Municipal Solid Waste Incinerator Residues (Studies in Environmental Science)

MUNICIPAL SOLIDWASTE INCINERATORRESIDUES This Page Intentionally Left Blank Studies in Environmental Science 67 MU...

Author: H.A. van der Sloot | S.E. Sawell | J. Vehlow | D.S. Kosson

69 downloads 1831 Views 46MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

MUNICIPAL SOLIDWASTE INCINERATORRESIDUES

This Page Intentionally Left Blank

Studies in Environmental Science 67

MUNICIPAL SOLID WASTE INCINERATOR RESIDUES The INTERNATIONAL c o m p r i s e d of

ASH WORKING GROUP,

(in alphabetical order):

A. John Chandler

David S. Kosson

T. Taylor Eighmy

Steven E. Sawell

A.J. Chandlerand Associates Ltd., Willowdale, Ontario, Canada University of New Hampshire, Ourham, New Hampshire, U.S.A.

Jan Hartl6n

Rutgers, The State University of New Jersey, New Brunswick, New Jersey U.S.A. Compass Environmental, Burlington, Ontario, Canada

Hans A. van der Sleet

Swedisch GeotechnicalInstitute, LinkEping, Sweden

Netherlands EnergyResearchFoundation, Petten, The Netherlands

Ole Hjelmar

JiJrgen Vehlow

VKI WaterQuality Institute, Hersholm, Denmark

1997 ELSEVIER Amsterdam

-

Lausanne

Forschungszentrum Karlsruhe GmbH, Institute of TechnicalChemistry, Karlsruhe, Germany

-

NewYork-

Oxford -

Shannon

-

Tokyo

ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands

ISBN 0-444-82563-0 © 1997 ELSEVIER SCIENCE B.V. All rights reserved. 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, Elsevier Science B.V., Copyright & Permissions Department, P.O. Box 521, 1000 AM Amsterdam, The Netherlands. Special regulations for readers in the U.S.A. - This publication has been registered with the Copyright Clearance Center Inc. (CCC), 222 Rosewood Drive Danvers, Ma 01923. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the U.S.A. All other copyright questions, including photocopying outside of the U.S.A., should be referred to the publisher. 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. This book is printed on acid-free paper. Printed in The Netherlands

PREFACE

The International Ash Working Group (IAWG) was established in 1989 to conduct an in-depth review of the existing scientific data and develop a state-of-knowledge treatise on MSW incinerator residue characterisation, disposal, treatment and utilisation. The topics of operator and worker health and safety, and health risk assessment were beyond the scope of this project, and therefore have not been addressed. Members of the IAWG had been involved in various research and development programs concerning MSW incineration residues for several years prior to establishing the IAWG. The IAWG has met regularly since its inception to discuss aspects of residue characterisation and management, as well as offering a forum for other researchers to provide their perspectives on the issues. The project soon grew beyond the original scope, due in part to the need to examine the ever increasing volume of published research data which became available in the early 1990's. In addition, the IAWG project was designated as an Activity under the International Energy Agency's (lEA) Bioenergy Agreement Task Xl - Conversion of MSW to Energy 1991 - 1994. This final treatise and the Summary Report represent the culmination of the IAWG efforts over the period from February 1990 through July 1996. The input of information from colleagues, along with other information available from the literature and personal contacts, was used to formulate the conclusions and recommendations summarised in this document. The results of this effort have been presented in extended seminars, in conjunction with both the WASCON '94 Conference (June 1994) in Europe and with the Municipal Waste Combustion Conference (April 1995) in North America. In addition, the IAWG co-sponsored and participated in the "Seminar on Cycle and Stabilisation Technologies of MSW Incineration Residues" along with the Japan Waste Research Foundation (March 1996) in Japan. Currently, the IAWG continues to operate as a sub-group of Thermal Conversion Activity under the IEA's Bioenergy Agreement Task XlV - Energy Recovery from Municipal Solid Waste.

This Page Intentionally Left Blank

vii

AUTHORS A. John Chandler A. J. Chandler and Associates, Ltd. Willowdale, Ontario Canada T. Taylor Eighrny University of New Hampshire Durham, New Hampshire United States of America Jan Hartldn Swedish Geotechnical Institute Linkoping Sweden Ole Hjelmar Danish Water Quality Institute H~rsholm Denmark David S. Kosson Rutgers, The State University of New Jersey New Brunswick, New Jersey United States of America Steven E. Sawell Compass Environmental Burlington, Ontario Canada Hans van der Sloot Netherlands Energy Research Foundation Petten The Netherlands Jtirgen Vehlow Forschungszentrum Karlsruhe GmbH Institute of Technical Chemistry Germany

... VIII

THE INTERNATIONAL ASH WORKING GROUP A. John Chandler A. J. Chandler and Associates, Ltd. Willowdale, Ontario Canada

Shin-ichi Saki (Since 1994) Environment Preservation Centre Kyoto University Japan

T. Taylor Eighmy University of New Hampshire Durham, New Hampshire United States of America

Steven E. Sawell Compass Environmental Burlington, Ontario Canada

Jan Hartl6n Swedish Geotechnical Institute Linkoping Sweden

Hans van der Sloot Netherlands Energy Research Foundation Petten The Netherlands

Ole Hjelmar Danish Water Quality Institute H~rsholm Denmark David S. Kosson Rutgers, The State University of New Jersey New Brunswick, New Jersey United States of America

JQrgen Vehlow Forschungszentrum Karlsruhe GmbH Institute of Technical Chemistry Germany

DISCLAIMER This report was prepared by the International Ash Working Group (IAWG). The work was sponsored by the agencies listed herein, who are not necessarily in agreement with the opinions expressed by the IAWG. Neither the sponsoring agencies (including its members), nor the IAWG, nor any other person acting on their behalf makes any warranty, express or implied, or assumes any legal responsibility for the accuracy of any information or for the completeness or usefulness of any apparatus, product or process disclosed, or accept liability for the use, or damages resulting from the use, thereof. Neither do they represent that their use would not infringe upon privately owned rights. The IAWG also does not, and never intended to, discuss or make recommendations with regard to health and safety issues concerning facility operators or workers. Furthermore, the sponsoring agencies and the IAWG hereby disclaim ANY AND ALL WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING THE WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, WHETHER ARISING BY LAW, CUSTOM, OR CONDUCT WITH RESPECT TO ANY OF THE INFORMATION CONTAINED IN THIS REPORT. In no event shall the sponsoring agencies or the IAWG be liable for incidental or consequential damages because of the use of any information contained in this report. Any reference in this report to any specific commercial product, process or service by tradename, trademark, manufacturer or otherwise does not necessarily constitute or imply its endorsement or recommendation by the IAWG and the sponsoring agencies or any of its members.

SPONSORING AGENCIES The IAWG is grateful for the financial and technical contributions made to this project by the following agencies/organisations/companies:

Major Sponsors Asea Brown Boveri (Switzerland) Danish Ministry of Energy Energy, Mines and Resources Canada Environment Canada European Commission Forschungszentrum Karlsruhe (Germany) International Energy Agency International Lead Zinc Research Organization Integrated Waste Services Association (USA) Japan Waste Research Foundation LAB (France) Management Office for Energy and the Environment (Netherlands) National Institute of Public Health and Environmental Protection (Netherlands) Swedish National Board for Industrial & Technical Development Takuma Co., Ltd. (Japan) United Kingdom Department of Environment United States Environmental Protection Agency Wheelabrator Environmental Systems (USA)

Minor Sponsors American Society of Mechanical Engineers Greater Vancouver Regional District (Canada) Kubota Corporation (Japan) Northeast Waste Management Officials Association (USA) New Jersey Department of Environmental Protection (USA) Waste Processing Association (Netherlands (WAV))

TECHNICAL CONTRIBUTORS The IAWG gratefully acknowledges the technical contributions made during the course of this project by: T. Aalbers - RIVM, Netherlands M. Adams - VROM, Netherlands I. H. Anthonissen - RIVM, Netherlands J. A t w a t e r - University of British Columbia,Canada J. Barniske - Umweltbundesamt, Germany J. B e r r y - Wheelabrator Environmental Systems Ltd., USA S. B i n n e r - V~lund, Denmark R. B o e h m - PBI, Netherlands H. Borrmann - Forschungszentrum Karlsruhe, Germany J. P. B o r n - VVAV, Netherlands R. Braam - PBI, Netherlands S. Burnley - Energy Technology Support Unit, United Kingdom D. C h a m b a z - BUWAL, Switzerland A. Chamberland - Tiru Inc. (formerly with Montenay Inc.), Canada W. Chesner- Chesner Engineering, P.C., USA B. Christensen - Environment Canada S. C o o k - Bermuda Biological Station R. C o m a n s - ECN, Netherlands S. Dalager- dk TEKNIK, Denmark A. Damborg - Danish Water Quality Institute C. Dent - AEA Technology, United Kingdom A. M. F~llman - Swedish Geotechnical Institute A. Finkelstein - Environment Canada J. Fraser - Wastewater Technology Centre, Canada M. G a l l o - Rutgers University, USA D. Goetz - University of Hamburg, Germany J. G r o n o w - United Kingdom Department of Environment T. Guest - Montenay Inc., Canada L. Gullbrand - Swedish National Board for Industrial and Technical Development G. Hansen - United States Environmental Protection Agency D. Hay - Environment Canada S. Hetherington - Compass Environmental Inc., Canada F. Hoffman - Rutgers University, USA G. Hoffmann - Umweltbundesamt, Germany R. H u i t r i c - LA County Dept. of Sanitation, USA L. Johansson - Swedish Geotechnical Institute B. J o h n k e - Umweltbundesamt, Germany T. Kimura - Kubota Corporation, Japan J. Kiser- Integrated Waste Services Association, USA R. Klicius - Environment Canada O. Knizik - Greater Vancouver Regional District, Canada M. K n o c h e - LAB, France K. Knox - Knox Associates, United Kingdom

T. Kosson - Rutgers University, USA H. K r u i j d e n b e r g - NOVEM, Netherlands P. Leenders- (formerly with VEABRIN - Netherlands) G. L u e r s - Corning Glass Ltd., USA T. Lundgren - Terratema AB, Sweden D. Mitchell - AEA Technology (formerly with Warren Spring Laboratory), United Kingdom K. Oberg- Swedish Environmental Protection Agency G. Owen - Environment Canada J. Pappain - Peel Resource Recovery Inc., Canada J. Pearson - AEA Technology, United Kingdom A. Petsonk- Swedish Environmental Protection Agency B. Putnam - International Lead Zinc Research Organization G. Rigo - Rigo & Rigo Associates, Inc., USA J. Robert- Energy, Mines & Resources, Canada F. Roethel - University of New York at Stoney Brook, USA H. Roffman - AWD Technologies, USA S. Sakai - Kyoto University, Japan M. Sheil - New Jersey Dept. of Environmental Protection & Energy, USA B. Simmons - California Board of Health, USA D. St~mpfli - formerly with EAWAG, Switzerland J. Stegemann - Wastewater Technology Centre, Canada L. Stieglitz - Forschungszentrum Karlsruhe, Germany M. Stringer - Greater Vancouver Regional District, Canada H. T e j i m a - Takuma Co., Ltd., Japan T. Theis - Clarkson University, USA B. T i m m - Swedish Environmental Protection Agency J. Tsuji - formerly with Environmental Toxicology International Inc., USA A. van Santen - Energy Technology Support Unit, United Kingdom J. F. Vicard - LAB, France J. Vogel - Heidelberger Zement, Germany H. V o g g - Forschungszentrum Karlsruhe, Germany S. Waring - AEA Technology, United Kingdom C. Wiles - National Renewable Energy Laboratory, USA M. Winka - New Jersey Dept. of Environmental Protection & Energy, USA J. W i t t w e r - Environment Canada D. Wexell -Corning, Inc., USA W. Wormgoor - TNO, Netherlands The IAWG also wishes to thank all of the other people not mentioned here, who in their own way assisted us in this endeavour. A special "thank you" to S t e p h e n H e t h e r i n g t o n for his patient efforts in revising and reformatting this document.

This Page Intentionally Left Blank

xiii

TABLE OF CONTENTS PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

V

CHAPTER 1 - INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 A BRIEF H I S T O R I C A L E X C U R S U S ........................... 1.2 THE D E V E L O P M E N T OF W A S T E I N C I N E R A T I O N ................ 1.3 O B J E C T I V E OF THIS TREATISE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES ...............................................

1 1 2 12 13

CHAPTER 2 - MUNICIPAL SOLID WASTE 2.0 INTRODUCTION ......................................... 2.1 DEFINITION OF M U N I C I P A L SOLID W A S T E . . . . . . . . . . . . . . . . . . . . 2.2 C O M P O S I T I O N OF M U N I C I P A L SOLID W A S T E . . . . . . . . . . . . . . . . . . 2.3 QUANTITY AND MANAGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.5 Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.6 The Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.7 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.8 Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.9 United Kingdom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.10 United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 CHEMICAL CONSTITUENTS ................................ REFERENCES ...............................................

15 15 17 21 22 25 26 28 31 32 35 37 37 39 41 51

CHAPTER 3 - MUNICIPAL SOLID WASTE INCINERATION TECHNOLOGIES . . . . . . . . . 3.1 FUEL RECEIPT A N D H A N D L I N G 59 3.2 AVAILABLE COMBUSTION ALTERNATIVES .................... 3.2.1 Mass Burning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . European Type Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . Grates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Furnace Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating Philosophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modular Incineration Systems . . . . . . . . . . . . . . . . . . . . . . . . Other Mass Burn Variants . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Refuse Derived Fuel Systems . . . . . . . . . . . . . . . . . . . . . . . . Semi-Suspension Burning Systems . . . . . . . . . . . . . . . . . . . . Stoker Fired Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

59 61 62 62 65 70 73 76 77 79 82 85

xiv Fluidised Bed Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HEAT RECOVERY SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IN-PLANT RESIDUE M A N A G E M E N T . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.2 Grate Siftings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.3 Heat Transfer System Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.3 3.4

85 87 89 90 92 94 95

CHAPTER 4 - AIR EMISSION CONTROL STRATEGIES . . . . . . . . . . . . . . . . . . . . . . 4.0 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 COMBUSTION CONTROL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Compensation for Fuel Variability . . . . . . . . . . . . . . . . . . . . . . Factors Controlling the Chemical Reaction Rate . . . . . . . . . . . 4.2 POST-COMBUSTION C O N T R O L . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 Unit Processes For Air Pollution Control . . . . . . . . . . . . . . . . Particulate Matter Control Systems . . . . . . . . . . . . . . . . . . . . Electrostatic Precipitators . . . . . . . . . . . . . . . . . . . . . . . . . . . Fabric Filter (Baghouses) . . . . . . . . . . . . . . . . . . . . . . . . . . . Gaseous Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wet Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dry Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metals Control in Dry Systems . . . . . . . . . . . . . . . . . . . . . . . Mercury Control with Activated Carbon . . . . . . . . . . . . . . . . . NOx Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 TYPICAL APC INSTALLATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Hogdalen, Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Munich South, Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Warren County, New Jersey, USA . . . . . . . . . . . . . . . . . . . . 4.3.4 Zirndorf, Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.5 Vestforbr~nding, Copenhagen . . . . . . . . . . . . . . . . . . . . . . . 4.3.6 Lausanne, Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.7 Bremerhaven, Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.8 Stuttgart, Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97 97 97 97 98 98 101 103 103 104 106 109 109 111 112 113 115 118 118 118 122 122 125 125 128 128 131

CHAPTER 5 - REGULATION OF MSW INCINERATORS . . . . . . . . . . . . . . . . . . . . . 5.1 EXISTING MSW INCINERATOR OPERATING GUIDELINES . . . . . . . . 5.1.1 Furnace Temperature and Residence Time . . . . . . . . . . . . . . 5.1.2 Combustion Efficiency and Carbon Monoxide . . . . . . . . . . . . 5.1.3 APC Temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4 Other Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 AIR EMISSION STANDARDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Chronological Changes in Emission Standards . . . . . . . . . . . 5.2.2 Emissions of Combustion Products and Acid Gases . . . . . . .

135 137 137 139 139 140 140 141 144

XV

H y d r o g e n Chloride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Particulate Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S u l p h u r Dioxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . O x i d e s of Nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carbon M o n o x i d e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total H y d r o c a r b o n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H y d r o g e n Fluoride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trace Metals Emission S t a n d a r d s . . . . . . . . . . . . . . . . . . . . 5.2.3 5.2.4 Trace O r g a n i c Emission S t a n d a r d s . . . . . . . . . . . . . . . . . . . . 5.3 CURRENT ASH AND RESIDUE DISPOSAL PRACTICES .......... 5.3.1 Disposal of Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . Canada ....................................... Denmark ...................................... France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G e r m a n y and Switzerland . . . . . . . . . . . . . . . . . . . . . . . . . . Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sweden ....................................... United K i n g d o m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 Disposal of Fly Ash and A P C Residues . . . . . . . . . . . . . . . . Canada ....................................... D e n m a r k & the Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Germany ...................................... Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sweden ....................................... Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Utilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Canada ....................................... Denmark ...................................... France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Germany ...................................... Sweden ....................................... Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

144 144 144 145 145 145 145 145

CHAPTER 6 - ISSUES RELATED TO INCINERATOR ASH SAMPLING . . . . . . . . . . . . 6.0 INTRODUCTION ........................................ 6.1 THE C O N C E P T OF THE REPRESENTATIVE SAMPLE ........... 6.1.1 Waste T y p e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 T y p e of Incinerator/APC S y s t e m . . . . . . . . . . . . . . . . . . . . . 6.1.3 Residue S t r e a m s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 O B J E C T I V E S OF M A T E R I A L S A M P L I N G P R O G R A M S ...........

167 167 167 168 169 170 171

147 149 149 149 149 150 150 150 152 152 152 153 153 153 153 154 154 155 155 155 155 156 156 156 157 157 158 161 161 161

xvi 6.3 6.4

AVAILABLE SAMPLING PROTOCOLS . . . . . . . . . . . . . . . . . . . . . . . . SAMPLING CONSIDERATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 Increment Collection Classification . . . . . . . . . . . . . . . . . . . . 6.4.2 Bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.3 Precision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Number of Increments in Composite Sample . . . . . . . . . . . . . 6.4.4 Size of Increments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.5 Collection Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.6 Sampling Streams Other Than Bottom Ash . . . . . . . . . . . . . . Grate Siftings and Heat Recovery Ash . . . . . . . . . . . . . . . . . APC Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Storage Piles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sampling from Trucks or Containers . . . . . . . . . . . . . . . . . . . 6.4.7 Sample Preparation Concerns . . . . . . . . . . . . . . . . . . . . . . . Sample Size Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preservation of Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample Containers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laboratory Sample Preparation . . . . . . . . . . . . . . . . . . . . . . Laboratory Sample Subdivision . . . . . . . . . . . . . . . . . . . . . . Drying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Size Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Balance of Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SAMPLE COLLECTION R E C O M M E N D A T I O N S . . . . . . . . . . . . . . . . . 6.5 6.5.1 Generic Bottom Ash Testing Protocol . . . . . . . . . . . . . . . . . . 6.5.2 Generic Boiler Ash Sampling Protocol . . . . . . . . . . . . . . . . . 6.5.3 Generic APC Residue Sampling Protocol . . . . . . . . . . . . . . . 6.5.4 Documentation of Sampling and Preparation Procedures . . . . E X A M P L E S OF SAMPLING STRATEGIES . . . . . . . . . . . . . . . . . . . . . . 6.6 6.6.1 Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Regulatory Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.2 Grate Siftings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.3 Boiler/Economiser Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . Regulatory Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.4 Air Pollution Control System Residues . . . . . . . . . . . . . . . . . Regulatory Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

174 175 175 177 178 178 179 180 181 181 182 182 183 183 183 185 185 186 186 186 187 187 188 188 188 191 192 193 194 194 194 196 197 198 198 199 199 199 200 200

CHAPTER 7 - CHARACTERISATION METHODOLOGIES . . . . . . . . . . . . . . . . . . . . . 7.1 PHYSICAL TESTING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.1 Visual Observation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

203 203 203

xvii

7.2

7.3

Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fly Ash and APC Residue . . . . . . . . . . . . . . . . . . . . . . . . . . Particle Size Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.2 Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dry Sieve Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fine Particle Analyses Methods . . . . . . . . . . . . . . . . . . . . . . 7.1.3 Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bulk Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Specific Gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laboratory Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.4 Absorption Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Test Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.5 Water Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.6 Proctor Compaction Test . . . . . . . . . . . . . . . . . . . . . . . . . . . Standard Proctor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modified Proctor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.7 Strength and Strength Development . . . . . . . . . . . . . . . . . . . 7.1.8 Bearing Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.9 Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Soundness Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LA Abrasion Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Freeze-Thaw Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.10 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHEMICAL COMPOSITION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.1 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Size Reduction Techniques . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.2 Inorganic Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Digestion Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Specialty Methods for Specific Elements . . . . . . . . . . . . . . . . 7.2.3 Analytical Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . Destructive Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-Destructive Analytical Methods . . . . . . . . . . . . . . . . . . . 7.2.4 Loss on Ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.5 Total Carbon, Carbonate, Sulphur and A m m o n i a . . . . . . . . . . 7.2.6 Acid Neutralisation Capacity . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.7 Organic Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample Preservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHEMICAL SPECIATION M E T H O D S . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.1 Separatory Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . .

203 203 205 205 205 206 206 206 207 207 207 208 208 208 208 209 209 210 210 211 212 212 213 213 214 214 214 221 221 221 223 223 225 226 226 229 232 234 235 236 236 237 237 238 238

xviii Sample Drying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Particle Size Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetic Separation Techniques . . . . . . . . . . . . . . . . . . . . . Density Separation Techniques . . . . . . . . . . . . . . . . . . . . . . Selective Phase Dissolution Methods . . . . . . . . . . . . . . . . . . 7.3.2 Impregnation, Thin-Sections, and Thin-Foil Methods . . . . . . . 7.3.3 Analytical Methods for Solid Phase Chemical Speciation . . . . Transmitted Light Microscopy . . . . . . . . . . . . . . . . . . . . . . . Scanning Electron Microscopy ...................... Petrography (Morphology) . . . . . . . . . . . . . . . . . . . . . . . . . . Scanning Tunnelling Microscopy . . . . . . . . . . . . . . . . . . . . . X-Ray Powder Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . Petrography (Mineralogy) . . . . . . . . . . . . . . . . . . . . . . . . . . . Scanning Electron Microscopy/X-Ray Microprobe Analysis . . Scanning-Transmission Electron Microscopy/X-Ray Microprobe Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . Auger Electron Spectroscopy ....................... X-Ray Fluorescence Spectroscopy . . . . . . . . . . . . . . . . . . . X-Ray Photoelectron Spectroscopy . . . . . . . . . . . . . . . . . . . Secondary Ion Mass Spectroscopy . . . . . . . . . . . . . . . . . . . Electron Energy Loss Spectroscopy . . . . . . . . . . . . . . . . . . . X-Ray Adsorption Spectroscopy and Extended X-Ray Adsorption Fine Structure . . . . . . . . . . . . . . . . . . . . . . . Nuclear Magnetic Resonance . . . . . . . . . . . . . . . . . . . . . . . . Infrared Spectroscopy and Raman Spectroscopy . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

................. MECHANISMS CONTROLLING THE FATE OF ELEMENTS . . . . . . . . 8.1.1 Fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.2 Processes in the Combustion Chamber . . . . . . . . . . . . . . . . Physical Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chemical Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sintering and Related Processes . . . . . . . . . . . . . . . . . . . . . Physicochemical Transformations . . . . . . . . . . . . . . . . . . . . . 8.1.3 Mechanisms in the Boiler . . . . . . . . . . . . . . . . . . . . . . . . . . . Condensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.4 Mechanisms in the Dust Removal System . . . . . . . . . . . . . . 8.1.5 Mechanisms in the Air Pollution Control System . . . . . . . . . . MASS STREAMS IN A MUNICIPAL SOLID WASTE INCINERATOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LITHOPHILIC ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3.1 Fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CHAPTER 8 - FATE OF ELEMENTS DURING INCINERATION

8.1

8.2 8.3

240 240 241 242 242 243 245 247 247 247 249 249 250 250 250 251 251 251 252 252 253 253 254 254

263 264 264 265 265 266 272 277 280 280 281 284 285 285 287 287

xix 8.3.2 8.3.3 8.3.4

8.4

Alkali M e t a l s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Earth-Alkali M e t a l s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heavy Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chromium ...................................... Nickel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper ........................................ VOLATILE ELEMENTS ................................... 8.4.1 Halogens ...................................... Chlorine ...................................... Fluorine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B r o m i n e and Iodine ..............................

288 290 291 292 294 295 295 296 296 297 298 300

Sulphur ....................................... Nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4.2 Volatile M e t a l s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mercury ....................................... Cadmium ...................................... Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arsenic ........................................ Antimony ...................................... O t h e r Volatile E l e m e n t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5 CARBON AND SELECTED CARBON COMPOUNDS ............. 8.5.1 Total C a r b o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.2 P o l y c h l o r i n a t e d D i b e n z o - p - D i o x i n s and -Furans . . . . . . . . . . . 8.5.3 Polychlorinated Biphenyls .......................... 8.5.4 Polychlorinated Benzenes .......................... 8.5.5 Polychlorinated Phenols ........................... 8.5.6 Brominated Hydrocarbons .......................... 8.5.7 Polycyclic A r o m a t i c H y d r o c a r b o n s . . . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

300 302 303 304 305 307 308 310 311 312 312 312 314 320 322 323 324 324 326

CHAPTER 9 - BOTTOM ASH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1 PHYSICAL CHARACTERISTICS OF BOTTOM ASH .............. 9.1.1 Gross C o m p o s i t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reject Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Visual Classification .............................. Water Content .................................. Ferrous C o n t e n t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Loss on Ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D i s s o l v a b l e Solids C o n t e n t . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1.2 Gravimetric Characteristics ......................... Specific Gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Absorption .....................................

339 342 342 342 343 345 346 347 351 351 353 353

XX

Unit Weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percent Fines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1.4 Durability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Soundness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abrasion Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1.5 Geotechnical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proctor Compaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field Compaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . California Bearing Ratio (CBR) . . . . . . . . . . . . . . . . . . . . . . . Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1.6 Influence of Combustor Type and Operation on Physical 9.1.7 Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1.8 Influence of Aging on Bottom Ash Physical Characteristics .. PARTICLE MORPHOLOGY, MINERALOGY, AND ALKALINITY OF BOTTOM ASH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.1 Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.2 Mineralogy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.3 Alkalinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.4 Influence of Combustor Type and Operation on Bottom Ash Surface Area, Mineralogy and Alkalinity . . . . . . . . . . . . . . 9.2.5 Influence of Aging on Bottom Ash Surface Area, Mineralogy and Alkalinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . INORGANIC CHARACTERISTICS OF BOTTOM ASH . . . . . . . . . . . . . 9.3.1 Elements Present in Bottom Ash . . . . . . . . . . . . . . . . . . . . . 9.3.2 Major Matrix Elements (> 10,000 mg/kg): O,Si,Fe,Ca,AI,Na,K,C . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.3 Minor Matrix Elements (1,000 to 10,000 mg/kg): Mg, Ti, CI, Mn, Ba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.4 Other Minor Elements (1,000 to 10,000 mg/kg): Zn, Cu, Pb, Cr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.5 Other Trace Elements Including Oxyanionic Elements (<1,000 mg/kg): Sb, V, Mo, As, Se . . . . . . . . . . . . . . . . . 9.3.6 Other Trace Elements (<1,000 mg/kg): Sr, Ni, Co, Cd, Ag, Hg . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.7 Other Trace Elements Continued (<1,000 mg/kg): B, Br, F, I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.8 Elements Related to Biogeochemical Cycles: C, S, N, P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.9 Exotic Elements, Lanthanides, Actinides . . . . . . . . . . . . . . . 9.3.10 Isotopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.11 Role of Particle Size in Element Distribution . . . . . . . . . . . . . 9.3.12 Influence of Combustor Type and Operation on Bottom Ash Inorganic Characteristics . . . . . . . . . . . . . . . . . . . . . 9.3.13 Influence of Aging on Bottom Ash Inorganic Characteristics .. 9.1.3

9.2

9.3

356 357 359 360 360 361 362 362 364 364 366 366 367 368 368 370 372 374 374 376 377 379 383 385 388 391 396 397 400 400 400 401 405

xxi

O R G A N I C C H A R A C T E R I S T I C S OF BOTTOM ASH . . . . . . . . . . . . . . 9.4.1 Organics Present in Bottom Ash . . . . . . . . . . . . . . . . . . . . . . 9.4.2 Dioxins and Furans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4.3 Chlorinated Benzenes and Chlorinated Phenols . . . . . . . . . . 9.4.4 Polyaromatic Hydrocarbons and Polychlorinated Biphenyls C H A R A C T E R I S T I C S OF GRATE SIFTINGS . . . . . . . . . . . . . . . . . . . . 9.5 C H A R A C T E R I S T I C S OF C O M B I N E D ASH AND S C R U B B E R 9.6 RESIDUE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

9.4

..

406 406 406 408 408 408 409 413

CHAPTER 10 - CHARACTERISTICS OF HEAT RECOVERY SYSTEM ASH . . . . . . . . . 10.1 ASH DEPOSITION M E C H A N I S M S . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 PHYSICAL C H A R A C T E R I S T I C S . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2.1 Particle Size Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2.2 Particle Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3 C H E M I C A L C H A R A C T E R I S T I C S . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.1 pH and Acid Neutralisation Capacity . . . . . . . . . . . . . . . . . . . 10.3.2 Solubility in Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.3 Chemical Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.4 Heavy Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.5 "Volatile" Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.6 Organic Contaminants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PCDD/PCDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Chlorinated Organics . . . . . . . . . . . . . . . . . . . . . . . . . Polycyclic Aromatic Hydrocarbons (PAH) . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

419 419 422 422 424 425 426 426 428 431 433 435 435 437 437 438

CHAPTER 11 - CHARACTERISATION OF AIR POLLUTION CONTROL RESIDUES . . . . 11.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fly Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dry System Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Semi-dry System Residues . . . . . . . . . . . . . . . . . . . . . . . . . Wet Scrubber Sludge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 MAJOR FACTORS INFLUENCING THE C H A R A C T E R I S T I C S OF APC RESIDUES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3 PHYSICAL C H A R A C T E R I S T I C S OF APC RESIDUES . . . . . . . . . . . . . 11.3.1 General Appearance and Behaviour . . . . . . . . . . . . . . . . . . . 11.3.2 Particle Size Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3.3 Geotechnical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.4 PARTICLE M O R P H O L O G Y AND M I N E R A L O G Y . . . . . . . . . . . . . . . . 11.5 WATER SOLUBILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.6 LOSS ON IGNITION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

441 441 441 441 442 442 442 443 444 444 444 447 450 452 453

xxii 11.7

CHEMICAL CHARACTERISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.7.1 pH and Acid Neutralisation Capacity . . . . . . . . . . . . . . . . . . . 11.7.2 Chemical Composition: Inorganic Constituents . . . . . . . . . . . Ranges of Elemental Composition of APC System Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Major Elements (>10,000 mg/kg): O, CI, Ca, Si, Mg, Fe, AI, K, Na, Zn, S, Pb . . . . . . . . . . . Trace Elements (< 1,000 mg/kg): Hg, Cd, Sb, Cr, Sr, Ni, As, V, Ag, Co, Mo, Se . . . . . . . . . 11.7.3 Role of Particle Size in Element Distribution . . . . . . . . . . . . . 11.7.4 Chemical Composition: Organic Constituents . . . . . . . . . . . . Organics Present in APC System Residues . . . . . . . . . . . . . Chlorinated Benzenes and Phenols . . . . . . . . . . . . . . . . . . . Polychlorinated Biphenyls . . . . . . . . . . . . . . . . . . . . . . . . . . Polycyclic Aromatic Hydrocarbons . . . . . . . . . . . . . . . . . . . . Dioxins and Furans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phthalates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.8 COMPOSITION OF WASTEWATER FROM WET SCRUBBER APC SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CHAPTER 1 2 - PHYSICAL ASPECTS OF LEACHING . . . . . . . . . . . . . . . . . . . . . . . 12.1 AN INTRODUCTION TO LEACHING . . . . . . . . . . . . . . . . . . . . . . . . . 12.1.1 Physical Aspects of Leaching . . . . . . . . . . . . . . . . . . . . . . . . The Leaching System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Particle Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fluid Flow Past Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . The Local Equilibrium Assumption . . . . . . . . . . . . . . . . . . . . 12.1.2 Chemical Aspects of Leaching . . . . . . . . . . . . . . . . . . . . . . . Equilibrium Versus Kinetic Systems . . . . . . . . . . . . . . . . . . . Influence of pH on Dissolution . . . . . . . . . . . . . . . . . . . . . . . Influence of Complexation on Dissolution . . . . . . . . . . . . . . . Influence of Oxidation-Reduction Potential on Dissolution . .. Influence of Sorption on Leaching . . . . . . . . . . . . . . . . . . . . 12.1.3 Leaching Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.1.4 Leaching Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.1.5 Unified Approach to Leaching . . . . . . . . . . . . . . . . . . . . . . . 12.2 THE SOLID PHASE/LEACHANT/SOLUTE LEACHING SYSTEM . . . . . 12.2.1 The Leaching System Concept . . . . . . . . . . . . . . . . . . . . . . . 12.2.2 A Multiphase Heterogeneous System . . . . . . . . . . . . . . . . . . 12.2.3 Leaching Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.3 RESIDUE PARTICLES AS A SOLID PHASE . . . . . . . . . . . . . . . . . . . 12.3.1 Particle Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.3.2 Particle Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

454 457 460 461 464 470 473 473 473 474 476 476 476 476 476 478

483 483 485 485 486 486 487 487 488 488 488 489 489 489 490 491 491 491 491 493 495 495 496

xxiii 12.4

FLUID FLOW, DIFFUSION AND MASS TRANSFER . . . . . . . . . . . . . . 12.4.1 Fluid Flow Through Residues . . . . . . . . . . . . . . . . . . . . . . . . 12.4.2 Fluid Flow Past Particles in Suspension . . . . . . . . . . . . . . . . 12.4.3 Diffusional Processes and Internal Mass Transfer Considerations in Residues . . . . . . . . . . . . . . . . . . . . . . 12.4.4 External Mass Transfer Considerations in Residues . . . . . . . 12.5 THE LOCAL EQUILIBRIUM ASSUMPTION . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

...................... LEACHING CHEMISTRY FUNDAMENTALS . . . . . . . . . . . . . . . . . . . . 13.1.1 Thermodynamic Equilibrium Models Versus Kinetic Models .. 13.1.2 Ionic Strength, Ion Activity, Activity Coefficients . . . . . . . . . . . 13.1.3 Cation/Anion Balances . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.1.4 A Note on General Equilibrium Constants . . . . . . . . . . . . . . . SOLUTION COMPLEXATION AND SPECIATION . . . . . . . . . . . . . . . . 13.2.1 Solution Complexation Equilibria . . . . . . . . . . . . . . . . . . . . . 13.2.2 An Example of Lead Complexation in a Hypothetical Leaching Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.2.3 Leaching Solution Speciation . . . . . . . . . . . . . . . . . . . . . . . . DISSOLUTION/PRECIPITATION REACTIONS . . . . . . . . . . . . . . . . . . 13.3.1 Heterogeneous Dissolution/Precipitation Equilibria . . . . . . . . . 13.3.2 Oversaturation/Undersaturation and the Ion Activity Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.3.3 Metastability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.3.4 An Example of Lead Dissolution/Precipitation as PbSO4(s): A Very Simple Leaching System . . . . . . . . . . . . . . . . . . . 13.3.5 An Example of Lead Dissolution/Precipitation as Pb(OH)2(s): The Role of Solution Phase Complexation and Amphoterism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.3.6 An Example of Lead Dissolution/Precipitation as PbCO3(s): The Role of CO2(g) in Controlling Metal Carbonate Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.3.7 Solubility Control and the Coexistence of Multiple Solid Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.3.8 An Example of Lead Dissolution/Precipitation When Both Pb(OH)2(s ) and PbCO3(s ) are Present . . . . . . . . . . . . . . . 13.3.9 Solid Phase Stability in a Redox-Variable System . . . . . . . . . CHEMICAL WEATHERING AND AGING . . . . . . . . . . . . . . . . . . . . . . 13.4.1 The Mineral Surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.4.2 Weathering Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.4.3 Surface Reaction-Controlled Dissolution . . . . . . . . . . . . . . . . 13.4.4 Weathering Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.4.5 Aging Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CHAPTER 13 - CHEMICAL ASPECTS OF LEACHING

13.1

13.2

13.3

13.4

498 498 500 501 503 504 505

507 507 507 511 517 520 522 524 525 530 531 532 534 536 538

540 543 544 546 548 550 551 551 553 555 556

xxiv

13.5

SORPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.5.1 Surface Functional Groups . . . . . . . . . . . . . . . . . . . . . . . . . 13.5.2 Activity-Based Sorption Models . . . . . . . . . . . . . . . . . . . . . . 13.5.3 Electrostatic Surface Complexation Models . . . . . . . . . . . . . . 13.5.4 Adsorption Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.6 A UNIFIED APPROACH TO LEACHING . . . . . . . . . . . . . . . . . . . . . . . REFERENCES .............................................

558 558 561 564 567 567 570

CHAPTER 1 4 - LEACHING TESTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.1 PURPOSE OF LEACHING TESTS . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.1.1 Classification of Leaching Tests . . . . . . . . . . . . . . . . . . . . . . Extraction Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Agitated Extraction Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-agitated Extraction Tests . . . . . . . . . . . . . . . . . . . . . . . Sequential Chemical Extraction Tests . . . . . . . . . . . . . . . . . . Concentration Buildup Tests . . . . . . . . . . . . . . . . . . . . . . . . Dynamic Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Serial Batch Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Flow Around Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Flow Through Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.1.2 Leaching Test Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Leachant Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Method of Contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Liquid-to-Solid Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contact Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Leachate Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.1.3 Compilation of Leaching Tests . . . . . . . . . . . . . . . . . . . . . . . 14.2 A UNIFIED APPROACH TO LEACHING TESTS . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

579 579 579 580 581 582 583 584 584 585 585 586 588 588 589 591 592 594 594 595 595 599 606

CHAPTER 15 - LEACHING MODELLING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.1 EQUILIBRIUM MODELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.1.1 Thermodynamic Equilibrium Models . . . . . . . . . . . . . . . . . . . 15.1.2 Use of the Geochemical Thermodynamic Equilibrium Model MINTEQA2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.1.3 Verification of MINTEQA2 . . . . . . . . . . . . . . . . . . . . . . . . . . 15.1.4 Recommendations for Utilising MINTEQA2 to Model Leaching Behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.1.5 Modelling US, pH, and Redox Control of Leaching . . . . . . . . 15.1.6 Modelling US, pH, Redox and Complexation Control of Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

607 607 608 609 611 615 617 619

XXV

15.1.7 15.1.8

Modelling Sorption Reactions Influencing Leaching . . . . . . . . Modelling Solid Phase Control of Leaching in Conjunction with Solid Phase Speciation Studies . . . . . . . . . . . . . . . . 15.1.9 Modelling Solid Phase Control of Leaching in Dynamic Flow-Through Systems . . . . . . . . . . . . . . . . . . . . . . . . . . 15.1.10 Modelling Field Leaching Behaviour . . . . . . . . . . . . . . . . . . . 15.2 D Y N A M I C M U L T I C O M P O N E N T M O D E L S . . . . . . . . . . . . . . . . . . . . . 15.2.1 Dynamic Multicomponent Models . . . . . . . . . . . . . . . . . . . . . 15.2.2 Modelling Solid Phase Dissolution . . . . . . . . . . . . . . . . . . . . 15.2.3 Modelling Solid Phase Reprecipitation and Solubility Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.3 FUTURE D I R E C T I O N S IN M O D E L L I N G . . . . . . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

CHAPTER 1 6 - LEACHING DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.1

16.2

16.3

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.1.2 Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.1.3 Data Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TOTAL S O L U B L E F R A C T I O N A N D A V A I L A B I L I T Y OF ELEMENTS ........................................... 16.2.1 Total Soluble Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.2.2 Total Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.2.3 Sequential Chemical Extractions . . . . . . . . . . . . . . . . . . . . . SOLUBILITY AND R E L E A S E OF E L E M E N T S . . . . . . . . . . . . . . . . . . . 16.3.1 Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arsenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cadmium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chromium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molybdenum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nickel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.3.2 APC Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Boron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Barium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cadmium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chromium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mercury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molybdenum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

621 621 626 627 628 628 632 632 633 634

637 637 637 637 638 638 638 639 641 642 646 648 650 650 650 650 653 653 653 653 653 657 657 657 657 657 657 657 657

xxvi 16.4

16.5

16.6

16.7 16.8

GEOCHEMICAL MODELLING OF LEACHING EQUILIBRIA . . . . . . . . . 16.4.1 Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calcium and Sulphate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnesium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Silicon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aluminum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manganese . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sodium, Potassium, Bromide and Chloride . . . . . . . . . . . . . . Cadmium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molybdenum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.4.2 Modelling of APC Residue Leachability . . . . . . . . . . . . . . . . . 16.4.3 Application of Geochemical Modelling Results . . . . . . . . . . . . RELEASE RATES OF ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.5.1 Release As A Function Of Liquid To Solid Ratio . . . . . . . . . . Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . APC Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.5.2 Diffusion Controlled Release . . . . . . . . . . . . . . . . . . . . . . . . RESIDUE LEACHING IN THE CONTEXT OF REGULATORY LEACHING TESTS AND WASTE FROM OTHER SOURCES . . . . . . . . 16.6.1 Regulatory Tests and pH Dependent Leaching . . . . . . . . . . . German DIN 38414 (1984) and French AFNOR X-31-210 (1988) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Japanese Leaching Test . . . . . . . . . . . . . . . . . . . . . . . . . . . Swiss TVA (1988) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . USA, California WET Test . . . . . . . . . . . . . . . . . . . . . . . . . . US EP Toxicity Test (1980), the Toxicity Characteristic Leaching Procedure (TCLP) (1990) and the Regulation 309 (now 347) Leach Procedure . . . . . . . . . . . . . . . . . . . Alternative Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.6.2 Systematic Leaching Behaviour Among Different Incinerator Residues Streams And Other Wastes . . . . . . . LEACHING OF ORGANIC CONSTITUENTS . . . . . . . . . . . . . . . . . . . . EFFECTS OF INCINERATOR OPERATION ON LEACHING . . . . . . . . 16.8.1 Combustion Efficiency (Burn out) and Facility Operation . . . . 16.8.2 Waste Feed Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.8.3 Seasonal Variations In Leaching . . . . . . . . . . . . . . . . . . . . . Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molybdenum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nickel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

658 659 659 659 659 662 662 662 663 663 663 666 666 666 668 668 670 671 673 679 682 684 684 684 686 686 686

686 687 687 700 702 702 702 708 708 711 711 711 711

xxvii 16.8.4 Quench Water Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EFFECTS OF RESIDUE P R O C E S S I N G AND M A N A G E M E N T ON LEACHING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.9.1 Size-Reduction And Size Fractionation ................ 16.9.2 Storage And Aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.9.3 Comparison Of Laboratory Data To Field Measurements REFERENCES ..............................................

711

16.9

...

CHAPTER 17 - SEPARATION PROCESSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

DEFINITION OF PROCESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . P H Y S I C A L SEPARATION T E C H N I Q U E S . . . . . . . . . . . . . . . . . . . . . . 17.2.1 On-site Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fly Ash and APC Residues . . . . . . . . . . . . . . . . . . . . . . . . . 17.2.2 Metal Separation from Bottom Ashes . . . . . . . . . . . . . . . . . . 17.3 P H Y S I C O - C H E M I C A L AND C H E M I C A L S E P A R A T I O N TECHNIQUES .......................................... 17.3.1 Washing Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bottom Ashes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Air Pollution Control Residues . . . . . . . . . . . . . . . . . . . . . . . 17.3.2 Acid Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Filter and Boiler Ashes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.3.3 Ion Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Principles and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.3.4 Hg Recovery from Flue Gas Scrubbing Solutions . . . . . . . . . 17.3.5 C rystallisation/Evapo ration . . . . . . . . . . . . . . . . . . . . . . . . . . Principles and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NaCI Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CaCI 2 Production from Dry/Semidry APC System Residues Gypsum Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.3.6 HCI Recovery by Distillation . . . . . . . . . . . . . . . . . . . . . . . . . Principles and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proposed Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.3.7 Electrochemical Processes . . . . . . . . . . . . . . . . . . . . . . . . . Principles and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chlorine Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

735 735 736 737 737 738 739

17.1 17.2

CHAPTER 1 8 - SOLIDIFICATION & STABILISATION . . . . . . . . . . . . . . . . . . . . . . . 18.1 18.2

DEFINITION OF P R O C E S S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EFFECTS OF SOLIDIFICATION/STABILISATION . . . . . . . . . . . . . . . .

713 713 718 722 728

..

741 741 741 741 743 745 745 746 749 749 751 751 751 751 752 752 755 755 755 755 755 758 759

763 763 764

xxviii

18.2.1 Physical Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.2.2 Chemical Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.3 EVALUATION OF SOLIDIFICATION/STABILISATION PROCESSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.3.1 Physical Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.3.2 Chemical and Leaching Tests . . . . . . . . . . . . . . . . . . . . . . . 18.3.3 Other Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.4 REVIEW OF AVAILABLE PROCESSES . . . . . . . . . . . . . . . . . . . . . . . 18.4.1 Stabilisation Without Additives . . . . . . . . . . . . . . . . . . . . . . . 18.4.2 Solidification/Stabilisation with Binders . . . . . . . . . . . . . . . . . Cement-Based Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Waste Pozzolanic Systems . . . . . . . . . . . . . . . . . . . . . . . . . Chemical Stabilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.4.3 Stabilisation With Organic Additives . . . . . . . . . . . . . . . . . . . 18.4.4 Macro-Encapsulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.4.5 Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

764 766

CHAPTER 19 - THERMAL TREATMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.1 DEFINITIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.1.1 Vitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.1.2 Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.2 GLASS COMPOSITION AND METALS RETENTION IN GLASS MATRICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.2.1 Glass Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.2.2 Constituent Retention Mechanisms . . . . . . . . . . . . . . . . . . . . 19.2.3 Chemical Attack and Leaching Mechanisms . . . . . . . . . . . . . 19.2.4 Factors Influencing Vitrified Ash Leaching . . . . . . . . . . . . . . . Chemical Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Waste Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.3 PROCESSING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.3.1 Processing Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.3.2 Energy Requirements and Costs . . . . . . . . . . . . . . . . . . . . . 19.4 THERMAL TREATMENT PROCESSES UNDER DEVELOPMENT 19.4.1 Overview of Reported Processes . . . . . . . . . . . . . . . . . . . . . 19.5 VITRIFICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.5.1 Vitrification of APC Residues by Corning, Inc . . . . . . . . . . . . . 19.6 FUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.6.1 Fusion of Filter Ash by ABB Deglor Process . . . . . . . . . . . . . 19.6.2 Japanese Fusion Processes . . . . . . . . . . . . . . . . . . . . . . . . 19.7 SINTERING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

791 791 791 792

767 771 773 775 776 776 777 777 780 782 783 785 786 786 788

....

793 793 796 799 801 801 801 801 802 803 803 803 803 803 805 811 817 819 822 830

xxix 19.7.1

Integrated RDF Combustion with Sintered Aggregate Production (Neutralysis Process) . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

CHAPTER 20 - LEACHING OF PRODUCTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.1 I N T R O D U C T I O N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.2 P H Y S I C A L A N D C H E M I C A L F A C T O R S W H I C H E F F E C T CONSTITUENT RELEASE ................................. 20.3 T E S T M E T H O D S FOR M O N O L I T H I C AND C O M P A C T E D GRANULAR PRODUCTS ................................. 20.3.1 A N S 16.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.3.2 NEN 7345 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.3.3 Compacted Granular Leach Test . . . . . . . . . . . . . . . . . . . . . 20.4 I N T E R P R E T A T I O N OF DIFFUSION C O N T R O L L E D R E L E A S E . . . . . . 20.4.1 Characteristic Release Behaviours . . . . . . . . . . . . . . . . . . . . Diffusion Controlled Release . . . . . . . . . . . . . . . . . . . . . . . . Depletion of Leachable Species . . . . . . . . . . . . . . . . . . . . . . Delayed Release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surface Wash-Off . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wash-Out of Mobile Species (Dissolution) . . . . . . . . . . . . . . . Change in Chemical Conditions . . . . . . . . . . . . . . . . . . . . . . 20.4.2 Definition of Leaching Parameters . . . . . . . . . . . . . . . . . . . . 20.4.3 Calculation of Effective Diffusion Coefficients from Cumulative Release Data . . . . . . . . . . . . . . . . . . . . . . . . 20.4.4 Alternative Release Models for Monolithic Materials . . . . . . . . 20.5. R E L E A S E FROM P R O D U C T S C O N T A I N I N G I N C I N E R A T O R RESIDUES ............................................ 20.5.1 Total Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.5.2 Effective Diffusion Coefficients, Physical Retention and Chemical Retention . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cadmium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sodium, Sulphate and Chloride . . . . . . . . . . . . . . . . . . . . . . 20.5.3 Solubility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.6 I N T E G R A T E D I N T E R P R E T A T I O N OF pD e A N D A V A I L A B I L I T Y . . . . . . Cadmium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sodium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chloride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

832 836

841 841 842 843 844 844 845 845 845 845 845 847 847 847 847 848 854 858 860 860 861 861 861 863 863 863 874 878 887 887 887 890 890 890 893

xxx CHAPTER 21 - UTILISATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.1 I N T R O D U C T I O N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2 C U R R E N T AND P L A N N E D P R O J E C T S ...................... 21.2.1 Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.2 Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.3 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.4 The Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.5 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.6 United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 C U R R E N T R E G U L A T O R Y F R A M E W O R K . . . . . . . . . . . . . . . . . . . . . 21.3.1 Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3.2 Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3.3 The Netherlands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3.4 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3.5 United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.4 T E C H N I C A L R E Q U I R E M E N T S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.5 UTILISATION LIFE C Y C L E A N D E N V I R O N M E N T A L CONSIDERATIONS ...................................... 21.5.1 Ash Type Selection and Elements of Concern . . . . . . . . . . . . 21.5.2 Ash Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.5.3 Physical Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.5.4 Stockpiling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.5.5 Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.5.6 Use Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.5.7 Reuse and Disposal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.5.8 Economic Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES ..............................................

895 895 896 896 896 897 897 898 899 901 902 905 905 911 911 914

CHAPTER 22 - DISPOSAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.1 I N T R O D U C T I O N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2 C H A R A C T E R I S T I C S OF I N C I N E R A T O R R E S I D U E LANDFILL LEACHATES ........................................... 22.2.1 Overview of Incinerator Residue Leachability . . . . . . . . . . . . 22.2.2 Bottom Ash Leachate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2.3 Fly Ash and Acid Gas Scrubbing Residue Leachate . . . . . . . 22.2.4 Combined Ash Leachate . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.3 D I S P O S A L S T R A T E G I E S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.3.1 General Philosophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lifetime of Active Systems . . . . . . . . . . . . . . . . . . . . . . . . . . Waste Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adaptation of Landfill Design, Operation and Siting to Strategy and Waste Types . . . . . . . . . . . . . . . . . . . . . Ultimate Fate of the Leachate . . . . . . . . . . . . . . . . . . . . . . .

931 931

916 916 917 918 918 919 919 922 926 926 926

932 932 934 937 940 940 941 942 943 945 946

xxxi 22.3.2

General Disposal Strategies . . . . . . . . . . . . . . . . . . . . . . . . . Total Containment or Entombment . . . . . . . . . . . . . . . . . . . . Containment and Collection of Leachate . . . . . . . . . . . . . . . . Controlled Contaminant Release . . . . . . . . . . . . . . . . . . . . . Unrestricted Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.4 DESIGN AND OPERATIONS ISSUES . . . . . . . . . . . . . . . . . . . . . . . . 22.4.1 Siting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.4.2 Liners and Leachate Collection Systems . . . . . . . . . . . . . . . . 22.4.3 Caps and Top Covers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.4.4 Geotechnical Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.4.5 Abatement of Noise, Odour and Fugitive Dust Problems . . . . 22.4.6 Monitoring of Leachate Quantity and Quality . . . . . . . . . . . . . 22.4.7 Monitoring of Groundwater and Surface Water Quality . . . . . . 22.4.8 Leachate Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.5 DISPOSAL PRACTICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.6 DISPOSAL R E C O M M E N D A T I O N S FOR INCINERATOR RESIDUES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.6.1 Bottom Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.6.2 APC Residues (Fly Ash and Acid Gas Scrubbing Residues) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.6.3 Combined Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

947 948 948 950 950 950 951 951 953 953 954 954 955 957 957 959 960 961 962 962

This Page Intentionally Left Blank

CHAPTER 1 -INTRODUCTION 1.1 A BRIEF HISTORICAL EXCURSUS Since the dawn of human existence, people have produced what could be described as waste. But the issue of managing these wastes did not appear until human culture evolved past the stage of nomadic hunter-gatherer into a settled agricultural-based existence. During these times, waste management was not a real cause for concern. The relatively small populations lived in vast areas, and there was no shortage of places to discard food-scraps and excrement. Food-scraps even provided a source of nourishment for livestock or were used as fertiliser to enrich the cultivated soil. Broken implements, pottery or other materials which were no longer useful were simply discarded on scrap piles in outlying areas without further thought. Today, archaeologists are able to gather detailed information on the state of human cultural development and prosperity by excavating ancient dump sites. As human culture advanced, increasing population densities created a very real need to manage both the solid waste and excrement being generated. Evidence exists that the first culturally complex cities in Mesopotamia and Indus established means of disposing of excrement through underground sewage pipes and practised separate collection of solid waste. Even in the Bible, we find reference to a waste disposal site. In the time of King Solomon's successors Asa, Hisia and Josia, the waste from Jerusalem was brought to the Kidron Valley, where it was incinerated in open fires (Anonymous, First Book of Kings). The ashes from these fires were then brought to Bethel, or were scattered over the graves of the nearby Jerusalem cemetery (Anonymous, Second Book of Kings). Even with all that had been achieved in some of these civilisations, waste management took a turn for the worst in the cities of medieval Europe. Excrement, food waste and other materials were merely cast into the streets to be dealt with by the rain, wind, sun and any unfortunate passerby. The situation became so serious that solutions had to be found. Examples of these, albeit simple solutions, began to appear during the latter half of the 13th century. There is evidence that streets constructed in Hamburg, Germany, were designed with inclines leading to a central gutter to allow the rains to wash away wastes more easily. During the 15th century, increasing population densities forced municipalities to introduce waste collection systems. In some areas, prisoners of war, slaves and criminals were often used to carry out these onerous tasks, whereas other municipalities hired people to do the job. These people were either paid directly by the authorities or special taxes were collected. In 1473, there was a new twist in waste management. An enterprising commercial hauler actually paid the City of Amsterdam for the right to collect putrescible wastes from the City and then sold the material to farmers in the surrounding countryside as fertiliser (Erhard, 1964). Although this form of waste management soon caught on in

bigger cities, the eventual glut of material made it difficult to market the putrescibles to farmers, and municipalities were finally forced to pay for hauling of the waste. The commercial viability continued to decline until the 17th century, when individual towns began to takeover waste management operations themselves. With the industrial revolution in the 19th century, came the generation of new types of waste materials in previously unheard of quantities. Although many of these new wastes were not biodegradable, they were considered a problem due to their noxious nature. In North America, the marketing of putrescible waste to farmers was a common and relatively successful practice, however to reduce the quantities of other waste materials requiring disposal, other potentially valuable materials were separated using a three level collection system. This involved installing three-bin collection systems in homes. One bin was used to collect only organic kitchen waste, such as food scraps, which was then used for food for livestock. In some instances, milkmen performed a double duty by collecting this waste as well as their normal delivery duties. In major cities like New York, Boston and Chicago, fat was collected and recycled in special facilities, some of which remained in operation into the early 1900's. Another bin was used to collect potentially marketable materials such as textiles, footwear, glass, metals and wood. The remaining bin was used to store the ashes generated from fireplaces and stoves. This form of waste collection system was also in evidence in some parts of Europe in the latter part of the 19th century (de Fodor, 1911). However, most European cities managed their waste by transporting it out of the city and dumping the material in designated areas. Animals foraged in the piles for food and the less fortunate were allowed to pick through the waste for whatever was salvageable (see Figure 1.1 ). Despite the efforts of municipalities, waste management could not keep pace with the rapidly growing population and the resulting burgeoning quantities of waste. By the early 1900's, annual per capita waste generation rates were growing. For example, it was estimated that New York City's annual rate was 540 Kg/person, London's exceeded 300 Kg/person, whereas Budapest, Munich and Zurich all exceeded 230 Kg of waste/person/year (de Fodor, 1911 ).

1.2 THE DEVELOPMENT OF WASTE INCINERATION Although the relationship between hygiene and human health was first recognised in England during the 19th century, it took the high profile efforts of physicians like Max von Pettenkofer, Louis Pasteur and Robert Koch to emphasise the fact that epidemics of disease, such as cholera and typhoid, were the result of bacteria spread by unsanitary conditions, and not acts of God. Based on this need to manage waste in a sanitary fashion and the growing need to quench industries' thirst for fuel, the first waste incinerators were developed in England. Needless to say, the first attempts at the process were not very successful. In 1870, attempts were made to burn waste in a retrofitted coal-burning furnace in Paddington.

Figure 1.1 Women Sorting the Waste of Vienna at the end of the 19th Century

de Fodor, 1911 The incinerator was designed so that coal was burned on a grate to provide the major source of heat to initiate combustion of the waste on a separate grate located above the coal grate. Unfortunately, the wet nature of the waste and poor heat transfer in the system resulted in what could best be described as a smouldering effect. Only the persevering demands of physicians to "disinfect" waste kept this, and other small units operating against heavy public opposition. But even in these early attempts, it was noted that high temperatures in the furnace not only reduced the odour of flue gases, but they also generated an ash which was suitable for reuse as a building material. The first fully functional municipal solid waste incinerator was constructed in Manchester, England in 1876. The unit design included an induced draft fan which helped to maintain higher burning temperatures on the hearth. The facility operated for 27 years and the generated ash was used as a building material. The next major development was based on the need to further reduce the odours of the flue gases. In 1885, the incinerator in Egling was equipped with what was dubbed "a cremator." The "cremator" consisted of a coal-fired grate in which the flue gases from the waste furnace were passed before being released into the atmosphere. Soon after, the energy release from waste combustion came under the study of scientists like Lord Kelvin. It was found that 1 Kg of waste could generate 1.5 Kg of steam, which subsequently led to the design and construction of the world's first combined waste incinerator/electrical power generation facility, in the municipality of Shoreditch in London, in 1897.

Early waste incinerators were operated in a batch-wise mode. The units were usually fed by hand, as illustrated in Figure 1.2. Removing the slag and clinker was also done by hand, although proved to be a difficult task at times due to the extensive slagging on the grates. An example of this is illustrated in Figure 1.3. The Figure also illustrates the design of the facility. Note that the individual units were located in close proximity to one another, to facilitate better heat transfer and to help start-up new fuel beds after a unit was cleaned out. Design efforts soon began to focus on mechanical feeding and cleaning of the incinerator grates. Sequential feeding and deslagging provided a more homogenous temperature profile and hence improved combustion (Figure 1.4). Even at this early stage, the basic principles of waste incineration were comparable with those of today. For example, it was not merely satisfactory to disinfect the waste, but it had to be done in a manner which prevented the emissions of malodorous flue gases. In addition, it was preferable to recover the energy released during the process and make use of the ash which remained. In 1910, 194 English towns made use of waste incinerators. The proliferation of the practice also led to further innovations, such as the practice of using the waste feed as a gas tight seal for the furnace, which is still used today (Figure 1.5). One of the most successful incinerator designs of the day was the Horsefall cell-type incinerator, developed by the Horsefall Destructor Company of Leeds. This was further modified by Heenan & Froude Ltd. of Worcester. These systems were generally operated as the final stage in a separation and reclamation system which was equipped with drum screens, magnetic separators and a hand sorting line. Figure 1.6 provides a cross-section schematic of this incinerator design. The refuse was charged onto the grate through an overhead hopper above the furnace chamber. The grates were stoked manually as the bottom ash was raked into a clinker channel below the grate. The only air pollution control device was a simple baffle system which kept the coarse particles and flaming paper from exiting the stack. Although these systems were operated without the benefit of a fan, surprisingly, some units were still in operation into the late 1960's (Tanner, 1967). The development of waste incineration outside of the UK was slow to start, and was based mostly on the English technology. In 1892, the City of Hamburg decided to construct a municipal solid waste incinerator to cope with a growing waste management problem spurred by the reluctance of farmers to accept more waste, and more important, a major cholera epidemic. The facility went into operation in 1895 and had the capacity to burn a total of 16 tonnes of waste/day in two back-to-back units. Shortly thereafter, incinerators began to spring up all over Europe, especially in Germany, and major cities such as Brussels, Stockholm and Zurich (de Fodor, 1911). The latter facility was constructed at the Josephstrasse in 1904, and was a Horsefall design with 12 furnace cells and two boilers which generated 150 KW of electricity. It was replaced with a Heenan & Fourde design in 1928 which remained in regular use until 1969.

Figure 1.2 Manual Feeding of a Unit at the Glasgow Incinerator Facility

Adapted from de Fodor, 1911

Figure 1.6 Sectional Diagram of One Single Horsefall Cell (a) and Longitudinal Section of Horsefall Cell Furnace (b)

Adapted from Tanner, 1967

The early batch mode incinerators were based on a flat grate design, which resulted in relatively poor burnout of both the waste and the flue gases. Performance was greatly improved prior to the First World War by the development of mechanised inclined reciprocating grates which promoted steady-state combustion at elevated temperatures, which were facilitated by much better primary air distribution. An early design, still in use in modern incinerators, is the Martin reverse-acting grate shown in Figure 1.7, which was taken from the German patent 458540 of 1926. Its first application in waste combustion dates back to 1931. Another design which is still in use today is the system developed by V~lund of Copenhagen. It consists of a reciprocating primary grate, followed by a rotary kiln. The original design of the incinerator is shown in Figure 1.8, and albeit with some modifications over the years, was in operation until 1993. Figure 1.7 Original Patent Drawing of the Martin Reverse-Acting Grate

~~

A~.x.

e A&h.~.

Abb. j .

u

Martin, 1926

.~,

w"

(,, i~

I?'AA-p

b'///,.

Figure 1.8 Design of the Vslund Incinerator at Gentofte, Denmark (commissioned in 1932)

Adapted from Tanner, 1967

10 During the early 1900's, the practice of waste incineration was generally confined to urban areas. Alternatively, rural regions mainly practised composting and direct landfilling as a means of waste management. Especially during the two World Wars and the Great Depression, sorting and recycling were practised out of a need to conserve resources. Since the little waste that was generated was reused, the heating value of waste during these times was extremely low, and consequently there was little impetus to build new incinerator facilities. However, the economic boom of the postSecond World War era brought about many changes in lifestyles, including the stockpiling of prepared foodstuffs and an unprecedented increase in consumerism. Because these products were manufactured, shipped and stored in mass quantities, packaging was required to maintain the integrity of the products, and this packaging evolved slowly over the years. Glass and tin cans were soon replaced by paper and plastics. This, coupled with the fact that waste avoidance and recycling were not a major concern, eventually resulted in a huge increase in the heating value of municipal solid waste. Consequently, many new incinerators were constructed in the 1960's and 70's to deal with the increasing volumes of waste. Most of these new incinerators were based on the designs of the 1920's. In 1957, Yon Roll built the Borsigstrasse incinerator in Hamburg, and this design (Figure 1.9) provides a typical example of the type of incinerators built during the post-war era. Although the combustion efficiency and reliability of the operation were improved over the 1920's designs, there was little effort placed on reducing emissions to the atmosphere or improving the quality of the ash. The poor flue gas quality, especially the relationship to dioxin production, inevitably led to the vociferous public opposition against MSW incineration in the late 1970's and early 1980's. As a consequence, the installation of new incinerator facilities took a dramatic decline, exemplified by the temporary moratorium placed on the operation of facilities in Sweden in 1986. The incinerator industry was quick to react to the problems. As noted in Chapters 3 and 4, the designs of incinerators were changed to further enhance the operating conditions and reduce emissions of contaminants to the atmosphere. The most notable progress has been in the air pollution control technologies. This is illustrated by a comparison of emission values over the years (Table 1.1 ). Table 1.1 Relative Improvements in the Reduction of Incinerator Emissions (Mg/m 3 ex ;ept PCDD/PCDF given in TE ng/r Time APC I Dust HC 1 SO2 Technology 1900 none 5,000 1,000 500 300 1,000 I 0.5 I <1970 cyclones 500 1,000 500 300 1,000 0.5 1970-80 ESP's 100 1,000 500 300 500 0.5 1980-90 ESP + APC 50 100 200 300 100 0.2 > 1990 latest APC <10 i <10 <10 <0.05 <50 <100 N/A = not available Adapted from Vogg, 1988

NO'l CO I "~ i N/A N/A N/A 10

<0.1

Figure 1 . 9 Design of the Von Roll Borsigstrasse Incinerator at Hamburg (1957)

1 2 3 4

= charging shaft

= drying grates = main grate = clinker generator

Adapted from Tanner, 1967

5 = boiler

6 = electric filter

7 = generator chain 8 = clinker chain

12 The improvements in the quality of atmospheric emissions over the past ten years are evident. The various methods available for flue gas cleaning have become well established, and the selection of equipment can be based on the individual needs to meet certain emission limits. The improvement in air emission quality has been so dramatic that the World Health Organisation no longer considers the emissions from modern, well operated and maintained MSW incinerators to be a hazard to human health or the environment (Suess, 1989). This has resulted in a shift of concerns from air emissions to the management of incinerator residues. 1.3 OBJECTIVE OF THIS TREATISE In many countries, the potential impacts of these residues on the environment have come under close scrutiny and more stringent regulation. However, many of the concerns regarding the disposal of MSW incinerator residues have become controversial in nature, mostly due to inconsistencies based on: the different quantities and qualities of incinerator residue streams due to the type of incinerator and APC systems the heterogenous nature of the residue streams the lack of standardised sampling and analytical protocols the difficult process of comparing data from different studies the different conclusions drawn from different evaluation protocols the variations in regulations for residue management between different countries Furthermore, the situation has become confused due to the broad scope of the technical information required to understand the characteristics of the residues, the fragmented data base and the uncertainty over the long-term behaviour of the residues in the environment. Recognising there was a need to compile and evaluate the available information, the International Ash Working Group (IAWG) was established to conduct an in-depth review of the existing scientific data and develop a state-of-knowledge treatise on MSW incinerator residue characterisation, disposal, treatment and utilisation. The project was designed to:

1)

Define uniform protocols for the sampling and full characterisation of MSW incinerator residues, including chemical, physical and leaching properties;

2)

Describe the fate and behaviour of contaminants during the incineration process, including documenting the effects of different incinerator designs, air pollution control systems, incinerator operations and refuse feedstocks;

13

3)

Evaluate and develop recommendations with regard to the current or proposed disposal, treatment, utilisation and recovery practices;

4)

Identify areas for further research and development.

The following chapters have been prepared to provide a detailed review of the existing information related to MSW incinerator residues. It is hoped that this information will place the issue of incinerator residue characteristics into perspective, and ultimately provide a framework for making sound decisions regarding their management in an environmentally acceptable manner. REFERENCES

Anonymous. The Holy Bible, First Book of Kings, 15,13; Second Book of Chronicles, 15,16,30 & 14. Anonymous. The Holy Bible, Second Book of Kings, 23,4. Erhard, H. "Aus der Geschichte der St~dtereinigung", in MelI-Handbuch, G. H(~sel, W. Schenkel and H. Schurer, Erich Schmidt Verlag, Berlin, Vol. 1, Kennziffer 0110, 1964. de Fodor, E. Elektrizit~t aus Kehricht, K.u.K. Hofbuchhandlung von Julius Benk(~, Budapest, 1911. Martin, J. "Verfahren und Rost zum Verbrennen fester Brennstoffe", Patentschrift Nr. 458 540, Deutsches Reich, Reichspatentamt, January 30, 1926. Suess, M.J. "WHO Viewpoint on Health Impact From Municipal Waste Combustion", Proceedinqs of the International Conference on Municipal Waste Combustion, Vol. 1, Hollywood, Florida, April 1989. Tanner, R. "The Development of the von Roll Refuse Incineration System", 1967. Taylor, R. Personal communication with staff of von Roll, 1993. Vogg, H. "Von der Schadstoffquelle zur Schadstoffsenke - neue Konzepte der M011verbrennung", Chemie-ln.qenieur-Tech .nik, 60, 217, 1988.

This Page Intentionally Left Blank

15

CHAPTER 2 - MUNICIPAL SOLID WASTE 2.0 INTRODUCTION

In many developed countries, incineration of municipal solid waste, either with or without energy recovery, is considered one of the key components in an integrated waste management strategy. Incineration is viewed as a means to destroy pathogens and organic matter, and typically results in a 90% reduction in volume and 60% reduction in weight of the waste being combusted. Since the reduction concentrates the mineral and elemental components of the original waste, fundamental knowledge of the characteristics of the MSW stream is essential to understanding the characteristics these materials impart on incinerator residues. Consequently, this chapter provides a summary of the nature of MSW, including the gross physical and chemical composition of MSW in different countries and the strategies employed by those countries to manage their waste. 2.1 DEFINITION OF MUNICIPAL SOLID WASTE

In order to better compare the quantities of MSW generated in different countries, waste production figures are generally transformed into per capita waste generation rates. The generation rates for ten different countries are presented in Figure 2.1. Although these values appear to be directly comparable, the list of waste materials included in the calculated values may vary widely. Clearly, there are a number of factors which contribute to the variations within these national generation rates, including societal customs, socioeconomic conditions and geography, however, the basic question remains "What is MSW?". Unfortunately, the definition of MSW is open to administrative interpretation which varies widely between countries, agencies and even local jurisdictions. Definitions may range from: "Solid waste includes residential, light industrial, commercial and institutional waste that is collected by a municipality or by contracted collectors on behalf of the municipality." (one definition taken from "The State of Canada's Environment", (Government of Canada, 1991 )), to "Any garbage, refuse, sludge from a waste treatment plant, water supply treatment plant, or air pollution control facility and other discarded material, including solid, liquid, semisolid, or contained gaseous material resulting from industrial, commercial mining, and agricultural activities, and from community activities..." US Resource Conservation and Recovery Act (RCRA, 1984).

16 Figure 2.1 Per Capita Waste Generation Statistics

Source: Environment Canada While these definitions may represent the extremes, they illustrate the potential difficulty in directly comparing MSW related data developed by agencies or groups with different goals, responsibilities or motives. In 1989 for example, the US EPA estimated that between 11 - 13% of the MSW stream was being recycled. However, some groups insisted that automobile scrap should be included in the calculation, thereby effectively doubling the nationally quoted recycling rate (OTA, 1989). On the other hand, Japan does not include the quantity of materials which have been recycled in the definition of MSW because these materials are not considered waste. In view of the lack of a standard definition for MSW, the following definition is suggested: "The solid waste generated at residences, commercial establishments (e.g., offices, retail shops, restaurants), and institutions (e.g., hospitals and schools), but does not include construction/demolition debris, automobile scrap or medical/pathological waste," United States Office of Technology Assessment (OTA, 1989). The fact that this definition says more about what is not included in MSW than what is considered MSW, merely emphasises the heterogeneous nature of the waste stream. For the purposes of this document, this general definition will be considered an appropriate compromise, since it excludes materials which are not likely to be processed through MSW incinerators. Nevertheless, it should be noted that the statistics cited in this chapter may have been developed based on different criteria in each of the different countries.

17

2.2 COMPOSITION OF MUNICIPAL SOLID WASTE Municipal solid waste, as generated from residential and/or commercial sources, is heterogeneous with respect to both physical and chemical composition. Although the composition of MSW has changed throughout history, the most dramatic changes have occurred during the last 50 years, characterised by an accelerated proliferation of waste organic matter, paper and plastics (Table 2.1 ). Table 2.1 Changes in Waste Compositio n over Time in the UK (% by weight) 1982

1986

1988

7

8.8

6

7.5

31.1 25.2

29

22.8

33

25

21.3

3 5 . 5 28.3

25

23.7

20

22.8

9

8.5

5.3

7.2

8

9.6

8

13.4

9

8

9.5

9.3

11.8

10

9.6

9

3.5

57

39

22

19.8

1 2 . 3 13.9

14

16.7

10

13.4

2

3

2

3.5

1.7

2

3

2.6

4

7.6

3

6.9

1.8

5.9

4

6.2

10

5.8

Waste Category

1935

1963 1968 1974 1975 1978 1 9 8 0

Plastic

-

-

1

2.9

Paper

14

23

37

26.8

Putrescibles

14

14

18

Metals

4

8

Glass

3

Dust/Cinders Textiles

Other 5 4 Adapted from Baker, 1990

3

5.7

The size of the various components of MSW can range from small dust particles to large bulky items such as packaging materials, tires, furniture and appliances. In addition, non-combustible and inert materials, such as metals, glass and ceramics, form a significant portion of the MSW stream. As a result, this heterogeneity creates numerous problems when characterising MSW. Similar to the problems associated with developing a broad definition of MSW, the various choices available for the categorisation of various components in the waste have created some confusion over the exact definition of terms applied to materials in various countries. For example, garbage and garden waste are typically combined in Swedish data (SAPCSWM, 1988) and the term is assumed to cover kitchen scraps and yard waste. Yet, these materials are normally defined and reported separately in the North American data, while the British have combined these materials under the term putrescibles. The variation in definitions has prompted some calls for uniform definitions of each category (at least in Europe) (Williams, 1984):

18 Putrescibles - kitchen, vegetable and yard waste; Paper and Card - paper, paper/plastic laminates, newsprint, card, corrugated cardboard; Plastics - all plastic materials including film and dense materials; Metals - all metallic materials; Textiles - natural and manmade fibres; Miscellaneous combustibles - wood, shoes, leather; Miscellaneous non-combustibles- ceramics, stone, cinder; Glass; and Fines - < 20 mm size materials. However, even with these suggested categories, identification of various components may become problematic due to the multi-component content of the materials themselves. For example, material such as laminates would be classified depending on the major component in the material, (i.e., paper or plastic) however, it may be difficult to determine which material is the major component. Furthermore, there is potential for confusion over items such as carpeting which could be considered a textile, but is classified under miscellaneous combustibles. Similarly, although bone could be classified as food waste (putrescibles), in some countries it is classified under miscellaneous non-combustibles. In addition to the differences in national classification, there are regional differences. Management practices such as bottle-return legislation, recycling, waste reduction initiatives, and social customs result in differences in classification. At the present time, several agencies including, the International Energy Agency, UN/ECE, OECD and EUROSTAT, are attempting to develop a standardised MSW classification scheme, however, in the interim it is suggested that definitions of waste categories be sufficiently detailed to permit others to extract information for direct comparison. An example of a detailed list of materials and the higher order categories is given in Table 2.2. Once there is an understanding of the various potential categories of waste, it should also be noted that there are inherent differences between residential/household, institutional, commercial and light industrial waste (Table 2.3). Industrial waste contains less organic material than the other source streams but higher levels of glass, wood, construction waste and miscellaneous materials. Commercial waste has the highest level of paper products. The composition of institutional waste is very similar to residential waste with the exception that institutional waste was found to contain little, or no, wood or construction waste. Seasonal variations, including both mean temperature and precipitation, can affect refuse composition. As an example, the effects of seasonal variations on Canadian MSW composition are shown in Table 2.4, with the most notable change being in yard waste. The data indicates that a higher percentage of yard waste is disposed in spring and fall compared to summer and winter, mostly resulting from the maintenance of landscaped residential properties. Weather conditions can also vary the moisture

19

Table 2.2 Comprehensive List of Waste Component Categories 1. Paper & Cardboard

1.1 1.2 1.3 1.4 1.5 1.6

1.7 1.8 1.9 1.10 2.1

2. Glass

2.2 2.3 2.4 2.5 2.6 2.7 2.8 3.1 3.2 3.3 3.4

3. Ferrous Metals

3.5

containers - beer 2.1.1 clear 2.1.2 green 2.1.3 brown 2.1.4 other containers - soft drink (colors) containers -liquor (colors) containers - wine (colors) containers - food (colors) containers - other (colors) plate glass other

8. Organics

8.1 8.2 8.3 8.4 8.5 8.6

food waste yard waste garden waste textiles leather rubber 8.6.1 tires 8.6.2 other

9. Misc. Inorganics

9.1 9.2

ceramics asbestos

10. Household Hazardous Waste

10.1 10.2

automotive batteries other batteries 10.2.1 carbon 10.2.2 ni-cd 10.2.3 mercury 10.2.4 other water based paints other paints solvents waste oils pesticides/herbicides other (acids/antifreeze/etc.)

5. Non-Ferrous Metals

5.1 5.2 5.3 5.4 5.5 5.6

containers - beer containers - soft drink containers - food other packaging die cast other

6. Plastics

6.1

film 6.1.1 LLDPE 6.1.2 LDPE 6.1.3 HDPE 6.1.4 other container - beverage 6.2.1 PET 6.2.2 polyprop., etc container - food (types) container - household product other (toys/furniture/etc.)

6.5

Waste

7.2 7.3 7.4 7.5 7.6

10.3 10.4 10.5 10.6 10.7 10.8

containers - beer containers - soft drink containers - food strap metal and banding other (toys/tools/etc.)

containers - beer containers - soft drink containers - food electric motors other

6.3 6.4

construction debris 7.1.1 gypsum & plaster 7.1.2 wood 7.1.3 concrete/ stones /rubble 7.1.4 fibre glass white goods furniture automobile parts bicycles/motorcycles other

7. Oversized & Bulky

4. Bimetallic Items 4.1 4.2 4.3 4.4 4.5

6.2

7.1

newsprint fine paper computer ledger/office magazine coated (waxed/plastic/foil) box board brown kraft other corrugated cardboard mixed (junk mail/flyers/etc.)

11. Household Biomedical 12. Fines

11.1 11.2 11.3

diapers bandages, etc. sharps

20 Table 2.3 Comparison of Institutional, Commercial and Light Industrial Waste Composition (% by weight) Component

Residential

Institutional

Commercial

Industrial

Paper Metal Plastics Organics

36 7 7 31

41 5 15 29

50 5.5 15 18

40 4.5 13 6

Glass Wood Construction Waste Miscellaneous Adapted from Government of

7.5 6 4 2 0 3 2 0 0.5 7.5 4 4 Ontario, 1992; Gore and Storrie, 1992

9 12 2 13.5

Table 2.4 Seasonal Variations in Waste Composition (% by weight) Category

Component

Spring

Summer

Fall

Winter

Combustibles

Paper, cardboard Food waste Yard Waste Plastic, rubber Textiles

33.9 28.4 12.3 4.7 3.7

38.9 24.6 4.4 5.4 4.3

36.6 24.9 6.1 5.7 4.4

36.5 32.9 0.8 5.1 4.7

Wood Misc. Organics

2.7 0.9

5.0 0.9

5.7 1.7

3.5 1.2

7.6 5.8

9.7 6.8

8.1 6.8

8.1 7.2

Noncombustibles

Glass, ceramics Metals Adapted from Bird and Hale, 1977

21 content of the waste, both on a seasonal and a daily basis. This underlines the problem of calculating the proportionate weight of each category based on an "asdiscarded" or "as-received" basis. The mixing of waste can cause the moisture levels of the components to vary from their "as-discarded" level. For instance, wet food wastes may transfer moisture to paper and textiles during storage and transportation (Table 2.5). Table 2.5 Estimated Percent Moisture in Waste Component % H20 in As-received Paper Metal Plastic Yard Waste Food Waste Textiles Glass Wood Leather & Rubber Miscellaneous Niessen, 1970

24.3 6.6 13.8 37.9 63.6 23.8 3 15.4 13.8 3

% H20 in As-discarded 8 2 2 55 70 10 2 15 2 2

In light of the numerous problems associated with characterising MSW, there are some guidelines which should be followed when reporting MSW data. In general, the more comprehensive the list of materials, the better the definition of MSW. Secondly, seasonal variations should be considered when reporting MSW data. Even comparisons within the same jurisdiction should be made on data collected under the same conditions, i.e., seasonally adjusted figures or data collected at the same time of year, under similar weather conditions.

2.3 QUANTITY AND MANAGEMENT Typically, quantities of MSW produced in various countries are calculated based upon the sum of materials handled through controlled MSW processing facilities, although nationally reported data is seldom accompanied by a description of the procedures used to develop the statistics. Although the quantities are usually reported on the basis of wet weight, in some instances it is difficult to determine whether the influence of moisture has been taken into account. Some current estimates of per capita discards of MSW range from 0.5 kg/day in China to 1.7 kg/day in Canada (Figure 2.1), (Environment Canada, 1989). In view of the numerous waste material control measures being implemented in various countries, it is expected that generation rates

22 will decrease slightly or remain the same during this decade. However, the trends in some countries have been to the contrary. For example, the United States is anticipating a waste/capita generation rate increase of 1 to 2% per year up to the year 2000 (Franklin, 1988), and the Japanese were experiencing an increase of about 4% per annum earlier in the decade (Eller, 1992). Irrespective of any national waste minimisation policies, MSW will continue to be generated and require management employing the options available under the "Waste Management Hierarchy": separation for reuse, separation for recycling, separation for composting, incineration, or, direct landfilling. All of these methods are utilised to some extent, however, in many countries landfilling continues to be the most utilised means of disposal. Each method has limitations, most of which are the result of the nature of the material, including quality, durability and practicality of collection. Other considerations in the selection of alternatives are economic viability, available space, geographical location and terrain, and public opinion. The following is a synopsis of the current MSW situation in several developed countries and the management strategies employed therein. Since the main focus of this publication is MSW incinerator residues, emphasis has been placed on detailing the use of incineration technology. 2.3.1 Canada

The most recent statistics available (1992) indicate that it costs Canadians about $2.2 billion (Cdn) per year to manage the estimated 30 to 35 million tonnes of nonhazardous waste that is generated in Canada annually (Environment Canada, 1996). More than half of the 30 million tonnes is MSW (Figure 2.2), an estimated 1.7 kg of waste per person per day. The results from recent studies suggest that the composition of MSW in Canada is similar to that of the United States (Figure 2.3). In Canada, although the federal government does maintain some regulatory authority over MSW management, regulation is generally the responsibility of municipal and provincial levels of government. As a result, regulations have varied from province to province based on regional differences. In order to unify federal and provincial policies, the Canadian Council of Ministers of the Environment (CCME) was established to deal with issues concerning resources and the environment. Within the CCME framework, committees consisting of representatives from both levels of government develop guidelines and standards for specific environmental issues. The rationale behind this concept was to provide the individual provinces with the ability to develop draft guidelines cooperatively, which then could be readily adopted as provincial legislation.

23 Figure 2.2 MSW Management in Canada - 1992 (million tonnes)

Recycling 15-19% Residential Commercial Institutional Light Industry 21%

Municipal Solid Waste

Incineration 5-6%

16% Landfill 75%

Environment Canada, 1996 Figure 2.3 Canadian MSW Compositional Statistics

Environment Canada, 1995

Energy Recovery 5%

24 In 1990, CCME set a national objective of 50% diversion of waste from landfill by the year 2000 using the hierarchical approach of reduction, reuse, recycling and recovery (CCME, 1990a). This was followed up by a National Packaging Protocol, which set a target of 50% reduction in packaging sent for disposal by the year 2000, using the approach of source reduction and reuse to achieve at least half of the diversion and recycling for the remainder (CCME, 1990b). The aim of the initiatives is to drastically reduce the reliance on landfill, which ultimately accepts about 82% of the currently disposed MSW. Approximately 15 - 19% of the MSW stream (excluding the C&D waste fraction) is recycled at the present time, and although there are some facilities in the planning stages, there are few major composting plants in operation. About 5 - 6% of MSW is incinerated, most of which (more than 92%) involves energy recovery. Most of the operating energy-from-waste (EFW) facilities are situated in the most densely populated areas of Canada, namely, the lower mainland of British Columbia, south-central Ontario and southern Quebec. Two major EFW facilities are operating in the maritime provinces. Several small modular incinerators (without energy recovery) are located on Vancouver Island and in Newfoundland, some of which are in the process of closing. It was estimated that there was a potential capacity to incinerate approximately 11 to 12% of the 16 million tonnes of MSW generated in Canada per annum in 1991 (Finkelstein, 1991), however, only about 50% of the potential capacity was realised. Although a new facility was constructed in Ontario (Peel) in 1992, one large incinerator in Quebec (Montreal) was closed in 1994, reducing the capacity dramatically. A summary of operational and proposed EFW plants is given in Table 2.6. Table 2.6 Summary of Energy-from-Waste Facilities in Canada Incinerator Type

Number of Facilities 5 11 1

Proposed

With Energy Recovery* 3 5 *1

Total Capacity (tonnes/day)* Mass Burn 2 1800 Two-stage 0 1000 Semi-suspension 0 550 Total 17 2 9 3350 * - not including proposed facilities ^ - energy recovery system not used Adapted from Sawell, 1992; Environment Canada, 1994 Under the umbrella of CCME, the Municipal Solid Waste Incineration Sub-Committee released the "Operating and Emission Guidelines for MSW Incinerators" in June of 1989 (CCME, 1989). The performance standards recommended by CCME were supported by a large test data base and information generated directly through Environment Canada's National Incinerator Testing and Evaluation Program (NITEP).

25 Based on the scientific evidence provided by NITEP and the stringent standards recommended by CCME, the federal government presently views modern incineration technology (with energy recovery) as a viable option for reducing the burden on landfills under the "4R's" hierarchy (reduction, reuse, recycling and recovery). This view was further supported by the Province of Ontario in 1995, when it rescinded the highly publicised four year old ban on constructing new MSW incinerators and adopted new strict atmospheric emission limits. 2.3.2 Denmark

The latest statistics available (1993) indicate that approximately 9.6 million tonnes of nonhazardous waste are generated in Denmark each year by its 5 million inhabitants (Danish National Agency of Environmental Protection, 1995). About 23% of the total is considered MSW from residential sources (Table 2.7), representing about 1.2 kg of MSW generated per person per day. It was estimated to be increasing at the rate of 1.5% annually in the early Ninety's (Haukohl, 1991), however, this has been revised and is now expected to decrease. The composition of MSW in Denmark is similar to that of other European countries with the bulk of MSW being food and yard waste (around 40%) (Figure 2.4). Table 2.7 MSW Management. Statistics i.n Denmark (millions of tonnes) Category

Incinerated

Ut!lised/recycled

Landfilled

Total

Residential

0.9

0.1

0.2

1.2

Yard Waste

0.04

0.16

0.2

0.4

Commercial

0.6

0.9

0.8

2.3

Residential Bulky

0.03

0.05

0.22

0.3

1.21 .

1.42

4.2

Total (1985) Total MSW (1990)

....

1.57

.

2.6

Projected %* 25 ........... 50 25 100 * = 5 year projection based on nonhazardouswaste - sewagesludge, demolitionwaste & coal ash Adapted from Hjelmar and Johannesen, 1992

26 Figure 2.4 Danish MSW Compositional Statistics

Danish National Agency for Environmental Protection, 1990 Incineration plays a major role in MSW management in Denmark with more than 48% of the MSW stream (including household, trade, bulky and garden wastes) being combusted at 29 large incinerator facilities (Rijpkema et al., 1992). A total number of 31 facilities were operating in 1993 (Hjelmar, 1996). Energy is recovered at each facility. Landfill and recycling/utilisation play an equal role in the management strategy. Projections are that the dependence on landfill will decrease during the next five years due to increased recycling or utilisation of wastes, whereas the quantities of MSW incinerated will remain stable. It is assumed that the 43,000 tonnes of yard waste sent for composting in 1990 (Vestforbraending, 1990) would be included in the recycle/ utilisation category. The Danish Government instituted an "Action Plan for Waste and Recycling for 1993 1997" which set targets for the various waste management options, including 54% recycling, 25% incineration and 21% landfill by the year 2000. In addition, a target was set for reducing the use of PVC packaging by 82% by the year 2000, which is anticipated to help reduce the reliance on landfill and incineration (Danish Environmental Protection Agency, 1995). 2.3.3 France

Approximately 18 million tonnes of MSW (19.5 million tonnes including bulky residential waste), or 0.88 (0.95) kg/person/day, were generated by the 56 million inhabitants of France in 1990 (Baltzinger, 1991). The composition of MSW in France is illustrated in Figure 2.5 (Rousseaux et al., 1988). Similar to most other countries, the largest component in MSW consists of paper products. Also note that the moisture content data for the different component fractions is also provided in Figure 2.5.

27

Figure 2.5 French MSW Compositional Statistics

Rousseaux et al., 1988 Of the 18 million tonnes, 0.6 million tonnes were collected separately for recycling (mostly glass and paper) and the remainder was sent for treatment and disposal. Incineration plays a very major role in MSW management in France. As illustrated in Figure 2.6, about 42% of the MSW collected for treatment and disposal is incinerated, 10% composted and the remainder landfilled. Of the total amount incinerated, 75% is processed through incinerator facilities with energy recovery capabilities. A summary of French incinerator facilities is given in Table 2.8. Note the number of French facilities quoted by sources ranges from 170 to 315 (Knoche, 1992; Baltzinger, 1991; Rijpkema et al., 1992). Table 2.8 offers an extensive breakdown by facility capacity.

Figure 2.6 MSW Management in France 18 million tonnes Recovery/Recycli ng 34 million 42.3% Incineration tonnes 16 million tonnes Treatment/Disposal

9.2% Composting

i 48% Other !

43.7% Landfilling

Adapted from Beltzinger, 1991

28.6% Incineration with heat recovery 13.7% Incineration no heat recovery

28 Table 2.8 Summary of Incinerator Facilities in France (1990) Capacity (tonnes/day)

Number of Facilities

With Energy Recovery

Total Capacity (tonnes/day)

>720

7

6

10,368

>480- <720

9

9

5,400

>240 - <480

30

23

9,708

<240

152

50

11,952

88

37428

Total 198 Adapted from Baltzinger, 1991

2.3.4 Germany Since the records regarding waste management within the former East German infrastructure were difficult to interpret, most of the information given here pertains to the former West German data and post reunification data. In 1990, the West German population of 62 million generated approximately 35 million tonnes of MSW, whereas approximately 43.3 million tonnes were generated by Germany's 90 million inhabitants in 1993. This includes, residential, bulky residential, commercial, light industrial (similar to household) and yard/garden waste. Therefore, the estimates that the new eastern provinces would generate between 3.6 and 5 million tonnes of MSW (Umweltbundesamt, 1990/91; Reimann, 1991 ) were reasonably accurate. In 1986, the federal government passed the Waste Management Act which set out national policy on waste management. Avoidance of waste was given the highest priority which is to be facilitated by mandating:

1) 2) 3) 4)

reuse of beverage containers through deposit/return laws, an aggressive Packaging Ordinance (1990) which set targets for recovery and reuse/recycling of all packaging materials, a Paper Ordinance (1991) designed to divert paper products from landfill, and a Plastic Ordinance which set a target of 80% reuse/recycling of plastics.

A decade later, the "Closed Cycle Economy Law" was adopted to refine the Waste Act by regulating the use and disposal of materials, and thereby helping to promote an environmentally compatible and sustainable economy. More important, the new law makes the recovery of energy from waste materials equal in priority to materials recycling within the "Hierarchy" (Vehlow, 1996).

29 The collective result of these initiatives is estimated to be a 45% diversion of waste from landfill via reuse/recycling. The largest single component of the MSW stream is kitchen waste (Figure 2.7) which represents a substantial potential for composting, although an emphasis has been placed on producing quality compost (marketable) by maintaining strict control of the source materials. The Waste Act stipulates that the remaining combustible MSW must be treated thermally (preferably with energy recovery) prior to disposal, as evidenced by the Technical Directive for Residual Waste which severely restricts the organic content of waste destined for landfill (see Table 2.9). As a result, incineration will continue to play a major role in German waste management. Figure 2.7 German MSW Compositional Statistics

Vehlow, 1992 Table 2.9 Regulations under the Technical Directive for Residential Waste and the LAGA Parameter

Landfill Class I

Landfill Class II

Loss on Ignition %wt

3

5

1

3

Total Organic Carbon %wt Adapted from Vehlow, 1996

LAGA

1

30 About 11 million tonnes (36%) of MSW were incinerated in 51 incinerator facilities, all of which recover energy (Reimann, 1991; Rijpkema et al., 1992) (Table 2.10). Although recycling and composting strategies must be exhausted prior to incineration, it is estimated that the current incineration capacity will have to increase by more than 80% within the decade to keep pace with the increase in waste and still adhere to the targets set within the Waste Act (Umweltbundesamt, 1990). At the present time, four facilities are under construction and there 19 incinerator facilities in the planning stages, representing a potential capacity of 3.5 million tonnes per year. Some of the shortfall is expected to be taken up by modifying existing facilities to accommodate an increase of half a million tonnes of MSW/year. A summary of existing incinerator facilities in Germany is given in Table 2.11. Table 2.10 Municipal Solid Waste Treatment per Category in Germany Component

Municipal ktonnes/year

Solid Waste %

Recycling

4000

16

Composting

500

2

Landfill

11500

46

Combustion

9000

36

Total 25000 Adapted from Rijpkema, 1992

100

Table 2.11 Summary of MSW Incineration Facilities in Germany Capacity (tonnes/day)

# of # with BA Facilities Utilisation

# with Dry APC Systems

# with Wet APC Systems

Ash Generation (tonnes/day)

>1200

13

10

9

6

4,937

>720- <1200

15

13

3

12

3,100

>480- <720

7

4

3

4

966

<480

16

11

6

10

1,060

Total 51 37 21 32 10,063 Note: total # of facilities & capacities include those capable of incinerating sewage sludge and hospital waste Adapted from Barniske, 1989; UBA, 1991 Vehlow, 1996

31

2.3.5 Japan Approximately 50.2 million tonnes of MSW were generated in Japan in 1992. This translates into a 1.1 kg/person/day generation rate which was 1.1% lower than the previous year (Sakai, 1996), reversing a trend of a 4 -6.5% increase per year between 1988 and 1991 (Eller, 1992). This decrease is due mostly to regulations enforcing reuse and recycling of materials. The latest statistics on the composition of the Japanese waste stream indicate that the largest components of MSW are waste paper (37%) and food waste (32%) (Figure 2.8), whereas metals, glass, plastics, etc., combined make up only about 30% of the MSW stream. Figure 2.8 Japanese MSW Composition Statistics

Eller, 1992 Only about 15 % of the total MSW stream was landfilled in 1992 (Sakai, 1996). Since landfill capacity in the country is becoming more scarce (see Figure 2.9), MSW management in Japan is increasingly reliant on incineration (Patel and Edgcumbe, 1992). Almost 85% of MSW was incinerated in 1992 (Sakai, 1996). The most common practice in Japan is to separate MSW into combustible and non-combustible fractions. The non-combustible fraction is not included in national generation statistics, but is estimated to represent about 30% of the total amount of waste generated. Only a small fraction of waste (about 3%) is composted (Eller, 1992). The bulk of the 48 million tonnes of combustible material (about 70-77%)is incinerated in 1,893 incinerators around the country (Government of Japan, 1990; Tsukamoto, 1991; Patel and Edgcumbe, 1992). The remainder is landfilled or sorted to remove metal and glass.

32 Figure 2.10 Summary of Japanese Electricity Production from MSW Incineration 19851991 .~ "0 Or) t,.,.. 0 (0 t-,

200000

m

capacity of incinerators which produce electricity

I--1

capacity of all incinerators

40

30

150000

r v

0 ~0 (0 c}. tO

0

20

10oooo

(l) t'o r

50000

0

0 85

86

87

88

89

90

Year Ministry of Health and Welfare, 1994 2.3.6 The Netherlands

Approximately 28 million tonnes of what is defined as "Priority Waste" is generated each year by the Netherlands' 14 million inhabitants (Government of the Netherlands, 1988 and 1991 a). About 12 million tonnes (43%) of the total could be classified as MSW and less than 6 million tonnes (<22%) are considered household waste (Folmer, 1991 ). Although the quantities of domestic waste discarded annually have increased at the rate of 5 - 10% (Government of the Netherlands, 1988; Folmer, 1991), it is anticipated that the implementation of new policies will effectively drop the rate to a net 1.4% increase during each of the next 25 years (Government of the Netherlands, 1992). The results from compositional studies indicate that the largest component in household waste is food and yard waste followed by waste paper (Figure 2.11 ). At the present time, it is estimated that about 55% of the "Priority Waste Stream" generated is landfilled, 35% is reused or recycled and about 10% is incinerated. This translates into about 35% of the household waste being incinerated (Rijpkema et al., 1992). Although there is an initiative to compost only vegetable, fruit and garden wastes, there is no data available on the success of the program. Previous studies have shown that compost from unsorted municipal waste is rendered unusable due to the high contents of trace metals (Table 2.13).

33 Figure 2.11 Dutch MSW Compositional Statistics

Government of the Netherlands, 1991 Table 2.13 Average Concentration of..Trace Metals in Compost from Different Sources (IJg/g) Trace Metal

Food and Yard Waste

Municipal SolidWaste

Cadmium

1

2

Chromium

20

120

Copper

32

120

Lead

73

450

Nickel

9

45

Zinc 149 Adapted from Government of the Netherlands, 1991b

600

The current national policy for MSW management was set out in the publication "Memorandum on the Prevention and Recycling of Waste" (Government of the Netherlands, 1988) and is similar to that used in most other countries, namely a hierarchical approach of prevention, reuse, utilisation (recycling), incineration and landfill. The projected target is to decrease the amount of material landfilled by more than 70% by the year 2000, relying mostly on recycling/utilisation and incineration (Figure 2.12).

34

1~

Figure 2.12 Projected Targets for MSW Management in the Netherlands 28,000,000 ton

prevention ~

~

! t

~

r

.

1

=

6

2

-

-

i

7~1

recycling and useful a p p l i c a t i o n

6ot 50-

1986

12000

I

i

407 304

J

incineration

20I 10J

landfill

i

O-

Government of the Netherlands, 1991 With respect to incineration, the Dutch government's new policy is toward promoting the use of energy recovery. Previously, many of the facilities were not capable of recovering energy, whereas now about 6 of the 8 currently operating facilities recover energy (Folmer, 1991). The new trend is also aimed at modernising or replacing some of the older facilities with an increased waste capacity. A summary of Dutch incinerator facilities is given in Table 2.14. Table 2.14 Summary of Incinerator Facilities in the Netherlands Furnace Type

Capacity (tonnes/day)

Facilities

Energy Recovery

Mass Burn

>300 > 100 - <300

4 4

4 2

Tonnes of MSW Processed (1990) 2075000 648,000

2 1

0 1

111,000 75,000

11

7

2909000

<100 Semi-suspension <100 TOTAL Adapted from VEABRIN, 1991

35 2.3.7 Sweden

Approximately 2.7 and 3.2 million tonnes of household waste or MSW were generated in 1990 and 1991, respectively, by Sweden's 8.5 million inhabitants (Nilsson, 1991; RVF, 1994). These statistics do not include the waste materials recovered or reused such as bottles, cans, newsprint or cardboard. The latest statistics indicate that about 63% (375,000 tonnes) of waste newsprint and paper, and 94,000 tonnes of glass were recovered for recycling in 1991 (RVF, 1994). Based on these data, it is anticipated that the amount of MSW sent for treatment or disposal will decrease, especially in light of the 1994 legislation stipulating producer responsibility for collection of packaging waste. The goal of the legislation is to recycle 75% of paper by the year 2000. The largest components of MSW are waste paper, and food and yard waste, which constitute up to 45% and 35% of the weight of total discards, respectively (Figure 2.13).

Adapted from Nilsson, 1991 After separation of recyclable materials, it is estimated that the bulk of MSW discarded is incinerated (55%) or landfilled (40%) (Rijpkema et al., 1992) (Figure 2.14). Only about 5% of the MSW stream is currently sent for separation and composting. This figure is down from 1988, when it was estimated that 10% of MSW was sent for separation and composting (Bergstr6m, 1988; SAPCSWM, 1988). However, the demand for higher quality materials from the separation process has limited the success of this process. The market for composted material is also very limited (about 30% of the total produced is sold) due to the content of trace metals. Most compost produced is only used for landfill cover. In the future, it appears that the Swedish government is going to emphasise more utilisation of wet kitchen waste and other non-

36 combustibles for landfilling to enhance landfill gas production. This may decrease the amount of MSW incinerated, however, the decrease will be made up with alternative heating value fuels (Nilsson, 1991).

Adapted from SAPCSWM, 1988 All 21 of the operating incineration plants in Sweden are capable of recovering energy from the 1.8 million tonnes of household and industrial waste incinerated annually (RVF, 1994). About 98% of the energy recovered is used to generate steam for district heating purposes, representing about 13% of the country's district heating requirement (Nilsson, 1991). Most of the remaining 3% is used for generating electricity. A summary of incinerator facilities is given in Table 2.15. Table 2.15 Summary of Incinerator Facilities in Sweden Furnace Type Capacity Facilities (tonnes/day) Mass Burn

Fluidised Bed

>200 <200 - > 100 <100 <140 - >50

<50 - > 14 TOTAL Adapted from Nilsson, 1991

Total Capacity (tonnes/day)

7 4 7 2

3, 710 540 380 195

3 23

96 4921

37 2.3.8 Switzerland More than 3 million tonnes of MSW were generated by Switzerland's 6.6 million inhabitants (1.4 kg/person/day) in 1991 (WRI, 1992). The largest components of MSW are food/yard waste and waste paper products, comprising 30 and 33% of the total MSW stream, respectively (Tabasaran, 1984). The current Swiss policy on MSW management is outlined in the "Ordinance Relating to Treatment of Waste" and is based on the hierarchical approach taken in most other countries, namely reduce, reuse, recycle and recover. However, the policy includes stipulations that every waste treatment process must "produce materials that either are recyclable or have final storage quality." In addition, the processes used must also be economically viable and must result in a net benefit to the environment, i.e., create less pollution than the disposal of the waste or the production of the product from virgin material (Swiss Environmental Protection Agency, 1988). These stipulations have had an impact on the methods used to manage waste. Any component of MSW stream which is deemed non-useable and reactive, must first be inertised by thermal treatment prior to disposal. Consequently, the Swiss incinerate approximately 80% of the MSW generated annually, which is the highest rate of MSW incineration in the world (WRI, 1992). Latest substantiated data indicate 30 incinerator facilities are in operation. Approximately 50% of the incinerator facilities are capable of recovering heat (Rijpkema et al., 1992). 2.3.9 United Kingdom Approximately 19- 20 million tonnes of household (residential) waste were generated per year in the UK between 1987 and 1994 by a population of 58 million, including 5 million tonnes of yard waste and bulky wastes (Krol and Dent, 1988; UK Dept. Of Environment, 1995). This number increases to 30 - 35 million tonnes if similar waste from the commercial and light industrial sectors are included (Baker, 1991; UK Department of Environment, 1995). The latest data available indicate that putrescible (food and yard) wastes (almost 42%), and paper (28%) comprise the bulk of the household waste stream in the UK (Figure 2.15). The current range of estimates on the methods of MSW disposal are that 85 - 88% of the total quantity is landfilled, 4 - 9% is incinerated, and 3 - 11% is disposed or processed by other means including recycling and production of refuse derived fuel (Krol and Dent, 1988; Hinchcliffe, 1992). Recent developments, including the introduction of the Environmental Protection Act (EPA) 1990 (HMIPC, 1990), the EC Landfill Directive and proposed Landfill Tax, are likely to raise both standards and the costs associated with landfill in the UK. The EPA also obliges the local authorities to consider recycling options for MSW. In addition, the 1990 White Paper on the Environment (HMSO, 1990) set a target for recycling 25% of all household waste by the

38 end of the century. This level is being achieved by some local authority initiatives, but the general level of recycling remains at about 3%. Figure 2.15 UK MSW Composition Statistics

Baker, 1990 Although it seems likely that landfill remains the dominant disposal route for the nonrecycled waste stream, the recent introduction of financial incentives for energy-fromwaste projects may lead to the building of new incinerators over the next few years. Facilities in London, Birmingham and Cleveland have been or are in the process of being constructed. Many facilities will be phased out once the EC emissions standards come into force in December of 1996. Consequently, eight further facilities are in the planning stages and some of these will make up for the lost capacity. A summary of existing MSW Incinerator Facilities in the UK is provided in Table 2.16. Table 2.16 Summary of MSW Incinerator Facilities in the UK Capacity Number of With Energy (tonnes/day) Facilities Recovery >720 5 2 >480- <720 9 >240 - <480 10 <240 12 Total 36 Adapted from Baker, 1990

2 2 1 7

Total Capacity (tonnes/day) 4,320 4,512 3,168 1,956 13956

39 2.3.10 United States

Between 300 and 320 million tonnes of residential, commercial and light industrial waste were generated per year in the United States in 1990 - 1993, half of which (163 million tonnes) was considered MSW (US EPA, 1991; Steuteville, 1994). The quantity of MSW generated (1.8 kg/person/day) is anticipated to grow at a rate of 2% per year (Franklin, 1992). The major component of the MSW stream is waste paper which accounts for about 72 million tonnes (40% by weight) of the MSW discarded (Figure 2.16). If food and yard wastes were classified together, they would comprise about 30% of the total MSW stream. Figure 2.16 US MSW Composition Statistics (1990)

US EPA, 1992 The federal policy regarding MSW management in the US is based on the "4R's", reduce, reuse, recycle and recover. The latest statistics indicate that between 62 - 67% of MSW was disposed in landfill, 10- 16% was incinerated and 21 - 23% was recycled. The quantity of MSW sent for composting was negligible. Although there have been national targets set, including 20% incineration, 25% recycling and 55% landfill (US EPA, 1991), regulation and management of MSW falls directly under the jurisdiction of state and regional governments. Consequently, the management strategies vary widely depending on the jurisdiction. Nationally, it is projected that the quantities of MSW destined for landfill (net discards) will decrease during the decade as existing landfills close and greater percentages of the waste are diverted using recycling, composting and incineration processes (Figure 2.17). As of September 1991, 189 MSW incinerator facilities were operating in the

40 United States, representing a total design capacity of about 92,000 tonnes per day. Many of the facilities are located in the densely populated northeastern states. A summary of incinerator facilities is given in Table 2.17. Figure 2.17 US Trends in MSW Management 240 200160c O pco

:s

1208O 4O I

1

1960 _

l

[--|

1

|

1965

J Net Discards ~

I

I

1

I

I

1970

I'"1

'1"1

1

I

1

I"1

I

1

1

1

i

1

1975 1980 1985 Year

Incinerated ~

I

I

I ' I

I

1

I

'1

I

I

1

I

1 9 9 0 1 9 9 5 2000

Composting ~

Recycled

US EPA, 1991 Table 2.17 Summary of MSW Incinerator Facilities in the United States Incinerator Type Mass Burn

Capacity (tonnes/day) >500 > 100 - <500 <100

Two-stage

>100

16

16

RDF (Semi-suspension)

<100 >500 <500

34 18 9

33 18 8

176

135

Total Adapted from Waste Age, 1992

Number of Facilities 44 25 30

With Energy Recovery 39 21 None

Total Capacity (tonnes/day) 51275 6760 950 3145 1957 25320 2200 91607

41 2.4 CHEMICAL CONSTITUENTS The various components in the waste stream all have a unique chemical composition. Most of the materials have varying quantities of carbon, hydrogen and nitrogen in their basic composition. During the incineration process organic-based materials such as paper, kitchen waste and plastics are generally oxidised to H20, CO2, CO and minor constituents. Conversely, the elements, inorganic compounds or mineral phases in the waste feed either:

1)

remain as solid particles and are trapped in the various residue streams;

2)

are volatilised and carried in the flue gas stream until sorbed or condensed out onto particles; or

3)

are discharged with the flue gases.

Since most of the potential environmental effects of waste disposal are related to the chemicals in the waste, developing data on the chemical composition of the waste stream is important. In relation to this document, the trace chemical composition of the waste is an important issue which needs to be examined, since it ultimately influences the chemical composition of incinerator residues and hence their eventual impact on the environment. For the most part, analyses of MSW have generally been performed for the purposes of ascertaining the feasibility of energy from waste treatment alternatives and centres on determination of the burning characteristics of the waste. Bulk waste samples are the most frequently analysed for this purpose after care has been taken to make the samples representative of the mass of waste being tested. Examples of these tests are:

1)

Proximate Analysis - to determine the percentages of volatile matter, fixed carbon, moisture and ash in the waste;

2)

Ultimate Analysis - to determine the percentage of moisture, C, H, S, N, 02 and ash in the waste; and

3)

Heating Value - to determine the amount of energy available from combustion of the waste.

Tables 2.18 and 2.19 provide examples of these analyses of MSW from different Canadian cities and Figure 2.18 shows a phase diagram of the proximate analysis results from a German study on 400 samples of MSW. MSW contains both high heating value components (plastic, rubber, paper and wood), and low heating value components (food and yard wastes). Notice that at a 50% confidence interval, the German waste consisted of between 29 - 43% moisture, 21 to 35% ash and 29 - 38% combustibles, which is consistent with most of the Canadian data.

42 Table 2.18 Examples of Ultimate, Proximate, and Heating Value Analysis Results for MSW from Three Canadian Studies Analysis

Parameter

Ultimate

Carbon Hydrogen Nitrogen Oxygen Sulphur Chlorine Moisture Ash

Proximate

Volatile Matter Fixed Carbon Moisture Ash

(%)

(%)

Hamilton, ON

Charlottetown, PEI

Quebec City, PQ

38.5 2.9 0.5 23.0 0.1 NA 25.0 30.0

25.6 3.1 0.3 21.2 0.1 0.4 35.3 14.0

25.6 3.6 0.4 13.9 0.2 0.8 31.2 24.2

38.5 6.5 25.0 30.0

42.4 8.2 35.3 14.1

41.0 3.6 31.2 24.2

Heating (kJ/kg) 11,165 10,527 Value Adapted from Walls, 1982; Environment Canada, 1985 & 1988 Table 2.19 Summary of Higher Heating Values of Various Components in MSW Waste Component Paper Plastic/paper laminates Food waste Wood (dried) Yard and garden waste

Heating Value (MJ/kg) 12.2 - 18.6 17.1 4.1 - 38.3 15 - 17 4.8 - 18.6

Tires

32.1

Rubber

26.1

Plastics

22.7 - 45.8

Textiles Adapted from Haley, 1990

16.1 - 18.5

9,792

43 Figure 2.18 Tanner Diagram of the Proximate Analysis of MSW w

~0~1

~'%

~_

=~,=~~

r

~ o.-6-

%

-o/ ..# 0

j ~

-.-

--

9 80 % C o n f i d e n c e Interval w' = 20.5 - 48.5 % a' = 17.0 - 46.0 % b' = 25.0 - 44.0 %

....

50 % C o n f i d e n c e Interval w"= 29.0 - 43.0 % a"= 21.0 - 35.0 % b"= 29.0 - 38.0 %

Reimer, 1975

(A=Ash, B=Combustibles, W=Water)

The sources of trace elements in MSW are diverse. For example:

1)

consumer products such as batteries and circuit boards may contain cadmium, lead, mercury and zinc;

2)

pigments used in printing inks, paints, glass and plastic may contain cadmium, chromium, lead and zinc; and

3)

preservatives or fungicides used in paints and lumber products may contain arsenic, copper or mercury.

Several other examples of trace metals used in consumer products are given in Table 2.20.

44 Table 2.20 Some Potential Uses of Selected Trace Metals in Various Consumer Products Element Compound Use or Product Cadmium cadmium benzoate plastic stabiliser (chloride scavenger) yellow pigment for plastic cadmium sulphide yellow, red and green pigments Chromium various chromate leather tanning agent compounds lumber/finished wood preservative (fungicide) Copper copper arsenate phthalocyanine Cu blue pigment yellow pigment Lead lead chromate white pigment lead sulphate lead crystal lead oxide yellow pigment Zinc zinc chromate paper processing bleaching agent zinc dithionite widely used, e.g., paints, rubber products, zinc oxide inks, soaps, plastics, textiles, etc. zinc sulphide luminous materials, fluorescent lights Adapted from Rousseaux, 1988; Hammond, 1992; Chandler et al., 1992 Although bulk samples of MSW have been collected to determine their trace metal composition, obtaining a relatively small representative sample (1 to 10 grams) from a large quantity of heterogeneous material such as MSW is extremely difficult. Examples of reported concentrations of a wide variety of elements and trace metals in bulk MSW were determined during Environment Canada's NITEP Program and are provided in Table 2.21. In general, the statistics indicate that there is a high degree of variability in measurements for some metals, especially cadmium, chromium, lead and mercury, and certainly between sample sets from different locations. In addition to bulk analysis, there are two other methods that have been used to evaluate and characterise the trace metals in different components which make up the MSW stream, namely, a direct and an indirect approach. The direct approach is extremely labour intensive since it involves actual sampling and analyses of the various components of the waste stream. Typically, this method requires a large number of samples to generate statistically valid data, but is capable of producing an accurate picture of specific local waste streams. It should also account for all the various components in the waste stream, specifically the organic, soil and fines fractions. Alternately, the indirect approach is based on material flow models which involve examination of production data and estimation of product life expectancy. The data generated from this method is quite useful for providing benchmark data on a national or regional level and monitoring long-term trends which are free from seasonal variation

45 and sampling biases. However, the data is often not specific enough to develop local waste management strategies. There are numerous other advantages and disadvantages to both approaches, some of which have been summarised in Table 2.22.

Table 2.21 Summar~i of Elemental Concentrations in Various Fractions of MSW

Element AI

Charlottetown, PEI (Combustibles) n=12

Quebec City, PQ (Combustibles) n=12

x

x

I

sl

m

s

I1

m

12,050 4,060 12,225 5,530 1,740 5,233 Ba 1.2 4.O4 0 147 66.9 145 Ca 5,140 1,375 5,125 20,060 6,470 17,250 Cd 0.75 1.14 0 8.06 7.39 5.5 Cr 21.8 16.3 16.5 172 215 112 Co 0.17 0.58 0 3.71 1.76 2.8 Cu 48.3 25.4 41 430 660 108 Fe 2,365 1,830 1,960 6,050 1,385 5,970 Pb 82.4 47.4 79 732 1,080 255 Hg 0.17 0.39 0 1.23 1.04 0.74 Na 3,040 2,070 2,400 2,470 1,010 2,170 Ni 4.25 2.22 4 45.1 8.01 43 Sn 14.2 5.51 12.5 54.9 108 20 Zn 146 61.5 134 429 243 39 x = arithmetic mean s = standard deviation m = median Adapted from Environment Canada, 1985, 1988 & 1991

Hartford, CT (RDF) n=12

x

I

s

I

m

72,220 19,930 66,200 385 130 395 76,260 14,460 79,700 30.3 10.1 29.0 433 495 275 52.8 34.2 42.9 8,930 17,130 1,720 31,930 20,205 23,050 2,760 2 , 1 5 5 1,820 0.11 0.10 0.08 71,960 63,550 53,000 442 407 280 889 337 875 5,870 9 , 0 5 0 2,560

Examples of data generated from three studies which used the direct approach for evaluating MSW composition are given in Figures 2.19, 2.20 and Table 2.23. The data generated from the French study conducted under the R&D Programme on Recycling and Utilisation of Waste were based on a survey of results from 26 smaller studies of MSW from five different countries, namely, France, Germany, Netherlands, Switzerland and Sweden (Rousseaux, 1988). The results from the Waste Analysis, Sampling, Testing and Evaluation (WASTE) Program were based on 31 separate samples of MSW from Vancouver, British Columbia (Chandler et al., 1992). The Dutch study of household waste from the Bilthoven area was based on analysing several 700 kg samples of waste. All of the programs used similar classification systems to define the waste. The distribution of elements within the different waste categories from the studies indicates that the organic, fines and battery fraction of MSW contribute substantial proportions of many elements to the waste stream.

Figure 2.19 Distribution of Trace Metals in North American MSW

Al

As

B

Ba

Cd

Cr

Cu

Fe

Hg

Mn

Pb

Se

Sn

Zn

Total

Total W a s t e Stream Ferrous plastic

Chandler et al., 1992

Fines

I paper

Glass

Inorganic

Non-Ferrous

Organics

Figure 2.20 Distribution of Trace Metals in European MSW

Cd

Cr

Hg

Cu

Pb

Ni

Zn

Total W a s t e S t r e a m Non-Ferrous

KG

paper

Adapted from Rousseaux, 1988

Fines

Li

~lastics

Em

Glass Ferrous

a

Organics Batteries/Other

48 Table 2.22 Comparison of Direct and Indirect Approaches to MSW Characterisation Approach

Advantages

Disadvantages

Direct

Accurate for site-specific studies which are useful for developing local solid waste management strategies Accounts for all materials in the stream including organics (kitchen and yard), soils and fines fractions Can account for variations in waste caused by waste source, recycling initiatives, season & climate Samples can be fully characterised, includincj potential leachability

Large number of samples required to be statistically valid Susceptible to generating skewed data if methodology selects atypical samples Labour intensive & expensive Data is time dependent & must be repeated if comparisons are to be made Requires very careful sample preparation and analyses

Indirect

Provides composition data on a broader context than direct, i.e., national data Can account for variations caused by imports and exports Useful for developing national or regional management strategies by ability to track trends in production

Model must include data on organic, soil & fines fractions Model must be comprehensive and account for all significant input of metals Data is not site, nor source, specific, i.e., may not be accurate for specific city or apply to specific residential or commercial waste streams Product life-span estimates are generalised and susceptible to biases Stability of waste components and other component characteristics cannot be measured Adapted from Rousseaux, 1988; US EPA, 1992; Chandler et al., 1992

An example of the indirect approach is a study commissioned by the US Environmental Protection Agency to identify potential sources of various trace metals in the waste stream by examining production data and material flows (Franklin, 1989). The material flow model used scenarios of 80 and 90% recovery of batteries for recycling to estimate the contribution of lead acid batteries and other materials to the waste stream. Based on the result from this study, lead acid batteries were believed to be responsible for contributing more than half of the lead in the waste stream (Figure 2.21). This is inconsistent with the data from the WASTE Program which indicated that the organics fraction contributed the greatest proportion of lead to MSW and underlines the importance of accounting for the composition of the organic, soil and fines fractions of the waste stream in the material flow model.

Table 2.23 Distribution of Trace Metals in Dutch Household Waste Component Organics Paper Plastic Glass Ferrous Nonferrous Textiles Bread Ceramics Wood

Adapted from Beek et al., 1988

50 Figure 2.21 Estimated Contributions of Lead to MSW in the US (90% Battery Recycling)

IEI, 1990

51 REFERENCES

Baker, L.E. "Municipal Refuse Incineration in Great Britain", Ener.qy from Waste State0f-the-Art Report - 1st Edition, ISWA Working Group on Waste Incineration, Swedish Association of Solid Waste Management, Malm(~, Sweden, November 1991. Baltzinger, D. "Solid Waste Incineration in France", Ener.qyfrom Waste State-of-the-Art Report- 1st Edition, ISWA Working Group on Waste Incineration, Swedish Association of Solid Waste Management, Malm5, Sweden, November 1991. Barniske, L. "State of Waste Incineration in the Federal Republic of Germany", Umweltbundesamt (Federal Environmental Agency), Berlin, FRG, 1990. van de Beek, A.I.M.,A.A.J. Cornelissen and Th.G. Aalbers. Fysisch en Chemisch .0nderzoek aan Huishoudelijk Afval, Pub. Rijksinstituut Voor Volksgezondheid en Milieuhygiene Bilthoven, Rapportnr. 738505005, June 1988. Bergstr6m, R. "Disposal of Domestic Refuse and Industrial Refuse in Sweden 1987", Report for Stockholm Energi, Sweden, 1988. Bird and Hale Ltd., Clarke Derszko and Associates Ltd., L.S. Love and Associates Ltd. "Municipal Refuse Statistics for Canadian Communities of Over 100,000 (1976-1977), Report prepared for Environmental Protection Service, Environment Canada, 1977. Canadian Council of Ministers of the Environment (CCME). "Operating and Emission Guidelines for Municipal Solid Waste Incinerators", Government of Canada Publication, CCME-TS/WM-TRE003, Winnipeg, Manitoba, June 1989. Canadian Council of Ministers of the Environment (CCME). "Waste Management Committee Recommendation on Waste Reduction Objectives", unpublished Government of Canada Report, April 1990a. Canadian Council of Ministers of the Environment (CCME). National Packaging Protocol, Government of Canada Publication, CCME-TS/WM-FS020, Winnipeg, Manitoba, 1990b. Chandler, A.J., H.G. Rigo and S.E. Sawell. "Trace Metals and Municipal Solid Waste The Environmental Management Challenge", procee .din.as of the 14th Canadian Waste Mana ciement Conference, Regina, Saskatchewan, October 1992. Danish National Agency for Environmental Protection. Orienteering fra Milj~stryrelsen, Copenhagen, 1986.

"Waste in Denmark"

Danish Environmental Protection Agency. Waste Mana.ciement in Denmark, Ministry of Environment and Energy, Copenhagen, Denmark, 1995.

52 Eller, R. "Solid Waste Management in Japan". Article in WasteAqe_, pp 53 -64, May 1992. Environment Canada. The National Incinerator Testin.q and Evaluation Program Twosta.qe Com.bustion (Prince Edward Island), Environment Canada Reports EPS 3/UP/l, Vols 1-4, September 1985. _

_

Environment Canada. The National Incinerator Testin.q and Evaluation Pro.qram: Environmental Characterization of Mass Burn Technolo.qy at Quebec City, Environment Canada Reports EPS 3/UP/5, Vols 1-7, June 1988. .

.

.

.

Environment Canada. "Comparison of Per Capita Waste Generation Rates", Waste Management Branch, Government of Canada, 1989. Environment Canada. The National Inci.nerator Testing and Evaluation Program: The Environmental Characteri..zation of RDF Technology (Mid-Connecticut), Environment Canada Reports - Waste Management Series WM/14, Vols 1-6, March 1991. Environment Canada. The State of Canada's Environment, Government of Canada Publication, Chapter 25, Ottawa, Canada, 1991. Environment Canada. The Waste Analysis, Samplin.q, Testin.q and Evaluation (WASTEJ Pro.qram" The Effect of Waste Stream Characteristics on Mu..nicipal Solid Waste Incineration-. Fate and Beh.aviourof Trace Metals, Government of Canada Publication, Report EPS 3/HA/10, December 1995. .

.

.

.

.

.

.

.

Environment Canada. An Assessment of the Physical, Economic and Ener.qy Dimensions of Solid. Waste Mana,qement in Canada, Government of Canada Publication, Report EPS 21UPI21E, March 1996. Finkelstein, A. "An Overview of Environment Canada's National Incinerator Testing and Evaluation Program", proceedings of the .Municip.al Waste Combustion Conference, Tampa, Florida, April 1991. Folmer. D.T. "Structure of waste processing in the Netherlands", Ener.qy from Waste State-of-the-Art Report- 1st Edition, ISWA Working Group on Waste Incineration, Swedish Association of Solid Waste Management, Maim6, Sweden, November 1991. Franklin Associates, Ltd. "Characterization of Municipal Solid Waste in the United States, 1960 to 2000 (Update 1988)", final report prepared for the U.S. EPA, Prairie Village, Kansas, March 1988. Franklin Associates, Ltd. "Characterization of Products Containing Lead and Cadmium in Municipal Solid Waste in the United States, 1970 to 2000", Final Report prepared for US EPA, Municipal Solid Waste Program, Prairie Village, Kansas, 1989.

53 Franklin Associates, Ltd., 1992. "Characterization of Municipal Solid Waste in the United States, 1960 to 2000 (Update 1992)", prepared for the U.S. EPA. Prairie Village, Kansas, March 1988. Gore and Storrie, Ltd. "Composition of Residential Waste in Ontario", Report to Ontario Ministry of the Environment, 1992. Government of Japan. "Quality of the Environment in Japan 1989", Environmental Agency Report, Tokyo, 1990. Government of the Netherlands. "Memorandum on the Prevention and Recycling of Waste", Joint publication of the Ministries of Housing, Physical Planning and Environment, Economic Affairs, Agriculture and Fisheries, and Transport and Water Management, October 1988. Government of the Netherlands. "National Environmental Policy Plan Plus", Joint publication of the Ministries of Housing, Physical Planning and Environment, Economic Affairs, Agriculture and Fisheries, and Transport and Water Management, 1991a. Government of the Netherlands. "Holland Waste Handling - From Refuse to Reuse", Joint publication of National Institute of Public Health & Environmental Protection, and Netherlands Agency for Energy and the Environment, published by DHV Environment and Infrastructure, 1991. Government of the Netherlands. Personal communication with staff from the Dutch National Institute of Public Health and Environmental Protection (RIVM), August 1992. Government of Ontario. "The Physical & Economic Dimensions of Municipal Solid Waste Management in Ontario", Report to Ministry of the Environment, prepared by CH2M Hill Ltd., 1992. Haley, C.A.C. "Energy Recovery from Burning Municipal Solid Wastes: A Review", Resources, Conseryation and Recycling, 4, pp 77-103, 1990. Hammond, C.R. "The Elements", Handbook.0f C.hem..istry and Physics, 72nd Edition, Ed. D.R. Lide, pp 4.1-4.34, 1992. Haukohl, J. "Solid Waste Incineration in Denmark", Energy from W.aste State-of-the-Art_ Report - 1st Edition, ISWA Working Group on Waste Incineration, Swedish Association of Solid Waste Management, Malm(~, Sweden, November 1991. Her Majesty's SO, British Government Publication, United Kingdom, 1990. Her Majesty's Inspectorate of Pollution Control. "Environmental Protection Act 1990". British Government publication, 1990.

54 Hinchcliffe, P.R. "The Role of the Department of the Environment in Improving Technical Standards", proceedin.qs of the Harwell Waste Mana.c]ement Symposium New Developments in Landfill, Harwell, Oxfordshire, UK, May 1992. Hjelmar, O. and L.M. Johannesen. "Groundwater Protection and Landfill in Denmark", Proceedin.qs of the Harwell Waste Mana.qement Symposium - New Developments in... Landfil.~.__JI,Harwell, Oxfordshire, UK, May 1992. Hjelmar, O. "Waste Management in Denmark", Proceedin,qs of the Seminar on Cycle and Stabilisation Technolo.qies of MSW Incineration Residues, Kyoto, Japan, March 1996. Industrial Economics, Incorporated (IEI). "Potential Human Exposures From Lead in Municipal Solid Waste", Report for the Lead Industries Association, Inc., May 1991. Knoche, Mireille. "Residues of Municipal and Hazardous Waste Incineration - Situation in France", Waste Mana,qement International - Volume 2, Berlin, Germany, November 1992. Krol, A. and C Dent. "Municipal Solid Waste Conversion to Energy: A Summary of Current Research and Development Activity in the UK", Report for the Internationa..[ Ener.ay A,qency - MSW Conversion Activity, Task IV, published by UK Atomic Energy Authority, Harwell, 1988. Ministry of Health and Welfare. "The Waste Treatment and Disposal in Fiscal 1991 in Japan", Government of Japan (in Japanese), 1994. Niessen, W.R. and S.H. Chansky. "The Nature of Refuse", Proceedin.as from the 1970 .ASME Incineration Conference, ASME, p.1, New York, 1970. Nilsson, K. "Solid Waste Incineration in Sweden", Ener.qyfrom Waste State-of-the-Art Report - 1st Edition, ISWA Working Group on Waste Incineration, Swedish Association of Solid Waste Management, Malm(~, Sweden, November 1991. Office of Technology Assessment (OTA). Facin,q America's Trash - What Next for Municipal Solid Wa,s.te?, Publication of the Congress of the United States, OTA-O-424, Washington, D.C., October 1989. Patel, N.M. and D. Edgcumbe. "Some Observations on Municipal Solid Waste Management in Japan", Trip Report for the UK Department of Trade and Industry, September 1992. Reimann, D. "National Information on Waste Incineration (in Germany)", Enerqy from Waste State-of-the-Art Report- 1st Edition, ISWA Working Group on Waste Incineration, Swedish Association of Solid Waste Management, MalmO, Sweden, November 1991.

55 Resource Conservation and Recovery Act (RCRA). US Public Law 98-616, Hazardous and Solid Waste Amendments of 1984. Rijpkema, L.P.M., G.W. Krajenbrink, P.W.A. Stijnman and J.L.B. de Groot. Survey of Municipal Solid Waste Combustion in Europe - Data for 17 Europe_an Countrie.S, Report prepared by TNO Environmental and Energy Research, Reference Number 92-304, August 1992. Reimer, H. "Grundlagen der M011verbrennungstechnik". In M011- Handbuch (G. H0sel, W. Schenkel and H. Schnurer, Eds), Berlin: Erich Schmidt Verlag, Kennzahl 7140, Lfg. X1/75. Rousseaux, P. "Heavy Metals in Household Refuse - Origins, Chemical Forms, Contents", Report to Commission Des Communautes Europeennes, Ministere Delegue a L'environnement (France) and Agence Nationale Pour La Recuperation Et L'elimination Des Dechets, December 1988. RVF. Waste Manaclement in Sweden, Swedish Association of Waste Management, Maim0, Sweden, 1994. Sakai, S. "Municipal Solid Waste Management in Japan", .P_rocee_din.qsof the Seminar on Cycle and stabilisation Techn01o.qies of MSW Incinerator Residues, Kyoto, Japan, March 1996. Sawell, S.E. "Municipal Solid Waste Conversion to Energy: An Update of Research Activities in Canada", Report prepared for the International Energy Agency, Biomass Conversion Annex VII, Conversion of Waste to Energy, 1992. Steuteville, R. "The State of Garbage in America", Biocycl.e, 36(4):54 - 63, 1995. Swedish Association of Public Cleansing & Solid Waste Management (SAPCSWM). "Solid Waste Management in Sweden", Government of Sweden publication, February 1988. Swiss Environmental Protection Agency. "Ordinance Relating to Treatment of Waste", Government of Switzerland publication, 1988. Tabasaran, O. "Composition of Waste in Switzerland", M011u. Abfall, S. 17, 1984. Tsukamoto, S. "Solid Waste Incineration in Japan", Ener.qyfrom Waste. State-of-the-Art Report - 1.s.tEdition, ISWA Working Group on Waste Incineration, Swedish Association of Solid Waste Management, Maim6, Sweden, November 1991.

56 Umweltbundesamt. "Stellnwert der Hausmuellverbrennung in der Abfallentsorgung (The Relative Importance of Incineration in the Management of Municipal Refuse)", German Federal Environment Agency, 1990. United Kingdom Department of Environment. piqest 0f En.vironmental Statistics, No 17, HMSO London, 1995. United States Environmental Protection Agency (US EPA). "Characterization of Municipal Solid Waste in the United States: 1990 Update". Report for Office of Solid Waste, EPN530-SW-90-042, June 1991. United States Environmental Protection Agency (US EPA). "Characterization of Municipal Solid Waste in the United States: 1992 Update". Report for Office of Solid Waste, EPA/530-R-92-019, July 1992. Veabrin. Personal communication with P. Leenders and J.P. Born, 1991. Vehlow, J. "Thermische Zerst6rung Organischer Verbindungen", Proceedin_as of II. AbfalI-Wirtschafts-Symposium, Berlin, Germany, November 1991. Vehlow, J. "Municipal Solid Waste Management in Germany", proceedin.c]s of the Seminar on Cycle and Stabilisation Technolo.aies of MSW Incinerator Residues, Kyoto, Japan, March 1996. Vestforbraending. "Annual Statistics 1990", Report published by Vestforbraending, A Danish waste management company, 1990. Walls, T.B. "Energy Recovery from Mass Burning of Municipal Solid Waste", Proceed!n.as of Resource Recovery from Solid. Waste Conference, Florida, May 1982. Waste Age. "The 1992 Municipal Waste Combustion Guide", Developed by the National Solid Wastes Management Association, Waste A_qe Ma.aazine, November 1992. WASTE Program. "Waste Analysis, Sampling, Testing and Evaluation Program: Effect of Waste Stream Characteristics on MSW Incineration - The Fate and Behaviour of Metals". Phase I final draft report prepared for Environment Canada, US EPA and the International Lead Zinc Research Organization, unpublished, 1993. Waste Resources Department. "Personal communication with Departmental personnel, Energy Technology Support Unit, Harwell, UK, 1995.

57 Williams, H.E. "Recommendations for the Classification and Description of Household Waste and Refuse Derived Fuel (RDF)", The.Final Report of the Co.-0rdination Group on Classification and Analysis.of Househo[.d Wast..e. and the Speci.fi.cation of Refus.e Derived Fuel, Prepared by the author at Ecotec Research and Consulting Limited, Birmingham, England. World Resources Institute. Environmental.....Almanac 1992, Ed. A. Hammond, Pub. Hougton Mifflin, Boston, 1992.

This Page Intentionally Left Blank

59

CHAPTER 3 - MUNICIPAL SOLID WASTE INCINERATION TECHNOLOGIES

Incineration offers a means of managing MSW, thereby reducing landfilling requirements, and recovering the energy present in the materials being burned. Incineration technology has evolved dramatically during the past 15 years with the introduction of new system designs. Each modification of these systems has the potential to influence the physical and chemical nature of the residue streams. An appreciation of the differences in technology and how these differences can affect residue quality is necessary to developing an understanding of MSW incinerator residues and the options available for managing them in a sound environmental manner. The technology review in the next two chapters provides an overview of the various combustion and air pollution control alternatives currently in use, along with a discussion of their effects on residue streams and the handling of the waste products. MSW incinerator facilities typically contain several process sections: a waste receiving and storage area a waste feed system to charge the incinerator a combustion system a boiler to convert the heat of combustion to usable energy an air pollution control (APC) system an ash handling system. These processes can be arranged as shown in Figure 3.1. Each part of the system has a unique function and several sections of the plant are responsible for the generation of residue streams: The combustion unit produces the bulk of the residues, generally termed bottom ash. The heat recovery and APC systems generate smaller portions of the residue. The waste storage area and the ash quenching areas generate wastewater. The APC systems emit stack gases after cleanup. Residue characteristics can be affected by operations in all parts of the process. The descriptions that follow detail each operation and its potential effects on residues. 3.1 FUEL RECEIPT AND HANDLING

MSW is normally delivered to an incineration facility in the trucks used for local pick up. These receipts might be augmented by larger loads from intermediate collection stations where waste is off-loaded from the local vehicles, processed to remove selected materials such as corrugated paper and construction debris, and reloaded into

Figure 3.1 Generic Incinerator Plant Schematic

1. 2. 3. 4. 5. 6.

Refuse Coflec!ion Vehicle Refuse Storage Pi Refuse Handling Crane Feed Hopper Feeder Grate

Courtesy Martin GmbH

o m

7 . Forced-OraR Fan 8. Undergrate Air Zones 9. Furnace 10. Boiler 11. Ash Bunker 12. Superheater

13. Economiser 14. Dry Scrubber 15. Fabric Fitter Baghouse 16. Induced-Drafl Fan 17. Stack 18. Fly Ash Conveyor

61 larger transfer trailers. Under either circumstance, most incineration facilities require storage facilities because they only receive waste for a limited period during the day but operate around the clock. The storage facility provides a continuous source of material to the process. Commonly, storage facilities hold up to five days of fuel. This is sufficient to continue plant operation during holiday periods when there is no waste pick up. The majority of plants utilise a storage bunker in the form of a pit. This normally runs the full width of the incinerator installation. Trucks are off-loaded into the pit and a grapple crane is used to transfer materials from the pit to the charging hopper of the incinerator. An alternative used in smaller plants is the flat dumping floor. In this case a large open area is used to store waste which is stacked by front-end loaders to a height of 4 to 5 meters. The same loaders are then used to transfer waste to the incinerators. Steady operation of the incinerator requires continual loading of the unit with a relatively uniform fuel. Since most incinerators burn "as-received" MSW, an inherently heterogeneous material, the waste must be mixed to reduce the variability of the charged fuel. The operators perform this function in the storage area and at the same time remove large non-combustible components such as appliances and furniture. There is one major difference between the pit and the flat floor. During the mixing and retrieval operation of a pit system the finer fraction of the waste stream is sifted to lower levels in the pit. The nature of the grapple precludes removing all this material from the pit and some build-up of fines is inevitable. Fines, as discussed in Chapter 2, contain elevated levels of trace metals and charging higher quantities of these materials will produce variations in residue chemistry. With the flat floor system the waste mixing operation tends to mix the fines back into the bulk of the waste and the potential for segregation of fines and variations in residue chemistry is reduced.

3.2 AVAILABLE COMBUSTION ALTERNATIVES Although there are numerous systems available for the incineration of MSW and the generation of usable energy, combustion systems can be divided into two broad categories: mass burning: the "as-received" MSW is fed directly into the furnace and burned on a grate or hearth without any pretreatment such as size reduction, shredding or material separation prior to burning. refuse derived fuel (RDF): a fuel of a more homogeneous nature is prepared on-site and either burned in a "dedicated" furnace at the same location or sold to outside customers who utilise the fuel in their furnaces

62 Mass burning was adopted in Europe at the turn of the century and has evolved favourably over the past twenty years. Removal of oversized material is critical to the smooth operation of mass burn facilities. The RDF process involves the separation of certain materials from the waste to improve the combustion characteristics of the material. Various levels of processing are possible but they all involve some basic operations. The MSW is usually shredded to reduce the size of the material, sorted to remove non.combustibles and burned in semisuspension or suspension fired furnaces. Ferrous metals may be recovered using magnetic separators, and glass, grit and sand may be removed by screening. Further processing using air classifiers, rotary drums or advanced separation techniques can remove additional non-combustible materials and certain plastics and aluminum materials. During processing, the material is mixed to improve its homogeneity. In fact, some facilities process waste before feeding it to conventional grate type incinerators and consider the extra cost to be warranted because the system runs more smoothly. Any processing of the waste stream has the potential to change the nature of the fuel and thus change the residue streams. Later in this chapter, consideration is given to the changes processing can induce in the physical and chemical nature of the residues. There is no simple answer as to which incineration method is better. Each situation has to be considered on in its own merits, taking into consideration the institutional, environmental and economic issues. A summary of the key technical features of the systems follows.

3.2.1 Mass Burning Systems Mass burning is a well-established technology. Two types of mass burn systems are available: the European type system and the modular type system.

European Type Systems

The European systems have proven to be rugged as well as reliable and have been constructed in sizes ranging from 100 to 840 tonnes per day (Mg/d). This mass burning technology can be applied in almost all situations, however, it does not compete well with other incineration systems at design capacities below 300 Mg/d because of the high capital cost per tonne of waste burned. The European mass burning incinerator can be either of the refractory lined or the waterwall design. In a refractory lined furnace, combustion temperatures are regulated by using high excess air rates (100 to 200% excess air). In a waterwall furnace, the combustion temperature is maintained by circulating water in closely-spaced tubes located on the furnace walls. Most waterwall furnaces operate at a lower excess air

63 rate (about 80%), than refractory lined furnaces. This results in a reduction of both the furnace volume and the size of the air pollution control equipment. The basic combustion process in European mass burn furnaces, as described in detail in a later chapter, consists of layered burning of the waste on a grate that transports the material through the furnace (as shown in Figure 3.2). The fuel passes through various temperature regimes while on the grate. On the initial grate section both under-fire air supplied to the furnace and radiant heat from the furnace combine to dry the waste. Once dry, the waste begins to pyrolyse prior to burning. The pyrolysis and combustion process at this stage consumes the waste but generates significant quantities of hydrogen, carbon monoxide and unburned hydrocarbons. Additional air is required to complete the conversion to carbon dioxide and water vapour. This air is supplied above the material on the grate (over-fire air). The last section of the grate completes the reaction, driving the balance of the combustibles from the bed material. The material leaving the burnout section of the grate passes through a quench tank before being dewatered and conveyed to a bottom ash storage bunker. To maintain high combustion efficiency, sufficient time must be allowed for the last stage of combustion to go to completion thereby reducing residual carbon levels in the residue. Excessive feed rates, or insufficient air in the final stage will result in incomplete burnout and elevated carbon levels in the bottom ash. The balance of under-fire to over-fire air, the waste nature and the control of the system can influence the way the material burns and affect both ash quality and the air emissions. In order to promote good combustion, manufacturers of furnaces try to compensate for the natural variability of MSW by using different grate configurations and specialised air control systems. The grates agitate and move the waste through the furnace. Air control systems provide varying flows of air to different regions of the grate and to different areas of the zone above the grates. Maintaining uniform conditions reduces the possibility of operational problems caused by ash slagging or corrosion in the combustion zone. The manufacturers achieve their goals in various ways but the main variables they try to control are: bed coverage: maintaining a uniform distribution of waste on the grate combustion air flow: adjusting the initial combustion zone air to match the burning characteristics of the solids furnace configuration and the location of over-fire air ports: developing good mixing above the bed and enhancing the combustion effectiveness. Generally, the grate manufacturers provide the grate system and design the furnace configuration above the grate. When combined with their proprietary air control systems, the grate systems can meet guarantees of the appropriate level of combustion for the waste being burned. The energy recovery system downstream of the furnace can be supplied by any one of several boiler manufacturers. Companies manufacturing mass burning furnace systems that utilise either waterwall or refractory wall designs include:

64 Figure 3.2 Schematic of the Combustion Zone, Mass Burn Incinerator

Courtesy Martin GmbH

65 9 9 9 9 9

Alberti Bruun and Sorensen CEC de Bartolomeis Detroit Stoker

9 9 9 9 9

Deutsche Babcock EVT Heenan Nichol Martin SteinmQIler

9 9 9 9

VKW V~lund Von Roll Widmer and Ernst (now ABB W+E Umwelttecknik AG)

Differences in grate and furnace design or operating philosophy are based on the manufacturer's experience. To illustrate how different manufacturers address these issues, mass burn incinerator grates and furnace configuration are discussed in some detail.

Grates

The grate forms the bottom of the furnace and supports the burning bed of waste as it moves through the furnace. In designing the grate, care is taken to ensure that high temperature and fine ash do not affect its operation. Air flow through the grate acts to cool the grate bars and protect them from the high temperatures encountered in the furnace. The action of the grate, regardless of its design, will cause some sifting of the finer material downwards. Depending on the grate design, the degree of movement of this fine material (grate siftings or riddlings) into the under-fire air plenums can vary. Grate siftings are typically removed from these plenums and combined with the bottom ash in the ash extraction system. The efficiency of the grate system, as defined by consumption of carbon, depends upon its ability to provide combustion air to all the waste by means of a revolving and agitating movement. Manufacturers have various means of adjusting the flow of air through the grate. The provision of adjustable dampers and splitters that distribute air evenly to all parts of the grate is important. More important is ensuring a good pressure drop through the grate itself so that any variability in waste loading on the grate does not cause a shift of air away from a particular part of the grate. The net result of good combustion efficiency is a reduced level of carbon in the bottom ash residue stream. Most grate systems are some variation of one of the three forms of grates: rocking grates, reciprocating grates or travelling grates. Other alternatives, such as the roller grate are also marketed.

a)

Rocking Grates The grate sections, Figure 3.3a, are placed across the width of the furnace. Alternate rows are mechanically pivoted or rocked to produce an upward and forward motion, advancing and agitating the waste.

b)

Reciprocating Grates This design, Figure 3.3b, consists of sections that span the width of the furnace but are stacked above each other. Alternate grate sections slide back and forth while the adjacent sections remain fixed. Waste tumbles off the fixed portion and is agitated and mixed as it moves along the

66 grate. Numerous variations of this type of grate exist some with alternating fixed and moving sections, others with combinations of several moving sections to each fixed section. In the latter case, the moving sections can either move together or at different times in the cycle.

c)

Travelling Grates The travelling grate, Figure 3.3c, which consists of a continuous metal belt conveyor or interlocking linkages, moves along the longitudinal axis of the furnace. With a reduced potential to agitate the waste because it is only mixed as it is transferred from one belt to the next, the travelling grate system is seldom used in modern facilities.

d)

Roller Grates The roller grate, Figure 3.3d, consists of a perforated roller that traverses the width of the grate area. Several rollers are installed in series and a stirring action occurs at the transition when the material tumbles off the rollers.

The majority of grate systems in use in modern facilities are reciprocating and the quality of the burnout achieved by these systems is generally excellent. As noted earlier, inappropriate loading rates contribute to higher combustibles in bottom ash. It is worth discussing several variations of the reciprocating grate to illustrate how waste movement and combustion can be controlled. The Martin grate, Figure 3.3b, is a reciprocating grate which operates in the reverse direction to the flow of material on the grate. The grate is sloped from the feed end toward the residue discharge end and is comprised of fixed and moving grate steps. The moving grate steps perform a slow stirring stroke ensuring that the burning refuse layer is continually rotated and mingled to form an even depth of bed, and red-hot mass is pushed back toward the feed end of the grate. The intense fire builds up at the front end of the grate, with all combustion phases taking place simultaneously. The grate is divided longitudinally into several zones to enable under-fire air to be controlled. This air enters through both the bars and the narrow gaps between adjacent rows of bars. The gaps are restricted to ensure a high pressure drop across the grate. The Von Roll grate, Figure 3.4a, features a push forward block arrangement installed at a steep angle. In the case of low calorific value material the grate can be modified with steps to improve the separation of clumps of refuse. The grate is designed to operate with a high pressure drop for the under-fire air supply thereby ensuring good air distribution regardless of bed depth. The Widmer & Ernst (now ABB Umwelttecknick AG) grate, Figure 3.4b, is installed horizontally and the waste is transported along the grate solely through the "overthrust" action of the grate bars. The double motion tends to cause ignited particles to drop over the face of the fixed grate sections as the motion starts and then the grate below the fixed bar forces the burning material under the unburned waste above, thus prompting ignition from the bottom of the bed. Air flow is controlled through a high pressure drop through the grate.

67 Figure 3.3 Diagram of Various Grate Types Figure 3.3a Rocking Grate

Jjl

I

Figure 3.3b Reciprocating Grate (Martin)

|

68 Figure 3.3c Travelling Grate

II

m

9 @ ",,1~/ o

Figure 3.3d Roller Grate

69 Figure 3.4 Types of Reciprocating Grates a) Von Roll

. . . . . .

CROSS-SECTIONAL VIEW

LONGITUDINAL VIEW

adapted from Von Roll Schematic b) W + E

c) de Bartolomeis

70 The de Bartolomeis grate, Figure 3.4c, is unique in that it can be installed at any angle from horizontal to 21 ~, the angle being determined by the nature of the waste. The unit is a forward thrust grate. The three part design of the system incorporates a scraper that cleans the lower surface plate and controls the air flow to the grate. Variations on these grate designs are used to feed waste materials into the furnace. Waste is commonly fed into these furnaces through a chute. The chute is kept full of waste to minimise infiltration into the furnace. At the bottom of the chute a feeder system is used to meter the waste into the furnace. Figure 3.5 provides several examples of feeder systems. Figure 3.5a shows a travelling grate feeder; Figure 3.5b, a hydraulic ram type unit. In both cases, the furnace controls provide a uniform feed of material to the grate with the rate of feed being governed by the quality of the waste. Unlike grates inside the furnace, waste feed grates are not equipped with air supply systems. At the discharge end of the grate, the ash is transferred to a water quench system that serves to seal the discharge end of the furnace. Some manufacturers modify the end of the grate to extract the slag and transfer it to the quench tank; others merely allow the material to fall into the quench tank. Since manufacturers of all types of systems view bottom ash handling in a similar manner, the in-plant handling of residue is discussed after the various furnaces are reviewed.

Furnace Design

Mixing the waste on the grate and controlling the distribution of air to both the undergrate and over-grate regions are important factors in achieving the desired rate of combustion efficiency and ensuring minimal trace organic emissions. The furnace configuration also plays an important role in the ease with which combustion can be controlled and the quality of the ash leaving the grate. The path that the combustion gases take after they leave the burning waste is very important in ensuring uniform and complete combustion. There are several general configurations for the furnace. These are classified both as a function of the location of the furnace throat with respect to the grate and the flow direction of combustion products in the furnace relative to the waste flow. These classifications are based upon Deutsche Babcock Anlagen (DBA) studies that developed furnace geometries for particular types of wastes (Seeker et al., 1987). Figure 3.6 shows three possible furnace configurations. The parallel flow situation has the gases moving with the waste due to the presence of a hood or arch over the drying portion of the grate. This arrangement is recommended for highly volatile waste. At the other end of the scale, the contra flow system has the hot gases from the volatilisation zone flowing over the waste on the drying grate and increasing the removal of moisture from this material. The extra drying capability makes the contra system suitable for wet waste or waste with a high ash content. The drying effect of the furnace can be enhanced by adding heated air to the under-fire area. Both the wet

