, and other materials dropped into the reactor area by helicopter. The presence of zircon, together with (Zr,U)Ox and UOx phases and globules of Fe(Cr,Ni) metal, proved to be important in determining the sequence of events. Based on a consideration of the phase relations, Burakov et al.116 suggested that in the initial phase of the accident, nuclear fuel interacted with Zr metal as temperatures increased to 2500  C just prior to the explosion, followed later on by silicate lava flow and emplacement of the melt below the reactor at temperatures of <1700  C. Geisler et al.117 conducted a detailed electron microprobe and micro-Raman spectroscopic investigation of the compositionally zoned Chernobyl zircons and reported that the UO2 content ranges from 0.8 to 15.8 wt%, equivalent to 0.005–0.115 U atoms per formula unit. Because the compositions of these remarkable zircons are confined to the system Zr1xUxSiO4, they provide an important benchmark for the analysis of observable Raman bands. To summarize briefly, the frequencies of the SiO4 stretching modes decrease by 0.67–0.75 cm1 per formula unit of U as a direct

582

Minerals and Natural Analogues

result of increasing Si–O bond length with increasing U content. Similar results are found for the lattice modes, but the SiO4 bending modes remain relatively constant in terms of the Raman frequency shift. Line broadening is significant for the lattice modes due to the ionic size difference between Zr4þ (rVIII ¼ 0.072 nm) and U4þ (rVIII ¼ 1.00 nm) and the resulting microscopic strain fields induced by substitution of Zr by U. A detailed analysis of the lattice vibrational modes indicates that the microscopic strain is larger in the (001) plane than along the c axis, consistent with the structure of zircon.117 In a pioneering study of natural zircons from several different localities, Hurley and Fairbairn118 demonstrated that the mineral experiences a transformation from the crystalline to the amorphous state and they determined that the diffraction angle of the (112) reflection decreased from 35.635 2y to 35.1 2y up to 0.4  1016 a per milligram. The data were found to follow an exponential function that related the ‘fractional disorder’ to the a-activity, the number of atoms displaced per a-decay event, and an annealing parameter. From the data, Hurley and Fairbairn determined a value of B ¼ 2.3  1016 mg per a-particle and calculated that 4500 atoms were displaced by each a-decay event. Closely following this work, Holland and Gottfried119 published their classic paper on the density, refractive indices, and unit cell parameters of zircon as a function of dose. Using a suite of samples from Sri Lanka, they showed that the density decreased systematically from <4.70 to 3.96 g cm3 at a dose above 1.2  1016 a per milligram. These data showed that the density of zircon decreased by 16%, the largest change of any potential waste form material. Refractive indices also decreased systematically and approached a single isotropic value of 1.81 over a similar dose range defined by the density measurements. XRD work revealed that the a and c cell parameters both increased rapidly, but anisotropically, as a function of dose before leveling off at 0.6  1016 a per milligram. Holland and Gottfried concluded their paper with an analysis of the fractions of crystalline zircon, an intermediate phase, and amorphous zircon as a function of dose. Analysis of the data for the crystalline fraction showed that B ¼ 3.8  1016 mg per a-particle. Murakami et al.120 showed that certain XRD peaks of zircon samples from Sri Lanka could be separated into Bragg and diffuse scattering components with increasing dose. This procedure enabled a more accurate determination of the anisotropic expansion with increasing dose, giving expansions of 1.5% along

the a axis, 1.8% along the c axis, and a lattice volume expansion of 4.7% (total volume expansion is 18%). Further work, including density measurements and TEM observations, delineated three stages of damage in the natural zircon samples. At dose levels below 0.3  1016 a per milligram, the damage is characterized by the accumulation of point defects, unit cell expansion, and lattice distortion. Within an intermediate dose range of (0.3–0.8)  1016 a per milligram, there is a progressive overlap of a-recoil tracks to produce larger amorphous domains with increasing dose. Above 0.8  1016 a per milligram, the zircon is completely X-ray and electron diffraction amorphous. Weber et al.121 have also suggested a long-term annealing rate of 1  109 year1 for zircon from Sri Lanka, a value that is similar to that reported for natural zirconolite. Changes in the mechanical properties of zircon as a function of dose are extremely important. This is not surprising in view of the large total volume expansion and anisotropic unit cell expansion documented earlier. Using a single specimen from Sri Lanka, Chakoumakos et al.122 provided a dramatic illustration of the fracture properties of a zoned zircon sample from Sri Lanka. Even though the total ThO2 þ UO2 concentration only varied by about 0.4–0.5 wt% between the 5–400-mm thick growth zones, the variation in dose was sufficient to cause microfracturing of the more brittle, low dose zones. The fractures were pinned in the high dose zones, indicating an increase in fracture toughness for these actinide-enriched layers. Following this work, Chakoumakos et al.123 revisited the sample and examined in detail the changes in chemistry and mechanical properties using electron microprobe data and a mechanical properties’ microprobe. Results of this study demonstrated that the hardness and elastic modulus of natural zircon decreased by 40% and 25%, respectively, for a-decay doses ranging from 0.3  1016 to 1.0  1016 a per milligram. Zircon samples from Sri Lanka were also used in an important study of the energetics of radiation damage as a function of dose.124 In this study, the results of temperature calorimetry revealed that the enthalpy of annealing at room temperature follows a sigmoidal trend with increasing dose, reaching a saturation value above 0.5  1016 a per milligram with DH ¼ 59 kJ mol1, consistent with structural changes at the subnanometer scale. The magnitude of this value exceeds DH for tetragonal ZrO2 þ SiO2 glass by 18–59 kJ mol1 and baddeleyite þ quartz by 33 kJ mol1, indicating that these assemblages may form upon thermal annealing of

583

Minerals and Natural Analogues

fa ¼ 1  eBD

½4

The best fit of eqn [4] to the data gives a value of B ¼ 2.7  1016 mg per a-particle, consistent with amorphous cascades having radii of 2.5 nm. Although there has been some controversy surrounding the damage accumulation model, a recent article by Palenik et al.129 indicates that sensitive measurement techniques, such as diffuse X-ray scattering, IR spectroscopy, and NMR spectroscopy, are capable of direct determination of the amorphous fraction and provide support for damage accumulation via the direct impact model. Using available data, the components of volume expansion are plotted in Figure 12, where we see that expansion of the crystalline fraction (e.g., unit cell) dominates at lower dose levels, but saturates at about 5 vol.%. Subsequent expansion is dominated by accumulation of amorphous domains. The structure of the amorphous state in natural zircon has been examined using EXAFS–XANES by Farges and Calas130 who found that the average Zr–O distance 0.1 A˚ less than that of crystalline zircon. They were also able to determine that the coordination number of Zr decreases from 8 to 7, indicating that O atoms are displaced from the ZrO8 coordination sphere during a-decay damage. Although long-range periodicity is lost, the Zr–Zr distances are still observed

Total Unit cell Difference

15

DV/ V0 (%)

amorphous zircon (see McLaren et al.125 for a detailed study and implications for age dating by ion microprobe techniques). Significant progress has been made in the understanding of a-decay damage in zircon through a comparison of natural samples and synthetic specimens doped with short-lived 238Pu.120,126,127 It was initially reported that the critical dose for amorphization of 238 Pu-doped zircon is 1.0  1016 a per milligram, and a multiple cascade overlap model was proposed to explain the accumulation of amorphous domains.126,127 Detailed analysis of the dose dependence of the crystalline fraction, derived from XRD analysis of the Bragg peak intensities, gives a value of B ¼ 5.8  1016 mg per a-particle for the amount of material damaged per a-decay event.127 The volume expansion is 16% (Figure 1), similar to the natural zircons. Until recently, the fraction of amorphous material as a function of dose has never been directly measured. Rı´os et al.128 accomplished this by directly measuring the amorphous fraction by careful determination of the diffuse scattering component in a suite of zircon samples from Sri Lanka, wherein the dose dependence of the amorphous fraction fa is shown to follow a direct impact model of amorphization:

10

5

0.1

0.2 Dose

0.3

0.4

(1016

0.5

0.6

a per milligram)

0.7

Figure 12 Plot showing the components of volume expansion in zircon doped with 238Pu as a function of a-decay dose. Unit cell expansion dominates at low dose, but saturates at about 5 vol.%. At higher doses, the total macroscopic swelling is dominated by volume expansion due to amorphous domains in the material.

in amorphous zircon, but on average they appear to decrease by 0.3 A˚ relative to crystalline zircon. Interestingly, these authors also found that Hf and Th retain eightfold coordination in amorphous zircon, but U appears to exist in a sixfold coordination geometry. Later work on synthetic zircon samples doped with trivalent 238Pu and 239Pu (t1/2 ¼ 2.42  104 years) revealed that the amorphous 238Pu-doped samples retained a ‘distorted zircon structure and composition’ on the subnanometer scale after 18 years, equivalent to D ¼ 2.8  1016 a per milligram.131 Thermal annealing of these samples in air revealed that the zircon structure is restored for T 1200  C, but with oxidation of Pu3þ to Pu4þ, together with minor PuO2 formation. Below this temperature, the samples recrystallize to a mixture of ZrO2, SiO2, and PuO2, generally consistent with the thermodynamic data of Ellsworth et al.124 The 239 Pu-doped samples remained crystalline after a cumulative dose of 0.012  1016 a per milligram due to the longer half-life of 239Pu. Thermal annealing of these samples at 1200  C in air also resulted in oxidation of Pu3þ to Pu4þ, decreased lattice distortion, and formation of some PuO2 in the ceramic. Wayne and Sinha132 were among the first investigators to show that the cracks in zircon, caused by radiation damage and differential swelling of different compositional zones, serve as pathways for

584

Minerals and Natural Analogues

migration of aqueous fluids. These fluids were able to penetrate into the zircon crystals, resulting in preferential leaching of radiation-damaged zones at temperatures of 450–500  C during deformation of the host rocks. Furthermore, Geisler et al.133 have provided a detailed description of the hydrothermal alteration of 619 Ma zircon from a posttectonic granite, Eastern Desert, Egypt, at temperatures on the order of 100–200  C. The zircon crystals exhibit oscillatory zoning and cracking arising from anisotropic, differential volume expansion. These cracks provided pathways for fluid migration and chemical exchange with the solid zircon, resulting in preferential alteration of the U and Th rich zones. Geisler et al.133 have determined that the more heavily damaged, higher dose zones, are enriched with Al, Ca, Mn, Fe, light Ln (e.g., La–Nd), and H2O species, and have lost Zr, Si, and radiogenic Pb. Due to the importance of zircon in geological age dating, numerous experiments have been conducted on both natural and synthetic samples. Pidgeon et al.134 conducted hydrothermal experiments on X-ray amorphous natural zircon from Sri Lanka containing 0.6 wt% U. They showed that up to 61% of the original amount of radiogenic Pb was lost after treatment with an aqueous solution of 2 M NaCl at 500  C and 100 MPa fluid pressure. Later studies carried out by Sinha et al.135 show that crystalline zircon loses 35–51% Pb and 30–47% U when treated with 2 M NaCl at 600  C and 600 MPa. Under the same conditions, partially damaged zircon lost 87% and 52% of the original amounts of Pb and U, respectively. Experiments were also conducted using the crystalline sample in a fluid with 2% HNO3 at 600 MPa and two different temperatures. In these experiments, the zircon lost 13–25% Pb and 0.5–4.3% U at 300  C and 30–33% Pb and 5.7–12% U at 600  C. A study by Rizvanova et al.136 demonstrated that Pb and U are released from amorphous zircon at temperatures as low as 200  C at 100–500 MPa in 2 M Na2CO3; however, crystalline zircon required higher temperatures of at least 650  C to produce significant levels of Pb and U loss in this aqueous system. At such a high temperature, crystalline zircon was replaced by baddeleyite in the silica-undersaturated solution. Using natural, radiation-damaged zircon samples from Sri Lanka,133,137–140 conducted a series of hydrothermal experiments at temperatures between 75 and 650  C in 2 M AlCl3, 2 M CaCl2, 0.1 M HCl, 0.1 and 3 M KOH, pure water, and complex aqueous solutions at pressures between 50 bar and 2.5 kbar.

