Groundwater Science and Policy: An International Overview

Groundwater Science and Policy An International Overview Groundwater Science and Policy An International Overview Ed...

Author: P. Quevauviller

54 downloads 2619 Views 72MB 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

Groundwater Science and Policy An International Overview

Groundwater Science and Policy An International Overview

Edited by Philippe Quevauviller European Commission, DG Environment, Brussels, Belgium

ISBN: 978-085404-294-4 A catalogue record for this book is available from the British Library r The Royal Society of Chemistry 2008 All rights reserved Apart from fair dealing for the purposes of research for non-commercial purposes or for private study, criticism or review, as permitted under the Copyright, Designs and Patents Act 1988 and the Copyright and Related Rights Regulations 2003, this publication may not be reproduced, stored or transmitted, in any form or by any means, without the prior permission in writing of The Royal Society of Chemistry, or in the case of reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK, or in accordance with the terms of the licences issued by the appropriate Reproduction Rights Organization outside the UK. Enquiries concerning reproduction outside the terms stated here should be sent to The Royal Society of Chemistry at the address printed on this page. Published by The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 0WF, UK Registered Charity Number 207890 For further information see our web site at www.rsc.org

Foreword One of the major questions that faces the world at the beginning of this century is the threat to our water resources, both in terms of quantity and of quality. There are today 6.2 billion inhabitants on Earth, 14% of whom are already suﬀering from hunger, a number that has unfortunately been increasing for the last ﬁve years, and 20% of whom lack an adequate drinking water supply. In 2050, there will most likely be 9 billion people. To provide food and an adequate domestic water supply to that many people, with their evolving diets, we will need roughly 75% more water than we use todayw, mostly for agriculture, at the same time as we will need to safeguard a biodiverse and healthy series of natural ecosystems, indispensable for the ecological equilibrium of the planet. In addition, our current urban life, our industrial activity and our agricultural practices increasingly generate sources of contaminants that are spread in the atmosphere, in surface water, on soils and in groundwater, threatening water quality. Finally, these threats must be examined in the context of climate variability, which most likely will be enhanced by impending climate changes, as indicated by the recent Fourth Assessment Report of the Intergovernmental Panel on Climate Change released on 29 January 2007z. Not only will more water be needed, but it will have to be available even during droughts. The world has always experienced great climate variability, such as the seven years of fat cows and lean cows in the Bible. For instance, some archaeological studies conducted simultaneously in Greece and China seem to show that a major drought occurred in these two countries around AD 400; in 1876–1878, and in 1898– 1900, there were severe droughts simultaneously in Brazil, China, India and Ethiopia, causing dramatic faminesy. In 1998, following a strong El Nin˜o event, there were large deﬁcits in the grain production in China and Indonesia. These two countries were able to import food from the world stocks, and no major consequences were felt, but the global food stocks fell to a very low level. It is w

See, for example, Les Eaux Continentales, coordinated by G. de Marsily, EDP Sciences, Paris, 2006; M. Griﬀon, Nourrir la Plane`te, Odile Jacob, Paris, 2006; International Water Management Institute, Water for Food, Water For Life: The Comprehensive Assessment of Water Management in Agriculture, Colombo, Sri Lanka, report to be published 2007. z Intergovernmental Panel on Climate Change, Fourth Assessment Report, 2007. y See, for example, A. Sen and J. Dre`ze, Omnibus, Oxford University Press, New Delhi, 1999; M. Davis, Late Victorian Holocausts, El Nin˜o Famines and the Making of the Third World, Verso, London, 2001 (also available in French: Ge´nocides Tropicaux, La De´couverte, Paris, 2003 and 2006).

v

vi

Foreword

most likely that such events will occur again and severely aﬀect water resources simultaneously across diﬀerent continents. As emphasised by the Stern Report released on 30 October 2006z, taking action now to combat climate change is of utmost importance, if severe economical and social crises in the course of this century are to be avoided. But this applies also to the protection of the environment and the management of water resources, as shown, for example, by the Millennium Ecosystem Assessment Report8. This book addresses one of the central issues of these concerns: the science and policy of groundwater resource management. Groundwater is and will indeed be a major world resource for both irrigation and domestic use. Aquifers supply, in fact, about one-fourth of the ﬂow of all the rivers in the world, about 90% of it in the low-ﬂow season, about 75% of the drinking water supply in Europe, about 50% worldwide and a majority of the irrigation water in the world; aquifers are also the major natural means of storing water during wet years and make it available during droughts. Finally, almost 20% of the world freshwater reserves are contained in the aquifers, while surface waters only contribute 0.5%; the remaining 80% are the continental ice sheets, not readily usable. Managing and protecting our groundwater resources is thus a very urgent and important task. By contrast, aquifers, because they are not visible and their functioning in general poorly understood, are currently very poorly managed worldwide, or rather not managed at all, and often exploited at the highest possible level without a thought spared for their protection. In fact, very little time is left to learn how our aquifer systems operate, what their reserves are and how to protect and to manage them in order to be able to meet tomorrow’s challenges. This book is thus a timely contribution to this endeavour. With the 2000 Water Framework Directive**, the European Community took the lead in Europe in addressing the issue of re-establishing the ecological and chemical quality of our continental waters, and with the new daughter directive on groundwaterww, adopted on 12 December 2006, the European Community is again setting out the framework for the management of groundwater in Europe. Philippe Quevauviller, who was the leader at the European Commission for the development of the daughter directive on groundwater, was in an ideal position to assemble the necessary multidisciplinary team of specialists who have contributed to this book. Groundwater is indeed a topic that requires the interaction of many disciplines for its management: geologists, hydrologists, geochemists, geobiologists, agronomists and farming experts, health specialists, z

www.sternreview.org.uk. Millennium Ecosystem Assessment, Ecosystems and Human Well-being: Synthesis, Island Press, World Resources Institute, Washington, DC, 2005 (www.maweb.org). ** Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities, L 327, 22.12.2000. ww Directive 2006/118/EC of the European Parliament and of the Council on the protection of groundwater against pollution and deterioration, L 372, 12.12.2006. 8

Foreword

vii

economists, law-makers, social scientists, etc. About 70 such experts from 13 European countries have contributed to this book, with one author from the US Environmental Protection Agency giving some perspectives on groundwater protection in the USA. With eleven diﬀerent sections, the book in a coordinated fashion covers most of the topics of importance in the management of groundwater, both from the technical side and the management aspect. Bringing together often opposite views of aquifer management, that of the scientist and that of the policy maker, is a rare achievement, for which the authors and the editor must be congratulated. Science–policy integration, regulatory frameworks and stakeholder involvement are covered at the same time as groundwater characterisation, monitoring, risk assessment, remediation, modelling and management at the basin scale. The book ends with a remarkable concluding section on further policy and research objectives that need to be addressed in the coming few years, in order to put Europe at the forefront of groundwater resource management, and to meet the social, economic and ecological challenges of our water supply in the 21st century. Ghislain de Marsily French Academy of Sciences and University Paris VI, Paris, France

Preface Having been educated in geosciences, it was somehow logical that my career path would, after 25 years, bring me back to a scientiﬁc sector which opened my eyes as a researcher. Even if my actual research activities primarily focused on environmental analytical developments and applications, and later on quality assurance matters, I never lost interest in geochemistry and geology, and this is certainly what decided me to move from science to policy when the Environment Directorate-General of the European Commission searched for an oﬃcer who would develop a new groundwater directive responding to the requirements of the Water Framework Directive (WFD) 2000/60/ECw. The new Groundwater Directive (2006/118/EC)z was adopted on 12 December 2006 and it opens a new era for groundwater protection. Policy developments are also ﬂourishing in other parts of the world as illustrated by the new groundwater rules also adopted in December 2006 by the US Environmental Protection Agencyy, and the recently published FAO legislative study on groundwater in international lawz. These regulations obviously represent progress, but we should not overlook that groundwater data are still very scarce in comparison to data gathered in the surface water sector over the last 40 years. Groundwater is a ‘‘hidden’’ resource which has essentially been monitored in the light of its uses, mainly for human consumption, over the past decades. It is only recently that the environmental value of groundwater has been put forward as a key issue, and this has been reﬂected in the orientations of the European Union (EU) legislation. Much work, therefore, remains ahead of us to get a better appraisal of risks aﬀecting the qualitative and quantitative status of groundwater, perform representative monitoring programmes and establish management plans that will enable measures to be identiﬁed and operated for the sake of prevention of deterioration and enhancement of groundwater quality and quantity. This is speciﬁcally the purpose of the WFD and its ‘‘daughter’’ Groundwater Directive. This implementation work, however, will only be eﬃcient if it is backed up by the best of scientiﬁc and technological state-of-the-art. One would think, therefore, that it would then be natural that scientiﬁc experts work hand-inhand with policy-makers to identify the most appropriate tools and methods w

Oﬃcial Journal of the European Communities, L 327, 20.12.2000, p. 1. Oﬃcial Journal of the European Communities, L 372, 12.12.2006, p. 19. See Chapter 3.2 of this book. z FAO Legislative Study no. 86, 2005 (ISSN 1014-6679). z y

ix

x

Preface

that would best serve policy implementation. This is happening to a certain extent, but the situation to date is far from being satisfactory. In other words, the dialogue and interactions between the scientiﬁc and policy-making communities are not as straightforward as one could expect. In the sector of groundwater, a bridge between these two communities has developed within the last ﬁve years, and greatly contributed to EU policy design and development. It is now time to put sciences and technologies at the service of policy implementation, and this is reﬂected by the content of this RSC book Groundwater Science and Policy. This book has been written by internationally recognised experts who have gathered experiences in policy or research developments, in particular in the framework of projects funded by the EU Framework Programme for Research and Technological Developments. It represents a unique experience of operational links among the policy and science worlds. Philippe Quevauviller

Contents xxxiii

List of Contributors 1. General Introduction Chapter 1

General Introduction: The Need to Protect Groundwater Philippe Quevauviller 1.1 1.2

Introduction The Scientiﬁc Background 1.2.1 The Hydrogeological Cycle 1.2.2 Waters in Aquifers 1.2.3 Groundwater Flows 1.2.4 Groundwater Quality 1.3 Groundwater Deterioration Risks 1.3.1 Quantitative Aspects 1.3.2 Links to Associated Ecosystems 1.3.3 Groundwater Pollution 1.4 Groundwater Risk Assessment: Implications for Policy 1.4.1 Groundwater Protection Needs 1.4.2 Assessment, Prevention and Control 1.4.3 Monitoring 1.5 Conclusions References

3 4 4 5 6 6 7 7 9 10 13 13 14 16 18 18

2. Science–Policy Integration Needs Chapter 2.1 Science–Policy Integration for Common Approaches Linked to Groundwater Management in Europe Philippe Quevauviller 2.1.1

Introductory Remarks on Science–Policy Integration Needs

xi

21

xii

Contents

2.1.2

Science Integration in the Light of Groundwater Management 2.1.3 Examples of Projects in Support of Groundwater Policy 2.1.3.1 Risk Assessment 2.1.3.2 Groundwater Remediation 2.1.3.3 Diﬀuse Pollution 2.1.3.4 Management Issues and Information Tools 2.1.3.5 Relevant Networks 2.1.4 A Project ‘‘Tailor-Made’’ to Support the New EU Groundwater Directive: BRIDGE 2.1.5 Conclusions: Some Research Needs References

24 25 25 26 26 26 27 27 28 29

Chapter 2.2 Transferring Scientiﬁc Knowledge to Societal Use: Clue from the AQUATERRA Integrated Project Philippe Ne´grel, Dominique Darmendrail and Adriaan Slob 2.2.1 2.2.2 2.2.3

Introduction Overview of The AQUATERRA Project Environmental Policies 2.2.3.1 Four Generations of Environmental Policies 2.2.3.2 Science in the Four Generations of Environmental Policies 2.2.4 AQUATERRA Case Studies 2.2.4.1 The Ebro Case Study 2.2.4.2 The Meuse Case Study 2.2.5 Discussions of the Contributions of AQUATERRA to Societal Use Related to Groundwater Resources Management 2.2.5.1 Stakeholder Involvement 2.2.5.2 Learning Approaches in Policy-Making 2.2.6 Conclusions References Chapter 2.3

31 33 35 35 36 38 38 43

51 54 55 56 56

Groundwater Management and Planning: How Can Economics Help? Jean-Daniel Rinaudo and Pierre Strosser 2.3.1 2.3.2

Introduction Economic Methodologies and Tools: Four Possible Ways of Supporting Groundwater Management and Planning 2.3.2.1 Economic Characterisation of Water Uses

58

60 60

xiii

Contents

2.3.2.2 2.3.2.3

Methods to Assess Environmental Costs Economic Methods for Appraisal of Groundwater Projects and Policies: Cost-eﬀectiveness and Cost–Beneﬁt Analysis 2.3.2.4 Economic Behavioural Models 2.3.2.5 Selected Illustrations 2.3.3 Assessing and Simulating Current and Future Socio-Economic Impact of Groundwater Deterioration 2.3.4 Cost Eﬀectiveness Analysis of Groundwater Protection Measures: Finding the Least Costly Way to Reduce Nitrate Pollution to Groundwater 2.3.5 Cost–Beneﬁt Analysis of Groundwater Protection: Finding the Economically Optimal Level of Groundwater Protection 2.3.6 Integrating Economic and Groundwater Models for Simulating Nitrate Pollution in the Upper Rhine Valley Aquifer 2.3.7 Designing Economic Instruments for Groundwater Management 2.3.7.1 Environment Taxes and Charges 2.3.7.2 Tradable Groundwater Licences and Rights 2.3.8 Conclusions References

61

62 63 64

64

66

69

71 73 74 76 77 79

3. Groundwater Regulatory Framework Chapter 3.1 European Union Groundwater Policy Philippe Quevauviller 3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 3.1.6 3.1.7

Introduction The 1980 Groundwater Policy Framework Preliminary Assessment (1982) The Groundwater Action Programme (1996) The Groundwater Policy Framework Under the WFD The New Groundwater Directive 2006/118/EC Policy Integration 3.1.7.1 Nitrates Directive 3.1.7.2 Urban Wastewater Treatment Directive 3.1.7.3 Plant Protection Products Directive 3.1.7.4 Biocides Directive 3.1.7.5 IPPC Directive 3.1.7.6 Landﬁll Directive 3.1.7.7 Sewage Sludge Directive 3.1.7.8 Other Directives

85 86 88 89 91 97 98 98 99 100 101 101 102 103 103

xiv

Contents

3.1.8 Conclusions: The Need for Worldwide Cooperation References

104 105

Chapter 3.2 US Drinking Water Regulation: The Ground Water Rule Crystal Rodgers-Jenkins 3.2.1 3.2.2 3.2.3

Introduction Federal Statutory Authority Challenges in Developing the GWR 3.2.3.1 US Groundwater System Demographics 3.2.3.2 Occurrence Data 3.2.3.3 Public Health Risks 3.2.4 The Risk-Targeted Approach 3.2.5 Conclusions Acknowledgements References

107 108 108 109 110 112 113 116 116 116

4. Stakeholder Interactions Chapter 4.1 Principles of the Common Implementation Strategy of the WFD: The Groundwater Woking Group Philippe Quevauviller, Johannes Grath and Andreas Scheidleder 4.1.1

The Need for Multi-stakeholder Involvement in the Environmental Policy Development and Implementation Process 4.1.2 The WFD Common Implementation Strategy 4.1.2.1 General Principles 4.1.2.2 Supporting Activity: The Pilot River Basin Network 4.1.3 The CIS Working Group on Groundwater 4.1.3.1 Objectives 4.1.3.2 Leadership and Network 4.1.3.3 Achievements from 2003 to 2006 4.1.3.4 Perspectives for 2007–2009 4.1.4 Perspectives References

121 122 122 123 123 123 124 124 125 127 127

Chapter 4.2 The Pilot River Basin Network: Examples of Groundwater-related Activities Lorenzo Galbiati and Giovanni Bidoglio 4.2.1 4.2.2

Introduction Science–Policy Integration in the PRB Exercise Linked to Groundwater Management

128 129

xv

Contents

4.2.3

Managing Groundwater Bodies in the Shannon PRB for the Implementation of the WFD 4.2.3.1 Groundwater Body Delineation 4.2.3.2 Groundwater Management for the WFD 4.2.3.3 Example of Risk Assessment Methodology for Diﬀuse Groundwater Pollution in the Shannon PRB 4.2.4 Groundwater Natural Background Levels and Threshold Deﬁnition in the Tevere PRBs Under the BRIDGE Project 4.2.4.1 Tevere PRB: The Colli Albani Case Study 4.2.4.2 Groundwater Status Evaluation by Threshold Values 4.2.4.3 Case Study of Salone-Acque Vergini System 4.2.4.4 Case Study of the Protected Area of Castelporziano 4.2.5 Conclusions References

130 131 131

131

133 134 135 136 136 140 141

Chapter 4.3 The Harmoni-CA Initiative Geo E. Arnold, Wim J. de Lange and Michiel W. Blind 4.3.1 4.3.2 4.3.3

Introduction The Harmoni-CA Initiative Process of Bridging the Gap Between Research and Policy/Water Management 4.3.3.1 Harmoni-CA Forums and Conferences 4.3.3.2 CatchMod/Harmoni-CA Workshops 4.3.3.3 Conclusions and Lessons Learned from the Conferences and Workshops 4.3.4 Products/Tools of Harmoni-CA 4.3.4.1 WISE-RTD Web Portal 4.3.4.2 Guidance Documents, Synthesis Reports and Summaries 4.3.5 SPI-Water 4.3.6 Groundwater Directive References

142 143 143 144 145 145 146 146 147 148 149 149

Chapter 4.4 Linking Public Participation to Adaptive Management Claudia Pahl-Wostl, Jens Newig and Dagmar Ridder 4.4.1 4.4.2 4.4.3

Introduction Adaptive Management and Public Participation Rationales and Requirements for Eﬀective Participation in Groundwater Management

150 151 155

xvi

Contents

4.4.3.1

Rationales and Goals for Public Participation 4.4.3.2 Public Participation Provisions in the WFD 4.4.3.3 German Experiences with Public Participation 4.4.3.4 Example: Regional Participation in Groundwater Protection from Agricultural Nitrate 4.4.4 Social Learning in Public Participation: Support for Adaptive Management 4.4.4.1 Appropriate Framing Conditions 4.4.4.2 Well-Designed Process Management 4.4.4.3 Well-Selected Methods and Tools 4.4.4.4 Leadership Issues 4.4.4.5 An Example of Participation-based Measures in Groundwater Management 4.4.5 Conclusions References

155 157 158

159 161 162 163 164 165 167 169 170

5. Groundwater Characterization and Risk Assessment Chapter 5.1 Groundwater Characterization and Risk Assessment in the Context of the EU Water Framework Directive Andreas Scheidleder, Johannes Grath and Philippe Quevauviller 5.1.1 5.1.2 5.1.3 5.1.4 5.1.5

5.1.6 5.1.7 5.1.8 5.1.9

5.1.10

Legal Background Groundwater Body Identiﬁcation and Delineation Initial Characterisation Further Characterisation Additional Requirements of the WFD 5.1.5.1 Transboundary Groundwater Bodies 5.1.5.2 Groundwater Bodies with Lower Objectives 5.1.5.3 Interaction with Aquatic and Terrestrial Ecosystems Conceptual Model/Understanding Identiﬁcation of Driving Forces and Pressures Identiﬁcation of Signiﬁcant Pressures Assessing the Impacts of Pressures 5.1.9.1 Tools to Assist 5.1.9.2 Scaling Issues Evaluating the Likelihood of Failing to Meet the Objectives

177 180 182 183 184 184 184 184 184 186 187 188 188 189 190

xvii

Contents

5.1.11 Reporting on the Characterisation and Risk Assessment References

191 192

Chapter 5.2 Groundwater Quality Background Levels Emilio Custodio and Marisol Manzano 5.2.1 5.2.2

Introduction Rationale to Establish the Groundwater Quality Baseline 5.2.3 Methods to Establish the Natural Baseline Quality of Groundwater 5.2.3.1 Study of Major and Trace Inorganic Component Chemistry 5.2.3.2 Organic Component Chemistry 5.2.3.3 Hydrogeochemical Modelling 5.2.3.4 Tracers and Temporal Scales 5.2.3.5 Study of Natural Baseline Trends 5.2.4 Conclusions Acknowledgements References

193 195 198 200 202 206 208 211 214 215 215

Chapter 5.3 Groundwater Age and Quality Klaus Hinsby, Roland Purtschert and W. Mike Edmunds 5.3.1 5.3.2

Introduction Groundwater Age Estimation 5.3.2.1 The Deﬁnition of Groundwater Age 5.3.2.2 Environmental Tracers for Absolute Age Estimation 5.3.2.3 Geoindicators: Estimating Environmental Change and Relative Ages 5.3.2.4 Numerical Modelling of Groundwater Age 5.3.3 Groundwater Age and Water Quality and Quantity Issues 5.3.3.1 Groundwater Quality as a Function of Age 5.3.3.2 Groundwater Age and Monitoring 5.3.3.3 Groundwater Age and Water (Over)exploitation 5.3.4 Groundwater Age and the Water Framework and Groundwater Directives 5.3.4.1 Groundwater Age and Derivation of Natural Background Levels and Threshold Values

217 219 219 219 223 225 226 226 227 228 228

228

xviii

Contents

5.3.4.2

Groundwater Interaction with Dependent Ecosystems 5.3.5 Case Studies 5.3.5.1 Examples of Danish Case Studies 5.3.5.2 Examples of Swiss and German Case Studies 5.3.5.3 Example of a British Case Study 5.3.6 Conclusions References

228 229 229 231 234 235 236

Chapter 5.4 Characterisation of Groundwater Contamination and Natural Attenuation Potential at Multiple Scales Thomas Ptak and Jerker Jarsjo¨ 5.4.1 5.4.2