71 Figure 3.5 Types of Feed Grates a) Travelling Grate

1\ ~

'~ \

~~~~ . \

,,

iI

,

.,

1

b) Hydraulic Ram

% r-----7

r-------7 r - - - - - ~

Beseitigung von Abfallstoffen durch Verbrennung

r-T.

}{

]

Figure 3.6 Variations on Furnace Geometry

a ) parallel flow Dusseldorf #6

Beseitigung von Abfallstoffen durch Verbrennung

b) contra flow Frankfurt Nordweststadt

c) centre flow

73 and high ash wastes have a lower heating value than the material suited to the parallel configuration. The centre flow arrangement is more flexible than the other two having the throat located over the zone of volatile thermal decomposition. Additional air must be added to the gases leaving the fuel bed to complete the combustion process. Typically, 60 to 80% of the air added to the furnace comes from the under-fire system. With limited over-fire air, its addition must be carefully controlled to achieve the desired mixing. Proper design of the furnace throat and the over-fire air injection system ensures proper air mixing and leads to the control of organic contaminants in the flue gas and the APC residues. The throat causes a flow constriction, enhancing turbulence and providing the best location to ensure the complete mixing of the over-fire air and the combustion products. It is important that the over-fire air does not short-circuit or create temperature depressions if low emissions are to be achieved.

Operating Philosophy

The location and configuration of the over-fire air ports are made more critical by variations in the waste and the need to follow the steam load curve in EFW plants. Several control philosophies are employed: Von Roll monitors the steam production rate and controls the ram feeder frequency and the amount of primary air to the middle region of the grate, the pyrolysis region, to maintain the correct steam rate. Von Roll also monitor the furnace temperatures in the radiant region to control the secondary airflow rates. If the temperature drops, the secondary air can be reduced to restore temperatures to the correct level. Martin uses 02 levels in the flue gas to control the refuse ram feeder rate and the grate speed, thus controlling the MSW feed rate. A second control loop monitors steam rate and adjusts the under-fire air to control the steam production rate. Most of these combustion control measures are aimed at maintaining low organic emission rates from the furnace. However, optimizing these conditions can influence the trace metal partitioning between the furnace and the APC system. These influences are illustrated by data collected during Environment Canada's 1986 National Incinerator Testing and Evaluation Program (NITEP) study of the Quebec City facility in Canada.

Quebec City Modifications and Ash Quality

The NITEP tests run on the Quebec City incinerator provide an indication of the effects of design and operation on general ash parameters (Environment Canada, 1988). The Quebec City incinerators, based upon a Von Roll design, were installed by Dominion

74 Bridge in 1974. The units were modified in 1978 by the addition of a lined waterwall arch over the burning grate and a refractory chicane over the end of the burning grate. The configuration was very similar to the centre flow furnace arrangement discussed earlier. After the modification, it was found that the gas flow rates up the rear wall of the furnace were very high and this resulted in gas flow stratification in the furnace. This was judged to be unacceptable and further modifications were undertaken. These modifications, completed in 1986, lowered the roof over the burning and finishing grates and added both lower and upper bull nose sections on the rear wall of the furnace (Figure 3.7). This was intended to improve the ash quality by increasing the radiation reflection onto the burning and finishing grates. The modifications also served to reduce the flow of combustion gases leaving the finishing grate zone and enhance the burning of the volatile gases. The new arrangement could be be described as a contra flow furnace. During the latter set of modifications, the grate system was modified to provide dampers and flow monitoring/controlling systems to permit automatic control of the air split to the grates. This was prompted by observations that significant flow variations lead to unstable operation during periods of varying bed depth. The modifications allowed independent and automatic flow control to each of the grate hoppers thus maintaining the desired proportions on each grate section, and the correct total primary air flow to maintain good combustion and steam flow rate. Additional control modifications included a grate hydraulic control system that varied the grate operating frequency with respect to the steam flow, oxygen monitor feedback to the grate control system, primary airflow control by steam set-point and excess air level, and controlling both secondary air flow and front to rear air ratio as a function of temperature readings in the upper part of the radiation chamber. The modifications are reported to have resulted in the following changes in the ash: increasing the air to the first stage of the burning grate reduced the finishing grate bed thickness and improved ash quality through greater burnout. Raising the velocity through the bed increased the carry-over into the boiler, raised the level of combustible materials in the boiler ash, and particulate matter emission rates at the stack. Other studies have shown that these conditions can result in distinctly different chemical and leaching characteristics in the ash. It was concluded that control of primary air flow rates is critical in achieving low CO levels and low particulate emissions. increasing the total under-fired air to the finishing grate lowered upper radiation chamber temperature and produced high excess air levels. Reducing the air flow to the front of the finishing grate resulted in reducing the burning rate in this area and increasing the bed depth. This resulted in a decline in ash quality.

75

Figure 3.7 The Quebec City Incinerator Cross-Section r"-r

9COMPUTER CONTROL

ll

t-"

WATERWALL MEMBRANE

I

LARGER CHAMBER /

i

~

GAS

NEW CLADDING FRONT "BULL NOSE"

Ikk

NEW CLADDING ~ ,

!

FEED i GAS l

-----

HOPPER

E

NEW PRIMARY AIR SYSTEM

~ ~ .,

GA S

PIIVlSHING

Environment Canada, 1988

I

76 refuse bed depth indirectly establishes the amount of primary air required but also influences the extent of burnout. For a specific steam requirement, a thick refuse bed requires less primary air to supply the energy necessary to meet the steam demand. With the increased bed depth the primary air decreases, resulting in incomplete combustion and an increase in the amount of unburned material in the ash. while automatic grate speed control had been incorporated into the modifications test data suggested that manual control of the finishing grate speed resulted in improved ash quality. Clearly, there are many factors that influence ash quality from mass burn facilities, and both the designers and the operators need to become familiar with these cause/effect relationships if plants are to be run under optimum conditions.

Modular Incineration Systems Modular incineration systems are usually designed to burn MSW in an "as-received" state without size reduction or any other pretreatment operation. Typical modular incinerators for MSW applications range in capacity from 10 to 100 tonnes per day. The modular incinerator, also referred to as the controlled-air incinerator, makes use of a two-stage combustion process. It usually consists of a primary chamber and a secondary combustion chamber. The mode of operation of the primary chamber is used to classify controlled-air incinerators as either excess-air or starved-air (substoichiometric) units. The difference in these two modes of operation are summarised below:

a)

Starved-Air Incinerator The primary chamber of this incinerator is run without sufficient air to complete the burning process (below the stoichiometric requirement). Typically 30 to 80% of the stoichiometric requirement is provided. Without sufficient air, pyrolysis gases are formed in the primary chamber. Excess air is provided in the secondary or afterburner section of the incinerator to complete the combustion process.

b)

Excess-Air Incinerator The primary chamber in these units has more than the stoichiometric requirement of air. Typically 60-200% excess air is supplied to these units and this promotes almost complete combustion in the primary chamber (in the order of 90-95%). Gas-phase combustion is completed in the secondary chamber where additional air is added on an as-required basis.

Of the two types of controlled-air incinerators, the starved-air unit appears to be the more widely used. The success of the starved-air design has in large part been due to its ability to reduce the entrainment of particulate matter in the flue gas. This has been attributed to:

77 minimising disturbance of the fuel bed by limiting the number of grates maintaining a slow rate of volatilisation by reducing air flow into the chamber consuming any liberated particles in the secondary chamber. Most starved-air modular systems feature a stepped series of solid hearths with limited air injection points. This is different from European mass-burn and excess-air modular units that feature air introduction through the grate and numerous moving grate sections. Units are normally batch fed using a hopper/ram assembly or double ram system to minimise the infiltration of air into the primary chamber during charging. The waste is moved through the starved-air furnace by transfer rams placed along the stepped bottom of the furnace. This system retains a large mass of partially combusted material in the furnace at all times, thereby effectively equalising the energy release rate from heterogeneous waste streams. The controlled-air concept provides faster response to temperature fluctuations, with easier operating control than large conventional mass burning units. Historically, the limited disturbance of the bed and low air levels in the starved-air system resulted in poorer burnout conditions in the furnace and higher residual energy levels in the ash than commonly found in the ash from other technologies. Changes to the latest generation of starved-air units have included supplying air to the last hearth in the furnace to improve ash burnout. This has reduced the ash volume and lowered the unburned carbon levels to below 6% (Peel Resource Recovery Inc., 1992). This facility has noted increased particulate flows to the secondary chamber compared to operational experience at other facilities. It is not clear if this has resulted in changes to the concentrations of inorganics in the heat recovery and APC residues because comparsion data are limited. Modular incineration systems have a lower capital cost per daily tonne of waste burned compared to mass burning operations. However, the energy recovery efficiency is also lower; typically 55 to 60% compared to 65 to 70% for mass burning. The quantity of bottom ash generated from these unit is normally higher than from grate type systems, due in part to the lower burnout. Modular MSW incineration systems also have relatively low excess air requirements which can reduce the size of APC equipment. Controlled-air incinerators are manufactured by several vendors including Basic, Consumat, Morse Boulger, Smokatrol, and Simonds. A schematic of the Consumat modular incinerator are provided in Figures 3.8. These units are supplied as standard models but can be modified to suit the specific needs of a customer. Other Mass Burn Variants Several other mass burning technologies are in limited use throughout the world. Among these technologies are variations of the rotary kiln.

Figure 3.8 Schematic of Consumat System

Environment Canada, 1985

79 Rotary kilns can be of waterwall or refractory wall design and can also include ignition grates. Systems in current operation or under development include the V~lund rotary furnace, and the Westinghouse/O'Connor rotary kiln. A schematic of a rotary kiln system incorporating a drying grate and an ignition grate is presented in Figure 3.9.

a)

Velund System The V~lund system utilises a refractory wall rotary kiln design. The rotary kiln is used in conjunction with ignition grates located upstream of the kiln, although there are some cases where this may be reversed. Primary combustion air is introduced at the entrance of the combustion unit, while secondary air is injected at the exit end of the kiln. This technology is said to improve burnout of high moisture level materials. This is accomplished largely through increased residence time in the tumbling action of the rotary combustor section.

b)

Westinghouse/O'Connor System In the Westinghouse/O'Connor system, a waterwall rotary kiln is used as the main combustion unit. With this particular waterwall design (Figure 3.10) the rotary kiln has a cylindrical pinhole grate mounted over water tubes. During operation, the refuse travels downwards through the furnace as it burns. Combustion air is introduced along the entire length of the grate through air plenums between the boiler tubes. The unit operates at significantly lower excess air levels than conventional mass burn facilities.

3.2.2 Refuse Derived Fuel Systems Unlike the mass burn systems, Refuse Derived Fuel (RDF) systems fire a waste that has had its physical characteristics altered. The first step in this alteration is usually size reduction and this can be followed by various stages to remove non-combustibles and further reduce the size, or alternatively produce a more dense material. The American Society for Testing Materials (ASTM) Committee E-38.01 on Resource Recovery Energy (Seeker et al., 1987) defines seven categories of RDF processing: 1. oversize material removed 2. size reduction to minus 15 centimetre mesh with or without iron removal 3. shredded (minus 5 centimetre) with metal, glass and other inorganics removed 4. powdered form minus #10-mesh (0.2258 centimetres square) 5. densified into briquettes, pellets, etc. 6. processed to liquid fuels 7. processed to gaseous fuels.

Figure 3.9 Schematic of Valund Rotary Kiln Incinerator

Radian Corporation, 1989

81 Figure 3.10 Schematic of Westinghouse/O'Connor System

Seeker et al., 1987

82 The unit processes incorporated in RDF systems include primary shredding, ferrous metal recovery, screening or air classification to remove non-combustibles, secondary classification and storage (Figure 3.11 ). The net effect of RDF systems on ash quality is a reduction in the quantity of ash per tonne of waste introduced into the facility since many of the non-combustibles are removed in the process. The removal of materials such as ferrous and fines changes the composition of the ash. Furthermore, removal of some materials from the feed stream can change the characteristics of the combustion process and result in changes in partitioning as waste moves through the furnace. A number of demonstration and commercial scale RDF systems have been established in North America within the last decade. RDF systems typically have lower capital costs per daily tonne of waste processed than other energy-from-waste systems; however, operating costs are considerably higher. In addition, there are currently a number of problems associated with the establishment and operation of RDF facilities, including limited market for RDF and breakdowns in the waste handling/processing system. Furthermore, RDF may cause problems in storage (bridging or spontaneous combustion) and due to the typically high ash content, may overload the ash handling system on suspension-fired boilers. Densified or pelletised RDF has been suggested as a substitute for coal in the power generation field, however there are concerns about the variability in combustion characteristics of the material, the potential for increased air emissions, and the potential for increased fouling problems particularly with small boiler units (Glen and Howarth, 1988). Successful operating history has been limited by the many problems that these systems have encountered The RDF fuel is generally fired in suspension, stoker or fluidised bed incinerators. The most common of these are the suspension and stoker arrangements.

Semi-Suspension Burning Systems

Typically, semi-suspension RDF burning systems require size reduction and the removal of non-combustible materials from the waste stream prior to burning. These units are also used to burn wood waste. Operating experience with MSW on a large scale has been limited until recently when Combustion Engineering started the Hartford facility (Figure 3.12). The fuel is injected into the furnace through wall ports. Once in the furnace it ignites and burns while falling to the grate. These furnaces are generally equipped with a travelling screen grate system where final burnout occurs. The major design consideration with these systems is to ensure that the fuel is injected in such a manner that it builds an even bed across the grate, similar to the desire for uniform bed characteristics in the mass burn system. To accomplished this, the designers ensure heavier materials travel further across the furnace before they fall to the grate, and they design the injection system to spread the injected material across the grate.

Figure 3.1 1 Schematic of an RDF Processing Facility

FERROUS METAL

MUNICIPAL SOL10 WASTE

FLAILMILL

MAGNETIC SEPARATION

-

TROMMEt SCREEN

, RESIDUE

SHREDDER

FUEL

-8

--

u RDF STORAGE

Environment Canada, 1991

84 Figure 3.12 Cross-Section of a Semi-Suspension Combustion Unit (Mid-Connecticut)

RDF Distributors i

,9 ~n? ii

i

Grate Surface

~, o = m = /.

.

l. ,

I

Tangential Overflre Air

.

,

,

!,

.

|

I

9

am

L i"

,o,

o

i

'

Drive Shaft

Ib

Water Seal Idler Shaft

SiRing Screw Conveyor Undergrate Air Compartment

[ i

Environment Canada, 1991

85 Typically RDF burning offers higher energy recovery efficiency, lower excess air requirements and lower capital cost expenditures than mass burning systems. Despite these advantages, semi-suspension burning has limited economic benefits below 400 Mg/d.

Stoker Fired Systems

Stoker fired boilers are common in the utility industry. Their adaptation to burning MSW or combined MSW and coal was an early development that benefitted both the utilities by supplying fuel and the municipalities looking for MSW disposal options. This older technology is not used widely, but still represents a portion of the installed capacity. In stoker fired systems the RDF is injected into and distributed in the furnace by intercepting the fuel in the feed chute with a stream of high pressure air. The stoker system also contains a travelling grate and fresh fuel is injected onto the clean part of the grate. The distribution across the grate is controlled by the distribution air nozzle. Unlike older semi-suspension grate systems where the under-fire air is supplied by one plenum, the typical modern spreader stoker has several plenums to enhance air distribution on the grate. Over-fire air introduction into these systems is accomplished through a tangential entry system.

Fluidised Bed Systems

The fluidised bed reactor (Figure 3.13) is capable of destroying a wide range of wastes including sewage sludge, petroleum waste and paper industry waste. The units have been adapted to fire RDF materials and have shown promise in both Europe and Japan. Two new systems are under development in the United States. Sweden has five operating fluidised bed systems, but they tend to be small, handling 4,500 to 11,200 tonnes per annum (RVF, 1988). The reactor usually consists of a vertical refractory lined steel vessel containing a bed of granular material such as silica sand, limestone, alumina or ceramic material. The bed material is supported by a refractory lined grid. This grid is perforated to allow air to be injected through diffusers located below the grid. The air passing through the grid expands the bed by 80 to 100%, causing it to become fluidised. Wastes can be injected into the bed pneumatically, mechanically or by gravity. The constant moving action of the fluidized bed causes quick uniform mixing of wastes and bed material, resulting in good combustion conditions, and relatively high heat transfer rates. Furthermore, the movement in the bed and its inherent thermal storage capacity increases the burnout of material and minimizes bottom ash generation. This movement and the generation of fine particle in the bed leads to substantial quantities of fine ash being carried out of the bed by the air movement in the furnace. Thus, these systems generally require additional particulate removal devices in the gas stream ahead of boilers and air pollution control systems.

86 Figure 3.13 Schematic of a Fluidised Bed System

Refuse

Waste gas / k

Refuse constant feeder

~r"

V ,4

Incinerator

Non-combustible d~scharge conveyor Fluidizing air Vibrating screen

Sand circulating elevator

87 A typical fluidised bed reactor has a height to diameter ratio of 1.25:1 with the expanded bed occupying about 20% of the height. The bed material functions as a heat sink capable of absorbing large amounts of heat generated during the combustion process. Bed temperatures are typically maintained in the range of 760~ to 870~ which is lower than the operating temperatures of other types of systems. These lower gas temperatures and a lower excess air requirement minimizes the formation of nitrogen oxides. Fluidised bed systems may require auxiliary burners located either above or below the bed to maintain bed temperature, however other options are available to maximize thermal efficiency. The reactor can be operated either as a cold windbox in which the fluidising air is injected directly into the reactor or as a hot windbox in which the air is preheated in a heat exchanger or recuperator prior to injection depending upon the nature of the waste and the need to supply additional heat. Because of its simple design concept, the fluidised bed reactor has a low capital cost, a relatively long service life and low maintenance costs. In addition, this unit can tolerate large fluctuations in both waste composition and the rate of feed due to the high thermal inertia of the fluidised bed, typically in the order of 596,000 kJim 3 (16,000 Btu/ft3). Some of the potential problems and special considerations of the fluidised bed incinerator include the build up and removal of residual material from the bed, the formation of eutectic mixtures that fuse in the furnace, and bed degradation.

3.3 HEAT RECOVERY SYSTEMS Although incinerators can be used solely to reduce the volume of MSW requiring disposal, the economic and operational benefits of recovering energy and lowering flue gas temperatures has resulted in most modern MSW incinerator facilities being designed with boilers. Boilers installed downstream from the furnace are used to transfer heat from the flue gases to water and the hot water or steam is then used for space or process heating or converted to electricity using steam driven turbines. In North America, energy generation can account for up to 25% of the revenue for an incinerator facility. With the exception of some facilities in Japan and a few older facilities in North America, most incinerator plants now in operation recover the heat energy generated during incineration. As noted in the preceding sections of this chapter, the introduction of air into the bed of fuel can cause solid material to be released into the flue gas stream. This fine material is carried downstream until either a sharp change in flow direction or a drop in gas velocity causes it to be removed from the flow stream. Materials that volatilise on the grate also travel with the gas stream until the gases cool and the volatilised materials condense. Both these removal processes occur in the boiler where the gas flows past banks of tubes set perpendicular to the flow direction. As a result, residues are trapped in the heat recovery systems.

88 Several operational aspects of the heat recovery systems influence the nature of the residues collected in the heat recovery sections of the plant. These are addressed in this section, whereas the mechanisms responsible for residue deposition in boilers are discussed in a later chapter. A typical heat recovery system (Figure 3.14)includes: a radiant section where heat is recovered by radiation heat transfer from the flame zone in the furnace. The mechanism of heat transfer does not rely on convective heat transfer at this point. Generally this heat is removed by the water in the walls of the large mass burn facilities. However, it is possible to have the superheater discussed below in this section of the furnace as well. the convection section where the boiler tubes are perpendicular to the flow direction. This section recovers the majority of the available heat and produces saturated steam. the economiser section, which is placed after the convection section, and is similar in construction to the convection section. This section operates at a lower temperature generally to heat the feedwater for the boiler. the superheater section, used to add additional heat to the steam generated in the convection section of the furnace. It can be placed either before the convection section or before the economiser section. The selection of the location is a function of mass flow and the amount of energy the designer wants to remove from the gas stream. In most installations there are hoppers installed under of the heat recovery system to collect particulate matter. The hoppers are commonly equipped with double valves to allow the material to be removed without introducing air into the boiler during the ongoing operation of the unit. Heat recovery systems are operated to maximise heat recovery while minimising operating problems. The most common operating problem is boiler tube fouling caused by high flue gas temperatures entering the boiler. At higher temperatures, volatilized metals in the gas stream condense onto the tube surface and form a hard, tough deposit that resists removal by all but the most aggressive means. Gas inlet temperatures below 900~ reduce the degree of fouling on the tubes. Lowering the temperature results in the condensation of the volatilised materials in the gas stream and the transport of the fine particles further into the boiler where they do not appear to be as prone to forming a hard deposit on the tubes. Condensation reactions are not the only removal process occurring in the boiler. Impaction, settling and other deposition processes also remove materials from the gas stream through a combination of velocity and momentum changes. Some of the

89 materials settle out of the gas stream and fall to the bottom of the boiler. Material impacted, but not condensed on the boiler tubes can be removed during operation by mechanical cleaning. This is accomplished by injecting high pressure steam or air onto the tubes or by mechanically rapping the tubes to dislodge the deposits. The dislodged materials settle to the hoppers located at the bottom of the boiler and can be removed while the system is operating. The nature of the boiler residue is dependant upon the temperature in the various sections of the boiler as this influences the condensation process. The typical temperature profile in the boiler ranges from inlet values of 900~ to economiser exit gas temperatures in the region of 180~ 3.4 IN-PLANT RESIDUE MANAGEMENT

Solid residues, commonly referred to as ash, slag, or fly ash, are generated at various points in the incineration process. These materials require handling and eventual disposal. To complete the general introduction to residues, this section examines the common types of waste handling systems employed in MSW incineration plants and the nature of the material from each of the discharge points. For ease of presentation, the incinerator residue streams are defined as follows: bottom ash or slag grate siftings heat recovery system residues. Figure 3.14 Typical Heat Recovery System

FURNACE

90 3.4.1 Bottom Ash

The non-combustible fraction of the waste charged to the furnace forms a residue on the hearth. This material is generally referred to as bottom ash but is also called slag, grate ash or clinkers. Fluidised bed furnace bottom ash is commonly referred to as bed material. Removing the bottom ash from the incinerator must be done in a manner that minimises the ingress of air, maintaining control of the combustion process. The seal on the furnace is generally provided by a column of water. Since bottom ash may still contain carbon that would continue to smoulder after leaving the grate, the water serves to extinguish any remaining combustibles and cool the ash. Furthermore, large pieces of clinker fracture when quenched, reducing their size. Material discharged from the quench tank is normally transferred to an ash storage bunker where further dewatering takes place before a crane is used to transfer the material to containers or vehicles destined for utilisation or disposal sites. Bottom ash is normally wet when it leaves the plant thereby minimising fugitive dust emissions. To minimise leakage of contaminated water onto roads and into streams, the containers or vehicles are typically covered and designed watertight. The major design difference between various facilities is the mechanism used to remove the bottom ash from the quench tank. Different extraction devices influence the characteristics of the bottom ash. All extraction devices need to be designed to minimise operational difficulties induced if large non-combustibles are fed into the furnace and jam the ash removal mechanism. Several types of ash removal systems exist: drag-chain conveyors plate conveyors hydraulically ram type systems. The drag-chain conveyor, Figure 3.15, consists of a series of scrapers attached to a chain. The scrapers cover the width of the quench tank and move along the floor of the tank before riding up a slope to the discharge point. They then return overhead or under the tank back to the input end of the tank where they again enter the water. Ash discharged to the tank settles to the bottom and the scraper moves it as it passes through the tank. The inclined discharge chute allows the free water to drain back to the quench tank. Floating ash tends to build up on the surface of the quench tank in these systems and only limited amounts of this material are removed with each operational sequence of the conveyor. Eventually the system must be shut down and drained to clean the tank. Plate type conveyors, Figure 3.16, are similar in concept to the drag-chain conveyor. However, the plates are fully immersed in the quench tank and ash landing on the plates is transported out of the tank.

91 Figure 3.15 Diagram of Drag-Chain Conveyor System

Grate "

Siftings

'\X-x,X-X~\

II

II

Ash

Ii _~. _j, Ti\

Beseitigung von Abfallstoffen durch Verbrennung Figure 3.16 Diagram of a Plate Conveyor

Grate

I Ash

I

~ -T T- ....7

/

~ ~T--.il ~ ~

Beseitigung von Abfallstoffen durch Verbrennung

/

92 Hydraulic rams and sweeps are becoming a preferred technology, particularly on larger incinerators. In most cases these robust devices are capable of overcoming the problems associated with large non-combustible components charged to the furnace. Martin offers their ash discharger, Figure 3.17, for use on numerous services including MSW and coal-fired systems. The unit connected to the discharge of the furnace by a vertical inlet chute that is partially filled with water, consists of a curved bottom plate and inclined discharge chute. The ram inside the discharger continuously reciprocates along the bottom plate pushing the slag ahead of itself. The slag is compressed through the discharge chute, facilitating dewatering and reducing the size of any large clinkers or large compressible materials. The moisture content of the ash leaving this discharger is substantially lower than that leaving most drag-chain conveyor systems. The Simonds hydraulic sweep design, Figure 3.18, consists of a sweep arm that traverses the bottom of a curved tank during the discharge cycle and returns to the starting point of the cycle by a shorter radius route. This path minimizes disturbance in the tank thereby allowing ash to settle to the bottom. The design of the quench tank and ash removal system can influence the moisture of the bottom ash leaving the quench tank; however, the ash itself has a major influence on moisture levels. High residual carbon content in bottom ash will enhance the moisture holding capacity. For example, ash from drag chain conveyors on controlled air incinerators has been found to contain up to 50% moisture. Under any circumstances the quenched bottom ash is wet enough to negate fugitive dust emissions within the facility or during transport. Metal and glass make up a significant portion of the total ash mass. Ferrous metal in bottom ash can be recycled, so it is not common for the ferrous material to be extracted using magnetic separation prior to the bottom ash being discharged to the ash bunker. This reduces the mass and volume of material that needs to be managed. Bottom ash can also be screened to remove the oversize or fine materials and enhance the physical properties of the ash, facilitating utilisation. In some cases the material may need to be crushed prior to screening to remove the large size fraction that is unsuitable. In Europe, ash destined for use as an aggregate is stockpiled outdoors and allowed to age for a period before placement. Ash destined for disposal is frequently discharged to an ash bunker where a crane is used to transfer the material to containers for transport to the disposal site. Bottom ash is normally wet when it leaves the plant and, to prevent leakage of contaminated water onto roads and into streams, the trucks are typically covered and watertight.

3.4.2 Grate Siftings The material that passes through the openings in the grate, either because of its size or because it melts, and is trapped the under-fire air plenums located below the grate

93

Figure 3.17 Diagram of the Martin Discharger

Ash

i

Courtesy Martin GmbH Figure 3.18 Diagram of the Simonds Discharger /7

/ Y\

1

/% / \ ///~% \\ / / \ / / \

\

,<,,

\

",~ !

/ ~.,oo.o b

d _ \

Courtesy Simmonds

\

~

ri tl

k

,.~,~,~,,o,

,i3

94 is referred to as grate siftings. The siftings are normally conveyed to the quench tank for mixing with the bottom ash. During normal shutdowns it is necessary to clean the plenums to remove materials that freeze on the cooler surfaces. Aluminum and lead are two metals commonly found in grate siftings.

3.4.3 Heat Transfer System Ash As discussed earlier, some of the metals volatilised from the waste on the grate pass through the furnace and are condensed on the cooler surfaces of the boiler and other heat transfer equipment. This material acts as an insulator reducing heat transfer rates and must be removed to maintain process efficiency. When removed, it falls into the hoppers under the boiler along with other materials discussed previously. The ash collected in the hoppers of the boiler is dry and can be removed through air lock valves. In many plants the boiler ash is transported by screw conveyors to the bottom ash discharge sump but in newer plants it is being discharged separately or mixed with the APC residues for treatment and management. When collected separately, the material is transferred to containers for shipment to disposal sites. Boiler/economiser ash removed during operation is generally a fine material, smaller than sand. The material can be sticky or tacky by nature and may be difficult to handle. Although on-line cleaning can remove a substantial amount of residue from the boiler sections, over a period of time the ash coats the boiler tubes and results in a decrease in heat transfer capabilities. When heat transfer is impeded, exit gas temperatures rise and steam production falls, necessitating unit shutdown and cleaning. When the boiler has cooled, workers enter the boiler chamber and using a combination of air, water and mechanical cleaning procedures or explosive charges remove the residue coating the tubes. In large mass burn facilities this process can occur as often as quarterly but generally once or twice a year; in smaller starved-air facilities it could be as frequent as every 6 to 8 weeks. The material removed during maintenance activities tends to be sintered into large pieces. Several precautions should be taken in handling these materials. The heat recovery ash is fine and can easily become airborne. This material poses a potential health and safety problem to the workers in the plant and at the disposal site and to members of the public located adjacent to the plant, landfill or transportation routes. To minimise the potential for such fugitive emissions, it is common to wet the fly ash to the 5 to 15 % moisture level. The material is transported in covered trucks to minimise release to the environment and once at the disposal site is compacted and covered quickly.

95 REFERENCES Environment Canada. The National Incinerator Testin,q and Evaluation Pro.qram: Twostac]e Combustion (Prince Edward Island). Environment Canada Report EPS 3/UP/1 Vol 1, September 1985. Environment Canada. The Natj.0nal Incinerator Testin.q and Evaluation Pro.qram: Environmental Characterization of Mass Burn Technolo.(:w at Quebec City. Environment Canada Reports EPS 3/UP/5 Vols 1 - 7, June 1988. Environment Canada. The National Incinerator Testin,q and Evaluation Pro.qram: The Environmental Characterization .of RDF Technolo.qy (Mid-Connecticut). Environment Canada Reports Waste Management Series WM/14 Vols 1 -6, March 1991. Glen, N.F., J.D. Isdale, W.R. Ewart and J.H. Howarth. "Gas-side Fouling - The Limiting Factor in Recovering Energy From Waste?" Ener.(:w Recovery Throu.qh Waste Combustion Edited by Brown, Evemy and Ferrero. Elsevier Applied Science, 1988. Peel Resource Recovery Inc. Ash and Quench Water Testinc] Report, Volume I Main Report. Prepared for the Region of Peel by Air Testing Services, Toronto, Ontario, July 1992. Radian Corporation..Municipal Waste Combustion Multi-Pollutant Study, Emission Test Report, Mass Burn Refractory Incinerator, Mont.qomery County South, Ohio, Volume I, Summary of Results. Report prepared for the U.S. Environmental Protection Agency EPA-600/8-89-065a, August 1989. Svenska Renh~llningsverks-F6reningen (RVF). Solid Waste Mana.qement in Sweden. The Swedish Association of Public Cleansing and Solid Waste Management, Malmo, Sweden, February 1988. Seeker, W.R., W.S. Lanier and M.P. Heap. Municipal Waste Combustor Study: Combustion Control of Or.qanic Emissions. Report prepared for the U.S. EPA by Energy and Environmental Research Corporation, Irvine, Ca. EPA/530-SW-87-021C, 1987.

This Page Intentionally Left Blank

97 CHAPTER 4 - AIR EMISSION CONTROL STRATEGIES 4.0 INTRODUCTION

Significant improvements have been made in the air emission control systems of MSW incinerators [MSWI] since the early 1980s. These improvements, prompted by concerns over air emissions, were the result of significant research efforts that identified the mechanisms responsible for the release of various pollutants. This work provides an understanding of the processes that in turn allowed formulation of design and operating guidelines to limit the release of these pollutants (Seeker et al., 1987). The material in the gas stream leaving the heat recovery system consists of: 9 gaseous products of combustion including carbon dioxide, hydrogen chloride, sulphur dioxide and oxides of nitrogen 9 vapour forms of metals and organics 9 solid particulate matter. Testing programs on production installations have demonstrated that effective control of the emissions of trace metals, particulate matter and acid gases is possible with the new generation of equipment (Environment Canada, 1986; LIRPB, 1992). The environmental implications of the application of this technology, particularly with respect to changes in residues captured in the systems, need to be addressed. Successful control of MSWI air emissions involves a combination of two processes: 9 combustion control to limit conventional and trace contaminant emissions; 9 post-combustion control to reduce the amount of material leaving the stack. Discussions of air pollution control strategies must address both approaches because no modern plant can rely exclusively on one alternative. Good combustion control is the corner stone of most MSWI regulations issued in the past 6 years. Its application is standard in all new plants. This has led to a change in certain residue stream characteristics. Further changes have resulted from post-combustion control measures which, regardless of the degree of combustion control present in any plant, remove materials from the flue gas stream. To provide a basis for discussion of the changes in ash and residue quality a brief discussion of combustion control follows. 4.1 COMBUSTION CONTROL 4.1.1 Theory

Combustion control must compensate for: 9the natural variability in fuel quality; and, 9the controlling factors that govern the rate of chemical reactions.

98

Compensation for Fuel Variability

MSW is not a homogeneous material. As shown in Chapter 2, MSW is a mix of paper, plastics, other organic materials and non-combustibles. Each component has its inherent energy content and this must be matched with sufficient oxygen to ensure proper combustion. Because the mix of components changes from season to season, week to week and indeed from bag to bag, some means must be provided to allow the system to handle this variability. The first step in most facilities is to ensure that the arriving waste is well mixed before being charged to the furnace. The operators thus attempt to combine both the wet leaves and the plastic discards from a commercial establishment and average the energy level of materials entering the furnace. The well-mixed charge will still vary, but this variability can be handled in a well-designed furnace. The European mass burn systems generally utilise combustion control systems to compensate for the variability in the fuel. These systems sense the rate of heat release in the furnace and adjust the supply of combustion air to compensate for especially high or low heat-release rates. The means by which the designer accomplishes the adjustment is generally seen to be the most critical aspect of the control of the emissions of trace organic compounds. Modular systems generally rely upon maintaining a large quantity of fuel in the primary chamber to damp out the variability in the waste quality. The amount of fuel on the bed is typically greater than in either of the other designs. The production of RDF provides the ultimate control for fuel variability. The nature of RDF is different: the sizes of individual pieces of waste are more uniform; the non-combustibles have been removed; and, in the process the materials are thoroughly mixed. These characteristics make the control of the furnace less critical than it is in other combustion processes. Differences in the waste fed to the systems and the degree of control the operator has over the combustion process are reflected in variations of the residues streams resulting from the different types of systems. These aspects are addressed later in this chapter.

Factors Controlling the Chemical Reaction Rate

The thermal destruction of organics is not a simple process. Many intermediate steps are involved in the oxidation of long chain hydrocarbon materials to the products of complete combustion namely carbon dioxide (CO2) and water. The reaction that occurs on the bed of any incinerator is one of gasification, with the garbage being exposed to air while being heated. The amount of air added under the bed of material and the degree of agitation of this material controls the rate of gas generation. The gases that leave the bed are rich in carbon monoxide and hydrogen and contain many unburned hydrocarbons. When provided with additional air, these gases burn readily. This additional air, referred to as overfire air, is supplied above the bed. The degree to which the combustion process is completed is a function of how well the air and the gases are mixed. The amount of carbon dioxide generated defines the extent of combustion completion and is generally referred to as the level of combustion

99 efficiency. Complete combustion will result in the generation of CO2 and water vapour, however, 100% combustion is difficult to achieve in most combustion processes. Carbon monoxide (CO) is the most refractory species in the oxidative chain from hydrocarbon to carbon dioxide and water. The oxidation of CO to CO2 is accomplished much faster in the presence of hydrogen. Miller and Fisk (1987) suggest the dominant reaction in the chain is: CO + OH --> CO2 + H The concentration of hydroxyl radicals is very important in the reaction. However, the reaction between hydroxyl radicals and hydrocarbons is faster than that between CO and OH and it is necessary to consume all the hydrocarbons before the system can maximise the conversion of CO to CO2. Thus high levels of CO are generally correlated with higher levels of hydrocarbons, illustrating the rate-limiting steps in the reaction. If excessive air is present in the furnace, the combustion temperature and the concentration of hydroxyl radicals are reduced. In turn, the organics react with the OH radicals and the CO oxidation does not occur. Conversely, insufficient air can lead to pockets of fuel rich gas that lacks sufficient oxygen to oxidise the CO. It is possible to establish an appropriate range for the concentration of oxygen in any system. A typical curve is shown in Figure 4.1 and illustrates the limited extent of the appropriate or "good" operating region. Operation in this zone minimises the release of CO and thus also minimises trace organic releases. The establishment of this range is most important because, once determined for a system, it can be used for the purposes of ensuring that the system is operating at its most efficient level. Figure 4.1 Relationship of CO and 02 for Appropriate Operating Regions

(from DBA, 1986 as presented by Seeker et al., 1987)

100 Good combustion conditions leading to reduced organic emissions are those that: 9ensure complete mixing of the fuel and the air; 9maintain sufficiently high temperatures in the presence of sufficient oxygen; 9prevent the formation of quench zones or low temperature pathways that would allow partially-reacted solids or gases to exit from the combustion chamber. These design conditions must be combined with good operating conditions to ensure that the performance is maintained and organic constituents are reduced to basic elements. Combustion control addresses the destruction of organic compounds but it also has an influence on the downstream partitioning of inorganic materials in the incinerator. Higher temperatures in the furnace can influence the compounds formed in the furnace and the degree of volatilisation that occurs. Trace metals are neither created or destroyed in the combustion process; their form and speciation can be changed by the reaction and thus their eventual partitioning in the incinerator can be influenced. Higher temperatures and more complete combustion result in trace metals being found further down the system, particularly in the APC residue stream. These situations do not change the handling of MSW incinerator residues from the furnace or the heat recovery sections but they can have an effect on the ultimate management of these streams as the chemical composition changes. The conditions that lead to a reduction in organic emissions also can cause an increase in the generation of NOx. The formation of NOx is attributed to two mechanisms: the oxidation of the fuel nitrogen to NO~; and, the combination of nitrogen and oxygen in combustion air at high temperatures, the thermal NO~ portion. The conversion of fuel nitrogen to NO~ is dependent upon the local oxygen availability to volatile species, the amount of fuel-bound nitrogen and the chemical structure. The thermal NO~ reaction is strongly temperature dependent because it is formed by the combination of radicals of the two species. It has been shown that the conversion of fuel nitrogen can range from 5% to 50% controlled largely by the extent of mixing and the amount of oxygen present. Some combustion systems have been developed to reduce the levels of NOx created. These include reburning where flue radicals, primarily CH species, can reduce NO to molecular nitrogen. These processes were discussed by Seeker et al. (1987), and, more recently it was suggested that this technique can improve the destruction of trace organics (Seeker et al., 1991). Combustion control arrangements for NOx reduction in MSW incinerator systems have not progressed beyond the demonstration stage at this time. Reburning technology is most effective in those conditions were the use of the extra fuel is justified by improved thermal performance of the system but cost effectiveness may be improved when used in combination with other techniques. These other methods of NOx reduction include catalytic and non-catalytic techniques which are addressed later in this chapter.

101 If gas temperatures are not reduced in the APC system, volatile metallic species may be released from the stack. Regardless of the combustion control effectiveness, postcombustion control is necessary to limit metallic species emissions. Improved combustion control, aimed at reducing both emissions and operational problems such as high temperature corrosion in the reducing zones of the furnace, has placed more emphasis on combustion uniformity. This has improved the residue characteristics exhibited by newer units.

4.2 POST-COMBUSTION CONTROL Post-combustion control through the use of air pollution control (APC) systems will remove unwanted contaminants such as trace metals and various acid gases from the flue gases. Trace organics can also be reduced through the use of such systems (Environment Canada, 1986). There are three key aspects to the operation of these systems: the reagent used; the temperature control level; and, particulate removal efficiency. All have the potential to influence the characteristics of the solids found in the APC residue stream. APC systems rely upon both physical and chemical unit processes involving different solids removal and chemical conversion steps to effect control of unwanted emissions. These processes are combined to achieve the desired flue gas quality at an acceptable capital and operating cost. Different types of systems can change the quantity of residues resulting from the flue gas cleanup, thereby influencing disposal costs. Figure 4.2 provides a schematic of various APC system options as devised by Fl~lkt (1991). Table 4.1, adapted from Fl~kt (1991), provides a generalised comparison of the variations between the alternatives. The order of use of the various processes is governed by the selection of the process steps. These tools provide a relative comparison between different options and should not be used to selection purposes. For instance, apparent anomalies such as wet scrubber performance compared to other systems relate to Fl~kt's concerns about the need to clean the effluent from wet scrubbers. The heat potential category relates to the desire to use available heat for other purposes and in this case, the wet scrubber offers the best potential. However, if NOx control is to be used, some of this potential may not be realised. APC system development has changed over the years. The systems and methods discussed in this chapter represent the most current technologies in use throughout the world. Older technologies may be used in some locales but are being phased out. Since information on the characteristics of residues from these systems is not as readily available, only limited comment is provided. Individual unit processes are discussed in the following sections. These discussions outline the benefits and weaknesses of various types of devices for each of the process steps, with emphasis on the residues generated.

102 Figure 4.2 Comparison of Various APC Alternatives

Dry Cleaning System

I INCINERATORI

Wet-Dry Cleaning System Wet Cleaning System

!INCINERATORI

I INCINERATORI

stino,

ip I

9

i~X,. e. ,c,;l

Conversi~

I"eat~ec~

~

I COndensing I Stage

I "eat~ec~

Heat Pump

I ~e.ea~n~ I \

After Fl~kt, 1991

IIScrubber wet II )Water I Trca!mcnti

103 Table 4.1 Comparison of Operating Features of Various APC Alternatives CLEANING PRINCIPLE

DRY

INVESTMENT COSTS

WET

low

WET-DRY medium

usually high

lime consumption

medium

low

low

soda consumption

none

none

medium (if use specialso X stage)

energy consumption

low

medium

medium

very low

low

medium

dust

very high

very high

high

HCI and HF

very high

very high

very high

SOx

medium

medium

high (with specialstage)

low

low

none

heavy metals

very high

very high

medium

hydrocarbons

very high

very high

medium

high

medium

very high

OPERATING COSTS

maintenance requirements COLLECTION EFFICIENCY

NOx (withoutspecial add-onstage)

HEAT POTENTIAL After Fl~kt, 1991

4.2.1 Unit Processes for Air Pollution Control Particulate Matter Control Systems In the previous chapter, reference was made to various particulate removal mechanisms that result in the collection of particles in the boiler hoppers of MSW incinerators. A major physical mechanism responsible for this removal is settling. Settling requires low gas velocities and its effectiveness is limited by the size of the chamber required to remove fine particles from the gas stream. To increase the removal, early incinerator systems were equipped with cyclone separators that employed inertial forces to separate materials down to approximately 5 IJm in size. Other mechanisms were then utilised to remove the finer particles from the gas stream.

Liquids can be used to enhance the removal efficiency of fine particles from the gas stream. Two principal mechanisms remove aerosols:

104 9wetting the particles by contact and having the wetted particle impinge on a surface from which they are subsequently removed by the liquid; and, 9impinging dry particles on a wetted surface and then washing them off the surface. This removal mechanism depends upon establishing intimate contact between the particles and the wetting solution. This is accomplished either through impingement by spray droplets; diffusion; or, condensation. Impingement relies upon spraying the gas stream with liquid; its removal efficiency being proportional to the number of droplets and the force with which they are generated. Diffusion relies upon Brownian motion to bring dispersed particles in contact with liquid surfaces. Increasing the gas stream's turbulence will increase the removal efficiency and thus the frequent use of venturi devices combined with water sprays for high removal efficiencies. Condensation processes occur when the gas is cooled below its dew-point. In this condition, the particulate matter in the gas stream serve as nucleating sites for droplet formation and the particles are removed with the liquid. Conventional wet scrubber systems are effective at removing particles down to 1 pm. The removal of smaller particles requires another generation of control devices, the electrostatic precipitator (ESP), or the fabric filter (FF). The use of these devices has become dominant in modern incineration facilities.

Electrostatic Precipitators

Electrostatic precipitators (ESP) operate as follows: 9The gas stream passes through a series of discharge electrodes. These highly charged units impart a negative charge to the particles in the gas stream. 9A grounded surface, or collector electrode, is placed adjacent to the discharge electrode. The charged particles collect on the grounded surface. 9Particles are removed from the plates usually by rapping on the collectors. The removed particles drop into a hopper and are periodically removed. It is critical that material be removed from the collectors to prevent it from forming an insulator over the collector and reducing collection efficiency. Figure 4.3 shows a typical ESP system. The efficiency of ESPs is sensitive to variations in gas temperature and humidity and to changes in the nature and resistivity of the particles. Velocities through the precipitator usually range from 0.7 to 1.3 m/s. At these rates, the removal effectiveness is a function of the number of banks of precipitator fields used in the design. Manufacturers will quote efficiency factors as high as 99%+ if sufficient fields are used.

105

Figure 4.3 Typical ESP Installation

BUS DUCT ASSY

HIGH VOLTAGE SYSTEM RAPPE

INSULATOR ~COMPARTMENT VENTILATION SYSTEM ~I.C.V.S. CONTROL PANEL

INSULATOR COMPARTMEN

'RANSFORMER/RECTIFIER REACTOR

RAILINGs,.

-PRIMARY LOAD RAPPER WTROLPANEL ELECTRICAL EQUIPMENT PLATFORM

HIGH VOLTAGE SYSTEM UPPERSUPPORT FRAME

COLLECTING SURFACES

HIGH VOLTAGE ELECTRODES WITH WEIGHT CASI NG

COLLECTING SURFACE RAPPERS

24 in. MANHOLE

vL Typical wire-weight electrostatic precipitator with top housing. (Courtesy of Western Precipitation)

(from U.S. EPA, AP 40)

106 The ESP is efficient in the collection of material in the 0.1 to 10 micron (pm) size range, although it does have some limitations in the 0.1 to 1.0 pm range. Removing finer material also requires the use of extra fields. High efficiency ESP installations will remove 100% of the >50 pm material, >99% of the 5 pm material and 98% of the 1 pm and smaller material. Unlike fabric filters, emissions from ESPs can vary depending on the concentration of particulate matter in the flue gas stream. ESPs can be designed for temperatures as high as 375~ although more commonly they operate at about 200~ Some volatile metals will escape capture in ESPs operating at elevated temperatures because they will not condense and form particles that can be removed. Without an appreciable temperature drop volatile organic and metallic species will flow through the system. This is particularly the case for mercury as shown in Environment Canada's NITEP studies (Environment Canada, 1986). In addition, since trace metallic and organic species tend to condense on surfaces and the ESP has only limited fine particle control, the removal effectiveness for species condensing on the surfaces of the fine particles is reduced. Furthermore, ESPs operating at elevated temperatures have been shown by Vogg et a1.(1990) to be likely to have enhanced PCDD/PCDF in their residue streams. This finding suggests that de novo synthesis of these compounds occurs at the operating temperatures found in some units. On the other hand, ESPs operating at low temperatures (<200~ can experience premature failures due to corrosion damage. The most marked difference between the ESP and other types of high efficiency particulate matter control devices is cost. The initial cost of a high efficiency ESP unit will be substantially greater than that of most other types of control devices with similar performance, however, the operating costs are much lower. Pressure drops through ESP units are very low and thus the power expended for removal is low. With few moving parts, ESP maintenance concerns are usually a function of the type of materials being removed from the flue gases; sticky or corrosive fumes tend to raise the maintenance requirements. Although, as mentioned above, operating temperatures and flue gas composition must be considered if corrosion is to be avoided.

Fabric Filter (Baghouses) Fabric filters, or baghouses, are used in all types of industrial applications. They are essentially a set of permeable bags which allow the passage of gas but not the particulate matter entrained in the gas. They are effective for removing particles in the submicron range and their removal efficiency is typically better than that of the ESP. The fabric filter, Figure 4.4, consists of a large housing which may or may not be divided into a series of compartments. Each compartment contains a set of long cylindrical woven bags fitted over a wire mesh support. The bags and their support frame are fastened to a support rack which effectively seals the inlet gas stream from the exhaust or filtered stream. Gas is introduced into the inlet side of the compartment,

Figure 4.4 Typical Fabric Filter System

(courtesy Asea Brown Boveri Inc.)

108 passes through the bag material where the particulate material is filtered out by a combination of processes and is then exhausted from the outlet side of the compartment. As the process continues, the particulate matter removed from the gas stream forms a layer on the surface of the bag, generally called the cake, and the pressure drop across the filter increases. Periodically this cake must be removed from the filter either by mechanically shaking the bags or through the use of reversed or pulsed air flow. The residue released from the filter falls into a hopper under the unit and can be removed for disposal. The mechanisms governing removal on the fabric filter include: 9direct interception; 9inertial impaction; 9diffusion; 9electrostatic forces; 9weak molecular interaction forces; and, 9gravity. The presence of the cake on the bags enhances the removal efficiency of the system because it improves the interception of particles. Thus, the removal efficiency of fabric filter systems with newly installed bags can be lower than that of systems that have operated for several months. Some of this deficit can be overcome by using bags with natural fibres such as wool which provides a greater surface area to intercept particles than that found with bags woven of man-made fibres. The latter systems have better durability to acid gases and therefore, typical materials used for fabric filters in the MSW applications are glass fibre or Teflon. Typically an air-to-cloth ratio of 4:1 (m3/m2) is specified for pulse jet systems, but other types of systems incorporate a range of airto-cloth ratios from 2:1 to 10:1. The air-to-cloth ratio defines the flow rate through the filter material. A ratio of 4:1 implies that 4 cubic metres of gas per minute are filtered by every square metre of cloth in the filter compartment. The nature of the cake on the filter fabric is important to the performance of the filtering system. Even after cleaning, the bag will continue to retain a significant quantity of residue; bags can retain 10 to 20 times as much material as they filter out during any given filter cycle. This is beneficial when these systems rely on chemical and physical transformations for pollutant control. When unspent reagent is retained on the bag it provides additional acid gas control capability. Unlike the ESP, the fabric filter also has an ability to cope with fluctuations in both the particulate matter Ioadings and gas flow rates through the system. The fact that the emissions from typical MSW incineration processes are not consistent (since the feed material is not uniform) suggests that the fabric filter is a preferred particle control device if high removal efficiencies are required. Numerous factors can affect fabric filter performance. These include:

109 9flow rate; 9operating temperature; 9moisture levels; 9particle size range; and, 9particle characteristics. As noted earlier, typical flow rates for pulse jet systems are in the range of 4 m3/minute/m 2 of fabric. Higher flow rates increase the pressure drop and reduce bridging potentials, whereas lower flow rates reduce impaction processes. Gas temperature control is important to the removal of volatile materials. This is often accomplished by evaporating water in the gas stream, however, excess humidity in the flue gas can lead to filter blockage problems. Since fabric filters are frequently used in conjunction with lime addition for acid gas control, the residue removed on the bags can have a high percentage of hygroscopic salts. If the temperature and the humidity are not balanced properly, the nature of the residue layer on the bags will change. What should be a light, fluffy material that is easily removed during cleaning can become a sticky, dense layer that raises pressure drops across the system and leads to other operational difficulties such as: 9increases in energy costs due to higher pressure drops; 9a decrease in flow rate; and, 9reduced bag life due to adhesion and the abrasion properties of the particles. Current performance specifications for fabric filters on MSW incinerators suggest that the systems should have a routine performance capability below 0.008 grains per standard cubic foot of gas. This translates to an emission level of 18-20 mg/Rm 3 (dry gas at 25~ 101.3 kPa, and 11% 02).

Gaseous Controls

Particulate matter control is only one aspect of post-combustion control through the use of APC systems. Acid gases, particularly HCI, SO2 and HF require control and volatile materials, both organic and metallic, require special consideration if they are to be successfully controlled. The removal of gaseous phase acidic components can be accomplished by neutralisation. This requires the addition of chemical reagents either in liquid or dry form, hence two types of systems generally exist: 9wet systems, where the gases are passed through a liquid solution; and, 9dry systems, where the gas stream does not reach saturation. Special measures are required for NOx control.

Wet Systems

There are numerous types of wet collectors, ranging from modified cyclones containing water sprays to venturi scrubbers and packed tower units. Considered separately,

110 none of these systems is capable of performing at the level required of a modern APC system; rather they are frequently used in combination with other devices to achieve the required performance. An example of such a system is the dual wet scrubber required for acid gas control. Because wet systems operating at a low pH to remove HCI, HF, HBr and Hg do not remove SO2 effectively, a second stage, operated at neutral pH levels, is required. The advantage of wet systems is that they operate close to stoichiometry thereby reducing the quantity of residue generated in comparison to dry/semi-dry systems. The major limitation of many wet systems is their inability to remove fine particles effectively unless the energy expended in the system is sufficiently high. This is why wet systems are commonly used downstream of a conventional dust removal device. Corrosion of the materials in the quench stage of the wet system can lead to major maintenance requirements, but is frequently overcome by using fibreglass reinforced plastic (FRP) vessels and ceramic nozzles and cladding. As a single component system, the packed tower is a common wet system encountered in modern incinerator facilities. As its name implies, the packed tower is a vessel with a bed of granular or fibrous material which traps the particles. The liquid used in the system washes the material off the packing and prevents re-entrainment. Coarsely packed granular material will remove particles that are larger than 10 pm and the velocity through the bed should be below 2 m/s. Finely packed beds can remove particles in the 1 to 5 pm size range but the velocities through the bed must be kept below 0.25 m/so Finely packed beds are prone to plugging with particles. These limitations tend to result in packed towers being applied to neutralise the gases after primary particle removal systems. Examples of liquids introduced into these systems include water, lime slurry (calcium hydroxide) and caustic (sodium hydroxide). Hydrogen chloride and the other halogens show a high affinity for water, readily combining to form hydrochloric acid and similar substances. The introduction of lime slurry results in the removal of both HCI and SO2. Caustic easily combines with SO2 resulting in high removal efficiencies but is less efficient for HCI. The alkali addition leads to the formation of an aqueous salt/slurry, which increases waste volumes and can present some difficulties with particulate matter emission rates. Very fine particles of salt in the gas stream cannot be removed by the conventional inertia separators used to control moisture release from wet scrubbers. Downstream removal of these fine particulates requires sophisticated control devices. Introducing liquids into the gas stream reduces the temperature of the gas stream and evaporates substantial quantities of liquid. Thus, the stack gases have high levels of moisture. This situation can cause rapid cooling of the plume resulting in loss of buoyancy and local impingement which can lead to reduced visibility, icing, or elevated odour levels. Plumes with high moisture levels are often interpreted by the public as polluting the environment. To overcome these limitations, wet systems can include a

111

condensation stage to remove moisture and a reheat stage to heat the gases before they exit the stack. Wet scrubber systems can consume substantial amounts of scrubbing solution. The removal of the various contaminants from the gas stream converts an air emission concern into a potential wastewater treatment concern. This can be significant because most jurisdictions do not permit discharge to sanitary sewers without pre-treatment. Some discussion of pre-treatment systems is included in the example systems presented later in this chapter. Alternatives do exist that limit the need to treat the liquid effluent. In some cases the effluent from the wet scrubbers has been re-injected into the gas stream to promote cooling prior to the primary particulate control devices. These units, electrostatic precipitators in most cases, remove the particles formed during the evaporation process and they are collected in the normal manner. To prevent damage to the particle control systems, the gas temperature entering the evaporation zone is normally higher than it would be for normal energy recovery, thus there is some potential loss of energy efficiency in the system. Dry Systems In the context of this report, dry systems will be defined as those that release a gaseous effluent that is not saturated with water. This reduces the concerns about corrosion, visible emissions and plume sag associated with wet scrubbers. The operation of dry systems may involve humidification of the flue gases in an evaporative cooling section; however, the gas stream does not become saturated. Several types of dry systems exist: 9pure dry sorbent addition into the furnace, duct or reactor; 9semi-dry systems where the sorbent is: -injected in dual nozzle atomisers that mix water with dry sorbent; or, -injected as a slurry; and, 9combination systems utilising a gas conditioning system with a wet spray humidifier followed by dry sorbent injection in a venturi reactor. Furnace sorbent injection has only limited application in the MSW field although it is considered commercially proven given significant application on coal fired utility boilers. Typically, limestone or calcium hydroxide are injected into the high temperature zone of the furnace where the temperature causes the material to calcine to produce lime (CaO). This lime then combines with HCI, HF or SO2, to produce various salts. The particulate salt is removed in the particulate matter control device. Furnace injection systems have been shown to be effective for SO2 removal but not for HCI and HF removal (Zmuda and Smith, 1991). Removing these latter compounds with lime requires lower operating temperatures. With its low capital cost and ease of retrofit this method could be appropriate for upgrading facilities. Its drawbacks include low removal efficiency and significant increases in particulate matter loading to the control devices.

112 Several North American facilities, (Claremont, NH and Davis County, Utah) inject sorbent directly into the gas ducts after the furnace. These are called Duct Sorbent Injection [DSI] systems. At Claremont, lime is the sorbent of choice but the removal efficiencies are less than 70% for both HCI and SO2. Recent tests at the Davis County facility have included the use of water injection to cool the gas stream prior to sodium sesquicarbonate injection. This sorbent is readily available in Utah. The lower temperature results in higher removal efficiencies and lower reagent consumption. Ease of installation and lower capital cost are the main advantages of DSI systems; high sorbent use and high waste volumes are disadvantages. The choice of sorbent is a local decision based upon costs for supply and disposal of waste materials, and the removal efficiency of the particular material. Both sodium bicarbonate and sodium sesquicarbonate provide a higher removal efficiency at lower equivalent injection rates than lime, but they can be more costly. The semi-dry injection systems are commonly used in the U.S. and Europe and on hazardous waste facilities. The addition of the lime sorbent in a slurry combines the effects of moisture addition and sorbent addition and tends to result in slightly smaller installations than the combined systems discussed in the next paragraph. The chemical reactions are similar to those discussed in previous paragraphs. The third type of system has been installed in numerous North American and European plants and could be considered "purpose-built" duct sorbent injection system. The gas is cooled in an evaporative cooling tower by the injection of water and the cooled gas then passes through a venturi reactor vessel where lime is added to the gas stream in the highly turbulent zone created by the venturi. This type of system, along with the slurry injection system, produces higher removal efficiencies at lower sorbent injection rates than do the standard duct or furnace injection systems. Regardless of the system chosen, there will be dry powder in the gas stream. This dry powder combines readily with the HCI in the stream to form CaCI2 which can be trapped in the particle removal system. As mentioned earlier, fabric filter systems provide an additional benefit over ESPs when considering acid gas removal. The cake formed on the fabric filter contains unreacted lime which can combine with the residual HCI to increase the neutralisation efficiency. The fabric filter cake acts as a polishing step for HCI removal and allows an average lime addition rate to be used to maintain control during periods of higher than normal flue gas HCI concentration.

Metals Control in Dry Systems

Volatile metals and organics demonstrate a high affinity for surface sorption, thus introducing a finely divided acid gas reagent into the flue gases provides an additional benefit for the removal of trace metallic and trace organic species by providing a large surface area in the form of fine particles in the APC system. This is illustrated by reduced emissions of volatile species from modern APC systems, however, some

113 limitations exist. ESP installations are less efficient in collecting fine particles and have relatively higher emission rates than in comparable fabric filter installations. ESP systems also suffer from changes in particle characteristics. If the particle resistivity changes, the removal efficiency can be influenced. Furthermore, the adsorption effects are temperature dependent and regardless of the particulate matter control device, higher temperatures reduce the capture efficiency. Lastly, in some dry injection systems, the recycling of the fabric filter dust to the venturi reactor to increase SO2 removal efficiencies and lower reagent costs can result in an increase in the concentration of salts and trace species in the residue stream. The advantage of recycling the residue streams is that the partially expended sorbent is more capable of removing SO2 from the gases than the raw hydrated lime. Recent studies suggest that this might be because the SO2 removal process requires more moisture to be effective and the partially expended sorbent contains CaCI2 which has a high affinity for moisture.

Mercury Control with Activated Carbon Regardless of particulate matter control device or operating temperatures, mercury removal efficiencies are fairly low in most systems. This metal has a high vapour pressure and mercury compounds can remain in the gas phase at temperatures less than 180~ Indeed, NITEP testing at Quebec City in 1986 addressed the influence of temperature on the removal of mercury in the wet spray humidifier, dry lime injection, fabric filter APC system (NITEP, 1986). These data were instrumental in the CCME committee's recommendation for APC inlet temperature found in the MSW guidelines (CCME, 1989). Lowering the temperature was effective, introducing at least a 40% reduction in mercury emissions but further reductions were deemed necessary. In Europe, wet systems have added TMT reagent to trap the mercury and remove it from the gas stream. To meet the emissions limits in its permit, the Burnaby B.C. facility used sodium sulphide injection into the APC for a period of time. This was replaced by activated carbon injection to reduce the risks to plant workers who were handling the sorbent (Guest and Knizek, 1991). Activated carbon has been used for mercury removal in a number of European facilities and was introduced into the APC system of the new Amsterdam plant to reduce PCDD/F emissions. One European manufacturer has developed a fixed activated carbon bed scrubber to be installed at the end of the APC system to polish the gases and remove trace organics and mercury. The activated carbon in this system must be periodically removed and disposed either by incineration or by another method. Activated carbon appears to be a very effective mercury sorbent and improve removal efficiencies to the high 90% range. An additional benefit reported in some test data is an increase in PCDD/PCDF removal rates when activated carbon is used. During development of the new regulations in the US, the EPA conducted tests on numerous control techniques to establish their performance and efficacy. Two

114 activated carbon injection tests were conducted at the Stanislaus County facility in California and at the Camden County facility in New Jersey. Initial data from Stanislaus showed that 80-percent mercury reduction was achievable and subsequent testing produced higher Hg reductions with increased the carbon feed levels. This analysis concluded that, at a carbon injection rate of approximately 100 mg/dscm of flue gas (0.33 kg carbon/Mg MSW), the proposed limit of 57 mg/Rm 3 or 85 percent reduction would be achieved (US EPA, 1995BID). Using these and other data, Rigo (1993) developed an equation to relate carbon charge rate to mercury removal efficiency. This suggested that approximately 250 mg of carbon/DSCM of flue gas should achieve a five test average removal efficiency of in excess of 90% with outlet concentrations of 57 pg/Rm 3 given typical inlet data. Burnaby, which has a recirculation system with a fabric filter, consumes between 13,000 and 17,500 kg/year of carbon (VanWaters & Rogers, Pers. Comm.). This translates to a usage rate of approximately 2.4 kg/h, or 0.08 kg/t of waste and agrees well with the feed rates, 2 kg/hr, used during testing of the Burnaby facility in 1993. At that time the outlet mercury concentration was measured at 16-22 iJg/Rm3, or in excess of 95% removal efficiency (Guest, 1993). These data were collected in February, when the amount of yard and garden waste at the Burnaby facility would be anticipated to be low. Thus, consistent performance at these values should not be anticipated; the mercury emission levels will depend upon the amount of mercury in the waste. Besides the EPA tests, five U.S. MSW incinerators that began using activated carbon injection technology after 1994 (Union County, Lee County, Onondaga County, Falls/Bucks County, and Hennepin County) are meeting the proposed US limits for mercury (US EPA, 1995BID). The limitations of powdered activated carbon relate mainly to typical chemical reaction parameters: contact time; temperature of the flue gases; dosing rates; and, the type of carbon (Heath, 1995). The process loses efficiency if the stack gases are not cooled below 200~ (NORIT, Pers. Comm.). This reduction has not been extensively charted but data recently collected at the ESP equipped facility in Davis County, Utah suggests that the removal efficiency difference between 300~ and 150~ is about a factor of 2. That is, less than a 50% reduction was seen when 30 kg/hr of PAC was added to the high temperature gases but when the temperature was reduced the removal efficiency increased to over 90% (ASME, 1996). Residence time in the system is also important and fabric filter systems have a distinct advantage because PAC will be retained in the cake and improve the utilisation of PAC thereby maximising removal efficiency at any given addition rate. PCDD/F removal also occurs when activated carbon is introduced into the APC system. Heath (1995) reports removal efficiencies of 77-80% at an ESP equipped facility. Licata (1994) reports on the Wurzburg tests and notes that PCDD/F emissions were reduced 200 fold. Preliminary evaluation of the Davis County tests suggests that, at low temperatures, in excess of 80% of the PCDD/F is removed from the gas stream.

115 The US EPA (1995BID) concluded from all the tests it evaluated, 12 units at 5 new plants that are equipped with SD/FF/SNCR/CI controls, that an additional 50-percent or greater reduction of PCDD/F emissions can be achieved with carbon injection. The characteristics of the APC residues generated by the use of activated carbon are also of concern. The removal of additional mercury from the gas stream will raise the concentration of mercury in the APC residues. Limited testing has shown that the mercury removed on the activated carbon is fixed to this medium and less environmentally available, at least when tested with certain regulatory leach test procedures, than the material trapped in lime only injection systems. Lanier (pers. comm.) has reported that in limited tests of PAC injection at a hazardous waste incineration facility, the increase PCDD/F's in the APC residue was greater than the amount of PCDD/F removed from the gas stream. This suggests that the addition of PAC may increase the production of PCDD/F's in the APC systems. This is similar to performance noted by Sierhuis and Born (1994) when they examined test data from the new Amsterdam facility. The residues collected from the dry systems contain all the contaminants normally measured in stack gases. These systems have removal rates for most compounds in excess of 99%. The residue is finely divided and has the appearance of lime powder, unless high levels of particulate carry-over from the boiler occurs, causing the colour to change to buff or grey. The residue has a high percentage of CaCI2 and thus is hygroscopic in nature. Care must be taken to maintain the temperature of the material above the dew point to minimise condensation and the formation of solid blocks of material in the storage hoppers and transfer lines. All transfer lines should be sealed and insulated to minimise problems with moisture in the residue.

NO, Removal

As noted earlier, many of the combustion control techniques used to minimise trace organic emissions can lead to increases in emissions of oxides of nitrogen (NOx). Conventional lime-based scrubber systems will not control emissions of NOx; however, other techniques are successful. These techniques include reburning that occurs in the combustion chamber; and post-combustion technologies including selective catalytic reduction (SCR) processes and selective non-catalytic reduction (SNCR) processes; or wet chemical techniques. Reburning involves a modification of the combustion process in the furnace to reduce NO• generation. The SNCR and SCR control processes are based upon the reduction of NOx through the addition of ammonia either at high, furnace temperatures, (SNCR); or at lower temperatures where a catalyst is used, (SCR). The wet chemical techniques involve ozone and alkali absorption, complex absorption or reactions with sulphite solutions and organic compounds. The latter methods produce products that must be disposed of in an additional process step. Utilisation of the by-products as fertiliser is prohibited due to acid salts and ammonia salts which

116 are acidic in nature. The SCR and SNCR processes result in the generation of N2 and only trace amounts of reaction products, and thus are preferred methods. Care has to be taken, however, to prevent the production of N20. Experimental rebuming of the flue gas with natural gas has proven to be successful in reducing NOx levels by 65% (Seeker et al., 1991 ). The process has not progressed to commercial application. An 85% reduction in NOx can be achieved when re-burn was carried out with NH3 injection into the furnace. The latter is a form of SNCR. The rebuming process has been applied successfully to commercial utility boilers firing coal in the United States, Japan and the Ukraine as well as natural gas, oil and MSW combustion processes (Karil et al., 1992). The Japanese appear to be the most advanced in the application of this process. Karil et al. report on a recent MSW demonstration of this concept in Malm(~, Sweden. During the Malm(~ project the furnace was divided into three zones. In the first the main fuel, MSW, is burned under fuel lean conditions and provides 80 to 85% of the released heat. The NOx formed in this zone is due to fixation of atmospheric nitrogen or oxidation of fuel nitrogen. In the second zone, natural gas is introduced into the furnace to produce a slightly fuel rich reburning zone where NOx formed in the first zone is reduced. This fuel supplies 15 to 20% of the energy input to the system. The third zone of the furnace received additional air to complete the combustion process by oxidising CO and fuel fragments and producing an overall fuel-lean condition. Some difficulties were experienced during the testing with the production of higher CO levels. The authors conclude that both the lower furnace and reburning zones will require careful optimisation to achieve the ultimate goal of low CO and low NOx emissions. Overall they predict a 50% reduction in NOx is possible. Thermally based SNCR technologies include Exxon's DeNOx process. This process utilises NH3, injected with a carrier gas, either air or steam, at a point specifically selected to provide optimum reaction temperature and residence time. The injection is accomplished with specially designed nozzles strategically placed in the walls of the furnace to achieve adequate mixing (Hurst and White, 1986). SNCR technologies operate most effectively at flue gas temperatures of 850~ and furnaces that do not have temperatures in this ranges are unsuitable for this application (Hartenstein & Licata, 1996). McDonald et al., (1991) provide a discussion of three applications of the Thermal DeNOx technology at California plants. Several issues are raised, including the need for: stable conditions in the furnace; good mixing; and the right temperature, to ensure good performance. Difficulties experienced at three plants included the generation of false particulate matter in the stack testing samples. The presence of ammonia and water lead to the formation of an alkaline solution in the sampler that removed residual SO2, HCI and NO2 and caused the formation of ammonia salts. Furthermore, any excess ammonia combines with the HCI present in the stack gases to form NH4CI which condenses in the atmosphere and results in the generation of a visible plume. Because

117 there is a possibility of forming ammonia salts in the process, trace increases in these compounds could be found in the APC residues from plants equipped with this technology. Lastly, it was evident that the injection location needed to be carefully selected to ensure that the ammonia was not burned in the furnace, thus increasing the emissions of NOx. Thermal DeNOx is estimated to be capable of achieving reductions in NOx levels of 50 to 85%. Other SNCR technologies which have been applied or proposed include: urea injection (NOxOut developed by EPRI and licensed to NALCO FUELTECH); cyanuric acid (RAPENOx) and ammonium sulphate. NOxOut has been applied on a demonstration basis in the MSW incinerator in Frankfurt (Hofmann et al., 1990). With selective catalytic reduction (SCR) technology, a catalyst is used to increase the low temperature reaction of ammonia and oxides of nitrogen to create nitrogen and water vapour. Efficiencies of up to 80% have been recorded in thermal power plants although operating experience on MSW incinerator applications suggests it is lower. However, on the Munich plant Fl~kt have recorded emission values well within the German standard (184 mg/Rm 3 @ 11% 02). Difficulties with catalyst-based systems include: 9sintering where the microsurface disappears as the catalyst's pores collapse due to elevated temperature reformation of titanium; 9poisoning when a molecule or atom of an alkali metal permanently attaches to an active site; 9plugging by capillary condensation or dust blockage; or, 9erosion due to HCI attack. The latter situation can be overcome by placing the systems after the HCI removal device but this requires reheating the gases to the catalyst's optimum operating temperature of range of 300~ to 410~ Operating at temperatures above 350~ is not recommended because of limitations induced by catalyst brittleness and ash softening. Outside the temperature range the efficiency of the conversion process suffers. To overcome removal efficiency limitations more catalyst can be used, however cost and installation space pose problems. These have partially been overcome by recent developments in Europe where a new generation of Low Temperature SCR [LTSCR] systems are being employed. These systems have increased vanadium pentoxide concentrations in the catalyst and operate with more effective conversion efficiencies at temperatures that range from 150 to 180~ (Hartenstein & Licata, 1996). They are not without their limitations the most significant of which is the control of sulphur species in the gases introduced in the LTSCR. There are currently 15 operating APC trains equipped with LTSCR technology in Europe. The presence of ammonia in the gas stream results in the generation of ammonia bisulphate. This material limits the potential for reuse of the APC residues. The extent of this problem is illustrated by increases in the ammonia content of the APC residue

118 of 10 to 20 mg/kg with only a 5 ppm of ammonia being present in the gas stream. This level of ammonia, known as the 'Ammonia Slip' is not uncommon because precisely matching the requirements with the addition of the reagent is difficult. Furthermore, the ammonia slip will increase as the performance of the catalyst decreases. As a guide, Fl,~kt recommends replacement of the catalyst when the catalytic activity drops below 60% (Herrlander, 1990). Catalyst manufacturers will accept used catalyst for recycling purposes.

4.3 TYPICAL APC INSTALLATIONS As noted earlier, it is possible to combine APC processes to enhance emission control. The most common APC systems involve the use of sorbent injection, with or without gas conditioning, followed by some type of particulate matter control device, either an ESP or a fabric filter. It is possible to include SCR type systems after the particulate matter control device and lower the NOx emissions. On the other hand, systems incorporating an initial particulate removal stage followed by various stages of wet scrubbing and condensation and SCR have also been applied. This section will provide a brief overview of several systems as they are installed in MSW incinerator facilities, emphasising the different effects these combinations will have on the resulting residues and wastewater streams.

4.3.1 Hogdalen, Sweden An example of an early wet spray humidifier/dry reactor/fabric filter APC system is the installation at the Hogdalen plant in Stockholm, designed by ABB Fl~kt. A schematic of this plant is shown in Figure 4.5. The system is similar to that installed in four Canadian plants and used in numerous European facilities. While similar, the majority of installations in the United States utilise slurry injection systems for sorbent injection. There are several places in the Hogdalen system where dry residues are collected. In the cooling tower, heavy particles settle into the hopper at the base of the unit. From here the material is transferred to the residue treatment system. A limited amount of residue is also collected at the base of the dry reactor. This material and the residue from the fabric filter are directed toward either the recycle silo or the disposal system. The system at Hogdalen incorporates a residue treatment system, which involves storage of the spent sorbent in a silo until it is mixed on-site with cement and formed into a thick slurry for transport and disposal.

4.3.2 Munich South, Germany Fl,~kt was also involved in the Munich South installation in Germany that started operation in January 1990. A schematic of the system is provided in Figure 4.6. In this case, the conditioning and dry reactor are combined in one tower, a schematic of which is shown in Figure 4.7. Note the addition of sodium sulphide to the reactor for the

II.Dust silo12. Dust humidifier 13. Cement silo 14. Mixer

(adapted from Hogdalen Information published by Stockholm Energi)

Sodium

Figure 4.6 Process Flow Sheet Munich South Plant

I

L--------------------------------------------:

tesy Flakt Review, 1990)

I I

121 Figure 4.7 Schematic of CDAS Reactor

- o,4. 9

filter

".~ ~::

9:..'-;',::

, .:,'-.~ ;. ; ;_;."

.. ~,.:>..:. . . :';;=

.. ;~.. ~.,. .,..~~.; .-

Reaction

additives

Mixing

From

boiler

Cona~tioning

9-

Plu'lddd,~l,~e

Flue g a s

pretreatment

CDAS reactor. The flue gas is first cooled with water to suitable temperature, then mixed with dry sorption agent.

(courtesy Fl~kt Review, 1990)

122 control of mercury, which causes the mercury to condense out in the form of mercuric sulphide. The back end of this system is equipped with an SCR reactor for NOx control. To operate the SCR system, the gas passes through a heat exchanger and burner to raise the gas temperature to the appropriate operating temperature. The NH3 is added and the gas passes through the SCR reactor prior to being discharged to the stack. No dust control is installed after the SCR reactor and ammonia salts will not be present in the plant residues. The fabric filter dust from the APC system is recirculated similar to the Hogdalen application until such time as it transfers to the waste silo. From here the material is wetted only to prevent dusting during transport to the disposal site.

4.3.3 Warren County, New Jersey, USA Typical of many facilities in both the US and Europe, the Warren County facility utilises a spray dry absorber (SDA) for acid gas and trace compound removal, Figure 4.8 (Jorgensen et al., 1991). The system was supplied by Environmental Elements Corporation. Lime slurry is introduced into the SDA where it neutralises the acid gases and serves as nucleating sites for the condensation of trace materials. The particulate matters in the gas stream are then removed in a fabric filter. Typical of many U.S. installations, the APC residues are combined with the bottom ash before being transported to the disposal site.

4.3.4 Zirndorf, Germany The schematic for this plant's APC system, Figure 4.9, illustrates the use of a wet scrubber system in a retrofit situation (Beckert and Jungmann, 1992). The retrofit was performed by ABB W+E Umwelttechnik in combination with ABB Fl~kt-Umwelttechnik. The existing ESP and HCI scrubber were upgraded with the addition of a wet SO2 scrubber and an absorption stage. The residue from the absorption stage is returned to the incinerator where any organics are destroyed and residual mercury is liberated to be trapped in the organic sulphides in the HCI wet scrubber. A heat exchanger removes heat from the gases before they pass to the HCI scrubber and transfer the heat to the gas leaving the scrubber system. The HCI scrubber uses water as the scrubbing media in a contact tower. A bleed from this system maintains the pH of the solution at an appropriate level. Organic sulphides are added to the scrubber liquid to improve the separation of trace metals, particularly mercury. The sump of the scrubber is the eventual sink for all trace metals in the system with the bleed from this circuit combined with that from the SO2 scrubbing circuit. In the wastewater treatment system the pH is adjusted with a lime slurry and forms gypsum and organic sulphides which precipitate and are filtered from the effluent.

Figure 4.8 Process Flow Sheet -Warren County

I Boiler Ash

Mixer

Waste Product Waste to Disposal

The absorption stage is returned to the incinerator where any organics are destroyed and residual mercury is liberated to be trapped in the organic sulphides in the HCI wet scrubber

Jorgensen et al., 1991

A

h,

W

Figure 4.9 Process Flow Sheet - Zirndorf, Germany

Beckert and Jungrnann, 1992

125 The SO2 scrubber is a packed tower system. Sodium hydroxide in softened water is used as the scrubbing liquid. A bleed from this system maintains the salt content of the scrubber liquid at an optimal level. The gases are reheated before passing to a compact reactor where fresh lime and powdered activated carbon are added to the gas stream to remove trace organics and residual trace metals such as mercury. The sorbents are then captured in a fabric filter where the filter cake serves to polish the acid gas removal process and remove trace organics and trace metals. The filter cake is recirculated in the system and then eventually disposed by introducing it into the high temperature zone of the furnace where trace organics are destroyed. The only residue from this system is the filter cake from the wastewater treatment plant.

4.3.5 Vestforbra~nding, Copenhagen The APC system at this plant represents one of the latest wet scrubber concepts, Figure 4.10. The system was designed by G5taverken of Sweden. The existing hot side ESP has been maintained and is followed by an economiser, gas/gas heat exchanger, quench system, and HCI and SO2 scrubbers. The latter scrubber is designed to operate at 50~ to generate hot water and recapture energy while condensing much of the moisture out of the stack gases. The gases are reheated before exiting to the stack. Wastewater treatment in this system includes limestone coarse pH adjustment, lime slurry fine pH adjustment along with organic sulphide to bind trace metals, and polymer addition to aid settling and separation of the suspended particulate matter. The sludge from the system is mixed with the ESP residue and landfilled.

4.3.6 Lausanne, Switzerland The LAB system installed at Lausanne is one of a range of systems manufactured by this company. The systems are applied with different levels of equipment to achieve the emissions limits required. Figure 4.11 shows a typical installation of the EDV 7000 variant. In this case the scrubbers are open vessels without packing. Proprietary nozzles are used to generate a high density water curtain in the vessel that neutralises the gases. The gases then pass to electrofiltering modules where the particles are charged. Particles are removed from the gas stream by a water spray. The system is completed by a water droplet removal device based upon centrifugal force. The wastewater treatment system is similar to those seen in other plants outlined above, and produces a sludge requiring disposal.

Figure 4.10 Process Flow Sheet - Vestforbranding, Copenhagen

(courtesy Gotaverken Miljo AB)

127 Figure 4.11 Process Flow Sheet - Typical LAB EDV 7000

(courtesy LAB S.A.)

128

4.3.7 Bremerhaven, Germany Detailed flow sheets for two German installations, Figures 4.12 and 4.13, show variations in APC strategy (Lange, 1992). Figure 4.12 illustrates the system used at Bremerhaven. The use of SNCR NOx control through the injection of ammonia into the furnace is similar to that shown in other facilities. The unique part of this installation is the steam stripping of the scrubbing solution from the wet scrubber to recover ammonia which is then reused in the NOx control system. The quantities illustrate the low ammonia level in the flue gas. An ESP is used to remove the particulate matter from the flue gas leaving the furnace and clean it before it enters the scrubber. A wastewater treatment facility is noted, but no details of the system are available.

4.3.8 Stuttgart, Germany This facility utilises a spray dryer/absorber followed by an ESP to treat the scrubber solution from the APC system, Figure 4.13, after Lange (1992). The SCR NOx control system on the plant is similar to the installation shown for Zirndorf.

Figure 4.12 Process Flow Sheet - Bremerhaven, Germany

temperature

:

30

OC

to waste water

Flow sheet of the waste gas purification plant with the SNR process (municipal waste incineration plant Bremerhaven) L

(Lange, 1992)

Figure 4.13 Process Flow Sheet - Stuttgart, Germany

heat exchanger

blower

waste gas heating

-

SCR reaktor

. Flow sheet of the waste gas purification plant with the SCR process I municipal waste incineration plant in Stuttgart )

(Lange, 1992)

131 REFERENCES

ASME. Retrofit of Waste-to-Energy Facilities Equipped with Electrostatic Precipitators. An ASME Research Report prepared by H.G. Rigo and A.J. Chandler under the direction of the ASME Research Committee on Industrial and Municipal Waste. CRTD Vol. 39. 1996. AWMA. Air Pollution EnQineerin.q Manual/Air & Waste Mana.qement Association. Edited by A.J. Buonicore and W. Davis. Van Nostrand Reinhold, New York, NY., 1992. Beckert, P. and G. Jungmann. "Retrofit of the Flue Gas Cleaning System of the Waste to Energy Plant Zirndorf". A paper presented at the 7th IRC conference Berlin, Nov. 1992. In Wa.ste Management International, Karl J. Thome-Kozmiensky (ed.), Berlin: EF-VerI. fur Energie- und Umwelttechnik, ISBN 3-924511-64-0, 1992. Environment Canada. "The National Incinerator Testing & Evaluation Program (NITEP) Air Pollution Control Technology". Environment Canada Report EPS 3/UP/2, September 1986. Fl~kt. "Cleaning Flue Gases in Energy from Waste Plants". A sales document from Fh~kt, Sweden, 1991. Guest, T.L. and O. Knizek. "Mercury Control at Burnaby's Municipal Waste Incinerator". Proceedings of the 84th..Annual AWMA Meetin& Vancouver, B.C. Paper 91-103.30, June, 1991. Guest, T.L., 1993. Mercury Control in Canada. Proceedings of the 86th Annual AWMA Meeting. Paper 93-WP-109. 01. Denver, Colorado. June. Hartenstein, Hans-Ulrich and Anthony Licata. "The Application of a Low Temperature Selective Catalytic Reduction System for Municipal & Hazardous Waste Combustors". Proceedings of the 17th Biennial Waste Processing Conference. ASME. 1996. Heath, Patrick B., 1995. Design and Installation of Powdered Activated Carbon Storage and Injection Systems for Municipal Solid Waste Incinerators. Proceedings of the 88th Annual AWMA Meeting. Paper 95-RP-147B.01. San Antonio, Texas. June. Herrlander, B. "Recent developments in de-NOx technology". An article in Fl~kt Review No. 73, Stockholm, Sweden, February 1990. Hofmann J.E. et al. "NOx Control for Municipal Waste Combustors". A&WMA Annual Meeting, Pittsburgh, PA, 1990.

132 Hurst, B.E. and C.M. White. 'q'hermal DeNOx: A Commercial Selective Non-Catalytic NOx Reduction Process for Waste to Energy Applications". A paper presented at The ASME 12'h Biennial National Waste Processing Conference, Denver, Colorado, June, 1986. Jorgensen, C. et al. "Two and a Half Years Operating Experience at the Warren County Energy Resource Recovery Facility". A paper presented at the 2nd International Conference on Municipal Waste Combustion, Tampa, Florida. Published by A&WMA Association VIP-19 Municipal Waste Combustion, Pittsburgh, PA, April, 1991. Lange, M. "Legal Regulations and Technical Measures Limit the Emissions of Waste Incinerators in the Federal Republic of Germany". A paper in A Selection of Recent Publications (Volume 3) prepared by Umweltbundesamt, Berlin, Germany, 1992. Licata, A., M. Babu, and L-P Nethe, 1994. An Economic Alternative to Controlling Acid Gases Mercury, and Dioxin form MWC's. Proceedings of the 87th Annual AWMA Meeting. Paper94-MP-17.06. Cincinnati, Ohio. June. LIRPB. Lon.q Island ReQional Pla.nnin.q Board, The Potential for Beneficial Use of W_aste-to-Ener.c]v Facility Ash. Seven volume report. Hauppauge, NY, 1992. McDonald B.L., G.R. Fields and M.D. McDaniel. "Selective Non-Catalytic Reduction (SNCR) Performance on Three California Waste-to-Energy Facilities". A paper presented at the 2nd International Conference on Municipal Waste Combustion, Tampa, Florida. Published by A&WMA Association VIP-19 Municipal Waste Combustion, Pittsburgh, PA, April 1991. Miller, J.A. and G.A. Fisk. "Combustion Chemistry". C&ENAug. 31, 1987, pp. 2248. Rigo, H.G., 1993. How Good are Today's Mercury Test Methods and Controls? A paper presented at the Ash VI Conference, Arlington, Va. November. Seeker, W.R., G.C. England and R. Lyon. "Advanced Pollution Control in Municipal Waste Combustors Using Natural Gas". A paper presented at the 2nd International Conference on Municipal Waste Combustion, Tampa, Florida. Published by A&WMA Association VIP-19 Municipal Waste Combustion, Pittsburgh, PA, April 1991. Seeker, W.R., W.S. Lanier and M.P. Heap. "Municipal Waste Combustor Study: Combustion Control of Organic Emissions". A Report prepared for the U.S. EPA by Energy and Environmental Research Corporation EPA/530-SW-87-021C. Irvine, CA, 1987.

133 Sierhuis W.M. and J.G.P. Born. November, 1994. PCDD/F Emissions Related to the Operating Conditions of the Flue Gas Cleaning System of MWI- Amsterdam. A paper from Dioxin '94 Kyoto, Japan, published in Organohalogen Compounds Volume 19. U.S. EPA. "Air Pollution Engineering Manual/U.S. Environmental Protection Agency" AP-40. Prepared by the Los Angeles Air Pollution Control District, 1973. U.S. Environmental Protection Agency, 1995BID. Municipal Waste Combustion: Background Information Document for Promulgated Standards and Guidelines -- Public Comments and Responses. Emission Standards Division, U.S. Environmental Protection Agency Office of Air and Radiation, Office of Air Quality Planning and Standards Research Triangle Park, North Carolina 27711. October. Vogg H., A. Merz, L. Stieglitz, F. Albert and G. Blattner. "Zur Rolle des Elektrofilters bei der Dioxin-Bildung in Abfallverbrennungsanlagen", Abfall.wirtsch.aft Journal, 2, p.529, 1990. Zmuda, J.T. and P.V. Smith. "Retrofit Acid Gas Emission Controls for Municipal Waste Incineration: An Application of Dry Sorbent Injection". A paper presented at the 2nd International Conference on Municipal Waste Combustion, Tampa, Florida. Published by A&WMA Association VIP-19 Municipal Waste Combustion, Pittsburgh, PA, April 1991.

This Page Intentionally Left Blank

135

C H A P T E R 5 - REGULATION OF MSW INCINERATORS

During incineration, most organic-based materials are destroyed by complete oxidation to carbon dioxide and water vapour but inorganics remain substantially untouched. Inorganics either partition within the various residue streams or are entrained into the flue gas stream. As noted in the previous chapter, considerable work has been undertaken in the past 10 to 15 years to reduce the level of emissions to the atmosphere through the stack. This has resulted in the application of new APC technologies incorporating both new operating philosophies and new equipment. These changes have influenced the characteristics of the residue streams. For instance, the use of lime to remove acid gases has increased the mass of the fine residue generated in the systems which, in turn, has increased the alkalinity of this material, thereby influencing the potential solubility of trace metals in the disposal site. APC technologies based on wet scrubbers generate less solid residues, but they generate a wastewater stream requiring treatment. To address the regulatory concerns, various jurisdictions have developed specific MSW incineration regulations governing air emissions, residue disposal, and wastewater treatment. This chapter reviews these regulations. Regulations represent an evolving set of limits on the operation of MSW incinerators. Newer limits are more stringent and comprehensive in response to requirements that are aimed at controlling an increasing number of substances (contaminants). Each jurisdiction has adopted standards they consider appropriate for their circumstances. Typically, both the toxicity and persistence of contaminants in the environment are considered during the evaluation process. Examples of the contaminants that could be considered are shown in Table 5.1. This list was derived from the Canadian MSW incinerator operating guidelines (CCME, 1989). The contaminants listed cover the broad range from combustion related gas emissions to trace metallic and organic compounds. As noted in an earlier chapter, changes to the operational characteristics of a facility can influence atmospheric emissions and consequently influence the characteristics of the residues. Regulations governing the design and operation of MSW incineration facilities to minimise air emissions address both: ~ operational factors such as combustion conditions; and, ~ the quantity of materials released through the stack. Emission standards exist in many jurisdictions but they are not presented on a consistent basis. Both temperature and diluent concentration can vary between jurisdictions. The values in this report have been corrected to 11% 02 and reference conditions of dry gas at 25~ and 101.3 kPa pressure. (Note: The comparative diluent and temperature correction used in the United States is 7% oxygen and 68~ Thus, while the Ontario and US emission concentrations are equivalent, when comparing the two standards with their normal diluent basis, the Ontario values would appear to be

136 T a b l e 5.1 List of Air E m i s s i o n C o n t a m i n a n t s Acid Gases 9 9

Hydrogenchloride Oxides of nitrogen

9 9

..

Hydrogenfluoride Oxides of sulphur

Trace Metals *Cd - C a d m i u m Fe - Iron *Be - Beryllium *Pb - L e a d Mo - Molybdenum *Cr - C h r o m i u m Ca - C a l c i u m *Ni - Nickel

*V - V a n a d i u m Si - Silicon AI - A l u m i n u m Ti - T i t a n i u m Mg - M a g n e s i u m B - Boron Ba - Barium P - Phosphorous

K *Hg Na *As *Zn *Sb Mn Bi

- Potassium - Mercury - Sodium - Arsenic - Zinc - Antimony - Manganese - Bismuth

Co *Se *Cu Te Ag Sn

- Cobalt - Selenium - Copper - Tellurium - Silver - Tin

*metals selected, bY the committee as most important for health and the environment. Organics Polychlorinated dibenzo-p-dioxin (PCDD) homologues TCDD PeCDF HxCDF HpCDD OCDD

tetra penta hexa hepta octa

Polychlorinated dibenzo-furan (PCDF) homologues TCDF PeCDF HxCDF HpCDF OCDF

tetra penta hexa hepta octa

C h l o r o b e n z e n e s (CB) CI-2 benzenes CI-3 benzenes CI-4 benzenes CI-5 benzenes CI-6 benzene Polychlorinated B i p h e n y l s (PCB) C h l o r o p h e n o l s (CP) CI-2 phenols CI-3 phenols CI-4 phenols CI-5 phenol

Polyaromatic Hydrocarbons (PAH) Acenaphthylene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Chrysene Benzanthracene Benzo(e)pyrene Benzo(a)pyrene Benzo(b,k)fluoranthene Perylene o-Phenylenepyrene Dibenzo(a,h)anthracene Benzo(g,h,i)perylene

Other Particulate matter Opacity After C C M E , 1989

Carbon monoxide Total hydrocarbons

Oxygen or Carbon dioxide

137 about 70% of the U.S. levels.) Wet standard levels are converted to dry, assuming the average moisture level will be 20%. Residue treatment and disposal regulations exist in many jurisdictions. The scope of these can vary from requirements for disposal under controlled conditions to quality standards such as the level of carbon in the material being disposed. Some jurisdictions have developed regulations governing the utilisation of residues. Both these issues are discussed later in this chapter. Wastewater treatment aspects of the disposal of wet APC residues is not discussed because, for the most part, these regulatory requirements do not specifically address wastewater discharges from MSW incineration facilities.

5.1 EXISTING MSW INCINERATOR OPERATING GUIDELINES Operational guidelines have been developed to assist regulators in standardising the design parameters of incinerator facilities. A summary of typical conditions is provided in Table 5.2. The intent of these standards is to ensure good combustion conditions and minimal organic emissions. The design guidelines are based upon theoretical calculations of the combustion process. In some cases, the applicability of these calculations is influenced by new trends in technology. In addition, some regulations stipulate the objectives for incineration. For example, German regulations mandate the recovery of energy from MSW incinerators; no systems are used solely to incinerate waste.

5.1.1 Furnace Temperature and Residence Time The operating temperature concept used in North America was developed from hazardous waste incinerator test data and some now suggest that such standards may be inappropriate for MSW systems. Because the reactions needed to take the combustion process to completion occur very quickly, if the temperature is sufficient and there is enough oxygen, these restraints on temperature and residence time may be artificial and can limit the development of new technologies. Thus, while there is a regulatory move to relax the prescriptive design aspects for MSW incinerator regulations, staff charged with the responsibility of making decisions on applications still find the guidance of these standards helpful. When considering temperature values in the table it should be remembered that some of the apparent differences result from conversion between various temperature units. The value 982~ represents 1800~ a common value for the United States whereas, Canada and the European countries specify temperatures in ~ and choose 1000~ as a suitable number. Both numbers are within the accuracy obtainable from conventional measurement systems.

Table 5.2 Incinerator Operating Conditions Jurisdiction

Temperature ?c)

Residence Time (Seconds)

Minimum Excess Oxygen

CO Level (mglRm3@ 11% 0,) Averaging Times

1 hour

I 4 hours

U.S. 1991

see note California Guidelines

982

1

300 (none specified)

Pennsylvania BAT Criteria

982

1

300 (8 hour)

New Jersey Guidelines

815

1

6

1000

1

6lRDF 3

Peel, Ontario Perml

1000

1

6

B.C. Guideline

1000

1

1000

11

CCME Guidelines

Ontario A7 Guidelines 1995

1

1850

Eurowan Communitv Guidelines Germany

1

850

12

1

I

I

I

1

90% of readings el50 (24 hours)

850

2

6

Denmark

850

2

6

1

I

850

12

France

850

2

7

Belgium

800

1

6

Italy

950

2

Norway

800

1.5

Sweden (<6 tph) (>6 tph) Japanese Guideline 1990

850

850

1 800

see note

I

APC Operating Temperature (DC)

17 mar. test temD.

6IRDF 3

United Kingdom

Netherlands

I 1

24 hours

loo

I

100

620 (10 minute) 825 (1 minute)

loo

2

6 6

12

16

Notes: Under carbon monoxide the US Regulations distinguish between incinerator types. Under 4 hour averaging times: 40 (35 ppmdv) refers to modular units; 80 (70 ppmdv) to mass burn waterwall or refractory wall, and, fluidised bed units; and, 120 (105 ppmdv) to RDF mixed with pulverised coal units. Standards for both new and existing facilities are identical. Under the 24 hour averaging time category: existing units are limited to 80(7O ppmdv) for mass bum rotary refractory units; 160 (140 ppmdv) for RDF spreader stokers with or without coal and 200 (175 ppmdv) for mass bum rotary waterwall systems. For for new plants the 24 hour average for RDF systems except Mass Bum Rotary Waterwall are limited to 120 (105 ppmdv) for CO with the MBWW Rotary units having a limit of 80 (70 ppmdv).

139 Residence time specifications vary from the stringent 1 second residence time at 1000~ specified in Ontario Guideline A-7, to a more flexible approach such as that favored by some European countries. The Ontario guideline provides a definition of the zone where 1000~ could be expected to occur. Europeans require a longer time at lower temperatures to allow some flexibility in design. Germany has gone one step further in the latest documents, (17 BImSchV; November, 1990). That regulation states that if testing data collected at the specified conditions is the same as that collected at lower temperatures or shorter residence times, the facility can be operated at the lower level.

5.1.2 Combustion Efficiency and Carbon Monoxide Carbon monoxide (CO) operating levels are used as a surrogate for good combustion conditions. They also are used to calculate various combustion efficiency factors. In some jurisdictions combustion efficiency may be defined as the ratio of CO to the sum of CO and CO2; in others it may be the ratio of CO to CO2. Because CO can be measured directly, the trend is towards setting a standard for CO levels at the boiler exit. CO standards are contained in Table 4.3. Averaging time variations are evident in the CO standards. The Danish standards are for 10-minutes and 1-minute respectively. The latest German standards specify three values based upon differing averaging times for CO. The 24-hour or daily mean must be 50 mg/m3; with an hourly mean of 100 mg/m3; and, 90% of all readings in the 24 hour period being less than 150 mg/m 3. The German values are not considered emission values but rather used as an operational parameter. Where additional values are shown under the 4-hour category in Table 5.2, they are explained in the footnote. Notable by its absence is the lack of a carbon monoxide standard in the new Ontario Guideline A-7. It should be recognized that under the operating procedures in the province, each new facility will have a specific Certificate of Approval that will specify operating parameters of this nature. These numbers can be tailored for different combustion technologies thereby reducing the complexity of the new guideline.

5.1.3 APC Temperatures Performance of the air pollution control systems has been found to be correlated with temperature at the inlet to the APC (NITEP, 1986). Some jurisdictions are recommending operating temperature restrictions on these systems. The aim is to lower the temperature to increase the trace contaminant and acid gas removal efficiency while maintaining temperatures high enough to minimize corrosion in the system and blinding of the fabric filters. At low temperatures the hygroscopic nature of the reaction products of the sorbent injection systems can lead to formation of 'mud-

140 like' material that coats the bags and turns to a solid non-porous mass. Canada included a maximum inlet operating temperature recommendation in the CCME (1989) document. The 1995 US regulations specify that the facility cannot be operated with an inlet temperature to the last particulate control device in the APC that is any higher than 17~ more than the 4-hour block average during the most recent successful dioxin test period.

5.1.4 Other Aspects There are other design related parameters such as capacity, (throughput) and auxiliary fossil fuel fired burner capacities included in some regulations. During operation, incinerators are subject to upset if the loading rate is too variable or too far from the design point. Limitations on feed rate are included in some standards including the 1995 U.S. standards which limit loading to no more than 10% above that used during the last test series. This is to prevent excessive particulate entrainment and potential trace organic formation downstream of the combustion chamber (US EPA, 1995BID). The German, Dutch and English regulations and others require that the system be equipped with auxiliary burners to enable the furnace to reach operating temperature before MSW is added, and to maintain combustion efficiency in the event of a drop in fuel level in the system. Several jurisdictions have added staff/operator training requirements to regulations. The new US EPA rules require ASME certification of senior staff; British Columbia MOE's proposal requires that all staff be trained to a level acceptable to the Ministry. Owners of MSW incineration facilities attempt to hire staff who have previous experience at similar facilities. Lead operators generally have such experience and others on staff are promoted to positions of increased responsibility if they have appropriate qualifications and sufficient time in the facility.

5.2 AIR EMISSION STANDARDS The operating emissions from a facility are related to both the type of APC system installed and the nature of the MSW received at the facility. Generally, the greater the efficiency of the APC system, the lower the emissions, although the presence of commercial or industrial type waste in the fuel stream may raise trace contaminant levels in the emissions. Regulators set performance standards based upon their desire to minimise emissions; however, these standards vary from jurisdiction to jurisdiction based upon the perceived requirements in that area and their regulatory interpretation of the issues discussed earlier. This section reviews existing regulations for several groups of emissions:

141 9conventional combustion products and acid gases 9trace metals 9trace organics. Regulatory limits can be based upon either the absolute emission number or a removal efficiency determined by the ratio of emissions to input to the APC device. Other variations in the standard setting process include the use of various averaging protocols based either on the time of sampling or the number of samples taken. While the following summarises some of these issues, no attempt has been made to address the differences in sampling times, rather the data are normalised to standard temperature and pressure and a standard basis diluent concentration of 11% 02. 5.2.1 Chronological Changes in Emission Standards

As noted in the introduction, newer standards tend to be more stringent. Table 5.3 illustrates the trends for emission standards of conventional pollutants in those countries where the data are easily traced. Regulations in the United States have had numerous changes since 1989. They are extremely complex, given that they were written to address different technologies and sizes of facilities, but all exhibit the same decreasing trends in emissions. Table 5.3 Chronology of Municipal Solid Waste Incinerator Emissions Limits - Combustion products and Acid Gases (Values expressed as mg/Rm 3 @ 11% 02) ..... Jurisdiction (Country/State/Prov.)

Hydrogen Chloride

France

1982 1986

155

Nethedands

1984

1080

Switzerland

Hydrogen Fluoride

Sulphur Dioxide

Oxides of Nitrogen

120

23

690

290

Particulate Matter

Carbon Monoxide

Hydrocarbons (as CH4)

200

1530

11

78

1900

120

1989

9

1

37

65

5

46

1986

28

4

460

460

46

92

9

1991

18

2

46

74

9

46

18

1986

46

2

90

460

32

92

18

1990 mean 24 hour

9

1

46

184

9

46

9 (as carbononly)

Denmark

1986

83

2

240

33

83

1991

60

2

276

37

92

1994 Proposed

50 10

2 1

240 50

30 10

100 50 (daily)

Germany

Sweden

* = as total carbon for old plants

200

20 * 10 (totalcarbon)

142 Four major standards are currently enforced: the Canadian (CCME, 1989); European Economic Community (EEC, 1989); German (17 BImSchV, 1990); and the U.S. (EPA, 1995). A summary of these standards and those from other jurisdictions is presented in Table 5.4. As noted above, the US EPA standards have evolved since the late 1980's and those in the table represent the latest edition. These regulations are currently being challenged in the U.S. court system and may be subject to revision. The EEC Directive sets out minimum standards for MSW incinerators in all countries of the EEC. All new facilities, as of December 1, 1990, are required to meet the standards, whereas existing large facilities have until December 1996 to comply. Interim standards on facilities smaller than 6 tonnes per hour applied in December 1995, forcing local standards to adjust by 1996. The Germans have added other requirements to the EEC Directive noting that existing installations needed to comply by March 1994, and absolute compliance must occur by December 1996. The rule allows local jurisdictions some discretion in requiring tighter controls where necessary to protect the environment, while still allowing flexibility in operating conditions where it can be demonstrated that the alternative operating conditions do not have a detrimental impact on the quality of air emissions. The newest U.S. regulations were passed into law in late 1995. They apply across the country and set a minimum performance standard in much the same manner as the European Directive. These standards require compliance at existing facilities by 2000, but states have the ability to accelerate compliance by including these standards in state regulations. The current court challenge relates to the distinction between different sizes of facilities included in the standards and the absolute limits that will apply to some of the existing smaller facilities. The Canadian CCME guideline was developed by a joint federal/provincial committee and was meant to act as a basis for provincial regulations for new MSW incinerators in Canada. The new Ontario guideline applies only to new facilities built after December, 1995. The guidelines do not apply to existing facilities in Ontario unless they undergo modification or expansion. Currently these facilities have specific operating permits which contain marginally higher emission standards. As mentioned earlier, these specific permits allow more stringent standards to be set for specific facilities if necessary. As is the situation in Canada, all countries tend to view national standards as minimum operating levels. Local jurisdictions can apply more stringent standards. These are reflected in facility specific operating permits and thus there appear to be a plethora of standards in some countries when, in fact, the operating limits at a particular facility reflects the requirements of the local area.

Table 5.4 Municipal Solid Waste Incinerator Emissions Limits - Combustion Products and Acid Gases (mglRm3@ 11% 0,) Jurisdiction (Country/State/Prov.) European Economic Community 1991 Italy 1991 United Kingdom 1992 (new plants) Belgium 1991 Netherlands 1989 Sweden 1986 Norway 1992 Switzerland 1991 Austria 1989 Germany 1990 mean 24 hour Germany 1990 1R hour max Denmark 1991 mean 24 hour U.S.A. NSPS 1995 New Facilities Existina . ... ... . .Facilities - -...- ..>35 .. tod .r - 8 . . <225 -- . tod ~ r 2225 tpd Canada CCME Guidelines 1989 Alberta 1983 British Columbia 1991 Ontario Guideline A-7 1995

Hydrogen Chloride 46 46 46 46 9 80 110 18 9 9 55 60 27 (95%)

..

-

Hydrogen Fluoride 2 2 2 2 1 1 2 1 1 4 2

261. (50%) --

33 (95%j' 75 (90%) 110 70 27

3

Sulphur Dioxide 276 276 276

37 190 330 46 46 46 183 276 56 (80%) 147'(50%) 58 (95%j 470 250 55

Oxides of Nitrogen

92 65 320 74 92 184 366 197 (daily) exem~t 263-329

350 207

Particulate Matter 28 28 46 (92) 28 5 17 11 9 14 9 55 37 17 49 19 20 20 17

Carbon Monoxide 92 92 92 92 46 80 110 46 46 46 92 92 various

Hydrocarbons (as CH,) 18 18 18 18 9

18 18 9 35 18

5711 14

55 40 permit permit specific specific Japan 1984 778 570 89 69 Note: 1. " where percentage values are provide in bracket following the emission level they refer to a minimum removal efficiency required by the jurisdiction In most cases these conditions are enforced as the lesser of the two conditions either 7 mglm3or 95% removal so that if 95% corresponds to 10 mglm3it would be judged satisfactory performance as would 7 mglm3with only 90% removal. 2. Various refers to levels for different types of systems as outlined in Table 5.2.

144

5.2.2 Emissions of Combustion Products and Acid Gases The conventional by-products of combustion from most fuels include, water vapour, carbon dioxide, sulphur dioxide, oxides of nitrogen, particulate matter, carbon monoxide and hydrocarbons. Since MSW contains chlorinated plastics and other materials which may contain low but measurable concentrations of chlorine, such as wood, plant and vegetable matter and related products (e.g., paper products), hydrogen chloride is another by-product of combustion. Other constituents of MSW can contain fluorides and bromides which give rise to other halogenated gases. Reviewing the standards by compounds, it is evident that most jurisdictions place limits on HCI and dust (particulate matter) from MSW incinerators. Carbon monoxide, sulphur dioxide and oxides of nitrogen standards are the next group of frequently regulated substances with only a few jurisdictions regulating hydrogen fluoride and total hydrocarbons (expressed as an equivalent amount of methane, CH4 or as elemental carbon).

Hydrogen Chloride HCI standards are expressed as either removal efficiency requirements or emission limits. A removal efficiency of 90% is required in most jurisdictions using this approach. However, U.S. regulation require 95% if the concentration exceeds 27 mg/Rm 3 @ 11% 02. Promulgated emission limits differ in both averaging time and concentration. The lowest values are 24-hour averages of 9 mg/Rm 3 in the Netherlands and Germany. The U.S. value of 27 mg/Rm3 represents the average of three, one-hour test results, or a three-hour average. Continuous emission monitoring [CEM] for HCI can be required in Germany and some Canadian jurisdictions. In Ontario those facilities with CEM systems operate with 24 hour average emission limits equal to the standard. Particulate Matter Until the late 1980s, the emission requirements for particulate matter were decreasing to the 20 - 30 mg/Rm 3 @ 11% 02 range, based on the results of standard sampling methods that provide a 2 to 4 hour composite sample. More recently, North American standards have dropped to the lower end of this range, whereas European standards are now half this concentration. The newer standards reflect test data and the development of newer APC systems which are capable of consistently limiting particulate emissions to lower levels. Sulphur Dioxide Sulphur dioxide (SO2) standards range widely from 300 mg/Rm 3 @ 11% 02 to the newer European standards of 37 to 46 mg/Rm3. The lower range reflects the increased application of more complex APC systems. The U.S. national standard for new facilities contains a removal standard of 80% if emissions exceed 55 mg/Rm 3 @ 11% 02, but

145 some states have set higher performance targets. For example, the Commerce facility in California is licensed for 30 mg/Rm 3 @ 11% 02 based on a 24-hour average.

Oxides of Nitrogen Most MSW incinerators generate NOx emissions in the 300 to 450 mg/Rm 3 range, thus emission controls are required if the facility is to operate at regulated levels ranging from 65 to 190 mg/Rm3 @ 11% 02. NOx control systems are capable of meeting the 70 mg/Rm 3 @ 11% 02 level required by some European regulations. To avoid problems inherent with operations at this level, namely "ammonia slip" and visible plumes, increased efforts are being placed on monitoring residual ammonia levels after the NO,( control section and adjusting ammonia flow to reduce the slip. Carbon Monoxide Carbon monoxide levels were discussed in the previous section. Generally, well operated modular facilities will be operated at 4-hour average levels lower than 40 mg/Rm 3, whereas some incinerator variants may record 200 mg/Rm 3 values due to quenching of smouldering char or limitations in air distribution. Different jurisdictions use different averaging times for CO limits, but all appear to produce fairly consistent standards in the 50 to 100 mg/Rm 3 range. Total Hydrocarbons The total hydrocarbon [THC] standard is also related to combustion performance since higher CO values are normally accompanied by higher THC values. The standards range from 10 to 40 mg/Rm 3. Furnace operations must be adjusted to minimise this indicator since few back end control strategies effectively reduce these emissions. Auxiliary burners have resulted in higher THC emissions at some facilities since they do not perform as efficiently as the furnace does when burning waste. Hydrogen Fluoride In recent years, more jurisdictions have developed specific hydrogen fluoride (HF) standards for incinerators. Earlier, non-specific regulations, such as the U.S. EPA Prevention of Significant Deterioration (PSD) standards and the U.S. National Emissions Standards for Hazardous Air Pollutants (NESHAPS) provisions, limited general HF releases. Incinerators equipped to control acid gases generally meet these absolute standards. Of those jurisdictions with specific standards, the majority are in the range of 1 to 4 mg/Rm 3 @ 11% 02 and the average is 2 mg/Rm 3 @ 11% 02. 5.2.3 Trace Metals Emission Standards Generally, although it is possible to set absolute numbers for emission levels of most individual contaminants, the task is much more difficult when the synergistic effects of

146 combinations of contaminants are considered. In the case of individual species, scientists and regulators determine the level at which effects are observed in sensitive receptors (i.e., microbes, invertebrates, higher aquatic organisms, vegetation or mammals) and then use this information to set maximum chronic or acute exposure levels. Conversely, synergistic effects can only be examined within a finite set of combinations. Table 5.5 provides a summary of trace metal emissions limits. The metals in the classes vary by country as shown in the note below the table. Table 5.5 Municipal Solid Waste Incinerator Emissions Limits - Trace Metals (Values expressed as mg/Rm 3 @ 11% 02) ,,,

Jurisdiction (Country/State/Province) European Economic Community 1991

II 1.0

5.0

0.11

France 1991

0.05*

0.05*

5.0

Netherlands 1993

0.05"

1"1"

5.0

Sweden 1986 (Recommended) t

0.08 0.22

1.0

5.4

0.05 Hg, 0.05 Cd + TI

0.5***

NA

Germany 1990 Denmark 1991

3.2

0.20

1.0

5.0

0.1 Hg, 0.05 Cd

2.0 Pb+Zn+Cr

0.5 As+Ni+Cu

0.1"

1.0"***

NA

0.014 Cd 0.07 Cd 0.028 Cd

0.14 Pb 1.12 Pb 0.34 Pb

0.056 Hg 0.056 Hg 0.056 H~

none

none

none

British Columbia 1991

0.2 Hg/0.1 Cd

0.004 As/0.001 Cr

0.05 Pb

Burnaby permit 1983

0.2

1.0

5.0

Austria United Kingdom U.S.A. 1995 MACT Rule (NSPS) New >35 tpd Existing >35 & <225tpd .>225 tpd Canada CCME Guidelines 1988

Note:

I

0.20"

Italy

Switzerland 1986

9

Trace Metals by Cate.qory

Ontario Guideline A-7 1995 0.014 Cd 0.14 Pb 0.057 Hq Generally, Hg and Cd are in Class I but Sweden has Hg only and the old German and British Columbia standards include TI in Class I. Class II has As and Ni in the EC; Class III for the EC is Pb, Cr, Mn and Cu; in the Netherlands & Switzerland Pb and Zn; elsewhere the class contains Pb and Cr. * the French regulations adopted the EC Directive but tightened the cadmium and mercury emission levels. ** represents total for each compound Hg and Cd. *** the German standard combines As, Co, Cr, Ni, V, Sn, Sb, Pb, Cu, and Mn. .... the U.K. standard combines As, Cr, Ni, Sn, Pb, Cu, and Mn. NA no standards for Category III. 1" Individual for each plant. 1"1" Total of Pb, Sb, Cr, Cu, Mn, V, Sn, As, Co, Ni, Se, Te should not exceed 1.0 mg/Nm 3 (max hourly average)

147 The regulation of trace metal emissions from incinerators is a new development in North America. The U.S. NSPS (1995) regulations are based upon the Maximum Achievable Control Technology, i.e., those demonstrated by the five best plants in operation, [MACT] criteria and these values have been adopted by the province of Ontario, Canada. Previously, Ontario had applied the Point of Impingement dispersion calculation procedure to determine acceptable stack emissions. The province of British Columbia adopted the TA Luft (1986) values in setting permit conditions for the Burnaby facility. Previously, in the U.So, NESHAP standards exist for lead, mercury and beryllium, along with PSD limits for lead with some state and local standards also existing (Jordan, 1987).

5.2.4 Trace Organic Emission Standards Standards for organics are listed on Table 5.6. These are more difficult to develop than inorganic standards because there are differences in toxicity within a family of organic compounds. For example, there are 210 congeners of polychlorinated dibenzo-pdioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) divided into eight homologues for each compound. Since each congener has a wide range of toxicity, scientists developed a scaling factor (the Toxic Equivalency Factor, TEF) for the 17 congeners that are considered to be the most toxic. A Toxic Equivalent (TE) value is then calculated by:

mE=~ (XixTEF,) i=1 ,n

where

X~ = concentration of a specific congener; and, TEFi = the toxic equivalency factor for that congener.

This cumulative value can then be used to estimate the potential total toxicity relative to a single congener, namely 2,3,7,8-TCDD, which is considered the most toxic congener of the PCDD/PCDF group. The most common scheme for applying PCDD/PCDF toxicity factors is shown in Table 5.7. Not all legislation has adopted the ITEQ approach. The Eadon equivalency factors used in some European legislation result in a value that is approximately twice the ITEQ value for the same sample. This implies that a value of 0.1 Eadon is equivalent to 0.05 ITEQ. With that in mind, the Danish value is close to the CCME Canadian level. Furthermore, while not considered particularly significant, there are subtle difference in the sampling and analytical methods used for these compounds in Europe and strict comparisons to North American numbers may lead to invalid conclusions. The German value results from measurements expressed on a wet basis and for a considerably longer averaging time than that conventionally used in North America. The net effect of this is that the German value equates to approximately 0.3 ng ITEQ/Rm 3 @ 11% oxygen. On the basis of toxic equivalents, the Swedish standard would appear to be the lowest at 0.05 ng/Rm 3.

148 Table 5.6 Municipal Solid Waste Incinerator PCDD/PCDF Emissions Limits Jurisdiction (Country/State/Province) PCDD/PCDF (ng/Rm 3) (Toxic Equivalents) Netherlands 1993

0.1

Sweden 1986 (Recommended)

0.1 Eadon

Germany 1990

0.1

U.S.A. 1995

9 (total) 88 (total) 21 (total) exceptESPequippedfacilities42 (total)

New >35 tpd Existing >35 & <225 tpd >225 tpd Canada CCME Guidelines 1988 British Columbia 1991 Ontario Guideline A-7 1995 Japan 1990

0.5 0.5 0.14

0.55

Table 5.7 Toxicity Equivalency Factors (TEFs) for Specific PCDD and PCDF Congeners Positions Equivalency Homologue Chlorinated Factor Dioxins TCDD 2,3,7,8 1 PeCDD 1,2,3,7,8 0.5 HxCDD 1,2,3,4,7,8 0.1 1,2,3,6,7,8 0.1 1,2,3,7,8,9 0.1 HpCDD 1,2,3,4,6,7,8 0.01 OCDD 1,2,3,4,6,7,8,9 0.001 Furans TCDF 2,3,7,8 0.1 PeCDF 1,2,3,7,8 0.01 2,3,4,7,8 0.5 HxCDF 1,2,3,4,7,8 0.1 1,2,3,7,8,9 0.1 1,2,3,6,7,8 0.1 2,3,4,6,7,8 0.1 HpCDF 1,2,3,4,6,7,8 0.01 1,2,3,4,7,8,9 0.01 OCDF 1,2,3,4,6,7,8,9 0.001 Note: * When only homologue test data are available, then the most conservative (largest) equivalency factor should be applied. NATO CCMS, 1988

149 The U.S. EPA 1995 standard for PCDD/F does not involve the use of toxic equivalents; rather compliance with the standard specified must be determined through the use of a specific sampling and analytical method that quantifies the amount of each of the 2,3,7,8 substituted congeners (isomers). These are totalled and the total is compared to the standard. This effectively uses the same congeners that are included in the ITEQ scheme, but applies no weighting to the various levels of these material. 5.3 CURRENT ASH AND RESIDUE DISPOSAL PRACTICES The bulk of the residue generated at an MSW incinerator consists of grate ash. In Canada, most European countries and Japan, bottom ash is handled separately from the other residue streams, whereas the current trend in the United States is to combine all the residue streams and dispose of this material in dedicated landfills. In most jurisdictions, bottom ash is disposed by landfilling and regulations governing this activity have been developed. However, utilisation of the material as a substitute lightweight aggregate has emerged as a viable option to landfill disposal and experience has led to various regulations being developed specifically for utilisation practices. Since the regulations and management practices for ash are evolving in all countries, the current trends and approaches are described below. 5.3.1 Disposal of Bottom Ash Canada In Canada, ash is currently handled as two separate streams, bottom ash and fly ash/APC residue. The CCME guidelines (1989) recommended ash be handled in this manner to prevent contamination of the bulk of the material by the high trace metals concentrations in the fly ash. Furthermore, the guidelines suggested that bottom ash could be disposed in a conventional municipal landfill, which is the current procedure in British Columbia, Ontario and Quebec, although the bottom ash from the Burnaby, B.C., facility is used for road construction purposes within a landfill. Denmark Since 1984, Denmark has utilised a very large portion of the bottom ash generated at its incineration facilities. In 1993/94, Denmark utilised between 400,000 to 450,000 tonnes of processed bottom ash which represents almost 100% of the total amount generated (Hjelmar, 1994). The ash is processed by screening and removal of ferrous materials to generate an upgraded product. Previously, approximately 26% of the total amount produced was disposed in dedicated monofills. The term monofill refers to segregation of ash from other waste materials, including MSW. A further 36% was used for fill or land reclamation purposes (Ludvigsen, 1991). The impetus for wide spread use of ash stems from the imposition of a State tax on disposal which was initiated in 1987 (Hjelmar and Ludvigsen, 1993).

150

France

France has a landfill regulation (Law on Waste Disposal, 1975, revised 1992) which suggests that landfilling is the last resort after all recyclable uses have been made of the material. This encourages incineration of waste after recycling and utilisation of residues where appropriate. Any material landfilled must contain less than 5% organic matter and the TOC of the leachate is limited. The act stipulates categories of materials according to its solubility. A material can be recycled if the solubility is less than 3%. There are two classes of landfills: Class 1, Hazardous Waste with a solubility greater than 5% and less than 10%; and Class 2 where the material has a solubility less than 5%. All landfills must be lined. Bottom ash is the only material that is currently considered appropriate for re-use, however, given the limits on recycled materials, most residues will require some treatment before they can be used or they will require disposal.

Germany and Switzerland In Germany and Switzerland, landfill disposal of materials requires that the residues meet a loss on ignition criteria (a measure of the unburned material in the ash) of less than 10% and contain less than 10% soluble salts. Furthermore, leachate from the residue must meet criteria for various trace metals based on elution with distilled water. In 1993, the German Bundesministerium f0r Umwelt issued a new directive on landfills used for both MSW and incinerator residue disposal. This legislation defines two classes of landfills based on the total organic carbon, loss on ignition at 550~ leachate quality as defined by DEV $4, and solubility. The new directive promotes utilisation as the preferred option for bottom ash. If there are no available markets for utilisation, disposal should be at a Type 1 landfill. After simple in-plant treatment, bottom ashes from properly operated incinerators will be able to meet the criteria (Schneider et al., 1994). These are summarised in Table 5.8. The objective of these standards is to reduce the reactivity of materials being placed in landfills.

Netherlands In the Netherlands, the Soil Protection Act of 1987 provides statutory authority for protection of the environment through limiting the pollution of soil by anthropogenic activities. Under the act, construction materials are regulated to prevent contamination from industrial residues that may be used in construction. The regulations limit releases to the environment to a small percentage of the existing level of that contaminant in the first metre of underlying soil. The effect of these regulations is to emphasise the efforts of controlling releases of contaminants found in very low concentrations in the Dutch environment. This is a distinct difference from many other jurisdictions where release

151 Table 5.8 Summary of German Landfill Requirements Standard Not To Be Exceeded Type1 ... Type 2

Classification Criteria For Landfills TOC (%)

1"

3**

LOI (%)

3*

5**

pH

5.5- 13

5.5- 13

Conductivity (uS/cm)

10,000

50,000

TOC (eluate, mg/L)

20

100

Phenols

0.2

50

Solubility (%)

...... 3

.

6

Leachate Quality (mg/L) (from DEV $4 Procedure) As

0.2

0.5

Cu

1

5

Hg

0.005

0.02

Zn

2

5

Cd

0.05

0.1

Cr§

0.05

0.1

Ni

0.2

1.0

F

5

25

Pb

0.2

1

NH4

4

200

CN

0.1

O.5

AOX * = for new incinerators TA Siedlungsabfall, 1993

0.3

1.5 ** = for old incinerators

152 limits are based largely on the toxicity of the contaminants. Under the Soil Protection Act, a separate regulation deals with the disposal of wastes (Regulation for Disposal). In the Netherlands, a large portion (>80%) of the bottom ash produced is utilised in embankment and roadbase applications. Ferrous rejects are recycled.

Sweden It is estimated that Sweden produces 400,000 tonnes of bottom ash and 60,000 tonnes of fly ash and APC residues annually (F~llman and Hartldn, 1992). This quantity fills 250,000 m3 of dedicated monofill space in recently approved disposal sites. Each site has its own permit requirements which were approved by the Environmental Franchise Board. Furthermore, monofills that are used for both bottom ash and APC residues must dispose of these streams in separate cells. Current recommendations suggest that leachate be collected for the initial filling period and after this time infiltration should be kept below 50 mm/year by the use of proper soil covers. The Swedish regulators are currently monitoring disposal requirements developing in the rest of Europe with a view to amending their standards. Regardless of the standards imposed, local citizens are afforded an opportunity to review and comment on any landfill development plans during the approval stages. Efforts to develop a suggested re-use criteria are also under way in Sweden as discussed in the next section. United Kingdom In the United Kingdom, no special provisions exist for the disposal of ash from MSW incinerators, although the issue is under review. All ash generated in society, be it from residential, commercial or industrial establishments, is classified as a "controlled waste". Controlled waste must be disposed at approved licensed facilities that can handle the material. Licensing requirements reflect the need to preserve the environment and ensure neither the water resources nor public health are endangered by the disposal practice. The current practice is to co-dispose with MSW or to use the material as cover in older landfill sites. These sites are under reducing conditions and the theory is that they present a more stable environment for the containment of trace metals. The regulations governing ash disposal are expected to change when the new Air Emissions Regulations force facilities to install new APC systems in 1996. United States Up until the mid 1980s, most MSW incineration residues in the U.S. were disposed in co-disposal situations with MSW. Regulations for disposal varied by state and local situation, and considerable debate and confusion existed about the status of these materials with respect to the RCRA Subtitle C (hazardous waste) testing and management requirements. This was brought about by an exemption for "household waste" from the provisions of Subtitle C. However, in the Spring of 1994, the U.S. Supreme Court ruled that MSW incinerator ash was no longer exempt from testing using the Toxicity Characteristic Leaching Procedure (TCLP). Thus ash (combined or

153 separated bottom ash and APC residues) which passes the criteria associated with the TCLP can be landfilled or monofilled, however, ash which fails the criteria must be disposed as a hazardous waste. This involves disposal in a secure landfill with provisions including a series of liners and leachate collection and treatment facilities which are more stringent than the design criteria for Subtitle C landfills. Most new facilities are using monofills for combined residue disposal, but where space is limited, interest in utilisation is increasing. Although the majority of the facilities combine the residue streams, a small number segregate the bottom and fly ash/APC residue streams to facilitate treatment of the fly ash/APC residue. Furthermore, regulations regarding ash management still vary widely from State to State. For example, New York State requires semi-annual testing for ash and are developing a procedure to handle this material as a special waste (e.g., Bill 10780, State of New York), whereas the State of Florida has permitted the use of ash in artificial marine reef construction projects. Moreover, although some States actively discourage the practice of co-disposal with MSW, other States endorse the practice.

Japan In Japan, the Waste Disposal and Public Cleaning Law, which addresses all aspects of waste disposal, was thoroughly amended in 1991. Under that law, incinerated ash is classified as either bottom ash or fly ash. Bottom ash is treated as normal domestic waste and disposed directly into sanitary landfill sites. 5.3.2 Disposal of Fly Ash and APC Residues The options for handling and disposing of the finer ash streams from incinerators are more limited. Most jurisdictions treat the material as a hazardous waste.

Canada In Canada, the fine ash material must be handled as a hazardous waste. The disposal options include transfer to a hazardous waste disposal facility or treatment of the residues prior to disposal. Various treatment alternatives from disposal in secure landfills to solidification are being evaluated, but there are few regulations in place to evaluate the efficacy of a treatment process. The exception is in British Columbia, where the treated ash must pass a battery of laboratory tests prior to disposal in a conventional landfill. The testing protocol includes evaluating the treated residue using chemical, engineering, durability and leaching tests (Government of British Columbia, 1992). Denmark & the Netherlands In Denmark, APC residues from the dry or semi-dry processes and fly ash are currently classified as hazardous wastes and are disposed in dedicated monofills with leachate

154 collection systems and bottom liners, and often with impermeable cover layers. Wet scrubber sludges are generally monofilled alone or are mixed with fly ash residues. All of these measures are only considered temporary solutions until suitable treatment systems are made available. At sites in Denmark and the Netherlands, APC residues are stored in polyethylene bags in landfills that have leachate collection systems and bottom liners. Generally, APC residues in the Netherlands are sent to a hazardous waste landfill site, although nearly 40% of the ESP residues from Dutch facilities are currently utilised as a very small percentage filler in asphaltic concrete mixes, but this practice is waning. This residue stream is segregated from all other streams for this purpose.

France

The 1991 French law on MSW incineration adopted the EEC directive on air emissions but has tighter mercury and cadmium standards. This has resulted in an increased use of wet APC systems, and hence, more sludge from these systems. The changes in regulations have fostered increased study into ways of modifying residues to meet the disposal criteria mentioned above. Immobilisation of contaminants by solidifying with hydraulic binders is being practised in some areas, and four organisations are currently exploring vitrification alternatives. One manufacturer is utilising the wet scrubber system to modify the residue to meet the criteria (Knoche, 1992).

Germany

In Germany, the APC system has to be designed in a way to minimise the production of harmful residues (Bundesministerium, 1993). Heat recovery system ash is separated from dry/semi-dry scrubber residues in some facilities. The fly ash and APC residues are classified as a hazardous waste and requires disposal in either approved landfills or preferably in underground disposal sites such ash old salt mines or in special cells of municipal waste disposal sites. Mehlenweg (1990) estimates that 5% of the total fly ash/APC residue stream (210,000-240,000 tonnes/year)is deposited in underground sites, less than 1% is re-used and the balance goes to surface storage. To minimise the release of dust from surface stored materials, it is packaged in large bags or moistened. German regulations allow boiler/economiser and filter ashes to be modified to reduce the need for controlled disposal, however, few methods have been developed to the commercial scale. Heat recovery and filter ashes along with APC residues contain high quantities of water soluble salts and in Germany they are required to be disposed of in hazardous waste landfills. Limited work has been done to explore the options for treatment/re-use of these materials, but these processes have not progressed beyond pilot scale (Juritsch, 1990; Kurzinger, 1990).

155

Netherlands At the present time, APC residues and fly ash are considered hazardous wastes and are generally managed in a similar manner to that used in Denmark. Sweden In Sweden, APC residues are disposed separately from bottom ash. The Environmental Franchise Board is responsible for setting the requirements for these disposal sites. It has been found that the Swedish infiltration limit of 50ram/year is not suitable for limiting releases from APC residues and fly ash. A new practice is to stabilise these materials before disposal. This is done at one facility in Stockholm by adding 40% low calcium cement to the residue stream. This increases the volume of the stream but further retards the infiltration rate into the material. Other options for the disposal of APC residues and fly ash are currently being examined, including slurry deposition to achieve better compaction, advanced immediate compaction to reduce permeability and the use of plastic covers during deposition to reduce infiltration. Japan In Japan, fly ash and APC residues are treated as a domestic waste under special control. Before disposal, they have to be tested via a leaching procedure and compared to waste disposal standards. In order to treat fly ash, the Ministry of Health and Welfare specified four treatments: 9Melting and solidification 9Solidification by cement 9Stabilisation using chemical agents 9Extraction with acid or other solvent After all standards have been passed, the treated fly ash could be disposed directly into sanitary landfill sites with other domestic wastes.

5.3.3 Utilisation Two fundamental concerns with utilisation applications are that 1) the physical properties of the material are appropriate for the intended application (i.e., bearing capacity, compaction, etc.), and 2) the application does not lead to environmental degradation. The latter situation relates mainly to the leaching of metals and salts from the ash, since the potential loading of ash within a fill application may pose a potential problem. In Europe, the materials are used as a civil engineering material, largely as base and sub-base for roadways. Each country has considered the environmental implications of these uses and developed guidelines for implementation. While the subject of utilisation is discussed in more detail later in the document, a brief discussion of existing regulations governing the use of residues is included here.

156 Canada As mentioned previously, although no major efforts have been devoted to utilisation in Canada, the Greater Vancouver Regional District has evaluated bottom ash for utilisation applications, and currently uses the material for construction of roadways within a landfill site. The bottom ash undergoes ferrous removal prior to leaving the Burnaby facility, but no other processing is done other than compaction during placement.

One of the major impediments to bottom ash utilisation is that there has been little economic incentive to divert materials from landfill. Should sufficient regulatory criteria be put in place to allow the use of bottom ash as a lightweight aggregate, it is likely that the practice would be considered for ash from some of the major facilities. Denmark Although part of the bottom ash stream from incinerator facilities in Denmark has been used in sub-bases for roads, bicycle paths and parking lots since 1974, the first Danish utilisation requirements were not developed until 1983 (Statutory Order No. 568 of Dec. 6, 1983). Moreover, these requirements only applied to the use of small to moderate amounts of ash. Large scale applications (>30,000 tonnes or 5 m thickness) are regulated under the Environmental Protection Act (Disposal and Discharge Permit section). Additional guidelines for road sub-bases were developed in 1989 by the Danish Highway Department (Pihl et al., 1989 in Hjelmar, 1990). The Statutory Order is currently being reviewed. Ferrous material is removed from the ash by screening and then magnetic separation to generate an upgraded material for recycling purposes. The portion of the bottom ash stream which cannot be used is disposed in dedicated monofills.

A Danish testing protocol has been developed to determine the suitability of ash for utilisation based on chemical parameters (Hjelmar, 1990). The conditions include a pH >9 for a 1% slurry of the material, alkalinity of >1.5 eqv/kg, metals levels as determined from a HNO3 leach of Pb <3000 ppm, Cd<10 ppm and Hg<0.5 ppm. There are also restrictions on the placement of ash that passes the criteria. Under paved roads, the maximum average thickness allowed is one metre and the ash must be above the highest groundwater table and more than 20 metres from the nearest well. Under unpaved roads the same regulations apply except the thickness is restricted to 0.3 m. Sub-base material has specific size requirements (<45 mm with a specified fines level), must have been stored for at least one month, have a loss on ignition of <10% and a water content between 17 and 25%. France France currently utilises 45% of the bottom ash in roadbeds, however, this may change due to recently introduced regulations (Ministere de rEnvironnment, 1994). Bottom ash destined for a utilisation application must meet criteria in relation to combustible content

157 and leaching characteristics. The LOI content must not exceed <5% and the results from the leaching test AFNOR NF X 31-210 are compared to the list given in Table 5.9 The material will still require ferrous removal, screening and aging. Table 5.9 Requirements for Bottom Ash Utilisation - Leaching (mg/kg unless noted)

....

Total Soluble Solids

As

Cd

Cr.6

Pb

Hg

SO4

TOC

<5%

<2

<1

<1.5

<10

<0.2

<1.0%

<1500

Netherlands

The utilisation of ash in building materials is governed under the Soil Protection Act in the Netherlands. This regulation covers both granular (aggregate) and monolithic (solids >50 cms in size) and classifies materials according to chemical composition and leaching characteristics. If the material does not meet the requirements of one of the categories it cannot be placed in contact with the soil. Generally, material containing MSW incinerator ash of any form can be used if it meets the requirements but must be reclaimed after its useful life is over. Bottom ash must be placed at least 70 cm above the groundwater table and should be isolated from infiltration. Leaching requirements for material are relaxed for placement in the marine environment so that CI, F and SO4 leaching levels can be higher. The Dutch leaching limits are derived on the basis of a 1% increase in the average natural soil composition for the one meter of soil under the construction after exposure of 100 years. A process of certifying MSW incinerator residues for utilisation is currently being developed in the Netherlands. Other work has been ongoing to develop technical specifications for the use of bottom ash in roadway construction.

Germany

In Germany, the data on the reuse of bottom ash is not consistent, but they indicate a substantial amount is used in construction applications. A 1991 survey of the 48 German incinerator facilities indicated that the production of bottom ashes was approximately 2,400,000 tonnes, 60% of which (or 1,450,000 tonnes) were utilised (Krasse & Radenberg, 1994). Some large scale projects, each using tens of thousands of tonnes of material, were initiated during the late 1980's to demonstrate the capability of bottom ash use in road bases, embankments and noise protection walls (Toussaint, 1989). In March 1994, the federal states passed new legislation regarding ash management (LAGA, 1994). The legislation is intended to promote the reuse of as many residue streams as possible, but the main emphasis is on bottom ash. Bottom ash destined for reuse cannot contain heat recovery or fly ash and the ash requires pretreatment to

158 remove the metallic and unburnt materials from the mineral fraction. The recyclable ferrous scrap must contain >92% Fe with a grain size between 50-70 mm, with no less than 5% being under 5 mm in size. Aging has been found to increase the chemical stability of the ash by washing out readily leachable salts, forming carbonates via the sorption of CO2, and hydrating metallic and oxide forms of elements like aluminum. Aging also allows swelling reactions to occur prior to placement. The ash must be screened to remove material > 45 mm, which is then returned to the incinerator, and the remainder is aged for three months before qualifying for testing to determine its suitability for use. The tests involve an assessment of the loss on ignition (must be less than 2%), water soluble salt levels (less than 1%) and the generated leachate (for various ions and metals). At least one sample > 2 kilograms in weight is required per 10 m3 of ash, and there must be a least 10 grab samples per pile of ash. In addition to bottom ash, there are criteria in place for the use of by-products from APC systems, namely, NaCI, Na2SO4, gypsum and HCI. The criteria are given in Tables 5.10-5.13.

Sweden In Sweden, it has been suggested that the effect of bottom ash utilisation should be quantified in absolute terms and compared to the effects of using natural alternatives (Hartl~n and Lundgren, 1991). Consequently, a set of utilisation regulations is being developed based on a scheme to classify materials into at least three classes: inert, usable with restrictions, and unusable and requiring disposal. While Sweden has not utilised much bottom ash to date, the proposed guidelines are similar to those outlined above: 9 pre-screen and remove ferrous 9 maximum grain size 50 mm 9 no more than 10% smaller than 0.06 mm 9 loss on ignition less than 5% and age for three months. Hartl~n and Lundgren (1991) also suggest that aged ash should contain approximately 17% water, a condition they consider acceptable since it allows compaction with a 10 tonne vibrating roller to a 90% modified proctor. In addition, they suggest that after grading, 70% of the material is a suitable substitute for coarse aggregate. One-third of the rejects are coarse scrap that can be sold for recycling, whereas the rest must be disposed. The largest market for processed bottom ash is as a substitute aggregate. Limits are also proposed for metals in leachates (Hartl~n and Lundgren, 1991). The material can be used in embankments, under bicycle paths and low traffic roads, and under light buildings and floor structures. The thickness of the fill is generally restricted to a maximum of 3 metres. Furthermore, it must be placed above the groundwater table and below pavement.

159 Table 5.10 German Specifications for NaCI Production from APC Systems (IJg/g) Parameter NaCI Ca Mg SO4 K F Br I Sr Ba Fe Mn Ni Co Cr * = critical value

Limit

Parameter

Limit

> 95-96% 2% 0.2% 2% 1500 60 50 10 20 20 10 1" 1" 1" 1" ** = less critical

Cu Mo V Ti Zn Cd Hg Si Sn Pb As AI N TOC

5 1" 1" 10 1 1" 1 *** 1** 1** 0.5 1000 20 ****

*** = not critical **** = to be determined

Table 5.11 German Specifications for NazSO 4 Production from APC Systems Parameter

Limit

Parameter

Limit (#g/g)

Na2SO 4 CI H2SO4 pH Insoluble fraction Water Content Colour By-products COD

> 41.5% 1% 1% Neutral 0.05% 60% White None 100 mg O2/L

Fe Zn Mn V AI Sr Cr Mg Ca

50 250 2 10 10 10 10 25 50

160 Table 5.12 German Specifications for Gypsum Production from APC Systems PARAMETER

LIMIT

Moisture Content CaSO4.H20 pH Colour Odour Mean Grain Size By-products Soluble (water) MgO Soluble Na20 Soluble K20 Soluble CI Soluble SO2 CaSO3. 89 AI203 Fe203 SiO2 CaCO3 + MgCO3 NH3, NO3 TOC Trace Elements

10% > 95% 5- 9 > 80% White None > 60% passing 32 IJm 5% 0.1% 0.06% 0.06% 100 IJg/g 0.25% 0.5% 0.3% 0.15% 2.5% 1.5% none 0.1% Non-toxic concentrations, nonradioactive

Table 5.13 German Specifications for HCL..production from APC Systems (mg/L) Parameter

Limit

Parameter

Limit

HCI SO4 Heavy Metals Fe Cd Hg As TI CI2

30 - 31% 20 1 Total 10 0.1 0.1 0.1 0.1 10

HBr HI HF NO3 NH4§ TOC AOX PCDD/PCDF

25 10 10 10 1 5 1 1 ng I-TElL

161

Japan

Almost all the incinerated ash and APC residue in Japan (5,991,000 tonnes in 1991) are disposed in sanitary landfill sites. This value accounts for 35.6% of the 16,800,000 tonnes of landfilled material in 1991. A process of melting MSW incinerated slag is currently being developed and applied on a commercial scale. The Japan Waste Research Foundation pilot project is intended to determine the effectiveness of this melted slag as an additive to asphalt and as an upper and lower road bed material. Preliminary results indicate no appreciable difference in holding strength or road condition was found between the experimental slag and natural materials (Ando, 1993).

United States

Various utilisation alternatives have been investigated in the United States. Long Island Regional Planning Board (1990) reported that ash characterisation studies were sponsored by U.S. government agencies as early as the 1960s. These continued into the mid-1970s with particular emphasis on aggregate substitution in portland cement and asphalt paving (Kenahan et al., 1966; West Virginia University, 1971; Haynes and Ledbetter, 1975; Pavlovich et al., 1977; Pindzola, 1976; and Collins et al., 1977). The recommendations from these studies included limiting combustibles to less than 10% and storing materials for several months prior to use to reduce moisture and residual organic concentrations. The Long Island RPB (1990) report also refers to other studies in the United States where incinerator ash has been used for various aggregate purposes and suggests that the more current data show better performance. The report attributes this to improved combustion efficiency and less organic material in the ash. The authors note that the major difference between ash utilisation in the U.S. and Europe is the fact that, in the U.S., the emphasis has been on ash in a combined matrix such as concrete or bituminous products.

REFERENCES Ando, S. "Waste Management R&D by the Japan Waste Research Foundation", Proceedin.qs of 3rd Workshop on Solid Waste Mana.qement betw_een Japan and the ..Federal Republic of Germany, 1993. Born, J.G.P. Quantities and Qua!.ities of Municipal Waste Incinerator Residues in the Neth.erland.s. Environmental Aspect.s of..Construction with Waste Materials (J.J.J.M. Goumanns, H.A. van der Sloot, and Th. G. Aalbers, Eds), Elsevier Science Publishers, Amsterdam, pp. 633-644, 1994. Bundesministerium fer Umwelt, Naturschutz und Reaktorsicherheit Dritte AIIgemeine Verwaltungsverschrift zum Abfallgesetz (TA Siedlungsabfall). Technise Anleitun.q zur Verwertun~, Behandlun.q und sonsti.qe.n Entsorgun.q von SiedLun.asabf(~l vom ...14, . Bundesanzeiger Jahrgang 45, Nr. 99a, May 1993.

162 CCME. Canadian Operatin,q and Emission Guidelines for MSW Incin.erators. A publication of the Canadian Council of Ministers for the Environment, Ottawa, 1989. Collins, R.J., R.H. Miller and S.K. Ciesielski. "Guidelines for Use of Incinerator Residue as Highway Material". Federal Highway Administration Report No. FHWA-RD-77-150, Washington, D.C., 1977. Complin, P.G. "Witness Statement before the Consolidated Hearing Board in the matter of and application for Petro-Sun International Inc./SNC Inc. Consortium. Energy From Waste Facility, City of Brampton on Behalf of ARB." Ontario Ministry of the Environment, 1988. .

Dutch Regulation for Construction Materials. Staatscourant, 121. 26 June 1991. Dutch Regulation for Disposal. Staatsblad van het Koninkrijk der nederlanden, 55, 1993. EEC. Official Journal.. of the European Communities: Council Directive on the Prevention of Air P.ollution from New Municipal Wa.s.te Incineration Plants. 89/369/EEC, June 8, 1989a. EEC....Official Journal of the European Communities: Council Directive on the Prevention of Air Pollution from Existin.q Municipal Waste Incineration Plants. 89/429/EEC, June 21, 1989b. F~llman A.M. and J. Hartl~n. "New Perspectives on the Management of Residues from MSW Incineration in Sweden." 7th International Waste Management Conference, Berlin, Germany and contained in W.aste Mana.qement International, Karl J. Thom~Kozmiensky (ed.) Berlin:EF-Verlag fuer Energie-und Umwelttechnik ISBN 3-92451164-0, 1992. Government of British Columbia. Waste ManaQement Act 63/88 - Special Waste Schedule 4, amended B.C. Regulation 132/92, April 16, 1992. Hartl~n, J. and T. Lundgren. "Utilization of Incinerator Bottom Ash - Legal, Environmental and Engineering Aspects." Proceedinqs of WASCON '91, Maastricht, The Netherlands, November 1991. Haynes, J. and W.B. Ledbetter. Incinerator Residue in Bituminous Base Construction. U.S. Department of Transportation, Federal Highway Administration Report No. FHWARD-76-12, Washington, D.C., 1975. Hiraoka, M. "Formation and Control of Dioxins in Municipal Solid Waste Treatment" (in Japanese), Waste Mana,qement Research of the Japan Society of Waste Mana.qement Experts, Volume 1, No. 1, pp. 20-37, 1990.

163 Hjelmar, O. "Regulatory and Environmental Aspects of MSWI Ash Utilization in Denmark." ProceedinQs of the Ash III Conference., Arlington, VA, November 1990. Hjelmar, O. and K. Ludvigsen. "Management of MSWI Residues in Denmark", in Wast___.._ee ManaQement International (K.J. Thom~-Kozmiensky, Ed.), Berlin: EF-Verlag Fur Energie - und Umwelttechnik, pp. 133-144, 1992. Jordan, R.J. "The Feasibility of Wet Scrubbing for Treating Waste-to-Energy Flue Gas." JAPCA, Vol. 37, No. 4 August 1987, pp. 422-430. Juritsch, V. and G. Rinn. "Hydrogen Chloride Absorption and the Production of Hydrochloric Acid from Flue Gas of Incinerators." Recyclin.q International Edited by K.J. Thom~-Kozmiensky. Berlin: EF-Verlag, 1990, pp. 1230-1233. Kenahan, C.B., P.M. Sullivan, J.A. Ruppert and E.F. Spano. Composition and Chara.cteristi~ of Municipal Incinerator Residues, U.S. Department of Interior, Bureau of Mines, Report of Investigations No. 7204, Washington, D.C., December 1966. Knoche, M. "Residues of Municipal and Hazardous Waste Incineration - Situation in France." 7th International Waste Management Conference, Berlin, Germany and contained in Waste Mana.qement International Edited by K.J. Thom~-Kozmiensky. Berlin: EF-Verlag fuer Energie- und Umwelttechnik ISBN 3-924511-64-O, 1992. K0rzinger, K. and R. Stephan. "Hydrochloric Acid and Gypsum (Sulphuric Acid) as Utilizable End Products Obtained From the KRC Process for Cleaning Flue Gases From Incinerators", ..R.ecyclin.q In.tern.ational Edited by K.J. Thom~-Kozmiensky. Berlin: EF-Verlag, 1989, pp. 1224-1233. Krass, K. and M. Radenberg. Verwertungsraten von industriellen Nebenproduk und recycling - Baustoffen, Beihefte zur mull and Abffall, 31, pp. 11-15, 1994. LAGA. "Merkblatt Entsorgung von Abfallen aus Verbrennungsanlagen f0r Siedungsabf,~lle, verabschiedet durch die L~nderarbeitsgemeinschaft Abfall (LAGA) am 1, M~rz 1994", In M01[-Handbuch (Ht~se!,.G., Schenkel, W. & Schnurer, H., ed.) Berlin: Erich Schmidt Verlag, Kennzahl 7055, Lfg. 4, 1994. Leenders, P. "Municipal Solid Waste Residues in the Netherlands", Waste Materials in Construction. (J.J.J.M. Goumanns, H.A. van der Sloot, Th.G Aalbers, Eds), Elsevier Science Publishers, Amsterdam, pp. 593-600, 1991. Long Island Regional Planning Board. The Po.tential for Beneficial Use of Waste to EnerQy Facility Ash En.qineerinQ Property Data Report (Report No. 4) Long Island Regional Planning Board, Hauppauge, NY, June 1990.

164 Ludvigsen, K. "Nyttigg~relse af forbraendingsslagge." M.Sc. Civ. Eng. Thesis. Note 912. Institut for Veje, Trafik og Byplan. Technical University of Denmark, Lyngby, 1991. Ministere de I'Environment. Circulaire Relative ,~ la Valorisation de M&chefers d'lncineration de R~sidus Urbans en Techniques Routi~res, DPPR/SEI/BPSEID/FC no. 94-1V-1, Paris, France, 1994. M(~hlenweg, U. & Brasser, T. "Reststoffe bei der Hausm(~llverbrennung", Abfallwirtwirtschafts-Journal 2" 53-58, 1990. NATO CCMS. International Toxicity Equivalenqy Fact.ors (I/TEF) Method of Risk Assessment. for Complex Mixtures o.f...Dioxins and Related Compounds, Report 178, December 1988. Pavlovich, R.D., H.J. Lentz and W.C. Ormsby. Installation of Incinerator Residue as Base-Course Pavina Material in Wa.shin.qton, DC, Federal Highway Administration, Report No. FHWA-RD-78-114, Washington, DC, 1977. Pihl, K.A., P. Ahrentzen and K. Kalsmose. Subbase of Incinerator Residue. Guidance _Standard.. Specifications- General. Workin.q Proced.u_re, Milj~ministeriet/Skov-og Naturstyrelsen/Statens Vehjlaboratorium, Laboratorieapport nr. 66, Vejdirektoratet (in Danish), 1989. Pindzola, D. Larqe Scale Continuous Production of Fused A.q.qregate from !.n.cinerator Residue, Federal Highway Administration Report No. FHWA-RD-76-115, Washington, DC, 1976. Richtlijn Verbranden. Ministrie van VROM, Augustus 1989. Schneider, J., J. Vehlow and H. Vogg, "Improving the MSWI Bottom Ash Quality by Simple In-Plant Measures". In Environmenta! Aspects of Construction with Waste Materials, (J.J.J.M. Goumanns, H.A. van der Sloot, Th. G. Aalbers, Eds), Elsevier Science Publishers, pp. 605-620, 1994. TA Luft. First General Administrative Re,qulati.0n Pertainin.q to the Federal Emission Control Law (Technica.I Instructions on.Air Quality Control - TA Luft), as of 27 February 1986, GFMBI:202 p.95. West Germany Regulations, 1986. Toussaint, A. Verwertung yon Aschen aus der Abfallverbrennung, In ENVITE Kongressband (W. Klose & J. Vehlow, Eds), Essen: Verkan-Verlag, 61, 1989. U.S. EPA. New Rules for MWC Facilities Under the Auspices of the Clean Air Act Amendments, Washington, DC, Jan/Feb 1991.

165 U.S. EPA. New Source Performance Standards for MWC Facilities Under the Auspices of the Clean Air Act Amendments, 40 CFR Part 60, Washington, DC, 1995. Vehlow, J. "Management of Residues from MSWI in the FRG", Proceedings from the Ash III Conference, Arlington VA, November 1990. West Virginia University. "Study of Incinerator Gan gue," U.S. Bureau of Mines, Solid Waste Disposal Program Grant No. GO 190583 (SWD-22).

This Page Intentionally Left Blank

167

CHAPTER 6 - ISSUES RELATED TO INCINERATOR ASH SAMPLING 6.0 INTRODUCTION

Assessing the appropriate management options for MSW incinerator residues requires a knowledge of their physical and chemical characteristics and an understanding of how these characteristics contribute to their behaviour in the field. This type of information is generally derived from laboratory analyses of samples which are extremely small in size compared to the volume of material actually generated. "The initial ... and perhaps most critical ... element in a program designed to evaluate the physical and chemical properties of a solid waste is the plan for sampling." (U.S. EPA SW-846) Thus sampling can be an error generating process, (Gy, 1982; Pitard, 1986). Furthermore, the laboratory procedures have the potential to generate highly variable data if the inherent heterogeneity of the residues is not taken into account during the sample preparation stages. Consequently, the selection of an appropriate strategy for sample collection and preparation should be based on techniques which ensure an adequate level of representativeness is achieved. The information presented in this chapter places these issues into context with the overall objectives of a sampling program, the factors which can influence sample variability, and recommended approach for sample collection of the different ash streams. These discussions will draw on the information provided in Chapters 2, 3 and 4. 6.1 THE CONCEPT OF THE REPRESENTATIVE SAMPLE

A continuous theme throughout the U.S. legislation (US EPA SW 846 - 40 CFR, Part 261, Appendix II) is the concept of representativeness. The legislation also defines variability as the change of properties with time, whereas the variance is defined as the variability between samples which is attributable to laboratory procedures. Therefore, the overall error of any characterisation process is the sum of the sampling error and the analytical error. If results are to be considered representative, not only must the laboratory accuracy be high (the closeness of an analytical result to its true value), but also the sampling precision must be within reasonable limits (the closeness of repeated sample values). Irrespective of the accuracy of the laboratory analyses, data will still be of limited usefulness if the sample collection procedure is inadequate or inappropriate. Since repeating the analytical portion of a study is much easier than reconducting the sampling program, the importance of a sound sampling program cannot be overemphasised.

168 Although it would be ideal to collect and analyse the total of anyone residue stream from an incinerator, this is generally highly impractical, if not impossible. Moreover, the ash streams tend to exhibit a relatively high degree of heterogeneity. In view of these limitations, it is unlikely a truly representative sample can be taken, and therefore the definition of a "representative" ash sample becomes a sample collected based on an approach which minimises sample variability. Many factors will influence the "representative" nature of a sample of ash from a specific facility including: the origin/type of waste the type of incinerator/APC system and the operating conditions residue type and the means by which the residue stream is transported and stored

6.1.1 Waste Type One of the first considerations regarding ash sampling is to identify some general characteristics of the waste being fed to the incinerator. MSW characteristics will vary on a load-to-load, day-to-day and season-to-season basis. Unless the incinerator burns RDF, it is unlikely the waste will be adequately mixed in the storage pit to maintain uniform feed quality, which invariably means the chemical and physical properties of the residues generated from burning the waste will also vary. Since it is difficult to identify the origin of a load of waste fed to an incinerator and attribute its characteristics to that of a corresponding ash sample, it will suffice to note the general characteristics of the waste. For example: is the waste pre-processed into RDF or burned "as-received"? is the waste largely commercial or residential waste? is there a source separation or curbside recycling scheme in place? what day of the week and what time of year is the sampling being conducted? This type of information will not only assist with development of a sampling strategy, but also help with data interpretation and allow other researchers to make better comparisons with data from other facilities. For example, since the processing of RDF acts to "homogenise" the waste to a much greater extent than mixing "as-received" waste in a storage pit, the sampling frequency may not need to be as great as when sampling a mass-burn design incinerator. However, the carry-over of particles in the flue gas stream of the RDF system will increase the volumes of fly ash and APC residue generated and thereby may increase the number of grab samples required to generate a composite sample. Furthermore, commercial waste may be more homogenous than residential waste, but contain a greater proportion of larger items which will influence the size of the sample which needs to be collected. In addition, if

169 a source separation scheme is in place to remove recyclable materials, a much lower frequency of ferrous and glass in the ash may result in fewer preparation steps to generate samples for analysis. Overall, this type of assessment will assist with setting some of the initial parameters for sampling ash.

6.1.2 Type of IncineratorlAPC System In addition to waste feed variability, a sampling program developed for one type of incinerator system may not be suitable for another system because of differences in the operating regime of the two facilities. All of the different processes or systems should be factored into the sampling plan and each facility should be considered as a unique set of circumstances. Issues such as the ash discharge cycle must be taken into consideration. For instance, conventional large mass-burn units with moving grates discharge ash on a relatively continuous basis, but smaller controlled air systems, such as the Consumat technology, discharge residue to the quench tank on a periodic basis. This means the frequency of grab samples taken from a Consumat system should be geared toward collecting grab samples at intervals covering the period that ash is being removed from the quench tank. Another example relates to the manner in which samples can be taken from the different incinerator systems. Extracting a sample of dry grate ash prior to quenching from a mass-burn design can be accomplished by accessing the chute between the discharge end of the grate and the quench tank. However, obtaining dry ash directly from the discharge end of a two-stage system prior to quenching is complicated. The "starved-air" conditions in the primary chamber means the system is under pressure and backdrafts can occur when accessing the last discharge ram, thereby not only causing upsets in the operating conditions, but making this type of sample collection a dangerous task as well. The heat recovery and APC systems typically generate residues on a cyclical basis. Boiler soot blowers may operate on a timed cycle, i.e., only once every 6 to 8 hours, whereas the cleaning cycles for the fabric filter systems are generally initiated when the pressure differential across the bags is sufficiently high. While the differences in times are not a problem when sampling these streams in isolation of each other, the actual cleaning cycle's events must be considered when sampling combined ash streams, especially if they are entering the quench tank. Since these residues constitute a portion of the total ash being generated, the sampling events must be timed to ensure the appropriate proportions of these residues are included in the combined sample. Without this type of consideration, any wide variations in ash properties may be unaccountable. Finally, the operation conditions during the sampling must be recorded so any deviations from "typical" operations which might affect the ash characteristics, can be identified. For example, residues created during transition periods such as start-up and

170 shutdown will not be representative of normal operation. Similarly, excursions in furnace temperatures below or well above the "normal" operating temperatures (due to either wet waste or an unusually high proportion of highly flammable materials) should be noted to assist with data interpretation. 6.1.3 Residue Streams

MSW incinerator residue streams consist of particles of varied sizes and shapes, each of which can have different physical and chemical properties. Typical particle size distribution data for various residue streams are shown in Figure 6.1, although it should be noted that bottom ash data is based on samples screened to remove <50 mm material (e.g., metal cans, small appliances, large auto-parts, and large clinkers). Normally, metal cans and other materials larger than 150 mm would be anticipated to represent between 5 and 20% of the total bottom ash stream. The sampling of large clinkers requires special consideration particularly considering the weight, size and temperature of this material as it exits the system. Such materials should be sampled and analysed separately from the bulk of the bottom ash. Figure 6.1 Particle Size Distribution of Various Residues

~,,

100

~

ao

,,.,...

'

SILT

I L

2o

GRAVEL

SAND '=~lllll ~

--'

'

APC R E S I D U E ~

(

60

40

o

CLAY

I)' /

I / / , / 2_.

0.002

0.06

, 11

Z

,/LJ

,.-

GRAIN SIZE DIAMETER (mm) Hartl~n and Elander, 1986

,//Ii/

BOTTOM ASH

60

171 Sampling waste material with wide particle size distributions requires recognition of the physical processes that can affect the distribution of materials in a storage pile. If both fine and coarse particles are discharged to a storage pile, the coarse material tends to remain on the outside of the pile whereas the fines sift to the interior (Boomer, et al, 1988), see Figure 6.2. Furthermore, materials with different angles of repose will not intermix in a pile and surface friction effects can cause a separation of different materials as they are discharged from a chute (e.g. the high friction angle material falls nearer the chute). These problems can generally be addressed by employing pile sampling strategies which account for the differences in particle size, shape and density.

6.2 OBJECTIVES OF MATERIAL SAMPLING PROGRAMS Generally there are two principal objectives for sampling the MSW incinerator residues: quality control/regulatory monitoring - to confirm that the variability of the ash stream characteristics is within an acceptable range for a number of parameters research - to develop a better understanding of the mechanisms responsible for the physical and chemical properties of the streams, and to determine which treatment procedure can be used to render the streams acceptable for the chosen management method For each objective there is likely an ideal strategy that involves both the quantity of material collected and the frequency of the collection, but tradeoffs are likely to be made between the ideal sampling strategy and the practical or optimal strategy for a given case. If, for instance, the purpose of the sampling is quality control related to utilisation of the material, it is likely the optimal sampling strategy would involve rather large increments taken on a frequent basis from a pile, whereas regulatory sampling may only require small volume samples taken on a low frequency. Regardless of the strategy, if true statistics are to be developed, the frequency of sampling should always be greater than the rate of change of the characteristics. Although a sampling approach should be conducted within well defined limits, -a certain degree of flexibility in strategy is required. For example, if significant fluctuations in the material quality are found during low frequency sampling, the sampling frequency could be increased to achieve a better understanding of the average nature of the material. A statistically valid sampling plan attempts to sort out the inherent noise and overlapping cycles so a true estimate or value of a time-dependent characteristic can be obtained. Equations 6.1 -6.6 provide a means by which the representative nature of any sample can be evaluated (SW-846). These equations define the mean, X; standard deviation,

Figure 6.2 Examples of Material Heterogeneity

CASE #I

h

Fines sift to inside of pile, while coarse particles remain on the outside.

D :,*:,

@'.C.ra*..Ir!t

Black - - mrticles rhave a steeper angle-of-repose than white particles

~I*I lS I Dm ' ~ DD ~

CASE #3

.---,

H~ghfriction angle

CASE#2

h?'

a

*I.?

The material having the steeper angle tends to concentrate in the centre, while that having the flatter angle concentrates on the outside.

-.. m-*

d3)

Surface friction affects particle trajectories so that the higher-friction-anglematerial (black) falls nearer the chute

Angle-of-repose mechanism concentrates white particles at pile base.

173 S; standard error, Sx; a confidence interval based upon the standard error, (C I), and the appropriate number of samples, n, that need to be collected. The sampling protocol defines that a minimum of 4 samples must be collected regardless of the results of calculations based upon the equations:

Mean = X-

~x~

(6.1)

i=1

where n = the number of sample points

(~x~) 2

n

(6.2)

n

V a r i a n c e = S 2 = n=l

n-1

(6.3)

Standard Deviation = S = ~-~

S t a n d a r d D e v i a t i o n o f M e a n = S2 -

S

C o n f i d e n c e I n t e r v a l = CI = X + t.=oS2

nR

where:

=

t2S

.20

2

(RT - X

)2

(6.4)

(6.5)

(6.6)

RT = the selected acceptance level or Regulatory Threshold n, = the number of samples required ta = student's t for a suitable confidence limit (suggest 80% for residue sampling)

Since these equations are based upon the assumption that the parameters are normally distributed and caution should be used when applying these statistics. Most residue samples generate data which are log normally distributed rather than normally distributed. Therefore, it is necessary to assess the distribution of any data set. A simple test of comparison between the variance and the mean of the data should indicate a normal distribution. If the variance is larger than the mean, the data is

174 unlikely to be normally distributed and must be transformed before these equations can be used. The most common transformations are Square Root Transformation the transformed value is the square root of the original value with or

without an offset parameter;

Arcsine Transformationthe transformed value is the arc sine of the square root of the original

value divided by an offset parameter and expressed as a percentage.

All statistical calculations and tests of acceptability should be carried out on a transformed data set. 6.3 AVAILABLE SAMPLING PROTOCOLS

Although few full sampling protocols for incinerator residues have been published, the Swedish Geotechnical Institute has compiled a comprehensive discussion of the issue (SGI, 1993) and an earlier protocol was proposed by Fiesinger (1988). Moreover, many reports on ash testing generally contain a basic description of the sampling approach, however, these descriptions usually do not contain the rationale for the selection of a particular method. Consequently, the framework for developing an experimental design for residue sampling is generally adapted from published standards for sampling of other materials. Besides the SGI report, there are several other documents which can provide relevant guidance with respect to ash sampling and statistical interpretation: Test Methods for Evaluatin.q Solid Waste (EPA SW-846) from the U.S. EPA which covers sampling and analytical methods for a wide range of solid waste materials ASTM D346-78: The Collection and Preparation of Coke Samples for Laboratory Analysis. Documents sampling for Crushed or Powdered Materials. ASTM D1452-80: Soil Investi.qation and Samplin.q by AuQer Borings. Documents sampling for Soil-like Materials. ASTM D2234-86: Collection of a Gross Sample of Coal. Documents sampling of materials that are similar to bottom ash materials. ASTM E 122 - Choice of Sample Size to Estimate the Avera.qe Quality of .a Lot or Process; and,

175 ASTM E 177 - Practice for Use of the Term Precision and Bias in ASTM Test Methods Another important issue is handling the sample to obtain the necessary testing samples. This issue is covered in detail in: ISO 1988-1975 (E) Standard for Sampling Hard Coal. ASTM C702-87, Standard Practice for Reducin,q field Samples of AQQre.qate to Testing Size CEN/TC 154/TG 5 AQqre.aates: Samplin.q and Precision Consideration of various aspects of sampling aggregates is also addressed in the ZWL (1992) document that supports the CEN Aggregate Standard. Many of these sampling protocols utilise similar approaches and terminology when referring to collecting portions of the stream. Although the following general outline of the issues provides a common framework for discussing examples of sampling programs suitable for various residue streams and processes, it is recommended that further details be obtained from a direct review of the above-mentioned documents. 6.4 SAMPLING CONSIDERATIONS

To obtain a representative sample, it is normal practice to collect a definite number of portions, known as increments, from the whole mass of the material being sampled. Depending upon the purpose of the sampling, the increments can be stored for either individual analyses or they can be combined in the field to produce a gross (or composite) sample over a certain time period, e.g. day, week, month, or year. The method of collecting the increment usually influences the representative nature of the sample. 6.4.1 Increment Collection Classification

The conditions under which individual increments are collected, and the methods of spacing the increments (both spatially and temporally) from the whole stream are classified according to the descriptions given below. These designations can be used for sampling specifications and for descriptions of sampling programs and sampling equipment. Conditions of Increment Collection - These conditions relate the main body of the

residue relative to the portion withdrawn. Five conditions are recognised:

176 Condition A (Stopped-Belt Cut) - A loaded conveyor belt is stopped and a full cross-section cut with parallel sides is removed from the residue stream. Condition B (Full-Stream Cut) - A full cross-section cut is removed from a moving stream of residue. Condition C (Part-Stream Cut) - A portion, not a full cross-section, is removed from a moving stream of residue. Condition D (Falling Stream) - Material is removed by inserting the sampler into a falling stream of the residue being sampled. Condition E (Stationary Sampling) - A portion of residue is collected from a pile, container, truck trailer, rail car, barge, or ship's hold. Condition A is the one that is most likely to produce a representative sample. However, owing to the disruptions that can be caused by stopping a conveyor belt, Condition A is mainly used as a reference method to calibrate other methods. Theoretically, Conditions B and C should give samples with the same representative nature as Condition A, however, in practice it is difficult to remove material from a moving stream without introducing some disruptions that might bias results. Condition D is often referred to as the second best method, however its use requires care since the material being sampled must be relatively uniform and equipment should be robust enough to withstand the forces involved. Achieving a representative sample with a stationary sampling program (Condition E ) i s energy intensive and usually requires heavy equipment.

Spacing of Increments_- The spacing of increments pertains to the distances between

sample points, and the time interval between increments. Three methods of spacing the increments have been proposed (ISO 1988-1975(E)):

systematic sampling - the increments are spaced evenly in time or position over the unit random sampling - the increments are spaced at random in time or in position over the unit stratified random sampling - the unit is divided by time or quantity into a number of equal strata and one or more increments are taken at random from each stratum. From a statistical point of view (to ensure the highest precision and thus the most representative samples), random sampling or stratified random sampling should be used. The position of sampling points or the intervals used for sampling are

177 determined by using random numbers. Stratified random sampling should be used only if the stratification can be determined accurately enough, i.e. if the distribution of chemical contaminants in a waste is sufficiently known to allow an intelligent identification of strata and enough samples can be taken from each stratum. If this condition is not fulfilled, stratified random sampling is likely to give lower precision than simple random sampling. Many times, however, a proper random sampling can be difficult to carry out because of difficulties such as accessing the individual stream or too short a time interval within the process. Therefore, many random sampling plans turn out to be systematic random sampling in which the position and/or the point of time for the first sample is selected randomly and all subsequent samples are taken at a fixed space or time interval. Systematic sampling is usually chosen as it is easier to operate on a routine manner than random sampling. Systematic sampling can give samples of less precision, especially if the samples are collected in such a manner that cyclic variations have a great influence on the quality of the residue. Thus, it is essential that the influence of cyclical operations on the proportions of different residues be considered when sampling combined ash streams. Types of Increments - The types of increments are based on whether or not there is human discretion in the selection of the residue portions sampled. Type I - Specific pieces or portions are not subject to selection on a discretionary basis. This includes increments collected in precise accord with previously assigned rules on timing or location that are free of any bias. Type I selection increments generally yield more accurate results. Type II - Some measure of human discretion is exercised in the selection of specific pieces of residue of specific portions of the stream, pile, or shipment. 6.4.2 Bias Bias is defined as a tendency to obtain results which are consistently too high or too low. This may happen very easily during sampling. These conditions are often difficult to detect and care should be taken to prevent this occurrence during sampling. The two major causes of such bias are: selecting an unrepresentative part of the stream for sampling; or removing an unrepresentative portion of the stream when sampling. As noted earlier, sampling a pile consisting of particles with a wide size distribution is difficult because the material tends to segregate as a function of size and angle of repose. Therefore, it is necessary to collect increments from both the outside of the pile and the internal part of the pile.

178 The size of the device being used to collect the sample can influence the nature of the sample collected as well. For instance, using too small a sampler will automatically reject particles larger than the diameter of the sampler from the collected mass. Typically, bottom ash contains some ferrous material that is larger than 150 mm in size, and by using a 150 mm wide sample thief to collect bottom ash samples will result in selectively rejecting all the larger materials discharged from the hearth. In most cases, this larger material is composed of ferrous metals and relatively inert materials which are typically recorded and discarded from further analysis anyway. Consequently, this limitation would not be considered a severe detriment to the accuracy of the data. To eliminate the potential for bias, both the size of the sampling equipment and the mass of the increment should be in accordance with the maximum dimensions of the particles in the stream being tested. This means different criteria can be applied to bottom ash, heat recovery ash and APC residue streams. Bias can be improved by changing the sampling equipment or moving the sampling location, however, precision and accuracy of the sampling cannot be improved by increasing the size of the increment (ISO 1988-1975(E)). Similarly, while precision may be altered by adjusting the number of increments, this will have little, if any, effect on bias. The level of bias should be checked regularly, or at least when a new sampling program or new equipment is employed. A check procedure is provided in ISO 1088-1975(E).

6.4.3 Precision Equation 6.6 illustrates a relationship between the variability of sampling results and the number of samples required to ensure satisfactory results. Although a standard for sampling aggregates and coals (ISO 1988-1975(E)) suggests the precision should be within 10% for low values and 2% for higher values of ash or moisture, these values are derived by having a large base of data upon which to develop the requirement. Since only limited data are available for a wide variety of tests, it is not possible to develop precision guidelines for each residue type or each test. Rather, it is necessary to rely upon the guidance provided by various standards and an intuitive assessment of what is required to determine the number of increments to collect.

Number of Increments in Composite Sample ASTM D 2234 defines the objective of incremental sampling as the formation of a representative sample from the lot for analyses. This representative sample would have been within 1/10 of the average value of all the determinations for the parameter of interest for 95 out of 100 such samples. The standard also notes the number of composite samples that are collected defines the potential accuracy of the results. To reduce the error by a factor of two requires squaring the number of composite samples. Similarly to triple the accuracy takes 9 times as many composite samples. The ultimate is to analyse each increment, a situation that should provide one fifth of the error that would arise from analysing 1 sample composed from 24 increments.

179 ISO 1988-1975 (E) suggests the number of increments for sampling hard coal can range from 16 to 64 depending upon where the sample is taken and the type of coal. Conveyors provide a greater chance of sampling the full cross-section of the material and require fewer increments than stockpiles. Similarly, the variability in the material, i.e. washed versus unwashed coal, is reflected by the need to double the number of increments for the unwashed material. With bottom ash from MSW incinerators, experience suggests it is satisfactory to collect 24 increments although a minimum of 10 samples has been deemed to be more practical. Statistical relations should be used in all cases to verify that the number of increments taken is appropriate. If there is prior knowledge of the characteristics of interest in the residue, statistical relationships can be used to estimate the number of samples required. These estimates should then be confirmed using the results of the testing. 6.4.4 Size of Increments

The number and weight of the increments required for a given degree of precision depends upon the variability of the residue, and the parameters being tested. While the number of increments is a function of the precision required, the size of the increment is more a function of the size of the particles being collected. Several standards suggest some form of Equation 6.7 be used to determine the appropriate increment size for a given size of materials. Regardless of the result of applying the equation, most standards also specify minimum increment size requirements. Typically 0.5 kg is used as a minimum criterion. ISO 1988-1975(E) recommends a minimum increment size of 10 kg for materials >150 mm in size. This would be considered appropriate for an increment of wet bottom ash, which in practice may be 5 to 7 iitres in size.

P (kg) = 0.06 D (mm) where

(6.7)

P = the mass of the increment D = the nominal upper size of the material, i.e. the square mesh screen size such that no more than 5% of the material in oversize.

The ISO standard offers further guidance on sampling large size material. It suggests that the proportion of >150 mm material be determined for several increments and then the number of increments be adjusted to ensure that the >150 mm fraction is sampled in proportion to its mass in the overall stream. These large sized increments, which will weigh well in excess of the nominal increment weight, should be size reduced to <80 mm and then mixed and subdivided to reduce the mass to the required increment size before the increment is added to the composite sample. The standard also cautions that collecting larger than necessary increments can lead to problems with sample handling and reduction to the appropriate analytical sample size. Conversely, it notes

180 that under no circumstances should the number of increments be decreased if the mass is too large. 6.4.5 Collection Procedures

In order to obtain complete representation of all size fractions within an ash stream, it is preferable to have the sample increments withdrawn from the full cross section of the stream. This is accomplished using Condition A and D outlined earlier. The increment must be distributed throughout the lot to be sampled. This distribution is related to the entire volume of the lot, not merely its surface or any linear direction through it or over it. The spacing of the increments should be varied to ensure the sampling does not correspond to one single part of any cyclical operation which occurs in the facility. For instance, sampling should not occur every time fines are discharged to the bottom ash system from the boiler soot-blows or vice versa. This would bias not only the chemical composition of the samples, but also the particle size range for these samples. Similarly, if the boiler and bottom ash from the facility are mixed before the sampling point, ensure that within the range of samples collected, the number of increments includes an appropriate amount of the fine material that comes from the soot blow cycle. When sampling from moving streams, the sampling device should be designed to minimise disturbance of the residue, thereby avoiding separation of various residue densities and sizes, or both. Again the ISO 1988-1975(E) standard helps to define some important factors. When sampling from any stream, it recommends the minimum opening of the sampling instrument should be 2.5 times the upper size of the materials. This implies the minimum width of a stopped belt cut, moving stream cutter, and sampling scoop for trucks or stockpiles should be 2.5 times the material's upper dimension. As a note of caution, it does not recommend sampling material larger than 80 mm from a moving stream, therefore the largest opening size would be 200 mm. Several cautions have been noted previously on the need to ensure the size distribution of the samples is not biased. This is important when sampling the bottom ash from a vibratory conveyor. Vansant (1991) notes there is an increasing preference in North America for these devices on the bottom ash discharge since they are rugged and easy to service. The vibratory nature of the conveyor tends to cause both the segregation of certain size materials and the agglomeration of fines across the conveyor. This segregation continues into the ash bunker. Sampling the material off the conveyor is tedious and interrupts the plant operating cycle since the whole width of the conveyor would need to be carefully cleaned. Sampling from the bunker requires careful consideration of the potential for biased distributions.

181 All sampling equipment should be designed and used in a manner to prevent contamination of the samples so no biases are introduced into the results. Every sample must be collected carefully following the established procedure. As noted earlier, the errors introduced by poor samPling cannot always be corrected by repeating the work. In selecting the appropriate sampling procedure, local knowledge and input are also imperative. These provide an understanding of the operational variability of the furnace so compensating steps can be taken in the design of the program.

6.4.6 Sampling Streams Other Than Bottom Ash The nature of residues from other parts of the system (ie. heat recovery and air pollution control residues) are sufficiently different that many of the concerns outlined above are not as relevant. The materials are more homogeneous and consist of a much narrower size distribution based mainly in the sand or clay range (i.e. <2 mm). This reduces the variability and allows reduction of the size and number of both increments and composite samples required to obtain representative samples.

Grate Siftings and Heat Recovery Ash It should be recognised there is a potential for boiler ash to significantly change the characteristics of relatively large amounts of bottom ash. Clearly, the proportion of boiler ash in the total quantity of residue generated from a facility is <5% for mass burn systems, and much less for two-stage systems. For example, a recent study determined a 250 tpd mass burn incinerator generated approximately 150 kg of boiler ash per hour (WASTE Program, 1993), whereas only about 1 kg of boiler ash was generated over 12 hours in a two-stage combustion unit (Environment Canada, 1985). An issue that must be considered is the potential for the residue from these sources to accumulate in various sections of the incinerator/boiler/economiser during the operating cycle. The staff at the Burnaby EFW facility reported that during shutdown it is necessary to clean the grate siftings chutes to remove solidified materials from the inside surfaces of the hoppers. Similar experiences were observed during testing of the Connecticut Resource Recovery Facility where aluminum solidified in the screws used to remove siftings from under the grate. Boilers and economisers need to be cleaned on a periodic basis to remove solidified material from tubes and surfaces inside the units when heat transfer capabilities drop. In essence, the quantity of residues which are retained within these separate systems is highly dependent on the type of incinerator and the type of waste being burned. A thorough review of the maintenance problems and operating experiences will assist with making decisions regarding how to sample these residue streams. Therefore, boiler ash streams should be segregated from the bottom ash streams to ease the sampling requirements. In many cases, the addition of boiler ash to the bottom ash is not a continuous operation and the cyclical nature is difficult to compensate for in the sampling strategy.

182

APC Residues The quantity of APC residues from most mass burn systems represents about 10% of the total residue stream. Generally, given the fine and more homogeneous nature of this material, fewer samples collected by systematic sampling on a fixed time basis are adequate. However, modern plants tend to use enclosed transport systems that operate on a timed cycle and thus selecting the location and timing of sampling requires some preliminary investigation. For example, some dry lime injection systems recycle a portion of the spent residue to save on lime consumption. During operation the recycled lime is mixed with fresh reagent in the reactor and as it proceeds through the system, is collected in the fabric filter area. The material collected from the fabric filter is returned to the recycled lime storage bin where it is stored for up to 24 hours before it is re-injected into the reactor. Lime particles can go through the circuit up to 3 times before they are removed from the system and sent to the APC residue storage silo. The material that is bled from the circuit is generally removed from the base of the reactor. A possibility exists under this operating scenario for samples to be tainted by operational upsets that occur up to 24 hours prior to the sampling. In considering a sampling program, it may be appropriate to collect samples from this stream over longer periods of time to ensure all possible factors are assessed. Storage Piles In some instances, residue piles may need to be sampled and these piles present a unique problem for sampling. In forming the pile, segregation of the different size fractions can occur either due to density, size, friction angle or the angle of repose of the material. In considering conical piles, it is important to recognise that 70% of the mass is in the lower third of the pile, 26% in the middle third and only 4% in the top third. For tent-shaped piles, these proportions become 56%, 33% and 11%. When sampling such piles the samples should be taken in a proportional manner. If, on the other hand, the material in the pile has segregated, care should be taken to sample from each of the segregated areas in proportion to its quantity. Practically however, representative samples will be difficult to obtain. As noted elsewhere, sampling of piles should extend through the whole mass. During sampling, care should be exercised to avoid potential segregation of the material. The draft CEN/TC 154/TG 5 guidelines suggest driving a plank of wood or a steel plate into the pile above where the sampling is to be carried out and taking the sample immediately below the plate. The guideline also suggests the sample should be taken from a depth of at least twice the upper material size to ensure a representative sample, whereas the Swedish recommendation is at least 30 cm depth (SGI, 1993). Consideration of these issues relies heavily upon experience, however, scientific evaluation of the issues before initial sampling, and modification of the process after analysing the initial samples will allow collection of relatively representative samples from storage piles.

183

Sampling from Trucks or Containers In the case where an ash conveyor discharges directly to containers or trucks, it may be necessary to establish a systematic pattern for sampling the container. ASTM D346 provides guidance for sampling coke from rail cars. It suggests that, because there is a potential to segregate material by size in this situation, the surface is not the appropriate location to sample. A sample should be taken from 300 mm below the surface, and the hole to reach this location should be at least 3 times the diameter of the largest material in the shipment. In Sweden, it is generally accepted that the hole should be 5 times the diameter of the largest material. Given the practicality of these issues, it is recommended that a nominal diameter of 300 mm should be suitable for most bottom ash streams. Sample locations are not selected at random but rather at fixed locations based upon given measurements of the container. Divide the container length into 10 equal segments and divide the width into 6 equal segments. The intersection of the lines denoting l/6th, 1/2 and 5/6ths of the width and each length segment designate a potential sampling point. In a systematic fashion select sufficient sampling points to obtain the required number of increments.

6.4.7 Sample Preparation Concerns After samples are collected, it will be necessary to package them and ship them to the laboratory for processing. Care in sample handling is just as important as care in sample collection.

Sample Size Reduction In most cases, the volume of material collected is far too large to send to laboratories for analysis. Field methods need to be used to reduce the volume of the sample to a workable size. This is the first step in sample preparation. The amount of sample required for laboratory analyses depends upon the tests that will be run and the quantity of material required for each test. In most cases, no more than 20 L of bottom ash is required to undertake the testing. Correspondingly smaller quantities of the finer residue streams are required. The exceptions to this rule of thumb are special tests that require large quantities of materials such as column or lysimeter leaching tests. In this case, the total composite sample volume might be required. Suitable adjustments can be made in the procedures to allow for larger sample requirements. The samples collected in the field can be composites of the various increments or the increments themselves. Assuming the bottom ash sample is a composite formed of a number of increments, it will be necessary to reduce it in size to obtain the laboratory sample. It is recommended that the sample be divided by hand in the field. The procedure to be used follows that outlined in ASTM D346. (Figure 6.3)

184 Figure 6.3 Example of a Manual Sample Dividing Technique

..-.__

~lill'IIll~'lI'rl~ i|1

_

~

_

--

B

--

f

ii

A

i

250 Ibs. (113 kg) crushed to 3/4"(19 mm) and coned

Mix by forming long pile A- Spreading out first shovelful B- Long pile Completed

--.._.__

-..._

f

--.__

ill llll

Halving by alternate shovel method. Shovelfuls 1,3,5 etc reserved as A; 2,4,6 etc rejected as B

Adapted from ASTM D346-78

A

B i

i

Long pile divided into two parts; A- Reserve; B- Reject

185 Place the increment in a conical pile on a clean surface. Utilising a shovel, form a long pile as follows: Take a shovelful of residue from the conical pile and spread it out in a straight pile having a width equal to the width of the shovel and a length of 1.5 to 3 m. Spread the next shovelful directly over the top of the first shovelful, but in the opposite direction, and so on back and forth, pile, occasionally flattening the pile until all the residue has been formed into one long pile. This material can then be divided into two piles in the following manner: Beginning on one side of the pile at one end, and shovelling from the bottom of the pile, take one shovelful and set it aside; advance along the side a distance equal to the width of the shovel and take a second shovelful and discard it. Continue the process, retaining alternate shovelfuls, and reducing the long pile in a uniform manner on one side and then the other. This should produce a sample pile approximately half the size of the original pile. Continue the long pile and separation procedure until the sample is reduced to the appropriate volume.

Preservation of Samples

The issue of sample preservation is important. All environmental samples require some type of preservation if measured values are to be truly representative of the material. This is most important for wet bottom ash which is known to change its characteristics when it is stored. Indeed, as mentioned in Chapter 5, this is one of the reasons for the bottom ash storage requirements in European protocols directed toward the utilisation of bottom ash. Drying the samples slows these transformation processes, but can only be satisfactorily accomplished in the laboratory. Samples, after they have been reduced in size, must be stored and shipped in appropriate conditions.

Sample Containers

Containers can be either glass or plastic to limit the potential for interaction between the residue and the container, however, for residues that could potentially contain strong bases or HF, glass should not be used. Moreover, plastic containers are more durable. Irrespective of type, all sample containers should have tight fitting lids. Previously used containers should at least be washed with 10% nitric acid solution and rinsed with distilled water prior to use to avoid contamination from the previous sample. The hygroscopic nature of the heat recovery and air pollution control residue streams require that these samples be placed in dry, air tight containers immediately after

186 collection. Care should be taken in handling these materials to minimise the sorption of moisture during processing.

Sample Storage

Generally, the ideal storage temperature for ash samples is 4~ since most reactions that will change residue characteristics are limited at this temperature. Should mercury analyses be required, it may be preferable to oxidise the sample prior to storage to prevent the mercury compounds from reducing and volatilising. One important note however, these samples should be stored in glass containers and the oxidation preservation should be carried out in an ice bath to cool the container since the reaction between the oxidising acid and the residue will be highly exothermic.

Laboratory Sample Preparation

The normal analytical test series conducted on residue samples can be divided into three unique sets of data: physical, chemical and leaching. Preparatory to any of these analyses, the samples need to be processed to obtain a representative sample of the appropriate size for the specific analysis procedure. It is important to remember the sample size criteria related to particle size outlined earlier in this chapter should still be considered when attempting to reduce the volume of material to the laboratory sample. As a general rule, the finer the material, the smaller the sample required from the bulk sample.

Laboratory Sample Subdivision

Since bottom ash is known to change its chemical and leaching properties if it is stored wet, a drying step is introduced early in the protocol to stabilise the material for testing. Assuming a 20 L sample weighing on the order of 25 - 30 kg is sent to the laboratory, a large drying facility would be required to achieve reasonable turnaround for analysis. It is likely preferable to extract a representative sub-sample for drying (@ 105~ early in the process for chemical analysis (except Hg). Storing the balance of the sample in a refrigerated sealed container for several weeks before drying should not seriously affect the physical properties. Although conventional sample splitting devices such as whole or divided stream rotary sample dividers or riffle boxes are suitable for dry residue streams, they are unsuitable for dividing the wet bottom ash samples. Conventional splitters should be employed following the provisions of ISO 1988-1975 (E) or other appropriate standards. For the wet samples alternatives such as coring or fractional shovelling following the procedure in CEN/TC 154/TG 5 N 106 E, should be used. Coring should be accomplished by inserting a trier sampler or tube of at least 125 mm internal diameter at a slight angle to the vertical until it touches the bottom of the pail. The pail should then be tipped at an angle sufficient to allow

187 the probe to be withdrawn without loss of the sample. The sample extracted in this manner should be dried as described below. The fractional shovelling method may prove to be simpler to accomplish. Given the maximum particle size of the material a sample of approximately 3 kg is required according to equation 7. Given a mass (m) of approximately 30 kg in the laboratory sample, one tenth of the total would be required to provide the analytical sample. Assuming 10 sub-samples (n) will be created from the laboratory sample, a scoop capable of holding m/(30n) of the material should be used. The scoop should thus hold approximately 100 g of sample. The scoop should have a width of 125 mm and could be fashioned from a half section of pipe. The length of the scoop should be at least 100 mm to retain the large sized material that might be present in the laboratory sample. Take shovelfuls from the laboratory sample and add them to each of the n subsamples in turn, until the whole of the laboratory sample has been used. Using random numbers select the sub-sample or sub-samples required for chemical and leaching analyses. The balance of the sub-samples should be stored separately in sealed containers until such time as the laboratory has time to dry them. These samples may be required for physical testing purposes and storage time and transformations are not as critical for this testing.

Drying

To facilitate easier manipulation of the ash samples for further size reduction and subsampling, it is recommended to dry the ash. The laboratory sample should be placed in an open pan and put in a vented drying chamber to allow it to dry at 60~ overnight. Dry the material to a constant weight at 60~ noting the final weight of the sample. Please note the 60~ temperature should not be substituted for determining the moisture content of the sample, or for preparing a dry sample for submission for chemical analysis. For analytical purposes, these samples should be dried at 105~ prior to submission.

Size Reduction

Laboratory analyses are conducted on substantially smaller samples than that contained in the dried sub-sample discussed in the previous paragraph. Before a smaller sample can be created, it is necessary to complete size reduction on the dried sample. It is appropriate to remove ferrous material from the dried sample before size reduction using a magnet since milling this material could damage laboratory equipment. Weigh and note the amount of ferrous material removed.

188 The remaining ash should then be screened to <9.5 mm. Material trapped on the screen after shaking and stirring should be placed on a hard surface and subjected to three blows from a 3.5 kg. hammer dropped from approximately 0.3 m above the surface or alternatively passed through a standard jaw crusher. Material that will still not pass the 9.5 mm screen after attempting size reduction should be weighed, characterised and discarded. Typical sample preparation procedures are outlined in Figure 6.4.

Balance of Materials Although the previously mentioned steps have been used to obtain representative subsamples from the mass of material collected, there may be questions about the results of the analyses. To ensure sufficient material is available to repeat any analyses if required, the bulk samples should be stored for a period of 3 months after the test results are made available. During field sampling procedures, it is worthwhile to consider collecting and storing additional sub-samples of the total increment in the event materials are lost or contaminated in transit to the laboratory. 6.5 SAMPLE COLLECTION RECOMMENDATIONS Based upon a consideration of the issues in the preceding sections, recommendations for sampling of the various residue streams from MSW incinerators have been developed. These recommendations represent an ideal sampling situation for each of the residue streams, however, it should be recognised that sampling in this manner may not be possible in every case. Limitations in the ability to apply these recommendations can lead to samples that are not truly representative and the user must make a judgement that weighs practicality and necessity to arrive at the best solution for each case. Examples of successful sampling programs are provided later in this chapter.

6.5.1 Generic Bottom Ash Testing Protocol Ideally, each composite sample of bottom ash should consist of 24 increments, however, 10 increments should suffice. ISO 1988- 1975(E) suggests a minimum of 10 kg for materials whose maximum particle size exceeds 150 mm and this is deemed satisfactory for bottom ash. The procedure that follows will allow at least two samples to be collected in a working shift. The sampling of bottom ash should follow these steps: Identify the safest and most accessible location for obtaining samples of bottom ash

189 Figure 6.4 Schematic of Residue Collection and Sample Preparation

BOTTOM ASH COLLECT24

10 kg INCREMENTS FROM STREAM

ISCREEN I <50mm I ~ /

~, IWEIGH &l I RECORD/ [REJ~CTS/

........

\

AIR POLLUTION CONTROL SYSTEM ASH

COLLECT TOTAL STREAM

0.5 kg INCREMENTS FROM STREAM

TWO6 kg SAMPLES SEALEDIN CLEAN CONTAINERS

TWO6 kg SAMPLES SEALEDIN CLEAN CONTAINERS

........

COLLECT24

DISCARD

I TWO 20LSAMPLES

SEALEDIN CLEAN ----1 CONTAINERS ; . ~

STORE

~ i ~ ! ~ i A ~

ATLANT

DRY& CRUSH SAMPLE

GRATE SIFTINGS AND/OR HEAT RECOVERY SYSTEM ASH

STORE AT PLANT

SHIP TO LABORATORY

STORE AT PLANT

SHIP TO LABORATORY

SU3-

I ANA S'S I

190 Construct a sampling device (trough, bucket, shovel, thief etc.) to be used to gather a grab sample of the entire width of the belt conveyor, drag chain flight, vibrating conveyor or entire depth of the hopper, pile or truck load. Gather 10- 24 incremental samples weighing approximately 10 kg each. The sample should be collected from the entire width of the conveyor, etc., at the beginning of every ten-minute period for four hours. If trucks are to be sampled, then sample every other truck during the four-hour period, taking sufficient increments from each truck to achieve the desired number of increments. Screen the sample to minus 50 mm without employing any size reduction processes at this time. Examine the oversize material to characterise the type of components present (large pieces of glass, tin cans, steel etc.), then separate the oversize into two portions: metal and clinker. Weigh all three portions of the sample. This allows metal cans or other large material to be excluded from further examination. Mix the undersize fraction of the screened material thoroughly and subsample using procedures outlined in ASTM D346 or the previous section to obtain two 20 litre containers of ash. Discard the remaining material. Seal and label, both containers with the wet ash samples. Retain one at the facility and ship one to the laboratory. Place the oversize clinker material on a large clean hard surface and reduce it by dropping a 3.5 kg hammer on it from approximately 0.3 m height, up to four times. Screen the material through 50 mm mesh and place the minus 50 mm material in a separate labelled and closed container for shipment to the laboratory. Characterise and weigh the balance before storage at the plant. Note: the intent of the size reduction step is not to reduce all material to the required size regardless of form, rather it is to establish which portions of the sample may be largely composed of metallic materials that cannot be size reduced. Store the samples in an appropriate manner until shipped to the laboratory. This sampling procedure will result in the collection of between 100 and 240 kg of sample which will be reduced to a total mass of approximately 50 - 150 kg for subdividing and two sub-samples will be created. The generic procedure indicates that 4 hours are satisfactory for obtaining one composite sample, thus it is possible to collect

191 a second sample on each day. Not all sampling programs follow the generic plan and the 24 increments may in fact be collected over a different time frame. Where possible, it is recommended that a second four-hour composite be collected during the course of the work day, preferably during a different shift from the first composite. Weekly samples may be desirable for some purposes. Such weekly samples can be collected by developing a predetermined schedule for randomly selecting a shift and a starting time within that shift so that two samples are collected on each day and a total of fourteen samples are collected during the week. Monthly samples would be collected by randomly selecting one day per week and one shift within that day to collect the appropriate samples.

6.5.2 Generic Boiler Ash Sampling Protocol As noted earlier, this material tends to be smaller than 2 mm in size, although on occasion large chunks break off the boiler tubes. In most facilities, this material is transferred from the hoppers under the boiler to either the fly ash disposal system or the bottom ash disposal system. In either case, it is likely possible to intercept the flow of material from the boiler and collect all of the ash discharged during the test period. The recommended protocol assumes this is possible, and provides a method for subsampling this material. Caution is advised in handling this material as it can be extremely hot as it leaves the boiler and the fine particle size potentially poses a problem with fugitive dust in the sampling area. However, the methods used for generating a suitable laboratory sample should minimise the potential for the finely divided material being released to the air. The sampling of boiler ash should follow the steps below: Identify the most convenient location for obtaining samples of the ash to be evaluated Construct an appropriate diversion system to allow collection of all the material generated during the test period in a clean, sealed 210 litre barrel At the end of the test period, disconnect the barrel from the system, weigh, and move the barrel to a location suitable for sub-sampling Sub-sample the collected material by either the coring or the fractional shovel method discussed above. Each increment should contain approximately 0.5 kg of material. A total of 6 kg per composite sample should be created from 24 cores with alternate cores being used to create the two samples. Seal and label both containers with the boiler ash samples. Send one to the laboratory and retain the other at the facility

192 Coring should be accomplished by inserting a trier sampler or tube of at least 25 mm internal diameter at a slight angle to the vertical until it touches the bottom of the barrel. The barrel should then be tipped at an angle sufficient to allow the probe to be withdrawn without loss of the sample. The fractional shovelling method may accomplish the same results in some cases, but can give rise to significant dust levels. A scoop capable of holding approximately 250 g of sample should be used. The scoop should have a width of 50 mm and could be fashioned from a half section of pipe. The length of the scoop should be at least double its diameter. Take shovelfuls from the bulk sample and add them to each of 30 sub-samples in turn, until the whole of the bulk sample has been used. Using random numbers select the two sub-samples required for analyses. The generic procedure indicates that 4 hours are satisfactory for obtaining one composite sample, thus, it is possible to collect a second sample on each day. Timing issues are similar to those considered for bottom ash.

6.5.3 Genetic APC Residue Sampling Protocol APC residue volumes are considerably greater than those of boiler ash and it is not possible to collect the entire stream before sub-sampling. It is therefore necessary to collect increments over a fixed period of time to generate a composite sample for the test period. A minimum increment size of 0.5 kg is likely satisfactory for finely divided materials such as APC residues. It is recommended that 10 - 24 increments be collected. The resulting 6 - 12 kg of material can be divided into two samples, one for the laboratory, the other for storage at the plant. Caution is advised in handling this material as it is hot as it leaves the system and it is fine-grained. The methods used for dividing the sample into two parts should minimise the potential for releasing dust to the air. The sampling of APC residues should follow the steps below: Identify the most convenient location for obtaining samples of APC residue samples. Most systems use some means of conveying these materials from the hoppers under the particulate collection device to storage silos. Tapping into these transport systems is generally the most appropriate location for sampling. Construct a sampling device (trough, bucket, shovel, thief etc.) to be used to gather a grab sample of the entire width of the belt conveyor, or conveyor discharge chute. In the case of sampling pneumatic transport lines, the introduction of a manually controlled valve into the transport system piping will allow the discharge of material into a suitable, filter equipped collection barrel.

193 Gather a total of between 10 - 24 incremental samples at the start of every ten-minute period for four hours. Alternatively, if a pneumatic conveying system is to be sampled, open the valve for a fixed period of time every 10 minutes, and core the collected mass to obtain the composite sample. Divide the composite sample into two equal sized samples using coring, fractional shovelling or mechanical sample dividers. Seal and label both containers, one for the laboratory, the other for retention at the facility. 6.5.4 D o c u m e n t a t i o n of S a m p l i n g and P r e p a r a t i o n P r o c e d u r e s

The reporting of test results for residue samples should include: a complete description of the methods used to take the samples, including; - the type of increments - spacing of increments - number and weight of increments, and, - methods of creating composite samples. a description of the sample storage procedures, and, a description of laboratory preparation methods including any screening, separation or mixing procedures. To assist in developing a more useful database on the relationship of physical and chemical properties of residues to the various factors which influence them, it is suggested that the following information should be compiled (where practical) during an ash sampling program: Composition of the waste materials being burned, including the types of waste processes at the facility on the dates on which the sample was generated, and any waste composition analyses that may be available. Designation and description of the incinerator units and any other specific equipment within the facility which contributed material to the sample, including the heat recovery, pollution control and quenching equipment. Description of the operating conditions of the specified equipment prior to and during collection of the sample, including material throughput rates, material discharge frequencies, quench water flow rates, and gas stream temperatures.

194

6.6 EXAMPLES OF SAMPLING STRATEGIES To illustrate two methods of approaching the experimental design, this section provides examples of several protocols used for sampling residue streams. The examples cover all types of residues, except for wet scrubber systems, and are divided into regulatory testing protocols and research testing protocols to illustrate differences in the approaches. It is important to note that although it is possible to suggest general procedures for sampling, the exact protocol for any given facility and study cannot be outlined until issues of intent and safety are incorporated with the specific facility design and operating characteristics. As a general rule, consideration should be given to the potential influence of the various intermediate residue streams, siftings, boiler ash etc., to determine the best method for sampling. If the facility design permits segregation of the various residue streams, the sampling design will be simpler preventing the contamination of other streams. Moreover, regardless of the precise sampling protocol, good experimental design dictates all aspects of the operation be documented for the test period.

6.6.1 Bottom Ash Bottom ash can be collected in two ways, either quenched or unquenched. The normal process regime is to quench the ash and this is the form the material would take before discharge to the environment. Hence, testing quenched ash is more likely to be specified for regulatory purposes. The unquenched stream may be tested to examine the chemical differences induced by process variations, however, this is generally limited to research. The bottom ash sampling protocols that follow illustrate sampling of both the quenched stream and the unquenched stream. The former was completed for regulatory compliance testing (PRRI, 1992), the latter for the special testing associated with the WASTE Program (WASTE, 1993).

Regulatory Testing This particular sampling program (PRRI, 1992) was conducted for permit compliance purposes on a modular incinerator facility processing both raw MSW and a processed MSW fuel with the fines, glass and metals removed. The 100 tpd Consumat incinerators discharge bottom ash to a quench tank. The ash is removed via a drag chain system which conveys it up an incline allow free water to drain back into the quench tank prior to discharging the ash into a storage bunker. Ash stays in the bunker between 24 and 72 hours (depending upon the time of the week) before being removed with a front end loader and placed in a truck for haulage to a landfill site. The design of the facility precluded sampling the ash off the drag chain conveyor. Therefore, it was decided each bucket load from the front end loader would be sampled before the ash was placed into the truck. The following procedures were employed:

195 Each sample represented material from one segregated incinerator over the 24 hour operating period immediately preceding the sampling. As the ash was transferred from the storage areas to the haulage vehicle by the front-end loader, sampling personnel took two shovelfuls of ash from the centre of the front-end loader bucket and placed it into a 20 litre bucket. Ash samples were taken from each front-end loader lift. The average number of loader buckets needed to transfer the ash to the haulage vehicles was 12, thus 24 shovelfuls of ash were collected for each composite sample. This resulted in samples weighing between 110 and 200 kg. The use of this procedure excluded sampling large clinker type materials. However, one such mass was collected and several other samples of clinker that were more than 50 mm in diameter were isolated from the bulk samples and analysed. Sample preparation included screening the contents of each pail with a 50 mm screen composed of metal grating, separating the >50 mm material and metals into 2 different portions, and weighing all three portions. The <50 mm was then spread into a long pile and the pile was subdivided to provide two 30 kg samples. Another example of this type of sampling illustrates an approach to random sampling to assist with the statistical validity of the sampling for regulatory purposes (Rigo & Rigo (1989)). The basic premise of this sampling plan was to collect samples over a year long period. One front-end loader lift of ash contained approximately 1.8 tonnes of ash and the ash was hauled six days a week. The samples from any given day were judged to adequately represent a week's production. Thus, there was a need to select the day of the week for sampling and the individual lifts that would be sampled. Using a random number table, the day of the week was selected using the following procedure. 9

Each shipping day was assigned a range of numbers as follows:

Monday Tuesday

0-17 18-33

Wednesday 34-50 Thursday 51-67

Friday Saturday

68-83 84-99

A 2 digit random number was used to select the day of the week for each week of the sampling period. If the selected day fell on a holiday, the next number was used. The lifts to be sampled were determined on the sample day by: estimating the amount of ash in the bunkers and the total number of bucket loads that would need to be handled that day. the buckets which were to be sampled were identified by taking the decimal equivalent of the random number and multiplying it by the total number of loads, then rounding the answer up to a whole number.

196 24 random samples were then taken from the total number of buckets. This resulted in time periods which decreased between samples taken on days when less ash was generated, whereas the duration of an interval increased on the days with more ash produced. two shovelfuls of ash were taken from the centre of each of the designated buckets, or as close to the centre of the bucket as possible if too large a piece of material obstructed the penetration of the shovel. It was assumed since the front-end loader was digging through the pile, the sample met the criteria of a grab sample from the pile. Furthermore, the randomness of sample bucket selection satisfied statistical requirements without leading to problems associated with limitations on mixing the sample. This example could also be adapted to sampling combined bottom ash and APC system residues if necessary, as long as the cyclical operations of the heat recovery and APC system cleaning cycles are taken into consideration.