The authors observed inward penetrating, irregular alteration fronts, which resemble those found in natural zircon. The experimentally altered areas are characterized by a lowered backscattered and an increased cathodoluminescence intensity and sometimes show complex internal nonequilibrium textures. Their thickness was found to be dependent on temperature as well as on the duration and pH of the experiments. Nanosized baddeleyite could be identified in some reaction zones by TEM and IR spectroscopy. The experiments also documented the uptake of hydrogen (mainly as OH) and of cations such as Ca, Ba, Mg, and Al from the fluid, combined with release of variable amounts of Zr, Si, Hf, Ln, Pb, Th, and U from the altered zircon. The loss of trace elements and the degree of structural recovery (including nucleation and growth of new zircon from the amorphous phase and the removal of defects in the crystalline remnants) was found to be temperature-dependent. At experimental fluid temperatures between 75 and 200  C, recrystallization of the amorphous phase was not activated and loss of REEs, U, Th, and radiogenic Pb was severe, while at higher fluid temperatures, limited loss of trace elements have been observed. This observation was interpreted to reflect a competition between the kinetics of longrange diffusion and ion exchange and the kinetics of the short-range diffusion necessary for the structural recovery processes, which significantly reduce the molar volume of the reacted domains. This creates, on the one hand, stress that is partly released by fracturing and, on the other hand, a nanoporous microstructure, as observed by TEM, providing pathways for fast chemical exchange between the reaction front and the fluid. Based on these results, the authors postulated a ‘diffusion-reaction’ model for the alteration of radiation-damaged zircon whereby moving recovery/recrystallization fronts are driven by the diffusion of hydrogen species into the amorphous zircon structure. In an effort to determine the effect of radiation damage on the dissolution of zircon, Ewing et al.141 performed experiments using natural zircon samples at 87  C in an aqueous solution containing 5 wt% KHCO3. Results of this study indicate that the dissolution rate increases by nearly two orders of magnitude from 3  108 up to 2  106 g m2 day1 for a-decay doses up to 1.0  1016 a per milligram, for example, the zircon samples range from highly crystalline to completely amorphous. Helean et al.142 determined the forward dissolution rate of zircon at 120–250  C. Using the elemental release rate of Si as

Minerals and Natural Analogues

a guide, these authors found that the dissolution rate of zircon increases from 1.7  104 g m2 day1 at 120  C to 4.1  104 g m2 day1 at 250  C. The impact of self-irradiation damage in zircon on its aqueous durability was also studied by Geisler et al.143 These authors performed a hydrothermal experiment in a 2 M CaCl2 solution at 600  C and 100 MPa with 16 variably radiation-damaged, that is, amorphized natural zircon samples from Sri Lanka. They found a dramatic increase in the alteration rate, monitored by the penetration of Ca and a lowered backscattered electron intensity from the reacted areas, at two critical fractions of amorphous domains. Molecular-dynamics (MD) simulations showed that the recoil cascades consist of a core depleted in matter surrounded by a densified and polymerized boundary.144 Simulations of a second and third recoil event further revealed that the low-density regions form percolating regions inside the amorphous cascades.143 The existence of such a nonuniform amorphous structure was confirmed experimentally by low-angle X-ray scattering experiments. MD simulations of multiple events further reveal that strongly overlapping cascades produce connected regions of depleted matter (e.g., low atomic density), which likely serve as fast-diffusion pathways. The authors suggested that the first dramatic increase in the alteration rate marks the first percolation point where the amorphous domains form infinite clusters. However, at this stage, the polymerized boundary regions still exist as barriers to diffusion and have to be overcome by hydrolysis and hydration reactions during the alteration process. The second dramatic increase is interpreted to be a direct consequence of the formation of interconnected regions of depleted matter, allowing invasion-like penetration of Ca and water over macroscopic length scales. Geisler et al.145 have performed a comparative experiment with a radiation-damaged natural zircon from Sri Lanka and a synthetic 238Pu-doped zircon (4.7 wt% of 238Pu) in an acidic solution at 175  C. Both zircon samples have suffered a similar degree of radiation damage, as given by their degree of amorphization. XRD measurements of the experimental run products revealed that during the hydrothermal treatment, only the disordered crystalline remnants recovered in the natural zircon, whereas in the 238 Pu-doped zircon, the amorphous phase strongly recrystallized. Such a different alteration behavior of natural and Pu-doped zircon suggests two fundamentally different alteration mechanisms. The authors postulated that the alteration of natural

585

radiation-damaged zircon with a low doping level of U and Th is controlled by diffusion-reaction processes as discussed earlier, but that the high degree of recrystallization observed in the 238Pu-doped zircon doped is more compatible with the concept of an interface-coupled dissolution–reprecipitation process, where congruent dissolution is assumed to be spatially and temporally coupled to the reprecipitation of new zircon (poorer in Pu) at an inward moving reaction interface. 5.22.4.2

Thorite

Thorite (ideally ThSiO4) is isostructural with zircon and contains 17–19 wt% SiO2 and 55–75 wt% ThO2, depending upon the age and content of radiogenic Pb. In some natural systems, thorite may contain up to 4 wt% P2O5, 6 wt% CaO, 5 wt% ZrO2, 2 wt% As2O5, 21 wt% Ln2O3, 25 wt% UO2, and 5–16 wt% total H2O (see Farges and Calas,130 Foord et al.,146 and Lumpkin and Chakoumakos147 and references therein). Thorite samples from the Harding pegmatite, New Mexico, exhibit extensive solid solution toward the end-members, Ca0.5Th0.5PO4, Ca0.5Th0.5VO4, and to a lesser extent, YPO4. Rare, yellow thorites were also found with a significant Ca0.5U0.5SiO4 component in which the U is postulated to be hexavalent.147 Geologically young thorite crystals with a well-defined age of 6–7 Ma have been reported to be completely amorphous,146 providing an upper limit on the critical dose of 0.8  1016 a per milligram, consistent with data for natural zircon discussed in the previous section. However, Lumpkin and Chakoumakos147 found that the P and V thorites from the Harding pegmatite retained substantial crystallinity even after sustained doses of 40–120  1016 a per milligram (Figure 13). Based on a simple comparison of bond energies, they proposed that thorites rich in P and V have a lower energy barrier to recrystallization than samples that are closer to ThSiO4 in composition. Earlier work has shown that thorite is subject to extensive alteration in natural systems, generally resulting in hydration, loss of Si, and loss of radiogenic Pb, often forming secondary galena in the presence of S-bearing fluids.148,149 5.22.4.3

Titanite (Sphene)

Titanite, ideally CaTi(SiO4)O crystallizing in monoclinic space group C2/c, may incorporate minor Na, lanthanides, and low levels of actinides on the Ca site, together with Fe, Al, and Nb on the Ti site. Additionally, significant amounts of F and OH may

586

Minerals and Natural Analogues

Figure 13 Optical micrograph showing the presence of crystallinity in P and V-rich thorite from the Harding granitic pegmatite, Taos County, New Mexico. This image, taken with crossed polarizers, reveals crystalline domains as brightly colored spheroidal regions, suggesting recrystallization of radiation damage thorite over time. Width of image ¼ 1 mm.

replace O on the anion site that is not bonded to Si. The amounts of actinides incorporated in the structure are generally below 500 ppm Th and 3000 ppm U. Due to limited solubility of lanthanides and actinides in the structure, the titanite has seen limited use in waste forms apart from the titanite-based glassceramics developed for Canadian waste.150 Heavily damaged samples studied by Hawthorne et al.151 from the Cardiff mine, Ontario, Canada, indicate that the critical dose is somewhat higher than (0.3–0.4)  1016 a per milligram if we assume an age of 1000 Ma for this locality. This is in good agreement with a major study conducted by Vance and Metson,152 which showed that the critical dose is 0.5  1016 a per milligram and that the crystalline–amorphous transformation results in a density decrease of 8%. General features of a-decay damage in titanite with increasing dose include increasing thermal vibration parameters of the cations and anions, increasing unit cell volume up to about 3% in the most heavily damaged samples, and possible reduction of some Fe3þ to Fe2þ during electron transfer processes associated with the radioactive decay (see Section 5.22.2). 5.22.4.4

Allanite

Minerals of the epidote group conform to the general structural formula A2M3(SiO4)(Si2O7)(O,F)(OH) with A ¼ Ca, Sr, Pb, Mn, Th, Y, Ln, and U, and M ¼ Al, Fe, Mn, Mg, Cr, and V. In this section,

we are mainly interested in the mineral allanite, characterized by the presence of light Ln elements and Th on the A-sites together with Al, Fe3þ, Fe2þ, and Mg on the M-sites. Allanite typically occurs as an accessory mineral in felsic igneous rocks, granitic pegmatites, volcanic rocks, and metamorphic systems, among others. It has assumed some importance as a tracer of geochemical processes and has proved to be useful for geological age dating. Only limited data are available with regard to radiation damage effects in allanite; however, it is clear that the mineral becomes optically isotropic due to a-decay damage. The most consistent data sets also indicate that the density decreases by 8.5–10.5%, giving a reasonable indication that the volume expansion is well below that of zircon.153,154 Janeczek and Eby153 investigated three samples from different geological localities and provided a detailed assessment of the composition, microstructure, and annealing behavior. Based on careful EPMA, XRD, and TEM studies of samples with reasonably established geological ages, it appears from this work that the critical amorphization dose is somewhat >0.5  1016 a per milligram for two samples from the Appalachian orogen (200–400 Ma). A third sample from Arizona has an age of 1400 Ma and exhibits a slightly higher level of crystallinity in the bulk XRD pattern. The critical amorphization dose of this sample is slightly >5  1016 a per milligram, suggesting that a long-term annealing mechanism may be operative in allanite. Thermal recovery of allanite occurs above 500  C and the mineral decomposes above 850  C. Lattice parameter changes are anisotropic during recovery and the unit cell volume decreases by 2% at a temperature of 800  C. Allanite is subject to hydrothermal and low temperature alteration and breakdown to new phase assemblages. It is commonly replaced at low temperatures by fluorocarbonate minerals (e.g., bastnaesite, LnCO3F), clay minerals, and thorite. Breakdown during weathering to cerianite (CeO2), monazite, clays, and goethite has also been reported.154 Wood and Ricketts155 described in some detail the alteration of igneous allanite by low temperature (100–200  C) hydrothermal F and P containing fluids circulating in the Casto pluton, Idaho. The alteration occurs along the rims of crystals and along fractures and is accompanied by some Th enrichment and loss of Ln elements. The most severe alteration resulted in breakdown of allanite to fluorite, monazite, and other secondary phases. Figure 14 shows an example of hydrothermal alteration of natural allanite from the

Minerals and Natural Analogues

Figure 14 Backscattered electron image at low magnification showing progressive alteration of allanite from the Rutherford #2 granitic pegmatite, Amelia County, Virginia. The lower, lighter gray area is unaltered allanite. Darker gray zones in the middle of the image represent allanite–epidote–clinozoisite alteration, for example, chemically altered allanite possibly of hydrothermal origin. The heavily altered dark area in the upper part of the image consists of a complex assemblage of silicate, lanthanide, and iron minerals probably formed at lower temperature. Width of image ¼ 1 mm.

Rutherford #2 granitic pegmatite, Amelia County, Virginia. This image reveals an outer zone of progressive alteration of allanite to an assemblage of late stage silicate, lanthanide, and iron minerals.