Introduction The Integral Groundwater Investigation Method 5.4.2.1 Concepts and Principles 5.4.2.2 The Inversion Problem 5.4.2.3 Application of the Integral Investigation Method 5.4.3 Methodology to Consider Aquifer Parameter Uncertainty and to Delimit Contaminant Source Zones Using Integral Measurements 5.4.3.1 Principles 5.4.3.2 Decision Tree Approach 5.4.3.3 Example of Application at an Industrial Site 5.4.4 Quantiﬁcation of Natural Attenuation Rates Using Integral Measurements 5.4.4.1 Principles 5.4.4.2 Example of Application at a Former Gasworks Site 5.4.5 Multilevel Integral Investigation of Contamination 5.4.5.1 The Multilevel Integral Investigation Method 5.4.5.2 Example of Application 5.4.6 Conclusions Acknowledgements References

240 241 242 244 247

249 249 255 258 259 259 260 262 262 263 265 266 267

xix

Contents

Chapter 5.5 Improved Risk Assessment of Contaminant Spreading in Fractured Underground Reservoirs Christos D. Tsakiroglou 5.5.1

Introduction 5.5.1.1 Literature Review 5.5.2 Objectives and Approach of the TRACE-Fracture Project 5.5.3 Description of Ringe Site 5.5.3.1 Conceptual Fracture Network Model 5.5.4 Hierarchical Methods for the Determination of Transport Properties 5.5.4.1 Multiphase Transport Coeﬃcients of Single Fractures and Fracture Networks 5.5.5 Numerical Modelling of NAPL Fate in Unsaturated and Saturated Zones 5.5.6 Risk Assessment of Contaminated Sites 5.5.6.1 Site Remediation 5.5.6.2 In Situ Stimulation/Remediation of Contaminated Sites 5.5.7 Socioeconomic Relevance and Policy Implications Acknowledgements References

269 270 271 272 273 275 275 278 281 282 286 287 288 289

Chapter 5.6 Groundwater Risk Assessment at Contaminated Sites (GRACOS): Test Methods and Modelling Approaches Peter Grathwohl and Hans van der Sloot 5.6.1 5.6.2

Introduction Leaching Tests (Heavy Metals, Low-Volatility Organic Compounds) 5.6.2.1 Total Composition vs. Aqueous Concentrations 5.6.2.2 Percolation vs. Batch or Shaking Tests: Comparison to Field 5.6.2.3 Boundary Conditions Leading to Changing Release Rates 5.6.2.4 Release Kinetics 5.6.2.5 Column Tests: Standardisation and Design 5.6.2.6 Concluding Remarks for Leaching Tests 5.6.3 Groundwater Risk Assessment for Volatile Compounds 5.6.3.1 Vapour-phase Diﬀusion

291 293 293 294 299 299 302 304 305 305

xx

Contents

5.6.3.2

Coupled Models for Simulation of Field Sites 5.6.3.3 Concluding Remarks for Risk Assessment for Volatile Compounds 5.6.4 Modelling for Groundwater Risk Assessment of Inorganic Constituents 5.6.5 Conclusions/Recommendations Acknowledgement References

306 307 309 312 313 313

Chapter 5.7 INCORE: Integrated Concept for Groundwater Remediation Thomas Ertel and Hermann J. Kirchholtes 5.7.1 5.7.2 5.7.3

Motivation and Basic Concept Cycle I: Plume Screening Cycle II: Source Identiﬁcation 5.7.3.1 GC-MS Fingerprinting for Petroleum Hydrocarbons 5.7.3.2 Isotopic Fingerprinting for Chlorinated Hydrocarbons 5.7.4 Cycle III: Remediation Strategy 5.7.4.1 Remediation Scenarios 5.7.4.2 ISIRE: In situ Remediation Technologies— Decision Support 5.7.5 Implementation 5.7.5.1 Administrative Aspects 5.7.5.2 Cost Considerations 5.7.5.3 Implementation in Projects References

316 319 320 320 329 333 333 335 337 337 337 340 340

6. Groundwater Monitoring Chapter 6.1 Groundwater Monitoring in the Policy Context Johannes Grath, Rob Ward and Andreas Scheidleder 6.1.1 6.1.2

Introduction Monitoring Requirements: Legal Background and Objectives 6.1.3 General Principles 6.1.3.1 Conceptual Model 6.1.3.2 Three-dimensional Characteristics and Variability 6.1.3.3 Aquifer Types 6.1.4 Chemical Status and Trend Monitoring

345 347 349 349 349 351 352

xxi

Contents

6.1.4.1 Overall Objectives 6.1.4.2 Surveillance Monitoring Programme 6.1.4.3 Operational Monitoring Programme 6.1.4.4 Selection of Representative Monitoring Sites 6.1.4.5 Selection of Monitoring Determinands 6.1.4.6 Monitoring Frequency 6.1.5 Quantity (Water Level) Monitoring 6.1.5.1 Overall Objective 6.1.5.2 Monitoring Parameters 6.1.5.3 Selection of Monitoring Density 6.1.5.4 Monitoring Frequency 6.1.6 Review and Update References

352 352 353 353 356 356 359 359 360 361 361 362 362

Chapter 6.2 Screening Methods for Groundwater Monitoring Catherine Gonzalez, Anne-Marie Fouillac and Richard Greenwood 6.2.1

Groundwater Monitoring Requirements and Speciﬁc Issues 6.2.2 Environmental Variability: Spatial and Temporal Groundwater Quality Variability 6.2.3 Screening Methods Towards Groundwater Monitoring Needs 6.2.3.1 Common Physicochemical Methods used for Groundwater Assessment 6.2.3.2 Emerging Screening Tools 6.2.3.3 Potential Uses of Emerging Tools 6.2.4 Screening Methods and Priority Substances 6.2.5 New Trends and Perspectives References

363 366 368 368 369 371 374 375 376

Chapter 6.3 Quality Assurance for Groundwater Monitoring Philippe Quevauviller and Ste´phane Roy 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5

Need for Quality Assurance for Groundwater Analysis Within-Laboratory Quality Measures Statistical Control Comparison of Analytical Methods Interlaboratory Studies 6.3.5.1 Introduction 6.3.5.2 Scope of the Groundwater Interlaboratory Programme

378 379 379 380 380 380 381

xxii

Contents

6.3.6

Certiﬁed Reference Materials 6.3.6.1 General Principles 6.3.6.2 CRMs for Major Elements in Groundwater 6.3.6.3 CRMs for Trace Elements in Groundwater 6.3.6.4 CRMs for Bromide in Groundwater 6.3.6.5 Conclusions: Availability of Water CRMs 6.3.7 Assessment of Uncertainty Linked to Groundwater Sampling. A Case Study: The METREAU Project 6.3.7.1 Introduction 6.3.7.2 Groundwater Sampling: New Devices 6.3.7.3 Uncertainties Associated with the Sampling Stage 6.3.7.4 Conclusions Acknowledgements References

383 383 384 387 390 391 392 392 393 397 399 400 401

7. Groundwater Pollution Prevention and Remediation Chapter 7.1 Prevention and Reduction of Groundwater Pollution at Contaminated Megasites: Integrated Management Strategy and its Application on Megasite Cases Jeroen Ter Meer, Hans Van Duijne, Rob Nieuwenhuis and Huub Rijnaarts 7.1.1

Addressing Groundwater on Large (Former) Industrial Sites 7.1.2 Risk-based Approach for Contaminated Megasites 7.1.2.1 Integrated Management Strategy 7.1.2.2 Megasite Objectives 7.1.2.3 Concept for Megasites 7.1.2.4 Megasite Categories 7.1.2.5 Modelling and Monitoring 7.1.2.6 Risk Management Scenarios 7.1.3 Relevance of Risk-based Management of Megasites for the Groundwater Directive 7.1.3.1 Case Studies 7.1.4 Conclusions References

405 406 406 407 407 408 410 410 412 412 420 420

Chapter 7.2 Forecasting Natural Attenuation as a Risk-based Groundwater Remediation Strategy Ryan D. Wilson, Steven F. Thornton and David N. Lerner 7.2.1

The Nature of Groundwater Pollution from Point Sources

421

xxiii

Contents

7.2.1.1 7.2.1.2 7.2.1.3 7.2.2 Natural 7.2.2.1 7.2.2.2 7.2.2.3 7.2.2.4 7.2.2.5 7.2.3 Current 7.2.3.1

Sources and Plumes Risk Assessment: Is Clean-up Required? Options for Remediation Attenuation of Contaminants Why Natural Attenuation is Important Contributing Processes Plume Development and Life Phases Biodegradation in Detail Hydrogeology and Heterogeneity Natural Attenuation Assessment Practices Establishing Whether Natural Attenuation is Occurring 7.2.3.2 Establishing Whether Natural Attenuation is Suﬃcient 7.2.3.3 Steps in Natural Attenuation Assessment 7.2.4 CORONA: A New Natural Attenuation Assessment Philosophy 7.2.4.1 Core and Fringe Controlled Plumes 7.2.4.2 Preferred CORONA Site Instrumentation 7.2.4.3 CoronaScreen Models 7.2.4.4 Output Goals 7.2.4.5 Model Descriptions 7.2.5 Recommendations Acknowledgements References

421 422 423 424 424 424 427 429 432 433 435 437 439 442 442 443 443 444 446 450 451 451

Chapter 7.3 Diﬀuse Groundwater Quality Impacts from Agricultural Land-use: Management and Policy Implications of Scientiﬁc Realities Stephen Foster and Lucila Candela 7.3.1

7.3.2 7.3.3 7.3.4 7.3.5

Why is Agricultural Land-use the Greatest Challenge Facing the New EC Water Directives? 7.3.1.1 How Does Agricultural Land-use Impact on Groundwater Quality? 7.3.1.2 Are All Types of Groundwater Body Equally Threatened by Agricultural Practices? 7.3.1.3 What can be Done to Make Agricultural Cropping more ‘‘Groundwater Friendly’’? Guidelines on ‘‘Best Agricultural Practice’’ Reducing Overall Cultivation Intensity Constraints on Pesticide Manufacture, Sale or Use Improving Irrigation Water Use Eﬃciency 7.3.5.1 What are the Main Policy Implications for Groundwater Body Quality Protection?

454 457 460 462 462 463 464 465 466

xxiv

Contents

Acknowledgements References

469 469

8. Integrated River Basin Management Chapter 8.1 Integrated Management Principles for Groundwater in the WFD Context Philippe Quevauviller 8.1.1 8.1.2

Introduction Challenges Linked to Groundwater Management 8.1.3 River Basin Management Principles 8.1.3.1 Water and Its Environment 8.1.3.2 River Basin Management Objectives 8.1.4 Operational Management 8.1.4.1 Pollution Control 8.1.4.2 Voluntary Agreements 8.1.4.3 Cost Recovery 8.1.4.4 Institutional Structure 8.1.4.5 Infrastructure vs. Regulation, Financing and Empowerment 8.1.4.6 Decentralisation 8.1.4.7 Privatisation 8.1.5 Planning 8.1.5.1 Functions of Plans and Policies 8.1.5.2 The Planning Process 8.1.5.3 Planning Systems 8.1.6 Analytical Support 8.1.6.1 Analytical Support for Operational Management: Main Challenges 8.1.6.2 Analytical Support and the Strategic Level: New Directions 8.1.7 International River Basins 8.1.7.1 The Challenge 8.1.7.2 International Basins at the Global Level 8.1.7.3 International River Basin Organisations 8.1.7.4 Interbasin Co-operation (Twinning) 8.1.8 Public Participation 8.1.8.1 At European Level 8.1.8.2 At International Level 8.1.9 Conclusions References

473 474 475 475 476 476 477 478 479 479 480 481 481 482 482 482 483 483 485 486 487 487 488 488 489 489 489 491 491 492

xxv

Contents

Chapter 8.2 System Approach to Environmentally Acceptable Farming Ramon Laplana and Nadine Turpin 8.2.1 8.2.2

Introduction Integrated River Basin Management with BMPs to Mitigate NPS Pollution 8.2.2.1 What are Best Management Practices? 8.2.2.2 BMP Typology 8.2.2.3 How to Design Acceptable BMPs ? 8.2.3 How to Build a Cost-eﬀectiveness Grid of Bundles of BMPs 8.2.3.1 Eﬀectiveness Assessment 8.2.3.2 Implementation Costs 8.2.3.3 Building of the Comparison Grid 8.2.4 Conclusions References

494 495 495 497 497 502 502 505 507 508 509

Chapter 8.3 WATCH. Water Catchment Areas: Tools for Management and Control of Hazardous Compounds Thomas Track, Steve Setford, Sharon Huntley, Claudine Vermot-Desroches, John Wijdenes, Damia Barcelo´, Monica Rosell Linares, Peter Werner, Jens Fahl, Hans-Peter Rohns, Claudia Forner, Jesper Holm, Douglas Graham, Eckard Hitsch and Josef Lintschinger 8.3.1 8.3.2

8.3.3

8.3.4

8.3.5

8.3.6

Introduction Near-infrared Fuel Leak Sensor 8.3.2.1 Description of Results 8.3.2.2 Conclusion and Perspectives Analysis of MTBE and Related Compounds in Soil and Groundwater 8.3.3.1 Description of Results 8.3.3.2 Conclusions and Perspectives Immunoassay for Field-based Determination of MTBE 8.3.4.1 Description of Results 8.3.4.2 Conclusion and Perspectives Elucidation of Stimulating and Inhibiting Eﬀects on Biodegradation 8.3.5.1 Description of Results 8.3.5.2 Conclusions and Perspectives Development of an Early Warning and Management System for Groundwater Resources

511 513 514 516 517 517 519 519 520 523 524 525 527 528

xxvi

Contents

8.3.6.1 Description of Results 8.3.6.2 Conclusions and Perspectives Acknowledgement and Further Information

528 531 532

9. Groundwater Status Assessment Chapter 9.1 Methodology for the Establishment of Groundwater Environmental Quality Standards Dietmar Mu¨ller and Anne-Marie Fouillac 9.1.1

Introductory Remarks on Science–Policy Integration Needs 9.1.2 Towards an Integrated Management of Groundwater Resources 9.1.3 The Framework: An Integrated Characterisation Process 9.1.4 Criteria to Assess Quality and Status of Groundwater 9.1.4.1 Natural Background Levels 9.1.4.2 Generic Reference Values 9.1.4.3 Attenuation Criteria 9.1.5 Receptor-oriented Quality Standards Derived by a Tiered Approach 9.1.6 How to Determine a Threshold Value (Example: Surface Water) 9.1.7 Compliance Regime for Groundwater Quality Standards 9.1.8 Conclusions Acknowledgement References

535 536 536 537 538 538 540 541 541 543 543 544 544

Chapter 9.2 Pesticides in European Groundwaters: Biogeochemical Processes, Contamination Status and Results from a Case Study Christophe Mouvet 9.2.1 9.2.2

Introduction Major Processes Involved in the Transport of Pesticides from the Soil to and in Groundwater 9.2.2.1 Sorption and Degradation 9.2.2.2 Biodegradation in Microcosms with Solids from the Unsaturated Zone Below the Root Zone 9.2.2.3 Sorption on Solids from the Unsaturated Zone

545 546 547

548 550

xxvii

Contents

9.2.2.4

Biodegradation in Microcosms with Solids and Water from the Saturated Zone 9.2.2.5 Sorption on Solids from the Saturated Zone 9.2.2.6 Field Studies Involving both Sorption and Biodegradation 9.2.2.7 Synthesis of Published Results on Sorption and Biodegradation below the Root Zone 9.2.3 Status of Groundwater Contamination by Pesticides at European Scale 9.2.3.1 The Waterbase Data Base of the European Environment Agency, June 2006 9.2.3.2 The Indicator Fact Sheet of the European Environment Agency 9.2.4 Status of Groundwater Contamination by Pesticides in Selected European Countries or Regions 9.2.4.1 Status in Italy (Adapted from Refs. 60 and 61) 9.2.4.2 Status in Wallony (Belgium) 9.2.4.3 Status in Sweden 9.2.4.4 Status in France 9.2.4.5 Status in the UK 9.2.4.6 Synthesis of the Data at National and Regional Scales 9.2.5 A Case Study: The Bre´villes Spring 9.2.5.1 Brief Description of the System and the Methods Used 9.2.5.2 Main Results from the Piezometer Network 9.2.5.3 Main Results from the Monitoring of the Spring 9.2.5.4 Conclusions from the Bre´villes Case Study 9.2.6 Conclusions and Perspectives Acknowledgements References

550 553 553 555 555 555 558 559 559 561 564 565 568 570 570 571 572 573 576 576 579 579

Chapter 9.3 Evaluation of the Quantitative Status of Groundwater– Surface Water Interaction at a National Scale Hans Jørgen Henriksen, Lars Troldborg, Per Nyegaard, Anker L. Højberg, Torben O. Sonnenborg and Jens Christian Refsgaard 9.3.1 9.3.2

Introduction The National Water Resource Model (DK Model) 9.3.2.1 Conceptual Model 9.3.2.2 Processes and Data 9.3.2.3 Model Code

584 587 587 587 589

xxviii

Contents

9.3.2.4 Model Parameterisation and Calibration 9.3.2.5 Model Validation 9.3.2.6 A Few Examples of Model Results 9.3.3 Criteria for Sustainable Groundwater Abstraction 9.3.4 Model Results 9.3.5 Discussion and Conclusions 9.3.5.1 Appropriateness of Approach for the WFD 9.3.5.2 Linking the Modelling Approach with Monitoring 9.3.5.3 Strengths and Weaknesses of Approach 9.3.5.4 Novelty of this Work Acknowledgements References

589 591 591 592 596 600 600 601 602 604 604 604

10. Modelling Chapter 10.1 Conceptual Models in River Basin Management Antony Chapman, Jos Brils, Erik Ansink, Ce´cile Herivaux and Pierre Strosser 10.1.1 Introduction 10.1.2 Integrated Water Resource Management 10.1.3 Conceptual Models in the Context of River Basin Management 10.1.3.1 What are Conceptual Models? 10.1.3.2 What Role can Conceptual Models Play? 10.1.3.3 From Conceptual Models to Quantitative/Computer-based Models 10.1.4 Building Conceptual Models in the Context of River Basin Management: Some Principles 10.1.4.1 The DPSIR Framework as a Guide 10.1.4.2 Investigating the Dynamics of River Basin Systems 10.1.4.3 Stakeholder Integration and Response 10.1.5 Experience from the Aqua terra Research Project 10.1.5.1 Context and Objectives 10.1.5.2 Identifying the Main Environmental Issues Relevant to the Case Studies 10.1.5.3 Integrating Stakeholders’ Views 10.1.5.4 Developing Simpliﬁed Representations of the Systems Investigated 10.1.5.5 Preliminary Lessons from the Experience of INTEGRATOR 10.1.6 Conclusions

611 612 613 613 614 615 617 617 618 619 620 620 621 622 622 624 625

xxix

Contents

Acknowledgements References

626 626

Chapter 10.2 Modelling Reactive Transport of Diﬀuse Contaminants: Identifying the Groundwater Contribution to Surface Water Quality Hans Peter Broers, Bas van der Grift, Jasper Griﬃoen and Ruth Heerdink 10.2.1 10.2.2 10.2.3 10.2.4

Introduction Framing a Conceptual Model Building a Regional Model for the Kempen Area Verifying the Model: Setting up a Customised Monitoring System 10.2.5 Predictions of Groundwater Contributions to Surface Water Quality 10.2.5.1 Discussion 10.2.5.2 Policy Aspects 10.2.6 Conclusions Acknowledgement References

630 631 634 637 640 641 642 643 643 643

11. Conclusions: Further Policy and Research Needs Chapter 11.1 SNOWMAN: An Alternative for Transnational Research Funding Jo¨rg Frauenstein and H. Johan Van Veen 11.1.1 Introduction 11.1.1.1 Transnational Research 11.1.1.2 Soil and Groundwater Management 11.1.1.3 Forerunning Projects and Signiﬁcant Scientiﬁc Input 11.1.2 An ERA-NET Bridging the Gap Between National and European Research 11.1.3 The Way Forward 11.1.3.1 The Meaning of Cooperation 11.1.3.2 A Stepwise Approach Towards Cooperation 11.1.4 The Upcoming Coordinated Call 11.1.4.1 Objectives and Projects 11.1.4.2 The Principles of the Coordinated Call 11.1.5 Conclusions Acknowledgements References

647 647 649 651 654 654 654 657 662 663 664 668 669 670

xxx

Contents

Chapter 11.2 Incorporation of Groundwater Ecology in Environmental Policy Dan L. Danielopol, Christian Griebler, Amara Gunatilaka, Hans Ju¨rgen Hahn, Janine Gibert, F. Mermillod-Blondin, Giuseppe Messana, Jos Notenboom and Boris Sket 11.2.1 Introduction: Groundwater Science and the New Order 11.2.2 The New Groundwater Ecology: Its Interest for Water Management Projects and/or Water Policy Planners 11.2.3 The Groundwater Ecosystem Approach as a Framework for Planning Pollution Prevention and/ or Environmental Protection Strategies 11.2.4 Diversity of Groundwater Habitats and Organisms: Their Usefulness for Environmental Monitoring Programmes 11.2.5 Groundwater-Dependent Ecosystems: A Holistic Representation 11.2.6 Overview: The Expanded Order (Achievements and Future Needs) Acknowledgements References