Research Testing

One example of a research program designed to assess the quality of ash was the WASTE Program (WASTE, 1993), where simultaneous sampling of the waste feed and residue streams from a 250 tpd mass burn incinerator was conducted. Several issues were considered in developing the testing program including the desire to isolate the various streams, i.e., separate grate ash, grate siftings, heat recovery ash, and the residues from the different sections of the APC system. Safety issues and other difficulties made it impractical to sample quenched ash from the vibratory conveyor leading from the quench tank to the ash storage pit. Moreover, the grate ash, grate siftings and heat recovery system ash streams could not be isolated during the entire test period, and since all these streams were mixed in the quench tank, cross contamination of the samples was inevitable. Consequently, it was decided to sample unquenched grate ash as it fell off the end of the grate. A 1600 cm x 1000 cm chute located at the end of the grate leading to the quench tank was chosen as the area to collect the grate ash. The inspection door on the side of the cute was modified by adding four access ports across the width of the door. These ports allowed a sampling device to be inserted across the width of the grate. Their positions provided access to all parts of the discharged ash stream. The variability in ash quantity and size distribution across the discharge plume was determined prior to start-up of the project tests. Preliminary tests indicated both the port closest to the end of the grate and the one furthest away collected very little material during a fixed sample period, whereas a sample thief inserted into the middle two ports gleaned nearly equal volumes of ash and the size distribution appeared to be similar.

197 The intent was to gather a sample from the full width of the chute at regular intervals. One grab sample was collected every 30 minutes during each 4 hour test period by inserting the sample thief (a section of 100 mm diameter pipe with the top half removed) across the entire width of the drop-off area and allowing the sample chamber to fill with ash. The thief typically filled within 2 to 3 minutes. The physical size limitations of the sampling ports precluded sampling large materials falling off the end of the grate. Chunks of material caught on the sample thief which were too large to be removed through the sampling port were pushed off prior to removing the sampler. A 25 cm wide by 2 metre long steel trough was constructed to aid sample collection. After the sample thief was extracted from the inspection door, the trough was placed underneath the sample thief to receive the collected ash. The trough was then used to transfer the sample into an appropriately labelled 4.5 L steel container, sealed, weighed and allowed to cool. The individual grab samples were then composited for each of the 4 hour test runs to generate a sample for use in the laboratory. Another research program was aimed at evaluating the utilisation of ash in which a stratified random sampling approach was used for a continuous flow. The Long Island study (LIRPB, 1991) used the following method of obtaining samples from a continuous stream based upon modified ASTM protocols for sampling aggregate material: Conveyor sampling every 15 minutes, during one randomly selected hour in both the morning and afternoon. The samples were removed from the conveyor to form a composite sample. By assigning each 15 minute segment of the period a value, a random number table could be used to select the starting time. However, various alternatives to these methods can be considered, including: 9

random selection of flights from a conveyor based upon the total number of flights that must be sampled to obtain enough material, or, randomly pick times during the discharge cycle to sample.

In many cases it is sufficient to systematically sample the streams. For example, a constant discharge cycle can be sampled by collecting an increment every 10 minutes to provide the recommended 10 -24 samples for each composite sample in 4 hours.

6.6.2 Grate Siftings During the WASTE Program, the grate siftings were segregated to determine potential differences between these and the bulk of the bottom ash, and to determine how much was generated. The entire quantity of grate siftings generated during each test period was collected and sub-sampled by modifying the grate siftings hoppers underneath the

198 grates with diversion gates which directed the siftings into a duct leading to vented 200 L barrels. Since the grate siftings hoppers at this facility are purged at regular intervals under pressure, the vent on the barrel was fitted with a fabric filter bag to dissipate pressure and minimise fugitive dust. Prior to hooking up the sample barrels for the morning test runs, the grate sifting hoppers were blown clear. The barrels were then attached to the diversion chutes, the diversion gates were swung into position and the blowers were set back to automatic. Just prior to the end of each test run, the hoppers were blown clear and the barrels removed and weighed. The quantity of grate siftings generated for each test period was monitored by measuring the volume collected and determining the density of the material. Initial attempts to obtain composite core samples of grate siftings from the barrels proved to be very difficult and time consuming due to the density of the siftings. As a result, a 20 L composite sample for each test period was obtained by tilting the barrels on an angle and removing the contents down the side of the barrels with a shovel. The sub-samples from the two barrels from each test run were composited in a labelled 20 L plastic pail, sealed and weighed.

6.6.3 Boiler/Economiser Ash Depending on the configuration of the heat recovery system, boiler/economiser ash falls into hoppers under the equipment and can be periodically removed for disposal. The exact operating methods will dictate the appropriate sampling interval. It is important to collect samples of the material generated with and without operation of tube cleaning equipment, and careful documentation of these cyclical operations in relation to the sampling is essential. Heat recovery ash is generally more homogeneous than grate ash, therefore, it is possible to collect smaller sample sizes without affecting the quality of the data generated. The total volume generated in any given period may be so small the sampler will decide to collect all of the available residue. Such was the case in both the regulatory and the research test programs outlined above. Since the mass of material collected in both cases was greater than that required for representative sampling, site methods were used to generate sub-samples for analyses.

Regulatory Sampling

The boiler/economiser ashes from the Peel facility were collected at the end of every operating period marking the completion of a given set of operating conditions. Collected material included ash that settled in the hoppers during normal operation and the material dislodged during operation of the soot blowers. The entire quantity of the boiler and economiser ashes generated over the period was collected in separate 200

199 litre barrels for each boiler. Full depth tube cores were taken from the drum to produce approximately 5 litres of ash. Care was taken to ensure the samples do not absorb moisture during the cooling process since the material is slightly hygroscopic.

Research Sampling

One of the objectives of the WASTE Program was to determine the partitioning of trace elements within the temperature regimes of the incinerator. Therefore, samples of ash were collected from hoppers #1, #2 and #3 via diversion chutes fitted to each of the boiler ash discharge hoppers. The entire quantity of ash generated in each hopper over each four hour test period was collected in 200 litre barrels. The soot blowing cycle for the boiler was initiated 30 minutes prior to attaching the barrels to the diversion chutes in the morning and again 30 minutes prior to removing the barrels after each test run. Typically, the diversion chutes needed to be cleaned out by hand prior to removing the sample barrels. This provided a means to monitor the quantity of boiler ash generated during each of the test periods, i.e., the volume of ash collected was monitored and the mass was calculating using the density of the ash. Sufficient full tube core samples for each four hour test period were taken to fill a 4.5 L steel container, sealed, weighed and allowed to cool.

6.6.4 Air Pollution Control System Residues This section is more relevant to residue from ESP's, and dry and semi-dry APC systems than to residues from wet scrubber systems. Typically, the quantities of ash generated are too voluminous to collect the whole stream, although attempts to monitor the total volume generated should be made to assist with later interpretation and calculations of mass balances. It may also be necessary to collect various samples from this stream to assess the potential variability in quality brought on by process changes in the system. Grab samples should be collected at fixed periodic intervals, however, the weight of each grab sample does not have to exceed 1 kg in weight. If it is possible to obtain samples from the various locations in the APC system, research sampling studies may attempt to identify speciation differences in the APC residues at different stages of the process.

Regulatory Testing

The APC residue from the Peel facility was sampled from the pneumatic transport line that connects the bottom of the baghouse hoppers to the recycle and waste silos. The extraction pipe assembly, located after the last hopper of the second baghouse, was equipped with a ball valve to allow for periodic sampling. Samples were collected in a 20 litre bucket and transferred to a 200 litre drum for compositing. The collection of fly ash samples from the APC was accomplished by opening the ball valve every half hour and filling the bucket with material. This procedure continued for

200 an 8 hour sampling period. Full depth tube cores were taken from the drum(s) to produce approximately 5 litres of ash.

Research Testing During the WASTE Program, grab samples from both the dry scrubber reactor and fabric filter were collected via a gate valve fitted to the Depac unit at the base of the respective hoppers. The Depac purge cycle was turned off during the test periods to allow residue to build up in the hoppers. Initially, attempts were made to collect separate dry scrubber reactor grab samples once every 30 minutes, however, since very little residue was captured, the sampling frequency was changed to once every hour. All the residue generated during each 1 hour period was purged and collected from the Depac. The fabric filter residue samples were collected from the Depac unit every half hour. Samples were subsequently collected by opening the valve and manually pressurising the Depac unit to fluidise the residue. The ash was collected in 4.5 L steel containers, then sealed, weighed and allowed to cool. REFERENCES ASTM D346-78: The.. Collection and Preparation of Coke Samples for Laboratory Analysis. ASTM D1452-80: .S0il Investigation and Samplin.q by Au.qer Borin.(:]s. ASTM D2234-86: Collection of a Gross Sample of Coal. ASTM E 122: Choice of Sample Size to Estimate the Av.eraqe Quality of a Lot. or Process ASTM E 177: Practice for Use of the Term Precision and Bias in ASTM Test Methods ASTM C702-87: Standard Practice for Reducin.q field Samples of Aq.Qre.qate to Testin~ Size ASTM D 3665: Standard Practice for Random Samplin~ o.f Construction Materials, Samplinq In-place Pavina Materials Boomer, Bruce A., Thomas P. Dux and Daniel J. March. Plannin.q and Conductin.q Samplinq Surveys of Hazardous Wastes at Industrial Facilities. JAPCA, Vol. 38 No. 11, pp 1426-1432, November, 1988. CEN/TC 154/TG 5 "Aggregates: Sampling and Precision" Eighth working draft for the European Standard for Aggregates: Methods of Sampling. European Committee for Standardization c/o DIN Berlin, Germany, 1991.

201 Environment Canada, "The National Incinerator Testing and Evaluation Program: Twostage Combustion (Prince Edward Island) - Summary Report", Conservation and Protection, Ottawa, Ontario, Report EPS 3/UP/l, 1985. Eighmy, T. Incinerator Residue Samplin.q and Analysis Plan for Permitted Residue Disposal. Prepared for Wheelabrator Environmental Systems and Signal Environmental Systems Concord Co., Inc. Published by the Environmental Research Group, University of New Hampshire, Durham, NH, March, 1989. Fiesinger, Tom, "Second Generation Incinerator Residue Sampling Plan (Draft)". Proceedin.qs of Ash Research Seminar May 14-15, 1987, New En,qland Center, .University of New Hampshire. Edited by T.T. Eighmy and J.A. Minischiello, The Environmental Research Group, Department of Civil Engineering, University of New Hampshire, 1988. Gy, P.M. Samplin.a of Particulate Materials; Theory Practice, Elsevier Scientific Publishing Company, Amsterdam, the Netherlands, 1982. Hartl6n, J. and P. Elander, "Residues from Waste Incineration - Chemical and Physical Properties", Swedish Geotechnical Institute, Linkoping, SGI VARIA 172, 1986. Ingamells, C.O. and F.F. Pitard Applied Geochemical Analysis, John Wiley and Sons, New York, 1986. ISO 1988-1975 (E) Hard Coal Sampling First Edition 1975-03-01 published by the International Organization for Standardization, Switzerland. Long Island Regional Planning Board The Potential for Beneficial Use of Waste-toEner.av Facility Ash Samplin.q and Testi.0R Procedures Report No. 2. Hauppauge, New York, January, 1991. Pitard, F.F. "Sampling of Solid and Liquid Wastes, Note from a Training Course for the American Chemical Society", Denver, CO, August, 1986. PRRI. Ash and Quench Waster Testing Report Peel Resource Recovery Inc. Ener.av from Waste Facility. Brampton, Canada. July, 1992 Rigo & Rigo Associates, Inc. "Appendix J (Revised 3/31/89) Ash Sampling, Preparation and Reporting Procedures". An appendix to Performance Test Protocol for the South East Resource Recovery Facility, Long Beach, California. Berea, Ohio. March, 1989. SGI. Samplin,q and Characterization of Residual Products Interim Report, Part I Samplin.q of Residual Products Swedish Geotechnical Institute Report No. 1-253/92 Link0ping, Sweden, 1993.

202 Vansant, Carl. "WTE Equipment: Big Gains in a Short Time". Solid Waste & Power. Vol. V Number 33, June 1991, pp 12-16. US EPA SW 846 "Test Methods for Evaluating Solid Waste: Methods", US EPA Publication SW 846, July 1982 as amended.

Physical/Chemical

The WASTE Program. Effect of Waste Stream Characteristics on MSW Incineration: The Fate and Behaviour of Metals. Mass Bum MSW Incineration (Burnabv, B.C.) Final Report of Phase I of the Waste Analysis, Sampling, Testing and Evaluation (WASTE) Program. A report prepared by the Consortium of A.J. Chandler & Associates Ltd., Compass Environmental Inc., Rigo & Rigo Associates Inc., The Environmental Research Group of the University of New Hampshire, and the Wastewater Technology Center, 1993. ZWL. Final Report VC-26 Phase 1 and 2. Samplin.q of Granular Materials A report prepared by Zuidelijk Wegenbouw Laboratorium, Esscheweg 105, The Netherlands in support of CEN/TC 154/TG 5 Aggregates: Sampling and Precision, 1992.

203 CHAPTER 7 - CHARACTERISATION METHODOLOGIES

Both the physical and chemical characteristics of MSW incinerator residues must be evaluated to develop a proper management strategy for the various residue streams, including utilisation of the bottom ash. Information on the physical and chemical properties is also required when considering alternatives for handling, transportation and minimising occupational health risks. Since the physical properties of incinerator residues is highly variable, the following section provides an outline of the types of test methods available to determine key properties for evaluating disposal and utilisation scenarios for ash. 7.1 PHYSICAL TESTING 7.1.1 Visual Observation

Visual observation is used to define the relative quantities of recognisable materials in ash samples, coupled with parameters such as grain shape and grain size distribution. Recognisable materials in bottom ash include glass, ceramics, combustibles (char), ferrous metal, non-ferrous metals, minerals and ash. The proportion of each component provides an indication of the non-combustible fraction of material present in the waste fed to the incinerator and the potential extent of processing required to render the ash suitable for use. The data can also be used as a tool in assessing the impact on the quality and quantity of the ash resulting from recycling of ferrous, nonferrous and glass components within the solid waste stream. Furthermore, the visually identifiable combustible content provides a gross assessment of the relative efficiency of the combustion process. Fly ash and other flue gas cleaning residues are so fine-grained it is not possible to identify the individual particles. The noticeable parameters are colour, texture and the general appreciation of the particle size. Method Although, there is no specified ASTM procedure for cataloguing visual observations, the following section provides some guidelines. Bottom Ash Visual observation can be subdivided into two or more grain size ranges. The plus 50 mm fraction may be visually classified in the field during the sampling of the material. The material is then sieved through a 50 mm screen, and the individual components not passing through the screen are manually removed, weighed and categorised (i.e., ferrous, non-ferrous, brick, ceramic, motors, etc.). The minus 50 mm sample can be analysed in the laboratory in a dry state. Particles smaller than about 2 mm are difficult to visually identify.

204 The ferrous content can be determined by slowly passing a magnet over oven-dried ash. The ferrous content is expressed as a percent of removed weight related to the total dry weight of the ash prior to ferrous removal. Visual classification of bottom ashes from two incinerator plants in Sweden is given in Table 7.1 and show the following distribution, in the 2-16 mm fraction [Jacobsson & H6beda (1988a) and Jacobsson (1989)]. Table 7.1 Main Constituents of Bottom Ash in Maim6 and Link6ping Distribution of Components in Bottom Ash by Weight (%) Content

2-4 mm

4-5.6 mm 5.6-8 mm 8-11.2 mm 11.2-16 mm Total

Malmi~ Metal-magnetic

18.2

16.9

9.1

14.5

5.6

10.2

non-magnetic

58.0

38.6

30.7

18.0

13.1

22.5

Glass

23.1

41.6

55.4

64.2

73.3

62.0

Ceramics

0.0

0.4

2.2

2.8

5.0

3.2

Stone

0.7

0.7

1.2

0.0

2.6

1.4

..,.P.aper

0.0

1.8

1.4

0.5

0.4

0.8

Metal-magnetic

10.0

4.8

11.6

2.2

5.1

8.5

non-magnetic

67.0

51.4

33.2

23.1

25.2

54.8

Glass

20.1

39.9

50.1

69.9

57.4

32.6

0.0

1.9

3.4

3.7

10.2

1.7

1.2

2.2

2.4

Link~ping

Ceramic

Paper 2.9 2.0 1.6 Jacobsson & H6beda, 1988a; Jacobsson, 1989

205 Fly Ash and APC Residue Visual observation primarily concerns the colour and texture. For example, the colour of fly ash can range from a light brown through grey to almost black, whereas APC residues can range between white and dark grey. Darker coloured ash generally indicates poor combustion efficiency. Textures can range from a "dusty" fine-grained powder to agglomerated chunks. It is also a good idea to check the friability of the ash, since may have implications on interpreting the grain size distribution results. For example, the grain size distribution of agglomerated materials, such as some boiler ashes, may be much finer than indicated by the grain size distribution determination, and subsequent handling of the material may break the agglomerated chunks into much smaller pieces if the material is friable. 7.1.2 Particle Size Distribution

The sieve analysis test is used to determine the size distribution of the aggregates and is a suitable method for bottom ash. The grain size distribution gives the percentage by weight of different sizes of the particles, which can also be used to assess other physical properties such as shear strength, bearing capacity, permeability, workability, dusting and frost susceptibility. In general, well-graded materials (i.e., contain an even gradation of size fractions from coarse to fine) tend to be relatively stable, resistant to erosion, and can be readily compacted to a dense condition with a high bearing capacity. The strength generally also increases as the maximum size of the material increases. Materials deficient in fines will usually be less stable, despite compaction. The uniformity coefficient is defined as the ratio of the diameter of the particle size at the 60% fines fraction, to the diameter of the 10% fine fraction, and is sometimes used as a measure of the relative distribution of particle sizes in a soil sample. The material becomes frost-susceptible when the content of fines reaches a level high enough for capillary sorption of water to occur. In soils, this level is reached when the percent of minus No.200 sieve fines (<0.075 mm) exceed 10-15%. Bottom ashes seem to become frost susceptible at about the same level. Test Methods The test method selected is dependent on the type of residue. For coarse grained material, standardised sieve tests are generally adequate. This usually involves drying, screening and weighing the ash. The quantity of material retained on each sieve is weighed to determine the fraction of the size distribution. Sometimes the grain size distribution may be incorrect as finer particles stick to more coarse-grained particles or agglomerate due to electrostatic charges, although the latter primarily occur with fly ash samples. When the content of fines is substantial, washing the ash prior to sieving can be useful, although it should be noted this may result in solubilisation of some components in the ash.

206 The gradation curve is a graph of the percent weight passing a specified sieve size, plotted on the ordinate, and particle or sieve size, plotted logarithmically on the abscissa. The curve can be used to compare materials to specifications or to assess the gradation characteristic of the material. A steep slope indicates a poor gradation which is unfavourable for most engineering purposes. A gentle slope indicates a good gradation. Data can also be presented in the form of frequency by particle as well as by weight.

Dry Sieve Methods The standard test procedure (ASTM C136) is recommended for material sizes larger than the No. 200 mesh sieve (75 pro). Wet sieving, ASTM C117, is recommended for materials containing >10% particles <75 pm in size. After washing, the material is dried and sieved according to the dry sieving method. However, when a material containing high levels of salts (such as APC-residues) is washed, the grain size distribution is significantly altered, and therefore ASTM methods C136 and Cl17 are not recommended for APC residues or fly ashes. Fine Particle Analyses Methods As mentioned previously, particle size determination of fine sized particles may result in agglomeration or "bailing up" of particles when sieved through smaller sieves (100200 pro). Some of the static which causes this agglomeration can be eliminated by grounding the sieving device. For finer particles, test procedures such as sedimentation tests (ASTM D 422-63, DIN 18123, BS 1377:1975 and DIN 66115; or balance body method SS 027124) are not recommended since the ash particles can dissolve, whereas granulometric methods are appropriate for highly soluble ash matrices. Although ultrasonic devices can be used to disintegrate the agglomerated material, different solutions such as hydrogen peroxide, hexane and acetone are typically used. Grain size distribution of grains less than 200 IJm can be determined by dry methods such as the laser granulometer method used in the cement industry or by projected area diameter (Hinds, 1982).

7.1.3 Density Density is an important parameter when considering the homogeneity, porosity of the particles, the bulk pore volume, as well as the degree of compaction. The density is also used to determine if a mass is in a loose, medium or dense state. Specific procedures are followed to measure either the loose (uncompacted) or dense (compacted) state. The uncompacted density state is measured by pouring the material gently through a funnel into a known volume and weighing the material filling the volume. The compacted density is measured by compaction (see Section 7.1.6).

207 The term density involves a group of parameters, such as bulk density, dry density and specific gravity. The dry density classifies the individual aggregates as lightweight, normal weight or heavyweight. The dry density is used to evaluate the degree of compaction related to a compaction test (see Section 7.1.6).

Bulk Density

The bulk density expresses the mass per unit volume. The value includes solid particles, water (if present) and air. The dry density is defined by weight of material, divided by its original volume before drying. The dry density of a material reflects the void content and specific gravity of the particles. A test method is given in ASTM C 29. The dry density of a material is determined by drying a specified volume of a sample in 105~ in 24 hours or until constant weight is reached. The volume of the container depends upon the nominal maximum size of the material to be tested, i.e., the recommended sample size is 5 times greater than the nominal maximum particle size.

Specific Gravity

Specific gravity is defined as the ratio of the weight of a given volume of a sample to the weight of an equal volume of water. It is a dimensionless number. There are different types of specific gravity measurements. The purpose of the measurement is to evaluate the compactness (density) of the individual particles. Specific gravity provides an indication of the minerals, voids in the particles and existence of noncombusted material. It can be measured using a multi-volume pycnometer with helium gas.

Laboratory Testing

The bulk specific gravity (ASTM C 128) is a measure of the relative weight of the dry particle to the weight of a volume of water that includes the solid particle volume and the volume occupied by the internal pore space within each particle. The specific gravity, bulk specific gravity, and saturated surface dry (SSD) measurement (ASTM C 127 and C 128) is a measure of the relative weight of a wet aggregate (i.e., absorption has been satisfied) to the weight of a volume of water which also includes the solid particle volume and the volume occupied by the internal pore space within each particle. The apparent specific gravity (ASTM C 127, C 128), is intended to measure the actual or true specific gravity of the dry particle itself. It is the relative weight of the solid material making up the particles to the relative weight of an equal volume of water. The water volume in this measurement excludes the volume of the pore space within each particle. If there are any permeable pores in the sample, the apparent specific gravity

208 will have higher values than either of the two bulk specific gravities. Consequently, when evaluating the use of ash in construction applications, such as roads, it is recommended to determine the apparent specific gravity value to avoid potential problems (Eighmy et al, 1992).

Field Testing Density often has to be controlled in the field through compaction. For example, when residues are used in road construction, the degree of compaction has to exceed a minimum value. Another example is a sealing layer of natural soil is placed at a disposal site. The density is, to a large extent, the controlling factor limiting the permeability. The lower the density, the higher the permeability, and vice versa. The in-situ density can be measured by several methods, such as the sand-cone method (ASTM D 1556), the drive-cylinder method (ASTM D 2937), the rubber balloon method (ASTM D 2167) and by nuclear methods at shallow depths (ASTM D 2922). However, these methods are generally not applicable when the material contains particles coarser than about 40 ram. 7.1.4 Absorption Test The absorption test predicts the change in weight due to water absorbed in the pore spaces within the constituent particles. The result is based on the weight difference between the wet and dry ash. With bottom ash, those aggregate materials with higher absorptive capacity can also be expected to exhibit greater susceptibility to freeze-thaw weathering. The pores in lightweight aggregates may not satisfy the absorption potential during the period of this test (i.e., 24 h). Therefore, recorded 24 hour results for certain lightweight aggregates may be questionable.

Test Method A test method is presented in ASTM C 127 for coarse aggregate and in C 128 for fine aggregate. Typically, the material is immersed in water for a period of 24 hours prior to measurement. The absorption is normally only applicable for bottom ash. The values can change depending on the content of fines, and it may be of value to measure the absorption of both the coarse and the fine fractions to compare the difference. 7.1.5 Water Content Water content (moisture content) identifies the mass of free water which can be evaporated when exposed to a temperature of 105~ The water content affects the maximum density obtained by compaction in field and thus the internal stability and stiffness. Therefore, water content is an important parameter with respect to disposal

209 of ash, as well as in different utilisation applications. It must also be controlled within specified limits for selected construction applications such as Portland cement concrete and asphaltic concrete applications. For fly ash and APC residues, the water content is related to the amount of water used to limit the risk from fugitive dust emissions during transport. Normally, the water content must exceed 15% (geotechnical). On the other hand, water contents at levels higher than about 30% often results in a fluid material which may be difficult to handle. Care should be taken with the desiccation of lime scrubber residues, since they contain CaCI 2 which is hygroscopic and will create variable results depending on ambient humidity. Test Methods

Moisture content is determined by measuring the weight of material before and after drying at 105~ to a constant weight (ASTM D 2216). If parameters (especially mercury) require later analysis, the temperature of 105~ may cause volatilisation of the Hg, and consequently, the sample should be discarded after testing. The geotechnical water content (used in geotechnical engineering) is defined as the weight of the water divided by the weight of dry material, expressed as a percentage. The environmental water content (used in environmental, water quality, and solid waste management) is typically defined as the weight of water divided by the total wet weight of the sample. Thus:

Geotechnical water content wg=(weight watecweight dry material) Environmental water content we=(weight water:total weight wet material) Typically, the water content of an ash/residue is dependent on the ash handling equipment at a given facility. The types of quenching, drainage and storage prior to sampling can have a significant impact on the amount of moisture retained in the ash. Water contents of 30-45% in geotechnical terms (25-30% environmental) are commonly reported for bottom ash (Schmidt, 1984; VEABRIN, 1988; Hartl~n & Rogbeck, 1989; Chessner, 1990; Eighmy et al, 1992;). It is estimated that bottom ash requires a minimum of 17% moisture to prevent wind-borne fugitive dust emissions. 7.1.6 Proctor Compaction Test

The Proctor Compaction Test defines the relationship between bulk density and water content achievable within a specified compaction effort. This is achieved by

210 determining the moisture and compaction requirements of a soil-type material that will result in the maximum density of the material when compacted in the field. It is desirable, from an engineering viewpoint, to compact a material to a dense state to decrease future settlement, to increase shear strength and to decrease permeability. The water content in the ash affects the density that can be achieved. For example, too little water will inhibit compaction, whereas increasing water contents act as a lubricant and aids in compaction. However, too much water makes the material more fluid and elastic, preventing compaction. A greater applied compaction effort will generally result in an increase in soil density and a lowering of optimum water content. Data analysis generally includes a graphical plot of the water content versus the dry density. The density at the peak of the dry density curve is called the maximum dry density, and the water content at the peak density is termed the optimum water content. Many compaction specifications require a percent of the maximum density be achieved in field applications. The degree of compaction is the ratio of the density required in the field to the maximum density determined in the laboratory, expressed as a percentage. The required degree varies from 85 to 100% for granular soils as well as fine grained silts and clays. Standard Proctor

The Standard Proctor method (ASTM D 698) specifies the use of a 5.5 lb. (2.49 kg) rammer, dropped from a height of 12" (305 mm) onto a sample contained in a collared soil test mold. The rammer is dropped on the sample 20 times in a series of 3 "lifts" of the sample. Modified Proctor

The more commonly used Modified Proctor method (ASTM D1557) is similar to the standard method, but uses a 10-lb. (4.54 kg) rammer with an 18" (457 mm) drop. A series of samples is then prepared with increasing weight of water. After the sample is compacted in the mold, the dry density and the water content of the sample is determined. Widely varying compaction curves have been obtained for the same residue streams from different incinerator facilities, even though the configuration of the system is the same. This is probably a function of waste feed composition and the type of operating conditions. Consequently, this test should be conducted not only on samples from separate facilities, but also on series of samples from the same facility. It is often difficult to perform compaction tests on fly ashes. This is often revealed by the compaction curve, which may not show a defined maximum value. Moreover, the

211 powder-like nature of fly ash makes it impossible to compact at low moisture contents. Moisture contents below about 10-15% tend to result in the ash simply being driven out of the mold during compaction. The Standard Proctor test will result in lower maximum densities and higher water contents achieved at optimum compaction, due to the lower compaction energy. In addition, the modified proctor may crush ash particles due to the extra force exerted by the heavier rammer. Comparisons between laboratory and field compaction results indicate the standard laboratory proctor test is the most applicable test to estimate field compaction curves (Hartl~n & Rogbeck, 1989).

7.1.7 Strength and Strength Development Strength parameters are used in different circumstances, such as determining the maximum inclination of a landfill slope, the bearing capacity of a fill built up with ash, or the durability of cementitious material. The strength of coarse grained materials is dependent upon the friction between the particles, whereas the strength in fine grained materials primarily builds up by cohesion and cementation forces. This means different test methods should be used for bottom ash than those for APC residues and stabilised residues. The strength of aggregates, such as bottom ash, can be determined by direct shear tests or triaxial tests to evaluate the cohesion and angle of friction of the material. This type of test is not often made on bottom ashes as the knowledge of the grain size and its distribution may be enough to determine the strength. The angularity of the bottom ash particles results in a rather high angle of friction (35 to 50). However, the angle decreases with increasing stress level. Hence, ash subjected to high stresses (above 200 kPa) should be tested in the laboratory. In special cases, there may be a need to determine the shear strength parameters in detail. In these cases, the tests are performed using standard geotechnical equipment. As mentioned earlier, it is important to use a sample diameter five times the maximum particle size. It is also important to determine the strength of cementitious materials. Cementitious reactions that could occur are not only complex, but can change the character of the ash material with time. The most commonly used test is the uniaxial compression test (ASTM D 1633) on compacted samples. By performing the tests at different time intervals after compaction, the strength development can be determined. The pozzolanic type of reaction can be evaluated following test procedures allocated for "real" pozzolanic materials (ASTM C 311). Sometimes the tests are made on samples immersed in water (ASTM C 109 mod.). A test method to evaluate probable cementitious reactions is outlined in ASTM D 1557. The sample is compacted, sealed in plastic bags, stored and tested for compressive strength at intervals up to 28 days, in accordance with procedures outlined in ASTM 1633.

212 The uniaxial compression strength of a well-compacted sample can, under favourable conditions, reach 1 MPa. After stabilisation with additives, such as cement, it may be several MPa. It is important to record any swelling activity, since this generally indicates the ash strength will decrease over time.

7.1.8 Bearing Capacity The Californian Bearing Ratio (CBR) test is a special test used to evaluate the ability to meet road base specification criteria and to assist in defining the bearing capacity. The CBR test is a measure of the shearing resistance of a sample or aggregate material to that of a standard sample. CBR values of base, sub-base and sub-grade materials have been correlated to pavement performance. These correlations have resulted in criteria permitting the use of CBR as a strength and stability measurement. These correlations have been converted to design curves that can assist in the determination of requisite pavement thickness as a function of anticipated traffic and CBR values. The test procedure is outlined in ASTM D 1863. The sample is initially molded to the maximum density and optimum moisture content as determined from the compaction test (i.e. ASTM D1557). The molded CBR sample is subsequently soaked in water for a period of 4 days. The initial soaking is intended to simulate a condition of high water content encountered during the spring season. After 4 days, the soaked sample is taken from the water and placed into a load testing machine. A square loading piston is mechanically pushed into the sample at a defined rate, and the load at specified penetration distances are monitored. The CBR is usually selected at a 0.1" (2.54 mm) penetration, however, the highest CBR value is used, whether it is a 0.1" (2.54 mm), 0.2" (5.08 mm), or higher. The limited data base on CBR values for bottom ash generally show higher values than those of most natural soil materials, and hence fulfill the minimum specifications for use of the material as a road base. However, more laboratory data on ash are required to ensure the comparison of natural soil and ash is statistically valid. Another test to measure bearing ratios is the Swedish Earth Bearing Method (SEB). It was developed by the Swedish Road and Traffic Research Institute (VTI), and is outlined in their Report No. 31, 1973. The SEB method is similar to the CBR method, however, the load is only 0-500 kPa and is, according to Swedish experience, more relevant to incinerator residues.

7.1.9 Durability Durability testing is used to estimate the physical resistance of an aggregate to various weathering conditions. Soundness and freeze-thaw tests define freeze-thaw resistance, where as the LA abrasion test defines durability of ash particles.

213 Soundness Tests The sodium or magnesium sulphate test is a used method of soundness testing. The test procedure may follow ASTM C88 and involve testing for 5 cycles. After drying, salt crystals grow in the permeable pores and cause particles to disintegrate. After the last cycle, the sample is washed to remove all the salt and screened dry. The average loss of material from each specified sieve is used in calculating the soundness loss.

This test is considered very aggressive and may reject materials that in fact will perform well. Aggregates that readily break down during the test must therefore be carefully evaluated based on the deposition environment before being rejected. Typical results suggest most bottom ashes are not susceptible to freeze-thaw cycle breakdowns. ASTM C33 prescribes a maximum of 12 % loss when sodium sulphate is used and a maximum of 18 % when magnesium sulphate is used. The wet/dry weathering test covers procedures determining material loss caused by repeated wetting and drying of monolithic samples. It also covers the visual observation of disintegration of solid samples. The following test procedure is suggested by Stegemann (1991) for solidified ashes. Samples should be molded in disposable plastic molds, each having a 44 mm inner diameter and a 74 mm length, and then vibrated on a vibrating table or tamped/compacted with a compaction hammer to provide adequate settling in the mold. Samples molded for these tests require curing in a moisture chamber for 28 days. Reporting of data should include the water content of the specimens, the average cumulative, corrected relative mass loss after 12 cycles, the number of cycles survived (if the specimens did not survive 12 cycles of testing), and the results of visual observations after each cycle (physical deterioration). LA Abrasion Test The Los Angeles Abrasion Test measures the degradation of aggregates resulting from a combination of actions including abrasion, impact and grinding in a steel drum. The test method recommended is ASTM C131.

A 5000 g sample is placed in a drum with a charge of 6 to 12 steel balls which is tumbled for 500 revolutions. The percentage of wear is the percentage loss in weight during the test, and is that measured weight of material which will pass a standard No. 12 sieve (1.70 mm). Ash tested to date exhibit losses ranging from 10 (extremely hard) to 90 % (very soft). In the US the maximum acceptable loss for a material to be used in a road base course is 60 %. Eighmy et al (1992) considers this test to be overly aggressive for pavement applications.

214

Freeze-Thaw Tests This test covers procedures for determining material loss produced by repeated freezing and thawing of monolithic samples. It also covers the visual observation or disintegration of solid samples. Stegemann and Cote (1991) proposed a test procedure for solidified/stabilised waste material which was used during the solidification trials of APC residues under NITEP (Sawell & Constable, 1993). The test involved exposing cylindrical shaped samples to repeated cycles of freezing and thawing. Sample preparation is similar to the one used for wet/dry tests. The Swedish Traffic and Road Institute (VTI) has developed a modified freeze-thaw test procedure for residues (H0beda & Jacobsson, 1988b) in which the sample (diameter of 50mm and height of 100mm) is stored for a specified time, then saturated with water under a vacuum. Swelling is recorded based on a uniaxial compression test. Reporting of data should include the water content of the specimens, the average cumulative, corrected relative mass loss after 12 cycles, the number of cycles survived (if the specimens did not stand 12 cycles of testing) with the results of visual observations after each cycle (physical deterioration).

7.1.10 Permeability The coefficient of water conductivity (coefficient of permeability) is a parameter that explains how pervious a material is to water. It is essential to know the permeability when predicting the leachate flow in a disposed ash and forecasting the frost susceptibility and the drainage properties of an ash. The coefficient of conductivity, k, is defined as the ratio between the speed of transport and the hydraulic gradient. When the coefficient is higher than about 10.4 m/s the material is considered self-draining, whereas coefficients lower than 10.9 m/s are considered impervious (Sj0holm et al, 1994). Parameters that influence the coefficient of permeability are (Head, 1982): particle size distribution, particle shape and texture, mineralogical composition, void ratio, degree of saturation, fabric, nature of fluid, type of flow and temperature. To these factors, effects as cementation and specimen size may be added.

Test Methods Testing can be performed in the field or in the laboratory. Different types of laboratory equipment exist and the choice is dependent primarily upon the permeability. In low permeable material, the equipment shall have flexible walls to prevent leakage and to admit a procedure to saturate the sample. This pertains specifically for self-hardening residues. The test can be performed with falling or constant water pressure head.

215 It is important to ensure the permeability of the filter stones used is more than 10 to 100 times higher than that of the material tested.

Tube Permeameter

The tube permeameter utilises a sample with e.g. a diameter of 101 mm and a height of 124 mm contained within a rigid wall. This equipment is used primarily for granular materials, which are compacted in the cell, normally to a 90% degree of compaction. The following parameters are normally chosen: a gradient of 15, water saturation with back pressure for up to 1 day, an upward flow direction during testing and preferably a constant head. The total experimental time is about 48 hours (Sj6holm et al, 1994). This type of equipment is sensitive to coefficients of permeability between 10.2 - 10~ m/s.

Triaxial Cell Permeameter

Triaxial cell with flexible membrane is used to limit the risk of seepage along the circumference of the sample due to channelling. (Figure 7.1 ) The diameter is normally 50 mm and the height 30-200 mm. Larger diameters, up to 150 mm, can be used. This test type is especially suited for fine-grained, cohesive or stabilised material. The cell pressure to be used depends on the material tested, but may never be more than 200 kPa. It is recommended the gradient not be more than up to 30 and an upward flow direction be used. The time for saturation under back pressure is 1 to 10 days and the total experimental time will be 2 to 13 days. Saturation is an important consideration, since unsaturated samples will generate widely variable results. This is especially true for low permeable ashes, not the least for solidified samples (Sj6holm et al, 1994). This method is sensitive to coefficients of permeability between 10.6 - 1011 m/s. If a solidified waste is not monolithic, ASTM D2434 is recommended for determining the hydraulic conductivity of granular soils by a constant-head method. Stegemann (1991) has proposed the following procedure for monolithic samples. A triaxial cell is used which will accommodate a 3" (76.2 mm) high, 3"-inch (76.2 mm) diameter sample with a rubber membrane to enclose the sample. The test is carried out 28 days after sample preparation, using a monolithic cylindrical sample. The flow direction is downwards.

Rigid Cell Permeameter with Varying Sample Diameter

Undisturbed samples are taken by drilling in field for performance control. The sample is then enclosed in a rigid tube. In the laboratory, the same tube is used to reduce

216 Figure 7.1 Triaxial Cell Permeameter

INLET RESERVOIR

AIR PRESSURE

L m

m

GRADUATED METER

DE.AIRED WATER

m

I

n

~

l

~

/kH PRESSURE GAUGE

GRADIENT)

(

\ ~ VALVE j r (CELL-

IC

OL

PRESSURE)

RUBBER SHEATHIN~ -SAMPLE CELL

WATER

6

q

MEASURING BOTTLE

217 disturbance. The diameter of the sample may vary due to sampling disturbance. Any space between the sample and the wall should be filled with sealing material, such as bitumen or silicon-caoutchouc. Sample diameter may be 50-200 mm and height 30-300 mm. The same procedure given above for triaxial cell permeameter can be used.

In Situ Methods Laboratory testing of samples normally underestimates the permeability due to heterogeneities, variation in degree of compaction and variation in water content. Moreover, there is a need for field testing to monitor in-situ quality, using a double ring infiltrometer (ASTM D 3385-88) which is limited in application to materials with a coefficient of conductivity between 10 7 - 10.8 m/s. However, there is a need for methods that can be used for impervious material. Because of that, methods such as "The two staged borehole" have been developed (Boutwell & Tsai, 1992). In this method, the measurements are made in a 100 mm diameter borehole. Single ring infiltrometers have been developed. The problem is the flow is not very well defined and one has to consider both the vertical and the horizontal permeability. At the Swedish Geotechnical Institute, a test procedure has been developed, Figure 7.2. The equipment consists of one cylinder (inner infiltrometer ring) with stiff walls and with lid and stand pipe attached. The ring is placed on the surface of the material to be tested and loaded by weights. Outside this inner ring an outer ring is placed. Between the two rings, bentonite is placed and also loaded to utilise the swelling pressure of the bentonite. Water is filled into the infiltrometer ring and the stand pipe. The equipment is suitable for material with a hydraulic conductivity in the range of 10.5 - 10 1~ m/s. The calculation assumes full saturation. The hydraulic coefficient, K (m/s), can be calculated using the following two different equations. Assuming one dimensional flow according to Darcy modified to consider horizontal flow gives:

K =F

ARL In( Ht" +L A,(t(n+,)-t,) H,o" +L )

where F = correction factor, see Figure 7.3 = area of stand pipe AR & = area of infiltrometer L = thickness of layer = pressure head at time t (m) Ht t = time (s)

(7.1)

218 Figure 7.2 Single Ring Infiltrometer

II II

,E~ 16.5

STAND PIPE

SURCHARGE

;4,

I--'-'!

~ . ; ~t,:; ~;11 BENTONITE SEALING

295 INFILTROMETER

8oo

OUTER RING

SjSholm et al., 1994

SEALING TESTED LAYER

[mm]

219 An evaluation following Hvorslev (1949), assuming a half spherical flow, the equation becomes: rid2 In Ht" K 8D(t(n+l)-tn) (~t,+)

(7.2)

where d D Ht t

= = = =

stand pipe diameter (m) area of infiltometer (m 2) pressure head at time + (m) time (s)

Both these equations should be used parallel until further experience is achieved. Normally, they give results in the same order of magnitude. Remarks There are potential concerns to be aware of when carrying out permeability tests. The most important factor is to ensure the sample is saturated. In a triaxial cell permeameter, this can be done by using back pressure. Back pressure means both cell pressure and pore pressure inside the sample are increased to a high level simultaneously to solve the air. When using rigid wall permeameters, this procedure cannot be followed and the saturation not checked. This means rigid cells should not be used for ashes with a coefficient of conductivity lower than 10.8 m/s.

There is always a risk plugging or clogging may occur due to migration of small particles into interstitial pores, hence reducing the apparent hydraulic conductivity. To counteract this effect, the water shall have an up-flow direction. If the permeability decreases over time, the reason should be investigated to see if clogging or cementation are responsible. The coefficient of permeability decreases with increasing content of fines. This means heavy compaction may change the properties. Thus the compaction must reflect the compaction method to be used in field. Fly ashes, as well as APC-residues (fly ash mixed with flue gas cleaning residues), often show strength increase with time. As the strength increases, the permeability decreases. On the other hand, if the APC-residue is mixed with water and stored for some time before compaction, the permeability may be higher (F,~llman et al, 1989).

220 Figure 7.3 Correction Factor F

0.8

0.6

kx /k

;/...I.l

=1

0.8

0.6

0.4

/./

kx /ky = 10

/ 0.8

0.6

J

.

X / /

.

0.4

0.2 1 -DIH

kx /ky =100

0.5 T/H

Day and Daniel, 1985

221 The coefficient of conductivity is, as mentioned earlier, very dependent on the degree of saturation and thus of the water content. In field it is difficult to control the water content. This might be the main reason why large differences have been noticed when comparing field and laboratory data (Bryant & Daniel, 1985).

7.2 CHEMICAL COMPOSITION Some of the controversy over MSW incinerator ash characteristics is based on the variable chemical data generated on the different residue streams. In Chapter 6, the importance of exercising sound sampling practices to obtain representative samples was emphasised as a means to minimise the effects of the heterogeneity. In this section, the importance of following sound laboratory procedures to determine the chemistry of the residues is stressed as a key to better defining the characteristics of ash.

7.2.1 Sample Preparation With the exception of loss on ignition, all of the chemical tests covered in this section require particle size reduction of the various residue streams prior to testing and analyses. Reduction to less than #100 mesh sieve size (<149 IJm) is a minimal requirement for most analyses, whereas less than #200 mesh (74 pm) is preferable. The reason for recommending the finer particle is to facilitate more efficient chemical reactions by increasing the surface area to volume ratio of the ash particles, and minimise the sample heterogeneity by finely dividing and mixing the samples. In the case of the bottom ash and grate siftings residue streams, the mineral and silica content of the ash presents some unique problems.

Size Reduction Techniques Selection of size reduction techniques are highly dependent on the type of ash being processed and to a minor degree, perhaps the type of incinerator system generating the ash. The particle size data given in Figure 6.1 indicates bottom ash and grate siftings streams generally contain much larger sized particles than the heat recovery or air pollution control system residues and hence require a greater degree of processing than the finer sized residue streams. Bottom ash from RDF combustion facilities may contain a greater proportion of moderately sized particles, however, good laboratory practice dictates proceeding with the same sample preparation techniques used for the coarser bottom ash. After the samples of bottom ash or grate siftings have been sieved through a 7.5 mm mesh to screen out the reject material, it is recommended the sample be examined to remove ferrous (e.g., nails, paper clips, etc.) and non-ferrous metals (e.g., foil wrap, coins, etc.) since these materials typically cannot be size reduced. After removing

222 these malleable materials, the samples should be partially size reduced by passing the ash through a commercial grade laboratory jaw crusher or hammer mill. The crushing plates of the jaw crusher should be manufactured of a material which is capable of handling the tenacious mineral/silica based matrix of the two ash streams. Another sound laboratory practice is to avoid placing too much stress on the crushing device. Samples should be passed through the unit in succession, stepping down the aperture size after each pass to achieve the lowest particle size. After the ash has undergone the initial particle size reduction, final size reduction can be achieved by milling the ash in a motorised mortar and pestle. The mortar and pestle should be made from a suitably hard substance, such as agate or other such mineral, to be able to withstand the abrasive nature of the mineral matrix of the ash. Ceramic, or other types of material can be used, however, they tend to wear more easily and thus are not as effective as agate mills and may result in contamination of the sample. Furthermore, it is recommended the ash be ground in small batches. During the grinding procedure, processed ash should be screened through the required sieve size. Ash not passing the required sieve should be returned to the mill and ground. This should be repeated until all of the sub-sample will pass the proper sieve size. Of the remaining types of ash/residues (heat recovery and APC residues), the heat recovery ashes and perhaps residues from the early stages of the APC system should be ground in a motorised mortar and pestle. Conversely, APC system residues, especially fabric filter residues from mass burn or two-stage systems typically do not require grinding prior to testing and analyses, since the bulk of the particles are sized well below #100 and even #200 mesh sieve sizes. However, the type and age of incinerator should be taken into account, since older incinerators and RDF systems may carry over substantial quantities of larger sized particles than two-stage or modern mass burn systems, and hence should be processed accordingly. It should be noted careful sample preparation will only minimise potential variability due to the heterogeneity of the sample. Results from a Dutch study on carefully prepared bottom ash samples indicated reproducibility between sub-samples of prepared ash was difficult to achieve for trace elements such as lead, copper, cadmium and tin (Mammoet Project, 1990). It was speculated the analysis could be greatly influenced by fine metallic particles of these elements. One means of evaluating the reproducibility of results is to plot the concentration of a measured parameter versus the percent relative standard deviation of the results (% RSD). By overlaying the residual error created by the detection limit of the procedure, and the lower limit of determination (level of 10% precision), data falling outside the residual error zone indicates heterogeneity, whereas those falling inside the zone are reproducible results (Figure 7.4). The position of the DTL and the LLD can shift along the X-axis depending on the complexity of the matrix of the sample tested.

223 Figure 7.4 Statistical Interpretation of Analytical Precision 9 Lower Limit of loo J L

~

--.Detecti~- . .

,Limit

Determination

~NDomain of Heterogeneity

cO 00

(D E~ "O

ormal

Working Range

L

"o r 00 (o (9 n,

10

',l, ~

Csolid

. ~ Residual Error

or

J"

Cliquid

7.2.2 Inorganic Analyses Digestion Techniques

In an attempt to verify results, researchers compare their own data with data from other studies. As discussed in Chapters 9 - 11, total metal concentration results can vary widely, sometimes even between sub-samples. Typically, the inconsistencies are attributed solely to sample heterogeneity, especially in view of the fact that a very small quantity of residue is used for analytical purposes, perhaps as low as 0.1 grams. However, another factor which can strongly influence results is the sample digestion technique used. Although the US EPA provides a guidance document for choosing the best procedure for various types of samples (Chapter 2 of EPA's Test Methods for Evaluating Solid Waste, 1986), extra care must be taken in selecting the most appropriate procedure for digesting incinerator residue samples. One good example of how different digestion techniques can influence total concentrations is the experience gained during the first two NITEP studies. Subsamples of residues were submitted to two different laboratories for total metal analyses using two different digestion techniques (Environment Canada, 1985 and 1986). The results are presented in Table 7.2 in the form of a ratio of the concentration measured after an aqua regia digest versus the concentration measured after hydrofluoric acid/aqua regia/peroxide digestion (HF/AR/P) (see descriptions given below). The arithmetic mean values and standards deviations are provided. Values less than 1 indicate greater yield from the HF/AR/P digestion, values close to 1 indicate no difference and values greater than 1 indicate potential volatilisation of the metal during the HF/AR/P digestion.

224 Table 7.2 Comparison of Concentration Ratios Generated by Two Different Digestion Techniques on 24 Sub-samples of Ash Applicability Metal

Mean

Std Dev

Aluminum

0.72

0.13

Aqua regia

HF/AR/P *

Barium

0.31

0.29

**

Boron

0.65

0.67

*

Calcium

1.09

0.35

*

*

Cadmium

0.98

0.44

*

*

Chromium

0.52

0.47

Cobalt

1.14

0.78

Copper

0.77

0.41

*

Lead

0.57

0.32

**

Nickel

0.49

0.17

**

Sodium

0.85

0.38

*

Tin

2.68

2.19

**

Vanadium

321.05

241.66

**

** *

Zinc 0.91 0.37 * * * = suitable method ** = recommended method Adapted from Environment Canada, 1985 and 1987; Bridle and Sawell, 1986; Sawell et al., 1987 Based on these results, it is evident the type of digestion method should be chosen according to the type of metal and matrix being analysed. For example, it appears better recovery was achieved through an HF/AR/P digestion for metals associated with the silicate matrix such as B, Ba, Cr, Cu, Pb, V, and Ni. Both types of digestion methods work equally well for calcium, cadmium, sodium and zinc, whereas cobalt, tin and vanadium requires less rigorous digestion to achieve better recovery. Similar results were observed during the Mammoet Project (1990) and the EPA study of stabilised incinerator residues (1993). There are several methods routinely used for digestion of MSW incinerator residues. The following discussion includes a brief description of the method and the qualification factors surrounding the interpretation of the data generated by the specific digestion method:

225

EPA Method 3050- Acid Digestion of Sediments, Sludges and Soils (US EPA,

19ee)

This method involves a nitric acid and hydrogen peroxide reflux followed by a hydrochloric acid reflux for those elements compatible with hydrochloric acid. It is considered a strong acid leach and although it is moderately aggressive, it should not be expected to produce analyses considered "true total" concentrations for MSW residues. It provides limited dissolution of metals bound in the silica matrix of these materials and is limited in its ability to dissolve stable oxides (refractory materials) formed during incineration. The hydrochloric acid step is applied mainly for the adjustment of oxidation state of arsenic and selenium, thereby making it an appropriate digestion prior to analysis by hydride reduction.

Aqua-Regia Digestion

This method involves reflux of the solid in a mixture of HCI:HNO3 at a ratio of 10:3. It is very commonly used for the digestion of MSW incinerator residue samples. Although it is considered more aggressive than the EPA 3050, it is still only classified as a strong acid leach and as such will not likely produce "true total" analyses of these residues.

APHA Standard Method 30301

This method is widely used for "total" type digestions of solid materials. It involves a nitric acid/perchloric acid/hydrofluoric acid medium. This highly reactive combination should provide total dissolution of MSW wastes. Certain metal species however, may not be recovered quantitatively when subjected to this digestion. Silica will be lost through volatilisation. Chromium may also be volatilised due to the highly oxidative nature of this digestion. Either chlorochromate ion (CrO3CI) or chromyl chloride (CRO2CI2) may form and be driven off at the temperatures used during the digestion. These losses may be exacerbated in high chloride content solids such as incinerator residues. Other elements which may be lost during this digestion are: Ag, As, B, Ca, K, Mo, Sb, Se and V (see Table 7.4).

Wastewater Technology Centre (WTC) Hydrofluoric Acid/Aqua-regia/Peroxide Method

This method is also commonly employed by other laboratories and has a similar performance to SM 30301 (WTC, 1993). The major difference is the use of hydrogen peroxide which helps to keep the chromium in the reduced +3 state thereby preventing the losses described above.

Specialty Methods for Specific Elements Silica

For refractory elements such as silica, fusion techniques are recommended to obtain a "true total" analysis. Two common methods exist. The first method involves using

226 lithium metaborate as the flux agent mixed at a solid to flux ratio of 1:9, then heated to 950~ in a graphite crucible for 30 minutes. The molten "glassy material" is then dissolved in 4% nitric acid. This method is also acceptable as a digestion method for most metals with the exception of As, B and Se. The second method is an alkaline fusion using a mixture of sodium hydroxide or potassium hydroxide, which is effective for preparing ash samples for As and Se. The sample is mixed with the salt base (12:1 salt:sample) in a Zirconium crucible and heated to 1000~ or until the mixture is completely molten. After cooling, the residue is dissolved in hot 10% HCI. The analyses generated by using either of these fusion techniques are limited by the fact that the amount of the sample that can be used is minimal, consequently the detection limits become elevated and analytical precision suffers.

Mercury

The quantitative recovery of mercury from incinerator residue samples is highly dependent on maintaining oxidising conditions throughout the sample preparation step. EPA method 7471 is perhaps the most commonly applied method and involves digesting the sample with aqua-regia in a heated bath at 95~ followed by the addition of potassium permanganate. The potassium permanganate is added primarily to oxidise chloride to free chlorine which must be removed before the analysis step. This method has generally provided good recoveries, although in samples with relatively high carbon content, there is a tendency towards lower recoveries. This digestion protocol is designed for sample weights up to 0.2 grams. WTC developed another digestion method for mercury to permit using a greater mass of sample (up to 0.5 grams) and thereby enhancing the reproducibility of the data. The method is significantly more reactive than the EPA method since it involves subjecting the sample to a nitric acid/sulphuric acid/vanadium pentoxide digestion medium at 160~ At the present time, it appears the data generated by either method on low carbon content ashes are comparable.

7.2.3 Analytical Measurement The chemical matrix of MSW incinerator residues is very complex, and the measurement of the individual elements within the matrix can be equally complicated. Analysis can be divided into two major categories, 1) Destructive Methods and 2) Nondestructive Analytical Techniques.

Destructive Methods Metal species contained in the digestates generated from the digestion techniques mentioned above may be quantified through instrumental analysis. Either atomic

227 absorption or atomic emission spectroscopy can be used for the analysis. The methods under each type of analytical technique are summarised in Table 7.3 along with a list of suggested elements for which the method is suitable. Table 7.3 Summary of .Destruct!ve Spectroscopic Ana.!ytical Techniques Emission Absorption Method Application Application Method Flame AAS

Transition and alkaline elements

ICAP

Most elements

Graphite Furnace

Sb, As, Be, Or, Cd, Pb, Se, TI

DCP

Most elements

Cold Vapour

Hg

As, Se, Sb, Sn

Hydride Generation

As, Se, Sb, Sn

Hydride Generation

Flame Atomic Absorption Spectroscopy (Flame AAS)

Flame AAS is commonly used to measure most of the transition and alkaline elements in digestates. However, there are two areas of concern which must be considered when analysing a sample using this technique: 1) The choice of flame conditions (i.e., nitrous oxide/acetylene or air acetylene) on a per element basis is extremely important. Even the same sample run on the same instrument will produce widely variable results. Since some elements, such as AI, Ca, Mo and Si, have relatively low ionisation potentials, they are likely to undergo ionisation in the flame, thereby lowering the measured value. This is generally corrected by the addition of a concentrated element which has a lower ionisation potential than the element being analysed. 2) Most MSW incinerator residue streams contain high concentrations of salts, especially combined bottom/fly ash and air pollution control system residues. The high salt concentrations increase the viscosity of the sample and can potentially cause severe problems with aspiration of the sample into the flame. One way of remedying the problem is to dilute the sample, although this entails raising the method detection limit, perhaps beyond acceptable levels. Anion exchange can also be used to overcome the high salt concentrations, however, this approach requires extra QA/QC including analyte spiking checks.

228

Graphite Furnace Atomic Absorption Spectroscopy (GFAAS)

GFAAS is a method of quantification which is preferred in some instances since the detection limits are significantly lower than flame AAS, although the increased precision is gained at the expense of time. Typically, it is used for a limited number of trace elements, such as those listed in Table 7.3 (recommended under EPA Method 1620). Similar to flame AAS, analysis by GFAAS requires careful attention to the sample matrix, either through matrix modification, anion exchange, dilution or matrix duplication with standards. Another way of increasing the accuracy of the analysis is to use low porosity graphite tubes. There are a number of interferences encountered when using GFAAS to measure certain elements. Table 7.4 summarises the problem encountered with a number of metals and the corrective measure required to eliminate the interference. Table 7.4 Summary of Interference Problems and Correction Measure s for Specific Elements Element Interference Problem Corrective Measure Sb High chloride content produces Add an excess of 5mg of ammonium losses prior to atomisation nitrate to the sample in the graphite tube prior to drying and ashing As

Deuterium arc background correction insufficient compensation for high levels of AI and Fe

Conduct background correction with Zeeman or Smith-Hieftje correction

Be

Potential gas phase interaction with nitrogen

Do not use nitrogen as purge gas

Cr

Oxidation state may change

Add hydrogen peroxide to the acidified sample to ensure Cr is in trivalent state, or add 500 mg/L of Ca to reduce volatilisation (may cause further problems due to high TDS of sample)

CN band broadening interference

Do not use nitrogen as purge gas

Pb

Sulphate interference

Add lanthanum nitrate to sample to suppress the effect of sulphate

Se

Same as for As

Same as for As

Cold vapour AAS is another application of flameless AAS. It is used in the analysis of mercury with a primary working range of 0.1 to 5.0 ppb. Major interferences are from

229 gaseous species that also absorb at the 253.7 nm mercury wavelength. Chlorine and water vapour are the two main contributors to this type of error. After the rigorous digestions required for the MSW residues there is little chance of having free chlorine present and water vapour can be removed by passing the post reacted gas stream over concentrated sulphuric acid or other moisture trap. Hydride generation is another form of analysis that quantifies the analyte in the gaseous phase. Elements such as, Se, Sb, and Sn are routinely analysed by this method. It is adaptable to both absorption and emission spectroscopy. Detection limits as low as the part per billion range can be attained with good accuracy and precision. Interferences can occur in the generation of the hydride species which are due to high concentrations of metals such as copper, chromium, iron and nickel. Increasing the borohydride:sample ratio is an approach generally used to overcome this problem. It is also necessary to ensure the analyte species are in the proper oxidation state prior to the hydride reduction. This may be accomplished by heating the digested sample in hydrochloric acid before analysis. Atomic emission analysis can be performed by using either an inductively coupled argon plasma spectrometer (ICAP) or a direct current argon plasma spectrometer (DCP). ICAP is routinely used for metal analysis since it has the potential of being accurate and efficient with Method Detection Limits (MDL's) of 1 to 2 orders of magnitude lower than that of flame AAS. This is due to the very high operating temperature of the plasma. In addition, the sample is pumped into the flame rather than being aspirated, therefore less physical interference is generally encountered than during atomic absorption spectrophotometry. Similar arguments for DCP are also true. There are two main categories of interference which can occur with ICAP or DCP analyses, namely, spectral and physical interferences. These interferences are summarised in Table 7.5 along with the specific corrective measures required to eliminate or minimise the interference. There are trade offs which much be considered when choosing ICAP or DCP for analysis. Although the detection limits for DCP are higher than ICAP, DCP will handle the high total dissolved solids content of MSW incinerator digestates or leachates better than ICAP. Consequently, the choice of instrument is highly dependent on the requirements for a detection limit on specific elements.

Non-Destructive Analytical Methods

Instrumental Neutron Activation

In principle, Instrumental Neutron Activation (INA) measures primarily gamma radiation emitted by the radioactive isotopes produced by irradiating samples in a nuclear

230 reactor. The MDL's are generally higher than those associated with either AAS or DCP/ICAP, however, the technique has some advantages depending on parameter choices: 1) There is no chemistry required, hence the difficulties and limitations of trying to dissolve the sample do not exist; 2) It is a multi-element technique, capable of determining 35 elements simultaneously; and 3) A much larger sample mass (up to 30 grams) can be used thereby further minimising the effects of sample inhomogeneity. Table 7.5 Summary of Interferences and Corrective Measures for ICAP/DCP Analysis Category Spectral

Physical

Interference

Correction

Overlap of different elemental spectral lines

Install software packages designed to analyse the raw data and add further resolution

Unresolved overlap of molecular band spectra

Choose an alternate operational wavelength

Background contribution of stray light caused by line emissions from high concentration elements

Apply background correction adjacent to the analyte line

High dissolved solids content of the sample matrix causing fouling of equipment

Dilute the sample or ensure peristaltic pump lines are clean and free flowing

Salt accumulation at the tip of nebuliser and torch causing plasma drift, hence inaccurate readings

Constantly rinse the system with distilled water between readings and replace fouled parts

X-Ray Fluorescence

X-Ray Fluorescence (XRF) is another non-destructive mode of analysis that may be applied to MSW incinerator residues, however, the detection limits are generally higher than any of the aforementioned methods of analysis. This method is discussed in more detail in Chapter 7.3.

231 Since the detection limit required during analysis is quite often the determining factor when selecting analytical instrumentation, Table 7.6 provides a summary of the potential detection limits which can be achieved by the various methods. In addition, Table 7.6 provides a summary of acceptable methods for quantifying specific elements in the sample matrices of MSW incinerator residues. Table 7.6 Summary of Achievable Detection Limits by Different Analytical Instruments ICAP Element AI Sb As Ba Be B Cd Ca Cr Co Cu Fe Pb Mg Mn Mo Ni Se Ag Na Th Sn Ti V Zn k9

,..,,

"k~'__

-I- =

Wavelength (nm)*

MDL (ppm)**

INA

XRF

MDL (ppm)

MDL (ppm)

308.215 0.045 206.833 0.032 5.0 193.696 0.053 2.0 5.0 455.403 0.002 100 5.0 313.042 0.0003 249.773 0.005 226.502 0.004 317.933 0.010 10,000 267.716 0.007 10 5.0 228.616 0.007 5 5.0 324.754 0.006 5.0 259.940 0.007 200 220.353 0.042 5.0 279.079 0.030 257.610 0.002 5.0 202.030 0.008 5 5.0 231.604 0.015 50 196.026 0.075 5 328.068 0.007 5 588.995 0.029 500 190.864 0.040 0.5 189.989+ 0.030 100 5.0 334.941 0.003 292.402 0.008 5.0 213.856 0.002 50 5.0 Recommended for sensitivity and overall acceptance. Use of alternate wavelengths should be reported with data. Source: "Inductively Coupled Plasma-Atomic Emission SpectroscopyProminent Lines" (EPA-600/4-79-017). nitrogen purge used at this wavelength

232 7.2.4 Loss on Ignition

"Loss on ignition" (LOI) data has typically been used to provide an indication of the degree of "burnout" achieved during combustion or the "combustion efficiency". LOI is usually defined as the weight loss of a solid sample (previously dried for 24 hours at 105~ after exposure to 550~ in a muffle furnace for sufficient time to achieve a constant weight. Typically the results are expressed as a percentage of the dried sample weight. LOI =

WDA - WMA

(x 100%)

WDA where; WDA = weight of ash dried at 105~ in grams W ~ = weight of ash muffled at 550~ in grams Although the use of LOI values is an acceptable surrogate parameter for measuring "burnout" in bottom ash and grate siftings samples, the interpretation is not accurate when used in context with heat recovery system and APC residue streams. Since the temperature in the flue gases exiting the boiler are much lower than 550~ any flue gas reaction products which condense out or sorb onto particles, will be re-volatilised during muffling. In addition, a portion of the volatile material from lime-based APC residue is attributable to the loss of water of hydration from the excess lime (CaOH2 to CaO and water). A further bias which prevents direct comparison of LOI values between APC residues from different facilities can be attributed to the fact the stoichiometric ratio of lime addition varies from facility to facility. Hence, if the content of excess lime in the residue varies, so will the amount of water of hydration. Other methods, such as particulate carbon, should be used to determine the content of "combustible" material remaining in heat recovery and APC system residues. Much of the LOI data quoted in the literature are used to define combustion efficiency, however, caution should be used when attempting to compare LOI data from some sources. LOI data can also be administratively defined as a function of the total weight of ash leaving the facility, which includes accounting for the ferrous, non-ferrous and other non-combustible materials in the ash removed prior to testing (which were originally non-combustible). Consequently, comparison of laboratory LOI values and administrative values would generally indicate lower administrative values than the laboratory LOI values (by a factor equal to the proportion of reject material). The most commonly used methods of determining LOI are the APHA Standard Method 209E (1981) and ASTM Standard Method C25-88. Some of the different modifications incorporated into these methods include: 1) Reduction in Particle Size of Bottom Ash Samples - prior to muffling bottom ash, the samples are ground to minimise the potential for material loss due to

233 displacement via "popping" or violent cracking of particles which has been noted with some unprocessed bottom ashes. This phenomenon has been noted mostly on quenched bottom ash samples, not on those samples collected prior to quenching. Since there are blind pores or "pockets" within some ash particles and hydrogen can be liberated upon hydration of bottom ash, it is speculated quenched samples may contain small pockets of hydrogen gas or water trapped within clinker-like particles which are not liberated during drying at 105~ but exert sufficient pressure to escape when heated to 550~ Reducing the particle size also helps to minimise variations between bottom ash samples by acting to "homogenise" the sub-samples, thereby enhancing reproducibility. Consequently, it is recommended ground bottom ash and grate siftings samples be used for LOI determination. 2) Variations in the Quantity of Sample - the quantity of sample used in the determination varies (usually between 10 and 50 grams), and is typically based on the precision of the weigh scale being used and/or the quantity of sample available for testing. Generally, the less sensitive the scale, the larger the sample size required to provide an adequate degree of precision. For example, a minimum of 50 grams of sample should be used when employing a scale capable of sensitivity to a milligram in order to achieve a precision of at least 0.1% +0.05%. Weigh scales with the capability to measure to the 1/10 of a milligram will permit using sample sizes down to 10 grams. 3) Typically, the length of time for muffling the sample varies between 1 to 2 hours. However, the volume of the sample is a factor to be considered in deciding upon the length of time required for muffling. Since the density of some residue streams is less than 1 g/cm 3 (e.g., APC residues), a relatively large volume of low density ash (hence, increased insulating properties) will require more time to reach 550~ than a similar weight sample of a relatively high density ash (e.g, bottom ash >1.5 g/cm3). Consequently, it is recommended the time required for muffling should be increased to permit sufficient heating of the less dense residues (mixing the sample part way through the muffling process is also recommended). A period of two hours is recommended to ensure sufficient time for the temperature of most samples to reach 550~ This should also ensure adequate time for volatile material to escape the solid matrix. 4) The procedures stipulate the determination should be conducted to achieve constant weight. Although the ASTM method indicates precision of +4% difference between consecutive readings, experience indicates this level of precision will be difficult to achieve without weight sensitivity to 1/10th of a milligram. One repetition of the procedure should be sufficient to ensure constant weight has been achieved. 5) As an additional or alternate QA/QC procedure, it is recommended the test be conducted on triplicate sub-samples and the LOI values compared. If there is less than a +5% difference in values, the average of the triplicate values should be used.

234 There are other aspects of the procedure which require special attention to ensure generation of accurate data, namely, cooling and desiccation of the sample. Experience has shown MSW incinerator residues are hygroscopic, and hence are very sensitive to changes in humidity in the laboratory. Some precautions that should be taken to ensure better reproducibility include: 1) Employ the use of a vacuum vented drying oven for moisture content determination prior to muffling. This avoids the potential for a build up of humidity which can often occur in passive drying ovens, and can result in moisture being held in the "so called" dry sample for muffling. 2) After muffling, the samples should only be allowed to cool in the open air for a short period of time before being placed in a desiccator. Steps should also be taken to ensure the desiccant material is fresh and active.

7.2.5 Total Carbon, Carbonate, Sulphur and Ammonia Analysis of total carbon and total sulphur are generally conducted using the Leco induction furnace method. Since the sample mass used for the analysis can be relatively small (0.02 - 0.5 grams), a well homogenised sample plays a critical role in generating good data. Sulphate content is typically determined by analysing distilled water leachates of the residue samples, although this method will only detect the soluble sulphate compounds. It should also be noted sulphate solubility is limited, and several water extractions may be required to ensure most of the sulphate has been released from the solid sample. Sulphides can be measured by an iodometric method using Method 4500 S2 E (Standard Methods for Examination of Water and Wastewater (SMEWW). The speciation of sulphur-based compounds can be determined by chromatographic methods given in Streudel et al., (1989), and by the potentiometric procedures given in Satake et al., (1981). Examples of the achievable detection limits for the various sulphur compounds are given in Table 7.7. Table 7.7 Achievable Detection Limits for Sulphur Based Compounds Parameter

Detection Limit for Solid Sample (Pg/g) ......

Detection Limit for Liquid Sample (IJg/mL)

SO4 S2 SO32S2032" Sx2"

100

0.15 0.10 O.10 2.00 6. O0

235 Methods also exist to distinguish the different types of carbon present in a sample (i.e., total organic carbon and degradable organic carbon versus total carbon), which are given in VGB Arbeitsgruppe (1992). A further distinction between the specific organic carbon substances may also be warranted. For example, Method 5560 (SMEWW) provides for determination of organic and volatile acids, whereas Method 5510 (SMEWW) outlines the determination of sugars, humic and fulvic compounds. A precise method for determining carbonate contents in ash is important when evaluating the carbonation of bottom ash for utilisation purposes. Carbonate is often measured as total alkalinity based on Method 2320 (Standard Methods for Examination of Water and Wastewater). A more accurate value is obtained by quantitatively converting the carbonate to carbon dioxide via acidification. The carbon dioxide is then trapped in a NaOH solution which is subsequently titrated with HCI for the determination (Vogel, 1961 ). Several methods are available for the determination of ammonia which are described in Method 4500 NH3 Nitrogen (Standard Methods for the Examination of Water and Wastewater). These include the Nessler method, a phenate method, a titrimetric method and ammonia selective electrode method, although the latter method is less accurate for ammonia in the high salt content matrix of incinerator residue leachates.

7.2.6 Acid Neutralisation Capacity Acid neutralisation capacity (ANC) is defined as the capacity of a material to resist changes in pH. The ANC results are typically used to assist with determination of the potential for trace metal mobility, based on the variation in pH over time. The test is generally performed using one of either two methods. One method is a titration method based on the ASTM Standard Method C400-64. It typically involves slurrying a 1 gram sample of ground ash with distilled water at a 20:1 liquid-to-solid ratio and titrating to an end point pH with an acid solution, usually HNO3. Although this can be done either manually or by using an automatic titrator, there are some precautions which need to be taken to ensure generation of reproducible results. The ash should be ground to at least 150 pm (#100 mesh) particle size to hasten the chemical reactions. Experience has shown manual titration of ash can be very time consuming. When acid is added to the ash slurry, the pH may drop quickly, then slowly increase over a wide pH range. As the slurry moves closer to the end point pH, care must be taken to add the acid in drop-wise increments to avoid "overshooting" the target pH. Consequently, the length of time required to reach equilibrium at the end point pH is generally considerable (hours to days). Use of an automatic titrator saves on personnel time. The extended length of slurrying time may also be detrimental to the quality of data being generated. The longer the test takes, the greater the potential for uptake of CO2

236 from the air, which biases the measured ANC value. This can be avoided by running the titration in a covered vessel, or "closed" system, to limit the exposure of the slurry to the air. Another drawback to a titration is the difficulty in obtaining sufficient data to generate a pH/ANC curve, which can then be used in models to help estimate the potential for trace metal solubility over time. Even with an automatic titrator, obtaining a titration curve is difficult because the time required to reach equilibrium with these types of highly buffered materials is not conducive to a continuous titration process. The second method is a modified procedure which involves mixing individual 5 gram samples of ash in 30 millilitre acid solutions of varying strength. The solutions range in strength over 10 equal increments from distilled water to either 2.0 N HNO3 or 4.0 N HNO3. The respective ash and acid solution slurries are mixed for 48 hours in a rotary extractor before centrifuging and monitoring of the pH of each sample. This method is generally more cost effective since it is less time consuming than the straight titration. It also provides a closed system for the test, which reduces the potential for CO2 uptake to bias results. In addition, more pertinent information is derived, since a pH/ANC curve can be developed from the incremental analysis, which in turn can be used to assist with modelling of potential trace metal solubility. Furthermore, the filtered extracts from each increment can also be analysed for dissolved trace metal content to develop a pH/solubility curve which can also be used for modelling. One of the drawbacks to this method is it requires a rotational device and high speed centrifuge to conduct the determination, both of which can be costly. Another drawback is if the extracts are to be analysed for trace metals, the quantity of ash and acid solutions must be scaled up to accommodate analytical requirements. Scaling the quantities up may also require the containers to be "burped" during the mixing process to bleed off the build up of excess hydrogen generated by slurrying the ash. Based on the discussion above, it is recommended the incremental method of performing the ANC determination be used instead of the titration method. The benefits of the extra data generated from this procedure far outweighs the capital cost requirements, and is generally less susceptible to error caused by CO2 uptake.

7.2.7 Organic Analyses Sample Preservation

The prominence of the issues surrounding polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzofurans (PCDF) compounds in MSW incinerator ash warrants a brief discussion on the importance of proper sample collection and preservation prior to extraction.

237 It is suggested that due to the potential dynamic nature of organic compounds, all samples collected for organic analyses should be cooled as quickly as possible after collection, and a storage temperature of 4~ should be maintained until extraction. This is especially true for quenched ash samples. Samples submitted for analyses are typically processed on an "as received" basis and results should be specifically expressed either on a wet weight (give % moisture) or a dry weight basis. Should it be deemed necessary to dry and grind the sample prior to extraction, it is suggested the samples be either air dried or dried at temperatures less than 50~ to minimise potential volatilisation of some organic compounds.

Extraction

The extraction of ash samples is generally facilitated by decomposition of the matrix using hydrochloric acid. After digestion, the mixture is neutralised and lyophilised. The remainder of the extraction is conducted in a similar manner to air emission samples. Generally, standard solutions of organic compounds are then added to the digest solutions prior to extraction with toluene.

Chlorophenols

Potassium dichromate is mixed into the toluene extracts to extract the chorophenol fraction. The chlorophenols are then acetylated using acetic acid anhydride and further extracted using dichloromethane and concentrated by evaporation. The fractionate is then analysed using GC/MS.

Chlorobenzenes

Chlorobenzenes must be fractionated from the dioxins and furans prior to analysis. The toluene extract must be cleaned up removing the chlorophenols, alcohols and organic acids by treatment with potassium dichromate. The cleaned extract is then concentrated and fractionated by column adsorption chromatography in a aluminum oxide column. The fractionate is then eluted with benzene to generate chlorobenzenes as the first fraction, then with dichloromethane/hexane to give a dioxin/furan fractionate.

Analysis

Analysis of trace organic compounds can be conducted using gas chromatography/mass spectroscopy (GC/MS) preferentially, or a high resolution GC and MS (HRGC/HRMS). The HRGC/HRMS instrument provides lower detection limits and is considered state-of-technology at the present time. The actual analysis requires the skill of a highly trained technician to run the equipment and interpret the data being generated.

238

7.3 CHEMICAL SPECIATION METHODS The nature of the solid phase is very important when evaluating leaching phenomena, utilisation, and management of MSW combustion residues. As discussed in Chapter 13, the nature of the solid phase has a profound impact on fundamental leaching behaviour. It is therefore important to describe methods that can be used to characterise the morphology of solid phases, the mineralogy of the solid phases, elemental associations in the solid phases (particularly in less crystalline, more amorphous material), and valence states of elements in the solid phase. These issues comprise the nature of the chemical speciation of elements in ash particles. Elements at the surface of an ash particle (top atomic layer) exist in a markedly different environment than atoms situated just below the surface (near-surface environment), which in turn differ from elements situated in the bulk of the sample (bulk environment). These differences arise because the top atomic layer can have at least one bond direction where the coordination chemistry differs from bulk elements (Hochella, 1990). Additionally, external gases, liquids, or solid phases can react with the top atomic layer, causing disruption in bond strength among adjacent atoms in minerals and between crystal lattices. This disruption can extend into the near-surface environment but it dissipates with depth (Hochella, 1990), hence caution should be used when describing the type of solid phase analytical method employed to characterise bulk, near-surface, and surface elements for characterising their speciation. By drawing on the extensive experience used by metallurgists and by geochemists the mineral surfaces and mineral phases can be characterised, particularly for basaltic rock, evaporites, and sedimentary rock, since these minerals are close in composition to certain MSW ashes. Readers are directed to the recent reviews of Whan (1986), Hochella (1990), Coyne, McKeever and Blake (1990), and Hawthorne (1988) for very detailed reviews of all the spectroscopic methods available to characterise solid phases. The methods described in this section therefore are not considered to be inclusive of all the methods available for use.

7.3.1 Separatory Techniques The goal of this section is to discuss methods of "disassembling" the complex, heterogeneous constituents in ash samples so the resultant fractions can be better characterised for their solid phase chemical speciation. Disassembling can help to concentrate phases of similar magnetic or density characteristics. By concentrating the phases, problems with detection limits can be avoided. In one study on bottom ash, disassembly and subsequent concentration resulted in over a 500 fold increase in the concentration of chromium (Eighmy et al., 1993). Figure 7.5 depicts the type of sample processing to be discussed. Such schemes have been used in part or in total to characterise ash (Eighmy et al., 1992, Eighmy et al., 1993).

239 Figure 7.5 Schematic Showing the Type of Sample Processing that can be Employed to Disassemble Incinerator Residue for further Analysis

BOILER ASH, SCRUBBER RESIDUE, & FABRIC FILTER ASH <300 um Subsample of Working Sample Density Gradient Separation in Tetrabromoethane <2.95 g/cm3 Selective Etched

NonEtched

>2.95 g/cm3 Selective Etched

NonEtched

BOTTOM ASH, GRATE ASH, & GRATE SIFTINGS <300 um Subsample of Working Sample

Magnetic Separation Non-Magnetic Fraction Density Gradient Separation in Tetrabromoethane

Magnetic Fraction Density Gradient Separation in Tetrabromoethane

<2.95 g/cm3

>2.95g/cm3

<2.95 g/cm 3

Same

Same

Selective NonEtched Etched

>2.95 g/cma

Same

240

Sample Drying

It is sometimes advantageous to analyse ash particles in their native states. Almost all the described analytical methods discussed in this section require the use of dried samples. Additionally, some methods require fine powders, polished flat specimens, or particle thin sections. However, processes that dehydrate the ash samples or reduce the size of particles to unit mineral grains can have a profound impact on chemical speciation. Bulk specimens become so fractured that internal grain boundaries become atmosphere-exposed. The presence of oxygen or carbon dioxide can react with and alter these surfaces. Consequently, some of the efforts used to analyse solid phases have inherent compromises in them. A balance must be struck in the use and selection of analytical methods. Invariably, multiple methods should be used. Sample drying can arrest aging reactions that can occur in hydrated ash samples. Hydration, carbon dioxide uptake, and biological activity can be stopped. It is possible to use drying as a means of arresting aging reactions during a time course to look at "snapshots" of the aging progression. Drying methods, as outlined in Section 7.2.1, are sufficient. Care should be taken that temperatures are maintained at 60 to 70~ to prevent phase changes in some of the more heat-sensitive minerals or amorphous phases. Labile waters of hydration can also be removed from gels or crystal phases that may alter phase morphology or properties.

Particle Size Reduction

Ideally, investigations of the bulk particle can focus on the properties of all the particle constituents or the particle can be size-reduced into its component grains or phases. The latter provides the opportunity to better characterise its components. Thorough size reduction can even allow us to examine monophasic particles, provided fracturing occurs at grain or phase boundaries. In any case, questions can be investigated regarding the chemical speciation of the bulk, the component phases, the near surface environment, and the top atomic layer of particles or their component phases, keeping in mind the original ash particle size and the size of its constituent phases. This is more important for bottom ash and grate siftings than for all the other incinerator residues. Jaw crushing can be used to break apart bottom ash into smaller size material. Hutchinson (1974) details crushing methods. Jaw spacing can be adjusted to select for final particle size. Final sizes can range down to 1 mm. The use of a jaw crusher provides for more systematic final sizes than impact crushing. Riffle sampling can then be used to select subsamples. Further size reduction of all types of ash specimens can be accomplished using a number of devices. Caution should be used here. Further size reduction can help to produce a more homogeneous material. The degree of homogenisation also depends on the final particle size. In order to reduce fine-grained fly ash into its component

241 monophases, extensive grinding may be needed and a high degree of homogenisation may be obtained. Overgrinding introduces the risk of sample contamination from the grinder. Component monophases of a coarse bottom ash can be examined with less grinding and hence a low degree of homogenisation may be obtained. Grinding may be accomplished using hand-held mortar and pestles made of agate, mullite, corundum, or tungsten carbide. Porcelain or iron mortars can be too soft for certain ashes. Mechanised mortars and pestles can be used to provide more systematic grinding procedures. Grinding may also be accomplished using mechanised grinding plates, ball mills, or hammer mills. The use of tungsten carbide grinding surfaces is preferred; however, contamination from these surfaces is possible with certain bottom ash samples. The lower limit of particle size reduction using these mechanised method is about 10 pm. Micronising mills reportedly can produce this size range (Bish and Reynolds, 1989). For some fine grained ESP boiler, and fly ashes, it may be difficult to size-reduce to monophasic material. Extensive grinding can generate high quantities of heat. Alcohols and acetone can be used as lubricants for certain grinding processes (Bish and Reynolds, 1989); however, these solvents may have slight solubilising effects on certain highly soluble metal salts in ash residues. Control of temperature may be quite important for hydrated phases, but evaluation of potential solvent dissolution should be done.

Magnetic Separation Techniques

Incinerator residues, particularly bottom ash, grate siftings, and ESP ashes contain magnetic materials. Magnetic separation can be used to remove these fractions. This can allow for the more careful study of these fractions, or other phases intimately associated with these fractions. The non-magnetic residue can then be more carefully studied without the diluting effect of the magnetic fractions. Magnetic separation techniques have worked well with coal fly ash (Hansen et al., 1981 ) and with municipal solid waste ashes (Eighmy et al., 1992; St,~mpfli, 1992). Two types of magnetic separation can be employed. The first involves the use of either a hand-held magnet or a rotary ferromagnetic separator. The hand-held magnet can be covered with glycine weighing paper or done through a glass petri dish to facilitate separation of particles from the magnet. The rotary separator is a more systematic method (Hutchinson, 1974). Both methods remove hematite (Fe203), magnetite (Fe304) and iron metal fragments. These are routinely found in bottom ash at concentrations up to 35% on a dry weight basis. Isodynamic or barrier separators are available that can adjust the amperage applied to an inclined vibrating plate. The amperage is used to control the intensity of the magnetic field. More than sixty types of minerals can be diverted (Hutchinson, 1974).

242 A sample is run through the apparatus after setting the amperage. The passed through nonmagnetic materials are saved and the magnetic fraction is then collected. The passed through materials are then run through the apparatus again at a higher amperage. The process is repeated until the applied amperage is about 1.5 amps. Repetitive separations at the same settings can also be used to effect a very thorough separation. This method has been used on bottom ashes (Eighmy et al., 1992; St~mpfli, 1992) and on ESP ashes (Eighmy et al., 1993). A fine particle size should be used. Dry particles less than 150 IJm and greater than 90 pm in size are most readily processed (Hutchinson, 1974). The schemes in Figure 7.5 are also applicable for particles between 225 pm and 50 pm.

Density Separation Techniques

Density separation can be an extremely useful technique, particularly if it is performed on samples previously separated by magnetic separation methods (although the procedure could be reversed). Mineral grains can range in density from about 1.6 to 22.5 g/cm3 with typical values of aluminosilicates below 3.0 cm3 and heavy metal salts above 3.0 g/cm3 (Carmichael, 1989). Other phases in the ash can also vary in density. The method can further concentrate materials by density and remove materials that can dilute the sample during analysis. The principle behind separation is mineral grains less dense than the separating liquid will be buoyant and mineral grains more dense will sink through the liquid. Density gradient separation has worked well with coal fly ash (Furuya et al., 1987), and MSW ashes (Eighmy et al., 1992). Hutchinson (1974) and Zussman (1967) discuss methods for selection of dense liquids and use of various separatory apparatus. Typical liquids used are 1,1,2,2tetrabromethane (2.967 g/cm3 @ 20~ and diiodomethane (3.325 g/cm 3 @ 20~ These solutions are both toxic and expensive and very careful procedures for use in safety hoods are needed (Hutchinson, 1974). Diluents such as acetone and N,Ndimethyl formamide can be used to adjust fluid densities to lower values (Zussman, 1967). Detailed methods are described for separating coarse-grained material and fine grained materials (Hutchinson, 1974). Depending on the type of separatory process (separatory funnel, centrifuge tube, density gradient column), low particle settling velocities can be a concern. There are also lower limits on particle size that approach 10 IJm (Zussman, 1967) because of particle aggregation. Procedures have also been established to process larger volumes of samples (Hutchinson, 1974). Riddick et al. (1986) provide specific gravities of organic solvents. Carmichael (1989) lists the grain densities of elements, alloys, and minerals.

Selective Phase Dissolution Methods

After magnetic separation into magnetic and nonmagnetic fractions and further separation by density, selective dissolution techniques can be used to partially solubilise phases within a fraction to further concentrate the non-solubilised phase and remove phases that contribute to dilution effects. Many of the techniques that can be

243 used here are similar to the ones used for the sequential chemical extraction leaching procedure discussed in Chapter 15. The principle behind selective dissolution is certain phases can be removed by dissolution through the use of extractants. Such techniques have worked well for removing amorphous glassy phases from coal fly ash (Hulett and Weinberger, 1980), or surface phases from a variety of coal ashes, scrubber solids or oil ashes (EPRI, 1985) or surface phases from sewage sludge ash (Theis et al., 1984). Reviews by Theis et al. (1984), EPRI (1987), and Tessier and Campbell (1988) discuss the relative merits of selective dissolution. Some controversy exists in the use of sequential selective extractions (Nirel and Morel, 1990; Tessier and Campbell, 1991). Principle problems include the lack of exhaustive validation of the procedures (Nirel and Morel, 1990), validation with doped artificial phases that do not resemble natural solid phases (Tessier and Campbell, 1991 ), and post-extraction adsorption phenomena which can cause an underestimation of element mass association with specific phases (Nirel and Morel, 1990). More careful design of spiking experiments with natural materials suggests the adsorption problem is not as severe as the other issues (Belzile et al., 1989). Table 7.8, adapted from Theis et al. (1984), depicts the types of extracting agents that can be used to remove selective phases from a sample. Some methods remove surface soluble phases, others dissolve insoluble precipitates, others oxidise reduced phases, others reduce highly oxidised phases, and some can very selectively extract glassy phases or adsorbed metal species. Caution should be used to evaluate the efficacy of the desired extraction. Quantification of dissolved constituents in the extraction fluid, coupled with careful solid phase characterisation before and after the extraction, are usually needed to assess the efficacy of the extraction, as well as its selectivity to the desired phases in the solid. The work of Hulett and Weinberger (1980) has shown the utility of this method in dissolving the amorphous glassy phases from coal fly ash with hydrofluoric acid so mullite and quartz phases could be characterised for chemical speciation.

7.3.2 Impregnation, Thin-Sections, and Thin-Foil Methods As discussed in Section 7.3.1, it is possible to disassemble combustion residue particles into more defined fractions. Many of the chemical speciation analytical techniques that can be used require the use of fine powders. Some techniques can be used on the intact ash specimen. Frequently, both powders and intact ash particles must be further processed to make them amenable for analysis. Impregnation with a polymer so the samples can be polymerised, hardened, thin sectioned and polished to produce a petrographic thin section is frequently done. Fine powders can be manufactured into highly polished thin foils. Ion bombardment

244 techniques can then be used to ion-etch the particle surface to remove the top atomic layers. Table 7.8 Commonly Used E.xtractincj Agents Extracting Agent H20 Acetic Acid H2CO3 H3PO4 Citric Acid Dilute Strong Mineral Acids (e.g. HNO3, H2SO4, HCI, HF) Conc. Strong Mineral Acids (e.g. HNO3, H2SO4, HCI, HF) HNO3-HCI-HF NaOH BaCI2, MgCI2, CaCI2 KNO3, KF, NH4Ac EDTA

Na4P207 Ammonium Oxalate (acid) Citrate-DithioniteBicarbonate Hydroxylamine Hydrochloride H202

Primary Chemical Interaction Hydration Acidic Acidic Acidic Acidic Acidic

Major Application

Mild Dissolution Mild Dissolution Mild Dissolution General Metal Recovery General Metal Recovery General Metal Recovery HCIDissolution of Metal Oxides Acidic Dissolution of Major Mineral Forms (except silica) General Metal Recovery Acidic Complete Dissolution Basic Dissolution of Polar Organic Fraction, Dissolution of Fe, AI Oxides Solvating Power due Exchangeable and Physically to high ionic strength Adsorbed Species Complexometric Exchangeableand More Strongly Bound Metals, Dissolution of Carbonate, Recovery of Metals from Organic Phases Complexometric Recoveryof Metals from Organic Phases Complexometric Dissolutionof Amorphous Oxides of AI and Fe Reductive Dissolutionof Fe and AI Oxides Reductive Dissolutionof Mn Oxides Oxidative

From Theis et al., 1984 with permission of the author

General Metal Recovery Organic Digestion

245 Petrography and optical mineralogy frequently make use of thin slices of rock material that are subsequently glued to a glass slide, ground, and then polished to a 30 IJm thickness for subsequent analysis (Nesse, 1991). Usually, specimens must be impregnated with a heat-activated polymerising resin that solidifies the specimen. Intact ash particles or powders can be imbedded in a mold, polymerised, and then thin sectioned with a thin sectioning machine prior to grinding and polishing. Hutchinson (1974) describes two methods for ash-like material. The polymerising solution is added to the dry ash specimen, and then placed in a vacuum. Under a vacuum, pores loose gas and non-viscous epoxy infiltrates the pores. Eventually all pores are filled and the specimen can be hardened to produce a monolith. The second method uses a pressurisation technique to cause complete impregnation. These methods have been used for various ashes (Eighmy et al., 1992). Similar methods have been used for bottom ash (Lichtensteiger, 1992). A number of techniques are available to cut the polymerised monolith. Usually a microtrimming machine provides good samples. A diamond hand saw can be used to trim specimens. Hutchinson (1974) describes methods for sawing and lap grinding with corrudum or alumina powders or silicon carbide abrasive papers. An epoxy resin is usually used to mount the ground specimen to a glass slide. Finally the specimen is polished to very thin thicknesses using alumina or diamond pastes. A final specimen thickness of 30 IJm is required; it is a standard thickness for petrographic analysis and evaluation of refractive properties of light require this thickness. Grinding and polishing can heat the sample and cause debonding between the epoxy polymer and the specimen in the thin section. A number of methods are available to make thin foils. Tighe (1976) presents methods where petrographic thin sections are cored to collect small circular flat specimens for use in transmission electron microscopy. Powders can also be embedded in an epoxy amalgam that contains conductive silver (Eighmy et al., 1992). The molds used are the same size as transmission electron microscopy (TEM) grids. This allows direct insertion into the TEM microscope. The samples are ground with a dimpler that abrades through the centre of the specimen. After ion milling, the margins can be examined.

7.3.3 Analytical Methods for Solid Phase Chemical Speciation Table 7.9 depicts a variety of analytical methods used to look at element speciation in the bulk phase of intact particles, in the bulk phase of particle cross sections (e.g. thin sections or thin foils), in the near surface environment of intact particles, and in the top atomic layer of particles. Here it is assumed the specimens are, for the most part, particle size-reduced with fracturing at grain or phase boundaries so the analysed particles are mostly monophasic. In some instances, entire ash particles are analysed. Care will be taken to identify differences where appropriate.

hr

Table 7.9 Analytical Methods for Examining Solid Phase Chemical Speciation Type of Analysis

P

Q,

Element Location

Morphology

Mineralogy

Element Association1 Composition

Element Bonding and Valency

Bulk (Intact Particle)

-Transmitted Light Microscopy (TLM) -Scanning Electron Microscopy (SEM)

-X-ray Powder Diffraction (XRPD) -Differential Thermal Analysis (DTA)

-Scanning Electron MicroscopylX-Ray Microprobe (SEMIXRM) -X-Ray Fluorescence (XRF)

Bulk (Particle Thin Section)

-Petrography

-Petrography

-Scanning Electron MicroscopylX-Ray Microprobe (SEMIXRM) -Scanning-Transmission Electron MicroscopyIXRay Microprobe (STEMIXRM) -X-Ray Fluorescence (XRF)

-Electron Energy Loss Spectroscopy (EELS) -Extended X-Ray Adsorption Fine Structure (EXAFS) -Nuclear Magnetic Resonance (NMR) -Electron Energy Loss Spectroscopy (EELS) -Infrared Spectroscopy (IRS) and Raman Spectroscopy (RS)

-Auger Electron Spectroscopy (AES) -X-Ray Fluorescence Spectrometry (XFS) -X-Ray Photoelectron Spectroscopy (XPS) -Secondary Ion Mass Spectroscopy (SIMS)

Near Surface (Intact Particle, Particle Thin Section, Thin Foil)

Top Atomic Layer (Intact Particle, Particle Thin Section, Thin Foil)

-Scanning-Tunnelling Microscopy (STM)

-Electron Energy Loss Spectroscopy (EELS) -Infrared Spectroscopy (IRS) and Raman Spectroscopy (RS) -Extended X-Ray Adsorption Fine Structure (EXAFS) -Extended X-Ray Adsorption Fine Structure (EXAFS)

247 The methods in Table 7.9 provide information on particle morphology. This can include identification of crystal facets, twinning, and crystal structures. The methods in Table 7.9 also can identify minerals and estimate mineral formula which again work well for crystalline phases. Other methods in the table give elemental abundance or phase association where interrogating beams impinge the sample. Such methods can show elemental composition of mineral grains that have been characterised by other mineralogical methods. Finally, some methods are provided, that can infer an element's valence state or the nature of the bonding environment about an atom. The methods shown in Table 7.9 are not exhaustive, all have been used in matrices similar to combustion residues. Each method is briefly described; their relative advantages and disadvantages are discussed where appropriate. References are provided describing each method in detail. A number of the techniques used for quantitation and surface microanalysis are comparatively evaluated in Table 7.10. The table discusses radiation sources, emissions, depth of analysis, resolution, detection limits, and some relative advantages and disadvantages for each method.

Transmitted Light Microscopy (TLM)

Transmitted light microscopy (TLM) can be used to view the morphology of small ash particles or powders as individual grains on spindle stages or as mounts on microscope slides. This method provides morphological information on intact bulk powders; it can be used to obtain cross-sectional morphology of particles in thin section. Reviews by Zussman (1967) and Nesse (1991) provide details of these methods.

Scanning Electron Microscopy (SEM)

Scanning electron microscopy (SEM) can also be used to examine the morphology of ash particles. Like TLM, it provides morphological information on intact particles; however, it can be used to obtain cross-sectional morphology of particles in petrographic thin section. SEM utilises the secondary and backscattered electrons to derive a reflected energy beam picture of the sample. The beam penetrates the sample to a depth of 2 to 3 pm. Frequently specimens must be sputter coated with gold or palladium to make them electron dense or reflective. Precise detail can be observed up to magnifications of 5,000X. Reviews by Wenk (1976) and Blake (1990) provide further details about this method.

Petrography (Morphology)

Petrography can be used to examine phase morphology of intact ash particles imbedded in epoxy resins in 30 pm thin sections. The method can give exterior particle morphology by examination of the particle outline. It can also be used to look

248 at phase twinning, alteration, and association in the particle (Nesse, 1991). This method can also be coupled with modal analysis to quantify the relative abundance of phases in a particle cross section (Hutchinson, 1974). Table 7.10 Comparison of Surface and XRF x-ray .Interrogating Beam x-ray .Emitted Radiation D,G,P ,Sample Prepb 20-30 mm ,Typical Analysis Diameter (-) ,Diameter (minimum) 10-30 IJm Depth 9 of Analysis ( 1000 pm) Na to U .Detectable Elements ppm ,Detection Limits ,Data Established Reduction for Quantitation No ,Dot Mapping Capability Some ,Specimen Damage ,Advantages -Bulk Analysis

Bulk Microanalysis Methods AES XRM XPS electron x-ray electron, x-ray

SIMS ion

x-ray

photoelectron

Auger Electron

ion

D,P,C

D,CS

D,CS

D,G

3-5 IJm

1-5 mm

1-5 IJm

1-5 IJm

(1-2 IJm)

(150 IJm)

(0.03 IJm)

(1-2 pm)

1-3 IJm

10-50 .&

10-50 A

Varies

Be to U

Li to U

Li to U

Li to U

ppm

high ppt

high ppt

ppm

Established Improving

Improving

Established

Yes

Yes

Yes

Yes

Some

Some

Some

Yes

-High -Surface Resolution Sensitive

-Surface Depth Sensitive Profiling -High Resolution -Not Surface -High Vacuum -High Vacuum Charging Disadvantages 9 -Low Sensitive -Poor -Destructive Resolution -Not Surface -Destructive Resolution Sensitive aCompiled from Hochella (1988), Bancroft and Hyland (1990) with permissionfrom the Mineralogical Society of America bD = dried, G = ground, P = polish, C = sputter coat, CS = clean surface

249

Scanning Tunnelling Microscopy (STM)

Scanning tunnelling microscopy (STM) is one of the few analytical tools that can look at the microtopography of a surface down at the atomic layer (Hochella, 1990). The method has recently been developed and utilised for examining geological specimens. This method must be applied to smooth surfaces of intact particles, it may also be considered for looking at exterior surfaces or phase boundaries in particle thinsections. The reader should refer to Hochella (1990) for additional details of this method.

X-Ray Powder Diffraction (XRPD)

X-ray powder diffraction (XRPD) is a very important tool for characterising the mineralogy of crystalline phases in incinerator residues. Detailed principles of this method are given in Zussman (1967), Reynolds (1989), Bish and Post (1989), and Whan (1986). Care should be taken in selection of the target element to generate the characteristic monochromatic Ka x-ray because of the potential for adsorption of x-rays of certain wavelengths. Copper is usually used, but cobalt can be used when samples contain a high percentage of iron. Powder preparation is very important. Bish and Reynolds (1989), Hutchinson (1974), and Zussman (1967) discuss preparation methods. Particle sizes too large cause problems such as extinction and microabsorption. They can also interfere with the underlying assumptions of random orientation. For very coarse-grained or coarsephased crystals in ash, particle size production will create a particle size equal to a phase size. For very fine-grained or fine-phased crystals in ash, particle size may exceed phase size. It is best if powders are ground close to 10 IJm in diameter or that phase size is less than 10 tJm. Larger particle sizes give qualitative information, smaller particle sizes can be used for quantification (Snyder and Bish, 1989). Bish and Reynolds (1989) also discuss methods of sample mounting (smears, packed tubes, thin films), requisite sample thickness, characteristics needed for the sample surface, and procedures for optimising intensities. Because ash specimens contain numerous major mineral constituents as well as numerous minor mineral constituents, the identification of mineral phases can be quite complex using typical Hanawalt procedures. Computer-aided methods are available (Smith, 1989). However, verification of information from the computer search routines should be done using standard identification methods. XRPD can also be used for semi-quantitative analysis of crystalline phases in solids. This can increase sample analytical times from one to two hours to up to fifteen hours. Additional details are provided in Whan (1986). XRPD has been used successfully in characterising combustion residues (Vehlow et al., 1992; Eighmy et al., 1993; Stampfli, 1992; Kirby and Rimstidt, 1993).

250

Petrography (Mineralogy)

When light is transmitted through mineral grains or mineral thin sections, a number of characteristic properties of the mineral can be ascertained depending on the behaviour of the light. This is the basis of petrography or optical mineralogy. Light wavelengths can be shortened and slowed down. Light can be reflected, refracted, dispersed, and adsorbed. For isotropic minerals, polarised light will be adsorbed when the polarising filters are cross polar. Mineral identification of isotropic materials can be made by examining grain structure and comparing this information to data bases of isotropic minerals (Nesse, 1991). Anisotropic minerals exhibit characteristic uniaxial or biaxial nature using interference microscopy. Birefringence can also be used to characterise the mineral. Such properties are characteristic of anisotropic minerals. Coupled with refractive index measurements of mineral grains, such information can be used to identify minerals (Nesse, 1991). This procedure has been used to characterise the crystalline component of bottom ash (Lichtensteiger, 1992; Eighmy et al., 1993). It holds promise when coupled with XRPD and x-ray microprobe analysis (see below).

Scanning Electron Microscopy/X-RayMicroprobe Analysis (SEMIXRM)

As discussed in the scanning electron microscopy section, primary excitation beams can cause characteristic x-rays to be emitted. Using SEM, the technique can be coupled with x-ray microprobe analysis (SEM/XRM). Samples can be characterised visually and then quantified using microprobe analysis. The probe is calibrated with primary standards, and elemental quantification of surface material can be accomplished (it can also be done with petrographic thin sections) by counting the number of x-rays in discrete channels over a continuum of x-ray energies. Dot maps showing the location of emission of the characteristic x-ray can be used to spatially locate elements. There are techniques that use standardless calibrations. The technique must be used with caution given the beam spreading that occurs, however it does provide very useful analysis of element associations particularly if it is coupled with petrography. Kevex (1989), Wenk (1976), Blake (1990), and Whan (1986) provide detailed methods.

Scanning-Transmission Electron Microscopy/X-Ray Microprobe Analysis (STEMIXRM)

It is possible to expand upon the principles of SEM/XRM to make use of transmission electron microscopy analysis of thin foils. This is termed scanning-transmission electron microscopy (STEM) and it can be coupled with x-ray microprobe analysis to be able to quantify elements in thin foils (STEM/XRM). The excitation beam width in transmission systems is only 30A wide compared to 500A in SEM. Thus, greater resolution is obtained.

251 The use of a thin foil is critical to STEM/XRM. The thin specimen means characteristic x-rays are emitted without absorption or fluorescence. As with SEM/XRM, dot maps can be generated. Electron diffraction patterns can also be analysed. One utility is transmitted electrons can be imaged to examine phase microstructure. Wenk (1976), Blake (1990), and MacKinnon (1990) provide detailed methods.

Auger Electron Spectroscopy(AES)

The emission of Auger electrons of specified energy is characteristic of the elements emitting them. Auger electron spectroscopy (AES) can be used to quantify elements in the surface of a sample. It can be used on powders or polished specimens. The reader is referred to Hochella (1988, 1990), Browning and Hochella (1990), Mogk (1990), and Whan (1986) for further details. It has been used successfully in characterising coal fly ash (Farmer and Linton, 1984) and MSW combustion residues (Eighmy et al., 1992, Eighmy et al., 1993). AES is a relatively rapid method (10 to 15 minutes per sample). Its accuracy is limited to + 30% for most elements with published sensitivity factors. Electron beam charging can be a problem. Embedding particles in indium foil can overcome this problem. Typical sensitivities are 0.1 to 1.0 atomic percent. This method can also do depth profiling.

X.Ray Fluorescence Spectroscopy(XRF)

X-ray fluorescence spectroscopy (XRF) is a sensitive qualitative and quantitative analytical technique. The method can be used on bulk solids, powders, or fused material. Characteristic X-rays in the 1 to 60 KeV range are emitted from samples after excitation with an external energy beam, usually an x-ray. The interrogated sample depth may range from a few microns to millimetres depending on the x-ray energy of the impinging beam. Analytical times are relatively fast, in the order of 15 minutes per sample. The advantages of this method are its applicability to many types of solids (powders, polished specimens, fused specimens), the method is rapid, and the instrumentation is inexpensive. The disadvantages are detection limits in the part per million range and elements of lower atomic number than sodium cannot be readily quantified. More detailed information can be found in Whan (1986).

X-Ray Photoelectron Spectroscopy(XPS)

X-ray photoelectron spectroscopy (XPS), formerly known as electron spectroscopy for chemical analysis (ESCA), is a sensitive surface analysis technique. In XPS, X-rays are directed at a sample where ionisation of the surface atoms in the sample occurs (Bancroft and Hyland, 1990). As an atom ionises, a photoelectron is emitted from the

252 atom. The ejected photoelectron has an energy that is in part characteristic of its binding energy. Over a range of binding energies, emitted photoelectrons can be counted in discrete channels. Usually at least one peak or doublet appears for each element (except hydrogen). Peak intensity is used for quantitative analysis (Bancroft and Hyland, 1990). Peak shift, or chemical shift, is also an extremely valuable tool for documenting valency, nearest neighbour affects, and speciation. Detailed reviews are given in Hochella (1988; 1990) and Perry et al. (1990). XPS is usually conducted with an ion sputtering apparatus to allow for removal of adventitious carbon and oxygen. It can also allow for removal of the top atomic layers of a sample. The method is very slow, quantitative and chemical shift analysis can take many hours. Charging can be a problem with insulated samples. The large beam size means individual particles cannot be analysed. The method is, however, extremely valuable in identifying the relative abundance of elemental species in a sample (Eighmy et al., 1993). Whan (1986) provides more detail on the method.

Secondary Ion Mass Spectroscopy (SIMS)

Secondary ion mass spectroscopy (SIMS) is an excellent near-surface analytical technique. Elements can be quantified to trace levels. High degrees of spatial resolution are obtainable (Hochella, 1990). The method works by bombarding a specimen with an ion beam (Cs*, 02*, or 0) which sputters atomic and molecular ions from the sample into the vacuum of the instrument. The sputtered ions are then extracted into a mass spectrometer where they are quantified with magnetic and electrostatic sector analysers. Detailed methods are provided by Metson (1990). The method has been used to characterise coal fly ash (Farmer and Linton, 1984) and MSW combustion residues (Eighmy et al., 1993). Analytical times can be long (a few hours). Charging by insulated specimens is very problematic. Indium foil-mounted powders can be used to avoid this problem (Eighmy et al., 1993). It is sometimes difficult to quantify sputtering rate. Nevertheless, it is the most sensitive depth profiling method. Whan (1986) provides more detail.

Electron Energy Loss Spectroscopy (EELS)

Electron energy loss spectroscopy (EELS) is one tool to look at element valency and bonding, it is based on the principles of studying the effects of primary electron beam elastic scattering. The method is used with very thin foils. A spectrum of energy of the transmitted electron ranges from high energy loss to low energy loss. The energy loss spectrum has characteristic edges and x-ray emissions that describe the valency of an element and the bonding environment of an element. Krishnan (1990) provides details about the method.

253

X-Ray Adsorption Spectroscopy (XAS) and Extended X-Ray Adsorption Fine Structure (EXAFS)

X-ray absorption spectroscopy (XAS) is another analytical tool to characterise the bonding environment of an element. The method is rather new and requires a synchrotron radiation source that provides huge and energetic X-ray fluxes (as compared to fluxes seen in XRM or XRPD target bombardment). The method works at the 1000 ppm level and can provide detailed information of bond distance, number, and type of nearest neighbour. It works for almost all elements in crystalline or amorphous solids, liquids, and gases. It works both in bulk specimens and at interfaces.

The X-rays of precise incident energy from the synchrotron are absorbed by a sample with various degrees of absorbitivity. Very large absorption edges occur when an inner orbital electron is excited. This absorption occurs when the energy of the incident x-ray photon equals the energy required for excitation of the inner orbital electron. As x-rays of increasing energy are applied, absorption is a smooth function until a characteristic edge is encountered. Usually, multiple edges are seen, one occurring at high x-ray energy when a K-level (1 S) electron is excited and at lower x-ray energy when L-level electron excitations occur. The energy of these absorption edges is very precise. Subtle differences in the location of this edge occur because of element oxidation state, nearest-neighbour bonding, and geometry. After the absorption edge, absorbance decreases and shows modulation that is characteristic and comprises the extended component of XAS or extended x-ray absorption fine structure (EXAFS). Brown et al (1988) provide a detailed description of this complex method.

Nuclear Magnetic Resonance

Nuclear magnetic resonance (NMR) can be used to evaluate the speciation of silicon29 and aluminum-27 in solids. Kirkpatrick (1988) describes the method in detail, it has been used to examine the degree of hydration of silicon and aluminum in Portland cements (Ortego et al. 1991, Cartledge et al., 1990). The chemical shift of a 29Si nucleus in five silicate speciated states; e.g. SiO44 (QO), Si(OSi)O3 (Q~), Si(OSi)20~ (Q2), Si(OSi)~O (@), Si(OSi)4 (Q4), can be used to look at the change in condensation of silicon-oxygen tetrahedra and formation of chains of Q~ and Q2 from Q ~ units during hydration. Chemical shifts are also seen as hydration of the four-coordinate to the six-coordinate aluminate phase occurs for 2ZA1. These techniques hold promise for evaluating aging reactions in bottom ash. Analytical times range from 30 minutes to 48 hours. Care must be taken to remove magnetic material from the sample. Whan (1986) provides more details about the method.

254

Infrared Spectroscopy (IRS) and Raman Spectroscopy (RS) Infrared spectroscopy (IRS) and Raman spectroscopy (RS) involve the use of light to probe the vibrational behaviour of molecules in solid phases (McMillan and Hofmeister, 1988). Vibrational energies of molecules and crystals, in the range of 0-60 kJ/mole (or 0-5,000 cml), correspond to the wavelength of infrared light. The absorption of infrared light to promote vibration can be characteristic and is thus termed an infrared absorption measurement. Molecular vibrations can be evaluated with light scattering techniques. Changes in the energy of scattered visible light, caused by inelastic interactions with vibrational modes, promote Raman scattering. Both infrared and Raman spectroscopes generate absorption or scattering spectra as a function of energy. The spectra produced are vibrational spectra. It is dependent on interatomic forces and is therefore sensitive to microscopic structure and bonding. The methods work well for crystalline or amorphous solids, liquids, gases, and for elements of low atomic weight (McMillan and Hofmeister, 1988). Gas cells and cuvettes are used to hold gas or liquid specimens, though thin polished surfaces can be used. Packed capillary techniques can also be used for powders. Detailed preparation methods as well as interpretation of complex spectra are outlined in MacMillan and Hofmeister (1988). The method has been successfully applied to MSW fly ash (Henry et al., 1983).

REFERENCES American Public Health Association. Standard Methods for the Examination of Water and Wastewa.t_e.r, Fifteenth Edition, Washington, D.C. 1981. Bancroft, G.M. & Hyland. Spectroscopic Studies of Adsorption/Reduction Reactions of Aqueous Metal Complexes on Sulfide Surfaces. In Mineral-Water Interface Geochemist~ (M.F. Hochella, Jr. and A.F. White, eds.) Mineralogical Society of America, Washington, D.C., 1990, p. 511. Belzile, N., P. Lecomte and A. Tessier. Testing Reabsorption of Trace Elements during Partial Chemical Extractions of Bottom Sediments. En..vironmental Science and Technolo.qv, 23, 1989, p. 1015-1020. Bish, D.L. and J.E. Post. Modem Powder Diffraction, Mineralogical Society of America, Washington, D.C., 1989. Bish, D.L. and R.C. Reynolds Jr. Sample Preparation for X-ray Diffraction. In Modern Powder Diffraction (D.L. Bish and J.E. Post, eds.), Mineralogical Society of America, Washington, D.C., 1989, p. 75. Blake, D.F. Scanning Electron Microscopy. In Instrum.ental Surf.ace Analysis of Geoloclic Materials (D.L. Perry, ed.) VCH Publ., N.Y., 1990, p. 11.

255 Boutwell, G.P. and C.N. Tsai. "The Two-Staged Field Permeability Test for Clay Liners". Geotechnical News. Vol. 10, no. 2, 1992. Bridle, T.R. and S.E. Sawell. NITEP Phase I: Testin.q ..at the.Prince Edward Island Ener.av-from Waste Facility, Assessment of Ash Contaminant Leachability, Internal Environment Canada Report, 1986. Brown, G.E., Jr., G. Calas, G.A. Waychunas and J. Petiau. X-ray absorption spectroscopy and its applications in mineralogy and geochemistry. In Spectroscopic Methods in Mineralogy and Geolo.av (F.C. Hawthorne, ed.) Mineralogical Society of America, Washington, D.C., 1988, p. 431. Browning, R. and M.F. Hochella Jr. Auger Electron Spectroscopy and Microscopy. In Instrumental Surface Analysis of Geolo,qic Materials (D.L. Perry, ed.) VCH Publ., N.Y., 1990, p. 87. Bryant, J. and A. Bodocsi. precision and Reliability of Laboratory Permeability Measurements. US EPA Contract No. 68-03-3210-03, 1986. Carmichael, R.S. (1989) Practical Handbook of Physi.cal Properties of Rocks and Minerals, CRC Press, Inc., Boca Raton, Florida, 1989. Cartledge, F.K., L.G. Butler, D. Chalasani, H.C. Eaton, F.P. Frey, E. Herrera, M.E. Tittlebaum and S.-L. Yang. Immobilization Mechanisms in Solidification/Stabilization of Cd and Pb Salts using Portland Cement Fixing Agents. Environmental Science and Technology, 24, 1990, p. 867-873. Chesner, W. The Potential of Beneficial Use of Waste-to-Energy Facility Ash. Samplin.q and Testin,q Procedures. Report No. 4. Long Island Regional Planning Board, 1990. Coyne, L.M., S.W.S McKeever, and D.F. Blake. SPectroscopic Characterization of Minerals and Their Surfaces, American Chemical Society, Washington, D.C., 1990. Day, S.R. and D.E. Daniel. Fie.!d permeability test for clay liners. ASTM STP 874, 1985. Eighmy, T., D. Gress, X. Zhang, S. Tarr, and I. Whitehead. Bottom Ash Utilization Evaluation for the Concord, New Hampshire Was te-to-Ener.av Facility. University of New Hampshire, 1992. Eighmy, T.T., D. Domingo, D. St~mpfli, J. Krzanowski and J.D. Eusden. The Nature of Elements in Combustion Residues and their Stabilized Products, pp. 541-575. In: Proceedin.qs of 1992 Incineration Conference, Albuquerque, New Mexico, 1992.

256 Eighmy, T.T., D. Domingo, D. St,~mpfli, J. Krzanowski and J.D. Eusden. The Speciation of Elements in MSW Combustion Residues, pp. 457-478. In: The Proceedings.of the 1993 MSW Incineration Con.ference, Williamsburg, Virginia, AWMA, Pittsburgh, PA, 1993. Environment Canada..The National Incinerator resting and Evaluation Pro,qram: Twosta~e Combustion (P.E.I,). Report EPSl31upI1, Vols. 1 - 5, 1985. Environment Canada. The National Incinerator Testin.q and Evaluation Program Air pollution Control Technolocw. Report EPS 3/UP/2, Vols. 1 -6, 1986. . . . .

EPRI _Leachin.q Studies on UtilityS01id Wastes: Feasibility Experiments. EPRI, Palo Alto, California, 1985. EPRI Chemical F.on'n and Leacha.bi!.ity of Inor,qanic.Trace Elements in Coal Ash. EPRI, Palo Alto, California, 1987. F~llman, A-M, A. H0jlund and S. Kullberg. S.tabliserin.q och deponerinQ av r0k.Qasr.e.nin.(:lsprodukter fr~n sopf5rbr~nninq. Stiftelsen for V,~rmeteknisk Forskning, Stockholm, Br,~nsleteknik 370, 1989. Farmer, M.E. and R.W. Linton. Correlative surface analysis studies of environmental particles. En.vironmental Science and Technolo.qv, 18, 1984, p. 319-326. Furuya, K., Y. Miyajima, T. Chiba and T. Kikuchi. Elemental Characterization of Particle Size-Density Separated Coal Fly Ash by Spectrometry, Inductively Coupled Plasma Emission Spectrometry, and Scanning Electron Microscopy-Energy Dispersive X-ray Analysis. Environmental Science and Tec.hnolo.qy, ;?.1, 1987, p. 898-903. Hansen, L.D., D. Silberman and G.L. Fisher. Crystalline Components of StackCollected, Size-Fractionated Coal Fly Ash. Environmental Science and Technolo.qy, 15, 1981, p. 1057-1062. Hartl~n, J and J. Rogbeck. Sorted Incinerator Sla.q .U...sedas Fill Material. International Conference on Municipal Waste Combustion. Proceedings, Vol. 1, Hollywood, Florida, 1989. Hawthorne, F.C. Spectroscopic Methods in Mineralogy and Geolo.clV, Mineralogical Society of America, Washington, D.C., 1988. Head, K H. _Manual of Soil Laboratory Testinq. Soil Classification and Compaction Tests. Engineering Laboratory Equipment Limited, Vol. 1, London. Henry, W.M., R.L. Barbour, R.J. Jakobsen and P.M. Schumacher. Inor.qanic Compound Identification of Fly Ash Emissions From Municipal Incinerators, EPA600/53-83-095, U.S. EPA, Cincinnati, Ohio, 1983.

257 Hinds, W.C. Aerosol Technolo.qy. Wiley Interscience Publications, 1982. Hochella, M.F. Jr. Auger Electron and X-ray Photoelectron Spectroscopes. In Spectroscopic Methods in MineraloQvand Ge01oqy(F.C. Hawthorne, ed.) Mineralogical Society of American, Washington, D.C., 1988, p. 573. Hochella, M.F. Jr. Atomic Structure, Microtopography, Composition, and Reactivity of Mineral Surfaces. In Mineral-Water Interface .Geochemistry (M.F. Hochella, Jr. and A.F. White, eds.) Mineralogical Society of America, Washington, D.C., 1990, p. 87. Hulett, L.D. and A.J. Weinberger. Some Etching Studies of the Microstructure and Composition of Large Aluminosilicate Particles in Fly Ash from Coal-Burning Power Plants. E.nvironmental Science and Technolo.cw, 14, 1980, p. 966-970. Hutchinson, C.S. Laboratory Handbook of Petr.o.graphic Techniques, John Wiley and Sons, N.Y., 1974. Hvorslev, M. Time la,q in the Observation of Ground Water Levels and Pressur_es. US Army Waterways Experiment station, Vicksburg, Missouri, 1949. Jacobsson, T. V~,qtekniskI,~.qesrapportfr,~n provv~.amed sla.a.qrus.j ft~rst~rknin.aslaqer, SYSAV, Maim6, 1989. Jacobsson, T. and P. H5beda. Provv,~gsfOrsOkG~rstad 1987. Sla.a.qerfr~n k.o.I-och sopf(~rbr~.nnin.q - L~gesrapport 88-03. Statens V,~g-och Trafikinstitut, LinkSping, 1988a. Jacobsson, T. and P. HSbeda. Re.stprodukter fr~n trycksatt .virvelb,~dd som v~.amaterial. Statens V~g-och Trafikinstitut, Link6ping, 1988b. Kevex Instruments, Inc..Energy Disper.sive X-Ray Micr0analysis: An Introduction, Kevex Instruments, Inc., San Carlos, California, 1989. Kirby, C.S. and Rimstidt, J.D. Mineralogy and Surface Chemistry of Municipal Solid Waste Ash. Environ. Sci. Technol,, 27, 1993, p.652-660. Kirkpatrick, R.J. MAS NMR Spectroscopy of Minerals and Glasses. In _Spectro.,scopic Methods in Mineralo,qy and Geolo.ay (F.C. Hawthorne, ed.) Mineralogical Society of America, Washington, D.C., 1988, p. 341. Krishnan, K.M. Electron Energy Loss Spectroscopy. In Spectroscopic Characterizatj.on of Minerals and Their Surfaces (L.M. Coyne, S.W.S. McKeever, and D.F. Blake, eds.) American Chemical Society, Washington, D.C., 1990, p. 54. Lichtensteiger, T. Personal Communication, 1992.

258 MacKinnon, I.D.R. Thin Film Elemental Analyses for Precise Characterization of Minerals. In Spectroscopic Characterization of Minerals and Their Surfaces (L.M. Coyne, S.W.S. McKeever, and D.S. Blake, eds.) American Chemical Society, Washington, D.C., 1990, p. 32. McMillan, P.R. and A.M. Hofmeister. Infrared and Raman Spectroscopy. In Spectroscopic Methods in Mineralo,qy and Geolo.o,y(F.C. Hawthorne, ed.) Mineralogical Society of America, Washington, D.C., 1988. Metson, J. Secondary Ion Mass Spectrometry. In Instrumental Surface Analysis of Geolo.qic Materials (D.L. Perry, ed.) VCH Publ., New York, 1990, p. 311. Mogk, D.W. Applications of Auger Electron Spectroscopy to Studies of Chemical Weathering. In Reviews...ofGeophysics (G. Sposito, ed.) American Geophysical Union, Washington, D.C., 1990, p. 183. Nesse, W.D. !ntroduction to Optical Mineralo.qv, Oxford University Press, New York, 1991. Nirel, P.M.V. and F.M.M. Morel. Pitfalls of Sequential Extractions. Water Research, 24, 1990, p. 1055-1056. Ortego, J.D., Y. Barroeta, F.K. Cartledge and K. Akhter. Leaching Effects on Silicate Polymerization. An FTIR and 29Si NMR Study of Lead and Zinc in Portland Cement. Environmen..tal Scie.nce and Technology, 25, 1991, p. 1171-1174. Perry, D.L., J.A. Taylor and C.D. Wagner. X-ray Induced Photoelectron and Auger Spectroscopy. In Instrumental Surface Analysis of Geolo,qic Materials (D.L. Perry, ed.) VCH Publ., New York, 1990, p. 45. Reynolds Jr., R.C. Principles of Powder Diffraction. In Modern Powder Diffraction (D.L. Bish and J.E. Post, eds.) Mineralogical Society of America, Washington, D.C., 1989, p.l. Riddick, J.A., W.B. Bunger and T.K. Sakano. Organic Solvents: Physical Properties and Methods of Purffica.tion, John Wiley and Sons, New York, 1986. Satake, Hisano en Ikeda. The Rapid Determination of Sulphide, Thiosulphate and Polysulphide in the Lexiviation Water of Blast-furnace Slag By Means of Argentimetric Potentiometric Titration, Bull. Chem. Soc. Jpn., Vol. 54, No. 7, 1981. Sawell, S.E., T.R. Bridle and T.W. Constable. N!TEP Phase I1" Testin.q of the FLAKT Air Pollution Control Technolo,qy at the Quebec City M.unicipal Enerclv,from-Wa_ste .F.acility: Assessment of Contaminant Leachability, Environment Canada Report, Manuscript Series IP-70, 1986.

259 Schmith, N.B. Incinerator Residue as a Pavement Material III. Laboratory tests, 1984. Sj6holm, M., P. Carlsten and P. Elander, P. Best,~mning av permeabilitet hos restprodukter och jord, in situ och p~ laboratorium. Nordtest, Contract 1100-93, 1994. Smith, D.K. Computer Analysis of Diffraction Data. In Modern Powder Diffraction (D.L. Bish and J.E. Post, eds.) Mineralogical Society of America, Washington, D.C., 1989, p. 183. Snyder, R.L. and Bish, D.L. Quantitative Analysis. In Modern Powder Diffraction (D.L. Bish and J.E. Post, eds.) Mineralogical Society of America, Washington, D.C., 1989, p. 101. St~mpfli, D..Final Report - Cements and Bottom Ash Chemistry, ERG Report, 1992. Stegemann, J and P. Cot~. Proposed Evaluation. Protocol for Cement-based Solidified Wastes. Environmental Protection Series, Report EPS 3/HA/9, Environment Canada, 1991. Streudel, R., Holdt and G0bel. "Ion-pair Chromatographic Separation of Inorganic Sulphur Anions Including Polysulphide, J. of Chrom., 475, pp 442-446, 1989. Swedish Road and Traffic Research Institute (VTI), Report No. 31, 1973 (in Swedish). Tessier, A. and P.G.C. Campbell. Partitioning of Trace Metals in Sediments. In Metal Speciation: Theory, Analysis, and Application (J.E. Kramer and H.E. Allen, ed.s) Lewis Publ., Chelsea, Michigan, 1988, p. 183. Tessier, A. And P.G.C. Campbell. Comment. Water Res., 25, 1991, pp. 115-117. Theis, T.L., M. McKiernan and L.E. Padgett. Analysis and Assessment of Inciner.ated Municipal Sludge Ashes and Leachates. EPA-600/2-84-038, U.S. EPA, Cincinnati, Ohio, 1984. Tighe, N.J. Experimental techniques. In Electron Microscopy in Minera!o.qy (H.R. Wenk, ed.) Springer-Verlag, Berlin, 1976, p. 144. United States Environmental Protection Agency (US EPA). Test Methods for Evaluatin.q Solid Waste - Volume 1A: Laboratory Manual Physical/Chemical Methods, 3rd Edition, SW846, Office of Solid Waste and Emergency Response, Washington, DC, November 1986. VEABRIN. Veabrin Kwa!iteitskontrole van AVI-Slakken '87-'88, RAP-305/JJS/avd. 1988.

260 Vehlow, J., G. Pfrang-Stotz and J. Schneider. Behandlung, Verwertung, KFK report 5000, 1992.

Rest stoffe-Charakteriserung,

VGB Arbeitsgruppe. TOC Bestimmunq an Schlacken, Essen, July 1992. Vogel, A. Quantitative InorQanic Analysis, 3rd Edition, Published by Longmans, London, 1961. Wastewater Technology Centre (WTC). WTC Laboratory Analytical Methods, Internal Manual for WTC, updated 1993. Wenk, H.-R. Electron Micro.scopv in Mineralo.av, Springer-Verlag, Berlin, 1976. Whan, R.E. Metals Handbook, v. 10- Materials Characterization. American Society for Metals, Metals Park, Ohio, 1986. Zussman, J. Ph.ysicalMethods in Determinative Mineralogy, Academic Press, London, 1967. STANDARDS

ASTM C 29. Standard Test Method for Unit Weight and Voids in Aggregate. ASTM C 33. Standard Specification for Concrete Aggregates. ASTM C 88. Standard Test Method for Soundness of Aggregates by Use of Sodium Sulfate or Magnesium Sulfate. ASTM C 109. Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-in. or 55 mm Cube Specimens). ASTM C 117. Standard Test Method for Materials Finer than 75-m (No. 200) Sieve in Mineral Aggregates by Washing. ASTM C 127. Standard Test Method for Specific Gravity and Absorption of Coarse Aggregate. ASTM C 128. Standard Test Method for Specific Gravity and Absorption of Fine Aggregate. ASTM C 131. Standard Test Method for Resistance to Degradation of Small-Size Coarse Aggregate by Abrasion and Impact in the Los Angeles Machine. ASTM C 136. Standard Test Method for Sieve Analysis of Fine and Coarse Aggregates.

261 ASTM C 311. Standard Test Method for Sampling and Testing Fly Ash or Natural Pozzolans for Use as a Mineral Admixture in Portland-Cement Concrete. ASTM D 422. Standard Test Method for Particle-Size Analysis of Soils. ASTM D 698. Test Method for Laboratory Compaction Characteristics of Soil Using Standard Effort (12,400 ft-lbf/ft 3 (600 kN-m/m3)). ASTM D 1556. Standard Test Method for Density and Unit Weight of Soil in Place by the Sand-Cone Method. ASTM D 1557. Test Method for Laboratory Compaction Characteristics of Soil Using Modified Effort (56,000 ft-lbf/ft 3 (2,700 kN-m/m3)). ASTM D 1633. Standard test Method for Compressive Strength of Molded Soil-Cement Cylinders. ASTM D 1863. Standard Specification for Mineral Aggregate Used on Built-Up Roofs. ASTM D 2167. Standard Test Method for Density and Unit Weight of Soil in Place by the Rubber Balloon Method. ASTM D 2216. Standard Test Method for Laboratory Determination (Moisture) Content of Soil and Rock.

of Water

ASTM D 2434. Standard Test Method for Permeability of Granular Soils (Constant Head) ASTM D 2922. Standard Test Method for Density of Soil and Soil-Aggregate in Place by Nuclear Methods (Shallow Depth). ASTM D 2937. Standard Test Method for Density of Soil in Place by the Drive-Cylinder Method. BS 1377:1975. Methods of test for Soils for civil engineering purposes. DIN 18123. Bestimmung der Korngr6ssenverteilung. DIN 66115. Particle size analysis. Sedimentation analysis in the gravitational field: pipette method. SS 027124. Geotechnical tests - Particle size distribution - Sedimentation, hydrometer method

This Page Intentionally Left Blank

263

C H A P T E R 8 - FATE OF ELEMENTS DURING INCINERATION

The municipal solid waste stream consists of components containing an almost unlimited number of chemical compounds and virtually every element. The intent of burning this waste is to destroy the organic constituents and to convert the inorganic species into essentially inert materials. Since all chemical processes are controlled by equilibria, complete conversion may not be achievable in some cases. Furthermore, side-reactions may take place causing the production of unwanted by-products. Many transformations in the combustion chamber or in the resulting residues from an incinerator may produce compounds which have the potential to impact on the environment through contamination of air, soil or groundwater. These reactions have to be minimised by appropriate control of the combustion process and all other processes associated with MSW incineration and/or by adequate treatment of postcombustion residues. To develop suitable technical measures to minimise the potential environmental impact, the behaviour of the elemental species in each part of the incinerator should be understood. This chapter is a compilation of the present knowledge in this field. Much of the focus has been placed on the elements which are of greatest concern, either because of their concentration or their toxicity. The elements have been classified based on their partitioning between the different residue streams of an incinerator facility. Matrix or lithophilic elements are those elements which are considered tightly fixed in the waste fuel. Although they may undergo chemical or geochemical changes on the grate during combustion, the greatest proportion appears in the bottom ash. Volatile elements are subdivided into two groups: elements forming acid gases, like halogens or sulphur, and volatile trace metals. Both these groups of elements account for the major release of contaminants into the flue gas stream. The difference between matrix or volatile elements is caused by the different vapour pressure curves of the individual elements or their compounds, which implies that the most dominant influence in ash chemistry is the temperature in the fuel bed. A third group is represented by the organic compounds. Due to their thermal instability, organics are virtually totally destroyed during combustion. However, special trace byproducts may have the ability to synthesise new compounds during the cooling phase of the flue gases and thus may cause difficulties with contamination of the different output streams of an incinerator. Much of the data presented here are gleaned from several research programs or investigations of full scale incinerators. The attempt is made to interpret the observed

264 effects based on the chemical and physico-chemical properties of the elements and on the temperature dependent vapour pressure equilibria. 8.1 MECHANISMS CONTROLLING THE FATE OF ELEMENTS 8.1.1 Fundamentals

In order to adequately describe the chemical reactions and physico-chemical transformations which control the behaviour of elements and compounds during incineration, a fundamental knowledge of the physical parameters, chemical concentrations and the speciation of elements is required. Although most of the important physical parameters, such as temperature and pressure can be measured, not all of them can be measured to the degree of accuracy needed for physico-chemical calculations. Much of the attention given here is to the measurement of temperature, since it controls nearly all other physical parameters, especially the essential vapour pressures, as well as the chemical equilibria and the kinetics of chemical reactions. Temperatures in the gas phase of combustion chambers are well documented and are subject to legislative regulations. The bed temperature and the pertinent residence time of the bed material are the most important measurements for calculating the release rates of thermally mobile species. Both parameters influence the extent of immobilisation of elemental phases in the bottom ash. Unfortunately, there is a dearth of accurate data of fuel bed temperatures which are required to understand some of the reactions which take place. Although recent speciation data suggest that Iocalised areas of the fuel bed can reach substantially higher temperatures than 1,000~ (Eighmy, 1993), we have assumed that average temperatures on the burning section of the grate of a modern mass burning MSW incinerator is between 800 and 1,000~ Nearly all chemical compounds of importance are well characterised. Their behaviour is controlled by natural laws and can be predicted if certain parameters are established. The current knowledge is based on investigations in ideal systems, i.e., low quantities of one or a few compounds in well-defined closed systems. If concentrations or numbers of reactants or both increase, the exact correlations fail and the measured effects can only be approached by introduction of activity coefficients that try to describe unknown interactions. The concentrations of elements, especially their local distribution, show substantial variations. Waste is not only a physically but also a chemically inhomogeneous material. Consequently, the application of chemical and physical laws on elements in the field of municipal solid waste incineration is limited. Most of the following is based on best practical knowledge.

265 8.1.2 Processes in the Combustion Chamber Physical Parameters As described in Chapter 3, there are a great variety of incineration technologies in use today, all sharing some common processes. In most systems, the temperature of the untreated waste charged to the furnace is similar to the ambient temperature of the storage pit. As the waste moves down the grate, it heats up to temperatures above the point of ignition. The schematic in Figure 8.1 illustrates the pronounced temperature profile established along the grate and also between the top and bottom of the fuel bed on the grate. In most countries, the minimum gas temperature must exceed 800 to 850~ although the maximum temperature is normally kept below 1,100~ to prevent corrosion of the walls of the combustion chamber and the heat recovery system. Figure 8.1 Temperature Zones on the Grate

1" flue gas > 850 ~

feeding < 50~E:

The radiant heat emitted from the hot furnace walls and burning waste causes a temperature profile across the depth of the fuel bed with higher temperatures on top and lower temperatures at the bottom. As the waste is exposed to these temperatures, the moisture is driven off into the flue gas. Further heating of the dried waste increases the temperature of the fuel to some 100~ and pyrolysis of organic compounds starts. At about 500~ the ignition temperature is reached and the waste combusts. The remaining part of the grate is used to achieve a good burnout by extending the residence time of the waste to high temperatures. The maximum temperature in the fuel bed is generally found at the transition zone between the combustion and burnout zone. The maximum temperature in the gas phase is just above the top of the flame bed, downstream of the secondary air injection ports. For safety reasons, the combustion chamber is designed to operate under a

266 slightly negative pressure. Since all of the physico-chemical parameters are a function of pressure, the following calculations are based on standard pressure as a first approximation. Chemical Reactions Reactions of Main Waste Components The energy production in the combustion chamber is essentially caused by the exothermic oxidation of the carbon and the hydrogen inventory of the waste. An ultimate analysis of municipal solid waste is seen in Figure 8.2

Figure 8.2 Ultimate Analysis of Waste N 0.4%

H 3.7%

CI 0.8

too;

Adapted from Environment Canada, 1988 In the drying zone of the grate, no chemical reactions of importance are to be expected. The mean humidity of waste is in the order of 30% and the evaporated water adds about 60 g/Rm 3 to the water content of the flue gas. While the temperature increases, highly volatile organic and/or inorganic compounds may be transferred into the gas phase. However, these products will not survive when they pass through the hot zones of the incinerator and thus will not appear in the flue gas in their initial gas phase form. In the pyrolysis zone at temperatures above 200~ most organic compounds start to disintegrate. The decomposition products are volatile or semi-volatile organic compounds, tar, and elementary carbon. The volatile species are oxidised in the gas phase, at least in the post-combustion region and can account for the major energy

267 release in the furnace. The fixed carbon is burned on the grate in the combustion zone. Due to the overall oxygen surplus and the high stability of oxides of most elements, oxidation processes will be the predominant reactions in the fuel bed and in the gas phase of the combustion chamber. The most important oxidation reactions are those of carbon and organic carbon compounds. The most stable oxidation product up to very high temperatures is CO2. In all cases, the first oxidation step results in the intermediate product CO which has to be suppressed and will be oxidised to CO2 if an oxygen surplus can be maintained. As an example, the oxidation of graphitic carbon is given: 2 C graphite+ 02 " 2 CO

2 CO + 02 ,* 2 CO2

(8.1)

(8.2)

The equilibrium between CO and CO2 [equation (8.2)] is shifted toward CO with increasing temperature, but at 1,200~ (commonly not exceeded in waste combustion) only about 0.03% of the CO2 molecules are dissociated (Remy, 1961). Another process, the Boudouard reaction, may be responsible for enhanced CO concentrations in the flue gas: Cso,d + CO2 "* 2 CO

(8.3)

If equilibrium could be reached at 900~ about 97% of the gaseous compounds would be present as CO. Fortunately this reaction is kinetically hampered and at temperatures below 800~ without the presence of catalysts, reaction rates cannot be measured. However, in ash deposits in the boiler where diffusiophoresis (see Chapter 10) is a controlling factor, long residence times in combination with catalytic ingredients of the ashes may be responsible for a certain portion of the CO production. The equilibrium in equations (8.2) and (8.3) indicate that the production of CO can be minimised if care is taken to ensure proper oxygen distribution in the combustion zone and if the release of particulate carbon into the flue gas can be avoided. Under these conditions, virtually all of the carbon in the waste (about 25%) should be transformed to CO2 and result in concentrations of about 130 g/Rm 3 in the flue gas. In properly operated MSW incinerators, the CO emissions can easily be kept below 50 mg/Rm 3. The next most important element due to its concentration is hydrogen. Hydrogen is present in all organic compounds. The total inventory is transformed into the only stable oxidation product, H20. To quantify the production, a mean hydrogen concentration in the waste of about 36 kg/tonne (see Figure 8.2) corresponds with a C/H-ratio of about 7.5 (Tillman et al., 1989; Behrendt, 1992). The resulting water concentration in the flue gas is in the order of 65 g/Rm 3 which adds to the 60 g/Rm 3 of water evaporated in the drying zone of the grate.

268 The third most important element (compare Figure 8.2) is chlorine, which cannot be oxidised due to the similarity of its electronegativity to that of oxygen. During incineration, the chlorine favours the oxidation state of-1. The chlorine compounds leaving the combustion chamber are inorganic chlorides, or HCI. Since there is no elementary hydrogen available in the combustion chamber, the HCI is formed by secondary reactions like hydrolysis.

Reactions Controlling the Alkalinity

The water formed during incineration may be very aggressive to other compounds not altered by the oxygen attack resulting in secondary hydrolytic reactions. In most cases, those reactions result in the formation of hydroxides which contribute to the alkalinity of the solid residues. These reactions have a major influence on the fate of alkali and earth-alkali elements. Normally, the chlorides of these elemental groups are thermally very stable. However, in the presence of water, hydrolysis takes place. The earth-alkali chlorides are less stable than the alkali salts. CACI2"6 H20 can even be hydrolysed by its own crystal water at temperatures above 260~ if it is heated rapidly: CaCI2 + H20 .* CaO + 2 HCI

(8.4)

The product CaO is a strong base if brought into contact with water. Many compounds, especially heavy metal salts, are also sensitive to hydrolytic reaction at elevated temperatures. Other inorganic salts in the waste do not survive the combustion process due to their limited thermal stability. A well-known example in the building industry is the burning of lime: CaCO3 ,, CaO + CO2

(8.5)

With the exception of the alkali compounds, all other carbonates are sensitive to thermal decomposition, forming CO2 and the respective oxide. In the case of the earth-alkali carbonates, the thermal stability increases with the atomic number. MgCO3 and CaCO3 decompose at about 550~ and 897~ respectively, whereas more than 1,400~ is required to destroy BaCO3. This is another reaction which may contribute to the alkalinity of the combustion residues. Sulphates (with the exemption of the alkali compounds) can be thermally decomposed, forming SOx compounds in the flue gas. The most sensitive sulphates are those of trivalent metals, such as AI or Fe: 2 AI2(SO4)3 ,- 2 AI203 + 6 SO3

(8.6)

The disintegration temperatures of AI2(SO4)3 and Fe2(SO4)3 are about 750~ and 720~ respectively. If water is present in the flue gas of the combustion chamber, hydrolytic attack is also observed and the decay of these compounds may start at about 550~

269 In some cases, the high temperatures decompose sulphates and form SO2 and 02, instead of SO3. In the cooling regions on the grate, or in the flue gas stream, many of the described equilibria shift back, forming the original compounds. In most cases, however, the gaseous reaction product is no longer present and the reaction cannot be reversed. Oxides or hydroxides released as dust particles from the fuel bed have a certain probability to react with acid gas components like HCI, SO2, and CO2. Generally, the high concentration of CO2 drives the reaction toward the formation of carbonates, and hence provides a level of buffering capacity to the residues. This is consistent with evidence that ash particles from different ash streams (that have short residence times in the flue gas stream) have acid neutralisation capacities within the same order of magnitude (Sawell and Constable, 1988).

Reactions of Heavy Metals The complex behaviour of heavy metals and other trace elements during incineration has been modelled qualitatively based on Gibbs' fundamental equation (Borchers, 1989): G = H- T * S where:

(8.7)

G = molar free enthalpy H = molar enthalpy T = temperature S = molar entropy

Using thermodynamic data, the temperature dependencies of the standard formation enthalpies (&G ~ can be calculated using the Gibbs-Helmholtz equation: AG~

= AH~

+ fACp dT- T

AS0298 -

T [ACp/T dT

(8.8)

where: Cp = molar heat at constant pressure. The calculation must include all compounds of an element which can theoretically be formed under the respective conditions. The basic data needed for the calculation can be found in the literature (Barin, 1993). It is assumed that the residence time allows the system to reach equilibrium. The comparison of the calculated free enthalpies gives an indication of which compound is the most stable and will drive the thermochemical reactions. For example, the calculated free formation enthalpies of reactions for elemental lead and antimony in the temperature range between 700 and 1,100~ are shown in Figure 8.3. To concentrate on the most important reactants in the combustion chamber, only reactions with oxygen, H20 and HCI have been taken into account. Clearly, metallic lead and lead oxide prefer to react with HCI. PbCI2 is the most stable combustion product in the

270

temperature range illustrated. Although these calculations explain the behaviour of these lead compounds, similar calculations have to be carried out using all other species of lead present in the waste to provide an overall indication of the element's behaviour. Figure 8.3 Free Formation Enthalpies of Some Pb and Sb Compounds

O

~

l~) + 1/2 0~, ~--~PDO

!

Pb + 2 HC! +~/2 02 ~ PbCl2 + H20

2Sb +6UC! +3/2 02~---, 2SbCl3 + 3H20

-6

.EE

-5O0-

IE =..==

2 Sb + 3/2 02 -~ Sb203

2 SbCI3 + 3 HzO

~

Sb203 + 6 HCI

-1000-

.IE r" 0 0

,1~ - 1 5 0 0 -

- 2000700

PbO + 2 HC! ~--~

I

I

I

800

900

1000

1100

temperature in ~ In the case of antimony, the differences in formation enthalpies are not as high as lead, and therefore the prediction of its behaviour is more difficult. The enthalpies of the reactions with 02 and with CI are similar. The resulting SbCI3 may be hydrolysed, however, the actual residence times and the kinetics will limit the importance of this reaction. It must be assumed that SbCI3 and Sb203 will be the most dominant products of combustion. There is another type of reaction in the fuel bed with the potential to form volatile species. This is the solid state reaction between inorganic salts involving the exchange of cations. For example:

271 2 NaCI + ZnS04 ,~ ZnCI2 + Na2SO4

(8.9)

The reaction given in Eq. (8.9) transforms the immobile ZnS04 into ZnCI2 which can be vaporised in the combustion chamber. The formation enthalpy of this reaction is in the order of 8 kJ/mole at 400~ (Lieser and Elias, 1962). The probability of such reactions can hardly be calculated due to the unknown speciation as well as the unknown local distribution of these species. If the respective species come into contact, however, the yield will be high due to the long residence time on the grate. It should be mentioned that this type of reaction can also take place in the heat recovery system and in other parts of the flue gas system where elevated temperatures prevail. Thermodynamic data are useful to estimate the integral reactions. It must be emphasised that limitations arise because there is a lack of data about the kinetics of the reactions and the conditions for the chemical reactions (i.e. temperature and local concentrations) change substantially down the burning grate. A great deal of complex information would be required to accurately model the reactions which do occur under the variable conditions. Most reactions are prevented from reaching equilibrium. For example, if a solid piece of oxidisable metal is present in the combustion chamber, the oxidation starts at the surface by forming an oxide layer. If oxidation is to continue, the oxygen must reach the surface of the elemental metal first, which can only be facilitated by diffusion. Since the oxide layer is a dense structure, it inhibits the diffusion and thus protects the metal from further oxidation. Consequently, many metals and alloys, especially in bulky form, pass through the combustion chamber virtually unaltered by the process. Aluminum, antimony, copper, iron, nickel, titanium and lead all belong to this group. Although this is true for most metals, metals with relatively low melting points present an even more complicated set of reactions. For example, the respective melting point temperatures of aluminum and lead are 660.4~ and 327.5~ Even if a dense protective oxidised layer forms, the metallic nucleus may melt prior to further oxidation and result in molten material dropping through the grates to solidify as a metallic mass in the under-grate hoppers. A substantial amount of metallic aluminum and lead can be observed in grate siftings (Environment Canada, 1991; WASTE Program, 1993). The presence of metallic aluminum in grate siftings has two potentially detrimental affects. The first is that mixing grate siftings with bottom ash in the quench tank results in the formation of aluminum hydroxide and the liberation of hydrogen from the quench tank. As a result, most facilities provide adequate ventilation above the quench tank to avoid a dangerous build-up of the gas. The second potential problem occurs when grate siftings end up in a utilisation scenario, such as in a road base. The formation of aluminum hydroxide results in an expansion of the material, which could cause undesirable heaving in Iocalised areas. These disadvantages support recommendations suggesting that grate siftings be collected separately from bottom ash, especially if bottom ash is to be utilised (Vehlow et al., 1990).

272 If the primary combustion products are stable at the temperatures in the combustion chamber, the shape of their vapour pressure curves will ultimately dictate their further behaviour. The nonvolatile products may be incorporated into either the ashes in their original species or they may undergo further reactions forming compounds of greater stability. In most cases, these are reactions between metals and nonmetallic oxides, or solid phase reactions, which are controlled by solid state diffusion and have moderate velocities even at elevated temperatures. Nevertheless, they contribute substantially to the chemical stability of bottom ash. Volatile species are released from the fuel bed into the gas phase. Among these are pure metals which have low boiling points or at least vapour pressures high enough to be evaporated. Such is the case for aluminum. This is evident from the liberation of hydrogen from filter ashes contacted with acids or even with water (Oberste-Padberg and Schweden, 1990). A very simple example of the transfer of pure metals into fly ash is the burning of aluminum foil which results in entrainment of small pieces of foil in the flue gas stream. Pure metals such as cadmium (b.p. 765~ mercury (b.p. 356.6~ and zinc (b.p. 907~ are very likely to be volatilised during incineration. Although these metals should not be stable in the metallic state at high temperatures, a small fraction of metal vapour might survive and condensate on the dust particles at lower temperatures. The same is true for lead, however, the vapour pressure is about 1 hPa at a temperature of 950~ and volatilisation is less likely to occur. Another reason which might explain the presence of metallic forms of metals on the surfaces of fly ash is that these volatile compounds may decompose into radicals and/or atoms when passing the regions of highest temperature in the gas phase. Some iodides of heavy metals, some oxides (e.g. Pb304) and carbonyls (e.g. Ni(CO)4) are, to a substantial degree, thermally dissociated into the metals and other ionic or radical groups. In the case of iodides and carbonyls, quick condensation onto particles may enable some of the metals to be absorbed in the metallic state on appropriate surfaces. All these reactions are dependent on a number of variables, including the Iocalised temperature in the fuel bed, the local distribution of the element and the speciation of reactants. The inhomogeneity prevents the adjustment of chemical equilibria and thus only allows the prediction of gross trends. These trends, based on extended laboratory, semi-technical and full scale research programs as well as operational experience, are discussed in more detail for selected elements later in this chapter.

Sintering and Related Processes The conditions in the combustion chamber control the chemical reactions, which in turn dictate the reactivity or stability of the combustion products. To a certain extent, changes in the stability of solid residues can be achieved alone by changes in physico-chemical parameters without changing the nature of the chemical bond. The

273 best-known cases are phase transfers in single phase materials, induced by changes in temperature. For example, the sequence of stable SiO2 modifications with increasing temperature is: a-quartz

575~ 870~ 13-quartz ,-

1470~ 13-tridymite ,- 13-cristobalite

If the cooling down of those materials is fast, as in the case of bottom ash quenching, some high temperature modifications can be frozen. Residues from waste incineration are in most cases multi-phase, multi-component products. This applies especially to bottom ash. Although the temperatures in the fuel bed are generally below the melting points of the main constituents of bottom ash, a certain proportion of annealing can take place, resulting in changes to the physical and chemical properties of ash. In masses of powders and compacted porous material, the application of heat causes bonding of particles and diffusion transport of material across particle surfaces, ultimately resulting in: changes in pore structure and porosity strengthening densification crystal grain growth ("Oswald ripening") changes in electrical resistivity and other parameters. A fundamental scheme of the changes that can occur for some physical parameters is illustrated in Figure 8.4. In a porous or powdery material, the porosity starts to decrease at a certain temperature. Simultaneously the density and strength increase. Other effects, like the electrical resistivity, start to vary substantially, even at lower temperatures. This effect is called "sintering" and it is used industrially in (metal) powder technology and in the production of special ceramics. The definitions of sintering vary considerably between authors because of the complexity of the process, the different mechanisms which cause sintering and the different properties which are affected (Hausner, 1979). Furthermore, there are a number of different types of sintering, including cold sintering, pre-sintering, high sintering, pressure sintering, solid state sintering, liquid phase sintering, etc. This discussion will focus on sintering effects which are essential factors in the solid state reactions on the grate. If the original material is chemically homogeneous, sintering affects only the crystal lattice via self-diffusion of atoms or molecules across the boundaries of adjacent grains. However, if it consists of different chemical phases, solid state reactions between single components have to be considered.

274

Figure 8.4 Changes of Physical Properties versus Temperature

8

% I~" I" 4, " / ~ /

\2.,,

I ./

temperature

Hausner, 1979

The stepwise formation of silicates from a mixture of the single components CaO and SiO2 is provided in Figure 8.5 as an example of the fundamental sintering process. These compounds are by far the most common components in bottom ashes. The concentration of SiO2 in bottom ashes is in the order of 35 to 55%, whereas the concentration of CaO varies between 5 and 20% (Baccini and Brunner, 1985; Hjelmar, 1987; Eighmy et al., 1987; LOftier, 1989; Vehlow et al., 1992). Figure 8.5 Progress of Reaction Between CaO and SlO2 b)

ZaO--

1

SiO 2

Ca2SiO 4

a)

%Si% %Si97 C

SiO 2

Ca2 Si04 Ca3 Si20 7 CaSi03

Ca2SiO4 Ca3 Si20 7 CaSi03

275 The diffusion coefficient of CaO is higher than that of SiO2 and thus enables CaO to migrate into the SiO2, forming different phases. The first step is always the formation of the ortho-silicate Ca2SiO4. Depending on the local ratio of components, new phases may be formed. The final product is the most stable phase, the meta-silicate CaSiO3. The number of phases present in an equilibrium is regulated according to Gibbs' phase law and depends on the number and the molar ratio of the initial components. Although calcium is the predominant cation in bottom ash, other cations of earth-alkali and alkali metals can neutralise silicate anions. Thus, the principal reaction shown in Figure 8.5 produces a series of different silicates and meta-silicates with mainly calcium, magnesium, sodium and potassium. Aluminum and iron can also act as carriers of the positive charge. The different silicates are distinguished by different crystal lattices depending on the radii of the single ions. These radii and their ratios determine the coordination number and thus the type of lattice. Crystallographic investigations using microscopic and X-ray diffraction methods have revealed well-defined crystals of several silicates generated during combustion (see Bottom Ash Characteristics). In all defined crystals, the components of the lattice can be replaced by other ions if: these have a corresponding chemical behaviour. the ionic radii are similar. they have the same chemical valence. they tend to crystallise in the same type of lattice. Consequently, many trace metals can be incorporated into stable matrices. The data in Table 8.1 is a compilation of crystal ionic radii from the literature (e.g. Pauling, 1969; Handbook of Chemistry and Physics, 1992) which indicates the important heavy metals which could replace the main constituents in silicates. For example, cadmium may replace the calcium found in carbonates or silicates. Table 8.1 Crystal Ionic Radii (nm)

.

Na§ 0.097

K* 0.133

Ca2. 0.099

Fe2. 0.074

Fe3. 0.064

Cu§ 0.096

In§ 0.132

Ti 2. 0.094

Mg 2. 0.066

Cr3§ 0.063

TI§ 0.140

Cd2. 0.097

Co2. 0.072

Mn 3. 0.066

Ni2. 0.072

Co 3. 0.063

Cu2. 0.072

Ni 3. 0.062

Zn 2. 0.074 Adapted from Pauling, 1969; Handbo0k of Chemistry and Physics,-1993

276 The tetravalent Si 4§ ion in the silicate lattice has a crystal radius of 0.042 nm and can be replaced by AI3§ ions with a radius of 0.051 nm to form aluminosilicate. These compounds are widespread in the lithosphere. The most common group is the feldspars, found in bottom ash in concentrations between 5 and 10% (Vehlow et al., 1992). The same study measured other aluminosilicate, such as the melilithes, in concentrations above 10%. At the other end of the scale, some aluminosilicate found in bottom ash are artificial compounds that can only be formed within very small temperature ranges. This can be used to estimate the maximum temperatures in the fuel bed. To a certain extent, trace elements can also be incorporated into silicates (as well as other lattices) as impurities, bond to lattice defects or precipitate onto grain boundaries. Many silicates, especially double silicates consisting of sodium and calcium, show no defined transformation temperature when cooled down and do not crystallise, but convert into a glassy state. Glasses are defined by non-periodic lattices which have large areas without any symmetry. Nevertheless, their atoms are grouped and exhibit the same coordination number in their neighbouring spheres as they do in crystals. The mean composition of a "normal" glass can be defined with the formula: Na20 * CaO *6 Si02 Glasses generated on the grate during the combustion process are characterised by typical streak patterns and bubbles, which can be found in bottom ash in relatively large proportions of more than 10 % (Vehlow et al., 1992). Glasses have a greater potential to incorporate trace metal ions than crystals, mainly due to their lack of symmetry and the normal inhomogeneity of their structure. Substantial concentrations of metals in glass are generally denoted by distinct colours. Metals may also become embedded with small particles of other materials in the glass. If these particles become totally encapsulated, the potential for mobility through leaching is reduced significantly. Single oxides of metals are also present in bottom ash. Magnetite (Fe304) is the most prevalent compound (up to 10%). Fe203, FeO, Cr203 and AI203 are also present, but in concentrations below 5%. A special type of mixed oxide consisting of bivalent and trivalent metals is represented by the group of spinelles. These form defined crystals with the representative molecular formula: AB2 04 where:

A = bivalent metal B = trivalent metal.

Common spinelles can contain the trivalent ions of AI, Cr and Fe, and the bivalent ions of Mg, Mn, Fe, and Zn. Trace elements can be substituted for some of the main

277 components and thus become fixed in the solid phase. Many different spinelles have been observed in bottom ash, however, they typically represent only a small fraction of the total of mineral phases. In general, chemical reactions in the combustion chamber produce thermodynamically stable compounds. The elements which remain in the fuel bed and form bottom ash are mainly bound to Si, AI, and O. Similar mineral matrices can also be found in fly ash particles entrained in the flue gas stream (Hundesr0gge, 1990). The quantity of fly ash generated is influenced by the type of grate and the control of air flow within the combustion chamber. Modern incinerator technologies are based on combustion with lower air velocities through the fuel bed and enhanced mixing of the gas stream in the furnace which severely limits entrainment of particulates. This also has the benefit of reducing the production of polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzofurans (PCDF) in the heat recovery and APC systems.

Physicochemical Transformations

Most substances can exist as solids, liquids or gases, depending on the actual temperature, however, there are some notable exceptions. For instance, the liquid phase of iodine, carbon or carbon dioxide cannot be achieved under normal pressure conditions. Other compounds, such as all organics, heavy metal carbonates, several oxides (e.g., Pb304, HgO), and heavy metal sulphates, will simply decompose if the temperature is too high. In most substances, the transition between solid and liquid phases takes place at one defined temperature, the melting point, which is a function of pressure. Although solid and liquid phases of a species can only be in equilibrium at one defined temperature, both phases are in permanent equilibrium with the gas phase. Due to the statistical distribution of oscillation energy in the lattice or of kinetic energy in the liquid, some atoms or molecules are able to cross the interface and evaporate into the gas phase. This transition is stimulated by increasing temperature, until the substance reaches its boiling or sublimation point above which all molecules are present in the gas phase. Consequently, the gas phase will always contain molecules of all species present in the system, even at very low temperatures. These gas molecules move freely through the gas volume and have the potential to impinge on the surface of the liquid or solid and return, or condense back into the original matrix. Condensation is a function of the concentration of the respective molecules in the gas phase. Equilibrium is reached if the number of particles crossing the interface to evaporate is exactly the same as the number of particles striking the interface and returning into the condensed phase. The equilibrium is, like all chemical equilibria, a function of temperature and pressure. At each temperature, the equilibrium concentration of the chemical species in question causes a defined partial pressure of this species in the gas phase. The temperature dependency of this partial pressure, called the vapour pressure of the species, can be described using fundamental physico-chemical

278 properties of the species. The conditions for equilibrium between a condensate and a gaseous phase are based on 9

R log

P pO

.

G ~176

.

where:

.

T

G %`` (T)

5 G o (T)

T

T

.

.

R = gas constant G = molar free enthalpy o = standard conditions

(8.10)

p = (vapour)pressure T = temperature

Using the definition of free enthalpy given in equation (8.7): G=H-TS Equation (8.10) can be transformed to: R log--PP :- H%"(T) -H~ ....(T) +sOq.,(T) _sO~o.d(T)=- 6H~

pO

T

T

+DSO(T )

(8.11)

using C

Sgas (T) = Cpg~ iogT+[0 T og~ T dt+D1

(8.12)

and: C

Scond (T) =f0 T Pcood T dt +DI

(8.13)

where cp = molar heat under constant pressure Co = molar oscillation heat D = constant Simplified, the molar enthalpies and the integrals can be described by the linear functions of T (only valid for limited temperature ranges), and the temperature dependency of the vapour pressure can be approached by: A

1 ogp=--- +B+Cl o g T + D T T

(8.14)

279

where A, B, C, D = constant. This is the type of equation used to interpolate experimentally obtained vapour pressure data. In small temperature ranges the equation can be even further simplified to:

A

(8.15)

1 ogp=--- +B T

Figure 8.6 compiles the vapour pressure curves for selected species of some elements which are known to be more or less mobile during incineration. The curves are calculated using vapour pressure formulas from Bartels et al. (1960). Figure 8.6 Vapour Pressure Curves for Selected Metals in the Metal, Oxide and Chloride State 1000.0

oxide

r t,"

:3

chlodde 100.0

r l ,

I

9i

.I:

10.0 I.:.-~-.-:~:..-~..--,/-:

Q. =-.

/../ / / , ' i ~:!/,." ...."//

I I

1.0

0.1

,

, 9. . . .

u

"

100

i if:.// :

'r:,.":l

ii,://E 9: e :

300

i

500

i / _'~ I ....

..."Jr

....

/

I:,

! /"1

.....

).-'.,,-/

i ../i,'/!

.

/

1 / /

. . . . . . . .

i,.71 :

Yml

.

.

metall

i

:: .....

i

i .....

=l

!

Hg As Sb Cd ~

Zn

. I l l .

.

-

II

=l

=l

Pb

=,,

7 0 0 g 0 0 1100 1300 1500 Temperature in ~

If the exact chemical formula of a species in the gas phase at a given temperature is known, by means of the fundamental gas equation" pV=nRT where

n = number of moles.

(8.16)

280 The equilibrium concentration of this species in the gas phase can be calculated, providing an estimate of the potential partitioning. In a limited volume of gas, achieving vapour pressure equilibrium is a relatively rapid process. The major controlling factor is the time the evaporated atoms or molecules need to establish a uniform concentration in the gas phase by diffusion. In the combustion chamber of an incinerator, equilibrium will normally not be reached, due to the large volume of fresh air used to maintain the combustion process. Hence, the quantity of evaporated substances cannot be calculated precisely, even if all concentrations and the temperature distribution in the combustion chamber were known. Nevertheless, the shape of vapour pressure curves allows an estimate of the degree of evaporation of a substance, and thus indicates the potential partitioning of a substance between the bottom ash and flue gas stream. 8.1.3 Mechanisms in the Boiler Condensation The hot flue gases from the furnace enter the boiler at temperatures of <800~ and are cooled rapidly to about 200~ or lower. The decrease in temperature across the boiler prompts condensation of many gaseous species entrained in the flue gases. Condensation is also controlled by vapour pressure. The condensation process in a dust-free gas phase starts as soon as the temperature falls below the boiling point by forming a condensation nucleus. However, the condensation process is kinetically inhibited in a saturated atmosphere and nuclei begin to form at lower temperatures.

Supersaturation of the flue gas stream is minimised due to the loading of dust particles which act as condensation nuclei. Condensation takes place mainly at the surface of dust particles. Since smaller particles have higher specific surface areas, the finest fraction of fly ash particles is generally the most enriched with condensated material. Furthermore, the enrichment of special compounds in the fine dust fraction indicates the thermal mobility of these compounds under the operating conditions present in the combustion chamber. There are indications that fly ash from modern MSW incinerators is characteristic of enhanced enrichment of volatile species compared to fly ash from older facilities. The enhanced enrichment is probably due to a combination of high bed temperature, which promotes the vaporisation of volatile compounds, and a more "soft" combustion which causes a reduction in the overall mass of inert material carried over from the combustion chamber (e.g., Sawell and Constable, 1989; Horch, 1990). Although the boiler's primary function is to act as a heat exchanger, it also acts to a limited extent as a dust removal system. Particles which are either coarse enough or dense enough to fall out of suspension in the flue gas stream collect in the bottom of the boiler chambers. These ash particles are not as enriched with volatile substances as residue collected in APC systems. The main reasons for the difference are:

281 The higher temperature of the flue gas in the boiler limits condensation of volatile components. The condensation reaction favours enrichment of finer sized particles, namely the APC residues rather than the relatively larger sized boiler ash particles. Although reactions between the acid gases and alkaline dust particles take place in the boiler, they do not cause substantial changes in the overall partitioning of elements. However, some evaporated species may undergo special reactions with other reactants during transport through the boiler, resulting in a difference in speciation of the condensated compounds compared to those remaining in the flue gas. For example, H20 concentrations in the flue gas can reach values up to 170 g/Rm 3 and in cases of soot blowing of the boiler tubes using steam, levels can reach above 200 g/Rm 3. At the comparably high temperatures in the boiler, this water may cause hydrolytic reactions with organic compounds, but also with inorganic salts (see above: Reactions Controlling the Alkalinity). Corrosion Another reaction which is often neglected with respect to the mass transfer, but is of vital importance for the maintenance of the boiler, is corrosion attack of gases and salts on the heat exchanger tubes (F,~l~ler et al., 1968). The materials used for the tubing are non-alloy or low alloy steels. Waterwall tubes are often coated with a castable refractory like SiC. In the superheater region, sometimes high alloyed austenitic steels have been used (Reichel and Schirmer, 1989). Principally two temperature regions can be identified where severe corrosion of the ferrous tubing material is observed. Figure 8.7 depicts the dependency of the corrosion rate as a function of the wall temperature of the tubes. At temperatures below 150~ condensation of acids takes place and the low alloy steels suffer tremendous electrochemical corrosion attack. High alloy steels show pit corrosion and due to a combination of acid attack and mechanical stress loading, corrosion cracking may occur. This so-called "standstill" corrosion can be avoided by diluting the flue gases during shutdown procedures with air or by controlled cooling down of the incinerator using fuels with low chlorine and sulphur levels. A second region of potential corrosion attack is observed at temperatures above 300~ with increasing effects at increasing temperatures. The waterwall tubes in the combustion chamber and the superheater tubing are especially at risk. The flue gas temperature may exceed 800~ at these locations and the wall temperatures in the superheater sometimes reach values above 500~

282 Figure 8.7 Schematic Dependency of Corrosion Rates on Wall Temperature

~

s

I chemical corrosion

corrosion in the gas phase I

0

S

200

i

I

|

400

!

600

!

800

Temperature in ~

Although there are substantial amounts of corrosive gases present in the flue gas, corrosion effects caused by direct interaction with gaseous HCI or SO2 can only take place during the start up of a new boiler. After a short period of operation, the heat exchanger tubes are normally covered by layers of ash (see Chapter 10) and there is no free access of gases to the tube surfaces. The dotted curve in Figure 8.7 compares qualitatively the corrosion rates caused by pure gas corrosion with the corrosion rates actually found in boilers. Within the temperature regime in the boiler, corrosion attack on the covered tubes exceeds that which would be expected by gas corrosion. These corrosion effects must be induced by ingredients of or by reactions in the surface deposit layers. The deposits on the heat exchanger tube surfaces contain salts which may produce corrosive species by disintegration or by reaction with components of the flue gas. One example may be the formation of SO3 close to the tube walls by thermal decay of sulphates [equation (8.6)]. The theoretical production rate should be very low and no major corrosion damage could directly be traced back to this component. Sulphates with low melting points, especially complex alkali-iron sulphates can be formed in the temperature range 320 to 450~ following the reaction (Corey et al., 1965): Fe203 + 3 K2SO4 + 3 SO3 ,- 2 K3Fe(SO4)3

(8.17)

This reaction has been blamed for severe corrosion attack since it may induce a mechanism which recycles the corrosive agent (Figure 8.8).

283 Figure 8.8 Mechanism of High Temperature Sulphate Corrosion

Fe

55o_ { 2o3 }

+ K3Fe(SO4)3 700 ~C

KzSO4

1

+[eo o,J Nelson and Cain, 1960 Although a small proportion of corrosion damage may be explained by this mechanism, in most cases it is difficult to measure the respective corrosion products (F~131eret al., 1968). Furthermore, the formation of the complex sulphates requires SO3 concentrations exceeding 250 mg/Rm3, which is far above normal levels in the flue gas (<50 mg/Rm3). The high corrosion damage in boilers of MSW incinerators compared with those in coal fired power plants can be explained by the higher chlorine inventory in MSW than coal. The higher chlorine content of MSW flue gases has prompted development of a corrosion mechanism based on the reactivity of chlorine compounds, including elementary chlorine which is shown schematically in Figure 8.9. The basic corrosion reaction in the mechanism postulated above is the formation of FeCI2 at the metal surface and the direct recycling of elementary chlorine. Its presence at the corrosion front prevents the growth of the tightly bound, protective oxide layers and thus is responsible for the relatively high corrosion rates. According to this corrosion model, chlorides are the essential components in the deposits. The formation of elementary chlorine by sulphation is favoured at high temperatures. A second mechanism (Equation 8.18), is also a primary step in the de-novo synthesis of PCDD and PCDF (the Deacon Process), which is initiated at temperatures >350~ 2 CuCl2 + 02 " 2 CuO +2 CI2 2 CuO +4 HCI ,, 2 CuCl2 +2 H20

(8.18)

284 Figure 8.9 Chlorine Induced Mechanism of High Temperature Corrosion

N

N //

flue gas ash deposit (sulphates, chlorides, silicates) outer oxide layer (Fe203 + ash) inner oxide layer (Fe304 + FeS) corrosion front (FeC12)

S~ 2 ? 2

! C12 4- Na2804

SO2 + 02 +

<

~

~

3 Fe203 3 Cl 2 + Fe304

FeS + 4 ~'12+

2 NaCI

t/'2 0 2 + Fe304

< <

2 02 + 3 FeCI2 SO2 + 02 + 4 F~C12

Fe304

Cb. + Fe

>

/

FeC12

tube wall (low alloyed steel)

Reichel and Schirmer, 1989 However, due to the much higher concentration of alkali chlorides compared to CuCI2, this reaction should contribute little corrosive chlorine at the higher temperatures. 8.1.4 Mechanisms in the Dust Removal System

As mentioned in Chapter 3, the most commonly used dust removal system, if dry or semi-dry scrubbing is applied, is the fabric filter (baghouse), especially in North America. In European countries where very strict emission guidelines are in place, wet scrubbing is prevalent and dedusting is generally facilitated using electrostatic precipitators. In addition, cyclones are sometimes still used as an initial dust removal device prior to modern APC equipment. In most existing MSW incinerators, filter residues consist not only of the original particulate released from the grate, but also of additional substances which have been added for gas cleaning purposes. If dry or semi-dry scrubbing is applied, or if in wet systems the flue gas scrubbing solution is evaporated by spray drying upstream of the filter system, the dust particles are combined with these APC residues. Similarly, in other MSW incinerators, charcoal is injected into the flue gas for sorption of PCDD/PCDF and mercury, and subsequent removal in the APC system.

285 8.1.5 Mechanisms in the Air Pollution Control System This system is installed to remove volatile species from the flue gas by means of chemical reactions which are described in detail in Chapter 4. In all following calculations of mass streams for solid matter or single elements, only the stoichiometric quantities are taken into account, regardless of the scrubbing technology. 8.2 MASS STREAMS IN A MUNICIPAL SOLID WASTE INCINERATOR To estimate the partitioning of elements, data from the literature were used, despite the differences in quality. These data have been generated in different countries at different times and thus reflect different waste qualities, technological standards and operation parameters. The basic requirement of the calculations is the knowledge of the different mass streams. Figure 8.10 shows mean data for these streams which are compiled from modern mass burning systems (Baccini and Brunner, 1985; Schneider, 1986; Environment Canada, 1988; Reimann, 1989; Barniske and Johnke, 1990; G6ttlicher and Anton, 1990; MQhlenweg and Brasser, 1990). Figure 8.10 Streams of Solid Masses in MSW Incinerators

286 The anticipated air consumption of this typical mass burn system is estimated to be about 5,000 m3 per Mg of waste. Old incinerators operated more in the region of 6,000 m3/tonne, whereas new installations are about 4,000 to 4,500 m3/tonne. The bottom ash production rates differ substantially, between 250 and 420 kg/tonne of waste. These data include the grate siftings, which are generally combined with the bottom ash. The fraction of siftings is a function of the type of waste fuel and the type and age of grate being used. The generation rates for grate siftings are seldom measured, consequently there is a dearth of data. Observations from the WASTE Program (1993) estimate that grate siftings contributed about 3.8% of the total mass flux from a modern mass burn incinerator. Siftings from some facilities may increase the amount of unbumt matter and metallic phases in the bottom ash and consequently are collected separately and fed back into the combustion chamber (Vehlow et al., 1990). However, the recent WASTE Program data indicates that there is very little unburnt matter in grate siftings (<2% loss on ignition). Although recent data indicate that grate siftings contain enriched levels of metals such as aluminum, copper, lead and zinc, the concentrations of most metals are similar to the concentrations in bottom ashes (Schneider, 1986). For the purposes of this section, the grate siftings mass stream has not been accounted for in the valuation of element partitioning. The exceptions will be noted later in the chapter. As mentioned in Chapter 3, the quantity of boiler ash generated depends on the type of boiler and on the amount of particulate originally released from the grate. The variation of this mass stream is estimated between 2 and 12 kg/tonne of waste. In many MSW incinerators, the boiler ash is still combined with the bottom ash in the quench tank. In the future it should be treated together with the APC residues since its chemical characteristics are more similar to APC residues than to bottom ash. In some countries, this is already enforced by legislative regulations. Although the estimated mass flux of filter ashes used here is about 25 kg/tonne of waste, (based on a particulate loading of 5 g/Rm3), the operating conditions at most modern incinerators promote lower particulate loading (<2 g/Rm 3) (Environment Canada, 1988; Vogg et al., 1990 & 1991; Reeck et al., 1991 ), with corresponding APC residue production rates of about 10 kg/tonne (Faulstich et al., 1990). The mass of APC residues shows the highest variation of all residue streams. The 12 kg/tonne listed is a mean value for wet systems which operate near stoichiometry. The value includes the amount of dry neutral sludge (2 to 4 kg/tonne) and the soluble salts (5 to 12 kg/tonne). In semi-dry or dry systems, the amount increases because of unreacted additives which are removed along with the residues. In the calculated mass balances presented in the following sections, the residue streams from wet systems will be used. This limitation will not influence the balances substantially, since all existing scrubbing systems achieve about the same emission limits and remove about the same fraction of the specific species from the flue gas.

287 For calculating stack emissions, the residual particulate loads of about 10 mg/Rm 3 are used to comply with the emission limits of many countries. Modern MSW incinerators achieve emissions below 1 mg/Rm 3. In the following balances of single elements, the German emission limits are used, which are similar to those in many other countries (Bundesministerium, 1990). 8.3 LITHOPHILIC ELEMENTS 8.3.1 Fundamentals Some elements and their most common compounds have boiling points far above 1,500~ Consequently, these materials will remain in the bottom ash. However, this does not mean that they do not undergo changes in chemical speciation during the combustion process. In some cases, it can be expected that they gain stability, either through sintering or incorporation into other stable matrices (as discussed above). The most important members of this group are silicon and aluminum, which form very stable oxides and anions, i.e. the silicates and aluminosilicate. These are the major matrix compounds of bottom ashes, since they are relatively "heat stable". Their presence in APC residues is due to particulate matter carried over from the furnace in the flue gases. Some theoretically stable volatile compounds of metals are not formed under the conditions of waste combustion due to energetic or kinetic hindrance. These metals are preferably fixed in the crystal structures of silicates and aluminosilicate, or they may be transformed into more stable oxidic compounds during the quenching or condensation. The latter is true for chromium, manganese, iron and nickel, which are distinguished by similar crystal radii and are included under the category of lithophilic elements. Some metals form stable volatile compounds in the combustion chamber but the fraction actually vaporised is very low. Although these metals can exhibit a significant enrichment in the fine particle fraction of the APC residues, much of the material remains in the bottom ash, and therefore has been categorised here with the lithophilic elements. Copper is an example of a lithophilic element with volatilised species of importance for other reactions, i.e. its chlorides play an important role since they act as catalysts in the dioxin synthesis (Vogg and Stieglitz, 1986; Stieglitz and Vogg, 1987; Hagenmaier et al., 1987). Other members of this group are the alkali metals which are responsible for the alkalinity of the bottom and fly ashes. A certain proportion of these elements are also associated with the silicates, where they act as neutralising cations. The predominant metal in this group is calcium.

288 8.3.2 Alkali Metals

Sodium, potassium and to a lesser extent rubidium are considered major alkali metals present in ash. All these elements prefer the oxidation state +1 and exhibit a pronounced electropositive character, designated by very high negative potentials against a hydrogen electrode. Due to their high reactivity these metals are present in nature, and in the waste, bound to electronegative elements. The inorganic salts are relatively thermally stable, i.e. the boiling points are far above 1,000~ In the case of chlorides, the boiling points decrease with increasing atomic number: NaCI boils at 1,730~ KCI at 1,411~ RbCI at 1,383~ and CsCI at 1,303~ The extremely high transfer temperature of NaCI will only allow a very limited fraction to be volatilised in the combustion chamber. A large proportion of sodium and potassium are bonded to silicates, and hence partitions mainly to the bottom ash. This fraction is higher for sodium than for potassium. Although the concentration of sodium does not differ much in the various residue streams, filter ashes might show a certain enrichment, thus indicating a small transport of Na via the gas phase (Angenend, 1990). The thermodynamic data of the other alkali chlorides, especially the slopes of their vapour pressure curves, indicate a much more distinct volatilisation at temperatures found in the combustion chamber. This explains the pronounced concentration of potassium and rubidium in the fine fractions of filter ashes (see Clean Gas Dust column in Table 8.2) Table 8.2 Mean Concentrations of Alkaline Elements in the Waste and in Solid Residues (mg/g) Element

Waste

Bottom Ash

Boiler Ash

Fly Ash

Clean Gas Dust

Na

3 - 15

8 - 40

8 - 40

10 - 40

20 - 50

K

5

2 - 15

15 - 40

4 - 40

100

0.16

0.45

Rb 0.04 0.125 0.10 Data taken from literature, see respective chapters

Organic alkali compounds decompose during incineration, with the primary products being Na202 for sodium and K20 or K204 for potassium. Since these compounds are not stable in a reactive environment, reactions with other species present in the fuel will take place. The final products can include the respective chlorides, sulphates, carbonates and eventually hydroxides. The latter two compounds add to the alkalinity of the residues.

289 The alkali elements sodium and potassium are ubiquitous and show similar abundance in the lithosphere with about 2.6 and 2.4% respectively. Both are essential physiological elements and act to maintain the osmotic pressure of cells. The daily intake for humans is about 3 to 7 g of sodium and about 2 to 4 g of potassium. In the waste, the concentration of Na is about 10 to 15 kg/tonne, which is much higher than that of potassium (about 5 kg/tonne) (Schneider, 1986). In Figure 8.11, the partitioning of sodium is depicted on the basis of the cited data set. The concentration is more or less equal in all residue streams and compares favourably to data in other publications (Hjelmar, 1987; Angenend, 1990). Figure 8.11 Concentrations and Percent Partitioning of Na

As has been mentioned above, the fraction of volatilised chlorides should be higher for the heavier alkali metals than for sodium. The compilation of concentrations in the residue streams given in Table 8.2 supports this hypothesis for potassium. The only available data set which provides values for all residue streams depict a substantial fraction of potassium partitioning to the fly ash stream (Figure 8.12). It can be expected that the alkali metals rubidium and cesium behave more like potassium than like sodium. The key point seems to be the speciation of these alkali

290 metals in the feed. If they are mainly bound to silicates, they will not volatilise. Volatilisation can only occur with the organic matter or chloride bound metals. Figure 8.12 Concentrations and Percent Partitioning of K

8.3.3 Earth-Alkali Metals

The group of earth-alkali metals consists of beryllium, magnesium, calcium, strontium, barium and radium. With the exception of beryllium and radium, which are very rare in nature and in the waste, the earth-alkalis are among the most abundant elements in the lithosphere. The chemical properties of magnesium are more similar to those of aluminum than those of calcium, strontium and barium. The latter metals are the typical earth-alkalis and are always found in the oxidation state +2 and bound to electronegative elements. Generally, no earth-alkali salts can be thermally mobilised at the temperatures inherent in waste incinerators. The reactivity of earth-alkali salts is somewhat higher than that of alkali salts. During incineration, many earth-alkali compounds present in the waste are subject to hydrolysis or thermal decomposition. The primary reaction products are oxides or hydroxides which can be transformed into carbonates by the high levels of CO2 in the

291 atmosphere. Hence, the earth-alkali metals can substantially contribute to the alkalinity and buffering capacity of the bottom ash stream. The lithosphere contains about 3.6% calcium, mainly bound to carbonates like limestone, CaCO3, or to sulphates like gypsum, CaS04 o2H20. These calcium compounds are often found together with salts of other earth-alkali metals. A considerable amount of calcium can be found in the putrescible waste fraction contributing alkalinity to the bottom ash stream. Calcium is physiologically important with respect to bone structure, i.e. about 1.2 kg of calcium are present in the human skeleton. Although calcium is among the most abundant elements, there is a dearth of analytical data on its behaviour during incineration. The partitioning of calcium shown in Figure 8.13 is calculated from two data sets (Schneider, 1986; Belevi, 1993). Since no calcium compounds can be volatilised, it has to be concluded that the concentrations of calcium reflect the proportion of ash generated in each stream. However, if lime scrubbing APC systems are utilised, the calcium content in the APC residue will increase dramatically. Magnesium, strontium and barium behave in a manner similar to calcium (see Table 8.3). Table 8.3 Mean Concentrations of Earth-Alkali Elements in the Waste and in Solid Residues (mg/g) Element

Waste

Bottom Ash

Boiler Ash

5 - 20

Mg

Fly Ash

Clean Gas Dust

10 - 40

Ca

10 - 35

50 - 100

100

50 - 100

50

Sr

0.12

0.20 - 1.0

0.50

0.65

0.45

Ba

0.73

0.50 - 2.5

2.5

2.5 - 3.5

1.30

Data taken from literature, see respective chapters

8.3.4 Heavy Metals Most heavy metals in the group of lithophilic elements are present in the waste in relatively high concentrations (>100 g/tonne), however, some trace elements also belong in this category. Similar to other elements, there is a paucity of valid data sets for all residue streams containing these metals. Table 8.4 contains the mean concentrations of iron and other lithophilic elements in the waste input and in residue streams, taken from the literature.

292 Figure 8.13 Concentration and Percent Partitioning of Ca

Table 8.4 Mean Concentrations of Elements in Waste and Solid Residues (IJg/g except..* = %) Element Waste Bottom Ash Boiler Ash Fly Ash Clean Gas Dust Ti

1300

3500-8000

6500

7000-9000

4000

Cr

40-400

100-1200

200-800

100-1000

200-800

Mn

200-500

400-1000

700-1200

800-1500

900

Fe*

2.5-5.0

5-15

3.0-5.0

3.0-6.0

4.0

Ni

20-130

50-800

100-300

100-300

100-200

Cu 200-1000 250-4500 300-1500 Data taken from literature, see respective chapters

50-5000

300-3000

Iron

Iron is another common element in the lithosphere, as well as in municipal solid waste. The concentration of iron in the waste feed has been well documented in the order of 25 to 50 kg/tonne (Brunner and Zobrist, 1983; Dobberstein, 198:3; Brunner and Ernst, 1986; Schneider, 1986; Perrier-Rosset, 1989; Belevi, 1993). Today, the iron in most incinerators is introduced by the metallic fraction. Modern waste management strategies, however, promote metal recycling and this will reduce the input of metallic iron in future.

293 The most stable iron compounds are the oxides (Fe203 and Fe304) which can be found by mineralogical examination of bottom ash (Vehlow et al., 1992) and fly ash (Hundesr0gge, 1990). There are two iron chlorides which can theoretically be volatilised in the combustion chamber. The FeCI2 sublimates at 1,026~ and has a high vapour pressure at relatively low temperatures, but it can only exist in a reducing environment (compare the corrosion mechanism in Figure 8.9). FeCI 3 has a very low boiling point of 319~ but is immediately converted into Fe203 as soon as H20 is present. Therefore, both compounds do not contribute substantially to the transport of iron to the fly ashes. Since fly ashes show no enrichment of iron in the fine fractions, the iron is originally bound to particles carried over from the furnace. However, it should be noted that chlorides of iron can be formed in certain parts of the incinerator, (i.e. in the boiler), by the same types of reactions given in equation (8.9) or in the corrosion mechanisms outlined in Figure 8.9. Metallic iron present in the waste fuel cannot be oxidised easily. The relatively small surface area-to-volume ratio of larger particles prevents the rapid reaction. It is often protected against corrosion by zinc, tin or cadmium layers which will volatilise first. Consequently, ferrous material passes through the incinerator relatively unscathed. Concentrations of iron in bottom ash can vary widely for a number of reasons, including: if primary material was or was not removed from the sample prior to sample preparation and analysis if the sample was collected prior to or after magnetic separation of ferrous from the ash the effectiveness of source separation schemes for the waste fuel. The quantity of magnetically separated iron from bottom ash can reach 12 to 15% of the total weight of the ash (Pietrzeniuk, 1985). Typical iron concentrations in bottom ash from mass burn systems can vary between 9 and 15% (Hjelmar, 1987; Eighmy et al., 1987; Schneider, 1986; Belevi, 1993), although concentrations of up to 24% have also been reported (Baccini and Brunner, 1985; Brunner and MSnch, 1986; Cernuschi et al., 1987). Concentrations in boiler ash are in the order of 3 to 5% (Schneider, 1986) and mean values of 3 to 6% have been recorded for fly ash (Brunner and Zobrist, 1983; Baccini and Brunner, 1985; Schneider, 1986; Environment Canada, 1988). Figure 8.14 provides the partitioning of iron during incineration based on three different sets of data (Schneider, 1986; Brunner and MSnch, 1986; Belevi, 1993), which are consistent with the information provided above for individual residue streams. Virtually all of the iron in the waste remains in the bottom ash. Only negligible amounts (1 to 2%) are transferred in the fly ash particles and, if excellent dedusting is achieved, less than 0.01% of the total input is emitted through the stack.

294 Figure 8.14 Concentration and Percent Partitioning of Fe

Chromium Chromium is an abundant element in the lithosphere and an essential additive in all stainless steel alloys. In addition, layers of chromium are used for corrosion protection of iron materials. Chromium is an element of environmental concern, since its hexavalent compounds are distinguished by high oxidation potentials as well as by high solubility in water. The waste fractions with the highest enrichment of chromium include the metal fractions, batteries, green glass, construction debris and textiles (especially leather due to the tanning process) (WASTE Program, 1993). Significant concentrations have also been measured in wood and rubber fractions (Lorber, 1983). The overall concentration of chromium in waste is in the order of 40 to 400 mg/tonne (Schneider, 1986; Environment Canada, 1988; Reimann, 1989; Angenend, 1990). Chromium compounds are not considered thermally mobile during incineration and therefore should remain mainly in the bottom ash stream. A survey of the literature indicates that the chromium concentrations in the bottom ash are slightly higher than with other streams: up to 1% (Baccini and Brunner, 1985; Eighmy et al., 1987; Sawell and Constable; 1993). The raw particulate carried over from the furnace, even clean gas flue dust, contain substantial concentrations (100 - 1,200 IJg/g) (Schneider, 1986; Environment Canada, 1988; Reimann, 1989; Vogg et al., 1991; Sawell and Constable 1993), however, there is little evidence of significant enrichment in the fine fraction (compare Table 8.4). The partitioning of chromium is similar to that of iron (compare Figure 8.14).

295

Nickel

Nickel is a very common element and is used as an alloying component in stainless steel and other nickel-based alloys. Fine nickel powder (Raney nickel) is utilised as a catalyst in the mineral oil and chemical industry, and large quantities are used in rechargeable Ni-Cd batteries. Inhalation of nickel compound dusts or aerosols represent a potential cancer risk. The highest input of nickel into the waste is in the ferrous and nonferrous metal fractions. The partitioning is similar to that of chromium and iron (Table 8.4). The overall input concentrations are in the order of 100 g/tonne of waste (Schneider, 1986; Environment Canada, 1988; Brunner and Zobrist, 1983; Reimann, 1989; WASTE Program, 1993). There is one nickel compound [Ni(CO)4] which is gaseous at ambient temperature and which might theoretically be formed under the conditions in the combustion chamber. However, at the present time there is no data to support this hypothesis, since it is easily hydrolised. The experimental results categorise nickel as a lithophilic element. The concentrations in all solid residue streams are about 50 to 300 IJg/g (Baccini and Brunner, 1985; Schneider, 1986; Environment Canada, 1988; Reimann, 1989; Angenend, 1990).

Copper

Copper is considered a half-noble metal and has a strong resistance to oxidation. As a result, copper and copper alloys (with tin and zinc) are used for kitchen utensils and other tools. About 40% of the copper produced is used in apparatus and instrument construction today. Its ability to conduct electricity readily makes it an important material in the electronics industry. Furthermore, copper compounds are applied as pigments, catalysts, stabilisers and for acetate rayon production. Copper and its compounds are potentially toxic to special plants and microbes so they are used in herbicides and wood preservatives (e.g. copper arsenate). The copper concentration in waste is reported to be about 200 to 1,000 mg/tonne of waste (Brunner and Zobrist, 1983; Schneider, 1986; Brunner and MOnch, 1986; Reimann, 1989; Tobler, 1988; Angenend, 1990; Belevi, 1993; Dalager, 1993) and is generally enriched in the organic, electrical and paper board fractions (Lorber, 1983; Waste Program, 1993). The behaviour of copper during incineration is similar to that of iron, chromium and nickel. The only thermally mobile compound is CuCI 2, which has a boiling point of 993~ This decomposes into CuCI which has a boiling point of 1,490~ Consequently, CuCI2 does not contribute substantially to the transfer of copper into the gas phase. The mean concentration of copper in the individual residue streams is highly variable, especially in the bottom ash, but the mean concentrations are in the order of 1,000 IJg/g

296 (Table 8.4) (Baccini and Brunner, 1985; Schneider, 1986; Reimann, 1989; Brunner and M5nch, 1986; Tobler, 1988; Angenend, 1990; Sawell and Constable, 1993; Dalager, 1993). Investigations on the distribution of copper in fly ashes indicate a significant increase in concentration in the fine particle fractions, which in turn indicates a certain degree of volatilisation of either pure metal, copper chloride compounds or copper-based compounds in wood preservative (Vehlow, 1993). CuCI2 is thought to play an important role in the de-novo synthesis of PCDD/PCDF in the boiler (Vogg and Stieglitz, 1987; Stieglitz and Vogg, 1987; Hagenmaier et al., 1987). 8.4 VOLATILE ELEMENTS Many nonmetallic elements present in the waste are converted by the combustion process into compounds which form acids if contacted with water. Those compounds may be oxides, as for sulphur (SO2 and to a lesser extent SO3), or nitrogen (mainly NO). Other elements such as halogens have a strong tendency to react with hydrogen and are preferably converted into hydrides by the combustion process. Theoretically, hydrogen compounds may also exist in association with sulphur, selenium, tellurium, arsenic or antimony. Although these compounds should be of interest because of their toxicity, the limited data in literature indicate that the actual quantity of those compounds is probably very low.

8.4.1 Halogens The predominant and most stable oxidation state of the halogens is -1. Due to their reactivity, the respective hydrohalogenic acids are not found in nature. Instead, the natural form is as an alkali or earth-alkali halogenide. Most halogens are also essential elements for all plants and animals. During MSW incineration, the prevailing products are the acid hydrogen halides, which are treated in the APC systems. The stability of hydrogen halides against thermic dissociation according to: where X = halogen

2 HX = H2 + X2

(8.19)

decreases with increasing atomic number. Hydrogen fluoride cannot be decomposed. The degree of thermal dissociation of the other hydrogen halides is given in Table 8.5. The data demonstrate clearly that virtually all HCI is present as a molecule in the flue gas, whereas HBr and HI are to a substantial degree dissociated at elevated temperatures and may be involved in other reactions, especially bromination or iodation of organic compounds.

297 Table 8.5 Thermal Dissociation of Hydrogen Halides (%) Temperature ~

.

HCI

......

HBr

.

HI

300

0.3 x 10.6

0.003

19

1,000

0.014

0.5

33

The halogens enter the incinerator in various chemical forms. In inorganic materials they are present as hydrohalogenic salts. The alkali and earth-alkali salts, especially those of fluorine and chlorine, are very stable and possess high boiling points >1,400~ This limits their volatilisation from the grate, however, if H20 is present, hydrolysis may take place at much lower temperatures forming hydrogen halides. In most organic materials, such as plastics or textiles, the halogens are present in the form of covalent bonds. This permits formation of hydrohalogenic acid gases in the combustion chamber. Chlorine Chlorine is classified as a major element. Its concentration in the lithosphere is similar to that of carbon or chromium. Chloride ions are ubiquitous and essential to life on earth. Sea water contains about 2% chloride ions. The human body contains about 0.12 % chlorine, most of it in the form of chloride. Organic bound chlorine is converted at comparably low temperatures into HCl. Even from plastics, such as PVC, HCI is evolved at temperatures above 230~ Some inorganic chlorides tend to undergo hydrolytic decomposition at temperatures above 500~ In residues, organic compounds containing chlorine are present, namely, polychlorinated dibenzo-p-dioxins, dibenzofurans, benzenes, phenols and biphenyls. More detailed information on these compounds is included later in this chapter. Based on six data sets, the partitioning of CI in MSW incinerator residues is illustrated in Figure 8.15 (Brunner and M(Jnch, 1986; Schneider, 1986; Reimann, 1989; Angenend, 1990; Belevi, 1993; Dalager, 1993). Parts of the data given there, for example the concentrations in the gas phase, are consistent with other publications, not being referenced with respect to their almost unlimited number. On the basis of these data sets the partitioning of CI in a municipal solid waste incinerator is compiled in Figure 8.15. The chlorine concentration in the input is estimated to be 6.9 to 7.7 kg/tonne. Plastics, especially PVC, may contribute up to 50% to this value (Toetsch and Gaensslen, 1990).

298 Figure 8.15 Concentrations and Percent Partitioning of CI

Chloride concentrations in bottom ash range from about 2,000 to 3,500 pg/g. Approximately 8 to 13% of the chloride partitions to the bottom ash stream and the remaining fraction is entrained in the flue gas leaving the combustion chamber (Figure 8.15). Depending on the length of time the ash is exposed to the flue gas stream, chloride concentrations in boiler ash can reach upwards of about 20,000 IJg/g [median concentration for mass burn facility (Environment Canada, 1988)], representing only 0.3% of the total chlorine mass flux. Raw particulate from the flue gas stream has been found to contain up to 11% chloride (based on particulate loading of 1 to 2 g/m 3) (Horch et al., 1990; Vogg et al., 1991). Typical HCI concentrations in the raw flue gas stream are in the order of 1,000 mg/Rm 3. New regulations in some countries set HCI emission limits down to 10 mg/Rm 3. This stringent target can be achieved with modern APC systems, resulting in <1% of the chlorine input being emitted through th~ stack to the atmosphere. The balance of the HCI is trapped in the APC system. In wet scrubbing systems, about 50% of the chloride which was fed to the incinerator is discharged with the wastewater effluent (Dalager, 1993). Fluorine Due to its extremely high electronegativity, fluorine is a very reactive element. The only stable oxidation state is -1. Hydrofluoric acid is very aggressive and can dissolve SiO2

299 and other stable silicates. Fluorides are used to a limited extent in some industrial processes, however, the importance of the element has increased since 1955, when the production of fluorine containing organic compounds and plastics began. Large quantities of chlorofluorocarbons (CFCs) were used (and subsequently banned) as propellants for sprays, as refrigerants and as thermal insulating gases in polyurethane and polystyrene foams. Now, new plastics containing fluorine are being produced, such as PTFE, and higher concentrations should be expected in the waste. Organic fluorine compounds start to disintegrate at temperatures above 250~ forming HF. Only very high concentrations of fluoride prove harmful to humans. The only two data sets available on fluorine concentration in MSW were used to generate the mass balance given in Figure 8.16 (Brunner and MOnch, 1986; Reimann, 1989). Both studies found similar input concentrations of 140+60 g/tonne and 103 g/tonne. The bulk of fluorine is typically bound to alkali metals. Only a minor fraction is currently contributed by organic compounds, such as PTFE. Figure 8.16 Concentrations and Percent Partitioning of F

Bottom ash concentrations were measured at about 200 pg/g. Although Reimann (1989) describes a washing process for bottom ashes using a quench tank, only about 1% of the fluorine inventory was separated. This indicates that fluorine in bottom ash is tightly bound to earth-alkali elements which have very low solubilities in water.

300 The fluorine content in gas path residue streams varied substantially. Reimann (1989) measured concentrations of 45 -62 IJg/g in boiler ash and about 100 IJg/g in fly ash. Brunner and M5nch (1986) measured concentrations up to 2,000 IJg/g in fly ash. There is no discemable reason based on facility type, operation or sampling methods which could account for the discrepancy. It is speculated that one major source of error could have been the digestion and analysis of special fluorides, especially those of calcium or iron, however, this cannot be confirmed. The HF concentrations measured in the gas phase are consistent with each other, ranging from 3 to 5 mg/Rm3. This is also consistent with other findings in the range of 1 to 10 mg/Rm 3 (Jochem, 1989; Ruytenbeek and Braams, 1989; Vehlow et al., 1992). Regulations in several countries limit HF emissions to 1 mg/Rm 3 which can be met by modern air pollution control equipment (Reeck et al., 1991; Environment Canada, 1986), even if the HF concentration in the raw gas is substantially increased (Vehlow et al., 1992). Typically less than 1% of the total fluorine input is emitted to the atmosphere.

Bromine and Iodine Bromine and iodine are much less abundant than chlorine, although bromine typically is found associated with chlorine. The general properties of both elements are similar to those of chlorine. In sea water, the concentration of bromine is 200 times lower than the concentration of chlorine. The concentration of iodine is 2000 times lower. Bromine compounds are often used in plastics as flame retardants. Although chlorine and bromine are mainly present as halogenides, iodine is more frequently found in covalent bonds in organic compounds. Little is known about the behaviour of both elements during incineration, however, it is speculated that the main combustion products are hydrogen halides. Schneider (1986) reports a bromine concentration of 6.5 g/tonne in MSW. The concentration of bromine in bottom ash is, as expected, very low (< 10 IJg/g). The major portion is captured in the APC residues, where concentrations of about 250 pglg were measured. On a comparative basis, the partitioning of bromine will be similar to chlorine. Unfortunately, no data were available for iodine in MSW or ashes. Since HBr and HI will be present at very low concentrations in the flue gas and the partitioning is similar to HCI, no problems are foreseen with respect to their emission to the atmosphere. No emission limits exist for these compounds.

Sulphur About 80% of global sulphur production is used to produce sulphuric acid in the chemical and fertiliser industry. Elementary sulphur is also used in the tyre industry for vulcanisation of rubber. Sulphur can form a series of different organic and inorganic compounds. The most stable inorganic compounds are sulphates. The gaseous oxide

301 SO2 is by far the most prevalent combustion product of elementary sulphur, sulphides or organic sulphur compounds present in MSW. The formation of SO2 in the combustion chamber depends on the temperature in the fuel bed. Elementary, organic, and sulphidic bound sulphur is converted into SO2 at temperatures of about 300~ whereas sulphates of trivalent metals like iron and aluminum start to disintegrate at temperatures of more than 550~ (see 8.2.3). Sulphates of alkali or earth-alkali metals are stable at temperatures up to about 1,000~ Hence, the formation of SO2 from sulphates increases with increasing temperature, and the concentration of sulphates in bottom ash decreases. A rough estimate of sulphur partitioning during incineration compiled from several sources is depicted in Figure 8.17 (Baccini and Brunner, 1985; Hjelmar, 1987; Eighmy et al., 1987; Brunner and M(~nch, 1986; Lorber, 1983; Roffman, 1991; Environment Canada, 1988; Jochem, 1989; Ruytenbeek and Braams, 1989; Belevi, 1993; Dalager, 1993). About 30% of the 1 to 3 kg/tonne of sulphur in MSW originates from the organic fraction and the remaining 70% is found in paper, paperboard, plastics and fines. Figure 8.17 Concentrations and Percent Partitioning of S

(Concentrations in the gas phase given for SO2, the solid residue for SO42) * = raw gas concentration

302 Sulphur will be present in all ash streams in the oxidation state +6, which means sulphates of mainly alkali and earth-alkali metals. Sulphate concentrations in bottom ash have been recorded ranging from 1,000 to 20,000 pg/g. Based on these results, about 35% of the sulphur input remains in the bottom ash. Sulphur concentrations in boiler ash do not differ significantly from those in the fly ashes, although the length of time the ash is exposed to the flue gas will significantly influence the concentration (see Chapter 10). SO42-concentrations ranging from 12,000 to 120,000 IJg/g have been recorded in fly ashes. These data indicate that about 2% of the sulphur input partitions to the boiler ashes and about 25% partitions to the fly ash. Less than 40% of the sulphur input is released as sulphur dioxide into the flue gas. The corresponding SO2 concentration in the raw gas is about 200 to 500 mg/Rm 3. Most modern APC systems guarantee SO2 emissions of <20 mg/Rm 3 (Reeck et al., 1991; Environment Canada, 1986), thus emitting less than 5% of the total sulphur input.

Nitrogen

The oxidation product of nitrogen is NO, which is relatively chemically stable. During pyrolysis, amines, ammonia and other compounds can also be released, which are subject to further reactions in the hot zones of the combustion chamber. In dry MSW, the nitrogen content is about 1%. Complete oxidation of all the nitrogen in the waste fuel to NO would result in concentrations of about 1,500 mg/Rm 3 in the flue gas. However, the normal NO load is only in the order of 200 to 300 mg/Rm 3. It is speculated that the NO formed is produced only by the nitrogen compounds present in the waste. This is supported by data generated under different oxygen concentrations in the flue gas and at different combustion temperatures (Horch, 1987). NO emissions are limited in several countries to 50 to 200 mg/Rm3. These limits cannot be achieved without additional post-combustion abatement (see Chapter 4). Since NO is virtually insoluble in water, wet gas cleaning systems have very limited effects on the NOx levels in the flue gas. The addition of the Fe2§complex with the organic compound EDTA enhances the solubility, and if sulphites are present, NO can be reduced to N2 (Schuster, 1984). However, this process is no longer used since it is detrimental to achieving adequate control of mercury emissions. Studies indicate that amines and ammonia are able to react with NO to produce elementary nitrogen (de Soete, 1975). This process can be adapted for selective non-catalytic NOx reduction (SNCR) in MSW incinerators by means of ammonia injection (Hurst, 1983). There are some technical challenges in all SNCR processes: 9

to inject the agent at the proper temperature

9

to achieve a uniform distribution across the gas stream to guarantee a sufficient residence time in the appropriate temperature regime.

303 On the other hand, ammonia injection has the advantage of being performed at the front end of the flue gas system, and does not require special reactors or reheating of the gas. Although ammonia slip is a real problem, it can be minimised by the application of special organic nitrogen compounds (Dransfeld et al., 1992). Most applied or proposed abatement processes are selective catalytic reduction processes (SCR) using metal oxide catalysts often based on TiO2. Much of the drive for improvement of this technology has been in areas with chronic urban smog problems, especially Japan (Ando, 1985). Oxidative processes producing nitrates have been tested but are of little use since nitrates are water soluble and therefore are substances which are subject to landfill disposal regulations. Moreover, their contamination with heavy metals and other compounds precludes their use as fertilisers. The occurrence of oxidised nitrogen compounds in bottom and fly ashes is unlikely and has not been reported. The ammonia slip in SNCR and SCR processes may cause problems in APC residues, and more research needs to be conducted on the extent of the influence de-NOx systems have on ash quality.

8.4.2 Volatile Metals The main property of this group of metals is their volatility during incineration. The metals included in this group are not only those compounds of which vaporise readily, such as mercury, but also those semi-volatiles which cause a significant enrichment in fly ash even though a substantial portion may remain in the bottom ash stream. Reported concentrations of volatile metals in the waste stream and the residue streams are compiled in Table 8.6. Table 8.6 Mean Concentrations of Volatile Metals in Waste and Ash Element Waste Bottom Ash Boiler Ash Fly Ash g/Mg IJg/g Pg/g Pg/g Sb

10 - 60

10 - 80

20 - 60

Emissions Pg/Rm3

40 - 120

<5

As

3-9

1 - 80

20 - 60

40 - 120

<50

Cd

5 - 15

<0.5 - 10

50 - 150

50 - 1,000

< 10

Pb

400- 1,000

350- 5,000

2,000- 8,000

2,500- 12,000

<50

Hg

0.5- 5

<1

<5

1 - 30

<50

Se

0.8

0.4- 1

4

10- 20

<2

Sn

120

250

500

1700

-

TI

0.2

<0.5

<0.5

<0.5

<1

Zn

600 - 2000

800 - 6,000

5,000 - 10,000

5,000 - 80,000

<5

304

Mercury

Mercury is a rare noble metal, and most of its compounds are distinguished by their relatively high vapour pressures. Mercury is used in thermometers, barometers, diffusion pumps, batteries, fluorescent light tubes, electric switches and relays, and as cathodes in chlorine-alkali electrolysis. Other applications include dental alloys, dyes and herbicides. Mercury is considered to be potentially highly toxic and the use of Hg is quickly becoming restricted in most countries. Mercury is by far the most thermally mobile trace metal and mercury compounds can remain in the gas phase at temperatures <180~ This poses a unique problem with respect to MSW incineration, even though the concentration in the waste is very low (only 0.5 to 5 g/tonne) (Bergstr6m and Sundquist, 1984; Environment Canada, 1985 & 1988; Schneider, 1986; Brunner and Ernst, 1986; Reimann, 1989; WASTE Program, 1993). The mechanisms of release and removal for mercury have been studied in great detail (e.g. Lindquist et al., 1986; Nagase et al., 1986; Braun et al., 1986 & 1988; Metzger and Braun, 1987; Teller and Quimby, 1991; Guest and Knizek, 1992; Reimann, 1989; Richman et al., 1993). The mercury input depicted in Figure 8.18 is compiled from four of the data sets (Reimann, 1989; Brunner and MOnch, 1986; Angenend, 1990; Dalager, 1993). The mass balance indicates that almost no mercury remains in the bottom ash, since the reported mercury concentrations are normally <1 pg/g (Baccini and Brunner, 1985; Hjelmar, 1987; Eighmy et al., 1987; Schneider, 1987; Environment Canada, 1988; Ruytenbeek and Braams, 1989; Reimann, 1989; Angenend, 1990; Roffman, 1991; Sawell and Constable, 1993). In the combustion chamber, virtually all the mercury and its compounds are transformed into HgCI2, which has a boiling point of only 304~ Boiler ashes contain less than 5 pg/g of Hg and the concentrations in fly ash is also very low (1 - 30 pg/g). Therefore, the partitioning of mercury in the incinerator is about 2% in bottom ash and about 13% in fly ash. Concentrations of mercury in the raw gas have been measured at several hundred pg/Rm 3 (Environment Canada, 1986 and 1988; Braun et al., 1986; Reimann, 1989; Bma, 1991). It is estimated that about 70 to 80% of the mercury inventory in MSW is present in vapour form at the entrance to the APC system, and as outlined in Chapter 4, there are several methods of mercury removal. Since mercury has a very strong preference to adsorb to particulate carbon, most modern incinerators (with characteristically low carbon particulate content in the flue gas stream) must rely on special mercury removal methods. In dry or semi-dry systems the temperatures are kept below 150~ and the injection of activated carbon or sodium sulphide is recommended to guarantee sufficient removal. In the case of wet scrubbing systems, the acidity of the first scrubber has to be kept well below a pH of 1 to ensure adequate absorption of mercury as complex (HgCI4)2ions. Reducing compounds in the scrubbing solution may cause a partial reduction of the Hg 2. to Hg 1., which is absorbed in the acid environment, but also undergoes

305 spontaneous dissociation into Hg 2§ and Hg~ (Braun et al., 1986 & 1988). Metallic mercury can only be removed from the flue gas by means of activated carbon. Figure 8.18 Concentrations and Percent Partitioning of Hg

Additional problems arise in wet scrubbing systems when effluents have to be cleaned in wastewater treatment plants. The mercury concentrations in the order of 5 to 10 mg/I must be reduced to about 0.05 to 0.02 mg/L. Although this process is well established, it needs careful monitoring. Irrespective of the APC system, mercury emissions from modern MSW incinerators can be kept below 0.05 iJg/Rm3, which means that <10% of the mercury input is emitted to the atmosphere (Reeck et al., 1991; Brna, 1991; Braun and Gerig, 1990). Cadmium Environmental concerns over cadmium arise over its thermal mobility, potential solubility and toxicity in aquatic systems. It is classified as a rare element, and its chemical properties are very similar to those of zinc. In fact, cadmium is always present as a trace contaminant in zinc applications. Because of its corrosion resistance, cadmium is used in the metallic state as a corrosion protection layer on screws and other ferrous materials. A substantial amount of cadmium is also used in the manufacturing of rechargeable Ni-Cd batteries, as pigments in dyes and plastics, and

306 as chloride scavengers in certain plastics. In some countries, applications of cadmium compounds are restricted, suggesting that concentrations in MSW will decrease in the near future. According to thermodynamic calculations, cadmium and cadmium oxide are preferentially converted into CdCI2 during combustion (Borchers, 1989). Furthermore, Figure 8.6 indicates that not only the metal itself, but cadmium chloride and to a limited extent cadmium oxide can be vaporised during incineration. Consequently, the bulk of cadmium entering the incinerator will be entrained in the flue gas stream. Since the boiling point of CdCI 2 is about 975~ and the flue gas contains high levels of HCI, virtually all of the cadmium found in the flue gas downstream of the boiler will be found in the particulate matter as CdCI2. The partitioning of cadmium is given in Figure 8.19. It is based mainly on five different sets of data (Brunner and MSnch, 1986; Schneider, 1986; Reimann, 1989; Dalager, 1993; Belevi, 1993). Figure 8.19 Concentrations and Percent Partitioning of Cd

The concentration of cadmium in MSW ranges from 5 to 15 g/tonne in earlier studies (Schneider, 1986; Brunner and Zobrist, 1983; Dobberstein, 1983; Brunner and Ernst, 1986; BergstrSm and Lindquist, 1984; Reimann, 1989; Angenend, 1990; Belevi, 1993; Dalager, 1993), whereas two recent studies indicate that the quantities may have increased slightly (up to 13.6 g/tonne). Data indicate that although Ni-Cd batteries are a major source of the cadmium in the waste stream, plastic and organic fractions also contribute substantial proportions (20 to 50%) (Lorber, 1983; WASTE Program, 1993).

307 Cadmium concentrations in the bottom ash are normally well below 10 pg/g (Baccini and Brunner, 1985; Brunner and MOnch, 1986; Ruytenbeek and Braams, 1989; Reimann, 1990; Sawell and Constable, 1993; Belevi, 1993; WASTE Program, 1993), which represents less than 15% of the total input. Concentrations in any of the fly ash streams can range widely from 50 to almost 1,000 pg/g (Baccini and Brunner, 1985; Brunner and MOnch, 1986; Schneider, 1986; Hjelmar, 1987; Eighmy et al., 1987; Reimann, 1989; Horch et al., 1990; Angenend, 1990; Sawell and Constable, 1993; WASTE Program, 1993), depending on the length of time the ash remains exposed to the flue gas. The bulk of the cadmium fed into the incinerator partitions to the fly ash streams (about 80%). The quality of the dust removal system dictates the emissions of cadmium to the atmosphere. As mentioned above, volatile species are enriched in the finer size fractions of the APC residue. For example, concentrations of cadmium in clean gas dust range from 100 to 4,000 IJg/g (Brunner and Zobrist, 1983; Schneider, 1986; Brna, 1991; Vogg et al., 1991). Although these fine particles are difficult to capture, modern APC equipment can easily meet the most stringent emission limits for cadmium (e.g. values <10 iJg/Rm3 have been reported by Reeck et al., 1991; Brna, 1991; Vogg et al., 1991 ). Based on these emission limits, less than 1% of the cadmium inventory in MSW is emitted into the atmosphere. Zinc