5.22.5 Phosphates 5.22.5.1

Monazite

Like zircon and thorite, monazite also has ABO4 stoichiometry, but the crystal structure is monoclinic (space group P21/n) and consists of chains of alternating BO4 tetrahedra and AO9 polyhedral sites.156 These chains are cross-linked by edge sharing with the AO9 polyhedra, effectively closing off open tunnels and creating a structure that is 10% more dense than the zircon structure type. Orthophosphates with Nd, Pr, Ce, La, Am, or Pu on the A-site adopt the monazite structure; whereas, those with the heavier and smaller Ln elements Lu, Yb, Tm, Er, Ho, and Y adopt the tetragonal xenotime structure, which is isostructural with zircon. In natural systems, the formula of monazite is generally given as (Ln,Th,U,Ca)(Si,P)O4, representing a solid solution of the ideal end-members LnPO4, Ca0.5Th0.5PO4, Ca0.5U0.5PO4, and ThSiO4. Natural

587

monazite may contain up to 16 wt% UO2 and 52 wt% ThO2,156,157 and has recently been found in alkaline rocks containing over 8 wt% SrO, indicating solid solution toward an end-member of the form Sr0.5Th0.5PO4.42 Therefore, the mineral may be considered as a potential host phase for actinides and a range of fission products, including Sr. Natural monazite remains crystalline even up to a-decay doses approaching 7  1016 a per milligram, a feature that makes synthetic monazite-based materials very attractive for nuclear waste encapsulation. The radiation stability of synthetic (La,Pu)PO4 and PuPO4 doped with 8.1 and 7.2 wt% 238Pu, respectively, has been investigated by Burakov et al.158 These authors discovered that (La,Pu)PO4 remains crystalline up to a dose of (0.2–0.3)  1016 a per milligram, albeit with a decrease in the intensity of the measured XRD peaks. In contrast to this result, the PuPO4 ceramic sample is heavily damaged at a dose of only 0.1  1016 a per milligram, and exhibits substantial volume swelling and cracking. A number of studies have documented alteration of monazite during interaction with various hydrothermal fluids.159–164 An important study was conducted by Mathieu et al.162 on natural monazite occurring in Lower Proterozoic sandstones of the Franceville basin, Gabon (see Section 5.22.6.3). The results of this work demonstrate that monazite was altered to a microcrystalline Th-silicate phase by interaction with a low temperature (<200  C, 100 MPa) diagenetic brine (NaCl–CaCl2, with Li, Br, and SO4), resulting in loss of light Ln elements and U. Several monazite alteration mechanisms have been identified, including chemical exchange, dissolution–reprecipitation, dissolution and replacement by a different mineral, and, in rare cases, selective Th removal (see Figure 10). In their detailed study, Poitrasson et al.160 documented that while the light Ln are typically released from monazite, U, Y, and the heavy Ln were retained during hydrothermal alteration at temperatures of 260–340  C and in fluids with salinities ranging from 3 to 18 wt% NaCl equivalent. In a study of monazite from the Steenkampskraal mine, South Africa, Read et al.164 showed that the light Ln elements are retained in altered monazite, the heavy Ln and Y are being released and precipitated locally as secondary phosphate minerals, and U is released to the fluid phase and removed from the system. In general, Th is typically less mobile than the lanthanide and Y, and is concentrated into Th-bearing alteration products.165,166 More recently, Hetherington and Harlov167 demonstrated

588

Minerals and Natural Analogues

that monazite is subjected to a chemical refinement process during interaction with an evolving granitic pegmatite fluid. These relatively high temperature, H2O-rich fluids contained Na, K, F, and minor Cl and reacted with high Ca–Th–U–Si monazite via a coupled dissolution–reprecipitation mechanism to produce near end-member LnPO4 monazite and precipitates of thorite and uraninite. Experimental data indicate that monazite is highly insoluble in most hydrothermal and low temperature fluids; however, the solubility may be enhanced in aqueous fluids with low pH, low phosphate content, or high F concentrations which can lead to the formation of REE-fluoride complexes.168,169 Thermodynamic calculations suggest that an increase in pH from 3.5 to 5.0 will decrease the solubility of monazite by about two orders of magnitude at 300  C. A similar decrease in solubility is expected if the total PO4 concentration decreases from about 107.5 to 105.5 molal.169 At temperatures below 250  C, the solubility of monazite in aqueous solutions decreases with increasing temperature,156 and this has attractive implications for geological disposal. Recent work by Oelkers and Poitrasson170 provided important results on the steady-state dissolution rates of monazite at temperatures of 50–230  C and pH ranging from 1.5 to 10 with variable flow rate and surface area. Using a natural sample as the starting material, these authors show that the release rates of the REEs and U are essentially congruent for all experimental conditions. The Th concentration in solution was stoichiometric only in the basic solutions and was detected at lower than stoichiometric ratios in acidic solutions, probably due to precipitation of a Th-rich secondary phase. A few experiments have been performed at elevated temperature and pressure in order to assess the behavior of the U–Th–Pb system in monazite.171,172 Teufel and Heinrich,171 by means of hydrothermal experiments at 400–750  C and 300 MPa, demonstrated substantial Pb loss in monazite powder at temperatures as low as 400  C. The mechanism of Pb loss involved dissolution and reprecipitation of the monazite at these high temperatures. In contrast, Seydoux-Guillaume et al.172 found no evidence for Pb loss in their experiments in pure water at 800–1200  C and 700 MPa, even though some dissolution and recrystallization was observed on the margins of the monazite grains. Additional experiments were conducted at 1000  C using NaCl, CaCl2, SrCl2, and Pb-bearing fluids. Significant changes in the U–Th–Pb system were only observed in the CaCl2

and SrCl2 hydrothermal fluids, with a Pb loss discordance of 68% and 16%, respectively. Nevertheless, no Pb diffusion profiles were observed and the discordance observed in these experiments was attributed to the dissolution–reprecipitation mechanism. 5.22.5.2

Apatite Group

An important feature of this rather larger structural group is the ability to incorporate a range of substitutions on the Ca and P sites, including solid solution toward britholite that contains Si on the tetrahedral sites. The general formula of the group, consisting of more than 20 mineral species, can be expressed as M5(ZO4)3X with M ¼ Ca, Sr, Ba, Pb, Na, Mn, Y, and Ln; Z ¼ P, S, V, As, and Si; and X ¼ F, Cl, and OH.173 There are also more than 80 synthetic compounds that crystallize with the apatite structure. The common rock-forming minerals of the apatite group (fluor-, chlor-, and hydroxyapatite) are described by the formula Ca5(PO4)3(F,Cl,OH) and crystallize in space group P63/m (X ¼ F) or P21/b (X ¼ Cl, OH). In most rocks, these minerals generally contain low levels of Th and U, for example, <0.2 wt% Th þ U and have limited value for radiation damage studies. However, in more highly evolved rocks, there may be extensive solid solution toward britholite, ideally Ln4Ca(SiO4)3O, wherein the ability to substitute Si on the tetrahedral sites and O on the anion site normally occupied by halogens or OH groups allows for incorporation of actinides on the Ca/Ln sites. The rare apatite minerals of the pyromorphite– vanadinite–mimetite series, Pb5(PO4)3Cl–Pb5(VO4)3 Cl–Pb5(AsO4)3Cl, are also important as potential analogs for materials designed to encapsulate longlived radioactive 129I in nuclear wastes. Synthetic apatites can also contain Sr on the Ca-site and may be considered as a possible host phase for short-lived 90 Sr. It is of interest to note that fluorapatite occurs in the alteration zones of the Oklo natural fission reactors, where the mineral contains Nd and Sm isotopes of fission origin.174 One sample also showed a slightly elevated 235U/238U ratio, possibly due to incorporation of 239Pu derived from the fission reactors at the time of formation. This has been confirmed by Hidaka,175 who measured 235U/238U isotopic ratios of up to 0.0171 for apatite coexisting with uraninite with slightly depleted 235U/238U ¼ 0.0066 (see Section 5.22.6.3). Although technically a silicate mineral, britholite is discussed here as part of the group. Britholite from pegmatitic segregations in felsic igneous rocks of the

Minerals and Natural Analogues

Eden Lake complex in Manitoba, Canada, has been studied in some detail.176 These silicate apatites contain about 0.9–1.3 wt% ThO2 and 0.7–1.0 wt% UO2 together with 52–55 wt% Ln2O3, 12–17 wt% CaO, 1–5 wt% P2O5, and 2–4 wt% F. The crystals are optically isotropic but weakly crystalline according to XRD analysis; however, in this case, the observed diffraction could be due to inclusions of associated minerals. Assuming an age of 1700 Ma for the host rocks, the estimated a-decay dose is (4–7)  1016 a per milligram and probably represents an upper limit for the critical dose for amorphization. For comparison, the results of Carpe´na et al.177 on a series of 2100 Ma apatite–britholite samples from Ouzzal Mole, Algeria, indicate that the critical dose is 2  1016 a per milligram. This suggestion is reasonably consistent with the recent work of Yudintseva,178 who examined six natural britholite samples from alkaline rocks in Russia, ranging in age from 320 to 2600 Ma and containing up to 12 wt% ThO2 and UO2. Detailed XRD and TEM work on these samples places the critical dose for amorphization at a value close to 0.9  1016 a per milligram. Thermal annealing of fully amorphous samples indicated recrystallization temperatures of 500–600  C by XRD.178 Apatite is one of the few minerals studied by actinide doping experiments in the laboratory.179,180 For example, the synthetic britholite CaNd4(SiO4)3O becomes completely amorphous at a dose of 0.3  1016 a per milligram when doped with 1.2 wt% 244Cm. This is about a factor 3 lower than the critical dose required to render natural samples amorphous and is generally in keeping with observations on the other minerals discussed in previous sections of this chapter. The crystalline–amorphous transformation in this compound results in a bulk volume expansion of 8.0–8.5% (see Figure 1) in which the calculated damage volume B ¼ 200 nm3, corresponding to a spherical track radius of about 3.6 nm. DTA showed that the amorphous Cm-doped britholite recrystallized at temperatures of 600–700  C, releasing 130 J g1 of stored energy and activation energy of 3.1 0.2 eV.179 Boudreau and McCallum181 reported significant findings on the alteration of Ln-rich chlorapatite from the Stillwater igneous complex in Montana. This alteration event is associated with the hydrothermal alteration of the mafic host rocks as olivine was replaced by serpentine þ magnetite calcite during infiltration of meteoric water at low temperatures (100  C) and near-neutral pH. Under these conditions, the chlorapatite was partially replaced by hydroxyapatite and small grains of monazite. Based

589

on the results of electron microprobe analyses, the authors suggest that Si was lost from the magmatic chlorapatite and that CO2 and minor amounts of other elements (e.g., Fe) may be incorporated in the hydroxyapatite alteration product. Arden and Halden176 also reported that apatites from Eden Lake, Manitoba, exhibit evidence for hydrothermal alteration in which La, Ce, and F were lost and Cl, and possibly OH, were gained. The temperature during the hydrothermal alteration event may have been as low as 200  C. This study is interesting in view of the dissolution experiments performed on the Cmdoped synthetic britholite CaNd4(SiO4)3O.179 The tests described in this study were conducted on fully amorphous and recrystallized samples using deionized water at 90  C for 14 days. In general, the results indicate that radiation damage induced amorphization and volume swelling lead to an increase in the dissolution rate by approximately one order of magnitude. However, the release of Nd was below detection limits in all of the experiments, suggesting that it had reprecipitated on the surface of the specimens as a secondary oxide phase. Valsami-Jones et al.182 noted that, although phosphate mineral dissolution has not been studied in detail, they are known to have very low solubility products in aqueous solution. Based on a series of laboratory experiments on synthetic hydroxylapatite and natural fluorapatite at 25  C and pH ¼ 2–7, they proposed that apatite dissolves according to the following dissolution reactions: Ca5 ðPO4 Þ3 OH þ 7Hþ ¼ 5Ca2þ þ 3H2 PO 4 þ H2 O ½VI  Ca5 ðPO4 Þ3 F þ 6Hþ ¼ 5Ca2þ þ 3H2 PO 4 þF

½VII

Experimental data indicate that the dissolution rates of both phases increased with decreasing initial pH of the buffered solution, in accordance with the reactions shown earlier. A similar experimental study was recently reported for synthetic apatite having the composition Ca4.5Nd0.5(P2.5Si0.5O4)3F, conducted at 25  C and pH ¼ 3–12.183 Results of this work also showed a negative correlation between the dissolution rate and pH for pH ¼ 2–7, but a constant rate was observed for pH > 8. In these experiments, the authors discovered that the Nd release rates are slower than those of Ca, P, and F and attributed the result to the precipitation of a secondary phase, possibly rhabdophane, NdPO4nH2O. Apatite samples doped with Nd, Th, and U were examined by