671

672

674

680 683 684 686 686

Chapter 11.3 Towards a Science–Policy Interface (WISE-RTD) in Support of Groundwater Management Philippe Quevauviller 11.3.1 Introduction 11.3.1.1 General Needs 11.3.1.2 Diﬀerent Levels of Interactions 11.3.2 Introduction to WISE 11.3.2.1 What is WISE? 11.3.2.2 Why do we Need WISE? 11.3.2.3 What is the Objective of the WISE Process? 11.3.2.4 Does WISE Already Exist? 11.3.2.5 Who will Use WISE? 11.3.2.6 What will WISE be Used for? 11.3.2.7 What are the Next Steps in WISE Development? 11.3.3 EU RTD Funding Mechanisms 11.3.3.1 FP5 Research Projects 11.3.3.2 FP6 Targeted Research and Integrated Projects

690 691 691 693 693 694 694 695 695 696 697 697 697 698

xxxi

Contents

11.3.3.3 FP6 ERA-NET Projects 11.3.3.4 Projects Issued from the Scientiﬁc Support to Policies (SSP) Priority 11.3.3.5 Orientations of the 7th Framework Programme 11.3.3.6 LIFE: Demonstration Projects 11.3.4 An Operational Web Interface: WISE-RTD 11.3.4.1 The Harmoni-CA Initiative 11.3.4.2 The WISE-RTD Web Portal 11.3.5 Conclusions: Needs for an Overall Science–Policy Integration Framework References

698

Appendix I Outline of Water Framework Directive Appendix II Outline of Groundwater Directive

704 711

699 699 700 700 700 701 701 703

Appendices

Subject Index

716

Contributors List BLIND Michiel W. RIZA P.O. Box 17 NL-8200 AA Lelystad The Netherlands Email: [email protected]

ANSINK Erik Wageningen University Environmental Economics and Natural Resources Group P.O. Box 8130 NL-6700 EW Wageningen The Netherlands Email: [email protected]

BRILS Jos Netherlands Organisation for Applied Scientiﬁc Research (TNO) Built Environment and Geosciences Business Unit Groundwater and Soil P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected]

ARNOLD Geo E. RIZA P.O. Box 17 NL-8200 AA Lelystad The Netherlands Email: [email protected]. nl BARCELO´ Damia Consejo Superior de Investigaciones Cientı´ ﬁcas Instituto de Investigaciones Quı´ micas y Ambieltales de Barcelona Dept. de Quı´ mica Ambiental Jordi Girona, 18-26 E-08034 Barcelona Spain Email: [email protected] BIDOGLIO Giovanni European Commission, Joint Research Centre, Via E. Fermi l, TP 460 IT-21020 Ispra (VA) Italy Email: [email protected]

BROERS Hans Peter Netherlands Organisation for Applied Scientiﬁc Research (TNO) Geological Survey of The Netherlands Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected] CANDELA Lucila Technical University of Catalonia Dept. of Geotechnical Engineering & Geoscience Gran Capita`, s/n Ed. D-2 ES-08034 Barcelona Spain Email: [email protected]

xxxiii

xxxiv

CHAPMAN Antony r3 Environmental Technology Ltd c/o School of Horticulture and Landscape University of Reading TOB2, Earley Gate, Whiteknights Reading RG6 6AU United Kingdom Email: [email protected] CUSTODIO Emilio Technical University of Catalonia Dept. of Geotechnical Engineering Gran Capita`, s/n Ed. D-2 ES-08034 Barcelona Spain Email: [email protected] DANIELOPOL Dan L. Austrian Academy of Sciences Institute of Limnology Mondsee str. 9 AT-5310 Mondsee Austria Email: [email protected] DARMENDRAIL Dominique Bureau de Recherches Ge´ologiques et Minie`res (BRGM) 3, avenue Claude Guillemin FR-45060 Orl!ans c!dex 2 France Email: [email protected] DE LANGE Wim J. RIZA P.O. Box 17 NL-8200 AA Lelystad The Netherlands Email: [email protected]. nl

Contributors List

EDMUNDS W. Mike Oxford Centre for Water Research Oxford University Centre for the Environment South Parks Road Oxford 0X1 3QY United Kingdom Email: [email protected]

ERTEL Thomas Sachversta¨ndigen-Bu¨ro Boschstr. 10 DE-73734 Esslingen Germany Email: thomas@sv-ertel de

FAHL Jens University of Technology Dresden Institute of Waste Management and Contaminated Site Treatment Pratzschwitzer Str. 15 D-01796 Pirna Germany Email: [email protected]. de

FOSTER Stephen IAH President Int. Association of Hydrogeologists P.O. Box 9 Kenilworth (Warwick) CV8 1JG United Kingdom Email: [email protected]

FORNER Claudia Stadtwerke Du¨sseldorf AG Quality Control Water Wiedfeld 50 D-40589 Du¨sseldorf Germany

xxxv

Contributors List

FOUILLAC Anne-Marie Bureau de Recherches Ge´ologiques et Minie`res (BRGM) 3 avenue Claude Guillemin FR-45060 Orle´ans ce´dex 2 France Email: [email protected] FRAUENSTEIN Jo¨rg Federal Environmental Agency Dessau Unit II 4.3 Terrestrial Ecology, Land Management, Regional Protection Concepts Wo¨rlitzer Platz 1, 06844 Dessau P.O. Box 1406 DE-06813 Dessau Germany Email: [email protected] GALBIATI Lorenzo Age`ncia Catalana de l’Aigua Provenc¸a, 204-208 ES-08036 Barcelona Spain Email: [email protected] GIBERT Janine Universite´ Claude Bernard Lyonl UMR CNRS 5023 EHF Equipe d’Hydrobiologie et Ecologie Souterraines. Baˆt FOREL 43 Bd 11/11/1918 FR-69622 Villeurbanne cedex France Email: [email protected] GONZALEZ Catherine Ecole des Mines d’Ale`s 6 avenue de Clavie`res FR-30319 Ale`s Cedex France Email: [email protected]

GRAHAM Douglas DHI–Water and Environment DK-2970 Horsholm Denmark Email: [email protected]

GRATH Johannes Umweltbundesamt GmbH Spittelauer Laende 5 AT-1090 Wien Austria Email: Johannes.grath@ umweltbundesamt.at

GRATHWOHL Peter Centre for Applied Geoscience Universita¨t Tu¨bingen Sigwartstrasse 10 DE-72076 Tu¨bingen Germany Emai1: [email protected]

GREENWOOD Richard School of Biological Sciences University of Portsmouth King Henri Building King Henri I Street UK-Portsmouth PO1 2DY United Kingdom

GRIEBLER Christian GSF-National Research Center for Environmental and Health Institute of Groundwater Ecology Ingolsta¨dter Landstrasse 1 DE-85764 Neuherberg/Mu¨nchen Germany Email: [email protected]

xxxvi

GRIFFIOEN Jasper Netherlands Organisation for Applied Scientiﬁc Research (TNO) Geological Survey of The Netherlands Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: jasper.griﬃ[email protected]

GUNATILAKA Amara Department of Ecotoxicology Center for Public Health Medical University of Vienna Wa`hringer Strarse 10 A-1090 Vienna Austria Emai1: amarasinha.gunatilaka@ verbundplan.at

HAHN Hans Ju¨rgen Arbeitsgruppe Grundwassero¨kologie Universita¨t Koblenz-Landau, Campus Landau Abt. Biologie Im Fort 7, D-76829 Landau Germany Email: [email protected]

HEERDINK Ruth Netherlands Organisation for Applied Scientiﬁc Research (TNO) Geological Survey of The Netherlands Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected]

Contributors List

HENRIKSEN Hans Jørgen Geological Survey of Denmark and Greenland GEUS Øster Voldgade 10 DK-1350 Copenhagen K Denmark Email: [email protected] HERIVAUX Ce´cile BRGM Water Department 3 avenue Claude Guillemin BP 36009 FR-45060 Orle´ans cedex France Email: [email protected] HINSBY Klaus Geological Survey of Denmark and Greenland, GEUS Øster Voldgade 10 DK-1350 Copenhagen K Denmark Email: [email protected] HITSCH Eckard Salzburg AG Centre Wasser Hagenau 1 A-5101 Bergheim Austria HØJBERG Anker L. Geological Survey of Denmark and Greenland GEUS Øster Voldgade 10 DK-1350 Copenhagen K Denmark Email: [email protected] HOLM Jesper DHI – Water and Environment DK-2970 Horsholm Denmark Email: [email protected]

xxxvii

Contributors List

HUNTLEY, Sharon Cranﬁeld Centre for Analytical Science Cranﬁeld University Silsoe Bedforshire MK45 4DT United Kingdom Email: s.l.huntley@cranﬁeld.ac.uk

LINTSCHINGER Josef Salzburg AG Centre Wasser Hagenau 1 A-5101 Bergheim Austria Email: josef.lintschinger@ salzburg-ag.at

JARSJO¨ Jerker Stockholm University Dept. of Physical Geography and Quaternary Geology SE-106 91 Stockholm Sweden Emai1: [email protected]

MANZANO Marisol Technical University of Cartagena Paseo Alfonso XIII, 52 ES-30203 Cartagena Spain Email: [email protected]

KIRCHHOLTES Hermann J. Landeshauptstadt Stuttgart Amt fu¨r Umweltschutz Hermann Josef Kirchholtes 36-3.51 Gaisburgstr. 4 DE-70182 Stuttgart Germany Emai1: [email protected]

F. MERMILLOD – BLONDIN Universite´ Claude Bernard Lyou 1 UMR CNRS 5023 EHF Equipe d’Hydrobiologie et Ecologie Souterraines Ba´t FOREL 43 Bd 11/11/1998 FR-69622 Villerubanne Cedex France

LAPLANA Ramon CEMAGREF Unite´ Ader 50 avenue de Verdun FR-33612 Cestas France Emai1: ramon.laplana@cemagref. fr LERNER David N. Groundwater Protection and Restoration Group University of Sheﬃeld Kroto Research Institute Broad Lane Sheﬃeld S3 7HQ United Kingdom Emai1: d.n.lerner@sheﬃeld.ac.uk

MESSANA Giuseppe Istituto per lo Studio degli Ecosistemi CNR – ISE, Sede di Firenze Via Madonna del Piano IT-50019 Sesto Fiorentino/Firenze Italy Emai1: [email protected]

MOUVET Christophe Bureau de Recherches Ge´ologiques et Minie`res (BRGM) 3 avenue Claude Guillemin FR-45060 Orle´ans ce´dex 2 France Email: [email protected]

xxxviii

Contributors List

MU¨LLER Dietmar Umweltbundesamt GmbH Spittelauer Laende 5 AT-1090 Wien Austria Email: dietmar.mueller@ umweltbundesamt.at

NYEGAARD Per Geological Survey of Denmark and Greenland, GEUS Øster Voldgade 10 DK-1350 Copenhagen K Denmark Email: [email protected]

NEGREL Philippe Bureau de Recherches Ge´ologiques et Minie`res (BRGM) 3, avenue Claude Guillemin FR-45060 Orle´ans ce´dex 2 France Email: [email protected]

PAHL-WOSTL Claudia Institute of Environmental Systems Research University of Osnabru¨ck Barbarastrasse 12 DE-49069 Osnabru¨ck Germany Email: [email protected]

NEWIG Jens Institute of Environmental Systems Research University of Osnabru¨ck Barbarastrasse 12 DE-49069 Osnabru¨ck Germany Emai1: jnewig@ usf.uniosnabrueck.de

PTAK Thomas University of Go¨ttingen Geosciences Center Goldschmidtstrasse 3 DE-37077 Go¨ttingen Germany Email: thomas.ptak@geo. uni-goettingen.de

NIEUWENHUIS Rob TNO Netherlands Institute of Applied Geoscience –National Geological Survey Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected]

PURTSCHERT Roland Climate and Environmental Physics Physics Institute, University of Bern Sidlerstrasse 5 CH-3012 Bern, Switzerland Email: [email protected]

NOTENBOOM Jos Milieu- en Natuurplanbureau Netherlands Environmental Assessment Agency. Postbus 303 NL-3720 AH Bilthoven The Netherlands Email: [email protected]

QUEVAUVILLER Philippe European Commission DG Environment (BU9 3/142) Rue de la Loi, 200 BE-1049 Brussels Belgium Email: philippe.quevauviller@ec. europa.eu

xxxix

Contributors List

REFSGAARD Jens Christian Geological Survey of Denmark and Greenland GEUS Øster Voldgade 10 DK-1350 Copenhagen K Denmark Email: [email protected] RIDDER Dagmar Institute of Environmental Systems Research University of Osnabru¨ck Barbarastrasse 12 DE-49069 Osnabru¨ck Germany Email: [email protected]

ROHNS Hans-Peter Stadtwerke Du¨sseldorf AG Quality Control Water Wiedfeld 50 D-40589 Du¨sseldorf Germany Email: [email protected]

ROSELL LINARES Monica Consejo Superior de Investigaciones Cientı´ ﬁcas Instituto de Investigaciones Quı´ micas y Ambieltales de Barcelona Dept. de Quı´ mica Ambiental Jordi Girona, 18-26 E-08034 Barcelona Spain Email: [email protected]

RIJNAARTS Huub TNO Netherlands Institute of Applied Geoscience — National Geological Survey Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected]

ROY Ste´phane Bureau de Recherches Ge´ologiques et Minie`res (BRGM) 3, avenue Claude Guillemin FR-45060 Orle`ans ce`dex 2 France Email: [email protected]

RINAUDO Jean-Daniel Bureau de Recherches Ge´ologiques et Minie`res (BRGM) 1034, rue de Pinville FR-34000 Montpellier France Email: [email protected]

SCHEIDLEDER Andreas Umweltbundesamt GmbH Spittelauer Laende 5 AT-1090 Wien Austria Email: andreas.scheidleder@ umweltbundesamt.at

RODGERS-JENKINS Crystal U.S. EPA 1201 Constitution Avenue, NW MC-4607M USA-Washington, DC 20460 United States of America Email: Rodgers.Crystal@epamail. epa.gov

SETFORD, Steve Cranﬁeld Centre for Analytical Science Cranﬁeld University Silsoe Bedforshire MK45 4DT United Kingdom Email: s.j.setford@cranﬁeld.ac.uk

xl

SKET Boris University of Ljubljana, Biotechnical Faculty Dept. of Biology Vecna pot 111, PP2995 SI-1001 Ljubljana Slovenia Email: [email protected] SLOB Adriaan TNO Environment and Geosciences Dept. Innovation & Environment Van Mourik Broekmanweg 6 P.O. Box 49 NL-2600 AA Delft The Netherlands Email: [email protected] SONNENBORG Torben O. Geological Survey of Denmark and Greenland, GEUS Øster Voldgade 10 DK-1350 Copenhagen K Denmark Email: [email protected] STROSSER Pierre ACTeon Le Chalimont BP Ferme du Pre´ du Bois FR-68370 Orbey France Email: [email protected] TER MEER Jeroen TNO Netherlands Institute of Applied Geoscience –National Geological Survey Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected]

Contributors List

THORNTON Steven F. Groundwater Protection and Restoration Group University of Sheﬃeld Kroto Research Institute Broad Lane Sheﬃeld S3 7HQ United Kingdom Email: s.f.thornton@Sheﬃeld.ac.uk

TRACK Thomas DECHEMA e.V. Theodor-Heuss-Allee 25 DE-60486 Frankfurt am Main Germany Email: [email protected]

TROLDBORG Lars Geological Survey of Denmark and Greenland GEUS Øster Voldgade 10 DK-1350 Copenhagen K Denmark Email: [email protected]

TSAKIROGLOU Christos D. FORTH/ICE-HT Stadiou Street, Platani P.O. Box 1414 GR-26504 Patras Greece Email: [email protected]

TURPIN Nadine UMR Me´tafort-Cemageef-Agro Paris Tech-ENITA-INRA 24 avenue des Landais BP 50085 FR-63172 Aubiere Ce´dex France Emai1: [email protected]

xli

Contributors List

VAN DER GRIFT Bas Netherlands Organisation for Applied Scientiﬁc Research (TNO) Geological Survey of the Netherlands Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected] VAN DUIJNE Hans TNO Netherlands Institute of Applied Geoscience – National Geological Survey Princetonlaan 6 P.O. Box 80015 NL-3508 TA Utrecht The Netherlands Email: [email protected] VAN VEEN H. Johan SKB Bu¨chnerweg 1 Postbus 420 NL-2800 AK Gouda The Netherlands Email: [email protected] VAN DER SLOOT Hans Energy Research Centre (ECN) P.O. Box 1 NL- 1755 Petten ZG The Netherlands Email: [email protected] VERMOT-DESROCHES Claudine Diaclone l Bd Fleming-BP 1985 F-25020 Besancon Cedex France

WARD Rob Environment Agency—England and Wales Olton Court Solihull West Midlands B92 7HX United Kingdom Email: rob.ward@ environment-agency.gov.uk WERNER Peter University of Technology Dresden Institute of Waste Management and Contaminated Site Treatment Pratzschwitzer Str. 15 D-01796 Pirna Germany Email: [email protected] WIJDENES John Diaclone l Bd Fleming-BP 1985 F-25020 Besancon Cedex France Email: [email protected] WILSON Ryan D. Groundwater Protection and Restoration Group University of Sheﬃeld Kroto Research Institute Broad Lane Sheﬃeld S3 7HQ United Kingdom Email: r.d.wilson@Sheﬃeld.ac.uk

1. General Introduction

CHAPTER 1

General Introduction: The Need to Protect Groundwater PHILIPPE QUEVAUVILLER European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium

1.1 Introduction Groundwater constitutes the largest reservoir of freshwater in the world, accounting for over 97% of all freshwater available on earth (excluding glaciers and ice caps). The remaining 3% is composed mainly of surface water (lakes, rivers, wetlands) and soil moisture. Until recently, focus on groundwater mainly concerned its use as drinking water (e.g. about 75% of European Union (EU) inhabitants depend on groundwater for their water supply). Groundwater is also an important resource for industry (e.g. cooling waters) and agriculture (irrigation). It has, however, become increasingly obvious that groundwater should not only be viewed as a drinking water reservoir, but also as a critical aquatic ecosystem.1 In this respect, groundwater represents an important link of the hydrological cycle for the maintenance of wetlands and river ﬂows, acting as a buﬀer through dry periods. In other words, it provides the base ﬂow (i.e. the water which feeds rivers all year round) for surface water systems, many of which are used for water supply and recreation. In many rivers, indeed, more than 50% of the annual ﬂow is derived from groundwater. In lowﬂow periods in summer, more than 90% of the ﬂow in some rivers may come from groundwater. Hence, deterioration of groundwater quality may directly aﬀect other related aquatic and terrestrial ecosystems. Since groundwater moves slowly through the subsurface, the impact of anthropogenic activities may last for a relatively long time, which means that pollution that occurred some decades ago—whether from agriculture, industry or other human activities—may still be threatening groundwater quality today and, in some cases, will continue to do so for several generations to come. The legacy of the past is clearly visible at large-scale contaminated sites, e.g. industrial sites or harbour areas, where it is simply not possible, with 3

4

Chapter 1

state-of-the-art technology and a proportionate use of public and/or private money, to clean up the regional contamination encountered at these locations.2 In addition, the experience of remediation of the past 20 years has shown that the measures taken have in most cases not been able to completely remove all contaminants and that pollutant sources, even if partially removed, continue to emit for long periods of time (i.e. several generations).3,4 Therefore, an important focus should be on preventing pollution in the ﬁrst place. Secondly, since surface water systems receive a continuous discharge of inﬂowing groundwater, a deteriorated groundwater quality will ultimately be reﬂected in the quality of surface waters. In other words, the eﬀect of human activity on groundwater quality will eventually also impact on the quality of associated aquatic ecosystems and directly dependent terrestrial ecosystems if so-called natural attenuation reactions such as biodegradation in the subsurface are not suﬃcient to contain the contaminants. Finally, groundwater is a ‘‘hidden resource’’ which is quantitatively much more signiﬁcant than surface water and for which pollution prevention and quality monitoring and restoration are even more diﬃcult than for surface waters, which is mostly due to its inaccessibility. This ‘‘hidden’’ character makes it diﬃcult adequately to locate and quantitatively appreciate pollution impacts, resulting in a lack of awareness and/or evidence regarding the extent of risks and pressures. Recent reports, however, show that pollution from domestic, agricultural and industrial sources is, despite the progress in some ﬁelds, still a major concern, either directly through discharges (eﬄuents) or indirectly from the spreading of nitrogen fertilisers and pesticides or through leaching from old landﬁlls or industrial sites (e.g. chlorinated hydrocarbons, heavy metals). For example, around one-third of groundwater bodies in Europe currently exceed the nitrate guideline values.5 While point sources have caused most of the pollution identiﬁed to date, there is evidence that diﬀuse sources are having an increasing impact on groundwater. This chapter develops the elements discussed above as a general introduction to this book, which further elaborates issues related to groundwater policy, protection and remediation throughout the diﬀerent chapters.

1.2 The Scientiﬁc Background 1.2.1

The Hydrogeological Cycle

It is estimated that roughly 22% of freshwater is stored underground, representing some 8 million km3 of 37 million km3 of freshwater found on the planet. Excluding water from polar ice, groundwater constitutes some 97% of all the freshwater that is potentially available for human use on or beneath the earth’s surface. The remainder is stored in lakes, rivers and swamps.6 Groundwater recharge is essentially ensured by rain that inﬁltrates through the soil into underlying layers; this recharge is occasionally augmented by streams and rivers that lose water to underground strata. Once underground,

General Introduction: The Need to Protect Groundwater

5

groundwater ﬂows at rates which range from more than 10 metres per day to as little as 1 metre per year until it reaches an outlet, e.g. a spring or seepages at the ground surface (which actually keep rivers ﬂowing during dry periods). The time scales at which groundwater ﬂows hence considerably vary, depending on hydrogeological conditions. It may take years to decades for water to move through the soil to reach the water table, the level at which the ground is fully saturated, where it may remain underground for tens or even thousands of years before reappearing at the surface.6 Geological settings may also trap groundwater from both its source and its outlets. Finally, climate change may also lead to groundwater losses by depriving aquifers from recharge as it appears in a number of regions which turned into deserts. The level of available geological and hydrogeological information varies from area to area, and this has an eﬀect on the protection schemes to be developed.7 Where the information is adequate, a comprehensive scheme, based on hydrogeological concepts, is achievable. However, as mentioned below, aquifers are rarely homogeneous and their geological variability conditions the nature of groundwater ﬂowing through their respective lithologies and structures, which makes it diﬃcult to establish large-scale conceptual hydrogeological models.