In nature, zinc is often found in ores together with lead and cadmium. The greatest quantity of zinc produced is used to galvanise steel. Zinc is also an essential alloying element in brass and other alloys. Zinc oxide and other zinc compounds are used widely, for example in paper production, filler materials and pigments and catalysts in plastics. Zinc is generally not considered a toxicological concern. Zinc is not as volatile as mercury or cadmium. Figure 8.20 illustrates the partitioning of zinc during incineration. Although zinc is fairly ubiquitous in waste components, nonferrous, organic, plastic, fines, wood and paper materials contribute substantial proportions to the remarkably high concentration of zinc in MSW (Lorber, 1983; Beek et al., 1989; WASTE Program, 1993). Concentrations in the overall waste stream range between 600 and 2,000 g/tonne (Brunner and Zobrist, 1983; BergstrOm and Lindquist, 1984; Schneider, 1986; Brunner and M0nch, 1986; Horch, 1987; Environment Canada, 1988; Reimann, 1989; Angenend, 1990; Beveli, 1993; Dalager, 1993; WASTE Program, 1993). In the combustion chamber, zinc can either be transformed into an oxide or a chloride (Borchers, 1989). Due to the oxidative nature of the burn, the first reaction will be an almost complete oxidation. The second step, chlorination, is a relatively slow process. Since a substantial amount of the zinc is already present in the waste in the oxide form, the conversion into ZnCI 2 will be limited. Nevertheless, zinc shows the typical properties of a thermally mobile metal with enrichment on finer fly ash particles.

308 Figure 8.20 Concentration and Percent Partitioning of Zn

Although the ZnCI2 and eventually traces of metallic zinc are volatilised in the combustion chamber (see Figure 8.6), the bulk of the zinc remains in the bottom ash. Concentrations range from about 800 to 6,000 IJg/g (Brunner and MOnch, 1986; Schneider 1986; Hjelmar, 1987; Eighmy et al., 1987; Reimann, 1989; Angenend, 1990; Sawell and Constable, 1993; Beveli, 1993; WASTE Program, 1993). Estimates on the basis of these values result in approximately 65% of the zinc inventory staying in the bottom ash. The volatilised zinc species begin to condense onto the surfaces of fly ash particles in the boiler. Zinc concentrations in fly ash streams range between 5,000 and almost 60,000 IJg/g, depending on the length of time the ash remains exposed to the flue gas (Brunner and MOnch, 1986; Schneider, 1986; Reimann, 1989; Angenend, 1990; Sawell and Constable, 1993; WASTE Program, 1993). As is expected for a thermally mobile metal, zinc enrichment increases with decreasing particle size. It is estimated that about 30% of the zinc is partitioned into the fly ash streams. The emissions of zinc can be kept very low (down to less than 5 tJg/Rm3) if an efficiently operating filter is installed (Vogg et al., 1991). Lead

Lead is a relatively rare element, but it is enriched in nature in some large deposits, and because it can easily be reduced to the metallic form, it has been in use for centuries. Today, the main use of lead is in lead acid batteries (approximately 45% by weight), however, recent studies indicate that most of these are recycled. Lead is also used as a pigment and a stabilising agent in plastics. Dust and soil in MSW are typically

309 contaminated with lead from its use as an anti-knocking agent in gasoline [Pb(C2Hs)j, although the quantities of leaded gas being sold globally are decreasing annually. Furthermore, the use of lead in paints has been phased out. All mobile species of lead are potentially toxic. As indicated in Figure 8.3, the most preferential reaction for lead during the incineration process is to produce PbCI2 (Borchers, 1989), then vaporise. Although the metallic form itself is potentially volatile, it is speculated that the metal melts and drips through the grates before reaching the vaporisation temperature (WASTE Program, 1993)o Lead is an amphoteric metal, which is soluble in acidic or highly alkaline media. These phenomena must be considered when developing disposal strategies for MSW incinerator ash. Partitioning of lead during incineration is given in Figure 8.21. Older studies have reported data which vary between 400 and 1,000 g/tonne in MSW (Dobberstein, 1983; Brunner and M6nch, 1986; Schneider, 1986; Reimann, 1989), however, more recent studies indicate that lead in MSW is decreasing (see Chapter 2 and Angenend, 1990; Beveli, 1993; Dalager, 1993). In spite of the thermal mobility of the lead and lead chloride, relatively high concentrations of lead ranging up to 5,000 IJg/g were measured in bottom ash (Schneider, 1986; Hjelmar, 1987; Eighmy et al., 1987; Brunner and M6nch, 1986; Reimann, 1989), although more recent studies have indicated that these levels are decreasing (see Chapter 9). Nevertheless, bottom ash combined with grate siftings still contains the bulk of the lead (about 70%) which enters the facility and indicates that the chlorination reaction is kinetically hampered. Figure 8.21 Concentrations and Percent Partitioning of Pb

310 About 30% of the lead input is vaporised from the grate, but all lead compounds formed in the combustion process condense onto dust particles in the temperature ranges downstream of the boiler. Boiler ashes can contain lead concentrations between 2,000 and 8,000 pg/g depending on the residence time in the boiler, or about 1% of the lead inventory (see Chapter 10). Lead concentrations between 2,500 and 12,000 IJg/g have been recorded in fly ashes (Baccini and Brunner, 1985; Horch, 1990; Reimann, 1989; Sawell and Constable, 1993; WASTE Program, 1993), representing about 30% of the total input. Clean gas dust can be highly enriched in lead (up to 10%) but modem APC systems are capable of limiting emissions to <50 iJg/Rm3 (Brna, 1991 ). Arsenic Arsenic is classified as a semi-metal and is most stable in the +3 oxidation state. Its industrial use has been reduced to some special applications, such as an alloying additive and as a component of special semiconductors. In many countries the application of arsenic compounds in herbicides or drugs is prohibited, however, copper arsenate is still used as an antifungal treatment for wood. Due to the poisonous nature of all of its compounds, arsenic is an element of concern in MSW incinerator ash. Many compounds (e.g. chlorides and oxides) are distinguished by a high thermal mobility, but should not be affected in the fuel bed if present in the anionic state. The behaviour of arsenic during incineration is not as well documented as other metals. Three data sets have been used as the basis for the partitioning diagram in Figure 8.22 (Schneider, 1986; Angenend, 1990; Sawell and Constable, 1993). Since arsenic always creates analytical problems in the low concentration range, the given values should be regarded as first estimates. The waste concentration is reported to be between 3 and 9 g/tonne. This is surprisingly low given that the background concentration in natural soils is 2 g/tonne (Taggart, 1948). Concentrations in the bottom ash vary between 8 IJg/g and 32 IJg/g which is consistent with the variability reported for some Danish and U.S. ashes (1.3 to 80 IJg/g) (Hjelmar, 1987; Roffman, 1991). It is estimated that about 60% of the arsenic input remains in the bottom ash. Concentrations of arsenic in boiler ash and fly ash (including APC residues) are slightly enriched compared to bottom ash, varying from 20 to 60 IJg/g and from 40 to 120 IJg/g respectively. The overwhelming majority of the arsenic vaporised in the furnace condenses onto the finer fly ash particles which are trapped out by the APC control system. These data indicate that there are no gaseous arsenic compounds to be found in the temperature range below 200~ Based on a study of a modern incinerator, the cumulative emission of arsenic, cobalt, nickel, selenium and tellurium was about 60 pg/Rm 3 with total dust emissions of 12.8 mg/Rm 3 (Reeck et al., 1991), indicating that air emissions of arsenic do not pose a health risk.

311 Figure 8.22 Concentrations and Percent Partitioning of As

Antimony

Arsenic and antimony are closely related elements with respect to their chemical properties. Antimony is considered a semi-metal, however, it is also stable in its metallic form. In nature antimony is often found together with arsenic in copper, lead or silver ores. Antimony is used in special alloys and some of its compounds are used for pharmaceutical purposes. The toxicity of antimony is not as clearly defined as arsenic, but its compounds appear to be less harmful. The formation enthalpies of antimony compounds are shown in Figure 8.3, and indicate that antimony is converted in the combustion chamber to Sb203 and SbCI 3 (Borchers, 1989). Although SbCI3 has a relatively low boiling point and is the only compound which could potentially be vaporised, it is also sensitive to hydrolysis and can ultimately be oxidised (Sb203). In nature, antimony is much less abundant than arsenic, by a factor of 10 to 20 (Taggart, 1948). In waste, however, the opposite is true. Concentrations of antimony in MSW are reported to be between 10 and 60 g/tonne (Schneider, 1986; Environment Canada, 1988). The partitioning of antimony is depicted in Figure 8.23, and indicates an almost equal distribution of antimony between the bottom ash and fly ash streams. Hence, kinetic factors may promote chlorination of antimony and hamper its oxidation and/or hydrolysis.

312 Figure 8.23 Concentrations and Percent Partitioning of Sb

Other Volatile Elements There are other elements which are known to show significant enrichment in the fine particle fraction and hence are classified as volatile elements. Among these elements, tin, thallium, selenium and tellurium are present in quantifiable concentrations in MSW incinerator ash. Unfortunately, there is a dearth of data on their behaviour during incineration, which needs to be addressed in future research initiatives. 8.5 CARBON AND SELECTED CARBON COMPOUNDS 8.5.1 Total Carbon

Carbon is one of the key elements in MSW incineration. It not only contributes to the heating value of the waste, but it is also highly reactive and responsible for the generation of many by-products of combustion. An ultimate analysis of waste (see Figure 8.2) indicates that approximately 250 kg/tonne of waste is carbon (CCME, 1989). At the beginning of this chapter, a description of the oxidation reaction and resulting products of carbon (CO2 and to a limited extent CO) was given. This section builds on those reactions and addresses the issue of products of incomplete combustion. The concentrations of particulate carbon in bottom ash are in the order of 0.5 to 5%, depending on the combustion efficiency of the incinerator (Faulstich, 1991; Horch and Schneider, 1991; Vehlow et al., 1992). Fly ash concentrations also range from 0.5 to

313 5% (Vogg, 1984; Vogg et al., 1987; Faulstich, 1991; Horch and Schneider, 1991). These data indicate that at least 95% of the carbon compounds entering an incinerator should be thermally destroyed. Moreover, in modern incinerators, improved combustion conditions should result in even greater destruction efficiency. In modern incinerators with dust emissions of <10 mg/Rm3, CO and organic compounds are the major forms of emitted carbon. If an emission standard of 10 mg/Rm 3 for the organic carbon and 50 mg/Rm3 for CO is taken as a basis, about 300 g/tonne or 0.12% of the total input of carbon would be emitted via the stack in these forms. However, the application of the latest combustion technology guarantees a combined total emission of CO and organic carbon of <10 mg/Rm3, representing less than 0.02% of the total carbon input (Reeck et al., 1991). On the basis of these data, a carbon balance for products of incomplete combustion (i.e. not counting CO2) has been calculated and is depicted in Figure 8.24. The information provided in the Figures given in this section for organic compounds is based on a recently published overview paper (Vehlow and Vogg, 1991). Figure 8.24 Concentrations and Balance of C

314

8.5.2 Polychlorinated Dibenzo-p-Dioxins and Furans Unfortunately, not all carbon compounds present in incinerator ash or stack emissions are in the form of CO, particulate carbon or short chain hydrocarbons like methane. In 1977, the presence of polychlorinated dibenzo-p-dioxins and dibenzofurans was detected in fly ash from an MSW incinerator (Olie et al., 1977). This finding contributed considerably to the passionate discussions about the acceptability of MSW incinerators, a discussion which is still steeped in controversy. This group of compounds is symbolised by the catchword "dioxins" and consists of 75 isomers of polychlorinated dibenzo-p-dioxins and 135 isomers of polychlorinated dibenzofurans. All of the isomers which contain chlorine atoms in positions 2, 3, 7 and 8 are considered toxic, with the tetra isomer ranking the most toxic. Although PCDD/PCDF compounds are susceptible to photodegradation, they are persistent under normal environmental conditions and tend to bio-accumulate in fatty tissues of animals (Poiger and Schlatter, 1986). To compare PCDD/PCDF data on the basis of their toxicity, toxic equivalency factor (TEF) units are calculated based on weighting of the potential toxicity of each specific isomer in relation to 2,3,7,8 - tetrachloro-dibenzo-p-dioxin. Although the basic weighting is the same, there are several differences in the schemes between the several methods, including the Eadon, U.S. EPA, California, Canadian, German and Nordic methods. To avoid any confusion, an international scheme was presented by an expert group (NATO 1988) and tends to replace all other proposals. Table 8.7 summarises the international toxic equivalent factors (I/TEF) together with some other TEF schemes. The "tolerable daily intake" for humans is 1 to 10 pg (TE)/kilogram of body weight/day (WHO, 1987). In all residues from waste incineration the isomer distribution of PCDD/PCDF is more or less the same. The toxic equivalents are dominated by PCDF, especially the 2,3,4,7,8 -pentachloro-dibenzofuran, due its high equivalency factor. About 70% of the TE concentration is normally contributed by PCDFs, whereas only 30% is actually derived from PCDDs. Based on these considerations, the information provided here will focus on the major trends. In many cases, especially for data from old MSW incinerators where isomer specific concentrations of PCDD and PCDF are not available, the TE is estimated using the following approximation based on Vogg et al., (1989): E = 0 . 0 1 x ~ P C D D + 0.02 x ~ P C D F f o r ~ P C D D / P C D F E = 0.005 x ~ P C D D + 0.02 x ~ P C D F f o r ~ P C D D / P C D F

> 1.5 <

1.5

(8.20 (8.21

315 Table 8.7 Summary of Various Toxic Equivalent Factors I/TEF

Eadon

EPA

BGA

Nordic

1

1

1

1

1

1,2,3,7,8 - PentaCDD

0.5

1

0.2

0.1

0.5

1,2,3,4,7,8 - HexaCDD

0.1

0.03

0.04

0.1

0.1

1,2,3,6,7,8 - HexaCDD

0. t

0.03

0.04

0.1

0.1

1,2,3,7,8,9 - HexaCDD

0.1

0.03

0.04

0.1

0.1

1,2,3,4,6,7,8 - HeptaCDD

0.01

0

0

0.01

0.01

Octa CDD

0.001

0

0

0.001

0.001

0.1

0.33

0.1

0.1

0.1

1,2,3,7,8 - PentaCDF

0.05

0.33

0.1

0.1

0.01

1,2,3,4,7,8 - PentaCDF

0.5

0.33

0.1

0.1

0.5

1,2,3,4, 7,8 - HexaCDF

0.1

0.01

0.01

0.1

0.1

1,2,3,6,7,8 - HexaCDF

0.1

0.01

0.01

0.1

0.1

1,2,3,7,8,9 - HexaCDF

0.1

0.01

0.01

0.1

0.1

2,3,4,6,7,8 - HexaCDF

0.1

0.01

0.01

0.1

0.1

1,2,3,4,6,7,8 - HeptaCDF

0.01

0

0.001

0.01

0.01

1,2,3,4,7,8,9 - HeptaCDF

0.01

0

0.001

0.01

0.01

OctaCDF

0.001

0

0

0.001

0.001

Isomer 2,3,7,8 - TetraCDD

2,3,7,8 - TetraCDF

0-0.01 0.001-0.01 0 Other PCDD/PCDF 0 0 Adapted from: NATO, 1988; Barnes et al., 1986; BGA, 1985; and Ahlborg, 1989 Although PCDDs and PCDFs are ubiquitous (formed during burning of any material), in the environment their concentrations are very low. Industrial and thermal processes have increased the potential for PCDD/PCDF contamination of air, water and soils. Consequently, the waste entering an incinerator is contaminated with PCDD/PCDF as well. Published data consistently indicate that concentrations in waste are about 50 IJg(TE)/tonne (Environment Canada, 1985 & 1988; Tosine, 1985; Hagenmaier, 1989; Fricke et al., 1989; Lahl et al., 1991; Johnke and Stelzner, 1992).

316 Although uncontrolled waste incineration plants were considered to be among the major sources of PCDD/PCDF loading to the environment, the situation during the last five to ten years has changed dramatically. The understanding of formation and decomposition reactions for PCDD/PCDF has resulted in a reduction of these emissions from MSW incinerators by a factor of 10 to 100. In addition, methods to destroy PCDD/PCDF in APC residues have been developed to further reduce the potential problem. The PCDD/PCDF partitioning in old MSW incinerators (Figure 8.25) is based on mean data from numerous publications which have been checked for consistency (Vehlow and Vogg, 1991 ). Figure 8.25 Concentrations (TE) and Partitioning of PCDD/PCDF in an Old MSW Incinerator

The most important reactions and conditions for the production of PCDD/PCDF in MSW incinerators have been studied in great detail (Rghei and Eiceman, 1982; Griffin, 1986; Vogg and Stieglitz, 1986; Stieglitz and Vogg, 1987; Hagenmaier et al., 1987; Hinton and Lane, 1991). The findings can be summarised: PCDD and PCDF are formed mainly by heterogeneous oxichlorination of particulate carbon. The chlorine is supplied by alkali and earth-alkali chlorides. 9

An oxygen surplus in the atmosphere is essential.

317 Copper and to a minor extent other metal compounds act as catalysts. The temperature range is 250 to 450~ The critical temperature range for this de-novo synthesis occurs in the flue gases passing through the boiler and economiser, and consequently is the principal area for PCDD/PCDF formation. This mechanism has been confirmed not only in laboratory experiments, but also by the results of a great number of field tests. Investigations on full scale incinerators demonstrated a strong correlation between PCDD/PCDF levels and particulate carbon. Based on these findings, it was deduced that a reduction of the PCDD/PCDF levels in raw flue gas can be facilitated by excellent burnout of the waste. The formation of PCDD/PCDF in the coolest area of the boiler was verified by the analysis of flue gas sampled at different points across the boiler in a full scale incinerator (D0wel et al., 1990; Nottrodt et al., 1990) and by data on boiler ash discharged via separate hoppers at different temperature levels (Vogg et al., 1987; Mariani et al., 1990). An analysis of ashes sampled at different areas of an incinerator and an electrostatic precipitator elucidates the increase in PCDD/PCDF concentration with decreasing temperatures (Figure 8.26). The inlet temperature to the boiler was 950~ and the outlet temperature was 270~ No temperature profile across the boiler was reported. Only very low concentrations of PCDD/PCDF are present in the flue gas stream exiting the combustion chamber (<0.2 ng(TE)/Rm 3) (Environment Canada, 1985; L6ffler, 1989; DOwel et al., 1990; Mariani et al., 1990; Nottrodt et al., 1990). Concentrations in bottom ash from well-operated incinerators are typically <0.02 ng(TE)/g (Roffman, 1991; Johnke and Stelzner, 1992; Morselli et al., 1989; Sawell and Constable, 1993). Less than 5 pg(TE)/tonne of the combusted waste partitions to the bottom ash stream and less than 1 pg(TE)/tonne of waste is released into the flue gas. This means that at least 90% of the PCDD/PCDF fed into incinerators with the waste are destroyed by the combustion process. There is no data currently available on whether or not PCDD/PCDF formation can occur in the bottom ash as it cools, although a survey of PCDD/PCDF levels in German reports indicates that concentrations in quench tank water vary between 1 and 98 ng(TE)/L (Johnke and Stelzner, 1992). It should be noted that no details were given with respect to the operation of the incinerator or the quench tank (i.e. make up water addition). Furthermore, these concentrations may be the result of measuring the PCDD/PCDF content of suspended carbon particles introduced into the quench tank by the boiler ashes and accumulated during long periods of operation. In the raw flue gas of old incinerators sampled directly downstream of the boiler, mean PCDD/PCDF levels of about 30 ng(TE)/m 3 have been measured (Environment Canada, 1985, 1986 & 1988; Merz et al., 1989; DOwel et al., 1990; Mariani et al., 1990; Nottrodt, 1990; Vogg et al, 1990; Johnke and Stelzner, 1991; Vogg et al., 1991 ). Based on these data, PCDD/PCDF loading of 150 IJg(TEF)/tonne of waste can be calculated, which in

318 turn means that the raw flue gas downstream of the boiler carries three times more PCDD/PCDF than was fed into the incinerator. The PCDD/PCDF concentrations in the boiler ashes given in the cited studies ranged from about 0.1 to 5 ng(TEF)/g of ash. Figure 8.26 PCDD/PCDF in Ashes Sampled from Different Points of an MSW Incinerator 1~.000_~'~ P C D D 100.I~

mmm

1

t= 9t= o

10.000

~,

1.000-

o~

lil iaull

"l N" """ ,~"

PCDF

O ~

_ _

_ _

0.010-

mmiim]' ~

0.100-

A

B1

B2

B3

B4

r. []

-4 ~.4

i/

""

m m t

iii./

B5

B6

mum

in |am~u

m t.~ ""

N s

"

L" ~ C

D1

D2

sampling point

Since most of the PCDD/PCDF load present in the raw gas of older incinerators is bound to particulate matter, especially to carbon particles, it will be partitioned to the APC system residues. A sampling program in different German MSW incinerators resulted in a mean value of 10 ng(TEF)/g of ash, i.e. about 250 pg(TE)/tonne of waste combusted (Hagenmaier, 1989; Johnke and Stelzner, 1992). Highly chlorinated isomers are typically formed over the lesser chlorinated isomers. These data are consistent with other data found in the literature. Although temperature is the most important controlling factor for trace metal condensation, the types of dust removal system and the operating temperature are both important considerations when examining organic compounds. In the past, electrostatic precipitators were often operated at temperatures in the region of 300~ This obviously resulted in increased PCDD/PCDF formation, which has been confirmed by several studies (see below). As an example, a 20 tonne per hour incinerator fitted with an electrostatic precipitator operated at a temperature of 320~ resulted in an increase of seven times the PCDD/PCDF concentrations measured in the entrance to the precipitator (Vogg et al., 1989).

319 Recently published studies provide evidence that in some cases there may be another mechanism for PCDD/PCDF production in lower temperature precipitators as well (200 to 220~ (Hanay et al., 1986; Vogg et al., 1990; Kilgroe et al., 1991; Hiraoka et al., 1991). The underlying reaction is not yet fully known, however, it may be the influence of the electromagnetic charges which enhance formation in the device. The net formation of PCDD/PCDF between the boiler exit and the APC ash discharge hopper, given in Figure 8.25, should reflect these effects if an electrostatic precipitator is in use. The minimisation of PCDD/PCDF formation has been investigated as extensively as the formation mechanisms. Much of the data in this area point to improved combustion conditions. In addition, new designs to combustion chambers were found to be successful in limiting formation of the compounds. Now proper operating conditions and modern furnace designs are capable of reducing raw flue gas concentrations of PCDD/PCDF to <0.2 ng(TE)/m 3 (Environment Canada, 1988; Horch, 1990; Reeck et al., 1991). A few years ago, this value was impossible to achieve even in clean gas of old facilities. Moreover, the concentrations in the different ash streams were also considered very low, i.e. bottom ashes 0.007 ng/g, boiler ashes 0.023 ng(TE)/g, and APC ashes 0.213 ng/g. Although proper operating conditions and adequate design of the furnace are requisites for reduced PCDD/PCDF formation, efficient removal and/or destruction technologies are also warranted. Spray absorbers in combination with fabric filters (Environment Canada, 1986) and even electrostatic precipitators (Nielsen and M~ller, 1989) have been optimised. Activated charcoal filters have been developed with a high absorption potential (Dannecker and Hemschemeier, 1990). Similarly, activated carbon is now injected into the flue gas to sorb PCDD/PCDF and mercury (Mosch and Gottschalk, 1991; Regler, 1991) and catalytic oxidation has been successfully demonstrated (Hiraoka et al., 1989; Hagenmaier and Mittelbach, 1990; Fahlenkamp et al., 1991 ). The use of different reactive additives injected into the flue gas has also been examined in a semi-technical test facility (Lenoir and Hutzinger, 1989; Vogg et al., 1990a). Figure 8.27 illustrates the partitioning of PCDD/PCDF during incineration, based on a state-of-the-art incinerator facility. The contamination of bottom ashes is generally lower than levels found in some soils (about 5 ng/kg). Moreover, the gaseous emissions do not add substantially to today's background concentrations in ambient air (50 to 300 fg/Rm3). Furthermore, the PCDD/PCDF discharged with the boiler and filter ashes can be thermally destroyed by various treatment methods. Two processes have been developed and tested, and both guarantee excellent destruction (>99%). The first one applies a thermal treatment in a rotary kiln at about 400~ under oxygen deficient conditions, then copper salts are added as catalysts. A full scale facility is in operation (Schetter et al., 1990). A second process, the 3R Process, utilises the combustion chamber of the incinerator itself to decompose PCDD/PCDF in extracted and compacted filter ashes (Merz et al., 1989). All vitrification processes proposed for APC residues may also be appropriate for treatment of PCDD/PCDF as well.

320 Figure 8.27 Incinerator

Concentrations (TE) and Partitioning of PCDD/PCDF in a Modern

Based on the information given above, a modern MSW incinerator which is well operated and equipped with adequate APC devices is capable of meeting the most stringent emission regulations for PCDD/PCDF. In some European countries, this limit is 0.1 ng(TE)/Rm 3, however, even lower limits have been achieved during recent pilot plant studies (<0.01 ng/Rm 3) (Cleve, 1989; Vicinius and Knoche, 1991). Overall, modern incinerators act as net destroyers of PCDD/PCDF.

8.5.3 Polychlorinated Biphenyls (PCBs) PCBs are a class of noncorroding, highly stable, nonflammable chemical compounds which were used widely in electrical transformers, other electric applications, heat exchange equipment and to a lesser degree in inks, oils, sealants and caulking compounds. Their use was banned in 1977 due to their persistence and their propensity to bio-accumulate in fatty tissues of animals (Government of Canada, 1990). Although PCBs are suspected of being toxic, the toxicity is much lower than PCDD/PCDFs. In 1983, the estimated mean intake rate per person in industrialised countries was about 350 ng/kg of body weight/day (Bennett, 1983).

321 There has been a tendency to extend the toxicity equivalent factor defined for PCDD/PCDFs to these compounds (Bol et al., 1989; Safe, 1990) and first estimates suggest that emissions and residues from MSW incineration might increase the toxicity equivalence proposed for PCDD by approximately 10% (van Bavel et al., 1992). Data for PCB concentrations in MSW have ranged widely from below detection to 700 mg/tonne of waste, reflecting the sporadic presence in MSW. (Environment Canada, 1985, 1988, & 1992; Tosine et al., 1985; Fricke et al., 1989; Lahl et al., 1991 ). Concentrations in bottom ashes are very low, typically below detection limits (Environment Canada, 1993). The concentrations depicted in Figure 8.28, which attempts to provide an idea of partitioning for PCBs, are compiled from Italian and U.S. data and might overestimate the residual burden of these compounds (Morselli, 1989; Roffman, 1991). Although PCBs have not been detected in boiler ash (<5 ng/g) (Environment Canada, 1993), concentrations in the raw flue gas downstream of the boiler range between 20 ng/Rm3 and 1,000 ng/Rm3 in mass burner systems (Environment Canada, 1986 & 1988). In well operated incinerators, concentrations <12 ng/Rm3 could be achieved (Reeck et al., 1991). The levels reported for filter ashes of 10 to 270 ng/g are consistent with the raw flue gas concentrations (Sawell and Constable, 1993; Gonzales et al., 1991 ). Figure 8.28 Concentrations and Partitioning of Polychlorinated Biphenyls

322 The partitioning of PCBs during incineration indicates that modern incinerators are net destroyers of PCBs. Based on the fact that these compounds have been banned from use and their concentrations should decrease in the future, this group of compounds does not pose a potential problem with regard to MSW incineration. The concentrations of emissions from well-operated incinerators are typically <100 ng/Rm 3 (Environment Canada, 1986; Cleve, 1989; Morselli et al., 1989; Reeck et al., 1991), which means that these emissions are insignificant and do not add a substantial portion of PCB to the background intake.

8.5.4 Polychlorinated Benzenes Polychlorinated benzenes (CBs) are widely used in the chemical industry as intermediate products in special syntheses and as solvents for fats and resins. CBs are among the most thermally stable organic compounds. Under the specific conditions of elevated temperatures in the presence of water vapour, they can form PCDDs and PCDFs, hence they have a certain relevance in the waste incineration. These compounds are of relatively low toxicity, however, they bio-accumulate in fatty tissue of humans and animals. The acceptable daily intake of hexachlorobenzene for humans is in the order of 4 IJg/kg of body weight/day which is far greater than any possible dose which could be received through exposure to MSW incinerator emissions or ash (Greim, 1990). Since CBs are ubiquitous in the environment, they are present in MSW. The data from three separate studies indicate that concentrations in waste vary widely, e.g. 5 to 200 mg/tonne of waste (Environment Canada, 1985 & 1986; Tosine et al., 1985). Data on bottom ash concentrations from Environment Canada's NITEP Program (Environment Canada, 1993) and one American facility investigation (Roffman, 1991) are relatively consistent ranging from below detection to 80 ng/g. CBs can be formed at boiler outlet by the same type of de-novo synthesis which is responsible for PCDD/PCDF formation (Schwarz et al., 1990). Concentrations in the flue gas downstream of the boiler measured in mass burn units during the NITEP Program ranged from 4 to 12 IJg/Rm3 (Environment Canada, 1985 and 1986). In a new German incinerator, the concentration was only 15 ng/Rm 3 (Reeck et al., 1991 ). The partitioning of CBs during incineration (during good and poor operation) is given in Figure 8.29. Concentrations in boiler ash and APC residues ranged from below detection to 4.2 IJg/g (Environment Canada, 1985; Sawell and Constable, 1993). The data included samples collected during a wide range of operating conditions. "Good" operating conditions generally resulted in much lower concentrations of CBs being measured in all ash streams (Environment Canada, 1985, 1986, 1988 and 1991 ). Similar to the phenomenon of enhanced PCDD/PCDF formation in electrostatic precipitators, the concentrations of CBs also appear to increase due to the electrical charges (Hanay, 1986). Concentrations downstream of the precipitator increased by

323 a factor of about five. The partitioning of CBs during proper operating conditions indicates that modern MSW incinerators are net destroyers of these organic compounds. Figure 8.29 Concentration and Partitioning of Polychlorinated Benzenes

8.5.5 Polychlorinated Phenols The properties of polychlorinated phenols (CPs) are similar to those of polychlorinated benzenes. They are thermally stable and are used mostly as biocides, herbicides and fungicides, and bio-accumulate in fatty tissue. The partitioning of CPs during incineration is given in Figure 8.30. Concentrations of CPs in MSW between 3 and 2,000 mg/tonne of waste have been reported (Environment Canada, 1985, 1986, 1988 and 1991 ). Like PCDD/PCDFs and CBs, CPs can be formed at the boiler outlet and a second formation mechanism may occur in an electrostatic precipitator. Emissions can be kept far below 200 ng/Rm 3, especially if the best available technology is applied. Although carcinogenic effects can be realised at very high doses, the emissions from incinerators pose no significant health risk (Greim, 1990).

324 Figure 8.30 Concentrations and Partitioning of Polychlorinated Phenols

8.5.6 Brominated Hydrocarbons Since the chemical properties of chlorine and bromine are similar, the same mechanisms that form chlorinated organic compounds should also form brominated and mixed brominated/chlorinated compounds during incineration. The stability of hydrogen halides decreases with increasing atomic number (see Table 8.5), making it easier to produce elementary bromine from HBr or bromides than it is to form chlorine from HCI or chlorides. Although oxybromination might be a favoured reaction over oxychlorination, bromine concentrations are only 0.5% that of chlorine, hence the overwhelming bulk of formation compounds will be chlorinated. Brominated compounds, even brominated dibenzo-p-dioxins and dibenzofurans, have been detected in fly ash (Sch~fer and Ballschmiter, 1986; Oehme et al., 1987; Schwind et al., 1988). There are 3,500 isomers of mixed halogenated dibenzo-p-dioxins and dibenzofurans, and appropriate standards are only available for a few isomers making quantitative analysis of these compounds extremely difficult.

8.5.7 Polycyclic Aromatic Hydrocarbons (PAH) The group of polycyclic aromatic hydrocarbons (PAHs) consists of 16 planar condensated aromatic compounds, eight of which are suspected of being carcinogenic

325 (Menzie et al., 1992). PAHs are formed in all kinds of smouldering or pyrolytic reactions, thus their presence is an indicator of poor combustion conditions. The main sources of PAHs in the environment include car exhausts, domestic heating and atmospheric emissions from industry and sewage sludge. Figure 8.31 illustrates the partitioning of PAHs during incineration. Similar to the other organic compounds, concentration of compounds in the waste, the type of incinerator and the operating conditions under which the samples were collected will all greatly influence the concentrations of PAHs measured in the ash and the flue gases. Consequently, the data from the NITEP Program are widely scattered. The concentrations of PAHs in the waste varied from 10 to 10,000 mg/tonne of waste (Environment Canada, 1985, 1986, 1988 & 1991). Data from a German study indicate about 100 mg/tonne (Fricke et al., 1989). Figure 8.31 Concentrations and Partitioning of Polycyclic Aromatic Hydrocarbons

Since PAHs are considered semi-volatile, the highest concentrations are generally measured in bottom ash. Concentrations as low as 100 ng/g seem to be achievable with modern grate systems (Fricke et al., 1989). Concentrations of up to 700 ng/g have been measured in mass burn systems (Sawell and Constable, 1988; Morselli et al., 1989), up to almost 7 IJg/g in modular incinerators (Environment Canada, 1985) and 19

326 IJg/g in old RDF systems (Environment Canada, 1993). Boiler ash is normally only moderately contaminated with PAHs (20 to 500 ng/g) and concentrations measured in APC residues are typically even lower (Environment Canada, 1985 and 1988). Although a recent study has focused on the potential mobility of these compounds after deposition of ash by means of a special extraction procedure (Bauw et al., 1991 ), the low solubility of the compounds indicates these compounds do not pose a risk of contaminating groundwater. Although relatively high emissions (5 to 30 tJg/Rm3) have been measured during interrupted operation of an incinerator (Benestad et al., 1990), well-operated MSW incinerators are capable of achieving emissions of PAHs below 200 ng/Rm3 (Environment Canada, 1985, 1986, 1988 & 1991). PAH concentrations measured in stack gases during a test run of a charcoal filter were only 12 ng/Rm 3. The partitioning of PAHs during incineration indicates that although these semi-volatile compounds behave differently from the other chlorinated compounds discussed earlier, modern MSW incinerators act as a net destroyer of PAHs. REFERENCES

Ahlborg, U.G. "Nordic Risk Assessment of PCDDs and PCDFs." Chemospher .e..19: 603, 1989. Ando, J. "Review of Japanese NOx Abatement Technology for Stationary Sources." NO.x--Symposion Karlsruhe 1985, Proceedinqs Edited by O. Rentz, F. 1131eand M. Weibel. Universit~t (TH), Karlsruhe, A1, 1985. Angenend, F.J. and L. Trondt. Kraftwerkstechnik 70: 36-42, 1990.

"Schadstoffinput

- Schadstoffoutput." VGB

Baccini, P. and P.H. Brunner. "Behandlung und Endlagerung von Reststoffen aus Kehrichtverbrennungsanlagen." Gaz - Eaux- Eaux usees 65: 403, 1985. Barin, J...Thermochemical Properties of Inor,qanic Substances. Springer-Verlag, Berlin-G5ttingen-Heidelberg, 1993. Barnes, D.G., J. Bellin and D. Cleverly. "Interim Procedure for Estimating Risks Associated with Exposures to Mixtures of Chlorinated Dibenzodioxins and Dibenzofurans (CDDs and CDFs)." Chemosphere 15:, 1985, 1986. Barniske, L. and B. Johnke. "Die Abfallverbrennung im Spannungsfeld einer modernen Abfallentsorgung." M(~llverbrennun.qund Umwelt 4 Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin, 15, 1990.

327 Bartels, J., P. ten Bruggencate, H. Hausen, K.H. Hellwege, KI. Schafer and E. Schmidt Landolt-B0rnstein. Zahlenwerte und Funktionen aus Physik, Chemie, Astronomie~ Geophysik und Technik, sechste Aufla.qe, I1 Band: Ei,qenschaften der Materie in ihren A.a.qregat~ust,~nden, 2. Teil Gleic.h.qewichte aul3er Schmelz.qleich,qewichteq. Springer-Verlag, Berlin - G0ttingen - Heidelberg: 1, 1960. Bauw, D.H., G.A. de Wilde, G.A. Rood and T.G. Aalbers. "A Standard Leaching Test, Including Solid Phase Extraction, for the Determination of PAH Leachability from Waste Materials", Chemosphere 22: 713, 1991. Beek, A.I.M. van de, A.A.J. Cornelissen and Th.G. Aalbers. Fvsisch en Chemisc.h.. Onderzoek aan..~Huishoudelijk Afval, Pub. Rijksinstituut Voor Volksgezondheid en Milieuhygiene Bilthoven, Rapportnr. 738505005, June, 1988. Behrendt, Th., H. Dittrich and A. Lubecki. "Modellansatze zum Betriebsverhalten einer Rostfeuerung am Beispiel TAMARA." Symposium 25 Jahre LIT.., 5 Jahre TAMARA, KfK report 5000: 38, 1992. Belevi, H. "Wie wird die Schlacke erzeugt?" Entwicklun.q und Umsetzun.q neuer Qualit,~tsa.nforderun,qen in der Abfallwirtschaft Kartause Ittingen, September 14-16, 1993. Dokumentation Vertiefungskurs Vl/93, EAWAG, Z0rich: 54-74, 1993. Benestad, C., L. Hagen, A. Jebens, M. Oehme and H. Ramdahl. "Emissions of Organic Micropollutants from Discontinuously Operated Municipal Solid Waste Incinerators." Waste Mana.qe....mentand Research 8 193, 1990. Bennett, B.G. "Exposure of Man to Environmental PCBs - an Exposure Commitment Assessment." The Science of the Total Environment 29: 101, 1983. Bergstr6m, J. and J. Sundquist. "Emission Control for Municipal Solid Waste Incineration in Sweden." EWPCA-ISWA '..84, 6th European Sewa.qe and Refuse Symposium EWPCA Munich, 21 to 25 May 1984, Documentation, EWPCA, Ed., GFA, St. Augustin: 885, 1984. BGA. "Sachstand Dioxine." Bericht des Umweltbundesamtes 5/85: 264, 1985. Bol, J., M. van den Berg and W. Seinem. "Interactive effects of PCDDs, PCDFs and PCBs as Assessed by the E.L.S.-Bioassay." Chemosphe.re 19: 899, 1989. Borchers, H.-W. "Thermal Behaviour of Heavy Metals in Exhausts of Waste Burning Plants." Recyclin.q International Edited by K.J. Thom~ Kozmiensky. Berlin EF-Verlag: 1461, 1989.

328 Braun, H. and A. Gerig. "Mercury Emission Monitoring on Municipal Waste Combustion." Municipal Waste Combustion, Conference Papers and Abstracts from the 2nd Annual International Speciality Conference Tampa, FL April 15 - 19. Air and Waste Management Association, Pittsburgh, PA: 798, 1991. Braun, H., H. Metzger and H. Vogg. "Zur Problematik der Quecksilber-Abscheidung aus Rauchgasen von MQIIverbrennungsanlagen." 1. Teil, M011und Abfall 18: 62, 1986. 2. Teil, MQII und Abfall 18: 89, 1986. Braun, H., H. Metzger and H. Vogg. "Die Verbesserung der Abscheidung von Quecksilber aus Rauchgasen der MQIIverbrennung." Chemie-ln,qenieur-Technik 60: 135, 1988. Braun, H. and A. Gerig. "Entwicklung eines Quecksilberanalysators zur kontinuierlichen Messung in Rauchgasen." AbfallwirtschaftsJournal 2: 819, 1990. Bma, Th.G. "Toxic Metal Emissions from MWCs and their Control." Municipal Waste Combustion, Co..nference Papers and A.bstracts, Tampa, FL, April 15- 19, 1991. Air and Waste Management Association, Pittsburgh, PA: 145, 1991. Brunner, P.H. and J. Zobrist. "Die M011verbrennung als Quelle von Metallen in der Umwelt." MQII und Abfall 15: 221, 1983. Brunner, P.H. and W.R. Ernst. "Alternative Methods for the Analysis of Municipal Solid Waste." Waste Mana,qement & Rese..arc.h4: 147, 1986. Brunner, P.H. and H. M6nch. "The Flux of Metals through Municipal Solid Waste Incinerators." W.aste Mana.qement and Research 4: 105, 1986. Bundesministerium fQr Umwelt. "Naturschutz und Reaktorsicherheit 17." Verordnung zur DurchfQhrung des Bundesimmissionsschutzgesetzes, ( 1 7 . BlmSchV), BundesQesetzblatt, Teil 1, 2 545, 2: 832, 1990. Canadian Council of Ministers of the Environment (CCME). 0peratinQ and. Emission Guidelines for .Municipal Solid Wast.e Incinerator.s Report CCME-TS/WM-TRE003, June 1989. Cernuschi, S., M. Giuliani and R. Marforio. "Kurzfristiges Umweltverhalten von Flugasche und Schlacke aus stQdtischen FestmQIIverbrennungsanlagen." Messen und Analysieren. an Abfallbehandlunqsanlaqen, Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 308, 1987. Cleve, U. "Stand of Technical and Economical Application of Activated Carbon for Flue Gas Washing Plants with Waste Incineration." Recyclin.q International Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 1265, 1989.

329 Corey, R.C., B.J. Cross and W.T. Reid. "External Corrosion of Furnace-WalI-Tubes-II. Significance of Sulfate Deposits and Sulphur Trioxide in Corrosion Mechanism." Transactions ASME 67: 289, 1965. Dalager, S. "Danske Massebalanceundersegelser." Konference affaldsforbraendin,q under nye betin.aelser, Kopenhagen, September 3, 1993.

om

Dannecker, W. and H. Hemschemeier. "Level of Activated-Coke Technology for Flue Gas Dust Collection Behind Refuse Destruction Plants Looking at the Problem from the Special Aspects of Dioxin Separation." In Or.aanohalo.qen Compounds, Vol. 4, Dioxin 90 - EPRI-Seminar Edited by O. Hutzinger and H. Fiedler. Ecoinforma Press, Bayreuth: 267, 1990. de Soete, G.G. 15th Symposium Int. on Combustion, Tokyo, August 1974. Proceedings: 1093, 1975. Dobberstein, J. "Energie und Schadstoffe im Hausm011." Mell und Abfall 15: 305, 1983. Dransfeld, P., H. Hunsinger and H. Vogg. "Die thermische NOx-Emissionsminderung mit Melamin und verwandten Verbindungen bei M011verbrennungsanlagen." VG__..BB Kraftwerkstechnik 72 995, 1992. D(3wel, U., A. Nottrodt and K. Ballschmiter. "Simultaneous Sampling of PCDD/PCDF Inside the Combustion Chamber and on Four Boiler Levels of a Waste Incineration Plant." Chemosphere 20: 1839, 1990. Eighmy, T.T., M.A. Guay, S.A. McHugh, P.H. Thompson, N.E. Kinner and T.P. Ballestero. "Codisposal of Municipal Solid Waste Bottom Ash and Wastewater sl ud ge~. - " ..Municipal Waste Incineration, Proceedings, October 1 - 2, 1987, Montreal, Quebec. Edited by Environment Canada. 349, 1987. Eighmy, T.T. Personal communication concerning petrographic data from the WASTE Program, Phase II, Speciation of Elements in MSW and Incinerator Residues, Progress Review, August 1993. Environment Canada. .The National Incinerator Testin.a and Evaluation Pro.aram" Twosta~e Combustion (Prince Edward Island) Environment Canada Report EPS 3/UP/l, 1985. ,

Environment Canada. The National Incinerator Testin.q and Evaluation Pro.aram: Air Pollution Control Technolo.qy Environment Canada Report EPS 3/UP/2,, 1986. Environment Canada. National Incinerator Testin.q and Evaluation Pro.aram: Environmental Characterization of Mass Burnin.a Incinerator Techn01o.av at Quebec City, Summary Report. Environment Canada Report EPS 3/UP/5, 1988.

330 Environment Canada. The Environmental Characterization of RDF. Combustion Technol0.qy, Summary. Report Environment Canada Report, 1991. Environment Canada. The National Incinerator Testin,q and Evaluation Pro.qram: A Summ..aryof the Characteriza.tion and Treatment Studies on Res.idues from Municipal Solid Waste Incineration Prepared by S.E. Sawell and T.W. Constable for Waste Management Branch. Environment Canada Report, 1993. Fahlenkamp, H., G. Mittelbach and H. Hagenmaier. "Katalytische Zerst5rung von Dioxinen." M011verbrennun.q und Umwelt 5 Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 469, 1991. F~l~ler, K., H. Leib and H. Sp,~hn. "Korrosionen an M(311verbrennungskesseln." Mitteilun.qen der VGB 48:126, 1968. Faulstich, M., A. Freudenberg, G. Kley and P. Kt~cher. "Thermodynamische und mineralogische 0berlegungen zur Inertisierung fester R0ckst,~nde aus der Abfallverbrennung." MQIIverbrennuna und Umwelt. 4 Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 339, 1990. Faulstich, M. "Feste R0ckstande aus Abfallverbrennung und Rauchgasreinigung." MQIIverbrennun.q und Umwelt 5 Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 297, 1991. Fricke, K., H. Vogtmann, J. Jager and M. Wilken. "Organische Schadstoffe in Bioabfallkomposten." Mell und Abfall 21" 472, 1989. Gonzales, M., M. Fernandez, L. Jimenez and L. Hemandez. "Application of Multivariate Analysis to the Determination of PCB in Fly Ash from Municipal Incinerators." Chemosphere 18: 2213, 1989. G5ttlicher, R. and P. Anton. "Reststoffe AbfallwirtschaftsJournal 2: 376, 1990.

aus

der

Mellverbrennung."

Government of Canada. PCB Fact Sheets Health and Welfare Canada, 1990. Greim, H. "Toxicological Evaluation of Emissions from Modern Municipal Waste Incinerators." Chemosphere 20: 317, 1990. Griffin, R.D. "A New Theory of Dioxin Formation in Municipal Solid Waste Combustion." Chemosphere 15, 1987, 1986. Guest, T. and O. Knizek. Mercu.ry Control at Burnaby's Municipal Waste Incinerator, 1992.

331 Hagenmaier, H., M. Kraft, H. Brunner and R. Haag. "Catalytic Effects of Fly Ash from Waste Incineration Facilities on the Formation and Decomposition of Polychlorinated Dibenzo-p-Dioxins and Polychlorinated Dibenzofurans." Environmental Science and Technolo.av 21 1080, 1987. Hagenmaier, H. "Polychlorierte Dibenzodioxine und polychlorierte Dibenzofurane Bestandsaufnahme und Handlungsbedarf." In Halogenierte organische Verbindungen in der Umwelt, VDI-Berichte VDI-Verlag, DQsseldorf. 745: 939, 1989. Hagenmaier, H. and G. Mittelbach. "Versuche zum katalytischen NOx- and Dioxinabbau im Abgas einer Hausm011verbrennungsanlage." VGB Kraftwerkstechni .k 70: 491, 1990. Hanay, Y., T. Katou and H. Ide. "Chlorination Process in the Refuse Incinerator's Equipment." Yok0h.a.ma Kokuritsu Daiqaku Kank.yo Ka,qaku Kenkyu Senta Kivo13: 37, 1986. Handbook of Che..mistry and Physics 72nd Edition. CRC Press, Cleveland, OH, 1992. Hausner, H.H. "Discussion of the definition of the term 'sintering'." Materials Science Mono.araphs, 4 .......Si..ntering - New Developments Edited by M.M. Ristic. Amsterdam, Oxford, New York, 1979. Hinton, W.S. and A.M. Lane. "Characteristics of Municipal Solid Waste Incinerator Fly Ash Promoting the Formation of Polychlorinated Dioxins." Chemosphere 22: 473, 1991. Hiraoka, M., N. Takeda, S. Okajima, T. Kasakura and Y. Imoto. "Catalytic destruction of PCDDs in flue gas." Chemosphere. 19: 361, 1989. Hiraoka, M., T. Fujii, K. Kashiwabara, K. Teyama and M. Kondo. "The Removal Efficiency of Dioxin in Flue Gas of MSW Incinerator - Comparision of a Fabric Filter with an Electrostatic Precipitator." Ch.emosphere 23: 1439, 1991. Hjelmar, O. "Leachate from incinerator ash disposal sites." Municipal Wa.ste Incineration, Proceedin.as, October 1 - 2 , 1987, Montreal, Quebec. Edited by Environment Canada. 287, 1987. Horch, K. "Optimierung von Feuerungen mit dem Ziel der Schadstoffminderung." T.hermische Abfallbehandlun.a in Entsorgun.askonzepten VDI-Berichte Nr. 637, VDI-Verlag, DQsseldorf, 173, 1987. Horch, K., A. Christmann and G. Schetter. "Zukunftorientiertes Feuerungskonzept zur Abfallverwertung." MQII und Abfall 22" 288, 1990. Horch, K. and D. Schneider. "Das Brenn-Schmelz-Verfahren -Verglasen von RQckstQnden aus der Abfallverbrennung." M011verbren..nun.aund Umwelt 5 Edited by K.J. Thom~-Kozmiensky. EF.-Verlag, Berlin 341, 1991.

332 HundesrQgge, T. "Phasenanalytische Untersuchungen MQIIverbrennungsan agen. Aufschlul~ 41 281 1990. I

,,

.

an

Filteraschen

aus

,

Hurst, B.E. Technoloav Advances in Selective Non-Catalytic NOx-Reduction (Therma! DeNOx) Exxon Research and Engineering Company, Houston, TX, 1983. __

Jochem, M. "Stand der Rostfeuerungstechnik Entsor.aun.aspraxis Spezial No. 10: 28, 1989.

und Entwicklungstendenzen."

Johnke, B. and E. Stelzner. "Results of the German Dioxin Measurement Programme at MSW Incinerators." Chemosphere 10: 345-355, 1992. Kautz, K. "Korrosionsursachen in HausmQIIverbrennungsanlagen." Mitteilun.qen der VGB 51 398, 1975. Kilgroe, J.D., W.S. Lanier and T.R. van Alten. "Montgomery County South Incinerator Test Project: Formation, Emission and Control of Organic Pollutants." Municipal Waste Combustion, Conference Papers and Abstracts, Tampa, FL, April 15- 19, 1991. Air and Waste Management Association, Pittsburgh, PA: 190, 1991. Lahl, U., L. Wilken, B. Zeschmar-Lahl and J. Jager. "PCDD/PCDF Balance of Different Municipal Waste Management Methods." Chemosphe.r.e.23: 1481, 1991. Lenoir, D. and O. Hutzinger. "PCDD/F Verhinderung der Entstehung chlorierter Dibenzodioxine und -furane bei der M011verbrennung." UWSF-Z. Umweltchem. Okotox 4: 3, 1989. Lieser, K.H. and H. Elias. _Z. Anor.a. AII.a. Chem. 316: 208, 1962. Lindquist, O., K. Puromaki and P. Schager. "Mercury from Waste Combustion." Appendix to Residues from Waste Incineration -Chemical and Physical Properties by J. Hartl~n and P. Elander, Swedish Geotechnical Institute, 1986. LOftier, H. "Wiener Anforderungen an eine umweltvertrQgliche Entsorgung von MVA-Reststoffen." In Reststoffe aus der thermischen Abfallbehandlung, VDI-Berichte Nr. 753, VDI-Verlag, DQsseldorf: 51, 1989. LSffler, H. "MQIIverbrennungsanlagen und polychlorierte Dioxine und Furane." Beitr,~.ae zum Umweltschutz Heft 19, Magistrat der Stadt Wien, 1989. Lorber, K.E. "Die Zusammensetzung des MOlls und die durch MQIIverbrennungsanlagen emittierten Schadstoffe." MQIJverbrennun.q und RauchQasreinigun.q Edited by K.J. Thome-Kozmiensky. EF-Verlag, Berlin: 559, 1983.

333 Mariani, G., E. Benfenati and R. Fanelli. "Concentrations of PCDD and PCDF in Different Points of a Modern Refuse Incinerator." Chemosphere 21 507, 1990. Menzie, C.A., B.B. Potocki and J. Santodonato. "Exposure to Carcinogenic PAHs in the Environment." Environmental Science and Technology 26 1278, 1992. Merz, A., K. Horch, J. Schneider, L. Stieglitz, J. Vehlow and H. Vogg. "Thermal Destruction of PCDD/PCDF Compounds by 3R Product Recycling into the Solid Incineration Furnace." International Conference on Municipal Waste Combustion, Hollywood, FL, April 11 - 18, 1989. Vol. 2, Supply and Services Canada, 7A/37, 1989. Metzger, M. and H. Braun. "ln-situ Mercury Speciation in Flue Gas By Liquid and Solid Sorption Systems." Chemosphere Volume 16: No. 4, 1987. Morselli, L., S. Zappoli, A. Liberti, M. Rotatori and E. Brancaleoni. "Evaluation and Comparision of Organic and Inorganic Compounds between Emission and Immission Samples from Municipal Solid Waste Incinerators." Chemosphere 18: 2263, 1989. Mosch, H. and J. Gottschalk. "Filsorption - Kombinierte Filtration und Sorption bei Gesamtkonzepten for die Rauchgasreinigung." M011verbrennun.aund Umwelt 5 Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 227, 1991. M0hlenweg, U. and Th. Brasser. "Reststoffe bei der Hausm011verbrennung." M011und Abfall 22: 53, 1990. Nagase, H., Y. Ose, T. Sato and M. Yamada. "Mercury Methylation by Ash from Refuse Incineration." Science of the Total Environment No. 53, 1986. NATO, CCMS. "International Toxicity Equivalency Factors (I/TEF) Method of Risk Assessment for Complex Mixtures of Dioxins and Related Compounds." Report 178, December 1988. Nelson, W. and C. Cain. "Corrosion of Superheaters and Reheaters of Pulverized Coal-Fired Boilers. part 1." Transactions ASME Series A, 83: 194, 1960. Part 2, Transactions ASME Series A, 84: 468, 1960. Nielsen, K.K. and J.T. M~ller. "Reduction of Chlorinated Dioxins and Furans in the Flue Gas from Incinerators with Spray Absorbers and Electrostatic Precipitators." Chemosphere 19: 367, 1989. Nottrodt, A., U. DOwel and K. Ballschmiter. "The Influence of Increased Excess Air on the Formation of PCDD/PCDF in a Municipal Waste Incineration Plant." Chemosphere 20: 1847, 1990.

334 Oberste-Padberg, R. and K. Schweden. "Zur Freisetzung von Wasserstoff aus M5rtlen mit MVA-Reststoffen." Wass.er, Luft und Boden 6: 61, 1990. Oehme, M., S. Mane and A. Mikalsen. "Formation and Presence of Polyhalogenated and Polycyclic Compounds in the Emissions of Small and Large Scale Municipal Waste Incinerators." Chemosphere 16: 143, 1987. Olie, K., P.L. Vermeulen and O. Hutzinger. "Chlorodibenzo-p-Dioxins and Chlorodibenzofurans are Trace Components of Fly Ash and Flue Gas of Some Municipal Incinerators in the Netherlands." Chemosphere 6: 455, 1977. Pauling, L. The Nature of the. Chemical Bond, 3rd Ed. Cornell University Press, 1969. Perrier-Rosset, A. "Valorization or Disposal of Incineration Residues in France." SARDINIA 89, Second International Landfill Symposium, Porto Conte (Alqhero), Italy. Proceedin,qs - Vol.I, C.I.P.A.s.r.I., Milan. LIX 1, October 9-13, 1989. Pietrzeniuk, H.J. "Stand der Verwertung der R0ckstQnde aus M011verbrennungsanlagen." Thermische Verfahren der Abfallwirtschaft Statusbericht 1985, VDI-Gesellschaft Energietechnik, D0sseldorf: 153, 1985. Poiger, H. and C. Schlatter. "Pharmacokinetics of 2, 3, 7, 8-TCDD in Man." Chemosphere 15: 1489, 1986. Reeck, G., W. Schr(~der and G. Schetter. "Zukunftsorientierte Abfallverbrennung in der MVA Ludwigshafen." M011und Abfall 23: 661, 1991. Regler, H. "Flugstromverfahren mit Kalk-Aktivkoks-Gemischen." MQIIverbrennu.nQ und Umwelt 5 Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 245, 1991. Reichel, H.H. and U. Schirmer. "Waste Incineration Plants in the FRG: Construction, Materials, Investigation on Cases of Corrosion." Werkstoffe und Korrosion 40: 135, 1989. Reimann, D.O. "Heavy Metals in Domestic Refuse and Their Distribution in Incinerator Residues." Waste Mana,qement and Research 7" 57-62, 1989. Remy, H. Lehrbuch der AnorQa.nischen Chemie, Vol. Verlagsgesellschaft Geest, and Portig KG, Leipzig 883, 1961.

II

Akademische

Rghei, H.O. and G.A. Eiceman. "Adsorption and Thermal Reactions of 1, 2, 3, 4-Tetrachlorodibenzo-p-Dioxin on Fly Ash from a Municipal Incinerator." Chemosphere 11" 569, 1982. Richman, M., D. Fickling and J. Hahn. "Mercury Removal Studies at a Municipal Waste Combustor in Marion County, Oregon." Report by Ogden Martin Systems Inc., 1993.

335 Roffman, H. "Major Findings of the U.S. EPA/CORRE MWC-Ash Study." Municipal ..Waste Combus.tion, Conference Papers and.Abstracts Tampa, FL, April 15- 19, 1991. Air and Waste Management Association, Pittsburgh, PA: 96, 1991. Ruytenbeek, A. and P.M. Braams. "Einflu6 der Vorwegnahme von organischen Abf~llen auf die Abfallverbrennung." MQIIverbrennun.q und Umwelt 3_ Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 85, 1989. Safe, S. "Polychlorinated Biphenyls (PCBs), Dibenzo-p-Dioxins (PCDDs), Dibenzofurans (PCDFs), and Related Compounds: Environmental and Mechanistic Considerations Which Support the Development of Toxic Equivalency Factors (TEFs)." CRC Crit. Rev. Toxicol. 21" 51, 1990. Sawell, S.E. and T.W. Constable. The Natio&al Incinerator Testin.q and Ev_aluation ProQram: the Combustio...n. Characterization of Mass Burning Incinerator Te.c.hnolo.qy Quebec City,.....Volume V!., Assessment of Con.taminant Leachability.. from R.esidue.s Environment Canada Report, Manuscript Series" IP-82, 1988. Sawell, S.E. and T.W. Constable. "The National Incinerator Testing and Evaluation Program: Characterisation of Residues from a Refuse Derived Fuel Combustion System and a Modular Municipal Waste Incinerator." Proceedin.qs of the International Conference on Municipal Waste Combustion Hollywood, FL, 1989. Sch~fer, W. and K. Ballschmiter. "Monobromo-polychloro-derivates of Benzene, Biphenyl, Dibenzofurane and Dibenzodioxin Found in Chemical-Waste Burning." Chemosphere 15 755, 1986. Schetter, G., K. Horch, R. Stutzle, H. Brunner and H. Hagenmaier. "Low Temperature Thermal Treatment of Filter Ash from Municipal Solid Waste Incinerators for Dioxin Decomposition on a Technical Scale." Or.qanoh.alo.qenCompound_s Vol. 3, Dioxin 90 EPRI-Seminar. Edited by O. Hutzinger and H. Fiedler. Ecoinforma Press, Bayreuth: 165, 1990. -

Schneider, J. "Determination of Elemental Waste Composition by Analysis of Incineration Residues." Recyclinq International V o l . 1.1. Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 318, 1986. Schuster, H. "Minderung der NOx-Emissionen aus VDI-Berichte Nr. 495 VDI-Verlag, DQsseldorf: 143, 1984.

Kraftwerksfeuerungen."

Schwarz, G., L. Stieglitz and W. Roth. "Formation Conditions of Several Polychlorinated Compound Classes on Fly Ash of a Municipal Waste Incinerator." Or.qanohalo.qen Compounds Vol. 3, Dioxin 90 - EPRI Seminar. Edited by O. Hutzinger and H. Fiedler. Ecoinforma Press, Bayreuth: 169, 1990.

336 Schwind, K.-H., J. Hosseinpour and H. Thoma. "Brominated/Chlorinated Dibenzo-p-Dioxins and Dibenzofurans. Part 1 Brominated/Chlorinated and Brominated Dibenzo-p-Dioxins and Dibenzofurans in Fly Ash from a Municipal Waste Incinerator." Chemosphere 17" 1875, 1988. Stieglitz, L. and H. Vogg. "On Formation Conditions of PCDD/PCDF in Fly Ash from Municipal Waste Incinerators." Chemosphere 16: 1917, 1987. Taggart, A. Handbook of Mineral Dressin.q, Ores and Industrial Minerals J. Wiley, New York, 1948. Teller, A.J. and J.M. Quimby. "Mercury Removal from Incinerator Flue Gas", 1991. Tillman, D.A., A.J. Rossi, and K.M. Vick. Incineration of Municipal and. Hazardous Solid Wastes Academic Press: 29, 1989. Tobler, H.P. "Behandlung und Verfestigung von ROckst~nden aus M(~llverbrennungsanlagen in der Schweiz" Entsor~un,qspraxis Special 1988 188, 1988. Toetsch, W. and H. Gaensslen. Polyviny!c.hlorid Verlag T0V Rheinland, K61n: 78, 1990. Tosine, H.M., R.E. Clement, V. Ozwacic and G. Wong. "Levels of PCDD/PCDF and other Chlorinated Organics in Municipal Refuse." Chemosphere 14: 821, 1985. van Bavel, B., I. Fangmark, S. Marklund, G. Soderstrom, K. Ljung and C. Rappe. "Quantitative Analysis of Organic Trace Compounds by Isotope Dilution" Formation of the Non-ortho PCBs 77, 126, 169 during Incineration." OrQanohalo,qen Compounds 8: 225, 1992. Vehlow, J., H. Braun, K. Horch, A. Merz, J. Schneider, L. Stieglitz and H. Vogg. "Semi-Technical Demonstration of the 3R Process." Waste Mana.qement a..nd Research 8: 461, 1990. Vehlow, J. and H. Vogg. "Thermische Zerst6rung organischer Verbindungen." M~311verbrennun,qund Umwelt 5 Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin: 447, 1991. Vehlow, J., K. Jay, C. Rittmeyer, L. Stieglitz and W. Vielohr. "Disposal of Rigid PUR Foams in Municipal Waste Incinerators." Alternatives to CFCs and Halons~L International Conference Berlin, 24th to 26th February 1992. Organized by BMU, Umweltbundesamt, U.S. EPA. Proceedings, Umweltbundesamt, Berlin: 543, 1992. Vehlow, J., G. Pfrang-Stotz and J. Schneider. "Reststoffe - Charakterisierung, Behandlung, Verwertung." .S.ymposium 25 Jahre LIT, 5 Jahre TAMARA, KfK report 5000: 124, 1992.

337 Vicinius, J. and R. Knoche. "Abscheidung von Dioxinen, Furanen und Schwermetallen in einer Gasreinigung." Me ll und Abfall 23: 229, 1991. Vogg, H. "Verhalten von, (Schwer-)Metallen bei der Verbrennung kommunaler Abf~lle." Chemie-ln,qenieur-Technik 60: 740, 1984. Vogg, H. and L. Stieglitz. "Thermal Behaviour of PCDD/PCDF in Fly Ash from Municipal Incinerators." Chemosphere 15: 1373, 1986. Vogg, H., A. Christmann and K. Wiese. "Das 3R-Verfahren, ein Baustein zur Schadstoffminderung bei der Abfallverbrennung." Recyclin.q von Haush.altsabf~llen Edited by K.J. Thom6-Kozmiensky. EF-Verlag, Berlin, Vol. I 471, 1987. Vogg, H., M. Metzger and L. Stieglitz. "Recent Findings on the Formation and Decomposition of PCDD/PCDF." Municipa.! Solid ,VVas.te In.cinerati.on, Waste Mana.aement and Researc_h 5: 285, 1987. Vogg, H., A. Merz, L. Stieglitz and J. Vehlow. "Chemical Process Engineering Aspects of Dioxin Reduction in Waste Incineration Processes." VG..BKra.ftwerkstechnjk 69: 795, 1989. Vogg, H., H. Hunsinger and L. Stieglitz. "Contribution to Solving the Problem of Dioxins Generated during Waste Incineration". Chemj.cal En,qineering Technology 13: 221, 1990. Vogg, H.,A. Merz, L, Stieglitz, F. Albert and G. Blattner. "Zur Rolle des Elektrofilters bei der Dioxin-Bildung in Abfallverbrennungsanlagen." AbfallwirtschaftsJournal 2: 529, 1990. Vogg, H., H. Hunsinger, A. Merz, L. Stieglitz and J. Vehlow. "Head-end-Techniken zur Dioxinminderung." Proze~f0hrunq und Verfahrenstechnik der .M0!lverbrennun~ VDI Berichte Nr. 895, VDI-Verlag, D0sseldorf: 193, 1991. Vogg, H., A. Merz and E. Becker, E. "Permanent Monitoring of the Emission of Highly Ecotoxic Substances from Waste Incineration Plants." Proceedin.as of the 1991 Incineration Conference Knoxville, TN 555, May 1991. WASTE Program. Effects of Waste Stream Characteristics on MSW Incineration: The Fate and Behaviour of Meta.ls, Final Report Report prepared for Environment Canada, U.S. Environmental Protection Agency and International Lead Zinc Research Organization, 1993. WHO. Dioxins and Furans from Municipal Incinerators Environmental Health Series 17, World Health Organization, Regional Office for Europe, Copenhagen: 25, 1987.

This Page Intentionally Left Blank

339

CHAPTER 9 - BOTTOM ASH

Bottom ash is the most significant residual by-product from MSW incineration. As noted in Chapters 3 and 8, it accounts for 85 to 95% of all the residues produced during combustion. Bottom ash is usually comprised of grate ash and grate siftings. Occasionally, economiser ashes can be added to the bottom ash process stream. Unless noted otherwise, the data presented in this chapter are for bottom ash. By virtue of combustion temperature and grate design, bottom ash is a slagged material containing lithophilic elements with low vapour pressures. The slagged material is granular. It can be intermingled with ferrous and non-ferrous metal and other uncombustibles if these constituents are present in the MSW prior to combustion. In Europe, bottom ash is generated as a separate waste stream in the combustion facility. This material is also widely utilised as an aggregate substitute. Consequently, a relatively large data base exists on the physical and chemical properties of bottom ash (Hartl6n and Elander, 1986; Ludvigsen and Hjelmar, 1992; Stampfli et al., 1990; van der Sloot, 1992; Vehlow, 1992; Vehlow et al., 1992). Recent interest in bottom ash utilisation in the United States (Chesner et al., 1988; LIRPB, 1993; Eighmy et al., 1992) as well as fundamental ash characterisation studies for a variety of incinerator types in Canada under the NITEP and WASTE programs (Sawell et. al., 1989a; 1989b; 1990a; 1990b; Sawell and Constable, 1988; 1993; Bridle and Sawell, 1985; PRRI, 1992; WASTE Program, 1993) have also produced large data bases on physical and chemical properties. Table 9.1 describes the data base that was used to compile data on bottom ash characteristics from around the world. In the table are data from combustors in Canada (seven facilities), Denmark (eight facilities), Germany (three facilities), the Netherlands (five facilities), Sweden (eight facilities) and the United States (eight facilities). The table depicts the type of combustor that is used at each facility, the grates used, the grate manufacturer, the facility capacity (in tonnes per hour) of the combustor units, any bottom ash processing that takes place at the full scale level and various procedures that were used to collect the samples evaluated in this chapter. In all, 39 facilities were evaluated and generated data. Where possible, time-dependent data on bottom ash characteristics from the Concord, New Hampshire facility in the United States is provided, hereafter referred to as the Concord study (Eighmy et al., 1991 ). This study, in part, assessed the variability in ash properties collected from 18 sampling events over a 548-day (1.5 year) period. Each sampling event produced four consecutive hourly composite samples and a daily composite. These data allow the reader to assess hourly, daily and seasonal variation in the bottom ash process stream for a typical modern mass burn facility.

Table 9.1 Bottom Ash Database Comb. Unit" Typea Grateb ManufC Rating

Bottom Ash' Stream Processing

Siftings1 Samp.' Riddlings' Period

Country

Facility

Canada

-GVRD -GVRD

MB MB

RE RE

1 1

10 10

Q Q

Yes No

-PEI -LVH

TS TS

RM RM

1.4 3.8 5.4 10.4 4.1 48

Q Q Q Q

NO No No No

Denmark

FB Stoker -SWARU -QUC MB RE -3M TSIRK -Amagerforbraedingen MB MGIRK

2 2 3 1 4 5

Q Q,T,*45mm

No Yes

Germany

-Vesforbraedingen -KARA, Roskilde -Kolding II -REFA, Nykobing -ASA,Sanderborg -Middlefart -KAVO, Slagelse A

MB MB MB MB MG MB MB MB

MGlRK MG MG MG MG MG MG RG

5 5 6 5 5 6 6 7

50 20 8 7 10 4 10 10

Q,T,<45mm Q.T,<45mm Q,T,<45mm Q,T,<45mrn Q,T,<45mm Q,T,<45mm Q,T,<45mm Q,T,M,<40mm

Yes Yes Yes Yes Yes Yes Yes Yes

1985 Hjelrnar, 1992 1991 1991 1991 1991 1991 1991 1988+Vehlow,1992;

C -Den Haag -AVR

MB MB MB MB MB

MGIRK RE RG RE RE

5 8 9 1 1

12 25 21 7 12.5

Q,T,M,<40mm Q,T,M,<40mrn Q,T,M,<40mm Q,T,M,<40mm Q,T,M,<40rnrn

Yes Yes Yes Yes Yes

1988+ 1985 TAUW, 1988; 1985 Stoelhorst, 1991 1988 1989

MB MB

RE MG

1 10

45 2.5

Q,T,M,<40mm

Q

Yes YES

1988 1985

FB MB

RE

11 1

4.5 10

Q Q

YES YES

1985 1985

MB MB

MG MG

8 1

31 25.2

Q Q

YES YES

1985 1985

Netherlands

Sweden

-AVIRA -Rotterdam -Amsterdam -Bollmora -Bolln& -Halrnstad -Linkbping -Malmb

-

Reference

1988192 Sawell et al., 1990a; WASTE Program, 1993 1992 WASTE Program, 1993 1984 1088 1988 1987

NITEP, 1985 Sawell et al., l989a Sawell et a!., l989b Sawetl & Constable, 1988

1989 Sawell et al.. l990b 1985191 Lundvigsen 8

Hartlbn & Elander, 1986; SGI, 1993

Country

--

-

USA-

Comb. Unit' Type' Grateb Manufc Rating

Facility

Bottom Ashe Stream Processing

Siftings1 Samp.' Riddlingsf Period

-Hbgdalen

MB

RG+RE

12

37

YES

1985

-UmeA

RE

YES

1985

MG

7,13 14.5.6

16.8

-Unasala

MB ME

31

Q

YES

1985

-SW Brooklyn, NY -Concord, NH

MB

RE

8

17

Q

Yes

MB MB ME

RE MGlRK RE

8

4

Q

7

8

19

? Q

Yes Yes Yes

-Glen Cove. NY -Dry Scrubber 1 -Dry Scrubber 2 -Dry Scrubber 3

-

-Babylon, NY a

1 2 3 4

1 20 Q -Massburn MB RE Combustor T v ~ eMB : (mass bum). FB (fluidised-bed).TS [two-staae mass burn). (&tary K kiln)," MG' (moving gate). RG (roier grate). RE Grate T ~ ~ ~ : ' R (reciprocating grate), V (Vibratory Grate), RM (Ram), TB (travelling bed) Grate manufacturer. (See numeric legend) Nominal rating of a unit at the facility in tonnes per hour. Process stream treated by quenching (Q), tromrnel (T), grizzly (G), magnet (M). Are siftings and riddlings blended with the bottom ash that was sampled? Martin Consumat Tricil Enercan

5 Vslund 6 Bruun 8 Ssrensen 7 Deutche-Babcock 8 von Roll

9 10 11 12

'

1

Reference

Chesner, et al., 1988 1990192 Eighmy et al., 1992

-

LIRPB, 1992a,b

1986 1986 1986

LIRPB, 1992a,b LIRPB, 1992a,b LIRPB, 1992a,b

-

LIRPB, 1992a,b

Yes 1989 Kosson et al., 1992 How were samples processed; S (sieving), C (crushing), D (drying). Where were sahpl& collected; G (knd orgrate), L lift);^ (con;ey6r), DC (drag chain), B (bunker), P (pile). Year of sample collection. Type of sampling regime; G (grab), C (composites of grabs). T (timedependent sampling), Z (sampling during altered combustor operation).

Dijrr Kablitz Generator VKW + Martinwehrle

13 KtK ofenbau 14 Widmer + Ernst

342

9.1 PHYSICAL CHARACTERISTICS OF BOTTOM ASH 9.1.1 Gross Composition It is important to evaluate the gross composition of bottom ash so that comparisons can be made to soil and other aggregate-like materials. Gross composition is usually comprised of procedures that involve defining a reject fraction, visual classification of the bottom ash, evaluation of the water content of the bottom ash, assessing the concentration of ferrous material in the bottom ash and evaluating the degree of bumout of the combustor and the resulting loss on ignition (LOI) of the bottom ash. All of these parameters are considered gross descriptions of bottom ash and indicate a general environmental and physical quality of the bottom ash.

Reject Fraction Bottom ash process streams at most modern incineration facilities are processed prior to subsequent utilisation or disposal. This treatment can entail quenching, size separation (trommeling, grizzly separation, magnetic separation) and storage at the incineration facility. Frequently, the trommeling, grizzly separation and magnetic separation of bottom ash produces a reject fraction that comprises some percentage of the total bottom ash process stream generated at the facility. The proportion of the reject fraction depends on materials in the waste feed and the utilisation requirements for the bottom ash. For instance, in Europe, a cutoff size of 40 mm is frequently used. This automatically creates an operationally defined reject fraction that contains large slag and metal pieces. The metal material is usually recovered and recycled. The slag material may be crushed or disposed. The weight percentage of the reject fractions in Denmark and the United States are shown in Table 9.2. The oversized material reject fraction is typically comprised of ferrous material, nonferrous metallic material, large pieces of slagged bottom ash material, construction debris and unburned MSW. Frequently, this fraction contains enough ferrous to have economic value. The lower limit of the particle size of the reject fraction is determined by the separation system being employed by the facility. In facilities in Denmark, the reject fraction tends to be greater than 45 mm. In the United States, the reject fraction will most likely be greater than either 50 mm or 19 mm, depending on the type of utilisation that is envisioned for the bottom ash. The quantity of reject material that is generated is dependent on the particle size that is being excluded from the bottom ash process stream. In Denmark, at the seven facilities that were evaluated, an average of 6 to 9% (range of 1.3 to 22.7%) of the total bottom ash process stream is rejected by trommels that are removing the greater than 45 mm fraction. This reject fraction contains metals that are both ferrous and nonferrous in nature. It also contains ceramic and glass-like material, as well as slag and unburned material (Ludvgson & Hjelmar, 1992). In the U.S., where the size cut-off tends to be lower for bottom ash utilisation, larger quantities of reject fractions are observed (see Table 9.2). Typically, in the United States, mean reject fractions can be 3.2 to 32.9% (range of 2.4 to 66%) of the total bottom ash process stream. These U.S. data are

343 based on a small percentage of MSW combustion facilities in the U.S. and are not based on mass balances conducted at full scale.

Table 9.2 Bottom Ash Reject Fraction Country

Facility

Reject Fraction, % Min Max

Denmark

Amagerforbraedigen

KARA, Roskildea Kolding IIa REFA, Nykobinga ASA, Sonderborga Middlefart" KAVO, Slagelsea United States Southwest Brooklyn, NYb

Mean

Median

7.5 12.2 1.3 9.2 3.5 9.4 7.6 9.7 12.1

Reference n

7

Lundvigsen &

7 Hjelmar, 1992 7 7 7 7 7 170 Chesner et al., 1988

12.6 22.7 13.1 12.7 11.6 16.4 29.5

19.3

-

Concord, NHc

21.0 66.0

32.9

31.0

72

Eighmyet al., 1992

Dry Scrubber 1b

6.1 23.8

16.9

-

6

LIRPB, 1992a

Dry Scrubber2 b 2.4 4.0 3.2 4 Dry Scrubber 3=' 5.9 12.5 8.4 5 Dry Scrubber3 b 9.8 15.6 12.4 5 a Rejected fraction is the > 45 mm fraction removed from the raw bottom ash; based on wet weight b Rejected fraction is the> 50.8 mm fraction removed from the raw bottom ash; based on wet weight c Rejected fraction is the > 19 mm fraction removed from the raw bottom ash; based on wet weight.

Figure 9.1 shows how the reject fraction percentage can vary over time for the Concord, facility (Eighmy et al., 1992). A reject size of greater than 19 mm was selected to comply with the use of bottom ash in paving applications. This is a smaller particle size cut-off than those typically used in Europe (40 to 50 mm). The data indicates that the reject fraction is relatively constant. The long term variability seen over a 1.5 year period is as variable as the values seen from four consecutive hourly composites within a single sampling day.

Visual Classification

Table 9.3 provides information on the visual classification of bottom ash fractions that are collected for either storage, disposal or utilisation. Data are provided for the Netherlands, Sweden and the United States. The methods for visually classifying bottom ash involves the quantification of the percent composition of metal material, slag material, stone and ceramic material, glass, and organic material. In the data that is available (see Table 9.3), the majority of the fraction that is passed through the size cut-offs is comprised of metal (3 to 46.9%) and slag (27 to 61.8%) material. This

344 Figure 9.1 Bottom Ash Rejects as a Function of Time

Mass Rejected, %

55 50

45 )

40

.

!

i

-

i

I

35

I

.... i~K_

i i

30

t ,II nl

J

iI

.!i

25

\\\ /

I\

20

15

Overall Average

Hourly Average

10

.,

0

I

I

I

I

I

I

I

I

I

2

4

6

8

10

12

14

16

18

Sampling Day The vertical bars are the 95% Confidence Interval After Eighmy et al., 1992

20

345 combined fraction is typically 50% of the passing material. Often glass is found in concentrations of 10.8 to 44.9% and therefore comprises a major fraction of the material. Ceramics, stones, and organics frequently comprise much lower fractions of this bottom ash process stream. Table 9.3 Visual Classificatio n Gross Compositional Analysis of Bottom Ash Country Netherlands

Sweden

Facility

Reference

% Composition

/

Ferrous %

Slag

Stone

Glass

Ceramic

Organic

Amsterdam b

3

27

34

27

5

4

Dordrecht t)

4

29

26

27

7

6

Den Haag b

7

38

21

21

9

4

Rotterdam b

6

32

30

20

5

7

Maim0 c

-

45.0

8.7

44.9

1.3

0.1

Maim(5 (j

-

55.6

2.0

40.0

2.4

-

-

50.6

10.1

35.9

3.4

-

United

Maim0 e Dry Scrubber 21

16.4 (95)

53.5

-

25.3

4.0

1.0

States

Dry Scrubber 31

41.5 (98)

37.3

-

16.4

4.4

0.4

Dry Scrubber 31

46.9 (97)

33.5

-

17.3

2.4

0.0

Babylon, NY f

24.3 (86)

61.8

-

10.8

a Visual Classification. b Fraction larger than 50 mm.

c Fraction between 5.6-8 mm. d Fraction between 8-11.2 mm.

Stoelhorst, 1991

Hartl~n & Lundgren, 1991 LIRPB, 1992a

3.3 0.0 I Fraction between 6.35-50.8 mm.

Water Content

The geotechnical water content (or moisture content), as described in Chapter 7, is the weight of water in a sample relative to the oven dry weight of the sample; it is expressed as a percentage. Table 9.4 provides information on the water content of bottom ash materials. Data are presented for both Sweden and the United States. The water content of bottom ash is frequently dependent upon the type of quenching that is utilised to cool the bottom ash. Some incineration facilities utilise a spraying device to quench the ash and this tends to reduce the water content of the bottom ash that is generated. Wet ram systems also produce a drier ash. Other incineration facilities utilise quench tanks with drag chains. This tends to create a bottom ash that has a relatively high water content. The data in Table 9.4 show that typical mean water contents in bottom ash range from 9.4 to 58.4%. Water content is an important parameter of gross compositional analysis of bottom ash, particularly with regards to its transportation, storage and processing for either disposal or utilisation. It also has an economic impact on transportation and disposal. High levels of water content can increase the weight of material that is being transported and the degree of leachate drag out that is generated at a facility. Water content is usually controlled for utilisation of bottom ash either in granular fill applications, road sub-base applications or asphalt paving applications.

346 Table 9.4 Bottom Ash Water Content Country Facility Sweden

Malmt) Maim0 United States S.W. Brooklyn, Nua.d Concord, NHb,d Dry Scrubber 1c,d Dry Scrubber 2c,d Dry Scrubber 3c,~ a Fraction less than 50.8 mm. b Fraction less than 19 mm. c Fraction less than 50.8 mm.

Water Content, % Min Max Mean Median 11.3 23.6 41.0 53.6 15.7

16.4 22.9 72.5 27.9 36.6 65.2 37.8 54.8 46.9 63.1 58.4 26.4 19.9 d ASTM D2216 " Aged Ash. f Fresh Ash. .

Reference n 13 5 62 65 4 2 10

Hartl6n& Rogbeck, 1989 Hartl6n& Rogbeck, 1989 Chesneret al., 1988 Eighmyet al., 1992 LIRPB,1992 LIRPB,1992 LIRPB,1992

Frequently, the drying of bottom ash and the aging of bottom ash outdoors allows for evaporative losses of water contents that can reduce the water content levels. Work by Hartl~n and Rogbeck (1989) has shown that the water content can drop from 23 to 16% upon aging. Frequently the heat of hydration that is generated while the ash ages is high enough to cause significant evaporative losses to take place. Dewatering can also cause water loss. Some water can be taken up into crystalline structures during aging reactions. These tend to be less important loss or transformation reactions than evaporation. Maintaining some level of moisture allows for the bottom ash to remain in aggregated form and prevents fugitive dust problems. Recent work by Chesner (1993) has indicated that moisture contents above 15% are helpful in preventing fugitivity in outdoor storage piles.

Ferrous Content Table 9.5 provides information on the ferrous content of the bottom ash fractions that are considered for disposal or utilisation. Data are shown for the Netherlands and the United States. In spite of trommelling, the bottom ash passing the reject cut-off can contain appreciable ferrous material as determined by magnetic separation techniques. The evaluation of the ferrous content of bottom ash is important with regards to economics and its potential utilisation, and also with regards to evaluation of a recyclable component of the bottom ash. The ferrous content of bottom ash fractions that are considered for utilisation or disposal tends to range from 0.4 to 43.5% for all samples. Typical mean values are 1.3 to 25.8%. The variation is due to MSW feed composition and to methods in determining the ferrous content. Figure 9.2 provides information on the grain size distribution of ferrous particles in bottom ash (Eighmy et al., 1992). As can be seen, the ferrous content tends to predominate in the coarser fractions, suggesting that larger particles are predominately

347 ferrous materials. The relative predominance of ferrous material in bottom ash is dependent upon the MSW feed composition and the size cut-offs that are used for producing bottom ash process streams. Figure 9.3 provides information on how the ferrous content varies as a function of time at the Concord facility (Eighmy et al., 1992). As can be seen in the figure, the ferrous content seems to be relatively constant over a 548-day period. The long-term variability seen over a 1.5 year period is as great as the variability seen over four consecutive hourly composites within a single sampling day. Table 9.5 Bottom Ash Ferrous Content Country Netherlands

Facility

Ferrous Content, % Min

Max

AVI I a,b

0.4

4.4

AVI 2

0.6

AVI 3

0.9

Mean Median

Reference n

1.3

-

29 TAUW, 1988

8.7

3.18

-

26

3.6

2.32

-

26

AVI 4 0.6 3.3 United States Concord, NH~ 13.4 43.5 a Fraction less than 45 mm. ~ b Method based on magnetic separation, d

20 1.59 72 Eighmy et al., 1992 25.8 25.1 Fraction less than 19 mm. Method based on magnetic separation.

Loss on Ignition Loss on ignition (LOI) is the weight fraction (expressed as a percentage) of material that is lost during ashing at 550~ for two hours. Table 9.6a provides information on the LOI values for bottom ashes from Canada, Denmark and the United States. LOI is an important parameter to measure in bottom ash because it is a gross measure of the relative burnout of the combustor that is generating the bottom ash. Data from Switzerland (St~mpfli, 1992) suggests that up to 60% of the LOI is organic carbon; the remainder is carbonates and tightly bound water of hydration. LOI values can be a function of the type of MSW feed, combustor type and its operation, the relative moisture content of the MSW that is being combusted and the type of combustor operations that are taking place. As can be seen for data presented in Table 9.6a, typical mean values for bottom ashes in Canada range from 3.5 to 29.2%. Denmark mean values range from 1.9 to 6.3%. In the United States mean values range from 3.7 to 6.4%. Values of approximately 3% or less indicate a high degree of burnout (Brunner et al., 1987). Based on data gleaned from the NITEP, the WASTE Program and other Canadian studies, significant differences in LOI are found as function of combustor type. As shown in Table 9.6b, the LOI content of bottom ash collected from poorly operated twostage combustion systems was much higher (20 to 30% LOI values) than ash from

348 Figure 9.2 Particle Size Distribution of Ferrous Materials from Bottom Ash

100

Percent

Passing,

%

80-

60-

40-

20-

0 0.01

I

I I I~-F~"-'--t

~

I

I I I Itll

0.1

I

I I I IIII

1

Particle

After Eighmy et al., 1992

l

Size, mm

I

10

I

I ! JIH

100

349 Figure 9.3 Bottom Ash Ferrous Content as a Function of Time

Ferrous Content, %

45

40

35 /

\\\\1 II \\ _H__:-U-': //I

30

25

20

I I

~ m

\~\\ i/

__i

/~\

_

/\,

I'

/

-..,..

'

=

i

i

15

10

Overall Average

Hourly Average

0 0

I

I

I

I

I

I

I

I

I

2

4

6

8

10

12

14

16

18

Sampling Day The vertical bars are the 95 % confidence intervals After Eighmy et al., 1992

20

350 Table 9.6a Bottom Ash Loss on Ignition Country Canada

GVRD PEI

" b c d

Max

Mean

2.5

2.6

2.65

4

28.4

8

Bridle & Sawell, 1986 Sawell et al., 1989b Sawell et al., 1989a

23.6

33.4

Median

n

Min

LVH

-

-

29.2

2

SWARU

-

-

4.9

2

Sawell et al., 1990a

-

-

3.5

12

Sawell & Constable, 1988

3M

18.7

22.1

20.4

4

Sawell & Constable, 1990b

Amagerforbraending ",f

6.1

6.5

6.3

KARA, Roskilde "'f

2.6

2.8

2.7

ASA, S~nderborg "'f

2.4

2.8

2.5

REFA, Nyk~bing "'f

3.1

3.8

3.4

Kolding II "'f

2.2

2.3

2.2

KAVO, Slagelse "'f

1.7

2.2

1.9

Middlefart "'f

3.8

4.5

4.1

QUC Denmark

Reference

Loss on Ignition, %

Facility

Luvigsen & Hjelmar, 1992

United

Southwest Brooklyn,NY b,g

1.4

10.5

4.3

62

Chesner et al., 1988

States

Concord, NH c'Q

3.2

10.7

6.4

6.2

72

Eighmy et al., 1992

Dry Scrubber 1d'h

3.5

4.0

3.7

-

4

LIRPB, 1992a

Mass burn e'g

4.4

4.9

4.6

-

3

Fraction Fraction Fraction Fraction

Mid-Conn less than 45 mm. less than 50.8 mm. less than 4.75 ram. less than 50.8 mm.

0.6

-

Kosson et al., 1992 Sawell et al., 1991

1.5 9 Fraction less than 2.0 mm, with crushing of over size material. f Loss on ignition at 550~ g ASTMC114. h ASTM D2974. -

-

Table 9.6b Bottom Ash Loss on Ignition Combustor Type

LOI Content (%)

Reference

Two-stage (poor operation)

18.7 - 29.2

Sawell & Constable, 1993

Two-stage (good operation)

12.6- 16.5

Sawell & Constable, 1993

Older mass burn

2.5-3.5

Sawell & Constable, 1993

Modern mass burn

0.1 -1.7

WASTE Program, 1993

Older RDF Modern RDF

4.9

Sawell & Constable, 1993

0.6- 1.5

Sawell & Constable, 1993

351 either mass burn or semi-suspension combustion systems (LOI values less than 5%). Two-stage systems employ semi-pyrolytic operating conditions in the primary chamber to produce energy-rich flue gases which are then passed into a highly-oxidative secondary combustion chamber. As a result, the bottom ash contains a relatively high proportion of uncombusted material. It should be noted that although some of these data were collected from poorly operated systems, even well operated two-stage systems generate bottom ash with LOI contents in the range of about 10% without special primary chamber modifications. The results also indicate that modern mass burn and RDF systems are capable of achieving low LOis (<2%). Figure 9.4 shows the loss on ignition of bottom ash as a function of bottom ash particle size (Eighmy et al., 1992). As can be seen, there are two maxima, one at a larger particle size (10 mm) which is uncombusted char material and paper. The other tends to be at very small particle sizes (< 0.25 mm) and reflects the fact that very fine materials in bottom ashes can be organic materials. Figure 9.5 shows how bottom ash LOI changes as a function of time at the Concord, New Hampshire facility (Eighmy et al., 1992). The data indicate that the LOI content from 18 sampling events over the 1.5 year period is relatively constant and is as variable as the variability seen within four consecutive hourly samples within a single sampling day.

Dissolvable Solids Content Some limited data are available on the dissolvable solids content of bottom ash. The content is directly related to the soluble mineral content of the residue. The data come from the NITEP and WASTE program studies (Sawell and Constable, 1993; WASTE Program, 1993). The sequential batch extraction procedure was used to evaluate the leaching properties of various bottom ashes from a variety of North American facilities. One aspect of the extraction test is that total dissolved solids can be determined in the leachates. They can be summed for the four sequential extractions and related to the initial weight of ash. The data show that two-stage bottom ashes have the lowest dissolvable solids content (2.5 to 4.5%). Modern mass burn bottom ashes range from 3.0 to 14%. One RDF facility had bottom ashes with values ranging from 6.5 to 7.5%. The data indicate that more complete mineralisation (e.g. better burn-out) increases the soluble mineral content of the bottom ash. 9.1.2 Gravimetric Characteristics Gravimetric characteristics of bottom ash are important in evaluating the civil and geotechnical properties of bottom ash. These calculations are needed in designing Portland cement and asphalt job mixes as well as in evaluating field densities of these materials.

352 Figure 9.4 Bottom Ash LOI as a Function of Particle Size Loss On I g n i t i o n , %

24

|

0

i

i

i

i

1

#If)

#20

#40

#80

#200

l

1/2 " 3;8"

#4

Pan

P a r t i c l e Size

After Eighmy et al., 1992 Figure 9.5 Bottom Ash LOI as a Function of Time 12

Loss o n I g n i t i o n , %

11 I0 9

!

'i li ....;ou~

0

2

4

..i.. 6

8

..... i,,,~ .... .... i " 10

12

S a m p l i n g Day

The verticle bars are the 95% intervals

After Eighmy et al., 1992

14

16

. 1 18

30

353

Specific Gravity

Specific gravity is defined as the ratio of the weight of a given volume of a sample to the weight of an equal volume of water at standard temperature and pressure. It is reported as a dimensionless number. Specific gravity is usually a very important parameter to measure in bottom ash when consideration is given to utilising bottom ash in civil engineering construction. The procedure for determining specific gravity is usually conducted on two specific size fractions of granular materials. Usually the fine fraction is material less than 4.75 mm in diameter. The coarse fraction is material larger than 4.75 mm in diameter. Table 9.7a provides data on bulk specific gravity of bottom ash materials from facilities in the United States. The data from the Concord facility show that the fine fraction has a mean value of 1.86, with ranges of 1.49 to 1.86. The coarse fraction has a mean value of 2.19, with ranges of 1.82 and 2.43. There seems to be good uniformity among the determinations of the bulk specific gravity from the facilities that were evaluated within the United States. These values classify bottom ash as a lightweight aggregate. Table 9.7b provides information on the saturated surface dry (SSD) specific gravity of bottom ash from a number of facilities in the United States. The data from the Concord facility show that the fine fraction has a mean value of 2.13, ranging from 1.81 to 2.37. The coarse fraction has a mean value of 2.35, with a range for individual values of 1.90 to 2.41. There are good agreements between the bulk specific gravity determinations (SSD) between the facilities within the United States. Table 9.7c provides information on the bottom ash apparent specific gravity from a number of facilities in the United States. The data from the Concord facility show that the fine fraction has a mean value of 2.55 with a range of 2.11 to 2.93. The coarse fraction has a mean value of 2.51 with a range of 2.28 to 2.93. Again, there is good agreement between the different facilities. Figure 9.6 shows the variation of bulk specific gravity as a function of time at the Concord, New Hampshire facility (Eighmy et al., 1992). As can be seen in the figure, the coarse bulk specific gravity does exhibit some variability as a function of time over the 1.5 year sampling period. The fine bulk specific gravity is less variable; the variation seen within any four consecutive hourly composite samples in a day is similar to the variation seen over the long term. Evaluation of apparent specific gravity and bulk specific gravity saturated surface dry shows similar behaviours as a function of time.

Absorption

Absorption is used to calculate the change in weight of an aggregate due to the absorption of water into permeable pore spaces within aggregate particles. Bottom ash

354 T a b l e 9.7a Bottom Ash Specific Gravity Country Facility

United States

Concord, NH a'c Dry Scrubber 2 b'c Dry Scrubber 3 b'c Dry Scrubber 3 b,c

Bulk Specific Gravity Fine ~ Max Mean Median

1.49

2.13 1.86

-

-

n 72 1 1 1

Min 1.82

Max Mean Median n 2.19 72 2.43 2.19 2.11 1 2.17 1 2.23 1

Bulk Specific Gravity (SSD) Fine d Coarse e Min

Max Mean Median

Concord, N H a,c 1.81 2.37 2.13 Dry Scrubber 2 b,c Dry Scrubber 3 ~,c Dry Scrubber 3 b,c

T a b l e 9.7c Country Facility

1.88

1.81 1.70 1.76

-

T a b l e 9.7b Country Facility

United States

Coarse e

Min

-

-

.

2.12 2.03 1.98 2.04

n

Min

72 1 1 1

1.90 .

Apparent Specific Gravity Fined Coarse e

Min Max Mean 2.11 2.93 2.55 United Concord, NHa,c States Dry Scrubber 2 b,c Dry Scrubber 3b,c Dry Scrubber 3b,c Fraction less than 19 mm. Fraction less than 50.8 mm. ASTM C127, C128.

Max Mean Median 2.33 2.35 2.21 2.26 2.30

2.41 .

-

n 72 1 1 1

Reference

Median n I Min Max Mean Median n 2.56 7212.282.83 2.51 2.50 72 Eighmy et al, 1992 2.31 11 ! 2.35 1 LIRPB, 1992a 2.38 2.38 1 2.45 2.41 1 d FracUonless than 4.75 mm. e FracUongreater than 4.75 mm. I

355 Figure 9.6 Bottom Ash Bulk Specific Gravity as a Function of Time B u l k (I)ry) S p e c i f i c G r a v i t y

B u l k (Dry) S p e c i f i c G r a v i t y 2"5 I

Fine Ash

2.4

2.2

t tltttttt1tlltttttit

1.8

1.6

2.

1.4

Coarse Ash

1

1.9

1.2 ..... Hourly Average 1

0

2

Overall Average

"

t

,

t

J

I

,

I

t

~

4

6

8

10

12

14

16

18

20

Sampling

Day

..... Hourly 1.7

0

Average

Overall

Average

i

i

t

2_

1

[

[

2

4

6

8

10

12

14

Sampling

Day

___L____:

16

......... 18

20

The vertical bars are the 95 % confidence interval After Eighmy et al., 1992 is a highly porous aggregate, therefore it has the tendency to absorb water. The absorption of water is also a useful predictor for the potential for bottom ash to absorb asphalt during the manufacturing of asphaltic cement or to retain water during field compaction. Table 9.8 provides data on absorption characteristics for bottom ashes from United States facilities. As with specific gravity, absorption is evaluated for both fine and coarse fractions. For the fine fraction, a mean value of 14% is seen with minimum values of 7.6% and maximum values of 27.9% for all measurements. For coarse fraction, a mean value of 5.7% is seen, with minimum and maximum values ranging from 1.7 to 13.4% for all measurements. There is good agreement between values from the different facilities in the United States. The fact that absorption values are higher for the fine fraction again supports the contention that the fine fraction is a highly porous material with a high surface area to volume ratio that has the capacity to absorb large quantities of water. Most natural aggregates have lower absorption values. Figure 9.7 shows the variation in bottom ash absorption as a function of time for the Concord, New Hampshire facility (Eighmy, et al., 1992). The long-term variability seen over a 1.5 year period is slightly greater than the variability seen within four consecutive hourly composites within any single sampling day.

356 Table 9.8 Bottom Ash Absorption Country Facility

Absorption, % Reference Fineb Coarsec Min Max Mean Median n IMin'Max Mean Median n ' 5.4 72 Eighmy et al., 1992 United Concord, NHa 7.6 27.9 14.7 14.0 7211.7 13.4 5.7 12.0 1 I " " " 4.7 1 LIRPB, 1992a States Dry Scrubber 2" Dry Scrubber 3a 17.0 1 I " 4.1 1 Dry Scrubber 3" 16.1 ~13.2 1 b Fraction < 4.75 mm. c Fraction > 4.75 mm. "ASTM C127, C128.

Figure 9.7 Bottom Ash Absorption as a Function of Time 30

Absorption,

%

14

Absorption,

F i n e Ash

H!I

25

20

%

..... Hourly Average

Overall Average

C o a r s e Ash

tjIIttttLIIjtt tII tt ttittittittitF1 ..... Hourly Average

0

Overall Average

I

1

[

I

t

I

2

4

6

8

lO

12

.~ampllng

_ ~ L

14

[

_._1~__

16

18

20

Day

0

I

I

1

I

I

I

1

[

I

2

4

6

8

10

12

14

16

18

Sampling

Day

20

The vertical bars are the 95 % confidence intervals After Eighmy et al., 1992

Unit Weight

Table 9.9 provides information on bottom ash unit weight from a number of facilities in the United States and Canada. Typical mean values range from 955 to 1,420 kg/m 3. Typical minimum and maximum values are 732 and 1,510 kg/m 3 for all measurements, respectively. There is reasonably good agreement about bottom ash unit weight values for each of the facilities that were evaluated. These unit weight values show that bottom ash is a lightweight aggregate.

357 Table 9.9 Bottom Ash Unit Weight Country Facility Canada GVRD

Unit Weight, kg/m 3 Reference Min Max Mean Median n 1,3701,5101,420 4 WASTE Program, 1993

Southwest Brooklyn, Nu a,b 732 1,229 1,054 62Chesner et al., 1988 Concord, NH ",c 1,039 1,234 1,157 1,159 20 Eighmy et al., 1992 Dry Scrubber I a'b 1,090 1,183 1,152 4 LIRPB, 1992a Dry Scrubber 2 a'b 955 956 955 2 Dry Scrubber 3a,b 1,102 1,378 1,215 5 Dry Scrubber 3",b 1,150 1,312 1,234 5 c Fraction less than 19 mm. a ASTM C29. b Fraction less than 50.8 mm.

United States

Figure 9.8 shows bottom ash unit weight as a function of time at the Concord, New Hampshire facility (Eighmy et al., 1992). As can be shown in the figure, bottom ash unit weight was relatively constant over the 1.5 year sampling period.

9.1.3 Gradation Figure 9.9 shows a typical bottom ash grain size distribution. The distribution is classified as well-graded, meaning that there is equal abundance of coarse and fine material. Such uniform gradation is important to the compactability of bottom ash and the potential to utilise bottom ash as an aggregate substitute. Table 9.10 provides information on bottom ash effective size and uniformity coefficients. Effective size is determined by calculating the grain diameter where 10% of the material is passing. Coefficient of uniformity is the ratio of the grain diameter in millimetres corresponding to 60% passing by weight to the effective size of that material. As can be seen in Table 9.10 the mean effective size of bottom ash is 0.293 mm. Minimum and maximum values for all measurements are 0.127 and 0.508 mm. The mean uniformity coefficient was 21.6 with minimum and maximum values of 12.8 and 38.0. The data that are shown in Table 9.10 are from only one facility in the United States. The effective size and uniformity coefficients indicate that bottom ash is a wellgraded gravelly sand. Table 9.10 Bottom Ash Effectiv e Size and Uniformity Coefficients Country Facility Effective Size, mm I UniformityCoefficient Reference Min Max Mean Median n I Min Max Mean Median n USA Concord, ... NHa 0.1270.5080.2930.25472112,82138.02!.:6820.6872 Eighmy et al., 1992 " Fraction less than 19 mm.

358 Figure 9.8 Bottom Ash Unit Weight as a Function of Time 1300

Unit

Weight,

kg/m3

1200

1100

1000

..... 95~0 C. I.

Average

900 0

1

i

i

t

i

L

i

2

4

6

8

10

12

14

Sampling

_.!

i~

16

18

Day

After Eighmy et al, 1992 Figure 9.9 Bottom Ash Grain Size Distribution 100

Passing,

Percent

0 0.01

i

J

iJ~lllJ

%

i

....

i

Jl~Jlll

0.1 Particle

After Eighmy et al., 1992

|

,

,Jl

1 Size. mm

100

359 Figure 9.10 shows how bottom ash effective size and uniformity coefficients vary as a function of time for bottom ash samples obtained from the Concord facility. The data indicate that both effective size and uniformity coefficient were reasonably constant over the 1.5 year sampling period. Figure 9.10 Bottom Ash Effective Size and Uniformity Coefficient as a Function of Time 0.6

Effective Size, mm

40

Uniformity

Coefficient

35

0.5

30

0.4

,-.. li

i. , ,:,'I

25

0.3

20

. i!/!

0.2 0.1 ..... Hourly Average

0

..... Hourly Average

Overall Average

1

i

1

i

i

i

i

i

i

2

4

6

8

10

12

14

16

18

Sampling

Day

0

20

0

Overall Average

i

t

i

i

i

i

i

i

i

2

4

6

8

10

12

14

16

18

Sampling

Day

After Eighmy et al., 1992 Percent Fines The concentration of fine material in bottom ash is an important consideration when bottom ash is to be used as an aggregate substitute. The percent fines can frequently create problems because that fraction is highly absorptive for water, asphaltic cement and Portland cement. Frequently, high fine contents create a material that has a tendency towards freeze-thaw susceptibility and durability failure. The values for percent fines shown in Table 9.11 are from facilities in Germany, the Netherlands and the United States. The fraction denoting fines can be different in Europe compared to the United States. In Europe a fine fraction is usually denoted as that material passing a 63 pm mesh sieve. In the United States, a fine fraction is denoted as that material passing a 75 pm mesh sieve. Nevertheless, mean values for fines range from about 1.9 to 7.4%. Minimum and maximum values for all measurements are 1.0 and 10.1% respectively. There is good agreement between the values seen from the different facilities in the different countries.

360 Table 9.11 Bottom Ash Percent Fines Country Facility Min Max a Germany A 2 7 Bla,b 3 8 B2,.c 2 7 C" 2 6 Netherlands AVI 1' 6.2 9.5 AVI 2" 3.7 7.2 AVI 3" 4.8 9.6 AVI 4' 4.3 10.1 United Southwest Brooklyn, Nud,e 1.0 3.2 States Concord,NH~'~ 2.17 6.57 Dry Scrubber 3d,~ 2 3 Dry Scrubber 3d,~ 1 3 a Finesdenoted as fraction passing 63 pm. d b Unit 1 e c Unit 2

Reference Fines, % Mean Median n 4 5 Vehlow, 1992 5 4 5 4 4 4 7.3 29 TAUW, 1988 5.9 26 7.4 26 7.3 26 62 Chesneret al., 1988 1.9 72 Eighmy et al., 1992 3 . 9 6 3.92 10 LIRPB, 1992a 2 2 10 Finesdenoted as fraction passing 75 pm. ASTMC136 (dry sieving technique).

As bottom ash exhibits some friability, the production of fines during processing operations may occur. The levels of fines in bottom ash are somewhat problematic with regards to the utilisation of bottom ash in civil engineering construction applications because fines increase water capillarity and promote frost susceptibility. This means that under certain utilisation scenarios this fine fraction may need to be removed from the bottom ash process stream. At certain ash processing facilities in Europe, fines are removed by trommeling processes using agricultural trommels. Figure 9.11 shows how percent fines vary as a function of time at the Concord facility (Eighmy et al., 1992). The data indicate that the fine fraction was relatively uniform over the 1.5 year sampling period. The variations seen over time were as great as the variations observed within four consecutive hourly sampling events.

9.1.4 Durability The assessment of the durability of bottom ash is an important characterisation when considering the utilisation of bottom ash. Frequently, bottom ash is considered as an aggregate substitute and it is important to characterise how durable bottom ash is in comparison to natural aggregates.

Soundness The sodium or magnesium sulphate soundness test is generally the accepted method for aggregate soundness testing. Data are provided in Table 9.12 on percent losses

361 observed to bottom ash samples subjected to soundness testing. The data are from facilities in the United States. Soundness testing usually involves evaluation of both fine (< 4.75 mm) and coarse (> 4.75 mm) fractions. The data shown in Table 9.12 indicate that for the fine fraction, mean percent losses ranged from 1.6 to 11.91. The coarse fraction mean values were 2.6 and 2.9. As with sorption, the fine fraction is more susceptible to expansive fragmentation compared to the coarse fraction. This is because bottom ash fine material is more porous than the coarse material. There is a wide variation in values for the fine fraction seen between facilities within the United States. It is not clear as to why this variation exists. Figure 9.11 Bottom Ash Percent Fines as a Function of Time Percent

Passing,

%

..... H o u r l y

0

--

0

i

2

Average

Overall A v e r a g e

t.

P

~

~

L

_l

J

4

6

8

10

12

14

16

Sampling

Day

18

20

The vertical bars are the 95% confidence intervals After Eighmy et al., 1992 Abrasion Resistance The Los Angeles abrasion test measures the ability of an aggregate material to maintain its physical integrity under defined abrasive conditions. The test is conducted on two different size fractions, a coarse and a fine fraction, termed "B" and "C", respectively. The test is considered to be highly aggressive with respect to evaluating lightweight porous aggregate materials. Table 9.12 provides data on LA abrasion resistances for bottom ash B and C fractions from facilities from the United States. Typical percent losses observed for both fractions are around 40 to 45%. These values are considered to be high, however they are typical for porous lightweight aggregate materials.

362 Table 9.12 Bottom Ash Durability Country Facility

United States

Concord, NHc Dry Scrubber I d Dry Scrubber 3 d

Soundness % Lossa Fine Coarse Min Max Mean Median n Min Max Mean Median 2.63 10.38 14.32 11.91 11.48 4 2.51 2.76 2.63 . . . . 1.7 3.4 2.7 4 2.9 1.1 2.4 1.6 5 1.6 4.0

Country Facility n

" b

Concord, NH c Dry Scrubber 14 Dry Scrubber 24 Dry Scrubber34 Dry Scrubber34 ASTM C88. ASTM C131.

5

LA Abrasion, % Loss b C Fraction

B Fraction United States

n 4

Min

Max Mean Median

4 46.4 48.2 - 57.8 58.5

47.3 58.2

. . . . 5 57.6 61.0 - 46.0 61.8

. 59.6 54.9 c d

47.3 -

n

Min Max Mean Median

n Reference

2 4

42.6 44.2 43.4 40.4 40.7 40.6

43.4 -

2 E i g h m y e t al., 1992 2 LIRPB, 1992a

5

43.1 46.1 44.6 44.9 48.3 46.6

-

2 5

6

41.3 47.5 45.0

-

6

Original fraction less than 19 mm. Original fraction less than 50.8 mm.

9.1.5 Geotechnical Properties Many of the utilisation scenarios envisioned for bottom ash involve the use of bottom ash as an aggregate substitute subjected to compaction. Despite the potential problems of durability with bottom ash, bottom ash is a highly compactible material that upon compaction has high levels of E-modulus and strength. The fine fraction in bottom ash, the retentive capacity of bottom ash for holding water and the porosity of bottom ash mean that careful attention must be given to bottom ash water content prior to compaction.

Proctor Compaction The compactibility of a granular material is frequently assessed in the laboratory using Proctor compaction testing. Figure 9.12 shows a typical Proctor compaction curve with an optimum moisture content at maximum dry density. Table 9.13 provides data on bottom ash Proctor moisture optimums as well as maximum Proctor densities. Data are provided from facilities from the Netherlands, Sweden and the United States. As can be seen in the table, mean Proctor densities range from 1530 to 1739 kg/m 3. The minimum and maximum values that are observed are 1242 to 1838 kg/m 3 for all measurements. Good agreement is seen between facilities from the different countries. The geotechnical moisture content, at which maximum compaction occurs, tends to range from about 9.6% to 20%, with typical mean values of 13 to 16%. These moisture optimums are similar to those seen for gravelly sands in their ability to allow compaction to occur. There is good agreement between facilities from different countries with regard to moisture optimums that are observed.

363 Figure 9.12 Bottom Ash Proctor Compaction Curve Dry Density,

1.9

1000 k g / m 3

Typical Ash Zero Air Voids

1.6

--

8

I

I

,I

12

16

20

Moisture

After Eighmy et al., 1992

Content, %

Table 9.13 Bottom Ash Proctor Moisture and Proctor Density Compaction Country

Facility

Proctor Density (kg/m 3) Min

Max

Proctor Moisture (% W C )

Mean Median n Min Max Mean Median

Reference n

Netherlands AVI 1"

1,513 1,665

1,602

-

29 11.9 16.5

13.3

-

AVI 2"

1,543 1,630

1,573

-

26 10.9 16.0

13.0

-

16

AVl 3'

1,445 1,620

1,530

-

26 10.6 18.7

14.2

-

26

20 9.6 16.5 1,825 1 1,748 18 12.0 16.0 1,345 1 2 15.0 17.0 5 14.3 14.8 5 20.8 21.7

12.9 15.5

..... -

20 1 Hartl6n & Ro~lbeck, 1989

15.4 -

16.0 12.8

18 Eighmy et al., 1992 1 LIRPB, 1992a

16.0 14.6

-

21.3 -

16.6

Sweden United States

AVl 4" 1,475 1,630 1,530 Maim6 ~ ......... Concord, NH c 1,619'""'1,838 ""'1,739 Dry Scrubber 1d Dry Scrubber 2 d 1,242 1,298

1,271

Dry Scrubber 3 d 1,500 1,588

1,545

Dry Scrubber 3 d 1,550 1,580

1,566

Mass burn d

Standard Proctor. Standard Proctor.

-

-

-

c d

2,463

1

-

-

29 TAUW, 1988

2 5 5 1 Kosson et al., 1992

ASTM D1557 (Modified Proctor). ASTM D1557 (Modified Proctor).

The E-modulus can be evaluated for bottom ash materials as a function of Proctor compaction. Work by Hartl~n and Elander (1986) shows very high E-modulus values for ashes compacted in the fresh form and aged. The E-modulus values that are observed indicate that bottom ash is a strong aggregate material when it is in a compacted state.

364

Field Compaction Field compaction can be an aggressive, energetic process. Frequently, the compaction methods that are used in the field can break down particles in bottom ashes. Figure 9.13 shows how bottom ash particle size distributions become finer after field compaction efforts. The data, provided by Hartl~n and Rogbeck (1989), show that most full scale field compactors will fracture and fragment bottom ash and change the grain size distribution. At this time it is not clear if the degree of grain size redistribution is problematic with respect to civil engineering structural fill applications. Many bottom ash utilisation studies have shown that bottom ash can be successfully used as a compacted aggregate material in road sub-bases or in wind barriers and embankments. However, the fines content may require control as this influences frost susceptibility in cold climate applications. Figure 9.13 Bottom Ash Particle Size Distribution after Field Compaction Siltl

"

I

sand

Gravel

.... I

Silt

-,.., r r i._ C~.

100

0.06

0.2

0.6

2

90

6

20

./;y

/M

80

70 V e-

/

50

*~9

so

9

70

0

10

~

0 0.063

1

).2

0.6

4

Mesh Opening, mm

2

6

20

60

/,Y 7

30

S

20 10

0.5

Gravel

/,#"

*-~11 0.25

I

,/

-~

20

San~

1 0 0 0.06

~

0 r

~>

I

Grain Size d. in mm

Grain Size d. in mm

r

_

1

16

4

16

63

Mesh Opening, mm

2 months old

1 year old

The shaded area represents the original grain size distribution The solid line (A) shows the new grain size After Hartldn and Rogbeck, 1991 California Bearing Ratio (CBR) The CBR is a determination of the strength and stability of a compacted material. Values greater than 100% for CBR are seen in bottom ash. Table 9.14 provides information on CBR values at 0.1 inches and 0.2 inches for bottom ash samples obtained from facilities in either the Netherlands or the United States. Mean values seen at 0.1 inches range from 51.8 to 79.7% with minimum values at 22 and maximum values of 112.5 for all measurements. The CBR at 0.2 inches is higher. Typical mean

365 values are 39.0 to 154.5%, with minimum and maximum values ranging from 32.0 to 167.3% for all measurements. There is not good agreement between the data from different countries, but it is not clear why. Table 9.14 Bottom Ash Penetration Resistance Country

Facility

CBR @ 0.1 Inches a

CBR @ 0.2 Inches a

Reference

Min

Max

Mean Median n

Min

Max

Mean Median n

Netherlands AVI 1

24.0

65.0

52.0

-

29

32.0

76.0

62.4

-

AVI 2

38.0

62.0

51.8

-

26

52.0

74.0

61.2

-

26

AVI3

22.0

42.0

31.7

-

26 28.0

50.0

39.0

-

26

46.0

34.1

42.1

AVI4

27.0

United

Concord, NH

63.0 112.5

States

Dry Scrubber 1

.

.

.

.

1

Dry S c r u b b e r 2

.

.

.

.

2

121.0 158.7 139.9

-

2

Dry Scrubber 3

.

.

.

.

5 122.7 167.3 154.5

-

5

Dry Scrubber 3

.

.

.

.

5

-

5

79.7

80.0

20

34.0

58.0

20

92.0

136.5 110.2

29 T A U W , 1988

-

-

38.7

-

126.0

20 107.5 2 0 E i g h m y e t a l , 1 9 9 2 146.0

90.1

1 LIRPB, 1992a

Figure 9.14 shows how CBR varies as a function of time at the Concord facility. The data show that CBR exhibits some variability as a function of time. Figure 9.14 Bottom Ash CBR as a Function of Time CBR, %

CBR

a t 2.54

mm Penetration

CBR '

9O

130 1 120

a t 5.08 m m P e n e t r a t i o n

@~

f" ...........

.

ll0t. -

--

70

i

100 I

Average 50

i

~ _ L

0

2

4

6

L

8

i

..... 95% C. I.

10

L___

12

S a m p l i n g Day

After Eighmy et al., 1992

~I___~L 14 16

i 18

80

70

I

0

Average

..... 05% C I

L

t

i

I

I

I

2

4

6

8

10

12

q9a m p l i n g

Day

14

16

18

,

366

9.1.6 Permeability The permeability of bottom ash, or its ability to transmit water via percolation, is an important component with regards to characterising the hydraulic regime to which bottom ash can be subjected. Because bottom ash is a well-graded material and can be compacted to high densities, it is expected that under compactive efforts the permeability of bottom ash will be quite low. Frequently, an assessment of the permeability of bottom ash is needed to model leaching of bottom ash, to model water balances of water moving through bottom ash and to assess the ability of bottom ash to freely drain. Utilising permeability testing apparatus, permeabilities that have been observed in bottom ash are usually in the low 106 cm per second range. Table 9.15 shows some bottom ash permeabilities obtained in studies conducted in both Denmark and Sweden. At maximum density, it appears that bottom ash permeability can range from about 0.2 to 10.0 x 106 cm/s. Such permeabilities are considered to be relatively low for wellgraded materials and reflect the presence of fine material which increases the tortuosity within bottom ash. Such low permeability values suggest that bottom ash may be subject to some infiltration but could also create some surface runoff. Table 9.15 Bottom Ash Permeability Permeability 106 cm/s

Reference

-

3.5-4.4

Geoteknisk Institute, 1992

Malm5

0.2-10.0

Hartl~n & Elander, 1986

Country

Facility

Denmark Sweden

9.1.7 Influence of Combustor Type and Operation on Physical Characteristics There has not been a great deal of study conducted on the influence of combustor type and combustor operation on the physical properties of bottom ash. The comprehensive N ITEP program was the only large scale study that has been conducted to date that has looked at the influence of poor combustor operation and combustor type on ash characteristics. The only data that is available at this time is information on the loss on ignition content for a variety of facilities operated under both good and bad conditions. As shown in Figure 9.15, two-stage systems operated either under good or bad conditions produce bottom ash with significantly higher LOI values than mass burn or RDF systems.

367 Figure 9.15 Influence of Combustor Type on Bottom Ash LOI LOI

30-

r... o ~

20-

tO O) O) 0 _J

10-

0 -

[; 2_Stage

I I

Mass_Burn

!

I

RDF

9.1.8 Influence of Aging on Bottom Ash Physical Characteristics There have been some studies conducted in Sweden, Germany and the U.S. on the influence of aging on certain physical characteristics of bottom ash. In Sweden, it has been shown by Hartl~n and Rogbeck (1989) that the E-modulus of bottom ash will increase as bottom ash ages over time when compacted at optimum moisture under Proctor compaction testing. This increase in strength is attributable to the formation of mineralogical phases that increase particle interlocking within the bottom ash. Additional studies on aging in Sweden from bottom ashes at the Malmo facility have shown that when ash is aged for almost a year, the Proctor compaction characteristics are much better than when ash is freshly collected (Hartl~n and Elander, 1986). Again, aging is thought to increase the formation of certain mineralogical phases that increase the durability of the residue and interlocking characteristics of the particles in the residue. Studies in Sweden have also evaluated the influence of aging on the gross composition gradation of bottom ashes generated from the Malmo facility. There do not appear to be any significant differences between the nonmagnetic fraction, the glass fraction, the ceramic material, stone material and organic material in either fresh or aged fractions (Hartl~n and Lundgren, 1992).

368 Studies have been conducted in Germany to look at the influence of aging on a number of civil engineering properties. The data, presented by Vehlow (1992), show the influence of aging on leachable solids in bottom ash as bottom ash ages. The concentration of leachable solids in bottom ash decreases during aging. This is particularly true for facilities A and C (see Table 9.1). Facility B did not exhibit the same trends. The susceptibility to freeze/thaw fracturing has also been evaluated for aged materials in German facilities and the data suggests that the susceptibility to freeze/thaw erosion decreases with aging. This is particularly true again for facilities A and C. The data from facility B does not support this observation. Also evaluated in the German study was the raw density of aged material compared to fresh material at the three facilities. In all cases the raw density tended to increase with aging. This phenomenon is not presently understood.

9.2 PARTICLE MORPHOLOGY, MINERALOGY, AND ALKALINITY OF BOTTOM ASH Particle morphology, mineralogy and alkalinity of bottom ash play important roles in both the physical and chemical characteristics of bottom ash. The particle morphology of bottom ash is an important component in its physical characteristics and performance because of the angular nature of bottom ash particles. Bottom ash also tends to be a rough-textured material and this is an important property with regards to its physical performance. The mineralogy of bottom ash is thought to be important to understanding the leaching behaviour of bottom ash; however, the mineralogy also plays an important role in the compactibility and strength development of bottom ash as it ages with time. Finally, buffer capacity is an important component for both physical and chemical performance because of the role of the carbonate buffer system in bottom ash and the influence that has on strength development, particle aging and leaching.

9.2.1 Morphology Figure 9.16 provides some scanning electron microscopy micrographs and petrographic thin section micrographs of bottom ash particles. As can be seen in the SEM micrographs, bottom ash is an angular material. Slag-like material can be seen; the slag material is porous and contains vesicles. The petrographic thin section of bottom ash clearly shows that bottom ash possesses a high degree of internal porosity that is connected to the exterior of the particles. This vesicle porosity provides a large surface area for chemical reactions to take place and for leaching phenomenon to occur. As shown in Table 9.16, a number of researchers have looked at the specific surface area of bottom ash using BET absorption isotherms. Typically, bottom ashes have surface areas of 3 to 46 m2/g dry weight of bottom ash. These are considered to be very high degrees of surface area for a granular material. For instance, traditional

369 soils have surface areas that are orders of magnitude less than that. Mercury porosimetry is also used to look at internal pore diameters as can be seen in Table 9.16. A number of the pores in the material are considered to be quite small in nature. Values of less than a tenth of a micron are seen. The data also suggest that combustor type has an influence on the type of surface area and pore diameter that is found in bottom ash. The reader should refer to Chapters 12 and 13 for further discussion on the role of surface area in chemical leaching phenomenae. Figure 9.16 SEM Micrographs (a,b) and Petrographic Thin Section Micrographs (c,d) of MSW Bottom Ash

370 Table 9.16 Bottom Ash Surface Area BET Country Facility Surface Area m2/g

Pore Diameter" IJm

Reference

RDF b 4.605 Gardner, 1991 PRF b 3.286 0.0947 Gardner, 1991 RDF b 9.469 0.0786 Gardner, 1991 RK-MB b 28.184 2.117 Gardner, 1991 Unknown 9.4-46.3 Theis & Gardner, 1990 MB b 3.2 0.0342 Kosson et al., 1992 Mercury porosimetry. RDF=Refuse-Derived Fuel, PRF=Processed Refuse-Derived Fuel, RK=Rotary Kiln, MB=Mass Burn

United States

In comparing the petrographic thin sections of bottom ash with the surface area and porosimetry data provided in Table 9.16, it is clear that the internal porosity that is connected to the exterior accounts for a great deal of the surface area measured by nitrogen BET absorption isotherms. Compared to most aggregates, bottom ash is considered to be a light-weight porous aggregate with more angularity and more surface roughness and textures than many traditional aggregates.

9.2.2 Mineralogy The mineralogical characteristics of bottom ash play a very important role in the leaching behaviour of bottom ash. It is estimated that bottom ash contains numerous mineral phases. Such diversity complicates our understanding of the leaching behaviour of bottom ash and more research needs to be conducted on the role of mineralogy in bottom ash aging and strength development. Nevertheless, there are at least four studies that have been conducted that have examined the mineralogy of bottom ash. The studies have been conducted by St~mpfli (1992), Vehlow et al. (1992), Kirby and Rimstidt (1993) and Eighmy et al. (1994). All four of these studies have used rigorous procedures employing x-ray powder diffraction (XRPD) and other methods as precise procedures for estimating the nature of the mineral phases. St~mpfli (1992) has examined bottom ash for the presence of those mineral phases that are associated with strength development as bottom ash ages. St~mpfli used XRPD to determine the mineralogy. Minerals identified include SiO2, CaCO3, Fe304, Fe203, Fe, FeO, Ca~,I(OH)7.6.5 I--120,Na2Si20~, and CaSO4. Others are shown in Table 9.17.

371 Table 9.17 Mineral Phases in Bottom Ash (in relative order of decreasing abundance) St~impfli (1992)b Vehlow et al., (1992)o Kirby and Rimstidt Eighmyet al. (1993)d (1994)e SiO2 CaCO3 Fe304

Fe304 SiO2 (Ca, Na)2(AI,Mg)(Si,AI)207

SiO2 CaSO4 o 2 H 2 0 3(AI203)~ TiO2

Fe203 Feo

CaCO3 KAISi308

Fe203 FeO

FeAI204 SiO2

FeO

NaAISi308 CaAI2Si208

CaSO4 KCI

CaG(PO4)2 Fe203

FeCr204 Ca(Mg,Fe)Si206 Fe2SiO4 Cr203 Fe203 CaMgSiO4

NaCI

CaSO4 CaO AI(OH)3 NaCI ZnCI2 NaAISi308

Ca2AI(OH)T.6.5H2 O Na2Si2Os CaSO4 (Ca, Na)(Al,Si)2Si8 NaAISi308

AI203 Ca(OH)2 CaSO4

b Basedon XRPD c Basedon petrographyand XRPD

Ca2AI2SiOT MgCa2Si207 Fe304

AI2SiOs TiO2 Based on XRPD Basedon petrography,XRPD,XPS, SEM/XRM

Vehlow et al. (1992) have conducted extensive characterisations of bottom ashes from three facilities in Germany. They used XRPD and petrography. Data are presented in Table 9.17. The principle phases found in ash from the German facilities are glass, magnetite, quartz, melilite and feldspar. A number of other minor phases were also identified. Agreement was seen in the relative presence of major phases amongst the three facilities. Vehlow et al. (1992) also looked at aging effects. Kirby and Rimstidt (1993) studied bottom ashes containing small quantities of fly ash. They used XRPD as well. Principle minerals include (% abundance) Fe203 (3.7%), CaCO3 (3.5%), NaCI (0.5%), SiO2 (2.3%), magnetic spinel (3.5%), TiO2 (1.1%), and CaSO4 o2H20 (1.8%). The majority of the non-LOI mass of the sample was amorphous glass and minerals present below the detection limit for XRPD. Table 9.17 provides further information. Eighmy et al. (1992) examined the characteristics of bottom ash using XRPD, petrography, SEM/XRM and surface microanalytical techniques for samples from a U.S.

372 facility. The bottom ashes were ground and separated using magnetic and density gradient separation procedures. Table 9.17 provides a summary of the data. Many of the phases found in the U.S. bottom ashes are similar to the ones identified by the other studies.

9.2.3 Alkalinity The buffer capacity of bottom ash is an important component in the leaching characteristics of bottom ashes. The acid neutralising capacity of the residue is a measure of how many milliequivalents of nitric acid are required to reduce the pH of one gram of residue to a value of 4.3. The endpoint of the titration can vary. Some researchers use a value of 7.0, while others use the more traditional carbonate alkalinity endpoint. To put this measure into perspective, one gram of residue would need to be leached with 45 litres of acidic precipitation to reduce the pH from 12.0 to 7.0. Table 9.18 provides some information on the acid neutralising capacity as well as the initial pH or the inherent pH of bottom ashes for samples collected from Canadian and U.S. facilities. Typically, bottom ash has an initial pH ranging from 10.5 to about 12.2. This is in part due to the presence of calcium hydroxides produced from CaO hydrolysis in the bottom ash. The acid neutralising capacity of bottom ash ranges from about 1.2 to 4.1 milliequivalents per gram. This means that bottom ash is reasonably well-buffered. Such buffering capacity indicates that bottom ash can moderately resist changes in pH. Table 9.18 Bottom Ash pH and Acid Neutralising Capacity Country

Facility

Canada

LVH SWARU QUC

Initial pH

.

ANC, meq/g Reference

10.20

3.05

Sawell et al., 1989b

-

4.11

Sawell et al., 1989a

11.39

2.15

Sawell and Constable, 1988

United States Concord, NH 10.5-12.2

1.2-3.0

Eighmy et al., 1992

Figure 9.17 provides a typical titration curve for bottom ash. The data indicate that there are a number of locations in the titration curve where a slight degree of buffering takes place. These buffers tend to occur at a pH of around 10, 8 and 5. Such locations for buffers are attributable to the carbonate system. Figure 9.18 shows the change of botom ash acid neutralising capacity as a function of time for bottom ashes collected from the Concord facility (Eighmy et al. 1992). The acid neutralising capacity is relatively variable over time.

373 Figure 9.17 Bottom Ash Titration Curve I0.000 8.000

Ave A N C - 2.1 m e q / g

6.000 "i-

4.000 2.000 0.000

. . . .

I

i

J

,

1.0

0.0

9

i

. . . .

!

2.0

. . . .

!

3.0

. . . .

I

4.0

. . . .

5.0

!

. . . .

!

6.0

. . . .

i

7.0

. . . .

8.0

m e q / g dry bottom esh After Eighmy et al., 1992 Figure 9.18 Bottom Ash ANC as a Function of Time 4.5 0

First Hour 9 Second Hour Z~ Third Hour A Fourth Hour 9 Daily Composite i l l

4.0 ol O" Q)

3.5 9

E 3.0

0 z <:

9

/

2.5 2.0

01 o L. 1.5 1.0

l_ l

/\J

9

0.5 0.0

0

I

I

I

!

|

I

I

!

I

n

50

1O0

150

200

250

300

350

400

4.50

500

ELAPSED TIME (DAYS) After Eighmy et al., 1992

550

9.0

374

9.2.4 Influence of Combustor Type and Operation on Bottom Ash Surface Area, Mineralogy and Alkalinity There is not a great deal of information available on the influence of combustor type and operation on the bottom ash particle morphology and mineralogy. Some of the data presented in Table 9.16 on bottom ash surface area suggests that newer mass bum facilities that bum at higher temperatures produce bottom ashes of higher relative surface areas. This may be a function of the quenching system and the degree of gas trapping that occurs when bottom ash slag is rapidly cooled. As shown in Figure 9.19, two-stage bottom ash has less soluble mineral phases than mass bum or RDF systems. Using total dissolved solids as a measure of the presence of soluble mineral phases, two-stage systems are lower. This is believed to be because of poorer burnout and less complete oxidation of organics in the system. As shown in Figure 9.20, the pH of two-stage bottom ash is generally lower than that of mass burn or RDF systems. This is believed to be due to the less efficient degree of oxidation of organic matter, the relatively low amounts of metal oxides and the reduced likelihood of forming alkaline hydroxides from oxide phases.

9.2.5 Influence of Aging on Bottom Ash Surface Area, Mineralogy and Alkalinity At present there are four studies that have evaluated the influence of aging on bottom ash mineralogy. The first study is by Stampfli (1992). As shown in Figure 9.21, x-ray powder diffraction traces of fresh and aged ashes from the Horgen facility in Switzerland suggest that some transformations occur during a four-month aging process. St~mpfli notes that certain mineral phases appeared in the aged residues. These include the minerals CaSO4 ~ (gypsum), Ca6AI2(SO4)3(OH)12 ~ (ettringite), Na2Si2Osand Ca2AI(OH)z .6.5 !-120. The former two are related to strength development. The work by Vehlow et al. (1992) suggests that aging does not drastically transform major mineral phases in bottom ash. The aging phenomenon was examined at threeand six-month increments. The data suggests that some mineral phases disappear with time; however, these mineral phases tend to be minor phases in the residues. The phases that tend to disappear with aging appear to be the phases FeCr204 (chromite), Ca(Mg,Fe)Si206 (diopside) and possibly Ca(Mg,Fe,AI)(Si,AI)206 (augite). Zevenbergen et al. (1993) have identified a clay-like structure that forms when glassy phases in bottom ash are weathered. An illite-like 10 ~, basal spacing in a weathered rim region was observed. The presence of such a mineral can increase the ability of bottom ash to retain metals that can ion-exchange onto clay surfaces.

375 Figure 9.19 Bottom Ash Total Dissolveable Solids as a Function of Combustor Type TDS ~5-

lO-

O

I

..... I O

5

i l

I

i

o

2_Stage

Mass_BuPn

Mass_Burn_GS

RDF

Figure 9.20 Bottom Ash pH as a Function of Combustor Type pH 13-

m~

9-

7-

I 2_Stage

Mass_Burn

RDF

376 Figure 9.21 XRPD Diffractograms of Changes in Mineral Phases in Bottom Ash as a Function of Aging

1200

--

,

1200

J

i

|

Ash

Fresh

Aged Ash 1000 -

8OO

800

-

o o.

600

600 -

8 4O0

400-

200

10

20

30

40 Diffraction

50 Ingle

60 {29}

,

i

,

70

80

90

-

O-

|

,

,

i

i

10

20

30

40

50

Diffraction

1 angle

60 (2~l

i

i

i

70

80

90

After St~mpfli, 1992 Kluge et al. (1981) documented hydrogen gas evolution from road sub-bases made of bottom ash. The bottom ash aged in situ. During aging, hydrogen evolution caused cracking and blistering of the standard asphalt wearing course. Gas analyses obtained from blisters revealed hydrogen gas. The authors speculated that the following redox reaction was occurring: 2 A 1 ~ + 6 H 2 0 -- 2 A I ( O H ) 3 ( s )

+ 3 H 2(g)

(9.1)

They hypothesise that other metallic components with valence states of zero (copper, zinc) could also promote similar redox reactions that promote hydrogen gas evolution. 9.3 INORGANIC CHARACTERISTICS OF BOTTOM ASH Elemental distribution in bottom ash plays a very important role in the chemical behaviour of bottom ash. The chemical makeup of bottom ash also plays an important role in the physical and mineralogical characteristics.

377 9.3.1 Elements Present in Bottom Ash

The relative partitioning of elements into bottom ash is dependent on the chemical makeup of the MSW feed to the incinerator, the volatility of the element, the type of incinerator and grate system used to combust the waste, and the operation of the combustion system. These processes are presented in Chapters 3 and 8. Typically, over fifty elements can be identified in bottom ash using total compositional analytical methods. Some elements are present as major constituents (>10,000 mg/kg), some are present as minor constituents (>1,000 but <10,000 mg/kg) and most are present as trace constituents (<1,000 mg/kg). As shown in Table 9.19, when the entire range of concentration of elements in bottom ash is compared to typical constituents in either the lithosphere or in soils, it is apparent that some of the major matrix elements in bottom ash are very similar. The data provided in Table 9.19 (adapted from Lindsey, 1979) show that a number of the trace constituents in bottom ash are relatively enriched. Nevertheless, the composition of elements in bottom ash is for the most part comparable to those materials found in the lithosphere or soils. Table 9.1 provides information on the incineration facilities within the various countries that generated the bottom ash that was evaluated in this section of the chapter. Many of the samples that were collected in each of the studies contain either grate siftings or economiser ash; however, the data that is to be presented here suggests that there is still similarity in the composition of bottom ash within each country, particularly for less volatile, lithophilic elements. The composition of bottom ash has been broken down into various defining categories. The categories include major matrix elements, minor matrix elements, other minor elements, trace elements including oxyanionic elements, other trace elements, elements related to biogeochemical cycling and exotic elements, lanthanides and actinides. The designation of major, minor and trace descriptors is based on whether or not the concentration of the element is typically present at greater than 10,000 mg/kg (major), at concentrations ranging between 1,000 to 10,000 mg/kg (minor), and elements present at concentrations less than 1,000 mg/kg (trace). These designations by concentration are somewhat arbitrary but useful. It should also be noted here that all of the information provided on element concentration is based on the analysis of bottom ash using total quantification techniques. Such techniques include the use of neutron activation analysis or total digestion, followed by some spectrophotometric quantification. Therefore, the data available world-wide on bottom ash is rather limited because not all researchers have used total compositional analysis. In this chapter, it was felt that it was very important to only look at total composition. Many of the other analytical techniques that are based on partial analysis tend to recover only 50 or 60% of the element in the material and are not considered to be total compositional analyses.

378

Table 9.19 L i t h o s p h e r e , Soil a n d Bottom A s h C o m p o s i t i o n Element Content in C o m m o nRange Ave for Soils Lithosphere (ppm)' for Soils (ppm)" (ppm)" Ag AI As Au B Ba Be Br C Ca Cd CI Co Cr Cs Cu F Fe Ga Ge Hg I K La Li Mg Mn Mo N Na Ni O P Pb Rb S Sc Sb Se Si Sn Sr Ti V Y Zn Zr Lindsay, 1979

0.007 81,000 5 10 430 2.8 2.5 950 36,000 0.2 500 40 200 3.2 70 625 51,000 15 7 0.1 0.3 26,000 18 65 21,000 900 2.30 28,000 100 465,000 1,200 16 280 600 5 0.09 276,000 40 150 6,000 150 80 220

0.001-5 10,000-300,000 1-50 2-100 100-3,000 0.1-40 1-10 7,000-500,000 0.01-0.70 20-900 1-40 1-1,000 0.3-25 2-100 10-4,000 7,000-550,000 5-70 1-50 0.01-0.3 0.1-40 400-30,000 1-5,000 5-200 600-6,000 20-3,000 0.2-5 200-4,000 750-7,500 5-500 200-5,000 2-200 50-500 30-10,000 5-50 0.1-2 230,000-350,000 2-200 50-1,000 1,000-10,000 20-500 25-250 10-300 60-2,000

0.005 71,000 5 10 430 6 5 20,000 13,700 0.06 100 8 100 6 30 200 38,000 14 1 0.03 5 8,300 30 20 5,000 600 2 1,400 6,300 40 490,000 600 10 10 700 7 0.3 320,000 10 200 4,000 100 50 50 300

Range in Bottom Ash World-Wide 0.29-36.9 21,900-72,800 0.12-189 <0.20 38-510 400-3,000 1.4-150.2 10,000-60,000 370-123,000 0.3-70.5 800-4,190 6-350 23-3,170 1.0-2.0 190-8,240 200-1,100 4,120-150,000 10 0.02-7.75 2-10 750-16,000 2-20 400-26,000 83-2,400 2.5-276 110-900 2,870-42,000 7-4,280 400,000-500,000 1,400-6,400 98-13,700 40-50 1,000-5,000 3-6 10-432 0.05-10.0 91,000-308,000 2-380 85-1,000 2,600-9,500 20-122 10 613-7,770 200

379

9.3.2 Major Matrix Elements (> 10,000 mglkg): O,Si,Fe,Ca,AI,Na, K,C Figure 9.22 is a grouping of box plots depicting the concentration of the major matrix elements in bottom ash by country. The box plot depicts all of the data that is available for each of the countries. The central box for each of the elements in each plot extends from the first quartile to the third quartile, with a horizontal line to indicate the median value. The first quartile denotes the twenty-fifth percentile, the median denotes the fiftieth percentile and the third quartile denotes the seventy-fifth percentile. The height of the box equals the interquartile range. Lines are sometimes drawn out from the quartiles to adjacent values, defined as those data points less than 1.5 times the interquartile range beyond the first or third quartiles. Values more than 1.5 times the interquartile range are considered to be outliers and are denoted by individual circles in each of the plots. The use of box plots allows the reader to compare the databases that are available for each of the countries. In this chapter, the interquartile range is used for comparative purposes. As can be seen in many of the plots presented in this section, many of the data have outliers that denote a positive skewness in the data. Such skewness means the data are not normally distributed. The use of box plots is therefore the most appropriate measure for comparing populations of data that are not normally distributed. Oxygen is the most prevalent element in bottom ash. Its concentration is estimated to be between 300,000 to 500,000 mg/kg. It is present in bottom ash as oxides of silicon, iron, calcium, aluminum and carbon. Bottom ash concentrations are found to be similar to oxygen concentrations in either lithosphere or soil materials (see Table 9.19). Silicon is present in bottom ash in concentrations ranging from 91,000 to 308,000 mg/kg. The data presented in Figure 9.22 show that there is a wide degree of variability in silicon concentration in the United States. Data from Switzerland, Denmark and the Netherlands are more closely grouped together at higher concentrations. Median values tend to be fairly close for each of these populations, with the United States, Switzerland, Denmark and the Netherlands having median values of 200,600, 237,000, 270,000 and 196,000 mg/kg respectively. The silicon concentration is similar to those observed in lithospheric and soil materials. Iron is the next most abundant element in bottom ash. Typical values seen for bottom ash range from 4,120 to 150,000 mg/kg. Such concentrations are also similar to those found in soils and lithospheric materials. The range of data presented in the United States is found to be more variable. Data presented for Canada and the European countries tend to fall within the range presented for the United States. Median values by country are 79,000, 14,700, 73,000, 27,000, 28,150 and 42,400 mg/kg for the U.S., Canada, Switzerland, Denmark, Sweden and the Netherlands respectively. Calcium is the next most abundant element in bottom ash. It is typically found at concentrations of 370 to 123,000 mg/kg. Such concentrations are similar to those observed in lithospheric and soil materials. As can be seen in Figure 9.22, median

Figure 9.22 Major Matrix Element Distribution by Country

te:

;

t :::

11.1

lOO~L0

." aoooo

:60000

3 100000

y

:50000

.oooo

20000 0

S i l i c o n C o n c e n t r a t l o n I n Bottam Ash Dy Country

20000

I r a n C o n c e n t r a t i o n I n Bottom Ash D y Country

0

A l u m ~ n u m C a n c e n t r a t l o n I n Bottom Ash O v Country

Country

8 0

Calc~urn C o n c e n t r a t ~ o n I n Bottom Ash b y Country

0

Soalum C o n c e n t r a t ~ o nI n Bottom Asn ~y

-

j

PotaSS~umC o n c e n t r a t l o n In B o t t o m Asn b y Country

The keys for each plot read left to right by line and correspond to the boxes moving left to right

381 values for the United States, Canada, Switzerland, Denmark and the Netherlands are 77,700, 39,400, 103,000, 60,500, and 82,900 mg/kg respectively. Aluminum is the next most abundant element in bottom ash. It is found in concentrations ranging from 21,900 to 72,800 mg/kg. It is similar in concentration to those seen in lithospheric and soil materials. Median values seen for the United States, Canada, Switzerland, Denmark and the Netherlands are 51,100, 45,000, 58,000, 49,000 and 32,900 mg/kg respectively. Sodium is the next most abundant element in bottom ash. It is found in concentrations ranging from 2,870 to 42,000 mg/kg. Like many of the other major elements in bottom ash, its concentrations are also similar to those seen in lithospheric and soil materials. Median values observed for the United States, Canada, Switzerland, Denmark and the Netherlands are 34,300, 17,850, 36,000, 27,000, and 30,740 mg/kg respectively. The next most abundant element in bottom ash is inorganic carbon. There is not a great deal of information available on the concentration of inorganic carbon in bottom ash. Data to date suggest that concentrations range between 20,000 to 40,000 mg/kg. The last major matrix element that is present in bottom ash is potassium. Potassium is present in concentrations ranging from 750 to 16,000 mg/kg. Such concentrations are very similar to those seen in lithospheric and soil materials. Median values observed for the United States, Canada, Switzerland, Denmark, Sweden and the Netherlands are 9,000, 11,800, 13,000, 11,000, 1,685 and 9,880 mg/kg respectively. As discussed in Chapter 8, many of these elements exhibit very low vapour pressures and are therefore not subjected to variations in combustion temperature. However, there is the possibility that the composition of the MSW feed going into the incinerator could alter the concentrations seen in bottom ash. As an example, Figure 9.23 is a plot of aluminum concentration in bottom ash for the various facilities in the United States. These facilities include the Concord facility in New Hampshire, the Dry Scrubber 1, Dry Scrubber 2 and Dry Scrubber 3 facilities, the EPA facility and the mid-Connecticut facility. As can be seen in Figure 9.23, the median values observed for aluminum in bottom ash from each of these facilities tends to be fairly close, ranging from 48,100 to 53,750 mg/kg. Such data suggest that for aluminum at least, there seems to be reasonably good agreement between different types of facilities with different types of MSW feed material for data presented for the United States. Figure 9.24 shows how a major matrix element (aluminum) varies in bottom ash as a function of time. The data is presented for the Concord facility. Over the 1.5 year sampling time frame, the 18 sampling events show that the variation in total aluminum concentration is usually less than the variation seen within four consecutive hourly composites from any one sampling day. This data suggests that aluminum is a very conservative element in bottom ash. Most of the other major matrix elements exhibit similar behaviours.

382 Figure 9.23 Major Matrix Element Distribution (Aluminum) within United States Facilities

con E~ ds3

~ dsl E~ epa

~ ds2 ~ mid

80000 -

[D3 C3~

6 0 0 0 0

E

-

E

0 -r-I -l-J

C

(D

40000 -

U cO U

20000

-

Aluminum Concentpations in Bottom Ash From U.S.-Facilities

Facility

n

rain

max

mean

rnedian

con

23

34,400

79,500

52,634

53,700

dsl

6

21,900

51,800

42,000

45,050

ds2

8

43,300

72,800

54,187

53,750

ds3

6

44,500

50,700

47,800

48,100

epa

12

49,000

54,400

51,208

51,200

midconn

3

50,000

56,000

53,000

53,000

383 Figure 9.24 Major Matrix Element Concentration (Aluminum) as a Function of Time 8.0E4 ~7.0E4

I

E

v6.0E4

I

I

...;I I

Z 5.0E4 I!

-n <: 4.0E4 -

II

<: F-- 5.0E4 o 2.0E4

,

,

0

,

,

J

5

.

.

.

.

0 First Hour Composite 9 Second Hour Composite /k Third Hour Composite A Fourth Hour Composite n~ll Daily Composite I

,

,

i

i

10

i

15

/

,

,

20

SAMPLING EVENT After Eighmy et al., 1992

9.3.3 Minor Matrix Elements (1,000 to 10,000 mglkg)" Mg, Ti, CI, Mn, Ba Figure 9.25 presents information on the minor matrix element distribution in bottom ash by country. Magnesium is the most prevalent minor matrix element in bottom ash. It ranges from 400 to 26,000 mg/kg. This range is similar to those seen in lithospheric and soil materials. Median values for the United States, Canada, Switzerland, Denmark and the Netherlands are 9,450, 4,850, 16,000, 8,400 and 11,400 mg/kg respectively. As can be seen in Figure 9.25, the magnesium concentration in bottom ashes by country does not exhibit variability. The variability seen within each of the countries is similar in the countries presented. The next most abundant element is titanium. Titanium is present in concentrations ranging from 3,000 to 9,500 mg/kg. These are similar to lithophilic soils. Median values for the United States, Canada, Switzerland, Denmark and the Netherlands are 6,500, 8,000, 5,500, 4,100 and 4,150 mg/kg respectively. Chlorine is the next most abundant minor element in bottom ash (Figure 9.25). Median values for the United States, Switzerland, Denmark and the Netherlands are 2,100, 3,500, 1,800 and 1,900 mg/kg respectively. These values are enriched relative to soils. Manganese is also a minor element in bottom ash. Data are reported for the United States, Canada, Denmark, Sweden and the Netherlands. Median values are all around 1,000 mg/kg. These are similar to soils.

Figure 9.25 Minor Matrix Element Distribution by Country

'

0

Magnesium Concentration i n Bottom

0

Asn b y Country

0

Manganese Concentratlon I n Bottom

T ~ t a n l u mConcentration i n Bottom

asn

by Country

a

Asn

b y Country

Barlum Concentratlon l n Bottom

Asn

by Country

ChlOrlne Concentrat~on ~n Bottom

~ s nby

Country

385 The next most abundant element in bottom ash in barium. Barium is present in concentrations ranging from 380 to 2,652 mg/kg. Such concentrations are similar to those seen for lithospheric or soil materials. The median values for the United States, Canada, Denmark and the Netherlands are 900, 600, 1400 and 1603 mg/kg respectively. As shown in Figure 9.25, the variability and range of data for each country are very similar. It is interesting to compare the concentrations of barium in bottom ash from facilities within a single country. As shown in Figure 9.26, the concentration of barium is depicted in facilities from the United States. Six facilities are shown. As can be seen, the barium concentrations in bottom ash is remarkably similar for each of the six facilities. It is also important to examine how variable the concentration of barium is within bottom ash collected at the same incineration facility over time. In Figure 9.27, the barium concentration in bottom ash is depicted for the Concord facility. The barium concentration over time tends to be as variable as the variation seen within four consecutive hourly sampling events from the one sampling day. This suggests that barium concentration can be somewhat variable in the process stream at the incinerator.

9.3.4 Other Minor Elements (1,000 to 10,000 mglkg): Zn, Cu, Pb, Cr Other minor elements in bottom ash are depicted in Figure 9.28. Zinc is a prevalent minor element in bottom ash. Its concentrations range from 613 to 7,770 mg/kg. Typically, it is enriched relative to lithosphere or soil materials. This is because zinc is widely used in manufactured goods. The median values seen for the United States, Canada, Germany, Denmark, Sweden and the Netherlands are 3,490, 2,420, 2,650, 2,200, 2,825 and 2,130 mg/kg respectively. As shown in Figure 9.28, the zinc concentration in bottom ash within the various countries is remarkably similar. Another minor element in bottom ash is copper. Its concentrations range from 290 to 8,240 mg/kg. Like zinc, it is enriched relative to concentrations seen in lithospheric or soil materials. Copper is also widely used in manufactured goods. Median values observed for the United States, Canada, Germany, Denmark, Sweden and the Netherlands are 1,880, 2,286, 1,500, 2,500, 1,900 and 1,687 mg/kg respectively. As with zinc, the copper concentration in bottom ash is similar amongst the countries that are shown in Figure 9.28. Another minor element is lead. Lead ranges in bottom ash from 98 to 13,700 mg/kg. It is very enriched relative to lithospheric and soil materials. Lead is also widely used in manufactured goods. It is a common soil contaminant near roadways. Median values for the United States, Canada, Germany, Denmark, Sweden and the

386 Figure 9.26 Minor Matrix Element Distribution (Barium) within United States Facilities

ET# con E~ ds3

@ dsl E~ epa

@ ds2

E~ mid

2000 -

o]

1500 -

I

c~ E

I

d

o -t-i

1000 -

ro

T .__--[-

c.

c (1) t] C o 0

500 -

_

Bapium C o n c e n t r a t i o n

in

Bottom

Ash From U.S. F a c i l i t i e s

Facility

n

min

max

mean

median

con dsl ds2 ds3 sav epa mid

32 6 9 6 1 1 3

570 417 383 481 830

1,680 1,620 1,030 1,220 1,300

981 1,156 719 832 1,060

905 1,195 754 818 1,500 520 1,050

387

Figure 9.27 Minor Matrix Element Concentration (Barium) as a Function of Time 2000 ',F u~

. . . . . . . . . . ._ . . . . . . . 0 First Hour Composite 9 Second Hour Composite A Third Hour Composite A Fourth Hour Composite

1 800 1 600

E

i

1400

~

iii2

T

I1--11

Daily Composite

/,/-.

1200

/m-m. m..m.m-m

cD 1000 <~ 9 F--

800

F-

II

600 400

0

....

, . . . . . . . . . 5 10

, 15

20

SAM P LI N O DATE After Eighmy et al., 1992 Figure 9.28 Other Minor Element Distribution by Country den

gooo

~,

,,we

net

rE)o

6000

o o ~,

6000

8 o

-II

~

~

I

I

o Zinc Concentration

15000

r

i n Bottom Ash by C o u n t r y

o

s,e

Lead C o n c e n t r a t i o n

o

3000

o o

;L o

o l•

i n Bottom Ash by C o u n t r y

Z'"

Sc'0," 8 o

1oooo

Copper C o n c e n t r a t i o n

4~

.J._

in Bottom Ash by C o u n t r y

tooo

,.,

~

o

o Chromium

Concentration

in Bottom Ash by Country

388 Netherlands are 2,400, 1,935, 1,250, 1,500, 780 and 1,295 mg/kg respectively. What is interesting to note in Figure 9.28 is that the median values for lead in bottom ash from each of the countries is very similar; however, almost all countries show outlier lead values that are high in concentration relative to the median values. Such outlier values are probably attributable to the analysis of small pieces of slag lead in the bottom ash. If the outliers are removed from the data bases, the data do show that their variation in lead concentrations in bottom ash are similar by country. Chromium is another minor element in bottom ash. It ranges from 23 to 3,170 mg/kg. It is only slightly enriched relative to lithospheric or soil materials. Median values for the United States, Canada, Denmark, Sweden and the Netherlands are 1,072, 806, 290, 245 and 57 mg/kg respectively. As shown in Figure 9.28, the chromium concentration in bottom ash between countries is fairly uniform. It is important again to evaluate how variable the other minor element concentrations can be in bottom ash from samples collected from facilities within the same country. As shown in Figure 9.29, the chromium concentration in bottom ashes from facilities in the United States exhibits a fair degree of variability. Such variability is probably attributable to the presence of automotive wastes in the MSW feed going to the incinerators. Facilities that utilise refuse processing tend to have less chromium in the bottom ash. It is also important to evaluate how variable the chromium concentration can be in bottom ash as a function of time. As shown in Figure 9.30, the chromium concentration in the bottom ash from the Concord facility is remarkably uniform as a function of time.

9.3.5 Other Trace Elements Including Oxyanionic Elements (<1,000 mg/kg): Sn, Sb, V, Mo, As, Se The other trace elements, including oxyanionic elements, have been grouped together largely because of their behaviour during leaching; these elements tend to leach as oxyanions. Figure 9.31 shows box plots for how these elements are distributed in bottom ash by country. Tin, which is not an oxyanion, ranges in concentration from 2 to 380 mg/kg. Its concentration is very similar to those seen in lithosphere or soil materials. Median values for the United States, Canada, Denmark and the Netherlands are 254, 10, 200 and 10 mg/kg respectively. As can be seen in Figure 9.31, the data for the United States and Denmark are remarkably similar in their distributions, but these two data populations differ from the data available for Canada and the Netherlands. Antimony is the most abundant trace oxyanionic element in bottom ash. It is present in concentrations ranging from 10 to 432 mg/kg. It is enriched in bottom ash relative to its concentration in the lithosphere or in soils. Median values for the United States,

389

Figure 9.29 Other Minor Element Distribution by Country @ con

@ ds~

E~ d s 3

2000

ds2

E~ mid

E~ e p a

o

-

1500

-

I000

-

-7

!

' _..L_ 500

-

EL~

0 -

Chromium C o n c e n t r a t i o n

in Bottom

Ash From U.S. F a c i l i t i e s

Facility

n

min

max

mean

median

Con

31

867

2,004

1,457

1,378

Dsl

6

185

358

277

298

ds2

9

23

205

130

153

ds3

6

245

329

295

297

epa

12

-

-

-

724

mid

3

1,070

1,260

1,183

1,220

Figure 9.30 Other Minor Element Concentration (Chromium) as a Function of Time 4000 CD

,,, v

E

3000 2500

TT

1500

0 .< I--o

1000 500 0

,

|

i

1, s

9

,

i~ . . . .

|

9

9

i ,.....

0 First Hour Composite II Second Hour Composite /k Third Hour Composite A Fourth Hour Composite m~m Doily Composite

3500

O~ 2000

.....J

m

. m_n-m-m._ n'|" m'm.m-mm ~/re.m-m-m m-m -m" m .

.

.

.

i

5

.

.

.

.

!

.

.

.

.

.

10

SAMPLINO EVENT After Eighmy et al., 1992

!

15

,

,

20

Figure 9.31 Trace Oxyanionic Element Distribution by Country

-

0

-TT

Tin Concentration lo Bottom Ash Dy Country

-

p

"

I00

200

a

Antimony Concentration

0;

in Bottom I s n Dy

Country

::::

-

0

0

0 0

*9Molyoaenvm Concentrat~on in Bottom Asn b y Country

Arsenlc Conceotrat~oni n B o t t o m Asn by Country

vanaalum Concentrat~on i n Bottom ASn ny country

391 Canada and the Netherlands are 125, 69 and 15 mg/kg respectively. As shown in Figure 9.31, there is some variability between countries in the concentrations of antimony present in bottom ash. Vanadium is the next most abundant trace oxyanionic element in bottom ash. It is found in concentrations ranging from 20 to 122 mg/kg. It is relatively difficult to recover this element from bottom ash. These concentrations tend to be lower than the concentrations found in the lithosphere or soils. Median values for the United States, Denmark and the Netherlands are 49, 68 and 34 mg/kg respectively. The concentrations of vanadium in bottom ash between countries is remarkably similar. The next most abundant trace oxyanionic element is molybdenum. It is found in concentrations in bottom ash ranging from 2.5 to 276 mg/kg. It is enriched in bottom ash relative to concentrations seen in the lithosphere or soils. Median values for the United States, Denmark and the Netherlands are 39.5, 27.5 and 17.0 mg/kg respectively. As shown in Figure 9.31, concentration of molybdenum in bottom ash between countries is very similar. The next most abundant trace oxyanionic element in bottom ash is arsenic. Typical concentrations range from 0.12 to 189 mg/kg. Such concentrations are slightly higher than those seen in the lithosphere or soils. Median values for the United States, Canada, Denmark, Sweden and the Netherlands are 25, 26.2, 21.5, 3.0 and 13.0 mg/kg respectively. The concentrations of arsenic in bottom ash are remarkably similar for the countries that are presented in the figure. The next most abundant trace oxyanionic element in bottom ash is selenium. Typically selenium range in bottom ash from 0.05 to 10.0 mg/kg. These concentrations tend to be slightly enriched relative to lithospheric or soil materials. Median values for the United States, Canada and the Netherlands are 1.44, 0.30 and 6.00 mg/kg respectively. As shown in Figure 9.32, arsenic distributions in bottom ashes from facilities from the United States are similar. Median values tend to range from 19.1 to 38.0 mg/kg. It is also important to understand how variable trace oxyanionic elements can be in bottom ash in process streams from an incinerator. As shown in Figure 9.33, arsenic concentrations in bottom ash are depicted for bottom ash produced at the Concord facility. The arsenic concentrations tend to exhibit some variation with time; the variation tends to be about the same as that seen in concentrations in four consecutive hourly samples from the one sampling day. The samples were collected over a 1.5 year period.

9.3.6 Other Trace Elements (<1,000 mgikg): Sr, Ni, Co, Cd, Ag, Hg Figure 9.34 shows box plots for additional trace element distributions by country.

392 Figure 9.32 Trace Oxyanionic Element Distribution (Arsenic) Within the United States @ con ds3 200

@ dsl

ET#ds2

E~

~

sau

mid

-

150 -

100 -

50-

I i

O-

Apsenic

Concentpat]on

]n

Bottom

Ash

From

U.S.

Faci]]ties

Facility

n

min

max

mean

median

con

31

16.7

189

39.53

33.90

dsl

5

13.7

20.5

17.98

19.10

ds2

12

4.0

28.8

17.13

19.20

ds3

6

24.5

34.5

30.21

31.85

Saugus

1

-

-

-

38.0

m idconn

3

13.6

20.3

18.0

20.1

Figure 9.33 Trace Oxyanionic Element Concentration (Arsenic) as a Function of Time IO0

E

!

80

~ 7 i,I u)

6O

~ <

4O

A

m-ii

__J

<

9

;-, 9"

/

2O

.

.

.

I

5

i

.

9

I

",m -m, /%

n

F--

.

i 9

i

,

,

,

,

10

SAMPLING EVENT After Eighmy et al., 1992

.

.

.

.

0 First Hour Composite i Second Hour Composite /k Third Hour Composite A Fourth Hour Composite m~l Daily Composite

m "m-m"

!

15

2O

Figure 9.34 Other Trace Element D~stributionby Country

I

Strontium Concentratlon I n Bottom Ash Dy Country

N l c k e l Concentratlon I n Bottom Asn by Country

Conalt Concentration I n Bottom d s n b y Country

,.I m-1 I

Caamlum Concentratlon I n Bottom Ash Dy Country

0

S l l v e r Concentratlon I n Bottom bsn ~y Country

Mercury C o n c e n t r a t l o n in Bottom Ash Dy Country

394 Strontium is present in bottom ash at concentrations ranging from 85 to 1,000 mg/kg. Such concentrations are similar to concentrations seen in lithospheric and soil materials. Median values for strontium in bottom ash for the United States and Denmark are 500 and 220 mg/kg respectively. As shown in Figure 9.34, the concentrations of strontium seem to be higher in the United States bottom ashes compared to those concentrations seen in Danish bottom ashes. At this time, no explanations can be provided for this phenomena. Nickel is found in bottom ashes in concentrations of 7 to 4,280 mg/kg. Median values for the United States, Canada, Denmark, Sweden and the Netherlands are 470, 584, 155, 55.5 and 50.0 mg/kg, respectively. The concentrations of nickel in the North American countries tends to be higher than those seen for the European facilities. Cobalt is found in bottom ash at concentrations ranging from 6 to 350 mg/kg. Values found in lithospheric and soil materials tend to be lower. Median values for the United States, Canada, Denmark and the Netherlands are 31.5, 238, 26.5 and 8.5 mg/kg respectively. The concentrations in Canadian facilities tends to be much higher than those observed in either the U.S., Danish or Dutch facilities. Cadmium is found in bottom ash at concentrations ranging from 0.3 to 70.5 mg/kg. Such concentrations are very enriched relative to lithospheric or soil materials. Median values for the United States, Canada, Germany, Denmark, Sweden and the Netherlands are 13.9, 5.0, 11.0, 1.3, 25.0 and 0.5 mg/kg respectively. Silver concentrations range in bottom ash from 0.29 to 36.9 mg/kg. Such concentrations are very enriched relative to lithospheric or soil materials. Median values for the United States, Canada and Denmark are 6.75, 5.00 and 10.55 mg/kg respectively. As shown in Figure 9.34, the concentrations of silver in bottom ash is fairly similar between the three countries that are depicted. Mercury concentrations in bottom ash range from 0.02 to 7.75 mg/kg. Such concentrations are much higher than those seen for lithospheric or soil materials. Median values for the United States, Canada, Switzerland, Denmark, Sweden and the Netherlands are 1.08, 0.02, 0.20, 0.52, 0.08 and 0.20 mg/kg respectively. As can be seen in Figure 9.34, the concentrations of mercury in North American facilities is much higher than those seen for facilities in Europe. It is important to understand how the concentration of some of these additional trace elements can vary in bottom ash samples from facilities within the same country. As shown in Figure 9.35, the concentration of cadmium in bottom ashes from the United States facilities is found to be quite variable. At this time it is not certain as to why. It may be a function of both the waste composition that is being combusted and the temperature of combustion.

395 Figure 9.35 Other Trace Element Distribution (Cadmium) within United States Facilities

ET#con E~ ds3

ET#dsJ E~ epa

~ ds2 ELTamid

80-

ID)

7-

60-

Er) E

E

o

n3

40-

[_ c(.3 CO r.._.)

L

20-

1

..1

__1_

I

~1

I

_

Cadmium Concentnation in Bottom Ash Fpom U.S. F a c i l i t i e s

Facility

n

min

max

mean

median

con dsl ds2 ds3 epa mid

15 6 6 6 11 5

2.62 15.2 11.0 8.5 20.2 5

70.57 62.5 45.4 14.4 45.8 8

16.15 36.7 20.0 11.28 32.4 6.2

6.06 31.1 15.6 11.0 32.9 5.0

396 It is also important to understand how variable trace elements can be in bottom ashes produced from the same facility over time. Cadmium has been selected as an example. As shown in Figure 9.36, cadmium can be quite variable in the bottom ash process stream over time. Sampling occurred at the Concord, New Hampshire facility over a 548-day period. It is not clear why cadmium is so variable in bottom ash. It may be due to waste stream effects and the fact that cadmium is a very volatile element. Figure 9.36 Other Trace Element Concentrations (Cadmium) as a Function of Time 100 o~

90

E7~

8O

E

70

.

.

.

.

i

.

.

.

.

i

.

.

.

.

i

9 Replicate Analyses of Daily Composite m~m Daily Composite

m

60 5o < 0

3o

~

2o

._J

9

0

9

0

I

9

I

5

i

i

i

l-m i

.m

I

10

i

mm 9

SAMPLING EVENT

n-n-m '

f

15

20

After Eighmy et al., 1992

9.3.7 Other Trace Elements Continued (<1,000 mglkg)" B, Br, F, I Concentrations of some other trace elements in bottom ashes by country is depicted in box plots shown in Figure 9.37. Boron is found in concentrations ranging from 38 to 510 mg/kg. Such concentrations tend to be slightly enriched relative to lithospheric or soil materials. Median values for the United States and Canada are 140 and 171 mg/kg respectively. As shown in Figure 9.37, boron concentration is quite similar in bottom ashes from the United States and Canada. Bromine is found in bottom ash in concentrations ranging from 1.4 to 150.2 mg/kg. Such concentrations are higher than those seen in lithospheric or soil materials. Median values for bromine in the United States and the Netherlands are 40.75 and 2.40 mg/kg respectively. There is a wide difference in concentrations of bromine between the United States and Dutch bottom ashes.

397 Fluorine is found in bottom ashes in concentrations ranging from 200 to 1,100 mg/kg. Such concentrations tend to be less than those seen in lithospheric or soil materials. Median values for Switzerland, Denmark and the Netherlands are 300, 500 and 385 mg/kg respectively. As shown in Figure 9.37, the concentrations between the three European countries is similar. Iodine is also found in bottom ash. The data are very limited, he data suggests that iodine is present in concentrations of around 1,000 mg/kg.

9.3.8 Elements Related to Biogeochemical Cycles: C, S, P, N Figure 9.38 shows distribution of elements by country for those elements related to biogeochemical cycling. Organic carbon is the most prevalent element in bottom ash related to biogeochemical cycling. Very limited data are available on the concentrations of organic carbon in bottom ash. Those data that are available suggest that organic carbon is present in concentrations of 10,000 to 20,000 mg/kg though higher concentrations are possible. Sulphur is the next most abundant element related to biogeochemical cycling in bottom ash. Sulfur is found in concentrations ranging from 1,000 to 5,000 mg/kg. Such concentrations are less than those typically seen in lithospheric or soil materials. The majority of the sulphur that is present in bottom ash is present as sulphate. Median values for sulphur in bottom ash for the United States, Canada, Switzerland and Denmark are 6,100, 4,070, 2,000 and 2,100 mg/kg respectively. As can be seen in Figure 9.38, the total sulphur concentrations in bottom ashes from around the world are fairly similar. The next most abundant element in bottom ash relative to biogeochemical cycling is phosphorus. Phosphorus is found in concentrations ranging from 1,400 to 6,400 mg/kg. Such concentrations are similar to those seen in lithospheric and soil materials. Median values for the United States, Canada, Switzerland and Denmark are 2,900, 3,350, 5,000 and 3,450 mg/kg respectively. The total phosphorus concentrations in bottom ash are fairly similar between countries around the world. The next most abundant element in bottom ash relative to biogeochemical cycling is nitrogen. Nitrogen is found in concentrations ranging from 110 to 900 mg/kg. Such concentrations are less than those seen in lithospheric or soil materials. Median values for Switzerland and Denmark are 400 and 490 mg/kg respectively. As can be shown in Figure 9.38, such concentrations are similar between the two countries.

Figure 9.37 Other Trace Element Distributions by Country

LOO

-

0

-

0

Boron Concentration i n bottom Ash by Country

0

4

F l u o r i n e Concentration i n Bottom Ash by Country

Bromine Concentration i n Bottom Ash by Country

Figure 9.38 Elemental Distributions by Country of Elements Related to Biogeochemical Cycling

""I

0

Sulfur

'

0

Concentration i n Bottom

Ash by Country

Phosphorus Concentration in Bottom Ash by Country

0

Nitrogen concentration i n Bottom ~ s hby Country

400

9.3.9 Exotic Elements, Lanthanides, Actinides There are a number of exotic elements, lanthanides, and actinides that have been measured in bottom ash. As shown in Table 9.19, the elements scandium, rubidium, cesium, lanthanum, tantalum, gold and indium are found in bottom ashes at concentrations that are usually less than 10 mg/kg. There is very little information available on the concentrations of these elements in bottom ashes. Data are available from Eighmy et al. (1992). These exotic elements play a minor role in ash physical properties and chemistry. Nevertheless, these elements are present. There are a number of lanthanide elements that are present in bottom ashes. These include cerium, neodymium, scandium, europium, terbium, tellurium, dysprosium, and ytterbium. These exotic elements are present in concentrations less than 1 mg/kg. They play a negligible role in ash characteristics. Actinides are also present in bottom ash. Thorium and uranium are the two elements that have been found. Their concentrations tend to be less than 1 mg/kg. Like the lanthanides, these actinides play a negligible role in ash characterisation. They are not enriched relative to soils.

9.3.10 Isotopes Some limited information is available on the presence of radioactive isotopes of elements in bottom ash. The data show that the levels seen in bottom ash are the same as those seen for natural soils. Data reported by van der Sloot (1992) show the following activities for ~Ra, =STh, =SRa and 4~ respectively (Bq/kg): 18.6, 15.2, 13.3 and 277. Soils have values of 10 to 40, 10 to 50, 10 to 50 and 200 to 500 for those same isotopes (Ackers et al. 1985).

9.3.11 Role of Particle Size in Element Distribution Stegemann and Schneider (1991) have examined how copper, nickel, chromium, zinc and lead vary as a function of particle size (<0.4 mm, 0.4-2 mm, 2-8 mm, >8 mm). Zinc was the only element that exhibited enrichment as a function of decreasing particle size. Figure 9.39 presents the data. Conversely, data from the NITEP program (Sawell et al., 1990a), relative to the GVRD facility, do show that lead is enriched in the finer fractions (e.g. > 1.0 mm to < 4.0 mm and the <1.0 mm fractions). The reason for the relative enrichment of certain metals in the fine fraction is not clear, though it may be related to how these metals, as pure solids, melt to form smaller particle sizes.

401 Figure 9.39 Element Distribution by Particle Size 5000

4500

~

4000 ~3500

o2

rnm

E 3000 2500 2000 1500

........................................... i! ........i ..........................................

t'

!

1000

.......

'

500

Cr

h1

Cu

Zn

l

After Stegemann and Schneider, 1991 9.3.12 Influence of Combustor Type and Operation on Bottom Ash Inorganic Characteristics The NITEP and WASTE programs provide the best information to date on the influence of combustor type and operation on bottom ash inorganic composition. The NITEP program is noteworthy because of its comprehensive evaluation of a variety of combustor types and combustion conditions on ash quality (Sawell and Constable, 1993). The WASTE program is noteworthy because of its evaluation of the influence of MSW composition on bottom ash composition (WASTE Program, 1993). Aluminum concentrations in bottom ashes are shown in box plots in Figure 9.40. Generally, the concentration is lower in two-stage bottom ashes, possibly because of the dilution effect from unburned material in the bottom ash. Copper (Figure 9.41), nickel (Figure 9.42), zinc (Figure 9.43) and chromium (Figure 9.94) also show similar behaviours to that of aluminum. All of these elements have relatively high vapour pressures and would be expected to remain in the bottom ash fraction during combustion (see Chapter 8). Elements that are much more volatile, such as cadmium, show a different behaviour than less volatile elements. Cadmium concentrations in bottom ashes are shown in Figure 9.45. The highest concentrations are found in bottom ashes with the lowest hearth temperatures. Thus, the two-stage bottom ashes have the highest levels. There is relatively close agreement amongst the samples from mass burn and RDF facilities.

402 Figure 9.40 Aluminum Concentration in Bottom Ashes and Grate Siftings as a Function of Combustor Type Al_conc 60000 -

T

T

ITI

A = 2-Stage, Boiler B = 2-Stage, Economiser (Bridle & Sawell, 1986; SaweUet al., 1989; Sawell & Constable, 1990) Peel 2-Stage, Boiler D = Peel 2-Stage, Economiser (PRRI, 1992)

C =

E = Mass Bum, Boiler (Waste Program, 1993)

40000 -

__L_

I 20000 A

F = Mass Bum, Economiser (Sawell & Constable, 1988)

T

G = RDF, Economiser (Sawell et al., 1991)

1 B

C

O

E

F

6

Figure 9.41 Copper Concentration in Bottom Ash and Grate Siftings as a Function of Combustor Type Cu_conc 26000

-

A = 2-Stage, Boiler B = 2-Stage, Economiser (Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990) C = Peel 2-Stage, Boiler D = Peel 2-Stage, Economiser (PRRI, 1992) E = Mass Burn, Boiler (Waste Program, 1993)

13000 -

F = Mass Burn, Economiser (Sawell & Constable, 1988)

T

0 -

i ' ---,-rl A

T~ L

..j_ B

1 i ~_

7_L_

G = RDF, Economiser

m

L_..._I

I _b !

__L_

(Sa~U et al., 1991)

I

403 Figure 9.42 Nickel Concentration in Bottom Ashes and Grate Siftings as a Function of Combustor Type ~a NLconc 4000 0

A = 2-Stage, Boiler B = 2-Stage, Economiser 3000 -

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser

2000 -

E = Mass Burn, Boiler F = Mass Burn, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

G = RDF, Economiser

(Sawell et al., 1991)

o

1000 "F0 -

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

A

~

T

-r-

l

I _._.1

B

C

O

E

1

o ~

l

l

-r__,___

'i ]

I '....L_ 1 F

G

Figure 9.43 Zinc Concentration in Bottom Ash and Grate Siftings as a Function of Combustor Type Zn_conc

A = 2-Stage, Boiler B = 2-Stage, Economiser (Bridle & Sawell, 1986; Sawell et al., 1989; Saweil & Constable, 1990)

7000 -

C = Peel 2-Stage, Boiler D = Peel 2-Stage, Economiser (PRRI, 1992)

w

E = Mass Burn, Boiler (Waste Program, 1993)

5000 -

F = Mass Burn, Economiser (Sawell & Constable, 1988) 3000 -

!

I

T

_L

1000 A

B

C

D

G = RDF, Economiser (Sawell et al., 1991)

T

I

=

1 E

F

G

404 Figure 9.44 Chromium Concentration in Bottom Ashes and Grate Siftings as a Function of Combustor Type ~CP_conc

-I-

1200 -

-I-

I I ....,I__

A = 2-Stage, Boiler B = 2-Stage, Economiser (Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

900 -

C = Peel 2-Stage, Boiler D = Peel 2-Stage, Economiser

600 -

(PRRI, 1992) T

!

300 -

T

-T-

I

..J.--'--

0 -

I

E = Mass Bum, Boiler (Waste Program, 1993) F = Mass Burn, Economiser (Sawell & Constable, 1988)

_1_

~ "-'--

G = RDF, Economiser (Sawell et al., 1991)

O

A

B

C

O

E

F

G

Figure 9.45 Cadmium Concentration in Bottom Ash and Grate Siftings as a Function of Combustor Type

~a Cd_conc

60-

T

A = 2-Stage, Boiler

B = 2-Stage, Economiser

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

C = Peel 2-Stage, Boiler

(PRRI, 1992)

E = Mass Bum, Boiler F = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

G = RDF, Economiser

(Sawell et al., 1991)

D = Peel 2-Stage, Economiser

40-

20-

5

_1_

I,II

"-F-

B

-F-

C

I

I ! "-'r--"

O

E

_L F

G

405 Elements which melt easily and are capable of dripping through the grate bars into the siftings hoppers show a different behaviour. Lead data for bottom ashes and siftings and riddlings are shown in Figure 9.46. High values are found in the siftings. Generally, lead values in bottom ashes from two-stage, mass burn and RDF facilities are similar. Figure 9.46 Lead Concentration in Bottom Ashes and Grate Siftings as a Function of Combustor Type A = 2-Stage, Boiler B = 2-Stage, Economiser

e~ Pb_conc 34000 -

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser E = Mass Burn, Boiler F = Mass Burn, Economiser G = RDF, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988) (Sawell et al., 1991)

..--.,.-.

~7000-

o 0 -

-T

r ~ ,

--r--

Lw~

A

B

~

'•

i__

_

,,..,L.., C

D

E

'

F

T..

I __.__ I '~6

Results from the WASTE program study suggest that some cause and effect relationships can be observed between waste feed composition and bottom ash elemental composition. The addition of lead acid batteries increases lead concentrations in grate siftings rather than bottom ash. The lead tends to melt and become siftings and riddlings. The addition of cadmium benzoate solutions does not influence cadmium concentrations in bottom ash, rather the cadmium partitions to the fabric filter residue stream. Similar observations were made with cadmium pigment additions.

9.3.13 Influence of Aging on Bottom Ash Inorganic Characteristics Data presented by Vehlow (1992) examine the role of aging in the composition of lead, cadmium, cooper and zinc in bottom ashes from three German facilities. No significant changes in the concentration of these elements were seen after nine months of aging.

406 9.4 ORGANIC CHARACTERISTICS OF BOTTOM ASH 9.4.1 Organics Present in Bottom Ash

The organic carbon content of bottom ash tends to range from 2 to 4% in well burnedout bottom ashes. The majority of this carbon is not characterised at this point in time. Examination of bottom ash samples with scanning electron microscopy suggests that much of this carbon is unburned municipal solid waste. It is possible that this carbon is cellulose, plant fibre or plastic in nature. Trace organics of potential human health concern have been quantified in bottom ash. These include the polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzo-p-furans (PCDFs) as well as potential precursors for these compounds under certain reaction conditions. The potential precursors are chlorinated benzenes (CBs), chlorinated phenols (CPs) and polychlorinated biphenyls (PCBs). Another group of toxic compounds that may attract regulatory attention in the future are condensed polyaromatic hydrocarbons (PAHs). At present, there are 16 compounds that fall under this category; some are carcinogenic. These also have been characterised in bottom ashes. Chapter 8 discusses their presence in combustion residuals and their fate during combustion. Compared to the many detailed data sets on physical properties or inorganic characteristics of bottom ash, the information on trace organics is very limited. Careful review of the data that are available indicate that many studies have a paucity of information about combustor operation, sampling technique and analytical methodologies. Therefore, selected data sets are presented. For the most part, the combustors listed in Table 9.1 were the basis for these data. 9.4.2 Dioxins and Furans

Table 9.20 summarises compiled data on total PCDD and PCDF concentrations found in bottom ash. Also included are total toxic equivalents according to the NATO-CCMS toxic equivalency method. The data show relatively low concentrations of PCDDs and PCDFs. Most of the isomers accounting for the typical PCDD values of 0.03 to 0.04 ng/g come from the hepta and octa groups. Most of the isomers accounting for the typical PCDF values of 0.05 to 0.1 ng/g come from the tetra and penta groups. These values are all associated with well operated mass burn facilities. Data from the RDF mass burn facility in the United States had slightly higher levels of both PCDDs and PCDFs. Other data on PCDD and PCDF levels in bottom ash have been reported by Hiraoka et al. (1991), Klein and Tscheschlok (1989), Morselli et al. (1989) and Roffman (1991). These levels are similar to or higher than the values reported in Table 9.20. It is difficult to determine if combustor operation influenced the observed levels. The addition of boiler ash to the bottom ash process stream may have occurred.

Table 9.20 Trace Organic Concentrations (nglg) in Bottom Ash Countrv

Facili

GVRD PEI LW SWARU QUC Germanv A

Canada

Total Dioxins Total Furans l/TEQ Total Total Total Total Reference (PCDD) (PCDF) Chlorophenols Chlorobenzenes Polyaromatic Polychlorinated (cp) (CB) Hydrocarbons Biphenyls (PAH) (PCB) NDa ND 9.0 ND 181 ND Sawell et al., 1990a ND 1,800 ND ND ND 20 2,190 ND Sawell et al., 1989b 34.1 6.7 ND ND Sawell et al., 1989a 19,000 8 0.4 ~0.2 164 4.0 125-968 ND Sawell & Constable, 1988 1448 6.0-1 3.5 ND ND 0.036-0.039 0.096-0.1 02 0.001 8

C 0.025-0.029 0.054-0.068 0.0008 Mid-Conn 0.04-0.31 0.10-0.50 Dry Scrubber 1 Dry Scrubber 2 Dry Scrubber 2 Dry Scrubber 3 a ND indicates not detected.

United States

4-5 73 120 36 83

ND 18 ND ND 36

13-29

ND

Sawell et al., 1992 LIRPB, 1992b

408 Most modern facilities are able to achieve total PCDD and PCDF levels below 0.5 ng/g. This represents a total toxic equivalent (1/TEQ) of about 10 ng/g. These levels are of the same magnitude as those seen in forty German soil samples (Hagenmaier, 1989).

9.4.3 Chlorinated Benzenes and Chlorinated Phenols Table 9.20 also summarises compiled data on total CB and CP concentrations found in bottom ash. Typical levels are between 9 and 164 ng/g for CP and 4 and 36 for CB. These levels are considered low.

9.4.4 Polyaromatic Hydrocarbons and Polychlorinated Biphenyls Table 9.20 also summarises PAH and PCB concentrations found in bottom ash. Typical concentrations for total PAHs range from 13 to 2,190 ng/g. Typical concentrations for total PCBs range from below detection to 8 ng/g. PAHs are a measure of the quality of the combustion process. Significantly higher concentrations are seen in two stage systems and poorly operated mass burn facilities. Well operated incinerators can easily produce bottom ashes with total PAH concentration less than 100 ng/g. These are typical for soils in rural areas (Menzie et alo, 1992). For PCBs, regardless of which technology is employed, levels of less than 10 ng/g can be achieved. This is considered to be a low level. Higher values reported in the literature (Morselli et al., 1989; Magagni et al., 1990; Roffman, 1991, Morselli et al., 1992) may be caused by the inclusion of boiler ash in the bottom ash process stream.

9.5 CHARACTERISTICS OF GRATE SIFTINGS Grate siftings comprise a small mass fracture of the combustion residuals produced in incineration facilities (see Chapter 8). They are usually collected in hoppers beneath the grate and added to the grate ash to produce the bottom ash process stream. Very little information is available on the physical or chemical characteristics of grate siftings. Size ranges can vary given the spacing between grate bars on the hearth. Molten aluminum, zinc, copper and lead can drip down into the hoppers and reagglomerate as small stalagmites on the hopper walls. Visual classification shows that molten metals, fine glass and fines comprise the bulk of this material. No published information exists on the physical properties of this residue. Data are available on the inorganic characteristics of this process stream. Recent efforts under the WASTE Program (1993) have resulted in detailed characterisation of

409 the material. Elemental abundances from a single composite sample are shown in Table 9.21. Some of these elements were quantified over a one- week period. The range of data is shown in Table 9.22. Similar efforts were also made for an RDF facility (Table 9.23). The bulk of the data shows that grate siftings are enriched in AI, Cu, Pb and Zn. High levels of Si are also found. Mass balances on grate siftings, grate ash and bottom ash suggest that grate siftings account for up to 40% of the total Pb loading in bottom ash. These metals can be problematic with respect to H2 evolution during bottom ash aging, and the removal of this process stream from the bottom ash process stream warrants further consideration. It should be noted that the practice is already in use at some European facilities. 9.6 CHARACTERISTICS OF COMBINED ASH AND SCRUBBER RESIDUE Although the various residue streams are generally separated in most countries, combined bottom ash and APC residue is presently the most prevalent waste stream from U.S. incineration facilities. There are a number of U.S. studies on the physical and chemical characteristics of combined ash. The two most recent reports (Kosson et al., 1993; LIRPB, 1993) provide a detailed description of sampling, testing and analytical procedures, as well as the data evaluation techniques used. The LIRPB (1993) study cited data on combined ash and bottom ash from 5 different facilities, whereas the study conducted for the U.S. EPA (Kosson et al., 1993) evaluated separate streams of bottom ash, APC residues and combined ash from one U.S. facility. These studies indicated that many of the physical properties of combined ash are similar to bottom ash, whereas many of the chemical properties of combined ash can be determined through proportioning the contributions from bottom ash and APC residues. Other studies and reviews were conducted by the U.S. EPA on combined ash from a number of U.S. facilities and the reader is referred to those studies for further information on combined ash streams (U.S. EPA, 1987 & 1990).

410 Table 9.21 Elements Detected in Mass Burn Grate siftings Element

Concentration (mg/kg)

Element

Concentration (mg/kg)

Ag

64.8

Mn

1,500

AI

57,300

Mo

48

As

59.7

Na

38,000

Au

0.2

Nd

13.4

Ba

740

Ni

691

Br

0.0

Pb

9,680

Ca

68,200

Rb

25.6

Cd

22.2

Sb

305

Ce

28.6

Sc

6.2

CI

2,110

Se

0.0

Co

24.1

Si

243,700

Cr

1,304

Sn

1.5

Cs

0.6

Sr

0.0

Cu

3,041

Ta

0.7

Dy

0.0

Tb

0.7

Eu

0.4

Th

2.4

Fe

82,300

Ti

7,000

Hf

5

U

2.0

Hg

0.0

V

79.2

In

330

W

8.3

I

0.0

Yb

0.5

K

7,900

Zn

2,140

La

13.0

Zr

266

411 Table 9.22 Range of Concentration of Elements in Mass Burn Grate Siftings Concentration Range (mg/kg) Element AI

38,000-63,000

As

2-65

B

98-232

Ba

1349-2629

Be

<3.0

Cd

<5.0-14.6

Cr

278-562

Cu

2,347-25,215

Hg

<0.02-5.39

Ni

169-468

Pb

56,000-34,000

Sb

130-570

Se

<0.25-8.6

Sn

171-946

Zn

2,450-5334

412 Table 9.23 Range of Concentration of Elements in RDF Grate Siftings Concentration Range (mg/kg) Element Ag

<0.9

AI

36,230-62,300

As

8-13

Ba

150-990

Be

<0.9

Bi

500-1170

Ca

51,790-92,190

Cd

5.3-12.9

Co

28-170

Cr

230-460

Cu

740-11,530

Fe

25,150-36,510

Hg

0.2-2.35

In

<0.9

Mg

7,860-12,780

Mn

490-1,170

Mo

40-120

Na

31,800-41,600

Ni

210-1,140

P

70-1,560

Pb

2,140-20,390

Sb

6-59

Se

0.3-1.8

Si

49,070-99,380

Sn

450-2,000

Te

<2.3

Ti

4,390-8,990

V

60-170

Zn

1,150-6,730

413 REFERENCES

Ackers, J.G., J.F. DeBoer, P. DeJong and R.A Wolschrijn. "Radioactivity and Radon Exhalation Rates of Building Materials in the Netherlands." Sci. Total Environ 45: 151156 (1985). Benoit, J. and T.T. Eighmy. Methods of Placement a.nd Stability Analyses for Ash/Sludqe Mixtures.. Environmental Research Group Final Report. UNH Durham, NH (1989). Bridle, T.R. and S.E. Sawell. "NITEP Phase I: Testing at the Prince Edward Island Energy-From-Waste Facility, Assessment of Ash Contaminant Leachability", Internal Environement Canada Report, 1986. Brunner, P.H., H. Moench and S. McDow. "Organic Carbon in the Residues of Waste Incineration". EAWAG News, 22/23: 1718, 1987. Chesner, W.H. Personal Communication, 1993. Chesner, W.H., R.J. Collins, and T. Fung. Assessment of the Pote.ntial Suitability of Southwest. Brooklyn incinerator Residu..e in Asphaltic Concrete Mixes. New York State Energy Research and Development Authority Report 90-15, Albany, NY (1988). Eighmy, T.T., D. Gress, X. Zhang, S. Tarr and I. Whitehead. Bottom Ash Utilization Evaluati .on for the Concord, New Hampshire Waste-t0.-Ener.qy Facility. Environmental Research Group Interim Report, UNH Durham, NH (1992). Eighmy, T.T., J.D. Euseden Jr., K. Marsella, J. Hogan, D. Domingo, J.E. Krzanowski and D. St~mpfli. "Particle Petrogenisis and Speciation of Elements in MSW Incineration Bottom Ashes". In Environmental Aspects of Construction with Waste Materials Edited by J.J.J.R. Goumons, H.A. van der Sloot and Th. G. Albers. Elsevier Science B.V., Amsterdam, p. 111, 1994. Environment Canada..The National Incinerator .Testing and Evaluation Program: Two: Staqe Combust.jon (Prince Edward Island) Environment Canada reports EPS 3/UP/l, vols 1-4, Ottawa, Canada (1985). Gardner, K.H. Characterization of Leachates from. Municipal Incinerator Ash Materials. Masters Thesis, Clarkson University (1991). Geoteknisk Institut. Laboratorieunder s~.qelse: hvidovre.. Aued~revaerket. sag 160 04776, rapport 1, 1992-01-17 (1992). Hagenmeier, H. "Polychlorierte Dibenzodioxine und Polychlorierte DibenzofuraneBestandsaufnahme und Handlungsbedorf." VDI Berichte 745 939-978 (1989).

414 Hartl6n, J. Personal communication (1992). Hartl6n, J. and P. Elander. Residues from Waste Incineration-Chemical and Physical Properties. Swedish Geotechnical Institute report SGI Varia 172. Linkoping, Sweden (1986). Hartl6n, J. and T. Lundgren. "Utilization of Incinerator Bottom Ash-Legal, Environmental and Engineering Aspects." In Waste Materials in Construction Edited by J.J.J.R. Goumans, H.A. van der Sloot and T.G. Aalbers. Elsevier Scientific Publishers B.V., Amsterdam, p. 207 (1991). Hartl6n, J. and J. Rogbeck. "Sorted Incinerator Slag used as Fill Material." In Proceedin.qs of the International Conference on Mun.icipal Waste Combustion Hollywood, Florida, April 11-14. AWMA, Pittsburgh, Pennsylvania, p. 5B-1 (1989). Hiraoka, M., N. Takeda, K. Tsumura, T. Fujiwara and S. Okajima. "Control of Dioxins from Municipal Solid Waste Incinerator." Chemosphere 19:323-330 (1989). Hjelmar, O. Personal communication (1992). Kirby, C.S. and D.J. Rimsstidt. "Mineralogy and Surface Properties of Municipal Solid Waste Ash" Enviro. Sci. Technol. 27:652-660, 1993. Klein, H. and K. Tscheschlok. "Thermische Aufarbeitung yon Flug-und Filterst~uben aus M011verbrennungsanglen durch Drehstrom-Plasmatechnik". In M011verbrennunq und Umwelt Edited by K.J. Thom~-Kozmiensky. EF-Verlag, Berlin, 3:823 (1989). Kluge, G., H. Saalfeld and W. Dannecker. Untersuchun.qen ..desLan,qzeitverhaltens von.. M0!lverbrennun.qsschlacken beim Einsatz in Stra Benbau Forschungsbericht Nr. 103 03 006, Umweltforschungsplan des Bundesministers des Innern, Berlin (1981). Kosson, D.S., T.T. Kosson and H. van der Sloot. U.S. EPA Pro.qra.m for Evaluation of Treatment and Utilization of Municipal Waste Combustor ......Residues - Phase 1. Laboratory Testin.q of Solidification/Stabilization processes. U.S. EPA, Cincinnati, OH (1992). Kosson, D.S., T.T. Kosson and H . A . van der Sloot, Evaluation of. Solidification/St.abilization Treatment Processes for Mu.nicipal Waste Combustion Residues. U.S. EPA Report NTIS PB 93-229 870/AS), 1993 Lindsay, W.L. C_hemical Equilibria in Soils, J. Wiley & Sons, New York, 1979. Long Island Regional Planning Board (LIRPB). The Potential for Beneficial Use of Waste-to-E.ner,qy Facility Ash - Draft En.aineerin,q Property Data Report. LIRPB/NYSERDA (1992a).

415 Long Island Regional Planning Board (LIRPB). The Potential for Beneficial Use of. Waste-to-Energy Facility Ash - Draft Chemical and Environmental Property Data. Report. LIRPB/NYSERDA (1992b). Long Island Regional Planning Board (LIRPB). The Potenti..al for Use of Waste-to: Energy Facility Ash, final report (8 volumes), LIRPB/NYSERDA, 1993. Ludvigsen, K. and O. Hjelmar. Vurderin~ af sla.q,qe fra affaldsforbr.aendin,qsanlae.q. Vandkvalitetsinstituttet Udkast 1992-04-15 Horsholm, Denmark (1992). Magagni, A., G. Boschi and V. Cocheo. "Emissions of a MSW Incinerator Equipped with a Post-Combustion Chamber, Dry Scrubber and ESP." Chemosphere 20:1883-1890 (1990). Menzie, C.A., B.B. Potocki and J. Santodonato. "Exposure to Carcinogenic PAHs in the Environment." Environ. Sci. Technol 26:1278-1284 (1992). Morselli, L., S. Zappoli, A. Liberti, M. Rotatori and E. Brancaleoni. "Evaluation and Comparison of Organic and Inorganic Compounds Between Emission and Immision Samples from Municipal Waste Incinerators." Chemosp.here 18:2263-2273 (1989). Morselli, L., S. Zappoli and T. Tirabassi. "Characterization of the Effluents from a Municipal Solid Waste Incinerator Plant and of their Environmental Impact." Chemosphere 24:1775-1783 (1992). Peel Resource Recovery Incorporated (PRRI). "Ash and Quench Water Testing Report." Report prepared for the Region of Peel, Brampton, Ontario, July 1992. Roffman, H. "Major Findings of the U.S. EPA/CORRE MWC-Ash Study." In Proceedin,qs of the Municipal Waste Combust.i0n Conferen.ce April 15-19, Tampa, Florida. AWMA, Pittsburgh, PA pp. 96 (1991). Sawell, S.E. and T.W. Constable. NITEp Pha.se liB: Assessment of Contaminant .Leachability from the Residues of a Mass Burnin.q Incinerator Environment Canada, EPS manuscript series IP-82, vol. VI, Ottawa, Canada (1988). Sawell, S.E. and T.W. Constable. The National !n.cinerator Testin,q and Eyaluati_on Pro.qram. A Summary of the_Characterization and Treatment Studies on Residues From Municipal Solid W.a.ste Incineration. Environment Canada Publication No. EPS 3/UP/8, Ottawa, Canada (1993). Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testin.q and .E.valuation Pro.qram: And Evaluation of Contaminant Leachability from Residues Collected at a Refuse Derived..Fuel Municipal Waste Combustion Facility. Environment Canada Report, EPS manuscript series IP-96, Ottawa, Canada (1989a).

416 Sawell, S.E., T.W. Constable and R.K. Klicius. The Nation.al Inc.inerator Testin.q and Evaluation Pr0.qram.:.Characterization of Residues from a Modular Municipal Waste Incinerator with_Lime-Based Air Pollution Control. Environment Canada Report, EPS manuscript series IP-101, Ottawa, Canada (1989b). Sawell, S.E., T.W. Constable and R.K Klicius..The Natioqal Incinerator Testin,q and Evaluation Pro.,qram; Characterization of Residues from a. Mass Burnin,q Municipal Waste Incinerator with Lime-Based Air Pollution .Control (Burnaby, B.C.) Environment Canada Report, EPS manuscript series IP-110, Ottawa, Canada (1990a). Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator TestinQ and Evaluation Program: Chara.cterization of Residues from a Two-Sta,qe Incinerator with Rotary Kiln (3M Canada) Environment Canada Report, EPS manuscript series IP-119, Ottawa, Canada (1990b). Sawell, S.E., T.W. Constable and R. K. Klicius. The National Incinerator Testin.q and Evaluation Pr0,qram: Charact.erization of Residues from a Refuse-Derived Fuel Combustion F.a.cility (Mid-Conn.ecticut). Environement Canada Report, Manuscript Series, 1991. St,~mpfli, D., H. Belevi, R. Fontanive and P. Bacchini. Reactions of Bottom Ash from Municipal Solid Waste In..cinerators and Const.ruction Waste Samples with Water. EAWAG/AWS, project 3335, Dubendorf, Switzerland (1990). St~mpfli, D. Personal communication (1992). Stoelhorst, D. 'q'he Use of Waste Materials in Civil Engineering: AVI Slag Can Replace Gravel in Concrete Production." Waste Materials in Construction Edited by J.J.J.R. Goumons, H.A. van der Sloot and T.G. Aalbers. Elsevier Science Publishers B.V., Amsterdam, p. 71 (1991). Stegemann, J.A. and J. Schneider. "Leaching Potential of Municipal Waste Incinerator Bottom Ash as a Function of Particle Size Distribution." Waste Materials in Construction Edited by J.J.J.R. Goumons, H.A. van der Sloot and T.G. Aalbers. Elsevier Science Publishers B.V., Amsterdam, p. 135 (1991). Swedish Geotechnical Institute. SGI Database, Link6ping, Swden, 1993. TAUW. Veabrin kwaliteitskontrole van AVI-slakken '87-'88 Deventer, the Netherlands (1988).

RAP-305/JJS/avd,

Theis, T.L. and K.H. Gardner. "Environmental Assessment of Ash Disposal." CRC Crit. Rev. Environ. Contro) (1990).

417 U.S. EPA, Characterization of MSW Ashes and Leachates from M..SW Landfills, Monofills, and Co-Disposal Sites, EPA 530-SW-87-028, Washington, D.C., 1987 U.S. EPA, Characte..r.ization of Municipal Waste Combustion Ash, Ash Extracts, and Leachates, EPA 530-SW-90-029, Washington, D.C., 1990 van der Sloot, H.A. Personal communication (1992). Vehlow, J. Personal communication (1992). Vehlow, J., G. Pfrang-Stotz and J. Schneider. "Restoffe-charakterisierung, behandlung, verwertug." In _Symposium25 Jahre LIT 5 Jahre TAMARA, Forschunq und Entwicklun.,_q in Kernforschun,qszentrum Karlsruhe zur Hausm011verbrennun.q Kfk, Karlsruhe, Germany, p. 124 (1992). WASTE Program. Waste Analysis, Sampling, Testing and Evaluation Program Final Report of th.e Mass Burn MSW Incineration Study (Buraby, B.C.).. Report Prepared for Environment Canada, U.S. EPA and the International Lead Zinc Research Organization. Vols. 1-4 (1993). Zevenbergen, C., et al. "Weathering as a Process to Control the Release of Toxic Constituents from MSW Bottom Ash." Geoconfine 93 Edited by Arnould, Barr~s and C6me. Balkema, Rotterdam, The Netherlands, p. 591 (1993).

This Page Intentionally Left Blank

419

CHAPTER 10 - CHARACTERISTICS OF HEAT RECOVERY SYSTEM ASH

In some jurisdictions, the consideration of energy recovery may be secondary to incineration as an MSW management tool. Nevertheless, the economic benefit of recovering energy and the technical benefit of being able to control flue gas temperatures prior to the air pollution control system has resulted in virtually all new MSW incinerator facilities being designed with energy recovery capabilities. As explained in Chapter 8, the presence of the heat recovery units adds an obstacle to the path of the flue gas stream, thereby creating the potential for another ash stream to be collected. The existence of separate heat recovery ash streams is dependent on the configuration of the incinerator (see Chapter 3). In some incinerator systems, the boiler banks (such as semi-suspension combustion systems) are directly above the furnace bed and the only way to access this material is to collect ash impinged on the boiler tubes or the walls during furnace maintenance periods. In other systems, the heat recovery units are quite separate from the furnace and access to these ashes is relatively simple. Since boiler ash is usually mixed with bottom ash in the quench tank, comparatively few studies have been conducted on the characteristics of separated ashes from heat recovery systems. In order to understand some of the characteristics of heat recovery system ashes, it is important to review the mechanisms by which deposition occurs. Some of these mechanisms are described in Chapter 8. Because the mechanisms represent a complex interplay of physical and chemical processes, they are explained in further detail in this chapter. Since each type of incinerator offers a unique set of conditions under which deposition occurs, the characteristics of heat recovery system ashes are also unique. The physical and chemical data presented here focus on the influence that incinerator type and operation have on ash quality. Wherever possible, the data have also been separated into the constituent ash streams within the heat recovery system, i.e. boiler ash versus economiser ash. It is important to note many of the inferences made in this chapter draw upon the discussions of heat recovery systems given in Chapter 3 and the fate of elements in Chapter 8. 10.1 ASH DEPOSITION MECHANISMS

The flue gas stream exiting the combustion chamber of an MSW incinerator contains particles in a vapour, liquid or solid phase. Along its path, the flue gas stream encounters boiler tubes which disrupt its flow. The heat transfer surfaces are much cooler than the prevailing temperatures in the flue gas stream. On a macroscopic scale, the denser particles entrained in the flue gas tend to fall out of suspension as the velocity of the flue gas permits and collect in the hoppers below the boiler. Some particles also collect on the walls of the boiler, while others impinge or condense on the

420 heat transfer surfaces. Only the finest sized particles remain entrained in the flue gas stream and are carried over into the air pollution control system. A study at three German mass burn facilities was conducted to monitor the hold up of particulate matter in the boiler between soot blowing events (i.e. loosely impinged fly ash on boiler tubes and boiler walls). The data given in Table 10.1 indicates with lower waste loading rates, and hence lower flue gas velocities, the proportion of fly ash retained in the boiler increases. It was also noted there was a threefold increase in particle loading to the flue gas stream during soot blowing, which was carried downstream to the APC system. This is corroborated by operational experience at another modem German mass burn facility where the ratio of fly ash released into the flue gas stream and fly ash collected in the boiler hopper during soot blowing was about 2:1. Table 10.1 Summary of Boiler Ash Hold Up in Three Different Inc!nerator .Facilities Facility Waste Feed Rate Fly Ash Loading % Captured in (Mg/hr) .. (kg/Mg of Waste)* Boiler A

12 10 10 10 Vehlow and Schneider, 1991

20 2O 30 5

7.4 12.5 11 27

Typically, ash from the heat recovery system of modern, well-operated incinerators is a dust-like or fine granular material, ranging from black-grey to beige-tan in colour. The particles range in size from submicron to greater than 400 microns in diameter. In some instances, much larger chunks of boiler ash can be observed. These chunks are formed on the boiler tubes by agglomerated particles and condensed constituents from the flue gas stream which "freeze" onto the heat transfer surfaces, forming a hard, sometimes rocklike material which is difficult to remove via normal on-line cleaning processes (soot blowing or rapping). The presence of the heat transfer surfaces provides a unique set of circumstances for a number of complex physical and chemical deposition mechanisms to occur. While the degree of deposition (or "fouling") which occurs varies depending on the type of incinerator, the basic principles behind the formation of deposits are the same. One major study conducted by the National Engineering Laboratory (NEL)in Glasgow investigated gas-side fouling and developed models to predict deposition within an incineration system based on the integration of a number of chemical and physical processes (Ewart, 1988; Glen et al., 1988; Glen and Howarth, 1988).

421 As mentioned in Chapter 8, fouling is generally classified into four major categories, 1) condensation, 2) chemical reactions, 3) corrosion and 4) particulate. Vapour, fluid or solid particles entrained in the flue gas stream can be deposited on the heat transfer surfaces through complex interactions of the physical mechanisms defined in Table 10.2. Table 10.2 Physical Mechanisms Responsible for Deposition of Particles on a Solid Surface Mechanism Definition deposition due to particles too large to follow the flow of the flue Inertial impaction gas around the solid surface Interception impaction

deposition caused by particles of a finite size which follow the stream flow around an object and are intercepted by another solid object

Brownian forces & eddy diffusion

deposition caused by random velocities of small particles which deviate from the stream flow and increases potential contact with the solid surface

Fickian diffusion

deposition of a vapour from the gas stream onto a surface that has a temperature below the prevailing dew point

Gravitation

deposition of particles which are sufficiently dense to settle out under a given velocity

Thermophoresis

deposition by the force exerted by the thermal gradient caused by gas particles on the higher temperature side of a particle having greater kinetic energy, thereby exerting a greater force on the gas molecules on the cooler side of the particle

deposition by the net force exerted on a particle by the flux of a condensing vapour toward a cool surface Glen and Howarth, 1988 Diffusiophoresis

Typical operational experience has shown the upstream or leading edge of boiler tubes can become coated with a very hard, fused material, whereas the trailing edge is generally encrusted with a more friable dust-like material. These differences are attributed to the degree to which any one deposition mechanism will occur, given the orientation of the heat transfer surface in relation to the flow of the gas stream (Glen and Howarth, 1988). For example, the leading edge of a boiler tube would be susceptible to thermophoretic deposition in the early stages of use. However, inertial impaction could progressively become the most dominant deposition mechanism over time as the layers built up on the leading edge. The downstream edge of a boiler tube would consist of more particles deposited through interception and diffusion than the other mechanisms, thereby imparting the friable characteristic on the material.

422 The degree to which fusing of material occurs in smaller incinerator systems, especially two-stage systems, is much greater than that which occurs in mass burn or semisuspension combustion systems. This is due to the difference in their operational modes. The thermal gradient between the flue gas and the heat transfer surfaces in a two-stage system is much greater than the gradient in either a mass burn or RDF combustion system. The collection of radiant heat in the waterwall boilers on most mass burn and RDF systems results in much lower temperatures of the flue gas entering the boiler (<800~ compared to the temperature of the flue gas stream entering the boiler of a two-stage system directly from the secondary combustion chamber (>900~ The higher thermal gradient in the two-stage systems results in a much higher degree of deposition occurring due to thermophoresis and diffusiophoresis impaction than any of the other mechanisms. As a result, the fused material becomes a condensed, compacted layer which is too hard to be removed via conventional online cleaning operations. The particulate loading in the flue gas stream from a two-stage system is much lower than the other systems. Consequently, the layer of fused material on the leading edge of the boiler tubes thickens mostly through continued thermophoresis and diffusiophoresis impaction rather than inertial impaction. The hardening of the fused layer is further compounded as the layer thickens, due to the increased temperature of the outer leading edge which increases the potential for melting and fusing of material (Glen and Howarth, 1988). The built-up layers also increase the surface area disturbing the gas stream, thereby increasing the turbulence and enhancing deposition of particulate via interception and diffusion (Brownian and eddy) impaction mechanisms. The overall result is a substantial build-up of material within a relatively short operating period. Two-stage systems are generally shut down for boiler maintenance once every 8 to 12 weeks, whereas mass burn or RDF combustion systems may operate normally for up to a year or more prior to scheduled cleaning maintenance of the heat recovery units. Since the build-up of hardened layers in the latter two types of systems is relatively thin (e.g. between 3 and 10 mm in a modern mass burn incinerator), even after long periods of operation these layers are typically left on the tubes as a coating to protect against further corrosion (Vehlow, 1991 ). Another unique characteristic of boiler ash from two-stage systems is that the material is generally more tacky or sticky in nature than ashes from large mass burn or RDF combustion systems, especially on the trailing edges of the heat transfer surfaces. This stickiness is mainly due to the enhanced sulphate concentrations in the residues (see below and Chapter 8). 10.2 PHYSICAL CHARACTERISTICS 10.2.1 Particle Size Distribution As mentioned earlier, the size of particles captured in heat recovery systems can vary widely. For example, the particle size distribution data determined by weight for boiler

423 ash collected from a modern mass burning incinerator is given in Table 10.3. The optically determined frequency of the particles for the different size ranges is also given. The data indicate the most prevalent particle size of friable ash by population is the size range 5 to 20 microns, whereas the greatest proportion of particle mass was in the 40 to +200 size fraction. This inverse relationship of the lowest population of particles representing the greatest proportion by weight corroborates observations by Glen and Howarth (1988) (Figure 10.1 ). Table 10.3 Particle Size Distribution Data of Boiler Ash from a Mass Burning Incinerator % of Particles in Each Size Range (microns) 5 - 10 10 - 20 20 - 40 40 - +200 Test (on 3 reps) 2-5 By Mass

0 0 0

0 1 0

4 2 2

11 11 0

85 86 97

By Frequency

64 76 80

9 12 9

16 5 7

5 4 0

5 4 4

Average Mass

0

<1

3

7

90

10

10

3

3

Average 74 Frequency WASTE Program, 1993

The particle size distribution of boiler ashes can be influenced by a number of factors including: incinerator type operating conditions configuration of the boiler waste feed composition. The type of combustion system will influence the range of particle sizes normally carried over to the boiler. Because of the semi-pyrolytic combustion conditions in the primary chamber of a two-stage system, the flue gas velocities in the primary chamber are relatively low and granular sized particles are not likely to be carried over to the boiler. Conversely, because of the semi-suspension combustion of RDF, there is a greater potential for carry-over of a greater proportion of larger sized particles than in a mass burn system.

424 Figure 10.1 Particle Size Distribution of Boiler Ash from a Mass Burn Incinerator 2

%

iii

-

e.0_

Mass Fraction

.

r

Number Fraction

u_ m r

1..

o L

9

E z

9 .

/

I

i

0.1

I

i

1

/

/

/

/

/

/

/

/

\

I

10

Dp/um

Glen and Howarth, 1988 Changes in operating conditions can also affect the particle size distribution. For example, in mass burn systems, an increase in the under-fire air pressure and decrease in the depth of the fuel bed due to very wet garbage can increase the potential for carry-over of larger sized particles compared to normal operating conditions. Conversely, a decrease in the under-fire air to control the combustion of relatively dry garbage would result in a smaller proportion of larger sized particles being present in the boiler ash. The configuration of the boiler will also influence particle size distribution. By increasing the turbulence in the path of the flue gas stream, the potential to increase the proportion of finer sized particles increases due to interception and diffusion (Brownian & eddy) impaction mechanisms. Finally, the proportion of fine sized or dust-like material (<200 microns) in the waste can influence the ratio of particle sizes carried over in the flue gas stream. For example, higher ratios of fines, soil or construction debris versus organic materials (cellulose) would increase the potential ratio of finer to coarser particles collected in the boiler, since this type of material is readily suspended in the furnace by the turbulence created by primary and secondary air injection.

10.2.2 Particle Morphology During the WASTE Program, qualitative analysis was conducted on some of the boiler ash samples using scanning electron microscopy (SEM) to determine the morphology of the particles. The particles were categorised into the five types listed in Table 10.4.

425 Table 10.4 Summary of Morphological Characteristics of Boiler .Ashes CATEGORY DESCRIPTION Fused Spheres spheroids of various colours with particulate or gaseous inclusions Crystals

irregular in shape, similar to soil-like particles of calcite or quartz

Polycrystallines

dense agglomeration of irregular shaped particles

Opaques

single, large irregular shaped particles (<300 microns)

Char Black fibrous particles WASTE Program, 1993 In addition to the identification exercise, the relative percentage of the different particle types was determined optically (Table 10.5). The most common particle types were the polycrystalline and opaque irregular shaped particles, whereas crystals, fused spheres and particles of char were the least common forms. Glen and Howarth (1988) report SEM examination of samples indicated visibly discrete amorphous particles as well as spherical silicate and alumino-silicate particles in the 1 to 30 micron size range. These discrete particles were often encrusted with, or embedded in, a non-silicate matrix, representing much larger agglomerated particle sizes. Table 10.5 Estimate.d Relative Percentage of Parti.cle Types in Boiler Ash MORPHOLOGY Estimated. Relative % Size Range Fused Spheres Crystalline Polycrystalline Opaque Char WASTE Program, 1993

15 10 30 30 15

2 - 100 5 - 140 5 - 200 5 - 300 10 - 400

10.3 CHEMICAL CHARACTERISTICS Although there is a relative paucity of data on boiler ash in comparison to other residue streams, there are several sources of data which provide an excellent indication of boiler ash composition. For example, in addition to the data provided in Chapter 8, the NEL study generated a small database differentiating the composition of captured particulate matter versus the composition of fused deposits and dust-like deposits on the heat transfer surfaces. Furthermore, the data from the NITEP Program provide an excellent illustration of how the incinerator type influences boiler ash composition

426 (Environment Canada, 1985, 1988 & 1992; Bridle and Saweil, 1985; Sawell and Constable 1988, Sawell et al., 1989a & b, 1990a & b, 1991, 1992). The WASTE Program data also provide an indication of how the waste stream can affect ash characteristics (WASTE Program, 1993).

10.3.1 pH and Acid Neutralisation Capacity The pH of boiler and economiser ashes from three different types of incinerator systems are presented in Figure 10.2. The box plots have been generated using data from the NITEP and WASTE Programs and are based on the pH of an ash sample mixed in distilled water at a liquid-to-solid ratio of 20:1. The wide variability in the data can be attributed to the difference in incinerator systems. Two-stage systems tend to produce boiler and economiser ash which is neutral to mildly acidic in nature, whereas the RDF and mass burn systems generate ash which is highly alkaline (see Chapter 8). This is a direct result of the degree of enrichment of sorbed SO2 (sulphation) on the outer layers of particles from the flue gas stream (length of time the ash is retained in the system prior to removal), and the prevailing temperature of the flue gas stream within the individual heat recovery units (see Section 10.3.3). It is also interesting to note the pH of the combined boiler/economiser ash collected during one of the NITEP studies was slightly less alkaline than the ashes from the other mass burn system. This is probably due to the longer retention time of the ash during the eight-hour test runs used during the NITEP study versus the four-hour test runs used in the WASTE Program study. The degree of sulphation will also greatly influence the acid neutralisation capacity (ANC) of the ash. The ANC curves for heat recovery system ashes from a modern mass burn and semi-suspension combustion system are compared to those of bottom ash and fabric filter residue in Figure 10.3. These naturally alkaline ashes have ANCs similar to that of bottom ashes. In comparison, two-stage system boiler residues have very little buffering capacity [0.95 meq/g to endpoint pH of 4.4 (Sawell et al., 1989)] due to the prolonged retention and subsequent sorption of SO2.

10.3.2 Solubility in Water The overall solubilities of boiler and economiser ashes in water from the different types of incinerator systems are given in Figure 10.4. Once again, the data has been gleaned from the NITEP and WASTE Programs. The box plots indicate the ashes from two-stage systems are much more soluble than those from either mass burn or semisuspension combustion systems. This is directly related to the prolonged retention time of the ash in the two-stage system and is consistent with the discussion on ash deposition in high thermal gradient heat recovery systems. The moderately high solubilisation of the combined boiler/economiser ash from the NITEP mass burn study compared to the WASTE Program study is due to the longer retention time of the ash during the NITEP study (see Section 10.3.1).

427 Figure 10.2 pH of Boiler and Economiser Ash from Three Different Types of Incinerator Systems pH

13O

, ti-

~

I

@

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

A = 2-Stage, Boiler B = 2-Stage, Economiser

9-

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser

-I-

7 -

C = Mass Burn, Boiler D = Mass Burn, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

E = RDF, Economiser

(Sawell et al., 1991)

5 A

B

C

D

"

E

Figure 10.3 Comparison of Ash Stream Acid Neutralisation Capacity Values from Mass Burn and Semi-Suspension Combustion Systems 14 12 10

r

8

E) v

-I6 Q.

0

2

bottom

4

§

6 ANC (meq/g of Ash)

boiler

---

8

10

economiser - - = - fabric filter

Adapted from Sawell et al., 1989b; WASTE Program, 1993

12

428 Figure 10.4 Water Solubility of Boiler and Economiser Ash from Three Different Types of Incineration Systems

60 -

'~ L o s s "-'L---

,,

A = 2-Stage, Boiler B = 2-Stage, Economiser

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

C = Mass Burn, Boiler D = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

E = RDF, Economiser

(Sawell et al., 1991 )

40-

_L

20-

[

I I

1 _.__,._.,B

I

0 A

B

C

O

E

10.3.3 Chemical Composition Compositional data from the NEL study is summarised in Table 10.6 and indicates the discernable difference between the chemical composition of flue gas particles, the fused upstream layer and the downstream layer collected on the heat transfer surfaces. The particulate matter represents the composition of a raw particle entrained in the flue gas prior to any deposition or condensation mechanism occurring. In relation to the raw particulate matter, both the upstream and downstream layers become substantially enriched in sulphate, chloride and potassium, and slightly enriched in zinc (bolded in Table 10.6). One of the reasons for the enrichment is the prolonged exposure of the upstream and downstream deposits to the flue gas, which results in sorption of acid gases (SO2 and HCI) and other volatile compounds. The enrichment of sulphate in downstream deposits is greater than the enrichment in the upstream deposits due to the increased available sorptive surface area of the dust-like deposits. Note the increase in mass of both type of deposits resulted in a corresponding decrease in the aluminum oxide and silicon oxide content, and a slight decrease in the phosphate, calcium oxide and chromium oxide content of the deposits (particulate born).

429 Table 10.6 Statistical Summary of Foulin(:j Deposit Chemical Composition Data Particulate Upstream Deposit Downstream Deposit Mean % Mean % Std. D e v . Mean % Std. Dev. AI203 14.0 6.41 2.59 5.75 1.32 SiO2 28.1 11.1 4.69 12.9 2.87 PO3 2.36 1.20 0.73 1.47 0.75 SO3 7.25 22.1 9.24 26.9 6.54 CI

2.51

11.1

K20 3.72 10.3 CaO 26.7 23.8 Cr203 0.64 0.57 Fe203 5.2 3.40 ZnO 2.42 4.80 Adapted from Glen and Howarth, 1988

9.16

5.44

5.03

3.66 5.57 0.54 1.56 2.32

11.7 18.2 0.55 5.21 6.44

4.65 7.58 0.75 2.39 4.94

As discussed in Chapter 8, the sorption of SO2 in heat recovery system ashes is a more important reaction than the sorption of chloride since chloride has a lower dew point temperature. This is evident from the order of magnitude difference in the concentrations of chloride in the heat recovery system ash (Figure 10.5) in comparison to the concentrations of sulphate measured in the same ash samples (Figure 10.6). The concept of sulphate enrichment, or sulphation, of various compounds in boiler ashes which have prolonged exposure to the flue gas stream is supported by the data given in Figure 10.6. The box plots include data gleaned from the NITEP and WASTE programs, and are grouped by incinerator type. These data are compared on the basis of relative sulphate content as indicated by leaching with distilled water. As mentioned above, sulphate enrichment in boiler ash from two-stage systems is greater than soot blown boiler ash in mass burn and RDF combustion systems due to the greater retention time of the ash in the system. The 390 mg/g of sulphate released per gram of boiler ash from the LVH two-stage facility was the most highly enriched ash due to the fact that all of the boiler ash was held in the heat recovery system for four days prior to soot blowing to accumulate a sufficient sample size, whereas the other samples were collected over four-hour test periods. The same trend can also be seen in the data in Figure 10.7 in which the releases of sulphate (in IJg/g of ash) from boiler ashes in distilled water are compared for the different test runs of the WASTE Program. The data indicate the sample collected during the first test run of the series contained about five times the amount of sulphate

430 Figure 10.5 Relative Chloride Content of Boiler and Economiser Ashes from Three Different Incineration Systems

50000

E. c l -

A = 2-Stage, Boiler B = 2-Stage, E c o n o m i s e r (Bridle and Sawell, 1986; Sawell et al., 1989; Sawell and Constable, 1 9 9 0 )

40000 -

L

o

30000 -

C = M a s s Burn, Boiler ( W A S T E Program, 1993)

O C

o

D = M a s s Burn, E c o n o m i s e r (Sawell and Constable, 1 9 8 8 )

20000 -

E = RDF, Econoiser (Sawell et al., 1 9 9 1 )

w

u

c0

tO000 -

--rI"

0 -

A

" I B

, (~

D

o,

,

E

Figure 10.6 Relative Sulphate Content of Boiler and Economiser Ashes from Three Different Incineration Systems

400000 -

300000 -

O

200000 -

C O L C

w

u

c o

100000 -

t0000

-

e so4 A = 2-Stage, Boiler B = 2-Stage, Economissr

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

C = Mass Bum, Boiler D = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

E = RDF, Economiser

(Sawell et al., 1991)

__I j ' .

,

431 Figure 10.7 Comparison of Sulphate Release from Boiler Ash 80000-

- - i - - FF1 E3

--+-

FF2

~

FF4

X

60000

Z c~

4oooo 1 20000

Cycle

WASTE Program, 1993 as the subsequent test run samples. This is because the conditions under which the sulphate enriched sample was taken differed from the other tests. Although a soot blow cycle was initiated prior to starting and finishing a test run, the normal soot blow cycle at the facility was 12 hours. It was hypothesised since the soot blow used to finish the first test run was only four hours after the end of the previous cycle, it dislodged a large quantity of encrusted ash from the boiler tubes which would not have been dislodged in normal operation. In addition, the method of sampling involved manually removing some of the ash from the boiler hopper chute, effectively cleaning out the hopper ducts. The material collected contained ash which had been retained in the boiler for a much longer period of time than the subsequent four hour test runs, and therefore had undergone a much higher degree of sulphation than the other samples.

10.3.4 Heavy Metals The concentrations of the heavy metals chromium and nickel in boiler and economiser ashes are presented in Figures 10.8 and 10.9. The box plots represent a compilation of data from the NITEP and WASTE programs, as well as data from the regulation compliance testing conducted at the Peel Incinerator Facility near Toronto, Ontario (PRRI, 1992).

432

Figure 10.8 Comparison of Chromium Content in Heat Recovery System Ashes from Three Different Incineration Systems Cr' 1400 -

3

t050

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser

-

E

"-I

E = Mass Bum, Boiler F = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

G = R D F , Economiser

(Sawell et al., 1991)

S

O r tr

,,,-

Q tO ..e-I 4.,a s 4.a t'. =J u tro

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

A = 2-Stage, Boiler B = 2-Stage, E c o n o m i s e r

I

700 -

-I-

I

I I I

-]-

T

1

8

o I

350 -

I

I

_L

I

.J_

-T-

0 A

B

C

D

E

F

G

Figure 10.9 Comparison of Nickel Content in Heat Recovery System Ashes from Three Different Incineration Systems

2000

-

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

A = 2-Stage, Boiler B = 2-Stage, Economiser ol t500

-

w-i Q u z o

c o ..P!

t000

-

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser E = Mass Bum, Boiler F = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

G = RDF, Economiser

(Sawell et al., 1991)

J_

c

o tO

500 -

U o

o

o

o

T I

0 -

A

B

C

D

_.L_--

E

I

I

433 Based on the discussion in Chapter 8, the partitioning of chromium and nickel during incineration indicates these metals will be present mostly in the form of solid particulate matter carried over from the combustion chamber by the flue gas stream. Since the only mechanisms available for deposition of solid particulate are gravitational, inertial or interception impaction deposition, the concentration of these metals in the ash is a function of the carry-over of particulate matter from the combustion chamber and hence is influenced by the composition of the waste feed, the incineration system used and the incinerator operating conditions. Lower concentrations of these metals can be expected if the ash is allowed to remain exposed to the flue gas stream. This is based on the relative increase in mass of the ash due to the enrichment of sorbed volatile compounds onto the boiler ash particles. The lowest concentrations of chromium and nickel were measured in the samples from the two-stage systems, ranging from 130 to 385 IJg/g for chromium and 52 to 387 IJg/g for nickel. The next lowest range of concentrations for both metals was the combined boiler/economiser ash from the mass burn system. This is consistent with the low carry-over of particulate from the primary combustion chamber (two-stage systems) and the increase in mass due to enrichment with volatile compounds (both systems). Although the two-stage chromium and nickel data are closely grouped, the variation would increase slightly if the PEEL data (C & D) had been determined via hydrofluoric acid/aqua regia/peroxide digestion instead of just an aqua regia digestion. These metals are typically associated with silicate matrices, and therefore it is reasonable to expect the "true total" concentrations for both these metals would be higher than expressed (see Chapter 7.2). This is due to the Peel two-stage incinerators being modern units and their operation has been modified so some primary air is injected into the bed of the furnace to enhance burnout. As a result, the particulate loading in the flue gas stream is slightly higher than Ioadings observed with older two-stage systems. The highest concentrations of chromium and nickel were measured in the RDF combustion system economiser residues. This is due to the high particulate loading to the flue gas stream caused by the semi-suspension combustion conditions. 10.3.5 "Volatile" Metals

The concentrations of the volatile trace metals cadmium, lead and zinc measured in boiler and economiser ashes are presented in Figures 10.10 to 10.12. The data were taken from the same studies as the heavy metals data. The partitioning of volatile metals in the incinerator is a function of temperature (see Chapter 8) and hence type of incinerator and operating conditions. Based on the given data, boiler and economiser ashes from two-stage systems contain the highest concentrations of volatile metals which can be attributed to the low particulate loading of the flue gas stream in two-stage systems and a greater degree of enrichment due to prolonged exposure of ash to the flue gas stream.

434 Figure 10.10 Comparison of Cadmium Content in Heat Recovery System Ashes from Three Different Incineration Systems Cd 1300

~o

-

g75 -

"

T

I,_LI

S U

C o .4J r r 4./ C

325 -

U

tO r

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser

I• 1

650 -

,9-

o

A = 2-Stage, Boiler B = 2-Stage, Economiser

E = Mass Bum, Boiler F = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

G = RDF, Economiser

(Sawell et al., 1991)

_3._

o "-I-_ 0 -

A

B. . . . .

C

D

....E

F

G

Figure 10.11 Comparison of Lead Content in Heat Recovery System Ashes from Three Different Incineration Systems A = 2-Stage, Boiler B = 2-Stage, Economiser

i~ Pb 0

36000 -

m

o

c

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser

!

27000 -

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

o o

E = Mass Burn, Boiler F = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

G = RDF, Economiser

(Sawell et al., 1991 )

_1_.

tB000 -

o

L C U

c o

9000 -

0 -

_.LA

B

C

D"

E

F

6

435 Figure 10.12 Comparison of Zinc Content in Heat Recovery System Ashes from Three Different Incineration Systems A = 2-Stage, Boiler B = 2-Stage, Economiser

Zn BOO00

-

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

C = Peel 2-Stage, Boiler (PRRI, 1992) D = Peel 2-Stage, Economiser

60000 u C

E = Mass Burn, Boiler F = Mass Burn, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

G = RDF, Economiser

(Sawell et al., 1991)

N Q

g

40000

-

'

~

C U

cQ

o

I

20000

0 -

A

B

C

I __L_1

D

E

F

G

The data indicate both the two-stage and mass burn economiser ashes are more enriched in these metals than the boiler residues. This is a function of the lower temperature regimes in the economiser compared to the boiler, hence greater potential for condensation of these metals onto the surfaces of the impinged ash particles. 10.3.6 Organic Contaminants

As mentioned in Chapter 8, the last sections of the heat recovery system provide potential sites for formation of dioxins and furans. The accepted reactions mechanism is oxychlorination from particulate carbon which occurs at temperatures between 300 and 450~ Consequently, the ash collected from the hotter sections of the boiler contain very little if any detectable quantities of PCDD/PCDF, whereas the back end of the boiler and the economiser tend to contain higher concentrations. PCDDIPCDF

The formation of PCDD/PCDF was confirmed by two studies. One study was conducted on boiler ashes collected at temperatures above and below 400~ Ash collected from zones operating at >400~ contained very small quantities of PCDD (1.1 IJg/kg for the sum of all congeners), whereas the below 400~ samples contained concentrations of about 600 IJg/kg PCDD (Vogg, 1987).

436 A second, more detailed study was conducted at an Italian incinerator where ash was collected from eight separate sections of the boiler, including the inlet (temperature of 950~ and outlet (temperature of 270~ It indicated the concentrations of PCDD and PCDF increased with decreasing temperatures (Mariani, 1990). Detailed data are given in Chapter 8.6.2 along with a more comprehensive description of formation mechanisms. The range of PCDD/PCDF (total) concentrations in boiler and economiser ashes from the different types of incinerator systems are given in Table 10.7. The results indicate the type of incinerator system is less of an influencing factor on the concentrations of PCDD/PCDF than the following: the 9 the 9 the 9 the 9

combustion conditions of the incinerator particulate carbon loading in the flue gas stream ash deposition rates in the back end versus the front end of the boiler particle size distribution of the ashes.

Table 10.7 Summary of Total PCDD/PCDF Concentrations in Heat Recovery System Ashes from Different Types of Incinerator Systems (ng/g) System Ash Type Total PCDD Total PCDF Two-stage

Boiler Economiser

60 - 150 30 - 170

6 - 20 10 - 40

Mass Burn

Boiler/Economiser

0.09- 183

0.03- 120

Semi-suspension Economiser BD - 0.433 BD - 1.83 BD = below detection limit of 0.01 ng/g Environment Canada, 1985, 1988 and 1992; Vogg et al., 1987; Hiraoka et al., 1989; Mariani et al., 1990; Nottrodt et al., 1990; Benfenati et al., 1991 The first two points are interrelated. Modern incinerators, irrespective of type, operate with much greater efficiency and generate less PCDD/PCDF than older incinerator systems. Poor combustion conditions result in less efficient destruction of precursor compounds in the waste and result in greater concentrations of organic precursors in the flue gases, enhancing PCDD/PCDF production. Incinerator systems which generate a greater proportion of boiler ash in the high temperature area of the boiler will contain a relatively low mean concentration of PCDD/PCDF in the boiler ash, whereas higher mean concentrations will be realised in systems where a greater proportion of ash is collected in the 300 to 450~ area. The particle size distribution acts to influence all of the other factors in that the potential for PCDD/PCDF formation will increase with decreasing particle size due to the greater surface area to volume ratio of the particles.

437 It is important to note the data ranges given in Tables 10.7 through 10.9 include data from the NITEP program collected under "poor" and "good" operating conditions, and that the maximum levels are typically related to poor conditions.

Other Chlorinated Organics

A summary of the concentrations of chlorobenzenes (CB), chlorophenols (CP) and polychlorinated biphenyls (PCB) in heat recovery system ashes are given in Table 10.8. These compounds are formed in the same manner as PCDD/PCDF compounds. Consequently, measured concentrations are typically influenced by the same factors as listed above. Table 10.8 Summary of Total Chlorobenzenes, Chlorophenols and PCBs in Heat Recovery System Ashes from Different Types of Incinerators (ng/g) System Ash Type Total CB Total CP Total PCB Two-stage

Boiler Economiser

70- 940 50 - 240

230- 540 <0.1 - 10

<0.1 - 50 <0.1 - 30

Mass B u r n

Boiler/Econ

259 - 1570

59 - 110

<0.1 - 24

<0.1 - 15

<0.1

SemiEconomiser <0.1 suspension Environment Canada, 1985, 1988 and 1992

Polycyclic Aromatic Hydrocarbons (PAH)

A summary of PAH concentrations in heat recovery system ash is given in Table 10.9. Since PAH compounds are formed during semi-pyrolytic and pyrolytic processes, their presence in heat recovery system ashes generally provides a good indication of combustion efficiency. These semi-volatile compounds are typically carried over from the combustion chamber and their concentrations are mostly influenced by the particulate loading in the flue gas and the combustion conditions in the furnace. Table 10.9 Summary of PAH Concentrations in Heat Recovery Systems Ashes from Different Types of Incinerator Systems (ng/g) System

Ash Type

Total PAH

Two-stage

Boiler Economiser Boiler/Economiser

110 - 780 12 - 160 21 - 70

Mass Burn

Semi-suspension Economiser Environment Canada, 1985, 1988 and 1992

16 - 6,430

438 REFERENCES

Benfenati, E., G. Mariani, R. Fanelli and A. Farneti. "Synthesis and Destruction of PCDD and PCDF Inside a Municipal Solid Waste Incinerator." Chem0sphere 23, pp. 715-722, 1991. Bridle, T.R. and S.E. Sawell. NITEP Phase I Testin,q at the Prince Edward Island Ener.qy-From-Waste Facility, Assessment of Ash Contaminant Leachability. Internal Environment Canada Report, 1986. Environment Canada. The N.a.tional Incinerator TestinQ and Evaluation Pro.qram: TwostaQe Combustion (Prince Edward.. Island) Environment Canada Report EPS 3/UP/l" Vols. 1-4, September 1985. Environment Canada. The National Incinerator Testina and Evaluation ProQram: Environmental Characterization of Mass Bum Technolo.qy at Quebec City Environment Canada Report EPS 3/UP/5: Vols 1-7, June 1988. Environment Canada. The National Incinerator Testin.q and Evaluation Pro.qram" The Environmental Characterization of RDF Technolo.qy (Mid-Connecticut.) Environment Canada Report- Waste Management Series WM/14: Vols 1-6, March 1992. Ewart, W.R. "Obtaining Valid Fouling Data from Industrial Gas Streams." Publication C 118/88, National Engineering Laboratory, Glasgow, 1988. Glen, N.F. and J.H. Howarth. "Modelling Refuse Incineration Fouling", Publication C119/88, National Engineering Laboratory, Glasgow, 1988. Glen, N.F., J.D. Isdale, W.R. Ewart and J.H. Howarth. "Gas-side Fouling - The Limiting Factor in Recovering Energy From Waste?" Ener,qy Recovery Throu.Qh Waste Combustion Edited by Brown, Evemy and Ferrero. Elsevier Applied Science, 1988. Hiraoka, M. et al. "Control of Dioxins from a Municipal Solid Waste Incinerator Chemosphere 19, pp. 323-330, 1989. Mariani, G., Eo Benefati and R. Fanelli. "Concentrations of PCDD and PCDF in Different Points of a Modem Refuse Incinerator." Chemosphere 21, pp. 507-517, 1990. Peel Resource Recovery Incorporated (PRRI). "Ash and Quench Water Testing Report." Report prepared for the Region of Peel, Brampton, Ontario, July 1992. Nottrodt, A., U. D(~wel and K. Ballschmiter. "The Influence of Increased Excess Air on the Formation of PCDD/PCDF in a Municipal Waste Incineration Plant." Chemoshpere 20, pp. 1847-1854, 1990.

439 Sawell, S.E. and T.W. Constable. NITEP Phase liB: Assessment of Contaminant Leachability from the Residues of. a Mass Burnin.(:] .Incinerator. Environment Canada, EPS Manuscript Series IP-82: Vol. VI, 1988. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testi0.q and Evaluation Program: Evaluation of Contaminant Leachability From Residues.Collected at a Refuse Derived Fuel Municipal Waste Combustion Facility. Environment Canada Report, Manuscript Series, IP-96, 1989a. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testin..c] and Evaluation Pro.qram: Characterization Of Residues from a Modular Municipal Waste Incinerator with Lime-based Air Pollution Control. Environment Canada Report, Manuscript Series, IP-101, 1989b. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testinqand Evaluation Pro,qram: Characterization of Residues from a Mass Burnin,q Municipal Waste Incinerator with Lime-basedAir P.ollution Control (Burnaby, B.C.). Environment Canada Report, Manuscript Series, IP-110, 1990a. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testing and Evaluation Pro(:iram C.haract.erization of Residues from a Two-sta.qe Incinerator with Rotary Kiln (3M Canada). Environment Canada Report, Manuscript Series, IP-119, 1990b. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testingand Evaluation Program: Characterization of Residues...fr0m a Refuse Derived Fuel Combustion Facility (Mid-C.onne.c.ticut). Environment Canada Report, Manuscript Series, 1991. Sawell, S.E. and T.W. Constable. The National Incinerator Testin.q and Evaluation Program: A Summary of the Characterization and Trea..tment Studies on Residues, From Municipal Solid Waste Incineration. Environment Canada Report (in press), " 1992. Vehlow, J. and J. Schneider. "Hold Up of Fly Ash in the Boiler of an MSW Incinerator." Unpublished report of the Kernforschungzcentrum Karlsruhe, 1991. Vogg, H., M. Metzger and L. Stieglitz. "Recent Findings on the Formation and Decomposition of PCDD/PCDF in Municipal Solid Waste Incineration" Waste Mana.qement and Research 5, pp. 285-294, 1987. Waste Analysis, Sampling, Testing and Evaluation (WASTE) Program. Final Report of the Mass Burn MSW Incinera.tion Study (Burnaby, .B.C.) Report Prepared for Environment Canada, U.S. Environmental Protection Agency and the International Lead Zinc Research Organization, Vols. 1-4, 1993.

This Page Intentionally Left Blank

441

CHAPTER 11 - CHARACTERISATION OF AIR POLLUTION CONTROL RESIDUES 11.1 INTRODUCTION The most dramatic change in incinerator technology over the past decade has been related to the air pollution control (APC) systems. The desire to reduce the emissions of undesirable contaminants to the atmosphere has resulted in the development of a wide variety of APC systems. In turn, this has affected the characteristics and quantity of residues (and liquid waste streams) generated from these systems. In Europe, both dry/semi-dry and wet scrubber systems are commonly in use, whereas in North America, only dry and semi-dry APC systems have gained widespread use.

The information presented in this chapter builds upon the concepts discussed in Chapters 4, 8 and 10. The amount of information available on the properties of APC residues from MSW incineration is not as extensive as that available on bottom ash (see Chapter 9). This is probably partly due to the fact that APC residues generally are "newer" than bottom ash, and partly due to the apparently more limited potential of APC residues for utilisation. Therefore, most of the available information on APC residues is related to disposal or to APC system performance. In addition to illustrating the different physical and chemical characteristics of the various types of APC system residues, the contrast between residue composition and properties in relation to the APC process, as well as type of incinerator technology and operation are discussed. As a final note, all dry and semi-dry APC system residues referred to in this chapter are from mass burn incinerators and contain fly ash unless otherwise indicated. 11.1.1 Terminology

Historically, the term "fly ash" has been used as a generic descriptor for all types of finely-sized ash and sorbent material collected in APC systems. However, due to the very different chemical and physical properties of the different residue streams generated in an APC system, it is prudent, and much more accurate to use specific terms for residue streams to avoid any potential confusion. Generally, the residue is best described by the type of unit where the residue has been collected. Fly Ash As mentioned in Chapter 10, the specific term "fly ash" refers to the particulate matter carried over from the combustion chamber and removed from the flue gas stream prior to the addition of any type of sorbent material (such as at facilities without acid gas cleaning systems or from particulate removal systems which are located prior to the acid gas cleaning equipment). Although fly ash is an acceptable term, it is more accurate to identify the ash based on the type of unit the ash was collected in, such as cyclone arrestor ash or ESP ash. For acid gas cleaning processes without upstream removal of particulate, the term fly ash is used to describe the portion of the residue corresponding to the original particulate matter carried over from the boiler and is

442 generally mixed with the other components of the APC residue (reaction products and surplus reactants).

Dry System Residues

"Dry" lime-based APC systems operate by injecting powdered lime into the flue gas stream, either in conjunction with humidification of the flue gas stream, or alone. The injection is performed in a reactor chamber (dry reactor- DR) before the particles entrained in the flue gas stream pass into a separate dust collection unit, such as an ESP or fabric filter (FF). The residue generated from these APC processes (DP residues) usually consists of a fine-powdered mixture of fly ash, reaction products (predominantly calcium chloride) and unreacted lime, and is generally referred to as dry scrubber (DS) or fabric filter (FF) residue. Since there is usually no upstream removal of particulate, in most cases the fly ash can make up a substantial proportion of the collected residues. In some facilities, a portion of the collected residue may be recycled as a substitute for a portion of the injected fresh lime in order to improve the reaction stoichiometry. In some cases, sodium sulphide or activated carbon is added to the lime reagent in order to enhance mercury removal. In all of cases, it is important to convey as much information as possible when identifying the residue stream.

Semi-dry System Residues

"Semi-dry" APC systems operate in much the same manner as dry systems, however, the lime is usually mixed with water and is injected into the flue gas stream as a slurry. Residues from semi-dry processes (SDP residue) resemble the residue from dry processes, although there is usually a lower content of unreacted lime due to a better reaction stoichiometry. Although it is not common, part of the collected residue may be recycled through the reactor in order to improve the reaction stoichiometry. In some cases sodium sulphide or activated carbon is added to the lime reagent in order to improve mercury removal.

Wet Scrubber Sludge

Wet process (WP) residue is a sludge resulting from the treatment of the wastewater from a one-stage or two-stage wet scrubber system with lime and organic sulphides (often trimercaptotriazine or TMT). The wet scrubber process produces a stream of fly ash which is collected upstream of the acid gas cleaning process (either a cyclone or ESP unit), and a stream of acid scrubber effluent which is contaminated with inorganic salts (predominantly calcium and/or sodium chloride) and trace elements/heavy metals. The content of trace elements/heavy metals in the saline scrubber effluent must be reduced prior to discharge (to a sewer system or a surface water body), e.g. by lime precipitation and subsequent polishing with TMT. In a two-stage wet scrubber process, the scrubbing liquid in the second scrubbing stage is kept nearly neutral, usually by the addition of sodium or calcium hydroxide, in

443 order to enhance SOx removal efficiency. In the latter case, gypsum can be produced from this residue stream. Direct mixing of the two wastewater streams may cause problems due to precipitation of calcium sulphate (gypsum) and in some cases are treated or pre-treated separately. The (dewatered) sludge from the treatment of the wastewater stream(s) is often mixed with the fly ash prior to disposal. This chapter presents information on the properties of the wastewater treatment sludges from the wet scrubber process both with and without fly ash. Examples are also given of the composition of treated and untreated wet scrubber effluents. 11.2 MAJOR FACTORS INFLUENCING THE CHARACTERISTICS OF APC RESIDUES Although the APC technology is the single most important factor influencing the characteristics of the APC residues, several other factors also play a role. Beginning with the type of waste feed to the incinerator, everything which occurs prior to, and during, the collection of the APC residues may affect the characteristics and quantities of the captured residues (see Chapters 3, 4 and 8). Typical ranges of the amounts of residues and wastewater produced from the three major APC processes at mass burn systems under Danish/German conditions (typical averages: 5.2 Nm3 of stack gas per kg of waste at 10% 02, removal of 700 mg HCI/Nm3) are presented in Table 11.1. Table 11.1 Approximate Quantities of Residues and Wastewater Produced per Tonne of Refuse APC Process Waste Stream/Residue

Unit

Dry

Semi-dry

Wet

Fly ash

kg DM

(10 - 30)

(10 - 30)

10 - 30

Dry residue, including fly ash

kg DM

20 - 50

15 -40

Sludge from wastewater treatment

kg DM

1-3

Treated wastewater

m3

0.3 - 0.5

Chloride content in kg DM treated wastewater DM 9Dry matter Hjelmar et al., 1990; Rasmussen et al.., 1993; Vehlow, 1993

3.5- 10

444 Another major factor which influences the characteristics and volume of APC residue collected is the type of incinerator system. For example, average uncontrolled particulate levels in the flue gas stream of RDF stoker units have been measured at >3,000 mg/Rm3 (dry @ 25~ 101.3 kPa & 11% 02) prior to a fabric filter (Environment Canada, 1991), whereas particulate levels of <2,200 and <300 mg/Rm 3 have been recorded for mass burn and two-stage systems (CCME, 1989). Consequently, the volume of APC residue collected in RDF stoker systems is far greater per tonne of refuse burned than in mass burn systems, which in turn, is greater than in two-stage combustion systems. These factors should be considered when attempting to compare data on APC residues from different types of incinerator and APC systems.

11.3 PHYSICAL CHARACTERISTICS OF APC RESIDUES 11.3.1 General Appearance and Behaviour Freshly collected fly ash, as well as dry and semi-dry process residues, appear as fine, dusty materials with practically no water content. The colour may vary from almost white through various shades of grey and brown to almost black, depending on the composition and (for dry and semi-dry process residues) the content of fly ash and how well it has been combusted. Since CaCI2 is the major constituent of dry and semi-dry process residues, the material has hygroscopic properties, and therefore will gradually sorb moisture from the ambient air. Residues with high contents of calcium chloride may even "liquify" due to excessive uptake of water vapour. The residue from the wet scrubber process(es) appears as a wastewater treatment sludge which is usually dark brown in colour. The water content may vary widely and depends upon the dewatering equipment used (e.g. filter press, band filter, centrifuge). Typical water contents of filter cakes are 65 - 75% (w/w), (Rasmussen et al., 1993). Effectively dewatered filter cakes are easily handled whereas insufficiently dewatered sludge may create handling problems, in some cases even if it is combined with the fly ash collected prior to the wet scrubber.