590

Minerals and Natural Analogues

Terra et al.,184 who measured release rates of 4  104 g m2 day1 for Nd and 1.3  104 g m2 day1 for Th in single-phase Nd–Th samples in experiments with 104 M HNO3 at 90  C. In comparison, the release rate of U was found to be 2  102 g m2 day1 for Nd–U-doped apatite under the same conditions. This was attributed to the oxidizing conditions of the experiment and the presence of a second U phase in the Nd–U-doped material. 5.22.5.3 Kosnarite and Related NZP Materials Kosnarite, ideally KZr2(PO4)3, crystallizing in space group R3c, is a relatively rare mineral occurring in granitic pegmatites where it occurs in association with late stage, secondary phosphate mineral assemblages.185 A synthetic analogue of kosnarite, NaZr2(PO4)3, otherwise known as sodium zirconium phosphate (NZP), is the prototype for a large family of synthetic compounds that can be described by the general formula A1xB2(PO4)3 with A ¼ monovalent Na and K, divalent Ca, Sr, Cd, Ba, and Pb, or trivalent Ln elements; and B ¼ tetravalent Ti, Zr, Hf, Sn, Th, and U or trivalent Cr, Fe, Ga, and In, etc.186–189 These materials exhibit low thermal expansion properties and have potential applications as fast ion conductors, redox insertion/extraction materials, and as a host phase for nuclear waste immobilization. In the latter case, it is the crystal chemical flexibility and durability in aqueous solutions that have attracted the most attention to this interesting family of compounds.190,191 In comparison to the available data on other minerals and synthetic compounds summarized in this chapter, very little work has been published on a-decay damage in NZP compounds, and natural kosnarite has not been reported with detectable amounts of Th and U. However, Russian scientists have succeeded in the synthesis of NaPu2(PO4)3 containing 239Pu or 238Pu. After 2 years of storage, the 239 Pu sample accumulated a dose of 0.0091  1016 a per milligram and showed no evidence of degradation; whereas, the 238Pu sample became amorphous at a dose of 0.93  1016 a per milligram (conference abstract by Orlova et al. 1993, cited by Zyryanov and Vance192). There have been several studies of the aqueous dissolution behavior, one of the earliest was conducted by Roy et al.190 who prepared CsZr2(PO4)3 and reported a release rate of 0.2 g m2 day1 for Cs after a hydrothermal experiment in deionized water for 14 days at 300  C and 30 MPa. Analyses of

the solution following the experiment indicate the release rates on the order of 0.1 g m2 day1 for P and 2  103 g m2 day1 for Zr. A second experiment was conducted using a brine containing 5 wt% NaCl, 5 wt% KCl, 10 wt% MgCl2, and 10 wt% CaCl2 at the same temperature and pressure. In this case, the Cs release rate was higher by a factor of three, while the release rates of P and Zr were similar to those of the experiment with deionized water. Furthermore, in the latter experiment, baddeleyite was observed as an alteration product. Zyryanov and Vance192 conducted dissolution tests on undoped NaZr2(PO4)3, several NZP samples containing Cs, Sr, Y, Nd, Gd, and Ca, and a sample containing 20 wt% simulated Purex type waste. Although the samples were hot pressed, the tests were conducted on 37–63 mm powders at 90  C in an aqueous solution with pH ¼ 5 and a solid surface-area-to-solution-volume ratio close to 1.0. After 28 days, the test results gave elemental release rates of 0.002–0.03 g m2 day1 for Na, 0.1–3  107 g m2 day1 for Zr, and 0.002–0.1 g m2 day1 for P. The release rates of Cs and Sr were 0.002 and 0.003 g m2 day1, respectively, whereas Y, Nd, and Gd were released at rates similar to that of Zr. Based on the measured weight losses, the authors suspected that reprecipitation had occurred during the experiments, but this was not proven. In the sample prepared with simulated waste, release of Zr, P, Ce, Nd, and Ag were similar to those of Synroc-C, whereas other elements in the NZP showed release rates typically 1–2 orders of magnitude higher than the Synroc sample under the same conditions. More recently, Bois et al.193 synthesized La0.33Zr2(PO4)3 and LaPO4 (monazite) ceramics and determined the elemental release rates at a temperature of 96  C. For experiments conducted with a solid surface-area-to-solution-volume ratio of 0.1, the authors report a minimum release rate of 103 g m2 day1 for P, while the release rates of La and Zr were both <105 g m2 day1. In a parallel experiment, they determined that the release rates of La and P from the monazite ceramic were about one order of magnitude lower than the NZP type compound. 5.22.5.4

Xenotime

Ideally YPO4 with heavy lanthanide elements (e.g., Dy–Lu) replacing some of the Y, xenotime is an important accessory mineral in igneous and metamorphic rocks where it may be a very useful mineral for U–Pb age dating and for geothermometry–geobarometry in

Minerals and Natural Analogues

conjunction with coexisting monazite and garnet.194,195 Like monazite, xenotime has never been reported in the amorphous state even though it has the zircon crystal structure. Typical concentrations of Th and U in xenotime are 0.1–0.3 wt%; however, xenotime samples from evolved granitic rocks and pegmatites may contain up to 6.0 wt% ThO2 and 7.5 wt% UO2.167,196 Based on several analyses of xenotime with high Th and U concentrations from host rocks with reasonably constrained geological ages, we estimate that they have received a-decay doses of (1.4–14)  1016 a per milligram. The principal study of relevance to radiation damage is the accelerated radiation damage study of Luo and Liu,197 wherein flux grown crystals of LuPO4 were doped with 1.0 wt% 244Cm and stored for 18 years. In this time, the crystals reached a cumulative dose of 5  1016 a per milligram which is within the range of the high Th–U natural xenotime samples noted earlier, but they remained in a highly crystalline state according to electron microscope observations. However, the authors did observe 5–10 nm defect clusters with associated strain fields and 5–20 nm voids assumed to contain He due to the high a-particle dose. The importance of this study lies in the demonstration of radiation resistance in Yand lanthanide orthophosphates with the zircon structure type and, at least to a certain extent, substantiates the observations on the recovery of radiation damage in P-bearing natural thorites discussed earlier (Section 5.22.4.2). Hetherington and Harlov167 have also described a chemical refinement type of alteration in high Th–U–Si xenotime, wherein a fluid-driven coupled dissolution–reprecipitation mechanism produced near end-member (Y,Ln)PO4 xenotime, thorite, and uraninite.

oxidizing conditions is given in the next section, followed by two examples of U ore deposits that have been studied quite extensively as natural analogs: the natural fission reactors of Gabon and a description of U migration at Koongarra, Australia. 5.22.6.2 General Aspects of Uraninite Alteration in Natural Systems Uraninite, ideally UO2, is a cubic (fluorite structure, Fm3m) mineral crystallizing in granitic rocks, granitic pegmatites, volcanic rocks, metamorphic systems, and sedimentary environments. All natural uraninites remain crystalline to a-decay doses reaching values in excess of 100  1016 a per milligram. Natural uraninite is always slightly oxidized and can be described by the formula UO2þx with x < 0.3.199 This is achieved through the slight oxidation of U4þ to U6þ in natural samples. Other elements incorporated during crystallization include Th, Ca, and Ln elements with similar ionic radii to that of U. The amounts of radiogenic Pb accumulated in uraninite can be quite high, but this element is not compatible with the fluorite structure and is commonly depleted due to long-term diffusion or episodic geological events. Under reducing conditions, the solubility of uraninite is extremely low; however, this is moderated to some extent by the presence of minor U6þ and radiogenic Pb2þ in the structure, the latter increasing substantially with time. Perhaps the most important alteration mechanism of uraninite under reducing conditions is replacement by coffinite (USiO4) and this proceeds according to the reaction: UO2þx þ SiO2 ¼ USiO4 þ ð0:5xÞO2

5.22.6 Ore Deposits: Analogs for Spent Fuel 5.22.6.1

Preamble

Although this chapter mainly deals with Th–U minerals as natural analogs for alternative ceramic nuclear waste forms, some of these polyphase materials may contain significant amounts of UO2 in addition to the targeted phases. Considering this and the fact that direct disposal of spent fuel is still an option, the purpose of this section is to introduce the reader to some of the geological and mineralogical aspects of U ore deposits as natural analogs (see Wronkiewicz and Buck198 for a more detailed review of this topic). A brief discussion of the general aspects of the geochemical alteration of uraninite under reducing and

591

½VIII

From this reaction, we can see that an increasing silica activity promotes coffinite formation, whereas increasingly oxidizing fluids tend to stabilize uraninite. If dissolved H2S is present, then radiogenic Pb released from uraninite may lead to the formation of galena as a byproduct of eqn [VIII]. Under oxidizing conditions, the uranyl oxyhydroxides play an important role in the early stages of uraninite alteration in relatively acidic and dilute aqueous systems. According to Finch and Murakami199, this important uranyl mineral group can be described by the general formula Mn[(UO2)x Oy(OH)z](H2O)m with M ¼ Ca, Sr, Ba, Pb, and K. Schoepite and studtite are the main phases with n ¼ 0 and ianthinite is an interesting oxy-hydroxide

592

Minerals and Natural Analogues

in which M ¼ U4þ. In relatively basic aqueous systems with dissolved silica, there are several important uranyl silicates, including soddyite (UO2)2SiO4(H2O)2, members of the uranophane group Mnþ[(UO2) (SiO4OH)]n(H2O)m, and members of the weeksite group M(UO2)2(Si5O13)(H2O)4. In more complex aqueous systems, a number of relatively insoluble uranyl phases may form as alteration products of uraninite. The general formula of many of these minerals can be written as Mn[(UO2)(XO4)]2(H2O)m where M ¼ Na, Mg, K, Ca, Mn, Fe, Co, Ni, Cu, Zn, Ba, and Pb, and X ¼ P, V, or As. Minerals described by this formula include members of the autunite group, metaautunite group, a large number of vanadates, such as carnotite and curie´nite, and certain arsenates.49 The alteration of uraninite commonly proceeds through the early-formed metastable uranyl oxyhydroxides, such as schoepite, becquerelite, vandendriesscheite, and fourmarierite, to silicates such as soddyite, curite, and uranophane. Certain uranyl carbonate minerals (e.g., rutherfordine) may also be present as part of the assemblage of secondary U minerals provided that there is enough dissolved CO2 in the system. Uranyl phosphate minerals are commonly the last to form in this alteration sequence. Wronkiewicz and Buck198 presented an informative comparison between a laboratory experiment involving the dissolution of synthetic UO2 in an aqueous solution designed to simulate conditions imposed by volcanic rocks at the Yucca Mountain repository in Nevada and the uranyl mineral paragenesis observed in the Nopal I ore deposit, a geologically young (8 Ma) occurrence with low radiogenic Pb, hosted by a similar suite of volcanic rocks in the Pen˜a district, Chihuahua, Mexico. With minor variations, for example, the Na/Ca ratio of the experiment versus the groundwater composition of the natural system and the relative amounts of the uranyl phases Naboltwoodite and uranophane, the laboratory experiment (lasting 8 years) closely followed the observed mineral paragenesis at the Nopal I deposit. 5.22.6.3

Natural Fission Reactors in Gabon

The western African country of Gabon is host to a number of sedimentary U deposits that formed around 2000 Ma in the Franceville basin and remained unmetamorphosed, reaching a maximum temperature of 200  C during diagenesis (see Janeczek200 and references therein). There are six known U ore deposits, three of which (Oklo, Okelobondo, Bangombe´) are known to have sustained

nuclear fission in 16 different reactor zones soon after formation. These natural fission reactors may have operated for several thousand years, driving the 235 U/238U ratio of the uraninite ‘fuel’ to values as low as 0.0038. In a careful study of the mineral florencite occurring as inclusions in uraninite, Meshik et al.201 found elevated levels of fission Zr, Ce, Sr, Xe, and Kr. Based on analysis of the isotopic signature of Xe, they determined that the natural reactor operated in a cyclic pattern, 30 min on and 2.5 h off. Radiogenic heat production during this time may have caused the temperature to increase locally to values as high as 450  C. As a direct result of the elevated temperatures, hydrothermal alteration occurred locally around the reactors, leading to dissolution of quartz and growth of clay minerals together with small amounts of apatite (see Section 5.22.5.2) and zircon. The Oklo deposits are located stratigraphically near the upper part of a geological formation consisting of sandstone and conglomerate, and are overlain by black shales (typical reducing environment). Uraninite is the major U ore mineral in all of the deposits in the region. The Th content is very low (<0.3 wt% ThO2), the amounts of SiO2, TiO2, ZrO2, FeO, and Ln2O3 are generally <1 wt% each, and the maximum CaO content is 2–3 wt%. The major impurity element is radiogenic Pb. Most analyses show PbO contents of 4–7 wt%, but this may approach a value of 22 wt% in some samples. Based on an age of 2000 Ma for the uraninite mineralization, the calculated maximum amount of radiogenic Pb is about 22.5 wt% PbO; thus, the uraninites from these deposits have typically lost up to 70–80% of the expected quantity of radiogenic Pb. The localized hydrothermal alteration that occurred during the operation of the fission reactors is attributed to the presence of generally reducing and moderately saline aqueous fluids. A second period of hydrothermal alteration is associated with 1400–1600 Ma groundwater circulation in the Franceville basin, leading to mobilization of radiogenic Pb in the reactor zones and precipitation of Sb-rich galena and other sulfides (e.g., pyrite). The major phase of Pb loss from uraninite is related to the intrusion of dolerite dikes in the region at 780 Ma, during a third period of aqueous fluid flow through the rocks, resulting in widespread sulfide mineralization and precipitation of much of the radiogenic Pb in galena. For the most part, reducing conditions prevailed during the low temperature hydrothermal activity described earlier, with some dissolution and