1.2.2

Waters in Aquifers

The nature of aquifers, consisting either of unconsolidated materials such as sand or gravel or consolidated rock such as sandstone, has a considerable inﬂuence on groundwater ﬂows and hence on pollution pathways (see Section 1.4.3). On the one hand, unconsolidated materials, such as sands, can store up to 30% of their volume as water. On the other hand, consolidated materials may also store large volumes of water, depending upon their porosity, but groundwater ﬂow is usually very slow owing to the small size of the pores. In some types of rocks, the capillary attraction between the groundwater and the pore surface does not allow water to be released and hence to ﬂow. However, consolidated materials may also store water in fractures in the rock, which although they usually represent less than 1% of the total volume can be enlarged by dissolution in rocks such as limestones. Enlarged fractures enable the aquifer both to store large volumes of water and permit high groundwater ﬂows,7 which has an impact on pollution spreading (see Chapter 5.5). Aquifers are usually bounded above by an unsaturared zone, which contains both air and water, and below by an impermeable bed constituted, for example, of clay or rock. The boundary between the unsaturared and saturated zones (water table) is found at diﬀerent depths depending on the hydrogeological and climatic settings, e.g. as much as 100 m below the surface in arid areas and close to the surface in humid areas. Some aquifers are, however, bounded entirely by impermeable layers and contain groundwater under pressure (which enables water abstraction by artesian wells).

6

1.2.3

Chapter 1

Groundwater Flows

With aquifer characteristics in mind (in particular the type of materials containing the water), it is possible to better approximate groundwater ﬂows. Groundwater moves through aquifers as a result of diﬀerences in pressure or elevation of the water table within the aquifer. The groundwater ﬂow may be slowed down by various obstructions while moving from the point of recharge to its exit from the aquifer. In some cases, impermeable rock formations (known as aquicludes) such as shale stop completely the water ﬂow, while other geological strata (known as aquitards), such as clay lenses embedded with sand, may slow down the groundwater ﬂow. The groundwater ﬂow rate depends on the permeability and porosity of the aquifer, and on the pressure gradient. As an example, highly permeable aquifers such as limestones respond rapidly to changes in recharge and abstraction rates, and groundwater levels in such areas may ﬂuctuate by as much as 10 m a year and may change by up to 50 m a year.7 The greatest variations in groundwater ﬂow patterns occur where changes in rock types, e.g. limestone overlying sediments and a hard crystalline rock, induce discontinuities in ﬂow and may bring groundwater ﬂow to the surface on the junction between the two rock types. Variations in groundwater ﬂow may also occur within an unconsolidated alluvial aquifer, e.g. great lateral variations occur in the mix of gravel, sand and clay making up the aquifer matrix. In larger-scale alluvial aquifers, layers of sand or gravel-rich sediment interbedded with clay-rich layers induce lateral ﬂow following the more permeable sand- and gravel-rich zones. Needless to say that groundwater ﬂow rates are very small in comparison to those of surface water. In this respect, some groundwater from deep alluvial basins is likely to be thousands or even hundreds of thousands of years old. The slow movement of groundwater largely contributes to its purity since contaminants become highly attenuated during the usually long groundwater ﬂowing pathway to the surface. Groundwater may also become enriched with elements that are naturally present in rocks. Furthermore, saltwater intrusion may occur near coastlines, in particular where the water table is lowered due to abstraction, and this is likely to be accentuated by rising sea levels due to climate change.

1.2.4

Groundwater Quality

Most groundwater originates from water that has permeated ﬁrst the soil and then the rock below it. The soil removes many impurities and the rock through which the water then ﬂows, perhaps for thousand of years, ﬁlters and puriﬁes the water even further.7 It therefore usually reappears at the earth surface free of pathogenic micro-organisms. This is the reason for an increasing exploitation of groundwater resources (see Section 3.1). While groundwater is generally less easily/rapidly polluted than streams and rivers, it often contains high concentrations of dissolved elements from the rock

General Introduction: The Need to Protect Groundwater

7

through which it has passed. Another feature is that when groundwater is polluted, many processes occur during its pathway to the surface; in particular, pollutant loads may be attenuated by adsorption by the rock itself or biochemical transformation into substances that are less harmful than the original compounds. However, severe pollutions may aﬀect groundwater quality over long periods, i.e. once pollutants reach the water table, it may take a very long time before they are ﬂushed out from the aquifer. Furthermore, groundwater quality aﬀected by pollution may take a long time to recover since the water within the aquifer moves so slowly. Once polluted, aquifers are diﬃcult—and sometimes even impossible—to clean up. The process can be likened to trying to squeeze out the last traces of soap from a sponge.7 As stressed above, the complex nature of groundwater is compounded in the context of pollution and quality problems. Let us repeat that the chemical characteristics of aquifer materials and the way pollutants react with them vary greatly. In some cases, pollutants are ‘‘ﬁltered’’ out mechanically or through adsorption onto particles within the soil or aquifer matrix. In other cases, however, pollutants remain mobile and can rapidly spread throughout an aquifer. The aquifer matrix itself can become contaminated and pockets of pollutants can serve as continuous sources of contamination. For example, small pockets of organic solvents can remain as pollution sources virtually indeﬁnitely because of their low solubility in water. Furthermore, changes in pH or other groundwater characteristics can cause the release of toxic materials, such as ﬂuoride, from natural sources within aquifers. Given the hundreds of thousands of naturally occurring compounds in groundwater and aquifer materials, and the similarly large number of compounds present in waste water released to aquifers, understanding and managing pollution problems is a highly complex task. This illustrates the importance of preventing pollution of groundwater from the start rather than dealing with the consequences (Chapter 5.5). All the above considerations have an impact on the way groundwater background levels are evaluated and also on the assessment of groundwater quality either related to its use or to its environmental value. These aspects are discussed in various chapters of this book (Chapters 5.1–5.3 and 9.1).

1.3 Groundwater Deterioration Risks 1.3.1 1.3.1.1

Quantitative Aspects Over-exploitation

Groundwater is extensively used by humans throughout the world as a drinking water resource, with some countries depending almost entirely on it while others only partly using the groundwater resource for drinking water abstraction. Groundwater supplies are of obvious importance in arid areas but they are also extensively used in humid areas, largely because they provide water that requires little or no treatment and which can be cheaply produced.

8

Chapter 1

In addition, its supplies are not subject to abrupt change as a result of abnormal weather, i.e. a dry summer while aﬀecting reservoirs will have little eﬀect on groundwater levels. Finally, groundwater can often be tapped near to where it is needed while surface water must be either developed at the sites of natural dams or reservoirs, or piped at considerable distances to where it will eventually be used.7 However, groundwater should not be seen as a simple alternative to the use of surface water. The inadequate control of groundwater abstraction in many parts of the world usually results in some form of over-exploitation, which can lead to either reversible or irreversible damages (this consideration also applies to damages due to pollution, see Section 1.3.3). In the ﬁrst case, matters can be corrected and only the question of costs is involved. In the second case, costs are also involved but, in addition, sustainability issues arise since future generations are deprived of an important resource.7 Many aquifers are being over-exploited in the sense that water is abstracted faster than the average recharge rate. This is particularly problematic in the case of fossil groundwater. The control of the balance of groundwater levels (equilibrium between abstraction and recharge) is, however, diﬃcult to apprehend in that the recharge rate of groundwater resources is not constant and can vary considerably with the rainfall pattern. This means that what may be considered as over-exploitation in one year may be a perfectly acceptable rate of exploitation in another. To complicate matters, in some arid areas major recharge only occurs once a decade or even less frequently.7 In this circumstance, deﬁning a sustainable abstraction rate is diﬃcult. Adding to this, climate changes impact on the dynamic balances of groundwater resources, and these are not easily predictable.

1.3.1.2

Falling and Rising Water Tables

Over-abstraction and subsequent fall of the water table may lead to severe damage linked to ground subsidence, which is caused by water draining from the pores in underground strata, causing the rock to compact. Unconsolidated strata, especially clays which have high water content, are particularly susceptible to this phenomenon. Conversely, circumstances such as over-irrigation of land may lead to rising of the water table, leading to waterlogging of agricultural land which is often associated with salinisation. This is due to two causes: a rising water table that brings saline water into contact with plant roots; and the evaporation of irrigation water by sun, leaving the salt behind. Another occurrence of rising water table is observed in urban areas where urban recharge rates may be higher than natural (pre-natural) ones. This is not so problematic when cities consume large quantities of water (thus balancing the high recharge rates) but it may be so when groundwater is not abstracted any more, which increasingly happens owing to contaminated groundwater beneath the cities. Rising water tables under cities may lead to urban ﬂooding, with associated costs (need to pump water out, etc.).

General Introduction: The Need to Protect Groundwater

1.3.1.3

9

Saltwater Intrusion

Under natural conditions, coastal aquifers discharge freshwater into the sea. However, in case of (over)abstraction of groundwater in areas that are close to the coastline, this process may be reversed, leading to salt water moving inland and polluting the aquiferw. Examples of such occurrences can be found in many places of the world.7 The problem may be severe on islands where the freshwater aquifer is only a few metres thick (e.g. composed of highly permeable sediments) and surrounded by salt water; in this speciﬁc case, aquifer abstraction has to be particularly well managed.

1.3.2 1.3.2.1

Links to Associated Ecosystems Links to Associated Aquatic Ecosystems

In many areas, it is groundwater that makes the use of surface water sources possible during dry seasons. Groundwater provides the base ﬂow to many of the world’s rivers, and this ﬂow continues throughout the year, regardless of weather conditions. Many rivers would dry up in hot and dry summers if they would not be fed by groundwater. This is particularly important in both humid and arid regions where precipitation is highly variable. Between precipitation events, groundwater and return ﬂows from agricultural, domestic and other users are the primary source of ﬂows in rivers. Since return ﬂows generally have higher pollution loads than groundwater ﬂow, the groundwater contribution is important to both the quantity and quality of dry-season ﬂow in surface watercourses. An evaluation of quantitative aspects of groundwater–surface interactions is described in Chapter 9.3 of this book. Productivity in coastal ecosystems is also highly dependent on the balance between freshwater inﬂows from surface water, groundwater discharge and saline ocean water. Disruption of this balance through diminution of groundwater contributions to base ﬂow could have major eﬀects on the coastal environment.

1.3.2.2

Links to Dependent Terrestrial Ecosystems

Wetlands are some of the most productive and biologically diverse inland ecosystems. In many if not most cases, water availability in wetlands depends on high groundwater levels. Consequently, the fall of the water table may have a direct impact on wetlands as the land is drying out. In this respect, a number of the world’s major wetland areas, which are sensitive ecosystems supporting a large number of plant and animal species, are now under threat due to overabstraction. In addition, groundwater pollution also represents a major threat not only to the habitat of many rare species but also in aﬀecting the purifying role of the wetlands with respect to inland lakes. w

Note that the term ‘‘pollution’’ is appropriate when saltwater intrusion is indeed due to overabstraction, i.e. due to human activity.

10

Chapter 1

Besides wetlands, groundwater levels and quality directly inﬂuence surface vegetation communities. Phreatophytes, plants that derive a major portion of their water needs from saturated soils, can be the dominant vegetative species in ecosystems where groundwater levels are shallow. They often form critical wildlife habitat and may serve as important sources of food and timber. This vegetation also uses substantial amounts of water. Removing these species can reduce evapotranspiration and hence the demand on groundwater resources, which may cause levels to rise and thereby lead to waterlogging and other environmental problems.

1.3.3 1.3.3.1

Groundwater Pollution Introduction

Once polluted, groundwater is extremely diﬃcult to clean up owing to its inaccessibility, its huge volume and its slow ﬂow rates. The three major pollution threats, namely urbanisation, industrial activities and agriculture, are discussed in the following paragraphs. Let us distinguish here pollution impacts related to either human uses or the environment. In the ﬁrst instance, pollutants found in groundwater are listed according to their toxic impacts on drinking water quality according to, for example, WHO drinking water quality guidelines or EU legislation (Drinking Water Directive). Pollutants may also be distinguished according to their ecotoxicological impacts, i.e. substances which are detrimental to the environment such as those pollutants listed in the EU Water Framework Directive (see Chapter 3.1). This distinction is important in that the ‘‘pollution impact’’ should be assessed diﬀerently whether it is related to a particular use of the water resource or to an impact on the aquatic or terrestrial environment. As noted in Section 2.4, groundwater may contain high concentrations of chemical substances that are present naturally (due to interactions of the groundwater with the soil or surrounding rocks) and which as such does not correspond to a pollution (i.e. due to human activities) but which may hamper the use of the groundwater for drinking water abstraction; this does not mean, however, that the groundwater is of ‘‘bad environmental quality’’. These two aspects of groundwater quality and related needs for protection against pollution are further discussed in the policy context in Chapters 3.1 and 3.2. Furthermore, issues of natural attenuation, risk assessment at contaminated sites (including megasites), remediation and prevention are extensively described in Chapters 5.4, 5.6, 5.7, 7.1–7.3).

1.3.3.2

Urbanisation and Related Discharges

Urbanisation introduces many changes to the aquifers that lie under cities. Natural recharge mechanisms are modiﬁed or replaced and new ones are introduced. Leakages and seepages from mains water and sanitation systems become an important part of the hydrological cycle in the urban environment. In this respect, many sub-city aquifers are polluted with human wastes,

General Introduction: The Need to Protect Groundwater

11

particularly where there is insuﬃcient connection to mains sewerage. In Europe, the situation has improved with the implementation of the Urban Wastewater Treatment Directive, but in many developing countries, septic tanks, cesspits and latrines are still common in major cities. Septic tanks, when properly operated, produce an eﬄuent of acceptable quality in areas of low population density. In practice, they are, however, often overloaded and operate ineﬃciently. Eﬄuent is often discharged directly into inland waterways, whence pollutants ﬁnd their way into the underlying aquifer.7 This pollution leads to increasing occurrence of pathogens in groundwater (in particular helminths, protozoa, bacteria and viruses), which may have a direct impact on the bacteriological quality of water abstracted for human consumption (in particular when drinking water is provided by shallow private boreholes with insuﬃcient sanitary controls). Domestic eﬄuents are also responsible for increasing nitrate concentrations in groundwater. It should be noted, however, that while sewage and urban wastewater is generally regarded as a major source of pollution, it is also considered as a large and important resource, i.e. in many arid areas, it is used with minimal, if any, treatment to irrigate crops, including some intended for direct human consumption.7 The water used also supplies crops with essential elements such as nitrogen and phosphate which would otherwise have to be added as artiﬁcial fertiliser. This re-use is often debated as regard to its safety, and many experts consider that a better use for urban wastewater is probably to recharge the aquifer from which it came. During the recharge process, the water is considerably puriﬁed. If it is required for irrigation, it can then be abstracted either from irrigation wells or from streams whose ﬂows has been increased by the recharge. Letting sewage water stand in shallow surface ponds and ﬁlter down through the soil and the aquifer below can be an eﬀective means of treatment. The more slowly this is done, and the more that the surface ponds are rested between treatments, the more complete will be the treatment. Allowing ponds to dry out regularly leads to a breakdown of nitrates in the sewage, with the release of harmless nitrogen gas. With careful control, nitrogen concentration in the recharge water can be reduced to below 5 mg L 1. At the same time, most bacteria and protozoa are eliminated, and levels of organic compounds and phosphates are greatly reduced. This inﬁltration treatment, although it uses land areas, presents the advantage of providing a cheap underground storage system from which water can be pumped for non-potable uses.7 Solid waste disposals represent another source of major urban groundwater pollution. The worst risks occur where uncontrolled tipping, as opposed to controlled sanitary landﬁll, is practised, and where hazardous industrial wastes, including drums of liquid eﬄuents, are disposed of at inappropriate sites which are selected on the basis of their proximity to where the waste is generated rather than their suitability as landﬁll sites. Often no record is kept of the nature and quantity of wastes disposed of at a given site and abandoned sites represent a potential hazard to groundwater for decades. To make matters worse, disposal is often on low ground where the water table is high and direct contamination of shallow groundwater likely.7

12

1.3.3.3

Chapter 1

Industry

Nearly all industries produce liquid eﬄuents, which according to legislation have to be properly treated before they are allowed to be discharged to a water course. There are, however, still many illegal discharges (in particular from small industries producing paper and textiles, processing leather, metals and other materials and repairing vehicles, as well as small service industries such as metal workshops, dry cleaners, photo processors, etc.), e.g. of acids, oils, fuels and solvents which have a direct impact on water courses, particularly canals, or disposed into the ground and ﬁnding their ways to groundwater. Chlorinated solvents are particularly insidious pollutants because of their persistency, toxicity and the way they travel in aquifers. In this respect, many groundwater supplies are contaminated by such substances, a common cause of which is leaking storage tanks. Unfortunately, cleaning up a polluted aquifer— usually by removing contaminated soil and continuous pumping of the aquifer—is extremely diﬃcult, very costly and takes a great deal of time.7 Industrial eﬄuents also often contain high levels of metals such as iron, zinc, chromium and cadmium, many of which are highly toxic, even carcinogenic. A speciﬁc industrial pollution is related to mining and petroleum extraction. Quarrying and open-case mining, for example, remove the protective layer above an aquifer, leaving it more vulnerable to pollution. Deep mines or oil ﬁelds may produce ﬂuids that are disposed of at the surface and may therefore pollute shallow aquifers, and pollutants from spoil heaps may leach into groundwater. Finally, rising water levels in abandoned mines produce acid mine drainage with subsequent mobilisation of oxidised metal ores, leading to increasing concentrations of sulfate, iron, manganese and other metals, which can cause serious groundwater pollution. Details on risk assessment of industrial pollution can be found in Chapters 5.6 and 7.1.

1.3.3.4

Agriculture

Agriculture is responsible for one of the main pollution groundwater threats. The main source arises from the intensive use of nitrogen-rich fertilisers and of pesticides, a problem that has spread from the industrialised countries to developing ones. The high levels of nitrate and, in some areas, pesticides in groundwater are clearly linked to agricultural activities. This pollution is generally worse where the soil is very permeable, allowing agricultural chemicals to be quickly washed down to underlying aquifers. However, not all nitrates in groundwater are due to agriculture, as we have seen that much of it also originates from untreated sewage. Monitoring techniques can hardly distinguish between the diﬀerent agricultural and sewage nitrates (besides isotopic measurements). The leaching of nitrate from ﬁelds not only leads to pollution but is also a serious source of waste: nitrate that percolates down into aquifers has done nothing to stimulate plant growth. Other components of fertilisers, including potassium and chloride, also ﬁnd their way from ﬁelds to aquifers.

General Introduction: The Need to Protect Groundwater

13

Regarding pesticides and herbicides, substances currently in use are designed to be toxic and, sometimes, persistent. There is no doubt that pesticides are leached through the soil and carried down to underlying aquifers (see Chapter 9.2). In some circumstances, soils can adsorb or immobilise a large fraction of such agricultural chemicals. Many pesticides and herbicides, however, break down slowly under aquifer conditions and, as a result can persist over long time periods. In any case, groundwater pollution data are generally scarce, and the extent of pollution in Europe is hence not accurately known. Further considerations on diﬀuse groundwater impacts from agricultural land use are discussed in Chapters 7.3 and 8.2.

1.4 Groundwater Risk Assessment: Implications for Policy 1.4.1

Groundwater Protection Needs

Public and policy-maker perceptions of groundwater represent an important root cause of emerging problems.8 In many cases, regardless of the degree of formal education individuals have had, perceptions of groundwater resource dynamics are partial at best. Groundwater is often viewed, for example, as an inexhaustible resource, cleaned by the ﬁltering action of aquifers and held, as in a ‘‘bowl’’ or ‘‘lake’’, or ‘‘underground river’’. These perceptions do not reﬂect reality, and often result in use patterns that cause unanticipated problems. Most misunderstandings relate to the scale of aquifer systems, the distribution of groundwater within them and the timescales on which groundwater systems function (see Chapter 5.5). Groundwater protection relies on two closely interlinked components:7 (i) land surface protection, based on hydrogeological concepts and information particularly regarding aquifers and vulnerability; and (ii) groundwater protection responses for potentially polluting activities, giving guidelines on the acceptability of the activities, investigation requirements and, where appropriate, the likely planning or licensing controls. Groundwater protection schemes enable authorities to take account of (i) the potential risks to groundwater resources and sources; and (ii) geological and hydrogeological factors, when considering the control and location of potentially polluting activities.7 Although groundwater is one of the world’s key natural resources, it is still not suﬃciently well protected and poorly controlled. In many instances, the extent of groundwater pollution still remains to be evaluated or even detected, given the slow rates of groundwater movement and the volume of storage involved. In 1996, UNEP indicated that what we know of pollution levels in aquifers may only be the tip of an underground iceberg,6 which consideration is still valid ten years later. There is, in any case, no doubt that this important resource needs to be better protected from both over-exploitation and pollution. This is the aim of the developing groundwater legislation which is described in Chapter 3.1.