11.3.2 Particle Size Distribution Typical particle size distributions of fly ash and dry/semi-dry APC process residues from mass burn incinerators are shown in Figure 11.1. The dry and semi-dry process residues generally contain a higher proportion of fine material than fly ash without acid gas cleaning residue. Because fabric filters are more efficient at removing submicron size particles than ESPs, the proportion of fine material in fabric filter residues will be higher than in residues collected in ESPs (e.g. Environment Canada, 1986 and 1988; Carlsson, 1988; Hjelmar, 1992). Results have also indicated that semi-dry APC systems (lime slurry spray) generate residues with a finer particle size distribution than

445 dry systems where lime powder is injected (Hjelmar, 1992). Determination of particle size distribution is usually not relevant for the sludge-like residues from the wet scrubber APC processes. Figure 11.1 Typical Particle Size Distribution for APC Residues ~,~', 9 ~....

100 "O

v

80

.2 a.

"~

40

/

e--

~-

FA (ESP)

-'/ ~zI

60

0

20

/ f"

O

]"

/

S D + F A (ESP)

/ S D + F A (FF)

,J

/

0 0.001

0.01

0.1 Particle size, d (mm)

1

10

As an example, the particle size distribution data determined by weight for dry scrubber reactor (DSR) and fabric filter (FF) residues collected from a modern mass-burning incinerator are given in Table 11.2. The optically determined frequency of the particles for the different size ranges is also given. The data indicate the particle size distribution (by mass) for the dry scrubber reactor residue is in the 40 to >200 micron size range, whereas the fabric filter residue is predominantly 10 to 40 microns in diameter. Furthermore, although the most prevalent residue particle size by frequency is the range 2 to 5 microns, the greatest proportion of particle mass was in the 10 - 40 micron size fraction. The inverse relationship of the lowest population of particles representing the greatest proportion by weight is similar to findings on boiler ashes (see Chapter 10). It should also be noted that the amount of residue collected in the DSR normally represents only a small fraction of the total amount of residue captured within the entire APC system. In addition to the type of APC system and the type of particulate removal system, the particle size distribution of APC system residues can depend on other factors including waste feed composition incinerator type operating conditions presence of a heat recovery system

446 Table 11.2 Particle Size Distribution of Dry Scrubber Reactor (DSR) and Fabric Filter (FF) Residues from a Mass Burning Incinera!or Test By Mass

By Freq

2-5 0 0 0

FF

12 2 11 70 50 43

7 2 10 13 18 0

81 17 79 0 15 0

0 78 0 9 9 29

0 0 0 9 9 29

92 87 90

3 5 5

5 5 5

0 3 0

0

0

0 0 0 88 0 16 0

DS

FF Avg Mass

% of particles in Each Size Range (microns) .5 - 10 .. 10- 20 20 - 40 40 - +200 0 0 11 89 0 2 11 87 0 0 11 89

Sample DS

DS FF Avg Freq DS FF WASTE Program, 1993

8 54 90

6 10 4

1

59 5 5

11

26 16 1

The proportion of fine sized or dust-like material in the waste can influence the ratio of particle sizes carried over in the flue gas stream. For example, higher ratios of fines, soil or construction debris may increase the ratio of finer to coarser particles collected in the APC system, since this type of material may be readily suspended in the furnace by the turbulence created by primary and secondary air injection. As mentioned earlier, the type of incinerator system would also act to influence the range of particle sizes normally entrained in the flue gas. Because of the semi-pyrolytic combustion conditions in the primary chamber of a two-stage system, the flue-gas velocities in the primary chamber are relatively low, hence granular sized particles are not likely to be carried over, even to the boiler. Conversely, because of the semisuspension combustion of RDF, there is a greater potential for carry-over of a greater proportion of larger sized particles than in a mass burn system. Changes in operating conditions can also affect the particle size distribution. For example, in mass burn systems, an increase in the under-fire air pressure and decrease in the depth of the fuel bed can increase the potential for carry-over of larger sized particles compared to normal operating conditions. Conversely, a decrease in the under-fire air would result in a smaller proportion of larger sized particles being present in the APC system.

447 The presence of a heat recovery system will also influence the particle size distribution of APC system residues. Since the heat recovery system will act to remove a great deal of the larger sized particles entrained in the flue gas via the deposition mechanisms mentioned in Chapter 10, the proportion of finer sized particles carried through to the APC system would increase.

11.3.3 Geotechnical Properties The geotechnical properties of the APC system residues are important both in relation to the physical feasibility of storage and disposal, and in relation to the potential environmental impacts resulting from these activities. Geotechnical properties include: 9 9 9 9

Compaction properties Density Hydraulic conductivity Compressive strength

Most of these properties are interrelated and will be addressed together. For example, higher compaction is often associated with higher density and lower hydraulic conductivity. Proctor tests carried out on dry and semi-dry APC system residues from mass burn incinerators indicated that densities of 1.22 - 1.43 tonnes/m 3 could be attained at optimum water contents (30 -40% (w/w)) (Hjelmar 1992), whereas uncompacted densities of dry and semi-dry APC residues can range between 0.65 and 0.75 tonnes/m 3 (Sawell and Constable, 1988; Sawell et al., 1989a, 1991). Kullberg & F,~llman (1989) and Hartl6n & Elander (1986) have investigated the geotechnical properties of ESP fly ash and semi-dry APC process residues. Their results are shown in Figure 11.2 and are compared to similar observations for bottom ash. Semi-dry APC residues began hardening immediately after it was moistened, and tests showed that the compaction properties deteriorated and became poorer the longer the moistened residue had been stored prior to compaction (see Figure 11.2 and Table 11.3). The data in Table 11.3 indicate that the compressive strength decreased and the saturated hydraulic conductivity increased with decreasing compaction. This may have significant implications for a landfilling operation for semi-dry/dry APC system residues. Field measurements (using double ring infiltrometers) on landfilled and compacted semi-dry APC process residues at various depths gave infiltration rates of 7.6 x 10.6 m/s - 1.3 x 10.7 (VKI 1992). The ranges of hydraulic conductivity of fly ash, dry and wet APC process residues, bottom ash (slag) and various other ashes observed by Kullberg et al. (1989) are presented in Figure 11.3.

448 Figure 11.2 Compaction Densities of Various MSWI Residues, Including Bottom Ash or Slag

1.8

I /t-~,

1.7

I',[ !

--

w a),n,0

1./, 84

~-

'

ORY FGCR (~AINING

'

FLY ASHI

t

J

/ 2~.r-

1.3

j.-

1.2 %

-~_

%,

1.1

1.0

i

I /~0.5h

1.6

j

.

,

i

0.9

_ 0

t

2'0 30 t.O 50 60 'mUtlER c0mrID~ I PI~R CENT m' wEH~rl" I

Kullberg and F~llman, 1989 Table 11.3 Influence of Storage Time on Compressibility, Unconfined Compressive Strength (UCS) and Hydraulic Conductivity of Semi-dr)/APC System Residue Time hrs

Dry density tonnes/m 3

0.5 1.55 4 1.47 24 1.40 48 1.34 168 1.23 Kullberg and F~llman, 1989

Water content weight %

UCS kPa

Saturated hydraulic conductivity m/s

20 19 19 19 19

2200 1380 2100 900 660

1x 1x 2x 2x 4x

10 .9 10 .9 10 .9 10 8 10 8

449 Figure 11.3 Saturated Hydraulic Conductivity (Permeability) of Various APC System Residues and Other Ash Types SATURATED HYDRAULIC CONDUCTIVITY (m/s) lO 3

lo ~

10'

ld'

ld'

Ill

10-" Fly

10~

ld '~

10"

Bottom A~ , I Loose!Fill

,,, Coal Coal

I

SAND

- A P 1Residue

--WetlScrubber

I

..,

~ludge

Ash

la

SILT

, I

CLAY,,,.SOILS

I

CLAY

Kullberg et al., 1989 Kullberg et al. (1989) have tested the unconfined compressive strength (UCS) of molds made from dry APC system residues mixed with 45% (w/w) water and stored at different temperatures. The results are presented in Figure 11.4 and indicate that the UCS increases with increasing temperature during the first 28 days at all temperatures between -18 and 20~ No further strength development was observed at-18 and 7~ whereas the UCS continued to increase through day 130 at 20 ~ It is uncertain which chemical or physical-chemical reactions within the residues are responsible for the strength development and the other observed changes occurring with time. Considering the high content of soluble chlorides (20 - 4 0 % (w/w)), pozzolanic reactions are not likely to occur, and attempts at detecting possible pozzolanic activity have proven negative (e.g. Kosson et al., 1993, Hjelmar 1992). Some of the cohesive properties of the dry/semi-dry APC process residues may be explained by the high content of calcium chloride and the excess lime present. When moisture is added to the dry residue, CaCI2.4H20 may be formed in the porewater. The addition of water also causes a strongly exothermic hydration of any excess CaO still in the residue, which may generate high temperatures (local boiling of water has been observed). CaCI2~ has a melting point of 29.9~ and may be first crystallised and then melted during the hydration of CaO. When the residue begins to cool off, melted

450 CaCI2.4H20 will fill the pores and subsequently solidify, thereby providing some mechanical strength to the residue (Hjelmar, 1993). This could also explain why the compaction properties deteriorate if moistened residues are stored prior to compaction. Figure 11.4 Unconfined Compressive Strength of Testbodies of Dry APC System Residue (45% of Water) as a Function of Time and Temperature 20"C

3000

m

n v

t-

2000

7~

t-

co

18oc

._> u} C1.

~ooo

E o

0

9

2s

&

7's

~;o

i

~3o

Days

Kullberg et al., 1989 11.4 PARTICLE MORPHOLOGY AND MINERALOGY

During the WASTE Program (1993), qualitative analysis was conducted on some of the dry APC system fabric filter residue samples using scanning electron microscopy (SEM) to determine the morphology of the particles. The particles were categorised into the 5 types listed in Table 11.4. In addition to the identification exercise, the relative percentage of the different particle types was determined optically (Table 11.4). The most common particle types in the dry scrubber residue were the polycrystalline and opaque irregular shaped particles, whereas the overwhelming majority of the fabric filter residue were polycrystalline in structure. The dry scrubber residue characteristics were very similar to those of heat recovery system ash (see Chapter 10), whereas the morphology of the fabric filter residue was dominated by the powdered lime particles encrusted with flue gas condensation/reaction products.

451 Table 11.4 Summary of Morphological Characteristics and Estimated Relative Percentages of Particle Types in Dry APC System Residues Morphology Estimated Relative % Size Range Fused Spheres Crystalline Polycrystalline Opaque Char

DS

FF

DS

FF

15 10 30 30 15

5 <5 80 5 <5

5 - 150 5 - 150 5- 200 5 - 200 160- 240

5 - 60 10 - 40 2 - 90 5 - 40 60- 500

Fused Spheres: Spheroids of various colours with particulate or gaseous inclusions Crystals: Irregular in shape, similar to soil-like particles of calcite or quartz Polycrystallines: Dense agglomeration of irregular shaped particles Opaques: Single, large irregular shaped particles (<300 microns) Char: Black fibrous particles WASTE Program, 1993 Oberste-Padtberg and Schweden (1990) have shown that fly ash and dry/semi-dry APC system residues may contain elemental AI particles. When the residues are contacted with water, a highly alkaline leachate will occur and react with the elemental aluminum to form hydrogen according to the following reaction (see also Equation 9.1): AI + O H + 3H20 -' [AI(OH)4]'+ 15H2

[11.1]

Hydrogen evolution has indeed been observed, both during laboratory batch leaching tests at low L/S ratios with fly ash and dry/semi-dry APC system residues, and in connection with full scale conditioning of semi-dry APC system residues with water. In the latter case, it caused explosions which were subsequently prevented by ventilation of the conditioning equipment (Hjelmar 1993). The mineralogical characteristics of APC residues play an important role in determining their leaching behaviour, and hence a thorough understanding of the mineralogy of APC residues may be essential to the development of an environmentally sustainable long-term APC residue management strategy. Specific mineralogical phases have been shown to control the leachability of key contaminant trace elements from incinerator ash (Comans et al., 1993), and the mineralogy of an ESP fly ash (without scrubber residue) has been investigated by Eighmy et al. (1993). Both studies used X-ray powder diffraction (XRPD) on fly ash before and after leaching, and observed that some of the more soluble phases (e.g. K2ZnCI4, NaCI, CaCIO4, MgSO4o5H20) were removed from the residue during leaching, (see Table 11.5). Hjelmar (1992) and Eighmy (1993) have reported XRPD data on semi-dry APC system residues (including fly ash) and the results are given in Table 11.5. Both sets of data appear to indicate that two of the major components, CaCI 2 and the excess reagent calcium oxide/hydroxide, occur partly as various double salts. In a study of two semi-dry APC

452 system residues (Stuart, 1993), CaCIOH, CaCO3, CaSO4, KCI, NaCI, SiO2, and PbO2 were identified using XRPD. It should be noted that the data represents only a portion of the crystalline minerals that may be present in APC system residues. The presence of several other "exotic" or less common minerals have also been detected, but not confirmed. Further research using XRDP and other techniques (see e.g. Chapter 7.3) is needed to reach a better understanding of APC system residue mineralogy. Table 11.5 Likely Mineral Phases of APC System Residues Based on Interpretation of X-Ray Powder Diffraction Data ESP fly ash (Eighmy et al., 1993) Semi-dry APC process residue Unleached ash Leachedash Hjelmar(1992) Eighmy (1993) Likely minerals Likely minerals K2ZnCI4 K2H2P20~ NaCI KCIO4 CaAI407 Ca3AI6Si2016 CaAI2Si208 MgSO4~ CaSO3 KAISI308

KAI(SO4)2 (Sr, Ba)SO4 (Mg2SnO4) Ba0~Sr0~SO4 TiO2 Ca3AI206 Na2CrO4 PbSO4 Pb~(PO4)3CI

CaCI2Ca(OH)2oH20 NaCI KCI Na2SO4 ~10H20 CaSO4

CaCIOH Ca(CIO)2o4H20 Ca(OH)2 NaCI (Pb2Sb207) NaAsO2 CaTiSiOs

11.5 WATER SOLUBILITY Water solubility is a very important property of APC system residues because it strongly influences the options available for treatment, disposal and possible utilisation of the residues. Generally, APC system residues contain a much higher proportion of soluble components than fly ash or bottom ash. The bulk of the soluble portion of combined fly ash and dry and semi-dry APC system residues consist of inorganic salts, which are acid-gas reaction products (most notably the highly soluble calcium chloride), surplus reactants and flue-gas condensation products. A certain proportion of most of the minor components, such as trace elements, are also soluble. This may be particularly true at low L/S ratios where solubility is enhanced for several components due to high ionic strength. Dry and semi-dry APC residues without fly ash are typically extremely soluble, whereas the pre-collected fly ash will be only moderately to very soluble in water. For residues from the wet scrubber process, the water soluble part primarily consists of gypsum and salt containing water retained in the pores of the filter cake. If the wet residue is mixed with the fly ash, the combined residue will contain more soluble salts which originated from the fly ash.

453 Total water solubility of the APC system residues should be determined at high L/S ratios (e.g. 100 -200 I/kg) and preferably in a two-step procedure in order to minimise solubility restraints caused by interference between the dissolved components. When such data are not available, column leaching data or multiple batch leaching data covering a reasonably broad L/S range may be used. For mass burn systems, water solubilities of 21 - 23% (w/w) for fly ash from ESPs, 27 38 % (w/w) for dry and semi-dry APC process residues and 14 % (w/w) for a wet scrubber APC sludge mixed with fly ash, have been measured (Hjelmar, 1992 and 1993). Conversely, water solubilities of up to 65% (w/w) have been observed for ESP ashes, and dry/semi-dry APC process residues from both mass burn and two-stage systems (Sawell and Constable, 1988; Sawell et al., 1989a & 1991; WASTE Program, 1993). In one of these studies, the relatively small corresponding proportion of residue collected in the dry scrubber reactor was only about 20% (w/w) water soluble due to the higher proportion of coarser fly ash present in the residue. 11.6 LOSS ON IGNITION Loss on ignition (LOI) is defined as the weight fraction (expressed as a percentage) of material that is lost by heating at 550~ LOI values for APC residues vary depending on type of APC system as well as on incinerator type and operation. Although LOI is an important indicator of the degree of burnout for bottom ash (see Chapter 9), the parameter is more difficult to interpret for APC residues. Comparisons between LOI and total organic carbon (TOC) indicate that the largest portion of the LOI for APC system residues containing acid gas cleaning residues must be attributed to loss of chemically bound water, such as dehydration of calcium chloride and excess calcium hydroxide at 550 ~ other flue-gas condensation products and possibly to loss of inorganic carbon (char). For fly ash collected upstream of the acid gas cleaning system, the LOI may be considered an indicator of carbon content, although the content of condensation products would depend on the temperatures within the collection unit. ESP fly ash from modern mass burn facilities can contain a relatively low LOI content of 1.1 - 1.3% (w/w) (Hjelmar, 1993). However, higher values (3 - 6% w/w) have been reported for ESP fly ash from older or retrofitted mass burn facilities (Hartl~n & Elander, 1986; Sawell and Constable 1988). LOI values of 2.8 -4.9% (w/w) have been observed for dry and semi-dry APC system residues (Hartl~n & Elander, 1986; Hjelmar, 1992), whereas the data from the NITEP Program indicated a wide variation of 2.1 - 12% (w/w) under a variety of operating conditions (Environment Canada, 1993). An LOI of 11% (w/w) has been observed for a sludge/fly ash mixture from a mass burn incinerator equipped with a wet APC system (Hjelmar, 1992). The differences underscore the influence that incinerator operating conditions have on APC residues. Stuart (1993) investigated LOI as a function of particle size for 2 semi-dry APC process residues from mass burn systems with bulk LOI values of 6.0 and 9.7%. The results are shown in Figure 11.5 and indicate that the larger particles (often char) generally

454 have the largest LOI values. The fine particles in one of the residues also had a high LOI, however, this was attributed to loss of water of hydration and calcination (Stuart, 1993). Figure 11.5 LOI as a Function of Particle Size for Two Semi-Dry APC System Residues from North American Mass Burning Incinerators

Loss On Ignition

Loss On Ignition

Facility 1

,.-,.

=m 12-

.-g_ 1 2 (n

"~ ~

~ 6: g ~

}

2'

"~o

Facility 2

a)

I 2141

4260

61110

1 1 1 - 151- >230 150 230

Bulk

Particle Size (microns)

10-

6,-

O"

.)141

4260

61110

111150

151230

>230

Bulk

Particle Size (microns)

Stuart, 1993 LOI values ranging between 5.9 - 10% (w/w) have been recorded for dry/semi-dry APC system residue from two-stage incinerators (Environment Canada, 1993). In addition, older RDF semi-suspension incinerators were found to generate dry/semi-dry APC residues containing higher LOI values (11% w/w) compared to values ranging between 4.1 - 7.9 % (w/w) for similar residues from a modern RDF incinerator (Environment Canada, 1993). 11.7 CHEMICAL CHARACTERISTICS

Knowledge of the chemical characteristics, i.e. pH and acid neutralisation capacity and the chemical composition of the APC system residues and their dependency on various variables, is necessary for an understanding of the behaviour of these residues during handling, treatment, disposal and/or utilisation. In order to provide an overview of the chemical characteristics of the various types of APC system residues, a data base has been compiled using data from incinerators in Canada (6 facilities), Denmark (7 facilities), Germany (4 facilities), Jersey, Channel Islands (1), the Netherlands (6 facilities), Sweden (7 facilities) and the USA (8 facilities). In all, information has been collected on APC system residue chemical characteristics from 39 incineration facilities. Some background on the origin of the data is given in Tables 11.6, 11.7, 11.8 and 11.9.

455 Table 11.6 Origin of Data on Composition of Fly Ash (FA) from Mass Burning (BM) Incinerators Country Type of Incinerator Sampling Y e a r Reference Residue . . . . Point Sampled ..... Canada

FA FA

Denmark

FA FA FA FA

Germany

MB, Quebec City . MB, Quebec City

ESP ESP

1991 1986

MB, Vestforbraending MB, Amagerforbraending MB, Kolding II MB= Fasan

ESP ESP ESP ESP

1985+92 1985 1992 1992

FA FA FA FA

MB, Bamberg MB, G6ppingen MB, GSppingen MB, Oberhausen

ESP ESP ESP ESP

1982 1982 1984 1987

Schneideret al., 1983 Schneideret al., 1983 Schneider,1986 . Vehlow, 1988

Jersey, Channel Is.

FA

MB, Bellozane

ESP

1988

Hjelmar et al., 1993

The Netherlands

FA FA FA

MB, Amsterdam MB, Rotterdam MB, den Haa~l . . . .

ESP ESP ESP

1985186 Versluijset al., 1990 1985/86 Versluijset al., 1990 ! 985/86 Versluiiset al., 1990

Sweden

FA FA FA

MB, GRAAB, G6teborg ESP MB, Uppsala ESP . MB, Avesta.... ESP.

MB, Glen Cove, LI FA MB, Saugus, Mass. -FA ESP = Electrostatic Precipitator USA

ESP ESP

Eighmy,1992 ....Sawell &.Constable, 1988 Hjelmar,1987 & 1993 Hjelmar,1987 Hjelmar,1993 Hielmar, 1993 .

1988 1988 1988

SGI data base, 1993 SGI data base, 1993 SGI data base, 1993

1987 1989

LIRPB, 1993 Hjelmaret al., 1993

The data have been divided into groups representing ESP fly ash from mass burn incinerators (see box plots Table 11.6), dry and semi-dry APC system residues from mass burn incinerators (Table 11.7), wet scrubber APC system residues from mass burn incinerators (Table 11.8) and dry/semi-dry APC system residues from two-stage incinerators and RDF-fed incinerators (Table 11.9). Box plots depict all of the data that is available in the data base for each type of residue. The central box for the elements in each plot extends from the first quartile to the third quartile, with a horizontal line across the box to indicate the median value. The first quartile denotes the twenty-fifth percentile, the median denotes the fiftieth percentile and the third quartile denotes the seventy-fifth percentile. The height of the box equals the interquartile range. Lines are sometimes drawn out from the quartiles to adjacent values, defined as those data points less than 1.5 times the interquartile range beyond the first or third quartiles. Values farther than 1.5 times the interquartile range from the box are considered outliers and are denoted by individual circles in each of the plots. The width of each box in a plot is proportional to the number of observations it represents.

456 Table 11.7 Origin of Data on the Composition of Dry(DP) and Semi-dry (SDP) APC System Residues from Mass Burn (MB) Incinerators Country

R e s i d u e Incinerator

Sampling Year Point Sampled

Reference

Canada

DP + FA DP + FA DP + FA DP + FA SP + FA

MB, GVRD MB, GVRD MB, GVRD MB, Flakt/QUC MB, Flakt/QUC

FF FF FF FF FF

1991 1988 1991 1986 1986

Eighmy, 1992 Sawell et al., 1990 WASTE Program, 1993 Sawell et al., 1987 Sawell et al., 1987

Denmark

DP + FA DP + FA SP + FA SP + FA

MB, Nordforb. MB, REFA MB, Amagerforbraending MB, KARA

FF ESP FF ESP

1989 1989 1989/92 1989

Hjelmar, Hjelmar, Hjelmar, Hjelmar,

Germany

SP + FA

MB, DiJsseldorf

ESP

1982

Schneider et al., 1983

Netherlands

DP (-FA) MB, Incinerator 1

Sweden

DP + FA SP + FA SP + FA

MB, Ht)gdalen MB, Sysav MB, Linkt~ping MB, Karlstad

USA

DP + FA DP + FA SP + FA SP + FA SP+FA

MB, "Dry Scrubber 1" MB, "Dry Scrubber 2" MB, 3x750 tpd MB, "Dry Scrubber 3" MB, 500tpd

(-FA) + FA ESP

:

9 9

Without fly ash (precollected) Including the fly ash Electrostatic precipitator(s)

1992 1992 1992 & 1993 1992

van der Sloot, 1992 FF CY

FF FF FF CY

1985188 1988 1988 1988

SGI database, SGI database, SGI database, SGI database,

1988 1988/89 1989 1988/89

L I R P B ,1993 LIRPB, 1993 Kossonet al., 1993 LIRPB, 1993 Stuart, 1993

:

9

1993 1993 1993 1993

Fabric filter Cyclone system

Table 11.8 Origin of Data on the Composition of Wet Scrubber Residues from Mass Burning (MB) Incinerators Country

R e s i d u e Incinerator

The WP (-FA) MB, Incinerator2 Netherlands WP (-FA) MB, Incinerator 3 Sweden

Sampling Year Reference Point Sampled WWT WWT

1990/91 van der Sloot, 1992 van der Sloot, 1992

WP (-FA) MB, GRAAB, G0teborg WWT 1988 SGI database, 1993 WP (-FA) MB, GRAAB, GOteborg VVW3" 1989 Hjelmar, 1994 WP + FA MB, GRAAB, G0teborg WWT 1989 Hjelmar, 1994 WP (-FA) MB, Uppsala WWT 1988 SGI database, 1993 WP: Wet scrubber process residue + FA: Mixed with the precollected fly ash WWT: Wastewater treatment system (-FA): Not mixed with the precollected fly ash

457 Table 11.9 Origin of Data on the Composition of Dry APC System Residues from Two-Stage Mass Burn (MB) Incinerators and an RDF-fed Semi-Suspensi0n Incinerator Country Residue Incinerator

Sampling Point

Year Sampled

Reference

Canada

CT/FF CT/FF CT/FF

1992 1988 1988

PRRI, 1992 Sawell et al., 1990 Sawell et al., 1989b

USA CT/FF SS RDF

DP + FA DP + F DP + FA

MB, 2-stage, Inc. A MB, 2-stage, Inc. B MB, 2-stage, LVH

DP + FA SS, RDF, Inc. 1 CT/FF 1990 Sawell et al., 1991 Combined residue from conditioningtower and fabric filter Semi-suspension 9 Refuse-derived fuel

9

The use of box plots allows the reader to compare the data bases that are available for each type of residue. Several of the data sets have outliers that denote a skewness in the data. Such skewness means that the data is not normally distributed. As can be seen from Tables 11.8 and 11.9, data are available on fly ash and dry/semidry APC system residues from six or seven countries, covering both Europe and North America. In contrast, information on wet scrubber system residues is only available from Europe (Table 11.8) and information on APC system residues for two-stage and RDF-fed systems is only available from North America (Table 11.9). Only a few data sets are available on the characteristics of dry/semi-dry APC residues from systems with pre-collection of fly ash. Studies have shown that the content of most trace elements in the acid gas cleaning residues is very low and primarily originating from the lime or the process water if a very effective fabric filter precollection system for fly ash is employed upstream of the acid gas cleaning system (Vehlow, 1993). 11.7.1 pH and Acid Neutralisation Capacity The pH resulting from contact of an APC system residue with water is the most important factor controlling the solubility of various heavy metals and trace elements from these residues. Since the pH may change with time due to interaction with the surroundings (acid/base reactions and transport), the buffering capacity of a residue plays an important role in maintaining a certain pH level, or determines the rate of change of pH with time. Most APC system residues are highly alkaline (pH 11-12.5) and the buffering capacity is generally equal to the alkalinity. The pH of a residue is normally measured in a distilled water suspension at L/S ratios between 20 and 100. Based on a contact time of 0.5 hrs at 100:1 L/S ratio, pH ranges

458 of 7.0- 11.3 for ESP fly ashes, 12.1 - 12.5 for dry and semi-dry APC system residues, and 10.5 for wet APC system sludge mixed with fly ash have been recorded (Hjelmar, 1993). All these residues were from mass burn incinerators. Figure 11.6 shows pH ranges measured in leachates generated with water at L/S = 20 from ESP fly ash and dry/semi-dry APC system residues from mass burn incinerators, and dry/semi-dry APC system residues from a two-stage and a semi-suspension (RDF) incinerator (Sawell & Constable, 1988; Sawell et al., 1991; WASTE Program, 1993). All the dry/semi-dry APC residues are seen to be highly alkaline, whereas the ESP fly ash ranged from slightly acidic to moderately alkaline. Practically all dry/semi-dry APC system residues are highly alkaline due to the presence of excess lime. Normally, ESP fly ash also contains enough alkaline material (e.g. oxides and carbonates of calcium, sodium and potassium carried over from the boiler) to create an alkaline pH in equilibrium with water. The alkaline core of the ESP ash particles is, however, often coated by sorbed acidic condensation products. The relative amount of acidic condensation products to alkaline material depends on the conditions and temperatures in the ESP. In some cases an ESP ash generates an initially acidic pH when contacted with water, after which the pH gradually increases to an alkaline level (Hjelmar, 1987; Sawell & Constable, 1988). In other cases, the amount of alkaline material is insufficient to neutralise the acidic condensation products, resulting in a neutral or slightly acidic pH (as in Figure 11.6). Figure 11.6 Boxplots of pH at L/S = 20 for ESP Fly Ash and Various Dry APC System Residues

pH T

I,,Z

12-

I

8

tO-

8 -

A = 2-Stage, Boiler B = 2-Stage, Economiser

(Bridle & Sawell, 1986; Sawell et al., 1989; Sawell & Constable, 1990)

C = Mass Bum, Boiler D = Mass Bum, Economiser

(Waste Program, 1993) (Sawell & Constable, 1988)

E = RDF, Economiser

(Sawell et al., 1991)

6 A

B

C

[ "--T'-

I

I

!

, D

E

459 In column leaching tests with dry and semi-dry APC system residues, an initial depression of pH (9.7 - 10.8) in the first leachate has been observed, even under equilibrium conditions with high amounts of excess lime (Hjelmar, 1992). This was probably caused by the very unusual solubility conditions imposed by the high ionic strength ( >10 moles/litre or up to 500 g dissolved salts/litre). At higher L/S ratios and lower ionic strengths, the pH reached a level of 12.1 - 12.5. The alkalinity or the acid neutralisation capacity (ANC) of the residues may be measured either as a full titration curve (see Chapter 7), or as the amount of acid equivalents used to titrate the residue to a fixed endpoint, usually pH = 7 or 4.4. Some typical ANC titration curves for APC system residues are presented in Figure 11.7. Hjelmar (1992 and 1993) has reported ANC values (titration to pH = 7) for ESP fly ashes of 2.5 - 3.5 eqv/kg, for dry and semi-dry APC system residues of 5.6 - 12 eqv/kg, and for a wet APC system sludge mixed with fly ash of 4.5. This corresponds with the values shown in Figure 11.7. The ANC of the dry and semi-dry APC system residues reflect the amount of excess lime present which again reflects the operating conditions (i.e. the stoichiometric ratio of lime addition or whether or not the residue is recycled). The influence of recycling on the ANC of dry scrubber residues from the scrubber reactor and the fabric filters (the bulk of the material) is illustrated in Figure 11.8. The data indicate that the concentration of acid gas reaction products increases with increased exposure to the flue gas stream, thereby reducing the ANC. Figure 11.7 ANC Titration Curves for an ESP Fly Ash (FA) and Two Semi-Dry APC System Residues (SD) from Fabric Filters 14

11.2

FA from ESP

8.4 -r-

(3.

SD(1) from FF

5.6

o

SD(2) from FF 2.8

.

5

Acid Added Stuart, 1993; Hjelmar, 1994

(eqvlkg

10

dry residue)

15

460 Figure 11.8 Comparison of ANC Values for Dry APC System Residues with and without Lime Recycle 21 18

. . . . . . . . . . . . . . Fabric Filter (No R e c y c l e ) o...

9

,~

9 . "

~

e-

o

~ Z <

.-

...

,o~

. , ,

15

12

,o , .

"-----'2"-:.'.-22.....

"

==.m, "=~.

9

lk.

t~

6 "

t*

- -- ~ .~e)

t~

--~

|-

1'1

13

pH (units)

Sawell and Constable, 1992 11.7.2 Chemical Composition" Inorganic Constituents Although the compiled data base on elemental composition of APC system residues contains more than 100 individual data sets, the diversity of the APC residues due to differences in incinerator type and APC system technology is such that a statistically justifiable comparison of APC system residues by country similar to that shown for bottom ash in Chapter 9 cannot be made. Instead, the elemental compositions of APC system residues are compared by type irrespective of the country of origin, as shown in Table 11.10. The differences in composition between residues/products from the semi-dry (SP) and the dry (DP) APC processes are small, and these residues are therefore treated as one single category (SP/DP). One of the main criteria for acceptance of analytical data into the data base has been based on the analytical methods applied. The use of a total quantification technique, e.g. neutron activation, x-ray fluorescence or total digestion followed by some spectrophotometric quantification, was required. However, due to the limited availability of data, it has been necessary to accept data on the trace element composition of wet APC process residues without fly ash which were determined after aqua regia digestion rather than total digestion. The difference between trace element analyses based on total digestion and aqua regia digestion, respectively, is probably minor for this particular type of residue in which the trace elements primarily are associated with calcium or TMT-based precipitates rather than silicates.

461 Table 11.10 Categories of APC System Residues Considered Type of incinerator Type of APC residue Mass burn Mass burn Mass burn Mass burn Mass burn Mass burn, 2-stage RDF, semi-suspension

FA from ESP (no scrubber residue) SP/DP from ESP and FF (with fly ash) DP from FF (without fly ash) WP (without fly ash) WP (mixed with fly ash) DP from CT/FF (with fly ash) DP from CT/FF (with fly ash)

Ranges of Elemental Composition of APC System Residues Based on the data base, the total ranges of elemental composition of the various categories of APC system residues are presented in Tables 11.11, 11.12 and 11.13. Table 11.11 shows the ranges of composition for fly ash, for dry/semi-dry APC system residues (with fly ash) and for wet APC system residues without and with fly ash, all from mass burn incinerators. In Table 11.12, the ranges of elemental composition of dry APC system residues from an incinerator equipped with an upstream ESP are compared with the corresponding ranges of composition for SP/DP without pre-collection of fly ash (from Table 11.11 ) for a limited amount of elements. Table 11.12 shows that, with the exception of Hg, most of the trace elements are primarily associated with the fly ash. For most of the elements analysed, a substantial enrichment is seen in the residues containing fly ash relative to the fly ash free residues. Table 11.13 shows ranges of elemental composition of dry APC system residues from two-stage and RDF semi-suspension incinerators. The number of elements presented in each of the tables reflects both the availability of analytical methods, and the degree and contaminants of environmental concern. Traditionally, much emphasis has been placed on trace metals, therefore, trace metal analyses are available for nearly all types of APC system residues. Less interest has been directed at understanding the role of the major constituents, including the soluble salts which may indeed be the most important and problematic constituents, both in relation to environmental risk and to the options available for management/treatment prior to disposal/utilisation. Due to this lack of interest, comprehensive information on the contents of all relevant major elements and soluble salt is not available for all types of APC system residues. Similar to the treatment of the bottom ash data in Chapter 9, the elements have been divided into three groups, namely, those present in one or more of the APC system residue types as major constituents (>10,000 mg/kg), minor constituents (>1,000 mg/kg but <10,000 mg/kg), and trace constituents (<1,000 mg/kg). For some of the elements

462

Table 11.11 Composition of Fly Ash, Dry/Semi-Dry and Wet APC System Residues from Mass Burn Incinerators Element

R a n g e for Fly A s h (mg/kg)

R a n g e for D r y / S e m i dry A P C S y s t e m Residues (mg/kg)

R a n g e for W e t A P C S y s t e m R e s i d u e w i t h o u t Fly A s h (mg/kg)

R a n g e for W e t A P C System Residue/Fly Ash Mixture (mg/kg)

Ag

2.3 - 100

0.9 - 6 0

-

53

AI

49,000 - 90,000

12,000 - 8 3 , 0 0 0

21,000 - 39,000

71,000 - 81,000

As

37 - 3 2 0

18 - 5 3 0

41 - 2 1 0

130 - 190

Ba

330 - 3,100

51 - 14,000

55 - 1,600

330 - 1,900

Be

-

0.5

- 0.9

-

1.5

- 1.9

C Ca

74,000- 130,000

110,000- 350,000

87,000- 200,000

Cd

50 - 4 5 0

140 - 300

150 - 1,400

93,000-

220 - 270

110,000

48,000 - 71,000

CI

29,000 - 210,000

62,000 - 380,000

17,000 - 5 1 , 0 0 0

Co

13 - 87

4 - 300

0.5 - 20

14 - 2 2

Cr

140 - 1,100

73 - 570

80 - 560

390 - 660

Cu

600 - 3,200

16 - 1,700

440 - 2,400

1,000 - 1,400

Fe

12,000 - 44,000

2,600 - 71,000

20,000 - 97,000

15,000 - 18,000

0.1 - 51

2.2 - 2 , 3 0 0

38 - 390

K

22,000

62,000

5,900 - 40,000

810 - 8 , 6 0 0

35,000 - 58,000

Mg

11,000- 19,000

5,100- 14,000

19,000- 170,000

18,000- 23,000

Mn

8 0 0 - 1,900

200 - 900

5,000 - 12,000

1,400 - 2,400

Mo

15 - 150

9.3 - 29

1.8 - 44

2 0 - 38

N

-

-

1,600

-

Na

15,000 - 57,000

7,600 - 29,000

720 - 3 , 4 0 0

28,000 - 33,000

Ni

60 - 260

19 - 710

20 - 310

6 7 - 110 6,000- 7,400

Hg

0.7 - 30 -

O P

4,800- 9,600

1,700- 4,600

-

Pb

5,300 - 26,000

2 , 5 0 0 - 10,000

3,300 - 22,000

5,900 - 8,300

S

11,000 - 45,000

1,400 - 2 5 , 0 0 0

2,700 - 6,000

11,000 - 26,000

Sb

2 6 0 - 1,100

300 - 1,100

80 - 2 0 0

-

Se

0.4 - 31

0.7 - 29

-

12

Si

95,000- 210,000

36,000- 120,000

78,000

120,000

6 2 0 - 1,400

340 - 450

1,000

Sn

550

-

2,000

Sr

40 - 640

4 0 0 - 500

5 - 300

200

Ti

6,800 - 14,000

700 - 5 , 7 0 0

1,400 - 4 , 3 0 0

5,300 - 8,400

V

2 9 - 150

8 - 62

25 - 86

62

Zn

9,000 - 70,000

7,000 - 20,000

8,100 - 53,000

20,000 - 23,000

463 Table 11.12 Comparison of Trace Element Content of Dry/Semi-Dry APC System Residues with and without Prior Removal of Fly Ash Element

Range for Dry APC System Residue without Fly Ash from Mass Burn Incinerator (mg/kg)

Range for Semi-dry/Dry APC System Residue with Fly Ash from Mass Burn Incinerators (mg/kg)

As Ba Cd Co Cr Cu Hg Mo Ni Pb Zn

1.5- 3 44 - 48 10 - 15 1 3-5 13- 25 9.5- 24 6 1 110 - 270 220 - 680

18- 530 51 - 14,000 140 - 300 4 - 300 73 - 670 16- 1,700 0.1 - 51 9.3- 29 19- 710 2,500 - 10,000 7,000 - 20,000

Table 11.13 C o n c e n t r a t i o n of Selected Elements in Dry APC System Residues from T w o - S t a g e Mass Burn Incinerators and From an RDF-fed Semi-Suspension Incinerator Element

Range for Dry APC System Residue from Mass Burn 2Stage Incinerators (mg/kg)

Range for Dry APC System Residue from RDF-fed Semi-Suspension Incinerator (mg/kg)

AI As B Ba Cd CI Co Cr Cu Hg Mn Ni Pb Sb Se Zn

1,000 - 17,000 < 0.3- 140 3 5 - 360 57 - 210 100 - 820 97,000- 280,000 14 - 18 21 - 190 130 - 630 0.3 - 54 110 - 370 17 - 47 1,900 - 13,000 150 - 1,100 < 0.2- 10 5,100 - 47,000

43,000 - 53,000 6.3- 7.6 2 0 0 - 240 640 - 1,200 75 - 160 83,000- 120,000 32 - 67 240 - 420 620 - 760 34 - 84 890 - 1,500 280 - 650 2,800 - 5,200 330 - 580 0.69- 1.3 5,000 - 8,900

464 this means that even though they are listed as major elements due to their typical concentration levels in one or more types of APC residues, they may in fact be minor or even trace constituents in other types of APC residues. The elements in each group are discussed in approximate order of decreasing abundance in the residues.

Major Elements (>10,000 mglkg): O,Cl,Ca,Si,Mg,Fe,AI,K, Na,Zn,S, Pb

A comparison with section 9.4.2 on bottom ash shows that the number of major elements is larger for APC system residues than for bottom ash. This is partly due to the greater diversity of APC residues, i.e. the major elements may vary between the residues from different types of APC systems. Figure 11.9 is a grouping of box plots depicting the concentration of the major elements in three types of APC residues from mass burn incinerators, fly ash without scrubber residue (FA), semi-dry/dry APC system products including fly ash (SP/DP), and wet scrubber products without fly ash (WP). Oxygen is one of the most prevalent elements in all APC system residues. It is present as oxides of several of the other major, minor and trace elements, most notably silicon, calcium, magnesium, iron, aluminum, sodium, potassium, sulphur and carbon. Few analytical determinations of the total oxygen content of APC system residues are available. Based on a variety of alternative analytical methods, the oxygen contents of fly ash from an ESP ranged from 60,000 mg/kg (AES), to 120,000 mg/kg (XPS) to 480,000 mg/kg (SEM/EDS) (Eighmy et al., 1993). Table 11.14 shows the mean and median values as well as the 25-75 percentile ranges of the concentrations of the major components CI, Ca, Si, Mg, Fe, AI, K, Na, Zn, S and Pb in fly ash (FA), semi-dry/dry APC products (SP/DP) and wet scrubber products (WP) from mass burn incinerators. Calcium is present both as one of the major matrix constituents of the particulate FA carried over from the boiler and as part of the major reaction product in SP/DP, calcium chloride. It may also be present as unreacted calcium oxide/hydroxide. Therefore, calcium is most abundant in SP/DP and least abundant in the FA-free WP. Chloride is present as a major condensation product on FA particles and as part of the major reaction product in SP/DP, calcium chloride. The chloride concentration of SD/DP is therefore generally 2-3 times higher than that of FA without scrubber residue. The chloride content of the wet APC system product (WP) is much lower because the chlorides from a wet scrubber is discharged with (or recovered from) the wastewater stream. Numerous studies have shown that the chloride concentration in FA and SP/DP is a function of temperature in the APC system (e.g., Environment Canada, 1986 and 1988; Carlsson, 1988; Ettehadiah and Lee, 1989; Itaya et al., 1989; Chang et al., 1989;

Figure 11.9 Major Element Concentrations in Fly Ash (FA), Semi-Dry APC Products with Fly Ash (SPIDP) and Wet Scrubber Products without Fly Ash (WP-FA)

Calcium. Ca

400 1

400

-

Chloride. C1

1

Silicon. S i

400

i

0

100

Iron. Fe

Magnesium, Si 30 0

Potassium, K

1

Sodium. Na

466 Table 11 14 Concentrations of Major Elements (>1000,000 mg/kg) Measured in APC Residues from Mass Burn Incinerators Residue

Element

Mean (mg/kg)

Median (mg/kg)

25-75 percentile range (mg/kg)

n

FA

Ca CI Si Mg Fe AI K Na Zn S Pb

107,000 74,000 160,000 15,000 25,000 71,000 36,000 31,000 28,000 26,000 11,000

107,000 50,000 170,000 15,000 23,000 73,000 34,000 29,000 22,000 27,000 7,800

95,000 - 120,000 40,000 - 102,000 130,000 - 180,000 14,000 - 17,000 18,000 - 33,000 59,000 - 81,000 30,000 - 41,000 23,000- 38,000 16,000 - 35,000 21 000 - 33,000 6,300- 15,000

20 24 14 15 20 18 19 17 26 20 25

SP/DP

Ca CI Si Mg Fe AI K Na Zn S Pb

230,000 180,000 69,000 9,400 12,000 26,000 23,OOO 17,000 15,000 15,000 5,400

220,000 160,000 63,000 8,800 9,100 19,000 24,000 15,000 16,000 17,000 5,600

180,000- 280,000 91,000- 220,000 51,000 - 92,000 7,400 - 12,000 63,000 - 11,000 15,000- 29,000 15,000 - 31,000 12,000 - 20,000 12,000- 18,000 8,200 - 21,000 4,100 - 63,000

19 23 12 16 19 27 18 16 28 18 27

WP

Ca

150,000 36,000 78,000 75,000 54,000 28,000 3,900 1 900 31 000 4 400 11 000

160,003 38,000

87,000- 200,000 26,000 - 47,000

3 4

Cl Si Mg Fe AI K Na Zn S Pb

FA: Fly ash WP: Wet sludge without FA

SD/DP:

n:

-

36,000 45,000 25,000 2,300 1,700 29,000 -

-

19,000 - 170,000 20,000- 97,000 21,000- 39,000 810 - 8,600 720- 3,400 15,000 - 45,000 2,700

- 6,000

1

3 3 3 3 3 12 2

9,700 4,400- 19,000 12 Semi-dry/dry APC process products with fly ash Number of residues analysed

467 Klingspor et al., 1989;). This trend is illustrated in Figure 11.10 by the increase in chloride concentration of ESP ash with decreasing flue gas temperatures. The data is taken from a study at a mass burn incinerator and shows a marked difference in chloride concentrations with the inflection point between 220~ and 230~ (Sawell and Constable, 1988). It is important to note that there was no lime injection prior to the ESP unit. Figure 11.10 Influence of Temperature on Chloride Concentrations in ESP Ash 250000

E EL

rl

C~

13 PT5 GOOD

E

.2

2oo0oo

-

150000

-

o E E O (-J

~

O ..E (,_) 100000

!

190

!

200

~

210

ESP

13 PT9 GOOD PT4 POOR I

!

220

230

I"1 PT14. POOR

24-0

TemperGture (~

From Figure 11.9 it can be seen that calcium and chloride are the only major elements which are more abundant in SP/DP than in FA. The content of very soluble calcium chloride, which may account for up to 60 percent of the total mass of semi-dry/dry APC system residues (without prior removal of fly ash), is responsible for many of the difficulties involved in the management of these residues. Not only do the very high concentrations of calcium chloride in the leachate pose a risk to potable water, but it may also increase the solubility of other potential contaminants such as trace metals (see Chapter 13). Furthermore, the solubility and thermal instability of the calcium chloride are serious obstacles to solidification and thermal stabilisation of the residues without prior removal of the soluble salts (see Chapters 18 and 19). A substantial amount of the sulphur in APC residues is present as sulphate and sulphite. Between 46,000- 80,000 ppm of SO42- and 13,000 - 35,000 ppm of SO32 have been measured in semi-dry and dry APC system residues, and 110,000 ppm of SO42 and 9,000 ppm of SO32-in a wet APC system residue mixed with fly ash (Hjelmar, 1992). The sulphite is thermodynamically unstable under oxidising conditions, and if the APC residue is exposed to oxygen (e.g. atmospheric air), will gradually be oxidised to sulphate. In wet scrubber residues, part of the sulphur content is present in the form of organic

468 sulphides (e.g. trimercaptotriazine, TMT). The long-term stability of these sulphides is not known. Both Pb and Zn are present as major constituents in APC system residues. Because of its increased solubility at high pH values, high ionic strength and high chloride concentrations, Pb is of particular concern in relation to disposal and utilisation of APC system residues. Figure 11.11 presents box plots depicting the concentrations of the minor elements in APC system residues (FA, SP/DP and WP) from mass burn incinerators. Table 11.15 gives the mean and median values as well as the 25-75 percentile ranges of the concentrations of the minor components Ti, Mn, Ba, Sn and Cu for the three types of residues. Table 11.15 Concentrations of Minor Elements (1,000- 10,000 mg/kg)Measured in APC Residues from Mass Burn Incinerators Residue

Element

Mean (mg/kg)

Median (mg/kg)

25-75 percentile range (mg/kg)

n

FA

Ti Mn Ba Sn Cu

8,700 1,300 1,700 1,400 1,200

8,700 1,200 1,700 1,500 1,100

7,500- 9,400 1,000 - 1,600 940 - 2,600 890 - 1,800 930 - 1,300

17 19 18 15 25

SP/DP

Ti Mn Ba Sn Cu

3,300 480 540 890 710

3,200 440 450 840 630

2,600 - 4,400 280 - 680 320 - 660 770- 1,000 490 - 860

17 19 18 15 25

2,600 9,100 460 400 1,200

2,200 10,000 200 900

1,400 - 4,300 5,400 - 12,000 87 - 670 340 - 450 760- 1,700

3 3 11 2 12

WP

FA:

Ti Mn Ba Sn Cu Fly ash

SD/DP:

Semi-dry/dry APC process products with fly ash

WP:

Wet scrubber products without fly ash

n:

Number of residues analysed

Figure 11.I 1 Minor Element Concentrations in Fly Ash (FA), Semi-DrylDry APC Products with Fly Ash (SPIDP) and Wet Scrubber Products Without Fly Ash (WP-FA)

Titanium. T I

15000

m

ioooo

1

Y

\

m

E

5000

2500

I

Copper, Cu

Tin, Sn

1

3000 4000

Barium, Ba

Manganese, Mn

,OoO

1

470 Inorganic carbon in the form of carbonate is also one of the minor elements. Total concentrations between 16,000 - 33,000 ppm for CO32 in SP/DP and a value of 19,000 ppm for CO32 in WP have been reported (Hjelmar, 1992). Carbon is also present in APC system residues as elementary C (soot or char) carried over from the boiler. In addition, activated carbon may be injected with the sorbent in some dry/semi-dry APC systems in order to enhance mercury adsorption.

Trace Elements (< 1,000 mglkg): Hg,Cd,Sb,Cr, Sr, Ni,As,V,Ag,Co, Mo,Se

Figure 11.12 presents box plots showing the distribution of some trace elements in APC system residues (FA, SP/DP and WP) from mass burn incinerators. Table 11.16 shows the mean and median values as well as the 25-75 percentile ranges of the concentrations of the trace elements Hg, Cd, Sb, Cr, Sr, Ni, As, V, Ag, Co, Mo and Se for the three types of residues. The concentration of mercury is substantially higher in WP than in SP/DP, which in turn has a higher concentration of mercury than FA. Trimercaptotriazine (TMT) which is most commonly used for wet scrubber wastewater treatment is particularly effective for removal of mercury and is ultimately concentrated in the sludge. Several facilities equipped with dry or semi-dry APC systems inject sodium sulphide, or as mentioned above, activated carbon with the sorbent to increase the sorption of mercury vapour. Without mercury control, mercury tends to condense out as Hg2CI2 (mercuric II chloride) (Metzger and Braun, 1987). Mercury in this form can undergo either reduction or methylation on fly ash, and that elevated temperatures can greatly increase the speed of the reduction reaction (Lindquist et al., 1986). Nagase et al., (1986) observed that methylation of Hg2CI2 can occur at normal temperatures on fly ash. Several studies have been conducted on the removal efficiency of mercury from flue gas based on temperatures at the outlet of the APC system (e.g., Moiler and Christiansen, 1985; Clarke, 1986; Carlsson, 1986; Environment Canada, 1986). All of the studies demonstrated that the increased efficiency of mercury removal was achieved (>91% capture) at temperatures below 150 - 160~ At temperatures over 200~ mercury removal efficiency was negligible, if not nonexistent. However, the use of activated carbon or sodium sulphide has been demonstrated as effective control reagents for mercury (Guest and Knizek, 1991). Sodium sulphide injection results in effective formation of stable HgS. Consequently, mercury (as well as sulphate) concentrations are increased in the residues. Conversely, it has been speculated that mercury sorbed onto activated carbon is not as stable, since it is susceptible to reduction by carbon. Although this has not been confirmed, further study into the phenomenon should be conducted to ensure that there is no substantial release of elemental mercury into the atmosphere through reduction, or methylation reactions.

Figure 11.12 Trace Element Concentrations in Fly Ash (FA), Semi-DrylDry APC Products with Fly Ash (SPIDP) and Wet Scrubber Products without Fly Ash (WP-FA)

Antinony, c d

.

Cnroniun, c r

Y

a

1000

Vanadium, V

iOOD

400 200

S i l v e r . ~g 120 1

ioo

C O M l t . CO

1

472 T a b l e 11.16 T r a c e Elements (<1,000 mg/kg) M e a s u r e d in A P C Residues from Mass Burn Incinerators Residue

Element

Mean (mg/kg)

Median (mg/kg)

25-75 percentile range (mg/kg)

FA

Hg Cd Sb Cr Sr Ni As V Ag Co Mo Se

8.0 390 530 650 280 140 130 51 55 51 40 14

6.0 290 450 730 250 110 130 45 53 54 30 12

2 . 3 - 10 240 - 480 340 - 690 430 - 840 140 - 400 91 - 110 49 - 200 32 - 63 3 3 - 75 30 - 69 25 - 37 11 - 18

17 26 12 26 12 25 17 15 10 17 13 12

SP/DP

Hg Cd Sb Cr Sr Ni As V Ag Co Mo Se

15 300 790 180 460 94 170 33 22 9.6 15 8.2

12 260 820 140 400 30 170 31 14 9.0 15 7.0

8 . 4 - 18 190 - 360 630 - 940 110 - 220 4 0 0 - 500 23 - 60 120- 210 1 9 - 50 1.1 - 49 6.0 - 15 11 - 20 4.8 - 11

28 28 13 28 6 25 26 6 11 11 11 23

WP

Hg Cd Sb Cr Sr Ni As V Co Mo

650 630 140 240 104 62 89 47 9.8 12

660 660 210 85 36 72 31 9.7 6.0

2 4 0 - 790 290 - 880 8 0 - 200 130- 340 52 - 130 2 6 - 53 49 - 110 25 - 86 4 . 8 - 16 3.0 - 24

12 12 2 12 10 12 12 3 12 11

FA: Fly ash SD/DP:Semi-dry/dry APC process products with fly ash

WP: Wet scrubber products without fly ash n: Number of residues analysed

473

11.7.3 Role of Particle Size in Element Distribution Researchers (e.g. Ontiveros et al., 1989; Stuart, 1993) have investigated the variation of element distribution in fly ash and semi-dry APC system residues as a function of particle size. Based on acid extraction data, Ontiveros et al. (1989) have shown that more volatile elements such as cadmium and lead are enriched substantially in the smaller particle sizes of fly ash, whereas the opposite is true for the matrix elements, such as aluminum, barium, iron, manganese, nickel and potassium. The enrichment is ascribed to condensation of volatile metals on the surface of the smaller particles, since the small particles have larger surface area to volume ratios than larger particles. Similar results have been obtained by Stuart (1993) for semi-dry APC system residues when the results are corrected for the dilution effects caused by soluble acid gas cleaning reaction products. Stuart also found that a high percentage of the bulk APC residues investigated consisted of calcium complex species, which preferentially partitioned to the smallest size fractions. A significant increase in the concentration levels of cadmium, lead and zinc with decreasing grain size below 200 microns has been observed in ESP fly ashes (Hundesr0gge, 1990). However, between 600 and 1,000 microns (usually a small fraction), an increase in concentration with increasing grain size was observed for the same elements. This was ascribed to increased adsorption due to the high carbon content in this fraction.

11.7.4 Chemical Composition: Organic Constituents Organics Present in APC System Residues A number of studies have been performed using APC residues either to develop analytical techniques or to assess the formation of organic compounds during incineration, especially PCDD/PCDF. However, some of these studies contain insufficient information about the origin of the residues, incinerator and APC system technology, and sampling techniques. Other studies have addressed APC system residues only as a constituent of combined ash. Therefore, only a few selected data sets are presented, mostly in the form of concentration ranges. Most of these are based on the data base described in Tables 11.6, 11.7, 11.8 and 11.9. The data on total organic carbon (TOC) concentrations in APC system residues are presented in the box plots in Figure 11.13. The specific nature of the bulk portion of the TOC in the APC system residues has not been investigated. Median values for TOC of 7.7 g/kg and a total range of 2.7 - 17 g/kg for ESP fly ash from mass burn incinerators have been recorded (Hjelmar, 1987 and 1993; Hjelmar et al., 1992). A median value of 7 g/kg and a total range of 6 - 9 g/kg of TOC for semi-dry/dry APC system residues from mass burn incinerators have also been reported, whereas a single value of 7 g/kg was measured for a wet APC system residue mixed with fly ash from a mass burn incinerator. Some trace organics of potential concern for human health have been quantified in the various types of APC system residues. These include the polychlorinated dibenzo-p-

474

dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) as well as potential precursors for these compounds under certain reaction conditions. The potential precursors are chlorinated benzenes (CBs), chlorinated phenols (CPs), and polychlorinated biphenyls (PCBs). Another group of potentially toxic compounds that have been identified in APC system residues consists of condensed polycyclic aromatic hydrocarbons (PAHs). In addition, a limited amount of information is also available on the concentration of phthalates in some APC system residues. The presence of most the above-mentioned trace organics in combustion residuals and their fate during combustion is thoroughly discussed in Chapter 8. Table 11.17 summarises the data on trace organic concentrations in APC system residues. Figure 11.13 Concentrations of Total Organic Carbon (TOC) in Selected APC System Residues from Mass Burn Incinerators 20 -

TOC

15 -

10 -

I I _

O

_

FA

SP/DP

WP + FA

Chlorinated Benzenes and Phenols Table 11.17 indicates that total CB and CP concentrations are available for fly ash and dry APC residue from mass burn incinerators, dry APC residue from a two-stage mass burn incinerator and an RDF-fed semi-suspension incinerator. The values which range from 220 - 1,900 ng/g for CB and 860 - 3,200 ng/g for CP are ~1 to 2 orders of magnitude higher than those found for bottom ash (see Table 9.20).

475 Table 11.17 Trace Organic Concentrations in APC System Residues (ng/g) Parameter

Mass Burn Incinerators (MB)

DP

220 c

1,600 d

860 c

860 d

< 40 c

270 d

65 c

260 d

260

60 c

390 d

25-73 e 1.4-9.7 j

120

53 c

120 d

0.8-2.0 ~

2.8

SP/DP

CB

890" 800-1,900 b

0.03-0.4 j

CP

1,800 a 50-200 j 1,500-3,200 b

PCB

Nd" 6_16 b <20 i

ND j

PAH

110" 450-1,050 b

3(~

PCDD

580 ~ 590-1,040 b 115-140 f

14-32 e 0.7-3.6 j

PCDF

190" 190-280 b 48-69 f

TCDD-EQ

1.5-2.5 f

Phthalates

<500-4,200 ~

: : : : :

RDF,SS

DP

FA

FA SP/DP WP WP+FA RDF SS

MB, 2-Stage

WP

WP+FA

<20 i

1,600-10,000 i

Fly ash (from electrostatic precipitator) Semi-dry/dry APC system residue (with FA) Wet APC system residue (without FA) Wet PC system residue (mixed with FA) Refused derived fuel Semi-suspension combustion technology f : Hjelmar et al., 1993 Klicius et al., 1988 g : Hjelmar, 1987 Sawell & Constable, 1988 h : Hjelmar, 1993 Sawell et al., 1989a i : Milj~styrelsen, 1993 Sawell et al., 1989b j : Sawell et al., 1990 Hjelmar, 1992

476

Polychlorinated Biphenyls Determinations of concentrations of PCB in fly ash, dry APC residue and a wet scrubber product from mass burn incinerators, dry APC system residues from a twostage mass burn incinerator and an RDF-fed semi-suspension incinerator are shown in Table 11.17. The concentrations of PCBs are generally small or below the detection limit, except in the DP from an RDF-fed semi-suspension incinerator which has a PCB concentration of 270 ng/g. The lowest PCB concentration is found in the dry APC residue from the two-stage incinerator. This is probably a result of the very high temperature level in the secondary combustion chamber. In contrast, the bottom ash from these incinerators is generally poorly combusted and contains relatively high concentrations of PAHs. Polycyclic Aromatic Hydrocarbons PAHs are to a large extent a measure of the quality of the combustion process. For example, the higher the concentration, the poorer the combustion. The values found range from 65 to 1,050 ng/g. Dioxins and Furans The TCDD equivalents have been calculated according to Eadon's method. The ranges of concentrations of PCDDs and PCDFs in the APC system residues are between 1 - 1,040 n/g and 1.4 -280 ng/g, respectively. The data does not allow a clear distinction between PCDD and PCDF concentrations for the different types of APC system residues, although there is data available which indicates that ESPs can enhance PCDD/PCDF formation (see Chapter 8). The concentrations of PCDDs and PCDFs in APC system residues are generally several orders of magnitude higher than that in bottom ash (See Chapter 9). Phthalates One study has included a determination of the concentrations of phthalates in fly ash and wet APC system residues from a Danish mass burn incinerator (Milj~styrelsen 1993). As shown in Table 11.17, a phthalate concentration range of <500 - 4200 ng/g was found for the fly ash and a range of 1,600 -10,000 was found for the wet scrubber residue. The dominant phthalates were di-n-butylphthalate and di(2-ethylhexyl)phthalate. 11.8 COMPOSITION OF WASTEWATER FROM WET SCRUBBER APC SYSTEMS The wastewater from wet scrubbing systems normally has a high content of salts and a moderate to high content of metals and other trace elements. Table 11.18 shows the concentrations of metals in the wastewater streams from the first (acidic) and second (neutral) scrubbing stages of a two-stage wet scrubber system on a 4 x 12 tph mass burn incinerator (Vestforbraending 1993, Rasmussen et al., 1993).

477 The two wastewater streams constitute about 0.4 m3/tonne of the waste from stage 1, and 0.1 m3/tonne of the waste from stage 2. These are treated in the wastewater treatment system shown in Figure 4.10 in Chapter 4. The treatment consists of an initial adjustment of the acidic scrubber water to pH = 1.8 with limestone (calcium carbonate) followed by adjustment of the combined streams to pH = 8.8 with lime slurry (calcium hydroxide), addition of organic sulphide (TMT) to bind heavy metals and polymer to aid settling, separation of the suspended particulate matter in a lamellae tank separator, and sand filtration prior to discharge to the sewer system. The settled sludge is mixed with the ESP fly ash and landfilled. Bottom ash quench discharge water is used for slaking, and a stream of the treated wastewater is recycled and used for bottom ash quenching. The composition of the treated wastewater stream is presented in Table 11.18 in terms of ranges and average parameter values based on monthly sampling and analysis over a 12 month period (1993).

Table 11.18 Composition of Untreated and Treated Wastewater from a Two-Stage Wet Scrubber APC Syste m (mg/L) . . . . . . . . . . . . . . . . . . . . . . . Parameter

Wastewater from the first stage* (acidic)

Wastewater from the second stage* (neutral/alkaline)

_ Temperature (~ pH (units) Sulphate Chloride SS Ag Cd Co Cr Cu Hg Ni Pb Sn Zn Cyanide Phenols

0- 1 0.12 0.41 < 0.004 0.25 3.2 3.0 0.051 34 2.6 83 -

Approx. 7 0.037 0.25 < 0.004 0.031 0.74 0.79 0.020 2.0 0.46 22 -

= single determinations Vestforbraending, 1993; Rasmussen et al., 1993

Treated wastewater (combined from both stages) . Range

Mean

35 - 43 6.9- 10.2 620- 1,050 7,000- 12,000 15 - 110 < 0.02 < 0.001 - 0.020 0.14 - 0.36 0.10-0.47 0.002 - 0.015 0.03 - 0.27 <0.02 - 0.68 < 0.10 - 0.21 0.20- 3.2 < 0.01 - 0.11 < 0.05 - 0.53

840 9,600 53 0.008 0.23 0.19 0.005 0.10 0.18 0.90 0.05 -

478 REFERENCES

CCME, "Operating and Emission Guidelines for Municipal Solid Waste Incinerators", Canadian Council of Ministers of the Environment, Report CCME-TS/WM-TRE003, June 1989 Carlsson, K.B. "Removal of Acidifying Gases, Heavy Metals and Dioxins in Flue Gases from Waste to Energy Plants", in Proceedinas of the 1988 International Energy AQency Workshop, June 27-29, Cambridge, UK, 1988 Chang, J.C.S., T.G. Brna and C.B. Sedman, "Pilot Evaluation of Sorbents for Simultaneous Removal of HCI and SO2 from MSW Incinertor Flue Gas by Dry Injection Process", in Proceedin.q from the International Conference on Municipal Waste Combustion, Vol. 2, Florida, 1989 Clarke, M.J. "Emission Control Technologies for Resource Recovery", presented at Symposium on Environmental Pollution in Urban Area, Brooklyn Polytechnic University, 1986 Comans, R.N.J., H.A. van der Sloot and P.A Bonouvrie. "Geochemical Reactions Controlling the Solubility of Major and Trace Elements During Leaching of Municipal Solid Waste incinerator Residues". In (J. Kilgroe, ed.) Proceedinas 1993 International Conference on Municipal waste Combustion, March 30-April 2, Williamsburg, VA, AWMA, Pittsburgh, Pennsylvania, p. 457., 1993. Eighmy, T.T. Personal communication, 1992. Eighmy, T.T. Personal communication, 1993. Eighmy, T.T., D. Domingo, J.R. Krzanowski, D. St,~mpfli & D. Eusden. "The Speciation of Elements in Incineration Residues" In (J. Kilgroe, ed.) Proceedings 1993 International Conference on Municipal Waste Combustion, March 30-April 2, Williamsburg, VA, AWMA, Pittsburgh, Pennsylvania, p. 457, 1993. Environment Canada, "The National Incinerator Testing and Evaluation Program: Air pollution Control Technology (Summary Report)", Conservation and Protection, Ottawa, ON, Report EPS 3/UP/2, 1986 Environment Canada, "The National Incinerator Testing and Evaluation Program: Environmental Characterisation of Mass Burn Technology at Quebec City (Summary Report)", Conservation and Protection, Ottawa, ON, Report EPS 3/UP/5, 1988 Environment Canada, "The National Incinerator Testing and Evaluation Program (NITEP): A Summary of the Characterisation and Treatment Studies on Residues from Municipal Solid Waste Incineration", Report EPS 3/UP/8, 1993

479 Ettehadieh, B. and S.Y. Lee, "Reducing Acid Gas Emissions from Mass-Burn Waste-toEnergy Plants by Direct Lime Injection", in Proceedin,q from the International Conference on Municipal Waste Combustion, Vol. 2, pp. 82-26 to 86-44. Florida, 1989 Guest, T.L. and O. Knizek, "Mercury Control at Burnaby's Municipal Waste Incinerator", Proceedin,qs of the 84th Annual AWMA Meetinq, Vancouver, B.C., Paper 91-103.30, June 1991. Har[l~n, J. & P. Elander. "Residues from Waste Incineration - Chemical and Physical Properties". SGI Varia 172. Statens Geotekniska Institut, Link5ping, Sweden, 1986. Hjelmar, O. "Leachate from Incinerator Ash Disposal Sites". In Pjoceedings of the International Workshop on Municipal Waste Incineration,, NITEP, Montreal, Canada, October 1-2, 1987, pp. 287-318, 1987. Hjelmar, O. Restprodukter fra reQ.aasrensnin,q ved affaldsforbraendin,_q II, Eksperimentelle Underse.qelser. Miljeprojekt nr. 193, Miljestyrelsen, Kebenhavn, 1992. Hjelmar, O. Stofudvasknin,q fra flyveaske fra affaldsforbraendin.asanlae.a. Rapport til Miljestyrelsen. VKI, Hersholm, Denmark, 1993. Hjelmar, O., H. Thomassen & H. Hejmark Restprodukter fra re.a.qasrensning ved affaldsforbraendin,_q I, Udredning. Miljeprojekt nr. 146, Miljestyrelsen, Kebenhavn, 1990. Hjelmar, O., K.J. Andersen, J.B. Andersen, E.A. Hansen, A. Damborg, E. Bjernestad, A.H. Knap, C.B. Cook, S.B. Cook, J.A.K. Simmons, R.J. Jones, A.E. Murray, M.J. Lintrup, H. Schreder, F.J. Roethel Assessment of the Environmental Impact of Incinerator Ash Disposal in Bermuda, Final Report, February 1993. Prepared for Ministry of Works & Engineering, Hamilton, Bermuda, by the Water Quality Institute, Hersholm, Denmark, 1993. HundesrQgge, T. "Phasenanalytische Untersuchungen an Filteraschen aus MQIIverbrennungsanlagen", Aufschluss, 41, pp.281-285, 1990. Itaya, M., Y. Yamahata, Y. Hasebe, H. Harada and K. Mikawa, "Removal of Hydrogen Chloride Gas from MSW Fluidised Bed Incinerator Using Bag Filter", Proceedin.q from the International Conference on Municipal Waste Combustion, Vol. 2, Florida, 1989 Klicius, R., A. Finkelstein & D.J. Hay "The National Incinerator Testing & Evaluation Program (NITEP): Mass Burning Technology Assessment", In Proceedin,qs of the 5th International Solid Wastes Conference, Copenhagen, September 11-16, 1988 Klingspor, J.S., D.L. Roberts and I.A. Jefcoat, "Acid Gas Emissions: Results of Spray Dry Scrubbing Pilot Plant Study", in Proceedin.q from the International Conference on Municipal Waste Combustion, Vol. 2, Florida, 1989

480 Kosson, D.S., T.T Kosson & H.A. van der Sloot. Evaluation of .Solidification/S.tabilization Treat.m.ent Process for Municipal Waste Combustion Residues. Volume I. U.S. EPA, RREL, Cincinnati, Ohio, 1993. Kullberg, S. & A.-M. F,~llman. Leaching Properties of Natural and Stabilized Flue Gas Cleaning Residues from Waste Incineration. In Proceedings of the International Conference on Municipal Waste Combustion, Florida, p. 3B-21, 1989. Kullberg, S., A.-M. F,~llman & A. H0jlund. Stabilisern.Q och deponerin.q av. r0k~asreninasprodukter fr~n sopf0rbr,~nnin,q. Stiftelsen for varmeteknisk forskning, Br~nsleteknik 370, Stockholm, 1989 Lindquist, O., K. Puromaki and P. Schager. "Mercury from Waste Combustion". Appendix to "Residues from Waste Incineration- Chemical and Physical Properties" by J. Hartlen and P. Elander, Swedish Geotechnical Institute, 1986. LIRPB, Long Island Regional Planning Board. The Potential for Beneficial Use of Waste-to-Ener.qv Facility Ash - Chemical and En.vironmental Property Data Report. LIRBP/NYSERDA, 1993. Metzger M. and H. Braun. "ln-Situ Mercury Speciation in Flue Gas by Liquid and Solid Sorption Systems", Chemosphere, Vol. 16, No.4, 1987 Milj~styrelsen. Unpublished results, Copenhagen, 1993. Moiler J.T. and O. Christiansen. "Dry Scrubbing of MSW Incinerator Flue Gas by Spray Dryer Absorption: New Developments in Europe", presented at 78th Annual Air Pollution Control Association Meeting, June 1985. Nagase, H., Y. Ose, T. Sato and M. Yamada, "Mercury Methylation by Ash from Refuse Incineration", Science of the Total Environment, No. 53, 1986. Oberste-Padtberg, R. & K. Schweden. "Zur Freizetzung von Wasserstoff aus M0rteln mit MVA-Reststoffen". WLB Wasser, Luft und Boden, 34 (6), pp. 61-62, 1990. Ontiveros, J.L., T.L. clapp & D.S. Kosson. "Physical Properties and Chemical Species Distributions Within Municipal Waste Combustor Ashes". Environmental Pro.qress, 8 (3), pp. 200-206, 1989. Peel Resource Recovery Incorporated (PRRI), "Ash and Quench Water Testing Report", Report prepared for the Region of Peel, Brampton, Ontario, July 1992. Rasmussen, H. W., O. Hjelmar, A.O. Knudsen, T. Kristiansen & H. Birch. Sambehandlin.q .af ....restprodukter fra. affaldsforbraendin(:]. Arbejdsrapport fra Milj~styrelsen, Milj~styrelsen, K~benhavnl 1993.

481 Sawell, S.E. & T.W. Constable. NITEP Phase liB: Assessment of Contaminant Leach.ability from the Residues of a Mass Burning Incinerator. Environment Canada, EPS Manuscript Series IP-82, Vol. VI, 1988. Sawell, S.E. & T.W. Constable. The National Incinerator Testing and Evaluation Program" The Influence of Incinerator Design on Ash Characteristics. In Proceedinqs of t.h.e 3rd International Conference on Ash Utilization and Stabilization (Ash III), Arlington, Virginia, U.S.A., 1991. Sawell, S.E. & T.W. Constable. The National Incinerator Testin.q and E.valuation program: A Summary of the Characterization and Treatment Studies on Residues from Municipal Solid Waste Incineration. Environment Canada Report, 1992. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator TestinQ and Evaluation Program: Evaluation of Residues from a Modular Municipal Waste Incinerator with ....Lime Based Air Pollution Control. Environment Canada Report, Manuscript Series, IP-101, 1989a Sawell, S.E., T.W. Constable and R.K. Klicius. The National In.cinerator Tes.tin.q and Evaluation Pro.qram: Evaluation of Contaminants Leachability from Residues Collected at a Refuse Derived Fuel Waste Combustion Facility. Environment Canada Report, Manuscript Series, IP-96, 1989b. Sawell, S.E., T.W. Constable and R.K. Klicius. The National !.ncinerator Testin,q and Evaluation Pro.qram Characterization of Residues f.rom a Two-Sta.qe Incinerator with Rotary Kiln (3M Canada). Environment Canada Report, Manuscript Series, IP-119, 1990. Sawell, S.E., T.W. Constable & R.K. Klicius. The National Incinerator Testin~ and Evaluation P.ro.qram: Characterization of Residues from a Refuse Derived Fuel Combustion ..Facility (Mid-Connecticut). Environment Canada Report, Manuscript Series, 1991. Schneider, J., "Determination of Elemental Waste Composition by Analysis of Incineration Residues", Recvclin,q International, 1 318-326, 1986 Schneider, J., H. Walk, R. HQrtel and L. Scholz. Untersuchunqen an EFilterstaubproben. aus MQIIverbrennun.qsanla.qen, Prim~rbericht 4/83, Laboratorium fQr Isotopentechnik, Kernforschungszentrum Karlseruhe, 1983. Swedish Geotechnical Institute. SGI data base, Link5ping, Sweden, 1993. Stuart, B.J. "Characterization of Municipal Waste Combustion Air Pollution Control Residues as a Function of Particle Size". M.Sc. Thesis, Rutgers, The State University of New Jersey, New Brunswick, New Jersey, 1993.

482 van der Sloot, H.A. Personal communication, 1992. Vehlow, J., Personal communication, 1988. Vehlow, J. Personal communication, 1993. Versluijs, C.W., I.H. Anthonissen & E.A. Valentijn MAMMOET '85L Inte0rale evaluatie van de deelunderzoeken. Rapport: nr 738504008, Rijksinstituut voor Volksgezondheid en Milieuhygiene, Bilthoven, 1990. Vestforbr~nding. '~/andrensning 1993", Internal report, Vestforbraending, Glostrup, Denmark, 1994. VKI, Water Quality Institute. Unpublished field results. H~rsholm, Denmark, 1992. VKI, Water Quality Institute. Unpublished laboratory results. H~rsholm, Denmark, 1994. Waste Analysis, Sampling, Testing and Evaluation (WASTE) Program. Final Report of.. the Mass Burn MSW Incineration Study (Burnaby, B.C). Report Prepared for Environment Canada, U.S. Environmental Protection Agency and the International Lead Zinc Research Organization, Vols. 1-4, 1993.

483

C H A P T E R 12 - PHYSICAL ASPECTS OF LEACHING

Much of the concern over the management strategies for MSW incinerator ash is based on the perceived potential contamination of water resources through leaching of salts and trace metals from the ash. This focus has been driven by results from regulatory leaching tests designed to simulate a specific worst-case disposal scenario. These tests do not translate to most management practices for the residues. The leaching of MSW incineration residues is a highly complex subject which requires a great deal of background knowledge to interpret leaching data adequately. A more comprehensive approach to understanding leaching phenomena is needed. 12.1 AN INTRODUCTION TO LEACHING

When in contact with a liquid, many solid materials will dissolve to some extent. It is the degree of this dissolution in the contacting liquid that governs the relative strength of the leachate that is produced. Leaching can occur in the field in residues in piles, landfills or granular fill applications, or in the laboratory during column or batch leaching tests. There are a number of factors that can influence the rate at which the solid materials dissolve and these are discussed under two headings: physical aspects and chemical aspects. For physical aspects influencing leaching, some questions can be posed that help to define the leaching system: What is the leaching system being described? Is it landfilled material where an infiltrating leachant flows through the ash (the flow-through system), a batch leaching test where the ash and leaching fluid are agitated together and where the leachant always remains in contact with the ash (the agitated batch system) or is it a batch leaching test without agitation (the non-agitated batch system)? What is the leaching time frame (e.g. minutes to millennia)? What is the size of particle undergoing leaching? Does the particle have a high surface area relative to the volume of the particle, such as in fine-grained particles where the surface area available for dissolution is high? Is the particle composed of many different types of minerals or is it homogeneous? How does the liquid flow past the particle?

484 How fast does the particle dissolve? For a flow-through system, is the rate of dissolution faster or slower than the rate of fluid flow past the particle? For the agitated batch system, is the rate of dissolution faster or slower than the rates of turbulent mass transfer from the particle to the bulk solution? For the non-agitated batch system, is the rate of dissolution faster or slower than the rate of molecular diffusion of the dissolved constituents away from the ash particle? For chemical aspects influencing leaching, some questions can also be posed that help to define the leaching system further: In the flow through system, does the percolating liquid obtain a chemical equilibrium with the dissolving solid phase after some finite percolation distance or fluid velocity? In the agitated batch system does the liquid obtain a chemical equilibrium with the dissolving solid phase after some finite length of time and after some mixing energy is used to promote mass transfer between the solid and the liquid? In the non-agitated batch system, does the liquid obtain a chemical equilibrium with the dissolving solid phase after some longer finite length of time and after diffusion has resulted in mass transfer from the solid to the liquid? For all three scenarios, do some dissolution reactions never reach equilibrium? After initial dissolution, do some components reprecipitate? How does leachant pH affect dissolution? How do leachant complexing agents affect dissolution? How does oxidation-reduction potential affect dissolution? How do sorption processes redistribute back on the solid phase the originally dissolved solid phase components?

485

12.1.1 Physical Aspects of Leaching The physical aspects influencing leaching relate to the manner of contact between the liquid and the solid material. To introduce the concepts related to physical aspects of leaching adequately, the liquid and solid system, referred to here as the leaching system, must be described first.

The Leaching System

There are a number of scenarios that describe potential leaching systems. At one end of the spectrum, an example could be an uncapped landfilled residue which generates a leachate from rainwater percolating through the residue over a one hundred-year period under a field leaching scenario. At the other end of the spectrum could be three examples of leaching occurring in a laboratory setting: an artificial acid rain flowing through a residue in a column leaching experiment, an artificial acid rain and a residue agitated together in a flask in an agitated batch leaching experiment, and an artificial acid rain and a residue added to a batch flask under quiescent conditions in a non-agitated batch leaching experiment. While the ultimate goal may be predicting and modelling long term field leaching behaviour for disposed, treated or utilised residues, such data are not easily obtained under practical time frames. The use of batch and column leaching tests can simulate such long term scenarios. This is accomplished by increasing the throughput of liquid through columns or increasing the volume of liquid used in batch tests. Both procedures approximate exposure to naturally occurring precipitation over a longer time frame. Batch tests are more readily controlled and subjected to altering influences (temperature, gas partial pressures, complexing ligands, oxidation/reduction potential changes, etc.); column tests frequently better describe field condition fluid flow, mass transfer and dissolution mechanisms. Many leaching scenarios involve observational time frames of hundreds of years. Many of the leaching tests are operated on the order of hours, days or weeks. For leaching tests to simulate field conditions, the rate of dissolution must be speeded up. This is accomplished by (i) adding mixing energy to promote mass transfer from the solid to the liquid and (ii) using small masses of residues and large volumes of liquid so the solids dissolve into a large "sink" for dissolved constituents. Consider both the landfilled residue and the column leaching test in a simplified manner. A percolating liquid or leachant flows down through the residue. At the top of the residue, fresh leachant is not in chemical equilibrium with the dissolving solids.

486 However, as the liquid moves down through the ash material, more and more dissolution occur, increasing the strength of the leachate. At some point, further dissolution cannot occur because the leachate is saturated with respect to dissolved constituents. Residues subjected to a saturated, equilibrated solution will not leach. The scenarios can be made more complex by allowing the leaching front (the transition zone from the more dissolved region at the top of the column or pile to the undissolved region just below it) to move down through the material. Complexity is also introduced by acknowledging that some ash components that initially dissolve at the top of the pile or column can reprecipitate out from a saturated solution as another mineral further clown in the pile or column. Consider the two batch leaching scenarios in a simplified manner. One is agitated, and the other is static. When the materials are first added together, the liquid or leachant is not in chemical equilibrium with the dissolving solids. Unlike the flow-through system, the leachant and solids remain in contact with each other. There is no renewal, so the dissolving solid can continue to dissolve into the volume of liquid until equilibrium is obtained. The only difference between the agitated and static systems is the energy added to the system that promotes mass transfer from the solid to the liquid. In fact, the agitated system dramatically speeds up transfer; rates are one to two orders of magnitude higher than the flow-through system and many orders of magnitude higher than the static batch system. The scenarios can be made more complex by conducting batch tests sequentially. This is accomplished by removing the solid from one flask and adding it to a new flask with a fresh leachant that is not at equilibrium with the solid. Extrapolating further, it is easy to see how a sequential batch system can approximate a flow-through system. Particle Characteristics The particle surface area to volume ratio, the average particle size and internal pore structures in the residues all control the surface area where dissolution from the solid to the liquid can occur. Larger surface areas per mass or volume of ash can allow for more rapid dissolution at the surface. Frequently, ash particles exhibit widely different surface area to volume ratios, grain size distributions and internal pore diameters.

Invariably, the ash particles are composed of a number of major and minor minerals. Some are more soluble than others. Leachate pH and oxidation/reduction potential are governed by the predominant phases. Frequently, the dissolution of the more soluble major minerals governs the strength of the leachate. Fluid Flow Past Particles In the field and column leaching scenarios, liquid flow past the ash particles is typically facilitated by gravity. The porosity of the ash material and the amount of water above

487 the ash material attempting to flow through governs the velocity of flow of the leachant through the ash particles. Residues with low values of interparticle porosity and high degrees of impermeability will not transmit water. Moreover, the velocity of flow will be quite low. Taken to an extreme, a solidified, impermeable mass of particles will not conduct water; instead, water would be forced to flow around such a monolith. In an agitated batch leaching scenario, the relative rate of flow of a liquid past a particle is a function of the energy put into the system and the rate of fluid shear between adjacent parcels of water. Agitated systems where the particles remain in suspension and do not settle usually result in very high degrees of mixing and mass transfer. Fluid flow conditions for liquids that have not yet reached equilibrium with the dissolving solid play an important role in controlling the rate of dissolution by increasing rates of mass transfer from the particle to the bulk solution. For flow-through systems, the rate at which fresh leachant moves through the system can influence whether equilibrium is achieved. For conditions of slow fluid flow and highly soluble solid phases, the rate of dissolution can be faster than the rate that the dissolved constituents are advected or carried away and equilibrium between the liquid and solid can be achieved. Conversely, for conditions of fast fluid flow and relatively insoluble solid phases, the rate of advection is greater than dissolution and equilibrium will not be achieved. In completely static systems, where diffusion of constituents of dissolving solids are carried into the bulk solution by aqueous diffusive fluxes, the rate of diffusion can be the limiting process for controlling the rate dissolution, thus increasing the time required to achieve equilibrium.

The Local Equilibrium Assumption

For all of the identified leaching scenarios, the relative rates of dissolution and fluid flow or advection must be understood before determining whether there is equilibrium at the local level - between a dissolving particle and fluid in which the particle is situated. Fortunately, most leaching scenarios and tests are in equilibrium for the majority of the principal phases that can dissolve, This has great importance in how we mechanistically describe the chemical aspects of leaching.

12.1.2 Chemical Aspects of Leaching The chemical aspects influencing leaching relate to the fundamental processes controlling the solubility of solids, including: 9

the influence of pH on controlling solubility, the influence of solute-phase complexing agents that chemically bind to constituents from dissolved solids and hide them from forces promoting precipitation, thereby increasing dissolution, and

488

the role of oxidation-reduction potential in promoting solubilisation. Chemical aspects can also include reprecipitation processes or sorption processes whereby dissolved constituents return to the solid phase. Equilibrium Versus Kinetic Systems Many of the dissolution reactions that occur with ash are relatively quick. This permits for the use of equilibrium-based reactions and equilibrium reaction constants to describe the leaching system. Some reactions are relatively slow to extremely slow; equilibrium is never achieved and the reaction describing such a process is based only on kinetics. Kinetic systems are usually described as the rate of appearance of a solute in solution from a slowly dissolving solid as a function of various system parameters (temperature, pH, reaction stoichiometry). When a system is at equilibrium, it is possible to quantify the mass of an ash constituent that is in solution at equilibrium with the solid phase. As the system approaches equilibrium, the transfer of mass from the solid phase to the solution phase slows. The final equilibrated mass distribution between the two phases describes the equilibrium condition. Influence of pH on Dissolution Many solids exhibit a marked increase in solubility at both low and high pH values. Such solids are considered to be amphoteric. Other solids may exhibit maximum solubility at neutral pH values. The pH of the leaching system before leaching occurs is described as the initial pH of the leachant. The leaching system pH at equilibrium is usually governed by the dissolution of the major soluble phases in the ash particles. Frequently, the initial and equilibrium pH values differ widely, particularly if the liquid to solid ratio (L/S) or the mass of dry ash being leached to the mass of leachant, is low and the solid phase dominates the system. When L/S ratios are high, the reverse can hold true; the pH of the initial solution is close to the equilibrium pH because the solution phase dominates. As discussed below, other chemical aspects that influence leaching (complexation, redox, sorption) are influenced by system pH. Influence of Complexation on Dissolution Frequently, constituents in the solution phase of a system can bind to solid phase constituents and form a complex that remains in solution in a dissolved state. Such phenomena, governed by equilibrium reaction constants and susceptible to system pH, can increase the mass distribution of the solute in the solution phase and therefore promote solubilisation.

489 In the sequestered or complexed state, the bound solute is not available to participate in dissolution equilibria. The solid will then further dissolve to satisfy dissolution equilibria. The unbound free solute and complexed solute, when combined, increase the mass distribution in the solution phase. Such phenomena can occur when ash is disposed with MSW. Organic complexing agents in solution can bind to dissolving metals and increase the relative solubilisation of the ash solid. Influence of Oxidation-Reduction Potential on Dissolution Redox reactions behave like other chemical reactions with the exception that electrons, like protons (H* or H30*) or cations or anions, participate in the reaction. Reactions can be described thermodynamically or kinetically; a reduced solid is considered to be electron-rich. An oxidised solid is considered to be electron-poor. The presence or absence of oxygen or the presence or absence of oxidised or reduced mineral species can have an important influence on the stability of solids in ash. Under oxidising conditions, metal oxides remain an important phase for immobilising metals. Under reducing conditions, when oxygen is absent, metal sulphides become an important phase for immobilising metals. Metal sulphides are one of the most insoluble phases that can be created geochemically. The transition back and forth between oxidising and reducing environments can result in transitional redox states that promote the solubilisation of both oxides and sulphides. Influence of Sorption on Leaching Many of the solid phases in ash are very sorptive and are capable of immobilising dissolved constituents from the ash onto the ash surface via a number of sorption reactions. Sorption reactions can involve the formation of bonds that are relatively weak to those that are relatively strong. Strong binding implies that the adsorbing solute would rather stay on the particle surface than go back into solution. Many sorption processes are very pH dependent. For example, at a specified pH edge, a sorbed solute will be displaced by a proton or a hydroxyl ion. Below (or above) that pH edge, the solute remains sorbed. The distribution of trace constituents in ash is such that sorption can easily immobilise the dissolving solute. Frequently, aged ash has a high degree of sorptive surfaces which promote trace metal immobilisation. 12.1.3 Leaching Tests Leaching tests are employed to simulate field leaching scenarios. The three basic leaching concepts, the flow-through column, the agitated batch system and the non-

490 agitated batch system form the basis for most scientific and regulatory leaching tests that are used. In almost all cases, a number of tests are required to describe leaching behaviour. One concept that is important to note is that leaching data from many tests show a rather uniform leaching behaviour when ash leaching data are normalised. Raw leaching data in the form of concentration has little meaning until it is transformed into units which are comparable between leaching tests. The normalisation process requires that all data be reported as a function of the US ratio. Additionally, the mass of an ash constituent that is leached is also reported as a function of the dry weight of ash leached. This is usually termed "release" (mass of element leached per dry weight of ash) or "fraction leached" (mass of element leached per mass of element in the dry residue). Such normalisation processes are needed to look at mass fractions released from ash as a function of time or the L/S ratio. This then equates to "flux" (discrete mass of element released per unit time) and "cumulative flux" (total mass of element released per unit time). Leaching tests employ a number of reactor configurations and control measures for stabilising pH, pE (or Eh), temperature, etc. The tests are conducted for specific periods of time. A number of initial leaching solutions can be used. The researcher or regulator must select the types of tests that describe the leaching system to be studied and are amenable to leaching modelling efforts.

12.1.4 Leaching Modelling The ultimate goal for the selection of an appropriate leaching test or the evaluation of field leaching data is the ability to interpret the observed leaching behaviour mechanistically. By understanding field leaching behaviour, predictions can be made and modelled for how leaching behaviour will change over time and under various management scenarios. This ultimately provides the researcher or regulator with the ability to develop management options based on predicted leaching behaviour. Two principal approaches are used to model leaching. One involves the use of computer codes that use an iterative technique to simultaneously solve equilibrium expressions and mass balances. Such thermodynamic equilibrium models rely on the assumption that equilibrium will occur between the solid phase and the liquid. These models are valuable in predicting how the alteration of leaching system pH, redox potential, solution-phase complexation and sorption can control the equilibrium of the system. The second approach couples the equilibrium-based phenomenon of the first approach with an advecting fluid flow through the ash material. Such models usually employ mass transfer and dissolution/reprecipitation mechanisms. The multicomponent approach is particularly useful in predicting long term leaching behaviour.

491 It is anticipated that thermodynamic equilibrium models and multicomponent advection models will be merged with fate and transport models so as to provide better larger scale modelling capabilities for ash utilisation, storage or disposal sites.

12.1.5 Unified Approach to Leaching A unified approach to leaching involves the careful characterisation of the residues to be leached, the thoughtful selection of appropriate leaching tests, and the modelling of leaching behaviour. When such approaches are used, systematic leaching behaviour of waste materials is usually observed.

12.2 THE SOLID PHASEILEACHANTISOLUTE LEACHING SYSTEM 12.2.1 The Leaching System Concept Leaching can be broadly defined as the mobilisation, extraction or washing of soluble constituents from a solid phase by a contacting solvent. To discuss the phenomena of leaching appropriately, it is necessary to define the system and provide nomenclature. As shown in Figure 12.1, the residue constitutes the solid phase. The leaching of the solid phase by a solvent (or leachant) causes solid phase constituents to become solutes in the solvent. The solute-solvent system, initially described by the characteristics of the leachant, becomes more solute-dominated and takes on the characteristics of a leachate. Adherence to this style of nomenclature in the text will hopefully provide clarity and precision in the presentation.

12.2.2 A Multiphase Heterogeneous System This simplified system is much more complex in controlled leaching tests and infinitely more complex in dynamic flow-through field leaching scenarios. As depicted in Figure 12.2, the solid phase is a structurally and mineralogically complex heterogeneous material. Basic reactions such as acid-base chemistry and redox chemistry are governed by numerous interactive equilibrium-based reactions that are sometimes kinetically dissimilar. Precipitation, dissolution and sorption phenomena are also kinetically controlled and interactive. Kinetics can govern the appearance of solutes in the leaching solution. Both slow kinetic mechanisms and mass-transfer constraints can prevent equilibria from being attained. The internal porosity, tortuous path lengths of particles and internal reaction mechanisms constitute an internal resistance to diffusion of solutes or solvents into and out of the particle. The fluid boundary layer external to the particle constitutes an external resistance to diffusion of solutes of solvents between the particle surface and the bulk liquid.

492 Figure 12.1 The Leaching System

Leaching System Variables: pH, pE,L/S, I, T Leaching Solution (Leachate)

! i

Solvent ( (Leachant)

Weathering

Surface

siT!io

Reactions

Processes

9Complexation Solute

tl

1

Add-BaseReactions

Dissolution/ PrTlpltaj~on

J,,~ T

Phase

Cobol

coit+o'

Solution

AgingReactions

Solid Phase In fact, the leaching of incineration residues probably constitutes one of the more complex solid phase/leachant/solute systems that has ever been studied. For example, a typical leaching scenario can involve a solid phase containing multiple mineral and amorphous principal phases that exhibit differential solubility and are thus incongruent with each other. The leaching solution, in equilibrium with these phases, can contain a multitude of dissolved or colloidal constituents that are also in equilibrium with each other. The system can be exposed to dynamic hydraulic flow conditions where the velocity of the fluid relative to the particle varies and thus boundary layer thicknesses and mass transfer rates vary between the bulk liquid and the particle. The presence of a leaching solution can be intermittent because of the cyclic nature of precipitation or infiltration. The system pH and redox potential (pE or Eh) can dynamically vary. Given the fact that interest in this topic in the scientific and engineering communities increased only ten to fifteen years ago, much inference must be drawn from the work in fields of coal ash leaching, aquatic chemistry, marine geochemistry and terrestrial geochemistry to assist in the comprehension of leaching phenomena and in the design, selection and interpretation of leaching tests. This bridge can be made because the leaching system shown in Figures 12.1 and 12.2 is not that different from soil-porewater systems, soil-contaminated groundwater systems, marine sediments, mine tailing piles or coal ash piles. Central to leaching phenomena is the concept of solid phase control (see Figure 12.1 ). During solid phase control, the relative mass (or chemical poise) of the solid phase to that of the leachant is high. The inherent chemical characteristics of the solid phase dominate the solvent, producing saturated solutions. Under such situations, the residue is viewed as controlling leaching. Under certain leaching scenarios, solid phase control can transition into solution control (see Figure 12.1). Under solution

493 control, the mass (or chemical poise) of the solid phase to that of the leachant is small. The infinite volume characteristics of the solvent dominate the nature of leaching solution, acting as a sink for the leaching solutes. More precise descriptions of solid phase or solution control are discussed in Chapter 14. The other aspects of leaching shown in Figure 12.1 are solution behaviour, precipitation, dissolution, weathering and adsorption, as well as controls on that system. The pH, redox, liquid-to-solid ratio and time are the topics of Chapter 13. Figure 12.2 Schematic of Heterogeneous Complex Leaching System Key

0 MetalCation SlowInterstitial

O

SeepageVelocity o

y AnionicUgand

~ S u r f a c e

SorptlonSite

o

Y

Y

~

Y Yo

Preci!itatlon

Dissolution

<

O

Y

x Sorptlon

Complexation

~ SolutionPhase

After Kirkner et al., 1986

12.2.3 Leaching Scenarios There are a number of leaching scenarios that can be envisioned for incineration residues. As shown in Table 12.1, scenarios range from piled or landfilled material where some type of regulatory and engineering control is imposed, to utilisation applications where beneficial use is envisioned, to various leaching tests. For each scenario, typical leaching time frames, leaching regimes and leachants are shown. The scenarios depict whether the flow-through, agitated batch, or non-agitated batch systems apply. It is important to note that many potential leaching scenarios can assume a time frame of geologic time (e.g. 10,000 years). The examples in no way

Table 12.1

Leaching Scenarios for Incineration Residues Scenario

*Piled Residue (no cap, 100 year application

Leaching Leaching Regime Solid Phase or Solution Control Time Frame (years) Short Term Long Term 0-100 Flow Through System, Solid Phase Solid Phase Intermittent Saturation, Granular Material

lnlial Solvent (Leachant)

Initial Solvent PH

Precipitation

4.0-6.0

*Landfilled Residue (leachate removal, cap after 3-10 yr)

3-15

Flow Through System, Solid Phase Intermittent Saturation, Granular Material

Solid Phase

Precipitation

4.4-6.0

*MSW/Residue Codisposal (leachate removal, cap after 3-10 yr)

3-15

Flow Through System, Intermittent Saturation, Granular Material

Solid Phase

Solid Phase

Landfill Leachate

4.5-5.5

*Granular Fill (no cap, saturated groundwater 100 yr application) *Road Subbase Utilisation (15 yr application)

0-100

Flow Through System, Solid Phase Continuous Saturation, Granular Material Flow Through System, Solid Phase Very Intermittent Saturation, Granular Material

Solid Phase

Groundwater

6.0-8.5

Solid Phase

Road Run-off1 Groundwater

5.5-6.5 (run-off) 6.0-8.5 (groundwater)

-Road Paving Utilisation (15 yr application)

0-15

Very Intermittent Saturation, Monolithic Material

Solid Phase

Solid Phase

Precipitation

4.4-6.0

*Column Lysimeter

0-1

Flow Through System, Solid Phase Intermittent Saturation, Granular Material

Solid Phase

Artificial Precipitation

4.0

0-15

*Agitated Batch System

1-2 days

Agitated Batch System, Saturated, Granular Material

Solution Phase

Solution Phase

Artificial Precipitation

4.0

~Non-AgitatedBatch System

64 days

Non-Agitated Batch System, Saturated, Granular Material

Solution Phase

Solution Phase

Artificial Precipitation

4.0

495 represent all residue leaching scenarios, yet they illustrate how the systems can vary and how the perspective of time must be considered. 12.3 RESIDUE PARTICLES AS A SOLID PHASE

It is important to have a clear understanding of ash particle size, porosity and morphology and the role they play in influencing the ash leaching system. These particle characteristics influence the relative surface area available for chemical reactions to take place. They also influence the magnitude of the diffusion path length and the type of internal and external mass transfer resistances that occur as solutes or solvent diffuse into or out of particles. 12.3.1 Particle Characteristics

Theis and Gardner (1990) have assembled pertinent information from the literature on particle size, specific surface area, specific gravity and bulk density for MSW fly ashes and bottom ashes. The data are shown in Table 12.2. The fly ash residues are characterised as fine grained with high levels of specific surface area. The bottom ash residues are characterised as lightweight, medium-grained particles with very high levels of specific surface area. Table 12.2 Physical Properties of Ash Particles Property Particle Size Range, mm Specific Surface Area, m2/g Specific Gravity Bulk Density g/cm 3 Theis and Gardner, 1991

MSW Fly Ash

MSW Bottom Ash

0.01-230

295-9500

2.83-36.93

9.4-46.3

2.1-4.04

1.52

0.78-1.04

0.37-0.73

Gardner (1991) conducted detailed BET surface area analyses and mercury porosimetry studies on fly and bottom ashes from RDF, rotary kiln mass burn and starved-air mass burn incinerators. The data are presented in Table 12.3. The data indicate that the bottom ashes, though larger in particle size, contain significantly higher specific surface areas, larger total pore areas (based on mercury intrusion) and smaller average pore diameters. The bottom ashes studied typically had specific surface areas of 4 to 30 m2/g, total pore areas of 1.41 to 19.56 m2/g and average pore diameters of 0.07 to 2.11 IJm. These data strongly suggest that internal diffusional resistance and chemical retardation can be significant in bottom ashes.

496 Table 12.3 Physical Properties of Selected Fly Ashes and Bottom Ashes Property

MSW Fly Ash

MSW Bottom Ash

3-18

4-30

Specific Surface Area, m2/g Bulk Density, g/cm 3

0.61-1.69

0.7-1.2

Skeletal Density, g/cm 3

0.89-2.83

1.74-2.48

Total Intrusion Volume, cm3/g

0.23-0.71

0.21-0.74

Total Pore Area, m2/g

0.06-0.28

1.41-19.56

Average Pore Diameter, IJm Gardner, 1991

9.0-15.5

0.07-2.11

Fly ashes had lower specific surface areas (3.18 m2/g), significantly lower total pore areas (0.06 to 0.28 m2/g) and significantly higher average pore diameters (9.0 to 15.5 IJm). Such data suggest that internal diffusional resistances can be much less severe in fly ashes.

12.3.2 Particle Morphology Figure 12.3 is a composite of SEM micrographs of grate ash, grate siftings, boiler ash and dry scrubber residue. Figure 12.3 depicts the morphology and crystalline nature of a typical grate ash particle. The sample is from a large Canadian mass burn facility. The grate ash was collected prior to quenching. The particle is large and contains a high degree of externally-connected vesicles. These are formed by trapped gases that escape from material in a plastic state as it cools during quenching. Many of the vesicles have relatively large diameters, on the order of 100 to 1000 IJm in size. There are a number of crystalline phases on the particle surface and in the vesicle. Some of the phases exhibit high degrees of crystallinity, whereas some are more glassy, indicating that melting and rapid cooling have occurred. Coupling this information with mercury porosimetry data suggests that diffusion path lengths are long (up to 10,000 IJm) and highly tortuous. These data are consistent with the Gardner (1991 ) data which suggest a very high degree of internal diffusional resistance. Diffusional leaching processes are therefore much more complex in bottom ash. Figures 12.3 also depict the morphology of grate siftings, boiler ash and dry scrubber residue. The grate siftings are similar in characteristic to bottom ash; however, typical grain sizes are smaller and molten metal beads, glass fragments, soil, and metal pieces are commonly found. Slag particles are also present. The boiler ash and dry scrubber

497 Figure 12.3 SEM Micrographs of Grate Ash (a), Grate Sifting (b), Boiler Ash (c), and Dry Scrubber Residue (d) V=vesticle, S=spherical particle, C=crystalline particle, MM=melted metallic materials

498 residue are characterised by large specific surface areas because of their relatively small particle size and highly polycrystalline nature. The pore sizes are relatively large because of the agglomeration of very fine particles. Diffusion path lengths are relatively short (100 IJm).

12.4

FLUID FLOW, DIFFUSION AND MASS TRANSFER

In field and laboratory column leaching scenarios, the ash particle is stationary and a leachant flows through or around the ash particles and carries away dissolved ash constituents. In certain batch leaching scenarios, agitation is used to cause fluid to flow past particles and accelerate the dissolution of constituents in the ash. In other batch leaching scenarios, fluid flow or agitation is absent, which permits only strict molecular diffusion and Brownian motion to carry away dissolving constituents. The rate at which constituents are carried away (via advection) plays a fundamental role in influencing chemical reactions associated with leaching. An appreciation for the concept of fluid flow and advection is needed to characterise the interactive roles of diffusion and mass transfer in leaching phenomena.

12.4.1 Fluid Flow Through Residues In many field leaching scenarios, a leachant flows through or around a pile of ash. The flow can be approximated in the laboratory with columns or outdoors in small lysimeters. For most leaching scenarios, the flow of a leachant through or around incineration residues is governed by Darcy's Law. Typically, the flow is unsaturated. Examples include landfill leaching, road-base leaching, pile leaching and column leaching of residues. Flow, however, can be saturated under certain landfill, pile, road sub-base and column leaching scenarios. The factors describing saturated fluid flow in porous media are derived from Darcy's Law: Q= ~ where

~

(12.1)

Q is the quantity of water moving through an incremental volume per unit time (m3/d), K is the hydraulic conductivity constant; it is a function of porosity and leachant kinematic viscosity and density (m/d), A being the cross-sectional area of the incremental volume through which flow is occurring (m2), dh/dl is the hydraulic or gravitational gradient or headloss across a depth or length of the incremental volume through which flow is occurring (m/m).

499 Rearranging equation (12.1 ) produces: K -

Q

A

dl dh

dl dh

-x;

(12.2)

where v is a measure of saturated velocity or specific discharge. There are limits to which Darcy's law is valid for flow through porous media. For fine grained, impermeable materials there are threshold hydraulic gradients below which flow does not occur. At very high rates of flow, Darcy's law is also invalid. The dimensionless "Reynolds number" is used to determine when this happens. The Reynolds number (Re) is a ratio of inertial to viscous forces during flow. It is used to determine whether flow past a particle is laminar or turbulent. Typically mass transfer to or from a particle surface is much higher under turbulent conditions. The Reynolds number is defined as: Rewhere

pvd p

(12.3)

p is fluid density (kg/m3), v is fluid specific discharge (m/s), d is particle diameter (m), and IJ is dynamic viscosity(kg/m.s).

Darcy's law is valid when the Re is below values ranging from 1 to 10. A number of studies have characterised saturated hydraulic conductivities of combustion residues. Hartl6n and Elander (1986) documented saturated hydraulic conductivities (K) of 10 s to 10-7 m/s for bottom ashes, 10 .6.5 to 10 .9.5 m/s for fly ashes and 10 .8.5 to 10 1~ m/s for scrubber residues. The steady-state unsaturated flow of a leachant through and around porous ash particles can be determined from a modified form of Darcy's law: Q : mei

where

ho-z

z

+

-

I dh

[

(12.4)

Q is the quantity of water moving through an incremental volume per unit time (m3/d), K~ is the hydraulic conductivity under unsaturated conditions (m/d), A is the cross-sectional area of the incremental volume through which unsaturated flow is occurring (m2), (hc-z)/z is the capillarity or surface tension gradient (m/m), and dh/dl is the hydraulic or gravitational gradient through which unsaturated flow is occurring. (m/m)

500 Typically, K~ approaches K logarithmically as saturation approaches 100%. During unsaturated conditions the effective hydraulic conductivity of a residue is a function of particle size, capillarity and moisture content. There are no published effective hydraulic conductivities (Ks) for combustion residues. For typical hydraulic gradients and typical saturated conductivities, one would expect fluid velocities to range from 1 x 107 to 1 x 10-4 cm/s, which are low enough to maintain low Reynold's numbers (<5) and laminar flow during saturated conditions. For typical hydraulic gradients and typical unsaturated effective hydraulic conductivities, one would expect lower seepage velocities. Such values would range from 1 x 10~ to 1 x 10~ cm/s. These values are well within the range of low Reynold's numbers (10 .2 and less) and laminar flow conditions. The implication of laminar flow through and around combustion residues is that fluid and chemical boundary layers are large, constituting an external resistance to diffusion that may limit fast reactions. Such rate-limited reactions have been observed by Theis et al. (1992, 1993). For slow reactions at or inside particles, external diffusional resistance is probably not rate limiting. 12.4.2 Fluid Flow Past Particles in Suspension In non-agitated suspensions, particles settle at a terminal velocity according to their buoyant weight and the resisting forces resulting from drag. Stoke's Law describes particle settling velocities: v = where

(pp-pf) g d 2

18p

(12.5)

V is the settling velocity, p, and pf are particle and fluid density, respectively, g is the gravitational constant, and d is the particle diameter and p is the dynamic viscosity of the fluid.

Stokes law can be used to approximate velocities for tumbled agitation systems. In continuously agitated suspensions, the weight of the suspended particles is continuously supported by the fluid. Any inclination toward settling is counterbalanced by random upward movements of fluid parcels during turbulent flow. The intensity of turbulence, the scale of turbulence and kinematic eddy viscosity of the turbulent fluid parcels keep the particles suspended. Kinematic eddy viscosity is used to describe fluid shear between turbulent parcels and therefore between parcels and particles.

501 For most continuously agitated systems, kinematic eddy viscosity and shear is high. This means that fluid and chemical boundary layers are very compressed and not likely to constitute a resistance to diffusion or limit fast reactions at the particle surface. By increasing mixing energy, external diffusional resistances can be virtually abolished. 12.4.3

D i f f u s i o n a l P r o c e s s e s and Internal M a s s T r a n s f e r C o n s i d e r a t i o n s in Residues

It is useful to examine the role of bulk diffusion and internal mass transfer resistance on leaching (Hinseveld, 1991). As shown in Figure 12.4, Fick's second law of diffusion is applicable for leaching from a small unit area in a hypothetical ash particle: 8C 82C - D 8t e 8X 2 where

(12.6)

C is the concentration of a potential solute in the particle (mole/cm3), t is the leaching time frame (seconds), x is the distance from the leaching region in the particle to the particle shell (cm), and De is the effective diffusion coefficient (cm2/second).

As indicated previously, the diffusion path length in the particle may be tortuous and long and therefore should not be expressed as a radial distance. Assuming an area through which diffusion out of the particle is proportional to the porosity (e) in the particle, it is evident that the effective diffusion coefficient relates to the true molecular diffusion coefficient: m e

where

m

s

m .[

(12.7)

D m is the molecular diffusion coefficient (cm2/sec), e is the porosity in the particle (cm3/cm3), and r is a tortuosity factor (unitless).

Estimates of porosity (0.10 to 0.30) and tortuosity (1 to 1000) means the effective diffusion coefficients are up to 104 times smaller than molecular diffusion in dilute solutions. Li and Gregory (1974) provide a table of Dr, values for simple anionic and cationic species in dilute solutions at different temperatures.

502 Figure 12.4 Conceptual Model of Ash Particle Dissolution

Leached

Internal Diffuslonal Resistance ~

Boudary

~.r ~ ,~,,.~- I X Thickness ~'r

Thickness

1

,

~

-

~

External Dlffuslonal Resistance Flux of Reactants Inward

radius of core ra.,u.

of shell J

Hypothetical Ash Particle /'

'/' '

Surface Reaction

f I

,

I

I

, I I

Flux of Products Outward

Ci

Shrinking Core Concentration

Co

I

I Co

Radial Distance

After Hinseveld, 1991

Concentration of Inward Diffusing Reactant Species

Concentration of Outward Diffusing Product Species

503 An analytical solution to equation (12.6) is available in Crank (1975). As noted by Hinseveld (1991), the underlying assumptions of applying Fick's second law (flat element, infinite medium, one-dimensional diffusion, equimolar diffusion, constant surface concentration, constant De) make this an overly simplistic approach to modelling leaching. The bulk diffusion process can be modified for internal mass transfer resistance inside the ash particle in two steps (Hinseveld, 1991). The first part is accomplished through the use of a unitless linear distribution coefficient, KD," KD :

where

Cmobile

(12.8)

Cimmobile

Cmob,eis the concentration of the mobile solute in the particle and Cimmebi,eis the concentration of the immobile solute in the particle.

The new form of equation (12.6) is therefore: 8C

8t

De

(I +K d)

8 2C

(12.9)

8x 2

Second, equation (12.9) can then be further modified with a dissolution constant, k, for the immobile phase so that: 8C 82C 8Cimmobile + k ( C s -Cmobile ) (gt - me OX 2 Ot where

(12 10) "

k is dissolution constant (molal) and Cs is the surface concentration of the solute (molal).

Both linear distribution and dissolution decrease the internal flux of constituent solutes out of the ash particle. Hinseveld (1991) provides a solution to equation (12.10). The use of this approach theoretically demonstrates that internal diffusional resistances can act as a rate-limiting step to leaching phenomena. Unlike external resistances, these internal resistances cannot be abolished unless residues are dramatically size reduced. 12.4.4 External Mass Transfer Considerations in Residues

Consider the nature of the fluid flow past a hypothetical ash particle as depicted in Figure 12.4. Molecular diffusion, advection and hydrodynamic dispersion are three transport processes governing external mass transfer between the bulk fluid, the boundary layer and the particle surface (Zachara and Streile, 1991).

504 The Peclet number (Pe) is a useful parameter to investigate the relative role of fluid velocity and diffusion on transport processes. Pe is defined as: dv P e - Dm (1211) where

d is the average particle diameter (cm), v is fluid velocity (cm/s), and Dm is the molecular diffusion coefficient (cm2/s).

Rose (1977) makes use of the Peclet number for characterising the influence of fluid flow on the nature of transport mechanisms between particles and the bulk liquid. The five regimes are depicted in Table 12.4. Also shown in Table 12.4 are likely fluid velocities for APC residues of characteristic diameter of 0.01 cm (100 pm) and bottom ash residues of characteristic diameter of 0.1 cm (1000 pm) associated with each domain (assuming a typical solute diffusivity of 10 .7 cm2/s). The results indicate that molecular diffusion is the predominant transfer mechanism for anticipated fluid velocities through residues. Table 12.4 Peclet Numbers and External Mass Transfer Mechanisms Typical Seepage Velocities (cm/sec) Regime

Dominant Process

APC Residues a

Bottom Ash b

Molecular Diffusion

3x10 6

3x10 -7

Molecular Diffusion & Advection

3x10 1 to 5 x 10~

3x10 .7 to 5x10 6

5 < Pe < 1000

Advection But Molecular Diffusion Cannot be Neglected

5x10 -5 to l x l 0.2

5x10 -6 to 1xl 0 .3

1000 < Pe < 150,000

Advection

Not Likely

lx10 3 to 1.5x10 -1

Pe < 0.3 0.3

< Pe <

5

Pe > 150,000 Dispersion Not Likely Not Likely a Based on characteristic lengths of 0.01 cm, and on diffusion coefficients of 10.7 cm2/sec. b Based on characteristic lengths of 0.1 cm and on diffusion coefficients of 10.7 cm2/sec.

12.5 THE LOCAL EQUILIBRIUM ASSUMPTION It is important to note how fluid flow past a particle can influence our estimate of the local equilibrium assumption (LEA). Conceptually, for sorption reactions at surfaces, if a chemical reaction at the particle surface is fast relative to the rate of advection from fluid flow, then pointwise equilibrium may be obtained and the LEA is valid. If a reaction is slow, however, advection will sweep reactants away from reaction sites

505 (Jennings, 1987). This condition is likely to occur for many surface reactions processes. It is outlined in more detail in Chapter 13. REFERENCES Crank, J. The Mathematics of Diffusion. Clarendon Press, Oxford, U.K., 1975. Gardner, K.H. "Characterization of Leachates From Municipal Incinerator Ash Materials." Master's Thesis, Clarkson University, 1991. Hartl~n, J. and P. Elander. "Residues from Waste Incineration-Chemical and Physical Properties." Swedish Geotechnical Institute Report SGI Varig 172, Linkoping, Sweden, 1986. Hinseveld, M. "Towards a new approach in modelling leaching behaviour." Waste Materials and Construction. Edited by J.J.J.M. Goumons, H.A. van der Sloot and Th.G. Aalbers. Elsevier, Amsterdam, p. 331, 1991. Jennings, A.A. "Critical chemical reaction rates for multicomponent groundwater contamination models." Water Resources Res. 23, pp. 1775-1784, 1987. Kirkner, D.J., A.A. Jennings and T.L. Theis. M.ulti-solute Subsurface TransPort Modellin,q for Ener,qy Solid Wastes. DOE/EVIl0253-5, 1987. Li, Y.H. and S. Gregory. "Diffusion of ions in sea water and in deep-sea sediments." Geochim. Cosmochim. Acta. 38, pp. 703-714, 1974. Rose, D.A. "Hydrodynamic dispersion in porous materials." Soil Sci. 123, pp. 277-283, 1977. Theis, T.L. and K.H. Gardner. "Environmental assessment of ash disposal." CRC Crit. Rev. Environ. Control 20, pp. 21-42 1990. Theis, T.L., R. lyer and K.H. Gardner. "Dynamic evaluation of municipal solid waste ash leachate." EmerQin.qTec.hnolo.qies for Hazardou.s Waste Management Edited by D.W. Tedder. American Chemical Society, Washington, DC, p. 605, 1992. Theis, T.L., R. lyer and K.H. Gardner. "Dynamic evaluation of municipal waste combustion ash leachate." Proceedings of the International Specialty Conference on Municipal Waste Combustion, Air and Waste Management Association. Pittsburgh, PA, p.680, 1993. Zachara, J.M. and G.P. Streile. Use of Batch and Column Methodolo,qies to.Assess Utility Wast.e Leaching and Subsurface Chemical Attenuation. EPRI Report EN-7313, EPRI, Palo Alto, CA, 1991.

This Page Intentionally Left Blank

507

CHAPTER 13 - CHEMICAL ASPECTS OF LEACHING

The intent of this chapter is to: (1) Use a classical geochemical approach to understanding leaching. (2) Provide the fundamentals necessary to design and use appropriate leaching tests. (3) Mechanistically evaluate the leaching data. (4) Accurately model the leaching phenomena. Where appropriate, mention is made in the text to references in the literature that provide more detailed information on the concepts that are presented. 13.1 LEACHING CHEMISTRY FUNDAMENTALS

The information presented here draws on research that has been conducted in the areas of aquatic chemistry, terrestrial geochemistry, marine aqueous chemistry, sediment biogeochemistry, and coal ash leaching. Reference is made as to how these systems are similar to MSW residue leaching systems. The reader may wish to delve further into these subjects. Useful texts or monographs include the following: for aquatic chemistry principles, Butler (1964), Pankow (1991) and Stumm and Morgan (1981); for geochemistry principles, Lindsay (1979), Sposito (1981, 1989), Drever (1988), Garrels and Christ (1965), Pytkowicz (1983), and Nordstrom and Munoz (1986); and for coal ash leaching, the work of de Groot et al. (1987) and various monographs produced by the Electric Power Research Institute (Rai and Zachara, 1984; Rai, 1987; Krupka et al., 1988; Zachara and Streile, 1991). Throughout this chapter, the nomenclature of Pankow (1991) is generally used, as shown in Table 13.1. Activities are denoted as { } and concentration as [ ]. Generally, "M" stands for metal cation, "S" for base salt, "H" for proton, "OH" for hydroxyl, "A" for conjugate base of an acid and "L" for ligand. 13.1.1 Thermodynamic Equilibrium Models Versus Kinetic Models

The ability to understand leaching processes, utilise leach tests, interpret the data and ultimately model leaching requires an understanding of reaction kinetics. All leaching processes involve chemical reactions (Figure 13.1). Homogeneous reactions are those that take place in a single phase (e.g. solution complexation reactions). Heterogeneous reactions are those that take place in the presence of at least two phases (e.g. a solution containing a solid phase). Ash leaching systems are heterogeneous systems.

508 Table 13.1 Nomenclature Example Reactions ThermodynamicParameters Names 9GAS PHASE partial pressures Pi, Pj, pk .... activity coefficients 2H2 + 0 2 =2H20 Yi,g, Yj,g, Yk,g.... f,. fj. f. fugacities 9AQUEOUS PHASE [i], [j], [k] . . . . concentrations ....

Common Units atmospheres dimensionless atmospheres

mol/L (M) or mol/kg (m) H2CO3 = HCO3 + H+ Y~,Yj, Yk.... activity coefficients dimensionless {i}, 0}, {k} .... activities mol/L (M) or mol/kg (m) reprinted with permission from Pankow, 1991. Copyright Lewis Publishers, and imprint of CRC Press, Boca Raton, Florida. 9 Figure 13.1 Chemical Reactions

IICHEM'OA'"E'OT'ONSI LEVEL A

~v~

J 'SUFFICIENTLY FAST' I AND REVERSIBLE

I~ o ~ o ~ o o ~

I

i"~~176176

I

'INSUFFICINETLY FAST' AND/OR IRREVERSIBLE

I~ o ~ o ~ o o ~

I

I

I"~~176176

Adapted from Rubin, 1983 with permission, copyright by the American Geophysical Union Most chemical reactions are reversible, but some are relatively fast and others relatively slow. It is important to understand the specific time frame under which these reactions occur, i.e., from fractions of a second to millennia. Aqueous phase reactions such as acid-base reactions or complexation reactions are usually fast, in the order of fractions of a second to a few seconds (Stumm and Morgan, 1981). Surface sorption reactions such as lead or cadmium adsorbing to oxide surfaces are in the order of hours to days (Cole, 1983). Precipitation, dissolution, and redox reactions involving solid phases are slow, in the order of hours to years (Stumm and Morgan, 1981).

509 Diagenetic changes in crystal structure and mineral predominance and weathering reactions are very slow, in the order of months to millennia (Berner, 1981 ). In contrast, some leaching tests vary in their time frame from hours, to days, to weeks and occasionally years. Clearly, the time frame of the leaching scenario will dictate the type of reactions that can be observed. Figure 13.2 shows the change in solute concentration of AI, Ca and OH extracted from bottom ash from a Swiss incinerator by mixing the ash in a closed system. The reactions are heterogeneous in that a solution phase and a solid phase are both present. Initially, aluminum is inclined to dissolve from the solid phase and go into solution; however, the system is not yet in equilibrium. Further, calcium ions and hydroxide ions are inclined to precipitate out of solution until the solid is in equilibrium with the solution concentration of these components. As these reactions progress, high concentrations of hydroxide and calcium decrease to a constant equilibrium value (5.6 mm) while aluminum concentrations increase to a constant equilibrium value (5 to 6 mm) by about 10 to 100 hours. In the time frame of the observer, the equilibrium between the solution and the solid phase under the conditions of the system could be viewed as both fast (hours) and reversible (some reprecipitation is occurring at equilibrium). Consequently, if a leaching experiment or leaching test were designed to evaluate the equilibrium relations of calcium, hydroxide and aluminum with a bottom ash solid phase, and the time frame was greater than 10 to 100 hours, then the system could be said to be governed by fast, reversible reactions and that this condition could be described using thermodynamic equilibrium models. Figure 13.2 Thermodynamic Equilibrium Model of Swiss Bottom Ash CONCENTRATION (mmol/L) 14f

12~ OH" 10

81- Ca2+

2 ~- AI3+ 0-

0.02

-r

0.6

I

6

....

I

10

I

30

EXTRACTION TIME (hours) After Stampfli et al., 1990

60

600

510 Conversely, Figure 13.3 depicts a heterogeneous reaction where the change in molybdenum concentration in the aqueous phase in the presence of MSW bottom ash does not reach equilibrium over the time course of the experiment. The reactions were conducted in a closed system under batch-stirred conditions (Comans et al., 1993). By 200 hours, the aqueous phase concentration of molybdenum has clearly not reached a constant value. If the leaching tests or experiments were designed to evaluate the equilibrium relations of molybdenum with a bottom ash solid phase, and the time frame was less than 200 hours, then no confidence exists to state that equilibrium was achieved because the system is governed by relatively slow kinetics. It could, however, be modelled using a kinetic model. Figure 13.3 Kinetic Model of Molybdenum Dissolution from MSW Bottom Ash

Molybdenum 0.4 E 0 L_J e-'l~

03T

n .............................

71

(DE

O~ C 0 0

0

0

After Comans et al., 1993

,

100

,I

2O0

time (hours)

As explained previously, the ash system is made complex by the fact that the solid phase is multicomponent, containing many discrete phases each with different solubilities. Like coal ash systems (Dudas, 1981; de Groot et al., 1987), the MSW combustion residue leaching system is dictated by the initial release of readily soluble discrete inorganic salts (e.g. CaCI2, NaCI, KCI) admixed with much more complex glassy aluminosilicate solid phases that leach very slowly. The most widely used method for obtaining chemical speciation information is the thermodynamic-equilibrium approach (Pankow, 1991). With this method, it is assumed that the leaching system is at equilibrium; the system experiences no changes with time in chemistry, temperature or pressure, and moreover has no tendency to change. The

511 equilibrium state is controlled by the thermodynamic energy levels of all the chemical constituents involved, including all gases, dissolved species and residue solid phases. When modelling an actual leaching system, there are likely to be numerous chemical reactions. At system equilibrium, all of the corresponding equilibrium expressions and the various mass balance conditions governing that situation must be satisfied simultaneously. Such systems are usually complicated by the fact that many of the reactions involve species that are also participants in other simultaneously occurring reactions. In recent years, sophisticated computer codes that compute equilibrium speciation even for very complicated situations have been developed. These models will be described in Chapter 15. As discussed earlier, kinetic models are occasionally needed to represent natural water systems, but they are often difficult to apply. Their application requires a knowledge of specific initial conditions as well as all of the pertinent kinetic rate constants (Pankow and Morgan, 1981a, 1981b). There are some major problems with applying kinetic models to leaching (Pankow, 1991). Although the differential equations used for kinetic reactions can be solved using numerical integration techniques, most leaching systems involve complex coupled equations. In addition, rate constants are typically not known as well as equilibrium constants, and complex kinetic laws are often observed (Pankow, 1991).

13.1.2 Ionic Strength, Ion Activity, Activity Coefficients Many leaching solutions contain high concentrations of dissolved solids or ionic solutes. The presence of these solutes imparts an ionic strength to the leaching solution. Frequently, MSW incinerator residue leachates have ionic strengths (I) that exceed those found in seawaters and brines. At moderate to high ionic strengths, the leaching system departs from ideal behaviour since reactants can be blocked from colliding with each other, thus diminishing their apparent concentration in solution. Reactants can also be pushed closer together, facilitating collisions and increasing apparent concentrations. This departure from ideal behaviour should be understood because it influences the type of equilibrium constants we use to describe leaching and model leaching. Ionic strength, (I), is determined by I = 1 ~ mizi2 2 where

I = ionic strength in units of moles/litre, mi = the concentration in moles/litre of the ith ion zi = the charge of the ith ion (dimensionless).

(13.1)

512 Ionic strengths for leachates can range from 0.05M (high L/S tests of bottom ash) to 20M (low L/S leaching tests of ESP ash). The presence of concentrations of ions in a leaching solution can impact the thermodynamics of reactions. Typically, leaching solutions have sufficient ionic strength that the solution must be classified as non-ideal. Under such conditions, the concentration of solute, [i], must be corrected to describe the solute's activity. At low ionic strengths, the activity of the solute may decrease if other dissolved counter ions in solution shield the ion from participating in chemical reactions. Conversely, at high ionic strength, the activity of the solute may increase if the solute is "salted out" or pushed into closer association with a potential reactant than what would occur under dilute conditions. The activity of species i is represented as {i} and is defined as {i} = Yi[i] where

(13.2)

y~ is the unitless single ion activity coefficient [i] is the concentration of solute i.

Usually, y values are less than 1.0, depicting that a solute's activity has an effective concentration that is less than the actual concentration in solution. Under ideal conditions in very dilute systems, y values are 1.0 and {i} = [i]. There are a number of ways to calculate y values for ions in solution (Pankow, 1991; Pytkowicz, 1983). As shown in Table 13.2, at least five methods exist. They are very much dependent on the ionic strength range of the leaching solution. The first four are based on the electrostatic approach developed by Debye and Heckel; they are appropriate for more dilute solutions. The fifth is based on the work of Pitzer and coworkers (ie. Pitzer, 1979, 1981) who substantially modified the Debye-H(3ckel approach; it is appropriate for more concentrated solutions. The Pitzer equation contains a first, second, and third virial expansion term beyond the Debye-H0ckel term. The expansion terms account for complex ion interactions that occur with increasing ionic strength. A term for neutral species interactions Ok) has been added by Harvie et al. (1984). Equations are also presented by Harvie et al. (1984) for Yx and YN; single ion activity coefficients for anion X and neutral species N, respectively. In the Pitzer equation, mc and zc are the molality and charge of cation C. No is the total number of cations in solution. Additional terms include:

513 Table 13.2 Methods for Determining Activity Coefficients Applicable Ionic Strength Range

Name

Equation

Debye-H(~ckel a

log y~ = -Azi2~'l

I< 10 -2.3

Extended Debye-H(~ckel a

log y~ = - Azi2[,/l/(1 + BaZl)] a is a size parameter for ion i; it should not be confused with the activity

I < 10 1~

Gfintleberg a

log Yi = -Azi2[~l/( 1 + ~1)] equivalent to Extended form with an average value of a = 3

Davies a

log y, = -Az,2[(r

I< 10 1~

1<0.5

+ ~1))- 0.21]

I < 20M

PitzeP Na

ma(2BMa + ZCMa )

InYM = z2F + ~

a=l

Nc

+ ~

c=1 Na-1

+ Ea=l

+ IzMI

mc(2OMc + Na

Na

~

a=l

maqJMca)

mama/LPaa/M

I2 a/=a+l Nc

Na

c=1

a=l

mcmaCca

Nn

+ ~

n=l

mn(2~'ni)

" Equations for activity coefficient Yi where the activity a~ = yjm~. For the ionic strength ranges applicable for the equations, ~. = y= = y= where ~. and y~ are the activity coefficients on the mole fraction and molarity concentration/activity scales, respectively. The parameter A depends on T(K) according to the equation A = 1.82 x 106(eT)3~ where e is the temperaturedependent dielectric constant of water. B = 50.3(eT) 1~2. For water at 298 K (25~ A = 0.51 and B = 0.33. b The form of the Pitzer equation is for the single ion activity coefficient for a cation M (Harvie et al., 1984). Reprinted in part with permission from Pankow, 1991. Copyright Lewis Publishers, an imprint of CRC Press, Boca Raton, Florida. 9

514

F = -A~ Nc

+~

c=1

+

1 + 1.2~ Na

/

~[~ mcmaBca

a=l

Nc-1

+ ~

c=1

+

2 In (1 +1.2~)} 1.2

Nc

~

c/=c+l

Na-1

Na

~ a=l

~'~ a/=a+

mcma,(bcc,

mama,(I)aa/

where the 1.2 value is a universal empirical value. CMX = C~X/21ZMZxl1/2, Z = ~

i

I z ilm i, and

A ~ = l/3(2nNodwi 1000)l/2(r

3/2

where No is Avagadro's Number, dw is the density of water, e is static dielectric constant of water at temperature T, k is the Boltzrnann's constant, and e is electronic charge. A ~ is 0.392 for water at 25~ C. The second virial coefficients B are defined as: BMX

n (~ + P'MXY~ n(1) ..,cxMxV~ m) , + = P'MX

BMx/ = 13MXLj ../tO(Mx (1)" / ~ VV', I~

13(M2)Xg(12~)

+ ~'MXVn(2) "/(12~)/I

where the functions g and g' are: 2 )e -X)/x2 g /(X) = -2(1 -(1 +X+X-~ 2 g(x) = 2(1 -(1 +x)e -X)/x2 where X = CXMxVI which simplifies to 12VI. Further, if M or X is univalent, CXMx= 2.0. For 2-2 or higher valence pairs, aMx = 1.4. Simplified versions of the Pitzer equations

515 can be found in Whitfield (1975a,b) and Millero & Schreiber (1982). Pytkowicz (1983) reviews many ion association, ion interaction and activity determination methodologies. The Davies Equation is probably the most appropriate for use with dilute leaching solutions. It is semi-empirical with an applicable range that covers the values of I most often seen in like bottom ash leachates ash leaching solutions. Table 13.3 provides some values of Yi for solutes of valency 1 through 5 for ionic strengths appropriate to ash leaching systems. Table 13.3 Values of y~for Solutes of Valency 1-5, Based on the Davies E_quations Ionic Strength (I)

Valency, z~ 1

2

0.001

0.965

0.867

0.002

0.952

0.005 0.010

3

4

5

0.726

0.566

0.411

0.820

0.641

0.453

0.290

0.927

0.739

0.506

0.298

0.151

0.902

0.662

0.396

0.192

0.076

u

0.025

0.860

0.546

0.256

0.089

0.023

0.030

0.850

0.522

0.232

0.074

0.017

0.050

0.822

0.455

0.170

0.043

0.007

0.100 0.782 0.373 0.109 Data compiles from Lindsay, 1979.

0.019

0.002

Under conditions of higher ionic strength, the calculation of single ion activity coefficients is more complicated (Pytkowicz, 1983). Leachates from ESP ash or scrubber residues have ionic strengths ranging from 0.5 to 20 M. These solutions are more akin to seawater and brines in that (i) ionic strengths are very high, (ii) the solutions are complex and contain many electrolytes, and (iii) complex interactions like ion pair formation (i.e. Ca 2§ + C I ,, CaCI § ) can have a significant impact on reducing both overall ionic strength and the activity of the individual paired ions. Two approaches have been developed to calculate activities in complex electrolyte solutions (Stumm and Morgan, 1981; Drever, 1988). The first involves the concept of ion association, particularly the formation of ion pairs. Usually, ion activities are experimentally determined in various simple electrolyte solutions. This is obviously difficult to compare to complex electrolyte solutions like seawater or brines with many dissolved species. Consequently, a second approach involving specific ion interactions was developed. The extended Debye-H0ckel approach was made applicable for higher

516 ionic strengths by adding additional interaction terms or virial coefficients to account for short range interactions between oppositely charged ions (Guggenheim, 1966). Guggenheim's modifications were further changed by Pitzer & Brewer (1961), Scatchard (1968) and Pitzer (1973), who all refined the second virial coefficient. This equation was found to be valid up to an ionic strength of about 4M. Whiffield (1975a,b) modified Pitzer's equations to calculate single ion activity coefficients. Values for seawater-like solutions (up to 3.0 M) were calculated. These values are shown in Table 13.4 and are useful approximations for APC residue leachates.

Table 13.4 Values for Yi for Various Solutes in Brines Ion/Ionic 0.2 0.4 0.6 0.7 0.8 1 2 3 Strength H§ 0.744 0.716 0.71 0.711 0.714 0.724 0.826 0.995 Na§ 0.724 0.677 0.652 0.643 0.637 0.626 0 . 6 1 4 0.634 K§ 0.709 0.653 0.62 0.607 0.597 0.581 0.537 0.528 NH4§ 0.707 0.649 0.615 0.603 0.592 0.575 0.531 0.517 Be2. 0.206 0.128 0.092 0.081 0.071 0.058 0.029 0.019 Mg2§ 0.299 0.248 0.228 0.222 0.219 0.214 0 . 2 3 8 0.306 Ca2§ 0.284 0.23 0.206 0.199 0.194 0.188 0.19 0.224 Sr2§ 0.284 0.229 0.205 0.197 0.191 0.183 0 . 1 7 8 0.201 Ba2§ 0.273 0.215 0.188 0.179 0.172 0.162 0 . 1 4 4 0.148 Zn2§ 0.288 0.229 0.201 0.192 0.184 0.173 0 . 1 4 2 0.125 Cu2§ 0.273 0.215 0.188 0.18 0.173 0.164 0 . 1 4 9 0.156 Pb2§ 0.193 0.119 0.086 0.075 0.066 0.054 0 . 0 2 6 0.017 Fe 2§ 0.281 0.227 0.204 0.197 0.192 0.186 0 . 1 8 9 0.225 Co2§ 0.289 0.234 0.211 0.205 0.2 0.194 0 . 2 0 4 0.248 Ni2§ 0.29 0.237 0.215 0.208 0.204 0.198 0.21 0.259 Mn2§ 0.3 0.243 0.218 0.21 0.205 0.197 0 . 1 9 7 0.225 U022§ 0.312 0.261 0.242 0.237 0.234 0.233 0 . 2 6 7 0.345 F~ 0.694 0.628 0.587 0.571 0.558 0.535 0 . 4 6 8 0.435 cr 0.744 0.71 0.695 0.691 0.689 0.687 0 . 7 1 2 0.778 Br 0.753 0.726 0.718 0.717 0.718 0.723 0 . 7 8 6 0.892 I 0.768 0.751 0.753 0.757 0.762 0.777 0.889 1.059 OH0.711 0.658 0.628 0.618 0.609 0.597 0 . 5 7 8 0.596 H2PO4 0.662 0.577 0.523 0.502 0.483 0.451 0.353 0.299 NO2 0.69 0.624 0.585 0.571 0.559 0.539 0 . 4 8 8 0.469 NO3 0.718 0.665 0.634 0.622 0.612 0.596 0.551 0.533 SO420.235 0.169 0.136 0.125 0.116 0.101 0.064 0.05 HPO42 0.245 0.167 0.128 0.115 0.104 0.087 0 . 0 4 6 0.031 CO320.237 0.162 0.127 0 . 1 1 2 5 0.105 0.091 0.057 0.043 PO43 0.05 0.024 0.015 0.012 0.01 0.008 0 . 0 0 3 0.001 Based on the Pitzer Equations (from Whitfield (1975a, b) with kind permission from Elsevier Sciences Ltd., The Boulevard, Langford Lane, Kidlington OX5 1GB, UK)

517 Subsequent modifications (Pitzer & Mayorga, 1973; Pitzer & Kim, 1974) ultimately resulted in the addition of more interaction terms or virial coefficients to account for more complex ion interactions between ions of like charge and ions and water that occur at even higher ionic strength (up to 20 M) and thus define the Pitzer equations. The Pitzer equations have been modified by Harvie & Weare (1980) and Harvie et al. (1984) for complex electrolyte solutions like seawater. Interaction terms (13~ 131,C~', k) between ions are found in Pitzer & Mayorga (1973), Pitzer and Kim (1974), Pitzer, (1979), Harvie & Weare (1980), Harvie et al. (1984) and Kim & Frederick (1988). The model has been validated for many solutes under very high ionic strengths. Pitzer equations have been added to some geochemical thermodynamic equilibrium models (i.e. SOLMINEQ 88, PHREQPITZ, SOLTEQ) for use in modelling brine or cement geochemistry. Activity coefficients for aqueous neutral species can be determined using the following relationship (Helgeson, 1969): IogYi = all

(13.3)

Typically al is set at 0.1 (Felmy et al. 1984). The activity coefficients like those shown in Tables 13.3 and 13.4 are useful in understanding departures from ideal behaviour, in evaluating results from modelling efforts using geochemical thermodynamic models and in selection of appropriate equilibrium constants (see section 13.1.4). 13.1.3 Cation/Anion Balances

It is important to introduce the concept of cation and anion balances in leaching solutions. The sum of all positively charged ions (cations) must equal the sum of all negatively charged ions (anions). Leaching solutions can be evaluated for this balance; it provides information on the adequacy of an analytical protocol in quantifying all dissolved constituents in solution. This concept is also employed in modelling efforts described in Chapter 15. In aqueous systems, bulk solutions do not carry any charge. They are electroneutral. A modified electoneutrality equation can be used to depict this fundamental requirement. ~Z.C. i

where

I I(positiv e ions)

= ~ Z.C. i

I I(negativ e ions)

z~ = the charge on ion i in positive or negative integer values (unitless), q = the concentration of ion i in moles/litre.

(13.4)

518 The equation can be modified to employ equivalent solutions instead of molar solutions. Equation 13.4 becomes very valuable in assessing the quantitative success in accounting for all dissolved constituents in a leaching solution. Belevi and Bacchini (1989) have calculated cation and anion balances for bottom ash leachates (Figure 13.4) after a 50-hour extraction. Generally close agreement between the sums of cations and anions was seen. A more detailed calculation of cation-anion balances is provided by St~mpfli et al. (1990) for a wet bottom ash from a Swiss incinerator. Balances were conducted periodically over the 415-hour time course for the leaching of the bottom ash into a closed system (Table 13.5). As shown in the table, the largest discrepancy between cations and anions (1.7 meq/litre) accounts for only about 7% of the total amount of cations and anions in the system (up to 25.7 meq/litre). Table 13.5 Cation-Anion Balances Over Time Extraction Time Parameter

30 min

5h

9h

24h 48h

415h

Ca 2§

[meq/I]

10.1

9.3

9.3

9.8

10

9.4

[Ca(OH)] +

[meq/I]

1.2

0.7

0.7

0.5

0.5

0.5

Na §

[meq/I]

2.1

2.3

2.3

2.5

2.5

2.8

K§

[meq/I]

0.5

0.6

0.8

0.7

0.7

0.7

Cations

[meq/I]

13.9

13

13.1

14

14

13.4

[AI(OH)4]

[meq/I]

0.5

2.9

3.6

5.2

5

4

OH

[meq/I]

9.6

6.3

5.6

4.3

4

4.4

Cl

[meq/I]

2.2

2.5

2.6

2.9

2.8

3.1

[SO4]2-

[meq/I]

0.3

0.2

0.2

0.2

0.2

0.2

Anions

[meq/I]

12.6

12

12

13

12

11.7

1.3

1.1

1.1

0.9

1.7

1.7

Cations-Anions [meq/I] After St,~mpfli et al., 1990

Thus, the use of such balancing techniques can tell if the analytical scheme is accounting for all of the dissolved constituents of interest. It can also provide information as to the dominancy of the major cations and anions in solution. This can be coupled with mass balances on solids when a residue is leached, i.e. the total dissolved solids that are produced should equal the mass change in the residue when leached in a closed system.

Figure 13.4 Cation-Anion Balance

CONCENTRATION (meq/L)

14 1

After Belevi and Bacchini, 1989

I

520

13.1.4 A Note on General Equilibrium Constants It is necessary to understand the various methods for determining equilibrium constants as well as the nature of equilibrium constants in order to properly evaluate leaching behaviour, interpret data from field leaching scenarios or leaching tests and model leaching phenomena. The equilibrium constants in the literature or in thermodynamic models are routinely modified or updated during evaluation or modelling. Consider the following hypothetical dissociation reaction conducted at 25~ at 1 atmosphere under infinitely dilute conditions: {HA} ,= {H *} + {A-}

or

[HA]YHA ,- [H +]YH + [A-]YA

(298.15~ (13.5) (13.6)

Under ideal infinitely dilute conditions, activity coefficients approach unity and activity and concentration become equivalent. The activity equilibrium constant of 25~ under dilute conditions, designated as K, is defined as:

K = {H +} {A-} or [H +]YH" [A-]YA{HA} [HA]YHA

(13.7)

Sometimes the equilibrium constant K can be determined experimentally by measuring reactants and products at equilibrium (Rosotti, 1981) under infinitely dilute conditions at 25~ More often, K is determined from the Gibbs free energy of formation (AGf~ the standard free energy change of a reaction (AGr ~ the standard enthalpy of a reaction (AHr~ and the standard entropy of a reaction (nSr~ The standard free energy change accompanying a reaction (AGr ~ is the sum of free energies of formation (AGf ~ of the products in their standard state minus the free energies of formation of the reactants in their standard state: AGr ~ = ~ AGf ~ products - ~ AGf ~ reactants

(13.8)

AGr ~ can be related to K by the following: AGr ~ = RTInK

(13.9)

At 25~ R = 0.08314 K J/mole K and T = 298.15 ~ such that AGr ~ = -1.364 log K or

(13.10)

521 log K = - AGr~ 1.364

(13.11)

Therefore, for any reaction, published values for AGf ~ for the reactants and products can be used to calculate AGr ~ and therefore K. Occasionally, AGr ~ can be calculated from changes in the standard enthalpies of reaction (AHr ~ and changes in entropies of reaction (ASr ~ using the following: AGr ~ = AHr ~ - TASr ~ , where

(13.12)

AHr ~ = ~ AHf ~ products - ~ A H f ~ reactants and

(13.13)

~,S r~ = ~

o

o

S products - E S reactants

(13.14)

Calorimetric measurements of the heat of solution are used to calculate the heat of formation (AHf~ Calorimetric measurements of heat capacity are used to calculate entropy (AS ~ If these are known for the reactants and products, then AGr ~ and K can be calculated. Thermodynamic data bases containing various AGf ~ AHf ~ and AS o values for reaction constituents are found in the National Institute of Standards and Technology (formerly the National Bureau of Standards) Technical Note series 270 (Parker et al., 1971; Schumm et al., 1973; Wagman et al., 1968; 1969; 1971; 1981) and in Wagman et al. (1982). They are also found in a U.S. Geological Survey compilation (Robie et al., 1978, 1979;) and other compilations (Baes and Mesmer, 1976; Smith and Martell, 1976; Naumov et al., 1974; Lindsay, 1979). These databases, particularly Wagman et al. (1982), are deemed to be the most complete, annotated and accurate (Allison et al., 1990; Felmy et al., 1984; Krupka et al., 1988). They also form the basis for the geochemical source codes in the models presented in Chapter 15. Usually, three types of equilibrium constants are used in aqueous geochemistry applications. They are: an activity constant at infinite dilution, used most frequently because of its basis on AGf ~ K~

a concentration constant, where all reactants are expressed in terms of analytical concentration at some specified ionic strength I, and

K r" (or K')

a mixed constant, where all terms are given in concentration except H§ OH and e, which are given as activities.

522 Using equations (13.5) through (13.7) as examples, the three equilibrium values can be interrelated' K = {H*} { A - } _ [H*]YH+[A-]YA- KCYH.yAK myA= {HA} [HA]YHA YHA YHA

(13.15)

As indicated above, the thermodynamic databases mentioned above usually present K. Frequently, the literature contains Kc or Km values; the geochemical models also convert K to K~ when H§ OH or e are directly inputted as fixed species in the model. Equilibrium constants are temperature dependent. Most are reported at 25~ correct them for other temperatures, the Van't Hoff relationship is used: AHr ~

log K T = log KT, - 2 . 3 R

where

13.2

1

1__)

(-T - T r

To

(13.16)

KT = the constant at a specified temperature, KTr- the reference constant at 298.15 K (25~ AHr ~ = the enthalpy of reaction, T = the specified temperature (K), Tr = the reference temperature (K), R = the ideal gas constant.

SOLUTION

COMPLEXATION

AND SPECIATION

It is important to understand the role that complexation plays in controlling the speciation of dissolved elements in solution. The solution phase speciation of an element can impact its mobility in the environment, its bioavailability, and its ability to participate in precipitation or sorption reactions. Soluble complexation is also an important feature of modelling efforts presented in Chapter 15. In almost all leaching solutions, the solutes of interest (e.g. trace metal species such as Pb2.) are present as soluble complexes with various ligands (e.g. CI, OH, organic matter) such that the concentration of the uncomplexed or "free" solute (e.g. Pb 2.) is small compared to various complexes (e.g. PbCI, PbCI2~ Pb(OH), Pb(OH)2 ~ Pb(OH)3* ). This has ramifications in the interpretation of the leaching system and the mechanisms that occur in the system. In pure solution, most dissolved metal atoms are usually surrounded by six waters of hydration or "aquo" complexes. They form a defined geometrical structure around the atom (Drever, 1988). Complexation can be defined as the displacement of these coordinated water molecules by other ligands. Two types of soluble complexes can occur for elements in a solution where ligands exist (Evans, 1989). Outer sphere

523 complexes, or "ion pairs", can form. These are relatively weak electrostatic interactions between a cationic element (e.g. Pb2.) and an anionic complexing ligand (e.g. Br) to form a soluble complex (PbBr). With weak, outer shell complexes, one or both of the participants retains a hydration structure that solubilises the reactant. Inner sphere complexes are strong associations between an element and a ligand where a covalent bond is formed between the element of interest (e.g. Pb2.) and the ligand (OH) to form a soluble, but more tightly bound complex (Pb(OH)). These complexing ligands can be monodentate (forming one association) or multidentate (forming many associations, termed a chelate). Most ligands are conjugate bases of weak acids. The extent of complexation depends on the relative amounts of the elements and ligands, as well as the pH of the system. In general, complexes between monodentate ligands and metal ions from the first column of the periodic chart are weak. Figure 13.5 depicts some of the various complexes that can form (Stumm and Morgan, 1981). Figure 13.5 Metal Speciation in Solution =

I

filterable

:

membrane filterable

,,, 9

I

dialysable

,,,

=

It

in true solution

Free Metal Ions

Metal species Metal species in Metal species bound to high molecular the form of highly sorbed on wt. org. material dispersed colloids colloids

Organic Inorganic ion pairs; inorganic complexes, chelates complexes

Diameter range:

loo ~,

10 ,~,,

precipitates

organic particles

remains of living organisms

,1000 ,~

Examples:

Cu.2§ aq.

Cu2 (OHI~

Me. SR

Fe z+ aq.

Pb(CO 3 )0

Me-OOCR

Pb2§ aq.

CuCO3 AgSH

Me I~umic acid polymers

/CRy---- C~O INNz

"lakes" "Gelbstoffe" Me polysaccharides

O

CoOH § Zn(O,); Ag 2 S3 H~

I~

FeOOH

Me lipids

Me, (OH)y MeCO3, MeS Fe(OH) 3 etc. on clays. Mn(IV) oxides Mn7013- 5H 20 FeOOH or Na4 Mn14 O2T Mn(IV) on Ag2S oxides

2

From Stumm and Morgan, 1981. 9 permission of John Wiley & Sons, Inc.

by John Wiley & Sons, Inc. Reprinted by

When a leaching solution is quantified using ion chromatography, flame or graphite furnace atomic absorption spectrometry, inductively coupled plasma spectrometry or X-ray fluorescence, it tells us the analytical concentration of an element in all of its complexed forms and uncomplexed form in the leaching solution, CT,i. It does not

524 differentiate the uncomplexed from the complexed element. The actual concentration of an element that is in an uncomplexed form is usually much less. Yet it is this uncomplexed fraction of the CT.~that participates in precipitation/dissolution reactions and in many adsorption/desorption reactions in the leaching system. We can calculate the concentration of a solute i, [i], by:

[i]- CT'i a

(13.17)

where CT, i is the analytical concentration of the solute i and a is a term derived from the mass balance on an element or solute and various equilibrium relations describing how i is complexed. We can then calculate the activity of i, {i} by using equation (13.2).

13.2.1 Solution Complexation Equilibria The nomenclature used in depicting solution phase complexation reactions follows the description by Pankow (1991). Consider the formation reaction where metal cation M and ligand anion L combine to form the complex ML: M+L =, ML

(13.18)

The stability or formation constant, K, is described by: K=

{ME} {M} {L}

(13.19)

The reaction shown in Equation 13.18 can be written as a dissociation reaction where ML dissociates to M and L: ML ,- M + L

(13.20)

The instability or dissociation constant for this reaction is the inverse of the stability constant: I/K = {M} {L} {ML}

(13.21)

The strength of the complex that forms is a function of the value of the stability constant. High values (>108 ) denote strong complexes. Low values (<102 ) denote weak complexes. Table 13.6 depicts the relative strengths of stability constants and dissociation constants.

525 Table 13.6 Formation an d Dissociation Constant Strengths Type of Constant

Strength of Complex

,.

Weak < 10 2

Metal/Ligand formation (stability) constant

Moderately Strong --

10 s

Strong > 10 8

> 10 .2 ~ 10 -s < 10 8 Metal/Ligand dissociation (instability) constant Reprinted with Permission from Pankow, 1991. Copyright Lewis Publishers, an imprint of CRC Press. Boca Raton, Florida. 9 There are other types of stability constants for formation reactions that are summarised in Table 13.7 (Pankow, 1991). Knowledge as to the type of stability constant used in the evaluation of leaching data or in modelling leaching is important if it is to be done correctly.

13.2.2 An Example of Lead Complexation in a Hypothetical Leaching Solution As an example, a simple solution containing only Pb 2., H20, H*, OH, CI, Br and SO4 is assumed. Ionic strength effects are neglected for simplification. The concentration of ligands is set at 103M and {H*} is set at lx10 1~ (pH = 10). Using the nomenclature in Table 13.7, consider the following complexation reactions that can occur for lead in this very simplified system. Pb 2+ + H20 ~ PbOH + + H + (log*K = 7.70)

(13.22)

Pb 2+ + C I - ~

(13.23)

PbCI + o

Pb 2+

+

2CI - ~ PbCI 2

pb2+

+

Br- ~ PbBr +

pb 2~ + S O 2 - ~ PbSO4~

(log K 1 = 1.60) (log K2 = 1.78) (log K3 = 1.77) (log K, = 2.62)

(13.24) (13.25) (13.26)

526 Table 13.7 Nomenclature for Complexation Constants When a free, unprotonated ligand combines with a metal ion, a simple K (stability constant) is used with subscript (1,2,3, etc.) to identify the number of ligands: M+L ,, ML

K1 _

ML + L ,* ML 2

K2 =

ML 2 + L =* ML 3

K3 =

{ML} {M}{L} {ML 2} {ML}{L} {ML 3} {ML2}{L}

When the reactions presented above are added together, a 13notation is used to depict additive values of stability constants: {ML}

M+L ,=, ML

I~1 = K 1 -

M+2L =* ML 2

132 = K1K2 =

M+3L ,- ML 3

{M}{L} {ML 2} {M}{L} 2

133 : K, K2K3 :

{ML 3} {M}{L} 3

When the reacting ligand is protonated, an asterisk (*) is used to modify the stability constant, denoting the presence of protons: M + HL,=' M L + H *

ML + HL = ML 2 + H +

,K1 = {ML}{H*} {M}{HL}

{ML2}{H *} {ML}{HL}

527 Table 13.7 Continued {ML3}{H *}

ML 2 + HL ," ML 3 +H*

{ML2}{HL}

If the reactions presented above are additive, then: M + H L , " ML + H*

M + 2HL,,' ML 2 + 2H*

M + 3HL ," ML 3 + 3H *

*131 = *K1 = {ML}{H*} {M}{HL}

*132 =

"133 :

*K 1 K 2 =

*K, K2K 3 :

{ML2}{ H +}2 {M}{HL} 2 {ML3}{ H ,}3 {M}{HL} 3

For the reaction of a metal ion with OH, since the hydroxide is not protonated, a simple stability constant with the subscript H denoting hydroxide is u s e d

M

+

O H - ,,, MOH

KH1 =

{MOH} {M}{OH -}

When the reacting ligand is water and since H20 is the HL version of the O H ligand, the following notation is used: M + H 2 0 , - MOH + H ~

,KHI = {MOH}{H*} {M}

Reprinted with permission from Pankow, 1991. Copyright Lewis Publishers, an imprint of CRC Press, Boca Raton, Florida. 9

528 The equations depict a hydrolysis reaction [equation (13.22)], a monodentate and bidentate complexation reaction [equations (13.23) and (13.24)] and two monodentate complexations of various bond strengths [equation (13.25) and (13.26)]. An important consideration in defining solution speciation and the distribution of lead in the solution is as follows: what percentage of the total analytical lead in solution, CT,Pb, is the uncomplexed free ion (Pb 2*) and what is complexed as PbOH*, PbCI*, PbCI2 ~ PbBr § and PbSO4~ First, the mass balance equation for

CT,Pbmust be

identified:

CT,Pb = Pb 2* + PbOH § + PbCI § + PbCI2 + PbBr § + PbSO4~

(13.27)

By combining equations (13.22) through (13.27), equation (13.27) can be solved in terms of the various K values and {Pb2*}:

CT,Pb = {Pb 2+} + *K{Pb 2+}/{H +} + KI{pb2*}{CI -} + K2{pb2+}{CI-} 2 +

(13.28)

K3{Pb 2+}{Br-} + K4{Pb 2~}{SO2-} Dividing through by {Pb2*}, the equation reduces to:

CT,Pb

pb2§

- 1 + "K/{H *}

+ KI{Cl-}

+

K2{CI-} 2 + K3{Br- } + K4{SO2

(13.29)

Setting {cr}, {SO42-} and {Br} = 103M and {H § = 10 1~ (e.g. pH = 10), the equation reduces to:

CT, Pb - 1 + 10 -17"7 + 10 -14 + 10 -4"22 + 10 -1"23 + 10 -~ {Pb 2+}

(13.30)

or {Pb 2+} = 0.5683 CT,Pb

(13.31)

(or 56.8% of C T,Pb is Pb 2+). TO determine the relative activities of the other species, equation (13.28) and equations (13.22) through (13.26) can be used such that:

Table 13.8 Bottom Ash Outdoor Lysimeter Leachate Aqueous Element Speciation Percent Cornplexed AsC

Element or Conc. Parameter (mglL) Or

Value

Me

MeOH MeOH, MeOH, MeOH, MeSO, Me(SO,),

A13'

0.020

Sb5'

0.063 99.9

B a2+

0.10

99.5

Ca2*

590

73.8

Cu2+

0.59 55.4

2.8

0.050

97.4

Fe3'

2.4

22.6

55.7

MeHCO, MeCO, MeCl MeCI, MeCI, MeBrCl MeBr, MeBr,

18.7

26.0 19.6

5.3

5.6

2.7

4.4

9.5

20.8

2.2

0.0053 29.3

28.5

2.0

6.5

76.3

23.5

0.32

71.9

22.6

96.7

3.3

5.2 12.6

0.0006 220 2.2 740

97.6

4.4

100

Zn2+

0.068 65.4

Br C I-

30 1700 1700 300 29 6.41 0.08

SOP Organic C Total Alk.b

PH

US

---

4.9

61.3

18 8

1.1

(Speciated as H,SiO,)-

70C:for modelling purposes this was assumed to be valeric acid. shown

MeValerate

2.3 26.5

' m g l as CaC03.

2.8

1.6

3.1

'Metal or metaloid denoted as Me in either free ion form or as a complex. Note that the charge on the complex is no!

530 {PbOH +} = 10 -177 0 . 5 6 8 3 C T, eb

=

1 13x10-18C 9

T,Pb

(13.32)

{PbCI ~} = 10 -14 0 . 5 6 8 3 C T p, b = 2 20x10-2C 9 T,Pb

(13.33)

{PbCI2} = 10-4.22 0.5683Cm,Pb = 3.42X10-5CT,Pb

(13.34)

{PbBr +}

(13.35)

=

10-1.23

0.56830T,Pb

=

5.88X10-2CT,Pb

{PbSO4~ = 10-o.18 0.5683Cm,pb = 0.376Cm,Pb

(13.36)

It should be readily apparent that the complexed species with the largest K value (PbSO4 ~ plays the most significant role in complexing Pb in accordance with the hierarchy shown in Table 13.6. Given the fact that leaching solutions contain many solutes and that each metal or metalloid is capable of being complexed by 3 to 4 ligands, the tediousness of such calculations and the utility of geochemical models in computer codes is apparent. 13.2.3

Leaching Solution Speciation

Geochemical models such as MINTEQA2 can be used to examine solution phase speciation. The approach has been applied to two leachates: a bottom ash lysimeter leachate at a low L/S value (0.08) and a dry scrubber residue leachate also at a low L/S value (0.2). Table 13.8 shows the speciation of elements in the bottom ash leachate. For modelling purposes, the organic carbon was assumed to be valeric acid. As can be seen in the table, a number of complexes are present. Metal sulphate, metal bicarbonate and metal chloride complexes are the most predominant. Valerate will complex copper and lead. In some cases, the free ion is not the principal form of the element in solution. The ionic strength (I) of the leachate is 0.09 M. Table 13.9 shows the speciation of elements in the dry scrubber residue leachate. For modelling purposes, the organic carbon was assumed to be valeric acid. Unlike the bottom ash leachate, the scrubber residue is more alkaline and of much higher ionic strength (28 M). This extremely high value invalidates the use of Davies-based activity coefficients. This leachate is probably one of the strongest that can be produced under column settings. While the use of MINTEQA2 to model this leachate is inappropriate, the results are presented to show how metal chloride complexes are a predominant form of the aqueous species.

531 Models can also be used to do iterative calculations over a wide range of pH values. Figure 13.6 shows the speciation of lead over a large pH range for a complex system where O2(aq)is present. Figure 13.6 Lead Speciation as a Function of pH -5

.

.

.

.

.

.

.

.

.

.

.

-s .7

~

-8

--

Pbco~ "'

L

PbSO,O

|

PbCI"

~_

~

~:~ //)

,,.//

"'F/ 2

,b(OH)~ 7

,,o~,

t

PbF§ PbNO; / /

..,o

-13

!

....

/ I 3

4

# 5

' 6

7 pH

8

9

10

11

12

From Rai and Zachara, 1984 Copyright 9 Electric Power Research Institute. EPRI (EA-3356)"Chemical Attenuation Rates, Coefficients, and Constants in Leachate Migration". Reprinted with permission 13.3 DISSOLUTION AND PRECIPITATION REACTIONS

Frequently, the mineralogical characteristics of the ash solid phase play the dominant role in controlling all leaching processes. Knowledge as to how the solid phases dissolve or reprecipitate is crucial in evaluating leaching data (Chapter 16), designing or using leaching tests (Chapter 14), relating solid phase speciation data (Chapter 7) to leaching behaviour, and in correctly modelling leaching behaviour (Chapter 15). The interactions that occur in the ash leaching system (see Figure 12.2, Chapter 12) become more complex when a solid phase is included in the system. The system becomes a heterogeneous one, consisting of a solid phase and a solution phase. Under different conditions, some mineral phases dissolve and secondary mineral phases can form (Warren and Dudas, 1987). Dissolution and precipitation reactions are usually slower than the aqueous phase reactions discussed in Section 13.2. Nevertheless, the more leachable fractions in the ash residues can dissolve relatively quickly and it is relatively safe to adopt the local equilibrium assumption (LEA) as long as the system has approached equilibrium (e.g. on the order of days for many leaching

532 tests, weeks to months for field leaching scenarios). As discussed under the section in weathering and aging (Section 13.4), some of the solid phases in ash that dissolve very slowly (e.g. aluminosilicates) can only be examined using a kinetic basis and are not able to be evaluated under LEA. Precipitates that occur in residues can often be amorphous and metastable. Frequently, degrees of oversaturation or undersaturation can exist. Many of the precipitates are co-precipitates that become amorphous semi-solid gels. Pure solid phases do not always form. Isomorphic substitutions, or displacement of atoms in a crystal lattice with atoms of similar size, and the formation of solid solutions (noncrystalline solids) may decrease the activity of the solid phase. The presence of complexing ligands in solution can also increase solubility. Very fine particle sizes are also more soluble than larger crystalline structures. Nonreactive solutes in the solution phase can alter solubilities by their influence in controlling ionic strength. It is important to introduce some basic concepts describing equilibrium precipitation and dissolution reactions, the equilibrium constants that are used and the nomenclature that is employed. Section 13.3.1 presents that information.

13.3.1 Heterogeneous Dissolution/Precipitation Equilibria The approach of Pankow (1991) is used to differentiate nomenclature on dissolution/precipitation equilibria. The equilibrium constants for dissolution/ precipitation reactions that produce dissolved ions are usually referred to as Ksp values. The subscript "sp" is an abbreviation for "solubility product". The word product refers to the product of ion activities that appears in the expression for the solubility equilibrium constant. Another type of symbol used for solubility products is the K,o. In this symbol, the "s" in the subscript refers again to solubility. The zero, on the other hand, denotes the fact that the constant refers to the specific solubility product for which the identities of the ions in solution are exactly the same as in the solid (Pankow, 1991 ). Consider an equilibrium expression describing the precipitation/dissolution of a soluble metal salt (Pankow, 1991): MmSs(S) ~ mM *S(aq) + sS-m(aq)

(13.37)

At equilibrium, the stoichiometric relationship shown in equation (13.37) becomes: I~o = {M*m(aq)}s{s-s(aq)} m {MmSs(S)}

(13.38)

By convention, the activity of the solid phase is set equal to unity, and the equilibrium expression is simplified to:

Table 13.9 Dry Scrubber Column Leachate Aqueous Element Speciation Element or Parameter

Concentration (mg/L) or Value

AS

0.21

Cr

2.3

Organic Ca

45

Total Alkalinityb

540

us

0.2

Percent Complexed AsC Me

MeOH MeSO,

MeHCO,

MeCO,

MeCI,

MeCI,

Me-Valerate

(spec~atedas Hfis0, [100%]

(speciated as NaCrO, [31.4%], KCrO,' [68.6%]

bmeqV/L. "NVOC; for modelling purposes this was assumed to be valeric acid, 'Metal or metalloid denoted as Me in either free ion form or as a complex. Note that the charge on the complex is not shown.

534 Kso = {M +m(aq)}s{s-S(aq)}m

(13.39)

After conversion to a concentration based system, it becomes: CKso = (M +m(aq))s(s-S(aq))m/(ym)m(ys)S

(13.40)

where CK,o is the equilibrium constant specific for the reaction and the system it describes. Care must be taken as to which form of the constant is used from the literature. Use of incorrect constants for mechanistic interpretation can lead to gross errors in evaluation of a leaching system. A large number of sources are available that contain published solubility products. These include Sillen and Martell (1971), Feitknecht and Schindler (1963), Smith and Martell (1976), Robie et al. (1978), Baes and Mesmer (1976), Stumm and Morgan (1981), and Lindsay (1979). The selection of the actual value of a constant should be completed with care in calculations (Stumm and Morgan, 1981). Large differences in values are sometimes observed because (i) a system may be more complex than the simple systems used in determining the various solubility products, (ii) the activity of the solid phases are not always unity and (iii) aqueous phase complexation reactions are overlooked in the determination of published equilibrium constants (Stumm and Morgan, 1981 ).

13.3.20versaturationlUndersaturation and the Ion Activity Product (lAP) The identification of the relative degree of potential saturation of solutes in a leaching solution is an important tool in evaluating leaching data and in modelling leaching behaviour. This aspect becomes an important component in the modelling efforts described in Chapter 15. An introduction to the principles is therefore necessary if the tool is to be correctly used. Returning to equation (13.39), this equilibrium relationship can be used to help determine whether or not a leaching solution is undersaturated (tending towards dissolution) or oversaturated (tending towards precipitation) with respect to a solid phase. Table 13.10 (Pankow, 1991) depicts three conditions applicable to equation (13.39). By convention, the mathematical product of the metal ion activity, {M+r"(aq)}s, and the salt activity, {S-S(aq)}r", is termed the ion activity product (lAP). The lAP in a slightly modified form is also used in the geochemical models presented in Chapter 15. It is related to the saturation index according to the following (Allison et al., 1990): Saturation Index = log

lAP

Kso

(13.41)

535 Table 13.10 Conditions of Relative Saturation Condition State

Implications

{M§

{SS(aq)}m< I~o Undersaturated If solid phase is present, some will tend to dissolve; if solid phase is not present, it will not form. If solid phase is present, no more will {M+m(aq)}" {S-'(aq)} = K,o Saturated dissolve; if solid phase is not present, it will not form. {M+m(aq)}' {S"(aq)} > K~o Supersaturated If solid phase is present, more will tend to form; if solid phase is not present, some will tend to form. Reprinted with permission from Pankow, 1991. Copyright Lewis Publishers, an imprint of CRC Press, Boca Raton, Florida. 9 As noted by Pankow (1991), if the saturation index is negative, the mineral is undersaturated and will not precipitate out of solution. If the saturation index is positive, the mineral is supersaturated and can precipitate out of solution. If the index is zero, the mineral is present as a solid. The magnitude of the index describes the driving force towards remaining dissolved (a large negative number) or precipitation (a large positive number). The geochemical thermodynamic computer models presented in Chapter 15 utilise this principal in deciding which potential mineral phases will or will not form in a leaching solution. The largest positive indice will precipitate first in response to the driving force. This principle has been put to use in interpretation of ash leachate. St~mpfli et al. (1990) used this concept to evaluate relative degrees of saturation for bottom ash leachates. Their leachates tended towards undersaturation for certain calciumcontaining solid phases. Fruchter et al. (1990) use the same technique to evaluate the precise nature of the solid phase controlling {Ca 2§ and {SO42} in leachates from coal fly ash. Figure 13.7 depicts a useful graphical tool where the log of the activity of Ca =§ is plotted versus the log of the activity of SO42. The equilibrium relationship for two solid phases of interest, CaSO4 (anhydride) and CaSO4~ (gypsum), are shown as diagonal plots whose y intercepts are the log of the equilibrium K,o for each of the solid phases (or the lAP at equilibrium). Leachate analytical data (e.g. CT, C,,, CT, sO, ) were converted to activities (e.g. {Ca2§ {SO42-}) using a geochemical thermodynamic equilibrium model (MINTEQA2). The values of {Ca 2§ were plotted versus the corresponding sample analytical value of {SO2}. The data fall along the CaSO4o2H20 boundary, suggesting that some slight under- or oversaturation existed, but that the condition {Ca 2§ {SO42} = K,o(gypsum) was generally met. This becomes an important but simple analytical tool if one desires to identify the nature of the solid phase controlling leaching in the

536 leaching system. For example, Roy and Griffin (1984) have made extensive use of this approach to understand the nature of the solid phases controlling leaching in various coal fly ash systems as the system ages and weathers over time. From an MSW residue perspective, knowledge of the solid phase controlling solute concentrations in the leachate solution allows for modification or control of the system to either enhance or dissuade leaching. Figure 13.7 Plot of {Ca 2.} and {SO42-} for Coal Ash Leachates

o Site AB o Site CD Composite A Construction Layers CaSO4

O Leachate o Extracted Pore Fluid

§ m

o

o

o

v

v-

-3--

-4

-3.0

CaSO4-2H20

.

.

.

.

.

1 .....

-2.5

! . . . . . . . -2.0

-1.5

Reprinted with permission from Fruchter et al., 1990. Copyright 1990 American Chemical Society 13.3.3

Metastability

Frequently, solids that precipitate out of solution are still not thermodynamically stable. As the precipitate ages, the solid phase can change its mineralogical structure to become more stable. Such types of transformation, a component of aging reactions, are important in interpreting and modelling field leaching behaviour where solid phases play a strong role in controlling leaching. The concept of metastability of the solid phase, as depicted in Figure 13.8, is introduced by Stumm and Morgan (1981).

537 Figure 13.8 A Depiction of Metastability ity Limit of "active" Form ~hase i e Saturation =tion r of Stable lse ) I

Z 0 l-r~ F-Z LU

Z

0

UNDERSATURATED

pH From Stumm and Morgan, 1981. @1981 by John Wiley & Sons. permission of John Wiley & Sons.

Reprinted by

An active form of the solid phase, a very fine precipitate with an unorganised crystalline lattice, is generally formed from oversaturated solutions (Stumm and Morgan, 1981). This active precipitate can exist in metastable equilibrium with the solutes in the solution and may convert or age or ripen only slowly into a more stable inactive form (Stumm and Morgan, 1981). Measurements of the solubility of active forms of the solid phase in question would yield solubility products that are higher than those of the inactive forms. Inactive solid phases whose crystal structure is ordered can also be formed from solutions that are only slightly oversaturated (Stumm and Morgan, 1981 ). Hydroxides and sulphides often occur in amorphous and several crystalline forms. Amorphous solids may be active or inactive (Stumm and Morgan, 1981 ). As noted by Feitknecht and Schindler (1963), initially formed amorphous precipitates or active forms of unstable crystalline forms may transform via two processes during aging. The active form of the unstable form can become inactive or a more stable form can evolve. Deactivation of amorphous compounds may be accompanied by condensation reactions (Stumm and Morgan, 1981). If a metal oxide is more stable than the primarily precipitated hydroxide, dehydration may occur (Stumm and Morgan, 1981). When these reactions occur simultaneously, heterogeneous solids are formed upon aging (Stumm and Morgan, 1981). In dissolution experiments with such nonhomogeneous solids, the more active components are dissolved more readily; that is, the solubility results may depend more upon the quantity of a solid phase present (Stumm and Morgan, 1981).

538 Precipitates are frequently formed from strongly oversaturated solutions; the conditions of precipitation of the active compound rather than the dissolution of the aged inactive solid are often observed (Stumm and Morgan, 1981). Most solubility products measured for these solids refer to the most active component (Stumm and Morgan, 1981). Heterogeneous equilibria for natural leaching systems contain stable and inactive solids as dominant solid phases. Aging often continues for many years. Solid phases frequently form in nature under conditions of slight supersaturation (Stumm and Morgan, 1981). Thus, use of the correct equilibrium constants should be coupled with speciation methods (Chapter 7) to verify the nature of the solid phases in question.

13.3.4 An Example of Lead Dissolution/Precipitation as PbSO4(s): A Very Simple Leaching System To illustrate the use of precipitation/dissolution thermodynamics in interpreting and modelling leaching behaviour, it is necessary to start with simple systems to introduce approaches and concepts. This first example makes use of a very simple solid phase solution system. Ionic strength effects will be neglected so that activity, { }, is equal to concentration, (). Subsequent examples will add to the complexity of this first example so as to ultimately illustrate the roles of multiple solid phases in equilibrium with each other and with complexing ligands in solution. PbSO4(anglesite) is frequently found in ash residues. It accounts for some of the available or leachable fraction of lead in combustion residues. Its dissolution is, in part, responsible for the appearance of total dissolved lead (CT,pb) in leaching test or field leachates. PbSO4(s) exhibits a pH-dependent solubility because the conjugate base, SO42(sulphate), can exhibit pH-dependent protonation reactions that add protons to the base: HSO4(bisulphate) or H2SO4~ acid). Such reactions influence the availability of SO42- to precipitate out of solution with Pb2§ to form anglesite. Consider the following equilibrium reaction: I0~ K {PbSO~ ~ {Pb 2.} + {SO 2-}

-7.79

(13.42)

Equation (13.42) can be rearranged and put in log form such that: log {Pb 2~} = -7.79 - l o g {SO 2-}

(13.43)

It is then necessary to calculate the activity of SO42 as a function of pH for substitution into Equation 13.43. H2SO4 is a diprotic acid; it exists as the doubly protonated H2SO4 (sulphuric acid) at very low pH values (pH <-3), the monoprotonated HSO4 (bisulphate)

539 at low pH values (-3 > pH < 2) and as the unprotonated SO42 (sulphate) at moderately low to high pH values (pH > 2). By convention for a diprotic acid (Pankow, 1991) the activity of the SO42 when CT,so, is fixed at 1 x 103M, is calculated as:

{H +}2 + {H +}Ka,1 + Ka,lKa,2

)CTso4

(13.44)

where K~,1 is the first acid dissociation constant (1 x 10 3) and K~,2 is the second acid dissociation constant (1 x 102). By selecting various pH values, the SO42 concentration as a function of pH can be determined by substitution into equation (13.44): pH

{H +}

{SO42-}

14

1 x 10 "14

1 x 10 .3

12

1 x 10 12

1 x 10 .3

10

1 x 10 "1~

1 x 10 .3

8

1 x 10 .8

1 x 10 .3

6

1 x 10 .6

9.99 x 10 .4

4

1 x 10 .4

9.90 x 10 .4

3

1 x 10 .3

9.09 x 10 .4

2

1 x 10 .2

4.99 x 10 .4

1

1 x 10 1

9.09 x 10 s

0

1 x 10 ~

9.89 x 10 .6

If we substitute the log of the {SO42-} into equation (13.43) and solve for {Pb 2+} we get: pH

P{Pb 2+}

14

4.79

12

4.79

10

4.79

8

4.79

6

4.785

4

4.78

3

4.75

2

4.44

1

3.75

0

2.78

540 The values are plotted as a line in Figure 13.9 in a pC-pH plot. For this simple system, any combination of {Pb2*} and pH that is situated above and to the right of the solubility plot (line 1) means that the lead will precipitate out with sulphate to form anglesite. Any combination of {Pb 2*} and pH that is situated below and to the left of the solubility plot means that lead will stay in solution. Figure 13.9 p{Pb 2+} - pH Plot Showing PbSO4(s) in Equilibrium with {Pb 2*}

L

PbSO4(s)

p{Pb2+} 10 12 I

I

I

2

4

6

I

I

I

8

10

12

pH Concentration of SO4 is 103M 13.3.5 An Example of Lead Dissolution/Precipitation as Pb(OH)2(s): The Role of Solution Phase Complexation and Amphoterism

The second simple example expands on the one presented in Section 13.3.4. Again, ionic strength effects will be neglected and activity is equal to concentration. A solid phase that exhibits minimum solubility at neutral pH values and maximum solubility at low and high pH (an amphoteric solid) is in equilibrium with a solution which has ligands consisting of hydroxyl anions (OH). This example accurately portrays leaching data for a number of metals in ash (see Chapter 16). These ligands form soluble complexes which keep the metal in solution at low and high pH values. A very simple solid phase-leaching solution system consisting of the solid phase Pb(OH)2(s ) in equilibrium with Pb2* as well as the various solution phase hydrolysis complexes of lead, Pb(OH) § Pb(OH2)~ Pb(OH)3 and Pb(OH)42 will be examined. Consider the following family of equilibrium reactions:

541 lo.cl K {Pb 2+} + {H20} ~ {Pb(OH) t} + {H +}

-7.70

{Pb 2+} + 2{H20} ~ {Pb(OH)2 } + 2{H +}

-17.75

(13.46)

{Pb 2§ + 3{H20} ~ {Pb(OH)3} + 3{H +}

-28.09

(13.47)

{Pb 2+} + 4{H20} ~ {Pb(OH) 2-} + 4{H +} -39.49

(13.48)

{Pb(OH)2(s)} + 2{H +} ~ {Pb 2+} + 2{H20 }

(13.49)

(13.45)

8.16

Looking at the individual hydrolysis species in equilibrium with the solid phase, the result for the first reaction is Io~ K {Pb 2+} + {H20} ~ {Pb(OH) +} + {H +}

-7.70

(13.50)

8.16

(13.51)

+ {Pb(OH)2(s)} + 2{H +} ~ {Pb 2+} + 2{H20 } {Pb(OH)2(s) } + {H t} ~ {Pb(OH)+} + {H20 }

0.64

(13.52)

Rearranging equation (13.52) results in: Iog{Pb(OH) t} = 0.46 - pH

(13.53)

This is plotted as line 1 on Figure 13.10. Like the last example, any combination of {Pb 2+} and pH above and to the right of the line means that lead precipitates out as Pb(OH)2(s). Conversely, Pb(OH) + is situated below and to the left of the line. Taking the other possible hydrolysis species and completing the same analysis produces the following sets of relationships: Iog{Pb(OH)2} = -9.59

(13.54)

Iog{Pb(OH)3} = -19.93 + pH

(13.55)

Iog{Pb(OH)~,} = -31.33 + 2pH

(13.56)

These are plotted as lines 2 to 4 respectively in Figure 13.10.

542 Figure 13.10 p{Pb2*} - pH Plot Showing Pb(OH)2 in Equilibrium with {Pb2*}, {PbOH2~ {Pb(OH)3 }, {Pb(OH)4 2}

2-

1

4 p{Pb 2+}

68

10-

2

122

4

6

8

10

12

pH To finish the analysis, we also want to look at how {Pb 2.} is in equilibrium with Pb(OH)2(s). Taking equation (13.49) results in: 108"16 =

{Pb 2+}{H20} 2

{Pb(OH)2(s)}{H +}2

(13.57)

Rearranging and solving for {Pb 2§ produces: {Pb 2*} = 108.16{H+}2

(13.58)

Taking the log of the equation results in: Iog{pb2+} = 8.16 -2pH

(13.59)

which plots as line 5 in Figure 13.10. The data in Figure 13.10 are very illustrative. Lead hydroxide shows minimum solubility at about pH 10. By simultaneously solving all the equilibrium expressions for {Pb 2+} and {H*}, it follows that the solid phase is in equilibrium with a number of aqueous species; Pb 2§ when pH is low (0-8) and OH is not abundant (10 14 to 10-6M), Pb(OH) § when pH is near 8 to 10 and OH is more abundant (10 .4 to 106M), Pb(OH)2 ~ at pH 10, Pb(OH) 3 at pH 10 to 12 when OH is very abundant (10.2 to 104M) and Pb(OH)42 when pH is very high (12 to 14) and OH is extremely abundant (10 .2 to 1.0M). Because PbOH2(s ) is both acid- and base-soluble, this solid phase is considered amphoteric.

543 Such phenomena are seen with Cd, Zn, Cu, Ni, Co and Hg. The example illustrates how metals exhibit solubility minima when leaching tests are conducted at different pH tests, as shown extensively in Chapter 16.

13.3.6 An Example of Lead Dissolution/Precipitation as PbCO3(s): The Role of CO2(g) in Controlling Metal Carbonate Formation The third simple example expands on the two previous examples. Again, ionic strength effects will be neglected so that activity is synonymous with concentration. Additionally, the role of ligands such as OH will be ignored although they could be added to the analysis. In this case, the presence of a constant partial pressure of a gas, CO2(g), results in the formation of a carbonate solid. Such carbonation or mineralisation reactions are common in incinerator ashes as they age. PbCO3 (cerrusite) is frequently found in ash residues. Like anglesite, it accounts for some of the available or leachable fraction of lead in combustion residues. Also like anglesite, PbCO3(s) exhibits a pH-dependent solubility because the conjugate base, CO32 (carbonate), can exhibit pH-dependent protonation reactions. However, in the case of carbonate, the source of conjugate base is derived in part from atmospheric CO2(g) diffusing into a wet alkaline solid. The partial pressure of CO2(g) can control the amount of carbonate available to precipitate out with lead to form cerrusite. Frequently, because of biological activity, partial pressures exceed those seen in atmospheric air. This example will be considered as a closed system with an elevated partial pressure and an elevated CT,co, of 1 x 10"2M. In a closed system, such as a large covered ash pile, the uptake of CO2 from the atmosphere is minimal and the sole source of CO2(g) is from microbial respiration. Systems open to the atmosphere can also be evaluated. Consider the following equilibrium reaction: {PbCO3(s)} ,~ {Pb 2+} +

{CO~-} log ~p

(13.60)

-13.1

Equation (13.60) can be rearranged and put into log form such that: Iog{pb2+} = _13.1 - log{CO 2-}

Like the anglesite example, it is possible to calculate the activity of pH for substitution into equation (13.61). The same type of equation (13.44) can be used to calculate {CO32}; however dissociation constants have changed (K,,1 = 106-3s, K,,2 = 101~

(13.61)

of CO32 as a function relationship shown in the appropriate acid By selecting various

544 pH values, the CO32- concentration as a function of pH can be determined by substitution into equation (13.61 ):

pH

{H §

{CO32-}

14

1 x 10 14

9.99 x 10 .3

12

1 x 10 12

9.79 x 10 .3

10

1 x 10 1~

3.11 x 10 .3

8

1 x 10 8

4.55 x 10 s

6

1 x 10 6

1.44 x 10 .7

4

1 x 10 .2

4.67 x 10 15

2

1 x 10 .2

4.67 x 10 15

0

1 x 10 o

2.08 x 10 19

If we substitute the log of {CO32-} into equation (13.61) and solve for {Pb 2§ we get: pH

P{Pb 2§

14

-11.90

12

-11.90

10

-10.59

8

-8.75

6

-6.25

4

-2.41

2

1.23

0

5.58

The values are plotted as a line in Figure 13.11 in a pC-pH plot.

13.3.7 Solubility Control and the Coexistence of Multiple Solid Phases As discussed earlier, the hypothetical solid phases (Figures 12.3 to 12.6, Chapter 12) show that multiple solid phases can be present initially, possessing differing initial solubilities. Upon initial dissolution, reprecipitation can occur where the secondary mineral has to be in equilibrium with all other solid phases containing the element of interest. Pankow (1991) describes some important solubility laws which are presented below.

545 Figure 13.11 p{Pb 2.} - pH Plot Showing PbCO3(s) in Equilibrium with {Pb 2.}

p{Pb 2+} 10 12i

2

,, I

4

I

6

I

8

pH

I

10

I

12

Metals can form solids that are simple salts, hydroxides, oxides, oxyhydroxides, and carbonates. A variety of other types of more complex solid phases are also possible. This is particularly relevant to reprecipitation and secondary mineral formation that occurs during ash leaching. It also seems important to ask if both of these solids could be present and in equilibrium with the same metal ion aqueous activity. Based on the above discussion, it can be concluded that the criterion for identifying the governing (i.e. the solubility-limiting) solid for a metal ion can be stated using criteria established by Pankow (1991 ) Solubility-Limiting Criterion (Pankow, 1991): The solid phase that will limit the solubility of a metal ion will be that which gives the lowest activity of the metal ion at equilibrium for the conditions of interest. Three corollaries to the above criterion are: Solubility-Limitin.q Corollary 1 (Pankow, 1991): If more than one solid is in equilibrium with a given metal ion, then all of those solids must specify the same activity of the metal ion for the conditions of interest. It may be noted that an analogous situation exists for non-metal containing, anionic species like CI, C032-, etc (Pankow, 1991). That is, the solid that will limit the anion solubility will be the solid which prescribes the lowest equilibrium activity for the conditions of interest (Pankow, 1991). Similarly, if more than one solid is in equilibrium with and controlling a species like sulphate, then all such solids must specify the same activity for that species (Pankow, 1991 ).

546 Solubility-Limitinq Corollary 2 (Pankow, 1991): If, for the conditions of interest a given solid prescribes a value of the activity of a metal ion that is lower than that prescribed by all other possible solids, then it will also prescribe values for all other dissolved, metal-containing species that are lower than prescribed by all other possible solids for those same conditions. Just as any given aqueous solution can be characterised by only one value of {OH} or {CI}, etc., any given species will be characterised by only one activity coefficient (Pankow, 1991 ). Thus, the species that prescribes the lowest values of {M2*}, {MOH (= 1)+}, {M(OH)2(z-2)§ {MCI(Z-1)§ etc. will also prescribe the lowest values of [MZ*], [MOH (z ~)§ [M(OH)2(=-2)*], [MCI(Z~)+], etc. and therefore, also the lowest value of the total dissolved metal, CT.M(Pankow, 1991 ). Solubility-Limitin~ Corollary 3 (Pankow, 1991): If a given solid minimises the value of the activity of a metal ion, it will also minimise the total dissolved concentration of CT,M; if twO solids are coexisting in equilibrium, then they will specify the same value of CT,M-

13.3.8 An Example of Lead Dissolution/Precipitation When Both Pb(OH)2(s) and PbCO3(s) Are Present The fourth and final simple example expands on the previous three examples. Again, ionic strength effects and the role of complexing ligands will be ignored for simplicity. In this example, the presence of two solid phases will be examined. In ash, an element such as lead will be present in a number of mineral phases. However, as pH changes, the relative predominance of any phase can change. There are system pH values where more than one solid phase can stipulate the same activity of Pb 2§ in solution. Consider the two solids just addressed, PbSO2(s) and PbCO3(s). If we plot on the same graph the solubility plots for PbSO4(s) ( CT,so, = 1x103M ) and for PbCO3(s) ( CT,co, = lx02M ), then the plot shown in Figure 13.12 is obtained. At a pH of about 5.20, the two plots intersect. At that pH, both solids specify the same activity for Pb2§ This can be shown mathematically by rearranging and combining equations (13.42) and (13.60) to produce: {PbCO3(s)} ,- {Pb 2+} + {CO 2-}

K,o

10 T M

+ {SO 2-} + {Pb 2+} ,, {PbSO4(s)}

{SO 2-} + {PbCO3(s)} ,=, {PbSO4(s)} + {CO 2-}

107.79

10 -531

(13.62)

(13.63)

(13.64)

547 The new expression, which takes into account equilibrium constraints from both solids can be modified to produce equation (13.65) (the activities of the solids, by convention, set to equal 1). 10_5.31 = { 002-}

(13.65)

{so:-}

What the relationship in equation (13.65)is showing is that the two solids will be in equilibrium when the ratio of the activities of the carbonate and sulphate anions in solution at equilibrium is equal to 10s31. Simple trial and error substitution into equation (13.44) for sulphate and its equivalent equation for carbonate show that at about pH = 5.20, the ratio is 10s31. Figure 13.12 p{Pb 2+} -pH Plot Showing PbCO3(s)and PbSO4(s)

PbSO4(s)

p{Pb 2+} 10 12 I

2

I

4

I

6

I

pH

8

I

I

10 12

In actuality, ashes contain many solid phases that can undergo incongruent dissolution. For instance, metal carbonates may be present but thermodynamically unstable with respect to a metal hydroxide-governed system. Such incongruence can result in dissolution and reprecipitation in the more thermodynamically-favoured solid phase. While data suggest that ashes do dissolve and reprecipitate (Eighmy et al., 1994), some caution about the application of incongruent dissolution is required. Phases tend to be incompatible with each other only when in intimate contact with each. Thus, heterogeneous mixtures of incompatible phases may remain incompatible only at the local level.

548 13.3.9 Solid Phase Stability in a Redox-Variable System

The relative oxidising or reducing characteristics of the leaching system plays a crucial role in controlling the types of solid phases that can form. Some of the ashes that are studied are inherently reducing. Certain management scenarios may also cause the deposition environment to become reducing. Therefore understanding such processes becomes important. The principals of redox are used in the modelling efforts described in Chapter 15. These effects are also shown in data presented in Chapter 16. Many of the scenarios depicted in Figure 12.2, in Chapter 12 for ash leaching scenarios have the opportunity to undergo changes in redox chemistry of the leaching system as a consequence of (i) the presence of bacteria which consume 02 and cause anoxic, reducing environments to develop, or (ii) the presence of reduced mineral phases in the ash that formed under low 02 partial pressures in the combustion process. Such mineral phases would include iron-containing minerals where Fe is present as Fe(ll) (e.g. FeCO3) or where sulphur is present as S(IV) (e.g. CaSO3(s)), S~ (e.g. elemental sulphur) and S(-II) (e.g. PbS). The role of redox in controlling the predominance of solid phases becomes a function of the type of solid phase that is present and whether or not the solid phase contains a redox-sensitive element. As shown in Figure 13.13 (after Hering and Stumm, 1990), a number of important redox couples exist for elements that can participate in redox reactions. The figure shows the domain of measured pE-pH (or Eh-pH) values seen in natural waters (the shaded area). The dotted lines show the domain where values of pE-pH are likely to exist, but have not been measured, for the redox couples that are known to control redox in the environment. The various solid lines demarcate where an element is in a more oxidised state (above and to the right of the line) or a more reduced state (below and to the left of the line). Additionally, the element in question can undergo changes in valency because it is also redox-sensitive. A recent review by Hering and Stumm (1990) looked at the mechanisms and rates of oxidative dissolution of reduced phases and reductive dissolution of oxidised phases. Rates are controlled by complex and surface reactions that include transport phenomena to and from the surface of both reactants and products. Superimposing such solid phase stability issues onto a heterogeneous, multiphase system, it becomes readily apparent that the system becomes increasingly more complex. Regardless of the complexity, this approach is used by geochemists to look at mineral stability in terrestrial systems. The early pioneering work of Pourbaix (1966), Garrels and Christ (1965), and Hem (1967) have laid down the foundation for the construction of Eh (or pE) vs pH stability field diagrams. A more recent effort by Brookins (1988) has resulted in the simple stability fields for almost all the elements of concern to ash chemists. While Brookins prepared the stability fields for use in terrestrial geochemistry systems, they have applicability to the MSW ash leaching system. The methods of Brookins (1988) and Verink (1979) have been used to'

549 construct stability fields for lead in combined ash and scrubber residue. The field is shown in Figure 13.14. Details are provided by Eighmy et al. (1990). Figure 13.13 A pE - pH Diagram Showing Important Redox Couples and Domains 15

~

10

OXIC o2~2

9"-~

5

,,
P~ 0

~~

H~! ANOXIC

,, MnO2ls)/Mn"1

-5

-100

SUB-OXIC

2

4

6

pH

8

10

12

From Hering and Stumm, 1990 with permission from the Mineralogical Society of America

Pb211

Figure 13.14 An Eh - pH Stability Field Diagram for Lead in Combined Ash and Scrubber Residue 0.6 0.4

bs,Po,3c,-T L I

i Ps,o

0.2 Eh

32

PbCO

3

PbO

0 -0.2 _ ~b

-0.4 -0.6 -0.8

S 'e-

~1.p,S,Si,c.lX10-3 -2

CT,cI-lXl0 I

i

2

I

1

4

i

I

I

6 pH

After Eighmy et al., 1990

I

8

I , I

10

i

12

14

550 For a {Pb 2§ of 1 x 101~ predominant phases include PbO, PbCO3, PbSiO3 and PbSO4 under aerobic (oxic) conditions and PbS under reducing conditions. Brookins (1988) and Nriagu (1974) have constructed similar diagrams. Such diagrams are simplistic but helpful in understanding solid phase control. For instance, are these phases seen with speciation methods? Under specific leaching tests, do lAP values suggest that these solid phases control leachability? In modelling these systems, do the geochemical source codes predict that these phases are controlling leaching? By changing the system, can new phases form? These questions get to the very basis of attempts to understand leaching mechanisms. 13.4 CHEMICAL WEATHERING AND AGING

Chemical weathering can be defined as the dissolution of relatively insoluble minerals by the action of water and its solutes (Schnoor, 1990). Until now, this chapter has focused on the dissolution and precipitation (or reprecipitation) of relatively soluble minerals. The more soluble system can be described with a thermodynamic equilibrium approach. Weathering, on the other hand, is better modelled and interpreted mechanistically using a kinetic approach. This approach is postulated because many of the leaching scenarios involving insoluble solid phases such as aluminosilicates will be concerned solely with the forward reaction in dissolution. The backward reaction can be neglected because the solutes are not available in sufficient quantity under appropriate conditions of temperature and pressure to allow the backward reaction to occur. It is estimated that up to fifty percent of the solid phases in bottom ash are phases that weather like their geological counterparts in metamorphic rock. Many of the discrete solid phases in combustion residues are characterised as iron, magnesium and calcium-containing aluminosilicates or insoluble precipitates/amorphous gels. These phases can comprise a large fraction of the ash particle or solid phase. While their abundance is dependent on the type of residue, the slow dissolution of such phases may have implications under some leaching scenarios where behaviour of this phase is important (i.e. the long term release of lattice-bound or lattice-substituted elements). Chemical weathering has not been studied in MSW combustion residues. Some work has been done on coal fly ash systems (Warren and Dudas, 1984; 1985; 1986). The majority of the work has been conducted on geologic materials. Reviews by Schnoor (1990), Schott and Petit (1987), Stumm and Weiland (1990), Stumm and Furrer (1987), and Casey and Bunker (1990) are useful, in-depth summaries of dissolution phenomena of these relatively insoluble phases. This brief chapter section is presented as an introduction to the topic. Clearly more work is needed in this area, particularly with regard to evaluating processes for long term leaching applications (e.g. 100 to 1000 years).

551 13.4.1 The Mineral Surface

At the atomic level, the mineral surface might be conceptualised as shown in Figure 13.15 (Blum and Lasaga, 1987). The surface is three dimensional and contains terraces, steps (or edges), kinks in the edges, and adsorbed atoms (adatoms). The process of solvent interaction with the surface promotes slow dissolution of the surface by the solvent. Figure 13.15 Mineral Surface

V

>@.<

KINK

SURFACE

From Blum and Lasaga, 1987. @1987 John Wiley & Sons, Inc. Permission John Wiley & Sons, Inc

Reprinted by

The dissolution of the mineral surface is initiated at sites of high surface energy. These occur at terrace edges, adatoms and kinks at the atomic level and at cracks, scratches or holes in the mineral surface at a larger, sub-micron to micron scale (Schott and Petit, 1987). Ash particles, like geologic materials, have cracks, scratches and holes at the larger scale. 13.4.2 Weathering Reactions

Berner (1978) has visited the issue of rate control of mineral dissolution. Table 13.11 shows a variety of minerals exhibiting high to low solubility. Berner notes that there is a good correlation between the solubility of the mineral and the rate-controlling mechanism for dissolution.

552 Table 13.11 Dissolution Mechanisms Mineral

Solubility in Pure Water (moles/L)

Surface Reaction Control KAISi308

3 x 10.7

NaAISi308

6 x 10.7

BaSO4

1 x 10"s

SrCO3

3 x 10s

CaCO3

6 x 10s

Ag2CrO4

1 x 104

SrSO4

9 x 104

Opaline SiO2

2 x 10.3

Mixed Control PbSO4

1 x 10"4

Transport Control AgCI

1 x 10.5

Ba(IO3)2

8 x 104

CaSO4.2H20

5 x 10.3

Na2SO4 10H20 9

2 x 10"1

MgSO4.7H20

3 x 10~

Na2CO3 10H20 9

3 x 10~

KCI

4 x 10~

NaCI

5 x 10~

MgCI~ .6H~O 5 x 10~ Adapted from Berner, 1978 reprinted with permission of the American Journal of Science

Three domains are described. The first is the transport control domain where dissolution is so fast that the ability of the dissolved solute to diffuse out of the depleted surface of the mineral or out of the diffusion boundary layer around the particle rate limits dissolution. The second is the surface reaction control domain, where dissolution kinetics are so slow compared to diffusional transport that dissolution is the ratecontroller. The third domain is a mixture of the two processes.

553

13.4.3 Surface Reaction-Controlled Dissolution The dissolution mechanism that occurs at the mineral-water interface involves a number of reactants. These include the components of water, H20, H* and OH, as well as the ligands present in the leaching solution. The slow, kinetically-constrained dissolution involves both chemical and physical reaction steps. These steps are: the attachment of reactants at reaction sites where they polarise and weaken metal-oxygen bonds in the mineral lattice and the rate limiting detachment of surface metal species. This two-step reaction is depicted as follows (Stumm and Furrer, 1987): I. Surface (adatoms, edges, kinks) + (H*, HO ~ + reactants

fast ~ > (< .... ) slow

I1. Surface detachment of activated species

(< .... )

Activated surface species >

Aqueous metal ion

For the two-step reaction, backward reaction arrows are shown. When initial surface sites are consumed and new sites are created, steady state conditions can develop. Finally, if the system is far from equilibrium or if equilibrium is unattainable in the leaching system under consideration, then the back reactions can be assumed to be negligible and the dissolution rate becomes (Stumm and Furrer, 1987): Dissolution Rate =

(ReactantConcentration / at the Surface )

x

( Densityof /n Surface Sites)

where n is the number of mobile, solvent phase reactants required to activate the metal ion at the reaction site. The higher the surface energy at the reaction site, the faster the rate of dissolution. Dissolution may initially be incongruent. Basic cations such as Ca 2., Mg 2., Na*, K* may preferentially dissolve before AI 3. and Si 4.. Subsequent dissolution then becomes congruent when all elements are removed at the same rate (Schnoor, 1990). Weiland et al. (1988) have developed a model to describe surface reaction controlled dissolution. The rate law is: R where

= kXaPjSsite s

(13.66)

R is the proton or ligand-promoted dissolution rate (mol m2 sl), k is the rate constant (s-l), X, is the mole fraction of reaction sites (dimensionless), Pj is the probability of a finding a site for an activated precursor complex, Ss.es is the total surface concentration of sites (mol m2).

554 Schnoor (1990) provides a detailed description of a surface reaction-controlled dissolution reaction. The surfaces of oxides and aluminosilicate minerals in the presence of water are characterised by amphoteric surface hydroxyl groups. The surface OH group has a complex-forming O-donor atom that coordinates with H § and metal ions (Kummert and Stumm, 1980; Sigg and Stumm, 1981). The underlying aluminum or metal ion in the surface tetrahedra of the mineral and other cations are subject to coordination with O H groups, which, in turn, weakens the bond to their structural oxygen atoms. Detachment of an activated complex removes the coordination complex and renews the surface for further hydrolysis, protonation and dissolution. An example of hydrolysis, protonation, formation of the surface coordination complex and detachment to solution is shown in Figure 13.16 for an aluminum oxide. Figure 13.16 Dissolution of an Aluminum Oxide Surface

. / \ / \ / \ / \ / ~AI \ / AI\ / \AI / \ AI / \ o

o

0

o

0

0

A! H O

H O

H O

H O

o

+nH20

AI

0

oxide surface

H O

H O

H O

H O

H O

H O

\ / \ / \ / \ / \ / AI AI AI AI / \ / \ / \ / \ / \

H O +3H §

AI

hydrous oxide surface

H o

H o

oH

(~' ~)' H'

H o

H o

H o

H o

\ / \ / \ . / \ / \ / AI AI AI AI AI / \ / \ / \ / \ / \

H o ~ - ~

(AI-3H,

O ) ~§

+

protonated surface, formation of surface coordinated activated complex H O

H O

\ AI /\

H O

O

O

H O

H O

H O

/ \ / \ / AI AI /\/\/\ O

H O AI

O

detachment of activated complex H O

H O

\ AI /\.

H O

.

O

O

\../

H O

H O

H O

/ \ / \ / AI AI . / \ / \ / \

O

O

H O AI

O

renewed hydrous oxide surface

Schoor, 1990. & Sons, Inc.

9

John Wiley & Sons, Inc. Reprinted by permission John Wiley

555 A very important result of recent laboratory studies has been the fractional order dependence of mineral dissolution on solution phase hydrogen ion activity. If the dissolution reaction is controlled by hydrogen ion diffusion through a boundary layer, a first-order dependence on {H § would be expected. If the dissolution reaction is governed by some other factor like surface area, then the dependence on hydrogen ion activity should be zero-order. Rather, the dependence exhibits a fractional order for a wide variety of minerals, indicating a surface reaction controlled dissolution (Schott et al., 1981; Giovanoli, et al. 1989; Schnoor and Stumm, 1986; Busenberg and Plummer, 1982; Grandstaff, 1977; Furuichi et al., 1969). A strong dependency on the dissolution rate can be seen. Wollast and Chou (1985) have looked at the dissolution of silica from various mineral phases; minimal releases are seen at pH 3.0. This pH corresponds to the zero point of charge (ZPC), or pH condition where the negative site density of adsorbed OH equals the positive site density of adsorbed H§ Relative protonation or hydroxylation enhances site activation.

13.4.4 Weathering Rates There is quite a lot of data available on dissolution rates of minerals. While these rates have not yet been formally established for MSW residues, it is likely that similar rates will be shown. Table 13.12 summarises a few data (Schnoor, 1990) for silicon release from a variety of silicon-containing minerals. The rates of Si dissolution are all similar: approximately 5 x 1012 moles per cm 2 per second. Table 13.12 Measured .Weathering Rates Mineral

Weathering Rate (tool Si m"2 s1)

Plagioclase (oligoclase)

5 x 1012

Plagioclase (bytownite)

5 x 1012

Olivine

7 x 1012

Plagioclase, biotite Adapted from Schnoor, 1990. 9 permission of John Wiley & Sons, Inc.

6 x 1012 John Wiley & Sons, Inc.

Reprinted with

There are no direct data available on the weathering rates of glassy phases in various combustion residues. There are mineral phases in bottom ash, e.g. fine-grained quartz and feldspars, that will undergo weathering reactions according to models adopted by Berner (1981). A cursory examination of silicon leaching rates from bottom ash for a variety of less aggressive leaching tests shows that silicon leaching rates can range

556 from 5 x 1013 to 5 x 101~ moles per cm2 per second. These values fall within the same magnitude as those reported by Schnoor (1990). This suggests that silicon leaching is probably a kinetically slow weathering-like reaction in bottom ash.

13.4.5 Aging Reactions There are a number of aging reactions that can occur in hydrated ash specimens over time. The aging reactions are diagenetic in nature; the residue becomes more "rocklike" in its mineralogy and phase compatibility. The underlying processes that cause aging are the thermodynamic instability of, or incongruence between, mineral phases. The slow, irreversible conversion to more stable phases is the process that corrects the instability. There are at least six processes that occur during aging reactions: (i) the oxidation of elemental metals to oxides, (ii) the hydrolysis of oxides and liberation of exothermic heat, (iii) the uptake of carbon dioxide and the subsequent formation of carbonatecontaining minerals, (iv) the redox-driven oxidation of elemental metals (notably AI ~ and reduction of water to generate hydrogen gas, (v) the formation of certain mineral phases in bottom ashes that act as pozzolans for particle aggregation and strength development, and (vi) the formation of clay-like phases in residues. The oxidation of elemental metals to form metal oxides is a small, but not insignificant aging reaction that can occur in all residues. An example would be the conversion of elemental iron to iron oxides. The following reactions depict the conversion of ferric iron to various iron oxides found in bottom ash: Fe 3+ § 3/2 H20 ~ -

1 a-Fe203(hematite ) § 3H + 2

Fe 3* + 2 H20 ,= a-FeOOH(goethite) + 3H +

(13.67) (13.68)

Goethite is the most stable and least soluble form of iron oxide in oxidised systems (Schwertmann and Taylor, 1977). Hematite will convert to goethite via the following reaction: Fe203 + H20 ,~ 2a-FeOOH

(13.69)

557 The hydrolysis of metal oxides can be an aging reaction that releases substantial quantities of heat. An example reaction is the conversion of an oxide to a hydroxide: CaO + H20 = Ca(OH)2

(13.70)

This reaction, which may occur when APC residues are wetted, tends to release 400 to 500 joules of heat for every kilogram of CaO that is hydrolysed (Boynton, 1980). The uptake of CO2(g) from the atmosphere into alkaline residues is another significant aging reaction. Many hydroxide phases can be converted to carbonate phases as shown in the following reactions: Ca(OH)2

+ C O 2 ~=

CaCO 3 + H20

(13.71)

CaCO 3 + CO 2 + H20 = Ca(HCO3) 2

(13.72)

Kluge et al. (1980) have studied the redox reactions occurring during ash aging that result in the release of hydrogen gas. The most likely reaction involves elemental aluminum under alkaline conditions: 2 AI ~ + 6H20 ,= 2 AI(OH)3(s ) + 3H2(g)

(13.73)

Kluge (1980) was able to quantify hydrogen gas release from bottom ash used as road base. Hydrogen gas is probably evolved when bottom ash is quenched. Oberste-Padtberg and Schweden (1990) have measured H2 release under alkaline conditions when fly ash is stabilised with Portland cement. They also quantified methane and carbon monoxide evolution. The reaction for methane evolution was hypothesised to involve metal carbides which are found in the residues: Me4C3 + 6H20-~ Me203 + 4CH.

(13.74)

The next important aging reaction involves the formation of certain mineral phases associated with pozzolanic reactions. Such reactions can change leaching behaviour by altering particle-specific surface areas and reducing porosity. Such reactions have not been extensively studied in bottom ashes. Van der Wegen (1991) has found ettringite (3CaO.AI203~ in aged bottom ash specimens. SEM and XRPD were used to detect the mineral phase. Ettringite, formed from gypsum (CaSO4 2H20) 9 and tricalcium aluminate (Ca~,1206), is an important initial phase in the cementation process. St~mpfli (1992) identified gypsum in aged bottom ash. He was

558 not able to find tricalcium aluminate or the calcium silicates (Ca3SiO5 or Ca2SiO4) that are needed to complete the pozzolanic cementation process despite the monolithic nature of the sample. The last aging reaction to be identified is the weathering of glassy phases in bottom ash to form clay-like crystal lattice structures. Zevenbergen et al. (1993) obtained bottom ash from a 12-year old disposal site. A variety of analytical techniques were used to identify the clay-like structure. There are undoubtedly many other aging phenomena that have not been identified. As more researchers investigate these processes, more aging and weathering reactions will be identified. 13.5 SORPTION The solid phase in the ash leaching system (Figure 12.2, Chapter 12) can also play an important role in controlling the appearance and disappearance of solutes in the leaching solution through the process of sorption/desorption. Like precipitation/dissolution and solution-phase complexation, sorption is usually an equilibrium reaction. Many field and laboratory leaching data must be interpreted with sorption as another equilibrium process that must be simultaneously solved with other equilibrium expressions. Sorption plays an important role in the fine-tuning of modelling efforts introduced in Chapter 15. Most ash residues contain a number of metal oxides. These mineral phases include anatase (TiO2), corundum (a-AI203), magnetite (Fe304), rutile (TiO2) and quartz (aSiO2). Such mineral phases, as well as others that are present in combustion residues, constitute the sorptive surfaces in ash. The thermodynamics and modelling of sorption to mineral surfaces is complex. The following section provides a brief introduction to the topic. Adsorption processes undoubtedly play an important role in controlling metal leaching. However, the process has not been well studied in residues. The reader is encouraged to review the works of Sposito (1984), Leckie (1988), Sigg (1987), Schindler and Stumm (1987), Parks (1990) and Davis and Kent (1990). As research continues into a mechanistic understanding of leaching, it is anticipated this topic will become more important to understanding and modelling leaching behaviour. 13.5.1 Surface Functional Groups Metal oxides and hydroxides are surfaces where adsorption reactions take place. Oxides that become hydrated possess proton-bearing surface functional groups that can disassociate at high pH or reassociate at low pH. Aluminosilicate mineral phases

559 that do not have isomorphic substitutions, or substituted atoms in the crystal lattice, also can develop hydrated proton-bearing surface functional groups. As shown in Figure 13.17, surface hydroxyl groups comprise the functional groups on a variety of oxide and aluminosilicate mineral surfaces. Various types of surface hydroxyl groups are shown. The M (metal) and O (oxygen) function groups at the surface exhibit negative charge distributions that cause H20 to "chemisorb" to form surface hydroxyl groups. The chemisorption process usually involves a loss of a proton from the water molecule. Hydrogen bonding between the surface hydroxyl groups and either water vapour or liquid water causes layers of water to physically adsorb to the surface (Davis and Kent, 1990). These hydroxyl functional groups have different reactivities (Sposito, 1984); some hydroxyl groups dissociate easily to form the anionic oxygen ligand, other hydroxyl groups dissociate less easily. Another type of site that can develop is a Lewis Acid site (Davis and Kent, 1990). In such a site, H20 chemisorbs directly to a bare metal ion. In this case the site can only act as a proton donor. Figure 13.17 Different Types of Mineral Surface Hydroxyl Groups

From Davis and Kent, 1990 with permission from the Mineralogical Society of America

560 Table 13.13 provides some information on hydroxyl site densities for various minerals. Values range from 1.3 to 22 functional groups per nm 2. Goethite, a particularly adsorptive iron oxide (a-FeOOH) that is also a principal mineral phase in bottom ash, exhibits different numbers of adsorption sites depending upon the adsorbing cation (Table 13.14). Table 13.13 Density of Surface Functional Groups on Various Oxide and Hydrous Oxide Minerals Mineral Range of Site Densities (sites nm "2) Goethite (a-FeOOH)

2.6-16.8

a-Fe203 Rutile (TiO2) Gibbsite (a-AI(OH)3)

5-22 12.2 2-12

y-AI203 6-9 SiO 2 (Quartz) 4.5-12 Kaolinite 1.3-3.4 Adapted from Davies and Kent, 1990 with permission from the Mineralogical Society of America Table 13.14 Site Densities on Goethite (a-FeOOH) Adsorbing Ion Adsorption Maximum (sites nm2) OH/H § OH F SeO32 PO43

4 2.6 5.2 - 7.3 1.5 0.8

Pb2§ 2.6 - 7.0 Adapted from Davies and Kent, 1990 with permission from the Mineralogical Society of America The surface functional groups exhibit acid-base behaviour; they can donate or receive protons. The pH at which the mineral exhibits no net charge (or "zero point charge") in an electric field, the pHzp~, is useful in understanding at what pH mineral surfaces exhibit tendencies to adsorb solutes of opposite charge. At a pH < pHzp~, the surface would adsorb anions. At pH > pH=p~,the surface would adsorb cations. Table 13.15 depicts some PH=pcfor various adsorbent mineral phases.

561 Sorption phenomena are usually modelled using empirical, experimentally determined sorption constants based on solute activity or using mechanistic explanations of the electrostatic interactions that occur at the particle surface. Both approaches can be modelled in the modelling programs described in Chapter 15. Table 13.15 Estimates of pHD=e for Various Minerals Mineral

pHDzc

y-AI203

8.5

Anatase (TiO2)

5.8

Birnessite (5-MnO2)

2.2

Calcite (CaCO3)

9.5

Corundum (a-AI203)

9.1

Goethite (a-FeOOH)

7.3

Hematite (a-Fe203)

8.5

Magnetite (a-Fe304)

6.6

Rutile (TiO2)

5.8

Quartz (a-SiO2) 2.9 Adapted from Davis and Kent, 1990 with permission from the Mineralogical Society of America

13.5.2 Activity-Based Sorption Models The first activity-based sorption model is based on the distribution coefficient, Kd (Allison et al., 1990). Using the convention of M as a metal and SOH as a hydroxylated surface sorption site, consider the following reaction depicting m sorbing to the surface site: SOH + M -

SOH" M

(13.75)

The ratio of sorbed metal concentration to the total analytical metal concentration in solution at equilibrium is the distribution coefficient. It is analogous to an equilibrium constant: Kd = [SOH" M] [M]

(13.76)

562 It is more appropriate to examine the Kd relationship based on the activity of participating aqueous metal solute: K da c t _ -

By convention [SOH

{SOH" M}

(13.77)

{i}

9 M] = {SOH ~ M} and equation (13.77) becomes: act_

Kd

-

[SOH" M] u [M]

(13.78)

Many Kd"= values are tabulated in the sorption review document prepared by Rai and Zachara (1984). The second approach that employs activity in an empirical way to depict sorption is the Langmuir adsorption model (Allison et al., 1990). Again, using the SOH and M terminology, consider the following reaction: SOH + M ,= S O H - M

(13.79)

where at equilibrium K Lact

_ -

{SOH

9M }

(13.80)

{M} {SOH}

Conventionally, the Langmuir constant is derived experimentally using various quantities of SOH and M. To place ~"= into the more familiar context of the Langmuir adsorption isotherm, a mass balance on surface sites is needed (Allison et al., 1990): [SOH]T = [SOH

9 M] + [SOH]

(13.81)

Combining equations (13.80) and (13.81) produces:

[SOH

9 M] =

K act

L [SOH]T Ya[M] act

1 + "L I~"

Yi[ M]

(13.82)

The only difference between Kda~ and ~a~t is that the Langmuir equation assumes a finite concentration of SOH. Values for ~"~ can be found in Rai and Zacchara (1984).

563 The third activity-based empirical sorption model is the Freundlich model (Allison et al., 1990). Again using the SOH and M terminology, consider the following reaction: SOH + l/nM = SOH. M

(13.83)

where at equilibrium Kfact =

{SOH 9M} {M} TM {SOH}

(13.84)

Imposing the convention that {SOH ~ M} = [SOH ~ M], equation (13.84) becomes: [SOH

9M] = Kfact {MM+}TM

(13.85)

The 1In term is a mass action stoichiometric coefficient related to M. Kf"ct is similar to Kd~ if n=l. Kfact differs from ~act in its implicit assumptions about an unlimited supply of unreacted surface sites at equilibrium. Kf"~tvalues can be found in Rai and Zacchara (1984). The fourth empirical, activity-based sorption model is the ion exchange model (Allison et al, 1990). Again using the SOH terminology, but denoting the exchangeable metal as M and the sorbing metal as M2, consider the following SOH- M 1 - M 1 § M 2 ,, S O H - M 2

(13.86)

where at equilibrium, Kex =

{M1}{SOH'M 2} {M2}{SOH'M 1}

(13.87)

K.x differs from the other three constants by assuming a substitution reaction occurs at SOH. K,x values are found in Rai and Zacchara (1984). There are no conventions as to the applicability of these four empirical models (Allison et al., 1990). When modelling sorption processes, all four models can be tested (provided appropriate constants are available). The literature does suggest that mechanistic models based on electrostatic considerations do a better job at modelling sorption (Allison et al., 1990); frequently parameters for these models are not available (as discussed below).

564

13.5.3 Electrostatic Surface Complexation Models There have been a number of attempts to model adsorption to mineral surfaces while taking into account electrostatic interactions between charged surfaces and solutes. The models explain how both cations and anions adsorb as a function of pH, adsorbent site density and ionic strength (Westall and Hohl, 1980). Models such as the constance capacitance model (CCM) and the diffuse double layer model (DLM) were developed (Davis and Kent, 1990). A third model, the triple layer model (TLM), was developed to have multiple adsorption planes on the mineral surface to allow for outer sphere as well as inner sphere complexes to form (Leckie, 1988). The TLM is viewed as most applicable and is presented here. The TLM is schematically depicted in Figure 13.18 (Leckie, 1988). As described by Evans (1989), tightly bound inner sphere complexes reside in the inner or surface plane, ~. (or cx plane). Outer sphere complexes reside in the adjacent plane, )~ (or 13 plane). Noncomplexed species reside in the diffuse layer, A,~(or d plane). To maintain electroneutrality, the charge density, o, must equal the intrinsic charge density of the mineral such that:

(13.88)

Oint + Ois + Oos + 0 d = 0

Figure 13.18 Depiction of the Triple Layer Model % ~r=

o'p

I I I

Od

I I I

NO~

§

I

-r" ?~ 0 0

(I)

~:" ~:;!

:Z~

O~b*.... I

OebOH I

O

I I

I

I

....

I

....

,Np

....

,"9;

....

l~b 2.

~L~! a

....

Pt~OH§

, |

a~

NO~

I~a§ N

NO 3

pb 2+

I

i"L;: Ol'

~'#

Na* NO~

NO~ I

~--~-.~ O H +

~

NO~

Na* NO~ NO 3

N% pb 2+

NO3

CI_ ,, J

13

immobile layer

d diffuse

layer

Reprinted with permission from Leckie, 1988. Copyright Lewis Publishers, an imprint of CRC Press, Boca Raton, Florida. 9

565

Hayes (1987) has depicted the types of binding mechanisms that can occur in both the inner and outer spheres (Figure 13.19). The figure shows various inner sphere and outer sphere complexation reactions that can occur. Figure 13.19 Schematic Depiction of Coordinative Surface Complexes and Ion Pairs at Oxide Surfaces Metal

Oxygen

Other Examples

C~c,-b ~

",v..,,,>Water '~"" Molecules

C~a~1b

t3

I-. ~-. No;. ClO,-

Na* K + 2+ 2. , , Ca , Mg

uter-Sphere Complexes

c~,L co;k

"O"-~>__O_

pb.

Cu

f Divalent Transition Metal Ions

Monodentate

Divalent Transition Metal Ions

Bidentate / Inner-Sphere

"O~~

U~ap~exe s

F

,,O-'- H P

OH

Mononudear

SeO~;,~02

Binuclear

From Hayes, 1987 with permission of the author

566 Consider the following reaction for the monovalent metal ion M § (Allison et al., 1990): SOH

+

+

H s + M s ,- ( S O . M )

-

(13.89)

where SOH and Ms+ are the same surface binding site and sorbing metal and Hs§ is the sorbed proton that must deprotonate from SOH to allow for formation of the sorbed complex SO 9M. By convention,

and

{H~} = {H *} [e-~~

(13.90)

{ms} = {m *} [e -YI~FIRT]

(13.91)

+

where e "~F/RT is the Boltzmann factor for either the a or 13 planes depicted in Figure 13.19. Equations (13.86) through (13.88) can be written as an equilibrium expression (Allison et al., 1990): K = {SO

9M} {H *} [e-~~

(13.92)

{SOH} {M *} [e-~I3F/RT]

Other forms of equation (13.89) can be provided for a hydrolysis reaction of the type (Allison et al., 1990): SOH + M 2+ § H20 - 2H~ - S O . MOH

(13.93)

producing, K = {SO

9MOH} {H +}2 [e-I~~

2

{SOH} {M 2+} {H20 } [e-~I]F/RT]2

(13.94)

A similar approach can be taken for a sorbing monovalent anion (Allison et al., 1990): +

SOH + A s- + H s ,, SOH 2 A 9

(13.95)

producing, {SOH 2 A} 9 [e-~IBF/RT] K

._.

{SOH} {A-} {H *} [e-I~~

(13.96)

Finally, a similar approach can be taken for a sorbing divalent anion (Allison et al., 1990):

567 SOH + A 2- + H s =* SOH 2 A 9 t

(13.97)

producing K

=

{SOH 2 A-} 9 [e-e"F/RT]2 {SOH} {A 2-} {H +} [e-~oF/Rm]

(13.98)

The geochemical model MINTEQA2 has the capability to model sorption using estimated parameters for the TLM as well as the CCM and DLM. The TLM model has been widely used to model adsorption of cations and anions (Davis and Kent, 1990). For adsorption to oxides, Davis and Leckie (1978, 1980), Balistrieri and Murray (1982), Hsi and Langmuir (1985), Catts and Langmuir (1986), LaFlamme and Murray (1987), Zachara et al. (1987), Hunter et al. (1988), and Zachara et al. (1987) have successfully used the model. It has also been applied to non-hydrous oxide minerals like quartz, titanium and clay (Shuman, 1986). 13.5.4 Adsorption Data Davis and Kent (1990) have compiled some interesting data as to how cations and anions adsorb to mineral surfaces. As shown in Figure 13.20, there is a narrow pH range where a cation or anion goes from near zero adsorption to high levels of adsorption. This adsorption edge occurs at pHads, the pH where significant sorption occurs, which is close to the Prize:. Cations adsorb at high pH by forming inner sphere complexes with the deprotonated hydroxyl functional groups. Anions adsorb at low pH when forming inner sphere complexes with protonated functional groups. 13.6 A UNIFIED APPROACH TO LEACHING The information compiled in this chapter allows us to assemble an approach to characterising the leaching process. Understanding fundamental leaching behaviour of a residue such as ash requires the consideration of many factors (Figure 13.21). The speciation of the elements in the solid phase plays a fundamental role in controlling the nature of the leachate. This can then be related to that fraction of an element that is available for leaching. Particle morphology, porosity, and diffusion pathlength are also critical in assessing the role of diffusion in controlling reactions. Attempts to characterise those leaching processes that are kinetically based or thermodynamically based is an additional approach that is needed. The thermodynamic mechanisms can be modelled with geochemical codes. Additional data can be gleaned by conducting leaching studies to assess the effects

568 of pH, Eh and ligands on dissolution phenomena. The processes surrounding sorption cannot be ignored; these studies can be compared to the various sorption models that are presently available. Finally, the role of the L/S ratio and time must be considered. Figure 13.20 Adsorption Edges for Various Cations and Anions

From Parks, 1990 with permission of the Mineralogical Society of America Now that these concepts have been explained, they need to be applied. The next chapter sets out various types of leaching tests which can be performed to determine certain characteristics of ash under specific sets of leaching conditions.

569 Figure 13.21 Schematic of Fundamental Leaching Behaviour

Available

Fraction

Particle

Chemical Speciation

Therm odynamics

Morphology

FUNDAMENTAL LEACH IN G BEHAVIOUR

Kinetics

Influence of LS, Time

Influence of pH, pE & Ligands Sorption

570 REFERENCES

Allison, J.D., D.S. Brown and K.J. Novo-Gradac. MINTEQA2./.PR.0DEFA2.....A Geochemical Assessment Model for Environmental Systems: Version 3.0 User's Manual. Environmental Research Laboratory, U.S. EPA, Athens, GA, 1990. Baes, C.F. and R.E. Mesmer. The Hydrolysis of Cations. John Wiley & Sons, New York, 1976. Balistrieri, S.L. and J.W. Murray. The Adsorption of Cu, Pb, Zn, and Cd on Goethite from Major Ion Seawater. Geochim. Cosmochim. Acta 46, pp. 1253-1265, 1982. Belevi, H. and P. Baccini. Long-Term Assessment of Bottom Ash Monofill Leachates. In Proceedinas of the International Conference on Municipal Solid Wa.ste .C.ombustion. April 11-14, Hollywood, FL, USA, 1989. Bemer, R.A. Rate Control Of Mineral Dissolution Under Earth Surface Conditions. Am. J. Sci. 278, pp. 1235-1252, 1978. Berner, R.A. Kinetics of Weathering and Diagenesis. In Kinetics of Geochemical Processes (A.C. Edited by A.C. Lasaga and R.J. Kirkpatrick. Mineralogical Society of America, Washington, D.C., p. 111, 1981. Blum, P.J. and A.C. Lasaga. Monte Carlo Simulations of Surface Reaction Rate Laws. In Aquatic Surface Chemistry: Chemical Processes at the Particle-Water Interface Edited by W. Stumm. John Wiley and Sons, NY, p. 255, 1987. Boynton, R.S. Chemistry and Technolo.qy of Lime and Limestone. John Wiley and Sons, NY, 1990. Brookins, D.G. Eh-pH Diagrams for Geochemistry. Springer-Verlag, Berlin, 1988. Busenberg, E. and A.B. Plummer. The Kinetics of Dissolution of Dolomite in 002-H20 Systems at 1.5 to 65~ and 0 to 1 atm pCO2. Am. J. Sci. 282, pp. 45-78, 1982. Butler, J.N. _Solubility and pH Calculations. Addison-Wesley, Reading, MA, 1964. Casey, A.B. and A.B. Bunker. Leaching of Mineral and Glass Surfaces. In MineralWater Interface Geochemistry Edited by M.F. Hochella, Jr. and A.F. White. Mineralogical Society of America, Washington, D.C., p. 397, 1990. Catts, A.B. and A.B. Langmuir. Adsorption of Cu, Pb and Zn onto birnessite (6-MnO2). J. Appl. Geochim 1, pp. 255-264, 1986.

571 Cole, D.R. Theory and Application of Adsorption and Ion Exchange Reaction Kinetics to in situ Leaching of Ores. In Leachin.q and Diffusion in Rocks and Their Weatherin.(:i Products. Edited by S.S. Augustithis. Theophrastus Publishers, Athens, p. 3, 1983. Comans, R.N.J., H.A. van der Sloot and P.A. Bonouvrie. Speciatie van Contaminaten Tijdens Uitlo.qin.qvan AVI-Bodemas. Concept Eindrapport ECN-CX-93-XXX, April 1993. Davis, J.A. and J.O. Leckie. Surface Ionization and Complexation at the Oxide/Water Interface. I1. Surface Properties of Amorphous Iron Oxyhydroxide and Adsorption of Metal Ions. J. Colloid Interface Sci. 67, pp. 90-107, 1978. Davis, J.A. and J.O. Leckie. Surface Ionization and Complexation at the Oxide/Water Interface. 3. Adsorption of Anions. J. Colloid Interface Sci. 74, pp. 32-43, 1980. Davis, J.A. and D.B. Kent. Surface Complexation Modelling in Aqueous Geochemistry. In Mineral-Water Interface Geochemistry Edited by M.F. Hochella, Jr. and A.F. White. Mineralogical Society of America, Washington, D.C., p. 177, 1990. de Groot, G.J., H.A. van der Sloot and J. Wijkstra. Leaching Characteristics of Hazardous Elements from Coal Fly Ash as a Function of the Acidity of the Contact Solution and the Liquid/Solid Ratio. In Edited by P.L. C6t6 and T.M. Gillian.ln Environmental Aspects of Stabilization and Solidification of Hazardous and Radioactive Waste ASTM STP 1033, ASTM, Philadelphia, PA, pp. 170-183, 1987. Drever, J.l. The Geochemistry of Natural Waters. Prentice-Hall, Englewood Cliffs, NJ, 1988. Dudas, M.J. Long-Term Leachability of Selected Elements from Fly Ash. Environ. Sci. Technol. 15, pp. 840-843, 1981. Eighmy, T.T., S.F. Bobowski, T.P. Ballestero and M.R. Collins. Theoretical and Applied Methods of Lead and Cadmium Stabilization in Combined Ash and Scrubber Residues. In proceedin,qs of the Second International Conference on Municipal Solid Waste Combustor Ash Utilization. Edited by W.H. Chesner and T.T. Eighmy. November 8-9, Arlington, VA, p. 275, 1990. Evans, L.J. Chemistry of Metal Retention by Soil. Environ. Sci. Technol. 23, pp. 10461056, 1989. Feitknecht, W. and P. Schindler. Solubility Constants of Metal Oxides, Metal Hydroxides and Metal Hydroxide Salts in Aqueous Solution. Butterworths, London, 1963. Felmy, A.R., D.C. Girvin and E.A. Jenne. MINTEQ-A Computer Pro,qram for Calculatin,q Aqueous Geochemical Equilibria. EPA-600/3-84-032, U.S. EPA, Athens, GA, 1984.

572 Fruchter, J.S., D. Rai and J.M. Zachara. Identification of Solubility-Controlling Solid Phases in a Large Fly Ash Field Lysimeter. Environ. Sci. Technol. 24, pp. 1173-1179, 1990. Furuichi, R., N. Sato and G. Okamoto. Reactivity of Hydrous Ferric Oxide Containing Metallic Cations..Chimia 23, pp. 455-463, 1969. Garrels, R.M. and C.L. Christ. Solutions, Minerals, and Equilibria. Harper and Row, New York, 1965. Giovanoli, R. J.L. Schnoor, L. Sigg, W. Stumm and J. Zobrist. Chemical Weathering of Crystalline Rocks in the Catchment Area of Acidic Ticino Lakes, Switzerland. Clays Clay Min 36, pp. 521-529, 1989. Grandstaff, D.E. Some Kinetics of Bronzite Orthopyroxene Dissolution. Geochim. Cosmochim. Acta 41, pp. 1097-1103, 1977. Guggenheim, E.A. Applications of Statistical Mechanics, Clarendon Press, New York, 1966. Harvie, C.E., N. Mr and J.H. Weare. "The Prediction of Mineral Solubilities in Natural Waters: The Na-K-Mg-Ca-H-CI-SO4-OH-HCO3-CO2-H20 System to High Ionic Strengths at 25~ ''. Geochim. Cosmochim. Acta 48, pp. 735-751, 1984. Harvie, C.E. and J.H. Weare. The Prediction of Mineral Solubilities in Natural Waters: the Na-K-Mg-Ca-CI-SO4-H20 System from Zero to High Concentration at 25~ Geo.chim. Cosmochim. Acta 44, p. 981-987, 1980. Hayes, K.F. Equilibria, Spectroscopic and Kinetic Studies of Ion Adsorption at the Oxide Aqueous Interface. Ph.D. Dissertation, Stanford University, Palo Alto, CA, 1987. Helgeson, H.C. Thermodynamics of Hydrothermal Systems at Elevated Temperatures and Pressures. Am. J. Sci. 267, pp. 729-804, 1969. Hem, J.D. Equilibrium Chemistry of Iron in Groundwater. In Principles and Applications of Water Chemistry Edited by S.D. Faust and J.V. Hunter. John Wiley and Sons, New York, p. 625, 1967. Hering, J.G. and W. Stumm. Oxidative and Reductive Dissolution of Minerals. In Mineral-Water Interface Geochemistry Edited by M.F. Hochella, Jr. and A.F. White. Mineralogical Society of America, Washington, D.C., p. 427, 1990. Hsi, C.D. and D. Langmuir. Adsorption of Uranyl onto Ferric Oxyhydroxides: Application of the Surface Complexation Binding Site Models. Geochim. Cosmochim. Acta 49, pp. 1931-1941, 1985.

573 Kim, H.-T. and W.J. Frederick Jr. Evaluation of Pitzer Ion Internaction Parameters of Aqueous Electrolytes at 25~ Single Salt Parameters. J. Chem. EnQ. Data 33, pp. 177-184, 1988. Kluge, G., H. Saalfeld and W. Dannecker. Untersuchun,qen des LanQzeitverhaltens von M011verbrennun.qsschlacken beim Einsatz im Strassenbau. Unweltforschungsplan des Bundesministers des Innern, Forschungsbericht Nr. 103 03 006. Berlin, 1980. Krupka, K.M., R.L. Erikson, S.V. Mattigod, J.A. Schramke and C.E. Cowan. Thermochemical Data Used by the FASTCHEM Packa.qe. EPRI EA-5872, EPRI, Palo Alto, CA, 1988. Kummert, R. and W. Stumm. The Surface Complexation of Organic Acids on Hydrous u J. Colloid Interface Sci. 75, pp. 373-385, 1980. LaFlamme, B.D. and J.W. Murray. Solid/Solution Interaction: The Effect of Carbonate Alkalinity on Adsorbed Thorium. Geochim. Cosmochim. Acta 51, pp. 243-250, 1987. Leckie, J.O. Coordination Chemistry at the Solid/Solution Interface. In Metal Speciation: Theory, Analysis and Application Edited by J.R. Kramer and H.E. Allen. Lewis Publishing, Chelsea, MI, 41, 1988. Lindsay, W.J. Chemical Equilibria in Soils. J. Wiley and Sons, New York, 1979. Millero, F.J. and D.R. Schreiber. Use of Ion Pairing Model to Estimate Activity Coefficients of the Ionic Components of Natural Waters. Am. J. Sci. 282, pp. 15081540, 1982. Naumov, G.B., B.N. Ryzhenko and Khodakovsky. Handbook of Thermodynamic Data. US Geological Survey WRD-74-001. NTIS-PB-226 722/AS, Washington, D.C., 1974. Nordstrom, D.K. and J.L. Munoz. Geochemical Thermodynamics. Blackwell Scientific Publications, Palo Alto, CA, 1986. Nriagu, J.O. Lead Orthophosphates. IV. Formation and Stability in the Environment. Geochim. Cosmochim. Acta 38, pp. 887-898, 1974. Oberste-Padtberg, R. and K. Schweden. Zur Freisetzung von Wasserstoff aus M6rteln mit MVA-Reststoffen. Wasser Luft Boden 34, pp. 61-62, 1990. Pankow, J.F. Aquatic Chemistry Concepts. Lewis Publishers, Chelsea, MI,, 1991. Pankow, J.F. and J.J Morgan. Kinetics for the Aquatic Environment. I. Environ. Sci. Technol 15, pp. 1155-1164, 1981 a.

574 Pankow, J.F. and J.J. Morgan. Kinetics for the Aquatic Environment. II. Environ. Sci. Technol. 15, pp. 1306-1313, 1981b. Parker, V.B., D.D. Wagman and W.H. Evans. Selected Values of Chemical Thermodynamic Properties. Tables for the Alkaline Earth Elements (Elements 92 through 97 in the Standard Order of Arrangement). U.S. Nationa.I Bureau of Standards Technical Note 270-6, Gaithersburg, MD, 1971. Parks, G.A. Surface Energy and Adsorption at Mineral/Water Interfaces: an Introduction. In Mineral-Water Interface Geochemistry Edited by M.F. Hochella Jr. and A.F. White. Mineralogical Society of American, Washington, D.C., p. 133, 1990. Pitzer, K.S. Thermodynamics of Electrolytes. I. Theoretical Basis and General Equations. J. Phy..s. Chem. 77, pp. 268-277, 1973. Pitzer, K.S. Theory Ion Interaction Approach. In Activity Coefficients in Electrolyte Solutions. Edited by R. Pytkowicz. CRC Press, Boca Raton, FL, p. 157, 1979. Pitzer, K.S. Characteristics of Very Concentrated Aqueous Solutions. In Chemistry and Geochemistry of Solutions at Hi~..h Tempe.ratures and pressures. Edited by D.T. Rickard and F.E. Wickman. Pergamon, Oxford, p. 249, 1981. Pitzer, K.S. and L. Brewer. Thermodyn.amics, 2nd Edition, McGraw-Hill, New York, 1961. Pitzer, K.S. and J.J. Kim. Thermodynamics of Electrolytes. IV. Activity and Osmotic Coefficients for Mixed Electrolytes. J. Am. Chem. Soc_ 96, pp. 5701-5707, 1974. Pitzer, K.S. and G. Mayorga. Thermodynamics of Electrolytes. I1. Activity and Osmotic Coefficients for Strong Electrolytes with One or Both Ions Univalent. J.. Phys. Chem.. 77, pp. 2300-2308, 1973. Pourbaix, M. Atlas of Electrochemical Equilibria. Pergammon Press, Oxford, 1966. Pytkowicz, R.M. Equilibria, Nonequilibria, and Natural Waters, Vol. 1 and 2, John Wiley and Sons, New York, 1983. Rai, D. Inor.qanic and Organic Constituents in .Fossil Fuel Combustion Residues. Volume I A Critical Review. EPRI EA-5176, EPRI, Palo Alto, CA, 1987. Rai, D. and J.M. Zachara. Chemical Attenuation Rates, Coefficients and Constants in Leachate Mi,qration, Volume 1" A Critical Review.. EPRI EA-3356, EPRI, Palo Alto, CA, 1984.

575 Robie, R.A., B.S. Hemingway and J.R. Fisher. Thermodynamic Properties of Minerals and Related Substances at 298.1K and 1' Bar Pressure and at Hi.qher Temperatures. Geological Survey Bulletin No. 1452, U.S. Government Printing Office, Washington, D.C., 1978. Robie, R.A., B.S. Hemingway and J.R. Fisher. Thermodynamic Properties of Minerals and Related Substances at 298.15~ and 1 bar Pressure and at Higher Temperatures. Geolo.(]ical Survey Bulletin 1452. U.S. Government Printing Office, Washington, D.C., 1979. Rossotti, F. The Determination of Stability Constants.. McGraw-Hill Co., Inc., New York, 1981. Roy, W.R. and R.A. Griffin. Illinois Basin Coal Fly Ashes. 2. Equilibria Relationships and Qualitative Modelling of Ash-Water Reactions. Environ. Sci. Technol. 18, pp. 739742, 1984. Rubin, J. Transport of Reacting Solutes in Porous Media Relations Between Mathematical Nature of Problem Formulation and Chemical Nature of Reactions. Water Resources Res. 19, pp. 1231-1252, 1983. Scatchard, G. The Excess Free Energy and Related Properties of Solutions Containing Electrolytes. J. Am. Chem. Soc 90, pp. 3124-3127, 1968. Schindler, P.W. and W. Stumm. The Surface Chemistry of Oxides, Hydroxides and Oxide Minerals. In Aquatic Surface Chemistry. Edited by W. Stumm. John Wiley and Sons, New York, p. 83, 1987. Schnoor, J.L. Kinetics of Chemical Weathering: A Comparison of Laboratory and Field Weathering Rates. In Aquatic Chemical Kinetics" Reaction Rates of Processes in Natural Waters Edited by W. Stumm. John Wiley and Sons, New York, p. 475, 1990. Schnoor, J.L. and W. Stumm. The Role of Chemical Weathering in the Neutralization of Acidic Deposition. Schweiz Z. HYdrol. 48, pp. 171-193, 1986. Schott, J. and J.-C. Petit. New Evidence for the Mechanisms of Dissolution of Silicate Minerals. In Aquatic Surface Chemistry Edited by W. Stumm. John Wiley and Sons, New York, p. 293, 1987. Schott, J., R.A. Berner and E.L. Sj0berg. Mechanism of Pyroxene and Amphibole Weathering. I. Experimental Studies of Iron-Free Minerals. Geochmi. Cosmoch.im. Acta 45, pp. 2123-2135, 1981.

576 Schumm, R.H., D.D. Wagman, S.M. Bailey, W.H. Evans and V.B. Parker. Selected Values of Chemical Thermodynamic Properties. Tables for the Lanthanide (Rare Earth) Elements (Elements 62 through 76) in the Standard Order of Arrangement. U.S. National Bureau of Standards Technical Note 270-7. U.S. Government Printing Office, Washington, D.C., 1973. Schwertmann, U. and R.M. Taylor. Iron oxides. In Minerals in Soil Environments Edited by J.B. Dixon and S.B. Weed. Soil Science Society of America, Madison, WI, p. 145, 1977. Sigg, L. Surface Chemical Aspects of the Distribution and Fate of Metal Ions in Lakes. In Aquatic Surface Chemistry Edited by W. Stumm. John Wiley and Sons, New York, p. 319, 1987. Sigg, L. and W. Stumm. The Interactions of Anions and Weak Acids with the Hydrous Goethite (a-FeOOH) surface. Colloids Surf. 2, pp. 101-117, 1981. Sill~n, L.G. and A.E. Martell. Stability Constants of Metal-Ion Complexes. Supplement No. 1, Chemical Society, London, 1971. Smith, R.M. and A.E. Martell..Critical Stability Constants. Plenum Press, New York, 1976. Sposito, G. The Thermodynamics of Soil Solutions, Clarendon Press, Oxford, 1981. Sposito, G. Chemical Models of Inorganic Pollutants in Soils. CRC Crit. Rev. Environ. Control 15, pp. 1-24, 1984. Sposito, G..The Chemistry of Soils, Oxford University Press, N.Y., 1989. St~mpfli, D. Final Report: Cement and Bottom Ash Chemistry (CABAC). ERG Report. UNH, Durham, NH, 1992. St~mpfli, D., H. Belevi, R. Fontanive, and P. Baccini. Reactions of Bottom Ashes from .MSW Incinerators and Construction.Waste Samples with Water. EAWAG Project 3335, EAWAG, Dubendorf, Switzerland, 1990. Stumm, W. and G. Furrer. The Dissolution of Oxides and Aluminum Silicates; Examples of Surface Coordination-Controlled Kinetics. In Aquatic Surface Chemistry, Edited by W. Stumm. John Wiley and Sons, New York, p. 197, 1987. Stumm, W. and J.J. Morgan. Aqu.atic Chemistry.. John Wiley and Sons, New York, 1981.

577 Stumm, W. and E. Weiland. Dissolution of Oxide and Silicate Minerals: Rates Depend on Surface Speciation. In Aquatic Chemical KineticsLReaction Rates of Processes in. Natural Waters Edited by W. Stumm. John Wiley and Sons, New York, p. 367, 1990. van der Wegen, G. Orienterend Onderzoe.k. Naar O.orzaak Binding in een Monster AVISlakken. Rapportnumme, 91149, Intron, the Netherlands, 1991. Verink, E.D. Simplified Procedure for Constructing Pourbaix Diagrams. J. Educational Modules Materials Sci. EnQineer 1, pp. 535-560, 1979. Wagman, D.D., W.H. Evans, V.B. Parker, I. Halow, S.M. Bailey and R.H. Schumm. Selected Values of Chemical Thermodynamic Properties. Tables for the First ThirtyFour Elements in the Standard Order of Arrangement. _U.S. National Bureau of Standards Technical Note 270-3 U.S. Government Printing Office, Washington, D.C., 1968. Wagman, D.D., W.H. Evans, V.B. Parker, I. Halow, S.M. Bailey and R.H. Schumm Selected Values of Chemical Thermodynamic Properties. Tables for Elements 35 through 53 in the Standard Order of Arrangement. U.._S.National Bureau of Standards. Technical Note 270-4 U.S. Government Printing Office, Washington, D.C., 1969. Wagman, D.D., W.H. Evans, V.B. Parker, I. Halow, S.M. Bailey, R.Ho Schumm and K.L. Chumey. Selected Values of Chemical Thermodynamic Properties. Tables for Elements 54 through 61 in the Standard Order of Arrangement. U~S. National Bureau of. Standards Technical Note 270-5 U.S. Government Printing Office, Washington, D.C., 1971. Wagman, D.D., W.H. Evans, V.B. Parker, R.H. Schumm and R.L. Nuttall. Selected Values of Chemical Thermodynamic Properties. Compounds of Uranium, Protactinium, Thorium, Actinium, and the Alkali metals. U..S. N_ati0nal Bureau of Standards Technical. Note 270-8 U.S. Government Printing Office, Washington, D.C., 1981. Wagman, D.D., W.H. Evans, V.B. Parker, R.H. Schumm, I. Halow, S.M. Bailey, K.L. Churney and R.L. Nuttall. The NBS tables of Chemical Thermodynamic Properties. Selected Values for Inorganic and C1 and C2 Organic Substances in SI units. J. Phy. Chem. Ref. Data 11 (Supplement No. 2), pp. 1-392, 1982. Warren, C.J. and M.J. Dudas. Weathering Processes in Relation to Leachate Properties of Alkaline Fly Ash. J. Environ. Qual. 13, pp. 530-538, 1984. Warren, C.J. and M.J. Dudas. Formation of Secondary Minerals in Artificially Weathered Fly Ash. J. EnYiron. Qua!. 14, pp. 405-410, 1985. Warren, C.J. and M.J. Dudas. Mobilization and At..tenuati0n of Trac.e Elements in Artificially Weathered Fly .Ash. EPRI EA-4747, EPRI, Palo Alto, CA, 1986.