Minerals and Natural Analogues

replacement of uraninite by coffinite, but this does not preclude the possibility of more oxidizing conditions near the ore bodies due to radiolysis of water. Oxidizing conditions certainly prevail for ore bodies located nearer the surface during recent weathering of the region, resulting in U transport and precipitation as uranyl minerals or sorption on goethite, florencite, anatase, and chlorite. The main uranyl phases identified in and around the reactor zones are torbernite, fourmarierite, francevillite, franc¸oisite, and schoepite.200,202 Plutonium and lanthanide elements released by dissolution of uraninite were precipitated mainly in phosphate minerals. Other fission and decay products, such as Pb, Tc, platinum group elements, were predominantly released by diffusion, were fixed in sulfide minerals. In general, most of these relatively insoluble elements, together with the actinides, were probably retained locally with 10–20 m of the ore bodies. It appears that Cs, Rb, Sr, Ba, Xe, Kr, and I produced during the fission cycles were also lost by diffusion processes; however, these soluble elements were transported out of the system for unknown distances. 5.22.6.4 Uranium Migration in the Koongarra Ore Deposit The Koongarra U ore deposit is one of several located in the Northern Territory of Australia. Much like the deposits described earlier, Koongarra is of sedimentary origin having formed at about 1600 Ma, but experienced a significantly higher level of metamorphism. Due to the substantial influence of meteoric water circulation and weathering, this deposit has served as a case study for U migration in the environment.203,204 The tropical climate of this geographic region produces 1.7 m of rainfall annually, leading to a distinctive dispersion of secondary uranyl minerals around the primary ore body above and below the weathering baseline, lying at a depth of about 20–30 m below the surface. This pattern has developed over the previous 1–3 Ma and includes a dispersed fan of U associated with Fe minerals in the weathered zone, extending up to 200 m from the primary ore body. The groundwater flow pattern in the Koongarra area is primarily from northwest to southeast. A deeper penetration of groundwater occurs along the Koongarra fault and rises to the southeast to intermix with groundwater in the weathered zone after passing through the area of the primary ore body. Chemically, the groundwaters are neutral to slightly acidic and characterized by dilute concentrations of MgHCO3 and SiO2 due to the dissolution of

593

chlorite in the host rocks.205 Alteration of chlorite leads to the formation of vermiculite and ultimately to kaolinite and oxidized Fe minerals such as goethite in the weathered zone. Below the weathered zone, primarily between the Koongarra fault and the main ore body, there is a zone of dispersed secondary uranyl silicate mineralization, consisting of sklodowskite, Mg[(UO2)(SiO4OH)]2(H2O)6. Directly above the ore body and the secondary silicate mineralized zone, sale´eite, Mg[(UO2)(PO4)]2(H2O)10 is the major uranyl mineral where it replaces sklodowskite and apatite.199,203 However, the nature of the U association in the weathered zone changes with flow direction and distance away from the ore deposit as sale´eite gives way to secondary Fe minerals with significant adsorbed U contents. These Fe minerals are predominantly goethite (FeOOH), hematite (Fe2O3), and ferrihydrite (Fe5HO84H2O) and they occur as vein fillings, discrete nodules, and in intimate association with clay minerals. Detailed electron microscopy indicates that there is very little sorption of U on clay minerals. The highest concentrations of U are found in association with goethite in the nodules, but this is complicated by a correlation with minor metal cations such as P and Cu, suggesting that precipitation of a uranyl phase such as torbernite may also play a role in the fixation of U.204 Considerable experimentation has been reported with regard to the adsorption of U on Fe mineral surfaces as a function of pH and the availability of complexing ligands.206,207 In U sorption experiments performed directly on weathered schist from Koongarra, it was demonstrated, via a chemical reagent technique, that selective removal of Fe oxides caused a large decrease in the sorptive capacity of the material.206 This confirmed in later sorption experiments whereby the material was characterized by analytical electron microscopy,208 revealing selective U sorption on goethite at levels 20 times higher than associated clay minerals (Figure 15). Payne et al.206 also studied the effect of synthetic ferrihydrite transformation to hematite or goethite on both the adsorption and desorption of U. Using amorphous ferrihydrite with adsorbed 238U, sorption/desorption tests were conducted with 236U as a tracer in solution. Results of these tests revealed that crystallization of hematite and goethite reduced the sorptive capacity of the material and also immobilized a significant amount of the 238U initially adsorbed on the ferrihydrite. In a related study that is of particular relevance to Koongarra, the authors showed that the addition of phosphate to the system resulted in a significant shift to lower pH in the sorption of U by

594

Minerals and Natural Analogues

pyrochlore, zirconolite, and apatite doped with Pu or 244Cm. In the specific case of zircon, the effect of long-term annealing is certainly present, but appears to have less effect on the critical dose, which may be on the order of 6–7  1016 a per milligram or even higher. Thus, for most of these minerals, the observation of increased critical dose over time can be attributed to partial annealing of collision cascade damage on geological time scales (there is a difference in dose rate on the order of 107–109 year1 between the laboratory experiments and nature’s own work). The long-term annealing rate is consistently reported to be on the order of 109 year1 for natural samples of pyrochlore, zirconolite, perovskite, and zircon. Within this group of minerals, total volume expansion associated with amorphization is lowest for pyrochlore and zirconolite (5%) and greatest for zircon (18%). Monazite is the only potential waste form phase considered herein that remains crystalline to very high a-decay dose levels. Today, more than 30 years since the late ‘Ted’ Ringwood began to develop Synroc based upon his knowledge of minerals and geological processes, there have also been significant advances in research on the geochemistry of some of the minerals noted earlier. In terms of the low measured dissolution rate as a function of pH and minimal alteration in natural systems, zirconolite remains as a promising host phase for actinides and certain fission products. Experimental studies show that titanate pyrochlore has a slightly higher dissolution rate across the same range of pH and is also more susceptible to alteration by hydrothermal and low temperature fluids in nature. Nevertheless, quantitative retention of Th and U is a characteristic feature of pyrochlores from many different geological environments. Although zirconate pyrochlores do not occur in natural systems, we note that they may exhibit dissolution rates similar to zirconolite while having the attractive property of radiation resistance. However, these pyrochlores may lack the crystal chemical flexibility of the titanates (they also require higher processing temperatures). Monazite also has low dissolution rates in laboratory experiments, but several investigations of natural samples have presented evidence of alteration at relatively low temperatures in geological environments. In view of the radiation resistance of this structure type and the ability to incorporate actinides, lanthanides, and Sr, it would be useful to investigate the aqueous geochemistry in greater detail. Other phases such as apatite and hollandite have considerable potential for use in applications 238

Figure 15 Transmission electron microscopy image of goethite-clay intergrowths from weathered rocks near the Koongarra uranium ore body, Northern Territory, Australia. This sample came from outside the U dispersion zone. When these natural ‘substrates’ are used in laboratory U sorption experiments, transmission electron microscopy– energy dispersive X-ray analyses demonstrate that most of the U is adsorbed by the goethite phase. Similar observations have been made for samples taken directly from the U dispersion zone, for example, goethite and other Fe minerals adsorb U in the dispersion zone above the ore body. Width of image ¼ 2.4 mm.

ferrihydrite. The increased sorption of U was attributed to the formation of ternary surface complexes involving 2þ PO3 4 and UO2 on ferrihydrite. Similar results were reported for the uptake of U in the presence of humic acid but, in general, both types of complexing ligands may be affected by the U concentration and the ionic strength of the aqueous solution.207

5.22.7 Conclusions With regard to radiation damage effects, it is clear that many of the oxide and silicate minerals currently of interest as analogs for nuclear waste immobilization become amorphous due to the cumulative effects of a-decay damage on the structure. This includes minerals of the pyrochlore group, zirconolite, perovskite group, brannerite, various ABO4 and AB2O6 oxides, zircon, titanite, and silicate apatite (britholite). The critical amorphization doses of these minerals are typically around 1016 a per milligram (1 dpa) for samples with ages below 600 Ma. This compares with typical critical doses of 0.3–0.5  1016 a per milligram (0.2–0.4 dpa) for synthetic

Minerals and Natural Analogues

requiring immobilization of Sr, Cs, and I in isolation from actinides and other fission products. New materials based on these phases may be in considerable demand for Generation IV nuclear power systems. Studies of uranium deposits as natural analogs for spent fuel have provided valuable information bearing on the alteration of uraninite in hydrothermal and low temperature groundwaters. Recent investigations, in particular, have provided a new perspective on the kinetics and mechanisms of uraninite alteration in oxidizing aqueous systems. In nature, it appears that the higher U oxides such as U3O8 and U4O9 are kinetically inhibited in oxidizing fluids such that the initial alteration stage of uraninite involves replacement by uranyl oxyhydroxides. The oxyhydroxide phases are then replaced during successive stages of alteration, depending on (a) the activity ratios of the large metal cations in solution, [Mþ/Hþ] and (b) the activities of aqueous silicate, carbonate, phosphate, and other species. The uranium ore deposits in Gabon (e.g., Oklo) and in the Northern Territory of Australia (e.g., Koongarra) are very similar with regard to many of their geological features. However, conditions were ideal for cyclic natural fission reactions to occur in the former due to the scale of the ore bodies and concentration of U within them, the presence of water as a moderator, the host sandstone as a neutron reflector, and the absence of neutron absorbers. The natural fission reactors have provided an ideal laboratory for the study of fission and decay products in a sedimentary basin. In the case of Koongarra, differences in the water content and the amounts of neutron absorbers (e.g., Gd) present in the rocks probably played a major role in preventing the ores from reaching criticality. The extensive weathering and consistent groundwater flow pattern in the area resulted in a distinctive fan of secondary U mineralization and adsorption in the weathered zone extending up to 200 m from the primary ore deposit.

References 1.

2. 3.

Degueldre, C.; Kasemeyer, U.; Botta, F.; Ledergerber, G. In Scientific Basis for Nuclear Waste Management XIX; Materials Research Society Symposium Proceedings; Murphy, W. M., Knecht, D. A., Eds.; Materials Research Society: Pittsburgh, PA, 1996; Vol. 412, pp 15–23. Harker, A. B. In Radioactive Waste Forms for the Future; Lutze, W., Ewing, R. C., Eds.; North-Holland: Amsterdam, 1988; pp 335–392. Fielding, P. E.; White, T. J. J. Mater. Res. 1987, 2, 387–414.