14

1.4.2

Chapter 1

Assessment, Prevention and Control

Assessment and appropriate controls of major threats are highly necessary particularly in areas subject to irreversible side eﬀects such as saltwater intrusion and land subsidence (in the case of over-abstraction), but also in the case of high aquifer vulnerability, taking into account pollutant loads (in the case of pollution). Vulnerability assessments and controls require a number of legal and administrative steps, examples of which are described in the light of the EU Water Framework Directive and associated Groundwater Directive (see Chapter 3.1). In the case of pollution, there is a need to distinguish between point sources of pollution, such as landﬁlls and speciﬁc industrial discharges, and diﬀuse sources of pollution linked to agricultural activities and to a lesser extent atmospheric deposition. Eﬀorts should be made to reduce pollution from pollution sources, paying particular attention to practices in areas where aquifers are highly vulnerable. This means that land-use planning regulations have to be enforced in the most sensitive areas. This also concerns of course agricultural practices (see Chapter 8.2). A key step in assessing pollution risks is based on the analysis of the groundwater vulnerability (ease with which groundwater may be contaminated by human activities). It depends on the time of travel of inﬁltrating water (and contaminants), the relative amount of contaminants and the contaminant attenuation capacity of the geological materials through which the water and contaminants inﬁltrate (see further discussion in Chapters 5.4 and 9.2). As all groundwater is hydrologically connected to the land surface, it is the eﬀectiveness of this connection that determines the relative vulnerability to contamination. Groundwater that readily and quickly receives water (and contaminants) from the land surface is considered to be more vulnerable than groundwater that receives water (and contaminants) more slowly and in lesser quantities. The travel time, attenuation capacity and quantity of contaminants are a function of the following natural geological and hydrogeological attributes of any area: (i) the subsoils that overlie the groundwater; (ii) the type of recharge, whether point or diﬀuse; and (iii) the thickness of the unsaturated zone through which the contaminant moves. In general, little attenuation of contamination occurs in bedrocks via ﬁssures, and high attenuation will be possible in subsoils (sands, gravels, glacial tills (or boulder clays), peat, alluvial silts and clays). Groundwater is most at risk where the subsoils are absent or thin and, in areas of karstic limestone, where surface streams sink underground at swallow holes. Vulnerability may be mapped according to the elements in Table 1.1. Establishing such maps is an important part for deciding upon a groundwater protection scheme and an essential element in the decision-making on the location of potentially polluting activities.7 Firstly, the vulnerability rating for an area indicates, and is a measure of, the likelihood of contamination. Secondly, the vulnerability map helps to ensure that a groundwater protection scheme is not unnecessarily restrictive on human economic activity. Thirdly, the vulnerability maps help in the choice of preventive measures and enable developments, which have a signiﬁcant potential to contaminate, to be located

15

General Introduction: The Need to Protect Groundwater

Vulnerability mapping guidelines (adapted from Ref. 7).

Table 1.1

Hydrogeological conditions Vulnerability rating

Extreme High Moderate Low

Subsoil permeability (type) and thickness

Unsaturated zone

Kast features

High permeability (sand/gravel)

Moderate permeability (e.g. sandy subsoil)

Low permeability (e.g. clayey subsoil, clay, peat)

(Sand/gravel aquifers only)

(o30 m radius)

0–3 m 43 m N/A N/A

0–3 m 3–10 m 410 m N/A

0–3 m 3–5 m 5–10 m 410 m

0–3 m 43 m N/A N/A

– N/A N/A N/A

Risk to Groundwater

Hydrogeological Factors

Vulnerability to contaminants

Figure 1.1

Groundwater Value

Other Factors

Contaminant Loading

Preventive Measures

Factors under consideration for groundwater risk assessment.

in areas of lower vulnerability. This assessment strongly relies on proper characterisation of groundwater settings and related risks, an issue that is discussed in Chapter 5.1. In this respect, the risk depends on (i) the hazard aﬀorded by a potentially polluting activity, (ii) the vulnerability of groundwater to contamination, and (iii) the potential consequences of a contamination event. The hazard depends on the potential contaminant loading. The natural vulnerability of the groundwater dictates the likelihood of contamination if a contamination event occurs. The consequences to the target depends on the value of the groundwater, which is normally indicated by the aquifer category (regionally important, locally important or poor) and the proximity to an important groundwater abstraction source (e.g. a public supply well). The risk assessment encompasses geological and hydrogeological factors and factors that relate to the potentially polluting activity. This is illustrated in Figure 1.1. In the light of the above, prevention of groundwater contamination is of critical importance and must be a key aim for the following reasons:7 Once groundwater contamination occurs, the consequences last far longer than surface water contamination (months, years and sometimes

16

Chapter 1

decades) because groundwater moves slowly through the soil and unsaturated zone of the aquifer. Remediation is frequently not practical or is very expensive. Also, it is both impractical and a poor environmental strategy to provide comprehensive treatment to remove certain pollutants, such as pesticides and other trace organics. It is therefore preferable to prevent or reduce the risk of groundwater contamination than to deal with its consequences. Groundwater is an important resource used for drinking water, industry and agriculture, and should be protected for present and future usage. Groundwater provides the base ﬂow (i.e. the water which feeds rivers year-round, and upon which ﬂood ﬂows are superimposed) to surface water systems, many of which are used for water supply and recreational purposes.

1.4.3

Monitoring

The assessment and controls of groundwater quantity and quality have to be supported by representative and reliable monitoring data. Groundwater monitoring is largely a national responsibility but, because groundwater does not respect national boundaries, monitoring has to be conceived at national, regional and international levels. This is exactly the principle of the Water Framework Directive (see Chapter 3.1). Improved monitoring systems are needed to provide information not only on groundwater quantity and quality in the light of its use as drinking water resources, but also for the evaluation of its environmental quality in relation to associated aquatic and directly dependent terrestrial ecosystems. In particular, existing monitoring networks hardly provide early warning of pollution, and networks are needed to include monitoring of pollution loads, particularly in vulnerable recharge areas. Recommendations have been given by UNEP in this respect6 (see Table 1.2). Groundwater quality monitoring has at least four objectives that need to be carefully distinguished in the design of monitoring systems: deﬁnition of the extent of groundwater pollution (analysis of pressures and impacts); quality control of groundwater used as drinking water (drinking water supply surveillance); early discovery of groundwater pollution from a given activity (oﬀensive detection monitoring); and provision of advance warning of the arrival of polluted water at important sources of supply (defensive detection monitoring). Detailed monitoring guidelines in support of policy implementation, more speciﬁcally to the EU Water Framework Directive and new Groundwater Directive (see Chapter 3.1), have been developed and are summarised in Chapter 6.1.

a

FC: faecal coliforms.

Chlorinated hydrocarbons, other organic compounds, metals

Cl, NO3, NH4, SO4, FC

Domestic wastewatera Industrial eﬄuent

Pollution indicators

High

Bacteriological Monitor at water table

High

Chemical

Pollution risk

Early warning monitoring required

Hours to weeks

Fractured

Unconﬁned

Aquifer type

Cl, NO3, NH4, SO4, FC Metals, range of organic compounds

Monitor at water table

Moderate

High for mobile compounds

Days to months

Thin unsaturated zone

Granular

Persistent organics (metals)

Cl, NO3, NH4, SO4, FC

(1) Monitor unsaturated zone and (2) at water table

Low

High for mobile and persistent compounds

Years to decades

Deep unsaturated zone

Key elements in an early warning monitoring strategy (adapted from Ref. 6).

Travel time from surface to saturated aquifer zone

Table 1.2

Persistent organics (metals)

Cl, NO3, NH4, SO4

Monitor semiconﬁning layer and aquifer

Moderate for persistent compounds only Very low

Decades +

Semi-conﬁned

General Introduction: The Need to Protect Groundwater 17

18

Chapter 1

1.5 Conclusions Groundwater protection is essential in relation to the intrinsic water resource quality for various uses, and its environmental value. Basic environmental principles include the following: Sustainable development principles, seeking to ensure that economy and society can develop to their full potential within a well-protected environment, and with responsibility towards present and future generations and the wider international community. Precautionary principle, requiring that emphasis should be placed on dealing with the causes, rather than the results, of environmental damage and that, where signiﬁcant evidence of environmental risk exists, appropriate action should be taken even in the absence of conclusive scientiﬁc proof of cause. Polluter pays principle, of which the objective is to allocate correctly the costs of pollution, consumption of energy and environmental resources, and production and disposal of waste to the responsible polluters and consumer, rather than to society at large or future generations, which in turn provides an incentive to reduce pollution and consumption.

References 1. J. Gibert, in Groundwater Ecology: A Tool for Management of Water Resources, ed. C. Griebler, D. L. Danielopol, J. Gibert, H. P. Nachtnebel and J. Notenboom, European Commission, EUR 1987, 2001, p. 413. 2. S. Heidrich, M. Schirmer, H. Weiss, P. Wycisk, J. Grossmann and A. Kaschl, Toxicology, in press. 3. P. Grathwohl, Diﬀusion in Natural Porous Media, Contaminant Transport, Sorption/Desorption and Dissolution Kinetics, Kluwer Academic, 1998. 4. G. Teusch, H. Ruegner, D. Zamﬁrescu, M. Finkel and M. Bittens, Land Contamin. Reclam., 2001, 9, 1. 5. EEA, Groundwater Quality and Quantity in Europe, Environmental Assessment Report 3, Copenhagen, 2000. 6. UNEP, Groundwater: a threatened resource, UNEP Environment Library 15, Nairobi, Kenya, 1996. 7. Irish EPA, Groundwater Protection Schemes, 1999 (ISBN 1-899702-22-9). 8. J. J. Burke and M. H. Moench, Groundwater and Society: Resources, Tensions and Opportunities, United Nations, 2000 (ISBN 92-1-104485-5).

2. Science–Policy Integration Needs

CHAPTER 2.1

Science–Policy Integration for Common Approaches Linked to Groundwater Management in Europew PHILIPPE QUEVAUVILLER European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium

2.1.1

Introductory Remarks on Science–Policy Integration Needs

It is now well recognised that a better understanding of environmental problems requires an improved awareness of multidisciplinary scientiﬁc developments.1 Awareness by itself is, however, not suﬃcient, and a better research integration is also required at the various stages of policy developments (design, development, implementation and review). Ideally, the relevant research for any environmental policy should be feeding the policy-making process directly in a ‘‘tailor-made fashion’’ so that results may be used in the right way at the right time (in relation to the policy agenda). In many instances, however, this is still far from being the case. In the sector of European Union (EU) water policies, on-going discussions are taking place among the scientiﬁc community (representatives of research consortia and of the European Commission’s Research Directorate-General) and the policy-making community (representatives of EU member states’ environment agencies or ministries and the European Commission’s Environment Directorate-General) to examine research–policy coordination needs. These discussions have highlighted the key importance of improving/increasing the information and communication ﬂow within and between these w

The views expressed in this chapter are purely those of the author and may not in any circumstances be regarded as stating an oﬃcial position of the European Commission.

21

22

Chapter 2.1

communities.2 On the one hand, the implementation of existing policies and the development of new ones are constantly fed by scientiﬁc inputs, which must fulﬁl speciﬁc thematic and timing requirements in order to be of direct value to the policy-making process. On the other hand, research and technological development (RTD) activities aiming to support relevant environmental policies have to take into account the policy-making agenda and its speciﬁc needs to adapt the objectives and the scope of their related work programmes. Experience has shown, however, that this interrelationship is not as eﬃcient as it could/should be, owing to a lack of a clear coordination mechanism. In fact scientists view the end-user in the research project as the ‘‘legitimate’’ client for their research results, but on the ground there is a signiﬁcant lack of transfer mechanisms that would allow passing on of the relevant information to other stakeholders or policy makers.2 The latter often do not have the time, the capacity to translate research results into policy, or even simple access to speciﬁc technical journals and thus relevant information mostly remains within the specialised scientiﬁc community. Communication diﬃculties are also linked to the diﬀerent ‘‘jargon’’ used in the diﬀerent communities. One problem hampering proper science–policy integration is related to diﬀerent timing concepts of research and policy. Policy tends to operate on the short or medium term with clearly deﬁned milestones, while science is generally developed on a long-term basis. In addition, policy tries in most instances to achieve an acceptable (political) compromise (which is, sometimes somewhat critically, labelled as the ‘‘least common denominator’’), whereas the scientiﬁc community strives to obtain objective, scientiﬁc facts and wants to understand a phenomenon to the greatest possible detail. In addition, scientiﬁc career reviewing schemes rarely give credit to works for the integration of knowledge to fulﬁl policy objectives. As a result, scientists are not suﬃciently motivated to perform this type of work and in most cases they may not be aware about the speciﬁc research issues needed at a certain time to eﬀectively support policy development.2 In many instances, knowledge is available but is either not widely known or is insuﬃciently transferred from the scientiﬁc community to the other relevant parties. Furthermore, policy-makers, planners and implementers are often not suﬃciently ‘‘entrepreneurial’’ with respect to the selection and use of state-of-the-art research projects that would directly feed the policy objectives (which is essentially due to time pressures and lack of streamlined information). As is discussed below, this is often due to a lack of dialogue or an inappropriate communication of relevant information. In the best case, policy-makers are able to deﬁne a list of open questions and then, on this basis, open a discussion with the scientiﬁc community in order to understand if the questions are addressing the actual research needs or whether the subject matter has already been treated suﬃciently in the past so that responses may be obtained from readily available results. Such a mechanism, however, is certainly not straightforward and often not suﬃciently operational. The development of environmental policies mixes legal requirements with issues of technical feasibility, scientiﬁc knowledge and socioeconomic aspects and requires intensive multi-stakeholder consultations. In this context, the

23

Science–Policy Integration for Common Approaches POLICY IMPLEMENTATION POLICY DEVELOPMENT

RESEARCH, SCIENTIFIC PROGRESS, POLICY INTEGRATION

POLICY REVIEW

DESIGN OF POLICY

Figure 2.1.1

Integration of scientiﬁc progress into the policy-making process.

consideration of scientiﬁc progress represents one of the key aspects for the design of new policies and the review of existing ones. Within the EU, this consideration is fully embedded into the Sixth Environmental Action Programme which stipulates that ‘‘sound scientiﬁc knowledge and economic assessments, reliable and up-to-date environmental data and information, and the use of indicators will underpin the drawing-up, implementation and evaluation of environmental policy’’.3 This requires, therefore, that scientiﬁc inputs constantly feed the environmental policy process. This integration also involves various players, namely the scientiﬁc and policy-making communities but also representatives from industry, agriculture, non-governmental organisations etc. (Figure 2.1.1). The problem of non-coincidence of research and policy agendas depends upon the stage of development of the policy. It is more acute at the starting phase of policy design such as thematic strategies deﬁned by the EU 6th Environment Action Programme (e.g. Soil Thematic Strategy).3 In comparison, the Water Framework Directive (WFD),4 along with its related Common Implementation Strategy (see Chapter 4.1), provides a stable platform which allows the building of strong partnerships among policy and scientiﬁc communities. In the context of the WFD, we may distinguish R&D needed in support of policy development (e.g. background information required for designing ‘‘daughter directives’’ such as the new Groundwater Directive; see Chapter 3.1) and R&D directly feeding implementation (short- or medium-term research and demonstration projects supporting speciﬁc milestones such as e.g. monitoring, integrated management). Long-term R&D is of course also necessary for reviews and possible improvement of the legislation (such as the technical adaptations covered by the WFD and the review periods under the River Basin Management plans). At the present stage, eﬀorts are lacking for presenting results of research and demonstration projects in a form that policy-makers can easily use, e.g. ‘‘science-digested’’ policy briefs. Adversely, one may stress that the consideration of research results by the policy-making community is not straightforward, mainly for political reasons and diﬃculties in integrating the latest research developments in legislation. The diﬃculty is enhanced by the fact that the policy-making community is probably not deﬁning its role as ‘‘science

24

Chapter 2.1

customer’’ suﬃciently well. In other words, the dialogue and communication are far from being streamlined to allow for an eﬃcient ﬂow of information. In this respect, improvements could be achieved through the development of a ‘‘science–policy interface’’ based on a coordination of relevant programmes/ projects with direct relevance to the WFD implementation. This issue is discussed in Chapter 11.3. Integration in a broad sense, as perceived for the environmental policy sector, goes of course beyond the science–policy issues. It concerns the need to consider the environment as it appears in all relevant policies, interactions of various environmental compartments, socioeconomic aspects, etc. This chapter actually focuses on integration of scientiﬁc and technological progress into the policy-making and implementation process, in particular in the groundwater policy sector, illustrating the necessity and complexity of the knowledge-based approach. The above considerations are largely inspired by recent papers discussing this issue.2,5

2.1.2

Science Integration in the Light of Groundwater Management

When dealing with groundwater science and policy, the interdependent nature of hydrological, hydrogeological and water-use systems makes the continuum between data, information and knowledge of particular importance.6 Understanding systemic interactions is essential as a basis of identifying groundwater management options and generating suﬃcient social consensus to implement them. Developing this understanding requires a steady ﬂow of hydrological data as well as data on water use. It also requires continuous reﬁnement of the scientiﬁc foundation, both physical and social, upon which policy and management solutions rely. Examples of this type of progressive approach to groundwater information needs are quoted by Burke and Moench:6 e.g. the Groundwater Forum in the UK7 and the Groundwater Foundation in the USA represent initiatives to address, respectively, groundwater research needs and public education. Another example is the Working Group on Groundwater of the Common Implementation Strategy of the WFD, which is described in Chapter 4.1. Lack of data and scientiﬁc understanding of groundwater resources often represents a critical gap undermining the development of groundwater management approaches and institutions.6 The absence of data often limits the degree to which hydrogeologists are able to quantify and describe complex aquifer dynamics. Equally important are the ways in which raw data and information are treated, presented and used. Information is only useful if it used. For this to occur, the information must be accessible to potential users and presented in a manner they can understand. This is discussed in Chapter 11.3 with regard to the development of the EU Water Information System for Europe (WISE). In many instances, however, the absence of information (or the non-use of essential information) creates situations in which emerging

Science–Policy Integration for Common Approaches

25

problems and management options are poorly understood. As a result, the essential nature of basic data and research on hydrogeology should be clear. Generating the technical information required to meet emerging groundwater management needs depends on at least three types of scientiﬁc data collection and analytical activities:6 long-term baseline monitoring for understanding the dynamics of hydrogeological systems and providing warning of emerging problems; targeted research on basic processes; and site-speciﬁc analysis of problems and management options at local levels. Technicians often make a strong plea for more data, and when they obtain them, they frequently present tentative, diﬃcult-to-interpret scenarios and request ﬁnancing to collect more data in order to strengthen their interpretations.6 Policy-makers and the general public, on the other hand, often have unrealistic expectations regarding the ability of technical and scientiﬁc analyses to provide straightforward answers, particularly given the general paucity of groundwater data and the limited means with which to gather and analyse them. Hydrogeological systems are rarely simple. The importance of bridging this gap between the data available and their policy implications relates directly to some of the fundamental challenges facing the development of management institutions.

2.1.3

Examples of Projects in Support of Groundwater Policy

As highlighted in Section 1, the relevant research for any environmental policy should ideally be feeding the policy-making process directly in a ‘‘tailor-made fashion’’ so that results may be used in the right way and at the right time (in relation to the policy agenda).2 In many instances, however, this is far from being the case. This paragraph lists some RTD projects directly or indirectly related to EU groundwater policy and examines how they were related to policy implementation (Directive 80/68/EEC8 and/or Water Framework Directive) or development (proposal for the new Groundwater Directive described in Chapter 3.1).9 Details on the RTD funding mechanisms and on the priorities under which the projects discussed below have been funded are given in Chapter 11.3. It should be noted that the examples given below do not prejudge about the way project outputs were eﬀectively transferred to the user community (an impact assessment analysis has not been done so far to enable this issue to be concretely discussed), nor do they pretend to provide an exhaustive list of ongoing and terminated projects.

2.1.3.1

Risk Assessment

Research supporting improvements of groundwater (pollution) risk assessment has been ﬂourishing over the past few years (from 1998 onward). Examples are

26

Chapter 2.1

illustrated in several chapters of this book. Most of these projects directly or indirectly support groundwater characterisation needs as well as the future programmes of measures required under the WFD.4 They concern, for example, studies of contaminant spreading in fractured underground reservoirs leading to conceptual geological modelling and database development for organic pollutants in groundwater (TRACE Fracture project; contact: [email protected]) (see Chapter 5.6); and groundwater risk assessment at contaminated sites leading to guidelines (GRACOS project; contact: [email protected]).

2.1.3.2

Groundwater Remediation

Examples of projects of direct applicability to groundwater remediation concern, for example, integrated concepts for groundwater remediation, including guidelines and studies on natural attenuation of organic pollutants in groundwater (INCORE project; contact: [email protected]) (see Chapter 5.8); on-site remediation of groundwater contaminated by polar organic compounds using a new adsorption technology (OROGONATE project; contact: [email protected]); and protection of groundwater at industrially contaminated sites (PURE project; contact: [email protected]).

2.1.3.3

Diﬀuse Pollution

Linked to groundwater analyses of pressures and impacts, some projects examine issues such as, for example, actual status and scenarios of pesticides in European groundwaters (PEGASE project; contact: [email protected]), integrated soil and water protection from diﬀuse pollution (SOWA project; contact: [email protected]), etc. The two projects are described, respectively, in Chapters 9.2 and 5.7.

2.1.3.4

Management Issues and Information Tools

R&D projects also support groundwater management and information tools that are directly or indirectly in support of the WFD river basin management planning. Examples concern, for example, the development of an online portal in the form of a web-based information platform for soil, groundwater and contaminated land (EUGRIS project; contact: [email protected]); integrated management system (IMS) for the prevention and reduction of contamination at large-scale contaminated sites (WELCOME project; contact: [email protected]) (see Chapter 7.1); tools for management and control of hazardous compounds in water catchment areas (WATCH project; contact: [email protected]) (see Chapter 8.3); etc. The scientiﬁc basis for improved river basin management through a better understanding of the river– sediment–soil–groundwater system as a whole, at diﬀerent temporal and spatial scales, is studied in an integrated project, the AQUATERRA project (contact: [email protected]) (see Chapter 2.2).