578 Wieland, E., B. Wehrli and W. Stumm. The Coordination Chemistry of Weathering: III a Generatlization on the Dissolution Rates of Minerals. Geochem. Cosmochim. Acta 52:1969-1981 Westall, J.C. and H. Hohl. A Comparison of Electrostatic Models for the Oxide/Solution Interface. Adv. Coll. Inter. Sci. 12, pp. 265-294, 1980. Whitfield, M. An Improved Specific Interaction Model for Seawater at 25~ Atmosphere Total Pressure. _Marine Chem. 3, pp. 197-213, 1975a.

and 1

Whitefield, M. The Extension of Chemical Models for Sea Water to Include Trace Components at 25~ and 1 Atmosphere Pressure. Geochim. Cosmochim. Acta 39, pp. 1545-1557, 1975b. Wollast, R. and L. Chou. Kinetic Study of the Dissolution of Albite with Continuous Flow-Through Fluidized Bed Reactor. In T.he Chemis..try of Weatherin.Q, Edited by J.l. Drever, NATO AI Series C 149, pp. 75-96, 1985. Zachara, J.M., D.C. Girvin, R.C. Schmidt and C.T. Resch. Chromate Adsorption on Amorphous Iron Oxyhydroxide in the Presence of Major Groundwater Ions. Environ. Sci. Technol. 21, pp. 589-594, 1987. Zachara, J.M. and G.P. Streile. Use of Batch and Cplumn..Methodolo.qies to Assess Utility Waste Leachin.q and Subsurface Chemical Attenuation. EPRI EN-7313, EPRI, Palo Alto, CA, 1991. Zevenbergen, C., J.P. Bradley, T. van der Wood, R.S. Brown, L.P. van Reeuwijk, and R.D. Schuiling. Weathering as a Process to Control the Release of Toxic Constituents from MSW Bottom Ash. In Geoconfine 93 Edited by M. Arnoud, M. Borres and B. C6me. A.A. Balkema, Rotterdam, p. 591, 1993.

579

CHAPTER 14- LEACHING TESTS 14.0 LEACHING TESTS Now that the concepts behind leaching phenomena have been introduced, discussing leaching tests is appropriate. Invariably, these tests are involved in the regulation of residues, as well as in the interpretation of leaching phenomena. Careful consideration should be given to the specific "tools" that are selected to characterise ash. Clearly, it is preferable to use a number of tools, rather than a single tool, for work in both science and regulation. At the end of this section, a unified theory of leaching is presented. This will move away from the strict use of concentration data and toward normalisation of leaching data to release, fractions leached and fluxes from residues. The rationale for using this approach to develop models and management scenarios will be discussed. Generalised and detailed reviews of leaching tests are found in Jackson et al. (1984), F,~llman (1990), Environment Canada (1990), Zachara and Streile (1991) and van der Sloot et al. (1991; 1993). The information from Environment Canada (1990) provides much of the basis for the following discussion. 14.1 PURPOSE OF LEACHING TESTS In general, a leaching test involves contacting a solid material with a leachant to determine which components in the solid will dissolve in the leachant and create a leaching solution or leachate. To investigate the various processes governing the extent and rate of leaching, endless variations can be introduced by changing test variables, such as leachant composition, method of contact, liquid-to-solid (L/S) ratio, contact time and system control (pH, pE (or Eh), temperature). Leaching tests have a wide range of objectives, the most common of which are presented in Table 14.1. Leaching tests are typically used to provide information about the constituent concentration or the constituent release from a waste material under reference test conditions, or under conditions that more closely approximate the actual disposal site. This information may subsequently be used in mathematical models to predict long term leaching.

14.1.1 Classification of Leaching Tests For the purposes of this discussion, leaching tests have been separated into two broad categories on the basis of whether or not the leachant is renewed: 1) extraction tests (no leachant renewal), and 2) dynamic tests (leachant renewal).

580 Table 14.1 .Leaching Test Objectives Objective

Description

Identification of leachable constituents

Determine which constituents of a waste are subject to dissolution upon contact with a liquid.

Classification of hazardous wastes

Compare wastes against performance criteria for classification of wastes as hazardous or nonhazardous.

Evaluation of process modifications

Determine if modifications to a wastegenerating process result in a less leachable waste.

Comparison of waste treatment methods

Determine whether a given waste treatment method/process results in superior containment of contaminants.

Quality control in waste treatment

Verify the efficiency of a treatment process using a simple pass/fail criterion.

Design of leachate treatment systems

Obtain a typical leachate to perform treatability experiments.

Field concentration estimates

Express leaching over time (e.g. to be used as a source term in groundwater modelling).

Parameter quantification for modelling

Quantify partition coefficients and kinetic parameters to be used in transport modelling.

Risk assessment

Estimate potential impact of waste disposal on the environment.

The concept of leachant renewal is based on modifying the leaching system to promote solution control of leaching rather than solid phase control.

Extraction Tests Extraction tests include all tests in which a specific quantity of leachant is contacted with a specific quantity of waste for a certain length of time, without leachant renewal. (This definition does not include analytical extractions or digestion procedures which are used to measure the total constituent concentration in an ash sample). The leachate is separated from the solid and analysed either at various times during the

581 test, or, as in most extraction tests, at the end of the test. The analysis of leachates generated at various times can help determine the kinetics of the leaching process or if equilibrium has been attained. The underlying assumption in this type of test is that an equilibrium condition is achieved by the end of the extraction test (i.e. the concentrations of solutes in the leachate become constant). In this no-flow system, an equilibrium condition occurs when there is no net transfer of components from the solid phase to the leaching solution, or vice versa. Sampling in an extraction test over time to derive kinetic information or to monitor the attainment of equilibrium is difficult since it must be done without modifying the residueleachant interactions, which are a function of factors such as the L/S ratio and gaseous exchanges. This can be accomplished in three ways: nondestructive sampling and analysis of parameters such as pH, conductivity or specific ions removing small volumes (aliquots) that are negligible when compared with the total volume preparing as many parallel extraction tests as data points required and performing destructive analyses. Extraction tests can be further divided into four subcategories: agitated extraction tests non-agitated extraction tests sequential chemical extraction tests concentration buildup tests

Agitated Extraction Tests

Agitated extraction tests (Figure 14.1) are performed to reach steady-state conditions as quickly as possible. They measure the chemical properties of a waste-leachant system, as opposed to rate-limiting mass transfer mechanisms. Agitation ensures a homogeneous mixture, promotes contact between the solid and the leachant and reduces boundary layer thicknesses. Sample particle size reduction is often performed to increase the surface area to volume ratio of the solid to enhance liquid/solid phase contact and to eliminate mass-transfer limitations. Generally, this reduces the duration of the test by reducing the time required to reach a pseudo-equilibrium condition in the leachate. This procedure may also have the effect of overestimating the short-term release of constituents. A steady-state leaching environment can also be attained in a column apparatus by recirculating the collected leachate back into the column.

582 Figure 14.1 Agitation Extraction Test Crushed Solid Waste

Monolithic Solid Waste

o00 (:]DO ~

C]

OOO E]

Agitation

After Environment Canada, 1990

Non-agitated Extraction Tests A non-agitated extraction test is performed to study the physical mechanisms that are rate-limiting in leaching. The underlying assumption behind a non-agitated extraction test is that the physical integrity of the solid matrix and mass transfer constraints (both internally within the sample and externally in the boundary layer) affect the amount of contaminants that are leached during the test. Two types of non-agitated tests are illustrated in Figure 14.2. They can be performed on large particle-sized residue samples, concrete-type or monolithic samples. The disadvantage of running a non-agitated test is that a much longer contact period may be required to reach equilibrium conditions than is required in an agitated test. The advantage of this type of test is that rate-limiting mechanisms of leaching due to the physical integrity of the solid matrix are taken into account. These tests are presented in further detail in Chapter 20.

583 Figure 14.2 Static Leach Test

A) Static test with monolithic solid waste

A) Statictest with nonmonoUthic solid waste

After Environment Canada, 1990

Sequential Chemical Extraction Tests A sequential chemical extraction test is composed of a battery of non-agitated extraction tests (Figure 14.3). It involves performing sequential elutions of aliquots of a sample with different leachants (i.e. A, B, C, D and E in Figure 14.3), which are increasingly more aggressive in terms of chemical attack toward the residue (Figure 14.3a). One type of method assumes that each successive leachant also extracts the sum of contaminants extracted by all preceding leachants. The other type of method is conducted by subjecting the same aliquot of sample to each leachant (Figure 14.3b). The amount extracted in each elution is associated with a certain chemical form or mineral phase in the solid phase. The Sequential Chemical Extraction Procedure, originally compiled by Tessier et al. (1979), was adapted to sewage sludge incinerator ash by Fraser and Lure (1983), and then further modified for MSW incinerator ash (WTC, 1990). The test has been used in different studies (Wadge and Hutton, 1987) (Tessier Method); Environment Canada, 1993 (modified)), however, results by Khebohian and Bauer (1987) and discussion by Nirel and Morel (1990) (on the Tessier method) have shown that resorption and reprecipitation reactions can dramatically alter the mass fractions that are obtained in the different extractions. This limitation has been recognised and the latest studies using the modified method have basedmuch of their interpretation on the operationally defined extractions (e.g., peroxide extractable) rather than the implied chemical species (Environment Canada, 1993). Consequently, it appears that although the method is not appropriate for determining the chemical species, relating the operationally defined extractions to exposures under different leaching conditions (e.g., fraction available for leaching under acidic leaching conditions versus severe reducing conditions) is an appropriate set of interpretations.

584 Figure 14.3 Sequential Chemical Extraction Tests

Leachant a)

A

C

D

E

C

D

E

Wffh different waste samples

Leachant A b)

B

B

With the same waste sample and liquid/solid separation between elutions

After Environment Canada, 1990 Concentration Buildup Tests

In a concentration buildup test, an extraction is achieved at a very low cumulative L/S ratio. Aliquots of samples are successively contacted with the same leachant (Figure 14.4). The contact of leachate with fresh solid material can be considered as a model for an elemental volume of water flowing through a large body of residue and approaching saturation with respect to specific mineral phases. The purpose of this test is not to collect kinetic information, but to characterise a leachate saturated with soluble residue constituents. In some cases, this may simulate the actual pore water composition of a granular material in column leach tests or in outdoor disposal or utilisation scenarios. Dynamic Tests

Dynamic tests include all tests in which the leachant is continuously or intermittently renewed to maintain a driving force for leaching that is solution-controlled. The intermittent tests may be conducted by alternating leaching periods with dry periods to study the effects of desiccation or unsaturated flow conditions. Dynamic tests provide information about the kinetics of solid phase dissolution and contaminant flux. Information is generated as a function of time, and attempts are often made to preserve the residue's physical integrity. These two factors lend this category of leaching tests to the investigation of more complex mechanisms of leaching.

585

Dynamic tests can be further divided into subcategories according to how the interface between the waste and the leachant is defined. Tests in which individual waste particles are used to define the interface are called serial batch tests. The tests in which a characteristic dimension of the waste, (such as the external geometric surface area or the geometric surface area perpendicular to flow) is used to define the interface include flow-around tests and flow-through or column tests. Figure 14.4 Concentration Buildup Tests

1

2

N Discard

II

Agitation

After Environment Canada, 1990

Serial Batch Tests A serial batch test is conducted using a granular or crushed sample which is mixed with leachant at a given US ratio for a specified period of time (Figure 14.5). The leachate is then separated from the solids and replaced with fresh leachant until the desired number of leaching periods have been completed. The waste/leachant mixture is normally agitated to promote contact. Kinetic information regarding contaminant dissolution is obtained using the concentrations measured in the leachate from each of the leaching periods. Data from serial batch tests can be used to construct an extraction profile to infer the temporal release of leachable constituents. Flow-Around Tests In flow-around tests, a sample of residue is placed in the leaching vessel and the flow of fresh leachant around the residue provides the driving force to maintain leaching. The L/S ratio is modified to express the volume of leachant divided by the surface area

586 of the solid sample. Samples are usually monolithic, although non-monolithic or crushed residue may be used if it is confined in some manner. Agitation is generally not performed. Leachant flow is either continuous (Figure 14.6a), in which case it is sampled and analysed periodically, or it is intermittently renewed (Figure 14.6b). The latter method is generally simpler from an experimental point of view, but the renewal frequency must be sufficient to prevent a buildup of contaminants at the residue/leachant interface, which may inhibit further leaching by reducing the diffusional gradient. Figure 14.5 Serial Batch Tests

m

m

m

AgitaUon After Environment Canada, 1990

Flow-Through Tests In a flow-through or column test, an open container is packed with a porous solid and leachant is passed through, either continuously or intermittently. The effluent is sampled periodically and analysed for the parameters of interest. The results are used to examine contaminant removal in which the primary transport mechanism is advection. There are two basic types of flow-through tests characterised primarily by the shape and size of the container. The first type is a column test which is performed using a small cylindrical container (Figure 14.7a). The second is a lysimeter test which is conducted in a large rectangular or cylindrical container (Figure 14.7b). In general, the size of the sample used in a flow-through test tends to be large to minimise the effects of sample heterogeneity and wall channelling effects. The depth of waste in either type of test varies according to the individual experiment.

587 Figure 14.6 Flow-Around Tests leachant

@

leachant

a) continuous leachate renewal

9

9

9

b) intermittent leachate renewal

After Environment Canada, 1990 Figure 14.7 Flow-Through Tests leachant (downflow)

y_,

Y

leachant

leachant (upflow)

a) Columns After Environment Canada, 1990

I leachate b) Lysimeter

588 Columns may be operated either in an upflow or downflow mode, whereas lysimeters are always operated in a downflow mode. Flow through the solid depends upon its hydraulic conductivity, as well as the hydraulic gradient, and varies with the individual test. Mini-columns may be used to achieve a relatively rapid breakthrough of leached species. Since head losses may be large and a rapid breakthrough is desired, the leachant is usually delivered under pressure and at a constant flow rate. The advantages of minicolumns include: L/S ratios that are similar to those of real waste-leachant systems a known and easily varied average fluid velocity negligible axial dispersion or spreading of the solute a simple estimation of both equilibrium and kinetic coefficients automation permitting the rapid output of data. These tests are not applicable when large volumes of leachate are needed for a variety of analytical tests. Care should also be taken when conducting flow-through tests to avoid unnatural channelling of water and clogging by fine material or biological growth. In lysimeter tests, channelling cannot be avoided. It is a factor that occurs in the field, and its influence should be modelled in the laboratory, although quantifying it is difficult. Biodegradation of organics can also be a problem in columns, although in some cases experiments are intentionally set up to measure the effects of biological activity. Flowthrough tests can also be modified to examine other site-specific influences, such as vegetation on the surface of the container, or layered media, such as ash and geological material.

14.1.2 Leaching Test Variables

There are several experimental variables which are common to all extraction and dynamic tests. These variables need to be considered when designing a leaching test for specific purposes.

Sample Preparation

Depending on the nature of the waste and the test to be performed, the sample may require one of the following preparatory steps: 9 9 9 9 9 9 9 9

liquid/solid separation sub-sampling particle-size reduction surface washing compaction preservation curing aging

589 Liquid/solid separation may be performed on residues containing a free liquid phase. The leaching test is conducted only on the solid portion of the sample. The free liquid phase constitutes the initial leachate, which may be analysed separately to estimate the pore-water concentration or it can be included with the final leachate for analysis. Liquid/solid separation can be accomplished by various methods, including settling and decanting, centrifugation or pressure filtration through filter media of various types. Sub-sampling is generally required when several different tests or replicates are to be performed on the same sample. Waste samples should be thoroughly mixed before sub-sampling is performed. (See Chapter 6). Particle-size reduction is required for most extraction tests. The goal is to reduce the time required to reach steady-state conditions by increasing the contact surface area between the solid and the leachant. However, care should be taken to prevent the loss of volatile compounds in the solid if they are of interest. Particle-size reduction is usually carried out by grinding (e.g. mortar and pestle, centrifugal grinder or hammer mill). These issues are also discussed in more detail in Chapter 6 and 7. Surface washing may be performed prior to testing small monolithic samples in flowaround tests. The surface is washed to remove small detachable particles and readily soluble salts by quickly dipping the sample in an aqueous solution. Compaction or remolding is often required for flow-through tests. Reproducibility and field simulation considerations require that samples be compacted to a pre-specified density using methods such as vibration, proctor compaction or modified proctor compaction. Sample preservation is performed to avoid biological activity. This is a greater problem in tests of long duration, such as column tests. Various chemical treatments are available, such as the addition of sodium azide, however, none offer complete efficacy. Curing may be performed on samples that have been transformed into a solidified mass using various chemical additives, such as Portland cement. It allows the waste sample to gain physical and engineering properties, i.e. high unconfined compressive strength and low permeability, that are considered to be important in reducing leachability. Curing can be used to achieve a variety of chemical reactions within the waste, although this term usually refers to cement hydration reactions. Aging may be promoted on any type of waste sample to account for the physical, chemical and biological alterations that a waste might undergo in the field. Chapter 13 discusses classes of aging reactions that can occur in residues.

Leachant Composition

The release of contaminants from a waste in any leaching test may be strongly influenced by the initial leachant composition, especially at high L/S ratios, or with the

590 use of an aggressive solution. Chemical properties of the leachant that influence contaminant mobilisation are indicated in Table 14.2. Examples of three types of commonly used leachants, i.e. water, site liquid and chemical solution, are identified in Table 14.3. Several advantages and disadvantages of these leachants are outlined in Table 14.4. Table 14.2 Important Factors in Leachant Composition Factor Release Mechanism Affected Dissolution/precipitation of metals, speciation of inorganic species Adsorption/desorption of solutes Oxidation/reduction of inorganic species Ionic exchange of metals, speciation chemistry and solubility products

pH Eh, redox potential Ionic strength Chelating and complexing agents Buffering capacity After Environment Canada, 1990

Metal solubility All above properties

Table 14.3 Commonly Used Leachants Type of Leachant Water

Common Uses

Nonaggressive, baseline medium without buffering capacity Site liquid (real or Simulates site-specific synthetic) leaching conditions Chemical solution Examines metal speciation and organic compound binding After Environment Canada, 1990

Examples Distilled, deionised and tap water Rainwater, groundwater, surface water, landfill leachant, seawater Strong chemical solution (acidic, basic, reducing, oxidising, complexing, solvent,etc.)

591 Table 14.4 Advantages.and D isadvantacjes of Commonly Used Leachants Leachant

Advantages

Disadvantages

Pure water

Reliable, simple standard

Lack of background composition may result in dissolution of common ions

Site liquid

Best field case model

Requires characterisation (to obtain leaching results by subtraction)

Several synthetic liquids available

Results not comparable with other leaching studies Labour intensive (sampling and preservation)

Chemical solution Allows for the study of waste chemistry After Environment Canada, 1990

Aggressive, difficult to relate data to field conditions

Method of Contact Since a leaching test is primarily a system to study the transfer of contaminants from a residue to a liquid, it is important to consider the aspects of the test conditions that promote mass transfer, such as agitation, and to consider the effect of mass exchange with other components of the system, primarily the leaching vessel and the atmosphere.

Agitation of the leachant-solid slurry generally hastens reaching equilibrium conditions by maintaining maximum contaminant concentration gradients at the leachant-solid particle interface. Different methods can be used to agitate the waste, including shaking (wrist action or reciprocation) stirring (magnetic or paddle) tumbling gas bubbling. In static or non-agitated tests, the leachant-solid interface is usually the geometrical surface area of the solid form. There is usually no provision for mixing because diffusion of leached constituents within the leachate is assumed to be much faster than the rate of release by mechanisms such as dissolution from the surface or diffusion from within ash particles. Ensuring that the leachate is well mixed before sampling is important, however.

592 It may be important to identify and quantify exchanges of chemical species other than between the solid and the leachant. Exchanges between the leachant and the leaching vessel are always undesirable, whereas exchanges with the atmosphere depend in large part upon the objectives of the test, such as leaching with carbonic acid. To minimise exchanges with the leaching vessel, glass or stainless steel should be used for organic contaminants and plastic for inorganic contaminants. If the cost is not prohibitive, polytetrafluoroethylene is considered to be acceptable for both. For the purpose of verifying the mass of constituents adsorbed to the container wall, the emptied leaching vessel can be extracted with a strong solvent. The test system may be either open or closed to the atmosphere. The choice depends on the specific leaching problem being examined. For example, a closed system provides a better simulation of the saturated groundwater environment, whereas an open system models problems like a storage pile and unsaturated disposal environments more accurately. An open system facilitates sampling, leachant renewal and periodic or continuous adjustment of the pH or redox potential. However, a system that is open to the atmosphere allows for the loss of volatile compounds, including water and organics, and the introduction of CO2 and 02 from the air. Losses due to evaporation may have to be accounted for in an open system. Although volatile organics are generally not a concern with incinerator ash, there are several apparatus configurations that will prevent volatile contaminants from escaping. If there is no headspace in the leaching vessel, volatiles will remain in either the solid phase or the leachate. If there is a headspace, volatiles will be partitioned in the gas phase. Analysis of the headspace allows for an evaluation of this loss. Even for experiments carried out in closed containers and under controlled conditions, penetration of gases through plastic container walls can have a significant effect, especially over long durations. This is seen in reaction vessels kept at a low redox potential when oxygen diffuses into the vessel.

Liquid-to-Solid Ratio

The L/S ratio is the ratio of the amount of leachant in contact with the residue to the amount of waste being leached. Although this definition appears straightforward at first glance, it can become confusing because of the many ways in which the two variables in the ratio have been defined. The L/S ratio has been expressed as: volume of leachant/mass of solid mass of leachant/mass of solid and volume of leachant/surface area of solid (for monolithic material).

593 Furthermore, when using the first two expressions, the mass of solid being leached can be calculated on a wet weight or a dry weight basis. Another problem arises because of the various ways that the volume or mass of leachant can be calculated, depending on whether or not the liquid phase of the solid is included in the total leachant volume. The preferred way to report L/S is the mass of leachant to the dry mass of solid. Figure 14.8 illustrates how these various ways of defining the amounts of waste and leachant can give different L/S ratios for the same system. The three fractions shown in Figure 14.8 include the amount of leachant added, the liquid phase associated with the solid, and the solid phase. Figure 14.8 Liquid and Solid Fractions of the Waste Leachant System ,,,

A

Added Leachant

B Liquid Phase of Waste

C

Solid Phase of Waste

After Environment Canada, 1990 If the residue is dry, then the L/S ratio is simply A/C. If the residue is wet, there are three ways to define the L/S ratio: 1) the residue is the sum of the liquid and solid phases, i.e. L/S = N(B+C). 2) the leachant is the sum of the amount of leachant added plus the liquid phase associated with the solid, i.e. L/S = (A+B)/C 3) the liquid phase associated with the solid is excluded from the calculations, and residue is the solid phase only, i.e. L/S = A/C.

594 Leachate concentrations of highly soluble species (e.g. sodium, potassium) are generally inversely proportional to the L/S ratio of all of the species which have been removed from the solid. However, if the release of a species is limited by solubility, the final concentration is independent of the L/S ratio and simply equals the maximum solubility concentration. In general, the leachate concentration will be controlled by a number of competing factors, namely, the amount of contaminant available, solubility and kinetically controlled chemical reactions. Thus, the relationship between the L/S ratio and concentration is complex, and different for each species of interest. Selection of an appropriate US ratio depends on the objectives of the leaching test, the solubility of species of interest and analytical constraints. The ratio should be low enough to avoid dilution of contaminants to less than analytical detection limits. However, the ratio also must be high enough to prevent solubility constraints from limiting the amount of contaminants that can be leached from the waste. The selected ratio should be somewhere between these two limitations. Practical values for the L/S ratio range from 0.1 to 100:1. To place these values into perspective, most landfilled residues are exposed to L/S values of less than 3:1 during the operational life (10 to 20 years) of a disposal facility. After closure, the type and integrity of the cap will influence further increases in the L/S.

Contact Time

The total amount of time that a leachant is in contact with a solid sample before the attainment of equilibrium will influence the amount of contaminant released. In extraction tests, the contact time is equivalent to the duration of the test, whereas in dynamic tests, it is a function of the flow rate, or the number of elutions, in addition to the test duration. The contact time for extraction tests should allow equilibrium conditions to be reached for the contaminants of interest. This is generally in the order of hours to days for samples that have undergone particle-size reduction. For concrete-based or monolithic samples, it can be in the order of weeks to months. The contact time for dynamic tests should be sufficient to allow for observation of the processes of interest. Diffusion processes may be quantified within a few weeks, although several months may be required to study slow chemical reactions.

Temperature

Temperature affects the results of extraction and dynamic tests. Both the van't Hoff relationship, which applies to thermodynamic equilibrium constants and solubility products, and the Arrhenius relationship, which applies to kinetic processes such as adsorption and diffusion, indicate that properties or mechanisms relevant to leaching vary exponentially with temperature.

595 For convenience, most leaching tests are performed at room temperature. Higher temperatures may be used to accelerate the rate of leaching (although this may also change the properties of the waste) or to simulate the effects of biological activity in a landfill or the self-heating from exothermic reactions.

Leachate Separation

Leachates are commonly separated from agitated non-monolithic wastes by filtration using a 0.45 pm membrane filter (a convention used to define soluble species). However, very small colloid particles can pass through a 0.45 pm filter. A smaller pore size filter (0.2 pm) should be used if these particles are to be removed. The use of the smaller filter size should be reported with the data. Glass fibre filters are chosen when hydrophobic, low solubility organic molecules are expected in the leachate since they may have a high affinity for filters composed of an organic polymer. Membrane filters, such as cellulose acetate, should be used for metal species in place of glass. The same care used to select a leaching vessel should be applied when selecting the filter material. Filtering the leachate from non-agitated monolithic samples may not be necessary if the method of contact generates only dissolved species. This should be verified before sampling.

14.1.3 Compilation of Leaching Tests The leaching tests presented in Environment Canada (1990), Fallman (1990), and van der Sloot et al. (1991, 1993) serve as the basis for the compilation of various leaching tests presented here. The reader can refer to these references for precise details of each method. Table 14.5 summarises the various agitated batch extraction tests that are used for regulatory purposes or for research into leaching characteristics of waste. All the methods specify the type of leaching vessel to be used, the type of sample preparation that is required, the amount of sample that is needed, the type of leachant to be employed, the L/S ratio that is used, the type of agitation that is required and the duration of the test. Most methods also specify the type of filtration that is to be employed to allow for quantification of total dissolved constituents in the leachate. Table 14.6 specifies two non-agitated extraction tests that are commonly used to examine sequential dissolution of mineral phases in a solid or the fundamental dissolution and effective diffusion parameters of a solid dissolving under static conditions.

Table 14.5 Agitation Leach Tests Test Name and Proponent

Status of Leaching Development Vessel

Sample Preparation

Sample Mass

Leachant

US Ratio

-EP f o x U.S. EPA Method 1310

Standard regulatory method (1980)

Unspecified

Non-monolithic waste; phase separation Monolithic waste; particle-size reduction

100 g

Deionised water 0.5 N acetic acid (max. 2.0 meq H+lg solid)

20:l

-LEP MOE (Ontario)

Standard regulatory method (1985)

Wide mouth, 1250 mL cylindrical bottle

Phase separation by 0.45 pm membrane filter

50 g of d?, sol~ds

Distilled water Acetic acid (2.0 meq H+lg dry solids)

20.1

End over end (10 rPm)

24 hours

0.45 pm filtration

-TCLP U.S. EPA Method 1311

Standard regulatory method (1986)

Any material compatible with waste, zero head-space extractor (ZHE) for volatiles

Cuttinglcrushing and grinding Solidlliquid phase separation No structural integrity

100 g (25 g for ZHE)

Buffered acetic acid 1) pH = 4.93 2) pH = 2.88

20:l

End over end (30 rpm)

18 hours

0.6 to 0.8 pm borosilicate glass fibre filter combines liquid phase with extract

-Q.R.S.Q. MOE (Quebec)

Standard regulatory method (1987)

>1 L bottle

No phase separation 100 g dry Inorganics: buffered Grinding solids acetic acid (0.82 No structural integrity 50 g for meq H+lg dry volatiles solids) Organics: distilled water

10:l

End over end (10 to 20 rPm)

24 hours

30 min decantation, 0.45 pm filtration

-WET California

Standard regulatory method (1985)

Polyethylene or glass container

Milling, 0.45 pm filtration

50 g

0.2 M sodium citrate at pH 5.0

10:l

Table shaker Rotary Extractor

48 hours

Centrifugation 0.45 pm filtration

-X31-210 French Leach Test AFNOR (France)

Proposed standard for waste (Dec 1992)

Straight wall, 2 L bottle

Remove free liquid Reduce particle size to <4 mm

100 g

Demineralised water

10:l

Roller or shaker

16 hours

0.45 pm filtration

Agitation

Duration

Unspecified, 24 to 28 continuous hours

Leachate Separation 0.45 pm filtration

Test Name and Proponent

Status of Leaching Development Vessel

Published -EE Environment research method Canada

-ASTM D3987 ASTM

Standard research method

Published -MBLP Environment research Canada method

Sample Preparation

Inorganic: wide mouth, plastic sample bottle (250 mL); Organic: glass (500 mL)

Grinding (inorganic) Mortar and pestle (organic)

Round, wide mouth bottle

As received

Square, polyethylene or glass bottle, 1to2L

Remove free liquid Reduce particle size to <9.5 mm

Sample Mass

Leachant

lnorganics Distilled water : 408 Organ~cs: 80 g 700 g

Distilled water (ASTM Type IV)

Variable 1) Distilled water to fill 90% 2) acidic water of bottle buffer to pH 4.5 3) synthetic leaching media

US Ratio

Leachate Separation

Agitation

Duration

4: 1

NlST rotary extractor

7 days

0.45 pm vacuum screen

4: 1

Shaking

48 hours

0.45 pm filtration

4:l or 2: 1

Slow rotary tumbling

24 hours

0.45 pm filtration

-MCC-35 Materials Characterisat ion Center, England

Standard regulatory method (radioactive wastes)

Teflon container, 20mLto1 L

Crush waste form into two fractions: 74 to 149 pm 180 to 425 pm

>I g

Choice of high purity water, silicate water, brine, repository water

1O:l

Rolling and rocking

Variable: 28 days to several years

-DEVS4 Germany

Standard regulatory method

2L container

Ground specimen, <4 mm

100 g

deionised water

10:l

Table shaker

24 h

0.45 pm filtration

-TVA Switzerland

Standard regulatory method

Bottle

Not specified

1O:l

Bubbling, 100 mL

24 h

0.45 pm filtration

Standard regulatory method

1 L beaker

4h

0.22 pm filtration

-Total Availability NVN 7341 The Netherlands

100-200 g CO, saturated water

cod

minute Ground specimen, <300 pm

89

deionised water at pH 7 then pH 4 with nitric acid

100:1 Total

Magnetic stirrer

Table 14.6 Non-Agitated Extraction Tests Test Name Status of and Development Proponent -Sequential Chemical Extraction -Granular Diffusion Test The Netherlands

VI

CD

m

Leaching Vessel

Sample Leachant Preparation

LIS Ratio

Duration

Leachate Separation

Research

100 ml Teflon

<300 pm

Varies

Varies

Varies

0.22 pm

Under Development

2 L Plastic

Granular material

pH 4.0 Nitric

Varies, Sequential

64 days

0.22 pm

599 Table 14.7 identifies a number of serial batch tests that are used to provide for the successive renewal of leachant for a given mass of a solid material. Such tests can approximate flow-through leaching tests. Information is provided on the sample preparation, the type of leachant, the leaching vessel, the type of agitation, the sample mass and L/S ratio that is used, the contact time, the number of elutions that are generated, and the method of filtration. A number of researchers also employ a reverse cascade leaching test where the L/S ratio goes from 10 to 1 (F~llman, 1990). This is accomplished by renewing the solid to be leached rather than the leachant. Table 14.8 identifies a number of flow-around tests. These methods are used more for surfaces of monolithic-like materials. Information is provided on sample surface preparation, sample size, leachant, leaching vessel, volume to surface area and leachant renewal specifications. Table 14.9 provides details on a few flow-through or column leaching tests that are frequently employed by researchers. The type of column, the sample preparation, the sample mass, the cumulative L/S ratio, the means of providing the percolation, the times for sample collection and the filtration method are specified in the table. Table 14.10 discusses specifications that are used to establish outdoor field lysimeters by a number of institutions involved in ash research. Lysimeter construction, sample mass, and typical L/S ratios are specified in the table. 14.2 A UNIFIED APPROACH TO LEACHING TESTS

As indicated in previous chapters, examining leaching phenomena on a normalised basis is important. The normalisation process uses a ratio of instantaneous or cumulative mass of an element released to the dry weight of residue being leached. This is reported as a function of the L/S ratio which incorporates the dry weight of residue. The use of mass fractions leached and L/S ratios ultimately allow for examining mass fluxes from residues. Fluxes are more descriptive of leaching as a function of time than discrete measures of concentration. Because leaching behaviour can change dramatically over time and under different leaching conditions, it is recommended that more than one test (ie. more than one type of extract analysis) be used to provide the information necessary to characterise the potential leaching behaviour. A general evaluation framework, or a unified approach to leaching, includes use of the total elemental composition, an availability leaching test and pH-dependent leaching to provide reference levels for comparison. It is suggested that a series of basic characterisation leaching tests be selected, including determination of: the total availability for leaching; the elemental solubility as a function of pH; and the effect of increased L/S ratio or time on cumulative release.

Table 14.7 Serial Batch Tests Test Name and Prooonent

Status of Development

-MEP Standard test U.S. EPA method (1986) Method 1320

Leaching Vessel

Agitation

Sample Mass

US Ratio

Contact Time

Same as EP Acetic acid Tox Synthetic acid Distilled water

Same as EP Tox

Same as EP Tox

Same as EP Tox

Same as EP Tox

Same as EP Tox

10

Same as EP Tox

Particle-size Distilled water reduction to Site water <9.5 mm or structural integrity

Wide mouth sample bottle

Rotary tumbler

Unspecified

10:l

18 hours

4

Settling and filtration

Periodic gentle shaking (415 times daily)

300 g

213161121241481 96: 1

Until steadystate conditions attained

>7

Vacuum filtration

Sample Preparation

Leachant

Number Leachate of Separation Elutions

-MWEP U.S. EPA

Technical resource document (1986)

-Graded Serial Batch U.S. Army

Research method for waste and soil (1987)

-

-SEE D4793-88 ASTM

Standard method (proposed) (1988)

Drying, Phase separation

Reagent water (Type ll D l 193)

2 L, wide mouth bottle

None

100 g

20: 1

24 hours

10

0.45 pm membrane filter

-WRU Leach Test Harwell Laboratory United Kingdom

Standard method (1982)

Crushing Vacuum filtration

Distilled water Dilute acetic acid buffered (PH = 5)

50 mL, wide necked flask

Mechanical flask shaker

100 g

One bed volume (first five elutions) 10 bed volumes (more than six elutions)

2 to 80 hours Steady state

5

Vacuum filtration

Crushing1 sieving Dry

Distilled water 1L Nitric acid polyethylen e bottle (pH = 4.0)

Shakelroll

40 g

20: 1

23 hours

5

Settling and 0.45 pm filtration

-SLT Standard Cascade research Test method for SOSUV incinerator Netherlands residues (1984)

Distilled water Unspecified

Table 14.8 Flow Around Tests Test Name Status of and Proponent Development -MCC-4s MCC

Standard (1983)

Leaching Vessel

Volume: Surface Area

Leachant Renewal Rate

Pure water, bicarbonatesilicate water, repository water

1 L teflon cylinder

10 cm

Flow rate = 0.1. 0.001, and 0.001 mUmin

Demineralised water Site water

Compatible with sample and leachant

4 0 cm

Daily (1st week), weekly (8 weeks),monthly (6 months), biannually

= 0.5 to 0.001 Synthetic seawater

Deionised water

10to20cm

m2

Site water

Cylinder: unreactive, radiation resistant

1,3,7,10,14, 28, 35, 42, 72, 102 days

Cylinder, lengthldiamet er = 0.2 to 5.0

Distilled water

Cylinder: unreactive material, sized to immerse sample

10cm

2,7hours;1,2, 3, 4, 5, 14, 28, 43,90 days

Curing and surface Cylinder, 4.5 washing cm diameter x 7.5 cm long

Distilled water

Large, wide mouth jar

Allows contaminant detection, diffusion modelling

0.5, 1, 2, 4, 8, 12, 16, 32, and 64 days

Sample Preparation

Sample Size

Individual fabrication or surface washing or cut sample

400 mm2

-IAEA Dynamic Leach Test IAEA

Past standard One circular face is Cylinder, (1971) prepared for 5 cm x 5 cm leaching

-IS0 Leach Test IS0

Past standard (1986)

Surface polishing

-ANSI/ANS 16.1 ANS (American Nuclear Society

Standard ( I 986)

Surface washing

-NEN 7345 (NVN 5432) Monolith Leaching Test

Draft Standard (1993)

Surface area

Leachant

Table 14.9 Flow Through or Column Tests

DI o

N

Test Name, Proponent

Status of Development

Leaching Vessel

Sample Preparatio

Sample Mass

-Column Extraction Method, ASTM

Proposed method

10 cm x 30 cm column

field samples

500 g

-Column Method SOSUV, The Netherlands

Standard regulatory method

5 cm x 20 cm column

<3 mm

-Mini-Column, (Clarkson University)

Research method

1.5 cm ID, Piston controlled length

<300 pm

n

US Ratio

Leachate Collection

Leachate Separation

Simulate Pumped up field, 0-3 flow (cumulative)

At 1, 2, 3. 8 pore volumes

0.22 pm filtration

500-800 g

0-10 Pumped up (cumulative) flow

Periodically over 10 d

0.45 pm filtration

29

0-1.000 Pumped up (cumulative) flow, 0.1 to 2 mllmin

Periodically over hours to days (up to 1000 pore volumes)

0.45 pm filtration

Percolation

Table 14.10 Field Lysimeter Tests Test Name, Proponent

Status of Development

-Field Lysimeter, University of New Hampshire, USA

Research; Bottom Ash, Combined Ash

-Field Lysimeter, AVR, The Netherlands

Research; Bottom Ash

Sample Mass

US Ratio

Leachate Collection

Leachate Separation

2m x 3mX 4m Roll off lined with HDPE Liner, geonet, geofabric and drainage sand

5,00010,000 kg

0-5

Periodically pumped out after precipitation

0.22 pm

2.5m x 2.5m x 12m containers

100,000 kg

0-1

Periodic collection; pore water analysis; run-off analysis

0.45 pm

6,000 18,000 kg

0-4

Drain to automatic collection system

0.45 ym

Structure

3m x 3m x 1.5m or -Field Lysimeter, Research; VKI, Denmark Many Ash Types 3m x 3m x 2.5m prefab concrete with LDPE liners and drainage system -Field Lysimeter, SGI, Sweden

Research; Consulting; Many Ash Types

3m x 3rn x 1.2m or 4m x 4m x Im; lined with HDPE with geotextile drainage system

15,000 20,000 kg

0-2

Periodically drained to tipping bucket under argon gas

0.45 pm

-Field Lysimeter, KfK, Germany

Research; Bottom Ash, ESP Ash

cylindrical 0=63cm, h=lm cylindrical 0=80cm, h=lm acrylic columns

300-400 kg 140 kg

0-1 0-15

Periodically removed (-2 weeks)

0.45 pm

604 Consider the normalised plot of the pH dependent leaching of copper from an APC residue (dry scrubber) as shown in Figure 14.9a. On the y-axis is the mass fraction released in mg/kg. This can also be reported as concentration in mg/L. On the x-axis is pH of the leachate at equilibrium. For reference purposes, two horizontal lines are shown depicting the total and available copper concentration in the residue. A region is shown where the pH dependent data fall. The graph indicates the amphoteric leaching behaviour of the metal. Also note that at very low or high pH, the pH dependent test releases the same amount of copper as the availability test. The various single point data depicted in the plot are derived from regulatory agitated batch extraction tests, as shown in the key. These data are plotted as release versus final extraction pH. What is apparent is the uniformity of the leaching behaviour in spite of the various tests that were used. The regulatory test data fall within the range of the pH dependent data. Normalisation helps to reduce variability associated with concentration measurements and variable US ratios. Figure 14.9b brings together the principles shown in Figure 14.9a as well as from other tests in a unified plot. The y-axis shows either instantaneous or cumulative release of copper from an APC residue (dry scrubber, in mg/kg) as a function of either the instantaneous or cumulative US ratio. Depicted in the plot are the two horizontal lines for total and available copper in the residue. Both asymptotes are useful references for evaluating the results of various leaching tests. Also shown on the plot are cumulative release plots for serial batch extraction, upflow column leaching tests and the results from various regulatory agitated batch extraction tests. One important question to ask is how quickly or slowly these cumulative release plots approach the total available asymptote. By using an L/S of 100, the serial batch and upflow column tests have clearly removed most of the available Cu fraction, although this behaviour may be different for elements. Also shown on the plot are the ranges of data seen in the pH dependent leaching (Figure 14.9a) at an L/S of 20. The values close to the asymptote are the regions where copper is very soluble (pH 4 and 12). The values situated well below the asymptote are for pH values where copper is insoluble (pH 7-10)o A particularly useful approach is to relate the data shown in Figure 14.9b to actual field leaching data. Questions to pose include how the cumulative release plots approximate the field scenario? Alternatively, has the field plot also approached the total available asymptote? Data from many of the leach tests described in Section 14.1.3 can be incorporated into the unified plot of Figure 14.9b. The data can also be used with modelling efforts discussed in Chapter 15.

605 Figure 14.9 A Unified Approach to Leaching a) pH Dependent Leaching

1000

1L

I,,,~

Total Concntration Total Available Concentration (NVN 7341)

100 Release

+

10

mg/Kg

1.0

Key # Swiss TVA * German DIN @ French X31-210 o US TCLP x US EP-Tox % California WET + JLT 13

0.1 0.01

pH Dependent Leaching (pH stat) Leaching Region

I

2

I

4

I

I

6

I

8

I

10

12

pH b) cumulative leaching

1000 Release mg/Kg

Total Concentration Total Available Concentration (NVN 7341)

100 10 1.0 0.1

Y u

Leaching Region for Dutch Serial Batch Test (NVN 7349)

_~n~U~.owCo,u~oTosts~W ~ Z /

J

j

0.01

~

Y

0.1

1.0 L/S

(L/Kg)

J

~/

Key # Swiss TVA * German DIN @ French X31-210 o US TCLP x US EP-Tox % California WET + JLT 13 y pH-Dependent

I

10

100

606 REFERENCES Environment Canada. Compendium of Waste Leachin,q Tests. Report EPS 3/HN7, Environment Canada, Ottawa, Canada, 1990.

Environment Canada. The National Incinerator Testin.q and Evaluation Program (NITEP) A Summary of the Characterisation and Treatment Studies on Residues from Municipal Solid Waste Incineration", Report EPS 3/UP/8 prepared by S.E. Sawell and T.W. Constable, October 1993. F~llman, A.-M. International Seminar on Leach Tests. Geotekniska Institute, Link0ping, Sweden, 1990.

Report 317, Statens

Fraser, J.L. and K.R. Lum. "Availability of Elements of Environmental Importance in Incinerated Sludge Ash." Environ. Sci. Technol. 17 1-9, 1983. Jackson, D.R., B.C. Garrett and T.A Bishop. "Comparison of Batch and Column Methods for Assessing Leachability of Hazardous Waste." Environ. Sci. Technol. 18: 668-673, 1984. Kheboian, C. and C.F. Bauer. "Accuracy of, Selective Extraction Procedures for Metal Speciation in Model Aquatic Sediments." Anal. Chem. 59: 1417-1423, 1987. Nirel, P.M.V. and F.M.M. Morel. "Pitfalls of Sequential Extractions." Water Res. 8: 1055-1056, 1990. Tessier, A., P.G.G. Campbell and M. Bisson. "Sequential Extraction Procedure for the Speciation of Particulate Trace Metals." Anal. Chem. 51" 844-851, 1979. van der Sloot, H.A., D. Hoede and P. Bonouvrie. Comparison of Different Re,qulatory Leachin.q Test Procedures for Waste Materials and Construction Materials. ECN-C-91082, ECN, Petten, the Netherlands 1991. van der Sloot, H.A., O. Hjelmar, Th. G. Albers, M. Wahlstr0m and A.-M. F~llman. Proposed Leachin.q Test for Granular Solid Waste. ECN-C-93-012, ECN, Petten, the Netherlands, 1993. Wadge, A. and M. Hutton. "The Leachability and Chemical Speciation of Selected Trace Elements in Fly Ash from Coal Combustion and Refuse Incineration", Environmental Pollution, 48(2): 85-99, 1987. WTC (Wastewater Technology Centre). "Modified Sequential Chemical Extraction Procedure", WTC Laboratory Manual of Methods, Internal Report, 1990. Zachara, J.M. and G.P. Streile. Use of Batch and Column Methodolo.qies to Assess Utility Waste LeachinQ and Subsurface Chemical Attenuation. EPRI EN-7313, EPRI, Palo Alto, CA, 1991.

607

CHAPTER 15- LEACHING MODELLING

The principles introduced in Chapters 12 and 13, when combined with appropriate field leaching data or leaching test data introduced in Chapter 16, allow for the use of models to verify mechanistic leaching behaviour and to predict leaching behaviour as a function of time, residue management practice, or treatment scheme. There are numerous advantages associated with the use of models to help understand and simplify complex leaching systems. The fact that complex, heterogeneous, multiphase processes can be accurately described by models with mechanistic bases is, in its own right, an important breakthrough. Such information has been used to show the systematic leaching behaviour for many waste materials, including incineration residues. These models can be used to predict changes in leaching behaviour over much longer time frames of tens to thousands of years. Further, the models can also be used to predict changes in leaching behaviour under different management or treatment scenarios. When coupled with appropriate leaching tests for treated ash products, these models are useful in providing a high level of predictive capability. Researchers studying ash leaching have used two distinct modelling approaches. The first is based on geochemical thermodynamic equilibrium models that require the leaching system to be at or near equilibrium for all of its dominant solution phase reactions, solution-solid phase reactions, and solid phase-solid phase reactions. This approach describes leaching behaviour under non-flow-through conditions analogous to discrete "snapshots" in the geochemical domain (e.g., a specific pH, Eh, L/S, solution complexation conditions, sorption conditions). The second approach employs a dynamic multicomponent flow-through leaching model with mass transport, mass balance and mass action constraints. The model is described by boundary conditions and initial conditions for leachable mass fractions. Under plug flow conditions, the system is allowed to evolve and predict phase dissolution and equilibrium-bounded phase reprecipitation. This second approach differs from the first in that the flow-through system, as a function of pore volume or time, can predict longer term leaching behaviour. Again, the dynamic system can be viewed as a series of "snapshots" under changing regimes of pH, L/S and redox. Both approaches are described in detail below. The reader is provided with references to reviews that offer more detailed information on the two approaches. 15.1 EQUILIBRIUM MODELS

The leaching system (Figure 12.2 in Chapter 12) shows all of the equilibrium reactions (solution complexation, dissolution/precipitation, sorption) occurring amongst the hypothetical solid phase ash particles in a complex, heterogeneous system. To simplify

608 this system slightly, chemical weathering reactions are, for now, neglected because they are kinetically slow reactions that are unlikely to be at equilibrium in the leachates that are studied. This system may be described as typically containing many different mineral phases, some of which are oxides or aluminosilicates with surface adsorption sites of various affinities for a variety of adsorbing solutes. The leaching solution contains many different solutes in a non-ideal solution with high ionic strength. Each solute can undergo solution complexation reactions, sorption reactions and precipitation/dissolution reactions. This complex system can approach equilibrium where the free energies of each of these reactions collectively reach a minimum. Researchers are now attempting to employ geochemical thermodynamic equilibrium models to help mechanistically interpret leaching phenomena.

15.1.1 Thermodynamic Equilibrium Models Numerous reviews of thermodynamic equilibrium models have been published (Kincaid et al., 1984; Kincaid and Morrey, 1984; Morrey et al., 1986; Nordstrom and Ball, 1984). The utility, applicability and limitations of classes of these models are presented here. The chemical equilibrium condition of a complex system is an attempt to find the most stable state of the system under specified conditions of pressure, temperature and composition. In a thermodynamic sense, it is an attempt to find the Gibbs free energy minimum for the system. This can be done in two ways: by minimising a free energy function or by solving a set of nonlinear equations consisting of equilibrium constants and constraints on the mass balance of participating elements in the system (Nordstrom and Ball, 1984). The major disadvantage in utilisation of a free energy database is these values tend to be less reliable than directly measured equilibrium constants (Nordstrom and Ball, 1984). Invariably, caution is needed in selection of the appropriate model and database and in ensuring the database conforms to the constraints of the studied system. The numerical algorithms used to solve the nonlinear equations usually involve (i) pure iteration, (ii) Newton-Raphson iteration and (iii) integration of ordinary differential equations. The pure iteration method is a brute approach where the system approaches unity after successful back-substitution iterations. The integration approach is used to solve complex forward and backward kinetic reactions. The Newton-Raphson method is the most widely used; it sometimes does not converge as quickly as the iteration method. There are more than fifty programs that have been developed for aqueous phase systems. Some are now available for use on a personal computer, however, given their

609 complexity and iterative nature, mainframes are often used. Leachates derived from batch, column or field leaching studies are usually modelled. It should be determined that equilibrium is approached in the leaching test or in the field prior to collection of leachate: the concentration of major and minor constituents in the leachate should be constant over time. Most models input leachate pH, temperature, analyte concentrations, and sometimes Eh and ionic strength. Most models output aqueous phase speciation, species activity, species activity coefficients and saturation indices that describe a tendency for elements to precipitate or stay in solution. The solid phases that can precipitate are usually identified. More recent versions of the models such as MINTEQA2 include surface adsorption reactions and allow for the selection of various sorption models for use in modelling adsorption and desorption. Care is needed in the selection of adsorption constants. Some limitations to these programs involve the difficulty in which equilibrium constants in the database can be changed, the Gibbs phase rule violations that can occur when too many solid phase constraints are placed on the system, the inability of some models to converge easily and the inability of some models to contain algorithms that reprocess the system at different pH or pE values. The newest version of MINTEQ, MINTEQA2, and the model SOLGASWATER developed by Erickkson (1979) have this iterative capacity.

15.1.2 Use of the Geochemical Thermodynamic Equilibrium Model MINTEQA2 MINTEQA2 is an equilibrium model for dilute heterogeneous aqueous systems. The original MINTEQ, developed by Felmy et al. (1984), was a hybrid of MINEQL, developed by Westall et al. (1976) from REDEQL (Ingle et al., 1978) and WATEQ3, a code developed by the U.S. Geological Survey (Ball et al., 1981). The fundamental mathematical structure of MINEQL and the thermodynamic database of WATEQ3 were combined to produce MINTEQ. The most recent version of MINTEQ, MINTEQA2 (version 3.11) is presently administered by the Environmental Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Athens, Georgia 30613 (Allison et al., 1990). MINTEQA2 is a much more interactive program than its predecessors. There are more options and features available to the user, the methods of calculation have been improved and the thermodynamic database is more up to date and more readily modified. The database has been deemed to be the most complete, annotated and accurate of the existing current computer codes (Krupka et al, 1986; Morrey et al., 1986; Allison et al., 1990). For ash leaching scenarios the database must be modified to include certain constituent "components" and mineral phases found in residues. MINTEQA2 also uses an interactive program, PRODEFA2, to create input files for MINTEQA2. The information required to operate MINTEQA2 usually consists of chemical analyses of a leaching solution (at equilibrium) giving total dissolved analytical concentrations

610 (CT in units of mg/L, parts per million, molal or molar) and other measurements of the system, including pH, pE (or Eh), gas partial pressures, the presence of mineral phases and the capacity to allow sorption to occur. By specifying gas partial pressures (e.g. CO2) and finite (potential dissolvable) solid phases, source terms or supplies of these constituents are available to be pulled into the mass action equations. Alternatively, system pH and pE (or Eh) can be set and hypothetical minerals can be input to see if they will dissolve or remain insoluble and produce an aqueous solution similar to a leachate. MINTEQA2 uses the simultaneous solution of nonlinear mass action expressions (derived from equilibrium expressions) and linear mass balance relationships to solve for the system equilibrium conditions. During iterations it will transfer mass from gas phases or finite solids into these terms to satisfy equilibria. The system can be overconstrained when too many phases or set points are specified, removing degrees of freedom and ultimately causing failure of the program. MINTEQA2 uses an initial guess of the activity of each component to calculate the concentration of each species according to mass action expressions. The total mass of each component is then calculated from the concentrations of every species containing that component. The calculated mass is compared with the input mass; if it exceeds a tolerance level, a new iteration is conducted (hundreds of iterations can occur). A final aqueous phase concentration for all species is determined once tolerance levels are met. MINTEQA2 calculates saturation indices during or after the iterative process to see which potential phases can precipitate. The solid phase with the largest positive index can be allowed to precipitate in accordance with reaction stoichiometry. The reverse can also occur if an inputted solid phase is undersaturated with respect to the solution phase saturation index. The whole system is then re-equilibrated through successive iterations until there are no oversaturated possible solids or undersaturated existing solids. The iteration process can be operated for system set points for pH and pE. Sorption models can also be included. They are discussed below. MINTEQA2 has some system limitations. It is meant for temperatures below 100~ It uses the van't Hoff equation to modify equilibrium constants for temperatures other than 25~ It also employs the modified Debye-H0ckel and the Davies equations to determine activity coefficients. Thus, its effective range for l is from 0.0001 up to 0.5 molal. Higher ionic strengths require the use of alternative methods for determining activity coefficients, such as the Whitefield modifications to Pitzer's original work discussed in Chapter 13. While MINTEQA2 is not readily modified, models such as SOLMINEQ (Perkins et al.. 1990) and SOLTEQ (Batchelor and Wu, 1993) have these capabilities. Finally, MINTEQA2 also can include in the input some measure of total alkalinity to determine dissolved inorganic carbon [CO32, HCO3, H2CO3, CO2(aq)]. These are crucial components that are often neglected.

611 MINTEQA2 uses seven different surface adsorption models: on-Electrostatic Models Activity Kd adsorption model Activity Langmuir adsorption model Activity Freundlich model Ion exchange model Electrostatic Adsorption Models. Constant capacitance model (CCM) Diffuse layer model (DLM) Triple layer model (TLM) These models, described in Chapter 13, are a unique and important feature to MINTEQA2. An excellent source of sorption constants for each of these models can be found in Rai and Zachara (1984) or in Dzombak and Morel (1990). These databases report sorption parameters for a large number of sorptive surfaces, many of which can be modelled in ash systems. Again, caution should be used in the selection and use of these sorption parameters. 15.1.3 Verification of MINTEQA2

Morrey et al. (1986) conducted detailed evaluations of the MINTEQA2 precursor, MINTEQ, for its possible use as a basis for the EPRI FASTCHEM geochemical/ hydrodynamic solute transport code. Morrey and co-workers were intent on verifying the ability of MINTEQ to properly describe an equilibrated geochemical system containing the minerals malachite, gypsum, gibbsite, cerussite and goethite in equilibrium with coal fly ash leachate solution species. These minerals are important and dominant components of coal fly ash. One approach (the saturated solution approach) was to input the leachate concentrations of these elements as dissolved species in the presence of the five minerals. Could MINTEQ achieve equilibria and predict the continued presence of the five solid phases? The second approach (the solid phase approach) was to input the solid phases as finite (potentially soluble) phases and specify all potential dissolved components as present but at very low concentrations initially. In parallel, these phases were actually solubilised in a laboratory batch leaching test. Could MINTEQ achieve equilibrium and produce solution concentrations similar to the laboratoryderived analytical concentrations? Table 15.1 describes the master system and solution variables for the first approach. The constituents describe a leachate in equilibrium with the five dominant minerals;

612 other trace constituents were also present. MINTEQ was run two ways: 1) by allowing additional minerals to form and 2) not allowing additional minerals to form. In the first case, a phase rule violation caused a code termination when the program tried to reprecipitate hematite (Fe203) when goethite (FeOOH) was already designated as a finite solid. An attempt to precipitate the thermodynamically-preferred hematite caused the phase rule violation. Using the second approach, equilibrium was obtained by extracting 5.8 x 10 .3 moles of carbon from the infinite gas supply, dissolving small amounts of gypsum and precipitating small additional amounts of cerussite. Gibbsite, goethite and malachite were unchanged. As shown in Table 15.2, their saturation indices remained at 0, indicating their role as controlling phases. The trace constituents did not disrupt the achievement of equilibrium, indicating dominant phases usually control system equilibria. Table 15.1 Master System and Solution Variables for Evaluating the Saturated Solution Approach to Using MINTEQ Master Variables" Alkalinity pH CO2 pressure 02 pressure Temperature

= = = = =

4.3 meq/L 8.2 1.58 X 10.3 atm 0.2 atm (oxidising) 11~

Major Cations and Anions (mg/L) C a 2§

Mg*2 Na+ K*

=

= = =

400 50 75 20

Minor and Trace Constituents (m.q/L) AI = 0.01 As = 0.01 Cd = 0.03 Co = 0.01 Cr = 0.02 Cu = 0.02 F =0.50 After Morrey et al., 1986

Fe = 0.20 Mo = 0.01 Ni = 0.02 Se = 0.01 Zn = 0.03 Pb = 0.02

S O 4 -2

=

cr = NO3= B(OH)4 =

1,000 80 30 20

613 Table 15.2 Results from Saturated Solution Input Verification of MINTEQ System Variables pH Sum of Cations (molal) Sum of Anions (molal) Percent Difference (C-A/C+A) Ionic Strength

8.200 2.15 x 10.2 2.31 x 10.2 -3.86 3.86 x 10.2

Mineral Saturation Indices Aragonite, CaCO3 Boehmite, AIO(OH) Calcite, CaCO3 Diaspore, AIO(OH) Dolomite, CaMg(CO3)2 Hematite, Fe203 Gibbsite, AI(OH)3 Goethite, FeOOH Gypsum, CaSO4o2H20 Hunite, CaMg3(CO3)4 Magnesite, MgCO3 Tenorite, CuO Cupricferrite, (Cu, Fe2)O4 Cerussite, PbCO3 Otavite, CdCO3 Plattnerite, PbO2 Pb203, Malachite, Cu2(OH)2CO3

1.184 0.000 1.375 1.831 1.962 4.940 0.000 0.000 -0.002 -1.108 0.102 0.031 2.970 0.030 1.449 -0.444 -1.886 0.000

After Morrey et al., 1986 Table 15.3 denotes the master system and finite solid variables for the solid phase approach. In parallel, excess masses of ground malachite, gypsum, gibbsite, cerussite and goethite were added to distilled water to produce an equilibrated solution with solids present. Sodium hydroxide was used to control pH at 8.2. MINTEQ was operated with the variables shown in Table 15.3. These solids and their solid phase concentrations were inputted as finite solids while all possible dissolved components were listed but assigned very low concentrations. The addition of the sodium hydroxide titrant was also included as a system parameter.

614 Table 15.3 Master System and Solid Phase Variables for Evaluating the Solid Phase Approach to Using MINTEQ Master Variables:

Alkalinity pH Na* Total Hydrogen

Finite Mineral Phases

Malachite Gypsum Gibbsite Cerussite Goethite "From the NaOH used to control pH After Morrey et al., 1986

= = = =

9.83 mg/L as CO32 8.2 18.75 mg/L a 0.8277 mg/L a Cu2(OH)2CO3 CaSO4 o2H20 AI(OH)3 PbCO3 FeOOH

Table 15.4 denotes the results of the modelling run compared to the experimentally derived leachate. Fairly close agreement was seen. Some concern was expressed by the authors about the lack of complete agreement between the predicted and observed leachate concentrations. Experimental non-equilibrium was hypothesised for gypsum and gibbsite. Also a concern was the suspicion that 0.22 IJm filters allowed nanocolloid precipitates to pass through the filter, thus overestimating the true dissolved total concentration for some elements. Table 15.4 Results from Solid Phase Input Verification of MINTEQ Element MINTEQ Calculated mg/L ExperimentallyDetermined mg/L Sodium

18.75

11.0

Calcium

392.6

590.0

Carbon

9.61

9.93

Aluminum

0.02

0.31

Sulphur (SO42)

941.1

1,499.0

1.69 x 10.8

NAa

Copper

0.263

NA

Lead

0.320

NA

Iron

Ionic Strength pH aNA = not analysed After Morrey et al., 1986

2.847 x 10.2

NA

8.09

7.8

615 Recent efforts by many researchers (DiPietro et al., 1990, DiPietro, 1989; van der Sloot et al., 1992; Theis et al., 1992; Dzombak et al., 1992; Comans et al., 1993; Eighmy et al., 1993) have made use of MINTEQA2 or its precursors in modelling efforts. This model has also been selected by EPRI as a basis for their ECHEM geochemical thermodynamic code for modelling coal ash leachate advection and dispersion in FASTCHEM. Thus MINTEQA2 appears to be, for now, an appropriate geochemical code for modelling.

15.1.4 Recommendations for Utilising MINTEQA2 to Model Leaching Behaviour There are a number of recommendations that will assist users of MINTEQA2 in making use of this powerful tool for modelling leaching: For the Saturated Solution Approach Input all constituent concentrations in the equilibrated leachate solution; leachates should be from column tests, from longer duration batch tests at lower US ratios (to ensure equilibria) and from field leachates or pore waters. Set the appropriate temperature. Include measures of either total alkalinity or carbonate species (not both); this can also be left unspecified in initial trials if a complex system is being modelled. Fix pH to the measured leachate solution value; this can also be left unspecified in initial trials if a complex system is being modelled. Fix pE (or Eh) to an oxidised (or reduced) level; this can also be left unspecified in initial trials if a complex system is being modelled. Initially do not allow any solids to precipitate. Run the program with the intent of it converging to an equilibrium solution. Check cation/anion balances, subtly adjust simple cations (Na § K § or carbonates as most likely cause of discrepancy (assuming all constituents were quantified). Rerun the program; a number of trials may be needed to achieve a good ion balance.

616 Rerun the program after reasonable ion balances are achieved and allow the solid phase with the highest saturation indice to precipitate as a finite solid. The possible solid exclusion option may be needed to prevent MINTEQA2 from redissolving a more soluble phase before a more insoluble phase precipitates out or to prevent nonsensical solids from forming. Run the program until four or five dominant phases remain as precipitated solids with saturation indices at zero. Compare the identity of the precipitated solids with those seen in ash residues using methods described in Chapter 7. Try running the program with specified CO2 partial pressures to see how an infinite supply of CO2 alters equilibria (if this is appropriate). Try varying pH, pE (or Eh), gas partial pressures, solution-phase chelates (organics), the presence of new solid phases, etc. to see how the system responds. Attempt to run adsorption subroutines only after the more basic system is described and understood. .For the Solid Phase Approach Input all possible or suspected solid phases as finite solids; the thermodynamic database will most likely need modification to include phases not in the database. Data on phase equilibrium constants (e.g. I/K~ or K~ ~ because MINTEQA2 approaches all equilibria constants as dissociation constants so the inverse of the commonly reported Ks, or K,o is needed), phase enthalpy and reaction stoichiometry are needed. For inputted phases, estimates of the finite solid concentration will be needed. These are reported as moles available per one litre of solution. Total Availability Leaching Test data can be used here. Identify all possible component dissolved aqueous species of the solid phases and specify their starting concentration at a low value (e.g. 1 x 10-is molal). Include measures of either total alkalinity or carbonate species (not both); this can also be left unspecified in initials trials if a complex system is being modelled.

617 Input the total estimated hydrogen ion concentration (as well as its conjugate base concentration) as dissolved constituents to account for the acid used in the leaching test or fix the pH to the measured leachate solution value; this can also be left unspecified in initials trials if a complex system is being modelled. Fix pE (or Eh) to an oxidised (or reduced) level; this can also be left unspecified in initial trials if a complex system is being modelled. Run the program with the intent of it converging to an equilibrium solution, check cation/anion balances, subtlety adjust solid phase concentrations to correct the balance. Rerun the program until a balance is achieved. Compare the final solution composition with the leaching solution, examine solution-phase speciation. Try running the program with specified CO2 partial pressures to see how an infinite supply of CO2 alters equilibria (if this is appropriate). Try varying pH, pE (or Eh), gas partial pressures, solution-phase chelates (organics), the presence of new solid phases, etc. to see how the system responds. Attempt to run adsorption routines only after the more basic system is described and understood. These approaches are demonstrated in the following sections.

15.1.5 Modelling L/S, pH and Redox Control of Leaching The work of Collins and DiPietro (DiPietro et al., 1990; DiPietro, 1989) evaluated the effects of varying the liquid to solid ratio (L/S), pH and redox potential on leachability on combined bottom ash and boiler ash from a two-stage incinerator. They used the apparatus shown in Figure 15.1 to control redox in an ash/wastewater sludge system at either (i) reducing, intermediate, or oxidising conditions, (ii) pH values of 4, 7, or 10, or (iii) two L/S values of 5 or 10 under presumed equilibrium conditions. The sludge was used to make the matrix slurry initially reducing. The addition of oxygen or nitrogen was used to maintain appropriate redox ranges. After equilibrium was attained (4 to 5 days), leachate samples were collected, filtered and quantified for anions and metal and metalloid elements. The leachate compositional data and system values (pH, redox, ionic strength) were loaded into MINTEQ using PRODEF and the

618 saturated solution approach. After cation/anion balance adjustment, certain phases were allowed to precipitate in equilibria with the solution phase. These predicted precipitates were compared with data on solid phases as determined by X-Ray Powder Diffraction (XRPD). Figure 15.2 shows the results for lead. The data show how at low pH, lead is mobilised. At high L/S, the concentration is decreased, suggesting an extensive solid phase dissolution. Reducing conditions do lower aqueous phase concentrations. The calculated CT.pbfor the solution phase at equilibrium, as predicted by MINTEQA2, was in good agreement with observed data. Data for other elements showed agreements as well. There was not good agreement between the solid phases seen with XRPD and those predicted by MINTEQA2. Figure 15.1 Apparatus Used for pH and Redox Control ORP controller

J

Solenoid valve

Gas Exhaust ORP Probe \ f

Acid/Base input port

Diffuser Stone After DiPietro, 1989

pH Probe

619

Figure 15.2 Lead Leachability as a Function of L/S, pH and Redox

Lead /

~

.....

L/S-

5.1

I

L / S = 10:1

Con c , .

./

. . . . . . . . . .

1.5

~

-

/ 0"5 I / B D L ~

7pH

/BDL~

I

I

Batch

MINTEQ

REDUCED

~

OXIDIZED

Batch

MINTEQ

~

Middle ORP/pH 7

After Dipietro, 1989

15.1.6 Modelling L/S, pH, Redox and Complexation Control of Leaching Van der Sloot (1990) evaluated the effects of L/S, pH and solution phase complexation for bottom ash and ESP ash from a Dutch mass burn incinerator. Leaching tests based on the Dutch column percolation experiment and the Dutch multiple batch extraction test (NVN 2508) were used to evaluate three L/S ratios: 1, 5 and 100. pH was controlled over the entire pH range using either H2SO4 or CaO in the extraction solutions. The effect of aqueous phase chloride complexation was investigated by using a high chloride content fly ash that would presumably leach more chloride into the extraction solution than the bottom ash. The systematic leaching behaviour of cadmium, copper and zinc was investigated. The equilibrium concentration of leachates from the fly ash (L/S = 5) and the bottom ash (L/S = 1 or 100) were input into MINTEQA2 along with other system parameters (final pH, temperature) using the saturated solution approach. The solid phases were not allowed to precipitate in the model. The results for the leaching of cadmium, copper and zinc are shown in Figure 15.3. The results are presented as mass fractions leached (%) as a function of final leachate pH. In most cases, as discussed in Chapter 13, the presence of significant concentrations of chloride (as a complexing ligand) in the fly ash helped to increase the mass fraction released for cadmium and zinc. MINTEQA2 was able to accurately

620

predict the mass fraction released by successfully modelling significant chloride compiexation of Cd and Zn in the aqueous phase, thus promoting solubilisation. Both Cd and Zn exhibited minimum solubilities associated with hydroxide or carbonate phases. Figure 15.3 Cd, Cu and Zn Leachability F l y ash.

Slog

NITEP

-'-

--"

G"=+

LS : 5

LS: I 0 0

MINTEQ

Fly ash

Slag .~41- -

LS : 5

LS : I 0 0

r

..o..

0

IOO

MINTEQ --o--

I0r

Cd

Cu I(3

\

9

\

,

I

0.1 0.OI

\ O-OJ

o~

$

6

?

8

0

I0

II

Fly ash

:E

-r LS : 5

..J

12

13

o.ool

3;;;

NITEP

Slag ---ii- -

--&--

LS : I

LS = 2

9 "9~ , .

O~

/

MINTEQ 9- o.

9

Zn

0-01

0 "0013

4, ;

6l ,.

.

8 . +. . ,o. ,, ,2

pH

From van der Sloot, 1990 with permission of Waste Management and Research Copper, on the other hand, did not exhibit an increased solubilisation in the presence of increased solution phase chloride complexation. The presumed high levels of organic acids in the bottom ash tended to promote solubilisation of copper out of the bottom ash. Copper forms strong complexes with soluble organic carboxylic ligands. MINTEQA2 was less successful in predicting mass fractions released because organic ligand concentrations were not inputted into the modelling run.

621

15.1.7 Modelling Sorption Reactions Influencing Leaching Dzombak et al. (1992) evaluated the effects of solution pH and sorption for petroleum sludge fluidised bed fly ashes. The fly ashes were obtained from a venturi wet scrubber. Thus, the residues contained both fly ash and scrubber residues. XRPD was used to verify the presence of amorphous (FeOH)3 and AI(OH)3, SiO2, Fe304, CaCO3, CaSO4 and CaHPO4 in the solids. Leaching tests were conducted under batch-agitated conditions at an L/S of 100. The tests were run for 16 hours. The chemical equilibrium program MINEQL, a prototype to MINTEQA2, was used. The generalised diffuse layer model (DLM), a feature of both source codes, was employed to examine sorption/desorption in addition to precipitation/dissolution and solution phase speciation. The database generated by Dzombak and Morel (1990) for sorption of ions onto hydrous ferric oxide was added to MINEQL. The scrubber solids were considered to be dominated by metal oxides which were modelled as amorphous hydrous ferric oxide. Leachate analytes, solid phase concentrations, ionic strength and activity coefficients were inputted into the model using the solid phase approach. During modelling, solid phase formation was suppressed. The data for a variety of elements with strong dependence of pH are shown in Figure 15.4. As pH decreases, Mn, Zn, Co, Ni, Cd and Mg exhibit a plateau behaviour consistent with a sorption edge from an oxide surface that occurs near a pH of 3 to 6. For the most part, model predictions for solution phase concentration for Mn, Co, Cd, Zn, Ni and Mg were close to observed values. The data for a variety of elements with a weak dependence on pH are shown in Figure 15.5. As pH decreases, Pb, Sn, Mo, Sb, Be, Ba and Ag do not exhibit a strong plateau or pH dependency for dissolution. The model was not as successful in predicting behaviour, possibly because of the high carbonate content of the solid phase and the poor binding constant for ion sorption to carbonates.

15.1.8 Modelling Solid Phase Control of Leaching in Conjunction with Solid Phase Speciation Studies Comans et al. (1993) used MINTEQA2 to model the solid phase control of leaching of bottom ash as a function of pH. A pH stat system was used to control pH between 4 and 13. Three L/S ratios were investigated (2,5,10). A 24-hour equilibration period was used; kinetic studies revealed this was a reasonable time frame. They used a combined solid phase - solution approach to model the solid phase control. They inputted the equilibrated leachate constituents into the model and then tested a variety of minerals to see which mineral produced a solubility plot in agreement with the laboratory data. The solids were introduced as infinite solids in the model.

622

Figure 15.4 Sorption/Desorption Processes on Ash Surfaces for Elements Exhibiting a Strong pH Dependence 9

Mn

9

o

O

Co

-~

o

Cd

~

,e

E

Q. EL

o. 9 o

3

~ e'l

2

"o

-

l

I

1

i

I

I

i

~

i

1

1 2 3 4 5 6 7 8 9

12345

67

7O

15 13 E o. l l Q.

9

5O -

O m ,

,o o

o

~;, a..2

6

7

8

9

Mg

,o 9 "" o 9 ",o

30

Ni

E 10 90

E

7

4 5 pH

9O o

Zn

17

3

89

pH

19

2

I

pH

"'" O

7O

o

5

50

3

3O

1 123

4 56

pH

789

I;,3

4

5

6

pH

7

-9, 8

b "9

10 9

123

9 pH

Solid Line = Model-Predicted Distribution After Dzombak et al., 1992 with permission of author The data, some of it shown in Figure 15.6, show some solid phases clearly control the pH-dependent leaching behaviour. The controlling solids (e.g. CaSO4.2H20, Cd(OH)2, ZnO) are typically found in bottom ash. Eighmy et al. (1993) have utilised a number of spectroscopic and analytical methods to identify solid phases in a variety of ash residues. As part of that study, total availability leaching studies were conducted to allow leaching modelling of these residues with MINTEQA2. One aspect of the study was to use the solid phase approach and to input ESP ash solid phases and their molal concentrations into MINTEQA2 to see if the model would (i) dissolve the soluble phases in the ash, (ii) partially dissolve more insoluble phases in the residue, (iii) keep insoluble those nondissolving phases, (iv) reprecipitate new phases, and (v) produce a final equilibrium aqueous phase concentration of dissolved constituent close to the actual leaching test data. In effect, MINTEQA2 was asked to predict the observed leaching behaviour of solid phases determined by various spectroscopic methods such as X-ray powder diffraction (XRPD) or X-ray photoelectron spectroscopy (XPS).

623 Figure 15.5 Sorption/Desorption Processes on Ash Surfaces for Elements Exhibiting a Weak pH Dependence

o

.07

Pb

11 10

i i

1

.4

1

.01

.3

i

o

4

.02

0

E o.

i

5

2

pH

"O

i

3

o "o. - o .

.03 I

6

Ag

E .05 a..04

i

i

8 7

o

.06

Q

9 E (3. o.

!

.08

13 12

" i ~ " "O

i

1

~) 1

2

3

4

5

6

7

8

Mo

0 1

9

!

pH

,

I

i

9

i

,

pH

,o

E.1 Q. Q.

10 .09

0

.08

1.4

Be

.07

1.0

,O

.04

pH

1.1

i 9

.05

Ba

12

.06 E

O

1.3

9O

.9

.03

9

.8

02

E Q. Q"

01 , 1

I 2

3

4

. . . . 5 6 7

8

9

Sb

o

.7

9

.6 5

E),

.4

9

.3 .2

,Q "a.,~

-,.3,.b oo

"o

o

.1 1

2

3

4

5

6

7

pH

8

9

pH

Solid Line = Model-Predicted Distribution After Dzombak et al., 1992 with permission of author Suspected major mineral phases and their estimated molar concentrations were inputted into MINTEQA2 along with very low concentrations (1 x 10 15 molal) of all component species. Table 15.5 shows the inputted information. The databases to MINTEQA2 had to be adjusted to include new phases (K2ZnCI4, CdsCIAsO4). Inverse solubility products were calculated from thermodynamic or solubility data for modification of the Thermo.dbs and Type6.dbs databases.

624 Figure 15.6 Solid Phase Control of Bottom Ash Leaching as a Function of pH Cd ,..,

Cu

10

IO0 9

.EE

1

g

o.1

i o u

0.01

o

o

O

A.V~-| . I,.i~- I (}

9

AVI-1.L / S - 5

9

AVI-I, L/S-2 AVI-2.U3-10 AVII-2. US-2 OTAV1TE

0.001 0

2

4

0

8

10

12

14

~=

lO

g

1

.e ;

0.1

1=

o

AVI-2.US-tO

A

AVI-2,US'2

z

AVI-I,L/S,rS. ~r~lmn 55O"C TENORITE

0.0001

I~..d(OH)2

0

2

4

6

8

10

12

14

"

pH

Pb

- TENORITE.rood. Cu(OH):Z=q

Zn

10 =

x

AVI.I. L/S-2

o

0.001

pH

.E

x

AVI-I. L/S:IO AVI-1. US=,5

9

,,,3, ~,,..ef~i ,

0.01

9 9

1000

1

9

0.1 0.01 o

g o.ool

u

0.0001 0

2

4

6

8

10

12

14

A V I - I . IJ~-~10

9

AVI-I.

9

AVI-1. I./~a2

~'

oD

100

lO

~ e

~

o.ool

AVI-2. IJS-2

uo

0.0001 0.00001

Pb(OH)2

~ , ~ ~'

0.1 0.01

AVl-2.I./S- 10

9

A~-I, L~-I0

9

AVt,I. I.J~2

*

AVI-I. US,.5

O

AVI*2.L~-IO

A

0

.

:

~

:-

,

,

,

2

4

6

8

10

12

14

AVI-2.US-2 ZIN~TE

'

Zn,~03

pH

After Comans et al., 1993 Table 15.5 Solid Phase Approach to MINTEQA2 for ESP Ash Phase Source K2ZnCI4

XRPD, XPS

Moles/L" 3.53 x 10 .3

NaCI (Halite)

XRPD, XPS

5.50 x 10 .3

CaAI2Si208 (Anorthite)

XRPD, XPS

6.04 x 10 .4

CaCO3 (Calcite)

XPS

3.86 x 10 .3

XRPD, XPS

7.42 x 10 .4

Pb302SO4

XRPD, XPS

4.00 x 10 .4

CdsCIAsO4

XRPD, XPS

5.00 x 10 .6

CaSO4.2H20 (Gypsum)

XPS

3.85 x 10 .3

MgSO4~

(Epsomite)

Pbs(PO4)3CI (Chloropyromorphite)

XRPD

8.62 x 10 s

KAI(SO4)2 (K-Alum)

XRPD, XPS

4.00 x 10 4

PbSO4 (Anglesite)

XPS, XRPD

7.12 x 10 .4

ZnO

XPS, XRPD

8.80 x 10 .4

KAISi308 (Microcline) XRPD, XPS 8.81 x 10 .4 a Based on stoichiometry and total composition data, the model inputs the total moles of a mineral and assumes a volume of one litre. After Eighmy et al., 1993

625 The results of the modelling effort are presented in Table 15.6. Reasonably good agreement was seen between model-predicted equilibrium aqueous phase total concentrations and the analytical data. The lack of perfect agreement most likely reflects the presence of amorphous phases for CaCO3 or CaSO4 o2H20, or solid solutions for aluminosilicates with activities or solubilities that are not amenable to modelling with the present database in MINTEQA2. The solubility of gypsum had to be increased slightly and kaolinite had to be excluded to allow the program to converge to equilibrium. This usually took more than 400 iterations. Gypsum, chloropyromorphite and anglesite were the only finite solids to remain as solids. All the others, including NaCI and K2ZnCI4, dissolved in general agreement with XPS and XRPD data. Quartz (SiO2), diaspore [AIO(OH)] and alunite [KA13(SO4)2(OH)e] precipitated out. All six of the solids presented at equilibrium were detected with either XRPD or XPS in the leached ash. The final mass distributions of all solids were close to the insoluble fraction determined for the residue. The final equilibrium pH was close to the experimental value (4.69 vs. 4.00) and the cation-anion differences were virtually balanced (0.04 % difference). Table 15.6 Verification of Solid Phase Approach for ESP Ash Constituent Leaching Test Data MINTEQA2 Output (mg/L) (mg/L) AI

7.56 + 1.48

0.02

H4AsO4

0.815 + 0.60

0.72

Ba

0.14 + 0.01

not inputted

Ca

281.5 + 3.75

178.7

Cd

4.50 + 0.05

2.80

CI

620 + 0.0

694

Cu

4.92 + 0.05

not inputted

K

406.5 + 0.71

271.0

Pb

4.60 + 0.35

8.35

Mg

18.41 + 0.49

18.0

H4SiO4 SO42

107.4 + 0.15 890 + 10.0

9.40 424

Na

117.5 + 0.1

126.4

Zn 221.7 + 1.70 230.8 a All constituents in aqueous phase were entered at 1 x 10-1~M, except for NO3 and H§ (added at 1.6 x 10.3 M) as acid used to drop pH to 4.0 for the test. Phosphate was also added at low concentration (8.0 x 10.6 M) based on ICAP, SEM/EDS, and STEM/EDS analysis. After Eighmy et al., 1993

626 15.1.9 Modelling Solid Phase Control of Leaching in Dynamic Flow-Through Systems Gardner (1991) and Theis et al. (1992) employed small minicolumns to look at the dissolution of a variety of bottom ashes and fly ashes. A large throughput of leachant through the minicolumn (Figure 15.7) allows for high seepage velocities and high Peclet numbers in the system. The number of pore volumes pumped through the minicolumn was quite high (sometimes more than 1,000). Figure 15.7 Apparatus Used for Minicolumn IExperience

pH M e t e r pH Probe ~ , ~ ",&~

X=/..ORP Probe ~ ~ Flow-through Cell

Column

Fraction Collector

* T Reservoir

Pump

$

1/8" NPT FITTING

m

i.u O z _J D.

TOP CAP 1/8" BORE 10.8 X 2.6 mm CAPFE O-RING

THREADED REACTOR CYLINDER I

11/4"

I

BOTTOM CAP

I

21/4"

1/4" NPT I

3/4 X 1/8" VITON O-RING

After Gardner, 1991 with permission of author

627 The researchers measured minicolumn effluent pH and Eh. Periodically, effluent leachate samples were quantified for dissolved constituents. The geochemical code HYDRAQL was used to calculate the activities of the dissolved leachate constituents using the saturated solution approach, lAPs could then be calculated and compared to published solubility products. The data generated for fly ash (Figure 15.8) show the leachate Zn 2§ and SiO32 activity products fall within the region of available solubility products reported in the literature. The utility of the geochemical code in this application was to calculate activities and allow comparison of lAPs to solubility products. In this case, Zn 2§ solution phase activity and concentration were controlled by ZnSiO3. Figure 15.8 Log SiO32-versus pH as Compared to Theoretical ZnSiO3 Solubility Plots

r~

r"

-10

u~

g, -15

-20 --

i

I 6

i

I 8

i

I 10

i

I 12

After Gardner, 1991 with permission of author

15.1.10 Modelling Field Leaching Behaviour Fruchter et al. (1990) collected pore water samples from a combined bottom ash and fly ash disposal site. Ceramic cup suction lysimeters were used to collect leachates with depth in the disposal mass. The L/S ratio that the ashes were exposed to was considered to be less than one.

628 The collected pore water leachates were analysed for elements and anions. The analytical concentrations, along with pH, were input into MINTEQA2 using the saturated solution approach so the model could calculate ionic strength, ion speciation, single ion activity coefficients, single ion activities for aqueous solute species, ion activity products and saturation indices for mineral solids. Activity-pH or activity-activity plots were used to identify the solid phases controlling pore water chemistry. The results for aluminum, iron and calcium are shown in Figure 15.9. For aluminum, both AI(OH)3 (crystalline) and amorphous AI(OH)3 appear to control aluminum activity in the pore water. For iron, amorphous Fe(OH)3 appeared to play a limited role in controlling iron activity in the pore water. For calcium, gypsum rather than anhydride was the controlling phase.

15.2 DYNAMIC MULTICOMPONENT MODELS Many of the leaching modelling efforts described in Section 15.2 examine the role of precipitation/dissolution, sorption, and solution phase complexation under discrete conditions of L/S, pH and redox. As such, they mechanistically describe leaching behaviour as "snapshots" within the potential leaching spectrum. Frequently, the leaching tests used to generate data to be input into the model are done under agitated conditions so mass transfer limitations are reduced. A second generation approach to modelling leaching behaviour has been developed that incorporates, or can incorporate, elements of dissolution, reprecipitation, sorption, desorption and mass transfer phenomena on residues subjected to leaching under realistic seepage velocities. The dynamic multicomponent model developed by Theis and coworkers (Theis et al., 1992) is particularly attractive to this second generation approach because of the mechanistic basis it employs, the ease with which parameter estimation can be used to refine the parameters and the predicative capability the model employs if minicolumn leaching data over sufficiently large L/S ratios are used to challenge the model.

15.2.1 Dynamic Multicomponent Models Under many field leaching scenarios, a leachant is flowing through and/or around residue material. Accurate modelling requires the coupling of chemical behaviour with transport behaviour. Since field leaching involves both phenomena, a multicomponent model is needed. Transport and chemistry model components ultimately are coupled using source/sink terms of the time- and space-dependent mass balance equations for each chemical component. Simple partitioning coefficients describing dissolution/ reprecipitation and sorption/desorption are used in both the transport and chemistry components of the model (Theis et al., 1992).

629 Figure 15.9 Solid Phase Control of AI, Fe and Ca in a Field Leaching Scenario

,I

o S~to AB o Site CO ~ Composite ~ Construction Layers o Leac~ato

-6 ~ " ~ ~- ~ ~

-~. -26 -

-30

5

~

I 7

~

I 9

pH

11

-10 -12

-14

"~,,~o

-16 -18

~.~ .~.~

o Extracted Pore Fluid

OOOo

-24 -26

-28 -30 5

,

I

7

i

L

9

pH

11

a S~to~ o Sile CO Coexx~.'te , 9Comnrucz~oeLayers

~C~

-2--

0

I.R,lcl~ato

.3 B

~.o

After Fruchter et al., 1990

! -2.5

! -2-0 (SO2,1

"-1.5

630 The minicolumn leaching apparatus forms the basis for developing a dynamic multicomponent model. The device, shown in Figure 15.7, is modelled as a plug flow reactor with non-steady balances on residue components derived in the axial direction of the column (Theis et al., 1992): u

o3Ui

+

-~-

(l'erm 1)

where Ui

=

Pj =

Cs i "Cp =

t X

=

Np = Nc

-

~i E:

= --

V

=

~

j-~

Ad -

~

( T e r m 2)

--V

~

~

( T e r m 3)

+

~--L~-]

( T e r m 4)

+

(15.1)

( T e r m 5)

the total soluble concentration component [i] that is mobile, the total concentration of precipitate [j] that is immobile, the concentration of available component [i] on the ash particle surface, the average concentration of component [i] in the interior matrix of the ash particle that can be released by diffusive flux, time, axial distance, the number of discrete precipitates in the ash, the number of discrete components in the ash, the stoichiometry of component [i] in precipitate [j], bed porosity or porosity of the ash particles in the minicolumn, and interstitial or seepage velocity of leachant flow in the column.

The two terms on the left-hand side of equation (15.1) describe the change in mobile constituents (Term 1) and immobile precipitates (Term 2) with respect to time. Term 2 relates to fast dissolution and reprecipitation. The terms on the right-hand side of equation (15.1) describe the change in mobile constituents with respect to axial distance (Term 3), the change in leachable constituents on the surface of the ash particles with respect to time (Term 4) and the change in particle interior constituents with respect to time (Term 5). Terms 4 and 5 relate to sorption/desorption and slow dissolution, respectively. Mass balance constraints are placed in context with equation (15.1 ) that constrain mass over the iterative process that is used to provide time and space solutions to equation (15.1). The mass balances are: O~Csi

at

aCp c~

where a

=

k~=

De = ap

=

the the the the

= a K c (Cs~ - u~) f o r i -- 1 to N c, and

(15.2)

_ n2De ~ (Cri - ui) for i = 1 to N c Rp

(15.3)

ash particle specific area (area/volume), mass transfer coefficient, effective diffusion coefficient, and ash particle radius.

631 Equation (15.2) is a mass balance constraint on constituents leaving or going to the ash particle surface. Equation (15.3) is a mass balance constraint on constituents diffusing out of the ash particle itself. A chemical mass action constraint is also used to constrain equation (15.1 ) with regard

to reaction stoichiometry as the iterative process is used to provide time and space solutions. The chemical mass action constraint is

Kj

sO

Nc

= ~

[F(u i, u 2..... UNc, Aij ] for j = 1 to Np

i=1

(15.4)

where

SO

Kj

partition coefficient for precipitation/dissolution or sorption/desorption, and mass fractions for components involved in the partition.

F

Finally, boundary and initial run conditions are needed to start the iterative solution process and further constrain equation (15.1). These conditions are: u~(x, t=0) = 0

for

i = 1 to N o

(15.5)

t) = 0

for

i = 1 to N o

(15.6)

Pj(x, t=0) = 0

for

j = 1 to Np

(15.7)

Pj(x=0, t) = 0

for

j = 1 to Np

(15.8)

Vi(x=O

Csi(X ,

,

t=O)

= Csi 0

for

i = 1 to N c

(15.9)

Cpi0

for

i = 1 to N o

(15.10)

for

i = 1 to N o, and

(15.11)

i = 1 to N c

(15.12)

m

C,~(x t=O)

=

tI

1 Csi~ > Vss ~o q " utidt

1

Cpi~ > ~ss where Ut i

--

M s

=

q

=

t2

/

q " utidt

for

ti

column leaching curve from the minicolumn experiment for c o m p o n e n t [i], the volume of ash in the column, and the flow rate.

632 15.2.2 Modelling Solid Phase Dissolution Theis et al. (1992) used their approach to model minicolumn leaching experiments for a number of MSW incineration residues. Results for a soluble constituent (potassium) from an ESP ash are discussed further. The data are shown in Figure 15.10. Potassium in the ESP ash is readily soluble, it initially appears in high concentration in the minicolumn leachate and declines precipitously over time without any reprecipitation. The column was operated over 30,000 pore volumes during the one hundred-hour run. The model simulated the observed data quite well. Figure 15.10 Potassium Leaching and Modelling from Minicolumn Experiments

o

POTASSIUM CONC. VS TIME ak c = De

1.27E-3 t

1.18E-3

= 2.03E-8 *'- g.15E-9

l/sec cm ^ 2/sec

_.r -J'~r ~ - I

E Dmm,-1e,~,"i o O

Q --" 0.1 mL / min EXP. DATA

d

00 i

-3.5,o

M O D E L SIMULATION

" 16oo

2oo0

[]

3o'o0

4o'o0

TIME (in minutes)

[]

sdO0

6000

After Theis et al., 1992 with permission of the Air and Waste Management Association 15.2.3 Modelling Solid Phase Reprecipitation and Solubility Control The data for lead are shown in Figure 15.11. Lead in the ESP ash is less soluble than potassium. It appears in an initially high concentration in the minicolumn leachate and declines over time. The model simulates soluble lead in the leachate quite accurately.

633 However, in the case of lead, at two different positions in time and space, two solids are predicted to precipitate out and control lead dissolution. The first solid to appear is anglesite (PbSO4) early in the column run. The second, lead hydroxide [Pb(OH2)], controls later in the run. In support of these observations, earlier pC-pH plots of lead data from the column showed aqueous phase lead to be in equilibrium first with anglesite and then with lead hydroxide. Figure 15.11 Lead Leaching and Modeling from Minicolumn Experiments

LEAD CONC. VS Ti M E

-3.5

-4~ -4.5~"

m <

(s)

Q = 0.1 m L / m i n

-6-

(~ -6.5_.1

MODEL SIMULATION

/ / ..F. = 3.67E-5

-7De

-7.5 -8

EXP.DATA

I

-5-

-~ -5.5-

E

•/PbSO4

0

looo

20'00

=

+_. 3.33E-5

2.15E-8 ~

3000

9.33E-9

40'00

1/sec cm ^ 2 / s e c

sdoo

6000

TIME (in minutes)

After Theis et al., 1992 with permission of the Air and Waste Management Association The utility of the Theis model is that (i) many residues can be easily modelled, (ii) the basic tenets of equations 15.1 to 15.3 can be modified to include other partitioning reactions as well as soluble phase partitioning, and (iii) the model can be used to predict behaviour over geologic time (e.g. thousands of years). 15.3 FUTURE DIRECTIONS IN MODELLING

It is likely the next decade will see a better melding of ash solid phase and leaching solution speciation data with both equilibrium and dynamic multicomponent models. The speciation data on ash solids will allow better quantification of initial soluble

634 phases, insoluble phases, geochemically unstable phases, sorptive surface areas and surface adsorption site densities. The speciation data on solution phase ligands will allow for better quantification of binding constants in solution. Improvements in understanding sorption are also expected. All of these refinements in the knowledge base of ash will allow for more accurate modelling using both approaches. In fact, the dynamic multicomponent approach can be modified to include more components that exhibit equilibrium-like behaviour in the context of dynamic leaching. The multicomponent model is likely to be used for more predictive capabilities under various geochemical scenarios with ash management, treatment or disposal. Recent developments by EPRI have resulted in the creation of a dynamic-empirical leaching code for fossil fuel waste leaching, denoted as FOWL (Hoestetler et al., 1988). FOWL uses empirical observations on solubilisation as a function of pH as well as estimates of total available fractions to predict leachates generated from geometricallydefined ash masses. Additionally, the geochemical code ECHEM, a modification of the original MINTEQ, is the basis of a coupled leaching-hydrologic transport-geochemical reaction dynamic multicomponent model termed FASTCHEM (Morrey, 1988). ECHEM also contains dissolution kinetic rate constants for a number of relatively insoluble minerals. The approaches used in these two EPRI efforts are excellent bases for developing similar approaches with MSW residues. Finally, Batchelor and Wu (1993) have made extensive modifications to MINTEQA2 to model leaching from monolithic material. Their model, SOLTEQ, incorporates Pitzer-based activity coefficients and cement chemistries.

REFERENCES Allison, J.D., D.S. Brown and K.J. Novo-Gradac. MINTEQA2../PRODEFA2, A Geochemical Assessment Model for Environmental .Systems version 3.0 User's Manual. Environmental Research Laboratory, U.S. EPA, Athens, GA, 1990. Ball, J.W., E.A. Jenne and M.W. Cantrell. WATEQ3; A Geochemical Model with Uranium Added. U.S.G.S. 81-1183, U.S. Geological Survey, Menlo Park, CA, 1981. Batchelor, B. and K. Wu. Effects of Equilibrium Chemistry on Leaching of Contaminants from Solidified/Stabilized Wastes. In Chemistry and Microstructur.e of Solidified Waste Forms. Edited by R.D. Spence. Lewis Publications, Boca Raton, FL, 1993. Comans, R.N.J., H.A. van der Sloot and P.A. Bonouvrie. Geochemical Reactions Controlling the Solubility of Major and Trace Elements During Leaching of Municipal Solid Waste Incinerator Residues. Proceedin.qs of the 1993 Internat.i.onal Conference on Municipal Waste Combustion, March 30-April 2, Williamsburg, VA. Edited by J. Kilgroe. AWMA, Pittsburgh, PA, 1993.

635 DiPietro, J.V. Municipal Solid..Was..te C0mbustionResidue Leachate Composition: The effect of Selected Environmental Parameters and an Assessment of Geochemical Modelling Predi..c..ti0ns. Masters Thesis, University of New Hampshire, Durham, NH, 1989. DiPietro, J.V., M.R. Collins, M. Guay and T.T. Eighmy. Evaluation of pH and OxidationReduction Potential on Leachability of Municipal Solid Waste Incineration Residues. Pro_ceedin.as of the !.nternational C_onfere.nce on Municipal Waste Combustion, April 1114, Hollywood, FL, 2B, p. 21, 1990. Dzombak, D.A. and F.M.M. Morel. Surface Complexation...Mode!ing, J. Wiley and Sons, N.Y., 1990 Dzombak, D., F. Morel, J. Mundt, E. Hirschberg and G. Huff. Modelling the Leaching of Metals from Hazardous Waste Incineration Ash. _ Proceedings of the 1992 Incineration Conference, May 11-15, Albuquerque, NM, 1992. Eighmy, T.T., D. Domingo, J.R. Krzanowski, D. St,~mpfli and D. Eusden . The Speciation of Elements in Incineration Residues. In Proceedings of the...1993 Municipal Waste Combustion Conference, March 30 - April 2, Williamsburg, VA. Edited by J. Kilgore. AWMA, Pittsburgh, PA, 1993. Erikkson, G.A. An Algorithm for the Computation of Aqueous Multicomponent Multiphase Equilibria. Anal. Chem. Acta 112, pp. 375-383, 1979. Felmy, A.R., D.C. Girvin and E.A. Jenne. MINTEQ-A .....Computer P.ro~rarn for Calculatin.q Aqueous Ge0chemic.al Equilibria. EPA-600/3-84-032, U.S. EPA, Athens GA. NTIS, Washington, DC, 1984. Fruchter, J.S., D. Rai and J.M. Zachara. Identification of Solubility-Controlling Solid Phases in a Large Fly Ash Field Lysimeter. Environ. Sci. Technol. 24, pp. 1173-1179, 1990. Gardner, K.H..Characterization of.Leachates from Municipal In.cinerator Ash Materials. Masters Thesis, Clarkson University, Pottsdam, NY, 1991. Hoestetler, C.J., R.L. Erickson and D. Rai. The .F.ossil Fuel Combustion Waste Leachin.cl ('F.OWLTM) Code Version 1, User's Manual. EPRI EA-5742-CCM, EPRI, Palo Alto, CA, 1988. Ingle, S.E., M.D. Schuldt and D.W. Shults. A User's .quide for REDE.QL.EPA. A Computer ProQram for Chemical Equilibria in, Aqueous Systems. EPA 600/3-78-024, U.S. EPA, Corvallis, OR. NTIS, Washington, DC, 1978.

636 Kincaid, C.T., J.R. Morrey and J.E. Rogers. Geohydrochemical Models of Solute _.Mi,qration, Volume 1 Process. Description and .Computer Code Selection. EPRI EA3417, EPRI, Palo Alto, CA, 1984. Kincaid, C.T. and J.R. Morrey. Geohydrochemical Models of Solute Mi.qratio.n., Volume 2: Preliminary Evaluation of Selected Computer Codes. EPRI EA-3417, EPRI, Palo Alto, CA, 1984. Krupka, K.M., R.L. Erikson, S.V. Mattigod, J.A. Schramke, and C.E. Cowan. Thermochemical Data Use the FAST.CHEM Package. EPRI EA-5872, EPRI, Palo Alto, CA, 1988. Morrey, J.R., C.T. Kincaid and C.J. Hoestefler. Geohydrochemical Models for Sol..ute Mi.qration, Vol. 3: Ev.a.luationof Selected Computer Codes. EPRI EA-3417, EPRI, Palo Alto, CA, 1986. Morrey, J.R. FASTCHEM TM Package Volu.me 4: User's Guide to the ECHEM Equilibrium Geochemistry Code. EPRI EA-5870-CCM, EPRI, Palo Alto, CA, 1988. Nordstrom, D.K. and J.W. Ball. _Chemical Models, Computer ProQrams and Metal Complexation in Natural Waters. Edited by C.J.M. Kramer and J.C. Duinker. Martinus Nijhoff/Dr. J.W. Junk Publishing Co., Dordrecht, The Netherlands, 1984. Perkins, E.H., Y.K. Kharaka, W.D. Gunter and J.D. DeBraal. Geochemical Modelling of Water-Rock Interactions Using SOLMINEQ.88. Chemical Modellin.q of Aqueous Systems II. Edited by D.L. Melchior and R.L. Bassett. ACS, Washington, DC, 1990. Rai, D. and J.M. Zachara. Chemical Attenuati0.n..Rates, Coefficients and Constants. in Leachate MiQration Volume 1 A Critical Review. EPRI EA-3356, EPRI, Palo Alto, CA, 1984. Theis, T.L., R. lyer and K.H. Gardner. Dynamic Evaluation of Municipal Solid Waste Ash Leachate. Emer.qinqTe..chnolo,qiesfor_Hazardous Waste Mana.qement. Edited by D.W. Tedder. American Chemical Society, Washington, DC, 1992. van der Sloot, H.A., Leaching Behavior of Waste and Stabilized Waste Materials; Characterization for Environmental Assessment Purposes, Waste Man. Res., 8, pp. 215-228, 1990. Westall, J.C., J.C. Zachary and F.M. Morel. MINEQL A Computer ProQram for the Calculation of Chemi..cal Equilibrium Comp0sitio.n.of Aqueous Systems. MIT Tech. Note 18, MIT, Cambridge, MA, 1976.

637

CHAPTER 16- LEACHING DATA

16.1 INTRODUCTION 16.1.1 Overview The leaching of contaminants into the environment is the major concern associated with the management of incinerator residues (Chapter 1). Many types of leaching tests have been used for regulatory and research purposes to evaluate the leaching behaviour of residues. It is the apparently highly variable results from these different tests, and the misinterpretation of these results, which has resulted in a substantial portion of the controversy over incinerator residues. With the exception of research programs, the majority of leaching data that has been generated globally has been based on regulatory leach tests. These tests generally control the laboratory extraction conditions to simulate the various leaching conditions anticipated at disposal sites, and test results are compared to specified performance criteria. Consequently, ash has been labelled a "toxic" material without the benefit of understanding the intrinsic properties of the ash, the underlying mechanisms causing a particularly high or low release, nor an understanding of how the disposal or utilisation scenarios for the residues may effect environmental impact. The information presented in this chapter builds on the information presented in the previous four chapters and focuses on contaminant leachability data. Test results from a wide variety of test conditions have been compared to provide a uniform basis for data interpretation. The large amount of information collected to date allows a more rigorous treatment of data to generate a generic picture of residue leachability. The data evaluated was obtained from tests on ash collected at MSW incinerators in many different countries and facilities of varying design. Thus, the data set represents a global range of residue properties. 16.1.2 Data Sources

Bottom ash represents the bulk of the total residue stream generated by an incinerator (Chapter 8) and is also the most heterogeneous stream (Chapter 9). Leaching considerations are important to both disposal and to the utilisation of bottom ash in construction applications. In this chapter, results obtained by a wide variety of test methods applied in different jurisdictions have been collected to generate a better understanding of bottom ash leaching characteristics. Although there is a substantial amount of leaching data on bottom ash presented, it is important to recognise that "bottom ash" samples are generally a mixture of bottom ash, grate siftings and sometimes heat recovery system ash (dependent on the configuration of the incinerator). The presence or absence of these individual ash streams in a "bottom ash" sample may effect the leachability of some elements because of differences in the chemical characteristics of these streams. In addition, the refuse

638 feed stock for a particular incinerator facility may include or exclude components normally found in MSW, and thus have an effect on the leachability of elements in ash. The generation, collection and composition of different types of APC residues has been discussed in detail in Chapter 4 and 11. The data presented in this chapter represent data on different air pollution control (APC) residues including filter ash (electrostatic precipitator ash and fabric filter ash) and residues from acid gas cleaning (dry scrubbing, semi-dry and wet systems) (Born, 1993; Eighmy, 1993, Hjelmar, 1993, Environment Canada, 1993). The limited information which is available for heat recovery ashes is gleaned from some of these same studies. 16.1.3 Data Transformation

Leaching test results can be expressed either as leachate concentration (mg/I) or as constituent release (mg/kg residue). The basis selected for expressing leaching results should be based on the type of data comparison which is desired. Regulatory test results most often are expressed as leachate concentrations for comparison to performance criteria values (e.g., groundwater, drinking water), but do not consider the underlying basis for release phenomena which are observed. Results expressed as a concentration permits a comparison of contaminant solubility which reflects the chemical speciation of the elements and leaching solution conditions (e.g., pH). Transformation of measured concentrations into units of release is necessary for comparison of the data obtained at different liquid-to-solid (L/S) ratios and determination of availability. Release is defined as the ratio of the mass of a contaminant dissolved to the mass of residue leached (Environment Canada, 1986). Currently, there is a shift in some jurisdictions toward regulatory limits based on release, rather than leachate concentration, for evaluation of environmental impacts. Release fluxes (e.g., mg/m2/yr) describe release of a constituent as a function of geometric surface area and time. Much of the data presented in this chapter has been transformed from its originally reported basis into either leachate concentrations or release to facilitate valid comparison of the data. 16.2 TOTAL SOLUBLE FRACTION AND AVAILABILITY OF ELEMENTS 16.2.1 Total Soluble Fraction The total soluble fraction of a residue is an important consideration for evaluating potential groundwater impacts from disposal and the physical and environmental suitability for utilisation. Bottom ash is not highly soluble in water. Results from batch tests at a liquid-to-solid (US) ratio of 20 indicate that approximately 6% of bottom ash (including grate siftings) from mass bum and RDF incinerator systems can be readily solubilised in water. The majority of the constituents solubilised from bottom ash are potassium, sodium and

639 calcium chlorides and sulphates (see Chapter 11). This data is corroborated by data from column studies which indicate that the major components of percolated leachate are readily soluble salts which are flushed from the columns within 1 - 2 pore volumes without any noticeable retention. Bottom ash from two-stage systems appears to be slightly less soluble (3%). The reduced total soluble fraction most likely results from the higher content of uncombusted material which acts to dilute the salt concentrations remaining in the bottom ash. In addition, the char can act as a sorptive medium for potentially soluble metals. After conversion to mass leached relative to total mass of waste input, the overall quantities of total salts leached from two-stage, mass burn and RDF incinerators are similar. The high total soluble fraction of APC residues is a critical factor in management of these materials. In contrast with bottom ash, the total soluble fraction of APC residues, particularly scrubber residues, range from 30 to greater than 65%. This high solubility also is a major factor of environmental concern. In leaching experiments, the resulting high concentrations of dissolved can cause analytical problems (see Chapter 7).

16.2.2 Availability Availability for leaching is defined as the maximum quantity or soluble fraction of a residue constituent that can be released into solution under aggressive leaching conditions (NEN 7341, 1994). These conditions, in theory, should provide an estimate of the maximum mass of material that could leach under a 1,000 to 10,000 year time frame, except for mobile species such as highly soluble salts (e.g., sodium chloride) for which the availability can be reached in a matter of years. Under availability controlled conditions, the resulting solution is at a concentration less than the saturation condition for the element or species of interest. For example, availability in bottom ash typically excludes elemental species which are tightly bound in glassy matrices and in geologically stable mineral forms such as Si in SiO2 (quartz), Ca in Ca2AI2SiO7 (gehlenite), and Mg in MgCa2Si207 (&kermanite). Thus, the availability of a specific element can be significantly less than the total content of that element (e.g., Pb) or may be approximately equal to the total content (e.g., CI). Determination of availability, however, does not indicate whether or not this maximum quantity of a particular constituent will be released, or over what precise time interval the release will occur for the environmental exposure scenario of interest. NEN 7341 (formerly NVN 2508, 1987) was developed to quantify the availability of constituents in inorganic wastes and combustion residues. This leaching procedure is based on a pH-controlled extraction at pH=7 and 4, successively, using a total liquid to solid ratio (L/S) of 100. These conditions generally prevent solubility limitations during extraction.

640 The data obtained from the availability test assessed is not directly applicable for evaluation of actual environmental impact because the entire available fraction may not be released under specific environmental exposure scenarios or in realistic time scales. However, it does represent the mass transfer driving force for leaching and has been proven useful in modelling. (NEN7345, 1994). One exception to this is with respect to some amphoteric metal compounds, particularly Pb-based compounds. In some instances, the availability test data can underestimate the quantity of metals, such as lead, which are available for leaching under highly alkaline conditions. Table 16.1 summarises the ranges for total composition and availability for bottom ashes from USA, Canada, Denmark, Germany, Sweden and the Netherlands (Chapter 9). The availability for several constituents (e.g. Si, AI, Cr) is an order of magnitude less than the total content. When the availability of a specific constituent is very low with respect to the potential for environmental impact, further evaluation of the release of that constituent from environmental point of view is unnecessary. Availability results indicate that B, Ca, Cd, CI, Cu, Pb, Mo, SO4 and Zn are important constituents in bottom ash which warrant further evaluation. Table 16.1 Ranges of Total Content and Availability.for Bottom Ash Element Total content (mg/kg) Availability(mg/kg) Min Max Min Max M~or

Fraction available (-) Min Max

Ca CI K

50000 1000 7000

90000 3000 20000

20000 1000 1000

70000 3000 4000

0.4 1 0.14

0.8 1 0.2

Mg SO4 Trace As B Ba Cd Cr Cu Hg

10000 12000

30000 30000

1000 8000

6000 18000

0.1 0.6

0.2 0.7

5 80 500 2 200 1200 0.5

40 300 1800 25 1000 2500 1

0.3 50 50 0.5 2 50 0.01

5 200 200 5 10 200 0.1

0.06 0.6 0.1 0.2 0.01 0.04 0.02

0.13 0.7 0.15 0.25 0.01 0.08 0.1

Mo

5

30

1

4

0.1

0.2

Pb Sb Zn

1500 30 2000

3000 200 4000

50 1 50

300 2 500

0.03 0.01 0.03

0.1 0.03 0.13

641 Using a similar approach as for bottom ash, the availability of constituents from APC residues has been determined (Kosson et al., 1993; Versluijs et al., 1990; Hjelmar, 1993; Whitehead, 1990). Table 16.2 presents data on total composition and availability for filter ashes, scrubber residues and combined APC residues. In general, the availability is high (50%) for Pb, Zn, Ca, Mg and approaches 100% for several constituents (Cd, Na, K, Ca, CI, SO4). This indicates a potential risk posed by this type of residue if disposed in a landfill without adequate design considerations. Table 16.2 Ranges of Total Content and Availability for APC Residues Element Total content (mg/kg) Availability (mg/kg)

Fraction leached (-)

Min

Max

Min

Max

Min

Max

Ca

50000

200000

50000

100000

0.5

1

CI

8000

60000

80000

60000

1

1

K

20000

40000

10000

25000

0.5

0.6

Mg

10000

30000

4000

15000

0.4

0.5

SO4

30000

90000

30000

80000

0.7

1

As

30

100

1

2

0.02

0.03

B

30

200

30

150

0.6

1

Ba

100

3000

30

80

0.02

0.05

Cd

100

900

100

900

1

1

Cr

100

800

5

50

0.05

0.15

Cu

300

3000

1

20

0.01

0.05

Hg

5

20

4

10

0.5

0.9

Pb

4000

20000

100

5000

0.1

0.4

Sb

50

950

0.5

1

0.02

0.05

Zn

5000

40000

4000

20000

1

1

16.2.3 Sequential Chemical Extractions The sequential chemical extraction (SCE) procedure was originally derived to identify the association of trace contaminants with particular chemical phases in sediments (Tessier et al., 1979). The extractions are operationally defined and the test results do not necessarily reflect an association with the claimed phases (carbonate phase, iron and manganese bound, organic degradable or reducing phase) as shown by Gruebel et al., (1988). However, under the NITEP Program, a modified SCE procedure was

642 used on incinerator residues (WTC, 1993; based on Fraser and Lum, 1983). Moreover, the emphasis on interpretation of the SCE data shifted away from the associated chemical phases toward association with different leaching conditions within an MSW landfill over time (see Table 16.3). Results are usually expressed as a percentage contribution to the five operationally defined extraction steps. The procedure does provide a qualitative indication of the matrix association of specific elements. In Figure 16.1, SCE results for bottom ash, filter ash and economiser ash from the NITEP study (Sawell et al., 1988) are presented. In bottom ash, substantial fractions of the total element content are found in the fractions considered unavailable for leaching. This agrees with the observations made in the availability test. Table 16.3 Interpretation of the Sequential Chemical Extraction procedure Fraction

Interpretation

A B C D E

Immediately available for leaching Potentially available for leaching under acidic conditions Potentially available for leaching under severe reducing conditions Unavailable for leaching under normal leaching conditions Unavailable for leaching

16.3 SOLUBILITY AND RELEASE OF ELEMENTS

Measurement of solubility requires that sufficient contact time be allowed for the solution and solid phases to approach equilibrium. The most common time frame used in laboratory testing is 18 to 24 hours. Time dependent leaching studies have shown that equilibrium is approached for most constituents after about 24 hours for a particle size of less than 3 mm (Comans et al., 1993). For some constituents (Mg, Zn) equilibrium was not reached within 200 hours. In general, the time required to reach equilibrium may be shortened through reductions in particle size. The most dominant factor influencing the solubility of most elements from ash, especially trace metals, is the pH of the leaching environment. Numerous studies have underscored the importance of the endpoint pH of the leaching medium on the release of trace metals (DiPietro et al., 1989; Sawell et al., 1987, 1988 & 1989; US EPA, 1988; De Groot et al., 1987; van der Sloot et al., 1989, 1991a, 1991b; Kosson et al, 1993). Presentation of leachate concentrations as a function of pH is useful when solubility in the solution phase is limiting release (solubility control). This presentation also can be used to visualise the effects of complexation reactions of constituents and effects of reducing conditions.

643

Figure 16.1 Sequential Chemical Extraction (SCE) Results for Bottom Ash for Selected Metals

MSWI bottom ash lOO 90 c O .N O r .w cO L_ (i,) 13_

80 70 5o 4o 3o 20 lo

............

o A

C

O

E

100 90 80

A

B

C

O

E

lOO

Cu

9o 80

70

T-

60 50

Ni

7O 5O

40

4O

30

3O

-

20 10

1o

0

o

B

C

D

E

80

A

B

C

D

E

100

100 90

..,.

90

Pb

80

70

70

60

6O

50

50

40 30 20

29

~

10

,o o

0 A

8

C

D

E

Zn

Mmm B

C

SCE fractions

D

E

644 Figure 1 6 1 Continued

MSWI filter ash lOO

100

9o t0 0 t_

C t(1) 0 i.. EL

9o

8o

80

70

70

6o

6O

Cd

5o 40 3o

i

20 ~o o A

B

0 C

D

80

B

C

D

E

B

C

D

E

100

100 90

................... A

E

Cu

70 60

90 80

q--

Ni

70 60

50

50

40

40

30

30

20

20

10

10 0

0 A

B

C

D

A

E

100

100 90 80

90

Pb

80

70

70

60

60

50

50

40

40

30

30

20

20

10

10

Zn

~-_:_-_--_:--_-:-~

0

0 A

B

C

D

E

A

B

E--__----_-_-- ~:._-__-._-.-_~ ............... C

D

SCE fractions

E

645

Figure 16.1 Continued

MSW! economizer/boiler ash lOO

100

9o

90

t0

8o

8O

0 L

70

70

60

6O

C-G) 0 L--

(1) 13--

50

5O

4O

4O

30

3O

20

2O

10

10

0

0 A

B

C

D

80

A

E

B

C

D

E

8

C

O

E

100

100 90

CC

9O

Cu

8o

70

70

60

6O

50

5O

40

4O

Ni

30 20 lO

10

o

0 A

B

C

D

100

100 90 80

A

E

90

Pb

80

70

70

60

60

50

50

40

40

30

30

20

20

10

10

Zn

--F-

0

0 A

8

C

D

E

A

8

C

SCE fractions

D

E

646 16.3.1 Bottom Ash

Unified curves presenting the pH dependent solubility of various elements in bottom ash have been developed based on compilation of the data from a wide variety of sources (Versluijs et al, 1990; Sawell et al., 1987, 1989; Eighmy et al., 1989, Eighmy, 1993; Hjelmar,1991; Kosson et al., 1993; Kosson, 1992; F,~llman et al., 1992; Wahlstr(~m, 1992; WAV, 1988,1992; WASTE Program, 1993; van der Sloot et al., 1987, 1991a, 1991 b, Comans et al., 1993; Sakai, 1995). Data from the following test procedures (see Chapter 14) have been included: Acid Neutralisation Capacity Test DEV $4 Distilled Water Leach Test EP Toxicity MOE Regulation 309 (now 347) Leach Procedure pH Stat Test Serial Batch Tests (NVN 2508, now 7349) Swiss TVA TCLP California WET Test Japanese JLT-13 The data have been selected from studies in which either a wide range of materials was studied or in which particular focus was placed on the pH sensitivity of leaching. The pH related data compiled here does not represent the natural pH range of ash in the environment. Although aged bottom ash can exhibit a neutral to mildly alkaline pH in the environment, acidic to strongly acidic pH's on the solubility curve are strictly a function of the quantity of acid added during the test. It should also be noted that acidic to strongly acidic leaching conditions (pH <4.5) are not normally found in the environment. The results have been presented in the form of aqueous concentration as a function of pH to emphasise solubility control. The data used are all based on extractions at L/S =5, 10 or 20. No distinction was made with respect to incinerator manufacturer or type of grate, because such variations appeared to be of limited importance in comparison with the variability in the feed stock to the facilities. Care has been taken to select only data for bottom ash from mass-burn facilities, but the bottom ash may include grate siftings, depending on the facility and sampling protocol. Data from RDF facilities and other related installations are discussed in Section 16.4.4. The resulting leaching data set was sorted according to increasing pH and obvious outliers were omitted. Figure 16.2 presents the histograms for untransformed and log-transformed Pb and Zn data over a small pH interval. The results clearly indicate that the overall data set should be treated as log-normally distributed. Data over increments of 0.1 pH units were Iog-normalised, averaged and the standard deviation

Figure 16.2 Histograms of Raw (a,b) and Log Transformed Data (c,d) for Zn and Pb from Laboratory Batch Extractions Between pH 11.4 and 11.6

648 was determined. Figure 16.3 presents polynomial curves fit to the mean and +/- 1 standard deviation of data which has been log-transformed and averaged. Coefficients for the polynomial curves are provided in Table 16.4. All leaching data, with the exception of the California WET data and the Swiss TVA data, were omitted during regression. Data for As, Cd, Cr, Cu, Mo, Pb, Ni and Zn are presented. These curves indicate very consistent and well defined pH dependent leaching behaviour for the different elements studied. Table 16.4 Polynomial Coefficients for Unified pH Curves Coeff.

Cd

aO

0.59975

al

Cr

Cu

Mo

Ni

Pb

Zn

1.31282 2.19662

-4.99936

0.30126

2.31369

2.89268

-0.17105

-0.16596

0.29699

0.23601

0.37705

-0.17839

-0.60674

a2

0.26699

0.05870

-0.59716

0.36681 -0.42432

0.08597

0.51484

a3

-0.11127

-0.06533

0.15578

-0.1285

0.15610

-0.05557

-0.14276

a4

0.01573

0.01204

-0.01681 0.01861 -0.02478

0.00736

0.01489

a5

-0.00100

-0.00082

0.00079

-0.00123

0.00170

-0.00036

-0.00070

a6

0.000024

0.000019

-1.3E-05

0.00003

-4.2E-05

6.1E-06

0.000014

R 0.994 0.996 0.990 0.989 0.989 0.988 Conc. (mg/I) = aO + al(pH) + a2(pH)2 + a3(pH)3 + a4(pH)4 + a5(pH)~ + a6(pH) 6

0.995

The derived "unified pH solubility curves" can be used as a basis for evaluating consistency with, or deviation from, typical leaching behaviour. Thus, these curves are used as a benchmark for the interpretation of regulatory leach test results from different jurisdictions, comparison of the effects of different incinerator facility variables (e.g., design, country, operating parameters, etc.), and in modelling the chemical speciation of bottom ash to define the solubility controlling phases. The use of the unified pH solubility curves also may be beneficial for the operators of incinerators as a quality control tool for ash residues, as well as to regulators for defining acceptable and unacceptable behaviour of residues. Control tests can be derived with maximum sensitivity to the parameters chosen for verification of residue quality. Results are presented in Figure 16.3 for a selection of major elements, Ca, Si, AI, Mg, Fe, Mn, and sulphate, and the trace contaminants As, Cd, Cr, Cu, Ni, Mo, Pb and Zn. Solubility of Na, K and CI have not been presented because the solubility of these elements is independent of pH. Arsenic (Figure 16.3a)

The potential leachability of As from bottom ash is relatively small (around 0.1 mg/kg). No clear distinction of its leaching behaviour as a function of pH can be identified as

649

Figure 16.3ab Composition of Worldwide Leaching Data (n=460)of Unified pHSolubility Curves for MSWI Bottom Ash Selected Metals 100

As

r0

O

O

0

0

9

~,

4,

9

9

9

@,

X

8

@

X

9

(a)

X

X

x xx~~ x~oooo(xx

0,1 "~

0,01 4

I

I

i

t

I

I

5

6

7

8

9

10

x

x

I , 11

i

:

12

13

pH

10000 .. --

-

~

~

......... O o " " , OO~

.............................. "~.

1000

~ ,re .,o-,,I 100

0

O

~'

~~

v x -x

...~. " \

......

x x

*''*" ".. . X,~X.~_ '.. "'O . u C e 9 '"'O.

'"..

~

o

o""-.

2.. @

9

---I

1

.........

n n ~ T m

........................ _.~..................... o . . . . . . . . . = . . . . . ,!,',1

Cd

I

na

9

.........

0

(b)

X

". . , . O 0 ~ ' Ou - = "'O. ..... ~~.n

.

0,01

x

x

$ .....

8

0,1

x

X

2

t

I

I

I

I

I

I

I

I

t

3

4

5

6

7

8

9

10

11

12

pH

13

14

650 observed for coal ash (De Groot et al., 1987), in which a maximum leachability is observed at neutral pH. Arsenic is not regarded as an element of concern in bottom ash due to the small percentage of the already low concentrations in the ash (40 mg/kg). Hydrated iron oxide phases, which are present in abundance, are probably responsible for this retention (Van der Hoek, et al., 1993).

Cadmium (Figure 16.3b) The solubility of Cd ranges over more than 4 orders of magnitude in the pH range from 4 to 9 with very consistent behaviour. A possible cause for the wide scatter of data around pH 6 may be the result of complexation of Cd with CI in the ash. Release in the pH range from 0 to 4 reflects the availability of Cd in the bottom ash which is almost the total content. A significant difference in total Cd content may exist between different countries as well as between installations depending on the feed stock of MSW and the combustion efficiency of the incineration (see Chapter 9). Chromium (Figure 16.3c) Although the data set is still limited, the results are very consistent. At present, a validated explanation for the observed behaviour is lacking. In view of the fact that Cr III is less leachable than Cr VI at neutral pH, it may be that the shape of the curve reflects the presence of both species. Cr III would be leachable only at relatively low pH values (pH < 4), while a minor portion of the total Cr present is in the hexavalent state which is leachable (similar to other oxy-anionic species such as Mo). This speciation also would explain the depletion of a soluble Cr species as observed in column experiments (Section 16.2.4). Copper (Figure 16.3.d) The leachability of copper covers a relatively wide range. Cu leaching data obtained from the acid neutralisation capacity test (Kosson et al., 1993), in particular, showed wide variability. The pH static data are more consistent (Comans et al., 1993; Eighmy, 1993). The behaviour of copper has been studied in detail (van der Sloot et al., 1992; Comans et al., 1993) and various options for explaining the leaching behaviour have been presented. A relation between Cu solubility and the presence of organic matter exists. This aspect will be addressed in more detail in section 16.2.3 on chemical speciation of Cu. Molybdenum (Figure 16.3e) The leaching behaviour of Mo is characterised by fairly constant release in the pH range 7 to 13 and a decrease toward pH 4. This is consistent with the leaching behaviour of Mo from coal fly ash (van der Sloot et al., 1989) and may be indicative of a leachability controlling mechanism. Based on the relationship to column test results, it appears that Mo is an element which can be associated to a specific industrial input.

651

Figure 16.3cd 100000

.............. e...Q, e

10000

1000

-i

"-'. U c 0

100

I

''..,-~",.,.........

~o

........... (c)

10

t t Cr

" .........., ~ . ~ 0 ~ ' 9 9

...........

i

0,1

,

t

0

1

'

t

'

I

2

"

t

3

'

I

4

'

I

5

'

I

6

'

7

I

'

8

I

'

0

I

'

I

10

'

I

11

'

I

12

13

pH 1000000

I

/ !

....... o- ........

100000

0

-~oo ............. o ........"',.~ ...

~,,

t~t -1 cO

o

/

....... o

,,,,\

10000

9

-",....\ ',,.o

~o

.

9~ . .

100

o;"..... -

~..U

10

0

~

,......oO',, .

'..

lOOO

,,i

'....

~

~ ~

9

~

, /

e ~~,/ ..:"

- .......t... .~..,.-"~ ~o-,..... o.... ,..o..._...o.....o....o .......

~

\

~

o o"

9

o

"o

"

~" ~ ~

jO

"

~-Z....-g

.......

.............. 9..o....o....~~"'"o,, ,,

',

I

I

I

I

I

I

I

(

I

I

1

2

3

4

5

6

7

8

9

10

11

pH

i .......

o%~i

~

, o ~ O..... - ..... "_........... -.- .................. ;~

o

~

-

0

I

I

I

12

13

14

(d)

652

Figure 16.3ef 10000

-

Mo 1000

9

-

9

_

., .... Q ........... 0 ................................... ~1 HI

f"

....O"'&

_~ l o o -

A

9

li

9

9

q

"1

o

8 . D ~176 ~

"1

9 ..-""'"t

D

Q:!

(e)

rn

B

lo !

1

4

a

I

t

I

I

t

t

I

t

I

5

6

7

8

9

10

11

12

13

pH

100000

10000

&A 9

1000

0"-,. .,.

9

9 o'-... 9 "....

9

9 "-... ~L

T c 0

..

100

9,,, 9

o

n

10

D

--..

9 no

0

"

'

Ni

0,1 0

9

n

~

~

(f)

9 9

~

121

I

t

t

t

t

t

I

t

t

I

I

t

t

I

1

2

3

4

5

6

7 pH

8

9

10

11

12

13

14

653 Relatively high Mo levels occur in only a few specific urban locations, whereas most suburban communities produce waste with relatively low Mo levels. The element is crucial in the evaluation of utilisation in some jurisdictions because of its high mobility. The pH data presented could erroneously lead to a conclusion that leachability decreases at high pH. The leachability appears to be independent of pH in the range 7to 13.

Nickel (Figure 16.3f) The leaching behaviour of Ni is consistent based on the limited data available. The behaviour of Ni shows increasing solubility with decreasing pH, reaching a maximum solubility at pH less than 7. Lead (Figure 16.3g) The leaching behaviour of lead is surprisingly consistent in spite of the documented heterogeneity of Pb in of bottom ash. The amphoteric nature of Pb compounds is clearly evident, which results in increased solubility by several orders of magnitude at pH >11. Several test methods used for regulatory purposes are consistent with the general pH dependent leaching curve indicating the solubility control for Pb is very significant. Zinc (Figure 16.3h) The leaching behaviour of Zn as a function of pH is more consistent than any of the other elements, indicating of a high degree of solubility control. In the pH range below 6, the leachability reflects the amount of Zn available for leaching with the maximum leachability reached between pH 4 and 5. As in the case of Pb, the Zn leaching is characterised by a sharp increase in leachability at high pH due to formation of anionic hydroxide complexes.

16.3.2 APC Residues The release of constituents from APC residues as a function of pH is crucial for the understanding of the chemistry behind the observed release from APC residues. Figure 16.4 presents the solubility of B, Ba, Cd, Cr, Cu, Hg, Mo, Pb and Zn from APC residues as a function of pH. The data were obtained from (Eighmy, 1993; van der Sloot et al, 1992; Kosson et al., 1993; Environment Canada, 1993, Hjelmar, 1993). Characteristic solubility controlled release is indicated although insufficient data was available to develop unified leaching curves.

Boron The leachability of boron from filter ash (without scrubber residue) is consistent with that of bottom ash (Eighmy et el, 1993). At pH greater than 10 the solubility decreases.

654

Figure

16.3gh

1000000 -

.......o .................*...... .

~ ' ~ . o ".... --"......-~'.~"o

1 0 0 0 0 0 --

0

".

.... "'.

\..~..

"o

'

9

,

9

8

x

.o%~.

....,

1000

el

"~..-~..'~

100

10

Pb

0,1 I

0

=

:x

: .

"

"'",.

9

o

0

..... "

.

/

..... I

//: ! ,

~_=~,=.~ :-"

"o .....'- ......':...=..,"

(g)

"n"~

i

f

I

I

I

I

I

i

i

i

I

I

i

I

1

2

3

4

5

6

7 pH

8

9

10

11

12

13

14

..o ..........~ .............. o

~--~..~'~

-..9 ""-.. 41

9

100000

h

o"o~ ", o & ""-,, &

x

il 9 i/

x

/

9 ...~,~ i.., o,~

Xx

oh.

', 10000

9

1000

o (.3 100

/

........

.

"..,.

"....

c

9 .,9

X

9 .

9

I~

I

I

2

I

3

Ii

/

9

.....

c~q-

i

4

I

5

I

6

I

7 pH

I

8

""'-. 0 o ,,.i

I

9

I

10

I

11

(h)

i

.'

./"

9

/

/ i

.

il.

~..,~ ' .".....

10

1

/

X

x

"'...

%.\

Z~

/"/

I

",,

"....

0

9

.

xx /i /

Ipo/

9

,;

.:

,~9

9 ~

x

x

x

....""".,"............ , "..... " ............. ~

"

1000000 ..... o

"u"

9

..\."''., ._

I

x

"~x

9

...9" ~ ..o

1

,,, . x

x

.

u c

i

X

.....

10000

:3.

0

I

o

12

I

13

I

14

Figure 16.4 Comparison of pH Dependent Concentration Data for Selected Metals from APC Residue

656

"0 (D c" .m cO 0

(D L..

LI..

0 0

, ........ |

.,_ 0

9

4

~

q o I........

0 O. 0 0 T--

U u

, ........ u

,..

!;11

.,_ 9

0 w-.

m

0

O0

411-

o

9

41

, ........ I

v-. 0

4

0 ,-0

o

o

B

9

I........

4

ooO~

~uoo

,an c ~ o c : ]

, ........ ,

',-

O/Owl

~........

0 O. ~"

(I/6W) o u o o

I........

0 O. 0 w--

0 0

o

........

r~ , I

4e

c~

I........

,,-O. 0

o o o

0 O. 0

,. . . . . . . . !

@~o0

,. . . . . . . .

9

0 0 0 0

I. . . . . . . . '~-

0

0 O. 0 0 0

,

~~ n

I. . . . . . . .

"-

0 0 0 0 0 0

0 0

4

,........ |

9

~ ~ o

44

Ooc 3

0

e4

o

9

4

OeOo o

~o~ .4

,

9 0

4

4

,. . . . . . . .

I

,........

i

-'-. 0

i .......

0 O. ~

"ouoo

~--

I. . . . . . . .

.~

~ O. 0 w-

n o

1. . . . . . . . ~

"ouoo

1. . . . . . . .

~p

o0

* 0Q

%

(l16m)

Q

0

(l16~u)

o

657

Barium

The leachability of Ba does not show any significant dependence on pH except a slight decrease around pH = 12.

Cadmium

The solubility of Cd from APC residues follows a similar pattern to Cd in bottom ash, except that the solubility slope is shifted to higher pH caused by complexation with the high chloride content in these residues.

Chromium

Cr solubility measured with the ANC procedure (Kosson et al., 1993) shows a consistent pattern which would reflect Cr III leachability at pH < 4 and a chromate leachability plateau from pH 4 to 10. At pH greater than 10, the concentration of chromate also decreases.

Copper

The leachability data for Cu show a large amount of scatter. The pH stat data for ESP ash are fairly consistent indicating very low solubility of Cu corresponding to tenorite solubility. The irregular behaviour of Cu in APC residues is not well explained.

Mercury

The leachability of Hg indicates a pattern similar to many other metals but only limited data are available.

Molybdenum

The leachability of Mo is scattered. No distinctive trend in its behaviour has been observed.

Lead

The pH dependent leaching behaviour of Pb is consistent with only a few outliers. It is striking to note that the leachability of Pb at pH around 12 exceeds the leachability of Pb at pH = 4. This can result in significant release because the pH of many APC residues is greater than 11. Pb is considered one of the elements of concern in APC residues because results from regulatory testing exceed many current regulatory limits.

Zinc

The leachability of Zn is quite consistent considering the widely different origin of the

658 residues included in the figure. This may point at the presence of one important solubility controlling phase. 16.4 GEOCHEMICAL MODELLING OF LEACHING EQUILIBRIA In Chapter 15, several studies on modelling of leaching behaviour from incinerator residues have been discussed (Di Pietro et al., 1987; van der Sloot et al., 1987; Eighmy et al., 1993, Kirby and Rimstedt, 1993; Comans et al., 1993). Two complementary approaches have been used: Identifying and quantifying the solid phases in ash and running MINTEQA2 through a solid phase approach (Eighmy et al., 1993); and, Using aquatic chemistry to identify the solubility controlling phases using MINTEQA2 based on the measured leachate composition (Comans et al., 1993) Different sophisticated analytical techniques (see Chapter 7.2.7) are needed for the first approach to carry out species identification and quantification. Eighmy at al (1993) indicated fairly good agreement between test data and modelling results following this approach (See Table 15.5 and 15.6). The leaching test used for the comparison was carried out at controlled pH=4 using a high L/S value. The modelling results for AI and Si are significantly below the measured concentration levels because according to the model quartz (SiO2), diaspore (AIO(OH)) and alunite (KAI3 (SO4)2(OH)6) precipitated out. The second approach presented covers a wider range of pH conditions. Elemental concentrations have been predicted by assuming equilibrium between the leachates (at L/S=5) and potential solubility-controlling minerals in bottom ash. These minerals were selected on the basis of their saturation indexes in prior MINTEQA2 runs and their likeliness to be present or formed under the experimental conditions (Chapter 9). The thermodynamic data from the standard MINTEQA2 (version 3.11, Allison et al., 1991 ) database were used, unless indicated otherwise. As molybdenum was not present in the database, it was added as the component MoO42 together with equilibrium constants for aqueous species and solids reviewed by Rai and Zachara (1984). The model predictions are presented as total element concentrations, rather than free ion activities, in the leachate solutions at each pH. This enables presentation of model results together with the analytical leaching data in a graph of log-concentration as a function of pH, which maintains the characteristic shape and concentration levels of the unified pH leaching curves.

659 16.4.1 Bottom Ash Figure 16.5 presents the total dissolved concentrations of major and trace elements in leachates as a function of pH, as well as MINTEQA2 predictions assuming equilibrium with different mineral phases. In the following paragraphs, leaching curves and possible solubility controlling processes are discussed for each element. Calcium and Sulphate Ca and SO4 are the major dissolved components leached from the bottom ash samples, generally reaching concentrations of 3 g/L or greater. Since the solubility of Ca in fossil fuel combustion residues may be controlled by calcite (CaCO3), portlandite (Ca(OH)2) or gypsum (CaSO4*2H20) (Rai, 1987), the same mineral phases have been considered for MSW incinerator bottom ash. MINTEQA2 calculations indicate that the leachates are not in equilibrium with atmospheric CO2 (350 ppm) and that measured concentrations of Ca and CO3 are not in equilibrium with calcite. The steepness of the portlandite solubility curve indicates that the leachates are not in equilibrium with this mineral either. The Ca and SO4 concentrations between pH 4 and 10 do not depend very strongly on pH and follow the solubility curve for gypsum. This relationship is confirmed by plotting the data in a log Ca2*-activity versus log SO42"-activity diagram. Gypsum is a soluble mineral and does allow high concentrations of both Ca and SO4 to be leached from bottom ash. However, provided enough time, carbonation of the alkaline ash and the formation of calcite will further limit Ca leaching in the near-neutral pH range. Calcium concentrations at pH greater than 10 start decreasing at a lower pH and less sharply than predicted by portlandite solubility. The mineral phase controlling Ca solubility at strongly alkaline pH remains as yet unknown. Ettringite, which has been observed in field applications of compacted bottom ash, may play a role in the solubility of Ca at high pH. Stability data for ettringite is needed to be able to verify this possibility. Magnesium Mg concentrations are essentially independent of pH between pH 4 and 7. Normally magnesium also is largely controlled by carbonate minerals. The system is not in equilibrium with atmospheric CO2 as indicated for Ca already. At higher pH values, concentrations decrease sharply. MINTEQA2 calculations indicate this decrease to be consistent with the solubility line of brucite (Mg(OH)2), a mineral which has often been shown to control magnesium solubility. The data point at pH > 13 deviates due to the very high ionic strength of this particular sample. Silicon Dissolved silicon decreases between pH 4 and 10 and increases again at strongly alkaline pH. This leaching pattern is not in agreement with either amorphous or crystalline (quartz) SiO2, as calculated with MINTEQA2. Fruchter et al. (1990) have

660 Figure 16.5 Dissolved Metals in Bottom Ash Leachates as a Function of pH Compared with MINTEQA2 Predictions Assuming Equilibrium with Different Mineral Phases 10000

1000 m

I00

lO00

(a)

o o

~mm

9

I0

o O~

E

r 100 to

o

9

Mg

o,om

"

)

I

I

)

I

i

2

4

6

8

10

12

14

O,O00l

0

1000

I

I

I

I

I

I

I

2

4

6

8

l0

12

14

1000 nip

too

]oo

(c)

i~

i

E

~ 10 o co

(d)

l

13

Si 0,1

0

O,l

9

0

i

o,o,

t

i

)

i

i

I

)

2

4

6

8

l0

12

14

l(X)(x)

o

AI

O,OOl

0

t

I

I

t

I

t

I

2

4

6

8

l0

12

14

ICXZX:)

I000 (e)

(b)

|

0,01

Ca

0

q;I

O,l

9

99

99

99~

o

5

O

oo

~

on

0

9 q,e,,

lO0

ow

9

1

S04 l

0

~

)

)

I

)

)

i

2

4

6

8

10

12

14

pH

(0

m

E

"='Y,100 o cO 0

9

C03

O,l

0

t

I

t

I

I

I

I

2

4

6

8

10

12

14

pH

661 Figure 16.5 Continued

10-

1000 t

100

o

|

1

10

E

1

e0

0,1

(g) "~'

[]

(h)

0,1

0,01

0,01

Mo

Fe

0,001 i

t

t

t

,

i

t

2

4

6

8

10

12

14

o,oool 0

o,ool

0

i

i

t

i

f

:

:

2

4

6

8

10

12

14

100

1

(i)

"-:', o,1 co

I

:"

'

D

0,1

0

i

t

t

4

6

~

~

10

12

14

-

0

, :.

\ . . . . . ./ .. -....

I

I

I

i

I

1

1

2

4

6

8

10

12

14

/.

eZ~LAA\

100 n

X /

Cu

0,0001

8

ec]A

E d

:~ X .

0,001

2

:

x"

Cd

0,001

O)

X"

9

0,01

O,Ol

(k)

~ l i IN ]O 0 0

-

A

E

D:

IO

D '':

Al

I""

(0

0,1 0,1

C 0,01 0 0

"

0,01

''~

0,001

Pb

0,0001 0

*

~

0,001

,

t

t

I

t

t

i

2

4

6

8

10

12

14

pH

/ /

..

Zn

O,O00l

" "

[] '

..

.

.

0,00001 0

I

I

t

i"

I

I

I

2

4

6

8

10

12

14

pH

662 suggested wairakite (CaAI2Si4012*2H20) as the solubility controlling mineral for Si in coal fly ash leachates. Si data from the bottom ash leachates between pH 4 and 10 agree remarkably well with the solubility pattern of wairakite and strongly suggest this mineral phase controls silicon leaching. The data at higher pH deviate from wairakite solubility and suggest that another mineral controls dissolved Si under strongly alkaline conditions. The agreement between measured and calculated data appears to be better than the solid phase approach for this matrix component.

Aluminum AI leaching from bottom ash is strongly pH-dependent and results in the characteristic V-shaped Iog-concentration/pH curve which is typical for Al-(hydr)oxide solubility. Model calculations indicate that the leaching data closely follow the solubility curves of gibbsite (crystalline AI(OH)3) and amorphous AI(OH)3. The data below pH 7 are more similar to those calculated in equilibrium with amorphous AI(OH)3, whereas gibbsite seems to be the solubility controlling mineral at higher pH. Similar observations have been reported for coal fly ash (Fruchter et al., 1990). The formation of aluminosilicates may control the solubility of AI in the high pH ranges. Ettringite has been observed in field application of compacted bottom ash (INTRON, 1991 )(see also Chapter 9) and may also play a role in AI solubility control at high pH. Kirby and Rimstedt (1993) also suggest amorphous aluminum hydroxide as the most likely solubility controlling phase. Iron Fe is leached at relatively high concentrations at low pH, decreases strongly toward neutral pH values and remains essentially pH-independent at neutral to alkaline pH. Amorphous iron hydroxide, or ferrihydrite (Fe(OH)3), is the most obvious solubility controlling mineral for dissolved iron. The measured data at acid to neutral pH follow the calculated solubility line for ferrihydrite, but are up to two orders of magnitude higher in concentration. Similar observations have been attributed to the presence of colloidal iron or inaccuracy of published solubility data for ferrihydrite (Fruchter et al., 1990). The pH-independence at neutral to alkaline pH is not consistent with the calculated ferrihydrite solubility. A second mineral may be controlling Fe leaching in this pH range. Manganese Dissolved Mn also decreases strongly with pH, and reaches minimum values at approximately pH 10. A slight increase in concentration occurs at higher pH values. Although the solutions are not in equilibrium with pyrocroite (Mn(OH)2), the shape of the curves may suggest another less soluble Mn-(hydr)oxide to control Mn leaching. Recent work suggests manganite (MnO(OH)) as the solubility controlling phase. In view of the difference in concentration at the different LS values studied, Mn is probably not solubility controlled at pH below 8. Mn(hyrdr)oxide appears to control the solubility at pH >12.

663

Sodium, Potassium, Bromide and Chloride

The alkali metals Na and K, as well as Br and CI, are very soluble and are leached in high concentrations from the bottom ash in a pH-independent manner. No solubility controlling solids were found for these elements.

Cadmium

Cd is leached in high concentrations at low pH and decreases strongly toward pH values of 8-9. At higher pH, Cd concentrations are near the ICP detection limit of 1 IJg/L shows the stability lines for ottavite (CdCO3) and amorphous-Cd(OH)2. Ottavite stability was calculated using independent measurement of total carbonate (as was mentioned above, the leachates were not in equilibrium with atmospheric CO2). Cadmium concentrations in the leachates between pH 6 and 9-10 are close to values predicted on the basis of equilibrium with ottavite. Cadmium is also known to have a very high affinity for the surface of calcium carbonate (Comans and Middelburg, 1987). Although there is no evidence from solubility calculations that dissolved Ca and CO3 were in equilibrium with calcite, some calcite might nevertheless have been formed considering the high Ca concentrations, pH, and contact with the atmosphere. Co-precipitation or solid-solution formation with calcite may limit Cd concentrations to lower levels than would be predicted by the solubility of ottavite (Comans and Middelburg, 1990). It is uncertain at present what role this sorption process plays in controlling Cd solubility during bottom ash leaching and whether it may explain some of the cadmium concentrations below the solubility of ottavite. All leachates were at least two orders of magnitude under saturated with respect to amorphous Cd(OH) 2, which rules out the relevance of this mineral in systems containing sufficient carbonate. The role of carbonation and calcite formation during aging of bottom ash appears to be a significant factor for the retention of Cd in the ash matrix. Previous removal of chlorides, thus avoiding the formation of soluble Cd-complexes, will enhance this process. Accelerated carbonation may prove a means to improve ash characteristics.

Copper

Leaching of fresh bottom ash generally leads to high initial concentrations of copper (Figure 16.6). Cu concentrations decrease from pH 4 to 6 but remain constant at higher pH at levels (approx. 1 mg/L), depending on the L/S ratio. Tenorite (CuO) has been suggested as the controlling phase for Cu leaching from coal fly ash (Fruchter et al., 1990) and may also be relevant for bottom ash. However, the leachates are systematically over saturated with respect to this mineral, except at low pH where concentrations are lower than predicted by tenorite solubility. Substantial amounts of dissolved organic matter can often be released from the uncombusted fraction in the bottom ash. It is speculated that dissolved organic material may increase copper solubility because of the high affinity of this metal for organic species. The relevance of this process was investigated by giving an ash sample an 8-hour heat treatment at 550 ~ prior to the leaching experiments. This

F~gure16 6 Leachab~lltyof Copper from Bottom Ash

(b)

(a)

Demineralized water

-

-

B A - ~ & z ~, ~ 5% H, flushing q,

\

3

2

4

5

8

7

6

9

-400 2

1 0 1 1 1 1 1 3 1 4

(c)

I

.

-: I

J

6

5

r

J

7

1

(dl

a

J

s

l

9

1 0 1 1 1 2 1 3 1 4

pH

\ \\

0

01:

001

/ \\

j

CU

2

3

'

0

\

d

4

BA heated 1 0 5 5 0 S--, ~ ~ "

4

L

L

5

-

Prcah BA

Reducing

1 cond~tlons

0

'

6

L

7

-

8

pH

d-

4

j

BA + Orgmnic

U

abc-

'

1m

10

d

J

l

3

pH

loo C

3-

"

L

L

+& , c

//

/

/

BA from rotary ~ l n

.

9 1 0 1 1 1 2 1 3 1 4

6

I

U

cornp~ex.tion

01

II \\

I/

Ool

Z

,

3

,,

4

/

/

\\ \

/ >

5

6

7

8

,'CU(OH)X

-

'

9

1 0 1 1 1 2 1 3 1 4

/ 8

pH

Unified pH curve Redox potential in batch experiments under alr exposure and in a close container purged with 5% hydrogen Cu leaching from fresh bottom ash, same BA after addition of a mixture of organic acids, same BA after heating to 550°C to remove organlc matter, and Cu leachlng data point for a rotary kiln facility proposed factors and their pH dependence for leaching pf Cu from bottom ash under ambient conditions

665 procedure was assumed to destroy organo-copper complexes by removing any organic material left in the ash sample. As Figure 16.6c indicates, the leaching of Cu is reduced by the temperature treatment, especially between pH 6 and 11. At low pH, copper complexation by organic species is reduced by protonation of the acid functional groups on the organic material, whereas at high pH the formation of stable hydroxo-copper complexes may out compete Cu complexation by organic species. The treated samples followed the tenorite solubility pattern, but are under saturated with respect to this mineral. The copper solubility as predicted by MINTEQA2 calculations is, however, strongly affected by the stability of the aqueous Cu(OH) complex. Log K for the reaction Cu2§ + 2H20 - Cu(OH) + 2H § equals to -13.7 in the MINTEQA2 database and is more than two orders of magnitude higher than the value of -16.2 which has been reported by others (Rai and Zachara, 1984) as also been shown to be consistent with biological uptake of dissolved copper (Blust et al., 1991). If the value of-16.2 is used, the solubility of tenorite is calculated to be much lower, with leachates below pH 7 being under saturated, and above pH 7 over saturated with respect to this mineral. In summary, the solubility of Cu during bottom ash leaching is complicated by organic complexation and disagreement as to the stability of the aqueous Cu(OH) complex. In view of the above, the deviation of the heat treated ash sample, within an order of magnitude, from the Cu concentrations calculated in equilibrium with tenorite does not rule out this mineral as a possible solubility-controlling solid. Another approach that was taken was to extract the same bottom ash sample with a mixture of dissolved organic compounds. In Figure 16.6b, the results of the experiment are given indicating an increase in the release by about one order of magnitude after leaching with a mixture of acetic acid, propionic acid, butyric acid and valeric acid, which are present in leachates from domestic waste disposal. It appears that the addition of organic complexants increases the leachability of Cu, particularly in the pH range 9 - 10.5. It is interesting to note that a similar, but less pronounced, peak is observed in the unified pH plot for Cu at this same pH interval (Figure 16.6a). In Figure 16.6d, the role of different factors -inorganic copper (tenorite), organic acids, DOC controlling Cu leachability is schematically presented. The influence of reducing conditions on Cu leachability was addressed by using a recently developed redox stat (Comans, 1993) based on equilibration with H2 gas. The relation between Eh and pH under normal atmospheric conditions and forced reducing conditions is given in Figure 16.6b, whereas the consequences for the Cu leachability are presented in Figure 16.6c. A greater than one order of magnitude difference in leachability in the neutral to moderately alkaline pH range is observed. These results indicate that both changes in redox and removal of organic matter can result in decreased Cu release.

666

Molybdenum Mo is known to form very mobile oxy-anions, and is mainly present as MoO42 above pH 5. This element was measured only in the leachates from one installation (urban), with concentrations increasing strongly with pH between 4 and 7, and remaining virtually constant at higher pH. Of the Mo solids reviewed by Rai and Zachara (1984), which were added to the MINTEQA2 database for the purpose of this study, PbMoO4 and CaMoO4 are the most likely phases to control Mo solubility in the leachates. MINTEQA2 calculations showed the leachates not to be in equilibrium with PbMoO4. The Mo concentrations predicted from solubility control by CaMoO4 are shown in Figure 16.5h and could explain the leaching data between pH 7 and 10. The relationship is confirmed by plotting the data in a log Ca2*-activity versus log MoO42-activity diagram. CaMoO4 may, therefore, control Mo leaching in this pH range. This mineral has also been suggested to control Mo concentrations in high-Ca waters (Hem, 1985), but was calculated not to control Mo leaching from coal fly ash (Fruchter et al., 1991). Lead The Pb leaching patterns for the two bottom ash samples are clearly different from each other. While one bottom ash sample (urban) shows a strong pH dependency, with concentrations decreasing over three orders of magnitude from pH 4 to 10, dissolved Pb leached from the second sample (rural) remains high and is relatively pH-independent (Figure 16.7c). Equilibrium with Pb(OH)2 could explain the first ash sample data above pH 9 or 10, but the mechanisms behind the high Pb leaching from the second ash sample are unclear. Most dissolved Pb data from the first ash sample leachates between pH 7 and 10 are close to the ICP detection limit and do not allow conclusions as to the solubility controlling solid. Modelling studies on fly ash by Gardner (1991) indicate solubility control by PbSO4 in the lower pH ranges and Pb(OH)2 solubility above pH=7. Whether Pb solubility in bottom ash is partly controlled by sulphate remains to be proven. Zinc Similar to the other heavy metals discussed above, Zn is leached in high concentrations at low pH and decreases strongly (four orders of magnitude) with pH values up to 10 (Figure 16.7d). Complexation with hydroxide in solution increases dissolved Zn at strongly alkaline pH. The leaching pattern follows the solubility line for zincite (ZnO), with concentrations remaining down to 1 order of magnitude lower. A second possible solubility controlling solid is ZnSiO3. The solubility of ZnSiO3 in Figure 16.7b shows a pH-dependency similar to the leaching data and limits, except at very high pH, dissolved Zn to 1-3 orders of magnitude lower concentrations. Both zincite (at intermediate pH) and ZnSiO3 (pH 4-6 and pH > 12) may contribute in controlling Zn leaching from bottom ash. Gardner (1991) calculates Zn leaching from fly ash to be limited by the solubilities of Zn(OH)2 and ZnSiO3. Zincite is, however, less soluble than Zn(OH)2 and gives a better prediction of the data. It appears that the solubility control in bottom ash and fly ash is not significantly different for Zn.

Figure 16.7 Thermodynamic Activities from Column Experiments on MSWl Fly Ash -2 -4 A

+

N

D

f5 M 0 H

-6

(a)

;/ 0 0

-8

-2

-

on

-

0 0

0

-10 -

0

a ,

-I2 1

(b)

0

2

3

4

5

6

1 2 6

-

" " 8

'

" " 10

12

14

16

668

16.4.2 Modelling of APC Residue Leachability The modelling of APC residue leachability requires the application of corrections for the activity coefficient using Pitzer equations (Pitzer et al., 1973 & 1974) because the Davies equation and the Debye-H(3ckel equation are not valid at the ionic strengths present in leachates from APC residues. However, the general trend as observed for bottom ash is valid for APC residues. Extending the modelling efforts from bottom ash to APC residues opens a new area of research. The work of Gardner (1991) of modelling fly ash leaching from small columns and batch extractions is very much in accord with observations on bottom ash as described in the previous section. Figure 16.7 presents the solubility control of Pb and Zn from fly ash as a plot of element concentration against counter ion. This type of plot illustrates the boundaries of solubility control by specific mineral phases. It is clear that in different pH ranges, different minerals may control the leachability. It is, however, unlikely that many phases will control leachability at the same time and to the same extent. This implies that with the identification of the most relevant mineral phases (usually 2 or 3), the leachate composition should be adequately predicted. The task is now to sort out the most relevant phases in the relevant pH regions.

16.4.3 Application of Geochemical Modelling Results In combination with XRD and other surface techniques as discussed in Chapter 7, specific mineral phases can be identified. The modelling of the solution properties does not provide direct evidence for the existence of specific minerals in the bottom ash matrix, but the results of such studies support the hypothesis that geochemical reactions control the leaching of both major and trace elements from incinerator residues. A number of possible solubility controlling minerals and complexation processes in solution have been suggested that can to a large extent explain the observed leaching behaviour as a function of pH. Knowledge of these processes can be used to: Predict the long-term behaviour of incinerator residues in the environment Improve the interpretation and further development of regulatory leaching tests, and Chemically modify the ash and/or its environment in utilisation or disposal, to minimise contaminant leaching Further work is needed to clarify the behaviour of some of the elements discussed above. In a number of cases, the degree of agreement between model predictions and actual measurements is very promising. General leaching behaviour and the impact of waste feed on incinerator operating conditions can be identified based on modelling efforts. In Figure 16.8 some of these

669 Figure 16.8 Identification of pH Domains with Solubility Control and Availability Control for Cd, Cu, Mo, Mg, Pb and Zn in Bottom Ash 1 O0

1000

Variation with the availability for leaching

I \

\\

,

i= d

e0 0

\

\ k

0.1

\

0.01

\ /

.........

0

' .... 4

2

1000

100

8 o l u b l l l t y control by Inorganic cormtltuents

' .... 8

' .... ' .... ' .... 8 10 12 14

0.01

0

....

' .... "'"' 2 4

e

....

a v a i l a b i l i t y for leaching

O/0

%

/

,N ~ / / /

/

' .... ' .... ' .... 8 10 12 14

Region of aolublllty control

-O-

9

10

~

C)

9 I

~0 0.1

go

1

0

0

0

0.01

0.1

Mg 0.01

\ \

//

,' /

Vadation with the availability for leaching

Region Of solubility COntrol by bruclla

0

c o

o.,

~~//

Variation due to the T

o

"4', ~ 4,', I "4,

"

Cd

0.00001

vE

X\

\00\

",,,, o Ic~

0.0001

Region of solubility control by DOC 0

\

Xx

0.001

OI

\

\

"

0 0

\

X

/

10

\;~

0

\

Reglo~ Of solublllty control by chloi'Ido and omrboflite

/-I_ %.,U

~

10o O

p ~

OD

Vm'iatJoe with the availability for leaching

4

Ms

~o

I

J

z L'- T - -

T -- T--

T--'T~'Z

5

8

7

9

11

8

10

12

1:)

0.001

........ 0 2

10000

1000

100

Region of solubility control

Region of solubility r

100

0.1

go

0.01

Pb

0.001

,,,,I 0

.... 2

a,,,,j 4

.... 6

pH

! .... 10

0

:I

1

o.1

i .... 12

9

lO

o/

! .... 8

14

4~?o

1000

~O I I

12

Variation wllh the availability for leaching

~1 Varlmtlon with the I availability for leaching

O

' ......................... 4 6 8 10

14

0,01

0

9169

Zn

......... 2

' ......... 4 6

t I I I

' .... ' .... ' .... 8 10 12 14

pH

670 trends are presented, which indicate regions where leachate concentrations are controlled by availability and regions controlled by solubility as defined by geochemical reactions. Such general trends appear to be relevant for other types of residues as well. 16.5 RELEASE RATES OF ELEMENTS The observed release of a specific element over a time interval is a result of the element's availability, solubility and rate of mass transfer from the solid to the liquid phase. The rate of mass transfer between phases is most frequently controlled by diffusion through the porous solid phase, geometric dimensions of the solid phase, and the mode of liquid solid contact. In general, the following cases can be used to describe the release of elements and inorganic species from incinerator residues: Localised equilibrium is attained between the solid phase and the contacting leachate, resulting in a leachate which is saturated with respect to the element of interest, i.e., solubility is controlling release. This case typically occurs for inorganic species, except alkali metals and halogens (e.g., Na, K, CI), with percolation of infiltration through residues in the field and during column leaching tests or sequential batch extractions of small particles in the laboratory. For this case, cumulative release with respect to time reflects the product of the liquid to solid ratio (LS) with the elemental solubility at the specified conditions (e.g., pH and Eh). .

,

Equilibrium is locally attained between the solid phase and the contacting leachate, but the resulting leachate is not saturated with respect to the element of interest because of limited availability or flow channelling. This case typically occurs for alkali metals and halogens with percolation of infiltration through residues in the field and during column leaching tests or sequential batch extractions of small particles in the laboratory. For this case, greater than 50% of the availability is released at an LS = 1. Release is controlled by diffusion through the solid phase and the contacting leachate is not in equilibrium with the solid. This case typically occurs for most elements and species when flow is around either a monolithic material (e.g., solidified/stabilised residues) or a granular material compacted to low permeability such that is behaves as a monolith (e.g., residues used as a road base and overlain by an asphalt layer). For this case, cumulative release is a function of effective diffusivity, availability and release time interval.

The following sections present observed release rates for elements as a function of L/S to address Cases 1 and 2, above, and diffusion controlled release to address Case 3.

671 16.5.1 Release As a Function of Liquid to Solid Ratio

The leaching of granular residues is in many cases dictated by percolation, which can be represented by column experiments. This is valid both for utilisation and for disposal, unless measures have been taken to minimise infiltration drastically. In that case infiltration may be reduced to such an extent that diffusion becomes the rate controlling transport mechanism. It also is important to establish the relationship between column and batch testing results because while column tests more closely reflect field scenarios, batch tests are more efficient to carry out in the laboratory. A comparison of leaching data from both column and batch tests has been carried out for a wide variety of waste materials (WAV, 1988, 1992). The two approaches were agreement except for the cases where the initial release and depletion of one species subsequently resulted in altered release for a second species (van der Sloot et al., 1993). In addition, some long term processes, such as those caused by biological activity can be observed in column tests carried out at low flow rates over extended time periods, but cannot be addressed in short batch tests. Test results generally are expressed in mg/kg to allow a direct indication of release. Presentation of column leaching results in the form of cumulative release as a function of L/S may be used to discern between different types of release behaviour (Figure 16.9): Type I-

Rapid release of highly soluble species due to under saturation of the constituents in the leachate even at low L/S. This results in rapid wash out of these species in a percolation dominated system with release of the available quantity within L/S less than 1 or 2. The slope of the cumulative release as a function of L/S typically is greater than or equal to 1 at L/S less 1 followed by a slope of approximately 0, indicating depletion. Elements that exhibit this type of behaviour include alkaline metals and halogens.

Type II-

Release controlled by solubility in the aqueous phase which most often is a strong function of pH. Cumulative release as a function of L/S is approximately linear with the slope dictated by the elements solubility. Elements which exhibit this type of behaviour include Pb and Zn.

Type III-

Delayed release due to retention in the matrix by a second species controlling solubility which is depleted after a limited time interval. This behaviour is characterised by a transition from linear release at a lesser slope to a greater slope. An example of this behaviour is the release of sulphate which initially may be limited by barium until depletion of that element occurs.

Type IV-

Enhanced initial release due to the presence of a complexing agent which increases the solubility of the element of interest. This behaviour

672

Figure 16.9 Types of Release Identified from Column Tests or Sequential Batch Tests

Type I

[c]

tEj

~Jf---

Availability

..........................................................

o

2

LS

2

10

[c]

~

Type II

J

I

10

[El

T

/

/ 2

~,

2

LS

Type III

[c]

1.8

10

[El

/

j

2

L$

Type IV

1 2

LS

IO

'

2

10

[c]

......

10

[El

/]

2

LS

/

673 is characterised by a transition from linear release at a greater slope to a lesser slope. Unlike for Type I release, depletion does not occur. Examples of this behaviour are the initially increased release of copper in the presence of organic acids, and the initially increased release of cadmium from APC residues in the presence of high chloride concentrations.

Bottom Ash In several studies (Versluijs et al., 1990; Hjelmar, 1992; F~llman, 1992; Eighmy, 1992, van der Sloot et al., 1991) column leaching experiments have been carried out using bottom ash. The results of these studies, presented in Figure 16.10, can be used to compare release from ash generated in Canada, Denmark, Germany, The Netherlands, Sweden, and the United States. The release of the various constituents has been presented as a function of L/S with ranges for total content and availability indicated for comparison. Results for bottom ash from batch extraction tests with distilled water at L/S=20 are provided along with the column results in Figure 16.10. Cumulative release of Cd, CI, Cu, Mo, Na, Ni and sulphate observed for column tests at L/S=10 and release observed in batch tests at L/S=20 were similar, indicating that the leachable fraction of these elements was released at US less than 10. Solubility controlled release was observed for Ba, Ca and Pb between L/S=10 and 20 based on a linear increase in cumulative release from the column to the batch data.

Antimony

The release of antimony from bottom ash ranged from 0.0005 to 0.5 mg/kg. In contrast, the availability of Sb is approximately 5 to 10 mg/kg, which implies that Sb is largely retained in the ash matrix. Release patterns were very consistent and indicate slow dissolution with increasing L/S. The wide range in Sb release most likely was related to variability in the waste feed composition.

Cadmium

The release pattern for Cd indicated a small initial release of a highly mobile species followed by a slow increase with increasing L/S. The release patterns generally had a slope of 0 after L/S=1, which indicated that the soluble Cd-species was depleted. Possible species responsible for this release behaviour are anionic Cd chloride complexes or organically complexed Cd. The cumulative release at US=I 0 ranged from 0.005 to 0.05 mg/kg compared to availability which was between 0.5 and 5 mg/kg, indicating that Cd release was limited and significant retention in the matrix occurred.

674 Figure 16.10 Release (mg/kg dry ash) of Selected Metals as a Function of the Liquid to Solid Ratio 1oo

~ ~"

10000

As

1000 I

lO

Total

1

Availability

Total

~= 10010I

]~a

Availability ....I

0,1

--~ 0,01 n" 0,001

~

0,1 0,Ol

0,0001 =,

9

0,001

100

LS (llkg) 100

1 10 LS (llkg)

CI Total

A,,eilability

...........................................

j l P I l - n u-u-ll

~ ] - , * i B - - o

0,1

"==t

0,01 0,001

100

1OOOO

Cd

Total .1 Availability

0,1

E 1000

D-i~-a

.............................................

0,0001 0,00001

........

0,1

IOO0

.....

,,,|

Total

lOO "~

,

1 10 LS (llkg)

1

........

00

,

100

100

1 10 LS (llkg)

0,1 0000

Cr

1000 ~

Total

100 l

Avaiiabllity

,I

Availability

Cu

I

0,1 ~

0,01 0,001 0,0001

0,1 ~ 0,01

0,1

1 10 L8 (=lkg)

........ 100

0,0Ol

i

o,1

........

~

1 Le

........

(I/kg)

:

lO

........

lOO

675

Figure 16.10 Continued 100

10o0

Mo Total

10

t

100

Total

10_ u

9

9

Availability. . . . . ]

9

Availability

~ 0,1 0,01

Ni

_

E n ~

i

,

,

,

.....

,

.

,

,

u . , , l

1

0,1

LS (llkg)

,

_

10

0,001

i

. . . . . .

lO0

............ 1 10 100 LS (llkg)

0,1

100000

100o0 lOOO I

Total

lOO l

I

~b

S04

Total = Availability

.........................................

A~ailability

~, 10000 r

i

0,1

-~ iooo m.'

O,Ol 0,0Ol 0,0001

,

,

o,1

,

9 ......

1 lO LS (l/kg)

1000 Availability

10

Total

I

lO0 lO

...:,

0,1 B

0,01

B

!1 II Ii

0,001

,,

0,1

~1111 9

............................. 1 10 100 LS (llkg)

1000

"'=l

1

0,1

0O00

Sb

100

0,0001

100

:

lOO

~

II

Zn

Availability

~

o,1

I

f,~

D-I~--D

n-=---~

0,01 ........

i

1 10 LS (llkg)

i

,

i

. . . . ,i

100

0,001

o,1

1 lO LS (I/Vo)

lOO

676

Chromium

The release pattern for Or indicated depletion of a soluble species after L/S=1 (slope = 0). The fraction initially released may be a chromate species, reflecting the high mobility of Cr (VI) in comparison with Cr(lll). A steady increase in release with increasing L/S (slope = 1), indicating slow dissolution, was noted for only one case. Cumulative release at US=10 ranged from 0.005 to 0.5 mg/kg, in contrast to availability between 2 and 10 mg/kg.

Copper

The release of copper was very consistent among different sources of ash, as indicated by the parallel pattern of the release curves. The release of Cu was indicative of a highly soluble fraction that was washed out within L/S-1 (approximately 2 pore volumes). The concurrent release of a dissolved organic species which complex Cu has been postulated as an explanation for this initial release. The cumulative release at L/S=10 from many different sources was within a relatively narrow range of 0.3 to 10 mg/kg, which is a small fraction of the availability which varied between 50 and 200 mg/kg.

Molybdenum

Two distinct data groups can be distinguished for Mo. The lower release curve was related to input from typical domestic sources, while it appears that the relatively high releases of Mo observed for two facilities resulted from industrial contributions of Mo rich waste streams which were included with the incinerator feed. Cumulative release at L/S=10 ranged between 0.2 and 10 mg/kg, which was approximately equal to Mo availability. These results indicate that Mo is quite mobile and almost all leachable Mo can be depleted from bottom ash in a relatively short time span. The relatively high initial concentrations of Mo in leachates may be of concern in some jurisdictions.

Lead

The release pattern for Pb was reasonably consistent for most cases, with increasing cumulative release with increasing L/S (slope = 1)indicating solubility controlled release. In a few cases, however, depletion appeared to have occurred. This type of release behaviour may be explained by a decrease in pH during the leaching process, which resulted in decreased Pb solubility. Cumulative lead release was considerably variable between data sets. At L/S=10, the cumulative release varied between 0.005 and 10 mg/kg. The sensitivity of Pb solubility to pH was the primary reason for the wide range in release observed. However, the cumulative release at L/S=10 was several orders of magnitude less than the availability, which indicated considerable retention of Pb with the ash matrix.

677

Nickel

The release of Ni was similar to that of Cu. Both Cu and Ni have been demonstrated to form strong complexes with dissolved organic matter, which is hypothesised to have been responsible for initial release of both elements. The cumulative release at L/S=10 varied between 0.02 and 0.5 mg/kg, with only two cases having significantly less release. Ni availability was several orders greater than the observed cumulative release, indicating significant retention in the ash matrix.

Sulphate

The release of sulphate approaches the maximum leachability at L/S=100, which indicates that sulphate can be depleted from the bottom ash matrix. The quantity of soluble sulphate present in the ash is such that it becomes a constituent of major concern in relation to potential effects on local ground water quality.

Zinc

The release patterns for Zn were fairly consistent for a given data set, indicating very slow dissolution from slightly soluble phases. However, considerable variability existed between data sets. Cumulative release of Zn at L/S=10 varied from 0.01 to 3 mg/kg. The primary reason for the variability has been attributed to variable pH, because of the sensitivity of Zn solubility to small changes in pH. This was similar to the behaviour observed for Pb. However, the cumulative release at US=10 was a small fraction of the availability, which ranged from 50 to 500 mg/kg.

Alkali Metals and Halogens (e.g. Na, K, CI, Br)

Alkali metals and halogens, such as Na, K, CI and Br, were completely leached from the bottom ash at L/S=1-2. There was no retention of these elements in the matrix, which implies that most equilibrium batch extraction tests can be used to assess the potential release. The quantity of these elements present coupled with their rapid release should be carefully considered during the development of management practices for bottom ash. The column leaching test results discussed above demonstrate consistent release patterns reflecting depletion, dissolution or in some cases a delayed release due to changes in chemical conditions with time of leaching. For a specific element such as Pb or Zn, cumulative release may vary over several orders of magnitude depending on the origin of the particular bottom ash sample evaluated. This observed variation in cumulative release results primarily from variation in ash alkalinity which controls leachate pH and solubility of elements. This effect is substantiated through the presentation of measured concentrations of elements in leachates from column tests as a function of pH in Figure 16.11. The unified pH-solubility curves also are presented for reference. In general, the pH of leachate from column tests with bottom ash does

678

Figure 16.11 Release (mg/kg dry ash) of Cu, Pb and Zn as a Function of pH to Indicate Solubility Controlled Release in Column Leaching Test (n) as Compared to the Unified pH Curves (e) 100 -

Cu _0

o

o

O0

-= o ~ , O~ o

~o

1

o

.

.

.

.

t

10

.

11

oo oo 0 0

o

o B o O []

[] o

o

0,1

o _ oo u

,.,

o

.

.

.

I

,

'

'

I

13

12

pH

o

1000O o

o

~ ~/

~

10

o

o

1000

o

o

oO

oO. o w ... .~fi_ ~_~~oO~O o-

o

o

o 11

12

13

pH

14

100000 Zn ,-, ..... E o

10000

1000

o

O

._ E

8

10

o

n

100

J~4b@@~n -OOnLl

~ 0

0

11

0

DO

0

n

0

O00D'~D

12

pH

13

14

679 not vary over a wide range. Pb and Zn data from the column experiments are consistent with the unified pH curve, which illustrates the solubility controlled release as a function of pH for bottom ash. Measured concentrations and trends as a function of pH from column and batch tests are consistent. APC Residues Column leaching experiments have been carried out using APC residues in several studies. The release from ESP ash has been reported in two studies carried out in The Netherlands (Versluijs et al., 1990; Born, 1993). Additional studies have examined release from dry and semi-dry scrubber systems in Denmark (Hjelmar, 1991) and the United States (Theis and Gardner, 1991).

The results of these studies, presented in Figure 16.12, can be used to compare release from the different types of APC residues. The release of the various constituents has been presented as a function of L/S with ranges for total content and availability indicated. Results for ESP ash from batch extraction tests with distilled water at L/S=20 also are provided along with the column results for comparison. A typical feature of results from column leaching of APC residues is an initially moderately alkali pH at very low L/S ratios, followed by increasing pH with increasing L/S ratios. It is hypothesised that this behaviour results from the formation of a shell of acid gas reaction products surrounding a carbonate shell, which in turn surrounds a hydroxide interior of the APC residue particle. The carbonate shell subsequently dissolves during leaching with increasing L/S. This effect is greater for dry and semi-dry scrubber residues because of the injection of lime into the flue gas. The resulting changes in leachate pH also effect the release of elements for which solubility is a strong function of pH. From the leaching curves as a function of L/S it is clear that the cumulative release is constant or increases only slightly after an L/S=1 (As and Cr from two installations are the primary exceptions). This implies that the release within one L/S constitutes the amount available for leaching at the pH controlled by the residue and represents soluble chemical species that are washed out of the system with minimal retention. This mode of release is prominent for many constituents in ESP ash as evidenced by the release-L/S profiles. The variability in leachability of ESP residues within and between installations is substantial. It appears to be related more to the input to the incinerator and facility operation than in the case of bottom ash. Initial leachate concentrations and release of individual elements from APC residues during column tests typically are in agreement with the solubility as a function of pH derived from batch testing. Initial solubilities may increase relative to those observed for bottom ash because of the high ionic strength and chloride concentration of the solution. However for APC

680 Figure 16.12 pH and Release (mg/kg dry ash) of Selected Metals as a Function of the Liquid to Solid Ratio pFl

As 0,1

== 11

j

0,01

" 0,001

0,0001 0,01

0,1

1

10

100

.......

0,01

I

.......

0,1

LS 1000

........

I

10

1o0000O

Cd

100

I

1 LS

I

.......

100

CI

=

{,o

e == 1o000o

==

~

o,1 0,01

o

, 0,01

0000 0,1

1

10

100

. . . . . . . ; . . . . . . ; ........ = ....... 0,01

0,1

1 LS

LS 10

10

,oI

100

Cr

e .=

100

Cu

1.1-I-I

m-D--I

. . . . . . . . ',

....... I

....... i

0,1

1

==

.. 0,1 "

0,1

-t: [

0,01

................................. 0,01

0,1

1 LS

f 10

100

0,011

I

0,01

LS ( l l k g )

10

........ ; 100

681 Figure 16.12 Continued

100

100

Ni

Mo

10

lO

J

v

j

-

)

1

0,1

........ I

0,1

0,01

........ I

........ :

1

10

0,1

........

:

100

O,Ol

........

0,01

a

........

i

0,1

1000

10000 -~

m

........

i

100

Zn

Pb

1000

v

I

10

LS

LS

E

........

1

lOO

100

9 m

tr

lO ram=

r

1 0,01

........

I

0,1

........

:

1 LS

........

I

10

........

I

100

1 0,01

........ : 0,1

I

m

m m

........ : 1 LS

|

m

........ I 10

........ I 100

682 residues, availability is an important parameter controlling the release level with depletion occurring for many elements at L/S less than 1. As a consequence of these observations, column tests are not very functional for the evaluation of APC residue leachability. Availability and pH dependent solubility are much more a practical tools to assess APC properties. 16.5.2 Diffusion Controlled Release

Assessment of constituent release most frequently assumes that the predominant mode of leaching will be by percolation of water through the porous solid matrix. However, in several circumstances the release may not reflect percolation and solid-liquid equilibrium, but rather be controlled by diffusion through the porous solid. This situation may occur when a fine-grained material is placed in a surrounding matrix with a higher permeability, thus creating a preferential flow around, instead of through, the material (Environment Canada, 1990). A similar situation will be encountered when the material is compacted during placement to form a less permeable matrix, or when infiltration of water is minimised through use of low permeability barriers such as compacted clay covers. Evaluation of release under these circumstances requires measurement of diffusion controlled release fluxes. Measurement of diffusion controlled release fluxes for monolithic solids is achieved by refreshing the leachant in contact with a well-defined surface geometry at regular intervals. An analogous test for compacted granular material has been developed which maintains an undisturbed surface through use of a thin layer of glass beads placed on top of compacted granular material in an inert mold. This test was first applied in the framework of a USEPA study on stabilisation/solidification of incinerator residues (Kosson et al., 1993). After an initial delay in the release caused by the layer of glass beads, the release profile generated permits the calculation of tortuosity and chemical retention within the matrix (refer to Chapter 21). The intrinsic leaching properties obtained from this type of measurement allows prediction of release at longer time scales than the actual testing period. In addition, it provides an estimate of diffusive contribution to transport in a low flow column experiment. Table 16.5 presents the measured values of tortuosity and pD for several elements in untreated bottom ash, APC residue and combined ash (Kosson et al., 1993). Estimation of the pD values are based on using the availability of each element in the specific matrix as the driving force for diffusion. This approach results in a more accurate estimate of the release parameters than use of the total concentration as the driving force. Compaction of bottom ash and combined ash resulted in a reduction in permeability, which was reflected in the pD values for mobile species, such as Na, K, and CI. The difference in physical retention between bottom ash and combined ash compared to APC residue was substantial. The latter behaved as a loose powder without physical

683 Table 16.5 Diffusivities (-Iog(m2/s)) Measured using the Compacted Granular Leach Test for Residues Element

Bottom ash pD

Tortuosity

Std. dev.

24

Combined Ash

APC Residue

pD

Std. dev.

pD

13.6

0.2

As AI

14.06

0.09

13.63

0.19

Ba

12.3

0.08

12.04

0.08

Br

10.05

0.17

9.91

0.12

Cd

>15

Ca

12.68

0.06

12.75

CI

10.5

0.16

Cr

11.73

*

Cu

>14.8

Fe

14.2

Pb

16.17

>15

15.2

0.6

0.09

10.3

0.15

10.52

0.21

9.3

0.1

10.35

*

11.6

0.15

14.57

0.17

14.2

0.2

0.19

15.2

0.21

11.2

0.2

0.09

16.3

0.37

14.2; 11.8; 12.9

Li

11.93

*

11.69

*

Mg

14.66

0.72

15.1

1.06

Ni

>13

11.02

*

NO3

11.37

0.36

10.36

0.60

K

10.12

0.07

10.18

0.08

Si

14.49

0.12

13.48

0.06

Na

10.24

0.08

10.26

0.09

Sr

11.5

0.07

11.45

0.09

SO4

15.62

*

15.27

0.18

16.01

Zn

15.71 data. Kosson et al., 1993

Std. dev.

25

11.2

0.2

9.0

0.1

9.0

0.1

0.17

13.6

0.4

0.35

15.9

0.3

* single

restriction; only chemical retention was due to the chemical environment (high pH). The mobility data for combined ash and bottom ash are very similar, which is largely related to the fact that the pore water conditions (e.g. pH) are very similar. The low retention values for K, CI, Br, nitrate and Li are in agreement with the high release of these constituents observed in column experiments. The high retention values for

684 several other elements also are in agreement with release observed from column and batch tests. The difference between APC residue and bottom ash is significant for Ca, F e, Pb and minor for sulphate and Zn. In the case of Pb the difference is largely attributed to the higher pH in the APC residue versus the bottom ash. The large variability in Pb leachability from APC residue reflects the sensitivity to minor changes in pH. The further development and application of this procedure needs to be explored for predicting release from incinerator residues.

16.6 RESIDUE LEACHING IN THE CONTEXT OF REGULATORY LEACHING TESTS AND WASTE FROM OTHER SOURCES 16.6.1 Regulatory Tests and pH Dependent Leaching Regulatory leaching tests in most jurisdictions are batch tests which control the leaching conditions through definition of the leachant composition, L/S ratio, sample preparation (e.g., particle size reduction) and equilibration period. The final pH of the leachate is controlled to a large extent by the acid neutralisation capacity of the waste being tested. Measured leachate concentrations of elements and species are then compared to fixed regulatory thresholds. Figure 16.13 illustrates the impact of alternative presentations on the interpretation of results. Clearly, pH is the most important independent variable. However, most regulatory data that is obtained is focused on a narrow pH range dictated by the waste being tested, which limits the utility of results for defining characteristic release behaviour. This pH range may not reflect actual disposal or utilisation conditions, and may be in a range for specific elements where small changes in end point pH cause large changes in observed leachate concentration. Thus, the result is often a testing scenario that differentiates between the alkalinity of a waste, not the potential for release of specific elements and environmental impact. Figure 16.13 clarifies this effect through schematic presentation of the results of several regulatory leaching tests for Pb. The unified pH-solubility curve and the concentration range and threshold values for each test are presented. Deviations from a general leachability pattern can be related to changes in solubility controlling properties either imposed by the test, such as the Swiss TVA or the California WET test, or by changes in ash characteristics due to changes in input or operation conditions of the incinerator facility. None of the regulatory tests consider quantity of total soluble salts present in the material being evaluated.

German DIN 38414 (1984) and French AFNOR X-31-210 (1988) These tests are similar in that they are based on the conditions dictated by the material to be tested. The final leachate pH can vary widely depending on the acidic or alkaline properties of the material being tested. The pH range commonly observed for incinerator residues extends from 9 to 12. In this region, the solubility of some elements changes rapidly with minor changes in pH, especially metals which exhibit amphoteric behaviour. Consideration of this phenomenon should be given when interpreting test results, thus ensuring that the associated test conditions are relevant to actual management practices. For example, leachate pH from bottom ash disposed

685 Figure 16.13 A Comparison of Different Methods of Presentation of Regulatory Leaching Test Results for Lead

Regulatory limit

..........

200

.

10

.

.

.

.

.

.

.

.

.

e

~., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0

~

0 0

.

.

.

. Aft..}

A

J

g~

Regulatory limit

1

oo /

~

_

|

13

~

0 ......

0

I

I

I0

20

0.3

.

30

4

l

i

!

,

,

,

l

,

w

6

7

8

9

10

11

12

pH

LS

50 f .

. . . . . . . . . . Regulatory [ limit e 9

[ 0

~o e~

"

~

4

.......... 2'. . . . . . . . . . . . . . . . . . •. . . . . . . . . . . . . . . . . . . . . . . . .

10 L

v

t I

--v-+"

1

. . . .

.~

d

0

1 /.

I

9

10

OA~ . . . .

9

Ot 11

pH

I'\ I \

0

9 9 . _ %" o 0

ISS DIN

ro

0.1 ,TCLP

o

oo ~

+ I 12

_ 13

0.01

I

4

S

.I

1

6

7

.

1

1

I

l

I

8

9

10

11

12

pH

13

686 in landfills is rarely greater than 10 and test conditions manipulated higher than pH 10 will require different interpretation.

Japanese Leaching Test

The Japanese Leaching Test is based on extraction with water for 6 hours at L/S=20, however, the short duration of the test may not allow sufficient time for several elements to reach equilibrium. Testing results should resemble those obtained by the DIN and AFNOR procedures, although there is currently a paucity of leaching data on incinerator residues.

Swiss TVA (1988)

The Swiss test method generally produces results which are consistent with the pH dependent leaching curves discussed previously. The acceleration of carbonation provided by bubbling pure carbon dioxide through the leachant produces some side effects that are not representative of field conditions. This test is very sensitive to minor changes in pH because testing is carried out in the pH region where solubility changes quickly with pH. The test has also been shown to produce an exceptionally high release of oxy-anions. This is attributed to diminished retention capacity of the ash by the conversion of lime and basic calcium silicate phases to calcite, via additional carbon dioxide. The rate of retention of oxy-anions in other mineral phases is not as fast as the rate of release resulting from carbon dioxide injection, since these re-mineralisation reactions can be relatively slow. Slow sorption of oxy-anionic species, such as arsenate, selenite and molybclate, on hydrated ferric hydroxide phases is likely to occur in the field, but is too slow to be of significance during the laboratory leach test. This aspect needs to be considered for the relevance of the TVA test for oxy-anions.

USA, Califomia WET Test

The California WET test is a modification of the TCLP test which uses dilute citric acid as the leachant instead of acetic acid. This modification results in much greater complexation of metals than in TCLP. Results of the WET test indicate availability when evaluated on a release basis. Insufficient data is available to assess the comparability of WET test data for anionic species.

US EP Toxicity Test (1980), the Toxicity Characteristic Leaching Procedure (TCLP) (1990) and the Regulation 309 (now 347) Leach Procedure

The intended final pH of 5 of the EP Toxicity and TCLP tests is in a pH region where major changes in metal solubility can result with minor variations in pH. The final pH of the leachate from both tests is highly dependent on the acid neutralisation capacity of the waste, and incinerator residues typically require much more than the maximum 2 meq of acid per gram of ash addition to bring the pH down to 5. However, variation in buffering capacity leads to inherent variability in the test results.

687

Alternative Approach Since many of these tests expose ash to conditions which typically would not prevail in most disposal scenarios, an alternative approach is suggested for the regulatory evaluation of incinerator residues. The leaching properties of these materials are now well characterised. Testing of the materials should be focused on evaluating whether or not the sample being tested exhibits the same characteristic leaching behaviour for elements or species of concern. This would allow for implementation of management strategies which are most effective for the general material class and rejection of materials with characteristics beyond acceptable variability limits. An example of this approach would be to specify acceptable ranges of alkalinity and leachability in relevant pH domains, in particular at neutral pH. This would be valid for disposal and for utilisation. However, the criteria for acceptance would be different. 16.6.2 Systematic Leaching Behaviour Among Different Incinerator Residues Streams and Other Wastes Figure 16.14 presents a comparison of the release (mg/kg) of several constituents from different residue streams. It is striking to note that the curves show many similarities in spite the substantial differences in composition and origin. The solubility control in the pH region 6 to 11 is very pronounced for the metals. The primary difference observed between the residues is the total availability, which is generally higher for the combined ash than for bottom ash. For the APC residues, the availability can be one to two orders of magnitude greater than for bottom ash. The grate siftings typically indicate approximately an order of magnitude greater availability for Cu and Pb compared to bottom ash. The behaviour of Ba is very consistent for all incinerator residues. The solubility of Ba was independent of pH for most residues except from boiler ash, which decreases with increasing pH. The key to this behaviour is the influence of sulphate, which was not reported, because BaSO4 is the solubility controlling phase in most cases. The most significant difference between incinerator residues was noted for Cd, where the greater chloride concentrations in the APC residues results in increased Cd leachability at alkaline pH. For Cu, low content of organic matter in certain residues (e.g., ESP ash, boiler ash) tends to lead to low leachability of Cu. However, LOI is not a proper measure for the possible increased Cu leachability, because char alone does not affect Cu leachability (see section 16.2.3.1). Cr is more soluble in the boiler ash and SD-FF residue than the grate siftings and the D-FF residue. The reason for this is unclear but may be related to Cr speciation (e.g., Cr.6 is more soluble at neutral and alkaline pH than Cr*3). The leachability of Zn is the most consistent for all residues. Even the availability does not differ more than a factor of three to five. Table 16.6 provides a comparison of the total content, availability, and leachable fraction of several elements in the different residue streams. The availability is

688

Figure 16.14 Comparison of the Leaching of Selected Metals as a Function of pH from Bottom Ash, Combined Ash, ESP Filter Ash, Fabric Filter Ash, Grate Siftings and Boiler Ash 100 -

A Dry-FF= Grate sil'dngs o Boiler ash

10-

9Bottom ash o8o

~ o

t~t

E

r

I=

8

0,1

B

0,01

0,001

....

0

I ....

~ ....

I ....

I

2

4

6

8

....

I . . . .

I

10

....

12

I

14

pH 100 & Dry-FF a Grate siftings 9Boiler ash

10

9 AA,,

~

Ok

4>

=_.

~o

~

ee

A

9Bottom ash

E

G

0,1

O O

0,01

Ba

0,001 2

4

6

8 pH

10

12

14

689

Figure 16.14 Continued 1000 & Semi-dry-FF o Combined ash

106

+ESPash 9D n/-FF [] Grate siQngs

10

9Boiler ash J

o

9 0

*

8

o

0 (~) 0

0,0.%

~8 O8c~#A

(~U

0,01

....

a

I ....

I ....

4

9

+

O

2

9

o

9O

0,001

9

040

0,1

9Bottom ash

9

+ 9

9

+

O

I ....

6

I

pH

....

.~

+e

I ....

8

I

10

....

12

I

14

10000,00 9Bottom ash A Semi-dry-FF

1000,00

n Combined ash + ESP ash

100,00

00

9 00

Dry-FF [] Grate si~ngs

E (J

§176

10,00

9Boiler ash

+

O

cJ 1,00

O

O O ~

[] &

9

+

A

0,10

0,01

Pb

....

9

I ....

I ....

2

4

I ....

6

I ....

8 pH

+

+

I ....

10

I ....

12

1

14

690

Figure 16.14 Continued 100,00 I

A Semi-dry-FF & Dry-FF , Boiler ash + ESP ash o Combined ash

+ A , ~ A'~ d~A&

+

10,00

9

O

1,00 -~

9 & ~o t eSO,e~ + oo~ o" ~o ~

%

E

r

0,10

O

O

I

%0 * ~,,~,,

0,01

Cd

0,00

9 ~176 I

0,00 0

2

4

6

pH

8

10

12

14

100,00 9Bottom ash A Semi-dry-FF o Combined ash ,, Dry-FF o Grate sifdngs o Boiler ash

51,o~ o o

10,00

--~ 1,00 E

~

c

0,10

0

~%

9149

n

Cr

0,01

~ &

O

A

r-t r-k

l.Detection

0,00

. . . .

0

I ....

I ....

I ....

t

2

4

6

8

..................................

pH

~t,

r'n

limit {

....

I ....

10

I ....

12

I

14

691 Figure 16.14 C o n t i n u e d

10000,00 9Bottom ash 1000,00

qb Oo~

9 _=i_A&.~

9Semi-dry-FF

z~

oE~h

o Combined ash o Grate siltings

100,00 e

Ot

E

'-:'. U c O

& Dry-FF

0 9

e Boiler ash

0 9 o,t-

+ ESP ash

o

10,00

&

,&

&

cJ

00

1,00

+

[]

Zn

0,10

9

9

O

9~p

+ogL

O

0,01

....

0

I

2

....

i

4

....

I ....

6

r

....

8 pH

II I ....

10

I ....

12

I

14

Table 16.6 Total Content and Availability of Several Elements for Each MSWl Residue Stream Availability is presented on the basis of release per mass of ash, release per mass of MSWl incinerated, and fraction of total content Stream

As Min

Availabilitv (malMo M S ~ Bottom ash 0.09 Grate siflings Boiler ash Filter ash APC resiaue Total available mass per element

B Max

Min

Ba

Ca

Max

Min

Max

Min

Cd Max

Min

Max

21000 0.15 1050 <0.01

1.5 <0.01

1.5

15 0.75

60 3

15

60

6000 450

5

40

80

200

500

1800

50000

90000

30 50

100 100

30 30

80 80

1500 100

3000 800

50000 50000

200000 200000

Availability ima/ka ash) Bottom ash Grate siftings Boiler ash Filter ash APC residue Total content (malka ash) Bottom ash Grate siftings Boiler ash Filter ash APC residue

Fraction Available (-1 Bottom ash 0.06 0.13 0.63 1.00 0.10 0.11 0.40 Grate siflings Boiler ash 1.33 1.00 Filter ash 0.03 0.02 0.00 0.00 0.02 0.03 1.00 APC residue 1 .OO Total mass of individual streams (mglMg MSW): Bottom ash. 300;Grate siftings. 15:Boiler ash. 5;Filter ash. 20:APC residue. 12.

0.78 0.50 0.30

2 0.5 100 100 300 0.25 0.~0 0.05 l.pO

25 1 500 500 900 0.20 0.10 0.02 0.60

Table 16.6 Continued Stream Availability lmalMa M-sw Bottom ash Grate siftinas Boiler ash Filter ash APC residue Total available mass per element Availability (malka ash) Bottom ash Grate siflings Boiler ash Filter ash APC residue Total content Imalka ash) Bottom ash Grate siftings Boiler ash Filter ash APC residue

Min

CI

Max

Min

Cr

Max

Min

Cu

Max

Min

Hg

Max

Min

K

Max

300 3 f5 600 360 1278

1800 6 35 1000 720 3561

0.6 0.03 0.005 0.2 0.06 0.90

3 0.15 0.05 2 0.24 5.44

15 30 0.01 0.1 0.01 45

60 90 0.15 0.4 1.2 151

<0.01

0.03

300

900

0.02 0.05 0.07

0.06 0.12 0.21

10 200 120 630

20 500 360 1780

1000 200 3000 30000 30000

6000 400 7000 50000 60000

2 2 1 10

50 2000 1 5 1

200 6000 30 20 100

0.01

0.1

1000

3000

5

10 10 10 100 20

1 4

3 10

2000 10000 10000

4000 25000 30000

1000

3000

200

1000

1200

2500

0.5

1

7000

20000

30000 300 60000 400 100

800 800 500

750 300 800

2000 1000 2000

1

20

20000 20000

40000 40000

8000 30000

Fraction Available (-1 Bottom ash 1.00 2.00 0.01 0.01 0.04 0.08 Grate siftings Boiler ash 0.38 0.23 0.00 0.01 0.00 0.02 Filter ash 1.00 0.83 0.03 0.13 0.02 0.02 APC residue Total mass of individual streams (mgIMg MSW): Bottom ash, 300;Grate siftings, 15;Boiler ash, 5;Filter ash, 20;APC residue, 12.

0.02

0.10

0.14

0.15

1.00

0.15

0.50 0.50

0.63 0.75

Table 16.6 Continued Stream Availability [ma/Mo MSW) Bottom ash Grate siftings Boiler ash Filter ash APC residue Total available mass per element Availability lmolka ash) Bottom ash Grate siftings Boiler ash Filter ash APC residue Total content fmolka ash1 Bottom ash Grate siftings Boiler ash Filter ash APC residue

Min

Mg

Min

0.3

I .2

80

300 0.02

0.08

380

1200 0.32

1.28

900

Min

Mo

Max

300

Max

Pb

Sb

Max

Min

15 75 0.1 2 I .2 93.3

90 135 0.5 6 60 292

0.3

0.6

0.01

0.02

0.31

0.62

50 5000 20 100 100

300 9000 100 300 5000

I000

3000

I

4

4000

15000

I

4

10000

30000

5

30

1500

10000

30000

20 5

60 25

2000 4000 5000

Max

Min

so4

Max

Min

Zn

Max

2400 45 150 600 360 3555

5400 75 400 1600 1200 8675

15 30 15 I00 120 280

150 75 35 160 240 660

8000 3000 30000 30000 30000

18000 5000 80000 80000 I00000

50 2000 3000 5000 I0000

500 5000 7000 8000 20000

1

2

0.5

I

3000

30

200

12000

30000

2000

4000

I0000 8000 20000

150 50

500 950

30000

90000

10000 I0000 5000

20000 20000 40000

Fraction Available (-1 Bottom ash 0.10 0.10 0.20 0.13 0.03 0.10 Grate siftings 0.01 0.01 Boiler ash Filter ash 0.40 0.50 0.05 0.07 0.03 0.04 APC residue Total mass of individual streams (mglMg MSW): Bottom ash, 300;Grate siftings, 15;Boiler ash, 5;Filter ash, 20;APC residue, 12.

0.03

0.01

0.67

0.60

0.03

0.13

0.00

0.00

1.00

0.89

0.30 0.50

0.35 0.40

695 presented both on the basis of the content within the individual residue stream (mg/kg residue) and on the basis of the original MSW feed (mg/Mg MSW). Calculation of availability on the basis of the original feed permits comparison of the relative contribution of the availability from a specific residue stream to the overall release potential from all waste streams. Thus, the relative mass of each residue stream is considered. Using this approach, the following observations can be made. The total amount of Cd available for all residues is approximately 6 g/Mg MSW, of which 0.8 g/Mg is present in the bottom ash, 5 g/Mg is present in the APC residues, less than 0.001 g/Mg is present in the grate siftings and 0.03 kg/Mg is present in the boiler ash. Similarly, the total amount of CI available for all residues is approximately 1400 g/Mg MSW, of which 450 g/Mg is present in the bottom ash, 800 g/Mg is present in the APC residues, 3 g/Mg is present in the grate siftings and 20 g/Mg is present in the boiler ash. The total amount of Zn available for all residues is approximately 350 g/Mg MSW, of which 80 g/Mg is present in the bottom ash, 200 g/Mg is present in the APC residues, 40 g/Mg is present in the grate siftings and 20 g/Mg is present in the boiler ash. Consequently, separating the APC residue from the other streams would make sense, since the majority of the available elemental inventory is in the APC residue Conversely, the total amount of Cr available for all residues is approximately 3 g/Mg MSW, of which 2 g/Mg is present in the bottom ash, 1 g/Mg is present in the APC residues, 0.01 g/Mg is present in the grate siftings and 0.03 g/Mg is present in the boiler ash. The total amount of Cu available for all residues is approximately 80 g/Mg MSW, of which 30 g/Mg is present in the bottom ash, 0.8 g/Mg is present in the APC residues, 45 g/Mg is present in the grate siftings and 0.1 g/Mg is present in the boiler ash. The total amount of Pb available for all residues is approximately 150 g/Mg MSW, of which 45 g/Mg is present in the bottom ash, 30 g/Mg is present in the APC residues, 70 g/Mg is present in the grate siftings and 0.3 g/Mg is present in the boiler ash. Since the greatest inventory of these metals resides in the bottom ash or grate siftings, separation of these streams from other streams would not make much difference. However, segregating the grate siftings would be advantageous for reducing Pb and Cu Ioadings in bottom ash destined for utilisation. Figure 16.15 provides a comparison of release as a function of pH for Cd, Cu, Pb and Zn in six different waste materials of widely different origin (van der Sloot, 1994). The wastes compared include ash from a refuse derived fuel incinerator (RDF ash), bottom ash from mass burn systems, ESP ash (fly ash), waste from car shredders (Shredder waste), soil amended with sewage treatment sludge (Sewage amended soil), coal fly ash and a natural soil (Terra rossa). Although the general trends in release as a function of pH are very similar, the absolute release for each varies by several orders of magnitude in the pH regime where availability is controlled. It is clear that the major elements, dictating the leachate composition for the most part, also control the leachability of trace contaminants to a large extent. The differences between the behaviour of individual elements in different wastes can be attributed to specific factors,

696 Figure 16.15 Release of Cd, Cu, Pb and Zn as a Function of pH from Different Waste Materials in Comparison with MSWI Residues lO00

1000

100

100

Cd

10

lO

0.1 O.1 t~

0.01

0.01

0.001

I...4

o

0.001 ILRPUSB DHRIVBD

0.0001

9

3

|

4

,

|

5

.

,

6

|

.

7

PUHL ,

J

8

.

O.O00l

ASH |

9

.

i

.

,

.

,

COAL~LYASH i

.

10 11 12 13

3

1000

1000

100

100

i

4

a

,

I

5

J

6

,

|

~. .

7

|

.

|

.

!

,

i

,

i

,

8

9

10 11 12 13

................... 3 4 5 6 7 8

9

10 11 12 13

10

0.1 0.01

0.01

~,~

0.001

O.OO1

M S W I Fly ash ...................

0.0001 3

4

5

6

7

SHREDDER

0.0001 g

9

10

II

12

13

1000

1000

100

lO0

WASTE

Sewage Amended Soil

10

\,,"~

0.1

0.1

o.oi

0.01 O.OO1

0.001 MSW

0.0001 3

INCII'IBRATOR

BOTTOM

..................... 4 5 6 7 8

ASH

0.0001 9

10 11 12 13

3

...... i ............ [ 4

5

6

7

8

9

I0

II

pH

12

13

697 Figure 16.15 Continued 5000.00

'~

1000.00

1000

100.00

I00

10.00

10

Coal fly ash

o

9

a

0.10

1~

0.01

0

.

0.1 0.01 .--

0.00

1

3

i

4

|

J

5

|

f

6

|

i

~

7

,

|

~

g

i

~

9

l0

J

,

|

i

J

11 12 13

0.001 4

1000

1000.00

100

100.00

10

10.00

5

6

7

8

9

10

11

12

13

1.00 0.10

0.1 0.01

MSWI fly ash

0.001 4

5

6

7

8

9

0.01 i 10

i 11

12

Shredder waste ,

0.00

I

3

13

!

4

,

I

5

|

I

6

i

I

7

,

i

8

,

!

9

9

!

,

i

,

i

,

10 11 12 13

5000.00 1000.00

1000

100.00

loo

10.00

10

1.00

1

0.10

0.1

MSWI bottom ash

0.01 0.00

. . . . . . . . . . . . . . . . . . . 3

4

5

6

7

8

9

10 11 12 13

O.Ol 0.001

Sewage amended soil +

f F

L

i

t

J

l

Terra rossa ,

3

4

5

6

7

8

9

l

~

t.

10 11 12 13

pH

698

Figure 16.15 Continued 1000 100

Coal

fly ash

100 10

r I:~ o r~

Pb

10 1

1

0.1

0.1

§

a

o o

:A

\

: ~ 1

0.01

0.01

RDF ash 0.001

' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 3

4

5

6

7

8

9

10 11 12

9

0.001 13

3

1

,

I

4

.

I

t

.

5

6

,

!

J .

7

$

i

.

9

|

,

10

.

,

11

|

.

12 13

100 100 10

1

0.1

0.1

0.01

0.01

Shredder waste

MSWI Fly ash 0.001

.

.

3

.

.

5

.

.

.

7

.

.

9

,

0.001 11

13

3

I000

100

100

10

10

i

,

4

i

,

5

I

,

i

6

,

7

I

,

8

!

,

9

I

i

I

10

,

11

i

,

12

13

Sewage amended soil

1 A

0.1

0.1

0.01

0.01

MSWI bottom ash 0.001 4

A

l

J

,

I

I

J

I

l

5

6

7

8

9

10

11

12

0.001 13

3

I

I

1

t

I

I

I

1

t

4

5

6

7

8

9

10

11

12

pH

13

699

Figure 16.15 C o n t i n u e d 100000

50000

10000

~o

I0000

I000

1000

100

100

o

1

o o a~

0.1

Zn

0.1

O.Ol

R D F ash l

0.001 3

l

4

,

I

,

I

5

,

6

O.Ol I

~

7

I

,

8

I

,

l

9

,

l

!

I

Coal fly ash

l

10 11 12 13

50000

0.001 3

4"4.

*

'

'

~

~

'

'

J

4

5

6

7

8

9

'~

~

l0 11 12 13

50000

10000

Shredder waste

10000

M S W I fly ash

1000

I000

100

100 -

10

10

0.1

0.1

0.01

0.01

0.001

. 3

.

. 5

.

. 7

.

.

. 9

0.001

. 11

3

13

50000

i

i

I

i

f

l

4

5

6

7

8

9

i

i

i

10 11 12 13

50000 M S W I b o t t o m ash

10000

10000

1000

1000

100

100

I

Sewage amended

soil

10

0.1

0.1

0.01

0.01

0.001 3

. 4

. . . . 5 6 7

. 8

.

. 9

. 10 11 12 13

Terra rossa

0.001 3

I

I

I

t

I

1

!

1

4

5

6

7

8

9

I0

II

pH

,

I

12 13

700 such as the presence of high concentrations of dissolved organic matter (e.g., humic and fulvic substances) in sewage sludge amended soil. In the case of Cu, the higher leachability from bottom ash is related to the presence of dissolved organic matter (DOC) in bottom ash, which is absent in coal fly ash. The high leachability of Cu from shredder waste is directly related to the high amount of dissolved organic matter characteristic for this waste. Cd is another example of an element which behaves differently in different wastes. For Cd the influence of CI on release is substantial. The release of Cd from a variety of materials (RDF ash - Bottom ash - Fly ash - Shredder waste - Coal fly ash) with different CI levels shows the sensitivity to the CI level. The increase in Cd release at higher pH with increasing CI concentration in the leachate is very pronounced and can be modeled with geochemical speciation models (Allison et al., 1991 ). The relationship between bottom ash, fly ash and RDF ash form an interesting sequence in this respect, because of increasing CI content, respectively. The Zn leachability curves for different wastes are very consistent. A slight shift in solubility with pH occurs for RDF ash as a result of the high CI content in this waste. Some mobilisation of Zn through DOC complexation is noted for shredder waste.

16.7 LEACHING OF ORGANIC CONSTITUENTS The relevance of organic constituents in incinerator residues was discussed in Chapter 8. This Chapter follows that discussion along by describing the potential trace element complexation with organic compounds. Organic constituents in incinerator residues can be classified as: Organic contaminants of potential human health concern, such as PCBs, PCDDs, PCDFs, chlorophenols and PAHs Low molecular weight organic acids and related soluble organic species (e.g., phenols, sugars, etc.) Humic and fulvic type substances resulting from microbial activity in an ash repository Organic contaminants of potential human health concern can be grouped as polar (chlorophenols and chlorobenzenes) and non-polar compounds (PAHs, PCBs, PCDFs and PCDDs). The leaching behaviour of the relatively polar organic species can be described in a similar manner as inorganic species, where solubility and liquid-solid partitioning controls release. The significance of a leaching test with water is questionable for non-polar compounds with a very low water solubility because the main mechanism of release is not dissolution. Here the role of "facilitated transport" (McCarthy and Zachara, 1989) is important. To assess leachability of non-polar organic species by leaching with water the main focus should be on the potential by-

701 products of degradation of organic matter, which can then facilitate transport of nonpolar compounds. A method proposed for assessing the magnitude of this fraction is an extraction using 1 N KOH (van der Sloot, 1992). This releases a fraction of organic matter that may be considered readily available for degradation. It is speculated that the organic contaminants associated with this fraction can be used as a first estimate of potential long term release of non-polar organic contaminants, however, further work is needed to explore this approach. Organic contaminants of potential human health concern are present in greatest concentrations in the APC residues. The NITEP program included extensive testing for organic contaminants in different APC residues (Environment Canada, 1993). In a system containing consecutively a wet/dry spray reactor, a dry reactor and a fabric filter dust collector, the composition and leachability of PCDD, PCDF, PCB, PAH, chlorobenzenes and chlorophenols were measured. Organic contaminants were not detected in any of the leachates with the exception of chlorophenols, of which less than 10% of the total quantity was solubilised. After leaching, the leached residues were analysed again and the concentrations of organic contaminants were corrected for the loss of soluble solids. This resulted in virtually complete recovery of PCDD, PCDF, PAH, chlorobenzenes and chlorophenols from the solid, indicating that within the analytical limitations, no significant fraction of the organic contaminants were leached from the ashes. However, some data has indicated that there is another potential mechanism for release of organic compounds, namely colloidal transport. This process may explain some of the findings of water insoluble organic contaminants at distances from a site beyond the expectations from laboratory testing (McCarthy and Zachara,1989). Fortunately, the organic matter content of APC residues is very small, thus minimising the risk of release. Alternatively, bottom ash which can have a significant organic matter content, does not contain appreciable concentrations of the organic contaminants of concern (Sawell et al., 1986), thus the issue of leaching of organic contaminants from incinerator residues is not considered a problem. The concentration and significance of these organic constituents in the residue streams is different. Low molecular weight organic acids, humic and fulvic acids are most prevalent in bottom ash as a result of their presence either in the ash, the degradation of incompletely combusted organic matter, or through alkali hydrolysis of higher molecular weight organic matter. Leaching of these organic species can have a significant effect on the short-term leachability of specific elements, such as Cu. In addition to the influence of these compounds, the presence of unburned organic matter may also serve to generate reducing conditions within the ash matrix, which can decrease the release of heavy metals. The dissolution of low molecular weight organic acids and humic and fulvic compounds have been shown to be significantly increased at high pH due to alkali hydrolysis. At pH greater than 12, the extent of dissolution is greatest (Figure 16.6c). At pH values

702 below 9 the leachability of humic substances is probably insignificant, whereas the low molecular weight organic acids and fulvic substances may continue to play a role in the complexation of metals.

16.8 EFFECTS OF INCINERATOR OPERATION ON LEACHING 16.8.1 Combustion Efficiency (Burn out) and Facility Operation It appears that the variability in the waste stream composition masks most possible differences in leaching behaviour between installations. In a detailed study carried out in the NITEP program (Sawell et al., 1988, 1989, 1991)included investigation of the impact of waste feed rate, furnace temperature, air supply, temperatures in the flue gas system and lime feed rates on residue properties. Given that the leaching of most trace elements is dominated by the major element chemistry, the influence of facility operation within the normal bounds of good combustion on bottom ash leaching is a secondary effect. Overall, poorer burnout tended to reduce the release of most trace metals, either due to dilution from the increased mass of material, or through sorption onto activated carbon in the ash. Significant changes in APC residue pH (Figure 16.16), alkalinity and chloride content (Figure 16.17) were observed as a function of lime feed stoichiometry and operation (Sawell et al., 1988). This can have a significant influence on results observed for regulatory leaching tests because the extraction pH is controlled by the residue alkalinity, which in turn controls measured concentrations for many elements.

16.8.2 Waste Feed Composition The influence of waste feed composition on leachability is complex for several reasons: The speciation of a particular element in the waste feed can significantly effect the partitioning of that element between the bottom ash, grate siftings and APC residues. Although the overall inventory of an element may change, the change may only have a limited effect on the overall available fraction of that element across all residue streams, especially if the availability is only a small fraction of the total amount present in the residues. Thus, changes in the speciation of an element present in the feed may be as important as the total quantity in the input. Changes in the availability of an element may not effect the solubility controlled release of that element under relevant testing or field conditions. Thus, significant reductions in input of a specific element may have no noticeable effect on regulatory testing results or impact on field leachate quality.

703 Figure 16.16 Influence of MSW Operation on pH of ESP Ash 12

10

ESP Ash

:

tt~

8

ESP Ash Good Conditions

6

.

1

!

I

2

3

4

Cycle

Figure 16.17 Influence of ESP Temperature on the CI Level in Precipitator Residues 250 r

-

I Fr2

~O w,,,,t

GOOD D PT5000D

200

O O

r~ tD

150 C

O

L

r,.)

PT4 POOR~,

100 190

OOOD

200

I

i

210

220

ESP Temperature (~

PTI4 POORI ~

230

240

704 The most significant changes in leaching of trace elements may be achieved by modification of the major element composition or chemistry of the residues rather than by modification of the input of trace elements of concern. This is because the major elements typically control residue alkalinity, solubility and availability in many cases. A limited number of studies have been conducted using sorted waste as a feed to an incinerator to evaluate the impact on leachability. In the WASTE Program (1993) experiments were carried out where the waste feed to a mass burn incinerator was spiked with lead-acid batteries, Cd in the form of PVC pellets, Cd as a Cd-benzoate solution, and a combination of all these materials simultaneously. The various residue streams were sampled and subjected to both the Sequential Chemical and Sequential Batch Extraction Procedures. In general, although the total concentrations of these elements increased in certain residue streams due to the spiking, the solubility of the elements did not change appreciably based on results gleaned using the SCE procedure. For example, the increase of concentrations of lead in the grate siftings did not increase the fraction of lead available for leaching, because much of the increase was attributed to elemental lead which melted on and dripped through the grates. Since elemental lead (coated with a PbO layer)is relatively insoluble, the spiked lead did not contribute to the leachable fraction. Furthermore, results from a distilled water leach test indicated that the fraction of leachable Cd and Pb actually decreased in the fly ash residues from the spiked runs (see Figures 16.18 and 16.19). Consequently, the species of an element is the most important consideration when assessing change in the leaching characteristics of a residue. The impacts of three waste compositions were compared in a study carried out with the Hague incinerator (NOH/RIVM, 1991). The three waste compositions were: i.

MSW as collected from the curbside MSW with putrescibles, glass, paper and domestic chemical waste removed, resulting in a feed analogous to thorough source separation, and

iii.

MSW preprocessed in the same way as (ii) above, but also crushed on-site to create a more homogeneous feed

Each type of waste was incinerated over a three day period and bottom ash samples from each day of operation were leached using an L/S=20 with distilled water. The most significant impact on the leaching test results was a reduction in leachate pH from a range of 7.4-11.8 for Case i to 5.5-8.2 for Case ii. This change could not be attributed to the removal of waste components because the pH range for Case iii was 10.8 - 11.6. No significant impact was observed on the leachate concentrations for Cu, Cr, Pb or Mo. Results for Cu and Pb are presented in Figure 16.20. Cu leaching relative to the unified pH curve may have been reduced due to less soluble organic matter present in the ash, but the results were inconclusive.

Figure 16.18 Fraction of Cadmium Leached During the SBEP - Stabiliser Spike

Ash

Siftings

/ vA Stabiliser Spike 1 F=l

Pass

Scrubber

Filter

RESIDUE TYPE Stabiliser Spike 2

No Spike

I

Figure 16.19 Fraction of Cadmium Leached During the SBEP - Pigment Spike

5q (%) O3HOV37 NOI.LOVUJ Ash

/

Siftings

Pass

Scrubber

Filter

RESIDUE TYPE Stabiliser Spike 1

Stabiliser Spike 2

Concentrations expressed in microgramll in a batch test at LS=20.

I No ] Spike

707

Figure 16.20 Effects of Changes in MSW Feed on Leachability of Cu and Pb in Bottom Ash from the Hague Incinerator 10

1

~

Unified pH

curve

MSW.

=d 0,1

+

=t

.t4-

0

M S W w/o GFT

m

MSWw/oGFT andsizr reduced

+

3-year regular MSW

-I-

0,01

I 6

'

'

'

I 8

,

'

pH

'

I

'

'

'

10

I

'

'

'

I

14

12

10

Unified pH curve 4,

_o 0,1

o+ ~ +

8 0,01

"4"

o MSWw/oGFT

0 4...~:ft-I" +

0,001 ! . 4

9. 6

8

pH

10

12

MSW.

14

]l

M S W w/o G ~

+

3-year regular MSW

r~uc~d

and size

708 These relationships clearly indicate that knowing which factors have the greatest effect on leachability is important to facilitate significant improvements in leachate quality or release of critical contaminants during residue management. This does not imply that source separation of certain materials such as automobile and household batteries or Hg thermometers is a futile activity. Clearly, if a specific waste contributes a large fraction of a trace element to the overall loading, its separation is fully warranted. Moreover, if a specific waste contributes a large fraction of a potentially leachable metal, it should be targeted. However, the justification for source separation should not be focused primarily on reducing ash leachability. Some measures, such as removal of metallic lead (Pb ~ or copper (Cu~ may reduce the inventory of these elements, but may have no effect on ash leachability. Overall, the fate of constituents in waste components should be evaluated in more detail to identify their potential impact. The role of changes in bottom ash alkalinity due to changes in waste input also should be examined further because of the potential impact on release of many elements.

16.8.3 Seasonal Variations in Leaching The leaching data obtained from a 6 year waste characterisation program (WAV, 1988, 1992, 1993) have been evaluated for differences within installations and for differences over the seasons. Bottom ash samples were leached using a batch extraction method with distilled water at an L/S of 20:1. Differences between installations as a result of MSW input were generally more significant than variability in ash leachability over the seasons. Identification of the causes for differences in the feed to an incinerator, such as commercial waste, may provide clues to the possible reason for differences. The 6 installations clearly form two groups: those in the municipality of Rotterdam and the others (rural). Figure 16.21 presents the weekly variability in ash leachate pH and Pb concentration from one facility. Increases in Pb concentrations were correlated with decreases in leachate pH. Seasonal factors have been suggested to effect the MSW feed to an incinerator and possibly impact on the leaching of the residues. Results for the seasonal impact on the leaching of 8 elements are presented in Figure 16.22. Slight increases in pH (= 0.2 pH units) were observed during the summer in predominantly residential areas. This effect may have resulted from increased quantities of grass clippings and other garden waste, however, no significant seasonal effect was observed for As, Cd, or Cr. The following additional observations were made:

Copper

There appeared to be two clusters of installations, urban-based facilities with relatively higher levels of Cu leachability and those rural facilities with lower levels of Cu leachability. Bottom ash leachates from the 4 urban incinerators did indicate some

Figure 16.21 Variability in pH (A) and Pb ( 0 )Concentration Using a Serial Batch Extraction with Bottom Ash from One Installation Over a Period of Two Years

(1/5d) qd "3 u oo

Time (yr,wk)

Figure 16.22 Seasonal Variations in Concentrations of Selected Metals in Extracts of Bottom Ash from 6 installations

Wln Spr Sum AutAVO a11

Wia

Spr Sum AutAVGaII

Wln Spr Sum AnAVO all

0

W h Spt Sum AntAVO all

0

Win Spr Sum AutAVG all

concentrations expressed in pgll in a batch test at LS=20

Win Spr Sum A d A V O all

Wln

0

Spr Sum AntAVO all

Win Spr Sum AutAVG 111

711 seasonal effect with maximum Cu solubility observed in the winter months. The main factor here may have been the presence of dissolved organic matter capable of complexing copper to varying degrees.

Molybdenum

Some seasonal variability in the Mo leachability was observed for the urban facilities with a maximum in summer. The relatively high Mo input likely to be from a (small) industrial input to the incinerator. The use of Mo in lubricants has been postulated as a possible source of such inputs.

Nickel

Except for the autumn period the Ni leachability is very consistent between installations and between seasons. In autumn there are two installations with a significantly higher leachability in comparison with the other installations. The reasons for this higher mobility is unknown.

Lead

The Pb leaching data are fairly consistent over the seasons. Two installations show consistently higher Pb leachability than the others. This may be related both to input as well as to small differences in pH.

Zinc On average the highest levels of zinc leachability were observed in the summer periods for two installations. The difference between the installations is also significantly different based on an annual average. A seasonal trend indicating higher leachability in summer and low leachability in winter appears to exist for three of the installations from an urban origin. The cause of this effect is not clear. 16.8.4 Quench Water Quality Depending on the quantities of make-up water added, the quality of the quench water should be very similar in composition to the first leachate from a column leaching test on freshly collected ash. Generally, it has been observed that quench water contains high concentrations of salts and relatively high Cu levels. Table 16.7 compares quench water composition column data at low LS and field lysimeter leachate (Reimann, 1990; Chandler, 1993; AVR, 1995). The results indicate strong similarities between the quench water and leachate. A significant deviation is the high Ba content measured in data provided from Canada. The reason for the difference could be that the high carbon content in the ashes from this two-stage combustion system produced strong reducing conditions in the quench tank. As a result, sulphate was probably reduced to

712 sulphide which, in turn, increased the leachability of Ba. The exceptionally low Cd and Pb concentrations measured in this quench water are consistent with this hypothesis. The Cd concentration observed in quench waters also is directly related to the CI concentration and pH. The relatively high Pb and Zn concentrations in AVR quench water are consistent with the pH dependent leaching behaviour of bottom ash for these elements. For quench water quality, the pH (generally high) and the low LS are crucial factors for the concentration levels observed. Solubility control is already important in quench water, in spite of the limited contact time. Table 16.7 Comparison of Quench Water Quality, Laboratory Column Leachate and Field Lysimeter Pore Water (mg/I) .... Element

Quench water

Column leachate (L/S=0.2)

[Reiman, 1990] [Chandler, 1993] [Versluijset al, 1993] As

0.003

Ba

< 0.1

0.002

9.4

0.1

< 0.050

0.020

Lysimeter pore water Hjelmar,1991 0.006

Cd

0.15

CI

1540

6000

4500

Cr

0.1

0.01

0.005

Cu

0.26

0.02

0.3

F

1.7

1

Hg

0.015

Mo Ni

0.25

Pb

0.8

SO4

1500

< 0.200

0.001

2

0.9

0.4

0.18

0.02

0.02

5000

1500

Zn

0.1

0.15

pH

9.0

Eh

9.3 150

713

16.9 EFFECTS OF RESIDUE PROCESSING AND MANAGEMENT ON LEACHING 16.9.1 Size-Reduction And Size Fractionation Testing of bottom ash in the laboratory will usually require size reduction (see Chapter 7). Size reduction influences the leaching of bottom ash by exposing fresh surfaces which contains hydroxides which have not absorbed carbon dioxide and been converted to carbonates. This results in an observed increase in pH when the ash is extracted. Size fractionation of bottom ash also may effect leaching behaviour because of varying composition as a function of particle size. The influence of particle size on the leachability of bottom ash has been studied by testing different size fractions (SOSUV, 1989) which were prepared with and without size reduction. Ash samples were separated into the following sieve fractions: 4 - 31.5 mm 1 -4mm 0.5 -1 mm 0.125 - 0.5 mm <0.125 mm In addition, an aliquot of both the intact and size reduced (<3 mm) samples were subjected to a batch extraction at L/S=5 with distilled water. Analysis of chemical composition and the availability leach test was also carried out on sample (further size reduction of the respective fractions to <0.125 mm was required). Figure 16.23 presents results of the chemical analysis and leaching tests for Ca, Na, Cu, Mo, Pb and Zn in the different size fractions. Total sodium content increased with increasing particle size. This effect was not as pronounced for the other elements. Availability varied by a factor of 2 to 5 over a particle size range of more than two orders of magnitude (<0.125 - 31.5 mm), with the exception of Na which was depleted in the coarse fraction. In almost all cases, the leachability was lowest in the coarse fraction and highest in the fraction less than 0.125 mm. The nominal particle size in this fraction was well below 0.125 mm, which indicates that the general relation between particle size and leachability is related to external surface area. The final pH in the batch extractions varied from 11.5 to 12.5. The release of Pb in the batch extraction reflects the difference in pH as a function of particle size. Batch extraction results are presented as function of pH along with the respective pH-solubility curve for Cu, Mo, Pb and Zn in Figure 16.24. The primary effect of particle size reduction therefore may be a change in pH with its consequent effect on metal solubility. Laboratory test results on size reduced bottom ash (column dia. 5 cm, h=20 cm) has also been compared with data obtained from large columns (column dia. 65 cm, h=275 cm) containing untreated bottom ash, operated both in up flow and down flow

Figure 16.23 Influence of Size Fractionation and Size Reduction on the Leachability of Ca, Cu, Na, Mg, Mo, Pb and Zn from Bottom Ash Total sieved

P

6 Available sieved

A LS=S sieved

T o t a l reduced O m m

T o t a l as recieved

*Available as recieved

*Available reduced <3mm

A LS=5reduced O m m

A LS=5 as recieved

0,Ol

2

0,l

0.1

1 Particle sire (mm)

10

0,Ol

1 Particle size (mm)

0,l

1 Particle size (mm)

10

I

loo

10

I

loo

Figure 16.23 Continued

0,Ol

0,l

1 Particle size (mm)

10

0,Ol

0,l

1 Particle size (mm)

10

100

0,Ol

I

loo

0,Ol

0,l

0.1

1 Particle size (mm)

1 Particle size (rnm)

10

100

10

100

716

Figure 16.24 Corresponding Curve

I n c r e a s e in pH as a C o n s e q u e n c e of S i z e R e d u c t i o n a n d t h e C h a n g e in C o n c e n t r a t i o n of Pb a n d Z n C o m p a r e d to t h e U n i f i e d pH

10,& _

o,I E "-',

0,1-

I=

o

0,01 "

Pb 0,001

.... 10

0

. . . .

11

j 12

. . . . pH

i

i 14

13

100 -

10

E =_.

_

8 0,1-

2',n 0,01

. . . . 10

0 11

. . . .

~ 12

. . . . pH

I 13

'

I 14

717 configurations (Versluijs et al., 1990). Cumulative release at L/S=10 for several elements is presented in Table 16.8. An increase in pH of approximately 1.5 units was observed for the leachate from the small columns containing size reduced ash as compared to the leachate from the large columns. This result was similar to the effect observed for batch testing. This difference resulted in a different leachability in accordance with the pH - release relations established previously. The effect of the increased pH was most pronounced for Pb because the resulting initial pH for the small column was approximately 12, which is where the amphoteric Pb compounds become soluble. Table 16.8 Comparison of Cumulative Release (mg/kg) at L/S=10 for Small and Large Scale Laboratory Column Tests Element

Lab scale

Large scale

Availability

up flow

down flow

up flow

As

< 0.4

0.04-0.06

0.02-0.04

1.1

Ba

2.0

0.7

0.4-0.5

84

Cu

5.5

6.2

3.8

11.8

Cr

0.14

0.3-0.4

< 0.5

1.0

Mo

0.7

0.4-0.7

0.5-0.8

<1

Ni

0.01-0.1

0.08-0.40

0.2-0.5

1.0

Pb

0.4

0.01-0.05

0.01-0.05

24

Ca

5000

2500

3200

41000

K

490

975

950

800

Na

580

990

971

1100

CI

2040

2490

2560

2800

SO4

3990

5470

6750

13400

pH

12.1-11.8

10.3-8.7

9.8-8.5

4

The leaching data for size-reduced and size-screened ashes are consistent with the unified pH curves permitting conversion of the extraction data to comply with the pH of the actual ash sample. A crucial aspect of large column operation is avoidance of column effluent contact with air. Carbon dioxide uptake can result in a decreased pH in leachate samples and precipitation of elements such as Pb. This is especially a concern during collection of leachate from field lysimeters.

718

16.9.2 Storage and Aging The impact of ash storage and aging on leaching can be classified as: Lowering of pH due to uptake of CO2 from the air or biological activity Establishment of anoxic, reducing conditions including sulphide production resulting from biodegradation of residual organic matter Local reducing properties from evolution of hydrogen through the oxidation with water of reduced metals at high pH (e.g. AI, Zn, Pb), and Hydration and other changes in mineral phases, including cohesion of particles due to formation of calcite One of the most significant changes in bottom ash leaching behaviour as a function of aging is the change in pH from moderately alkaline (pH 9.5 - 10.0) to almost neutral as a result of uptake of carbon dioxide from the air and biological activity, as well as neutralisation reactions between acidic and alkaline components in the ash. At the same time the biological activity increases as a result of biodegradation of the residual unburned organic matter. This leads generally to a depletion of oxygen and the subsequent generation of reducing conditions in the ash. Sulphate is abundant to sulphidegenic biodegradation. These reactions can proceed for long periods of time. The presence native metals (AI, Cu, Zn, Pb) in the ash will lead to the formation of hydrogen, thus contributing to the reducing conditions is ash (Oberste-Padberg and Sweden, 1990). However, biological activity is generally considered to have the most pronounced effect on ash redox conditions. Recently the formation of clay minerals has been demonstrated in aged ash deposits (Zevenbergen et al., 1993). The resulting increase in cation exchange capacity may have an effect on the retention capabilities of ash. However, it will not affect the release of soluble salts and certain oxy-anions (e.g. Mo, B) and may be of minor importance relative to sulphide precipitation from biological activity. The changes observed in aged ash can for the most part be explained by changes in pH, redox properties of ash or complexation reactions. Re-mineralisation reactions may result in cohesion of ash particles which has been mistaken for pozzolanic reactions similar to those which occur in coal fly ash (Trondheim, 1990). It does provide some reduction in tortuosity as observed in testing compacted bottom ash (Kosson et al., 1993). Further work in this area is needed to be able to relate changes in ash leaching properties over the long term to parameters that can be quantified in through laboratory measurements. A comparison of release observed from fresh bottom ash and bottom ash which has been aged for 10 years is presented in Table 16.9 (Zevenbergen et al., 1993b). A

719 batch extraction at L/S=10 with distilled water was used to evaluate release. The pH of the ash decreased to neutral over the aging interval. However, significant release did not occurred based on the similarity in the Mo release, which is an element that should be readily depleted. Significant differences are only noted for Cu and Pb, which show an order of magnitude reduction in samples which were in the field for more than 10 years in comparison with fresh bottom ash. Table 16.9 Comparison of Batch Leaching Test Results (L/S=10 with Distilled Water) on Fresh Bottom Ash and Bottom Ash Aged for 10 Years in the Field (mg/kg) Element

Fresh bottom

Aged bottom ash (10 yr)

As

0.18

0.032

Cd

0.069

0.001 - 0.006

Cr

0.073

0.040 - 0.059

Cu

4.6

0.50

Mo

3.49

4.56

Ni

0.13

0.15

Pb

0.43

0.006 - 0.054

Sb

0.18

0.28

Zn

0.41

0.35

pH

9.8- 11

7-8

Stability of ash during laboratory storage is also an important consideration for sample analysis and archival. Table 16.10 provides a comparison of release data obtained from an ESP ash sample analysed in 1985 and again in 1992 (Versluijs et al., 1990; van der Sloot et al., 1992). The ash was stored dry, at room temperature in a sealed container. Total content and availability was measured at both times. The only significant change that occurred was a reduction in the availability of Pb. The reason for this effect is not known. The difference observed for Mo has been attributed to a modification of the availability test, which was introduced to optimise the release of oxy-anionic constituents. The relationship between column test data obtained on the same sample of ESP ash in 1985 and 1992 are presented in Figure 16.25. The difference in initial Cd and Pb leachability is related to the initial pH, which was lower in the first percolate in 1992 (pH 11.7) in comparison with the 1985 data (pH=11.9). This effect is consistent with a decrease in pH of approximately 0.1 unit on the respective pH-solubility curves for Cd and Pb.

Figure 16.25 Bottom Ash Landfill and Lysimeter Leachate Composition for Cd, Cr, Cu, Mo, Ni, Pb and Zn Compared with the Unified pH Curve for Bottom Ash

IU 4

o

Figure 16.25 Continued 10

-. E' =

4 .

Li

5

1 -r

0.1

0.01

-: -1

i 0,001

, ,

0

Ni c

~

~

:

2

c

s

4

c

~

:

c

~

8

6

..

c . . c : c ' c & r

~

6

:

~

10

~

~

c c

G

: *

12

c

~

14

0

4

2

6

pH

8

12

10

14

pH 1000

+ Un~fiedpH curve

100 -: 4 CORRE

US

6 AVR

F~eldtnal

=

10 -:

\

F

1 -;

V

Li

8 A Woodburn US

VKI DK

o,i; o,o1

0

-: Z"

,

0 0

0

1 2

,

~ 4

~

~

~

6

: 8

pH

~

~ 10

~

~ 12

:

~ 14

~

~

~

:

~

~

~

722 Table 16.10 Fly Ash Total Content and Availability Measured in 1985 and 1992 for the Same Sample (mg/kg) Element Ca

Available (pH=4) Available (pH=7 and 4)

Total 1985

1992

1985

1992

125353

124409

67350

82015

304

240

309

Cd Mo

30.5

34

1.0

3.7

Zn

14864

16315

9420

10654

Pb

5936

6477

816

337

Cu

647

713

213

251

The effects of aging on scrubber residues has not been studied. It is anticipated that the major changes with time will be (i) a decrease in pH due to uptake of carbon dioxide from air and (ii) the rapid release of soluble salts. The potential for release of very large quantities of very soluble salts and the attendant potential for impact on soil and water supplies warrants careful consideration during development of APC residue management practices.