595

4. Ringwood, A. E.; Kesson, S. E.; Reeve, K. D.; Levins, D. M.; Ramm, E. J. In Radioactive Waste Forms for the Future; Lutze, W., Ewing, R. C., Eds.; North-Holland: Amsterdam, 1988; pp 233–334. 5. Ball, C. J.; Buykx, W. J.; Dickson, F. J.; et al. J. Am. Ceram. Soc. 1989, 72, 404–414. 6. Vance, E. R. MRS Bull. 1994, XIX, 28–32. 7. Carter, M. L.; Vance, E. R.; Mitchell, D. R. G.; Hanna, J. V.; Zhang, Z.; Loi, E. J. Mater. Res. 2002, 17, 2578–2589. 8. Ewing, R. C.; Lutze, W.; Weber, W. J. J. Mater. Res. 1995, 10, 243–246. 9. Boatner, L. A.; Sales, B. C. In Radioactive Waste Forms for the Future; Lutze, W., Ewing, R. C., Eds.; NorthHolland: Amsterdam, 1988; pp 495–564. 10. Kinoshita, H.; Kuramoto, K.; Uno, M.; Yamanaka, S.; Mitamura, H.; Banba, T. In Scientific Basis for Nuclear Waste Management XXIII; Materials Research Society Symposium Proceedings; Smith, R. W., Shoesmith, D. W., Eds.; Materials Research Society: Warrendale, PA, 2000; Vol. 608, pp 393–398. 11. Ewing, R. C.; Weber, W. J.; Lian, J. J. Appl. Phys. 2004, 95, 5949–5971. 12. Subramanian, M. A.; Aravamudan, G.; Subba Rao, G. V. Progr. Solid State Chem. 1983, 15, 55–143. 13. Chakoumakos, B. C. J. Solid State Chem. 1984, 53, 120–129. 14. Smith, K. L.; Lumpkin, G. R. In Defects and Processes in the Solid State: Geoscience Applications; The McLaren Volume; Boland, J. N., Fitz Gerald, J. D., Eds.; Elsevier: Amsterdam, 1993; pp 401–422. 15. Krivokoneva, G. K.; Sidorenko, G. A. Geochem. Int. 1971, 8, 113–122. 16. Lumpkin, G. R.; Ewing, R. C. Phys. Chem. Miner. 1988, 16, 2–20. 17. Lumpkin, G. R.; Hart, K. P.; McGlinn, P. J.; Payne, T. E.; Giere´, R.; Williams, C. T. Radiochim. Acta 1994, 66/67, 469–474. 18. Salje, E. K. H.; Chrosch, J.; Ewing, R. C. Am. Mineral. 1999, 84, 1107–1116. 19. Greegor, R. B.; Lytle, F. W.; Chakoumakos, B. C.; Lumpkin, G. R.; Ewing, R. C. In Scientific Basis for Nuclear Waste Management VIII; Materials Research Society Symposium Proceedings; Jantzen, C. M., Stone, J. A., Ewing, R. C., Eds.; Materials Research Society: Pittsburgh, PA, 1985; Vol. 44, pp 655–662. 20. Greegor, R. B.; Lytle, F. W.; Chakoumakos, B. C.; Lumpkin, G. R.; Ewing, R. C. In Scientific Basis for Nuclear Waste Management IX; Materials Research Society Symposium Proceeding; Werme, L. O., Ed.; Materials Research Society: Pittsburgh, PA, 1985; Vol. 50, pp 388–392. 21. Greegor, R. B.; Lytle, F. W.; Chakoumakos, B. C.; et al. In Scientific Basis for Nuclear Waste Management X; Materials Research Society Symposium Proceedings; Bates, J. K., Seefeldt, W. B., Eds.; Materials Research Society: Pittsburgh, PA, 1987; Vol. 84, pp 645–658. 22. Greegor, R. B.; Lytle, F. W.; Chakoumakos, B. C.; Lumpkin, G. R.; Warner, J. K.; Ewing, R. C. In Scientific Basis for Nuclear Waste Management XII; Materials Research Society Symposium Proceedings; Lutze, W., Ewing, R. C., Eds.; Materials Research Society: Pittsburgh, PA, 1989; Vol. 127, pp 655–662. 23. Lumpkin, G. R.; Foltyn, E. M.; Ewing, R. C. J. Nucl. Mater. 1986, 139, 113–120. 24. Lumpkin, G. R.; Ewing, R. C.; Chakoumakos, B. C.; et al. J. Mater. Res. 1986, 1, 564–576. 25. Clinard, F. W., Jr.; Hobbs, L. W.; Land, C. C.; Peterson, D. E.; Rohr, D. L.; Roof, R. B. J. Nucl. Mater. 1982, 105, 248–256.

596 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

37. 38. 39. 40. 41. 42. 43. 44. 45. 46.

47. 48.

49.

50. 51. 52. 53.

Minerals and Natural Analogues Clinard, F. W., Jr.; Rohr, D. L.; Roof, R. B. Nucl. Instrum. Methods Phys. Res. B 1984, 1, 581–586. Weber, W. J.; Wald, J. W.; Matzke, H. J. Nucl. Mater. 1986, 138, 196–209. Strachan, D. M.; Scheele, R. D.; Buck, E. C.; et al. J. Nucl. Mater. 2005, 345, 109–135. Roberts, S. K.; Bourcier, W. L.; Shaw, H. F. Radiochim. Acta 2000, 88, 539–543. Zhang, Y.; Hart, K. P.; Bourcier, W. L.; et al. J. Nucl. Mater. 2001, 289, 254–262. Geisler, T.; Berndt, J.; Meyer, H. W.; Pollok, K.; Putnis, A. Mineral. Mag. 2004, 68, 905–922. Geisler, T.; Po¨ml, P.; Stephan, T.; Janssen, A.; Putnis, A. Am. Mineral. 2005, 90, 1683–1687. Geisler, T.; Seydoux-Guillaume, A. M.; Po¨ml, P.; et al. J. Nucl. Mater. 2005, 344, 17–23. Po¨ml, P.; Menneken, M.; Stephan, T.; Niedermeier, D.; Geisler, T.; Putnis, A. Geochim. Cosmochim. Acta 2007, 71, 3311–3322. Nasraoui, M.; Bilal, E.; Gibert, R. Mineral. Mag. 1999, 63, 567–578. Lumpkin, G. R.; Mariano, A. N. In Scientific Basis for Nuclear Waste Management XIX; Materials Research Society Symposium Proceedings; Murphy, W. M., Knecht, D. A., Eds.; Materials Research Society: Pittsburgh, PA, 1996; Vol. 412, pp 831–838. Lumpkin, G. R.; Ewing, R. C. Am. Mineral. 1992, 77, 179–188. Lumpkin, G. R.; Ewing, R. C. Am. Mineral. 1995, 80, 732–743. Lumpkin, G. R.; Ewing, R. C. Am. Mineral. 1996, 81, 1237–1248. Wall, F.; Williams, C. T.; Woolley, A. R.; Nasraoui, M. Mineral. Mag. 1996, 60, 731–750. Williams, C. T.; Wall, F.; Woolley, A. R.; Phillipo, S. J. Afr. Earth Sci. 1997, 25, 137–145. Chakhmouradian, A. R.; Mitchell, R. H. Mineral. Mag. 1998, 62, 769–782. Wise, M. A.; Cˇerny´, P. Mineral. Petrol. 1990, 43, 83–98. Ohnenstetter, D.; Piantone, P. Can. Mineral. 1992, 30, 771–784. Wall, F.; Williams, C. T.; Woolley, A. R. In Mineral Deposits: Processes to Processing; Stanley, C. J., Ed.; Balkema: Rotterdam, 1999; Vol. 1, pp 687–690. Lumpkin, G. R.; Payne, T. E.; Fenton, B. R.; Waite, T. D. In Scientific Basis for Nuclear Waste Management XXII; Materials Research Society Symposium Proceedings; Wronkiewicz, D. J., Lee, J. H., Eds.; Materials Research Society: Warrendale, PA, 1999; Vol. 556, pp 1067–1074. De Vito, C.; Pezzotta, F.; Ferrini, V.; Aurisicchio, C. Can. Mineral. 2006, 44, 87–103. Giere´, R.; Buck, E. C.; Guggenheim, R.; Mathys, D.; Reusser, E.; Marques, J. In Scientific Basis for Nuclear Waste Management XXIV; Materials Research Society Symposium Proceedings; Hart, K. P., Lumpkin, G. R., Eds.; Materials Research Society: Warrendale, PA, 2001; Vol. 663, pp 935–944. Burns, P. C. In Uranium: Mineralogy, Geochemistry, and the Environment. Reviews in Mineralogy; Burns, P. C., Finch, R., Eds.; Mineralogical Society of America: Washington, DC, 1999; Vol. 38, pp 23–90. Giere´, R.; Williams, C. T.; Lumpkin, G. R. Schweizerische Mineralogische und Petrographische Mitteilungen 1998, 78, 433–459. Bellatreccia, F.; Della Ventura, G.; Caprilli, E.; Williams, C. T.; Parodi, G. C. Mineral. Mag. 1999, 63, 649–660. Farges, F.; Ewing, R. C.; Brown, G. E. J. Mater. Res. 1993, 8, 1983–1995. Farges, F. Am. Mineral. 1997, 82, 44–50.

Lumpkin, G. R.; Smith, K. L.; Giere´, R. Micron 1997, 28, 57–68. 55. Lumpkin, G. R.; Colella, M.; Smith, K. L.; Mitchell, R. H.; Larsen, A. O. In Scientific Basis for Nuclear Waste Management XXI; Materials Research Society Symposium Proceedings; McKinley, I. G., McCombie, C., Eds.; Materials Research Society: Warrendale, PA, 1998; Vol. 506, pp 207–214. 56. Lumpkin, G. R.; Smith, K. L.; Blackford, M. G.; Giere´, R.; Williams, C. T. In Scientific Basis for Nuclear Waste Management XXI; Materials Research Society Symposium Proceedings; McKinley, I. G., McCombie, C., Eds.; Materials Research Society: Warrendale, PA, 1998; Vol. 506, pp 215–222. 57. Wald, J. W.; Offermann, P. In Scientific Basis for Nuclear Waste Management V; Lutze, W., Ed.; Elsevier Science Publishing: New York, NY, 1982; 369–378. 58. Clinard, F. W., Jr.; Peterson, D. E.; Rohr, D. L.; Hobbs, L. W. J. Nucl. Mater. 1984, 126, 245–254. 59. Strachan, D. M.; Scheele, R. D.; Buck, E. C.; et al. J. Nucl. Mater. 2008, 372, 16–31. 60. Oversby, V. M.; Ringwood, A. E. Radioactive Waste Mgmt 1981, 1, 289–307. 61. Ewing, R. C.; Haaker, R. F.; Headley, T. J.; Hlava, P. F. In Scientific Basis for Nuclear Waste Management V; Topp, S. V., Ed.; Elsevier: New York, NY, 1982; pp 249–256. 62. Rasmussen, B.; Fletcher, I. R. Geology 2004, 32, 785–788. 63. Hart, K. P.; Lumpkin, G. R.; Giere´, R.; Williams, C. T.; McGlinn, P. J.; Payne, T. E. Radiochim. Acta 1996, 74, 309–312. 64. Bulakh, A. G.; Nesterov, A. R.; Williams, C. T.; Anisimov, I. S. Mineral. Mag. 1998, 62, 837–846. 65. Williams, C. T.; Bulakh, A. G.; Giere´, R.; Lumpkin, G. R.; Mariano, A. N. In Scientific Basis for Nuclear Waste Management XXIV; Materials Research Society Symposium Proceedings; Hart, K. P., Lumpkin, G. R., Eds.; Materials Research Society: Warrendale, PA, 2001; Vol. 663, pp 945–952. 66. Giere´, R.; Williams, C. T. Contrib. Mineral. Petrol. 1992, 112, 83–100. 67. Pan, Y. Can. Mineral. 1997, 35, 105–118. 68. Malmstro¨m, J. C. Beitra¨ge zur Geologie der Schweiz. Geotechnische Serie 2000, 93, 130. 69. Po¨ml, P. Unpublished Ph.D. Thesis, University of Mu¨nster: Germany, 2008; p 87. 70. Szyman´ski, J. T.; Scott, J. D. Can. Mineral. 1982, 20, 271–279. 71. James, M.; Carter, M. L.; Watson, J. N. J. Solid State Chem. 2003, 174, 329–333. 72. Vance, E. R.; Watson, J. N.; Carter, M. L.; et al. In Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries V. Ceramic Transactions; Chandler, G. T., Feng, X., Eds.; American Ceramic Society: Westerville, OH, 2000; Vol. 107, pp 561–568. 73. Colella, M.; Lumpkin, G. R.; Zhang, Z.; Buck, E. C.; Smith, K. L. Phys. Chem. Miner. 2005, 32, 52–64. 74. Lumpkin, G. R.; Leung, S. H. F.; Colella, M. In Scientific Basis for Nuclear Waste Management XXIII; Materials Research Society Symposium Proceedings; Smith, R. W., Shoesmith, D. W., Eds.; Materials Research Society: Warrendale, PA, 2000; Vol. 608, pp 359–365. 75. Graeser, S.; Guggenheim, R. Schweizerische Mineralogische und Petrographische Mitteilungen 1990, 70, 325–331. 76. Lumpkin, G. R. J. Nucl. Mater. 2001, 289, 136–166. 77. Zhang, Y.; Thomas, B. S.; Lumpkin, G. R.; et al. J. Nucl. Mater. 2003, 321, 1–7. 54.

Minerals and Natural Analogues 78. 79. 80. 81. 82. 83.

84. 85. 86. 87. 88. 89. 90.

91.

92.

93. 94. 95.

96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107.