Science–Policy Integration for Common Approaches

2.1.3.5

27

Relevant Networks

Besides research projects, networks involving the scientiﬁc community, industrial stakeholders, representatives of international associations, NGOs, etc. are directly or indirectly contributing to networking activities that are relevant to groundwater policy development and implementation. Examples of such networks provide exchange of innovative know-how in the ﬁeld of applied research for contaminated land and groundwater issues (ANCORE network: www. ancore.org); develop technical recommendations for sound decision-making on the rehabilitation of contaminated sites in Europe (CLARINET network: www.clarinet.at); and discuss issues of industrially contaminated land in Europe (NICOLE network: www.nicole.org). Other networks are focused on training, e.g. on innovative management of groundwater resources in Europe (IMAGETRAIN project: www.image-train.net), or on the coordination of national research programmes, e.g. on sustainable management of soil and groundwater under the pressure of soil pollution and soil contamination (SNOWMAN project; contact: [email protected]) (see Chapter 11.1).

2.1.4

A Project ‘‘Tailor-Made’’ to Support the New EU Groundwater Directive: BRIDGE

The BRIDGE (background criteria for the identiﬁcation of groundwater thresholds) project was designed to develop a common methodology, intended for possible use by member states, on ‘‘how to derive groundwater threshold values’’. The project was developed in 2004–2006 and has just been concluded. It has been carried out at European level, involving a range of stakeholders and eﬃciently linking the scientiﬁc and policy-making communities. The diﬀerent objectives were: To evaluate and assemble scientiﬁc outputs to set out criteria for the assessment of the chemical status of groundwater. These criteria are data for characterisation of natural and anthropogenic pollutants, parameters indicative for pollution, data for characterisation of groundwater bodies as hydrologic and hydrogeological parameters. To derive a plausible general approach, how to structure relevant criteria appropriately with the aim to set representative groundwater threshold values scientiﬁcally sound and deﬁned at national river basin district or groundwater body level. To check the applicability and validity of this approach by means of case studies at the European scale, and to carry out an environmental impact assessment taking into account the economic and social impacts. The ﬁnal methodology for the derivation of environmental thresholds for pollutants at a national or regional level of groundwater bodies or river basins has been developed in close consultation with representatives from member

28

Chapter 2.1

states’ environment ministries and agencies, and stakeholders from the CIS Working Group on Groundwater. It had to take due considerations of the negotiation of the new Groundwater Directive which was running at the same time, which represented an additional challenge above the sole scientiﬁc one. The ﬁnal meeting was held in Paris on the 15 December 2006. The proposed method for deriving groundwater threshold values will now be directly communicated to the member state experts for policy discussions and expected adoption before summer 2007. The research will therefore fulﬁl one of the requirements of the new Groundwater Directive, requiring member states to establish groundwater threshold values by the end of 2008, following a common methodological approach. More details on the project are given in Chapter 9.1. The project was terminated at the end of 2006 and a CD-ROM provides a copy of the diﬀerent ﬁndings and workpackage reports to interested stakeholders. Contact: AnneMarie Fouillac (email: [email protected]).

2.1.5

Conclusions: Some Research Needs

Some basic research needs in support of groundwater management and policy have been expressed by Burke and Moench6 and are outlined below. Basic hydrological and hydrogeological data collection and research, including long-term monitoring of groundwater (levels and quality) and surface water (ﬂows and quality); the synthesis of background information (hydrogeological and socioeconomic data); and detailed investigations in areas where indicators signal the emergence of speciﬁc problems. Operational and basic research activities are also essential in order to document and evaluate the results of management attempts and in order to understand key elements of the hydrogeological processes that underpin various management options. Monitoring the resource use and potential sources of pollution. Basic information on how groundwater resources are being used and who is using them is as essential for management as scientiﬁc data on the aquifer system itself. Without understanding how groundwater resources are being used, particularly for irrigated agriculture, it is impossible to identify points of management leverage. Devising management systems requires detailed knowledge of such factors as the actual locations where groundwater extraction is occurring, the eﬃciency with which it is used and its role in agriculture and other water-use systems. Programmes for registering wells and estimating, either directly or indirectly, the amount of groundwater extracted are essential steps. Similarly, it is important to identify, as far as possible, other activities that could have substantial impacts on groundwater conditions. This includes the basic information on potential point and non-point sources of pollutants arising from industry, domestic sewage and agricultural chemical use.

Science–Policy Integration for Common Approaches

29

Besides these research needs, which are still valid at the time of publication of this book, other issues have been highlighted during the development of the new Groundwater Directive, in particular: research on groundwater ecosystems (see Chapter 11.2), and on interactions among groundwater and associated aquatic and dependent terrestrial ecosystems to strengthen the knowledge about environmental groundwater quality; research on groundwater conceptual modelling, georeferencing system, visualisation, vulnerability assessment, enabling to better assess quantitative and qualitative pressures on groundwater systems, as well as the eﬀectiveness of programmes of measures; improved risk assessment and characterisation methods (e.g. support to standardisation of assessment methods for pollutant mobilisation from diﬀerent types of pollution sources), integrated management approaches, innovative measures for a better protection of groundwater against pollution, etc.; R&D linked to improved programmes of measures, including integration of requirements of diﬀerent regulations (see Chapter 3.1), and considering identiﬁed pressures and the way to tackle them with technically feasible, cost-eﬀective and socially acceptable measures complying with current legislation; and development of new monitoring methods (e.g. in situ techniques, ﬁeld monitoring) and improvement of data quality and comparability, and pre-normative research related to operationally deﬁned operations (e.g. sampling) and parameters (e.g. extractable contents of chemical substances), etc. The EU 7th Framework Programme of Research and Technological Development10 will take some of these considerations into account in the Theme 6 ‘‘Environment’’. Research outputs will then need to be eﬃciently transferred to policy-makers, along the principles described in Chapter 11.3.

References 1. UK Department for Environment, Food and Rural Aﬀairs, Science Meets Policy in Europe, DEFRA Report, London, 2005. 2. Ph. Quevauviller, P. Balabanis, C. Fragakis, M. Weydert, M. Oliver, A. Kaschl, G. Arnold, A. Kroll, L. Galbiati, J. M. Zaldivar and G. Bidoglio, Environ. Sci. Pol., 2005, 8, 203. 3. European Commission, 6th Environment Action Plan 2001–2010, 2001. 4. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities, L 327, 22.12.2000, p. 1. 5. Ph. Quevauviller, J. Environ. Monitor., 2005, 7(2), 89.

30

Chapter 2.1

6. J. J. Burke and M. H. Moench, Groundwater and Society: Resources, Tensions and Opportunities, United Nations, 2000 (ISBN 92-1-104485-5). 7. D. R. C. Grey et al., Groundwater in the United Kingdom. A Strategic Study. Issues and Research Needs. Groundwater Forum Report FG/GF 1, Foundation for Water Research, Marlow, UK, 1995. 8. Council Directive 80/68/EEC of 17 December 1979 on the protection of groundwater against pollution, Oﬃcial Journal of the European Communities, L 20, 26.1.1980, p. 43. 9. Directive of the European Parliament and of the Council on the protection of groundwater against pollution and deterioration, 2006. 10. 7th Framework Programme for Research and Technological Development (2006–2013), European Commission, rue de la Loi 200, B-1049 Brussels, 2006.

CHAPTER 2.2

Transferring Scientiﬁc Knowledge to Societal Use: Clue from the AQUATERRA Integrated Project PHILIPPE NE´GREL,a DOMINIQUE DARMENDRAILa AND ADRIAAN SLOBb a

Bureau de Recherches Ge´ologiques et Minie`res (BRGM), 3 avenue Claude Guillemin, FR-45060 Orle´ans ce´dex 2, France; b TNO Environment and Geosciences, Dept Innovation & Environment, Van Mourik Broekmanweg 6, PO Box 49, NL-2600 AA Delft, The Netherlands

2.2.1

Introduction

Scientiﬁc knowledge and data are of signiﬁcant importance in a global information society. They should both be used for promoting innovation and economic development, for eﬃcient and transparent decision-making, particularly at the governmental level, and for education and training. However, scientiﬁc data and information should be as widely available and aﬀordable as possible. The more people are able to share them, the greater the positive eﬀects and returns to society will be. Scientiﬁc knowledge is a ‘‘public good.’’ The inherent function of scientiﬁc investigations is to carry out a comprehensive questioning of nature leading to further knowledge. It is this new knowledge that allows cultural and intellectual enrichment and leads to the technological advances and beneﬁts arising from science. Promoting fundamental research is a priority towards achieving development and progress. There can be no applied science if there is no science to apply and thus science is for knowledge; knowledge is for progress as stated by the World Science Forum (2003; http://www.sciforum.hu/index.php). In the ﬁeld of water resources management, policy plays a key role. The policy approach is becoming more holistic and the trend is to address the whole 31

32

Chapter 2.2

groundwater–river–sediment–soil system. This widens the quite narrow deﬁned issue of groundwater to a variety of related issues, like nature conservation, the use of space, and economic and social issues. It also confronts groundwater policy with new stakeholders that were not involved in groundwater issues before. So new stakeholders with interests from nature conservation to entrepreneurial interests are entering the policy arena. The until now usual technical approach of the groundwater issue does not ﬁt with these developments, so a new policy approach as well as new ways of involving science in policy are needed. River basins are facing diﬀerent types of pressures, such as impacts of agriculture, industries or urban waster water treatments. The drivers–pressures–state–impact–response (DPSIR) approach as illustrated by Figure 2.2.1 gives insight into environmental processes and the links between human activities and their impact on the environment. Economic activities (driving forces), such as industry, agriculture and tourism, lead to increasing pressures on the natural environment as these activities result in use of natural resources and/or emissions (accidental or controlled) of waste to water (surface and/or groundwater), soil and sediment. The use of resources and/or emissions will change the state of these environments in quantity and/or quality: sediment, water and soil resources are depleted and/or they are loaded (contaminated) with hazardous substances originating from economic activities. Above a certain level of depletion and/or contamination the environment may be impacted, i.e. loss of biodiversity, vulnerability to ﬂoods and landslides, decreased chemical and/or ecological water, soil or sediment quality and/or a shortage of these resources. Thus the DPSIR framework provides a helpful, conceptual framework for further improvement of system understanding.

Figure 2.2.1

Pressures on river basins.

Transferring Scientiﬁc Knowledge to Societal Use

33

In the past, water quality studies and monitoring in the major river basins1 have developed according to: (i) the water demand, which is increasing exponentially; (ii) the development of key issues, such as eutrophication, acidiﬁcation, salinisation; (iii) new pressures (radionuclides since the 1950s, pesticides since the 1980s, endocrine disruptors and pharmaceuticals more recently); and (iv) the development of new technical tools such as analytical chemistry, in situ monitoring or models. Management of river basins will certainly beneﬁt from an increased scientiﬁc understanding of the functioning of the biophysical system (i.e. water–sediment–soil interactions) and of its relation to the societal system like the application of scientiﬁc knowledge in river basin management and in policymaking and implementation. This chapter deals with the preliminary results issued from the 6th Framework Programme (FP) Integrated Project AQUATERRA and the transfers from the scientiﬁc research area to a policy and societal use.

2.2.2

Overview of the AQUATERRA Project

The European Commission (EC) anticipated the need to improve our common understanding of the functioning of river systems, especially in relation to changes in land use and climate. Hence, under their RTD Framework Programmes several research projects have been funded, aimed at improving this understanding, and at supporting the Water Framework Directive (WFD) implementation. AQUATERRA, the full title of which is ‘‘Integrated modelling of river– sediment–soil–groundwater systems: advanced tools for the management of catchment areas and river basins in the context of global change,’’ is one of the ﬁrst Integrated Projects within the 6th European Union (EU) Framework Programme (see website at http://www.eu-aquaterra.de/ for detailed information). AQUATERRA, including a multidisciplinary team of 45 partner organisations in 12 EU countries (researchers, but also practitioners and end-users such as policy-makers, river basin managers and regional and urban land planners), has been active since 1 June 2004 and will run until May 2009. AQUATERRA aims to provide the scientiﬁc basis for an improved river basin management through a better understanding of the river–sediment–soil– groundwater system as a whole, by integrating both natural and socioeconomic aspects at diﬀerent temporal and spatial scales.2 This should be applicable to European contexts facing modiﬁcations or changes due to climate change, land use and pollution of soil and water. The principal task of AQUATERRA is to provide the foundations for an improved understanding of the behaviour of environmental pollutants in order

34

Chapter 2.2

to better evaluate the evolution of water quality at diﬀerent scales, from local to global scales. New ﬁeld and laboratory as well as historical data will be assembled and addressed in four European river basins (Ebro, Meuse, Elbe and Danube) and a small French catchment (Bre´villes). Based on these biogeochemical, climatological and material ﬂux data, new simulation models will help to outline trends and pollutant transport behaviour with respect to soil functioning and the water cycle. These models will integrate key biogeochemical and hydrological processes from the laboratory to the river basin scale. AQUATERRA is divided in ten sub-projects with diﬀerent functions to provide information and logistical support to each other as illustrated in Figure 2.2.2. Note that the whole project is presented in a ﬂyer (available at http:// www.attempto-projects.de/aquaterra/uploads/media/Flyerﬁnal_13_10_05.pdf). Some sub-projects have a scientiﬁc focus (i.e. TREND, FLUX, COMPUTE, BIOGEOCHEM, BASIN, HYDRO, MONITOR) while others are more focused on the social sciences and European policies (i.e. EUPOL and INTEGRATOR). Finally, KNOWMAN has the role of knowledge transfer and dissemination. Some new developments have been done in INTEGRATOR and EUPOL sub-projects which combine hard (hydrogeology, chemistry, geochemistry, etc.) and soft (policy, socioeconomics) sciences. Using two of the ﬁve river basins studied in AQUATERRA as case studies, we will explore the new role of science in the societal information use and the development of the Environment Policy, and in particular the Water Resources legal framework.

Impact of Global Change on Soil and Water

TREND Future trends and impacts

HYDRO Global climate Water cycle

FLUX Intercompartement mass fluxes BIOGEOCHEM Key processes Transport functions

Figure 2.2.2

BASIN - Applications • Brevilles • Ebro • Meuse • Elbe • Danube EUPOL - Policies • EU policy framework • R & D requirements INTEGRATOR • Economic and social aspects • Stakeholder needs

Bench scale

MONITOR Screening tools Pollutants

Catchmentscale

COMPUTE Integrated soil-water numerical models

Basin scale

Scientific Methodology

KNOWMAN • Dissemination activities • Knowledge transfer

The ten sub-projects of AQUATERRA working on diﬀerent levels and scales.

Transferring Scientiﬁc Knowledge to Societal Use

2.2.3

Environmental Policies

2.2.3.1

Four Generations of Environmental Policies

35

Environmental policies have been developed continuously from the 1960s, when environmental problems were recognised. Various authors have described the development of environmental policy.3–5 In this chapter we distinguish four generations of environmental policies that developed subsequently.6 These generations are all needed to tackle environmental problems and thus are not replacing or competing with each other but are complementary. The ﬁrst generation started in the 1960s/1970s. Because of poor environmental conditions and health problems in some regions, an environmental consciousness awoke in the public, and governments adopted new sets of environmental regulations and rules. For the most part, environmental measures involved end-of-pipe techniques to reduce emissions. The focus of environmental policies in the ﬁrst generation was on regulatory measures: laws and rules. At the beginning of the 1980s, awareness increased that regulatory measures on their own were not enough to tackle the environmental problems adequately. This cleared the way for voluntary measures. The second generation of environmental policies focused strongly on prevention of environmental damage by a mix of voluntary measures and regulations. The focus of these voluntary and regulatory measures was, for example, on energy saving and waste reduction. Environment had become a topic in quality and safety systems of businesses and environmental management systems had become quite popular. So in the second generation of environmental policies, the attention shifted from end-of-pipe techniques to process-integrated measures designed to prevent pollution being emitted to the environment. In the 1990s, attention shifted from prevention within the factory to prevention of emissions within the whole production–consumption chain. Product oriented policies and chain management are typical examples of the third generation of environmental policies. The environmental aspects of products were measured during their entire life cycle and incorporated in chain management and the product design process. Major eﬀorts were put into the communication with external stakeholders concerning the environmental performance. The policy instruments that were being used were of a voluntary nature, sometimes supported by legislative and/or ﬁnancial measures. In the beginning of the 2000s even new strategies were needed because persistent environmental problems still were present. Some problems like acidiﬁcation had been tackled quite well, but more persistent problems, like the emission of CO2 or the spreading of chemicals in the environment, were harder to tackle. So in the beginning of the present century the need for a fourth generation of environmental policy was derived. Although this generation of environmental policies is still developing, we can see already contours of it. In the last Dutch Environmental Policy Plan 4 (2001) ‘‘transition management’’ is presented as a key concept to tackle persistent environmental problems. The policy approach is rather process-oriented instead of contentoriented and has an interactive character.

36

Chapter 2.2

Table 2.2.1 presents an overview of the four generations of environmental policies. When we examine the characteristics of the generations more closely, a clear trend becomes apparent. We can see an increasing complexity of the environmental problems which is reﬂected in an increasing complexity in environmental policies. Where in the ﬁrst generation goals are set quite straightforwardly, legislative instruments are being used and the number of actors is limited, in the fourth generation goals are discussed and formed with societal actors over a long period, participation is the most important instrument and the number of actors is numerous. Whereas the policy in the ﬁrst generation is directly aimed at cleaning the environment in a straightforward way, in the fourth generation this aim has shifted towards upgrading environmental quality by means of system innovations and societal change.

2.2.3.2

Science in the Four Generations of Environmental Policies

What does the development of environmental policies described above mean for knowledge production in the four generations of environmental policy? In the ﬁrst generation the demand for knowledge comes only from government, and is quite technical in nature. In the fourth generation, the strategy is directed towards societal change to diminish environmental damage and the number of actors involved is quite numerous. The demand for knowledge comes now from diﬀerent actor groups. Therefore, the new policy approach in the third, but especially in the fourth, generation will call for a new way of dealing with knowledge in the policy process. This ﬁts well with the ideas of Gibbons et al.7 who reﬂected on the role of science in society. According to them, science is undergoing a major shift from mode 1 science—the traditional way of production of scientiﬁc knowledge—to mode 2 science. In this mode 2, the societal context is very important for knowledge production. Mode 2 science has ﬁve characteristics that distinguish it from mode 1 science (see Table 2.2.2). Mode 2 science is generated in the context of application, whereas mode 1 science is produced in the ‘‘classic’’ academic environment. ‘‘The context of application describes the total environment in which scientiﬁc problems arise, methodologies are developed, outcomes are disseminated, and uses are deﬁned.’’8 Furthermore, mode 2 science is produced in a ‘‘trans-disciplinary’’ way, ‘‘by which is meant the mobilisation of a range of theoretical perspectives and practical methodologies to solve problems. But, unlike inter- or multidisciplinarity, it is not necessarily derived from pre-existing disciplines nor does it always contribute to the formation of new disciplines.’’8 Mode 1 science is produced in the academic setting (one place, a certain time), is organised in a hierarchical way and involves academic peers to control the quality of the output. Mode 2 science is developed close to the place of application, is organised in a heterarchical manner and quality control is performed by societal actors. Mode 2 science matches quite well to the fourth generation of environmental policy because of the social accountability, the

Staﬀ member

Legislation and external pressure New technology, registration, monitoring

Low

Actors involved

Drivers for actors

Complexity

Actions

Substances, emissions

Process changes, communication (internal and external) Moderate

Eﬃciency

Whole company

Processes

Regulation, voluntary measures

Legislation and regulation

Scope

Prevention

Second generation

Clean-up operations

First generation

Characteristics of four generations of environmental policies.

Means to reduce environmental damage Policy instruments

Table 2.2.1

High

Product design, balanced scorecard, covenants, etc.

Companies and stakeholders, consumers Strategic performance

Voluntary measures, ﬁnancial instruments, regulations Products, production chain processes

Chain management

Third generation

Very high

Sustainability, licence to operate Societal dialogue, institutional change, etc.

Sustainability, societal processes, system innovations Societal groups

Participatory instruments

Network management

Fourth generation

Transferring Scientiﬁc Knowledge to Societal Use 37

38

Chapter 2.2

Table 2.2.2

Characteristics of mode 1 and mode 2 science.7

Knowledge developed Knowledge production Place and way of knowledge production Organisation Quality control

Figure 2.2.3

Mode 1

Mode 2

Academic context (Mono)disciplinary Homogeneous (one place, a certain time)

Context of application Transdisciplinary Heterogeneous (knowledge developed close to the place of application) Heterarchical, transient

Hierarchical, preserves form Academic, peer review

Socially accountable, reﬂexive

General map of the Ebro basin and its distribution within the Communinades Automonas (FR, France; AND, Andorra; CV, Valencia; PV, Pais Vasco; CM, Castilla la Mancha; CN, Galicia; AR, Arago´n; CA, Catalun˜a; CL, Castilla y Leon; Na, Navarra; LR, La Rioja).

fact that it recognises the importance of stakeholder views and, more generally, because of the emphasis on the context of application.