16.9.3 Comparison of Laboratory Data to Field Measurements In Figure 16.26 a and b, the data obtained in field measurements (Cambotti and Roffman, 1993; Hjelmar, 1991, Hjelmar, 1992, 1993; van der Sloot and Hoede, 1991; Ratsma, 1991; Roffman, 1991; Ash pile project, 1992; RWS-DWW, 1992) are related to the concentrations measured in laboratory leaching tests. Data were gathered from the following studies: Ash type Bottom ash

Location Rotterdam

Country Netherlands

Bottom ash

Copenhagen

Denmark

Combined ash

Several monofills

USA

Combined ash Bottom ash Bottom ash

Woodburn, OR

USA USA Netherlands

HW 15

No field studies were available for release from APC residues only.

723 Figure 16.26 Reproducibility in Testing the Same Fly Ash after 7 Years and the Changes in Release Resulting from Aging 200000

20000 10000

Total

e~ooooo

1000

i

A

d'

"~ 10000

,D

100

[]

......... 10

1

60

3000

1 0.1

i

. . . . . . .

t

,

,

~

,

,,

1

Ioo

.

13 [] Q

o~ C d

13013 9

9

A

...................... 1

9

[] 9

.

.

.

.

60

10

.

.

.

100

.

-- - - " - - - ' - ' ~ "

0

[]

A DoE]

A

-

-

-

ZX

[]

10

tO

60

1 0.1

1000

100

100 10

10

Pb ,

. . . . . . .

i

,

L.,

, , , u l

10

1

0.1

Mo 10

6O

Cu A

0.1

..................... 1

~ -

9

800

0.1

,,]

1000

I0

0.01

?,

10000

I000 I

0.01 0.1

[] o .~.~.D.

10 1

Do n 1000 ~ 0.1

Zn

Availability

Ca

A

[] ol'l or-I [] ~ o

0.01 60

0.001 0.1

.................. 1

10

L i q u i d / s o l i d ratio (LS in l/kg)

60

724 The pH range in field measurements and lysimeter studies were significantly lower (pH=7 - 10) than the pH range found in laboratory tests on fresh ash (pH = 9.5 - 12). This also suggests that translation of laboratory data to actual field conditions requires adjustment based on pH-solubility curves, when it is desired to evaluate the long term behaviour of bottom ash. The pH-solubility curve for the sample to be evaluated can be obtained through use of pH static testing or similar tests. However, pH data generated in both field and lysimeter studies should be regarded with caution, since collection tanks or drainage systems usually are not isolated from contact with air. It is possible that the alkaline pH of collected pore water is neutralised during prolonged contact with air. The concentrations measured in field leachate and lysimeter studies are in equilibrium with the measured (lower) pH, which are not representative of the condition within the ash. The range of the concentrations measured in the field for most elements is within the range of results obtained in the laboratory. The Cu data from the field range about an order of magnitude below the lab data. This is related to Cu speciation as discussed in section 16.2.3.1 and the occurrence of reducing conditions in the field and in large scale lysimeters. The leachability of Pb deviates in some cases by an order of magnitude from the lab data. This has been attributed to the co-combustion of other waste streams with MSW. The Cr data in the field are up to an order of magnitude lower than the unified pH curve. Reducing conditions in the field may lead to reduction of chromate to trivalent chromium, which is less soluble than chromate in this pH region. With just a few exceptions (e.g. some Zn field data) the field data generally fall below the results obtained in the lab, which makes the laboratory evaluation a conservative approach. Leachate has been analysed from below Highway 15 in The Netherlands, where bottom ash was used as a road subbase (The Netherlands, RWS-DWW,1992). Results are provided in Table 16.11. The field measurements generally agree with laboratory data. Measured concentrations of CI were higher than laboratory observations, but was the impact of the lower L/S on very soluble species in the field. The observed release of TDS may be a concern during utilisation of bottom ash in many locations (see Chapter 22). Field experiments also have been carried out to assess the behaviour of bottom ash in a harbour filling operation in Rotterdam (van der Sloot and Hoede, 1991; Ratsma, 1991). The experiment consisted of setting up a container (2.5 x 2.5 x 15 m) with bottom ash under each of the following conditions: Closed on the top to prevent rainwater infiltration Closed at the bottom with runoff from rainwater drained by overflow, and Drainage through the bottom with percolate collected separately

725 Table 16.11 Composition of Leachate Collected Under Highway 15 (The Netherlands) where Bottom Ash was Used as Road Base Compared to Laborato.ry Column Test Data Element

Column test (L/S=0.2)

Field data (L/S = 0.04)

CI

6000

12000 - 16000

S O4

5000

3000 - 4000

F

1

0.5-1

Cu

< 20

20

Mo

2000

50 - 200

As

2

5

Pb

<20

<20

Cr

<10

10

Cd

5-20

6

Ba

<100

300

Ni

390- 500

Zn

120 - 75

50 200 - 1000

pH 9 4- 0.2 6.9 4- 0.1" *Reported pH may have been affected by carbon dioxide uptake from air during collection. The pore water composition within the ash layer near the top and bottom was measured over a 1.5 year interval. The dissolved species concentrations in the three containers varied only marginally between one another. However, considerable changes were noted over time. Table 16.12 provides the pore water composition for three representative periods. Changes in pore water pH were relatively small in the center of the containers. The pH values measured in collection tanks should be viewed with caution because of contact with air, which results in more neutral observed pH. Measured Cd concentrations were low due to the high pH. The high dissolved concentrations of Cu and Ni are noteworthy. Figure 16.27 presents the observed concentrations of As, Cu, Cr and Zn as a function of redox potential. Cu was the only element for which a significant difference in concentration was noted in response to changes in the redox potential. The field results for Cu were consistent with laboratory results which indicated a reduction in Cu release when the redox potential decreases to less than -100 mV (Comans, 1993). A significant difference between the top and bottom concentrations was observed for CI, sulphate, Ni and Mo indicating the washout of soluble components with progressive percolation.

Figure 16.27 Concentration of As, Cu, Cr and Zn as a Function of the Redox Potential in Pore Water Samples from Bottom Ash Field Studies

AVR, Rotterdam

727 Table 16.12 Pore Water Composition in Botto_m Ash as a Function of Time in the. Field Element

Unit

t=0

t=150d

t=300d**

t=400d

As

IJg/I

6

8

8

5

SO4

mg/I

1500

4800

6700

3500

Cd

pg/I

1

2

0.1

0.2

CI

mg/I

4500

5000 (3800)*

5500 (3500)*

5500 (3400)*

Cr

pg/I

5

<1

<1

<1

Cu

IJg/I

300

1000

10

60

Mo

IJg/I

900

1800

900

1400

Ni

IJg/I

180

240

4

150

Pb

pg/I

20

200

6

3

Sb

IJg/I

18

4

<2

<2

Zn

IJg/I

150

120

220

70

9.3

9-10

9-10

9-9.7

pH***

"k'A"~

Eh mV 150 120 -100 to -300 -100 to -200 Container with percolation (CI wash-out) Sharp decrease in redox potential In percolate and over flow water pH=7.2 (carbonate uptake from the air)

The concentrations of specific elements or species of interest in leachate from field sampling and lysimeter studies have been shown to equal to or less than the unified pH curve based on laboratory leaching tests. This indicates that the solubility controlling phases in the field are not significantly different from those in the laboratory. Cumulative release of specific elements under field conditions can be estimated based on knowledge of the unified pH curve, availability, anticipated pH in the field and the anticipated liquid to solid ratio in the field. These estimates can be refined further if the extent and effects of reducing conditions in the field are known.

728 REFERENCES

Allison, J.D., D.S. Brown, and K.J. Novo-Gradac. "MINTEQA2/PRODEFA2, A Geochemical Assessment Model for Environmental Systems: Version 3.0 User's Manual", EPAI60013-911021,U.S. Environmental Protection Agency, Athens, 1991. Ash Pile Monitoring Research Project. Volume II Site Hydrology and Unsaturated Zone Transport. PP&L, 1992. "AVI Bodemas in een Historische Opslag Situatie", IWACO report 10.3404.0, 1993. Barna, R., P. Moszkowicz, J. Mehu, H.A. van der Sloot. '~Vaste Solidification: Modelling and Simulation of Sodium Chloride Release from leached Concrete", Int. Symp.Waste. Manaqement 1992. Praag, pp. 65- 68, September 1992. Blust, R., A. Fontaine, and W. Decleir. "Effect of Hydrogen Ions and Inorganic Complexing on the Uptake of Copper by the Brine Shrimp artemia Franciscana", Mar. Ecol. Pro.q. Ser. 76:273-282, 1991. Born, J.P. W A V data from bottom ash quality control at 6 installations in the Netherlands over the period 1988 to present. Cambotti, R.K. and H.K. Roffman. "Municipal Waste combustion Ash and Leachate Characterization, Monofill- Fifth year study, Woodburn Monofill", Wood-burn, Oregon, Report prepared by AWD technologies, Pittsburgh, Pennsylvania, 1993 Chandler, A.J. Personal Communication regarding Data from Peel Incinerator, 1993 Comans, R.N.J. and J.J. Middelburg. "Sorption of Trace Metals on Calcite: Applicability of the Surface Precipitation Model," Geochim. Cosmochim. Acta 55: 2587-2591, 1987. Comans, R.N.J., H.A. van der Sloot, P. Bonouvrie. "Geochemical Reactions Controlling the Solubility of Major and Trace Elements During Leaching of Municipal Solid Waste Incinerator Residues", ...ProceedingsMunicipal Waste Combustion. VlP 32. Air & Waste Management Association Pittsburg, Pennsylvania, pp. 667- 679, 1993. de Groot, G.J., H.A. van der Sloot and J. Wijkstra. "Leaching Characteristics of Hazardous Elements from Coal Fly Ash as a Function of the Acidity of the Contact Solution and the Liquid/Solid Ratio", In: _Environmental Aspects of Stabilisation and Solidifi .cation of Hazardous and Radioactive Wastes, ASTM STP 1033, P.L. C6t~ and T.M. Gilliam, Eds., American Society for Testing and Materials, Philadelphia, pp. 170-183, 1989.

729 DiPietro, J.V., M.R. Collins, M. Guay, and T.T. Eighmy. "Evaluation of pH and Oxidation-reduction Potential on Leachability of Municipal Solid Waste Incinerator Residues," in Proceedin.as Intern_ational Conference on Municipal Waste C0mbu.sti0n., Hollywood, 1989. Eighmy, T.T., S.F. Bobowski, T.P. Ballestrero and M.R Collins. "Investigations into Leachate Characteristics from Amended and Unamended Combined Ash and Scrubber Residues", University of New Hampshire, 1989. Eighmy, T.T., D. Domingo, J.R. Krzanowski, D. St,~mpfli and D. Eusden. "The Speciation of Elements in Incineration Residues", Proceedin.gs Mun.icipal Waste Combustion. VIP 32. Air & Waste Management Association Pittsburg, Pennsylvania. pp. 457 - 478, 1993. Environment Canada. "The National Incinerator Testing and Evaluation Program (NITEP): A Summary of the Characterization and Treatment Studies on Residues from Municipal Solid Waste Incineration", EPS 3/UP/8, prepared by S.E. Sawell and T.W. Constable, October 1993. Environment Canada. "Compendium of Waste Leaching Tests", Environment Canada. Environmental Protection series. Report EPS 3/HA/7. May 1990. Felmy, A.R., D.C. Girvin, and E.A. Jenne. "MINTEQ--A Computer Program for Calculating Aqueous Geochemical Equilibria", EPA-600/3-84-032, U.S. Environmental Protection Agency, Athens, 1984. Fraay, A.L.A. "Fly Ash a Pozzolan in Concrete", Thesis TU Delft, The Netherlands, 1990. Fraser J.L. and K.R. Lum. "Availability of Elements of Environmental Importance in Incinerated Sludge Ashes", Environmental Scien_ce.and Technology, Volume 17 (1), pp. 52-54, 1983. Fruchter, J.S., D. Rai, and J.M. Zachara. "Identification of Solubility-Controlling Solid Phases in a Large Fly Ash Field Lysimeter," Environmental Science and Te.chnolo.ay. 24(8): 1173-1179 (1990). Gardner, K.H. "Characterisation of Leachates from Municipal Incinerator Ash Materials", M.Sc. thesis, Dept. of Civil and Environmental Engineering, Clarkson University, 1991. Gruebel, K.A., J.A. Davies and J.O. Leckie. "The Feasibility of Using Sequential Extraction Techniques for Arsenic and Selenium in Soils and Sediments", Soil Sci. Soc. Am. J., 52, pp. 390-397, 1988.

730 Hem, J.D. "Study and Interpretation of the Chemical Characteristics of Natural Water, 3rd ed.", U.S. Geological Survey Water-Supply Paper 2254, U.S. Geological Survey, Alexandria, 1985. Hjelmar, O. Leaching Data on a Velund Rotary Kiln Incinerator, Personal Communication, Denmark, May 1991. Hjelmar, O. "Field Studies of Leachates from Landfilled Combustion Residues", In Waste materials in Construction. Eds. J.J.J.M. Goumans, H.A. van der Sloot, Th.G. Aalbers, Elsevier Science Publishers, Amsterdam, Supplemetary issue. NOVEM 1991 Japanese Leach Test. "Methods of Assaying Metals, etc. Contained in Industrial Waste", Environment A.qency Notification No. 13, Revision No. 44, 1982. Kosson, D.S., H.A. van der Sloot, T. Holmes, C.C. Wiles. "Leaching Properties of Untreated and Treated Residues Tested in the USEPA Program for Evalustion of Treatment and Utilisation Technologies for Municipal Waste Combustor Residues", In: Waste Materials in Construction. Eds. J.J.J.M. Goumans, H.A. van der Sloot, Th.G. Aalbers, Elsevier Science Publishers, Amsterdam, pp. 119 - 134, 1991. Kirby, C.S. and D.Rimstedt. "Mineralogy and Surface Properties of Municipal Solid Waste Ash", .Environment.al Science & Technoloqy, 1993. Kirby, C.S. and D. Rimstedt. "Municipal Solid Waste Incinerator A s h A Geochemical Analysis", Proceedings Municipal Waste Combustion. VIP 32. Air & Waste Management Association Pittsburg, Pennsylvania. pp. 479- 511, 1993. Morel, F.M.M. Principles of Aquatic Chemistry, John Wiley and Sons, New York, 1983. NOH/RIVM. "Comparative Incineration Experiments with Different Domestic Waste Fractions. Parts I, II and II1." NOH/RIVM Report 10831-WPB 91-392, July 1991. NOH/RIVM. "Vergelijkende Verbrandingsproeven van Lange duur met Verschillende Fracties uit Huishoudelijk Afval", NOH/RIVM raport 10831-WPB 91-392, 1991. NOVEM. "Final Report of SOSUV Activities", Prepared for NOVEM by RIVM, KEMA, ECN, TNO, INTRON and TAUW, 1989. Oberste-Padtberg, R. and K. Schweden, "Zur Friesetzung von Wasserstoff aus Morteln mit MVA Reststoffen", WasserLLuft und Boden., 34,6, pp. 61-62, 1990. Pitzer, K.S. and L. Brewer. Thermodynamics, 2nd Edition, McGraw-Hill, New York, 1961.

731 Pitzer, K.S. and J.J. Kim. "Thermodynamics of Electrolytes. IV. Activity and Osmotic Coefficients for Mixed Electrolytes". J. Am. Chem. Soc. 96, pp. 5701-5707, 1974. Pitzer, K.S. and G. Mayorga. "Thermodynamics of Electrolytes. I1. Activity and Osmotic Coefficients for Strong Electrolytes with One or Both Ions Univalent". J. Phys. Chem. 77, pp. 2300-2308, 1973. Rai, D. and J.M. Zachara. "Chemical Attenuation Rates, Coefficients, and Constants in Leachate Migration. Volume 1" A Critical Review", EPRI EA-3356, Volume 1, Electric Power Research Institute, Palo Alto, 1984. Rai, D. "Inorganic and Organic Constituents in Fossil Fuel Combustion Residues. Volume 1 A Critical Review", EPRI EA-5176, Volume 1, Electric Power Research Institute, Palo Alto, 1987. Ratsma, K. Municipality of Rotterdam, Personal Communication, 1991. Reimann, D.O. "Beihefte zu Muhl und Abfall", Heft 29, pp. 12-16, 1990. Rigo, H.G., A.J. Chandler and S.E. Sawell. "Debunking Some Myth about Metals", Proceedin.qs Municipal Waste Combustion. VIP 32. Air & Waste Management Association Pittsburgh, Pennsylvania. pp. 609-643, 1993. Roffman, H.K. "Major Findings in the US-EPA/CORRE MWC-Ash Study", Municipal Waste Combustion Conference Tampa, Florida, VIP-19 Air & Waste Management Association, Philadelphia, pp. 96-123, 1991. Rotterdam. Contract Research for Municipality of Rotterdam (ECN-C-91-044). Prepared by H.A. van der Sloot and D. Hoede, in Dutch), 1991 "RWS-DWW Milieutechnisch Onderzoek Toepassing AVI Slakken in Rijksweg 15", Report D 0421-71-001, 1992. Salifu, L. Personal Communication, 1993. Sawell, S.E., T.R. Bridle and T.W. Constable. NITEP Phase II - Testin.q of the FLAKT Air Pollution Control Technolo.qy at the Quebec City Municipal Ener,qy-From-Waste Facility: Assessment of Ash Contaminant Leachability. Environment Canada, EPS Manuscript Series IP-70, 1986. Sawell, S.E., T.R. Bridle and T.W. Constable. "Assessment of Heavy Metal Leachability From Municipal Solid Waste Incinerator Ashes", Waste Mana,qement and Research, Vol. 6(3), 1988.

732 Sawell, S.E. and T.W. Constable. NITEP Phase liB: Assessment of Contaminant Leachability from the Residues of a Mass Burnin.a Incinerator. Environment Canada, EPS Manuscript Series IP-82, Vol. VI, 1988. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testin,q and Evaluation Pro.qram: Evaluation of Contaminant Leachability From Residues Collected at a Refuse Derived Fuel Municipal Waste Combustion Facility. Environment Canada Report, Manuscript Series, IP-96, 1989a. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testin.q and Evaluation Pro,qram Characterization of Residues from a Modular Municipal Waste Incinerator with Lime-based Air Pollution Control. Environment Canada Report, Manuscript Series, IP-101, 1989b. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testin.q and Evaluation Pro.(:iram: Characterization of Residues from a Mass Burnin.(:] Municipal Waste Incinerator with Lime-based Air Pollution Control (Burnaby, B.C.). Environment Canada Report, Manuscript Series, IP-110, 1990a. Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testing and Evaluation Pro aram Characterization of Residues from a Two-staQe Incinerator with Rotary Kiln (3M Canada). Environment Canada Report, Manuscript Series, IP-119, 1990b. ,,

Sawell, S.E., T.W. Constable and R.K. Klicius. The National Incinerator Testin.q and Evaluation Pro.c]ram Characterization of Residues from a Refuse Derived Fuel Combustion Facility (Mid-Connecticut). Environment Canada Report, Manuscript Series, in preparation, 1991a. Sawell, S.E., A.J. Chandler, H.G. Rigo, S.A. Hetherington, and J. Fraser. "The Waste Analysis, Sampling, Testing and Evaluation Program: Effect of Lead and Cadmium Spiking of Municipal Solid Waste on the Characteristics of MSWI Residues", Proceedin.qs Municipal Waste Combustion. VIP 32. Air & Waste Management Association Pittsburg, Pennsylvania. pp. 288-302, 1993. Steketee, J. "Improvement of Ash Quality by Washing", TAUW Report 51161.50, 1989. Stumm, W. and J.J. Morgan, Aquatic Chemistry, 2nd ed., John Wiley and Sons, New York, 1981. Sundstrom, D.W., H.E. Klei, B.A. Weir and A. J. Perna. "Leaching Behaviour of Residues from Mass Burn and RDF Incinerators", Municip.a.I Waste Combustion Conference Tampa, Florida, VIP-19 Air & Waste Management Association, Philadelphia, pp. 759-773, 1991.

733 Swedish Geotechnical Institute. SGI data base, Link5ping, Sweden, 1993. TNO. "Study of Incineration Parameters of Different Waste Processing Scenarios at ARN in Nijmegen", TNO-Report 91-236, 1992 van de Beek, A.I.M., A.A.J. Cornelissen en Th.G. Aalbers. "Fysisch en Chemisch Onderzoek aan Huishoudelijk afval", RIVM 738505005, 1988. van der Hoek, E.E., P.A. Bonouvrie, and R.N.J. Comans. "Sorption of As and Se on mineral components of fly ash: relevance for leaching processes," (submitted for publication) van der Hoek, E.E. and R.N.J. Comans, (unpublished results) van der Sloot, H.A., G.J. de Groot, J. Wijkstra, and P. Leenders. "Leaching Characteristics of Incinerator Residues and Potential for Modification of Leaching," in Proceedings International Conference on Municipal Waste Combustion, Hollywood, 1989. van der Sloot, H.A., O. Hjelmar and G.J. de Groot. "Waste/Soil Interaction Studies The Leaching of Molybdenum from Pulverised Coal Ash", In: .F..lueGas and Fly Ash, Eds. Sens, P.F. and Wilkinson, J.K., Commission of the European communities, Elsevier applied science, London, 1989. van der Sloot, H.A., G.J. de Groot, J. Wijkstra and P. Leenders. "Leaching Characteristics of Incinerator Residues and Potential for Modification of Leaching", Proceedings of th...e In.ternational Conference on Municipal Waste Combustion, April 11..14, Hollywood, Florida, 1989. van der Sloot, H.A. "Systematic Leaching Behaviour of Trace Elements from Construction Materials and Waste Materials", In: Waste Materials in Construction.. Eds. J.J.J.M. Goumans, H.A. van der Sloot, Th.G. Aalbers, Elsevier Science Publishers, Amsterdam, pp. 19 - 36, 1991. van der Sloot, H.A., D. Hoede and P. Bonouvrie. "Comparison of Different Regulatory Leach Test Procedures for Waste Materials and Construction Materials", ECN-C-91-082, 1991. van der Sloot, H.A. "Systematic Leaching Behaviour of Trace Elements from Construction Materials and Waste Materials," in W.aste Materials in Construction Proceedin,qs of .t..he International Conference on Enyironmental. Implica...tions of Cons.t..ruc.tion withWaste Materials, Elsevier, Amsterdam, 1991. van der Sloot, H.A., R.N.J. Comans, T.T. Eighmy and D.S. Kosson. "Interpretation of MSWI Residue Leaching Data in Relation to Utilisation and Disposal", Pr0ceedinqs of Int Recyclinq Conference, Berlin, November 1992.

734 van der Sloot, H.A., R.N. Comans, T.T. Eighmy, D.S. Kosson. "Interpretation of Municipal Solid Waste Incinerator Residue Leaching Data in Relation to Utilization and Disposal," R0ckstande aus.der M011verbrennun~q, Ed. Martin Faulstich, EF-Verlag f0r Energie und Umwelttechnik, GmbH, Berlin pp. 331-346, 1992. van der Sloot, H.A., R.N.J. Comans, T.T. Eighmy and D.S. Kosson, "Interpretation of MSWI Residue Leaching Data in Relation to Utilization & Disposal," in p_roceedin~qs International Recyclina Conference, Berlin, 1992. van der Sloot, H.A., O. Hjelmar, Th.G. Aalbers, M. Wahlstr6m, A.M. F,~llman. "Proposed Leaching Test for Granular Solid Wastes", ECN-C-93-012, 1993. van der Sloot, H.A. and D. Hoede. "Contract Research for Municipality of Rotterdam" ECN-C-91-044, in Dutch Versluijs, C.W., I.H. Anthonissen & E.A. Valentijn MAMMOET '85: Inte.arale evaluatie van de deelunderzoeken. Rapport: nr 738504008, Rijksinstituut voor Volksgezondheid en Milieuhygiene, Bilthoven, 1990. WAV. Quality Control Program for MSWl Bottom Ash for 1988. Personal Communication, 1993 Wahlstr6m, M. Personal Communication, 1992. Whitehead, I.E. "An Environmental Evaluation of Bottom Ash Substitution in Paving Materials", Thesis. University of New Hampshire. December 1992. Zevengeren, C. "AVI bodemas in een historische opslag situatie", IWACO Report 10.3404.0, 1993. Zevenbergen, C. "Natural Weathering of MSWI Bottom Ashes". Thesis. University of Utrecht, The Netherlands, 1994.

735

CHAPTER 17- SEPARATION PROCESSES

During the last decade, operating standards for MSW incinerators have rapidly become more stringent and the technology has generally followed suit. Many countries also enacted new legislation to improve the quality of incinerator residues for reuse or disposal in landfills. As a result, ingenious technical processes have been developed, generally to solve single problems, e.g., Hg or PCDD/PCDF emissions, or the leachability of APC residues. Each process typically created new problems, and often new, more concentrated residue streams. Consequently, compliance with the total set of imposed licence restrictions now requires a series of highly complex technologies to be added onto the incineration process. In some instances, these strategies may create unwarranted expenditures, and need to be considered in the context of Best Available Technology Not Exceeding Excessive Cost (BATNEEC). In general, the baseline concept of BATNEEC processes requires more focus on adequate process control and in-plant treatment techniques, which are relatively simple and robust. It is only when these options fail that secondary treatment is warranted. 17.1 DEFINITION OF PROCESS

In relation to residue treatment, the term "separation" includes all the techniques which are applied either to:

1)

separate mass streams of different origin and quality, or

2)

isolate single species from special residue streams in order to improve the quality of the respective residue and/or to recover the respective species

Separation processes are not only necessary to minimise potential environmental impact, but they are often necessary to render the residues suitable for use with respect to technical or engineering criteria. Based on the discussion of the physico-chemical processes which take place inside an incinerator (Chapter 8), the incineration process should be optimised toward both minimising emissions and residue quantity based on the following: a) the grate siftings, grate ash, boiler ash, filter ash, and scrubber residues streams should be collected separately to apply the best disposal or treatment option appropriate to each streams characteristics b) the grate or bottom ash must be inertised to the greatest extent possible to enable utilisation or simple disposal c) unavoidable hazardous products of combustion must be concentrated in the least voluminous side-streams

736 d) all treatment process steps must be tested to ensure ecological benefits e) the process design must weigh the benefits of the treatment against the costs involved The recently issued German guidelines on residential waste (TA Siedlungsabfall) are based on these fundamentals (Bundesministerium, 1993). It enforces the strict separation of all single residue streams. Only residues which are treated with the same process (e.g. boiler and fly ashes) are allowed to be combined. This is a simple and cost-effective means to enhance treatment of the residues. These techniques include on-site isolation of mass streams and post-incineration treatment by screening, and magnetic and eddy current separation. Physical separation methods have a limited effect on the ash quality since they are only able to isolate single components already present in the original mixture. They do not necessarily modify the chemical properties of the ash, but are capable of removing materials which are potentially detrimental to recycling the bulk of the material, i.e., as in the utilisation of bottom ash. Better separation efficiencies can be obtained by taking advantage of physico-chemical parameter changes for separation purposes, without feeding additional chemical agents into the system. These methods can be carefully directed to separate special compounds or classes of compounds, which may result in very pure and marketable products. For example, physical separation of ferrous metal and metallic aluminum can be easily achieved from bottom ash. Additional technologies have been proposed or are in use, ranging from washing salts out of bottom ash to the production of pure HCI from wet scrubbing solutions via distillation. Chemical reactions alter the chemical state of the species in question in order to obtain specific removal at high efficiencies. Applied or proposed chemical separation processes include leaching procedures using special media or precipitation processes to produce hydroxides, sulphides, or sulphates. But these techniques generally require the addition of special chemical reagents. According to the definition given above, electrochemical processes are distinctive in that they make use of both chemical and physico-chemical methods, but do not require the use of additional chemical reagents to produce a special material.

17.2 PHYSICAL SEPARATION TECHNIQUES Physical techniques consist of simple separation and/or classification of residues, and in most cases are limited to treatment of bottom ash. Due to the relatively large mass stream, one of the most effective means to modify bottom ash quality is to implement simple technical solutions. The following provides an overview of both primary and secondary measures which can be employed to treat the ash.

737

17.2.1 On-site Separation Bottom Ash The use of proper combustion control measures to enhance the completeness of burnout not only acts to minimise the organic contaminant content of the bottom ash, but also acts to enhance the partitioning of elements into the different mass streams of an incinerator. Bottom ash normally consists of grate ash combined with the grate siftings and in many facilities with the boiler ash. The grate siftings may contain a certain amount of incompletely combusted material and thus increase the total carbon content of the bottom ash. Furthermore, they may carry substantial quantities of metallic AI, Pb and Cu which are of concern if the ash is destined for most utilisation applications. However, the data on the quality of grate siftings is limited and varied. The results from an older study indicated there was no substantial difference in total concentration between grate ash and grate siftings (Schneider, 1986;). Even different separation and washing procedures used on ash (performed at the Swiss incinerator 'Zurich KVA 1') resulted in no significant deterioration of bottom ash quality by the inclusion of grate siftings (BUVAL 1990). Conversely, the NITEP and WASTE Programs both reported significantly higher concentrations of AI, Pb and Cu in grate siftings compared to grate ash (Environment Canada, 1991; WASTE Program, 1992). Furthermore, it is a common experience for considerable amounts of metallic AI and Pb to be found "frozen" in the chutes and collection hoppers underneath the grates. Based on operating experience and the latest data, there is sufficient justification to recommend separation of both residue streams and to feed the grate siftings back into the furnace to burn the remaining uncombusted material in the grate siftings. However, it is important to note that the metallic components, such as ferrous, AI, Cu and Pb, should be removed from the stream prior to reintroducing the uncombusted fraction back into the incinerator. This practice is being performed in a limited number of German facilities and is enforced by the new German regulation in cases where the grate siftings contain more than 3% of unburnt carbon (Bundesministerium, 1993). In the boiler ashes, especially in the ashes generated in the economiser, mobile heavy metals are found as well as semi-volatile organics (see Chapter 8). Hence in many countries, separation of this mass stream from the bottom ash is enforced by regulations (e.g., Bundesministerium, 1993). The Swiss research program cited above also recommends the separation of boiler ashes from the grate ash (Buval, 1990). The recommended practice of separate collection of the ashes from the heat recovery system is also supported by data generated during a Swiss investigation which involved conducting the Swiss leaching test on bottom ash after step-wise separation of the boiler ash and grate siftings. The results indicated that the grate ash was much less soluble than the combined streams. Moreover, the leaching behaviour of the pure grate ash is comparable to that of natural building materials like basalt, quartz, and gravel under the same test conditions (Vehlow et al., 1992).

738 Fly Ash and APC Residues Fly ash collected in the combustion chamber ("hot-side" fly ash) is characterised by very low concentrations of thermally mobile metals, which are still in the gas phase in this part of the incinerator (Environment Canada 1985 and 1988). Consequently, dedusting of the hot flue gases in the front part of the boiler at temperatures well above 500~ could potentially remove a high percentage of the inert material from the ash stream and thus reduce the quantity of ash collected further downstream. However, it has yet to be determined whether the quality, especially the leaching stability, of materials separated at high temperatures would permit mixing this material with bottom ash. Studies on the benefits of such a system would need to be conducted, especially in relation to the quality of bottom ash destined for application purposes. On the other hand, the separation of this fraction may be advantageous with respect to the formation of PCDD/PCDF in the boiler, which is highly dependent on the presence of particulate carbon in the fly ashes (see Chapter 8). The coarse hot-side fly ash fraction may contain the major proportion of carbon and its separation should result in lower levels of organohalogen compounds in the flue gas. It is speculated that the resulting increase in concentration of volatile trace metals, such as Cd, Zn, As, or Pb, in the fly ash and APC residues would be acceptable, since these residue streams already require special treatment prior to disposal (in most countries). Investigations into the behaviour of volatile trace metals under the conditions of hot gas filtration have been conducted by modelling a boiler-integrated cyclone in an electrically heated bypass duct (Borchers, 1989). A substantial reduction in concentrations was noted for some highly volatile metals in high temperature cyclone dust compared to those in downstream fly ashes. For example, maintaining higher temperatures (750~ in the hot cyclone dust reduced Cd by more than 70%, the Pb by more than 60%, and the Sn by approximately 55%. PCDD/PCDF concentrations in the cyclone dust were in the order of 12 - 15 ng/g (total), which is low compared to the PCDD/PCDF loading in fly ash. There is no full-scale application of such a technology. Although the integration of a cyclone into a boiler seems feasible, other problems may be encountered during steady state operation in that a number of components in fly ash, especially sulphates and phosphates, are sticky at high temperatures and tend to build up in layers on the surfaces of the unit (see Chapter 10). In addition to increasing the frequency of the conventional soot blowing cycles to prevent the gas channels from being closed off in relatively short period of time, this technology would probably require more frequent maintenance to remove the deposits formed via Fickian diffusion, diffusiophoresis and thermophoresis mechanisms. Because fly ash collected downstream of the heat recovery system contains characteristically high levels of volatile trace metals and semi-volatile organic compounds, it is recommended that these be collected separately from the bottom ash, as is the case in most countries. In several instances, there are additional regulations requiring the separate collection and handling of fly ash from the APC residues. For example, in Germany the access to different categories of landfills is regulated by a new federal guideline on the basis of results of the German DEV $4 leaching test

739 (Bundesministerium, 1993). This guideline classifies raw fly ash and residues from dry/semi-dry APC systems differently. As a result, the APC residues have to be disposed in underground disposal sites, whereas fly ash, at least after treatment, can potentially be placed in less costly aboveground landfills. In this case, separation of coarse fly ash prior to a lime-injection system or spray-dryer may be economically effective, and consequently cyclones or ESP's have been installed prior to the APC system in some incinerator facilities.

17.2.2 Metal Separation from Bottom Ashes The quantity of ferrous and nonferrous metals in bottom ash is highly dependent on the effectiveness of the source separation programs in a given region, or on the presence of automated magnetic and eddy-current source separators used in the production of RDF. Bottom ashes from incinerator facilities can contain 7 - 10% ferrous scrap (Schoppmeier, 1988)and approximately 1 - 2% nonferrous metals. The economics of recovering these scrap materials is in turn dependent on the fluctuating market for scrap steel, and ultimately the demand for finished steel. Normally, primary scrap must be "de-tinned" and "decontaminated"of organic material in order to concentrate the iron inventory prior to use. This additional refining is reported to cost the equivalent of about $10 US per tonne of waste (Reimann, 1992). Ferrous metals processed through an incinerator are effectively "de-tinned" and relatively clean of organic contaminants, and thus are generally highly sought after as recyclable ferrous scrap. In the case of bottom ash utilisation, the removal of ferrous and nonferrous metals must be completed prior to utilisation. Metals like AI, Fe, and Zn are susceptible to corrosion attack in an alkaline media resulting in the generation of hydrogen gas and corrosion products which may cause swelling, (especially metallic AI). Although it is common practise in most incinerator facilities to recover the ferrous scrap by means of magnetic separation, bottom ash utilisation requires its classification into specific grain sizes by screening, which is often combined with a magnetic scrap removal step (GSttlicher, 1990). This treatment is done either at the incinerator facility or at special treatment facilities. A schematic of a typical treatment technique is given in Figure 17.1. Generally, between 2 - 5 kg of nonferrous metals are present in a tonne of waste (see Chapter 2). In earlier times, some of these valuable materials were picked out of the ash stream by hand, however, a more efficient method of separating the material is the application of eddy current technology. The technology has not yet been widely applied, however, one example of a successful pilot-scale system was used to process the residues from a facility which employed a combination of pyrolysis, and high-temperature combustion of the pyrolysis gas and the carbon-rich fraction of the pyrolysis coke (at the 'SchweI-Brenn-Verfahren'). The eddy current system was used in combination with rotating permanent magnets to separate the ferrous metals, inert material, and nonferrous metals (like AI and Cu) (Berwein, 1990). A schematic of the applied separation technique is given in Figure 17.2.

740

Figure 17.1 Treatment of Bottom Ash by Screening and Magnetic Separation

riddle ..9

,." o. 9 .

crusner

magnets

sieve v

,6,

sieve

~

~ ~oar, escrap \ Adapted from G6ttlicher, 1990 Figure 17.2 Ferrous and Non-Ferrous Metal Separation by Magnets and Eddy Current rotating permanent magnets

,9, ot.~o.o- o.o, o,o

eddy current preseparation

l oj:.. O0

4~

811

mmml

v

m

...

I:I

9

D

0,~

-b,

m

Y,

p

~~.~~,~-"

-~

-"

0

v

lr

A

0

~

y

-Ao'A.-/ t

q

~,

non-ferrous

V

00

fe rro u s

9,

ferrous

Berwein, 1990

V~iO 0

0 0

>Y

~176 :~-

p ~z.~

inert

q

p

nonferrous

741

17.3 PHYSlCO-CHEMICAL AND CHEMICAL SEPARATION TECHNIQUES Physico-chemical and chemical separation methods described in this chapter are characterised by more or less complex technologies which entail high operating costs. The following discussion presents these special processes following an ascending order of complexity of technology, which is typically directly related to the cost of the system. No distinction between pure physico-chemical, pure chemical, and combined methods will be made.

17.3.1 Washing Processes Principles The quality of incinerator residues, including their environmental compatibility (especially for bottom ash), is currently evaluated based mainly on the leaching behaviour of trace metals. A second group of contaminants, the soluble salts, especially chlorides, are now being recognised as potential problems by regulatory authorities as well. The simplest means of separating the soluble compounds from the solids is by washing. Incinerator residues are generally highly alkaline. Contact with water typically generates a liquid with a pH ranging from 9.5 - 12. In this pH regime, most heavy and trace metals form hydroxides or other relatively insoluble compounds. Hence the efficiency of washing to facilitate metal extraction is very low, but a substantial removal of the soluble salts can be expected. If the pH exceeds values of 12, however, amphoteric metal compounds (such as some Pb and Zn compounds) are released and can remain in solution (van der Sloot et al., 1992; Environment Canada, 1993). Since the water consumption at most MSW incinerator facilities is severely limited, low L/S ratios are generally used to quench the ash. Consequently, there is a limited capability to remove the salts. This low US ratio, coupled with the short residence time of ash in the quench tank, typically prevents the process from reaching equilibrium. In recent years however, many investigations have tested the salt and metal removal efficiency of washing. Both options, the on-site washing (bottom ash only), as well as the secondary treatment in separate facilities have been tested from laboratory to full scale. Although there are benefits to these washing processes, they tend to generate a wastewater problem. If metals are dissolved, conventional wastewater technologies can be employed to reduce the concentrations of contaminants. However, the salts cannot be retained in wastewater treatment facilities. If their discharge is regulated, evaporation or complex membrane technologies must be applied. These technologies are expensive and, moreover, the question remains as to how and where the salt residues can be disposed.

Bottom Ashes

On-site Processes The following represents some examples of different types of systems in use at various incinerator facilities.

742 Most incinerator facilities quench bottom ash in a quench tank filled with water. Previously, the only make-up water added was to replenish the losses due to evaporation and discharge of water along with the bottom ash. The quench tank, however, can also be used as a washing device if the water throughput is enhanced. As briefly mentioned above (see Section 2.2.1), the Zurich incinerator "KVA 1" tests on bottom ash washing were conducted using an L/S ratio of 10:1 (Buval, 1990). Only 0.1 0.2% of the heavy metals were solubilised and 82% of the chlorides were removed. However, reduction of the L/S ratio to simulate on-site washing reduced the yield of soluble chlorides. It was estimated up to 500 L per tonne of waste is required to guarantee acceptance under the Swiss TVA limit for utilisation in road construction, which translates into 1 gram of CI per kilogram of ash (Bundesamt for Umweltschutz, 1988). The report also recommended the washing of fresh bottom ash was more effective than aged ash, since the chloride removal efficiency was significantly higher than from aged ash. Furthermore, the leaching of the ash with a CO2 enriched water also needs to be considered. Under the CO2 enriched conditions, some metals like Zn are easily mobilised, whereas other metals such as Pb and Cd form relatively insoluble carbonates and are not significantly affected by such treatment. It is speculated the overall intent of this application was to simulate the Swiss TVA leaching test, and thus allow facilities to meet the respective limits more easily. Whether this translates into an improvement of leaching stability in the natural environment has yet to be determined. -

About 350 L of water per tonne of waste feed are used to wash bottom ash (L/S ratio of 1:1) at one German incinerator facility (Reimann, 1992). Published data indicate about 50% of the chloride can be removed from the bottom ash, which corroborates the Swiss test results mentioned above, whereas the removal of fluoride and heavy metals were not significant. Unfortunately, the data for alkali metals were not available. Another process proposes to integrate treatment for all of the residues streams (the MR-Process), and involves washing of the bottom ash in a quench tank with an SO2 scrubber solution (Stubenvoll, 1989). This process is discussed in more detail later (see Chapter 3.3.3.2). The washing is used as a separation step to remove the soluble salts from the ashes. The original intent of the process was to desaturate the SO2 scrubber circulation fluid of sulphates by gypsum precipitation in the quench tank.

Secondary Processes

In some countries, the conventional conditioning required prior to bottom ash utilisation includes aging the ash, which involves storing the ash in piles exposed to the natural elements for at least a period of two weeks. This aging changes the chemical properties (see Chapter 9), but does not guarantee the ash will pass the regulatory leaching tests, even the German DEV $4 test. Therefore, consideration of washing the ash prior to use has been considered to enhance the aging process. Washing after, instead of prior to, aging was recommended in order to maintain the buffering capacity of the ashes, since alkalinity is regarded as an important parameter driving all of the

743 aging processes. In full scale experiments, aged bottom ash was washed at an L/S ratio of about 1:1 (Lahl, 1992). Although the aged ash was already suitable for utilisation, the washing improved the DEV $4 results, especially with respect to the chlorides and alkali ions. With respect to bottom ash weathering, a German study examined two different treatment options for bottom ash (Schneider 1994). Bottom ash from two facilities was stored outside in a pile (height 25 cm) for three months permitting an estimated 0.5 L/kg of precipitation to infiltrate the pile. Figure 17.3 illustrates the CI release from this material was reduced by the same order of magnitude than it was by washing directly after discharge. Subsequent washing of the aged ash resulted in only a minor reduction in the CI leaching, indicating little benefit was achieved by additional washing. In larger ash piles, the US ratio should be decreased substantially, and thus removal of CI as well. Consequently, washing rather than aging, was recommended as the preferable method to stabilise bottom ash, since it was more effective at CI removal and the CI rich washing solutions could be treated within the incinerator facility (e.g. as feed solution for the wet scrubber). Air Pollution Control Residues As mentioned previously, because APC residues and fly ash contain high concentrations of potentially mobile trace metals, soluble salts, and semi-volatile organic compounds, these residues are classified as "special" wastes in some countries. Since disposal strategies for hazardous wastes are generally expensive, and sometimes even limited within a specific region, efforts have been made to render the residues more suitable for conventional disposal strategies.

Swiss Fly Ash Treatment From a geological perspective, Switzerland has only a limited capacity for hazardous waste disposal, and research into developing alternate means of disposing fly ash were initiated in the mid-1980's. Since most Swiss incinerator facilities employ wet scrubbing technology with wastewater treatment, the focus of much of the research has been on washing the residues prior to stabilisation with cement (Tobler, 1988). These stabilised residues must then meet the stringent limits for salt and trace metal release based on the Swiss TVA leaching test before being classified as suitable for landfill (Bundesamt for Umweltschutz, 1988). Demonstration experiments were performed in a pilot-scale test facility where an L/S ratio of 2:1 was applied to remove the soluble salts (Dietler, 1990). The pH of most of the washed residues was about 10, resulting in a very minor release of trace metals into the wash water. However, some final pH values were mildly acidic (about pH 6). In these cases, the release of trace metals was sufficiently high to create problems in the wastewater treatment facility. In order to meet effluent discharge limits, especially for Cd and Hg, the organic additive TMT #15 was added to the solution to precipitate

Figure 17.3 Chloride Concentrations in the DEV S4 Test Solutions after Direct Washing and Washing After Aging of 3 Months

aged 1 month

aged 2 months

aged 3 months

washed directly

agedl washed

745 trace metals. (A more detailed discussion of this agent is found in Chapter 19.) The leaching stability of the washed fly ash alone has not been tested, but the solidified products pass the TVA test limits to permit disposal in a residue landfill. The process was deemed successful enough to be installed in a number of Swiss incinerator facilities. Some changes with respect to cement quality and quantity are discussed elsewhere (Ponto, 1993). Unfortunately, the main disadvantage to the process is the salts, including chlorides and sulphates, which pass through the wastewater treatment facility and are subsequently discharged into the river systems. In addition, the long-term stability of the organic contaminants remaining in the products, and trace metal compounds formed with TMT #15 remains unknown. Should the organic components begin to degrade, it is possible the trace metals will become available for leaching.

17.3.2 Acid Leaching The efficiency of washing processes to remove trace metals from MSW incinerator ash is limited, especially from fly ash and APC residues. Although trace metal concentrations in these residues are much higher than in bottom ash, the alkalinity of APC residues, especially from dry/semi-dry scrubber systems, is sufficiently high to promote the solubilisation of only amphoteric metals. Hence, extraction of other trace metals requires a significant addition of acid to reduce the pH to acidic levels. Some advanced processes, which are described in later chapters, combine separation with further treatment steps as an integrated approach to fly ash and APC residue treatment. However, the basis of extracting trace metals results in the release of salts, which creates a potential effluent discharge problem anywhere except marine environments.

Bottom Ash An acid extraction process for trace metals from bottom ash was tested at laboratory scale in the Netherlands. The primary objective of the process was to improve the leaching stability of the ash (Buijtenhek, 1989). The tests compared simple washing of bottom ash with acid extraction. Screening tests were also conducted to determine the removal capability of chelate-forming agents as well. Data from the multi-step extraction at a pH of 4 indicated promising results for some metals at high L/S ratios (i.e., >20:1), however, the overall extraction efficiency of the process was rather poor, since many metals in the bottom ash are solubility controlled. Therefore, the only tangible benefit achieved by the process is to treat the ash in the event it exceeds regulatory leaching criteria. Furthermore, the comparably high water consumption of the process may limit its applicability in many countries.

746 Filter and Boiler Ashes 3R Process The 3R Process is a two-stage treatment process which combines an acid extraction of soluble trace metal compounds from boiler and filter ash with thermal treatment of the compacted extraction residues in the combustion chamber of an incinerator (Vogg, 1984). Application of the process is generally limited to facilities which employ wet scrubber systems. The flow diagram given in Figure 17.4 illustrates the use the acid flue gas cleaning solution as the extraction medium.

The process requires the separation of Hg from the scrubbing solution prior to the extraction of other trace metals. This separation is done by ion exchange and is discussed in detail later in the chapter. The US ratio is in the order of 7 - 10:1, and the final pH adjustment in the extraction vessel was 3 - 4. The residence time in the continuous flow system is about 30 minutes, achieving extraction efficiencies of approximately 90% for Cd, 70% for Zn, and 20 -40% for other metals. Overall, about 20 % of the total solid matter is dissolved. The filtrate from the process must be treated prior to discharge, whereas the extracted residue is returned to the combustion chamber of the incinerator. If the economics are satisfactory, electrochemical or ion exchange processes can be used to separate out specific metals for reuse. When this is not cost effective, hydroxide precipitation is used to produce about 2 kg of concentrated inorganic residue per tonne of waste burned. The extracted residue is returned to the combustion chamber not only to destroy the PCDD/PCDF and other semi-volatile organic compounds present in the residue, but to further stabilise the heavy metals remaining in the residue through sintering (Vehlow, 1993). The process has been successfully tested in pilot-scale for the extraction step, and has also been installed at full-scale at a Swiss incinerator facility. The recycling step has been successfully tested at full-scale (Vehlow, 1990). The preliminary results from the full-scale facility on the extraction step in Switzerland corroborate the findings of the pilot scale tests (Quittec, 1993). This process has been found to increase the cost of waste incineration by approximately $4 - 6 US/tonne of waste burned. MR-Process Another multistage process already mentioned in the discussion of on-site bottom ash treatment is the MR-Process (Stubenvoll, 1989). A flow diagram of the process integrated into an incinerator is given in Figure 17.5. The aim of this treatment process, like that of the 3R process, is to improve the residue quality by concentrating trace metals for potential recovery and reuse, and the destruction of organic contaminants. The system operates by extracting the boiler and fly ash streams using the acid scrubbing solution from the first-stage wet scrubber. The washed and filtered residues are then heated in a rotary kiln for one hour at >600~ to destroy the PCDD/PCDF compounds and to volatilise the Hg. The off-gases are passed through an activated charcoal filter at 100~ where at least 95% of Hg is removed. The filtrate from the

Figure 17.4 Flow Diagram of the 3R Process

4

I

I I I I I

scrubb.

I

I I I I

scrubb.

bottom ash + 3R Product

- - - - m a - - -

solids liquids

recycling, evaporation or treatmnet

Figure 17.5 Flow Diagram of the MR-Process v scrubb.

scrubb.

*

I

I I I

I I

I

,

I

I I I I I I I I I , - , - - _ _ _ _ - - - - - - - - - - -

,

CaO t

I

1 I I

:

-

,- - - - - - - --:-I ' alkaline ; ; extraction l

gypsum

-

.

I

.

*

I

. - 4 . .

-r

I,

'

--

I

; :

.

neutralisation

4

f solids

------ . ......... . . . . .

Stubenvolt, 1989

ash

gases alkaline liquids acid liquids waste water

gas heated rota'y kiln

I

........

filtration

- ..... - .

extraction,

filtration

. . . . . . . . . . .....,............ treatment and recycling

749 extraction step is treated in a wastewater treatment facility. The effluents from the second-stage scrubber are used for on-site bottom ash washing, as well as for neutralising acidic effluent streams. The Hg laden charcoal must be treated or disposed of as a hazardous material since simply recycling the material back into the combustion chamber would only serve to release the Hg into the flue gas again. FLUWA Process The FLUWA process (Frey 1991 ), which is very similar to the extraction stage of the 3R process, has been tested in pilot and full-scale at two Swiss incinerator facilities. The acid flue gas cleaning solution is used (after Hg removal) to extract trace metals from the boiler and fly ash streams. At a typical Swiss gas cleaning L/S ratio of 4.5:1, an acid water stream with a final pH of 3 - 4 is generated. This solution is used to wash the residues which are then separated via vacuum belt filtration. Generally, the solid residues are capable of meeting the Swiss TVA leaching limits for residue landfills, while the extracted metals precipitated from the filtrate have the potential for recycling. ALS Process The ALS Process (Acid Leaching and Sulfide Process)is a metal extraction technology which is in commercial use in Japan. It consists of three stages:

slurrying of fly ash and wastewater at an L/S ratio of 5:1 adjusting the pH of the slurry to 6-8 with HCI to extract heavy metals. After the extraction, NariS solution is added to change soluble heavy metals into stable compounds adding a coagulant to the extracted slurry and dehydrating the slurry using a hydroextractor A process flow diagram of this system is provided in Figure 17.6.

17.3.3 Ion Exchange Principles and Results

Ion exchange is a process widely used in chemistry, both for analytical purposes as well as for industrial chemical production. The process involves isolating or enriching single ions or groups of ions from multi-component solutions. Many specific exchange materials have been developed on the basis of zeolites, as well as organic resins. In most cases, ion exchange can be highly specific by using the proper exchanger and optimising the operation conditions. Ion exchange is often used in industrial wastewater treatment to remove metals. However, the application of conventional acid ion exchange resins to MSW incinerator wastewater streams are not effective due to the high Ca concentrations. Hence, resins which carry chelate forming organic groups like iminodiacetic acid (trade names are Amberlite IRC-718, Lewatit TP207 and TP208, Ionac SR-5)or bispicolylamine (trade name Dowex XFS4195) are the most preferred.

Figure 17.6 Flow Diagram of the ALS Process

fly ash storage tank

4

1

I rnlxtr

reactor No. 1

Ul reador No 2

S I W ~ ~agulation

cake

tank

filtration

bavy remwal

Sakai, 1994

ng

adivated

removal

adsorption carbon

751 These resins have high removal efficiencies for specific metals like Co, Ni, Cu, Zn, Cd, and Pb (Jekel, 1992). In most cases, the ion exchange process is controlled by exchange equilibria which enables this method to be easily used to recover exchanged ions.

17.3.4 Hg Recovery from Flue Gas Scrubbing Solutions Recently, a method has been described which uses a special ion exchange resin which consists of thiourea (trade name Lewatit TP214) to remove Hg from the acid scrubbing solution of a two-stage wet scrubbing system. This process is an integrated part of the 3R Process and has also been incorporated into the APC system of a sewage sludge incinerator (Braun, 1993). The removal efficiency of Hg using the resin is excellent and is independent of the Hg concentration in the feed. Feed concentrations in the scrubber water can range from 1 - 30 mg/L of Hg, and at a pH of <1 produce effluents with residual concentrations of approximately 20 IJg/L. The maximum load is about 200 g of Hg on one litre of resin. The Hg can easily be recovered from the resin, and thus the process can be used to solve some of the Hg problems in incinerators equipped with wet APC systems.

17.3.5 CrystallisationlEvaporation Principles and Results The main constituents of wet flue gas scrubbing solutions are chlorides in the acid scrubber and sulphates in the neutral scrubbing stage. The related cation depends on the neutralisation strategy chosen at the facility, either Na or Ca based compounds. In some facilities, Na is used in the acid scrubber, whereas both Na and Ca are used in the neutral stage. In the latter case, gypsum can be produced as a substitute for natural gypsum. The disposal options available for chlorides are generally limited since they are characteristically highly soluble in water. Disposal in open landfills is not recommended, and discharge into sewers or fresh water is typically prohibited. While relatively safe long-term disposal can be achieved by deposition in old salt mines, this is a rather expensive measure, and discharge to a marine environment is not always possible. Hence different techniques have been proposed to isolate and purify NaCI or CaCI2 out of the salt mixture for reuse. In order to reuse the salts, trace metal contamination of the product must meet stringent standards. Although the isolation and purification techniques are not complex, they are generally multistage processes.

NaCI Production NaCl production in order to recycle parts of the chlorine inventory of the waste back into chlorine-alkali electrolysis has been proposed (Exner, 1989; Dettmann, 1990; Thom6, 1992), however, these processes have also been criticised for not really closing the chlorine balance (Karger, 1990).

752 The process shown in Figure 17.7 is designed to treat the effluents from a conventional wet scrubbing system neutralised with lime. The first step involves converting the Ca salts into Na salts. The generated CaCO3 is substituted back into the scrubbing system as a neutralising agent. Mg salts, as well as gypsum and trace metals, are removed prior to evaporation to produce a NaCI product. Since even the direct disposal of mixed salts requires neutralisation and solid separation, the economic benefits of this process are comparable to disposal of mixed salts. At one facility, the process is used to generate NaCI for on-site water conditioning (Dettmann, 1990). Other proposals have been made to recycle the NaCI, either as crystalline salt or as a concentrated brine, into the chlorine-alkali electrolysis to replace rock-salt. Both products can meet requirements limiting trace metal contamination. In Japan, a few facilities recover NaCI from wastewater discharged from wet scrubber systems. After removing Na2SO4 and heavy metals by using CaCI 2 and Na2S respectively, a NaCI slurry is crystallised by concentration through evaporation. The NaCI crystals are recovered from the solution by centrifugal separation generating a final product which is 99% pure and is sold to the chemical manufacturing industry as a raw material for NaOH production (Wada et al., 1981 ).

CaCI2 Production from DrylSemidry APC System Residues

Two different processes which combine washing and concentration of CaCI 2 have been proposed and tested in Canadian and Danish laboratory studies (Birch et al., 1993, Environment Canada, 1993). Although the results from the Canadian study were not available, the results from the Danish study indicate it is technically feasible to recover marketable CaCI2 from dry/semidry gas cleaning residues by countercurrent washing (using water). A schematic for the process is shown in Figure 17.8. However, if the scrubber residues contain fly ash, some of the trace metals may be dissolved, which need to be removed by the addition of hydroxide and TMT #15. The filtrate is then subjected to evaporation and precipitation to partly remove Na and K. The CaCI2o2H20 contains 3-7% by weight of alkali ions and can be commercialised. The filter-cake is then stabilised using cement, however, the effect of the hydroxide and TMT #15 on the sludge is not known.

Gypsum Production

The production of gypsum from the effluents of SO2 scrubbers is a widely used technology. The first experience with the process was acquired with the desulphurisation of coal-fired power station residues back in the early 1980's, and the systems are the same as those used for residues from coal-powered stations. The purification of gypsum in order to meet standards set for its utilisation is a proven technology.

753 Figure 17.7 Flow Diagram of NaCI Recycling

Salt conversion

Na2CO3

Metal precipitation

FGD - effluent

11

toFoo

Product to be dumped Evaporation Live steam

]

E!flue

Live steam condensate ~

NaCI ,'5

vapour steam to FGD

Exner, 1989

II

o

Figure 17.8 CaC12 Production from Dry and Semi-Dry Flue Gas Cleaning Residues

water b

1. washing

b

.

extract hydroxide precipitation

4

evaporation

heavy metal separation

heavy metal polishing

heavy metals

heavy 'metals

4 solution

Adapted from Birch, 1993

2. washing

residue

I

755

17.3.6 HCI Recovery by Distillation Principles and Results HCI present in the flue gases is trapped in the acid gas removal stage of a wet scrubbing system, forming dilute HCI. This acid is very volatile and distillation processes are an obvious option for HCI recovery. Under normal pressure, HCI forms an azeotropic mixture with water at a concentration of 20% HCI. To achieve higher concentrations additional treatment is required. Some of the typical difficulties encountered during distillation of HCI in flue gas scrubbing solutions are due to the HF, HBr, SO3, dust (trace metals), and organic compounds which are also present in the flue gas. Consequently, the primary HCI in the HCI absorber contains many unwanted contaminants. Different strategies have been developed to distil HCI from flue gas scrubbing solutions. Depending on the desired concentration of the product and on the operation conditions of the first scrubber, there are minor variations in technology.

Proposed Processes A relatively simple process applied in a two-stage wet scrubbing process produces marketable HCI in a secondary facility. A flow diagram of the system is given in Figure 17.9 (Juritsch, 1989). The process includes combined gypsum and Na2SO4 production from the effluents of the neutral scrubber. A potential problem with fluoride contamination of the HCI is addressed in a similar process combining distillation of HCI from the acid scrubbing solution and the production of gypsum from the neutral scrubber (K(~rzinger, 1989). In modern wet scrubber systems, pre-neutralisation of the acid scrubbing solutions is common practice to conserve water. This neutralisation is preferentially done using lime, which is relatively inexpensive. If HCI is to be distilled from pre-neutralised scrubber water, an intermediate step of salt conversion is required. A process dealing with this option is given in Figure 17.10 (Karger, 1990).

17.3.7 Electrochemical Processes Principles and Results Electrochemistry is a conventional technology widely used in industry to refine noble metals, produce chlorine at the industrial scale (chlorine-alkali-electrolysis), and to generate hydrogen from water. The main advantage to electrochemical processes versus chemical-based processes is that they require no additional chemicals, and thus create no new residue problems. According to the different electrochemical potentials of the specific chemical compounds, processes can be tailored to isolate and concentrate specific components out of even very dilute solutions. However, these advantages are accompanied by some disadvantages which include high operating and capital costs and the need for highly trained, experienced staff.

Figure 17.9 Flow Diagram of HCI Recovery

HCI Absorber Furnace

ESP

I

Heat Exchanger

Neutral Scrubber

Stack

Wet ESP

(DENOX)

...... . . - . -

b0

Filter Ash Treatment

Juritsch. 1989

El Production

NaOH, CaO

Production

Treatment

Figure 17.10 Flow Diagram of Hcl Recovery from Ca Containing Solutions salt conversion

Karger,

sedimentation

distillation

758 An example of an electrochemical process is based on the experience gleaned from the galvanising industry which could be used to develop an electrochemical Hg recovery system from flue gas scrubbing solutions. Theoretically this very specific process could generate a high yield, however, no full scale applications have been established.

Chlorine Recovery

On the basis of laboratory experiments, a proposal has been made to separate chlorine by direct electrolysis of the acid scrubbing solution (Volkman, 1991 ). In a bench-scale demonstration, current efficiencies of up to 60% could be reached. It is anticipated that a full-scale installation would consume about 3% of the energy produced by an incinerator to recover about 70% of the CI inventory of the waste. However, about 6 kg of soluble salts per tonne of waste remain in the solutions. Since these salts are not desirable, they must be disposed. Hence the process will not close the CI circuit completely, but it contributes to reducing the loading to local disposal sites. This proposed process has been designed as an integral part of the 3R Process for metal extraction out of filter ashes, and with gypsum production from the sulphate present in the second scrubber. A flow diagram of this combined process is given in Figure 17.11. Figure 17.11 Flow Diagram of a Combination of the 3R Process and Electrochemical Chlorine Recovery flue gas acid ~1 ~ scrubber

.,,

Hg

.,= H2, CI 2

Iprecipitation i g y p s u m ,.-j"

I chlorine recovery

filter '=1 3R ash ~" Process

II i,j

Volkman, 1991

! gypsum i-~ lime

separationI

i,a

neu,ra,1-"

scrubber

~evaporationl salts ,,~ it-

~!

, . lime hydroxide precipitati~ r~,=,~l~ " metals ~"

759 Common to all of the recovery technologies, is the need to evaluate the economic benefit/disadvantages based on the supply and demand for the recovered material within regional markets. For example, the gypsum market in most countries is generally not conducive to recovery from MSW incinerator residues due to the natural abundance of the material, and the availability from secondary sources, such as from coal-fired generating stations. In addition, the stigma associated with products derived from MSW incineration, may limit the potential use of the recovered material. In all cases, the recovery of a material from a residue stream should be compared against the expense of a direct disposal in an adequate disposal site. Given the costs of proper disposal and the problems associated with siting disposal facilities, recovery of materials from incinerators may yet prove to be economically viable.

REFERENCES Berwein, H.-J. & J. Erlecke. EinsatzmSglichkeiten der Verschwelung als Homogenisierungsstufe in der thermischen Abfallverwertung. In M011verbrennun.Q und Umwelt 4 (Thom6-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 225, 1990. Birch, H., O. Hjelmar & G.P. Lorenzen. _Genanvendelse af restprodukter fra a.ffaldsforbraendin,q med k.alksbaseret ro.qgasrensnin.q, Water Quality Institute, Report EFP 1323/91-006 og Genanvendelsesrodet, Horsholm, DK, 1993. Borchers, H.-W. & K.J. Thom6-Kozmiensky. Flugstaubabscheidung vor dem Kessel einer M011verbrennungsanlage - Erste Ergebnisse, AbfallwirtschaftsJournal, 1, 1, 28, 1989. Braun, H. Unpublished results from the Kernforschungszentrum Karlsruhe, 1993. Buijtenhek, H.S., J.H. de Zeeuw & J.J. Steketee. Improving the Quality of MSW Slags Using Extraction Processes. In Recvclin~ International (Thom~-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 1468, 1989 Bundesamt fer Umweltschutz. Entwurf zur Vernehmlas.sun.q einer Technischen Verordnunq 0ber Abfalle. (TVA), Bern, August 1988, revision 1991 Bundesamt fQr Umwelt, Wald und Landschaft (BUVAL)..Grundla.aen zur Festle.qun.q der QualitQtskriterien f0r KVA-Schlacke, im Hinblick auf ihre Verwendun~ im Strassenbau, Bern, Juli 1990 _

Bundesministerium fQr Umwelt, Naturschutz und Reaktorsicherheit. Dritte AIIgemeine Verwaltungsvorschrift zum Abfallgesetz (TA Siedlungsabfall), Technische Anleitung zur Verwe....rtung, Behandlun,q und sonsti.qen Entsor.qunQ von..SiedlunQsabfQIlen vom 14. Mai 1993, Bundesanzeiger Jahrgang 45, Nr. 99a, 1993.

760 Dettmann, P. Salze aus externer Waschwassereindampfung, Beihefte zu M011 und Abfall, 29, 37, 1990. Dietler, U. Auslaugverfahren mit nachgeschalteter Verfestigung gemaess TVA-Schweiz, Beihefte zu M011und Abfall, 29, 88, 1990. Environment Canada. Th.e National Inciner.ator.Testin.q and .Evaluation Program: Two-staqe Combustion (Prince Edward Island), Environment Canada Report EPS 3/UP/l, 1985. Environment Canada. The National Inciner..ator Testin.q and Evaluation Program: Environmental Characterization of Mass Burning Incinerator Technolo,qy at Quebec City, Environment Canada Report EPS 3/UP/5, 1988. Environment Canada. Development of a Treatment System for Residues from .Municipal Waste Incinerators, Terms of Reference for a study, internal document, Solid Waste Management Division, 1993. Exner, R., W. Hinsen & P. Sporer. Thermal Effluent Treatment for Flue Gas Treatment Systems in Refuse Incineration Plants from the Point of View of Residue Minimization by Recycling. In Recyclin.q In.ternational (Thom6-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 1372, 1989. Frey, R. Ein Verfahren zur Behandlung von Flugst~uben, AbfallwirtschaftsJournal, 3, 194, 1991. G6ttlicher, R & P. Anton. Reststoffe aus der M011verbrennung, AbfallwirschaftsJournal, 2, 376, 1990. Jekel, M., J. Ritz & C. Vater. Abw,~sser und deren Behandlung. In M011verbrennun.q und Umwelt 5 (Thom~-Kozmiensky, K.J., ed.), Berlin EF-Verlag, 375, 1992. Juritsch, V. & G. Rinn. Chlorwasserstoff- Absorption und die Gewinnung von Salzs,~ure aus den Rauchgasen von Abfallverbrennungsanlagen, In M011verbrennun.q und Umwelt 3(Thom6-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 1230, 1989. Karger, R. Verfahren zur Rauchgasreinigung AbfallwirtschaftsJournal, 2, 365, 1990.

bei

der

Abfallverbrennung,

K0rzinger, K. & R. Stephan. Hydrochloric Acid and Gypsum (Sulphuric Acid) as Utilizable End Products Obtained from the KRC Process for Cleaning Flue Gases from Incinerators. In _Recyclinq International (Thom~-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 1224, 1989. Lahl, U. Verwertung von MV-Schlacken - durch Optimierung konventioneller Aufbereitung, M011und Abfall, 24, 619, 1992.

761 Netherlands Standardization Institute (NNI) Dutch Pre-Standard NVN 2508, Determination of Leachin.Q Characteristics of Coal .Combustion Wastes, February 1988. Ponto, H.-U. Reststoffbehandlungsverfahren unter Einhaltung Schweizerischer Auflagen am Beispiel der KVA Horgen, VDI Bildungswerk, Handbuch Reststoffentsorqun~ BW 43-60-03, BW 1879, 1993. Quittek, C. & H. Suter. Babcock Rueckstandsbehand!un.qstechniken einschl. Ergebnisse des 3R Verfahrens der KVA Buchs, VDI-Bildungswerk ,Handbuch Reststoffentsor.Qun~q BW 43-60-03, BW 1879, 1993. Reimann, D.O. VerwertungsmSglichkeiten von M011verbrennungsschlacke, M011und Abfall, 24, 609, 1992. Sakai, S. Personal Communication with Assistant Professor of the Environmental Preservation Center at Kyoto University, Japan, September 1994. Schneider, J. Determination of Elemental Waste Composition by Analysis of Incineration Residues. In Recvclin.Q International, Vol. 1.1 (Thome-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 318, 1986. Schneider, J., J. Vehlow & H. Vogg. Improving the MSWI Bottom Ash Quality by Simple In-Plant Measures, proceedin.qs of the WASCON '94 Conference, Amsterdam: Elsevier Publishing, 605, 1994. Schoppmeier, W. Erfahrungen mit der Entsorgung, Aufbereitung und Verwertung fester Verbrennungsr0ckst~nde aus HausmOellverbrennungsanlagen, M011und Abfall, 20, 104, 1988. Stubenvoll, J. The MR-Process Treatment of MSW Residues with Reclamation of Heavy Metals. In Recycling International (Thom6-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 1476, 1989. Thom~, E. Rauchgasreinigung - Wertstoffr0ckgewinnung und Schadstoffminimierung. In M011verbrennun,q und Umwelt 5 (Thom6-Kozmiensky, K.J., ed.), Berlin: EF-Verlag, 197, 1992. Tobler, H . P . Behandlun,q und VerfestiQun.q von R0ckst~nden aus. M011verbrennun.qs.a.nlagen in der Schweiz, pg 188, Entsorgungspraxis Special 1988. van der Sloot, H.A., R.N.J. Comans, T.T. Eighmy & D.S. Kosson. Interpretation of Municipal Solid Waste Incinerator Residue Leaching Data in Relation to Utilisation and Disposal, Was..te Management International, (Thom~-Kozmiensky, K.J., ed.), Berlin EF-Verlag, Volume 2, 99, 1992.

762 Vehlow, J., H. Braun, K. Horch, A. Merz, J. Schneider, L. Stieglitz & H. Vogg. Semi-Technical Demonstration of the 3R Process, Waste .Mana,qement & Research, 8, 461, 1990. Vehlow, J., G. Pfrang-Stotz & J. Schneider. Reststoffe- Charakterisierung, Behandlung, Verwertung. In Svmp.osium 25 Jahre LIT, 5 Jahre TAMARA, KfK report 5000, 124, 1992. Vehlow, J. & H. Geisert. Characterization of the Leaching Behavior of 3R Products, 3r__dd Int. Conference on Municipal Waste Combustion, March 30 - April 2, 1993, Williamsburg, VA, 1993. Vogg, H. Verhalten von (Schwer-)Metallen bei der Verbrennung kommunaler Abf~lle, Chemie-lnaenieur-Technik, 60, 740, 1984. Volkman, Y., J. Vehlow & H. Vogg. Improvement of Flue Gas Cleaning Concepts in MSWI and Utilization of By-Products. In Waste Materials in Construction (Goumans, J.J.J., van der Sloot, H.A. & Albers, Th.G., ed.), Amsterdam: Elsevier Publishers, 145, 1991. Wada, A and A. Yoshida. "Making Salt from Waste" (in Japanese), PPM, Volume 12, No. 10, pp. 37-43, 1981.

763

C H A P T E R 18 - SOLIDIFICATION AND STABILISATION

One of the options generally suggested as a means to reduce the solubility of trace metals contained in incinerator residue streams, is solidification. Another related process which modifies the chemical properties of the residues is stabilisation. Although these terms are frequently used interchangeably, for the most part they are distinctly different, since they modify the behaviour of material in different ways. This chapter provides an overview of the various techniques available for use on MSW incinerator residues, and defines a protocol to assess the environmental and engineering suitability of solidified/stabilised residues. 18.1 DEFINITION OF PROCESS

The use of solidification and/or stabilisation techniques is widely considered to be a viable option for the treatment of many types of hazardous wastes. Development of specific formulations for different types of waste began in the late sixties and early seventies (Conner, 1990). The techniques were applied to stabilise either mine backfills, base courses for roads or most notably for the treatment of nuclear waste. The term Solidification/Stabilisation (S/S) is the generic term used to describe a wide range of techniques which can transform wastes into less environmentally problematic forms. This usually involves physical and/or chemical immobilisation of the waste constituents. US EPA has presented the following definition (Conner, 1990): Solidification refers to the techniques that encapsulate the waste in a monolithic solid of high structural integrity. The encapsulation may be of fine waste particles (micro-encapsulation) or of a large block or container of wastes (macro-encapsulation). Solidification does not necessarily involve a chemical interaction between the wastes and the solidifying reagents, but may mechanically bind the waste into the monolith. Contaminant migration is restricted by vastly decreasing the surface area exposed to leaching and/or by isolating the wastes within an impervious encapsulate. Stabilisation refers to those techniques that reduce the hazard potential of a waste by converting the contaminants into their least soluble, mobile or toxic form. The physical nature and handling characteristics of the waste are not necessarily changed by stabilisation. Frequently, both stabilisation and solidification processes are combined, thereby changing both the physical and chemical structure, and ensuring even if the monolith deteriorates that the contaminants will remain in the matrix. Numerous alternatives for accomplishing these changes have been proposed including:

764 mixing different waste types together adding hydraulic binding agents to the waste sequentially washing the waste then adding hydraulic binding agents adding organic binding agents Although many techniques have been proposed, and some are used commercially for treating hazardous industrial waste streams, few are in commercial scale operation for MSW incinerator residues.

18.2 EFFECTS OF SOLIDIFICATIONISTABILISATION 18.2.1 Physical Changes Solidification processes are typically designed to reduce the surface area to volume ratio of a waste, while at the same time decreasing the porosity and increasing the tortuosity of the material to minimise the release of contaminants to the environment. Consequently, the mere act of agglomerating finely divided materials into larger solid particles reduces the exposed surface area and thus potentially reduces the rate of release of contaminants from the mass. In addition, most solidification formulas are designed to generate physically strong and durable materials which maintain the integrity of the solidified matrix over geological time periods. The effectiveness of such physical changes can be influenced by many factors. For instance, cement- based systems are affected by curing temperatures and moisture content. Typically, it is generally acknowledged that higher temperatures hasten the setting reactions of the solidified material, whereas moisture contents above or below optimal moisture/density content will result in weaker matrix strengths and poorer durability. A Swedish study provides graphical proof of the effect of moisture and temperature on the leachability of lead from an APC residue mixed with water before compaction (Kullberg et al., 1989). The data indicate that although the release of lead normally decreases with time, the release rate of Pb is higher when less than optimal water content and lower curing temperatures were used (Figure 18.1 ). Furthermore, many chemical reactions can result in shrinking, swelling or changes in the time required for setting of the solidified materials. For example, the formation of ettringite, alkali-aggregate type reactions, and the oxidation of metals result in swelling of the matrix, which in turn, can reduce strength and durability (e.g., Neville, undated). In addition, the hydration of elemental metals can result in the liberation of gases (H2, CO, C2H2, etc.), especially in the presence of organic compounds (Soundarajan, 1989), which can form small voids within the solid matrix and potentially reduces its overall strength. Weathering, either due to freeze/thaw, wet/dry cycles or even erosion, can also result in a significant reduction in the physical integrity of the matrix, thereby increasing the exposed surface area and increasing the potential for contaminant release from the

Figure 18.1 Influence of Storage Temperature (t) and Water Content (w) on the Flux of Pb from APC Residue

28

49

Time (days)

Adapted from Kullberg et al., 1989

766 matrix. Another important consideration is the adequacy of the mixing process. The results from two studies have indicated that sufficient mixing time is required to permit some chemical reactions to occur which enhance the overall strength of the solidified material and hasten the setting time of the mixtures (Freidli and Brunner, 1989; Environment Canada, 1993).

18.2.2 Chemical Changes Solidification processes generally result in chemical changes that incorporate free water into the solid through chemical reactions, and can potentially bind metals into the matrix. These reactions can either reduce, or in some instances increase, the leachability of metals. The chemical binding of metals can include transforming soluble metal salts into insoluble silicates, hydroxides or carbonates, whereas other changes can result in the incorporation of a metal into a crystal adsorption mechanism. Pozzolanic reactions are examples of crystalline adsorption mechanisms offering potential long-term immobilisation of metals. Unlike metals, organic compounds do not readily lend themselves to such techniques since the larger organic molecules are not easily incorporated into a crystal structure. Furthermore, it is more difficult to form insoluble precipitates from organic compounds. However, it should be noted that most organic compounds present in MSW incinerator residues, such as PCDD/PCDF, are relatively insoluble in water, and consequently, applying solidification processes for the sole purpose of immobilising these compounds is unwarranted. Thermal treatment, sorption processes or physical containment are more appropriate processes to treat organic compounds. It is possible that the combination of certain binders/additives with APC residues can have a negative effect on the chemical and physical stability of a solidified material. For example, a Swedish study (Kullberg et al., 1991) indicated that some additives, such as gypsum, had negative effects. The presence of oils and greases, zinc, phenols and high concentrations of sodium have also been demonstrated to have a detrimental effect on the physical strength of solidified materials (Cullinane et al., 1987). The chemical stability of solidified/stabilised materials is dependent on the chemistry of both the matrix and the leaching solution. Dissolution and desorption are the most dominant processes which influence the mobilisation of contaminants, whereas immobilising and demobilising processes compete in typical non-equilibrium situations, such as under exposure to the natural environment. Since the rate of dissolution of trace metals is pH dependent, the release can be influenced by a reduction in the pH, either due to the dissolution and loss of calcium oxide from the matrix, or due to the creation of carbonates from the exposure to carbon dioxide. The physical constraints to the diffusion of ions out (or into) a matrix are dependent on the density of connected pores within the matrix (or the "tortuosity"). Generally, the weaker the product, the more "open" the pore structure (reduced tortuosity), and hence

767 the greater the potential release rate. Some inert species as such as salts can be used to determine the density of the network, since their mobility is generally only restricted by physical constraints. On the other hand, the chemical conditions in the matrix (pH, redox, complexation) dictate to what extent many trace metals can be released. The difference in effective mobility between the salts and the trace metals indicates the degree of chemical retention. These phenomena are discussed in more detail in Chapter 20. 18.3 EVALUATION OF SOLIDIFICATIONISTABILISATION PROCESSES

Since the aims of solidification/stabilisation processes include: enhanced physical properties increased size of agglomerated particles to reduce interaction with water solids with no free liquid improved handling characteristics reduced contaminant mobility Evaluating these effects clearly requires a wide range of protocols which address both the physical and chemical transformations. Physical properties such as size, density and strength are common tests used on solidified materials, whereas leaching tests are most common for stabilised materials. A number of protocols incorporating various standard engineering and chemical evaluation tests have been used to assess these solidified residues (e.g., Donnelly and Jons, 1988; Friedli and Brunner, 1989; Sawell et al., 1989; v.d. Sloot, 1990; Environment Canada, 1991). An example of a relatively comprehensive protocol is outlined in Figure 18.2. It is recommended that in addition to chemically and physically characterising the solidified material, leaching procedures should also be used to determine both the potential for contaminant release and the rate of release (see Chapter 20). Recognising the limitations with extrapolating laboratory data, it is also suggested that field verification studies should be undertaken (e.g., see Figure 18.3). Another protocol suggested specifically for cement-based solidified wastes is a 3 step testing procedure which is outlined in Table 18.1 (Environment Canada, 1991). The first level involves a characterisation of the untreated material with special emphasis placed on identifying species that may adversely affect the cement-based additives. An inert constituent, such as CI, is generally used to evaluate the tortuosity. The next level of testing assesses the effectiveness of the chemical fixation and the final stage involves evaluating the physical stability of the material.

Figure 18.2 Flow Schematic for Environmental Assessment of S/S Materials in Relation to Disposal EVALUATIONOF TEST RESULTS IN RELATION TO DISPOSAL

4

CALCULATE LEACH PARAMETERS - Torluosfly -Retention

MODIFY RECIPE

CALCULATERELEASEAT ARELEVANT TIME USING LEACH PARAMETERS C2 DISPOSAL

difference In ternprafure

INERT MATERIAL

Yes

I

I

van der Sloot et al., 1993

Deterrnlne wrn~resstvestrenqth after prolonged water exposure

I

1

I

I

iClPE OR APPLY MODIFY RI OTHER TREATMENT OPTIONS A

Figure 18.3 Framework for Assessing the Leachability of SIS Wastes

I SAMPLING. SAMPLE PREPARATION AND STORAGE

1

TOTAL WASTE ANALYSIS

1

SOLIDIFIED AND STABlLlSED WASTE MATERIALS

I L

TOTAL DIGESTION AND ANALYSIS OF INORGANICS

I

S A M P L I N G

1

/

A N A L Y S I S

1

/

EXTRACTIONAND ANALYSIS OF ORGANICS

I

I

AVAILABILIN TEST, ACID NEUTRALISATION CAPACITY AND REDUCING POTENTIAL

I

I

L E A C H I N G

I I

I

POTENTIAL LEACHABILITY

1 I

I

1

EXTRACTION OR VOLATlLlSATlON AND ANALYSIS (Hg,VOC

I

I

1

1

AVAlLABlLlN BY EXTRACTIONWITH ORGANIC SOLVENT(S) TO BE ESTABLISHED

I I

1-

1-1

1-1

TOTAL WASTE ANALYSIS

1-1

1

VOLATlLlSATlON

]

MODELLING RELEASE MECnAh SMS AhD ChEMICA. RETEhTlOh

SYSTEMATICSOF LEACHING I N DATABASE CERTIFICATION OF WELL CHARACTERISED MATERIALS

WASTE1SOlL INTERACTION

1

FIELD VERIFICATION

1

DIFFUSION TUBE MEASUREMENTSAT WASTE [WASTE OR WASTE1SOlL INTERFACES (STATIONARY AND DYNAMIC CONDITIONS) AND CONCENTRATION PROFILE ANALYSIS OF EXPOSED (WASTE) PRODUCTS

IMPROVEMENT OF PHYSICAL RETARDATION

SAMPLING AND ANALYSIS OF REPRESENTATIVESITUATIONS IN THE FIELD CONCENTRATIONPROFILE ANALYSIS, MODELLING AND EVALUATION OF DIFFUSION BARRIERS

MODIFICATIONOF CHEMICAL RETENTION

ENVIRONMENTAL IMPACT ASSESSMENT AND JUDGEMENT OF ACCEPTABILITY

Adapted from van der Sloot, 1990

1

Table 18.1 Test Methods to Investigate Cement Stabilised Materials Level

4

o

Testing Method

Recommended Method

0

Performance Indicator Process description

No testing required

NIA

0

Masslvolume change

Bulk density

WTC Protocol, Method 2

0

Porositylsaturation

Moisture content Bulk density Solids specific gravity

WTC Protocol, Method 4 WTC Protocol, Method 2 WTC Protocol, Method 3b

0

Organic content

Total organic content

glvrC, 1985)

3

0

Contaminant concentration

Digestionlextraction and analysis

(U.S. EPA, 1980)

3

1

Initial leachate concentration

Low U S ratio extraction

WTC Protocol, Method 9

4

1

Amount of contaminant available for leaching

Low pH extraction

Modified TCLP

1

Acid neutralisation capacity

Titration with acid

WTC Protocol, Method 11

3

2

Contaminant mobility in the matrix

Dynamic Leaching Test

ANSllANS 16.1

1

2

Hydraulic Conductivity

Constant Head Permeability

WTC Protocol, Method 5

4

2

Physical Strength (before1 after immersion)

Compressive Strength

WTC Protocol, Method 6

2

Weathering

Freezerrhaw or WeffDry Cycles

ASTM D 4842-89 ASTM D 4843-89

2

Biodegradability

Biological Growth

ASTM G 21-70 and G 22-76

Environment Canada, 1991

Minimum Number of Replicates NA 3

3

771

18.3.1 Physical Tests Since the solidification process changes the physical characteristics of the residues, it is appropriate to test the untreated residues, as well as the solidified materials. These results provide a comparative analysis of the characteristics before and after treatment. However, it should be noted that any test program should be related to the objectives of treatment process and the ultimate disposal scenario. Furthermore, not all of the tests are applicable to the both the untreated and treated residues. Consequently, it is recommended that several of the physical tests listed in Table 18.2 be performed on the untreated and treated materials (where applicable) to quantify physical properties. Table 18.2 Recommended Physical Tests for Evaluation of SIS Materials Category Test Material Reference Indices Grain Size Distribution Untreated ASTM D422/C136, DIN66115 Moisture Content Both ASTM D2216 Swelling Treated ASTM D3877-80 Carbon Content- LOI Untreated ASTM C25-93a, APHA 209E Density Bulk Both ASTM C29, E1109-86 Dry Both ASTM C29 Solids Specific Gravity Both A S T MD854-92, C127 & 128 Compaction Modified P r o c t o r Untreated ASTM D1557 Strength UnconfinedCompression Treated ASTM D1633 Triaxial Treated Hydraulic Constant H e a d Untreated ASTM D2434-68 Conductivity Flexible Wall Treated ASTM D5084-90 Permeameter Durability Freeze/Thaw Treated ASTM D4842-90 Wet/Dry Treated ASTM D4843-88 Immersion Treated The total carbon content of the untreated residue provides an indication of the relative amounts of solidifying agents required, or even the feasibility of the treatment process. Although no specific criteria have been established for this parameter, generally, the greater the carbon content, the greater the quantity of reagents which are required. The grain size distribution usually indicates the specific surface area which can assist with estimating the quantity of stabilising agent needed. For example, a material containing a narrow band of particle sizes is generally less stable and needs more solidification additives to fill in the voids to promote greater strength. Moisture content determinations on both the untreated and treated materials is necessary to estimate the

772 amount of water required during processing, and to calculate the weight increase factors of the treated material. To predict the optimum water content, a Proctor compaction test should be made on the untreated material. The stabilising reactions generally require access to free water, which is why the water content should be slightly higher than the Proctor compaction test indicates as optimum. The data on swelling provides an indication of the expansion which can be expected after treatment and deposition of the material. The strength of a treated product indicates to what extent chemical bonding has occurred. However, the bonding may be considered "temporary", since it is due more to capillary forces and salt precipitation than the actual formation of pozzolanic bonds. The unconfined compressive strength test can also be used to estimate the durability of a treated product, however, this should be conducted on samples which have been cured, then immersed in water prior to testing. The compressive strength test involves subjecting a SIS sample of a specified size to increasing pressure until the structural integrity of the solid material fails. Minimum compressive strengths are required for material deposited in a landfill. It is generally acknowledged that 350 kPa (50 psi) is necessary to withstand the operation of heavy machinery on top of the material in the landfill, however, other strength requirements have been recommended. For example, Austrian standards are based on compressive strengths exceeding 1,000 kPa (Verordnung proposed 1992), whereas Environment Canada (1991) has recommended a strength exceeding 350 kPa (after immersion in water) when placed in a monofill and 3,500 kPa in a sanitary landfill. Carlsson and Tuutti (1992) proposed that a SIS material with a strength exceeding 1,000 kPa should also be relatively durable. Determining the permeability of a material is as important in evaluating SIS residue as compressive strength. Permeability is generally defined as the rate at which a substance is able to pass through a solid body. In SIS materials, it relates to a factor which limits the interaction of a solvent with the solidified matrix, and thus limits the transport of contaminants out of the solid. The permeability of SIS materials is generally measured as hydraulic conductivity and compared to the hydraulic conductivities of soils. Typically, dense, compacted clays provide a coefficient of conductivity less than 10.9m/s, and it has been recommended that SIS materials have hydraulic conductivities less than 108 m/s (Environment Canada, 1991 ), 5 x 10.9 m/s (Austria, Proposed Verordnung, 1992) and 10.9m/s (Kullberg et al., 1989). Durability tests should be conducted on SIS materials to assess the response of a solidified matrix to the expansion and contraction of the solid matrix due to extremes in climatic conditions. Two tests have been used widely, namely the freeze/thaw and wet/dry weathering tests. These tests expose cubes or cylinders of SIS samples to alternating cycles of freeze/thaw, or wet/dry. Control specimens are generally exposed to alternating cycles of immersion in water and humidification. Consequently, weathering due to climatic extremes is expressed as a corrected weight loss value. This value is calculated by subtracting the cumulative weight loss through hydration

773 only (control) from the cumulative weight loss resulting from the exposure to freeze/thaw or wet/dry conditions (see Figure 18.4). The criterion generally used to assess sample durability is based on a corrected weight loss of less than 10% of their total weight through physical weathering. It is suggested that weathering due to dissolution in water or salt water (erosion) should also be considered as a criterion to provide an indication of whether or not the SIS matrix should be precluded from use in a marine reef or permitted to be deposited below the groundwater table. Many of these parameters provide valuable data on the physical properties of the material which indicate appropriateness of the treatment process for the eventual disposition of the material. Furthermore, it is recommended that the initial evaluation of a treatment option be conducted over a sufficiently long period of time to develop sufficient data to ensure the objectives of the treatment process are being met. Once this has been accomplished, selected surrogate parameters can be used for quality control. These selected parameters do not necessarily have to be a measure of the most important characteristics, but they should be sensitive enough to indicate changes in the treated product.

18.3.2 Chemical and Leaching Tests Since stabilisation and sometimes solidification processes are capable of changing the chemical nature of a waste, both the treated and untreated residues should be evaluated using chemical and leaching tests to assess the degree of those changes. A list of recommended tests is given in Table 18.3. Table 18.3 Recommended Chemical and Leaching Tests for SIS Materials Test

Material

Reference

Total metal concentration

Both

Acid Neutralisation Capacity

Both

Total Availability

Both

Tank Leaching Test

Treated

ANS 16.1, TVA or NVN 5432

Water Solubility Test

Both

Environment Canada, 1991

Environment Canada, 1991

The chemical composition of both the untreated and treated materials should be determined to compare the dilution effects of the SIS agent and to provide a benchmark for further evaluation. In addition to measuring the total concentration of metals, the data from a total availability test can be used in conjunction with data from leaching tests to estimate the potential release of contaminants from the SIS matrix.

Figure 18.4 An Example of Data Analysis from Weathering Tests

(%) ss07 s s e ~ e^!lelnwno

1

2

3

4

5

6 Cycle

7

8

9

10

775 Determination of the acid neutralisation capacity can also be useful for estimating the resistance to change in pH after deposition and estimating potential leachability under different disposal scenarios. Solidified materials are generally formed into monoliths with very low hydraulic conductivities and as such should be evaluated as a solid form as well. The release of contaminants from monoliths is generally governed by diffusion and not by percolation, therefore the ideal testing process should simulate diffusion in the environment. One method is the tank leaching procedure which permits an assessment of contaminant release from a solid monolith, whereas leaching tests conducted on ground samples (such as a total availability test or equilibrium leach test) provide an indication of potential contaminant release from a solid matrix which has disintegrated. As detailed in Chapters 14 and 20, tank leaching tests are conducted on monolithic cubic or cylindrical shapes. The monolith is submerged in either distilled water or an acidified water for various periods and the resulting change in contaminants in the solution is used to calculate the release. These data are presented either as a flux (mg/m2.day) or a diffusion coefficient (m2/s). Typical tank leaching tests are defined by either the ANS 16.1 (United States), TVA (Switzerland) or NVN 5432 (Netherlands) tests. Since the release rates will generally decrease with time, the cumulative amount released over time can be compared to the total amount available for leaching, then placed into context by generating an effective diffusion coefficient which takes into account chemical, as well as physical retention factors (see Chapter 20). Care must be taken in the interpretation of tank leaching data. The best comparison is the relative leachability between different samples. This comparison allows various formulations to be evaluated on a consistent basis. It should also be noted that if diffusion coefficients are used to estimate environmental impacts, the disposal conditions should be considered in relation to the manner in which the SIS material will be in contact with water. For example, tank leaching tests can be used to closely simulate reef disposal options where monoliths are used to create artificial underwater environments. Since land disposal is generally accomplished above normal groundwater levels, the monoliths are not continually immersed, and hence there is a much lower contact time between the solid and water. 18.3.3 Other Factors In addition to the factors which influence the effectiveness of SIS material discussed above, other factors such as the mixing and moulding equipment, the system used to transport the material to the disposal site, and the disposal method, can all influence the final properties of the material. Vibration during transportation can cause the constituent materials to separate, thus creating different chemical and physical properties within the separated layers. Continuous mixing prior to deposition is a simple means to avoid this problem.

776 In general, treated residues have a very low hydraulic conductivity which impedes percolation and promotes surface run off from disposed materials. Salts, especially chlorides, are very difficult to chemically bind within the most solidified matrices. The rate of release of chlorides can be improved by diminishing the total exposed surface area, although it should be noted that leaching of the exposed surface also results in a loss of solids which may increase the hydraulic conductivity. The loss of solids can also occur due to biodegradation of the organic carbon in the solidified matrix from the residues. Another consideration is the potential long-term influence of some of the additives used during the treatment of residues. For example, TMT#15 TM is used to precipitate trace metals from scrubber solutions, however, the long term stability of the heavy metal compounds formed with TMT#15TM is still not known. It has been speculated that if the organic structure degrades due to microbial activity, the associated metals may become available for leaching. In addition, the sulphides may not be stable over the long-term and oxidation may result in formation of metal sulphates which can be available for leaching. Consequently, it is recommended that the effectiveness of any solidification/stabilisation based treatment should be considered not only for the shortterm physical benefits, but the potential long-term effects of additives as well. 18.4 REVIEW OF AVAILABLE PROCESSES 18.4.1 Stabilisation Without Additives

Simply by mixing an APC residue with fly ash and water, a stabilisation effect may be achieved if the handling includes compaction at the optimum water content. The time elapsed between slurrying and placement in the disposal site must be minimised, or the permeability generally deteriorates (Kullberg et al., 1989). Although these residues from dry or semi-dry APC systems exhibit "temporary" setting qualities, these are not considered pozzolanic reactions which can bind metals into the matrix. Moreover, although this "freezing" of CaCI 2 and other salts provides some strength to the solid matrix, the material is highly susceptible to disintegration due to hydration. In contrast, many combined residues from wet scrubber systems show self-binding properties. One example of this is the Bamberg model (Reimann, 1990), where the ESP ash (26 kg/t) is mixed with 12 kg neutralising sludge (1 kg TS) from the first washing step in the APC system and 30 kg gypsum sludge (3 kg TS) from the second washing step. The mixed product becomes a paste, which increases in temperature to 65 ~ in a few hours. The stabilisation process requires weeks for proper curing. Although it has been found that the chlorides still leach, the metals become less leachable due mostly to their transformation to sulphides.

777

18.4.2 SolidificationlStabilisation with Binders Binders, both organic and inorganic, are commonly used as SIS reagents. Table 18.4 outlines typical binders in both categories. Cost plays a major role in the selection of binders because large quantities are normally required. Table 18.4 Typical Binder Reagents Wastes ...... Commercial

Bitumen

Polymeric

Cement kiln dust

Portland cements

Hot emulsion

Epoxy

Blast furnace slag

Lime

Cold emulsion

Polyesters

Lime kiln dust

Limestone

Polyolefins

Coal fly ash

Quicklime

Urea Formaldehyde

Inorganic binders are normally chosen because of their lower cost. Polymers are the most expensive binders (up to $500/tonne of processed waste), and do not react with the waste but rather encapsulate the materials to prevent their release. Combinations of inorganic and organic binders have been proposed, including (1) polystyrene and cement and (2) polymer gels and silicates with lime cement.

Cement-Based Systems The early development of cement-based fixation techniques was in the area of low level radioactive waste disposal. The development then proceeded into the area of high volume waste disposal, particularly with respect to sludges contaminated with heavy metals and to contaminated soils. Cement has also been used with complex wastes containing PCB, oils and oil sludges, wastes containing vinyl-chloride and ethylene dichloride, resins, asbestos, sulphides and other materials. The adding of cement generally decreases the hydraulic conductivity and porosity of the treated material, and increases the tortuosity, durability, strength and volume. Once the material has cured, the release rate of metal contaminants will usually be low, as long as the physical integrity of the material remains intact. Cement-based techniques have the following advantages: widely available at a reasonable cost the technique of mixing and handling is well developed the necessary equipment is readily available the technique is tolerant to chemical variations in the waste the strength and permeability of the final product can be varied by controlling the amount of cement added in the process

778 The disadvantages of cement-based techniques are: cement adds considerably to the weight of the waste pretreatment may be necessary for waste containing impurities that affect the setting and curing of cement cement/waste mixtures of low strength are often vulnerable to acid leaching solutions the matrices do not effectively bind chloride salts Cement based techniques generally use Portland Cement along with other additives to modify the properties of the cement. These have both positive and negative benefits such as increased shrinkage and reduced strength. The waste materials are mixed with Portland cement and sufficient water to ensure proper hydration reactions for bonding the cement. The waste is incorporated into the cement matrix and in some cases undergoes physical/chemical changes which further reduce contaminant mobility. Typically, metal hydroxides or carbonates can be formed which are less soluble than other ionic species of metals. The final products can vary from a granular, soil like material to a solid monolith, depending upon the amount of reagent added, the type of waste, and the water content. Metal salts including those of manganese, copper, lead, tin and zinc have been found to cause large variations in setting times and final strength. Sodium salts such as arsenate, borate, phosphate, iodine and sulphide have been found to act as retarders. Because high sulphur bearing wastes can result in sulphate attack on the cement matrix, special formulations (either Type II or Type V cements) may be preferable. Type II cement, as classified by ASTM C150, has a limited amount of C3A (tricalcium aluminate) and C3S (tricalcium silicate) and is used where moderate exposure to sulphate is anticipated. T y p e V is preferable for exposures where high alkalinity is expected and in structures exposed to seawater. Tricalcium aluminate is the compound most susceptible to sulphate attack and therefore minimum quantities (<5%) are present in Type V cements. Two major programs were undertaken to evaluate the effect of cement-based solidification processes on incinerator residues, namely the Environment Canada sponsored studies under NITEP (1988 through 1991 and summarised in Environment Canada, 1993), and a US EPA study (Kosson et al., 1993). Only the major conclusions from these studies are included here. More detailed information can be obtained from the referenced reports. The Environment Canada studies were conducted on APC residues from four different types of incinerator facilities. Portland Type II cement was used as the basis for the all formulations and one of three different types of waste pozzolans including:

779 9 9 9

cement kiln dust coal fly ash blast furnace slag

The study is outlined in greater detail in the next section. The US EPA study was conducted on bottom ash, APC residue and combined bottom ash/APC residue using 4 different vendor specific technologies which covered a range of processes and a control formulation using Portland cement as the only solidification additive. The 4 vendor specific formulations included: Portland cement, polymeric additives and other proprietary additives Portland cement, soluble silicates and dry carbonaceous material cement kiln dust and proprietary additives soluble phosphate additive (patented WES-Phix process) Some of the results from the two programs were very similar. First, none of the formulations tested were effective at reducing the potential release of salts from the residues, irrespective of residue type. This was not surprising, since it appears to be a typical result for most cement-based formulations used on wastes containing readily soluble salts. With respect to the physical strength of various formulations, the EPA study observed that the control specimens (cement additive only) resulted in unconfined compressive strengths (UCS) greater than or equal to the other solidification-based formulations with proprietary additives. Furthermore, the study also found that the formulations which resulted in the greatest UCS were also the most resistant to weathering. In addition, it was noted that the physical properties of cementbased formulations for bottom ash and combined ash could be greatly improved "by optimising process design based on results of multiple test criteria" (Kosson et al., 1993). All of the solidification-based formulations from the EPA study resulted in more physical, rather than chemical changes to the trace metals in the different types of incinerator residues. However, one stabilisation formulation was shown to have reduced the potential release of lead from the residues (see Stabilisation). Conversely, the Environment Canada studies indicated that some of the solidification formulations resulted in both chemical and physical retention of trace metals in APC residues. It should also be noted that both studies concluded that the rate of release of "potentially toxic metals" from SIS treated residues would be very slow. The disparity between the earlier conclusions regarding chemical retention merely emphasises that the effectiveness of individual solidification formulations can vary, and therefore should be evaluated on their own merits based on a battery of physical, chemical and leaching tests.

780

Examples

There are a number of facilities which make use of some type of cement-based SIS process to treat the different ash streams from MSW incinerators. "Monofill TM'' is a cement-based additive developed by Cementa Sweden. This material is used to stabilise the APC residue at the Hogdalen incinerator in Stockholm. The recipe calls for mixing 1000 kg of residue with 460 kg of the "Monofill TM'' additive and 1290 L of water (Sundberg, 1991). Approximately 8,000 tonnes of APC residue are treated annually. A German process was developed to stabilise residues with cement between 19821986 (Wewer & Maurer, 1990). The formulation was based on mixing 65% APC residue together with 35% cement. The water addition was 35% of the total dry weight. Although the technique was deemed feasible, it was decided not to build a full-scale system because less expensive disposal was available in old coal and salt mines. Furthermore, it was found the salts could not be effectively retained in the solidified matrix. In Germany, 19 incinerator plants are using the "UTR Technology" (UTR, 1992) to stabilise about 150,000 tonnes of ash annually. APC residues are stabilised with cement and proprietary additives prior to disposal in quarries. There is little other information available on the process. "Alinitcement" is a hydraulic chloride containing a binder which has been used to stabilise APC residues. The theoretical composition is 21 CaO - 6 SiO2 - AI203 - CaCI2 (RKW, 1992). "Alinitcement" can be used in mining as a back filling material and as stabilising agent for other wastes. The leachability of salts from this material is reduced by dilution to 50%. In Vienna, all types of incinerator residues are transported to a central processing facility where the magnetic material is separated and particles over 50 mm in size are removed (Magistratabteilung 48). The remaining material is stabilised with Portland cement and water. The mixture is 1,300 kg bottom ash, 350 kg fly ash, 200 kg cement and 120-150 kg water. The stabilised residue is used to build dikes around a conventional MSW landfill site. IVR-Techform (Inertiserung von RauchgasrOckst~nden) is used at a number of incinerator facilities in Switzerland. The APC residue is first washed at L/S of 2:1 before it is stabilised. The sediment from the washing process is mixed with 620 kg cement and additives, and 775 kg of water. A schematic of the system is given in Figure 18.5 which is used to treat 26,000 tonnes of APC residue annually. In Japan, about 110 facilities use some type of cement-based stabilisation process to treat APC residues (Wakamura and Nakzato, 1992).

Figure 18.5 The Principal Design of the IVR Stabilisation System r

Ash

Filter Press

dry matter 250 kglh water 250 kglh total 500 kglh

washed ash dry matter 2900 kglh water 2000 kglh total 4900 kglh

7200 K g h

4 - 5 blocks Per hour. Size 1 ma

Techform ENG AG, 1992

OR 7500 kglh of a pasty sludge filled into containersltroughs

Operation - 1 shift Iday Monday - Friday. Total 32 h Iweek

effluents (wastewater) 7.5 malh

782

Waste Pozzolanic Systems Pozzolanic materials are defined as siliceous substances that react with lime in the presence of water. The presence of high concentrations of silicates distinguishes these materials from the Portland cement or lime-based systems. Although they are much less expensive than cements, formulations generally require a higher proportion of these additives, compared to cements, to promote sufficient physical strengths. Waste pozzolans include coal fly ash, fluidised bed combustion material, cement-kiln dust and processed blast furnace slag. The major advantages with waste pozzolanic-based systems are: the binders are inexpensive and are generally widely available the processing equipment is widely available and simple to operate the chemistry of pozzolanic reactions is relatively well known However, processes using waste pozzolans can suffer from problems similar to cementbased processes, especially with regard to setting, curing and stabilising organic laden materials. For example, the decomposition of organic material in sludge stabilised waste can result in increased permeability and decreased strength. The major disadvantages of waste pozzolanic-based systems are: the binders add weight to the disposed materials formulations can be susceptible to acidification if there is insufficient buffering capacity curing and setting problems can occur due to inorganic salts in incinerator residues these formulations are temperature sensitive and may set very slowly at lower temperatures The final products generated from these formulations can vary from a soft fine-grained material to a hard monolithic block. As indicated previously, solidification formulations using waste pozzolans have been tested by both Environment Canada and the US EPA. Although most of the samples tested by Environment Canada were observed to swell during the curing period, this did not appear to be detrimental to the hydraulic conductivity or the physical strength of the samples. It was also noted that despite the fact that the weight of the total material (including water) increased by about a factor of 2, the density also doubled, resulting in very little, to no volume increase. All of the formulations generated material which had hydraulic conductivities as low as dense compacted clays (<6 x 10-8 m/s). All of the average compressive strength results indicated that the formulations exceeded the suggested criteria for placement in monofills (350 kPa) or landfills (3,500 kPa). Comparatively, the coal fly ash formulations resulted in the lowest strengths, however, the strength of the formulations tended to increase with the experience gleaned from each successive study.

783 The formulations in the Environment Canada studies were also subjected to freeze/thaw weathering to determine their durability. Formulations for two of the four residues tested generated solidified specimens which were very durable. However, the weight loss observed from all of the samples was due mostly to dissolution, rather than physical weathering. The reason for the lack of durability was the dissolution of salts and lime in the solid matrix, and sulphate attack on the pozzolanic bonds. The observed dissolution of salts is consistent with the findings from the EPA study. However, unlike the EPA study, there was no significant correlation between UCS and durability observed during the Environment Canada study. There are currently two known facilities using a cement-kiln dust based process designed by Energy Answers Corporation to treat APC residues, namely the SEMASS incinerator facility in Rochester, Massachusetts, and the SWARU incinerator facility in Hamilton, Ontario.

Chemical Stabilisation Frequently, elements such as cadmium, lead or mercury can be present in incineration residues as mineral phases that are available for leaching. These readily soluble mineral phases are usually metal salts such as CdCI 2, PbCI 2 or HgCI 2. Chemical stabilisation can convert these minerals to less soluble forms that reduce the environmentally available fraction for leaching. Successful chemical stabilisation requires the use of chemical additives that produce a more thermodynamically favoured solid phase (see Chapter 13). Three additives that have been used successfully at the full-scale level are sulphides, activated carbon and phosphates. Other additives such as ion exchange resins, clays and carbonates have also been used in laboratory studies. Sulphides are generally used as either a mercury sorbent in APC systems, i.e., either in dry or wet scrubber systems, or as a precipitating agent in treatment of wastewater from wet scrubber systems. The principal forms include inorganic additives (Na2S) or organic additives such as TMT#15 TM. The principal immobilisation reaction for the inorganic based system is: Na2S + HgCl 2 ~ HgS + 2NaCl

(18.1)

The principle immobilisation reaction for the organic based system is: H2+ HgCI 2 ~ HgS + 2H + + 2CI-

(18.2)

In the case of HgS (cinnabar), a precipitate forms within the inorganic system, which has an extremely low solubility product of 1 x 102s (KSP), indicating that HgS is an extremely insoluble precipitate. In the case of the sulphhydryl complex (TMT#15TM), the stability constant (K) is extremely high (1 x 103), indicating the organosulphhydryl

784 mercury complex is a stable complex with both the inorganic and organic additive systems. Caution must be used in interpreting the long-term geotechnical or biogeochemical stability of the precipitate or complex. Although it is possible for HgS to be oxidised and release of the Hg 2§ this is probably not a concern under normal disposal conditions. In the case of the organosulphhydryl mercury complex, the organic portion of the complex is susceptible to biodegradation. The sulphhydryl group also shows some susceptibility to oxidation. Powdered activated carbon has also been used to control mercury in dry scrubber systems. Unlike the sulphur based treatment systems, the activated carbon uses the principal of sorption to sequester mercury. The mercury forms an inner sphere complex with sorption sites on the carbon providing a relatively stable bonding environment. This process is used in a growing number of incinerator facilities around the world. With respect to residue treatment, the use of orthophosphate has been shown to be effective in controlling metal solubility in bottom ash, APC residue and combined ashes (Eighmy et al., 1989). Wheelabrator Environmental Systems has patented a process involving the addition of soluble orthophosphate to form insoluble metal phosphate minerals, such as lead phosphate (Pb3(PO4)2), chloropyromorphite (Pbs(PO4)3CI) and hydroxypyromorphite (Pbs(PO4)3OH). The principal immobilisation reaction to form simple metal (denoted as Me) phosphates is: 3Me 2§ + 2PO43 =~ Me3(PO4)2

(18.3)

The principal immobilisation reaction to form chloropyromorphite is: 5Pb 2§ + 3PO43 + CI'=~ Pbs(PO4)3CI

(18.4)

The principal immobilisation reaction to form hydroxypyromorphite is: 5Pb 2§ + 3PO2- + OH-=~ Pbs(PO4)3OH

(18.5)

All of these phosphate minerals have relatively low solubility products (10 12 to 1013 for the metal phosphates, 10.28for the pyromorphites) indicating these minerals are very insoluble. Unlike sulphide based minerals, phosphate minerals are geochemically stable (Eighmy et al., 1989). Wheelabrator is utilising this process at three of its mass burn facilities in the United States to treat combined bottom ash and scrubber residue. It is also under license to other incineration facilities.

785 Other additives have been considered to chemically stabilise metals in ash. One bench-scale study involved the use of the mineral trona (Na4(CO3)2) to provide carbonate to promote the formation of metal carbonates such as CdCO3 (otavite). This approach (Thompson, 1988) was found to reduce lead, cadmium and zinc leachability in column leaching tests on modular bottom ash/boiler ash blends. All of the above-mentioned chemical additive systems require good blending, mixing and moisture control to ensure intimate contact of the additive with the leachable fraction of the metals or elements in the residues. Moisture control can also be critical, particularly if some level of dissolution of the existing phase is required prior to reprecipitation as sulphides, phosphates or carbonates.

18.4.3 Stabilisation With Organic Additives Stabilisation with organics is a micro-encapsulation process which assumes that the waste material does not react chemically with the encapsulating material. Thermoplastic materials, such as bitumen, paraffin and polyethylene are the most common of the organic additives used to bind waste materials into a solid mass. One to two parts of dried waste can be mixed with one part bitumen at elevated temperatures (>100 ~ using specialised equipment to generate a hydrophobic solid material after cooling. Bitumen techniques are limited to the waste materials which do not contain high concentrations of salts such as nitrates or chlorides since they will react with the bitumen and can cause deterioration. In many cases, the bitumen-waste mixture is placed in a container such as a steel drum. The use of special equipment for heating and mixing restricts the use to small volumes and these processes are therefore relatively expensive. The encapsulation techniques utilising paraffin and polyethylene are similar to that used for bitumen. The thermoplastic material is mixed with a sludge at elevated temperatures and then allowed to cool. The resulting mixture is often containerised before disposal. The advantages of the thermoplastic technique include: the rate of contaminant release is generally much slower than for other techniques the thermoplastic material is relatively impervious to most aqueous solutions the thermoplastic material generally adheres well to incorporated materials The disadvantages are: the technique requires expensive equipment and skilled labour

786 the process poses a risk of volatilising some contaminants most thermoplastic materials are flammable wet sludges must be dried before mixing the technique is expensive it is not suitable for high volume wastes it is susceptible to chloride attack and is not recommended for APC residues At three different plants in Japan (Wakamura and Nakzato, 1992) the fly ash from incineration is stabilised with bitumen. The bitumen and the fly ash are carefully mixed, so all ash particles are covered. A sketch of the technique is shown in Figure 18.6.

18.4.4 Macro-encapsulation Macro-encapsulation involves enclosing the waste in a coating or jacket of inert material, thus placing an impervious barrier between the contaminant and the environment. Wastes treated in this manner include low radioactive materials, electroplating sludges and coal scrubber sludges. The technique has not been widely used, but its use can be warranted for specialised applications. Macro-encapsulation is also defined as the protection of disposed waste by a cover with a very low permeability. This cover may be of natural soil, like clay, or by a synthetic liner.

18.4.5 Costs Costs for any stabilisation/solidification process are site specific and depend upon the type of waste, pretreatment requirements, transportation distances, disposal criteria, regulatory criteria, health and safety requirements, assurance and quality control. In addition, the fluctuations in currency exchange rates makes it difficult to provide specific figures, consequently no costs are given here. It is recommended that the cost of any treatment process be evaluated based on the relative increase in cost it may cause on the tipping fee for the incoming waste.

Figure 18.6 Three Plants in Japan Using Stabilisation with Bitumen

Separation

Crushing

Mixing

Moulding

Bitumen

Moulded Material

I I I

Ferrous Scrap

Screened Material

j

I I I I I I I I I

.--------------

Wakamura and Nakzato, 1992

788 REFERENCES

Carlsson, B. and K. Tuutti. "Solidifiering Och Deponering Av Specialavfall", Stifelsen Reforsk, FoUnr 66, 1992. Conner, J. Chemical Fixation And Solidification Of Hazardous Wastes., van Nostrand Reinhold. New York, 1990. Cullinane, M.J., R.M. Bricka and N.R. Francingues. "An Assessment of Materials That Interfere With Stabilisation/Solidification Processes", Paper prepared for the US EPA Hazardous Waste Engineering Reseach Laboratory by the US Army Engineer Waterways Research Station under Contract DW96930146-01, 1987. Cullinane, M.J. and L. Jones. "Solidification and Stabilization of Hazardous Wastes. Part 2", Journal of Ha.z.ardous Materials Control, p. 24-58, March/April 1989. Demmich, J. RQckstanden.

Mechanismen und angewandte Verfah.ren .zur Verfesti.Qun.q von

Eighmy, T.T., S.F. Bobovski, T.P. Ballestero and M.R. Collins. "Theoretical and Applied Methods for Lead and Cadmium Stabilization in Combined Ash and Scrubber Residue", Proceedings of the Se.cond International Conference on Municipal Solid Waste Combustor. Ash Utilization, Resource Recovery Report, 1989. Environment Canada. Proposed Evaluation Protocol.. for Cement-based. Solidified Wastes, Prepared by J. Stegemann and P. Cote for Waste Management Branch, Report EPS 3/HA/9, 1991. Environment Canada. The National Incinerator Testing a.nd Evaluation Pro.qram: A Summary of the Charac.terization and Treatment Studies on Residues from Municipal Solid Waste Incineration. Prepared by S.E. Sawell and T.W. Constable for Waste Management Branch, Report EPS 3/UP/8, October 1993. Friedli, P. and P.H. Brunner. "Solidification of Filter Ashes from Solid Waste Incinerators", Proceedings of the Third Internat.jona! Conference on New Frontiers for Hazardous Waste M..ana~ement, US EPA, 1989. Kosson, D.S., T. Kosson and H.A. van der Sloot. US EPA Pro.qram for Evaluation of Treatment and Utili_zation of Municipal Waste Combustor Residues - Phase I, _Labor.atory Testin~ o_f Solidification/stabilization Processes, Report prepared for US EPA Risk Reduction Engineering Laboratory under CR 818178-01-0, Cincinnati, Ohio, 1993. Kullberg, S. A-M. F,~llman and J. Hartlen. A.usla.qung Unbehandelter und Stabilisierter Restprodukte Ein.. Modell zur Pro.qnos.tizierung von Umweltbelastun,qen durch Deponien. VDI Berichte Nr. 753, 1989.

789 Kullberg, S. A-M. F~llman and J. Hartl6n. Stabiliserin.q och Deponerin,q av Rok.qasrenin.Qsprodukter fran Sopforbrannin.q. Stifelsen for Varmeteknisk Forskning, Bransleteknik Nr. 370, 1991. Kullberg, S. and J. Hartl6n. Stabiliserin.q/solidifierin.q och Vitrifierin,q/for.qlasnin.a av Avfall- en International Oversikt. Stiftelsen Reforsk, FoU nr. 80, 1993. Magistratabteilung 48..Stadtreini,qun.q und Fuhrpark 1992. Personal Communication with Dipl.-Ing R. Siebenhandl. Austria, 1992. Ministry of Housing, Physical Planning and Environment. "General Administrative Order on Building Materials (Soil and Water Protection)". Netherland Government Gazette, No. 121, June 26, 1991. Neville, A.M. Properties of Concrete, Published by John Wiley and Sons, New York, undated. NUKEM GmbH. Personal communication with Dipl. Ing. H. Renetzeder, 1992. Reimann, D.O. "Filteraub-/Sorptionsschlamm - Mischverfahren - Bamberg Modell". Beiheft 29 zu Mull und Abfall (Hrsg. D.O. Reimann/J. Demmich), Erich Schmidt Verlag Berlin, s 65-68, 1990. Richlinien fer die Ablagerung von Abf~llen. Soundararajan, R. "Development of Organophilic Binders With Incinerator Ash for Stabilising Industry Organic Wastes", Proceedin,qs of the Municipal Solid Waste Technology, San Diego, California, Published for US EPA by Technical Resources, Inc., 1989. Sundberg, J. Stabiliserad rok,qasrenin,qsprodukt fran Ho.qendalenverket, deponering och miljopaverkan, 1991. Techform Engineering AG. Entsorgung von Rauchgasreinigungs - R0ckstanden aus Mellverbrennungsanlagen nach dem IVR Verfahren., 1992. UTR. Personal Communication with Mr. Poschwatta, Umwelt-Technologie und Recycling, Germany, 1992. van der Sloot, H.A., G.J.L. Wegen and E. Vega. Evaluate van Testmethoden en Criteria Voor de Beeerdelin.q van UitlooQ.q.qedra,qen Duurzaamheid van Immobilisaten, Stichting CUR. 1993. van der Sloot, H.A. Performance Tests of Solidified and Stabilized Waste Materials for Environmental Assessment and Quality Control. Workshop on Waste Characterization/ Classification, Sigtune, Sept. 23-24, 1992. Swedish Geotechnical Institute, March, 1993

790 Verordnung des Bundesministers f0r Land- und Forstwirtschaft 0ber die Begrenzung von Sickerwassermissionen aus Abfalldeponien. Proposal, 24 Febr 1992. Wakamura, Y. and K. Nakzato. 'Technical Approach for Fly Ash Stabilization in Japan. 85th An.nual Meeti0.q and Exhibition, Air and Water Mana.qement Association, June 2226 1992, Kansas City, Missouri, 1992. Wewer, H. and P.G. Maurer. Zementadditiv-Verfahren nach NUKEM. Restoffe aus der Rauschgasreinigung von Abfall - Sonderabfallverbrennun.qsanla.qen sowie von. Kohlenkraftwerken. ISBN 3 503 2888 9, 1990

791

C H A P T E R 19 - THERMAL TREATMENT

Thermal processing of both bottom ash and APC residue streams has been considered to reduce leaching of residue constituents, to further reduce the volume of the residues requiring disposal, and produce a treated material suitable for utilisation. In order to make a technically sound decision on the applicability of thermal treatment processes for MSW incinerator residues, fundamental knowledge of the various processes, including the process chemistry, temperatures and properties of the product generated are required. As a result, this chapter seeks to build on the information provided in Chapter 8, by describing the fate of elements during thermal processing of ash, and the potential applicability of these processes as treatment options. The major categories of thermal treatment are (i) vitrification, (ii) fusion and (iii) sintering and/or thermal binding. The principal differences between these categories relate more to the characteristics of the treated material, rather than the process itself. Definitions and descriptions of each of these thermal treatment categories are provided as follows: Vitrification - a chemical process whereby a mixture of glass precursor

materials and waste materials are melted or fused at high temperature to generate an amorphous, single phase glass product. Fusion - a chemical process whereby MWC residues are melted or fused at high temperature and form either a crystalline or heterogenous product. Sintering - a chemical process whereby MWC residues are heated to sufficient

temperature to allow reconfiguration of chemical phases present in the solid materials (refer to Chapter 8 for a detailed discussion of sintering). 19.1 DEFINITIONS 19.1.1 Vitrification

In a vitrification process, a residue or waste stream is incorporated with glass forming materials and/or other appropriate waste solids, by melting the mixture at a sufficiently high temperature to form a homogeneous liquid phase. Typical vitrification temperatures are between 1,100 and 1,500 ~ The molten material is cooled to form an amorphous, homogeneous, single phase solid. If cooling is not rapid enough, the molten material may crystallise or "devitrify." The generation of the monophased glass within a controlled composition range ensures consistent chemical and thermal stability of the glass product. The final glass product should not contain any undissolved refractory feed material or other crystals formed on cooling. Additionally, glass compositions should avoid the potential of liquid-liquid phase separation in the melting process which would result in a glass solid being formed with two or more intermingled glass compositions. In general, mixed phase glasses may contain one phase which is less durable in acid or alkaline environments, which may have a negative effect on the overall leaching characteristics of the glass.

792 Vitrification processes are distinguished by the physical configuration and operation of the melting furnace, and by the process chemistry. Typical vitrification furnaces or melters are refractory lined vessels heated by either fuel combustion (most frequently natural gas) or joule heating with submerged electrodes. Materials to be vitrified may be fed either batchwise or continuously. "Hot crown" or "hot top" melters function with the entire contents of the melting vessel in a molten state. "Cold crown" or "cold top" melters function with a continuous layer of unmelted feed material floating on top of the molten material being vitrified. Additional feed material is placed on top of the melter to maintain the unmelted layer as melting progresses. Size reduction or pre-treatment of the residues to allow efficient melting may be required. The objectives of vitrification processes are generally to either separate specific elements from the melted material through volatilisation or retention of specific elements within the vitrified product. However, the ultimate fate of potentially toxic constituents must be considered in all cases. Fluxing agents and glass precursors typically are added to the material to be vitrified to control retention of volatile metals during the melting process, and to control the treated material's physical and leaching properties. Process additives may be either virgin materials (silica, calcium carbonate, etc.) or recycled glass cullet. The relative quantity of additives used may vary from approximately 20 to 70% (by weight) of the total feed. Removal of excess sulphur, carbon, chloride or organics may be facilitated by appropriate thermal or chemical oxidations, or by thermal desorption at temperatures <1,000~ Vitrification units generally require an oxidising environment to cause combustion debris and contaminants to oxidise (ideally) to carbon dioxide and water, and to solubilise any trace metals present in the feed. This may also reduce the volatility of some metals during the melting process, depending on the speciation of the metals entering the melt and respeciation during the melting process. Products from vitrification can either be aggregate formed by "drigaging" (immediate immersion of the molten glass in water to form fractured, granular material), or formed into products such as fibre glass or molded shaped. Vitrified granular material has been considered for use as fill and as an aggregate substitute in pavement applications. Typical product specific gravities are between 2.4 and 2.9. Volume reductions achieved are a function of both the resulting product density and the quantities of additives used during the process. 19.1.2 Fusion

Fusion processes are similar in operation to vitrification processes but do not result in a homogeneous, amorphous glass single-phase product. Considerably less emphasis is placed on process chemistry during fusion processes. Typical melting scenarios may include the formation of a separate molten metal phase within the melter. Fused materials may be crystalline ("de-vitrified") with multiple crystal phases, or contain inclusions of unmelted feed. Products from fusion processes typically are being evaluated as aggregate substitutes.

793 19.2 GLASS COMPOSITION AND METALS RETENTION IN GLASS MATRICES

In view of the inherent variability in the chemical composition of incinerator residues, the need to evaluate the chemical characteristics and control the vitrification chemistry is considered highly important (Rabiger, 1992). The following introduction to glass structure and retention of metal constituents is summarised primarily from McLellan and Shand (1984), Wicks et al (1985, 1986) and USEPA (1992). 19.2.1 Glass Structure

Glass is a rigid, non-crystalline material of relatively low porosity, often composed primarily of silica, alumina, and oxides of alkali and alkaline earth elements. Although phosphate, sulphide, and oxynitride glasses are also important glass types, most glasses produced during thermal treatment processes are silicate glasses. Thermallyformed glasses are produced by fusing or melting crystalline materials and/or amorphous materials (e.g., previously formed glasses) at elevated temperatures to produce liquids. These liquids are subsequently cooled to a rigid condition without crystallisation. Silicate glasses are composed of three-dimensional networks. The basic structural unit of the silicate network is the silicon-oxygen tetrahedron in which a silicon atom is bonded to four oxygen atoms (Figure 19.1). The silicon tetrahedra are linked at the corners, where each shares one oxygen atom with another tetrahedron. Some, or all four, of the oxygen atoms from the tetrahedron can be shared with other tetrahedra to form a three-dimensional network. The extended 3-dimensional network is irregular and the Si-O-Si bonds randomly prevent tetrahedra from forming a crystalline network (McLellan and Shand, 1984). The shared oxygen atoms are called "bridging" oxygens. In pure silica glass, the ratio of silicon to oxygen is ideally 1:2 with all oxygen atoms acting as "bridges". Some atoms, such as sodium, are ionically bonded to oxygen when present in glass, and thus interrupt tetrahedra linking and network continuity. An oxygen atom ionically bonded to another atom is called "nonbridging". Appreciable amounts of most inorganic oxides can be incorporated into silicate glasses. Elements that can replace silicon are called "network formers". By replacing silicon in the glass network, some inorganic species (such as transition and other non-alkali metal elements) can be incorporated into a glass. Most monovalent and divalent cations (such as sodium, calcium and other alkali and alkaline earth [group la and 2a] elements) do not enter the network, but form ionic bonds with nonbridging oxygen atoms, and are termed network modifiers. Variation in the network integrity and the constituents of the glass are manifested in changes in glass properties such as softening point temperature and chemical durability (i.e., leachability and solubility) (McLellan and Shand, 1984).

794 Figure 19.1 The Basic Silicon-Oxygen Tetrahedron (a) and Linked Tetrahedron (b) Forming the Basic Structural Network in Silicate Glasses

(a)

,'

,'

,"

I

|

/

Oxygen

l

l

I

'

x

I '

II l

I

I', I',

',

',

',

~

Silicon ~ .

' Oxygen

"11 Oxygen

(b)

McLellan, 1984

795 The role of certain elements in glass, termed intermediates, may vary with conditions. For example, aluminum may be a network former or a modifier depending on the ratio of aluminum to alkali and alkaline earth ions. The role of iron depends on redox state or oxygen availability in the molten material. For example, Fe§ is a network former (McLellan and Shand, 1984) while Fe§ is not. The most common way of describing glass is to list relative amounts of oxides derived from the raw materials used in a glass formation, even though the oxides do not exist as such in the glass network. Many types of glass can be formed depending on the raw materials used. The glass industry prepares special formulations to obtain glasses with properties desirable for various uses. Important considerations for the thermal processing of incinerator residues include processing characteristics, such as melt viscosity and redox conditions, and product characteristics, such as physical and chemical durability. In order to decrease the viscosity of molten glass and lower the melting point of the raw material mix, it is often necessary to add a flux, or network modifier, that will soften the glass by generating nonbridging oxygen atoms. Alkali metals, such as sodium, are often used as fluxes in their oxide forms. Alkalis can be incorporated into glass as carbonates, or other salts that react (at elevated temperatures) with silica to form a siliceous liquid. However, adding alkali to the glass generally decreases its chemical resistance. At high alkali concentrations, the glass may become water soluble. Alkaline earth fluxes, instead of or in combination with alkali fluxes, may be used to decrease the aqueous solubility of alkaline glasses while maintaining lower melting points. Oxides of calcium and magnesium are the most common alkaline earth or stabilising fluxes. However, adding too much calcium can cause calcium silicates and aluminates to form, resulting in crystallisation (devitrification) on cooling. Soda ash (sodium carbonate) is commonly used in industry to supply alkali fluxes, while lime (calcium oxide) is commonly added to supply alkaline earth fluxes. Thus, glass made from silica, alkali and alkaline earth fluxes is commonly called soda-lime glass. Soda-lime glass is the most common type of glass, and is used in most container glass and window glass applications. In addition to soda-lime glass, two other types of glasses are typically used to immobilise waste constituents. These are borosilicate glasses and aluminosilicate glasses. Borosilicate glasses are formed through the addition of boron oxide (B203) to silicate glasses. Boron oxide forms a network of planer boron-oxygen triangles upon cooling from temperatures above its melting point (460 ~ In low alkali silicate glasses, boron oxide maintains a triagonal planar coordination and serves as a fluxing agent by reducing the cohesiveness of the silicate glass structure. Aluminosilicate glasses are formed when alumina (aluminum oxide [AI203]) is added to silicate glasses. Alumina serves as a network former and forms a tetrahedral coordination similar to silica. However, since alumina is trivalent, it reduces the number of non-bridging oxygens resulting in increased glass cohesiveness.

796 The typical composition of soda-lime glass, borosilicate glass and aluminosilicate glass are compared with the composition of glasses formed from the vitrification of incinerator residues and other wastes in Table 19.1. Figure 19.2 presents the compositions for chemically durable glasses formed from incinerator APC residues reported by Wexell (1992). Waste glasses generally contain less silica and more aluminum and iron than soda-lime glasses. The "aluminum-bearing glasses" are generally more typical of glass compositions produced in waste vitrification. The typical raw material used in industrial glass production consist of various formulations of the following main ingredients: 9S a n d

-

9Feldspar 9Dolomite 9Limestone9Soda Ash -

SiO2 KAISi3OB CaMg(CO3)2 CaCO3 Na2CO3

In addition, different trace metal compounds can be incorporated into the glass mix to provide coloration. Table 19.2 contains a list of commonly used metal compounds along with the resulting colours. Glasses typically can incorporate only small amounts (less than 1% by weight) of halogens, sulphur, nitrogen and carbon species, since a large proportion of these species are volatilised during melting. If large quantities are present in the mixture, they may form a separate phase in the melter. Elevated concentrations of these species in the product glass will result in decreased chemical durability of the material. In addition, carbon-based compounds can act as reducing agents and halogens can form more volatile metal complexes resulting in increased volatilisation of metal species during vitrification (also see chapter 8). Consequently, consideration of vitrification as a treatment technology for incinerator residues (especially APC residues) must include control of the off-gas emissions to minimise release of contaminants to the atmosphere. 19.2.2

Constituent

Retention Mechanisms

Vitrification can incorporate incinerator ash constituents in a glass matrix via two main reactions - chemical bonding and encapsulation. Certain inorganic species present in incinerator residues can be immobilised by chemical bonding with the glass-forming materials, particularly silica, present in the wastes to be vitrified. The most notable chemical bonding within a vitrified material occurs when certain metals or other inorganic elements bond covalently with the oxygen atoms in a silica network, becoming part of the network. Inorganic elements that interact in this way are considered network formers, since they essentially replace silicon in the glass network structure. Other inorganic species can behave as "network

Table 19.1 Typical Composition of Soda-Lime, Borosilicate and Aluminosilicate Glasses Compared to Glasses Formed from Vitrification of Incinerator Residues and other Wastes Element

Typical SodaLime Glass'

Borosilicate Glass2

Aluminosilicate Glass3

CaO BaO

6 - 10 nr - 0.5

nr 2-3

8 - 10 0-6

17.2 0.04

PbO

nr nr

0-6 nr

nr nr

0.2 0.9

Fe'203

Vitrified Vitrified MSWl APC flyash5 Residue (C~rning)~

Vitrified MSWl ash (Kubota)'

Durable Glass Compositions for APC Residue

Whatcom County Ash Glass Sample

24 nr

16 nr

9-31 nr

11.30 0.17

<0.005 1.8

0.03 4

nr nr

0.22 20.10

0-7 0-7

1.10 0.56

nr nr nr nr nr nr TiO, nr nr nr
798 Figure 19.2 Acid Durable Glass Compositions for APC Residues 10

~

O ~l~ll-~

0.7/- 0 , - ,

m

Si02

"2

~-k0.3

,"', " ~

k~

CaO

oy~ X;~ o,;~/+~., 2,/ +~.,

0.7

0.6

_

~

0.5

0.4 0.3 AI203

0.2

0.1

9 EPA Program Ash Glasses O

Misc. MSWAsh Glasses Sources land II 9 EPA Demo Glass

Na20

CONTENT Weight (%)

EPA Ash

2 - 14.8

i

O

Misc. MSW Ash EPA Demo

Wexell, 1992

1.7 - 12.8 5-7

0

799 Table 19.2 Inorganic Colorants used in Commercial Glasses Colour Produced

Material Under Oxidation Cadmium Sulphide None Cadmium Sulphide, Selenium None

Under Reduction Yellow Ruby

Cobalt Oxide Copper (11)Oxide

Blue-violet Greenish blue

Blue-violet Greenish blue

Copper (11)Oxide Cerium Oxide Chromium (111)Oxide Gold

Greenish Blue Titania Yellow Yellowish Green Ruby

Ruby Yellow Emerald Green

Iron (11)Oxide Manganese Dioxide

Yellowish Green Amethyst to Purple

Bluish Green none

Neodymium Oxide Nickel Oxide

Violet Violet in K20 glass

Violet Violet in K20 glass

Nickel Oxide Selenium

Brown in Na20 glass Fugitive

Brown in Na20 glass Pink

Sulphur Uranium

none Yellow with Green fluorescence

Yellow to Amber Green with fluorescence

Tooley, 1984

modifiers", forming ionic bonds with oxygen or other elements in the glass network. This ionic bonding incorporates the material into the glass, but disrupts the network continuity, thereby modifying the vitrified material's physical and chemical properties (see below). Ash constituents also may be immobilised without direct chemical interaction with the glass network. Constituents that do not interact chemically or have not completely entered solution upon melting can be surrounded by a layer of molten material and encapsulated, as the melt cools. This layer of glassy material can protect the encapsulated constituents from chemical attack, and limit their ability to escape from the fused product (McLellan and Shand, 1984).

19.2.3 Chemical Attack and Leaching Mechanisms Vitrified and fused materials are often thought of as being relatively "inert", because these materials exhibit high corrosion resistance compared with many other materials. However, all vitrified and fused products are chemically reactive to a limited degree.

800 There are two major forms of chemical attack on vitrified materials, namely, matrix dissolution and interdiffusion. Matrix dissolution is characterised by alkali attack, typically at solution pHs greater than 9, which begins by hydration of the silica network and may proceed to product dissolution. In pure silica glass, the matrix dissolution process can be described by the following equation: 2 N a O H + SiO 2 --, Na2SiO 3 + i-120

The resulting alkali silicate (Na2SiO3 in the above example) is usually water soluble, so as the silica network is attacked, the other constituents in the material are also released. Alkali attack is highly pH dependent. The rate of attack generally increases by a factor of 2 to 3 for each pH unit increase above 9. The influence of temperature on the rate of alkali attack follows an Arrhenius relationship with the rate of attack increasing by a factor of 2 to 2.5 for each 10~ temperature rise. Interdiffusion is typified by acid attack, typically at solution pHs less than 5, on vitrified materials. While alkaline attack (matrix dissolution) leads to surface dissolution of the vitreous material, interdiffusion is an ion exchange process which preferentially extracts elements present as network modifiers, leaving the silica structure primarily intact. Generally, interdiffusion involves the exchange of hydronium ions in solution for ionically bonded elements in the vitreous network (McLellan and Shand, 1984). The rate of acid attack on glass is generally proportional to the square root of time. Since the process is controlled predominantly by diffusion, the rate of leaching decreases as the thickness of the leached layer near the glass surface increases. However, this effect can be limited if the layer dissolves or sloughs off. Although the rate of interdiffusion is influenced by temperature in a relationship similar to that for alkali attack, the interdiffusion rate increases only by a factor of 1.5 to 2 for each 10~ temperature rise. While dissolution and interdiffusion describe leaching under many conditions, the leaching of many waste glasses appears to be modified by the formation of surface gel layers (Wicks et al, 1985). Layer formation is favoured in static or near-static solution conditions and where silica is present, as in many groundwaters. As matrix dissolution occurs, the surface layers, composed of insoluble glass components, arise. The formation of these layers proceeds in a three-stage process. Stage one is dominated by interdiffusion as network modifiers, such as sodium, diffuse out of the glass and into solution, and water diffuses in. The result is a modifier-deficient surface layer. During this stage, the pH of the leachant increases from the formation of alkali hydroxides in solution. Stage two is dominated by matrix dissolution. Stage three is characterised by the formation of surface layers from the precipitation and adsorption of insoluble compounds onto the surface of the glass. These compounds are the remaining more insoluble waste glass constituents. For example, these surface layers may contain substantial iron and manganese hydroxides. Where a surface layer forms, it can greatly reduce leaching of the waste glass underneath.

801

19.2.4 Factors Influencing Vitrified Ash Leaching The use of vitrification to treat high level radioactive waste has resulted in considerable knowledge about waste glasses and their production, particularly in terms of chemical composition, waste loading, temperature, time and pH.

Chemical Composition

Chemical composition plays an important role in product durability (Wicks et al, 1985). In general, as the ratio of oxygen to network formers (such as silicon) decreases, more bridging oxygens are produced, resulting in a more durable product. Network modifiers such as alkalis and alkali earth oxides tend to decrease glass durability. This occurs because these oxides increase the oxygen-to-network former ratio and produce more singly-bonded oxygen, thus breaking up the glass network. However, these elements do decrease the melt viscosity and lower processing temperatures, and are used as fluxing agents. In general, oxides with valences greater than 1 may increase glass durability. Composition of the waste feed can have enormous effects on product durability and processing parameters. Table 19.3 presents some of the effects of various inorganic oxides on processing and glass durability. Modification of the waste stream through additives and, or material removal can have dramatic impacts on processing and product characteristics. However, as Table 19.3 indicates, most additives have both desired and undesired effects.

Waste Loading

Increased incinerator residue loading does not necessarily increase product leachability (Wicks et al, 1985; Mendel, 1973). Research on borosilicate glass for the immobilisation of nuclear waste has indicated that glass leachability is reduced as the waste loading increases from 0 to 35% (by weight), with only small changes in leachability as the waste loading increases from 35 to 50% (Rankin and Wicks, 1983). In general, the incinerator residue loading in the product will be limited by the waste composition and the desired product physical and chemical properties. The amount of incinerator residue immobilised by borosilicate glass may not be limited by product durability, but by processing considerations. Some beneficial effects observed from increased waste loading on chemical durability results from the formation of surface layers during leaching that are made up of the major constituents found in the waste.

Temperature

Leachability of waste glass increases with temperature (Wicks et al, 1985). The mechanism of corrosion varies with temperature, i.e., at temperatures near ambient conditions, diffusion effects can dominate glass corrosion, but at temperatures near 100~ or higher, network dissolution can dominate. The exact temperature for the shift in mechanism varies with test conditions and glass composition.

802 Table 19.3 Effects of Waste Glass Components on Processing and Product Performance Fdt Processing Product Performance Components SiO 2 Increases viscosity greatly; reduces waste solubility increases durability B203

Reduces viscosity; increases waste solubility

Increases durability in low amounts, reduces durability in large amounts

Na20

Reduces viscosity and resistivity; increases waste solubility

Reduces durability

Li20

Same as Na20, but greater effect on a unit mass basis; increases tendency to devitrify

Reduces durability, but less than Na20

K20

Same as Na20; decreases tendency to devitrify

Reduces durability, but more than Na20

CaO

Increases, then reduces viscosity and waste solubility

Increases then reduces durability

MgO

Compared to CaO has much greater effect in increasing viscosity and a smaller effect in reducing viscosity; reduces tendency to vitrify

Same as CaO, but more likely to decrease

TiO 2

Reduces viscosity slightly, increases, then reduces waste solubility; increases tendency to devitrify Reduces waste solubility

Increasesdurability

ZrO 2, La203 Waste Components

Processing

Increases durability greatly Product Performance

AI203

Increases viscosity and has tendency to devitrify

Increasesdurability

Fe203 U3Os NiO

Reduces viscosity; is harder to dissolve Reduces tendency to devitrify Is hard to dissolve; increases tendency to devitrify

Increases durability Reduces durability Reducesdurability

MnO

Is hard to dissolve

Increases durability

Zeolite

Is slow to dissolve; produces foam

Increases durability

Sulphate

Is an antifoam, melting aid; increases corrosion of processing equipment

Too much causes foam or formation of soluble second phase

US EPA, 1992; Adapted from Plodinec et al., 1982

Time At a given temperature, the largest leach rates occur during the early stages of leaching (Wicks, 1985). Therefore, leach rates usually decrease over time. Two mechanisms appear to be involved in this leach rate decrease. First, under static or near static leachant conditions, the solution becomes saturated as elements are extracted from the glass and enter solution. Secondly, with time, the formation of surface layers can further inhibit leaching (Jantzen, 1988).

803

19.3 PROCESSING 19.3.1 Processing Equipment A melting furnace or kiln is the central component for most treatment processes. Melter design is derived from either the glass manufacturing or smelting industries. Melters may be designed for operation either in a hot crown or cold crown mode. Energy required for melting can be supplied by electrical resistance (joule) heating, electric arcs or fossil fuel combustion. Use of fuels may be either through burners placed in the melting chamber or addition of fuel to the melter feed. Diagrams of three melter designs are provided in Figure 19.3. Additional equipment required for thermal treatment processes included materials handling and air pollution control devices. Many processes propose recycling of the off-gases to the APC systems of the incinerator facility. In this case, a purge stream would be required to manage constituents (e.g., sulphates, halogens, etc.) that could not be incorporated into the fused or vitrified product.

19.3.2 Energy Requirements and Costs Costs associated with vitrification of residues that have been estimated are widely variable (Edwards, 1994). Significant components of overall processing costs include capital investment, fuel or energy requirements, periodic refractory replacement, electrode consumption (when required) and management of process residues. Processing costs also can be estimated based on similar processing costs associated with glass production. Processing costs estimates for waste vitrification typically range between $150 and $750 per tonne of product produced. Capital investment, fuel and feed preparation typically represent 20, 50 and 10 percent of the total costs, respectively. A realistic first estimate of vitrification costs for incinerator residues is considered to be between $200 and $250 per tonne of product. These costs do not include cost recovery from sale of product or cost of product or residue disposal. Thermal treatment of incinerator residues may not be economically attractive except under specific conditions. For example, if the residues are a substitute raw material for a glass in commercial production, processing costs associated with using incinerator residues may not be prohibitive. Other situations which may make it attractive are avoidance of extremely high disposal cost or use to meet specific regulatory requirements.

19.4 THERMAL TREATMENT PROCESSES UNDER DEVELOPMENT 19.4.1 Overview of Reported Processes Several thermal treatment processes are under development by vendors. In general, each of the processes can be classified based on the characteristics of the product

804 Figure 19.3 Schematic Diagrams of Typical Melters; Submerged Electrode Melter (a), Fuel Fired Glass Melter (b), Centrifical Plasma Arc Melter (c) Batch Blanket Fusion Area

Ele~es

20' - 40'

~.-

(a)

Furnace Melting End Regenerator , PORT Furnace Working End ~

,,

~

~

I Lill Ill

(b)

.

.

~

, I' '1 , .

.

7

II

FOREHEARTH ,I, 4 CONNECTION

.

I

DOGHOUSE 60'

-

100'

Water-Cooled Copper Electrode

(C)

Plasma Gas Injection Arc Termination \\ ~\

Nozzle Slag Bath ChPPatr

Spinning Reactor Well

1

--ilo Slag Removal

Eschenbach et al., 1989

"o

"-

i v

805 produced, the type of melter and the process chemistry involved. Table 19.4 provides an overview of the processes that have been reported. Most references contain a limited amount of technical details about the processing conditions and performance. Furthermore, most process demonstrations have been limited to batch (crucible) laboratory evaluations. In Japan, fusion/vitrification processes were originally developed for management of municipal sewage sludge and currently there are approximately ten full scale plants operating, some of which are processing bottom ash or bottom ash in combination with APC residues (Sakai & Hiraoka, 1994). Melting of APC residues only is under development but not in commercial application. A few processes (Kubota, Takuma, Corning, and ABB Deglor) have been demonstrated at pilot or commercial scale and are described in more detail in later Sections. A process that integrates RDF combustion with sintered aggregate production (Neutralysis Process) is also discussed later. 19.5 VITRIFICATION Several vitrification processes have been reported in addition to those developed by Coming, Inc. The majority of these processes have been carried out in such a manner that results in the separation of volatile trace elements (eg., cadmium, lead, chloride, sulphur, etc.) from the incinerator residues. The off-gases from these systems are passed either through the existing incinerator APC system or through an independent APC system. When process off-gases are recycled to the incinerator, a purge stream is required to avoid infinite accumulation. Similarly, treatment and disposal of residues from independent APC systems must be considered. Although the applications for bottom ash appear to be more feasible, the only advantage to vitrifying APC residues is the reduction in the mass requiring further management. The requirement for off-gas treatment tends to defeat the purpose of treating the APC residues if the metals are not retained in the melt. Combined ash from an incinerator was melted using joule heating and electric radiant heating in pilot unit being developed by Westinghouse Environmental Systems and Service Division (ESSD) (Westinghouse, 1990). The specific goals of these tests were to: Investigate the technical feasibility of combined ash vitrification Determine the emissions from the furnace while vitrifying combined ash Evaluate the vitrified ash material for leaching potential Determine the fate of the trace metals during vitrification Combined ash was fed into the reaction chamber by means of a hopper and ram charger system. No glass forming additives were added to the combined ash prior to melting. The reaction chamber consisted of two sections, an upper dome with resistance heater, and a lower, water-cooled tub with submerged, water-cooled electrodes. The roof of the dome was cooled by forced-air convection. The reaction chamber off-gases passed through several treatment steps prior to being discharged

Table 19.4 Summary of Thermal Treatment Processes under Development or in Use Process (Developer)

Short Description

Thermal Procedures SOLUR Procedure (developed by Sorg and Lurgi, marketed by Sorg, Germany)

glass melting kiln with directed heating by metal electrodes, unheated upper oven, glass byproducts are taken out

Temp 1300 1400°C

Feed Stream (MSWI) Residue and Additives

Mass Balance Volume Reduction

Energy Needs

Developmental Status

Filter ash and scrubber residue

adding 34 % glass waste

0.8-1.8 kWh1kg

32-36 M% SiO, 10-14 M% AI,O, 22-24 MOhCaO 2-5 M% CI

80% glass 10% gas 8.5% glass byproduct (melt crust) 1.5% Recycled

test installation 100 kglh (up to max. 1 tonnelhour)

cold crown oven

(shutdown)

volume of 50V O h + 5V % remainder

DEGLOR Process = decontamination and glassification of residues (ABB + W +E

continuous flow kiln with indirect electric heat by resister rods; hot crown oven, in larger plants additional bath electrodes

1300 1500°C

fully electric melting process (Melting technique Jodeit made in Jena, GmbH i, G)

glass melting kiln with direct heating by molybdenum electrodes, glass surface covered completely with feed (cold crown)

1400°C

filter ash

7540% glassified material 5-7% heavy metal salt concentrate

1-1.2 kWhlkg

experiment station

1.8-2.0 kWhlkg

experimental melting to form aggregate 1-2 tonneslday

volume reduced to 30°hV residues from hazardous waste incineration from SAV 'Schoneiche" (semi-dry)

M = Mass, V = Volume, M % = wt %, DM = Deutsche Mark (1 DM = $0.6)

7942% solids 7-10 % glass byproduct (slag)lcondensate 10 % vent produced gas

HinwillSwitz. lOOkglh since 1989

test facility under construction

Table 19.4 Continued Process (Developer)

Short Description

Temp

Feed Stream (MSWI) Residue and Additives

Mass Balance Volume Reduction

Energy Needs

Developmental Status

induction melting procedure

inductive (electrical) melting

1400°C

boiler and filter ashes

?

?

small scale test runs; data not available

staged combustion process with pulverised coal oxygen burners in melting cyclone

1400 1600°C

boiler ash, fly ash, fine fraction (
10% lime

--

(testing) station in Ksln-Porz (FRG) test facility

DBA and Steinmijller CORMIN process = continuous residue mineralisation (KHD Humboldt Wedag AG)

melting film develops on the cyclone wall and drains off

63-72% slag 10% dust 5-28% exhaust (vent gas) volume reduced to 25% of volume

burn-melting process (BABCOCK)

melting of bottom ash in gas fired kiln; produce mineral wool, condense some of heaw metals aaareaate

1300 1400°C

bottom ash, boiler ash, filter ash

81% glass 3% glass byproducts, heavy metal concentrate

1.2 - 1.8 kWhlkg

experiments unknown

Flame chamber meltdown procedures (example: TAKUMA procedure (Kubota, Japan) offered in Germany by LentjeslMannes m ann) M = Mass, V

twin cylinder melting process with 2 cylindrical surfaces, material added through a slot, heating oil or gas used as energy source

1300 1600°C

boiler and filter dust, i.e., slag

at the present:

2 kWhlkg

large scale plant process to 1 tonnelhour, 29 plants in Japan, different types of furnaces

55% slag 45% filter dust

costs per plant approx. 10 million DM

= Volume, M % = wt %, DM = Deutsche Mark (1 DM = $0.6)

Table 19.4 Continued Process (Developer)

Short Description

Plasma melting process (KruppMak)

Plasma reactor, heavy metals will be vaporised

RedMelt process (BAM, TU Berlin)

arc type furnace with 3 graphite electrodes

Temp

Feed Stream (MSWI) Residue and Additives

Mass Balance Volume Reduction

Energy Needs

Developmental Status

1400 2000°C

filter ash

70% glassified granular substance 25% salt and heavy metal residues 5% dust

1.8 kWhlkg

Pilot Plant in Stapelfeld Maximum 1.3 tonne Ihour, cost: 670 DMltonne full-scale under planning, (Kiel, FDR) plant costs approx. 20 million

1250 1500°C

bottom ash filter ash ash from sewage sludge incinerator

89-95 M% Product with minimal heavy metal content

0.8 kWhlkg

experimental plant batch type 100 kgltonne

heating by means of electrical resistance of melted product under reducing atmosphere 2-phase, metal and slag

HSR-Process 'Holderbank" Smelting-Redox

1. slag will be melted and fuel oxidised with pure oxygen in a blast furnace

(Holderbank1 Switzerland)

2. partial reduction of liquid slag with H-)HO , mixture

1300 1500°C

bottom ash filter ash ash from sewage sludge incinerator

M = Mass, V = Volume, M % = wt %. DM = Deutsche Mark ( I DM = $0.6)

4-8 M% molten metal phase

Pilot plant is being planned in Berlin

1-5 M% Pb, Zn containing metal phase as exhaust gas

incineration costs: -230 DMltonne

80% M% granular Fine material 0.4 M% heavy metal condensate (Zn, Cd, Mg) 0.1 M% heavy metal precipitate

0.7 kWhlkg

Laboratory experiments pilot (plant) planned cost: 160-270 DMA

Table 19.4 Continued Process (Developer)

(L8C Steinmuller)

Short Description

glass kiln fired with natural gas, combustion air enriched with 0, (90% 0,)

Temp

1300 1500°C

Feed Stream (MSWI) Residue And

Mass Balance Volume Reduction

Energy Needs

Developmental Status

slag filter dust sewage sludge incinerator ash

90 M% glassified product, contains heavy metals 2.5 M%

1.2 kWhlkg

experimental plant up to max of 100 kglh Pilot plant being planned

condensate of exhaust gas (Hg, Cd, Zn) 7.5 M% exhaust as total process carbonising (distilling) burning process

High temperature burning chamber with ---air supply fuel: gas and carbon from previous ----

1300°C

residues from the carbonising or distilling process up to 450°C. boiler and filter dust

Siemens-KWU M = Mass, V = Volume, M % = wt %, DM = Deutsche Mark (1 DM = $0.6)

Courtesy of Urnweltbundesarnt, 1992

--

---

Ulm, FRG, Test facility 200 kglh Full-scale facility planned in Forth, FRG

810 to the atmosphere. These treatment steps included particulate removal, cooling, and caustic scrubbing to neutralise acid gases produced during processing. An induced draft fan maintained a negative pressure throughout the reaction chamber and off-gas treatment system. The principal conclusions from test operations were: 1. The majority of the trace metals found in the ash, specifically lead and cadmium, volatilised from the glass melt. The metals which remain in the glass were susceptible to leaching. When furnace temperatures fell below 900~ carbon monoxide emissions increase appreciably. A commercial system functioning in this capacity may require an afterburner.

,

,

~

Carbon accumulated on top of the glass melt during the testing of "cold top" operations. This indicated that continuous operation in this mode may not be practical unless a method can be found to burn the carbon more efficiently. Pretest metal separation did not prevent molten metals from pooling on the bottom of the furnace. When designing a larger processing facility, the cost involved in improving the metal separation versus building a furnace with separate glass and metal discharges should be considered.

Champan (1989) has reported on laboratory studies involving vitrifying combined ash from the Whatcom County Incineration Facility, Washington, USA. This facility incorporates two Consumat CS-2000 incinerators with a nominal processing rate of 90 tonnes per day. The air pollution control system at the time of the laboratory research was a dry acid gas scrubber followed by a fabric filter unit. Samples of the combined ash were oxidised by isothermal treatment at 750~ for up to 24 hours. Oxidation was assumed complete when no further weight loss was noted after 2 hours. Glass was made from 100 to 300 grams of oxidised ash by heating in a crucible at 1500~ for 1 to 2 hours. Molten iron typically collected at the bottom of the crucible. The bulk of the melt was a homogeneous glass up to the top surface, which was typically composed of an unmelted rim. This rim was enriched with oxides of aluminum. Apparently, aluminum metal in the ash melted, beaded to the surface, and slowly oxidised around the periphery. This portion of the heated mass typically did not melt into the molten glass and a sulphur odour was associated with the crusty layer. After melting from 6 to 14 hours at 1500 ~C, the rim of enriched aluminum reacted with the bulk to form a homogeneous melt. The glass was very lustrous, reflective, darkblack material with densities between 2.75 and slightly more than 3 g/cm 3. The highly lustrous surface may have resulted from reducing conditions during the melting process and the formation of metallic iron or hematite on the surface. The high density was attributed to the high concentration of iron in the glass. The 100 poise melting

811 temperature of the resulting glass was about 1,325~ The resistivity for this glass was 12 to 13 ohm/cm, at a viscosity of 100 poise. Table 19.1 includes the composition of the vitrified combined ash. 19.5.1 Vitrification of APC Residues by Corning, Inc. The objectives of this vitrification process was to treat APC residue to form a homogeneous glass material as the principal product which was resistant to subsequent leaching and to retain potentially volatile metals in the glass (Wexel, 1992; Kosson, 1993). This process was one of several vendor processes evaluated under the USEPA program for evaluation of treatment and utilisation processes for incinerator residues. The APC residue used in this investigation was obtained from a mass burn incinerator incorporating a lime slurry spray drier (wet-dry) acid gas scrubber/fabric filter APC system. The bench-scale vitrification process consisted of three principal substituent processing steps (Figure 19.4), including: .

,

,

Two serial aqueous extractions of the APC residue followed by suspended solids recovery and drying at 500~ to remove chlorides and carbon from the APC residue Blending of the dechlorinated APC residue with glass forming additives, and Melting of the blended dechlorinated APC residue and additives into a homogeneous glass in a cold crown melter, producing a glass product

The temperature of the molten glass is a function of the specific glass composition, but typically is between 1,100 and 1,500~ The blanket of unmelted material served as a cooling layer for gases formed during the melting process and served to help reflux condensed volatile metals back into the reactor, however, the melting process step did result in the emission of off-gases. Glass products formed after the melting process were annealed to prevent stress fractures during cooling. The APC residue dechlorination was accomplished by batch extraction with water and then filtered to recover the solids. The moist filter cake was dried at 500~ for approximately 16 hours (overnight) to remove water and residual organic carbon, leaving approximately 67% of the original dry weight of the APC residue. Approximately 30% of the calcium, 60% of the potassium, 17% of the sodium, 29% of the lead and 19% of the sulphur present in the untreated APC residue were removed from the system during the dechlorination process.

Figure 19.4 Process Flow Diagram of APC Residue Vitrification by Corning, Inc.

Dl Water

Dl Water

Additives Water Vapour and Off Gases

I I

Transfer Line

A

I I

APC Residue I I

500 "C

,

I I I

Crucible

I I I

+

v

Dechlorination Dechlorination Extract 2 Extract 1

Dechlorination

Kosson, 1993

Glass Product

Blending w l Process Additives

Dl lrnpingers

Melting

813 16.7 kg of glass forming additives were blended with 18.0 kg of the dechlorinated APC residue and fed to the melter. The melter used to demonstrate the vitrification process consisted of a crucible placed in an electrical heating furnace. Feed was added to the top of the crucible as often as necessary to maintain a layer of unmelted material on top of the molten glass. Off-gases were collected from the melter and passed through a series of liquid-filled impingers and an activated carbon trap to collect volatilised metals and acid gases. The bottom of the crucible contained a reclosable orifice to allow semicontinuous collection of molten glass. Molten glass flowing from the melter was collected as patties in preheated graphite molds and then annealed at 575~ The glass collected was olive green in colour. Mass balances summaries for the dechlorination and blending steps of the process from both demonstrations are presented as oxide and elemental constituents in Table 19.5. The sum of the feeds includes the untreated APC residue and the vendor's additives. Using the sum of feeds and the removed material, the expected mass and composition were calculated based on the sum of the process feeds and the analysed extract composition. All water retained in the dechlorinated APC residue was assumed to be evaporated during the drying step. The calculated mass and composition were then compared to the measured mass and analysed composition. The mass balance summaries for the melting step of the process are presented as oxide and elemental constituents in Table 19.6. The only feed stream was the batch mix as weighed and added during the melt. Material was removed from the system during the melting step through gaseous emissions. Dust collected from the funnel accounted for 2.19% (by weight) of the feed. Unmelted feed remaining in the crucible was assumed to have the same composition as the feed batch mix. The weight of the crucible with the remaining material minus the tare weight of the crucible was assumed to be vitrified material having the same composition as the glass product. An additional assumption was that the entire inventory of carbon and sulphur in the ash would be volatilised at the temperature of the melt. The expected mass and composition of the glass product were calculated by subtracting all removed material from the feed and compared with the measured mass of glass formed and the composition analysed. Relative error was calculated as the calculated glass quantity minus the measured glass quantity divided by the measured quantity. Retention of cadmium, lead and zinc was calculated to be greater than 90% of the vitrified feed, although the >100% cadmium retention in glass was a result of cumulative analytical error. It should be noted that 29% of the lead originally present in the APC residue was removed with the dechlorination extracts. Corning, Inc. has indicated that subsequent process trials included precipitation of lead from the dechlorination extracts followed by addition of the lead precipitate to the melter feed (Wexel, 1992). Similar retention of lead in the glass product was claimed in this case.

Table 19.5 Mass Balance Summary for the Dechlorination and Blending Steps of the Corning, Inc. Process for Vitrification of APC Residue Element

Sum of Feeds (g)

Dechlor Extract 1

Dechlor Extract 2

Sample

Batch Mix (g) calculated

Batch Mix (g) analysed

% recovered

Summary of elements

25,000

17.6 %

13.1 %

2.5 %

17,000

18,000

106 %

1. Number in boxes represent material removed from the system, expressed as percent of feed, due to the process indicated 2. Percent recovered is calculated as follows: % recovered = Analysed BatchICalculate Batch

2 P

Table 19.5 Continued Element Dechlorination Extracts

Al

S

<0.1 %

Emission Losses (2)

Vitrified Products (3)

% of Batch

% of APC

% of Batch

% of APC

0.03 %

0.03 %

89 %

94 %

19 %

20 %

19 % 0.13 % 0.13 % 1. Removal of APC residue bv aqueous extraction for dechlorination 2. Losses from volatilisation during melting process 3. Retention of initial material in the final vitrification product Kosson, 1993

5

Table 19.6 Mass Balance Summary for the Melting Step of the Corning, Inc. Process for Vitrification of APC Residue Element

Feed (S)

lmpinger

lmpinger

1 &2

3

Funnel Rinse

Transfer Line Rinse

Funnel Dust

Material Lefl in Crucible (9)

Glass Product Calculated (9)

Glass Products Analysed

Q,

% recovered

(9,

1.900 NA NA NA NA 0.03% 1,200 950 79 % 730 0.00' 1 0 0.00% 0.00' 1 0 0.00% 0.04% 1.4 1.2 86 % 2.3 0.92 BaO 14 0.00% 0.00% 0.00% 0.00% 0.03% 4.5 45 % 10 3.6 CaO 6100 2,600 NA NA NA NA 0.03% 3500 3,400 97 % CdO 5.7 4.5 196 % 3.3 0.57% 0.04% 0.06% 0.09% 0.03% 2.3 CrzO, 7.4 4.5 110 % 3.4 0.00% 0.01 % 0.00% 0.00% 0.04% 4.1 CuO 14 5.7 0.21 % 0.00% 0.04% 0.04% 0.19% 8.0 7.5 94 % Fe,03 330 NA NA NA NA 0.03% 200 160 80 % 120 KzO 240 NA NA NA NA 0.10% 140 108 % 110 130 MOO, 2.8 0.00% 0.00% 0.00% 0.03% 0.00% 1.9 1.2 1 .O 63 % Na20 2,700 NA NA NA NA 0.82% 2,000 860 43 % 690 NiO 1.5 0.00Oh 0.00% 0.00% 0.00% 0.04% 0.83 0.92 111 % 0.69 PbO 71 0.81 ' 1 0 0.04' 1 0 0.17' 1 0 0.17% 0.50' 1 0 41 38 93 % 29 SiO, 18,000 11,000 8,700 NA 0.04% 6,700 79 % NA NA NA 200 180 90 % 0.00% 0.00% 0.00% 130 TiO, 330 0.00% 0.03% ZnO 470 0.05% 0.00% 0.01 % 0.01% 0.07% 250 290 116 % 210 Ci 390 310 83 27 % 0.68% 0.08% NA 71 NA 0.07% so3 2,300 0.06% 0.00% NA NA 0.07% 250 0 it] 250 111 26 C [!I30 NA NA NA NA 0.02% 0 It] 1.2 111 Sum of Elements 33,000 12,000 2.37% 0.18% 0.30 % 0.31 % 2.19% 19,000 15,000 79 % Notes: 1. Numbers in light box represent material removed from the system, expressed as percent of feed, due to process indicated. 2. Numbers in boldface and heavy box represent unmelted or uncollected glass, expressed in grams 3. Percent recovered is calculated as follows: % recovered = Analytical Batch ICalculated Batch [t]

Assumptions: C --> CO (g) and S --> SO, (g)

Kosson, 1993

817 Table 19.7 presents leaching test results from testing the vitrified APC residue produced by Coming, Inc. Leaching results indicated substantial reductions in leaching potential for all of the elements tested based on the availability leach test. 19.6 FUSION There are several examples of fusion technology which have been demonstrated at different scales. The "VS reactor' has been suggested as a method to combust waste and thermally treat residues into a single process (Kunstler et al, 1992). In this configuration, grate ash from a mass burn facility passes directly into a rotary kiln. Grate siftings are recycled back into the waste feed to the primary combustion chamber just prior to the rotary kiln entrance. The flue gas is also partially recycled. Operating conditions in the rotary kiln are adjusted to provide either sintering or ash fusion prior to discharge. Initial testing was carried out at the KVA Basel facility, unfortunately no process mass balances or leaching test results are available. Pauli (1992) suggests the use of refuse derived fuel combustion with an integrated ash fusion section prior to discharge from the combustion chamber. Testing with fuels other than RDF have only been carried out at pilot scale. The "FosMelt" process is designed for fusion of bottom ash with fly ash (Eisbein et al, 1992). The intent of the process is to separate volatile metals and inorganic constituents from the residue during fusion. Fossil fuels supply the energy necessary for fusion. The "RedMelt" process is similar in process objectives and design, but uses submerged electric arc heating rather than fossil fuel burners to achieve melting (Schumacher et al, 1992). Several investigators have examined the use of plasma torches for fusion of bottom ash and fly ash (Jimbo, 1992; Klein, 1992). Pilot scale tests indicated that plasma torches may be used to carry out the fusion. The more volatile trace elements (Pb, Cd, Zn, etc.) and inorganic species (CI, SO4) are volatilised during the process and the trace elements remaining in the fused material are relatively non-leachable. Phase separation of the fused material into a fused heterogeneous semi-crystalline, semiamorphous material and a metal phase has been observed. MSW pyrolysis followed by combustion and residue fusion is carried out in the "Siemens Thermal Waste Recycling Plant." The process incorporates an initial pyrolysis of size reduced waste followed by high temperature combustion and fusion (approx. 1,300~ of the pyrolysis products (Baumg~rtel & Berwein, 1992). Metals and particles larger than 5mm are separated from the pyrolysis residues for recovery prior to introduction into the combustion chamber. ESP dust and boiler ash are recycled into the combustion chamber where they are fused with the other combustion residues. The fused ash is water quenched and forms a granular material. Pilot-scale (0.2 tph) testing of this process was carried out at the UIm-Wiblingen facility. The elemental composition and leaching test results on the fused residue are provided in Table 19.8.

Table 19.7 Leaching Test Results for Testing of the Vitrified APC Residue Produced by Corning, Inc. Availability Leach Test (release basis, mglkg) TCLP (mgll) Vitrified APC Residue (2) Vitrified APC Residue Untreated Untreated APC Demo Demo Detection Regulator APC Demo Demo Detection % 1 2 Limit Y Residue 1 2 Limit Reduction Residue' Limit 0.20 0.19 0.02 I0 12 4 99.6 Arsenic Barium Cadmium Calcium

I

Chromium Copper Iron Lead Molybdenum Nickel Potassium Silicon Sodium

/ 1 /

BDL

BDL

0.06

5.0

BDL

BDL

12

1.60

0.23

0.21

0.001

100

24

24

0.2

73.8

BDL

BDL

BDL

0.01

1.O

BDL

BDL

2

>98.5

4,033

5.30

4.72

0.005

198,000

1,084

842

1

99.5

6.4

BDL

BDL

2.0

>96.9

197

BDL

BDL

2.0

>99.0

BDL

BDL

2.0

>93.4

BDL BDL

BDL BDL

20

-

>98.0

BDL

BDL

2

~77.3

14,000

600

460

0.2

96.2

0.22

BDL

BDL

0.001

0.07

0.03

0.01

0.01

0.06

0.17

0.17

0.01

32.6

BDL

BDL

0.1

0.28

BDL

BDL

0.02

BDL

BDL

BDL

0.005

712

1.6

1.50

0.01

0.29

1.3

1.50

0.01

5.0

5.0

I

1,010 11.1

Titanium

BDL

0.03

0.03

0.002

4.1

BDL

BDL

0.4

790.2

Zinc

4.33

0.73

0.78

0.006

7,910

24

82

1.21

99.3

5.9

-

PH

1

11.9 5.9 (1) Extraction Fluid 2 Wexel. 1993

I-

(2) Extraction Fluid 1

2 CQ

819 Relatively low total concentrations of trace metals (e.g. lead, cadmium) in the fused material compared to typical bottom ash values suggest that volatilisation of these elements may have occurred during fusion. Table 19.8 Composition and Leaching Results for Granular Fused Ash from the UIm-Wiblingen .Facility . . . . Element or Total composition 1Leachate 1Leachate Parameter (wt%) concentration by concentrationby TVA DEV-S4 (mg/I) (mg/I) pH 8.9 9.54 AI 8.3 0.64 0.44 Pb 0.080 <0.05 <0.05 Cd <0.0004 <0.002 <0.005 Cr 0.098 0.031 0.01 Fe 3.75 <0.63 2 Cu 0.19 0.08 0.14 Zn 0.2 <0.038 0.53 CI <0.08 1.3 0.5 1 Fused residue was size reduced to less than 200 IJm prior to testing BaumgQrtel and Berwein, 1992

19.6.1 Fusion of Filter Ash by ABB Deglor Process The DEGLOR process has been developed by ABB (Baden-DQttwil, Switzerland) as a thermal process for treatment of ESP ash from waste incinerators (Hirth et al, 1990; Plumley & Boley, 1990). This glassification or fusion process splits the ash into two fractions, a major glass fraction, and a small fraction in which trace metals are condensed. The flue gases generated during the process are fed back into the incinerator. Laboratory furnaces with capacities of 1 and 3 kg/hour were utilised during the development of the program. A demonstration installation with a capacity of 100 kg/hour was operated processing an ESP ash slip stream at the Hinwil municipal waste incinerator near ZQrich, Switzerland. A process schematic diagram and flow diagram are presented in Figure 19.5 (St~mpfli, 1992). The objective of the process is to separate specific metals through volatilisation and subsequent condensation to facilitate metals recycling. A black and green marbled, amorphous fused aggregate was the principal product. No additives were used in the melting process and the melting temperature was approximately 1350 ~C. Composition of the ESP dust feed and process output streams are provided in Table 19.9, based on multiple analyses of samples obtained under nominal steady-state process conditions. Cd, Pb, Zn, CI and S were substantially concentrated in the pre-

820 Figure 19.5 Process Flow Diagram (a) and Schematic Diagram for Mass Balances (b) for the DEGLOR Pilot Plant used to Fuse ESP Ash at the Hinwil MSWI, Switzerland

821 Table 19.9 ESP Dust Feed and Process Output Stream Compositions for the DEGLOR Plant Hinwil, Switzerland Element

D

R

P

C

G- F

Carbon

C

1.6

n.d.

n.d.

n.d.

2700

Chlorine

CI

1.2

0.27

7.9

10

220

Sulphur

S

1.1

0.090

2.4

1.0

1200

Zinc

Zn

0.33

0.13

0.62

0.88

0.28

F

0.18

0.18

<0.01

<0.005

2.3

Fluorine Lead

Pb

0.030

0.0016

0.32

0.44

0.15

Copper

Cu

0.013

0.0020

0.036

0.040

0.015

Cadmium

Cd

0.0022

0.000056

0.032

0.033

0.0066

Mercury

Hg

5.5E-05

<5E-07

1.5E-06

3.7E-05

0.065

Silicon

Si

5.0

6.4

0.0025

0.0050

n.m

Calcium

Ca

3.7

4.2

0.0050

0.012

n.m

Aluminum

AI

3.3

3.9

0.0015

0.0011

n.m

Sodium

Na

1.0

0.74

3.7

3.3

n.m

Potassium

K

0.82

0.54

6.6

6.4

n.m

Magnesium

Mg

0.74

0.86

0.00082

0.0033

n.m

Iron

Fe

0.34

0.27

0.0068

0.029

n.m

Titanium

Ti

0.25

0.29

n.m.

n.m.

n.m

Phosphorus

P

0.18

0.20

n.m.

n.m.

n.m

Boron

B

0.023

0.024

<0.005

<0.005

n.m

Manganese

Mn

0.018

0.022

0.0011

0.0016

n.m

Tin

Sn

0.016

0.0025

0.0011

0.077

n.m

Chromium

Cr

0.010

0.015

<0.001

<0.001

n.m

Barium

Ba

0.0095

0.017

<0.0004

<0.0004

n.m

Antimony

Sb

0.0025

0.0039

0.0012

0.023

n.m

Nickel

Ni

0.0024

<0.0009

<0.0009

<0.0009

n.m

Molybdenum

Mo

0.0017

0.0027

0.0024

0.0029

n.m

V

<0.001

<0.001

<0.001

<0.001

n.m

Vanadium

n.m. = not measured n.d. = not detectable (<0.02 mol/kg) Note: Element concentrations of the materials - mol/kg for D,R,P and C; mmol/m 3 for G-F St~mpfli and Baccini, 1992

822 condensate stream, whereas concentrations of Pb and Cd in the fused residue were one and two orders of magnitude less than in the E SP dust. A process mass balance on a unit feed basis is provided in Table 19.10. Closure of the mass balance was within the analytical error for all elements presented, with the exception of Cu and Zn, which appeared to accumulate within the melter. Table 19.11 presents the results of the Swiss TVA leaching tests on the fused residues, which were significantly below the Swiss regulatory requirements (Plumley & Boley, 1990). Table 19.10 Process Mass Balance for DEGLOR Pilot Plant, Hinwil, Switzerland (unit feed basis)

Mass flux trans, coef.

D (input)

R

P

C

G-F

Output

47

33

0.26

2.6

9.6 (28)*

46

0.70

0.0056

0.056

0.20

0.97 .

.

.

.

.

.

Elemental Transfer Coefficients - Average of Four Intervals Element

R

P

C

G-F

Output*.*

Carbon

C

sm.

sm.

sm.

0.99

0.99

Chlorine

CI

0.16

0.038

0.50

0.11

0.81

Sulphur

S

0.060

0.012

0.054

0.66

0.78

Zinc

Zn

0.26

0.010

0.15

0.0005

0.42

Fluodne

F

0.72

sm.

sm.

0.040

0.76

Lead

P

0.039

0.060

0.83

0.003

0.93

Copper

Cu

0.11

0.016

0.18

0.0007

0.30

Cadmium

Cd

0.017

0.080

0.82

0.002

0.92

Mercury *

Hg 0.70 sm. 0.0002 0.038 0.67 in parentheses the off-gas flux in m3/h (standardised to O~ 1 atm and dry); typically F= 120 m3/h ** equal to 1 with complete transfer; sm. = smaller than 0.0001 St~mpfli and Baccini, 1992

19.6.2 Japanese Fusion Processes Kubota Corp. has seven full-scale melting furnaces in operation and three under construction as of 1992 (Fujimoto et al, 1991, 1992). Three of the facilities are for melting incinerator residues while the others are for sewage sludge or other waste types. A schematic diagram of a typical installation is provided in Figure 19.6. Melting, typically of combined bottom ash and fly ash, is accompanied in an annular melting

823 Table 19.11 Swiss TVA Leaching Test Results on Fused ESP Dust from the DEGLOR Process Leachate

Limit Established by Swiss TVA

Aluminum

mg/I

1.0

1

Arsenic

mg/I

<0.005

0.01

Barium

mg/I

<0.1

0.5

Cadmium

mg/I

<0.005

0.01

Chrome III

mg/I

<0.005

0.05

Chrome VI

mg/I

<0.005

0.01

Cobalt

mg/I

<0.01

0.005

Copper

mg/I

<0.01

0.2

Lead

mg/I

<0.05

0.1

Mercury

mg/I

<0.001

0.005

Nickel

mg/I

<0.01

0.2

Silver

mg/I

<0.005

0.01

Tin

mg/I

<0.1

0.2

Zinc

mg/I

0.2

0.5

Ammonium

mg/I

<0.05

0.5

pH Value

mg/I

6.0

6-11

Fluoride

mg/I

<0.1

1

Nitrate

mg/I

<0.01

0.1

Sulphite

mg/I

<0.05

0.1

Phosphates

mg/I

<0.01

1

DOC

mg/I

0.95

20

3.9

5-10

AOX u/I Plumbley and Boley, 1990

824 Figure 19.6 Schematic Diagram of Kubota Melting Furnace (a) and Fusion Facility Design (b)

(a)

E~ ~lJ

Outer Cylinder

Inner

!r

Cylinder

Combustion

\

condary " "'" " .... """ " Combustion Chamber

Slag Pit

(b)

Melting

Stack

Heat Exchanger

Flue Gas Treatment

r162 n

.."

--I

SlagBunker Fujimoto et al., 1991

~~.

~ ----- ~-~1

i

825 furnace. The target alkalinity for Kubota melter slags is 1.0, as calculated by dividing the percent weight (CaO + MgO + Fe203) by percent weight (SiO2 + AI203), however, published average slag compositions from Kubota incinerators, are less basic. The melting point of the slag is 1,200 - 1,250~ and therefore the temperature in the combustion chamber is maintained at 1,300 - 1,400~ Slag volume is typically about 45% of that of the ash feedstock volume. Approximate percentages of Cd, CI, Pb, S, and Na20 in the feedstock which are retained in the slag are 5, 5, 15, 30 and 70% respectively (Fujimoto et al, 1989). Each plant generates a slag by-product (the largest plant has 5,850 tonnes/year slag output) that is approved by some of the municipal governments for use as substitute for landfill sand, pavement material for light-loaded roads, playgrounds and tennis courts, and bedding for pipes. Typical slag compositions appear in Table 19.1. Output from Kubota's Oyabe plant (1350 tonnes per year slag) is converted into decorative paving block. Bottom ash and ESP ash from a mass burn facility in Numazu, Japan have been melted together with waste plastic in a demonstration plant annular film melting furnace (Kubota) which has been in operation since 1979 (Fujimoto et al, 1991). The melting furnace capacity is 18 tpd and operates at a melting temperature of 1300 to 1400 ~C. Ferrous metal must be separated from the bottom ash prior to melting. Inclusion of crushed waste plastic (35% by weight) as fuel added to the incinerator ashes replaces the need for up to 200 litres of kerosene per tonne of ash melted. Fused ash is water quenched resulting in a residue weight and volume reduction of approximately 6 and 50 percent, respectively. Off gases from the melter are passed through an ESP. The resulting ESP ash is managed separately. The fused ash and ESP ash represent 73 and 16 percent, respectively, of the total mass of ash fed to the melter. Information on the composition and percent of mass for each element for each process stream is presented in Table 19.12. A 48 hour trial of melting only fly ash was carried out to evaluate operating condition and process mass and energy balances. Approximately 28 tonnes of fly ash was melted during the trial at a temperature of 1,330~ Utility consumptions reported were 260 litres kerosene, 235 kWh and 2 m3 water per tonne of fly ash melted. Mass and composition was determined for the fused ash, particulate remaining in the transfer ducts (to APC system) and ESP dust. The fused ash, ESP ash, transfer duct particulate, and flue gas represented 88.5, 8.7, 3.8 and 0.025 percent, respectively, of the total mass of fly ash fed to the melter. Information on the composition and percent of mass for each element for each process stream is presented in Table 19.13. The more volatile elements (Na, Cd, Pb, Zn) were volatilised from the melter and concentrated in the ESP ash. The process also resulted in greater than 99.98% destruction of PCDD/PCDF toxicity equivalents.

826

Table 19.12 Composition and Relative Masses in Feed and Output Stream during Fusion of Bottom Ash and ESP Ash fro m an MSW Mass Burn Facility by the Kubota Process Element

BottomAsh and ESP ash

Fused ash

ESP ash (From melter)

relative mass (%)

conc. (mg/kg)

relativemass (%)

relativemass (%)

S

100

2,900

-

-

CI

100

1,400

10

90

SiO2

100

330,000

100

0

CaO

100

260,000

90

10

AI203

100

115,000

90

10

Na

100

36,500

50

50

Fe

100

88,300

100

0

Cd

100

1.3

0

100

Pb

100

450

10

90

Zn

100

4,500

30

70

450

100

0

Cr 100 Fujimoto et al., 1991

Takuma Co., also has some experience with commercial scale fusion processes. The first facility began operation in 1981, and there are now four operating in Japan, with another under construction (Wakamura and Nakazato, 1994). These facilities only process bottom ash, however, some modifications are being developed to process combined bottom ash and APC residue. These four operating units are of the type known as "stationary single-slope surface-melting furnace." The bottom ash first undergoes a magnetic separation to remove ferrous material before it is charged into the furnace. In the furnace, a kerosene or fuel oil burner heats the ash and melts it from the top surface. The furnace temperature in the free standing units is approximately 1,300~ and their sizes range from 4.8 to 15 tonnes per day. After twostage magnetic separation, the molten bottom ash will consist of 92% slag (fused ash), 4% gas and 4% fly ash. The volume of ash is generally reduced by 50%. The 14.4 tpd unit built in 1985 is directly connected to the MSW furnace, and is designed to melt metals in the bottom ash. For this purpose, it is equipped with an oxygen generator (Pressure Swing Absorption) which produces 95% 02. The oxygen is added to the combustion air to raise its oxygen levels to 28%, which increases the furnace temperature to approximately 1,400 ~ A new 15 tpd facility currently under

Table 19.13 Composition and Relative Mass in Feed and Output Streams during Fusion of Fly Ash by the Kubota Process Element conc. (mglkg)

'relative mass (%)

Duct Particulate

Fused Ash

Fly Ash conc. (mglkg)

arelative mass (%)

conc. (mglkg)

arelahe mass (%)

conc. (mglkg)

arelative mass (%)

18,000

26,000

< I,000

ESP ash

C

59,000

S

5,600

100

3,300

46

33,000

20

18,000

24

CI

103,000

100

3,200

2.5

305,000

10

405,000

29

Si

113,000

100

142,000

100

1,000

0.03

0.00

Ca

137,000

100

192,000

110

13,000

0.33

2,000

0.11

Al

54,000

100

75,000

110

1,000

0.06

el00

0.00

Na

21,000

b87

7,000

21

202,000

29

178,000

55

Fe

15,000

100

21,000

110

2,200

0.09

3,800

0.05

Cd

46

100

3.0

5.2

270

20

420

68

Pb

1,600

100

110

5.4

8,000

17

15,000

70

Zn

4,900

100

1,200

19

14,000

9.7

25,000

38

Cu

440

100

1,300

230

2,200

17

2,500

42

Cr a b

42 0.78 35 0.3 1,000 200 400 100 = Relative mass (%) is equal to the percentage of the total mass of the element fed to the melter. = 13% of the sodium fed to the melter was in the form of NaOH used as a process additive

Fujimoto et al., 1992

828 construction will be equipped with opposed dual melting slopes for larger capacity and better melting efficiency. Figure 19.7 shows its structure and process flow. This design is based on a 5 tpd pilot plant which was used to study the melting characteristics of bottom ash only, combined bottom and fly ash, and bottom ash with crushed incombustibles. These incombustibles are the materials remaining from a bulky-waste crushing plant, where combustibles, ferrous and aluminum metals were removed from the waste stream. Certain portions of the material are combustibles (Nishigaki & Shibata, 1994). Table 19.14 provides some operating data. Table 19.14 Operating Data from Takuma Surface Melting Plant Unit Test Material

Test Material Treated Quantity

Gas-Air Temp.

Others

kg/h

Run A

Run B

B o t t o m BottomAsh + Ash incombustibles (10:6) 252 354

Run C Bottom Ash + Fly Ash (10:6) 248

Furnace Temperature

~

1432

1394

1385

Air Preheater Entrance

~

699

732

726

Air Preheater Exit

~

533

572

534

Bag Filter Entrance

~C

166

162

173

Combustion Air

~

378

389

385

Kerosene Used

I/h

71

55

81

Dust Produced

kg/h

1.5

2.0

8.3

Slag Produced

kg/h

222

247

214

Slag/Ash Ratio

I/tonne

282

155

327

When 6 parts fly ash were mixed with 10 parts bottom ash, the alkalinity of the molten slag increased to 0.6 from 0.3 for bottom ash alone. The kerosene consumption per tonne of ash also increased by 15%, but the slag production rate decreased from 94.5% to 86.4% due to the higher concentration of volatile matter in the fly ash. When 6 parts of "incombustibles" were mixed with 10 parts of bottom ash, the remnant combustibles within the crushed "incombustibles" worked as an auxiliary fuel, and the kerosene consumption went down by 45%. Conversely, although the high concentration of soil, sand and glass in the crushed incombustibles reduced the alkalinity to 0.25, the slag production rate also decreased to 77.4% as the volatile matter was driven off in the exhaust gas.

Figure 19.7 Flow Diagram of Takuma's Surface Melting Process

-

<

Ash from lnc~nerator

-

r Q ?

.

Ash Hoppers

-

7

Bag Filter

ExhaustGas Condenser

y-!=@Fl-

gs

W'

--

w

Melting Furnace

Fly Ash to Treatment

Induced Draft Fan

Stack

Slag to Pit Air Preheater

Slag Conveyor -

Forced Draft Fan

830 Table 19.15 shows the leaching test results of slag and bag filter dust from the test plant. There was no leaching of metals from the slag, but there was some leaching of Cd and Pb from the bag filter dust. Table 19.15 Leaching Test Results of Slag and Bag Filter Dust from Takuma Surface Melting Furnace Element

Unit

Run A

Run B

Run C

Slag

Fly Ash

Slag

Fly Ash

Slag

Fly Ash

Cd

mg/I

<0.01

17

<0.01

26

<0.01

39

Pb

mg/I

<0.05

16

<0.05

12

<0.05

170

As

mg/I

<0.02

0.02

<0.02

<0.02

<0.02

<0.02

T-Hg

mg/I

<0.0005 <0.0005 <0.0005

<0.0005 <0.0005

0.27

Cr

mg/I

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

19.7 SINTERING

Wainwright and Robery (1991) have reported on a pilot-scale process for the sintering of residues obtained from incinerator facilities in Rotterdam, The Netherlands and several facilities in the United Kingdom. The process involved removal of the ferrous and nonferrous metals followed by crushing the material to pass a 300 mm sieve and blending the material with clay. The pelletised material was then fired in a rotary kiln to produce a spherical aggregate. Tests were performed to obtain the optimum firing time, temperature and the percentage clay to give the strongest aggregate as determined by a splitting tensile strength test. A 15/85% blend of clay/incinerator residue fired at 975~ for 13 89 minutes was reported to be the best combination. The major oxide composition of the treated residues (before blending and firing) and of the aggregate, and some of the physical properties of the aggregate are shown in Table 19.16. It was indicated that sintering at 850~ for 3 minutes may not result in complete conversion of aluminum to alumina. Aggregate produced under these conditions resulted in hydrogen evolution during use in concrete, whereas aggregates produced at higher temperatures and longer sintering intervals showed no detrimental interferences when used in concrete. Sintering and fusion pilot tests have also been carried out in an oil fired rocking kiln with bottom ash (Ruegg, 1992). Kiln temperature, solids residence time and kiln atmosphere (oxidising or reducing) were the experimental variables. Table 19.17 presents treated ash (sintered or fused) compositions as a function of trial conditions. Other results indicated that the kiln atmosphere has a negligible effect of fused ash composition.

831 Table 19.16 Major Oxide Composition and Physical Properties of Pretreated Residues (Before Blending and Sinterincj) and the Resulting Acjgre(:jates Oxide Composition (%) %

Edmonton Residue Aggregate 50.23 54.14

SiO2

Rotterdam Residue Aggregate 57.50 63.42

Fe203

15.18

10.20

10.17

8.01

CaO

9.77

12.88

9.43

10.32

AI203

6.75

8.90

6.55

7.36

Na20

5.67

5.04

5.32

4.26

SO3

1.14

1.52

0.88

0.76

K20

1.04

1.47

1.19

1.56

Physical Properties

Bulk Density (kg/m3)

Edmonton Residue Aggregate 909

Loose Compacted

Rotterdam Aggregate Residue 1059

955

1121

Relative Density

2.18

2.29

Water adsorption (% SSD)

15.10

10.11

38.04

24.96

Porosity (%) Wainwright, 1991

Table 19.17 Effects of Thermal Treatment on Bottom Ash Composition Sintering with Element (mg/kg) U n t r e a t e d Sinteringwith reducing or test condition Bottom oxidising atmospheres Ash atmospheres 900 Temp (~ 950 Residence Time (rain)

Fused Residue >1200

30

32

22.5 1.9

Cd

26.4

8.0

4.2

Pb

2100

1700

1667

1200

Zn

3700

4600

3767

5200

Cr

680

1633

1467

1800

CI

640

C Ruegg, 1992

1.4

520 0.4

0.3

0.2

832

19.7.1 Integrated RDF Combustion with Sintered Aggregate Production (Neutralysis Process) The Neutralysis process involves firing a pelletised mix of refuse derived fuel, liquid waste and clay in a rotating kiln to produce a lightweight aggregate for use in the construction industry (Krol et al, 1991). In August 1988, a pilot plant capable of processing 25 tpd of MSW was completed in Brisbane, Australia. This facility has been used to demonstrate the process, and to provide aggregate for a testing program designed to assess its environmental acceptability and marketability. The major unit operations and process streams along with a solid's mass balance for a "typical" Neutralysis plant rated at 500 ton (RDF)/day are shown schematically in Figure 19.8. There are three main sections, including physical processing, kiln firing and gas cleaning. The actual mass balance will depend on factors such as the ultimate and proximate analyses of the MSW, and the desired RDF to clay ratio. The MSW is pulverised prior to removing the ferrous/nonferrous metals and some glass, to produce a "fluff" RDF. This RDF is mixed with a similar weight of milled clay and liquid waste, and the mixture is extruded to produce pellets of a size suitable for the required aggregate application. The rate of nonhazardous liquid waste (and/or water) addition is controlled to produce cohesive pellets. The pellets are passed through a rotary drum tumbler which, in conjunction with a trommel screen, removes fines which are recycled back to the mixer. A dryer using hot air from the product aggregate cooler is used to reduce the moisture content of the pellets. The partially dried pellets are passed through a series of rotary kilns. In the pyrolysis kiln, the pellets are heated to approximately 500~ under starved air conditions, producing porous pellets containing unburned carbon. In the oxidation kilns, air is added to burn off the carbon in the pores and raise the pellet temperature to approximately 900~ The porous pellets then pass into a sintering kiln where the temperature is raised to approximately 1100 ~ The off-gases from all three kilns are passed through an afterburner prior to entering the APC system. Aggregate discharge from the sintering kiln is air-cooled, providing heated air for the pellet dryer and combustion. The resulting aggregate is typically a matrix of SiO2 and AI203 with lower concentrations of other metal oxides such as Fe203, CaO and Na20. The total concentration of uncombusted carbon (as measured by the loss on ignition at 550~ is less than 1%. In addition, APC residues, including fly ash and spent/excess lime are produced which require further treatment prior to disposal. Table 19.18 presents total heavy metal concentrations and leaching test results for Neutralysis aggregate produced during a trial run and the range of values measured for other runs. Also shown are the concentrations of Cd, Cr, Pb and Hg in aggregate produced before and after the addition of metals into the kiln feed during one pilot plant run. The metals were added as an acetate salt solution and were equivalent to

Figure 19.8 Major Unit Operations and Process Streams Along with a Solids Mass Balance for a Typical Neutralysis Plant

Stack Gases

4 Electricity Steam

:

I I

I

I

I I I I I I

Residues to Secure Landfill Liquid Waste1 Water

:

I

I

ID Fan

I I I

I I I I

1

I I

I I I

I

Grit Recycle

v

I

v

I I I I

I

Pulverised MSW I I I I I I I

Air

i

Recycle or Dispose Steel Aluminum Non-Ferrous Metals Glass White Goods Tires

I

Lightweight Aggregate

Table 19.18 Compositional Analysis and Leaching Test Results for the Neutralysis Aggregate Element Total Concentration in Aggregate Leaching Test Results USEP TCLP Water Nitric Acid' Audit Range in Spiking Spiking (audit (audit run run (after A run other TCLP run) run) (before addition) runs* criteria addition) (mglkg) (mall) (mall) (mgll) (mdl) (mdkg) (malka) (malka) 5.0 0.015 0.011 ~0.1 Arsenic
MEP2 ( Ist extract)

MEP2 (2nd extract)

(mall) 0.89

(mall) 0.09

Barium Cadmium Chromium Copper Lead Mercury

none

1.29

<0.01

0.83-5.4

nm

nm

5.0

CO.01

CO.01

<0.01-0.23

0.32

0.16

0.2

<0.001

<0.001

<0.0002

<0.0005

~0.0005

nm

nm

Nickel Selenium Silver Zinc

3590 nm nm nm none 0.32 0.03 0.39-0.58 nm = not measured = Range of results obtained from 17 samples, various laboratories 1 = Range of results obtained from 2 runs, uncrushed aggregate extracted for 1 hour 2 = Results from single run Krol et al., 1991

835 approximately 120% more cadmium than in the MSW, 45% more chromium, 10% more mercury and 300% more lead. The results indicated that there are significant variations in concentrations of less volatile metals in the aggregate, which is, at least in part, a reflection of the variation in metal content of the MSW feed. The aggregate concentrations of Cd and Hg did not show a wide variation because these metals were almost totally volatilised to the kiln off gasses during processing. However, there were no mass balances reported. Several leaching procedures were carried out on Neutralysis lightweight aggregate: ~

2. 3. 4.

Toxicity characteristic leaching procedure (TCLP) A distilled water extraction test (ASTM D3987) A 10% (v/v) nitric acid extract at 20:1 liquid:solid ratio The US EPA Multiple Extraction Procedure (MEP, EPA Method 1320). This test involves initial use of the US EPA Extraction Procedure (EP, EPA Method 1310) followed by 9 subsequent extractions of the same solid using synthetic acid rain

The data presented (Table 19.18) indicate that in all cases the TCLP extracts were substantially below the U.S. EPA criteria. This also was the case when water, nitric acid and the EPA leaching media were used. Subsequent extractions in the MEP procedure continued to demonstrate a decline in leachate concentrations to below analytical detection limits. The aggregate produced during the trace metal spiking run also produced leachate concentrations of Cd, Cr, Pb and Hg which were below detection limits. Aggregate produced in the trial run was also subjected to organic analysis. The concentrations of indicator polynuclear aromatic hydrocarbons (PAHs) were all below the analytical detection limit (<0.01 ng/g), as were those of chlorinated benzenes (<0.1 ng/g), chlorinated phenols (<0.06 ng/g), polychlorinated biphenyls (PCBs) (<0.01 ng/g), 2,3,7,8 tetrachloro dibenzo-p-dioxins (<0.06 ng/g) and 2,3,7,8 tetrachloro dibenzofurans (<0.06 ng/g). The aggregate produced was reported to be less dense than conventional aggregate. Physical properties of a Portland cement concrete made from Neutralysis aggregate included compressive strengths up to 53 MPa at 28 days and cured concrete densities between 1,800-2,080 kg/m 3 (Krol et al, 1991).

836 REFERENCES

Baumg~rtel, G. and B. Berwein. "The Siemens Thermal Waste Recycling Plant, an Almost Residue Free Process to Convert Refuse into Recyclable Products", Waste Management International, vol 2, ed. K.J. Thom~-Kozmiensky, 1992. Champan, C.C. "Evaluation of Vitrifying Municipal Incinerator Ash", Nuclear Waste Manageme.nt, IV, 1989. Edwards, G.H. "Turning Furnace Waste Into Glass Raw Material", J. of AWMA, 1994. Eisbein, J., C. Jatzwauk, F. Lichtmann and W. Schumacher. "The LCS FosMelt Process for the Combined Fusion of Slag and Fly Ash", Waste Mana.qement International, vol 2, ed. K.J. Thom~-Kozmiensky, 1992. Eschenbach, R.C., R.A. Hill and J.W. Sears. "Process Description and Initial Test Results with the Plasma Centrifugal Reactor', Forum on Innovative Waste Treatment Technologies: Domestic and International, USEPA-540/2-89-056, pp. 263-279, 1989. Fujimoto, V.T., K . C . Shin and M. Shioyama, "Aufbereitung von Verbrennungsn3ckst,~nden mit dem Hochtemperaturschmelzverfahren", M011und Abf,~ll, vol 2, p. 64-70, Feb. 1989. Fujimoto, T., S. Abe, T. Kimura, F. Kanbayashi, K. Kawamoto and T, Ishimi. "Melting Treatment of Fly Ash from MSW Incinerators", Waste Mana.qement Internati0na.[, vol 2, ed. K.J. Thom~-Kozmiensky, 1992. Fujimoto, T., E. Tanaka and K. Taniguchi. "Melting Treatment for Incinerated Residue of Municipal Waste", Kubota Corp. doc., 1991. Hirth, M., J. Jochum, M. Jodeit and C. Weickert. DEGLOR -"A Thermal Process for the Detoxification of Filter Ash from Waste Incinerators", Asea Brown Boveri Corporate Research, Ch-5405, Baden Switzerland. Jantzen C.M., "Prediction of Glass Durability as a Function of Environmental Conditions", Materials Research Society, Symposium Proceedings, Vol. 125, 1988. Jimbo, H., S. Takenaka, T. Amemya, T. Kawase and T, Kikuchi. "Plasma Torch Melting Process for MSW Incinerator Residue", Waste Mana.qement International, vol 2, ed. K.J. Thom~-Kozmiensky, 1992. Klein, H., A. Hoffman and K. Tscheschlok. "Plasma Treatment of Toxic Residues", Wa.ste Mana,qement International, vol 2, ed. K.J. Thom~-Kozmiensky, 1992.

837 Kosson, D.S., B. Stuart and T.T. Kosson. "Vitrification of Municipal Waste Combustion Air Pollution Control Residues", Phase I, Corning Inc., 1992 DRAFT. Kosson, D.S. and B. Stuart. "Vitrification of Municipal Waste Combustion Air Pollution Control Residues - Phase I", Corning, Inc., Project Officer: Carlton Wiles, USEPA/RREL. DRAFT, 1993. Krol, A., K. White and B. Hodgson. "Production of Lightweight Aggregates from Wastes: The Neutralysis Process", Waste Materials in Construction, ed. J.J.J.R. Goumans, H.A. van der Sloot and Th.G. Aalbers, Elsevier Science Publishers, p. 459-466, 1991. Kunstler, H.J., Ch. Klukowsji and V. Grotefeld. "Combusting, Sintering and Melting in a VS-Reactor", Waste Management International, vol 2, ed. K.J. Thom~-Kozmiensky, 1992. McLellan G.W. and E.B. Shand, .Glass EnQineerinq Handbook, 3rd edition, McGraw-Hill Book Company, New York, NY, 1984. Mendel J., "A Review of Leaching Test Methods and the Leachability of Various Solid Media Containing Radioactive Wastes, PNWL-1765, Pacific NorthWest Laboratory, Richland, Washington, 1973. Nishigaki, M. and K. Shibata. "Mixed Melting of Fly Ash and Incombustibles by SurfaceMelting Furnace", The 5th Conference of the Japan Society of Waste Management Scientists, 1994. Pauli, B. "Waste Combustion on a Grate with Downstream Melting Section", ARS process, Waste Mana.(:iement International, vol 2, ed. K.J. Thom~-Kozmiensky, 1992. Plodinec, M.J., G.G. Wicks and N.E. Bibler, "An Assessment of Savannah River Borosilicate Glass in the Repository Environment, DP-1629, Savannah River Laboratory, Aiken, SC, 1982. Plumley, A.L. and G.L. Boley. "ABB Ash Treatment Technologies". Presented at ASHIII, 3rd International Conference on Ash Utilisation and Stabilisaton, Arlington, Virginia, November 13-14, 1990. Queneau, P., D.E. Cregar and L.D. May. "Application of Slag Technology to Recycling of Solid Wastes". REVIEWER COPY- 3rd DRAFT 12/27/1990. Rabiger, W., G. Meyer, G. Schetter, W. Scheffler, O. Frielingsdorf and R. Dudill. "Adaptation of Knowledge from Glass Production to Vitrification of Residues from Waste Incinerator Plants", ..Waste Manaqement. Internationa.l., vol 2, ed. K.J. Thom6Kozmiensky, 1992.

838 Rankin W.D and G.G. Wicks, "Chemical Durability of SRP Waste Glass as a Function of Waste Loading", J. Am. Cer. Soc., Vol 66, p. 417, 1983 Richards, R.S and J.W. Lacksonen. "Stir-Melter Vitrification of Simulated Radioactive Waste, Fiber Glass Scrap, and Municipal Waste Combustor Fly Ash", Nuclear Waste Mana.Qement IV, pp. 309, 313-314. Roos, C.E. and R. Quarles. "Process and Apparatus for Reducing Heavy Metal Toxicity in Fly Ash from Solid Waste Incineration", European patent application no. 91400533.5, 1991. Ruegg, H. "Municipal Refuse Incineration with Integrated Slag Treatment", Waste Mana.qement International, vol 2, ed. K.J. Thom~-Kozmiensky, 1992. Sakai, S. and M. Hiraoka. "Ash Management in Japan: the Properties of Ash From Municipal Waste Incineration and its Future Management Strategies", paper prepared for IAWG, March 1994. Schumacher, W., M. Bette and J.A. Gugat. "Integration of Fusion Plants in Refuse Incineration Systems", Waste Mana.qement International, vol 2, ed. K.J. Thom~Kozmiensky, 1992. St~mpfli, D. and P. Baccini. "Thermal Treatment of ESP Dust from MSW Incinerators: Material flux analysis and process evaluation", submitted to Environmental Science and Technology, 1992. Tooley, F.V. The Handbook of Glass Manufacture, Vol. 1, Third Edition. Ashlee Publishing Co., 1984. Umweltbundesamt, FGIII 1.3-G. Hoffman, Bismarkplatz 1, 1000 Berlin 33, October 1992 USEPA, "Vitrification Technologies for Treatment of Hazardous and Radioactive Waste", EPN6251R-921002,May 1992. Wainwright, P.J. and P. Robery. "Production and Properties of Sintered Incinerator Residues as Aggregate for Concrete" from Waste Materials in Constructi..on, ed. J.J.J.R. Goumans, H.A. van der Sloot and Th.G. Aalbers, Elsevier Science Publishers, p. 425426, 1991. Wakamura, Y. and K. Nakazato. "Recent Trend of Ash Management from MSW Incineration Facilities in Japan", National Waste Conference Proceedings, ASME 1994. Wicks G.G., W.D. Rankin and S.L. Gore, "International Waste Glass Studycomposition and Leachability Correlations", in Scientific Basis for Nuclear Waste .Management VII, (C.M. Jantzen, J.A. Stone and R.C. Ewings, eds.), 44, p. 171, 1985.

839 Wicks, G.G. "Structure of Glasses", in Encyclopedia of Materials Science and En.qineerin,q, 3, p. 2020, Pergamon Press, Oxford, 1986. Westinghouse Environmental Services, "1989 Test Report for the Electric Pyrolyzer Demonstration on Combined Ash Waste", U.S. Department of Energy, Bio-fuels and Municipal Solid Waste Division, Contract # DE-FC01-85CE30839, April 1990. Wexell, D.R. "Vitrification of MSW Flyash for Heavy Metal Stabilisation". 4th Annual AACS Div. of Industrial Engineering Chemistry Symposium on Emerging Technologies for Hazardous Waste Management, Atlanta, GA, September 1992.

This Page Intentionally Left Blank

841

C H A P T E R 20 - LEACHING OF P R O D U C T S

20.1 INTRODUCTION MSW incinerator residues are widely being considered as an aggregate substitute in a variety of construction applications. These construction applications result in the incorporation of incinerator residues in products such as asphalt pavement and cement structures, or direct use as a compacted granular aggregate after varying degrees of treatment or component separation. The evaluation of potential environmental impacts from the use of these products focuses on release of contaminants via leaching. As previously described in Chapter 12-15, there are different mechanisms which dominate leaching behaviour of granular versus monolithic materials (Figure 20.1). This chapter focuses primarily on the mechanisms of dissolution, surface wash off and diffusion through the product to the surface where the water-product interface exists. The information presented in this Chapter will serve as the framework for utilisation criteria discussed in Chapter 21. Figure 20.1 Comparison of Percolation and Diffusion Controlled Leaching

E:7

Percolation controlled

Diffusion controlled

842 20.2 PHYSICAL AND CHEMICAL FACTORS WHICH EFFECT CONSTITUENT RELEASE

The physical and chemical properties of a product and the interactions with the surrounding environment define the constituent release behaviour of the product (Figure 20.2). Several important physical properties of the product are related to the internal pore structure of the material. These include permeability, porosity and tortuosity. Permeability defines the rate at which water can percolate through the product. The difference in the hydraulic conductivity gradient between ash products and the surrounding soil layers generally results in flow around or runoff from a product instead of percolation through the material. Porosity indicates the void space within the material and the potential for water absorption. Tortuosity is the ratio between the actual and apparent geometric path length of a species diffusing through a material. Actual path lengths are longer than apparent geometric path lengths because of the irregular nature of internal pores. Tortuosity is used to estimate the physical restriction to leaching of a leachable constituent. Figure 20.2 Mechanisms of Release and External Influences During Diffusion Controlled Leaching Monolithic Stabilised Wastes and Construction Matedal CbC

Emanation VOC's

Chemical factors:

Physical properties: Permeability Tortuosity Porosity ( c l o s e d ) Waterrepellant properties

pH Redox Complexation Sorption

Release mechanisms: dissolution wash-off diffusion

Soluble salts Organic m a t t e r ~ (high pH)

Humic-and fulvic acids

~

o

0

9

o. Erosion

o o

External influences H§

C02

02

eg. C I

843 Durability of contaminant deterioration material with

a product also is a physical property that significantly influences release. A product which is not durable is subject to erosion and to a granular material over time and should be evaluated as a granular percolation (e.g., solubility) as the controlling release mechanism.

Some of the chemical properties of products which affect constituent release include pH, alkalinity, reducing potential, chemical complexation and sorption. For example, the internal pore water pH of a product greatly influences the solubility of ash constituents in the pore solution, whereas the alkalinity of the material determines the resistance to changes in pH by external influences. The most significant external influences on pH are external applications or acid or alkali, uptake of carbon dioxide, acids produced by biological activity, and acidic precipitation. Uptake of carbon dioxide by alkaline products such as cement blocks typically results in a lowering of pH to between 8 and 9 at the exposed surfaces. The reducing potential is important because subsequent exposure of reduced material to oxidising conditions may result in an increase in the mobility of metallic elements (ie. oxidation of sulphide minerals). Increased metal mobility may also occur when high concentrations of chlorides or organic acids act as complexing agents which sequester the metals and keep them in solution. Conversely, the sorption or cation exchange by an ash's products component may reduce the mobility of leachable constituents. The chemical retention factor is a relative indicator of the cumulative influence of the material's chemical properties on the mobility of contaminants. 20.3 TEST METHODS FOR MONOLITHIC AND COMPACTED GRANULAR PRODUCTS Test methods used to evaluate diffusion controlled release are frequently referred to as either "monolith leach tests" or "tank leaching tests." Two primary test methods are used for evaluation of monolithic materials. The first test method is the American Nuclear Society Method 16.1 (ANS 16.1, 1986). The second test method is a recent variant of ANS 16.1 that has been adopted in The Netherlands as NEN 7345 (1994). Both tests are operationally very similar (Figure 20.3). A molded and cured test specimen of defined geometry is immersed in distilled-deionised water (leachant). The leachant is replaced with fresh leachant after specified time intervals and the recovered leachant is sampled for chemical analysis. The principal difference between the tests is the basis for data reduction, and the calculation of effective diffusion coefficients and other parameters. A compacted granular leach test was developed to provide diffusion controlled leaching parameters from materials that would be compacted in place to a low permeability (Kosson et al., 1993). In addition, this test permits direct comparison of untreated ash and product properties.

844 Figure 20.3 Schematic Diagram of Tank Leaching Tests for Monolithic and Compacted Granular Material

J

a) Monolithic

b) Granular

20.3.1 ANS 16.1

This leach test uses a cylindrical monolith as the test specimen which is immersed in leachant for defined intervals. At the conclusion of each interval, the resulting leachate is collected and replaced with fresh leachant. A liquid to monolith surface area ratio of 10 is used. Each leachate is analysed individually for species of interest and the cumulative release is calculated as a function of the cumulative leaching interval. A leachability index (PDe) is then calculated based on the initial concentration in the monolith and the release rate for each species of interest. 20.3.2 NEN 7345

This leach test uses specimens with a minimum diameter of 40 mm which is immersed in leachant (demineralised water) using a closed vessel. The liquid renewal is carried out after 0.25, 1, 2.25, 4, 9, 16, 36 and 64 days. A liquid to product volume of 5 (l/dm 3 of product) is used which corresponds to a liquid to surface area of 80 I/dm 2. In the data processing, the leaching behaviour of constituents is verified to being diffusion controlled (slope -.. 0.5). Subsequently, the effective diffusion coefficient is calculated using the total availability of a contaminant as determined using the NEN 7341, (1994) as the driving force for diffusion. Corrections are made to account for surface wash-off effects. Release after a specified exposure time is expressed in mg/m 2. If the release is not diffusion controlled, alternative methods of assessing release estimates are provided.

845

20.3.3 Compacted Granular Leach Test Effective diffusion coefficients for diffusion controlled release from compacted granular materials can be estimated using a compacted granular leach test (Kosson et al., 1993). Granular materials to be tested are compacted at proctor optimum moisture content and a standard compactive effort into a cylindrical mold and allowed to cure prior to testing. The top face of the cylinder is covered with a thin layer of glass beads to avoid erosion during testing. The entire molded sample then is immersed in distilled water for specified time intervals (e.g., 1, 2, 4, 8, 16 and 32 days). The resulting leachate is decanted and replaced with fresh leachant at the conclusion of each leaching interval. Each leachate is analysed separately for each species of interest.

20.4 INTERPRETATION OF DIFFUSION CONTROLLED RELEASE 20.4.1 Characteristic Release Behaviours Data obtained from monolith or compacted granular leach tests can be used both to identify characteristic release behaviour and to calculate release rate parameters. Characteristic release behaviour can be identified by plotting the cumulative release as a function of contact time on a log-log scale (Figure 20.4). This form of data presentation will be referred to as a "release plot." The following characteristic release behaviours can be distinguished using the release plot: Diffusion controlled release Depletion of leachable species Delayed release Surface wash-off Washout of mobile species, and Changes in chemical conditions

Diffusion Controlled Release Leaching from most cement-based and asphaltic materials is controlled by diffusion through the solid matrix. Diffusion controlled release is indicated by a slope of 0.5 in the release plot. Diffusion controlled release may be preceded by surface wash off or delayed release. Depletion may occur after an interval of diffusion controlled release. Intrinsic leach parameters, including effective diffusion coefficients, tortuosity, and chemical retention can be derived allowing estimation of release at time scales considerably longer than the duration of the experiment. Depletion of Leachable Species Depletion may occur in the later stages of leaching when the concentration of the species of interest within the center of the test specimen is reduced significantly by the

846

Figure 20.4 Typical Release Profiles Obtained from Tank Leaching Tests for Different Release Mechanisms Diffusion controlled release

60000

Depletion of Leachable Species

5000000

~

1 000000

E

10000

Q~ (D (D

n,

J

f

00000

100o

+

9

9

f 300

~1

01

1'0

1 00

Delayed release

50000

1

10

100

Surface wash-off

30000

"'J

10000

10000 0.1

1 0000 -

1000-

100Jr

Jr

+

Jr

Jr

+

10-

1-

1000 300

9 0.'

100000

|

1

10

1 O0

Wash-out of mobile species

0.1, 0

~

lOO

Change in Chemical conditions

50000 10000

10000

+

9

I

000

+

$

100, J

.

+

Jr

-t-

e

9

o

1000

300 0.

+

"~

1;

loo

0.1

1

10

Time (days)

100

847 cumulative effect of leaching. When this occurs, the initial boundary conditions and experimental assumptions are not met. Depletion may occur when the approximate properties of the material being tested are not known in advance. A slope of less than 0.35 afte.._._Errelease in accordance with a slope 0.5 is an indication of depletion. This is verified easily by comparing the cumulative release data with the maximum release quantity obtained from the availability leach test. When more than 50% of the available mass present in the test specimen is leached within the time span of the test, significant depletion can be expected to have occurred.

Delayed Release

A material can be covered with a relatively insoluble or depleted surface layer as a result of process conditions or testing methods. This layer initially retards contaminant release during testing. This behaviour may be observed from the compacted granular leach test because of the presence of the surface layer of beads used to prevent material dispersion.

Surface Wash-Off

A material can be covered with a relatively soluble surface coating as a result of process conditions or condensation processes. This layer can dissolve rapidly during the initial phase of the monolith leaching experiment and is indicated by an initial slope of less than 0.35. In many cases, the subsequent release is diffusion controlled. This type of release is most common for slag type materials.

Wash-Out of Mobile Species (Dissolution)

The solubility of a large fraction of the matrix can be such that dissolution of material from the surface proceeds faster than diffusion through the pores of the matrix. This phenomenon has been observed in products containing very high gypsum Ioadings (van der Sloot, 1991). In the case of high gypsum content, calcium sulphate solubility is relatively high, which results in an initial theoretical slope of +1 in the release plot. At longer time intervals, the slope decreases as the extract solution becomes saturated with respect to calcium sulphate.

Change in Chemical Conditions

The release rate of a constituent may change during the course of testing because of changing chemical conditions either within the solid matrix or in the extracting solution. This includes changes in redox potential, pH or the depletion of a chemical species which may be limiting the pore water solubility of another species. An example of this type of behaviour is the leaching of Ba, which can be delayed clue to an initial high release of sulphate. This causes a low initial Ba leachability, after the sulphate peak Ba starts to leach at a higher level in diffusion controlled mode.

848 20.4.2 Definition of Leaching Parameters The leaching behaviour and release rates of constituents from products containing incinerator residues can be modified by changes in one or more of the following factors: Total availability, or the fraction of each element not tied up in silicate and relatively insoluble mineral phases (deemed to be not leachable). Availability referred to in this chapter is the quantity of an element extracted based on the NEN 7341 leach test. This quantity typically is less for products because of dilution and treatment effects. Physical retention, which is derived from measuring the release rate of an inert component (one which does not chemically interact with the product matrix) from the product matrix. Physical retention is equivalent to tortuosity for cases where the test sample remains physically intact (e.g., no significant cracking or disintegration). In these cases, tortuosity is an approximation of the ratio of the actual mean path length a species travels from within the monolith to the monolith surface for release to the mean direct geometric path length. For cases where the monolith does not maintain physical integrity, the physical retention factor is a relative reference index for the degree of species retention within the SIS matrix by physical encapsulation at the micro-scale. Sodium or potassium are most frequently used as the non-interactive component for estimating of physical retention, and Chemical retention, which is a function of each element's chemical interaction with the product matrix. This is derived from the measured release rate of a given component, its free diffusion coefficient in water and the tortuosity as obtained from the inert component release rate. However, differences in redox potential between the testing conditions and actual environmental exposure can result in different release rates The following parameters are used to quantify release rate information based on monolith or compacted granular leach test results: Availability (for NEN 7345) or total solid phase concentration (for ANS 16.1) Effective diffusion coefficient (for NEN 7345) and leachability index (for ANS 16.1 only) Physical retention factor or tortuosity, and Chemical retention factor Effective diffusion coefficients, leachability indexes, physical retention factors and chemical retention factors are calculated from data obtained during the monolith or compacted granular leach tests in conjunction with either the availability or total solid phase concentrations (van der Sloot, 1991, van der Sloot et al., 1989).

849 The ion flux through the geometric surface area of a product under diffusion controlled conditions is described by Fick's second law: aC _ D a2C at

(20.1)

~ ax 2

where De is the effective diffusion coefficient [m2/s] and C is the species concentration available in the solid phase. A one dimensional semi-infinite linear diffusion model can be applied to estimate De based on data obtained from both monolithic specimens and compacted granular material as long as the boundary conditions for the use of the one-dimensional model are fulfilled. The principal conditions applied to the leaching test are: 9

the material is uniform in composition

9

no depletion occurs over the duration of the test leachant replacement cycles are frequent enough to ensure that the concentration gradient between the solid being leached and the extractant is maximised (e.g., the species concentration in the extract is dilute), and monolithic samples maintain physical disintegration) during testing

integrity (e.g.,

no cracking

or

For cases where the physical integrity of the monolith is not maintained, estimated values of physical retention and chemical retention factors should be regarded as relative indexes and not be used for extrapolation and estimation of releases over longer time intervals or different physical geometries. The solution of Fick's law of diffusion for the above conditions was presented by Crank for the diffusion from a product with semi-infinite dimensions, in which the initial concentration is uniformly distributed in the product and the concentration on the surface between the product and the leachate is constant with respect to time (Crank, 1989):

c-C'=erfl X l

Co-C,

[2

(20.2)

where: C = C(x,t) is the concentration as a function of location within the solid test specimen and time C1 is a constant concentration at x=0 (test specimen surface) Co is the initial concentration (at t=0) in the product which must be uniformly distributed De is the effective diffusion coefficient [m2/s]

850 is time [s], and is the distance from the surface [m], positive values In the monolith leach test, the surface concentration will only be constant as long as no depletion occurs and the mean concentration in the solution does not deviate significantly from zero. These requirements are met by using a product for which the smallest dimension is greater than 5 cm, preventing depletion within the 64 day time frame of the experiment, and by refreshing the leachant at regular time intervals. This size requirement fulfills the assumption of semi-infinite media for most cases. Leachant replacement intervals should be using a geometric time progression:

t.=n2to where:

(20.3)

is the leachant replacement time for interval n [days] is the interval number (integer) [-], and is the initial leaching interval [days]

Laboratory experiments coupled with model estimations have indicated that use of the boundary condition C~=0 for t>0 is appropriate for these experimental conditions as long as the leachate remains dilute relative to leachate saturation element of interest (see Chapters 13 and 15). The resulting diffusion equation derived from Equation (20.2) for this boundary condition is: D= nM2 4t(PCo)2 where: D or De Mt t Co P

is the is the is the is the is the

(20.4)

diffusion coefficient for component x in the product [m2/s]1 cumulative release of the component [mg/m 2] contact time [s] maximum leachable quantity [mg/kg], and bulk density of the product [kg/m 3]

Co also can be interpreted as the "driving force" for diffusion in the solid. In the test method according to ANS 16.1, Co is equal to the total solid phase concentration of the product. This may not be accurate for evaluation of some elements in incinerator residue by-products because a fraction of these elements may be in mineral forms which render them unavailable for leaching. Defining Co as the total solid phase 1The diffusion coefficient (D) often is referred to interchangeable with the effective diffusion coefficient (De) in the literature.

851

concentration results in smaller values of D and thus can result in the underestimation of the rate of constituent release and the overestimation of the time to depletion. NEN 7435 corrects for this deficiency by defining Co as the experimentally determined availability. It is often convenient to express D as pD values where D is transformed by the mathematical operator "p" which results in

pD=-Iog(O)

(20.5)

When pD is calculated using the total solid phase concentration equal to Co according to ANS 16.1, the resulting value is also referred to as the "leachability index". Variability associated with the measurement of pD values for monolithic materials as a function of pD is presented in Figure 20.5. Greater standard deviations are associated with larger pD values because of slower rates of release and errors derived from measurement of extract concentrations approaching analytical detection limits. Greater variability associated with larger pD values are of limited concern because of the very slow release rates indicated. Figure 20.5 The Standard Deviation Associated with Measurement of pDe as a Function of pD e 1.50

tO

1.20

.n,.

(~

a

0.90 +

+ -t-

I,..

"O t-

0.60

,.I,,-I

0.30

0.00

-

8

+

-H-,

10

12

Pge

"

14

++ u

+

+

16

+

20

852 The effective diffusion coefficient can be divided into contributions from the free mobility of the element or species of interest, physical retention factor and the chemical retention factor:

D -D~

RT

where: D D R

is the is the is the is the

O,X

1;

(20.6)

diffusion coefficient of component x in the product [m2/s]; diffusion coefficient of component x in water [m2/s]; chemical retention factor of component x in the product [-]; and, physical retention in the product [-]

The physical retention or tortuosity reflects the extended path length of a diffusing ion in the pore structure of a product. Values may range from 1 to 10 for stabilised materials and up to 1000 and higher for very dense concrete and bituminous mixes with water repellant properties. A tortuosity value of 1 reflects no physical retention during the leaching process. Physical retention based on sodium as the inert tracer is calculated as:

DNa DNe o

Twhere: 1;

B~ DNa

is the tortuosity of the product is the diffusion coefficient of Na in water (pD~ = 8.88 at 22 ~ is the diffusion coefficient of Na in the product [m2/s]

(20.7)

[m2/s], and

The chemical retention factor quantifies the retardation of the release of a component relative to release of an inert species by chemical interactions of the diffusing ion with the product. The chemical retention factor (R) for the component of interest is calculated from:

R- Dx~ DxT

where:

R

BOx

Dx

t

(20.8)

is the chemical retention factor [-] is the diffusion coefficient for component x in water, [m2/s] is the effective diffusion coefficient for component x in the product, [m2/s], is the tortuosity of the product [-]

853 Figure 20.6 The Effects of Pore Water pH on Magnesium Chemical Retention and Observed Leachate Concentrations 106 ,"T"

~-.,... 10 s o o

10 4

r0

.=9

10 a

9

10 2

t-,.

o

o..,..

,..01

10 ~

E x: 0

0

Mg

10 0

10-~

9

I

f

8

I

9

l

I

l

I

10

12

11

pH

E tO

. m

(b 0 tO o t-

o

_J

A

10 2

Mg

"

e,

10 ~

10 0

O

10-1

10-2

I

'

I

10 pH

'

I

11

l

12

854 Pore water composition and the porewater pH are important factors in controlling chemical retention. In other work, relations between porewater pH and retention values have been established. In a number of cases, these relations are strictly based on the solubility as a function of pH of the specific species. Figure 20.6 illustrates the effect of porewater pH on the release of magnesium (de Groot, 1989). The element Mg is an example of an element whose solubility decreases monotonically with increasing pH. Thus, the retention value increases monotonically with the increasing pH. Reactive surface sites also may affect the diffusivity of components in the pore solution. Figure 20.7 presents typical chemical retention profiles from tank leaching tests for several elements. Chemical retention factors may range from 1 to more than 1,000,000, where a value of 1 reflects no chemical retention during the leaching process. The wide range of chemical retention factors reflects that substantial reductions in release rates can be obtained by chemical modifications. Measures taken to modify the chemical retention of specific components can result in the increased mobilisation of other matrix components. Multiparameter screening is necessary to insure that no adverse effects have been introduced. In summary, the larger the pD e at constant availability, the slower the release from the material. Measures to improve the environmental quality of products containing secondary materials can be targeted at any of the three factors (availability, physical retention or chemical retention). Changes in processing parameters can generally lead to a denser product and, consequently, influence the physical retention factor. However, within one category of materials, the range in tortuosity is relatively small. This implies that changes in chemical properties, which influence both availability and chemical retention, have a greater effect on release rates. 20.4.3 Calculation of Effective Diffusion Coefficients from Cumulative Release Data

For the determination of characteristic leaching behaviour, the logarithm of the cumulative release has been plotted versus the logarithm of time. Rearranging Equation 20.4 yields:

Mr=CoP!

(20.9)

and after logarithmic transformation: (20.10) The release of each component per time interval can be calculated from the monolith leach test results with the formula:

855 Figure 20.7 Typical Chemical Retention Profiles from Tank Leaching Tests for Several Elements

856

c'Vi

(20.11)

M,- I O00A

where:

Mi Ci

Vi A

is the is the is the is the

release during period i [mg/m 2] extract concentration of the component in the ith period [mg/I] volume of the contact solution [I], and geometric surface area of the specimen [m 2]

The measured release from previous periods is summed to obtain the measured cumulative release. This implies that deviations in a given period accumulate in the subsequent periods, which may hamper interpretation. The cumulative release until the i~ period can be calculated only from the release in the i~ period, assuming diffusion control in the i"~and previous time periods. These values can be used to check whether the release is diffusion controlled. If the measured and calculated cumulative releases are equal, the release is controlled by diffusion. The calculated cumulative release used for all N periods can be derived from:

M t i = Mj

where:

Mt.i

ti ti-1

(20.12)

~