Zhang, Y.; Lumpkin, G. R.; Li, H.; et al. J. Nucl. Mater. 2006, 350, 293–300. Mitchell, R. H. Perovskites: Modern and Ancient; Almaz Press: Ontario, 2002. Mitchell, R. H.; Chakhmouradian, A. R. Mineral. Mag. 1998, 62, 341–353. Mitchell, R. H.; Chakhmouradian, A. R. Can. Mineral. 1998, 36, 939–952. Mosley, W. C. J. Am. Ceram. Soc. 1971, 54, 475–479. Mariano, A. N. In Geochemistry and Mineralogy of Rare Earth Elements. Reviews in Mineralogy; Lipin, B. R., McKay, G. A., Eds.; The Mineralogical Society of America: Washington, DC, 1989; Vol. 21, pp 309–337. Banfield, J. F.; Veblen, D. R. Am. Mineral. 1992, 77, 545–557. Nesbitt, H. W.; Bancroft, G. M.; Fyfe, W. S.; Karkhanis, S. N.; Nishijima, A. Nature 1981, 289, 358–362. Chakhmouradian, A. R.; Mitchell, R. H.; Pankov, A. V.; Chukanov, N. V. Mineral. Mag. 1999, 63, 519–534. Kamizono, H.; Hayakawa, I.; Muraoka, S. J. Am. Ceram. Soc. 1991, 74, 863–864. Myhra, S.; Bishop, H. E.; Rivie´re, J. C.; Stephenson, M. J. Mater. Sci. 1987, 22, 3217–3226. Myhra, S.; Delogu, P.; Giorgi, R.; Rivie´re, J. C. J. Mater. Sci. 1988, 23, 1514–1520. Pham, D. K.; Neall, F. B.; Myhra, S.; Smart, R. C.; Turner, P. S. In Scientific Basis for Nuclear Waste Management XII; Materials Research Society Symposium Proceedings; Lutze, W., Ewing, R. C., Eds.; Materials Research Society: Pittsburgh, PA, 1989; Vol. 127, pp 231–240. Turner, P. S.; Jones, C. F.; Myhra, S.; Neall, F. B.; Pham, D. K.; Smart, R. C. In Surfaces and Interfaces of Ceramic Materials; Dufour, L. C., Monty, C., PetotErvas, G., Eds.; Kluwer: Dordrecht, 1989; pp 663–690. McGlinn, P. J.; Hart, K. P.; Loi, E. H.; Vance, E. R. In Scientific Basis for Nuclear Waste Management XVIII; Materials Research Society Symposium Proceedings; Murakami, T., Ewing, R. C., Eds.; Materials Research Society: Pittsburgh, PA, 1995; Vol. 353, pp 847–854. Zhang, Z.; Blackford, M. G.; Lumpkin, G. R.; Smith, K. L.; Vance, E. R. J. Mater. Res. 2005, 20, 2462–2473. Lumpkin, G. R. J. Nucl. Mater. 1999, 274, 206–217. Gong, W. L.; Ewing, R. C.; Wang, L. M.; Xie, H. S. In Scientific Basis for Nuclear Waste Management XVIII; Materials Research Society Symposium Proceedings; Murakami, T., Ewing, R. C., Eds.; Materials Research Society: Pittsburgh, PA, 1995; Vol. 353, pp 807–815. Malczewski, D.; Frackowiak, J. E.; Galuskin, E. V. Hyperfine Interact. 2005, 166, 529–536. Giere´, R.; Williams, C. T.; Wirth, R.; Ruschel, K. Chem. Geol. 2009, 261, 333–345. Tomasic, N.; Gajovic, A.; Bermanec, V.; et al. Phys. Chem. Miner. 2006, 33, 145–159. Lumpkin, G. R. Can. Mineral. 1998, 36, 339–353. Ewing, R. C. Geochim. Cosmochim. Acta 1975, 39, 521–530. Ewing, R. C.; Ehlmann, A. J. Can. Mineral. 1975, 13, 1–7. Lumpkin, G. R. J. Nucl. Mater. 1992, 190, 302–311. Nakai, I.; Akimoto, J.; Imafuku, M.; Miyawaki, R.; Sugitani, Y.; Koto, K. Phys. Chem. Miner. 1987, 15, 113–124. Greegor, R. B.; Lytle, F. W.; Ewing, R. C.; Haaker, R. F. Nucl. Instrum. Methods Phys. Res. B 1984, 1, 587–594. Kesson, S. E.; White, T. J. Proc. R. Soc. Lond. A 1986, 405, 73–101. Kesson, S. E.; White, T. J. Proc. R. Soc. Lond. A 1986, 408, 295–319. Cheary, R. W. Acta Crystallogr. 1987, B43, 28–34.

108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140.

597

Myhra, S.; White, T. J.; Kesson, S. E.; Rivie´re, J. C. Am. Mineral. 1988, 73, 161–167. Schreyer, W.; Fransolet, A. M.; Bernhardt, H. J. Contrib. Mineral. Petrol. 2001, 141, 560–570. Sinclair, W.; McLaughlin, G. M. Acta Crystallogr. 1982, B38, 245–246. Post, J. E.; Von Dreele, R. B.; Buseck, P. R. Acta Crystallogr. 1982, B38, 1056–1065. Pham, D. K.; Myhra, S.; Turner, P. S. J. Mater. Res. 1994, 9, 3174–3181. Speer, J. A. In Orthosilicates. Reviews in Mineralogy, 2nd ed.; Ribbe, P. H., Ed.; Mineralogical Society of America: Washington, DC, 1982; Vol. 5, pp 67–112. Aines, R. D.; Rossman, G. R. Am. Mineral. 1986, 71, 1186–1193. Anderson, E. B.; Burakov, B. E.; Pazukhin, E. M. Radiochim. Acta 1993, 60, 149–151. Burakov, B. E.; Anderson, E. B.; Galkin, B. Y.; Pazukhin, E. M.; Shabalev, S. I. Radiochim. Acta 1994, 65, 199–202. Geisler; et al. Eur. J. Mineral. 2005. Hurley, P. M.; Fairbairn, H. W. Geol. Soc. Am. Bull. 1953, 64, 659–674. Holland, H. D.; Gottfried, D. Acta Crystallogr. 1955, 8, 291–300. Murakami, T.; Chakoumakos, B. C.; Ewing, R. C.; Lumpkin, G. R.; Weber, W. J. Am. Mineral. 1991, 76, 1510–1532. Weber, W. J.; Ewing, R. C.; Meldrum, A. J. Nucl. Mater. 1997, 250, 147–155. Chakoumakos, B. C.; Murakami, T.; Lumpkin, G. R.; Ewing, R. C. Science 1987, 236, 1556–1559. Chakoumakos, B. C.; Oliver, W. C.; Lumpkin, G. R.; Ewing, R. C. Radiat. Eff. Defect. Solid. 1991, 118, 393–403. Ellsworth, S.; Navrotsky, A.; Ewing, R. C. Phys. Chem. Miner. 1994, 21, 140–149. McLaren, A. C.; Fitzgerald, J. D.; Williams, I. S. Geochim. Cosmochim. Acta 1994, 58, 993–1005. Weber, W. J. J. Mater. Res. 1990, 5, 2687–2697. Weber, W. J.; Ewing, R. C.; Wang, L. M. J. Mater. Res. 1994, 9, 688–698. Rı´os, S.; Salje, E. K. H.; Zhang, M.; Ewing, R. C. J. Phys. Condens. Matter 2000, 12, 2401–2412. Palenik, C. S.; Nasdala, L.; Ewing, R. C. Am. Mineral. 2003, 88, 770–781. Farges, F.; Calas, G. Am. Mineral. 1991, 76, 60–73. Begg, B. D.; Hess, N. J.; Weber, W. J.; Conradson, S. D.; Schweiger, M. J.; Ewing, R. C. J. Nucl. Mater. 2000, 278, 212–224. Wayne, D. M.; Sinha, A. K. Contrib. Mineral. Petrol. 1988, 98, 109–121. Geisler, T.; Zhang, M.; Salje, E. K. H. J. Nucl. Mater. 2003, 320, 280–291. Pidgeon, R. T.; O’Neil, J. R.; Silver, L. T. Science 1966, 154, 1538–1540. Sinha, A. K.; Wayne, D. M.; Hewitt, D. A. Geochim. Cosmochim. Acta 1992, 56, 3551–3560. Rizvanova, N. G.; Levchenkov, O. A.; Belous, A. E.; et al. Contrib. Mineral. Petrol. 2000, 139, 101–114. Geisler, T.; Ulonska, M.; Schleicher, H.; Pidgeon, R. T.; van Bronswijk, W. Contrib. Mineral. Petrol. 2001, 141, 53–65. Geisler, T.; Pidgeon, R. T.; van Bronswijk, W.; Kurtz, R. Chem. Geol. 2002, 191, 141–154. Geisler, T.; Pidgeon, R. T.; Kurtz, R.; van Bronswijk, W.; Schleicher, H. Am. Mineral. 2003, 86, 1496–1518. Geisler, T.; Seydoux-Guillaume, A. M.; Wiedenbeck, M.; et al. Am. Mineral. 2004, 89, 1341–1347.

598

Minerals and Natural Analogues

141. Ewing, R. C.; Haaker, R. F.; Lutze, W. In Scientific Basis for Nuclear Waste Management V; Lutze, W., Ed.; Elsevier: New York, NY, 1982; pp 389–397. 142. Helean, K. B.; Lutze, W.; Ewing, R. C. In Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries IV. Ceramic Transactions; Marra, J. C., Chandler, G. T., Eds.; American Ceramic Society: Westerville, OH, 1999; Vol. 93, pp 297–304. 143. Geisler, T.; Trachenko, K.; Rı´os, S.; Dove, M.; Salje, E. K. H. J. Phys. Condens. Matter 2003, 15, L597–L605. 144. Trachenko, K.; Dove, M. T.; Salje, E. K. H. Phys. Rev. B 2002, 65, 180102(R). 145. Geisler, T.; Burakov, B. E.; Yagovkina, M.; et al. J. Nucl. Mater. 2005, 336, 22–30. 146. Foord, E. E.; Cobban, R. R.; Brownfield, I. K. Mineral. Mag. 1985, 49, 729–731. 147. Lumpkin, G. R.; Chakoumakos, B. C. Am. Mineral. 1988, 73, 1405–1419. 148. Frondel, C. US Geol. Surv. Bull. 1958, 1064, 400. 149. Staatz, M. H.; Adams, J. W.; Wahlberg, J. S. J. Res. US Geol. Surv. 1976, 4, 575–582. 150. Hayward, P. J. In Radioactive Waste Forms for the Future; Lutze, W., Ewing, R. C., Eds.; North-Holland: Amsterdam, 1988; pp 427–493. 151. Hawthorne, F. C.; Groat, L. A.; Raudsepp, M.; et al. Am. Mineral. 1991, 76, 370–396. 152. Vance, E. R.; Metson, J. B. Phys. Chem. Miner. 1985, 12, 255–260. 153. Janeczek, J.; Eby, R. K. Phys. Chem. Miner. 1993, 19, 343–356. 154. Giere´, R.; Sorensen, S. S. In Epidotes. Reviews in Mineralogy and Geochemistry; Liebscher, A., Franz, G., Eds.; Mineralogical Society of America: Chantilly, VA, 2004; Vol. 56, pp 431–493. 155. Wood, S. A.; Ricketts, A. Can. Mineral. 2000, 38, 81–100. 156. Boatner, L. A. In Phosphates: Geochemical, Geobiological, and Materials Importance. Reviews in Mineralogy and Geochemistry; Kohn, M. J., Rakovan, J., Hughes, J. M., Eds.; The Mineralogical Society of America: Washington, DC, 2002; Vol. 48, pp 87–121. 157. Fo¨rster, H. J. Am. Mineral. 1998, 83, 259–272. 158. Burakov, B. E.; Yagovkina, M. A.; Garbuzov, V. M.; Kitsay, A. A.; Zirlin, V. A. In Scientific Basis for Nuclear Waste Management XXVIII; Materials Research Society Symposium Proceedings; Hanchar, J. M., StroesGascoyne, S., Browning, L., Eds.; Materials Research Society: Warrendale, PA, 2004; Vol. 824, pp 219–224. 159. Poitrasson, F.; Chenery, S.; Bland, D. J. Earth Planet. Sci. Lett. 1996, 145, 79–96. 160. Poitrasson, F.; Chenery, S.; Shepherd, T. J. Geochim. Cosmochim. Acta 2000, 64, 3283–3297. 161. Townsend, K. J.; Miller, C. G.; D’Andrea, J. L.; Ayers, J. C.; Harrison, T. M.; Coath, C. D. Chem. Geol. 2000, 172, 95–112. 162. Mathieu, R.; Zetterstro¨m, L.; Cuney, M.; GauthierLafaye, F.; Hidaka, H. Chem. Geol. 2001, 171, 147–171. 163. Read, D.; Williams, C. T. Mineral. Mag. 2001, 65, 589–601. 164. Read, D.; Andreoli, M. A. G.; Knoper, M.; Williams, C. T.; Jarvis, N. Eur. J. Mineral. 2002, 14, 487–498. 165. Cuney, M.; Mathieu, R. Geology 2000, 28, 743–746. 166. Hecht, L.; Cuney, M. Miner. Deposita 2000, 35, 791–795. 167. Hetherington, C. J.; Harlov, D. E. Am. Mineral. 2008, 93, 806–820. 168. Ayers, J. C.; Watson, E. B. Philos. Trans. R. Soc. Lond. A 1991, 335, 365–375.