2.2.4

AQUATERRA Case Studies

2.2.4.1

The Ebro Case study

2.2.4.1.1

Context

The Ebro river basin is located in northeast Spain (Figure 2.2.3). The river itself is approximately 930 km long and drains a basin of approximately 85 500 km2 in area, or 17.3% of the surface area of Spain. The total basin area is drained by 347 main tributaries with a combined length of approximately 12 000 km. The

39

Transferring Scientiﬁc Knowledge to Societal Use

river ends at the Ebro delta, one of the most important wetlands in Europe, covering an area of about 320 km2 of sediments with wetlands and coastal lagoons, which are valuables in terms of natural resources and related economic activities. The total mean annual runoﬀ is approximately 6.837 106 m3 and the total water storage capacity of the 187 reservoirs that regulate the water ﬂow is around 57%. The northern tributaries provide a greater contribution to the discharge than those in the south because the former drainage basins are located in higher rainfall areas. Monthly water discharge is quite irregular, with a signiﬁcant decrease during the 20th century attributed mainly to increased water use for human activities (agricultural irrigation, reservoirs, electricity production and domestic consumption). Large hydroelectric power plants in the Ebro system supply 50% of Spain’s electricity. Irrigation is responsible for an important hydraulic deﬁcit. An average of 300 m3 s1 is taken oﬀ the river: its natural ﬂow in the early 20th century (1914–1935) was around 590 m3 s1, whereas in the last few decades (1960–1990) it was 430 m3 s1, i.e. a 28% reduction.9 A summary of the present and potential future drivers and pressures acting upon diﬀerent economic areas in the Ebro basin (agriculture, industry and tourism) is shown in Table 2.2.3.

Table 2.2.3

Present Agriculture Industry

Tourism Other Future Agriculture

Summary of main drivers and pressures in the Ebro basin Drivers

Pressures

Increased productivity Abandonment of agricultural land in high altitude zones Increased use of hydroelectricity

Water supplies Soil quality

Chemicals, metals and manufacturing industries Inﬂux of tourists during summer months Recreational use of waters Population concentration in urban areas/ depopulation in rural areas Increased production of non-food crops Increase in agrochemical applications

Industry Tourism Other

Increase in irrigated area Growth in metals, chemicals, construction and manufacturing industries Potential increase in tourists Population concentration in urban areas/ depopulation in rural areas

Water supply downstream Sediment loads Water supplies Water quality Water supplies Water quality Water supply Land use Water supply Soil quality Water quality Water supply Water quality Water supply Water supply Water quality Water supply

40

Chapter 2.2

2.2.4.1.2

Management Structure at the River Basin Level

Overall responsibility for water resources in Spain belongs to the Ministry for the Environment. Individual river basins are controlled by the relevant river basin authority, the Confederacion Hidrograﬁca del Ebro, created in 1926 (website http://www.chebro.es/). Under the Water Law dated 1985, the responsibilities of the local water authorities are as follows: Planning: – implementation, monitoring and revision of the Hydrological Plan for the basin as appropriate. Management: – administration and control of public works and water supply, – administration and control of uses which are in the general interest or which aﬀect more than one Communinade Automona. Investment: – planning and construction of works appropriate to the organisation and approved by the state.

2.2.4.1.3

Problems Encountered in the Management of Groundwater Resources

The key issues in this river basin are the following. Water scarcity. The quantity of water available is insuﬃcient to meet the agricultural, industrial, commercial and domestic needs of the population within the basin. Many irrigation facilities in the basin are old and ineﬃcient, but the cost of improving them would not be oﬀset by increased yields and reduced water demand. The construction of large dams to retain limited water resources has altered the hydrological regime of the basin dramatically, which has resulted in many ecological impacts. Salinisation. The abstraction of ground and surface water and widespread irrigation has increased the deterioration of water and soil quality. High levels of salts in both the soil and underlying geology have resulted in high levels of salts in the available water supply. This is compounded by groundwater abstraction, which concentrates salts and increases invasion of aquifers by marine water. Water, soil and sediment quality. The surface waters of the Ebro basin are of poor quality due to high levels of industrial and agricultural chemicals from both point and diﬀuse sources. High levels of mercury and DDT are a particular problem, as are levels of nitrates in soil and groundwater. Soil erosion. Intense agriculture on saline soils leads to a deterioration in soil structure and quality, which in turn leaves the soil more susceptible to erosion. The areas most susceptible to soil erosion in the basin are in the Catalan Chain to the south of Fraga, Lleida, Alcaniz and Caspe.

Transferring Scientiﬁc Knowledge to Societal Use

41

Ecological issues. The ecology of the delta is under pressure from two key areas. First is the poor water quality as described above, which leads to a deterioration of habitats, with consequences for the wildlife which depends upon it for survival. Second are the reductions in water ﬂow and sediment transport caused by dam construction, which results in erosion of the Ebro delta, a nature reserve of European importance. This is of less importance for groundwater issues.

2.2.4.1.4

Main Contributions of AQUATERRA in the Ebro Basin

A workshop managed by AQUATERRA was held on 7 April 2006 in Barcelona with stakeholders from representatives of unions and water authorities, through academics, industry representatives and students, agricultural community and water distribution companies involved in the Ebro river basin. Aims of the workshop were (i) to discuss the key environmental issues aﬀecting the water–sediment–soil–groundwater system in the Ebro river basin; (ii) to discuss the evolution of these issues in the context of the DPSIR model used throughout AQUATERRA, with a view to establishing a suitable area for an economic case study; (iii) to inform stakeholders from the Ebro river basin about the work being undertaken by AQUATERRA within the basin area and to engage stakeholders in the work of AQUATERRA; and (iv) to interest them in further activities aimed at interaction with the project. Having identiﬁed the key issues for the basin, the concept of DPSIR was thus applied to assess what the drivers, pressures, impacts and responses to key environmental issues were and suggesting how they might change in the future. For the drivers (e.g. the factors of change), the main conﬁrmed ones were the WFD and other water-related policies, the planning of water management by the Confederation Hidrogra´ﬁca del Ebro and the demographic developments within and beyond the basin, especially on the coast, including the concentration of the population in the urban areas and the development of industry in zones with a limited water supply. For the pressures, agriculture is the main one within the basin, as it aﬀects both salinity and water quantity and quality problems. Some isolated pressures from local industries are also signiﬁcant in terms of water quality. A less important, but nonetheless signiﬁcant pressure is exerted by large urban areas and tourist zones on the coast, both within and beyond the basin due to water transfers. In this basin, a large range of contaminants are encountered and should be monitored at diﬀerent scales. For the impacts, those on ﬂora and fauna and ecology are the most signiﬁcant. For the responses, e.g. the measures taken or to be taken in order to respond to impacts, the reduction of water consumption in general would be the only eﬃcient, long-term, sustainable solution to the current situation. The activity most directly and immediately aﬀected by this change would be agriculture; industry and urban areas would also be aﬀected by water economy. Within AQUATERRA, the sub-project MONITOR aims to provide, develop and validate analytical tools necessary for the monitoring of organic and

42

Chapter 2.2

inorganic pollutants in water/sediment/soil compartments with the ﬁnal objective to identify those present at relevant concentrations. In the Ebro basin, studies started on the occurrence and the distribution of contaminants in the river basin. Numerous compounds have been analysed (trace metals, pesticides, PAHs, BTEX, PCBs, etc.). High concentrations were detected in some areas describing situations where the environmental impact is high (i.e. with ﬂuoranthene, benzo[a]pyrene, benzo[b]ﬂuoranthene). But in the ﬁeld of environmental monitoring, the information of interest is spread over time and space. Therefore the choice of monitoring locations can be a key issue in terms of global costs. Optimisation of the sampling network is currently done in an eﬀective and economic way (by improving on time intervals for sample retrieval and analysis of selected compounds). The environmental assessment of the Ebro river basin will be improved by the coupled use of chemometrics modelling of monitoring data, various sampler devices, Geographical Information System (GIS) and remote sensing information systems.

2.2.4.1.5

Problems to be Solved in the Future

The identiﬁcation and validation of major issues within the Ebro basin as a whole lead to the problems that should be solved in the future in the context of progress resulting from knowledge. One of the most important problems to be solved is that of water abstraction. A reduction of water consumption in agriculture would have positive consequence on both the quantity of water consumed and soil quality, as an excess of stagnant water left on ﬁelds after irrigation increases soil salinity through water evaporation. For agriculture the improvement of the water eﬃciency of irrigation systems should be urgently solved. One solution would be the reuse of aﬀected groundwater. Reuse of water would also improve the perception of agriculture by the population in general. A further alternative solution to the lack of water is desalinisation of seawater. Although this was only considered as a temporary solution, it was nevertheless considered preferable to water transfers from the Ebro to southern regions outside the basin. In the case of very acute localised problems, the compulsory purchase of agricultural land in order to avoid cultivation has been suggested, although this would clearly have a signiﬁcant economic impact for those directly aﬀected and is not a widely supported proposal for the long term. Changing to less water-intensive crops could also be undertaken, potentially at a large scale, although the economic viability of producing a diﬀerent crop must be considered. At a very large scale this measure should be carefully evaluated and would only be possible under technical and market conditions.

2.2.4.1.6

The Lessons Learned

Discussions have been conducted with the major stakeholders with a common agreement on the fact that drinking water is the priority use in the basin and that future planning should aim to maintain the quantity and quality of the

Transferring Scientiﬁc Knowledge to Societal Use

43

supply. More general problems arising in the management of the basin are (i) the fact that water is free or not charged at it true value, (ii) the implementation of water management policy is not optimal across the basin, (iii) the extent of irrigated areas is continuously increasing requiring an evaluation of the impact of these new areas before improving the irrigation systems of existing areas (to be done at the river basin level to be eﬃcient) and (iv) there is an imbalance in the quality of water treatment between that which is treated and exported to users outside the basin and that which is produced for use within the basin; consequently there is less investment in water within the basin. AQUATERRA and future research projects should investigate some of these areas in the future, like deﬁning decision-making tools for solving identiﬁed speciﬁc important problems relevant to the river–sediment–soil–groundwater system and socioeconomic analysis of the impacts and responses (like in the Coordination Action Risk-Base of the 6th FP). These tools could be relevant for a policy directive like the WFD. But in many ways, the River Ebro is unique, with its speciﬁc management administration system, its combination of distinct climatic issues and some speciﬁc problems (such as salinisation). Therefore the Ebro does not easily ﬁt into a standard ‘‘model’’ of a European catchment.

2.2.4.2 2.2.4.2.1

The Meuse Case Study Context

The Meuse river basin is one of the smallest international districts in Europe. The Meuse has its source in the Langres Plateau in France and ﬂows from Belgium and the Netherlands to the North Sea. A part of the Meuse river basin belongs to Germany and Luxembourg even though the river itself does not cross these countries. The Meuse is about 950 km long and its basin represents between 30 000 and 35 500 km2 according to the various information sources (Figure 2.2.4). The rivers of the Meuse basin are mainly plain rivers, characterised by broad valleys and weak medium slopes. Belgium is a central country for the Meuse management since it contains 41% of the total area of Meuse basin (5% in Flanders and 36% in Wallonia) and the basin represent 46% of the whole country area (Figure 2.2.4). The remaining area belongs to the Scheldt river basin. France is the second country, representing 26% of the Meuse river basin (9% in Champagne-Ardennes region, 3% in Nord-Pas-de-Calais and 14% in Lorraine region) whereas the Meuse basin only represents less than 1% of French territory. One-ﬁfth of the basin is located in the southern part of the Netherlands (14% in NoordBrabant region, 6% in Limburg and less than 1% in Zuid-Holland) and the Meuse basin represents 22% of the country. The German Nordrhein-Westfalen (NRW) land covers 11% of the basin (6% of Du¨sseldorf region and 5% of Ko¨ln) and the Meuse basin represents only 1% of the German area. Around 5 to 10% of the Luxembourg area is also included in the basin but it represents less than 300 km2.

44

Figure 2.2.4

Chapter 2.2

Meuse basin (source: IMC11,12).

The Meuse river basin can be characterised by a low runoﬀ coeﬃcient. Rivers are dominated by a rainfall–evaporation regime and the annual runoﬀ is generally about 258 mm per year in the River Meuse. The long-term average discharge is 250 m3 s1 and total volume per year is 7.6 km3 on average. However, since the Meuse is a rain-fed river, the maximum ﬂow can be as

Transferring Scientiﬁc Knowledge to Societal Use

45

high as 3000 m3 s1 in the winter and the minimum ﬂow as low as 10 m3 s1 in the summer. The main land uses in the Meuse river basin are distributed as shown in Figure 2.2.5. The most industrialised regions and those having the most intensive agriculture are also those having the highest population density. A summary of the present and potential future drivers and pressures acting upon diﬀerent economic areas in the Meuse basin (agriculture, forestry, industry and urban development) is shown in Table 2.2.4.

2.2.4.2.2

Management Structure at the River Basin Level

In the 15th and 16th centuries, the present Dutch and Belgian territories and some parts of France formed one country, the 17 Dutch provinces. But until the independence of Belgium in 1839, both countries began to formulate their own policies, particularly about transport, industrial and agricultural activities in the Scheldt and Meuse basin.10 Diverging interests arise concerning waterways and sharing availability of water (quality, quantity) for drinking water production. In the 20th century there were several attempts to have an international agreement concerning the Meuse basin and ﬁnally an agreement was signed in 1994. Thus, collaboration between the basin states of the Meuse is in its infancy compared to that of the Rhine.11 The main agreements concerning the Meuse basin signed to date are: 1992 1994

1995

2002

w

Agreement about the protection and use of transboundary watercourses and international lakes signed 17 March 1992. Agreement on the protection of the Meuse (France, the Walloon region, the Flemish region, the Brussels Region and the Netherlands) signed 26 April 1994 in Charleville-Me´zie`res. Meuse Discharge Treaty (Netherlands and the Flanders region). Equal sharing of water by both partners during low-water periods and a common responsibility for the Border Meuse.w Creation of the International Commission for the Protection of the Meuse (ICPM). The agreement on the ﬂood protection action plan for the Meuse. The ministers of the riparian states declared in Arles (France) the necessity to reduce the ﬂood risk and harmonise ﬂood-reducing measures. They particularly stressed the harmonisation of spatial planning, land use and water management in the river basin. International agreement on the Meuse (Germany, Belgium, the Brussels, Walloon and Flanders regions, France, Luxembourg and the Netherlands) signed 3 December 2002 in Gand.

For further information about the Border Meuse management, see ‘‘River management and low ﬂows in the river Meuse in the Netherlands’’.

46

Chapter 2.2

59% 28% 11%

1%

Figure 2.2.5

Table 2.2.4

Urban areas

Agricultural areas

Others

Water areas

1%

Forest areas

Land use in the Meuse basin (adapted from9).

Summary of main drivers and pressures in the Meuse basin.

Present Agriculture

Forestry Industry

Urban development

Others

Future Agriculture Industry Urban development Other

Drivers

Pressures

Increased productivity: cereals, maize and meadows Intensive farming systems like horticulture, glasshouse cultivation and breeding Increased sylviculture activities Increased industrial discharges Accidental and hazardous pollution Increased water abstraction volumes

Water supplies Water quality Soil quality

Energy, mining, chemistry, metals, construction and agro-food industries Population concentration in urban areas Increased water abstraction Water sanitation (discharge of wastewater) Fluvial transport (numerous navigable rivers and canals) Leisure and tourism versus discharge of wastewater Increased of agrochemical applications Increased in glasshouse cultivation Growth in some of the main industrial activities Population growth and concentrations, increased density Increased leisure demands

Water quality Sediment loads Water supply downstream Water supplies Water quality Water supplies Water supply Water quality Water quantity Water quantity Water quality Soil quality Water quality Water quantity Soil quality Water quantity and quality Water Supply Water sanitation Water quantity and quality

Transferring Scientiﬁc Knowledge to Societal Use

47

Actually, the International Meuse Commission (IMC) plays an important role in international consultations. Its purpose is to obtain sustainable integrated water management in the international basin of the Meuse. From its creation, the IMC has established three action programmes: the preliminary programme (1995–1997), the short-term programme (1998–2003) and the longterm programme (2003–2010). The important targets of the Meuse short-term programme included:12

reduction of urban and industrial discharges; reduction of diﬀuse pollution; prevention form accidental and hazardous pollution; monitoring water quality; sediment management; ecological improvement of water quality; and knowledge exchange between countries.

The present action program (2003–2010) also includes ﬂood prevention and protection and has a main target to contribute to the implementation of the WFD.

2.2.4.2.3

Problems Encountered in the Management of Groundwater Resources

The key issues in this river basin are the following. Water scarcity. The quantity of water available could be insuﬃcient to meet the needs of the diﬀerent activities within the basin (agriculture, industries and domestic uses). In some areas, water exploitation is concentrated in speciﬁc surface or ground waterbodies, causing large problems for water availability and localised environmental problems to the water ecosystems. A signiﬁcant example is the Tournaisis aquifer (Hainaut, Wallonia), where water is extracted at rate higher than the aquifer recharge capability. The Netherlands is also strongly dependent on the surface water of the Meuse for its drinking water production. Groundwater quality. Even if the groundwater system does not have a homogeneous network of measures all along the Meuse river basin, its quality seems to become degraded from the upper to the lower stretch of the river basin, in particular with nitrates and pesticides. Other contaminants are also brought by point-source pollution from industry (hydrocarbon fuels, solvents, but also chloride, iron bore and heavy metals from mines in France). Bad quality of the past mining reservoirs may aﬀect the groundwater quality of the Meuse Dogger aquifer: – by the rising of saline water from the mines reservoirs towards the bottom of the aquifer, and/or – by the inﬁltration of surface water contaminated by the waters out ﬂowing from the ﬂooded mines.

48

Chapter 2.2

The hydromorphology modiﬁcation of the Meuse. Since 1883, the Meuse has been largely managed by humans to support navigation, economic development and ﬂood protection. The major interventions involved the building of river control structures (weirs, dams and sluices, particularly in Walloon and Dutch parts) and banks (training walls, quays, etc.) and modifying the river bed by regular dredging of sediment. Such modiﬁcations to river imposed changes to the morphology (river fragmentation) and to the hydrological regime which have an impact on other factors including: – Flooding: the extent of zone liable to extreme ﬂooding (over the protection designed) is limited and hence the water volume that may be naturally stored in ﬂoodplains is also limited. – Sedimentation and transport: when the level discharge is low, the sedimentation transport is greatly reduced behind weirs. As the discharge increases, the deposited sediments are carried again and the volume load is greater than without weirs. According to IMC,11 the amount of sediments transported have tripled because of the hydromorphological modiﬁcation. – Ecological function: as a consequence, biodiversity became poor in certain stretches and migration routes for ﬁsh have been disturbed. Diﬀuse pollution. Agriculture, industrial activities and household discharges all contribute to diﬀuse pollution on soil and water bodies. The main pollutants are nitrates, fertilisers, manure and pesticides (like diuron and atrazine) which may induce the eutrophication of the river and thus that of the North Sea. The total concentration of diﬀuse pollutant in the water system is very important from a drinking water perspective as the River Meuse is a source of drinking water for more than 6 million inhabitants. Some new dangerous chemicals like endocrine disruptors, and new medicines from pharmaceutical industry have recently appeared. When looking at eﬀects on water quality, all pollutants should be considered together for understanding their impacts on the ecology of the system.

2.2.4.2.4

Main Contributions of AQUATERRA in the Meuse Basin

The main study objects in the Meuse basin are: (i) contaminated ﬂoodplain sediments; (ii) soil–sediment–groundwater–river interaction in the catchment of the Dommel (a tributary to the Meuse); (iii) groundwater quality eﬀects by river–groundwater interaction in the Belgian part of the Meuse system, more speciﬁcally near Lie`ge (Fle´malle cookery site) and in the Geer basin; (iv) quantiﬁcation of coupled ecotoxicological eﬀects of contaminants from sediment, suspended solids and freshwater on aquatic organisms at a number of Walloon locations; and (v) socioeconomic aspects of global change. This needs

Transferring Scientiﬁc Knowledge to Societal Use

49

integration of multiple disciplines, from geosciences, environmental engineering and chemistry to socioeconomic sciences, from the catchment to the regional scale, as is done in AQUATERRA. The main AQUATERRA contributions in the Meuse presented during the 2nd International Meuse Symposium in Sedan (18–19 May 2006) organised under the auspices of the IMC13 (http://www.meuse-maas.be/) concerned: metal speciation and bioaccumulation in ﬂoodplain soil and sediment; nitrate trends in the chalk aquifer of the Geer sub-basin in Belgium; breakdown of leaves as an indicator for sediment quality and ecosystem functioning; ecotoxicological eﬀects of contaminants on aquatic organisms in the River Meuse; metal availability in the Dommel sub-catchment; and measurement of metal concentrations (Cd, Zn, Cu, Fe, Ni) in soil and earthworms in the Dommel sub-catchment. In parallel with the work done in BASIN, the sub-project FLUX investigates the temporal and spatial trends of suspended matter pollution in the Meuse River. Spatial and temporal diﬀerences in suspended matter (SPM) quality in the Dutch part of the Meuse can be attributed to point sources contribution, variation in river discharge and sedimentation processes. The large ﬂuctuations in metal content during summer indicate an irregular contribution of point sources from the upstream area, most likely from the Liege industries; to control the high pollution levels of suspended matter, this contribution from Belgian industrial areas should be reduced. The spots of contaminated suspended matter deposition in the summer should be traced back to reduce the risks of higher contamination levels on ﬂoodplain areas by re-suspension of contaminated river bed sediment and possible long-term exposure of contaminated bed sediment in the river environment during dry periods.

2.2.4.2.5

Problems to be Solved in the Future

Among the major issues for the Meuse River, the sanitation in the Walloon part of the Meuse basin is one of the most important. Although 95% of the population is connected to a collective sewage system, collective treatment capacity (1.1 million IE) only represents 37% of the urban and industrial wastewater volume (respectively 2 and 1 million IE) that had to be treated in 2002. Figure 2.2.6 shows that numbers of treatment plants are under construction, granted or under study, but 18% remain non-existent. Moreover the whole Walloon stretch is considered as a ‘‘sensitive area’’ under the Council Directive 91/271/EECz: all treatment plants with capacity exceeding 10 000 IE have to ensure a treatment against phosphorus and nitrates (tertiary treatment). z

Council Directive 91/271/EEC of 21 May 1991 concerning urban wastewater treatment.