is the cumulative release of the component through period i [mg/m 2] is the contact time after the period i[s], and is the contact time after the period (i-1) [s]

After plotting the logarithm of the calculated cumulative release (Mu) against the logarithm of the time (t~) for i=1 to N, the slope of the relation over the complete time interval and three time intervals segments can be determined: Interval Interval Interval Interval

0 (total range) 1 (initial leach range) 2 (intermediate range) 3 (last range)

-

leaching leaching leaching leaching

extracts extracts extracts extracts

1 to 1 to 3 to 6 to

8 3 6, and 8

The mechanism of leaching during each interval can be derived from the slope of the data from the respective interval. Components dissolving from the surface (slope __0.8), short initial release of surface deposited components (initial stage slope _<0.4) and diffusion controlled release (slope = 0.5) can be distinguished. The meaning of the change in the slopes at different time-intervals is summarised below:

857

SLOPE Leachin~ Interval Initial

<0.35 Surface wash-off

0.35-0.65 Diffusion

Intermediate Last

Depletion Depletion

Diffusion Diffusion

>0.65 Lag time/dissolution Dissolution Dissolution

Only in the cases where the slope is approximately 0.5, can effective diffusion coefficient be calculated. When the slope in the entire plot is 0.5 + 0.15 and the slope in the last range is less than 0.6, data from all leaching extracts can be used for the calculation of the effective diffusion coefficient. If this is not the case, only those intervals in which the slope is 0.5 + 0.15 can be used for calculation of D(e). The effective diffusion coefficient (D(e)) for the component of interest then is calculated from each period using only those data points for which the slope is 0.5 + 0.15 by: nM 2

D~x= t ' (2Cop)2(ti-ti_l)

(20.13)

where:

Di,x Mi d C~ ti ti.1

is the effective diffusion coefficient of component x calculated from the release in period i [mS(2)/s] is the release in period i [mg/m 2] is the bulk density of the product [kg/m 3] is the maximum leachable quantity [mg/kg] is the contact time until period i [s], and, is the contact time until period i-1 [s]

The significance of the availability test for the judgment of the maximum leachability of a given component from waste materials is best illustrated by release data obtained for one product, where depletion of a very soluble component occurs, and the maximum leachable quantity, as determined by the availability leach test, is approached. The data obtained during the monolith leach test are expressed as (mg element leached per m 2 surface area) as a function of time. The availability leach test results are recalculated to the same units by taking the size and shape of the specimens into account. These figures indicate that the release of sodium, chloride, potassium and bromide asymptotically approached the release limit dictated by the availability leach test. Similar observations also have been made in column leaching experiments, indicating the practical significance of this test method as a screening tool for ultimate release at the very long term.

858 20.4.4 Alternative Release Models for Monolithic Materials

Batchelor and Wu (1993) reviewed a variety of alternative mathematical models for estimating diffusion controlled release from products. Simple models assume: 1. rectangular geometry 2. infinite bath, and 3. semi-infinite solid The rectangular geometry assumption implies that the surface of the product can be treated as a flat plate. Alternative solutions have been obtained for spherical and cylindrical geometries and coordinate systems. The infinite plate assumption implies that the leaching solution remains dilute with respect to the species or element of interest. This may not always be true and requires careful selection of laboratory conditions and consideration of field boundary conditions. The semi-infinite solid assumption implies that significant depletion of the species of interest does not occur. Several cases can be defined which meet the above assumptions. These cases include : (i) no reaction, (ii) equilibrium linear sorption, (iii) equilibrium precipitation, and (iv) reaction with a leaching solution component. The mass balance for all of these cases is:

ac =Do ,,,'' at

ax 2

-R

(20.14)

The solution for the case of no reaction is:

C(x,t) =Co err

Mt _ 1 r L

M0 LC0 Jo

(4D~t)o.s

x)

(20.15)

(Co-C) dx

(20.16)

(20.17) The solution for the case of equilibrium linear sorption is: C i=Kpc

(20.18)

859

Mt I Dobs =

0.5

nL 2 ) De

l+Kp

=FmD e

(20.19)

(20.20)

where"

a Mt

=

Mo

=

Dobs t L I% Fm

= = = = =

reaction term cumulative mass released initial mass present in the solid phase observed diffusion coefficient time characteristic length linear sorption coefficient fraction of the constituent that is mobile

The solution for the case of equilibrium precipitation is:

Mt-I4D~

Mo

D obs

0.5

nL 2 )

n(Fm-O'5Fm2) De =

(20.21)

(20.22)

if Fm < < 1.0, then

Dobs- nFmDe 2

(20.23)

The solution for the case of reaction with a leaching solution component, e.g., acetic acid ( H A c ) i s 2HAc + Ca(OH)2 = Ca 2§ = 2 A c + 2H20

860

Mr_( 4Do~tl0.5

(20.24)

Dobs- nCCHAcDe'HAc

(20.25)

Mo

nL 2 ) b

0

2nCc.oH

It is important to note that for each of the above cases, the function for cumulative release is of the same form, being related to the Dobsand time to the one-half power. A more complex model for contaminant release was developed by Hinsenveld (1992) and is referred to as the "shrinking core" model. This model was developed to model release from an alkaline SIS product being leached in an acid environment. The model assumes hydrogen ion diffusion into the product matrix and reaction at a boundary between an unleached core and a leached exterior shell. The reaction at the boundary results in release of product matrix components which, in turn, diffuse outward through the reacted shell into the leaching solution. Thus, the model includes time dependent moving boundary between the leached shell and unleached core. The core is depleted (and hence the shell is extended) as a function of the acid concentration in the leaching solution and the acid neutralisation capacity of the SIS product. Diffusion through the shell is assumed to be controlled by the matrix tortuosity. The reaction at the core-shell boundary is controlled by the kinetics of acid dissolution. The two applicable limit cases are (i) when diffusion through the shell is rate limiting, and (ii) when dissolution at the shell-core boundary is rate limiting. This model has been successfully applied to describe the rate of component release when a SIS product is leached using a variety of acid solutions (Hinsenveld, 1992).

20.5. RELEASE FROM PRODUCTS CONTAINING INCINERATOR RESIDUES 20.5.1 Total Availability The total availability of constituents from a product containing incinerator residues can be modified by the following processes: Dilution by treatment or product additives Filling of micropores and coating with hydrophobic materials, or Chemical respeciation Dilution occurs when incinerator residues represent only a limited fraction of the product material. Examples of significant dilution effects include partial substitution of grate ash as aggregate substitute in cement or asphalt materials. In these cases, the grate ash may represent from less than 10% to more than 50 wt% of the final material.

861 Filling of micropores and coating with hydrophobic materials reduces availability by providing a coating over soluble mineral phases that is resistant to aqueous dissolution. An example of when this case occurs is when grate ash (a porous material) is incorporated into asphalt. Hot asphalt cement is absorbed into the micropores of the ash. This asphalt coating may remain even after size reduction for availability testing. Chemical respeciation may occur as a result of chemical reactions during product production. Respeciation may be the result of a relatively passive process such as hydration and oxidation reactions that occur during ash weathering, or, may result from reaction with process additives. In addition, respeciation may occur through sintering or other thermal processes (e.g., vitrification) of the material during high temperature processing. In general, the most significant factor affecting availability for low temperature processes is dilution by process additives. Incorporation of trace elements (e.g., lead, zinc, etc.) into alumina-silicate matrices or respeciation during sintering is responsible for the reduction in availability by thermal treatment processes.

20.5.2 Effective Diffusion Coefficients, Physical Retention and Chemical Retention Typical tank leaching release profiles for several elements and sulphate from untreated and treated incinerator residues are presented in Figures 20.8 through 20.12 (Kosson et al., 1993). Copper (Figure 20.8) Release plots of copper from untreated and treated APC residues, fly ash and bottom ash generally form three groupings. The groupings, from least release to greatest release, are: (a) bottom ash incorporated into pavement blocks; (b) u,ltreated bottom ash, bottom ash treated by SIS processes and untreated APC residue; and, (c) APC residues treated by Portland cement or phosphate addition. In general, treatment of bottom ash tended to decrease release while treatment of APC residues tended to increase release. Cadmium (Figure 20.8) Release plots of cadmium from untreated and treated APC residues, fly ash and bottom ash generally also form three groupings. The groupings, from least release to greatest release, are: (a) bottom ash incorporated in paving blocks; (b) fly ash in asphalt and bottom ash treated with Portland cement or phosphate addition; and, (c) untreated APC residue and bottom ash treated with Portland cement. In general, while incorporation of fly ash into asphalt provided consistent results, the effectiveness of SIS treatment of bottom ash was very process specific.

862

Figure 20.8 Tank Leaching Release Profiles for Copper and Cadmium from Untreated Incinerator Residues and Products Containing Residues

"-

Cu

dk vs vbal

1.0E+04

-~----~P---- vs vba2 vs vba3

1.0E+03

&

vs vba4

.... Z~----- vs vab5

1.0E+02

-----O-~

ol

E

9

vs vba0 vs vfa0

1.0E+01

m

-"

vs vfal

~: 1.0E§

-"

vs vfa3 vs vfa4

1.0E-01

Fr fa R1 -----'-----

1.0E-02 0.10

1.00

10.00

100.00

Time (days)

mamm aspha FA

#

NL cm beton BA

r

N L beton BA

Cd 1000 9

NL aspha FA

9

NL asph fa

,~ --''-'-X~X~X'-'--X~X 100 -"

NL asph fa vs vfa0

,I,

vs Voal vs vba2

=

vs vba3 vs vab4

~ •

0.1

~

0.1

1

10 Time (days)

100

vs vab5 ~

NL beton ba

863

Lead (Figure 20.9) Release plots of lead from untreated and treated APC residues, fly ash and bottom ash indicate release behaviour very similar to that observed for copper. Generally, the three groupings formed, from least release to greatest release, are: (a) bottom ash incorporated into pavement blocks; (b) untreated bottom ash, bottom ash treated by Portland cement processes, and untreated APC residue; and, (c) APC residues treated by sis processes. In general, treatment of bottom ash tended to decrease release while treatment of APC residues tended to increase release. Zinc (Fig. 20.10) The three groupings observed from release plots for zinc, from least release to greatest release, are: (a) bottom ash incorporated into pavement blocks; (b) bottom ash treated by sis processes and fly ash used in asphalt; and, (c) bottom ash in asphalt and untreated APC residue. Unlike for all previous cases discussed, zinc release from bottom ash in asphalt does not appear to be diffusion controlled. The principal release mechanism appears to be dissolution from the asphalt binder, as suggested by the significant release from the reference material. Sodium, Sulphate and Chloride (Figures 20.11 & 20.12) Release plots for sodium, sulphate and chloride are provided only for bottom ash and fly ash in asphalt and corresponding reference asphalt materials. The release from both incinerator residues in asphalt appears to behave similarly for all three species, with greater sodium release from bottom ash than fly ash containing materials and both cases much greater than the reference materials. This result appears to be directly related to the total loading in the asphalt material. Much greater Ioadings of bottom ash than fly ash were incorporated in the materials tested. However, the release rates observed for asphalt materials are much slower than that observed for untreated residues and residues incorporated in cement matrices because of much greater tortuosity within the asphalt matrix. It is evident from the above discussion, that evaluation of release plots can provide important information and comparisons between processes and elements. However, release plots do not provide necessary information for translation of results from one scenario to another or a design basis for treatment process or product development and evaluation. Calculation of effective diffusion coefficients, physical retention factors and chemical retention factors provide this basis. Figures 20.13 and 20.14 present effective diffusion coefficients for several elements from untreated and treated (SIS with Portland cement) bottom ash and APC residue. The contributions of free diffusion, physical retention and chemical retention to the effective diffusion coefficient are indicated. The "+" symbol at the top of each bar on the figure indicates the effective diffusion coefficient calculated based on the total elemental content of the material instead of the elemental availability. Use of the total

864 F i g u r e 20.9 T a n k Leaching Release Profiles for Lead from Untreated Incinerator R e s i d u e s and Products Containing Residues

-

Pb

A

------o 1.0E+04

vs vba3 ~,

~ 1.0E+02

vs vba4 vs vab5

OI

(/)

vs vbal vs vba2

1.0E+03

E

dk

vs vba0

1.0E+01

vs vfa0

.==,

~: 1.0E+O0

vs vfal

1.0E-01 1.0E-02 0.10

1.00

10.00

Time (days)

100.00

-"

vs vfa3

"--

vs vfa4 Fr fa R1 mamm aspha

865 Figure 20.10 Tank Leaching Release Profiles for Zinc from Untreated Incinerator Residues and Products Containing Residues

Zn 1000

X~

_......_x~X~X

100 A

E

-'-'X~X

ol

E v

10

-"

NL aspha FA

~-

NL asph fa

A

NL asph fa

c}9

vs v f a 0

A

vs v b a l

/',

vs v b a 2 9

0.1

........

0.1

i

........

1

;

10

........

;

100

vs v b a 3

o

vs v a b 4

•

vs v a b 5

-~

N L beton ba

T i m e (days)

Zn 100 "r UNH 25%BA 0

UNH 50%BA UNH 75%BA

~ lO

[]

U N H ref NL 2 % F A 3 ' NL2 % FA2

A

NL 2 % FA1 NL asp ref

=me

(day

866 Figure 20.11 Tank Leaching Release Profiles for Sodium from Untreated Incinerator Residues and Products Containing Residues

Na 10000 UNH 25% A O4

E

UNH 50%

looo

=

E

B:

-----o -"

100

UNH 75 % UNH ref NL 2% FA1 NL 2% FA2

10 0.1

1

10

Time (days)

100

-"

NL 2% FA3

~9

NL asp ref

867 Figure 20.12 Tank Leaching Release Profiles for Sulphate and Chloride from Untreated Incinerator Residues and Products Containing Residues

Sulfate 10000

= r

UNH 25% UNH 50%

E

1000

E r

-$

100

t,

UNH 75 %

o

UNH ref

,t,

NL 2% FA1 NL 2% FA2

10

9-,,---.

~

0.1

-9

9 ...... 9

1

10

NL 2% FA3 NL asp ref

100

Time (days)

C! 10000

A

1000

-

UNH 25% BA

m__,

UNH 50% BA

-~

UNH 75 % BA

~

UNH ref

-"

NL 2 % FA1

"~

NL 2% FA2

O1

E

"~ lg

100

10

. . . . . . . .

0.1

:

1

. . . . . . . .

~

10

Time (days)

. . . . . . . .

100

NL 2% FA3

868 Figure 20.13 Effective Diffusion Coefficients from Tank Leaching Tests on Untreated Bottom Ash and APC Residue for Several Elements

869 Figure 20.14 Effective Diffusion Coefficients from Tank Leaching Tests on Bottom Ash and APC Residues SIS with Portland Cement for Several Elements

870 elemental content can result in under estimation of the effective diffusion coefficient by up to two orders of magnitude (or under estimation of PDe by two units). Thus, use of the leachability index described by ANS 16.1 is not recommended, because the driving force for diffusion is not specified by the L factor alone. For the same L factor, the net release at a given exposure time can be orders of magnitude different depending on the driving force for diffusion. Figures 20.15 and 20.16 provide a comparison of effective diffusion coefficients for untreated bottom ash, combined ash and APC residue and the same residues treated by five SIS processes (Kosson et al., 1993). The contributions of free diffusion, physical retention and chemical retention to the effective diffusion coefficient are indicated. The effective diffusion coefficient calculated based on the total elemental content of the material instead of the elemental availability also is presented. These figures illustrate that some SIS treatment may actually increase the rate of diffusion even if the availability is reduced. For example, the pDe for lead release from untreated APC residue (Figure 20.16) was greater than the pDes observed for the treated material. This type of evaluation also permits distinction between reductions in pD~ achieved through physical and chemical changes during treatment. Physical retention is defined by the physical matrix and therefore does not vary as a function of elemental release. It can however, be significantly influenced by the type and design of the treatment process. It should be noted that the treatment of APC residue presented in Figures 20.16 and 20.17 resulted in decreased physical retention compared to the untreated material. This result occurred because of the high soluble salt content (e.g., up to 50 wt% sodium and potassium chlorides and sulphates) present in the treated material. Wash out of the highly soluble salts resulted in a highly porous matrix. A correlation has been observed between incinerator residue loading (or, conversely the relative quantity of treatment additives) and physical retention (Figure 20.17). Increased incinerator residue loading resulted in decreased physical retention for cement-based products for both filter ash and bottom ash. This effect may be the result of increased loading of soluble salts. Increased physical retention was indicated for increased residue loading for asphalt materials containing bottom ash. This observation may be related to increased asphalt cement requirements associated with increased bottom ash content. Table 20.1 presents a comparison of tortuosity values observed for residues in different matrices and reference materials. In general, increased product density, decreased porosity and hygroscopic properties all result in increased tortuosity or physical retention factors. Chemical retention within a matrix is controlled by the chemical speciation of the material constituents and the pore solution composition. Thus, treatment processes may effect changes in chemical retention either by respeciation or alteration of the pore solution. The principal effect of the treatment process presented in Figures 20.15 and 20.16 is the most likely the result of increased matrix alkalinity and attention buffering of pore solution pH. In general, increased chemical retention values are associated

871 Figure 20.15 A Comparison of Effective Diffusion Coefficients from Tank Leaching Tests on Untreated and SIS Bottom Ash, Combined Ash and APC Residue for Cadmium and Lead 18 e~

E

t~

BOTTOM Ash

16

Cd

J

J

161

.---.,

,i

Pb

_....

9

o

_J

BOTTOM Ash

18

9

14

20!

14!

-

12

12

v

I-Z LU

10

r,.) ,..m

ii 1.1_ iii

0 (.) n,'

iii ii o3 Z < n,' i-._1 ,_1

ILl

> O

uJ "lF--

O

F-Z

O

0 18

2

3

4

5

COMBINED Ash

Cd

O o

1

20

2

3

4

COMBINED Ash 9

5 Pb

@

o

16

I

14 12

11

w

0

1

20

2

3

4

5 Cd

APC Residue

18

0 18 16

16 ,

14

!

i

l

t ,

12

,

3

4

5

Pb

-1

i I .

9

12 10

I]

I

0

2

7-'/

t

10

1

9 APC Residue

,

14 I

Fi

8

16

F-r

1

1

2

3

4

5

0 - untreated residue 1 - SIS with Portland cement and polymeric additives 2 - SIS with Portland cement and soluble silicates 3 - SIS with cement kiln dust 4 - SIS with soluble phosphates 5 - SIS with Portland C e m e n t only

0

1

2

3

PROCESS

4

5

872 Figure 20.16 A Comparison of Effective Diffusion Coefficients from Tank Leaching Tests on Untreated and SIS Bottom Ash, Combined Ash and APC Residue for Sodium and Sulphate

873 Figure 20.17 The Effect of Residue Loading on Tortuosity for Filter Ash and Bottom Ash Incorporated into Various Product Materials MSWI Filter Ash 100000< >

1000 .

n

O .-1 t~

#.

100 -

109 1 0.01

I

I

0.1

1

Waste Loading Table 20.1 Ranges of Tortuosity Values Measured for Untreated Incinerator Residues, Products Containincj Residues and Reference Materials Material Unconsolidated granular Waste 1 Compacted Bottom Ash 2 Compacted Fly Ash 2 Stabilised Bottom Ash 2 Stabilised Fly Ash 2 Stabilised Fly Ash 6 Stabilised Fly Ash 3 Pavement Blocks Containing Bottom Ash 4 Pavement Block Reference 4 Asphalt Concrete Containing Bottom Ash 4 Asphalt Concrete Reference 4 Asphalt Concrete Containing Fly Ash 5 Asphalt Concrete Reference s 1. ECN Work, Various Wastes 2. ECN Study 3. BCR intercomparison

Tortuosity 2-4 20 - 25 5-10 1-6 25 - 35 200 - 210 7-19 35 1900 - 3300 24000 450 - 1400 > 3000 4. Mammoet 5. UNH Whitehead 6. Mehu/vdSloot Record

874 with decreased aqueous solubility of the element of interest either as a function of pore solution pH or the presence of a solubility limiting species. Elements or species that are highly soluble (e.g., chloride, sodium) have relatively little or negligible chemical retention. Table 20.2 provides a comparison of chemical retention values observed for several elements in untreated and treated residues and reference materials. Note that relatively high chemical retention values are observed for untreated residues and low values are observed for asphalt matrices. Table 20.2 Ranges of Chemical Retention Values Measured for Untreated Incinerator Residues, Products Containing Residues and Reference Materials Chemical Retention Values

Material

a b c d e f g h i j k I

Cd

Cu

Pb

Zn

CI

n.a 5 x102 2 x102 1.6 x 108 3 x 108 40,000 >30,000 (220) 150-600 1 n.d.

32,000 8x104 7x102

3.1 x 106 1,300-9x104 2x102-2x106 13,800 70,000 - 98,000 13,900 1.5 x 106 (820) 1,300- 3,100 1 n.a.

4 x 105 1.5x102-4x106 3x106-3x107

2.9 1.2-1.5 1-5 2.5 1.5 - 2 2 92 (110) 2- 3 1 1.4- 2

n.d.

>2 x 104 (780) 17- 26 1 n.a. n.a.

a. Compacted Bottom Ash b. Compacted Fly Ash c. Stabilised Bottom Ash d. Stabilised Fly Ash e. Stabilised Fly Ash f. Stabilised Fly Ash n.d. - not detected

20.5.3

n.a.

790,000 7 x 102 (17,800) 100- 200 1 n.d. n.d.

n.d.

g. Pavement Blocks Containing Bottom Ash h. Pavement Block Reference i. Asphalt Concrete Containing Bottom Ash j. Asphalt Concrete Reference k. Asphalt Concrete Containing Fly Ash I. Asphalt Concrete Reference n.a.- Not analysed

Solubility

Cadmium solubility as a function of pH is presented in Figure 20.18 for untreated APC residue from a semi-dry scrubber/fabric filter unit and the same APC residue SIS treated with Portland cement and treated with soluble phosphate (Kosson et al., 1993). It can be seen that the effect of treatment on solubility was minimal. Similar information is presented in Figure 20.19 for chromium in untreated and treated APC residue and bottom ash, and for lead in Figure 20.20. The characteristic solubility curves for chromium were the same for untreated bottom ash, cement stabilised bottom ash and

875 Figure 20.18 The Effect of SIS Processes on the Solubility of Cadmium for Bottom Ash and APC Residues Bottom ash, untreated 100

-

--

APC residue, untreated

. . . . .

100 - ~

_ .... odla o dido

o%c~~ o ~

Oo

dB~

a)~

o

o

,,.,..

%

b)

o

0.1

0.1

Cr 0.01

0

o P"

2

4

6

pH

8

10

12

0.01

14

Bottom ash, phosphate treatment

0

2

4

6

8

:_:_

10

pH

!

12

14

APC residue, phosphate treatment

100

100

10

~Oo~ v

% o oo

o

% 0.1

0.01

0

2

4

6

pH

o%

0,1

oo

8

10

12

0.01

14

Bottom ash, Portland cement treatment

0

2

4

6

8

pH

12

14

APC residue, Portland cement treatment

lOO

100 OO

Q:~o

o

o

f)~ o.l.J O.Ol

10

~ c

e

~

.

.

pH

0.1 ~

~

~

0.01 -~ 0

o

c~ 2

........ 4

~

~. 6

,., 8

__ 10

12

.... 14

876

Figure 20.19 The Effect of SIS Processes on the Solubility of Chromium for Bottom Ash and APC Residues APC residue, untreated

Bottom ash, untreated 10

a>

oo

1 r

0.01

.

r

.

,

2

eo,~O

g

r

,

0

%

b) g

Cr

0.1

o Cr ]

O

q~ 8 "_ --_

.

-

4

.

6

.

pH

o

o

.

.

8

12

O.1

0.01

.

10

14

~176

. . . . . . . 2 4 6 8 pH

0

10

12

APC residue, phosphate treatment

Bottom ash, phosphate treatment

oo

o

c)

..

o

I

d)

oz

0

0.1

o.oi

.

0

.

2

.

.

4

o o

1

g

Ib

~3

...

.

.

.

.

6

.

pH

8

.

.

10

.

12

o.oi

14

Bottom ash, Portland cement treatment

c,~176

0.1

0

2

4

~ 8 6

pH

8

1

0.1

='-

_

_

.

2

.

.

4

.

.

.

6

.

.

pH

8

.

10

f) g o

m,~

.

12

_

i4

{oc;) o

c~

0

14

10

o

0.01

12

APC residue, Portland cement treatment

10 ~

.e) ~

10

Q~o o o d=~ O.1

0.01

% ~-

0

. . . . . . . . . . . . 2 4 6 8 10 12

pH

14

14

877 Figure 20.20 The Effect of SIS Processes on the Solubility of Lead for Bottom Ash and APC Residues

APC Residue, Untreated

Bottom Ash, Untreated 1,000

.

.

.

.

.

1,000

10o

100

o

0

~o

a)

o

0.1

0

2

4

6

o

oo

8

pH

% E

9

(~:)

10

12

b)

oo

9

10

JD

0.1

14

o

%

'~ 1

Bottom Ash, Phosphate Treatment

o:

Pb

0

o ,..,.,

2

4

6

pH

,=

8

10

12

14

APC Residue, Phosphate Treatment

-

1,000

6;~OOo

..-.,.

1 fO00

-

,

,

to O

100

c) ~ .....

o

10

10

0

.o 13.

%o

8o ~~176 o

tt

ooo

g~

Pb

0.1

0

2

. . . . . .

4

6

pH

8

. . . . .

10

Bottom Ash,. Portland Cement Treatment 1,000

~

.....

,

.o

l~.

10 1 0.1 1 0

o oO

wb 2

o

,r

4

0~

~ 6

pH

8

10

12 14

...........

2

4

6

pH

8

10

12

I

14

1,(X)O

[oPb)

o

0

APC Residue, Portland Cement Treatment

......

lO0

e) ~

0.1

12 14

100

f) ~

JO 0.

r

10 o o 1

0.1

Pb

0

"

" 2

. . . . . 4 6

-

pH

. . . . 8 10

"

12

14

878 cement stabilised APC residues. These curves exhibited a regime of intermediate solubility between pH 5 and 10, and sharply increasing solubility with pH decreasing below 5. The solubility between pH 5 and 10 may be indicative of chromate (Cr+6). The curves for untreated APC residue, APC residue treated with soluble phosphate and bottom ash treated with soluble phosphate indicate increasing solubility only at pH less than 5. This suggests that treatment with phosphate may have facilitated conversion of chromium to less soluble speciation, while cement-based treatment of APC residue facilitated the formation of more soluble chromium species. For lead, treatment of APC residue resulted in decreased solubility between pH 5 and 11 while treatment of bottom ash had limited effect. The solubility of an element or species of interest also can significantly effect the testing protocols required for estimating diffusion controlled release. A critical assumption in the tank leaching protocols is that the leachant remains dilute with respect to species of interest during each leaching interval. Figures 20.21 and 20.22 present the measured leachate concentrations for copper, cadmium, lead and zinc from several studies. The unified pH solubility curves for bottom ash (see Chapter 16) are provided for comparison. For the case of copper, all observed leachate concentrations were significantly less than the unified solubility curve. However, for most data sets, the observed leachate concentrations of Cd, Pb and Zn are below or equal to the solubility curve. In a few cases, data above the solubility curve have been measured. In these cases, the dilute solution criteria have not been met and the release data have to be interpreted with caution. Future testing for these cases should be carried out either with increased liquid to solid surface area ratios or with shorter leaching intervals. If release is controlled by solubility, other means of assessing long term release are needed. The use of release assuming an effective diffusion coefficient based on the highest pDe measured during testing is a worst case assessment of release. For the purpose of evaluating acceptability of release, this may already be adequate in view of the limit values concerned.

20.6 INTEGRATED INTERPRETATION OF pD,, AND AVAILABILITY Reductions in constituent release from a treated material can be achieved either through reducing the fraction of that constituent available for release, through reduction of the rate of release (increased pDe), or through modification of both critical parameters. In addition, the PDe can be modified either through physical or chemical effects of treatment. A useful mechanism for evaluating the combined effects of both availability and pD~ is needed. Estimation of diffusion controlled release from specific geometries and conditions over prolonged time periods can be accomplished through use of intrinsic leaching parameters and application specific geometries and exposure conditions. An important advantage to the use of intrinsic leaching parameters derived from the availability and diffusion release test data is the ability for prediction of release under conditions other

879

Figure 20.21 A Comparison of Tank Leaching Concentrations for Various Products as a Function of Solution pH and the Unified pH-Solubility Leaching Curves for Cu and Cd

Cu 10

9 r o

A

0.1 U C

o

US FA P3 US BA

9

O

O

G)

9

o.ol

e@

9

9

u

US BA

9

o.ool 0.0001

=

'

6

',

'

7

', .'

8

',

,

9

I

,

I

,

t

,

I-

,

P1

US FA P2 9

9

P4

US BA P9

A

...I

Unified pH MSWI BA

NL ASPHA 2% BA

I

10 11 12 13 14 pH

Cd

0.1 E c o

== u r o

0.01

9 ,&

0.001

9

0.0001

0.00001 6

7

8

9

10 11 12 13 14 pH

Unified pH MSWI BA France FA NL ASPHa 2% FA

880 Figure 20.22 A Comparison of Tank Leaching Concentrations for Various Products as a Function of Solution pH and the Unified pH-Solubility Leaching Curves for Pb and Zn

Pb A

NL asph Fa 9

0.1

0.01

0.001

R--

NL concrete BA

o

Unified pH MSWl BA

<>

US BA P1

9

US FA P1

9

US FA P2

9

US FA P3

9

US BA P4

m

0.0001

'

5

I

'

I

6

'

7

I

'

8

I

'

9

:

"

~

"~.-;

"

;

"

;

10 11 12 13 14

pH

US BA P9

9

DK FA

9

France FA

Zn 10 A

NL asph Fa 9

ol

E t-

u c o

-o

Unified pH MSWl BA

<>

US BA P1

9

US FA P1

0.1

o

0.01 O

0.001 9 4NIb

0.0001

9

~

6

"

',

7

"

:

8

"

I

9

'

I

10

pH

'

I

11

9 '

I

12

'

I

13

'

I

14

NL beton BA

a

9

US FA P2

9

US FA P3

9

US BA P4 US BA P9

881 than those studied in the laboratory. Many factors affect the translation of laboratory results to prediction of field behaviour. Field environmental conditions that are important include residue aging, contact with infiltration and precipitation frequency, temperature cycles, direct abrasion or erosion and the specific application scenario. Thus, estimates of field releases must be carefully derived. However, simplified models can be used to indicate relative releases and provide order of magnitude or limit case assessments. The availability and pDe leach parameters can be used to predict the release of contaminants during a given time period for a variety of application geometries. A 3-dimensional diffusion model enables one to take actual dimensions into account, so differences in leaching from a product with a cubic versus a flat rectangular shape can be described. With the 3-D model, release from only one side of the material also can be modelled. A 3-dimensional model is based on the analytical solution of the linear diffusion from a parallelepiped, which initially is at a constant concentration, to an infinite region outside with a constant surface concentration (Crank, 1989; de Groot, 1993). The diffusion profile is calculated in all three dimensions according to the equation: C0

13.3 ,=0 m.o n.o

XGOS

(2/+1)(2_m'-~2n+11

cos (21+1)nx 2a

(2m+ 1)nYcos (2n+ 1)nz exp(-tal, m,n) 2b 2c

where:

(20.26) a,,m,n=-~t 2~.--~1)2+(2% +1

+ 2n+1 2t

'C"

(20,27)

Integration of the constituent flux across the surface boundary with respect to time results in an expression for calculation of cumulative release (Crank, 1989):

Mt=fo

-n---~- p

~176 ~0 ~o (2/+1)2(2--~+1i2(2n+1)2

Application of Equation 20.10 permits estimation of the cumulative release of a constituent as a function of time. The cumulative release, expressed as fraction of the total leachable quantity (Rma,), can be calculated using the 3-D model for different product configurations and bulk applications based on the effective diffusion coefficient

882 measured during laboratory testing or in the field under well-defined boundary conditions. The usual boundary condition applied for field translation is that the surface concentration of the leaching constituent is effectively zero. The cumulative release, expressed as fraction of the total leachable quantity (Rm~), can be calculated using the 3-D model for different product configurations and bulk applications based on the effective diffusion coefficient measured under well-defined boundary conditions. The relative release from standard sizes with dimensions of 10 x 10 x 10 cm and 15 x 15 x 45 cm were calculated as a function of time for different effective diffusion coefficients ranging from PDe=9 to pDe=15 (PDe = -log De with De in [m2/s]). The cumulative release-time curves are provided in Figure 20.23. Between blocks of increasing size, the difference is largely a shift of the cumulative release for a given pD e to a longer timescale. It takes longer to reach the maximum leachable quantity, but the leachable quantity may ultimately be reached unless the chemistry or other release controlling factors change. A significant shift in the cumulative release curve is apparent for the roadbase simulation. For a base of 15 cm thickness, 50% of the highly mobile components (PDe=9) will be leached from the slab in less than approximately one year, assuming permanent contact with water. A 45 cm thick slab will reach the same level of relative release in about 6 years. It is important to consider that translation of lab data to field conditions further involves several factors such as corrections for the ambient temperature and degree of contact with water. Estimating release during utilisation must consider adjustments to the pD measured in the laboratory to reflect anticipated field conditions (Kosson et al., 1995). Temperature, the fraction of time the surface is wetted, and the degree of water saturation are important considerations. While the diffusion coefficient is a function of the diffusivity of the constituent of interest in water, tortuosity, and chemical retention, only diffusivity (Do) is significantly a function of temperature. The temperature dependence of diffusivity in dilute ionic solutions can be considered to be proportional to the absolute temperature over limited temperature ranges [20.28], e.g.,

DT2-

OTIT2 71

(20.29)

The above relationship assumes that the viscosity of the pore water (leachate) does not change significantly over the temperature range of interest. Alternatively, the effect of temperature on D has been correlated for release from cement stabilised products containing waste materials according to [20.29]:

PD1-PD2=0"71 '-~1-

(20.30)

The cumulative release for a series of wet dry cycles can be approximated based on Equation 20.24 by.:

883 Figure 20.23 Estimation of Fractional Cumulative Release as a Function of Time and Effective Diffusion Coefficient (pDe)

MONOLITHIC Block

.f

= o n-

~ E

= (3

0.80

~

~

.-

--

MATERIALS Block

cm

-~ ~-

"" /I / ..."~:.1= ",o-Q~/ ,," i11/ ..."'/

1.00

~= n" E

lOx10x10

CONSTRUCTION

. ..

~: ~r" n"

0.80

0.60

-/

0.20

,'

/

.."

../

r

.-"

-"

o.oo ~

10-2

/

.

'

~

10-1

/ ,a,.,a,.:

/

_ /'

'

~.- - ~ . . . . . . . . .

100

101

102

~

0.40

E

0.20

= (3

/

/

,

~.~ . ~ 10-1

10 .2

Time [years]

DIFFUSION

COMPACTED

/

/

0.60

/

0.40

// ~:>=,o'

0.20

/ /

1 )'2

//

/

10 "1

'

/

//

'

/

,//

/ / ~>=11// f /

100

/

/

~

/

I

..

)

9

/

."

."

."

....'"" .

/

."

/~=,..=

/ / /,,>/,,

,

100

.~.~..~.~. 101

102

103

101

Time [years]

/

/

/

//

/

/

//

/

/

/

/

=

~>,.12 .

/

102

cm

/

."" =0,.,3

...'"

45

1.00

," "

.'"

LAYER

Roadbase

f f---:: ~:>-,e /

/

GRANULAR

15 cm

1.00

0.80

/

//~=11/--I..-"

Time [years]

CONTROLLED

Roadbase

/

/

I

/r / .'" / /

0.00

03

,

//

,'

/

e=

/

/

/ /

e

,

cm

1.00

0.60 0.40

15x15x45

I

103

E n" n" ,,._. r e) r _.e rr

0.80

//,"' /

0.60

~>

0.40

E (~

0.20

0.00

10 .2

~=111

11//

10-1

100

101

Time [years]

102

103

884

Mr,,,1d

(20.31 )

or -o-,

=

=MtF,,~d

(20.32)

Note that the calculation of IVItassumed a continuously water saturated material. Figure 20.24 illustrates the effects of temperature and wet/dry cycles on diffusion controlled release from a 45 cm thick roadbase (Kosson, 1995). The cumulative effect of these conditions can be significant over long time intervals. Calculation of cumulative release curves for a variety of geometries and applications is impractical on a routine basis. A simple one-dimensional diffusion model, assuming a constant source, can be used as an approximation. The advantage of this approach is that cumulative release is only a function of availability, pDe and the exposed geometric surface area. The one-dimensional model is independent of application specific geometry. This approach is valid as long as the concentration in the material has not decreased substantially, avoiding species depletion. If depletion does occur over the time period of interest, the one-dimensional model will over predict release, providing a conservative estimate for decision making. However, the amount of a specific element released will not exceed the availability of that element for all cases. The one-dimensional model is based on the "semi-infinite slab" solution of the diffusion equation provided by Crank (1989):

M,=2pCo (-~) ''2 where: Mt Co D t r

= = = = =

(20.33)

cumulative release [mg/m 2] availability [mg/kg] effective diffusion coefficient [m/s 2] time [s] density [kg/m 3]

This is the same basis which was used to estimate the pD from tank leaching data. The initial and boundary conditions for this solution are (i) the initial constituent concentration is uniformly distributed in the matrix; (ii) the exposed surface for leaching has a liquid concentration which is maintained essentially at zero; and (iii) depletion does not occur. Figure 20.25 provides a comparison of cumulative release estimated using the three-dimensional model and the one-dimensional model applied to a 10 x 10 x 10 cm block and a 45 cm thick roadbase (one surface exposed) for two different

885 values of pDe. The effect of depletion on release is indicated by the greater predicted release by the one-dimensional model. However, in all cases the release estimated by the one-dimensional model will be equal to, or greater than the release predicted by the three-dimensional model. This confirms that the one-dimensional model provides a conservative estimate of release, provided that no significant changes in chemistry occur. Figure 20.24 The Effect of Intermittent Wetting and Temperature on Diffusion Controlled Release of Product Constituents

Road base 0.45 m

1.20

X

1.00

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

pD= 10, Sat., 20 ~

E r~

0.80

(D (/) (1)

0.60

pD= 10, Sat., 10~

-

_

pD= 10, Wet/dry, 10 ~

x

(D

rY .O >

0.40

l:::

0.20

~

/

/

o 0.00 -"- " ~ 0.1

--I--I-1

r'l'r'l--

"-

1

T

~

~ ~

~i

I

i

,

10

Time (years)

i

,~

,~

I

100

i

i

~ ~,

,~

1000

886 Figure 20.25 A Comparison of Fractional Cumulative Release for Products Estimated Based on the "Infinite Slab" Approximation and Accounting for the Product Geometry and Depletion Block 10x10x10 cm

1.20 . . . . . .

I

x

E E

rv" n,'

v

!

1.00 ................ ~.................

0.80

iIIll ,

pD=10 / /

I

(D (D n" ~

E

fO

0.60

iJ /

O 0. 0.20

~

j'

i

pD=14

, I

" .t

0.00 0.01

~ 1

0.1

' 100

10

1000

Time (years)

1.20 X

E E

n,, n,"

v I

1.00

Road base 45 cm

..................................................

0.80

~-.......

I

(D

(D n"

0.60 0.40

E O

./

0.20 0.00 0.001

0.01

0.1

1

10

Time (years)

/ 100

1000

887 Interpretation of the availability and pD information can be further simplified if a fixed time interval is defined over which to consider release. For example, a 100 year interval may be considered as the useful life for roadbase or other construction applications. Use of the one-dimensional model permits the development of charts which provide estimated cumulative release per unit area as a function of only availability and pD. These charts are referred to in this text as "cumulative release nomographs". Examples of cumulative release nomographs are provided for several elements in Figures 20.26 through 20.28. Availability and pD are presented on the x-axis and y-axis, respectively. Diagonal lines indicate lines of constant cumulative release (e.g., 10, 100, 1000, etc. mg/m 2 over a 100 year interval). Lines of constant cumulative release decrease in value from lower right to upper left of each figure. Data plotted above and to the left of a line of constant cumulative release provide less release than the indicated value; data below and to the right provide greater release than the indicated value. Cumulative release nomographs provide a straightforward method of interpretation for evaluation of data from laboratory testing of diffusion controlled release. An acceptable cumulative release can be defined for specific applications or locations based on evaluation of the potential impacts of that release. Data for applications which result in release less than the defined limit would be considered acceptable. Thus, diffusion controlled release information can be readily incorporated into a regulatory framework. Consideration of the cumulative release nomographs presented in Figures 20.26 through 20.28 also provide information about the effectiveness of various treatment processes and products for incinerator residues. Observations are summarised in the paragraphs that follow. Cadmium (Figure 20.26) Cumulative release for bottom ash incorporated into cement pavement blocks was between 0.1 - 1.0 mg/m 2. SIS treated bottom ash and fly ash incorporated in asphalt had release between 1.0 - 10 mg/m2. Untreated APC and SIS treated residue typically had cumulative release > 10 mg/m 2.

Copper (Figure. 20.26) Cumulative release for fly ash in asphalt, bottom ash in paving blocks and residue from the 3R process was between 10 and 100 mg/m 2. Untreated and SIS treated APC residue cumulative release was between 100 and 1000 mg/m 2. Untreated and SIS treated bottom ash and combined ash generally had cumulative release in excess of 1000 mg/m 2. Lead (Figure 20.27) SiS treated bottom ash, bottom ash in cement pavement blocks and fly ash in asphalt generally had cumulative release less than 100 mg/m 2. Untreated bottom ash,

888 Figure 20.26 Availability-pDe Plot of Release Parameters for Cadmium and Copper from Tank Leaching Tests on Untreated Incinerator Residues and Several Product Materials Containing Residues

1~l

/o.,

"F

/.

15

/".I:.

l"r ~

9

14

/.

/

l~ / ,

9

9

13 0.01

/,

,

e e

~

I !

9

0.1

i

10

100

1000

Availability (mg/kg) 16

~.~10

/,

15

,,,-,

14

,~. , , ~

§

j

s.

13

/100

..

./v

_p......

q) a a-12

11 10

[ j/..V , ooo /,oooo .

1

.

.

.

.

.

.

.

10

.

.

.

.

"

9

.. .. .. .. .. .. .. .. .. .. .

100

9

I

.

Ou cu

.

.

.

1000

Availability (mg/kg) Diagonal lines represent constant cumulative release (mg/m 2) estimated for 100 yrs using the "infinite slab" approximation

889 Figure 20.27 Availability-pDe Plot of Release Parameters for Lead and Zinc from Tank Leaching Tests on Untreated Residues and Several Product Materials Containing Residues

19/

,~1o /

18tPb

j ~oo s O " "~'-'.kd

17

1~

15

o.

9

13

~

|

I

12

~

11 10 9

1000 1

.

10000

10

-100000 1 O0

Availability 18

1000

10000

(mg/kg)

Zn / ' ~ y,/'~176 / '~176176 / /...'o / /

17 cI e~

/'

16

/

"/

/.;.. "J. ~

/

o0

15

14

/.

......

13 10

/,

77"-

100

.....

Z

1000 Availability

,o. ooo . . . . . . . . . . . . . . 10000

100000

(mg/kg)

Diagonal lines represent constant cumulative release (mg/m 2) estimated for 100 yrs using the "infinite slab" approximation

890 untreated combined ash, SIS treated combined, SIS treated fly ash, and fly ash in asphalt had cumulative release between 100 and 1000 mg/m 2. Untreated APC residue and SIS treated APC residue generally had cumulative release in excess of 10,000 mg/m 2. Phosphate treated APC was the exception to this with cumulative release between 100 and 1000 mg/m 2. Zinc (Figure 20.27) Bottom ash in cement pavement blocks had cumulative release between 10 and 100 mg/m 2. SIS bottom ash, SIS combined ash and fly ash in asphalt had cumulative release between 100 and 1000 mg/m2. Untreated bottom ash, combined ash and APC residue, and SIS treated APC residue generally had cumulative release in excess of 1000 mg/m 2. Sodium (Figure 20.28) Bottom ash in cement pavement blocks had cumulative release of approximately 10 mg/m2. Bottom ash in asphalt had release between 100 and 1000 mg/m2. All untreated residues and other treatment conditions had release in excess of 1000 mg/m 2. Several processes resulted in cumulative release in excess of 10,000 mg/m 2. It is important to recognise that because the PDe for sodium is less than 10.5 for most cases, the majority of the cumulative release will occur during the first ten years of application with subsequent depletion (see Figure 20.24). Chloride (Figure 20.28) Bottom ash in asphalt, fly ash in asphalt and bottom ash in cement paving blocks all had cumulative release of approximately 100 mg/m2 or less. Untreated and SIS treated bottom ash and combined ash generally had cumulative release between approximately 10,000 and 100,000 mg/m 2. SIS treated APC residue and fly ash had cumulative release in excess of 100,000 mg/m 2. As with sodium, the majority of the release will occur during the first few years of application. Figure 20.29 summarises the affects of various treatment processes on diffusion controlled release from incinerator residues. In general, incorporation of residues into asphalt decreases release through increased physical retention. Therefore, the process can be effective for both highly soluble salts and trace metals. Treatment of residues by SIS processes decreases release through chemical respeciation and modification of the matrix alkalinity. Therefore, SIS processes may be effective for reducing release of trace metals but are generally ineffective for reducing release of highly soluble salts, although release may be delayed. SIS processes also may be ineffective when highly soluble salt Ioadings are so great, as in the case of APC residues, that a highly porous matrix remains after depletion of the salts. Availability of all of the residue constituents may be reduced by any of the above processes by having limited residue loading in the final product (dilution effects).

891 Figure 20.28 Availability-pDe Plot of Release Parameters for Sodium and Chloride from Tank Leaching Tests on Untreated Residues and Several Product Materials Containing Residues 13

12

11

l::l

10

9

8

J,.oo:. .....

7 1 oo

"..a ........ ' 1 o00o

10oo Availability

12

11

/

'

'

" 1 ooooo

(mg/kg)

100

100

9 ,~

s Ill

10

/-

Ei

7 100

/

T/". ~,4

,.oooo..l ,,!oo7o

...... I

.............

1000

10000

Availability

100000

1000000

(mg/kg)

Diagonal lines represent constant cumulative release (mg/m2) estimated for 100 yrs using the "infinite slab" approximation

892 Figure 20.29 A Schematic Illustration of Treatment Process and Environmental Conditions Effects on the Availability-Effective Diffusion Coefficient Relationship

18

Poresealing Tortuosity increase by cementation Low effective watercontact Lower temperature

17 16

Decreasing availabilih/ by vitrification

15

(b C~ O_

14 13 12

Dilution with additives Matrix incorporation by recrystallisation

n

11 10

m

m

I I0

Desorption by remineralisation Less retention due to acidification I I O0 1000

Availability (mg/kg)

10000

893 REFERENCES

American Nuclear Society Standard Committee Working Group ANS 16.1. "American National Standard Measurements of the Leachability of Solidified Low-Level Radioactive Wastes by a Short-Term Procedure" American Nuclear Society, La Grange Park, IL, 1986. Barna, R., P. Moszkowicz, J. Mehu, H.A. van der Sloot. "Waste Solidification Modeling and Simulation of Sodium Chloride Release from Leached Concrete", Int. Symp.Waste_ ManaQement 1992. Praag, pp. 65- 68, September 1992. Batchelor B. and K. Wu. "Effects of Equilibrium Chemistry on Leaching of Contaminants from Solidified/Stabilised Wastes", In Chemistry and Microstructure of Solidified Waste Forms. Edited by R.D. Spence. Lewis Publications, Boca Raton, FL, 1993. Crank, J. The Mathematics of Diffusion, Oxford University Press, New York, 2nd edn., 1989. de Groot, G.J., D. Hoede. Verfiinin.q van de beschrijvin.q van de .uitlo,q.in~ van resstofprodukten, ECN-C-93-085. ECN publication. 1993. (3 D model) de Groot, G.J., H.A. van der Sloot and J. Wijkstra. "Leaching Characteristics of Hazardous Elements from Coal Fly Ash as a Function of the Acidity of the Contact Solution and the Liquid/Solid Ratio", In Environmental aspects of stabilization and solidification of hazardous and radioactive wastes, ASTM STP 1033, P.L. Cbte and T.M. Gilliam, Eds., American Society for Testing and Materials, Philadelphia, 1989, pp 170-183. de Groot, G.J., H.A. van der Sloot, P. Bonouvrie and J. Wijkstra. "Karakterisering van het uitlooggedrag van intakte produkten", Mammoet deelrapport 09, March 1990. Eighmy, T.T., D. Crimi, S. Hasan, X. Zhang, D.I. Gress. "The Influence of Void Change, Cracking and Bitumen Aging on Diffuional Leaching Behavior of Pavement Monoliths Constructed with MSW Combustion Bottom Ash", In: Proc. of the 74th Transportation Research Board Meeting, Washington, D.C., Jan. 1995. Hinsenveld, M. "Towards a New Approach in Modelling Leaching Behaviour", Waste Materials and Construction. Edited by J.J.J.M. Goumons, H.A. van der Sloot and Th.G. Aalbers, Elsevier, Amsterdam, 1992. Kosson,D.S., T.T. Kosson and H. van der S l o o t . Evaluation of Solidification/Stabilization Treatment Processes for Municipal Waste Combustioq Residues, NTIS PB93-229 870/AS, 1993. Milieutechnisch onderzoek AVI slakken toepassing Rijksweg 15. Rijkswaterstaat, Dienst Weg en Waterbouwkunde. Report D 0421-71-001. July 1992.

894 NEN 7341. Leaching characteristics of soil-, construction materials and wastes Leaching tests - Determination of the availability of inorganic constituents for leaching from construction materials and waste materials. NNI (Dutch Standardization Institute, Delft), 1994. NEN 7345. Leaching characteristics of soil-, construction materials and wastes Leaching tests - Determination of the release of inorganic constituents from construction materials, monolithic wastes and stabilized wastes. NNI (Dutch Standardization Institute, Delft), 1994. van der Sloot, H.A., and D. Hoede. AVl-bodemas als aanvulmateriaal. ECN-c-91-044, 1991. van der Sloot, H.A., G.J. de Groot and J. Wijkstra. "Leaching Characteristics of Construction Materials and Stabilization Products Containing Waste Materials", In: Environmental.Aspects of Stabilization and Solidification of .Hazardous and Radioactive Wa.stes, ASTM STP 1033, P.L. Cote and T.M. Gilliam, Eds., American Society for Testing and Materials, Philadelphia, 1989. van der Sloot, H.A., G.J. de Groot, J. Wijkstra, and P. Leenders. "Leaching Characteristics of Incinerator Residues and Potential for Modification of Leaching",in Proceedin.qs International Confer.ence on Municipal Waste Combustion.., Hollywood, 1989. van der Sloot, H.A. "Systematic Leaching Behaviour of Trace Elements from Construction Materials and Waste Materials", in Waste Materials in Construction~ Proceedin(:]s of the .International Conference on Environmental. Implications of Construction with Waste Materials, Elsevier, Amsterdam, 1991. Whitehead, I.E. "An Enviromental Evaluation of Bottom Ash Substitution in Pavement Materials", Master Thesis, University of New Hampshire, Durham, New Hampshire, 1992.

895

C H A P T E R 21 - U T I L I S A T I O N 21.1

INTRODUCTION

Utilisation of MSW incinerator residues is being conducted or considered for a variety of applications in many countries. Interest in utilisation principally is motivated by the potential for extending existing ash landfill capacity, and thus reduces disposal costs, and in some regions the substitution for natural aggregates. The relative importance of each of these factors varies considerably from country to country and between regions within a country. Primary applications include use as: an aggregate substitute in paving applications, including as compacted base, or in bituminous pavement, an aggregate in terrestrial Portland cement applications, including cement block and prefabricated or field erected forms an aggregate substitute in Portland cement-based marine applications such as artificial reefs and shoreline protection 9

daily cover for municipal waste landfills, or

9

granular fill material for embankments.

Bottom ash is the primary material being utilised or considered for utilisation in construction applications. However, there are some notable exceptions, for example, a small percentage of fly ash has been used as a fine aggregate filler in asphalt in The Netherlands and the use of combined ash has been considered in the United States. Almost all of these applications, except use as daily landfill cover, would involve some degree of ash processing, either physical and/or chemical. For example, most applications would require screening of ash to achieve a desired particle size gradation or would result in ash encapsulation in another matrix. Utilisation through recovery of chemical constituents (e.g., CaCI2) and recycling of ferrous and nonferrous metal is discussed in Chapter 17 (separation processes). A typical pavement consists of the following layers or a subset combinations of layers depending on design (listed from the top driving surface down): a shim/levelling course, a wearing~surfacecourse, a binder course, a base course, a sub-base course, a compacted subgrade, and a natural subgrade. The shim/levelling course is placed on the surface to level ruts and depression and typically consists of a fine grain sand. The wearing~surface course is the top 1 to 5 cm and the binder course is below the binder course. The binder course serves as the bottom portion of the roadbed if

896 needed. Otherwise, it will be placed between the wearing course and the base course. The base course is normally the lower portion of the pavement. However, a sub-base may be required and is place directly below the base. The pavement is built from the bottom to the top on a subgrade that has been prepared by compaction. The entire roadbed is placed on natural subgrade or fill. Applications in the marine environment include shoreline protection and artificial reefs. These involve the use of the residues mixed with Portland cement to form concrete structures. Shoreline protection is the process of creating physical resistance to disruptions, such as storm events, natural erosion, and boat wakes. Examples of shoreline protection are bulkheads, sea walls, breakwaters, jetties, and piers. Artificial reefs are constructed to provide structures for the growth of marine organisms and attraction of fish while additionally serving as shoreline protection. All of the above applications require a final product which has a high degree of physical durability. 21.2 CURRENT AND PLANNED PROJECTS

The following sections provide a summary of current and planned utilisation projects and testing programs in several countries at the time of writing. 21.2.1 Canada

Because of the availability of landfill and the modest production of incinerator ash, there is little incentive, either economic or environmental, to pursue ash utilisation applications in Canada. The only example of ash utilisation is the use of bottom ash from the Burnaby Incinerator facility to construct access roads within a landfill. As utilisation becomes more wide spread in other countries, it is expected that the practice will be considered more seriously in Canada. 21.2.2 Denmark

Danish incinerator facilities produce approximately 420,000 tons of bottom ash annually (including grate siftings and boiler ash which in most cases are mixed with the bottom ash) the overwhelming majority of which is utilised. Since 1974, screened and sorted bottom ash has been utilised in Denmark for civil engineering purposes, particularly as subbase material at parking lots, bicycling paths, and paved/unpaved residential and major roads (Hjelmar, 1992). As a subbase material, the bottom ash is usually substituted for the diminishing supplies of natural gravel in various parts of Denmark. Thus, the incentives for utilisation include both natural resource conservation and economic benefit. A substantial portion of the economic benefit is derived from the avoided costs of landfilling, which is typically $150 (US) per tonne.

897 21.2.3 Germany In 1991, the 48 German incinerators in operation produced 2.56 million tonnes (Tg) of bottom ash. About one half of the annual production was utilised (1.2 Tg) and almost 0.2 Tg of scrap ferrous was recovered for recycling. The remaining amount (about 1.2 Tg) was landfilled (Johnke, 1993). The bottom ash is sieved and ferrous metal is removed prior to utilisation. The primary uses reported were as aggregate for compacted roadbase and embankments, primarily in demonstration projects. The extent of utilisation varied considerably from state to state. There remains strong opposition against the use of bottom ash even though positive results have been obtained from demonstration projects. Following an inquiry of 176 municipalities with >10,000 inhabitants, only 6 make use of bottom ash on a regular basis, with another 11 making use of it on a limited basis. Air pollution control residues, including fly ash and scrubber residues, are undergoing pilot-scale and provisional evaluation for use in the coal mining industry as filling and sealing materials for excavation cavities, and as aggregate substitute in grouts (Plate, 1992). The hard coal mines of Ruhrkohle AG (RAG) consume approximately 1.5 million tonnes of grout per year and the potential capacity to incorporate APC residues in grouts is approximately 20,000 tonnes/yr. Approximately 50,000 tonnes of APC residue has been used for a pilot test of mine filling and sealing operations. Scrubber residues also are being considered for use in alinite cement (Oberste-Padtberg, 1992). In this case, scrubber residue is used as a substitute for lime as a raw material. Scrubber residues are pelletised with other raw materials and about 20% water, and then treated in a rotary kiln to form an alinite cement clinker which is subsequently ground into cement. 21.2.4 The Netherlands Approximately 600,000 tons of bottom ash and 80,000 tons of fly ash are produced annually (1988) in The Netherlands (Born, 1994). Ferrous scrap, representing about 70,000 tonnes/yr, is separated magnetically at all facilities and recycled in the steel industry. The government policy is to achieve utilisation of more than 80% of incinerator residues. In practice, approximately 95% of the bottom ash is currently utilised (Born, 1994). The principal motivation for utilisation in the Netherlands is the shortage of suitable natural aggregate and the lack of available landfill space. The primary use of bottom ash is in the following applications: road base material for roads and industrial sites material for embankments, noise and wind barriers aggregate in concrete and concrete products, and aggregate in asphalt concrete

898 A total of more than 2 million tons of bottom ash has been used in the listed applications. In addition, 30-40% of the fly ash produced since 1984 has been used as a fine aggregate filler in asphalt concrete. The following is a summary description of several major ash utilisation projects: Caland Wind barrier- This project carried out in 1985 used more than 650,000 tons of bottom ash in an embankment with a length of 700m and a height of 15m. The ashes are covered with a primary cover layer of 0.5m compacted clay with a sand drainage layer (0.5m) and top soil (1.0m) overlaying the clay layer. The slope of the compacted ash is 1 2 and 1:2.5. The mean compaction factor was 97.5% with a wet density of 1840 kg/m 3. Groundwater quality is monitored on both sides of the embankment. Highway A-15 Rotterdam - Approximately 400,000 tons of bottom ash was used in an embankment for this major roadway construction. Ash is covered with a compacted sand-bentonite mixture with a minimum thickness of 20 cm to reduce water infiltration. Road Base Material - Several projects with bottom ash as road base material have been carried out in Rotterdam and North Holland. In Rotterdam, primarily a mixture of ash, crushed rubble aggregate and additive (50%-50%-10%) has been used. The base thickness is 25-30 cm. Concrete Paving Blocks - A project in Keilehaven carried out in 1984 used more than 300,000 concrete paving blocks in which the coarse aggregate was replaced up to 40% with the 5-8 mm fraction from bottom ash. After five years of traffic, it was concluded that there was no difference in physical properties between standard concrete paving blocks and those using bottom ash as the coarse aggregate replacement (Leenders, 1988). Laboratory investigations of environmental properties of concrete paving blocks with 20% replacement of the coarse aggregate by bottom ash indicated no significant difference from conventional paving blocks (KEMA, 1986).

Hartel Canal Pilot Project - A pilot project with asphalt containing bottom ash was carried out in 1987 along the banks of the Hartel Canal. A length of approximately 50 m was coated with about 100 tonnes of asphalt containing 30% bottom ash. It was found that the mixture temperature was required to be 40~ higher and the bitumen content 3% higher than traditional material. 21.2.5 Sweden

Current utilisation of incinerator residues in Sweden is very limited due to the regulations for licensing the use of any residual products in a specified manner (under the Environmental Protection Act). This also applies to controlled tipping.

899 Uncertainty with regard to whether or not a license is obligatory under the Environment Protection Act for the utilisation of residues has resulted in essentially no full-scale utilisation of incinerator residues through 1993. However this uncertainty is supposed to be resolved within a short time frame (e.g., 2 years) with the overall intent to increase utilisation of waste materials. New regulations are likely to require quality control based on environmental parameters. It is unclear if incinerator bottom ash will be deemed acceptable since these parameters have not been established. The legal, environmental and engineering aspects of utilising incinerator bottom ash have been studied in a comprehensive project (Lundgren & Hartl~n, 1991 ). The results based on field and laboratory tests indicated that sorted bottom ash can be used as embankment fill, base course material in low traffic roads and under light buildings and floor structures. Recommendations included that the use of bottom ash should be restricted to applications where the ash is covered by a low permeability material, such as asphalt, and deposited away from the groundwater table or ground water catchment areas. It was also recommended that the thickness of a deposit should be limited to 3.0 m and the ash placed well above the ground water table until further experience is gained. 21.2.6 United States

In the United States, several factors have influenced the possible uses for incinerator residues in construction applications. The shortage of existing landfill space and the difficulty of securing new sites has created a situation in which either disposal fees are costly, disposal space must be sought in distant locations, or disposal will not be possible. Hence, recycling and reuse of residual wastes has been suggested as the preferred management option. A detailed summary of planned and ongoing demonstration projects utilising municipal solid waste incineration residues in the U.S. has been prepared by Hoffman (1993). The following paragraphs and Table 21.1 highlight summary information for project parameters. Type and Distribution of Application: Approximately half of the 23 identified projects utilising residues are concrete applications. Of those 11 projects, 5 involve using residues in concrete blocks for buildings, 4 marine application, and 2 are used for landfill functions. Six of the remaining 12 projects are asphalt road paving applications. The most common initially proposed was in the road wearing surface. Project plans have been substantially modified in several cases, resulting in more road sub-surface demonstrations. Other projects include utilisation of loose MSW residue aggregate as the base layer underlying a paved parking lot, commercial scale substitution of fly ash for a raw material in cement production and fill material for inactive salt mines. Plans for the remaining three projects do not contain specific identification of use.

900 Table 21.1 Summary of Incinerator Ash Utilisation Projects in the United States Project MSWl Facility Ash Type Type .Asphaltic Applications Mass Burn/DS-FF Combined ennepin County, Minnesota Pavement emonstration Hillsborough County Department of Solid Waste Municipal Incinerator Ash Reuse, Research Development and Demonstration Project, Florida

Mass Burn/ESP

Combined

McKaynite Demonstration, Acline Street, Florida

Mass Bum/ESP

Combined

McKaynite Demonstration, Ruskin, Florida

Mass Bum/ESP Mass Bum/DS-FF

Bottom Bottom

New Hampshire Bottom Ash Paving Project

Mass Burn/DS-FF

Bottom, Combined

NYSERDA- Phase Ila, New Jersey

Mass Burn/DS-FF

Bottom, Combined

Concrete/Cement Ash Management Building, OH (Montgomery County)

Mass Burn/ESP

Bottom

Center for Innovative Technology, VA

Mass Burn/DS-FF

Bottom, Combined

Commerce Refuse-to-Energy Ash Treatment and Reuse, Los Angeles County, CA.

Mass Bum/DS-FF

Combined

Fly Ash Stabilisation Building, OH (Montgomery County) Islip, Blydenburgh Landfill, Long Island N.Y.

Mass Bum/ESP

Bottom

Mass Bum/DS-FF

Combined

Pinellas County, Florida Artificial Reef

Mass Bum/DS-FF

Scrubber, Bottom

Residential Foundation, New York

RDF/ESP

Bottom

SEMASS Administration Building Project, MA

RDF/DS/ESP

Bottom

SUNY Artificial Reef Demonstrations

Mass Burn/ESP Mass Burn/DS-FF

Combined Combined

SUNY Boathouse Demonstration

Mass Burn/ESP

Bottom, Combined

Other Commercial Grade Cement Production (Tacoma, Washington) (shale replacement)

RDF/FF

Fly Ash

City of Albany, NY, Parking Lot Demonstration (loose aggregate)

RDF/ESP

Bottom

Hawaii: Field Tests of Use as Landfill Cover

RDF/ESP

Bottom

Metropolitan Washington, D.C. Demonstration

Mass Bum/DS-FF

Bottom, Combined

NYSERDA - Phase II B - New York City

Mass Burn/DS-FF

Bottom, Combined

RDF ESP

Refuse Derived Fuel Electrostatic Precipitator

DS FF

Dry Scrubber Fabric Filter

901

Geographic Distribution: The majority of projects cluster in two areas - Northeastern United States (New England, N.Y., and N.J.), and Florida. Additional locations include Los Angeles County, California and Honolulu, Hawaii. Factors influencing this distribution are lack of landfill space, availability of natural aggregate, and vendor and government interest in utilisation. Type of Incineration Facility: All but five of the sources of incinerator residue for the demonstration projects are mass burn facilities. Of those five, four employ refuse derived fuel (RDF) in a conventional boiler. The fifth also uses RDF, but co-combusts it with coal and wood waste in a fluidised bed burner. Slightly more than half of the incineration facilities supplying residue for the projects are equipped with a scrubber and fabric filter combination. One of this group is equipped with an additional lime injection system in the furnace, and one is fitted with a de-NOx treatment system. Eight are equipped with ESPs only.

Incinerator Residue Fraction: Six projects propose using only bottom ash, and one uses only fly ash or APC residues. The others either use bottom and combined ash in the experimental design or are debating which residue streams to employ. The trend is toward bottom ash use, rather than combined ash. Incinerator Residue Processing: In each project, the incinerator residue is processed to some degree before it is substituted for natural materials in asphalt and concrete media. Pre-combustion processing is employed in RDF: large non-combustible items and ferrous metals are removed, and waste is shredded. Post combustion processing frequently consists of ferrous and nonferrous metal removal and particle size control. In addition, stockpiling (aging) is done to improve the engineering performance of the residue. Because virtually all of the projects substitute incinerator residue for natural aggregate, efforts are made to supply the replacement material in a form as similar as possible to the natural material. Some companies have patented their aggregation processed and have applied trade name to their products (e.g., Ardellite, Boiler Aggregate, McKaynite, Permabase Plus, Rolite, etc.). A range of 5% to 90% substitution for natural aggregate in asphalt and concrete applications exists among the projects.

21.3 CURRENT REGULATORY FRAMEWORK The regulatory framework for utilisation of incinerator residues is evolving due to the current debate. Therefore, the following descriptions of regulatory frameworks are intended only to provide a summary of different approaches. Only Denmark has a regulatory framework which is not under revision.

902 21.3.1 Denmark

The utilisation of granular incinerator ash for civil engineering purposes in Denmark has been regulated since 1983 by rules issued by the Danish Ministry of the Environment (Statutory order No, 568 of Dec.6, 1983). These rules apply to the use of small and moderate quantities of incinerator ash for specified purposes. Large scale applications of incinerator ash involving more than 30,000 tons of ash and ash applied in layers thicker than 5 m are regulated under the Disposal and Discharge Permit Act (Section 5 of the Environmental Protection Act). In 1989, these regulations were supplemented by a set of technical guidelines for the utilisation of bottom ash as a subbase material, issued by the Danish Highway Department (Phil et al., 1989). The principles of the rules and guidelines regulating incinerator ash utilisation in Denmark are illustrated in the diagram shown in Figure 21.1. In principle, both bottom ash and fly ash or combined ash may be utilised. In practice, however, all fly ash and virtually all combined ash will fail to meet the conditions set on the heavy metal content of the ash. Therefore, the ability to collect the bottom ash separately from the fly ash at the incinerator is mandatory to ash utilisation. There are chemical composition requirements for each use. Each portion of incinerator ash, with a maximum 5,000 tons, intended for utilisation must be sampled. The sample, which may be collected on-line (e.g., from a conveyor belt) or from a stockpile, must be a composite of at least 50 sub-samples of 2 kg each. If the ash has not been screened prior to sampling, the composite sample is passed through a 45 mm screen to remove large objects. In order to facilitate the subsequent crushing, ferromagnetic material, pieces of nonmagnetic metals, and pieces of unburnt material (paper, fabric, etc.) may be removed from the screened and air-dried ash sample, which is then reduced to 5 kg by means of a riffle sampler. After crushing to <2mm (e.g., using a jaw crusher), the amount of sample is further reduced to 100 g, which is subsequently ground (e.g. in a mortar grinder) to 95% (w/w) <90 mm and analysed for pH (in a 1% slurry in demineralised water after stirring for 0.5 hours under cover), acid neutralisation capacity (alkalinity above pH = 7), and lead, cadmium, and mercury (metals determined by partial digestion with half-concentrated nitric acid for 0.5 hours at 1 atm followed by atomic absorption spectrophotometry, AAS). The results are expressed as concentrations in the ground samples. Utilisation of the 5,000 tons portion of ash is permitted if the results of the chemical analysis comply with the criteria shown in Table 21.2. The utilisation of ash that meets the quality requirements described above is further subject to quantitative and environmental protection/application related restrictions. The general conditions are the following: If incinerator ash is to be used under paved roads/squares, the following additional requirements must be met:

903 Figure 21.1 Flow Diagram of Guidelines Regulating Ash Utilisation in Denmark

INCINERATOR ASH BA

UTILISATION

OK

MEET CONDITIONS

(+FA)

NOTIFY AUTHORITIES

'1

LNO

PRETREATMENT AND CHEMICAL ANALYSIS

THAN

(5000 t PORTIONS)

YES NO

YES

II

THAN

NO FURTHER EIA OR STRICTER CONDITIONS

DISPOSAL

I UTILISATION MAY BE OK r'~

I

SUBMIT DESCRIPTION TO LOCAL AUTHORITIES WAIT 4 WEEKS

MAYBE PROJECT REFUSED

YES

NO

YES GO AHEAD

SPECIAL PERMIT UNDER DISPOSAL ACT NEEDED

904 Table 21.2 Testing Requirements for Utilising Ash in Denmark Parameter

Criteria

pH (1% Slurry) Alkalinity Lead

> 9.0 > 1.5 eqv/kg DW < 3,000 mg/kg DW

Cadmium

< 10 mg/kg DW

Mercury

< 0.5 mg/kg DW The distance to drinking water wells must be 20 m or more The ash must be placed above the highest groundwater table The maximum average thickness of the ash layer is 1 m, and the thickness of the ash layer must not exceed 2 m.

If the ash is to be used in unpaved single-lane roads and unpaved squares (maximum surface area of 2000 m2), the following additional requirements must be met: The distance to drinking water wells must be 20 m or more The thickness of the ash layer must not exceed 0.30 m. If a utilisation project of this nature involves less than 100 tons of ash of approved quality, it may proceed without any permit as long as the above conditions are met. If an ash utilisation project involves more than 100 tons but less than 30,000 tons of ash, a detailed description of the project must be submitted to the local authorities (county and municipality) in advance for approval. The applicant must then wait for 4 weeks. If he receives no (negative) reply within the 4 week period, the project is approved as submitted. Each county council may refuse the project if it is in conflict with environmental protection considerations or may ask the applicant to change the project or to provide an environmental impact assessment before resubmitting the project. Large scale applications of more than 30,000 tons of ash will usually covered by the legislation on disposal (Figure 21.1). When the ash is used a sub-base in road construction, it must comply with the following additional performance related conditions set by the Danish Highway Department: The bottom ash must not be mixed with fly ash (or any other materials); The bottom ash must be quenched immediately, and it must be stored for at least 1 month prior to utilisation; and, The bottom ash must be screened to maximum particle size of 50 mm, contain less than 9 percent (w/w) of fines below 0.075 mm, the loss on ignition (at 1000~ must be less than 10 percent (w/w), and the content of water must be between 17 and 25 percent.

905

21.3.2 Germany Guidelines for road construction have been developed (Hoesel, 1986), including options for the utilisation of bottom ash (grate ash only or grate ash combined with boiler ash) in the road surface (with or without binder), and use in road base and fill in areas such as parking lots, promenades, noise protection walls, etc. These guidelines serve as the basis for subsequent regulations. The raw ash has to be stored a minimum of 3 months to reduce water content (initially about 30%) and allow swelling to occur. Sieving and ferrous removal prior to utilisation is also required. The following properties are specified for bottom ash: grain size < 32 mm; splintering during freeze-thaw testing between 0.5-8.5 wt.%; proctor density of 1.5-1.9 Mg/m 3 at 11-18% moisture content; LOI < 5%; and pH of 8 - 12 in water. Bottom ash use should be at least 1 m above the groundwater table. In water quality protection areas, additional requirements recommended were: pure metals < 5%; unburnt material < 0.5%; LOI < 5%; particle size < 0.063 mm < 7%; soluble matter < 2%; leaching parameters based on DEV $4 test to include pH, conductivity, CI, sulphate, EOX, TOC, Pb, Cr, Cd, Cu, Ni, Zn. Later guidelines prohibited the use of ash in water quality protection areas. Monthly monitoring of ash quality is also recommended. Individual German states have issued regulations for ash utilisation. For example, in Hassia, bottom ash has to be pretreated according to the guidelines presented by Hoesel (1986) along with additional specifications (Hessisches, 1988). Ash must be aged for more than 2 months and have an LOI < 2%. Using the German standard leaching test (DEV $4), solubility measured as solid residue of evaporation must be < 1%, and limits for ions are (mg/I) NH4 0.4, CI 250, S04 600, F 3, Pb 0.1, Cd 0.004, Cr 0.04, Cu 0.5, Ni 0.04, Zn 0.5, and Hg 0.001. The moisture content, pH and conductivity must be recorded. Every 2 years a PCDD/PCDF analysis is required but no limit is provided.

21.3.3 The Netherlands The Netherlands currently has the most extensive framework proposed for utilisation of waste materials including incinerator residues. Management of these residues is regulated under the general framework established for solid wastes including dredge spoils, construction debris and other industrial and combustion residues (Eikelboom, 1992). The philosophical basis for the regulatory framework includes lifecycle management, (ii) marginal environmental burdening and (iii) user acceptance. The goal of lifecycle management is to maintain or modify the physical and environmental properties of residues to achieve the highest quality practical for recycling as granular construction materials, as many successive times as possible. The goal of marginal environmental burdening is to establish incremental increases in ambient soil and water

906 contaminant concentrations below which environmental impacts are negligible or acceptable. The goal of user acceptance is to allow routine residue utilisation in environmentally acceptable applications with public and product user confidence. These goals have resulted in the development of a detailed set of regulations and supporting research. The key aspects of the regulatory framework are: 9

Classification of waste substances and building materials, Establishment of target values for soils (including groundwater) and surface water, Establishment of standardised leaching tests, composition requirements and evaluation protocols for building materials, and

9

Certification of residues for use

Classification of waste substances and building materials is based on whether the materials are granular or monolithic, the use of additional emission controls (e.g., covers, liners, etc.), and the degree of contact with water. The target values for soils and groundwater are presented in Table 21.3. They were established based on a survey of soil and groundwater quality within the country. These target values then were used to determine acceptable marginal burdening levels for contaminants released from residues during use (Table 21.4). Marginal burdening is defined as "an increase of 1% in the level of pollution in relation to the target values in 1 meter of soil over 100 years." Soil composition was assumed to include 10% humus and 25% lutite. The assumed soil composition was used to estimate the distribution contaminants between assimilation by the soil and transport to groundwater. This approach indicated that marginal burdening of soil was also protective of groundwater. Contaminant release limits for building materials incorporating wastes are based on standardised column leach test for granular materials (NEN 7343) and a monolith leach test (NEN 7345) for molded construction materials (Aalbers, 1992). Detailed structured assumptions and extrapolations based on the physical structure of the material, application and leaching mechanisms have been developed to permit development of limits for laboratory tests. Utilisation of monolithic and granular wastes is classified based on contact with water and whether or not additional barriers (e.g., liners and covers) are employed: Type A -

Submerged or always in contact with water

Type B -

Primary contact with water is from precipitation (estimated contact with water 14% of time)

907 Table 21.3 Target Values for Soil and Groundwater Quality in The Netherlands Substance Ground Groundwater Surface Water mg/kg

pg/I

IJg/I

Cr

100

1

5

Co

20

20

NA

Ni

35

15

9

Cu

36

15

3

Zn

140

65

9

As

29

10

5

Mo

10

5

NA

Cd

0,8

0,4

0,05

Sb

(2.6)

NA

NA

Se

(1)

NA

(10)

Sn

20

10

NA

Ba

200

50

(200)

Hg

0,3

0,05

0,02

Pb

85

15

4

V

(68)

NA

NA

F

500

500

1500

CN-complex

5

10

NA

CN-free

1

5

(50)

(500)

150,000

100,000

20

300

8,000

S O4 Br

CI (200) 100,000 200,000 values from "Beleidstandpunt Over De Notitie Milbowa" (0.5-0.2-1992 with the exception of the values, which are in brackets). the values of chloride, fluoride, bromide and sulphate in surface water are limit values. NA = not available Sb, Se, and V from [13] Aalbers, 1992 -

908 Table 21.4 Maxirnum Acceptable Marginal Burdening Levels for Contaminants Released from Residues During Use in The Netherlands Substance

Ground

Groundwater

max. accept, immersions mg/m2 per 100 year

max. accept, immersion mg/m2 per 1 year

As

400

Ba

3000

Cd

10

Co

300

Cr-tot

1500

Cu

500

Hg

4

Mo

150

Ni

500

Pb

1000

Sb

35

Se

15

Sn

300

V

950

Zn

2000

Br

300

CI F

30000 7000

SO4

45000

CN-tot

70

CN-free

15

909 V1 building materials-

application without additional emission controls

V2 building materials-

application with additional emission controls (e.g., covers, liners, etc.)

The relationship between the maximum allowable release and the release observed during the standardised monolith leaching test is described by: where:

Im,x(J yr)= Em,• Imax(J yr)=maximum acceptable emission into the ground in a period of J year (mg/m2); Em~(64d)=maximum acceptable emission out of a material determined with the tank leaching test in 64 days (mg/m2); f~xt=extrapolation factor for the extrapolation from Em,~(64d) to Em~(J yr); fbev=Correction factor for wetting period; f~=correction factor for the difference between the laboratory temperature and the temperature in the field; f~so=isolation factor for V2 building materials (this factor is 1 for V1 materials).

The criteria for determining if diffusion is the controlling release mechanism is based on the slope of the cumulative release curve from the monolith leach test: Slope > 0.6-Emission is not diffusion controlled (more rapid) and the column leach test should be applied 0.35 < Slope < 0.6-Emission is diffusion controlled; Slope <0.35-Emission is not diffusion controlled (slower) and the 64 day emission can be used as an estimate or be based on the column test. Diffusion based emission is classified based on pD~ as: PDe > 12-1ow mobility 10.5 < pDe < 12-intermediate mobility pD~ < 10.5-high mobility

910 If the pDe is less than 10 for a species, depletion of that species can be anticipated to occur during the 100 yr assumed use interval. The release is estimated based on one dimensional diffusion from a flat plate. The quantity available for release is estimated based on the availability leach test (NEN 7340). The effects of weathering on the rate of diffusion are assumed to decrease the release rate. This implies physical durability and the occurrence of weathering mechanisms, such as carbonate uptake, which occur over the 100 year assumed use interval. The correction to account for this effect was to assume an increase in the PDe of 0.01/yr. The cumulative effects of the above derivation are summed in an overall extrapolation factor or multiplier to translate release from 64 day tank leaching results to a 100 yr release interval (Tables 21.5 and 21.6). Table 21.5 Extrapolation Factors for Determining Release in the Field from Laboratory Leaching Results for Cases where Depletion is Anticipated (The Netherlands) Layer thickness (m):

0.3

0.5

0.7

1.0

2.0

10

fuit

f,~

f,,t

fuit

fu,t

8 2

2

3

5

10

24

9 5

8

11

16

23

24

10 15

21

23

24

24

24

pD~ f,,

f,, : E(100yr)/E(64d)

Table 21.6 Extrapolation Factors for Determining Release in the Field from Laboratory Leaching Results with Geometric Considerations, Depletion and Assumed Effects due to Weathering Included (The Netherlands) Layer thickness (m): pO~

0.3

0.5

0.7

1.0

2.0

10

f,~

f,~t

f,~t

f,,~

.f,,~t

f,,~t

8

2

2

3

5

10

15

9

5

8

11

15

15

15

15

15

15

15

15

15

>10 fe~t= E(100yr)/E(64d)

Chloride and sulphate are considered exceptions to the above extrapolation factors because the maximum acceptable emission is defined to be that which would be released over a period of one year. This recognises the limited natural attenuation which occurs for these species and the potential direct impact on groundwater resources. The extrapolation factor, fe~, for chloride and sulphate is 2.4. The other

911 factors for translating release from tank leaching results to field conditions are to correct for differences in the frequency of wetting, temperature and isolation (e.g., barriers such as liners and covers). The correction factor, fb,,v, for always immersed conditions is 1.0 while for non-immersed applications it is 0.14. The factor of 0.14 is based on an average precipitation occurrence, and hence surface wetting, of 14% of the time in The Netherlands. While this is used as a multiplier to emission, a more appropriate approach would be to modify the release time interval using (0.14) ~ The correction factor for temperature, from,was based on an Arrhenius approach to diffusion kinetics and a mean laboratory and field temperatures of 20 C and 10 C, respectively. This resulted in ft~ equal to 0.7. The correction factor for isolation, fiso, was assumed to be 0.14 when addition barriers were employed and 1.0 when the material was directly exposed. Inversion of the above derivation permits estimation of acceptable laboratory results based on defined field conditions as Emax(64d) = Imax(100yr)/(fext*ft~m*fbev*fi,o) A similar laboratory to field extrapolation approach was applied for granular materials. In this case, the principal mechanism of contaminant release is assumed to be by percolation through the granular material. This is in contrast to when a granular material compacted in place results in a low permeability layer which may be treated as a monolith or diffusion controlled leaching. Examples of typical limit values for leaching test results on monolithic and granular materials are presented in Table 21.7. Since 1995, the limit value on composition $1 has been withdrawn following discussions between regulators and industry. 21.3.4 Sweden

In Sweden, both the utilisation and disposal of residues are treated in the same manner under The Environmental Protection Act (Hartl~n, 1989 and 1991; F~llman, 1992). Thus, individual regulatory reviews including local and regional authorities are required for each specific application (Figure 21.2). The general philosophy in evaluating utilisation applications is that the specific utilisation scenario should result in (i) improvement of general environmental conditions, and (ii) have less environmental impact than disposal. Examples of uses that may satisfy these criteria are use as cover in a municipal waste landfill, in road paving applications where the residue is covered with an asphalt layer, or where natural aggregate is in limited supply. To date, very little utilisation has occurred, and utilisation regulations are under review. 21.3.5 United States

The United States currently does not have national standards for the utilisation of residues. Requirements for the U.S. Environmental Protection Agency to develop

912 criteria for ash utilisation are being considered in pending legislation. In the absence of national guidelines, several states have developed applicable regulatory requirements. Florida and New York are the two states that have the most extensive requirements.

Table 21.7 Maximum Acceptable Limits for Leaching Test Results on Granular (mg/kg) and Monolithic (mg/m2) Materials (proposed, The Netherlands) Leaching

standard values oBB granular materials in mcj/kg

Substance

U1

U2

$1

U1

U2

$1

0.30

3.0

375

25

125

750

As

standard values oBB products in mg/m 2

Ba

4.0

40

7500

350

1750

15000

Cd

0.010

0.10

10

0.70

3.5

20

Co

0.20

2.0

250

15

75

500

Cr

1.0

10

1250

90

450

2500

Cu

0.35

4.0

375

30

150

750

Hg

0.010

0.050

5

0.30

1.5

10

20

250

Mo

0.050

0.50

125

4.0

Ni

0.35

4.0

250

30

150

500

Pb

0.80

8.0

1250

75

375

2500

Sb

0.030

0.30

50

2.5

13

100

Se

0.020

0.20

50

1.8

9.0

100

Sn

0.20

2.0

250

20

100

500

V

0.70

7.0

1250

60

300

2500

Zn

1.4

14

1250

125

625

2500

Br

0.20

2.0

500

20

100

1000

600

5000

5000

2250

11250

1000000

CN- complex

0.050

0.50

125

4.5

23

250

CN-free

0.010

0.10

25

0.90

4.5

50

5.0

50

4500

CI

F

SO 4 750 10000 25000 A = A-type application" B = B-type application

440

2200

9000

15000

45000

40000

913 Figure 21.2 Regulatory Reviews Required for Specific Applications of Ash Utilisation in Sweden APPLICANT

FRANCHISE BOARD

PLANNING I

COUNTY COUNCIL

LOCAL AUTHORITY

ENVIRONMENTAL PROTECTION BOARD

OTHER PARTIES

INFORMATIONAND PRELIMINARYDISCUSSIONS

APPLICATION TREATMENT

COMPLETION

CIRCUI.A,TION FOR COMMENT

, I

PUBLIC HEARING DECISION, TERMS OF CONCESSION

INSPECTION L OPERATION

r

io

I INSPECTION '~ PROGRAMME

,._sPECT,om

INSPECTION PROGRAMME INSPECTION

INSPECTION

Florida requires an ash management plan as part of the operating permit for an incinerator facility. These plans must be updated and reviewed at least once every five years (Florida reg. 17-702.400), and must address the methods, equipment and structures needed to control dispersion of ash during handling, processing, storage, loading, transportation, unloading and disposal. The plan must consider potential pathways for human and environmental exposure including inhalation and direct contact (human exposure) and migration to soil, groundwater and surface water (environmental exposure). Recycling of ash (utilisation) is explicitly discussed in the regulations (Florida reg. 17-702.600). The generator of ash, at least monthly must describe the chemical and physical properties of the ash which is to be recycled. Prior to the recycling of ash, the process and use of the ash must be shown not to cause discharges of pollutants to the environment. In addition, in order to utilise ash, the following steps must be completed: describe chemical and physical properties of the finished product line; identify the quantity of ash used in the product; identify the quantity of product to be marketed or used; demonstrate the process will physically or chemically change the ash residue so that any leachates produced after processing will not cause a violation of surface or ground water standards;

914 demonstrate ash or products will not endanger human health or the environment; performance standards need to be established as well as operational criteria. New York also requires the development of an ash management plan for each incinerator facility (New York reg. section 360-3.5). Ash generation, handling, storage, testing, transportation, treatment, and disposal or beneficial use plan must be included. Ash utilisation is regulated by a beneficial use petition. The party who desires to beneficially use ash must petition to utilise the ash. There is no permit involved directly with ash utilisation. As a result of no permit being required, there are no public hearings required. The party petitioning for ash to be used as an ingredient or as a substitute for a raw material must: demonstrate that the resulting material is not a waste requiring disposal; have a known market or disposition; 9

not accumulate the material speculatively; have contractual arrangements with a second person for use as an ingredient and this person has to have the equipment to do so; chemically and physically characterise the ash; identify the quantity and quality to be marketed; describe the proposed method of application or use, available markets and marketing agreements; demonstrate that the intended use will not adversely affect the public health, safety, welfare and environment; provide a description of each product mixture, if the use of the ash includes the mixing with different types of materials.

21.4 TECHNICAL REQUIREMENTS The three major categories of applications include use as a lightweight aggregate either for road base construction, as a fill for embankments, or as an amendment for Portland cement concrete or bituminous asphalt. Most physical utilisation criteria are based on standard engineering tests. Specific physical requirements will vary based on the type of application (e.g., asphalt pavement, cement concrete, structural fill, etc.) and local construction regulations. Table 21.8 provides a summary of the physical testing

915 requirements that may be required and typical acceptable values. Many of the tests traditionally specified for construction may not be directly applicable to bottom ash testing. However, they may be necessary for market acceptance. Most frequently, bottom ash can be blended with other aggregates to achieve specific design criteria.

Table 21.8 Physical Criteria and Property Ranges for Utilisation of Bottom Ash Requirement

Asphalt Pavement

Concrete

Particle Size Distribution

Specific 9 to location & application design Uniformity 9 coefficient (d60/dl0) can be specified total 9 content of fines (<601Jm) _ 10% (includinQ all materials*)

Specific 9 to location & application design Uniformity 9 coefficient (d60/dl0) can be specified _<10% fines (< 60pm)

Loss on Ignition

_ 3%" lower values are preferred

_<3%; lower values are preferred

_ 5%; lower values are preferred

Moisture Content

___15% (geotechnical); as dry as practical is preferred

_<15% (geochemical)

approx. 16-17% (proctor optimum for mod. compaction)

Durability

Specific 9 to location & application design LA 9 Abrasion California 9 Bearing Ratio Sodium 9 sulphate soundness

Specific 9 to location & application design; applicable to final product Unconfined 9 compressive strength

Specific 9 to location & application design Shear 9 strength California 9 bearing ratio for base course material

Expansion

Aging for ~ 3 months at 16% moisture

Aging for ~ 3 months at 16% moisture

Aging for ~ 3 months at 16% moisture

H2 Evolution

Metal recovery for ferrous and nonferrous metals recommended

Metal recovery for ferrous and nonferrous metals required

Metal recovery for ferrous and nonferrous metals recommended

....

Water retention and transmission

9

Structural Fill Specific 9 to location & application design Uniformity 9 coefficient (d60/dl0) can be specified 9 _<10% fines (< 60pm)

Hydraulic 9 conductivity capillary 9 suction This means that the fines content of the ash must be less than 10% to account for contributions from other materials

Some additional limitations have been applied to bottom ash materials being considered for utilisation. Organic matter in bottom ash may create problems with respect to subsidence in the specific construction application and the evolution of g a s e s as the organic matter degrades. In some European countries bottom ash s a m p l e s found to contain >3% LOI are deemed unacceptable for use and must be

916 deposited in a landfill. The ash must be processed to generate material with the proper grain size distribution (max. grain size 45-50 mm, <10% of the total weight <0.06 mm grain size). This may preclude a large portion of the bottom ash from RDF combustion systems which tends to consist of smaller grain sizes due to the preprocessing of waste feed. Pre-processing, such as screening and removal of oversize material and ferrous and nonferrous metals, is essential because metallic forms of these elements are detrimental to the utilisation of bottom ash as an aggregate. Separated metal can be directly recycled. 21.5 UTILISATION LIFE CYCLE AND ENVIRONMENTAL CONSIDERATIONS Development of a comprehensive program for utilisation of incinerator residues requires consideration of ash handling requirements and potential environmental impacts throughout the life cycle of residue utilisation, beginning with ash generation and ending with either permanent use or ultimate disposal. A utilisation program also must consider local climate, geography and sociology. The following sections are intended to provide an overview of relevant considerations and suggestions for regulatory approaches. The typical projected life cycle of incinerator residues during utilisation includes the following stages: .

2. 3. 4. 5. 6.

Ash generation (production at the facility) Physical processing Stockpiling Manufacture Use in designated application, and Post-utilisation management and disposal

Potential ash impacts and considerations are essentially common independent of utilisation application from the time of ash generation to the point of manufacturing. Subsequent stages in the life cycle are significantly more application dependent because of the nature of the material in which the ash will be used and exposure scenario during use. For example, utilisation in Portland cement applications will have different effects on contaminant release than utilisation in bituminous pavement. 21.5.1 Ash Type Selection and Elements of Concern Currently, bottom ash without grate siftings or boiler ash is considered to have the greatest potential for utilisation because it typically has the lowest content of leachable metals of concern (e.g., lead, cadmium, mercury, etc.) and soluble salts. In addition, this ash fraction has physical properties similar to lightweight natural aggregates and represents approximately 80 volume percent (70 wt %) of the total residues generated.

917 Grate siftings are excluded because of the content of fine particulates and relatively high contents of elemental lead and aluminum. Boiler ash is excluded because of the potential for relatively high content of more volatile metals such as cadmium and zinc. APC residues are excluded because of high soluble salt content (40-60 wt %) and relatively high contents of metals of concern such as cadmium, lead, zinc and mercury. The use of APC residues is also limited because of its high content of fine-grained particles which gives it high moisture holding capacity and therefore susceptible to frost expansion, and is difficult to compact. The chemical elements and species which are of potential environmental concern in the residues are Pb, Cu, Cd, Cr, Mo, Hg, Zn, total soluble salts (e.g., Na, K, CI-, SO42) and total soluble organic carbon. These elements and species were selected based on either current regulatory guidelines for drinking water or solid waste management, leachability, potential aquatic life toxicity or potential engineering effects. 21.5.2 Ash Generation

Ash generation is defined as the production of the residues to be utilised at the incinerator facility. This stage is the most critical for quality control. The intent at this stage should be to produce as uniform a product (ash) as possible that will permit utilisation after subsequent processing. This will minimise the amount of processed ash that would be rejected as unacceptable at later stages or require disposal. Development of an overall quality control plan that would minimise testing requirements while maintaining ash quality is needed. Critical testing parameters during this stage would be loss on ignition (LOI), alkalinity, total leachable salts, leaching potential or availability of key elements, and moisture. LOI serves as an indicator of combustion efficiency and residual organic matter. Alkalinity or acid neutralisation capacity provides a measure of the material behaviour in the environment because of leaching of potentially toxic metals is strongly a function of pH. Development of a pH titration curve also would permit estimation of the contributions of hydroxide, bicarbonate and carbonate buffer systems. Total leachable salts is an important parameter because total salt Ioadings can adversely affect soil and potable water resources. Total salt content can also adversely affect the durability of Portland cement-based products. The leaching potential, or availability, of key elements is important as threshold values for acceptance based on projected impact at the utilisation scenario. Availability is recommended rather than total concentration because fractions of each element of concern may be bound in mineral forms that would make it non-leachable or biologically unavailable under the normal extremes of environmental conditions. An example of this would be lead or chromium bound in a silicate matrix. Moisture content is important to insure that while excessive interstitial pore water does not exist (e.g., geotechnical moisture content should be less than 20%), enough moisture is present to prevent fugitive dust problems and allow proper aging of residues to proceed (>16%; see "stockpiling").

918 The frequency of testing to be carried out needs to be developed based on a statistical evaluation of the acceptable range and variation of critical parameters. Acceptance criteria should establish both the mean and the acceptable variance of analytical results that limit the quantity of material beyond the threshold limits that would render an entire lot of material unacceptable. Thus, testing at this stage would be for screening and quality control purposes and would be based on prior knowledge and detailed characterisation of the class of residue to be evaluated. Specific thresholds should be based on projected impacts for each utilisation scenario, since different utilisation scenarios may have different acceptance thresholds. For more broad scale requirements, grouping of utilisation scenarios by type of use with general site restrictions may be practical.

21.5.3 Physical Processing Physical processing of ash is defined as mechanical processing such as ferrous and nonferrous metal removal, and crushing and screening to control the particle size gradation of the material to be utilised. Removal of oversized material is necessary to facilitate subsequent processing into appropriate products (e.g., asphalt paving material or concrete forms) and would be based on the specific utilisation scenario. The principal environmental and occupational health impact concerns during this stage would be a consequence of fugitive dust. Removal of fines may be necessary to minimise fugitive dust, and attendant controls, during subsequent stages. These operations may occur either at the facility or at the stockpiling location. Physical processing operations should be carried out with practices intended to minimise fugitive dust and avoid potential occupational health hazards.

21.5.4 Stockpiling Stockpiling of ash is carried out for several reasons. First, during stockpiling, aging reactions occur within the ash which further stabilise the material. These reactions include oxidation, hydration and carbonation (fixation or uptake of atmospheric carbon dioxide) reactions. Oxidation of reduced metals typically results in less leachable forms. Carbon dioxide uptake results in a pH shift of the material from typically greater than eleven to more neutral pH, e.g., less than 9. This process also results in respeciation of some elements from hydroxides to carbonates. The net result of this process is a shift in the pH domain and speciation of the material to a less leachable regime for metals such as lead and zinc. Hydration reactions typically result in swelling of the material. These swelling reactions must be allowed to progress prior to utilisation to avoid detrimental effects on the structural durability of the final products. In the Netherlands, bottom ash is required to be aged a minimum of 6 weeks prior to utilisation (Born, 1994). Exact intervals required for sufficient ash aging have yet to be defined, but preliminary findings indicate a period between one and three months is sufficient.

919 A second reason for ash stockpiling is to allow for storage of the material because of seasonal demand. For example, most paving applications will be able to utilise the material only six to eight months out of the year depending on local climate. Potential environmental or health impacts from stockpiling can be a consequence of fugitive dust, precipitation runoff, leachate or site access. Fugitive dust can be controlled by limiting the fraction of fine material (less than 100 IJm) permitted in the stockpile and maintaining a minimum water content (greater than approximately 16 percent). Minimum moisture content also will facilitate ash aging processes. All runoff and leachate from the stockpile should be collected and treated if necessary. Applicable storm water and local regulations may be sufficient to address these concerns. Site access should be controlled to avoid unwanted human exposure. Minimum ash stockpiling intervals should be established based on ash aging requirements. Maximum storage intervals should be based on the local annual climatic cycle. Consideration must be given if the aging process period and the seasonal demand period are mismatched. For example, in the northeast U.S. ash generated in July may have to be stockpiled until the following April for paving applications. Maximum ash stockpile quantities can be based on either annual use or demonstrated prior agreements for use with the entity receiving the ash after stockpiling.

21.5.5 Manufacturing Manufacturing is defined as the processing of the ash into the final product form. This stage for paving applications would include ash drying and blending with asphalt at the asphalt plant, and placement of the pavement. This stage for Portland cement applications would include mixing with Portland cement, water and natural aggregate and forming into final structures such as blocks. Ash handling requirements during this and subsequent stages, should conform with standard handling procedures for materials which the ash is replacing. Potential environmental or health impacts during the manufacturing stage could result either from fugitive dust or drying process emissions. Asphalt production will require drying of the ash prior to blending with other aggregates and asphalt. Aggregate drying typically is carried out at approximately 200~ This temperature is not high enough to volatilise metals of concern, but may cause some entrainment of fine particles in the drying air stream, thus potentially increasing air emissions. Fugitive dust and air emissions impacts most likely can be minimised by limiting the fraction of fines in the ash stream.

21.5.6 Use Applications A general approach for selection of acceptable utilisation applications and overall control of utilisation can be classified by the following steps:

920 i.

Detailed ash characterisation Detailed evaluation of environmental impacts from proposed application

iii.

Ash screening and quality control

iv.

Verification of Ash characterisation, and

V.

Categorically approved utilisation for certain applications with limited restrictions (previously defined)

Detailed ash characterisation would require determination of the statistical variation of the ash to be utilised, including chemical, physical and leaching characteristics. Acceptance limits for key parameters and statistical evaluation methods need to be established to provide an indication of the material's performance during use. After the variability of key characteristics has been determined, only a reduced set of analyses for quality control would be required at the time of ash generation. Ash characteristics and behaviour would be verified prior to use. This approach is often referred to as ash "certification." Certification would be granted to a specific incinerator facility after the initial statistical verification confirmed that the facility consistently meets quality control criteria. An initial set of potential quality control parameters has been presented in the section entitled "Ash Generation". This approach is possible because it has been demonstrated that residues from similar types of facilities exhibit common characteristics. Thus, only quality control and screening for non-characteristic properties are required. Evaluation of impacts from proposed applications and potential approaches to criteria are presented in the sections that follow. For utilisation to be practical, extensive permitting should not be required for categorically approved applications which meet predefined restrictions. Predefined restrictions may include restrictions on location of ash utilisation and maximum quantities allowable. Record keeping should be required with regard to the location, quantity and nature of each utilisation application. Two primary routes for environmental impact require consideration for most applications. The first route is through particle transport followed by either incorporation into soil or sediment, or food chain uptake. Food chain uptake is a much greater concern for marine applications. The principal controls over particle transport are through limiting direct abrasion on surfaces containing ash and through requiring specific product durability. The second exposure route is through leaching, followed by impact on either groundwater, surface water or soil resources. Contaminant release through leaching can be viewed as consisting of two components, contaminant release potential and contaminant release rate. Establishing limits on cumulative contaminant release over a fixed time interval (e.g., 100 yr) is a potential approach for providing a uniform basis

921 for comparison of potential impacts from both utilisation and disposal. The cumulative contaminant release could be projected based on integration of release rate and release potential data for defined geometries. This projection could include application specific information such as mean temperatures and precipitation to provide translation of laboratory data to field scenarios. Details of this approach have been provided in the Leaching of Products Chapter. Cumulative contaminant release would be the most important parameter for elements or species of concern that accumulation in the surrounding environment. Release rate or flux would be the most important parameter for non-accumulating elements or species (e.g., sodium, chloride). Release potential would be the limiting parameter when the utilisation scenario is a permanent placement of the ash. This would be the case for marine applications. Results from estimating long term constituent release can be compared to natural materials performance or evaluated based on an environmental impact assessment. Alternative approaches that can be considered include the approaches of site specific risk assessments, marginal burdening or defining an acceptable release limit. Site specific risk assessments most likely would be overly burdensome for implementation without affording additional environmental protection, except for very large scale utilisation projects. The marginal burdening approach is limited by the need to define background levels for application locations or regions. Alternatively, it is recommended that the acceptable release limit should be defined as resulting in an impact which is a fraction (e.g., 10%) of the applicable soil or groundwater standard. In order to facilitate this assessment, the following assumptions are recommended: The impacted area used as the assessment basis for accumulating constituents (e.g., trace metals) should be 10 cm depth of soil below the application; The impacted area used as the assessment basis for mobile constituents (e.g., TDS) should be based on the annual flux of groundwater in the uppermost 1 m of the aquifer below the application. Thus, a dilution volume per m2 of application surface can be estimated; and, The time frame for impact should use either the design life of the application or a default interval of 100 yr. For example, if the soil quality standard for Pb was 100 mg/kg, a cumulative release that resulted in less than a 20 mg/kg increase in the 10 cm of soil below the application over 100 years would be considered acceptable. This approach provides for environmental protection while permitting regional flexibility with respect to climate and environmental standards. In addition, locations for utilisation should consider natural conditions, such as distance to sensitive natural resources and depths to groundwater. Life-cycle analysis for the

922 material also should be carried out to evaluate the fate of the material after the designated use period and to assure compatibility with standard product recycling or disposal practices. The result of the above estimation should be approval of routine application for utilisation projects with limited impact potential (e.g., structural fill less than a nominal thickness, aggregate in asphalt and landfill cover) while large scale applications (e.g., wind barriers, sound barriers, structural fill greater than a nominal thickness) would require site specific review and permit. pavin~ Applications. The primary paving applications considered for use of residues is in the wearing course (roadway surface), binder course, base and embankments. Options for use of ash which has been physically processed and aged are (i) as a compacted granular material (ii) directly incorporated in asphalt, (iii) further solidified or chemically stabilised and used as a compacted granular material, or, (iv) further solidified or chemically stabilised and incorporated in asphalt.

Examples

In Sweden, use of compacted bottom ash as roadbase underlying asphalt pavement and as embankment was evaluated in a test road segment located in Malmo (Hartl~n, 1994). Use of fresh bottom ash, bottom ash which had been aged for one year and natural aggregate was compared (Figure 21.3). All ash was screened to less than 20 mm and had ferrous metal magnetically removed prior to use. It was found that aged ash (moisture content 16%) was readily compacted while compaction of fresh ash (moisture content 23%) resulted in failure of bearing capacity during compaction. Both ash types had an optimum moisture content of about 14% for proctor compaction. Maximum compacted dry density for ash was between 1.79 and 1.82 tonne/m 3. Test results on the road indicated that the ash layers physically performed similarly to the natural aggregate. Environmental evaluation indicated that release of trace elements from bottom ash was similar to or less than that observed for natural materials (Table 21.9; Hartl~n and Lundgren, 1991). In the Netherlands, fly ash (ESP ash without scrubber residue) has been used as a fine aggregate filler in asphalt for approximately 10 years (Hudales, 1994). Approximately 40% of the 60,000 tonnes of fly ash produced annually is used in this application. A maximum substitution of 35% for natural materials may be achieved. Tank leaching tests (NVN 5432) of asphalt containing fly ash found that leaching met applicable contaminant release standards by greater than a factor of 3. Steketee and Urlings (1994) reported that substantial reductions in trace element and COD leaching was achieved by aging bottom ash in stockpiles for 12 months under moist, aerobic conditions. Leachate pH had decreased from 11.4 to 9.9 during this interval.

923 Figure 21.3 Test Road Built Up in Six Different Sections

1

2

3

35 MAB

95 AG

/

120 BASE COURSE GRAVEL

/

NORMAL SUBBASE COUSE 450

FRESH ASH 450

AGED ASH 450

4

5

25 MAB

35 MAB

50 AG

125

\

AGED ASH 500

I

80 m

I

80 m

145 AG 120 AGED ASH 520

AGED ASH 450

ko

SUBSOIL

80 m

95 AG 240

ADJUSTMENT LAYER SAND 200

I

6

/ o

I

80 m

I

80 m

I

80 m

I

Table 21.9 Observed Release of Trace Elements from Bottom Ash During Utilisation in a Paving Applicatio n (Sweden) Substance

Leachate from sorted BA

Background surface water

Localwater course

Contamination factor

Unit

CI

120

4

5.1

1.3

mg/I

SO4

300

15

17.8

1.2

mg/I

AI

50

30

30.2

1.01

pg/I

Cd

0

0.02

0.025

1.2

pg/I

Cr

4

0.3

0.34

1.1

pg/I

Cu

20

0.7

0.89

1.3

pg/I

Ni

25

0.5

0.74

1.5

pg/I

Pb

10

0.2

0.30

1.5

pg/I

Zn

100

2.0

3.0

1.5

pg/I

924 In 1993, a 350 m length of a secondary roadway in Leconia, New Hampshire with daily traffic from 5,000 to 20,000 vehicles per day was reconstructed using aged grate ash as an aggregate substitute in the binder course (Musselman et al., 1994). The binder course was 5 cm thick and employed 50% substitution of grate ash for natural aggregate. 7% asphalt cement was determined to be the optimum asphalt content for this substitution rate. A 300 m control section included only natural aggregate in the binder course for comparative purposes. Road physical performance to date has been considered the same as the control section. Detailed environmental monitoring is part of the testing program, but has not yet been reported. Applications using ash without further treatment as an aggregate replacement in asphalt used for binder or base courses should be given the highest priority. Significant reductions in potential contaminant release would be realised as a consequence of ash incorporation into asphalt. Further emphasis is given to these applications because both would have at least an impermeable asphalt layer above the utilised material, if not both above and below. Lower priority is given to use of ash in the wearing course because of potential abrasion and direct environmental exposure. Concern also exists about dust generated during milling of the wearing course during maintenance and repaving operations. Use of granular material directly as the wearing course is considered unacceptable. Use of granular material in embankments should be limited except in locations where salt release would not constitute a problem or if other precautions to limit salt release are established. Use of ash incorporated into asphalt for embankments is not considered a practical option because asphalt based materials are not typically used in that application. Use of treated ash (e.g., ash which had been further solidified or chemically treated) generally is ranked lower, except for use as compacted granular base, than use of untreated ash because additional processing requirements and economic concerns.

Terrestrial Cement Concrete Applications.

Terrestrial cement concrete based applications can be classified based on whether the structures are environmentally exposed or further isolated. Use of ash in concrete blocks, decorative paving blocks, or precast forms are typical applications. Isolation from exposure may be based on the application of sealants or use inside a building. The potential impact of salts (e.g., chloride) on concrete reinforcement must also be considered. Landfill Cover. Use of ash in municipal waste landfill applications present a unique environmental scenario because of the typical existence of environmental protection measures such as liners and leachate collection. Therefore, the principal exposure route of concern would be through fugitive dust.

Marine Applications. The marine environment can be considered more environmentally sensitive than the paving applications because of direct contact of marine biota with the application structure and the sediment and water column in the immediate vicinity of the application. The principal mechanisms for environmental impact in the marine environment would be through leaching and particle transport.

925 Particle transport is much more important in the marine environment than in the paving applications because it may be manifested through (i) erosion and biota uptake from the water column, (ii) erosion and biota uptake in local sediments or (iii) direct particle uptake by surface-attached biota. In all cases, food chain magnification of specific contaminants must be considered. Emphasis on particle transport also increases requirements for structural durability of ash containing materials. Definition of priority utilisation scenarios for the marine environment is complex because of the variety of marine environments and ecologically sensitive areas that exist. Primary variables are the salinity, intensity of wave energy, and the degree of water circulation or flushing. Lowest potential impact areas would be areas of high salinity and a high degree of water circulation. Sensitive areas such as coral reefs or highly productive estuaries should be avoided.

Application Restrictions. This type of potential restriction would limit the type of ash

to be utilised and the specific utilisation application. An example would be restriction of grate ash use to binder and base course layers in paving applications. For marine applications, an analogous limitation would be for use in defined structures either for shoreline protection or artificial reefs. The goal of this type of restriction is to facilitate development of highest priority applications first and to allow for revisions to criteria as more performance data becomes available.

Location Restrictions. This type of potential restriction excludes ash utilisation in environmentally sensitive areas. It also could be used to control site accessibility or provide for safety margins based on projected impacts. Examples for protection of sensitive areas include prohibiting ash utilisation in wetland areas or near coral reefs. Examples for control of site access or providing for additional safety margins are limiting use to applications on landfills or in industrially zoned areas, or, requiring a minimum distance from paving applications to groundwater supplies. Quantity Restrictions. This type of potential restriction is used to limit the maximum

quantity of ash to be used in a single location before more extensive review or permitting would be required.

Record Keeping and Monitoring Requirements. Records should be required to be maintained detailing each utilisation application. Information required should include ash characteristics, type of use and location of application. This would allow for future investigation of application performance. Routine environmental monitoring should be required only during pilot-scale evaluations of potential applications or for applications which exceed certain quantity restrictions. Extensive monitoring of every point of use is unwarranted and would be impractical and prohibitive.

926

21.5.7 Reuse and Disposal Paving Applications. Roads, parking lots and other paving applications are considered to have a finite application period. This period may be defined in terms of years or decades depending on the specific scenario. Asphalt pavement is frequently recycled into new paving material. Controls should be established that limit use of recycled ash containing materials to applications approved for initial utilisation. Disposal of ash containing materials should be in conformance with applicable guideline for similar materials not containing ash. Marine Applications. Shoreline protection installations and artificial reefs are considered permanent structures. Therefore, criteria for environmental acceptability should consider the utilisation scenario as the ultimate disposition of the material.

21.5.8 Economic Considerations Estimates of costs associated with utilisation of incinerator residues are considered very site-specific because of varied requirements for permitting, testing, transportation and cost offsets achieved through reduction in disposal and natural material costs. Generally, cost estimates have not been published. An additional consideration which has limited utilisation in some areas has the absence of clear definition of potential product liability. Especially in the United States, parties that may participate in ash utilisation are extremely reluctant until the extent of their potential product liability is defined and perhaps limited.

REFERENCES Aalbers,Th. G., P.G.M. de Wilde, G.A. Rood, A.I.M. van de Beek, M.B. Broekman, G. van der Meij, P. Masereeuw, Ch. Kamphuis, P.M. Dekker and E. Valentijn. Milieuhy.qienische kwaliteit van primaire en secondaire bouwmateriale.n in relatie to her.qebruik en bodembeschermin.a (Environmental quality for primary and secondary building materials in relation to recycling and Soil protection), Report no. 771402005, RIVM, 20 June 1992. Aitola, J-P. "Methodology for Size and Category Classification of MSW and the Downstream Effects," proceedings from Municipal WasteCombustion. Second Annual Int..ernational Specialty Conference. U.S. Environmental Protection Agency and Air & Waste Management Association, Tampa, FI. 1991. pp. 841 -850. Andrews Jr., J.C. "Incinerator Ash Disposal in the Tampa Bay Region," proceedings from Municipal Waste Combustion. Second Annual International Specialty Conference. U.S. Environmental Protection Agency and Air & Waste Management Association, Tampa, FI. 1991. pp. 284-298.

927 Benoit, J. and T.T. Eighmy. "Field and Laboratory Densities of Municipal Solid Waste Incinerator Ash/Wastewater Sludge Mixtures in a Codisposal Above-Ground Landfill," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 627-630. Benoit, J. "Compactability of MSW Ash and Wastewater Sludge Mixtures for Landfill Codisposal," proceedings from Municipal Waste Combustion. Second Annual International Specialty Conference. U.S. Environmental Protection Agency and Air & Waste Management Association, Tampa, FI. 1991. p. 609-626. Born, J.G.P. "Quantities and qualities of municipal waste incinerator residues in The Netherlands", in En.yironmentalAspects of Construction with Waste Materials, J.J.J.M. Goumans, H.A. van der Sloot and Th.G. Aalbers, eds., Elsevier Science B.V., Amsterdam, 1994. Chesner, W. "Environmental Issues Associated with the Use of MSW Combustor Ash in Asphalt Paving Mixes, (Abstract)" proceedings from Municipal Wast.e Combustion. Second Annual Interna.tional Specialty Conference. U.S. Environmental Protection Agency and Air & Waste Management Association, Tampa, FI. 1991. p. 608. Eikelboom, R.T. "Development of Environmental Criteria: Policy in The Netherlands", 1992. Fallman, A.-M. and J. Hartl~n. "New Perspectives on the Management of Residues from MSW Incineration in Sweden," .Waste Mana~qement International. Vol. 2. ed. K.J. Thom~-Kozmiensky., EF-VerI. fur Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1 992. pp. 125-1 3 2. Gress, D.L., X. Zhang, S. Tarr, F. Pazienza and T.T. Eighmy. "Municipal Solid Waste Combustion Ash as an Aggregate Substitute in Asphaltic Concrete," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 11 61 -1 76. Hartl~n, J. and T. Lundgren. "Utilization of Incinerator Bottom Ash Legal, Environmental and Engineering Aspects," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 207-214. Hartl~n, J. "Use of Incinerator Bottom Ash as Filling Material", XlII ICSMFE Conference, New Delhi, India, 1994. Heider, U. "The Use of Chloride-Containing Residual Materials in the Preparation of Backfilling Materials for Salt Mines," Waste Mana.qement International. Vol. 2. ed. K.J. Thom~-Kozmiensky., EF-VerI. fur Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 431 -432.

928 Hessisches Ministerium fQr Umwelt und Reaktorsicherheit. "Merkblatt 0ber die Verwertung von Schlacken aus Hausm011verbrennungsanlagen," Staatsanzei.qer fQr das Land Hessen, Nr. 28, 1514, 1988. Hjelmar, O. "Assessment of the Environmental Impact of MSWI Ash Disposal in Bermuda (Abstract)", proceedings from Municipal Waste Combustion. Second Annu .a.[ Internatio_nal Specialty Conference. U.S. Environmental Protection Agency and Air & Waste Management Association, Tampa, FI. 1991. p. 283. Hjelmar, O., and K. Ludvigsen. "Management of MSWI Residues in Denmark," Waste Mana.clement Internat.ional. Vol. 2. ed. K.J. Thom~-Kozmiensky., EF-Verl. for Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 133-144. Hjelmar, O. "Regulatory and Environmental Aspects of MSWI Ash Utilization in Denmark," presented at the 3rd International Conference on Ash Utilization and Stabilization (Ash III), November 1990, Arlington, VA. Hoesel, G., W. Schenkel and H. Schnurer, eds., "MQIIverbrennungsaschen (MV Asche)", Merkblatt for die Verwendung von industriellen Nebenproduckten im Strassenbau, Forschungsgesellschaft fQr Strassen- und Verkehrswesen, MQII-Handbuc.h, Berlin: Erich Schmidt Verlag, Kennzahl 8667, LFG. 5/91, 1986. Hoesel, G., Schenkel, W. and Schnurer, H., eds., "Verwertung von festen VerbrennungsrQckst~nden aus HausmQIIverbrennungsanalgen, L,~nderarbeitsgemeinschaft Abfalr', LAGA-Merkblatt: MQII-Han.d..buch, Berlin: Erich Schmidt Verlag, Kennzahl 7055, Lfg. 1/84, 1983. Hoffman, F.E., and D.S. Kosson. "Utilization of Municipal Waste Combustor Residues in the United States," Waste Mana,qement International. Vol. 2. ed. K.J. Thom~-Kozmiensky., EF- Verl. fur Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 161 -172. Hudales, J.B.M. "The Use of M.W.I. Fly Ash in Asphalt for Road Construction", in Environmental Aspects of Construction with Waste Materials, J.J.J.M. Goumans, H.A. van der Sloot and Th.G. Aalbers, eds., Elsevier Science B.V., Amsterdam, 1994. Johnke, B. and G. Hoffmann. "Verwertung von RQckst,~nden aus der thermischen Abfallbehandling aus der Sicht des Umweltbundesamtes", VGB Kraftwerkstechnik, 73, 903, 1993. Knoche, M. "Residues of Municipal and Hazardous Waste Incineration - Situation in France," .Waste Mana c]ement International. Vol. 2. ed. K.J. Thom~-Kozmiensky. EF-Verl. fQr Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 145-150.

929 Leenders, P. "Municipal Solid Waste Residues in the Netherlands," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 593-600. Mank, J.A., J. Brulot and W.J. Janssen van de Laak. "Incineration Slag in Road Construction," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 187-196. Mehan, D.G. "Processed Ash Demonstration Project," proceedings from Municipal Waste C.ombustion. Second .....Annual Internat_ional Specialty Conference_. U.S. Environmental Protection Agency and Air & Waste Management Association, Tampa, FI. 1991. pp. 627-641. Musselman, C.N., M.P. Killeen, D. Crimi, S. Hasan, X. Zhang, D.L. Gress and T.T. Eighmy, 1994. The Laconia, New Hampshire Bottom Ash Paving Project, in Environmental Aspects of Construction with Waste Materials, J.J.J.M. Goumans, H.A. van der Sloot and Th.G. Aalbers, eds., Elsevier Science B.V., Amsterdam. Nonneman, D.J., F.A. Hansen and M.H.M. Coppens. "The Use of Incinerator Slag in Asphalt for Road Construction," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 569-578. Oberste-Padberg, R., A. Roeder, J. Herbig and H. Motzet. "Alinite Cement Made From Incineration Residues," Was.re Mana.ciement International. Vol. 2. ed. K.J. Thom~-Kozmiensky., EF-VerI. fur Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 413- 420. Pihl, K.A., P. Ahrentzen and K. Kalsmose. "Subbase of Incinerator Residue. Guidance - Standard Specifications - General Working Procedure." MiljoministerieVSkov - og Naturstyrelsen/Statens Vejlaboratorium. Labortorierapport nr 66, Vejdirectoratet (in Danish). 1989. Sawell, S.E., H.G. Rigo, A.J. Chandler and S.A. Hetherington. "Downstream Effects of Lead and Cadmium Spiking of Municipal Solid Waste on Incineration Residues," Waste Mana.qement International. Vol. 2. ed. K.J. Thom~-Kozmiensky., EF-VerI. for Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp.59-68. Schmidt, M. and P. Vogel. "The Use of Industrial By-Products with Hydraulic Binders: Refuse Incineration Ashes as an Example," in W__asteMaterials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 177-186. Steketee, J.J. and J.H. De Zeeuw. "Certification of MSW Slags as a Road Construction Material," in Waste Mat.erials in Cons.tru.ction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 379-380.

930 Steketee, J.J., R.F. Duzijn and J.G.P. Born. "Certification of Municipal Solid Waste Incinerator Slags," Waste Mana.ciement International. Vol. 2. eds. K.J. ThomeKozmiensky., EF-Verl. fur Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 439-446. Steketee, J.J. and L.G.C.M. Urlings, 1994. "Enhanced Natural Stabilization of MSW Bottom Ash: A Method for Minimization of Leaching", in Environmental Aspects of Construction with Waste Materials, J.J.J.M. Goumans, H.A. van der Sloot and Th.G. Aalbers, eds., Elsevier Science B.V., Amsterdam. Teekman, G.A.O. "The INDAS Foundation, an Innovative Route for the Utilization of Industrial Ashes," in Waste Materials in Constr.uction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 655-658. Toussaint, A. "Verwertung von Aschen aus der Abfallverbrennung", in Technik for Um.weltschutz - Envitec'89 Kon.qre-band (eds. Klose, W., and Vehlow, J.) Essen, Vulkanverlag, 62. 1989. Van Houdt, J.J., E.J. Wolf and R.F. Duzjin. "Composition and Leaching Characteristics of Road Construction Materials," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 153-160. van den Berg, M.J.A., P.M. Eckhart and W.P. Bijl. "Standardization of Terminology, Characterization Methods, Acceptance Procedures and Leaching Tests for Waste Materials," in Waste Materials in Construction. eds. J.J.J. Goumans, H.A. van der Sloot, Th.G. Albers., Elsevier, The Netherlands. 1991. pp 265-274. van der Sloot, H.A., R.N.J. Comans, T.T. Eighmy and D.S. Kosson. "Interpretation of Municipal Solid Waste Incinerator Residue Leaching Data in Relation to Utilization and Disposal," Waste ManaQement International. Vol. 2. ed. K.J. Thom~-Kozmiensky., EF-VerI. fur Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 99-114. Vaquier, A. and S. Julien. "The Combined Use of Incinerated Household Rubbish Ash and Silicoaluminious Ash in Concrete," in Waste Materials in Construction. eds.. Goumans, van der Sloot, Albers, Elsevier, The Netherlands. 1991. p 631- 634. Wiles, C.C. "Research and Regulations on Management of Residues from Municipal Solid Waste Combustion," Waste Management International. Vol. 2. ed. K.J. Thom6-Kozmiensky., EF- Verl. f0r Energie und Umwelttechnik, ISB 3-924511-64-0, Berlin. 1992. pp. 151-160. Worner, T. and E. Westiner. "Characterization in Quality of Incineration Ash for the Use in Road Construction," Waste Management In.ternational. Vol. 2. ed. K.J. Thom6-Kozmiensky., EF-Verl. fur Energie und Umwelttechnik, ISB 3-924511 -64-0, Berlin. 1992. pp. 447-456.

931

C H A P T E R 22 - D I S P O S A L 22.1

INTRODUCTION

Waste management policy is often expressed in the following sequence of declining priorities (see also Chapter 2): Minimisation of waste production, hazard potential and energy consumption through substitution and cleaner technology measures 2.

Recycling or utilisation

3.

Incineration with energy recovery

4.

Landfilling

Although ranked third and fourth in this sequence of priorities, incineration and landfilling are playing and will continue to play important roles in waste management in many parts of the world. The incineration process itself is obviously not a final waste treatment stage, and the various incineration residues must themselves be utilised or landfilled. In accordance with the sequence of priorities, utilisation of residues is, in principle, preferred to landfilling, provided this does not give rise to unacceptable environmental impacts or health hazards. In practice, there are numerous factors which act as obstacles to utilising incinerator residues: existing regulations lack of economic incentives liability issues residue separation practices uncertainties concerning the functional properties of the residues uncertainties concerning the evaluation of the extent and acceptability of the environmental impacts and health hazards Therefore, landfilling or storage are the predominant MSW incinerator residue management options presently available in some countries. For example, in the U.S.A., most of the incinerator residues currently produced are landfilled as combined ash, whereas in some European countries (e.g. Denmark, France, Germany, The Netherlands), significant quantities (40 to 60% or more) of the bottom ash from the incinerators are being utilised for road construction and similar purposes (see Chapter 21). APC system residues are landfilled in most countries, although in The Netherlands approximately 50% of the fly ash generated is used as a filler in asphalt. Incinerator residues may, in some cases, have to be treated prior to or during disposal in order to reduce the risk of unacceptable environmental impacts.

932 The major potential environmental impacts of concern in connection with disposal of incineration residues are those associated with the formation and release of leachate and with fugitive dust emissions. Fugitive dust problems generally occur only during the relatively short period of actual deposition in a landfill. The use of covered or closed transport containers and maintaining an adequate moisture content in the residues, are generally considered to be effective measures in controlling fugitive dust problems. The formation of leachate may, in contrast, constitute both a short and long-term problem which should be minimised through the application of a proper disposal strategy, and through appropriate design and operation of the landfill. Incinerator residue or ash disposal is therefore discussed primarily in terms of leachate production, management and fate in this chapter. The emphasis of the chapter is placed on the relationship between the leaching behaviour of the residues and the applied or underlying disposal strategies, and specific design and operations issues are merely outlined and not discussed in detail. Interested readers are referred to textbooks and manuals on these subjects. 22.2 CHARACTERISTICS OF INCINERATOR RESIDUE LANDFILL LEACHATES 22.2.1 Overview of Incinerator Residue Leachability From a technical perspective, it is evident that the development of strategies for disposal of incinerator residues and management of the leachate should be based on extensive knowledge of both the short and long-term leaching behaviour of the particular residue streams in question. In this context, "short-term" may cover a time period of 25 to 50 or 100 years and "long-term" consequently represents the following several hundred to thousands of years. A reasonable amount of information is available on the short-term behaviour of most residues, and the evaluation of the shortterm behaviour may to a large extent be based on the results of laboratory and pilotscale leaching experiments and field observations. The long-term behaviour of incinerator residues is less understood. Due to the lack of direct observations, the evaluation of the long-term behaviour is more complicated and requires a synthesis of information obtained from laboratory testing of fundamental leaching behaviour, leaching tests simulating long-term disposal conditions, field measurements, and hydrogeochemical modelling of mineral changes and speciation. The results of ongoing and future research in this field may be expected to reduce the substantial degree of uncertainty with which predictions of the long-term behaviour of incinerator residues are currently made. Table 22.1 presents an overview of the maximum levels of concentrations of inorganic salts, trace elements and nonvolatile organic carbon (NVOC) observed in initial leachates from the major types of incinerator residues including: Bottom ash (usually including grate siftings and boiler ash)

933 Fly ash and mixtures of fly ash and acid gas scrubbing residues from the semidry process and dry lime injection process, and A mixture of fly ash and sludge from treatment of the wastewater from the wet scrubbing process with lime and trimercaptotriazine, TMT (Reimann, 1987) Table 22.1 Maximum Concentration Levels of Contaminants in Leachates from Various Incinerator Residues Typical maximum levels of concentration in leachate

bottom ash

10 - 100 g/I

fly ash and residues Mixture of fly ash from dry and semi- and sludge from dry APC processes wet APC process Na, K, Pb

1 - 10 g/I

SO42, cr, Na, K, Ca

100 - 1000 mg/I 10 - 100 mg/I

NVOC, NH4-N

NVOC, SO42, Zn

1 - 10 mg/I

Cu, Mo, Pb

Cu Cd, Cr

100 - 1000pg/I

Mn, Zn

As, Mo

10 - 1 0 0 IJg/I

As, Cd, Ni, Se Cr, Hg, Sn

1 - 10 #g/I

CI, Na, K SO42, Ca

NVOC, Mo As, Cr, Zn Pb

,,,< 1 IJg/I Cd, Cu, Hg Hg Hjelmar, 1991, 1992 and 1993; Thygesen et al., 1992; Hjelmar et al., 1993 The maximum concentrations shown in the table represent data from a number of laboratory leaching tests (mostly column leaching tests) and a few field investigations (Hjelmar, 1991, 1992 and 1993, Thygesen et al., 1992, Hjelmar et al., 1993). The maximum concentrations occur in the initial leachate for most parameters (note, however, the sulphate curve in Figure 22.1), and most of the concentrations in Table 22.1 have been observed in fractions of leachate collected at or below liquid/solid or leachate/waste ratio (L/S) = 0.5 I/kg. (e.g. expressed as cubic metres of leachate produced per tonne of waste or residue deposited). For a particular disposal site, S will be constant and L will increase as the leachate is formed. An L/S scale may therefore be transformed to a time scale if the rate of percolation or flow through the site is known. It is often practical to express field data and experimentally determined data on leachate quality as a function of L/S for each particular system in question. Such data may subsequently be used (with care) in conjunction with additional information (e.g. on pH and redox conditions) to provide estimates of leachate quality as a function of time at a disposal site which contains waste/residues with similar properties and for which the rate of percolation of water is known. At low L/S values, the leaching of

934 several contaminants, particularly trace elements, is solubility controlled and strongly influenced by the pH of the leachate (which in turn is governed by the major constituents of the incinerator residues and local conditions). As shown in Chapter 16, incinerator residues exhibit systematic leaching patterns, and the leaching behaviour of several contaminants is controlled by such factors as pH, redox potential, ionic strength, complexing inorganic ions and organics, the presence of various minerals, etc. Once the relationship between the controlled contaminants and the controlling factors has been established, it becomes important to be able to predict how these controlling factors may develop within an incinerator residue disposal site. The estimation of contaminant release during disposal of incinerator residues is discussed in detail by Kosson et al. (1996). 22.2.2 Bottom Ash Leachate

Table 22.2 shows the development over a period of 21 years of the quality of leachate from a Danish ash monofill, (Vestskoven), containing approximately 10,000 tonnes of ash (85% bottom ash, 15% fly ash). The data was derived from annual analyses of the leachate which is contained by a PVC bottom liner, conducted to a central pumping sump and pumped out at regular intervals. In 1994, a total of 7300 m3 of leachate had been removed from the monofill, and the historical characteristics have been described extensively elsewhere (Hjelmar, 1987, 1989, 1991 ). The concentrations of some of the major components and trace elements in the leachate are shown as a function of L/S in Figure 22.1. For bottom ash, the first leachate has a relatively high content of inorganic salts (chloride, sulphate, sodium, potassium, calcium). The content of dissolved organic matter (measured as NVOC) and ammonia are both associated with the residual uncombusted organic material and may vary considerably. The pH is usually slightly to strongly alkaline, depending on the degree of carbonation. The redox potential is low (reducing) due to microbiological degradation of the residual organic material. The concentrations of trace elements in the leachate are low due to the reducing environment (most of them form relatively insoluble sulphides) and a solubility limiting pH regime. The content of uncombusted organic material in incinerator bottom ash may be one of the key factors in controlling the hydrogeochemical conditions and hence the behaviour of trace elements within a bottom ash landfill. It may be a delicate balance, i.e., sufficient organic material should be present to support a certain level of microbiological activity, but the organic and degradation products should not be too high as to dominate the composition of the leachate. In order to minimise the mobility of trace elements, the amount of acidic degradation products must not be significant in comparison to the acid neutralisation capacity of the bottom ash. Dissolved organic material, per se, is an undesired contaminant, and it has further been shown to be associated with increased mobility of Cu in bottom ash (van der Sloot et al., 1992).

935 Figure 22.1 C o m p o s i t i o n of Leachate from a Danish Incinerator and Fly Ash Monofill as a Function of L/S over a Period of 20 Years 1o0oo0 E

r ~

E u 4-I r J= U r (9

10000 ]

~ ,~176176 .,~

..........................................

...........

1

100

0

'~

0.2

..........................................

0.4

0.6

0.8

1

100000

i

E O o,_ O t-t

E

0 O

10000

;~;~.~.,.,.~~.::.~.:;.~.:.:~:~~ .............................

1000

SO4

100

o 4-J

Ca IIB--

..................................... ;i ........ ............................. ......... ~.-".".q.~;~:::::~..~::.. ::-..,.~ ...... ~:~ ................................

10

O cO

"o

.J

0

0.2

0.4

0.6

0.8

'I

N H4"I~

u,~,loo

1

,,, i. O=

E

v

t~O .4m. l O CL

E

O. 1

......................................

G....___E~" •

(~.

J; "

""

"~

Cd

ool

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

~

O

0

As

tS-.-?~,:.~ ..............................

-~

-...: ..... .~..~;~_~...~::

0.001

Cu

................................................

o .J

Pb

ll=,~

0.0001

i

i

"1

0

0.2

0.4

"'

'

i

i

0.6

0.8

Average US (l~g)

Hjelmar, 1995

'

I

936 Table 22.2 Overview of a Time Series of Leachate Quality Measurements from a 21 Year Old Danish Bottom and Fly Ash Monofill Parameter

Unit

Variation 1973-94 (24 observations)

pH Alkalinity Redox potential, Eh Conductivity

Average Values 1973/74

1991/92

-

8.3

-

10.5

8.8-10.1

8.9-10.2

meqv/I

1.4

-

9.3

2.5

7.4

mV

-10

-

-290

-66

-

mS/m

1300

-

3900

3100

1900

BODs (from 1981 )

mg/I

<2

-

26

-

2

Sulphate

mg/I

2000

-

7200

3100

6100

Chloride

mg/I

1600

-

11400

9300

3300

Ammonia-N

mg/I

1.5

-

87

39

3.9

Na

mg/I

2600

-

7300

5600

3600

K

mg/I

600

-

4300

3900

800

Ca

mg/I

32

-

1000

670

58

As

mg/I

0.004

-

0.025

0.014

0.010

Cd

mg/I

<0.0001

-

0.001

<0.003

<0.0002

Cr

mg/I

<0.001

-

0.08

0.03

<0.002

Cu

mg/I

<0.0005

-

0.21

0.013

0.018

Fe

mg/I

<0.01

-

1.6

0.21

0.055

Hg

mg/I

<0.00005 -

0.003

0.00008

0.0004

Pb

mg/I

<0.0005

-

0.04

0.0013

0.007

Zn

mg/I

<0.01

-

0.59

0.05

0.09

gmol/I

0.18

-

0.48

0.41

0.23

-

0.730

0.027

0.602

Ionic strength

Accumulated L/S I/kg 0.017 The content of fly ash is estimated at 15%

There is a tendency for the concentrations of most salts in the leachate to decrease as the leaching progresses (Table 22.2 and Figure 22.1), however, sulphate concentrations actually increase due to the decrease in calcium concentrations. The pH remained between 8.5 and 10.5 (the lowest values probably reflect non-optimal sampling conditions) throughout the period. Reducing conditions have been maintained and the concentrations of trace elements have remained low during the 21 years of observation at this particular site. The low short-term concentration levels of most trace elements have been confirmed by observations at other ash landfills

937 (Hjelmar, 1989). In Table 22.3, the total amounts of some major and trace components leached from the Danish landfill site over the 21 year period from 1973 to 1994 are presented and compared to the estimated total amounts of those components initially present in the landfilled ash. The data indicate that approximately 40% of the chloride, 30% of the sulphate, 9% of the sodium, 6% of the potassium and 0.3% of the calcium have been leached, whereas the amounts of trace elements leached are negligible compared to the total amounts present. Table 22.3 Total Amounts of Major and Trace Components Leached from a 21 Year Old Danish Incinerator Ash Monofill Compared to the Estimated Total Contents of these Components . . . . . . . . . . . . . . Parameter

Unit

Amount leached in 21 years (US = 0.73 I/kg)

Estimated total content in the landfilled ash

Chloride Sulphate Na K Ca

tonnes tonnes tonnes tonnes tonnes

35 32 31 10 2.1

90 110 340 180 700

As Cd Cu Pb

kg kg kg kg

0.093 <0.008 0.25 0.065

270 400 39000 31000

22.2.3 Fly Ash and Acid Gas Scrubbing Residue Leachate An indication of the leaching properties of fly ash and acid gas scrubbing residues may be obtained from the results of laboratory studies (3 fly ash samples, 2 dry, 2 semi-dry APC residues and one wet process product, (Hjelmar, 1992 and 1993)). Field observations and large scale lysimeter leaching tests have subsequently confirmed these findings (Andersen and Boll, 1994) (See Chapter 11 and Chapter 16). The amounts of some major and trace components leached from the above mentioned residues in combined column and batch leaching tests (L/S = 0-25 I/kg) are shown in Table 22.4. The compositions of the first and last fractions of leachate collected in the laboratory leaching experiments on the dry/semi-dry and wet scrubbing process residues are shown in Table 22.5. The results indicated that the acid gas scrubbing residues from the semi-dry and dry processes contain 20 to 35% (w/w) readily soluble material. Most of the soluble material consists primarily of chlorides and hydroxides of calcium, sodium and

938 potassium, which appear in the first few fractions of leachate (L/S = 0 to 2 I/kg). The most leachable metals/trace elements are lead (2.3 to 65%(w/w) of the total content) and molybdenum (9 to 19% (w/w) of the total content). All other trace elements were less than 4% (w/w) leachable under these circumstances, and most of them far less than 1% (w/w)leachable. The high leachability of lead is caused primarily by chloride complexation, but the amphoteric behaviour of lead also plays an important role at pH values above 10. No releases of mercury, nickel and tin were observed. Concentrations of soluble matter are very high in the first leachate fractions from the dry and semi-dry products but level off to moderate and low values in later fractions (see Table 22.5). The leachability of the fly ash resembles that of the dry and semi-dry acid gas scrubbing residues, although the leaching of Ca and the alkalinity of the leachate is much smaller for the fly ash. The pH of the fly ash leachate is generally lower than that of the drylsemi-dry residue leachate which, under certain circumstances, may cause some changes in leachability of trace metals such as Cd, Pb, Zn, etc. Table 22.4 Total Amounts of some Major and Trace Components of Fly Ash and APC System Residues Leached in Combined Column and Batch Leaching Tests for L/S = 0-25 I/kg Parameter

TDS Sulphate Chloride Na K Ca As Cd Cr Cu Hg Mo Ni Pb Sn Zn NVOC

Unit

Fly ash

Residues from dry and semidry APC processes

Residue from wet scrubbing process mixed with fly ash .

g/kg g/kg g/kg g/kg g/kg g/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg

210 - 230 13 - 41 89 - 106 23 - 30 28 - 50 9.1 - 21 0.10 - 0.19 0.05 - 35 0.05 -0.12 0.03 - 0.31 < 0.01 7.1 - 14 < 0.01 17 - 370 < 1 < 2 - 580 14 - 46

290 - 380 0.47 - 3.1 120 - 200 12 - 17 17 - 29 61 - 109 < 0.02 - 0.04 0.03 - 0.44 < 0.6 - 2.4 0.13 - 22 < 0.003 1.1 - 3.0 < 0.2 2 2 0 - 3400 < 0.4 45 - 340 71 - 780

140 32 56 21 21 17 0.28 < 0.0006 0.52 < 0.004 < 0.001 12 < 0.2 < 0.001 <3 < 0.2 78

Hjelmar, 1992 and 1993

939

Table 22.5 R e s u l t s of C o l u m n a n d B a t c h L e a c h i n g T e s t s on R e s i d u e s f r o m t h e D r y a n d S e m i - D r y P r o c e s s e s ( R a n g e s for 4 p r o d u c t s ) a n d W e t S c r u b b e r P r o d u c t ( S i n g l e D e t e r m i n a t i o n ) Parameter

unit

Residues from Dry and Semi-dry Processes Residue from Wet Process (+FA)

First leachate fraction (L/S = O.0 - 0.2 I/kg) pH

-

9.8

-

10.2

9.0

mg/I

430000

-

520000

116000

Alkalinity

meqv/I

87

-

540

0.96

Chloride

mg/I

1900000

-

310000

65000

Sulphate

mg/I

150

-

430

2100

Ca

mg/I

110000

-

160000

5800

Na

mg/I

7500

-

37000

23000 23000

TDS

K

mg/I

19000

-

66000

Cd

mg/I

0.099

-

1.9

0.006

Cr*

mg/I

0.48

-

2.3

<0.003

Cu

mg/I

0.17

-

37

Hg

mg/I

<0.003

0.0009 <0.0005

Mo*

mg/I

0.31

-

0.61

7.9

Pb

mg/I

2.2

-

11000

0.0019

Zn

mg/I

0.02

-

730

0.02

NVOC

mg/I

24

-

640

8

Last leachate fraction (L/S = 5.0- 25 I/kg) pH TDS

-

12.1

-

12.5

10.6

mg/I

770

-

1900

1600 2.0

Alkalinity

meqv/I

13

-

42

Chloride

mg/I

20

-

36

2.8

Sulphate

mg/I

2.5

-

40

1100

Ca

mg/I

290

-

840

460

Na

mg/I

8

-

24

15

K

mg/I

10

-

41

19

Cd

mg/I

<0.00002

-

0.00006

<0.00002

Cr

mg/I

<0.004

-

0.004

0.023

Cu

mg/I

0.002

-

0.007

<0.0005

Hg

mg/I

Mo

mg/I

0.017

-

0.075

0.027

Pb

mg/I

4.3

-

8.6

<0.0001

Zn

mg/I

0.48

-

1.5

<0.02

na

na

NVOC mg/I 1.9 4.0 2.8 * - Combined fraction L/S = 0.0 - 1.0 I/kg analysed, na - Not Analysed Comparison of the First (L/S = 0.0 - 0.2 I/kg) and the Last (L/S = 5.0 - 25 I/kg) Leachate Fractions Hjelmar, 1992

940 The residue from the wet scrubbing process was shown to have a content of readily soluble material which is considerably smaller than that of the dry and semi-dry residues. The major ions in the initial leachate fractions are chloride, sodium, potassium, and to a lesser extent calcium. Sulphate from the wet scrubbing process is present in appreciably higher concentrations than in the leachates from the dry/semidry products. With molybdenum as the only exception, the concentrations of trace elements/heavy metals are low in all leachate fractions, probably due to the presence of the organic sulphide TMT. The long-term stability and effectiveness of this compound are not known. The highest concentration of lead measured in the leachate from the residue from wet condensing process was 1000 times lower than the lowest concentration measured in the initial leachates from the dry and semi-dry residues. 22.2.4 Combined Ash Leachate

In the US, it is common practice to manage and landfill incinerator residues as mixture of bottom ash and scrubber residue (combined ash). Some of the short-term effects on the leachate quality of adding the highly soluble residues from the semi-dry APC process to the bottom ash prior to disposal are seen in Table 22.6 which presents leachate quality data from a combined ash monofill (the Woodburn Landfill), in Oregon, US (Cambotti and Roffman, 1993). A comparison of Table 22.6 with Table 22.2 shows that the concentrations of soluble inorganic salts, particularly chloride and calcium from the acid gas cleaning residues, are substantially higher in the combined ash leachate than in the bottom ash (and fly ash) leachate. The higher concentrations of Cd (up to 0.6 mg/I as compared to a maximum value of 0.025 mg/I at Vestskoven) which have been observed at the Woodbum Landfill are consistent with the lower pH and the high (complexing) chloride concentrations. The lower pH at the Woodburn Landfill may be caused by organic acids produced by biodegradation of residual unburnt material in the bottom ash which may possibly be higher at Woodburn than at Vestskoven. This would also be consistent with the relatively high levels of iron observed in the leachate from the Woodburn Landfill. Leachate composition data from three other combined ash landfills in the US (Kosson, 1995) confirm that the parameter ranges shown in Table 22.6 are typical of combined ash leachate. 22.3 DISPOSAL STRATEGIES

Due to the potential leaching of contaminants, landfilling of incinerator residues such as bottom ash, grate siftings, boiler ash, economiser ash, fly ash and acid gas cleaning residues or combined ash may have long-term consequences for the surrounding environment. It is therefore important that the disposal solutions chosen for these residues are sustainable in terms of environmental impact, maintenance requirements and energy consumption. This may be achieved only through careful consideration of potential disposal strategies.

941 Table 22.6 Results of Analysis of Leachate from a Combined Ash Monofill, The Woodburn Landfill, in ..Oregon, USA over a Period of 5 Years Parameter

Unit

pH TDS mg/I TOC mg/I Sulphate mg/I Chloride mg/I Ammonia N mg/I Na mg/I K mg/I Ca mg/I As mg/I Cd mg/I Cr mg/I Cu mg/I Fe mg/I Hg mg/I Pb mg/I Zn mg/I TDS 9 Total dissolved solids TOC 9 Total organic carbon DL 9 Detection limit Cambotti and Roffman, 1993

Range of variation 1988 - 1993 5.7 - 7.5 1400073000 4 - 110 80 - 1500 7700 - 50000 < DL - 35 3000 - 9300 520 - 6900 1300 - 16000

22.3.1 General Philosophy The primary objective of landfilling as a waste management technique is to remove from general circulation materials/products that are no longer useful in any respect, and preferably to do this in a manner which eventually returns the basic constituents of the waste to the ecological cycle at levels similar to natural geologic materials. A second and equally important objective of landfilling of waste is to ensure that the landfilled waste does not cause any unacceptable short or long-term impacts on the environment or on human health. This should preferably be done in a sustainable manner, i.e. without excessive and/or prolonged maintenance or operation requirements, and without a need for long-term care. These objectives can be met by disposal strategies which observe the following generally accepted principles: Landfills should be designed to minimise the required lifetime of active environmental protection systems (i.e. systems requiring operation or maintenance)

942 Any disposal strategy should consider the intrinsic properties of the waste, and the potential health risks associated with a given disposal strategy Landfill design, operation and siting should be adapted to the admitted waste in such a manner that long-term emissions of leachate (and gas) become or remain environmentally acceptable. Any strategy for waste disposal at a landfill site must include consideration of the ultimate fate of the leachate and the residues remaining in the disposal site, as well as derived effects of disposal and leachate management. In the following, each of these principles are discussed in relation to landfilling of waste in general and landfilling of incinerator residues in particular. From a philosophical perspective, it could be argued that all expenses and resources required to ensure reliable and sustainable disposal must also be acceptable in terms of public health, long-term environmental impact, and energy consumption, and should be covered by those who enjoy the goods and produce the waste. Therefore, encapsulation/total containment strategies may not be considered adequate to meet these requirements. Since these only postpone the potential impacts until some time in the future when the encapsulation fails. For most waste types, the implementation of a strategy which reduces the contamination potential of the waste to a safe level, either prior to landfilling or gradually during the initial period after landfilling would be ideal. Alternatively, storage for potential recovery of resources sometime in the future could also be considered.

Lifetime of Active Systems Both active environmental protection systems (i.e. systems for leachate collection/removal, transport and treatment of leachate which require maintenance and input of energy) and passive environmental protection systems (i.e. systems such as geologically stable low permeable top covers and barriers and surface drains that do not require maintenance or operation) must have expected lifetimes that are long enough to ensure that they perform as intended. For a strategy based on restricting percolation of infiltrated precipitation through mineral wastes, such as most well combusted incinerator residues, this could in some cases require system lifetimes of several hundred years or more. It would be preferable (but for many types of wastes, not yet feasible) if no active environmental protection systems were required to function beyond a period of e.g. 30-50 years, after which the landfill could be left alone without any risk of unacceptable environmental impacts. A postclosure care period of 30-50 years may be a realistic target, both in terms of legal aspects and expected lifetime of technical systems, and would also be in agreement with the concept that each generation should care for its own waste.

943

Waste Properties

There may be fundamental differences between leachates produced by different types of waste. For example, leachate from relatively stable mineral waste (e.g. well combusted bottom ash) behaves very differently from that produced by more reactive, biodegradable types of waste (e.g. raw MSW such as domestic waste, garden refuse). These differences should be reflected in the disposal and leachate management strategy chosen for each type of waste. Types of waste that are incompatible in terms of disposal strategy should be directed to different categories of landfills. The normal course of landfill degradation of MSW dominated by organic, biodegradable waste may be described as an initial aerobic phase followed by the anaerobic acetogenic and methanogenic phases during which the organic compounds in the waste are broken down to simpler molecules such as short chain carboxylic acids and amino acids and eventually to ammonia, hydrogen, carbon dioxide and methane (Knox, 1992). The remaining, more persistent organic compounds may eventually be transformed into humic substances. These degradation processes normally occur in so-called "sanitary" landfills which have traditionally been used for the disposal of MSW. During the first several years after landfilling, i.e. during the aerobic and acetogenic phases and well into the methanogenic phase, the leachate generated will contain relatively high concentrations of organic compounds. Therefore, it makes sense, both from an environmental and a technical perspective, to collect the leachate and subject it to biological and perhaps chemical treatment prior to discharge, at least during the initial stages of waste degradation. This is a common practice for "sanitary" landfills. The biological wastewater treatment effectively removes a substantial part of the readily degradable organic contaminants from the leachate, and the degradation/mineralisation processes within the landfill will gradually reduce the pollution potential of the landfilled waste. It is generally assumed (but rarely verified) that traditionally landfilled MSW as well as several other wastes will become harmless in a relatively short-time, and that a landfill therefore may be safely abandoned and forgotten after a period of, e.g. 30 or 50 years. However, neither the criteria for determining whether abandoning a site is safe, nor the length of time needed to reach this point are generally well defined or known. Both depend strongly on the exact nature of the waste and on local hydrologic conditions, the landfill and the surrounding environment. Based on a number of assumptions, Belevi and Baccini (1989) have calculated that it may take 500 - 1700 years before the content of organic C in the leachate from a traditional "sanitary" landfill has been reduced to a level of 20 mg/l. They have also calculated, that it may take 55 - 80 years for the concentration of NH3+NH4§ to fall to 5 mg/I, 100 - 700 years for P to fall to 0.4 mg/I and 100 - 150 years for CI- to fall to 100 mg/l. In relation to groundwater and surface water protection, it is often the concentrations of ammonia which remain high over a considerable period of time that are of major concern. Inorganic waste types are not subject to biological degradation and mineralisation processes, although they may be influenced by biological activity in the residual

944 organic matter or co-disposed organic waste. The relevant processes (Chapter 13), include chemical reactions such as hydration, carbonation, reduction, oxidation, dissolution/leaching, precipitation, etc. Like biodegradation, most if not all of these reactions will only occur in the presence of water. In addition, they are influenced by such factors as pH, redox potential, the amount of water percolating or leachate generated (e.g. expressed in terms of the liquid/solid ration, L/S), temperature, the presence of complexing or chelating organics and by time. In the long-term, diagenetic changes in ash mineralogy will occur. The time horizon for these processes may be very long, from decades to perhaps several thousands of years. The processes which are significant in relation to the risk of emission from landfilled inorganic waste and contamination of the environment may, in some cases, proceed within a few decades and strongly reduce the risk of further contamination from the landfill. In other cases, it will be uncertain whether or not even prolonged exposure to leaching has reduced the risk of contamination and whether or not the contamination may not actually increase with time, e.g. if the local conditions changed. In a recent study (Hjelmar et al., 1995), a very simplified estimation indicated that for a 12 m high landfill and an assumed rate of infiltration/production of leachate of 200 mm/year, minimum periods of approximately 300 years and 100 years might be required for landfilled MSW and some inorganic waste types (e.g. bottom ash), respectively, to reach "final storage quality", i.e. a condition which allows the site to be safely abandoned without active environmental protection systems. An increased rate of infiltration may shorten these time periods, whereas a decreased rate of infiltration may lengthen them. Certain types of inorganic or mineral wastes will generate leachates which initially have a relatively high concentration of inorganic salts and a moderate to low content of trace elements. As the leaching progresses with time, both the salt content and the concentration of trace elements may gradually decrease to very low values. The content of organic substances in the leachate is often very low. The application of traditional "sanitary" landfilling techniques, i.e. installation of (multiple) bottom liners, collection and subsequent treatment of leachate at a biological wastewater treatment plant, to mineral wastes exhibiting such properties would generally not constitute an optimum or a sustainable solution for a number of reasons. First, the potential period of leaching may easily exceed the projected lifetime of the liners and the leachate collection system, and pumping the leachate for treatment (particularly biological) are likely to be both energy consuming and ineffective. For such wastes, a controlled contaminant release management strategy may in some cases be more appropriate. This strategy implies that the transfer of contaminants from the landfilled material into the surrounding environment is limited to an acceptable level. This may be accomplished by controlling the quality and/or quantity of the leachate which is generated and is subsequently released without being collected and treated. The disposal strategies applicable to organic and inorganic waste types are obviously very different and generally incompatible, both in the short and long-term. Co-disposal

945 of organic and inorganic types of waste is therefore generally not advisable. Different types of incinerator residues may also exhibit very significant differences in behaviour when landfilled, which means that it may often be advantageous to handle and landfill each type separately.

Adaptation of Landfill Design, Operation and Siting to Strategy and Waste Types

Since the avoidance of prolonged postclosure care is one of the objectives of landfilling, it may be necessary to design and operate some categories of landfills in two stages: An initial, relatively short and very active stage during which the contamination potential of the waste is reduced to an acceptable level, A subsequent, passive long-term stage during which the contamination potential remains at or below an acceptable level. During the active stage, mitigating measures may be taken to enhance the processes which reduce the contamination potential of the waste. These measures may include accelerated leaching due to increased infiltration of precipitation or use of irrigation systems. Active environmental protection systems such as leachate containment, collection and treatment systems must remain fully operative during this period. When the landfill changes from the active first stage to the passive second stage, the active environmental protection systems are no longer required, and they must cease to function or undergo changes to comply with a controlled contaminant release if this is warranted by the changes in leachate composition. Landfilling of organically dominated waste types must generally be operated in two stages; an initial mineralisation stage with active environmental protection systems and a subsequent second stage with passive environmental protection systems. The required duration of the active mineralisation stage is uncertain but may easily be several decades. Landfilling of inorganic, mineral wastes may be operated in one or two stages. For mineral wastes with a high content of soluble contaminants (like some APC residues), a two-stage landfill operation may be necessary: An initial stage based on (possibly enhanced) leaching of contaminants with active environmental protection systems followed by a second stage, based on controlled contaminant release and requiring only passive environmental protection systems. For mineral wastes containing only limited amounts of soluble contaminants, particularly trace elements, one-stage landfill operation based on controlled contaminant release with only passive environmental protection systems may be sufficient. For some mineral wastes which do not initially qualify for one-stage landfill operation, treatment prior to disposal may present an alternative to the first active stage and render one-stage landfilling based on controlled contaminant release feasible. The treatment could consist of extraction of soluble

946 contaminants from the waste and/or stabilisation with a binder. Extraction would reduce the total pollution potential of the waste, whereas proper stabilisation would also reduce the rate of release of contaminants substantially. When the disposal strategy is based on controlled contaminant release, a sufficiently slow rate of transfer of contaminants from the landfill to the surrounding environment may be achieved either by ensuring that the concentrations of potential contaminants in the leachate are sufficiently low or by restricting the rate of generation and emission of leachate to an acceptable (low) level. Properly designed measures aimed at controlling the composition of the leachate (e.g. extraction and/or stabilisation) are generally secure but they also tend to be relatively complicated and energy consuming. Passive environmental protection measures (e.g., geologically stable top covers and surface drainage systems) which are aimed only at restricting the quantity of leachate produced are technically much simpler, but are dependent on long-term durability and functionality of the systems. However, for passive systems, long lifetime expectancies are not unrealistic. The transformations of active environmental protection systems into passive systems are required when a landfill based on two-stage operation passes from the active initial stage, which may depend on collection and treatment of leachate, to the final, passive stage which may depend solely on a controlled rate of release of relatively benign leachate. In this case, an impermeable bottom liner apparently changes status from being an instrument of one strategy to become an obstacle to another strategy. One way to avoid this is shown in Figure 22.2. Drainage layers are built into the sides of the landfill, sufficiently high up to keep them inactive by maintaining a low level of leachate on top of the liner during the active stage of landfill operation. Once the removal of leachate is discontinued, the level will rise and the drainage layers will provide conduits for dispersion of leachate into an acceptable receiving environment (Johannessen et al., 1993). Proper siting is a crucial part of any disposal strategy, particularly if controlled release or passive discharge is envisaged from the start or during a later period of the existence of the landfill. Siting may strongly influence the criteria for acceptability of leachate in the surroundings. No disposal site should be placed on top of or immediately upstream of a valuable, sensitive aquifer or adjacent to a sensitive surface water body. However, if a saline (initial) leachate is expected, a location of a landfill near a wastewater conveyance or at the coast of a receptive marine environment would appear suitable (if available) both in case of controlled contaminant release and leachate collection, treatment and discharge. Ultimate Fate of the Leachate Regardless of whether the leachate is collected, treated and discharged, or it is allowed to disperse into the surrounding environment, an energy and resource consumption analysis should always be conducted. This analysis should also include an

947 environmental impact assessment covering the entire pathway of the leachate (until it is indistinguishable from the surroundings) for the entire leachate production period. Figure 22.2 Example of Design which Allows the Transformation of an Active Environmental Protection System into a Passive System

Clay

barrier

Embankment~k~

'~

~" ~ ~ ~176

:'~.!;~,.io'~~. ~

,,~\~,,,,~\~,,~\~,,~\~>-~,,\~:;,~\~,,,~,~,\~;,,,~,~ ,,~;,~\-~,,~-,r . High

permeable

9

Ground water table

Drainage layer .-.....:..,

~ . . . .

~,~,\~/~\,~X\,,,,~\,,~

Johannessen et al., 1993 Derived effects such as the environmental impact of producing the energy necessary to pump and treat the leachate for the prescribed period of time should be accounted for, and environmental impacts caused by, but remote from, the landfill (e.g., at the outfall from a wastewater treatment plant which receives the leachate) should also be considered. Any environmental protection measures at a landfill should therefore be subjected to a lifecycle analysis prior to implementation. This may help prevent the application of landfill solutions which are based on suboptimisation in space or time.

22.3.2 General Disposal Strategies Disposal strategies may be categorised according to the prescribed management and intended fate of the leachate. Some of the specific strategies which may be relevant or have been applied to the disposal of incinerator residues include:

948 Total containment or "entombment" (dry storage) Containment and collection of leachate Controlled contaminant release Unrestricted contaminant release Total containment and containment with leachate collection generally require active environmental protection systems, whereas the controlled contaminant release and unrestricted contaminant release strategies may require only passive environmental protection systems. Active environmental protection measures may be necessary during a first stage of landfilling in a number of cases but only strategies based on passive systems are sustainable in the long-term. The four leachate management strategy scenarios are presented schematically in Figure 22.3. Total Containment or Entombment

Total containment of incinerator residues will prevent any infiltration and percolation of water and, consequently, any generation and emission of a leachate provided the containment system remains intact (see Figure 22.3, A). The main weakness of this strategy is that the landfilled residues and hence the potential risk to the environment may remain virtually unchanged for a very long period, until the containment system finally fails and an uncontrolled plume of leachate may be released. This is particularly true for residue landfills equipped with (multiple) impermeable, artificial bottom and top liners and relying on zero discharge of leachate. The environmental risk is less pronounced for storage of untreated residues (such as air pollution control (APC) residues) in old salt mines which is practised in some parts of Germany. Underground storage is, however, a relatively costly solution which is not generally available. A total containment strategy may be acceptable for temporary, short-term storage of residues, but it should always include plans or development of plans for appropriate final disposal or utilisation of the residues. Containment and Collection of Leachate

A strategy based on containment with leachate collection and treatment is the standard method of designing and operating a sanitary or MSW landfill (see Figure 22.3, B). The leachate generated is contained by an impermeable or low permeable bottom liner, recovered and normally subjected to treatment prior to discharge to a body of water. The rate of leachate formation may be reduced by an impermeable top cover layer. Although commonly used, this strategy may not be optimal for incinerator residues since it requires maintenance and operation for a period of time exceeding the expected lifetime of the environmental protection systems. Also, effective treatment of the leachate is likely to be difficult and energy consuming. Although the strategy may not be a long-term solution to landfilling of incinerator residues, containment and collection of leachate may be applied as a first stage of operation, for example, in connection with measures designed to enhance leaching. However, eventually it

949 Figure 22.3 Disposal Strategies Categorised in T e r m s of L e a c h a t e M a n a g e m e n t and Fate

ttltltl GWT

ltttltl ','",. .. . . . . . .

J :.::

!

GWT

t

t

t

~`~!iiiii~iii~ii~iiiii`~i~ii~!iiiiiii~iii~;:i~!~`~i:`~i~:~`~i!:ii~i;~Lii~`i~i~:iii~i!iiii~`~`:i~ii~:~ GWT

t

t

t ii~i!iLLii':'~iiiiiiii

L----L----L-

A - Total Containment C - Controlled Contaminant Release GWT - Groundwater Table

~w~

B - Containment and Collection + Treatment of leachate D - Unrestricted Contaminant Release

950 should be replaced by a second stage of operation, based on a more long-term oriented strategy.

Controlled Contaminant Release The controlled contaminant release strategy implies that the release and discharge of contaminants are maintained at an acceptable level by controlling the quantity and quality of the leachate generated within the landfill (see Figure 22.3, C). The leachate is allowed to discharge into the surroundings or to be passively discharged to a wastewater conveyance as it is formed. An assessment must always be carried out to ensure that the impact of the emitted leachate on the environment is acceptable. Both the quantity and quality of leachate depend upon the characteristics of the waste, the design and operation of the landfill and the climatic conditions. Treatment of the waste may reduce both the contamination potential and the permeability. Installation of geologically stable, sloped top covers with surface drainage systems could ensure a very low rate of infiltration of precipitation and, consequently, a very low rate of release of contaminants from a disposal site in the short and long-term. This concept requires proper siting. Since contaminants are being removed from the landfilled material, a continuous reduction of the contamination potential will occur. The controlled contaminant release strategy may represent a sustainable and therefore preferable long-term solution to disposal of residues. For some residues, a final controlled contaminant release stage of operation should be preceded by a short-term, active stage of operation based on a different strategy, e.g. enhanced leaching and containment and collection of leachate. This strategy currently is applied to disposed residues in some countries. Unrestricted Leaching An unrestricted leaching strategy may simply be described as a landfill scenario where no precautions at all are taken to prevent or reduce the generation and emission of leachate (see Figure 22.3, D). This strategy represents in most respects the opposite of the total containment strategy. The environmental impact will depend on the leaching characteristics of the landfilled waste as well as on local physical and climatic conditions and the vulnerability of the surrounding environment. Since the strategy implies a total lack of control, it is generally unacceptable for landfilling of incinerator residues except in cases where the leachate has reached ambient quality. 22.4 DESIGN AND OPERATIONS ISSUES Although the disposal strategies employed for landfilling of MSW incinerator residues may (or should) differ somewhat from those employed for traditional landfilling of raw MSW, most of the general considerations concerning design criteria and operational conditions of landfilling are quite similar in both cases. A detailed discussion of design and operations issues is, as mentioned previously, beyond the scope of this treatise.

951 The reader is referred to the appropriate textbooks and manuals on this subject which is often covered in great detail by national or regional legislation and guidelines. The following is merely a brief listing of some of the more important issues that must be dealt with before, during and after the landfilling of incinerator residues (and most other wastes).

22.4.1 Siting Siting criteria are frequently utilised in the process of determining the location of disposal facilities. The type of disposal strategy being used (i.e. uncontrolled leaching, containment and leachate collection, controlled contaminant release, encapsulation/total containment) and the type of residue requiring disposal influence the siting criteria. Frequently criteria can include: Geotechnical Issues (slope, stability, seismic activity, soil bearing capacity) Hydrogeological and Environmental Issues (overburden soil type, overburden/bedrock interactions, groundwater direction and velocity, travel time, separation from potential groundwater receptors, background groundwater quality and vulnerability, separation from potential surface water receptors, background surface water quality and vulnerability) Civil Engineering Issues (traffic control, property abutment, site workability, site stormwater management).

22.4.2 Liners and Leachate Collection Systems Landfills for which a leachate containment and collection strategy is chosen must be equipped with bottom lining and leachate collection systems. Leachate containment may be achieved using mineral liners (natural and/or artificial), synthetic liners or composite liners. Regulations or guidelines may specify hydraulic conductivity, liner thickness, type, materials, multiple liner systems, etc., often depending on landfill category (the type of waste received) and/or the vulnerability of the environment. Leachate collection systems typically consist of a layer of coarse gravel, containing drainage pipes leading to collection sumps. The allowable head of leachate on the bottom liner is typically restricted (e.g. to 0.3 to 1.0 m). Sections of a landfill which are segregated in terms of waste types accepted should have separate leachate collection systems. The malfunction of drainage systems may occur due to clogging, and one separate system should not cover too large an area of the landfill. There is only limited data available on the long-term performance of liners (particularly synthetic liners), drainage systems and collection systems.

952 In Sweden, for example, several small, well-defined catchment areas exist. In the opinion of the Swedish EPA, it is possilbe to regard such areas as part of the leachate containment system without getting in conflict with the proposed European Union Landfill Directive (EU, 1995). According to the EU Landfill Directive each landfill shall be classified in one of the following classes: Hazardous waste, nonhazardous waste or inert waste. A landfill must be situated and designed so as to meet the necessary conditions for preventing pollution of the soil, groundwater or surface water. The geological barrier is determined by geological and hydrogeological conditions, see Table 22.7. In this table the corresponding vertical flow time has been calculated (hydraulic gradient = 1 ). Table 22.7 Geological Barrier Requirements According to the Proposed EU Landfill Directive and .Corresponding Calculated Vertical Flow Times. Landfill class

Permeability, K, m/s

Thickness, D, m

Vertical flow time, ty, year

Hazardous waste

_<1.0.10 .9

25

2 15.9

Nonhazardous waste

_<1 . 0 . 1 0

.9

21

2 3.17

Inert waste

_< 1 . 0 . 1 0

.7

21

2 0.051 (=18.5 days)

However, leakage often takes place in the unsaturated zone where the flow rate is much slower than it is in the saturated zone. Also, in the unsaturated zone, conditions are favour sorption and degradation of leachate components. In view of these circumstances it is proposed to assign a weighting factor of 10 to the vertical flow component. The total flow time through the geological barrier is then as follows: t = 10.tv + t. = 10 (D 9 ne)/(K~ 9 I)9 + (L n~)/(K. 9

I)9

(22.1)

where t = flow time, s D = vertical flow length, m K = hydraulic conductivity, m/s L = horizontal flow length, m I = hydraulic gradient, dimensionless ne = effective porosity of the groundwater aquifer, dimensionless (subscripts v and h denotes vertical and horizontal directions, respectively.) Nominal time values for groundwater are proposed: >200 years (hazardous waste), >50 years (nonhazardous waste) and >1 year (inert waste). Corresponding calculations for surface water show that they have little barrier effect.

953

22.4.3 Caps and Top Covers The capping of a landfill serves the dual purposes of modifying the infiltration of precipitation into the landfill and isolating the waste from the surroundings. The final capping of landfills with low permeability materials, such as clays, can greatly reduce the rate of leachate production. It is general practice to complete landfill sites with a layer of compacted clay or an artificial membrane, followed by soil and/or subsoil to support vegetation. In some cases (e.g. for MSW landfills) a gas drainage layer must be placed underneath the cap. Sloping surfaces, surface drainage systems and capillary barriers may be used to help reduce the infiltration of precipitation into the landfill if this is the part of the disposal strategy. It should be noted that synthetic cap materials, although impermeable in principle, may not be durable over a longer period of time. Furthermore, unlike most clay caps, a synthetic cap may (over a certain period of time) entirely eliminate infiltration, but it does not allow controlled infiltration at a predetermined rate which may be called for by the disposal strategy (e.g. leachate containment and collection in conjunction with waste mineralisation processes within the landfill or controlled contaminant release). A European review of clay cap performance (Knox, 1991) produced the following conclusions: Percolation through clay caps may range from 0 to 200 mm/annum depending on its quality and on materials and drainage arrangements above the clay layer. In many parts of Europe this represents a large reduction compared to effective rainfall. The performance of the cap is determined as much by what is put on top of it, as by the quality of the barrier layer itself. Percolation through the cap is extremely dependent on the efficiency of lateral drainage above it. For a given amount of effective rainfall, percolation is minimised when lateral drainage is maximised. To achieve a low percentage percolation, the ratio of hydraulic conductivity in the cap to that in the soil or drainage layer above it should be no greater than 104. The distance between field drains should be no greater than 20 m. Desiccation cracking of caps may lead to a large increase in percolation. To prevent desiccation of a cap placing sufficient soil or other material on top of it is necessary. Typically, at least 0.9 m of soil or subsoil may be needed under northern European conditions. In more arid climates greater depths may be needed to counteract desiccation.

22.4.4 Geotechnical Stability The geotechnical stability of the landfilled material (e.g. bearing capacity and slope stability) must in general be sufficiently high to accommodate trucks and various types

954 of waste moving and compacting equipment. In the longer term, it must be ensured that settlements, particularly differential settlements, do not hamper or interfere with the intended performance of liners, caps and drainage systems.

22.4.5 Abatement of Noise, Odour and Fugitive Dust Problems Proper precautions must be taken to minimise noise, odour and fugitive dust problems during the operation of a landfill. Keeping the residues moist or contained are among the measures used to avoid fugitive dust problems at residue landfills. In traditional landfilling, it is common practice to apply a daily cover of topsoil to the landfilled waste in order to minimise problems with odour, dust and vermin. Depending on the quality of the soil used, this may have the undesirable effect of creating a number of hydraulically isolated waste cells within the landfill.

22.4.6 Monitoring of Leachate Quantity and Quality Both the quantity and the quality of the leachate generated at any landfill equipped with leachate collection systems should be monitored during the period of operation and during the postclosure period until the waste in the landfill has reached final storage quality. Climatic data (e.g. precipitation and temperature) should also be monitored. The main objectives of monitoring the quantity of leachate produced at a landfill are 1): to evaluate through water balance calculations whether the leachate collection system is functioning as intended; 2): to check the efficiency of any systems intended to modify the rate of production of leachate and to obtain information on the actual rate of production of leachate at the landfill in question, thus enabling proper planning of leachate management; and 3): to provide information on the accumulated amount of leachate produced at a given landfill and hence allow an evaluation of the attained degree of leaching of the waste. The main objectives of monitoring the quality of leachate produced at a landfill are 1): to provide a basis for the selection of indicator parameters for the monitoring of groundwater (and surface water) monitoring; 2): to ensure that the leachate quality complies with the criteria in relation to on-site management and direct discharge, onsite treatment or treatment at a municipal wastewater treatment plant; and 3): to provide information on the progress of the waste stabilisation processes occurring within the landfill and, eventually, to provide background for an assessment of whether or not the waste has reached final storage quality. The leachate sampling frequency should be higher during the period of landfill operation than during the subsequent postclosure period. The analytical programme will depend on the type of waste present in the landfill and possibly also on the perceived environmental risks. A programme may consist of a routine analytical

-~

955 programme with relatively few parameters carried out on leachate samples collected relatively frequently and a more comprehensive extended analytical programme which is carried out on samples collected at longer time intervals. For instance, for leachates from incinerator residues, the simple analytical programme could comprise the determination of pH, conductivity, total dissolved solids (TDS), nonvolatile organic carbon (NVOC), chloride, sulphate, alkalinity and temperature. In addition to this, the extended programme could include measurement of adsorbable organic halogen (AOX), sulphide, total N, ammonia N, NOx N, Na, K, Ca, Mg, Fe, Pb, Cu, Cr, Hg, Ni, Zn and possibly As and Mo. The sampling point should be chosen carefully and the sampling procedure should allow for correct measurement of sensitive leachate quality parameters such as pH, redox potential, sulphide content, etc.

22.4.7 Monitoring of Groundwater and Surface Water Quality The main objectives of groundwater quality monitoring at landfills are 1): to provide information on background levels and natural variations of downstream groundwater quality prior to operation of a landfill; 2): to detect at the earliest possible time any unintended leakage of leachate from a lined landfill into the aquifer; and 3): to ensure that the anticipated impact on the aquifer is not exceeded if a controlled contaminant release strategy is applied to a landfill. The implementation of a groundwater monitoring programme which is highly dependent on local conditions includes the performance of geological and hydrogeological surveys to determine how many monitoring wells are needed, where they should be located and how they should be designed to ensure maximum likelihood of detecting any leachate plume at an early stage. It also includes the installation of such monitoring wells and implementation of proper groundwater sampling procedures. A detailed discussion of these subjects is, however, beyond the scope of this study. Several books and manuals are available on these issues. An overview is provided by Christensen et al. (1992). The analytical requirements for a groundwater monitoring programme should be based upon information on the compOsition of the leachate from the landfill in question and the background quality of the groundwater. The monitoring parameters should generally be chosen among components which are present in the leachate at significant concentrations, which are relatively mobile in the aquifer, and/or which are present in the groundwater at low background concentrations. The latter condition facilitates the detection of a leachate plume due to a large relative concentration contrast between contaminated and uncontaminated groundwater. In order to establish background levels, and the natural or seasonal variations of the groundwater to be monitored it is proposed that the samples from the monitoring wells are analysed 4 times a year for 2 years prior to commencement of landfill operation. The analytical programme carried out during this period should correspond to the

956 extended programme for leachate described in Section 22.4.6. When this has been accomplished, both the sampling frequency and the analytical programme should change. The sampling frequency should be high enough to ensure that a leachate plume will be detected before it has migrated significantly past a monitoring well. The sampling frequency should also be low enough to ensure that the water that is drawn into the well from the aquifer has been replaced between sampling events by water flowing from upstream. The sampling frequency will thus depend on the rate of flow of the groundwater which must be determined for each individual landfill site. It is recommended, however, that a groundwater sampling frequency of not less than once a year is adopted in any case. An analytical programme could consist of a basic set of parameters (pH, conductivity, TDS, NVOC, chloride, sulphate, ammonia, Na, K and Ca) which for each individual landfill may be supplemented with further parameters. The basic parameters are designed to provide an early warning for components which are present in most incinerator residue leachates. The supplementary parameters must be selected for each individual landfill based on knowledge of the actual waste accepted, its leaching properties, the composition of the leachate and the mobility and background values of various leachate components in the aquifer. Hjelmar et al. (1988) report that very early signs of a plume of bottom and fly ash leachate in an aquifer were observed as continuously rising concentrations of chloride and calcium (due to ion exchange between Na and K in the leachate and Ca in the soil) and other mobile salts in groundwater from a downstream monitoring well. The monitoring of the groundwater quality should continue during the postclosure period, possible at a reduced sampling frequency, until the waste in the landfill has reached final storage quality. The main objectives of surface water quality monitoring at landfills are 1): to provide information on background levels and natural variation of the quality of surface waters to which leachate is directly or indirectly discharged; and 2): to detect any unintended impact of leachate on a surface water system and to ensure that the anticipated impact on the surface water system is not exceeded if a controlled contaminant release strategy is applied to a landfill. It should be noted that monitoring of surface water systems generally is more difficult and much less likely to produce useful results than monitoring of groundwater. Surface water systems are usually much more diverse and dependent on local conditions than aquifers. There may be none or several surface water bodies downstream of a landfill and within the same catchment area or close enough to be affected by spillages of leachate. With the exception of stagnant waters, most surface waters systems generally have much higher dilution potentials than aquifers do. This means that leachate which may be discharged directly, may be removed rapidly or diluted by the

957 surface water to the extent that direct detection by chemical analysis is difficult or impossible. In cases where surface water quality monitoring is found desirable, the design of the chemical analytical programme may be based on the same principles and include similar ranges of analytical parameters as for groundwater while taking into account the nature of the surface water body in question. Analytical parameters such as chloride, sulphate, Na, K, Ca and Mg are, of course, irrelevant as contamination indicators in marine surface water bodies. Due to the low concentration levels of contaminants which must be expected for affected surface water systems, the monitoring programme may in certain cases alternatively be based on eutrophication indicators (e.g. chlorophyll-a) or on biological monitoring (in marine waters, e.g. placement and subsequent analysis of mussels). 22.4.8 Leachate Treatment

The most common leachate disposal route for landfills with active leachate removal systems is to sewer and subsequently to a biological wastewater treatment plant without any pretreatment. Biological treatment only affects the content of biodegradable organic contaminants (and perhaps ammonia) in the leachate. Leachates with high organic Ioadings may require aerobic biological pretreatment to remove organics and ammonia prior to discharge to sewer or surface water. Incinerator residue leachates, however, normally have relatively low contents of organics, but may have high contents of inorganic salts (e.g. CI, SO42, Na§ K§ Ca 2§ and variable contents of trace elements/heavy metals. Such leachates may be treated/pretreated by pH and/or redox potential adjustment, precipitation (e.g. with TMT)/filtration and/or adsorption on activated carbon to remove dissolved trace metals, suspended solids and non-degradable organics prior to discharge to sewer or surface water bodies. Preconcentration or removal of salts may be accomplished by reverse osmosis or evaporation, although this may not be advisable for this type of leachate from an economic or a life cycle perspective (Hjelmar et al., 1995). In most cases, biological wastewater treatment will have little effect on incinerator residue leachates dominated by inorganic constituents, but discharge to sewer may be convenient. If such leachates have very low contents of trace elements and organics, they are compatible with seawater and may be discharged directly into marine surface water bodies without pretreatment. Leachates from landfilling of poorly combusted residues may have high contents of organic material and may require (and benefit from) biological wastewater treatment. 22.5 DISPOSAL PRACTICES

Data have been compiled for a number of countries regarding present and future disposal practices for incinerator residues. In almost all jurisdictions, regulations have

958 been or are presently undergoing modifications. In Europe, for instance, a number of countries will have to modify regulations to come into compliance with an anticipated European Union (EU) waste disposal directive (European Union, 1995) and other EU waste management legislation. In the US, the Supreme Court ruling in May 1994 which lifted the exemption of MSW incinerator ash from being tested as a potentially hazardous waste may give rise to changes in some states (Supreme Court of the United States, 1994). Consequently, the situation is somewhat dynamic, and since the collected information may be outdated fairly rapidly, only a brief, general outline of the current disposal practices in some European and North American countries is presented in this section. Disposal practices for MSW incinerator residues vary widely across the world. Substantial variations in disposal practices are also found within countries consisting of federations of states or provinces such as US, Canada and Germany. Table 22.8 summarises the disposal strategies corresponding to the landfilling practices and/or policies of some countries. Table 22.8 MSW Incinerator Residue Disposal Practices in Various Countries Disposal strategy Bottom ash APC residues Combined ash Total Germany Canada US containment/dry Denmark (d) tomb The Netherlands (d) .. Germany Leachate Denmark Denmark (d) US containment and France France (a) collection Germany Germany (wet Sweden scrubber) Switzerland Sweden (a) The Netherlands Switzerland(a) Controlled Sweden Sweden (a) contaminant Denmark (b,c) release Unrestricted Canada contaminant release (a): Residue treatment or stabilisation required (b) Past practice (c) Planned future practice (pending legislation) (d) Temporary practice (awaiting improved treatment and disposal technology) Monofill disposal represents the most common method of bottom ash disposal in Europe and Canada, although co-disposal of incinerator residues with other wastes,

959 including MSW, does occur. The strategies employed for disposal of bottom ash cover the entire range from total containment over leachate containment and collection and controlled contaminant release to unrestricted contaminant release. At present, however, the prevailing disposal strategy for bottom ash is containment with some type of leachate collection. In Canada and several European countries, APC system residues are listed as hazardous or special wastes, and disposal at highly engineered monofills with extensive environmental protection systems corresponding to total containment or leachate containment and collection is normally required, often in conjunction with treatment/stabilisation of the residues. In France, for instance, stabilisation of APC residues is required prior to disposal at hazardous waste landfills, and in Germany most of the residues from dry/semi-dry APC processes are placed in underground storage in old salt mines. At one disposal site in Sweden, residue from the dry APC process is stabilised with approximately 30% (w/w) of a special type of cement; the material (37% residue, 16% cement and 47% water) is poured onto the site as a slurry in 2.5 m deep layers (cells) and allowed to cure. The median concentration of chloride in the leachate which consists primarily of surface runoff water is approximately 20,000 mg/I and the median concentration of lead is approximately 0.06 mg/I over a period from 2 to 6 years after the commencement of disposal (Sundberg and Tuutti, 1994). In the US, most of the incinerator residues generated are managed as combined ash. The combined ash is frequently disposed in landfills with relatively stringent design standards for leachate containment and collection, or total containment strategies. Most regulations evolve from the State level, provided Federal criteria are met. As of 1991,25 states required monofill disposal, 5 states allowed for monocells inside MSW landfills, and 16 states allowed codisposal of incinerator ash and raw MSW. According to the above mentioned Supreme Court decision, all MSW incinerator ash in the US must now be tested using the TCLP test (see Chapters 14 and 16). If the ash passes the requirements, it may be disposed in a Subtitle D landfill for nonhazardous waste. Conversely, if it fails, it must be placed in Subtitle C landfills for hazardous waste (as specified in the Resource and Recovery Act). The ash may be treated and combined prior to testing.

22.6

DISPOSAL RECOMMENDATIONS FOR INCINERATOR RESIDUES

Based on the data presented on incinerator residue leaching characteristics and the discussion of disposal strategy, a number of conclusions may be drawn concerning the feasibility of various disposal and leachate management options. The disposal strategies applicable to mineral wastes such as incinerator residues and organic waste types are, as previously discussed, very different and generally incompatible, both in the short and long-term. Co-disposal of incinerator residues and raw MSW is therefore generally not advisable. Different types of incinerator residues

960 may also exhibit significant differences in behaviour when landfilled, and separate management and disposal of, e.g. bottom ash and APC residues, is therefore recommended, since the opportunity of applying different disposal strategies to different types of incinerator residues when appropriate is lost if bottom ash and APC system residues are combined. This recommendation differs from the current practice in the US, where disposal of combined ash is frequently managed using very stringent landfill design standards (e.g. in double lined monofills). While this practice generally is environmentally protective, at least in a short-term perspective, it does require perpetual maintenance and can be costly, especially since the bulk of the combined ash (bottom ash) may not require the same level of perpetual care as a combined ash or APC residue fill. Overall, the optimal disposal strategy for the various incinerator residues may be considerably different from the traditional disposal strategies applied to raw MSW. The most feasible disposal options for bottom ash, APC residues and combined ash are briefly discussed below and summarised in Table 22.9. Table 22.9 Summary of Incinerator Residue Options and Recommendations Disposal strategy option

Bottom ash

APC..residue Combined ash

Total containment/dry No Possibly No tomb (e.g. salt mines) Leachate containment and Yes (a) Yes (a) Yes (a) collection Controlled contaminant Yes (b) Maybe (b) Maybe (c) release Unrestricted contaminant No (c) No (c) No (c) release (a) If requirements for controlled contaminant release are not met (e.g. as a first stage of disposal). (b) If requirements are met. May require prior or in-situ treatment of the residues or may be second stage of disposal. (c) Only after final storage quality criteria are met. 22.6.1 Bottom Ash

With proper siting (e.g. close to the sea or in an area without vulnerable aquifers), a disposal strategy based on controlled contaminant release seems appropriate and should be pursued for landfilling of bottom ash. Pretreatment (e.g. washing or stabilisation) or an initial disposal stage entailing containment, collection and treatment of leachate may in several cases be required (Belevi et al., 1992). The possibilities for controlling the geochemical and biogeochemical conditions within a bottom ash landfill

961 through ash quality requirements, ash treatment and landfill design should be investigated further. The rate of leachate production may be controlled partly through the design of the landfill. The construction of any bottom ash disposal site based on a controlled contaminant release strategy must be preceded by a thorough environmental impact assessment which ensures that the rate of release of contaminants into the surrounding environment will not exceed an acceptable limit, neither in the short nor long-term. If, for some reason, a solution requiring containment and collection of leachate is chosen, either temporarily or indefinitely, it becomes necessary to manage and dispose of the leachate. Leachate from bottom ash is generally accepted at wastewater treatment plants as long as it does not constitute a major proportion of the total input to the facility. It has in some cases been necessary to reduce the pH of the leachate (by addition of sulphuric acid) and/or to elevate the redox potential from a reducing level to an oxidised level (e.g. by addition of hydrogen peroxide) prior to treatment at a wastewater treatment plant. In most cases no pretreatment has been necessary. Dilution is practically the only beneficial effect of biological wastewater treatment on bottom ash leachate containing mostly inorganic salts and little or no organic degradable matter.

22.6.2 APC Residues (Fly Ash and Acid Gas Scrubbing Residues) A sustainable disposal solution for the APC residues, particularly fly ash and residues from the dry/semi-dry acid gas scrubbing processes, must eventually be based on a controlled contaminant release strategy and will almost certainly require extensive pretreatment of the residues. A two-stage treatment process involving removal and possibly recovery of the soluble salts (washing/extraction) followed by stabilisation, vitrification or fixation of the remnant may be appropriate for this purpose (Hjelmar, 1992). Considerable efforts are currently being spent on the development of such processes. In the meantime, disposal of APC residues must generally be based on less sustainable strategies involving total containment/entombment or containment and collection of leachate. The same requirements concerning proper siting and design of a landfill and performance of an environmental impact assessment as mentioned above for bottom ash apply to a controlled contaminant release disposal strategy for APC residues. The leachate produced at APC residue disposal sites based on containment and collection of leachate generally has a high concentration of inorganic salts and in some cases also relatively high concentrations of trace elements, particularly Pb and Cd. Such leachate is often accepted at municipal wastewater treatment plants without prior treatment, provided it does not constitute a major proportion of the total input to the plant. As mentioned for bottom ash leachate, the only beneficial effect of such a treatment is dilution. Leachates with a high content of heavy metals may have an adverse effect on the sludge from a biological treatment plant. If necessary, the

962 concentration of several trace elements in the leachate (e.g. Cd and Pb) may be reduced substantially by subjecting the leachate to pretreatment including adjustment of pH and sedimentation/flocculation with TMT. This treatment may be relatively expensive if large amounts of leachate are produced. Removal of the inorganic salts from the leachate (e.g. by evaporation) is not economically feasible or environmentally desirable under most circumstances. 22.6.3 Combined Ash

Although separate management and disposal of the different residue streams are believed to be technically and economically advantageous, both in the short and longterm, combined ash is still generated in the US. In principle, the disposal requirements for combined ash are similar to those described for APC residues. The proportion of residue requiring relatively stringent environmental protection measures when landfilled is increased substantially by the mixing of APC residue and bottom ash which also precludes utilisation and renders pretreatment or in-situ treatment of the residue more difficult and less efficient than it would be for separate ash streams. REFERENCES

Andersen, L. & J. Boll. "Leaching of APC System Residues from MSW Incinerators, Pilot Scale Experiments", In: Udvas.kning fra sla.q~er, lord o.q afraid, ATV-K0miteen vedrc~rende ,qrundva.ndsforurening, Lyngby, Danmark, pp. 85-106 (in Danish), 1994. Belevi, H. and P. Baccini. "Long-Term Behaviour of Municipal Solid Waste Landfills", Waste Management & Research, 7:43-56, 1989. Belevi, H., D.M. Stampfli and P. Baccini. "Chemical Behaviour of Municipal Solid Waste Incinerator Bottom Ash in Monofills", Waste ManaQement & Research, 10153167, 1992. Cambotti, R.K. & H.K. Roffman, "Municipal Waste Combustion Ash and Leachate Characterization. Monofill- Fifth Year Study, Woodburn Monofill, Woodburn, Oregon", Report prepared by AWD Technologies, Pittsburgh, Pennsylvania, 1993. Christensen T.H., P. Kjeldsen and J.L.C. Jansen. "Groundwater Control Monitoring at Sanitary Landfills." In (Christense.n, Cossu and Ste.Qemann, eds.): Landfillin.q of.Waste: Leachate. Elsevier, London, 1992. Comans, R.N.J., H.A. van der Sloot and P.A. Bonouvrie. "Geochemical Reactions Controlling the Solubility of Major and Trace Elements During Leaching of Municipal Solid Waste Incinerator Residues", in.. Proceedings of....th.e.. 1993 International Conference on Municipal Waste Combustion, March 30 - April 2, 1993, Williamsburg, Virginia, USA, 1993.

963 European Union. "Directive on the Landfilling of Waste", Brussels, Belgium, 1995. Hjelmar, O. "Leachate from Incinerator Ash Disposal Sites", in ProceedinQs of the International Workshop on Municipal Waste Incineration, Montreal, Canada, October 1-2, 1987. Hjelmar, O. "Characterization of Leachate from Landfilled MSWI Ash", in Proceedinqs of the International Conference on Municipal Waste Combustion, Hollywood, Florida, April 11-14, 1989. Hjelmar, O. "Field Studies of Leachates from Landfilled Combustion Residues", Presented at ...WASCON "91, Environmental Implications.of Construction with Waste Materials, Maastricht, The Netherlands, November 10-14, 1991. Hjelmar, O. "Municipal Solid Waste Incinerator Flue Gas Cleaning in Denmark: Residue Properties and Residue Management Options", in Proceedings of ISWA Specialized Conferenc_~..on Incineration and Biolo.qical Waste Treatment, Amsterdam, The Netherlands, September 1-3, 1992. Hjelmar, O. "Leaching Properties of Fly Ash from MSW Incinerators", Report for the National Agency for Environmental Protection, VKI Water Quality Institute, H~rsholm, Denmark, 1993 Hjelmar, O. "Disposal Strategies for Municipal Solid Waste Incineration Residues", J_ Haz. Mats., 47, pp. 345-368, 1996. Hjelmar, O., K.J. Andersen, J.B. Andersen, E.A. Hansen, A. Damborg, E. Bj~rnestad, A.H. Knap, C.B. Cook, S.B. Cook, J.A.K. Simmons, R.J. Jones, A.E. Murray, M.J. Lintrup, H. Schr~der, F.J. Roethel. "Assessment of the Environmental Impact of Incinerator Ash Disposal in Bermuda", Final Report, Prepared for Ministry of Works & Engineering, Hamilton, Bermuda, by the Water Quality Institute, H~rsholm, Denmark, 1993. Hjelmar, O., E. Aa. Hansen and A. Rokkjaer. "Groundwater Contamination from an Incinerator Ash and Household Waste Codisposal Site", In UNESCO Wor.kshop on Impact of Waste Disposal on Groundwater and Surface Wa.ter, Copenhagen, Denmark, 1988. \\

Hjelmar O., L.M., Johannessen, K. Knox, H.-J. Ehrig, J. Flyvbjerg, P. Winther and T.H. Christensen "Management and Composition of Leachate from Landfills", Final Report to the Commission of the European Communities, DGXl A.4, Waste '92, Contract No. B4-3040/013665/92, 1995. Johannessen, L.M., O. Hjelmar and J. Riemer. "A New Approach to Landfilling of Waste in Denmark", in Proceedin.qs of Sardinia "93, IV International Landfill Symposium, S. Margherita di Pula, Italy, 11-15 October 1993, 1993.

964 Knox, K. "A Review of Water Balance Methods and Their Application to Landfill in the UK". Report prepared for the UK Department of the Environment, DOE report No. CWM 03/91, 1992 Knox, K. "Control of Landfill Leachate", in Proceedin.qs of .the 8th International Conference: Water: Supply and Quality, Cork, Ireland, 1992. Kosson, D.S. Personal communication, 1995. Kosson, D.S., H.A. van der Sloot and T.T. Eighmy. "An Approach for Estimation of Contaminant Release During Utilization and Disposal of Municipal Waste Combustion Residues", J. Haz. Mats. 47, 1996. Lyons, M.R. "The WES-PHix Ash Treatment Process", Wheelabrator Environmental Systems Inc., Hamilton, NH, USA, 1995. Sundberg, J. and K. Tuutti. "Solidification of APC System Residue from H(~gdalenverket", Final Report 1994, Terratema ab, Link(~ping, Sweden (in Swedish), 1994 Supreme Court of the United States. "Syllabus: City of Chicago et al. v. Environmental Defense Fund et al.", Certiorari to the United States Court of Appeals for the Seventh Circuit. No. 92-1639. Argued January 19, 1994 - Decided May 2, 1994. Reimann, D.O. "Abwasserbehandlung aus M011verbrennungsanlagen", M011und Abfall, 19 (1):1-7, 1987. Thygesen, N., F. Larsen and O. Hjelmar. Environmental Risk Screenin.q of Utilization and Disposal of MSWI Bottom .Ash, Miljc~projekter 203, National Agency for Environmental Protection, Copenhagen, Denmark, (in Danish), 1992. van der Sloot, H.A., R.N.J. Comans, T.T. Eighmy and D.S. Kosson. "Interpretation of MSWI Residue Leaching Data in Relation to Utilization & Disposal", in Proceedin.qs of the Internation_al Recyclin(:] Conference, Berlin, Germany, 1992.

965

INDEX

abrasion resistance 361 absorption 353 as a function of time 355 accuracy 167 acid neutralising capacity 372, 454 as a function of time 372 acid gas cleaning residue 444 leaching 745 acid-base reactions 491 actinides 400 active environmental protection systems 942 leachate 942 activity coefficient 508, 511,553 activity-based sorption model 561 adatoms 551 adsorption edge 567 advection 487 aggregate 895 substitute 339 aging 550, 588 reactions 556 agitation 581 extraction tests 581 air controlled 66 excess 62 injection 73 over-fire 63 starved 76 under-fire 63 air pollution control residues 441 alkali metals fate in combustion 288 alkalinity 368 ALS Process 749 aluminum metallic 271 amphoterism 540 anion 517

ANS 16.1 843, 844 antimony 673 partitioning 311 apparent specific gravity 353 arsenic 648 partitioning 310 artificial reefs 896 ash deposition mechanisms 419 Brownian forces & Eddy diffusion 421 diffusiophoresis 421 Fickian diffusion 421 gravitation 421 inertial impaction 421 interception impaction 421 thermophoresis 421 asphalt 895 auger electron spectroscopy 251 availability 854 control 639 test 640 barium 657 base salt 507 batch tests 483, 684 BET surface area 370 analyses 495 bias 177 bidentate 528 binders blast furnace slag 777 cement kiln dust 777 coal fly ash 777 lime kiln dust 777 boiler ash 496 boiler tubes fouling 88 boilers convection 88 economiser 88 hoppers 88 radiant 88

966 superheater 88 boron 653 bottom ash 495 bottom ash removal drag-chain 90 plate 90 ram 90 bottom ash composition 401 bottom liner 934 Boudouard reaction 267 boundary layer 491 bromine containing organic compounds 300 partitioning 300 Brownian motion 498 bulk analysis 44 density 495 environment 238 specific gravity 353 as a function of time 353 cadmium 650, 657, 663, 673 partitioning 306 calcium CaCI2 production 752 partitioning 291 California bearing ratio 364 WET test 684 Canada 22 caps 953 carbon C/H ratio 267 CO formation 313 monoxide 139 oxidation reaction 312 partitioning 312 cation 517 cation/anion balances 517 CCME 22 chelate 523 chemical aspects of leaching 487 binding 766 composition 41 characteristics 425, 454 acid neutralisation capacity 426 cadmium 433 CB 437

chemical composition 428 chromium 433 CP 437 lead 433 nickel 433 PAH 437 PCB 437 PCDD 435 PCDF 435 pH 426 solubility 426 zinc 433 evaluation tests 773 acid neutralisation capacity 775 availability 773 tank leaching 775 total metal concentration 773 water solubility 773 retention 854 weathering 550 chloride 663 chlorinated benzenes 406 phenols 406 chlorine electrochemical recovery 758 partitioning 297 chromium 650, 657, 676 cleaner technology 931 closed system 592 co-disposal 944 cold-crown melting 792 collection procedures 180 column test 483, 585, 670 combined ash 409, 931 leachate 940 combustion products 272 zone 267 compacted granular leaching test 682 compaction 588 properties 447 complexation 488 complexing agents 489 compressive strength 447 concentration buildup tests 581,584 concrete 898 condensation process 280 congruent dissolution 553 contact time 579, 594

967 containment, collection and treatment of leachate 945 contaminants cadmium 154 lead 147 mercury 147 PCDD/F 147 contamination 944 control combustion 97 emission 118 fuel variability 98 mercury 113 metals 112 particulate 111 trace organics 101 controlled contaminant release 950 copper 650, 657, 663, 676 influence on PCDD/PCDF formation 319 partitioning 295 corrosion chloride induced 283 sulphate induced 282 criteria 915 crystal radii 287 structure 509 crystalline phases 496 cumulative flux 490 release 599, 854 curing 588 Darcy's Law 498 Davies equation 513 de novo synthesis 106 Deacon Process 283 Debye-Heckel equation 513 demonstration projects 897, 899 Denmark 25 density 447 separation 242 depth of analysis 247 diagenesis 509, 556 differential thermal analysis 246 diffusion 498, 841 coefficient 501,843 dioxins 406 direct approach 44 disposal 931

practices 957 strategies 940, 947 dissociation constant 524 dissolution 488, 841 reaction 488, 531 dissolvable solids content 351 DOC 7O0 documentation 193 dot maps 250 dry scrubber residue 442, 496, 679 dry system residues 442 durability 360 dynamic tests 579, 584 dynamic multicomponent flow-through leaching model 607 earth-alkali elements concentrations in waste and residues 291 fate in combustion 297 economic benefit 419 edge 551 effective diffusion coefficient 843 size 357 electron energy loss spectroscopy 252 electrostatic precipitator ash 441 surface complexation model 564 elements 377 association 246 bonding 246 distribution by country 380 related to biogeochemical cycling 377 elemental composition 460 embankment 897 emission standards 17 BImSchV 142 CCME 142 EEC 142 Ontario A-7 139 U.S. EPA 142 emitted radiation 248 energy requirements 803 entombment 948 environmental impacts 932 fugitive dust emissions 932 leachate 932 EP Tox 596 equilibrium 487

968 constant 520 pH 488 systems 488 ettringite 764 EU Landfill Directive 952 evaporative cooling 111 exotic elements 400 extended Debye-Heckel equation 513 x-ray adsorption fine structure 253 external resistance 491 extracting agents 244 extraction tests 579, 580 fabric filter residue 442 factors influencing representativeness 168 residue streams 170 type of APC system 169 type of incinerator system 169 waste type 168 FASTCHEM 611 fate 946 ferrous content 346 Fick's second law 501 field compaction 364 investigations 933 fill material 895, 899 filtration 595 final storage quality 944 flow-around tests 585 flow-through tests 585, 586 flue gas stream 420 fluid flow 486, 498 velocity 487 fluorine partitioning 299 FLUWA Process 749 flux 490 fly ash 441 and acid gas scrubbing residue leachate 937 formation enthalpies 269 PCDD/PCDF 317 formation constant 524 fouling chemical reaction 421 condensation 421

corrosion 421 particulate 421 FOWL 634 France 26 French X31-210 leaching test 596 fuel fired melter 804 fugacity 508 fugitive dust 954 furans 406 furnace configuration centre flow 73 contra 70 parallel 73 fluidised bed 85 rotary kiln 77 walls refractory 63 water 63 fusion 791,792, 817 gas bubbling 591 gas-side fouling 420 generation rates 15 geochemical modelling 658 thermodynamic equilibrium models 609 geochemistry 507 geological barrier 952 geotechnical stability 953 German DIN leaching test 684 Germany 28 Gibbs free energy of formation 520 minimisation 608 Gibbs' fundamental equation 269 Gibbs-Helmholtz equation 269 glass fraction 367 leaching 799 glassy phase 496 gradation 357 grain size distribution 346 granular material 584 (testing) 843 grate ash 339, 496 feed rates 63 manufacturers 63

969 reciprocating 65 rocking 65 roller 66 siftings 65, 339, 496 travelling 66 grinding 222 gross composition 342 guidelines 97 design 97 operating 97 GCmtleberg equation 513

intrinsic properties 942 ion activity 511 activity product 534 exchange 746 pairs 515 iron partitioning 293 separation 739 isodynamic separation 241 isotopes 400

halogens characteristics 296 thermal dissociation of hydrogen halides 297 HCI recovery 755 heat transfer surface 420 heating value 41 heterogeneous reactions 507 historical excursus 1 homogeneous reactions 507 Horsefall cell-type incinerator 4 hydraulic conductivity 447, 499 hydrodynamic dispersion 503 hydrogen evolution 451 hydroxyl 507

Japan 31 Japanese leaching test 686 jaw crushing 240

impregnation 243 Incongruent dissolution 547 increment collection classification 175 ~ndirect approach 44 ~nfiltration 492 influence of aging 367 of combustor type and operation 374 infrared spectroscopy 254 injection systems duct 111 furnace 111 reactor 111 inner sphere complexes 523 inorganic characteristics 376 constituents 460 salts 957 intended lifetime 942 internal porosity 491 resistance 491

kinetic model 507 systems 488 kinks 551 L/S 933 laboratory-field translation 881 landfill disposal APC residue 149 fly ash 149 grate ash 149 landfill design 945 operation 945 siting 945 lanthanides 400 leachant 491,579 composition 579, 589 renewal 579 leachate collection systems 951 management 948 treatment 957 leaching 905 behaviour 637 modelling 490, 622 regime 493 scenarios 493 solution 492 system 485, 491 tests 489, 509, 510, 579 time frame 493 lead 653, 657, 666, 676, 711 partitioning 309

970 ligand 507 liquid-to-solid ratio 488, 579, 592 lithophilic elements 339 heavy metal concentrations in waste and residues 292 local equilibrium assumption 487 Los Angeles abrasion test 361 loss on ignition 347, 453 lysimeter 586 macro-encapsulation 763 magnetic separation 241 major elements 464 matrix elements 379 management aspects 899 Martin reverse-acting grate 8 mass burn incinerators European 62 modular 62 mass balance 813 burn incinerator 704 streams in a MSW incinerator 285 transfer 498 constraints 491 rate 492 maximum concentration levels 933 melt structure 369 mercury 657 chloride induced 304 Hg recovery from flue gas scrubbing solutions 751 porosimetry 495 metal cation 507 metallic components in fly ashes 272 separation from bottom ashes 739 metastability 536 micro-encapsulation 763 MINEQL 609 mineral predominance 509 waste 943 mineralogy 250, 368, 450 minicolumns 626 minor elements 469 matrix elements 383

MINTEQ 6O9 MINTEQA2 530 monitoring 954, 955 monofill 149, 934 monolithic material 592 (testing) 843 morphology 247, 450 MR-Process 746 MSW management tool 419 MSWl APC residues 679 boiler ash 687, 695 bottom ash 673 economiser ash 642 ESP ash 679 grate siftings 637 semi-dry scrubber residues 679 wet scrubber residues 638 multiphase heterogeneous system 491 municipal solid waste definition 15 near surface environment 238 NEN 7345 843, 844 Netherlands 33 neutral species 512 nickel 653, 677, 711 partitioning 295 NITEP 24 nitrogen oxide emission limits 302 oxide formation 302 rfate in combustion 302 non-agitated extraction tests 581-583 non-selective catalytic reduction (SNCR) Exxon's DeNOx 116 NOxOut 117 RAPENOx 117 NOx removal 115 nuclear magnetic resonance 253 NVOC 932 on-line cleaning 420 open system 592 organic additives 783 characteristics 406 constituents 473 other

971 minor elements 385 trace elements 388 outer sphere complex 522 oversaturation 534 oxidation-reduction potential 489 PAH 324 partial pressure 508, 543 particle loading 420 characteristics 486, 495 morphology 368, 424, 496 char 425 crystals 425 fused spheres 425 opaques 425 polycrystallines 425 size 642 distribution 422, 444 reduction 221,581 particulate control 111 cyclone 103 dry systems 111 electrostatic precipitators 111 fabric filter 106 settling 103 wet systems 109 matter 441 passive environmental protection systems 942 paving applications 899, 911 PCB 320, 700 PCDD/PCDF abatement strategies 317 characteristics 314 concentrations 317 influence of burnout 317 of chlorine 314 of copper 319 thermal destruction 319 toxic equivalents 314 Peclet number 504 penetration resistance 365 percent fines 359 percolation 485, 842 permeability 366, 842, 950 petrography 247 pH 372, 454, 486, 488

static test 724 pH-dependent leaching 599 phase transfer of SiO2 273 philosophy 941 PHREQPITZ 517 phthalates 474, 476 physical aspects of leaching 485 characteristics 342 evaluation tests 771 durability 772 grain size distribution 771 moisture content 771 permeability 772 Proctor compaction test 772 total carbon content 771 unconfined compressive strength 772 Pitzer equation 512 plasma arc melter 804 polishing 245 polyaromatic hydrocarbons 406 polychlorinated benzenes 322 biphenyls 320, 406, 474, 476 dibenzofurans 474 dibenzo-p-dioxins 473 phenols 323 polycyclic aromatic hydrocarbons 324, 474, 476 pore area 495 diameter 369, 486 porosity 486, 842 postclosure care 942 potassium partitioning 289 pozzolanic reactions 766 precipitation reaction 531 precision 178 predictive capability 607 preservation 588 pretreatment 957 processing 895, 901 RDF 79 waste 62 Proctor compaction test 362 density 362 moisture 362 tests 447

972 PRODEFA2 6O9 proton 507 proximate analysis 41 pyrolysis 265 Raman spectroscopy 254 RDF incinerators semi-suspension 82 stoker 82 reaction stoichiometry 488 REDEQL 609 redox 488 potential 725 reaction 508 regulations utilisation 137 regulatory issues 901 leach test 648 reject fraction 342 release 590, 860, 910 models 858 nomographs 887 representative sample 167 Reynolds number 499 road base 670, 897 salt mines 948, 959 sample drying 240 preparation concerns 183 preservation 185 size reduction 183 storage 186 sampling considerations 175 protocols 174 strategies 194 APC system residues 199 boiler/economiser ash 198 bottom ash 194 grate siftings 197 sanitary 943, 944 saturated solution approach 611 surface dry specific gravity 353 saturation index 534 scanning electron microscopy 247, 424 transmission electron microscopy 246

tunnelling microscopy 249 seasonal variations 18 secondary ion mass spectroscopy 252 selective non-catalytic reduction 115 phase dissolution 242 semi-dry process residues 442 separation processes 735 crystallisation 751 definition 735 distillation 755 electrochemical 755 evaporation 751 Ion Exchange 749 on-site 737 physico-chemical and chemical 741 sequential chemical extraction tests 581,583 serial batch tests 585 shaking 591 short-term effects 940 sintering 272, 791,830 site density 555 siting 951 skeletal density 496 slag 339 sodium NaCI production 751 partitioning 289 sulphate soundness test 360 SOLGASWATER 609 solid phase control 492 phase approach 611 solidification 764 SOLMINEQ 88 517 SOLTEQ 517, 610 solubility 874 control 544, 642 product 532 solution control 492 solvent 491 soot blowing 420 sorption 112, 488, 489 reactions 489 surface 112 soundness 360 specific gravity 353, 495 stabilisation 682, 764 stability constant 524

973 standard enthalpy 520 entropy 520 free energy change 520 static leach tests 591 stirring 591 stockpiling 918 Stoke's law 500 storage ash 63 waste 59 submerged electrode melter 804 sulphate 659, 677 attack 778 gypsum production 752 thermal stability 268 sulphation 426 sulphur partitioning 301 surface adsorption models 611 area to volume ratio 486 reactions 505 sorption reactions 508 wash off 841,847 washing 588 Sweden 35 Swiss fly ash treatment process 743 Switzerland 37 tank leaching tests 843 TCLP 596 TDS 724 technical requirements 914 temperature fuel bed 265 gas phase 265 zones on the grate 265 thermal treatment 803 thermodynamic equilibrium model 507 thin foils 245 section 243 time series of leachate quality measurements 936 time-dependent data 339 TMT (trimercaptotriazine) 442, 933 TMT#15 776 top atomic layer 238

covers 953 tortuosity 501,842 tortuous path length 491 total amounts of major and trace components leached 937 available concentration asymptote 604 containment 948 intrusion volume 496 toxic equivalency factor 314 equivalent 147 trace elements 377, 470 oxyanionic elements 377 transmitted light microscopy 247 triple layer model 564 tumbling 591 TVA leaching test 597 ultimate analysis 41 underground storage 948 undersaturation 534 unified approach 491 leaching 653 pH curves 648 uniformity coefficients 357 unit weight 356 United Kingdom 37 United States 39 unrestricted contaminant release 948 utilisation 638, 895 valency 246 Van't Hoff relationship 522 vapour pressure of metal compounds 272 vesicles 496 visual classification 343 vitrification 791,805 vitrified ash composition 801 leaching 801 volatile metals concentrations in waste and residues 303 fate in combustion 303 Volund Incinerator at Gentofte 9

974 Von Roll Borsigstrasse Incinerator at Hamburg 11 washing 741 air pollution control residues 743 bottom ashes 741 waste categories combustibles 18 fines 18 glass 18 metals 18 non-combustibles 18 paper and card 18 plastics 18 putrescibles 18 textiles 18 waste composition 17 management policy 931 WASTE Program 45 WATEQ3 609

water content 345 solubility 452 weathering 493, 764 erosion 764 freeze/thaw 764 wet/dry 764 WES-Phix process 779 wet process (WP) residue 442 scrubber residue 442 x-ray fluorescence 230 photoelectron spectroscopy 251 powder diffraction 249 XRPD 451 zero point of charge 555 zinc 653, 657, 666, 677, 711 partitioning 307

Studies in Environmental Science Other volumes in this series 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 33

Atmospheric Pollution 1978 edited by M.M. Benarie Air Pollution Reference Measurement Methods and Systems edited by T. Schneider, H.W. de Koning and L.J. Brasser Biogeochemical Cycling of Mineral-Forming Elements edited by P.A. Trudinger and D.J. Swaine Potential Industrial Carcinogens and Mutagens by L. Fishbein Industrial Waste Management by S.E. Jorgensen Trade and Environment: A Theoretical Enquiry by H. Siebert, J. Eichberger, R. (3ronych and R. Pethig Field Worker Exposure during Pesticide Application edited by W.F. Tordoir and E.A.H. van Heemstra-Lequin Atmospheric Pollution1980 edited by M.M. Benarie Energetics and Technology of Biological Elimination of Wastes edited by (3. Milazzo Bioengineering,Thermal Physiology and Comfort edited by K. Cena and J.A. Clark Atmospheric Chemistry. Fundamental Aspects by E. Meszaros Water Supply and Health edited by H. van Lelyveld and B.C.J. Zoeternan Man under Vibration. Suffering and Protection edited by (3. Bianchi, K.V. Frolov and A. Oledzki Principles of Environmental Science and Technology by S.E. Jorgensen and I. Johnsen Disposal of Radioactive Wastes by Z. Dlouh~] Mankind and Energy edited by A. Blanc-Lapierre Quality of Groundwater edited by W. van Duijvenbooden, P. Glasbergen and H. van Lelyveld Education and Safe Handling in Pesticide Application edited by E.A.H. van Heemstra-Lequin and W.F. Tordoir Physicochemical Methods for Water and Wastewater Treatment edited by L. Pawlowski Atmospheric Pollution 1982 edited by M.M. Benarie Air Pollution by Nitrogen Oxides edited by T. Schneider and L. Grant Environmental Radioanalysis by H.A. Das, A. Faanhof and H.A. van der Sloot Chemistry for Protection of the Environment edited by L. Pawlowski, A.J. Verdier and W.J. Lacy Determination and Assessment of Pesticide Exposure edited by M. Siewierski The Biosphere: Problems and Solutions edited by T.N. Veziro~lu Chemical Events in the Atmosphere and their Impact on the Environment edited by (3.B. Marini-Bettblo Fluoride Research 1985 edited by H. Tsunoda and Ming-Ho Yu Algal Biofouling edited by L.V. Evans and K.D. Hoagland Chemistry for Protection of the Environment 1985 edited by L. Pawlowski, (3. Alaerts and W.J. Lacy Acidification and its Policy Implications edited by T. Schneider Teratogens: Chemicals which Cause Birth Defects edited by V. Kolb Meyers Pesticide Chemistry by (3. Matolcsy, M. Nadasy and Y. Andriska Principles of Environmental Science and Technology (second revised edition) by S.E. J~rgensen and I. Johnsen

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 63 64 65 66

Chemistry for Protection of the Environment 1987 edited by L. Pawlowski, E. Mentasti, W.J. Lacy and C. Sarzanini Atmospheric Ozone Research and its Policy Implications edited by T. Schneider, S.D. Lee, G.J.R. Wolters and L.D. Grant Valuation Methods and Policy Making in Environmental Economics edited by H. Folmer and E. van lerland Asbestos in Natural Environment by H. Schreier How to Conquer Air Pollution. A Japanese Experience edited by H. Nishimura Aquatic Bioenvironmental Studies: The Hanford Experience, 1944-1984 by C.D. Becker Radon in the Environment by M. Wilkening Evaluation of Environmental Data for Regulatory and Impact Assessment by S. Ramamoorthy and E. Baddaloo Environmental Biotechnology edited by A. Blazej and V. Privarova Applied Isotope Hydrogeology by F.J. Pearson Jr., W. Balderer, H.H. Loosli, B.E. Lehmann, A. Matter, Tj. Peters, H. Schmassmann and A. Gautschi Highway Pollution edited by R.S. Hamilton and R.M. Harrison Freight Transport and the Environment edited by M. Kroon, R. Smit and J.van Ham Acidification Research in The Netherlands edited by G.J. Heij and T. Schneider Handbook of Radioactive Contamination and Decontamination by J. Severa and J. B~.r Waste Materials in Construction edited by J.J.J.M. Goumans, H.A. van der Sloot and Th.G. Aalbers Statistical Methods in Water Resources by D.R. Helsel and R.M. Hirsch Acidification Research: Evaluation and Policy Applications edited by T.Schneider Biotechniques for Air Pollution Abatement and Odour Control Policies edited by A.J. Dragt and J. van Ham Environmental Science Theory. Concepts and Methods in a One-World, Problem-Oriented Paradigm by W.T. de Groot Chemistry and Biology of Water, Air and Soil. Environmental Aspects edited by J. T61gyessy The Removal of Nitrogen Compounds from Wastewater by B. Halling-S~rensen and S.E. Jorgensen Environmental Contamination edited by J.-P. Vernet The Reclamation of Former Coal Mines and Steelworks by I.G. Richards, J.P. Palmer and P.A. Barratt Natural Analogue Studies in the Geological Disposal of Radioactive Wastes by W. Miller, R. Alexander, N. Chapman, I. McKinley and J. Smellie Water and Peace in the Middle East edited by J. Isaac and H. Shuval Environmental Oriented Electrochemistry edited by C.A.C. Sequeira Environmental Aspects of Construction with Waste Materials edited by J.J.J.M. Goumans, H.A. van der Sloot and Th. G. Aalbers. Caracterization and Control of Odours and VOC in the Process Industries edited by S. Vigneron, J. Hermia, J. Chaouki Nordic Radioecology. The Transfer of Radionuclides through Nordic Ecosystems to Man edited by H. Dahlgaard Atmospheric Deposition in Relation to Acidification and Eutrophication by J.W. Erisman and G.P.J. Draaijers Acid Rain Research: do we have enough answers? edited by G.J. Heij and J.W. Erisman Climate Change Research: Evalution and Policy Implications (in two volumes) edited by S. Zwerver, R.S.A.R. van Rompaey, M.T.J. Kok and M.M. Berk Global Environmental Biotechnology edited by D.L. Wise

Municipal solid waste incinerator residues

Read more

Waste Materials in Construction: Proceedings (Studies in Environmental Science)

Read more

IRJET- Processing and Disposal of Municipal Solid Waste Management

Read more

Environmental Education and Solid Waste Management

Read more

Incineration of Municipal Solid Waste Impact on Global Warming

Read more

Environmental education and solid waste management

Read more

Incinerator

Read more

Incinerator

Read more

Improving Municipal Solid Waste Management in India: A Sourcebook for Policymakers and Practitioners (WBI Development Studies)

Read more

Environmental Contamination (Studies in Environmental Science)

Read more

Environmental Radioanalysis (Studies in Environmental Science)

Read more

Environmental and Health Impact of Solid Waste Management Activities

Read more

Solid Waste Engineering , Second Edition

Read more

IRJET- Understanding the Soil Pollution Surrounding Municipal Solid Waste Landfill Site in Southwest Ahmedabad

Read more

Climate Change Research (Studies in Environmental Science)

Read more

Algal Biofouling (Studies in Environmental Science)

Read more

Waste and Environmental Policy (Routledge Explorations in Environmental Economics)

Read more

Solid Waste: Assessment, Monitoring and Remediation, Volume 4 (Waste Management)

Read more

Municipal Solid Waste to Energy Conversion Processes: Economic, Technical, and Renewable Comparisons

Read more

Municipal solid waste incineration: requirements for a successful project, Volumes 23-462

Read more

Fluid Waste Disposal (Environmental Science, Engineering and Technology)

Read more

Handbook of Solid Waste Management, 2nd Edition

Read more

Emerging Contaminants from Industrial and Municipal Waste: Removal technologies (The Handbook of Environmental Chemistry)

Read more

Handbook of Solid Waste Management, 2nd Edition

Read more

Photochemistry on Solid Surfaces (Studies in Surface Science and Catalysis)

Read more

Environmental Science

Read more

Environmental Science)

Read more

Environmental Science

Read more

Environmental Science

Read more

Environmental Science

Read more

Recommend Documents

Municipal solid waste incinerator residues

MUNICIPAL SOLIDWASTE INCINERATORRESIDUES This Page Intentionally Left Blank Studies in Environmental Science 67 MU...

Waste Materials in Construction: Proceedings (Studies in Environmental Science)

WASTE MATERIALS IN CONSTRUCTION PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON ENVlR ONMENTAL I MPLICATI0 NS 0 F CO NST...

IRJET- Processing and Disposal of Municipal Solid Waste Management

https://www.irjet.net/archives/V6/i10/IRJET-V6I1090.pdf

Environmental Education and Solid Waste Management

1 E IR A0 LID T EE ag K. ·zaya u a .. LmRDNMEI'Al EDUCATION SDUD WASil MIDGEMIII This page intentionally left...

Incineration of Municipal Solid Waste Impact on Global Warming

ROYAL GEOGRAPHICAL SOCIETY with IBG GLOBAL WARMING WHAT DO I CARE? A Symposium held at Rheged, Penrith, Cumbria Saturda...

Environmental education and solid waste management

1 E IR A0 LID T EE ag K. ·zaya u a .. LmRDNMEI'Al EDUCATION SDUD WASil MIDGEMIII This page intentionally left...

Incinerator

Incinerator

Improving Municipal Solid Waste Management in India: A Sourcebook for Policymakers and Practitioners (WBI Development Studies)

Da Zhu P. U. Asnani Chris Zurbrügg Sebastian Anapolsky Shyamala Mani Improving Municipal Solid Waste Management in Indi...

Environmental Contamination (Studies in Environmental Science)

ENVIR 0NME NTAL C 0NTAMINAT10N A selection ofpaperspresented at the 5th international Conference on Environmental Con...