169. 170. 171. 172. 173.

174. 175. 176. 177. 178.

179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192.

193. 194.

195.

Wood, S. A.; Williams-Jones, A. E. Chem. Geol. 1994, 115, 47–60. Oelkers, E. H.; Poitrasson, F. Chem. Geol. 2002, 191, 73–87. Teufel, S.; Heinrich, W. Chem. Geol. 1997, 137, 273–281. Seydoux-Guillaume, S.; Paquette, J. L.; Wiedenbeck, M.; Montel, J. M.; Heinrich, W. Chem. Geol. 2002, 191, 165–181. Pan, Y.; Fleet, M. E. In Phosphates: Geochemical, Geobiological, and Materials Importance. Reviews in Mineralogy and Geochemistry; Kohn, M. J., Rakovan, J., Hughes, J. M., Eds.; The Mineralogical Society of America: Washington, DC, 2002; Vol. 48, pp 13–49. Bros, R.; Carpe´na, J.; Sere, V.; Beltritti, A. Radiochim. Acta 1996, 74, 277–282. Hidaka, H. J. Nucl. Radiochem. Sci. 2007, 8, 99–103. Arden, K. M.; Halden, N. M. Can. Mineral. 1999, 37, 1239–1253. Carpe´na, J.; Kienast, J. R.; Ouzegane, K.; Jehanno, C. Geol. Soc. Am. Bull. 1998, 100, 1237–1243. Yudintseva, T. S. In Scientific Basis for Nuclear Waste Management XXIX; Materials Research Society Symposium Proceedings; Van Iseghem, P., Ed.; Materials Research Society: Warrendale, PA, 2006; Vol. 932, pp 1049–1055. Weber, W. J. Radiat. Eff. 1983, 77, 295–308. Weber, W. J.; Matzke, H. Mater. Lett. 1986, 5, 9–16. Boudreau, A. E.; McCallum, I. S. Am. Mineral. 1990, 75, 687–693. Valsami-Jones, E.; Ragnarsdottir, K. V.; Putnis, A.; Bosbach, D.; Kemp, A. J.; Cressey, G. Chem. Geol. 1998, 151, 215–233. Chaı¨rat, C.; Oelkers, E. H.; Schott, J.; Lartigue, J. E. J. Nucl. Mater. 2006, 354, 14–27. Terra, O.; Dacheux, N.; Audubert, F.; Podor, R. J. Nucl. Mater. 2006, 352, 224–232. Brownfield, M. E.; Foord, E. E.; Sutley, S. J.; Botinelly, T. Am. Mineral. 1993, 78, 653–656. Sugantha, M.; Varadaraju, U. V.; Subba Rao, G. V. J. Solid State Chem. 1983, 111, 33–40. Buvaneswari, G.; Varadaraju, U. V. J. Solid State Chem. 1999, 145, 227–234. Hawkins, H. T.; Spearing, D. R.; Veirs, D. K.; et al. Chem. Mater. 1999, 11, 2851–2857. Bykov, D. M.; Gobechiya, E. R.; Kabalov, Y. K.; Orlova, A. I.; Tomilin, S. V. J. Solid State Chem. 2006, 179, 3101–3106. Roy, R.; Vance, E. R.; Alamo, J. Mater. Res. Bull. 1982, 17, 585–589. Scheetz, B. E.; Agrawal, D. K.; Breval, E.; Roy, R. Waste Manag. 1994, 14, 489–505. Zyryanov, V. N.; Vance, E. R. In Scientific Basis for Nuclear Waste Management XX; Materials Research Society Symposium Proceedings; Gray, W. J., Triay, I. R., Eds.; Materials Research Society: Pittsburgh, PA, 1997; Vol. 465, pp 409–415. Bois, L.; Guittet, M. J.; Carrot, F.; Trocellier, P.; Gautier-Soyer, M. J. Nucl. Mater. 2001, 297, 129–137. Harrison, T. M.; Catlos, E. J.; Montel, J. M. In Phosphates: Geochemical, Geobiological, and Materials Importance. Reviews in Mineralogy and Geochemistry; Kohn, M. J., Rakovan, J., Hughes, J. M., Eds.; The Mineralogical Society of America: Washington, DC, 2002; Vol. 48, 524–558. Spear, F. S.; Pyle, J. M. In Phosphates: Geochemical, Geobiological, and Materials Importance. Reviews in Mineralogy and Geochemistry; Kohn, M. J., Rakovan, J., Hughes, J. M., Eds.; The Mineralogical Society of America: Washington, DC, 2002; Vol. 48, pp 293–335.

Minerals and Natural Analogues 196. Fo¨rster, H. J. Am. Mineral. 1998, 83, 1302–1315. 197. Luo, J. S.; Liu, G. K. J. Mater. Res. 2001, 16, 366–372. 198. Wronkiewicz, D. J.; Buck, E. C. In Uranium: Mineralogy, Geochemistry, and the Environment. Reviews in Mineralogy; Burns, P. C., Finch, R., Eds.; Mineralogical Society of America: Washington, DC, 1999; Vol. 38, pp 475–497. 199. Finch, R.; Murakami, T. In Uranium: Mineralogy, Geochemistry, and the Environment. Reviews in Mineralogy; Burns, P. C., Finch, R., Eds.; Mineralogical Society of America: Washington, DC, 1999; Vol. 38, pp 91–179. 200. Janeczek, J. In Uranium: Mineralogy, Geochemistry, and the Environment. Reviews in Mineralogy; Burns, P. C., Finch, R., Eds.; Mineralogical Society of America: Washington, DC, 1999; Vol. 38, pp 321–392. 201. Meshik, A. P.; Hohenberg, C. M.; Pravdivtseva, O. V. Phys. Rev. Lett. 2004, 93, 182302. 202. Janeczek, J.; Ewing, R. C. Mineral. Mag. 1996, 60, 665–669. 203. Murakami, T.; Ohnuki, T.; Isobe, H.; Sato, T. Am. Mineral. 1997, 82, 888–899. 204. Sato, T.; Murakami, T.; Yanase, N.; Isobe, H.; Payne, T. E.; Airey, P. L. Environ. Sci. Technol. 1997, 31, 2854–2858. 205. Yanase, N.; Payne, T. E.; Sekine, K. Geochem. J. 1995, 29, 1–29. 206. Payne, T. E.; Davis, J. A.; Waite, T. D. Radiochim. Acta 1994, 66/67, 297–303. 207. Payne, T. E.; Davis, J. A.; Waite, T. D. Radiochim. Acta 1996, 74, 239–243. 208. Lumpkin, G. R.; Day, R. A.; McGlinn, P. J.; Payne, T. E.; Giere´, R.; Williams, C. T. In Scientific Basis for Nuclear Waste Management XXII; Materials Research Society Symposium Proceedings; Wronkiewicz, D. J., Lee, J. H., Eds.; Materials Research Society: Warrendale, PA, 1999; Vol. 556, pp 793–800.

Appendix: List of Mineral Names and Compositions Oxides Aeschynite, (Ce,Nd,Y,Fe)(Ti,Nb,Ta)2(O,OH)6. Anatase, Brookite, and Rutile, TiO2. Rutile, in particular, may contain V, Cr, and Fe. Baddeleyite, ZrO2. Brannerite, (U,Th,Ca)Ti2O6. Calciotantite, CaTa4O11. Calzirtite, CaTi2Zr5O16. Cerianite, CeO2. Columbite–tantalite, (Mn,Fe)(Nb,Ta,Ti)2O6. Crichtonite, (Sr,La,Ce,Y)(Ti,Fe,Mn)21O38. Davidite, (La,Ce,Ca)(Y,U)(Ti,Fe,Mn)20O38. Euxenite, (Y,Ca,Ce)(Ti,Nb,Ta)2O6. Fergusonite, (Y,Yb,Gd,Nd,Ce)(Nb,Ta)O4. Ferrihydrite, Fe5HO84H2O. Fersmite, CaNb2O6. Goethite, FeOOH. Hematite, Fe2O3.

599

Hollandite, (Ba,Pb,K,Na)(Mn,Fe,Ti,Al)8O16. Ilmenite, FeTiO3. Kassite, CaTi2O4(OH)2H2O. Liandratite, UNb2O8. Lucasite, (La,Ce,Nd,Pr)Ti2O6x(OH,F)xH2O. Magnetite, Fe3O4. Perovskite, (Ca,Ba,Na,La,Ce,Nd)(Ti,Nb,Fe)O3. Polycrase, (Y,Ca,Ce,U,Th)(Ti,Nb,Ta)2O6. Pyrochlore, (Na,Ca,Ce,Th,U)(Nb,Ta,Ti,Zr,Sn)2O6. Priderite, (K,Ba)(Ti,Fe)8O16. Samarskite, (Y,Dy,Er,Yb,U,Fe)(Nb,Ta,Ti)O4. Uraninite, UO2. Zirconolite, (Ca,Nd,Gd,Th,U)(Zr,Hf,Y)(Ti,Nb, Fe,Al)2O7. Silicates Allanite, (La,Ce,Ca,Y)(Al,Fe)3(SiO4)3(OH). Chlorite, (Mg,Fe,Al)6Si3AlO10(OH,O)8. Clinozoisite, CaAl3(SiO4)3(OH). Coffinite, USiO4. Epidote, CaFe3(SiO4)3(OH). Kaolinite, Al2Si2O5(OH)4. Olivine, (Mg,Fe)SiO4. Quartz, SiO2. Serpentine, (Mg,Fe)3Si2O5(OH)4, there are several polymorphs. Thorite, ThSiO4. Titanite, CaTiSiO5. Vermiculite, (Mg,Fe,Al)3(Si,Al)4O10(OH)24H2O. Zircon, ZrSiO4. Phosphates, Arsenates, Vanadates Chlorapatite, Ca5(PO4)3Cl. Florencite, (La,Ce,Nd,Pr)Al3(PO4)2(OH)6. Fluorapatite, Ca5(PO4)3F. Hydroxyapatite, Ca5(PO4)3OH. Kosnarite, KZr2(PO4)3. Mimetite, Pb5(AsO4)3Cl. Monazite, (La,Ce,Nd,Pr,Ca,Th,U)(P,Si)O4. Pyromorphite, Pb5(PO4)3Cl. Rhabdophane, (La,Ce,Nd,Pr)PO4nH2O. Vanadinite, Pb5(VO4)3Cl. Xenotime, (Y,Dy,Er,Yb)PO4. Carbonates, Fluorocarbonates, Fluorides Bastnaesite, (La,Ce,Nd,Pr)CO3F. Calcite, CaCO3. Fluorite, CaF2.

600

Minerals and Natural Analogues

Sulfides Galena, PbS. Pyrite, FeS2. Uranyl Minerals (Section 5.22.6) Autunite, Ca[(UO2)(PO4)]2(H2O)10–12. Becquerelite, Ca[(UO2)6O4(OH)6](H2O)8. Boltwoodite, (K,Na)(UO2)(SiO3OH)(H2O)1.5. Carnotite, K2[(UO2)(VO4)]2(H2O)3. Curie´nite, Pb[(UO2)(VO4)]2(H2O)5. Curite, Pb1.5þx(UO2)4O4þ2x(OH)32x(H2O). Fourmarierite, Pb(UO2)4O3(OH)4(H2O)4. Francevillite, Ba[(UO2)(VO4)]2(H2O)5.

Franc¸oisite, (Ce,Nd,Sm,Y)(UO2)3(PO4)2O(OH) (H2O)6. Ianthinite, U4þ(UO2)O4(OH)6(H2O)9. Meta-autunite, Ca[(UO2)(PO4)]2(H2O)6. Rutherfordine, UO2CO3. Sale´eite, Mg[(UO2)(PO4)]2(H2O)10. Schoepite, (UO2)8O2(OH)12(H2O)12. Sklodowskite, Mg(UO2)2(SiO3OH)2(H2O)6. Soddyite, (UO2)2(SiO4)(H2O)2. Studtite, UO44H2O. Torbernite, Cu[(UO2)(PO4)]2(H2O)8. Uranophane, Ca(UO2)2(SiO3OH)2(H2O)5. Vandendriesscheite, Pb1.57(UO2)10O6(OH)11(H2O)11. Weeksite, (K,Na) (UO2)2Si5O13 (H2O)4.