50

Chapter 2.2 non existent plants 18% plants in study 9%

existing plants 37%

projected plants 1% granted plants 24%

Figure 2.2.6

plants in construction 11%

Treatment plants in the Walloon stretch in 2002: distribution of IE to be treated.

Treatment capacity rates vary within the Walloon stretch from 13% in the ‘‘Meuse aval’’ sub-basin to 73% in the ‘‘Vesdre’’ sub-basin.

2.2.4.2.6

The Lessons Learned

Discussions have been conducted with the major stakeholders of the Meuse river basin and have shown that building a global scenario for managing the basin at the river basin scale remains a challenge for several reasons: The stakeholders have limited knowledge and information on possible future changes in the key drivers and their trends. The identiﬁed drivers, the economic activities, are really comprehensive and need to be disaggregated to be more issue-speciﬁc (e.g. for agriculture, irrigation drainage, fertilisers and pesticide uses) to be suﬃciently analysed and used for decision-making. The complexity of the management system needs to be simpliﬁed by developing accurate indicators based on basic simple ﬁgures, still to be identiﬁed. There are diﬀerent decision-making levels (local, regional, national and international in the case of the Meuse river basin) and the important issues at one of these levels are not necessarily viewed as important at another, in particular at the river basin scale. Stakeholders concerned by environmental problems have a perception coming from societal pressures that diﬀers from the scientiﬁcally based approach based on detailed factual information. Even the time horizon taken into consideration by both is diﬀerent, being shorter for the stakeholder issues. Developing scenarios for the sediment–soil–water system is very complex, with multiple possible entries: (i) system components (soil, sediment, groundwater and surface water); (ii) drivers and factors of change (climate, economic and policy changes); (iii) spatial scale of analysis (local/small area, regional, national, sub-basin and basin levels);

Transferring Scientiﬁc Knowledge to Societal Use

51

(iv) temporal scale (short, medium and long term); (v) potential issues at stake (organic pollution, diﬀuse pollution by nitrate, sediment contamination, water pollution by heavy metals, etc.). Therefore, deﬁning longterm scenarios for some speciﬁc issue at case study scale should be more appropriate than global management scenarios.

2.2.5

Discussions of the Contributions of AQUATERRA to Societal Use Related to Groundwater Resources Management

As shown in the case studies discussed above, human activities have greatly modiﬁed the water resources systems through climate change, land use changes, water engineering and releases of wastes and pollutants in the aquatic systems (Table 2.2.5). Nevertheless, aquatic systems have also high natural spatial heterogeneity. Therefore, understanding and modelling the degree of heterogeneity of water systems will be diﬀerent and will depend on the scale of interest (Table 2.2.6). The sediment–soil–water issues have to be addressed at diﬀerent levels, i.e. local, regional, national or international. An important issue at the regional/ national level is not necessarily viewed as an important issue at the river basin scale. The response and management concerns may also vary between the diﬀerent levels, especially for a transboundary river basins such as that of the Meuse. This complexity of the studied system induces some key elements for the new modern ways of knowledge production. Knowledge production should be socially accountable and should acknowledge the multiple rationalities and diﬀerent viewpoints that are brought in by the variety of stakeholders that are involved. This means that the research methods should contain the following key elements. Multi- and/or transdisciplinary research methods – Acknowledge the disciplines that should be involved in the research to give more insight in the policy issue, the problem framing and the solutions. – Work together in multidisciplinary teams and emphasise the process of interaction between the disciplines and evaluate these processes. – Take time to develop new theoretical multidisciplinary scientiﬁc frameworks. Involvement of the stakeholders in the research process – Develop methods to involve stakeholders and to deal with their viewpoints and values in the research process. – Appreciation of the values, interests and viewpoints of involved stakeholders ask for development of new research methods to develop

x x x x

x

x X

x x

x

x

A

X x x x

x x x

x

x

x x

x

x X X x

x x

x

X

x x x

x

X

C

x

B

x

x

x x

x

x

x

X X x

X

D

x

x

x

x

x

x x x x

E

A: human health; B: hydrological cycle balance; C: water quality; D: global carbon balance; E: ﬂuvial morphology; F: aquatic biodiversity; G: coastal zone impacts.

6. Irrigation / water transfer

5. Urban waters

4. Industrialisation and mining

3. River damming and channelling

2. Land use changes

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

1. Climate variability and changes

development of non-perennial rivers segmentation of river networks changes in ﬂow regimes development of extreme ﬂow events changes in wetlands distribution/function changes in chemical weathering changes in soil erosion salt water intrusion in coastal groundwater salinisation through evaporation wetland ﬁlling or draining changes in water pathways changes in sediment transport urbanisation alteration of ﬁrst-order streams nitrate and phosphate increase pesticides increase nutrient and carbon retention retention of particulates loss of longitudinal and lateral connectivity creation of new wetlands increases of heavy metals and POPs Acidiﬁcation of surface waters salinisation sediment sources nitrate and phosphate increase enhancement of waterborne diseases organic pollution heavy metals and POPs increase partial to complete decrease of river ﬂuxes salinisation (evaporation and percolation)

Local to regional changes of environmental states

Global impacts

x

x

X

X x

X x

x x

x x

x x x x x

F

Major global pressures on continental aquatic systems and the mapping of local- to regional-scale impacts.1

Pressures

Table 2.2.5

x x

x x

X

x x x x x X x

x x x x x

x

G

52 Chapter 2.2

53

Transferring Scientiﬁc Knowledge to Societal Use

Table 2.2.6 Scale level Plot scale

Heterogeneity of the water systems, natural and anthropogenic issues according to the scale level. Natural issues

distribution of root sys

Small catchment scale

tems lateral transfer of water

groundwater aquifer upstream–downstream ﬂow structure

biogeochemistry parameters of environment compartments

Anthropogenic issues

land cover changes diﬀuse atmospheric pollution

land

River basin scale

river ﬂow regimes, links

between groundwater and surface waters sediment supply and transfer water quality linked to climate

cover and use changes land management micro-climate changes artiﬁcial modiﬁcations of water bodies diﬀuse and local pollutions

water uses and demands impacts of urbanisation impacts of human activi-

ties such as agriculture, industries, mining

and share knowledge. The instrument of joint fact ﬁnding14 is already available, but it should be developed further. – Testing of scientiﬁc results with stakeholder panels is another direction of development.

Emphasis on learning in policy – Learning between stakeholders, between scientists, between stakeholders and scientists, etc., and on diﬀerent levels (individual, team and organisation) have a central role in the fourth generation and deserve special arrangements. – Organisation of reﬂection, feedback and evaluation of the goal achievement of the policy is needed.

As ground water policy is getting involved in a much broader policy ﬁeld, one can expect that these trends will also relate to groundwater policy. In eﬀect, the AQUATERRA research project is an example of the above mentioned trend of multidisciplinary research. As stakeholder involvement and policy learning are other policy trends, the next sections will go into these topics.

54

Chapter 2.2

2.2.5.1

Stakeholder Involvementy

For stakeholder involvement with respect to groundwater policy there are several arguments. Apart from the basic fact that stakeholders have an impact on the quality and quantity of groundwater, the main arguments can be grouped into three themes: obstructive power, enrichment and fairness. The early involvement of stakeholders reduces the risk of the use of obstructive power by them, and thus the policy not being carried out. On the other hand, stakeholders possess resources, like money or knowledge required for the design planning and implementation of sophisticated policies, which governments do not possess. Knowledge is distributed among several stakeholders and government. So stakeholders should be involved to get all the (pieces of the puzzle) together. Stakeholder involvement can provide good ecological practices in this way. The last argument for stakeholder involvement is fairness. It is fair to involve actors aﬀected by a certain policy, and give them a say in the decision-making process. But who are the stakeholders and how can one involve them? Derived from Slob et al.,15 we can make this deﬁnition: (stakeholders are all those people or organisations who have an eﬀect on or are aﬀected by groundwater policies.) We can make a distinction between two types of stakeholders: Organisations and people that have a direct impact on groundwater quality or are directly aﬀected by the relevant policies. This group includes: industries using groundwater, farmers, water authorities, regulators on the local, regional, national and international level and citizens that are directly aﬀected by the measures planned or taken. Organisations and people that have an impact on the relevant decisionmaking. This group covers citizens, landowners, homeowners, insurance companies, NGOs such as Greenpeace and the WWF, scientists and drinking water companies. A process of stakeholder involvement requires an independent chairperson or process manager. The ﬁrst step for the process manager in the organisation of stakeholder involvement is to ﬁnd out which stakeholders should be involved. Next, it is vital to collect information about the goals, ambitions and problem deﬁnitions (from the various perspectives) of the stakeholders. The process manager should ensure that all these interests are heard and acknowledged in the course of the process. The mobilisation of the stakeholders is an important issue. It is the duty of the manager to let stakeholders realise what the beneﬁts are. Why should they join the process? A sound and deliberate consideration of interests might persuade less interested parties to join and will be a signal to dominant forces not to overreact. To create more certainty into the process, the process requires ‘‘rules of the game,’’ that contain rules for entering the process in later stages, y

This section is an excerpt of a chapter about stakeholder involvement in one of the (SedNet) books.15

Transferring Scientiﬁc Knowledge to Societal Use

55

how decisions are made, how information is brought into the process, etc. These ‘‘rules of the game’’ should be discussed and should be approved by the involved stakeholders. The process of involvement can be arranged with diﬀerent goals: information: providing information to the stakeholders; consultation: ascertaining what stakeholders think must be done; advising: letting stakeholders advise on the policy and taking their recommendations into account; co-producing: stakeholders are regarded as equal policy-makers but decision-making remains in the political domain; and co-deciding: decision-making power is handed over to stakeholders. Every situation is unique and therefore the level should be chosen that ﬁts the speciﬁc situation.

2.2.5.2

Learning Approaches in Policy-making

With the whimsicality of the fourth generation of environmental policies, the policy approach that ﬁts best for the (local) groundwater issue is not ‘‘standard’’ and not known beforehand. This means that the policy approach required should be developed with the involved stakeholders speciﬁc for the local groundwater issue, and should be ‘‘tailor made.’’ To do this eﬀectively an adaptive approach should be followed, in which actions are deliberated with the stakeholders, and will be followed up by their implementation. Then, to see the results of the actions, it is needed to follow quite closely how the groundwater system responds to these actions with all stakeholders involved. If the actions do not give the desired results, new or extra actions should be deliberated and undertaken. In this manner the way the system responds will be followed and the understanding of the system will grow. Continuously monitoring of the groundwater system plays a key role in sustaining the ‘‘learning approach’’ of groundwater policy and to ﬁnd the best local approach to the groundwater issue. A summary of this learning approach could be as follows. Discussion on the groundwater issue with involved stakeholders, the problem (for instance with the DPSIR model) and the shared ambitions to solve the problem. Discussions on the actions that should be undertaken to reach the ambitions. Monitoring of the results and the way the system responds to the actions. Discussion on the results: does the system respond to the actions as was foreseen? Are new or extra actions needed? And so on. Implementation of these (extra) actions and monitoring of these results. Then the cycle starts all over again. In this way actions will be found that match best with the articulated ambitions.

56

2.2.6

Chapter 2.2

Conclusions

In this chapter, we present the preliminary results of the Integrated Project AQUATERRA in the framework of policy and societal use. AQUATERRA ﬁts fully with part of the general conclusions given at the end of the World Science Forum16 (http://www.sciforum.hu/index.php). For example, conclusion 1 stated that ‘‘Scientiﬁc research is having a more immediate societal inﬂuence and facing an increasing set of requirements on the part of the public. As a result of internal scientiﬁc development and societal need, new research priorities emerge, requiring the cooperation of various disciplines. Therefore an integration of the natural and social sciences is taking place for problem-solving. Such an integration reinforces the need for establishing interdisciplinary frameworks. This is to be reﬂected in the institutional structures of science and of science policy as well.’’ This was clearly evidenced throughout this chapter. In the not too distant future, the main developments in AQUATERRA will certainly agree with another conclusion of the World Science Forum (2003) that stated that ‘‘the development of science and the demands of society will remove the rigid boundaries between theoretical and applied research, between the academic and innovation sector, . . . and scientiﬁc communities shall communicate the achievements of science and shall support decision-making . . . ,’’ especially within the COMPUTE sub-project (by setting a modelling tool box to assess impacts on water quantity and quality at diﬀerent investigation scales: bench, catchment, river basins) and INTEGRATOR sub-project (by elaborating operational tools for the stakeholders, including a ﬁrst assessment of social and economic impacts of policies in a scientiﬁc basis decision-making system). Finally, we showed that such an Integrated Project is in full agreement with the most important conclusion of the World Science Forum16 stating that ‘‘There exist appropriate scientiﬁc guidelines for improving the quality of life. Solutions oﬀered by science are generally delayed by socio-economic patterns and inappropriate information transfer. It is the common responsibility of politicians, scientists and decision-makers to ensure the proper implementation of knowledge to improve the quality of life.’’

References 1. M. Meybeck, Hydrol. Proc., 2005, 19, 331. 2. M. H. Gerzabek, D. Barcelo´, A. Bellin, H. H. M. Rijnaarts, A. Slob, D. Darmendrail, H. J. Fowler, Ph. Negrel, E. Frank, P. Grathwohl, D. Kuntz and J. A. C. Barth, J. Environ. Manag., 2007, 84, 237–243. 3. F. Boons, L. Baas, J. J. Bouma, A. de Groene and K. LeBlansch, The Changing Nature of Business, International Books, Utrecht, The Netherlands, 2000. 4. H. Spliethof and J. van der Kolk, Environmental management: developments and interactions between technology and organisation, in Technology and Environmental Policy, [Technologie en milieubeheer],

Transferring Scientiﬁc Knowledge to Societal Use

5. 6. 7.

8. 9.

10. 11. 12.

13.

14.

15.

16.

57

ed. A. P. J. Mol, G. Spaargaren and A. Klapwijk, SDU, The Hague, 1991 [in Dutch]. G. Keijzers, J. Clean. Prod., 2000, 8, 179. L. Simons, A. Slob, H. Holswilder and A. Tukker, Environ. Qual. Manag., 2001, 11, 51. M. Gibbons, C. Limoges, H. Nowotny, S. Schwartzman, P. Scott and M. Trow, The New Production of Knowledge, Sage Publications, London, 1994. H. Nowotny, P. Scott and M. Gibbons, Minerva, 2003, 41, 179–194. N. Geilen, B. Pedroli, K. Van Looy, L. Krebs, H. Jochems, S. Van Rooij and T. Van Der Sluis, Final Report of the IRMA/SPONGE Project, no 9, 2001. R. J. Batalla, C. M. Gomez and G. M. Kondolf, J. Hydrol., 2004, 290, 117–136. IMC, Proceedings of the First International Scientiﬁc Symposium on the River Meuse, Maastricht, The Netherlands, 2002. IMC, Rapport interme´diaire du programme d’action ‘‘Meuse’’: Mise en oeuvre de la 1e`re phase et pre´paration de la 2e`me phase. Commission Internationale de la Meuse, Lie`ge, 2001. J. Joziasse, J. Vink, D. Slijkerman, E. Foekema, J. Brils, J. Batlle Aguilar, P. Orban, S. Brouyere, A. Poot, E. Bleeker, C. van der Wielen and M. He´mart, AquaTerra research activities in the Meuse river basin, 2nd International Meuse Symposium, Sedan, 18–19 May 2006. J. R. Ehrmann and B. L. Stinson, Joint fact-ﬁnding and the use of technical expertise, in: The Consensus Building Handbook, ed. L. Susskind, S. McKearnan and J. Thomas Larmer, Sage Publications, London, 1999. A.F.L. Slob, L. Gerrits and G. J. Ellen, Sediment management and stakeholder involvement, in Sediment Management on the River Basin Scale, ed. Ph. Owens, Elsevier, in press. World Science Forum, Budapest, 8–10 November 2003; available at http:// www.sciforum.hu/index.php2003.

CHAPTER 2.3

Groundwater Management and Planning: How Can Economics Help? JEAN-DANIEL RINAUDOa AND PIERRE STROSSERb a

Bureau de Recherches Ge´ologiques et Minie`res (BRGM), 1034 rue de Pinville, FR-34000 Montpellier, France; b ACTeon, Le Chalimont, BP Ferme du Pre´ du Bois, FR-68370 Orbey, France

2.3.1

Introduction

Public water managers and stakeholders involved in the design of groundwater management are increasingly aware of, and concerned by, the economic implications of the technical choices they promote. The deﬁnition of sustainable groundwater quality objectives as well as the identiﬁcation of technical actions needed to achieve these objectives require investigating trade-oﬀs between economic and environmental considerations. It also requires balancing interests between economic sectors through proportioning constraints imposed to sectors generating pollution or depletion with costs incurred by sectors depending on groundwater quality. Although the economic (and political) dimension of groundwater management might be taken into consideration implicitly by water planners in the large majority of European countries, it is rarely explicitly considered and quantitatively assessed. The promulgation in 2000 of the European Union (EU) Water Framework Directive (WFD) has drastically improved this situation by clearly integrating economics into water management and policy. This directive promotes the application of economic principles such as the polluter pays principle. It recommends applying economic methods such as cost-eﬀectiveness analysis to support the identiﬁcation of measures to achieve the environmental objectives of the directive. And it calls for a wider consideration of economic instruments (e.g. water pricing, charges and taxes) to provide adequate (ﬁnancial) incentives for reducing pressures exerted on water resources.

58

Groundwater Management and Planning: How Can Economics Help?

59

Clearly, the recognition of the need for more systematic and robust economic assessments to support water management programmes is not the sole result of the changing regulatory context. Indeed, it reﬂects an increasing demand by actors and economic interests for more robust economic justiﬁcations to water protection decisions, choices and orientations. More attention is given to the eﬃcient use of limited ﬁnancial resources allocated to environmental protection. And the assessment of the economic impacts of constraints imposed on economic sectors by environmental regulation (e.g. restriction or ban of dangerous substances in industry) is increasingly called for: as it may result in competition distortion with non-EU producers and may generate losses of competitiveness, income and employment. In this context, the assessment of these costs, the proof that they are minimised and fairly distributed between sectors and, in certain cases, the demonstration that they generate proportionate social beneﬁts, are elements that can help policy makers to justify decisions and gain social acceptance. Groundwater planners can expect three distinct types of contributions from economists. The ﬁrst one is the description of the hydro-economic system to better capture the interactions between, and dynamics of, economic activities and groundwater resources. This involves the construction of indicators characterising the economic signiﬁcance of water uses (in terms of employment, turn over, added value, etc.). It also requires understanding plausible futures for economic activities and resulting pressures on groundwater resources. In some cases, this description might beneﬁt from the assessment of ﬁnancial ﬂows between economic sectors and water users, or the evaluation of the damage costs imposed on third parties because of the degradation of groundwater resources (pollution or depletion). The second type of contribution from economists consists in assessing and comparing the economic cost and/or beneﬁts of diﬀerent groundwater management options. The role of economics can either be to identify the least costly way to achieve a given environmental objective (cost-eﬀectiveness analysis) or to assess the net beneﬁt (i.e. total beneﬁts minus total costs) of alternative management options (cost–beneﬁt analysis). It is important to point out that the ranking between diﬀerent policy options based on economic criteria is only one element provided to support a policy decision, not the decision itself. The third type of contribution consists in the design of economic instruments such as prices, charges (abstraction/pollution) or taxes that might inﬂuence water users’ behaviour. The need to inﬂuence users’ behaviour for enhancing the sustainability of groundwater resources is increasingly recognised by water planners. And they call for the design and implementation of economic instruments aimed at regulating water demand or pollution. The main purpose of this chapter is to show how practical economics can help provide some light on these diﬀerent demands. In its ﬁrst section, the chapter demystiﬁes key economic principles, concepts and methods. It then presents four concrete applications of economic methods applied to groundwater management. It then discusses the potential role of economic instruments to enhance the sustainability of groundwater resources. The chapter concludes

60

Chapter 2.3

by discussing the needs for, and challenges in, integrating economics with technical expertise and knowledge, an element that is essential to the relevance of economics for policy-making in the ﬁeld of (ground-) water management. The conclusion also identiﬁes gaps between policy demand and the economics toolbox where further research is needed to enhance the potential eﬀectiveness of economics in supporting groundwater management and policy decisions.

2.3.2

Economic Methodologies and Tools: Four Possible Ways of Supporting Groundwater Management and Planning

Economics as a science, its underlying theoretical principles as well as its operational tools and methodologies are not always given due considerations in groundwater management discussions and processes. This section presents a short overview of the economic methods and tools that can be mobilised to support groundwater management and to help answer the groundwater policy demand brieﬂy presented above.

2.3.2.1

Economic Characterisation of Water Uses

Characterising the economic dimension of groundwater use has both static and dynamic dimensions. First, assessment can be made to identify how important groundwater use is today for the economy, e.g. what are the economic weights in terms of turnover, value added, employment, economic value, etc., that one can attach to diﬀerent abstractors and polluters of groundwater resources. Integrated with information on pressures and/or impacts on the state of the groundwater aquifer, it helps obtaining ﬁrst insights into the trade-oﬀs between economic development and groundwater protection—and into possible economic impacts that might occur as a result of stricter groundwater protection. The information can be used to justify how essential groundwater protection is for key economic sectors for which development is directly linked to groundwater quality—or because groundwater of adequate quality has a high economic value that justiﬁes its protection. Second, economics can help capturing the dynamics of the system and identifying future groundwater protection challenges. Indeed, static situations in terms of pressures and impacts on groundwater are rare. Changes in sector policies, the internal dynamics of economic sectors (inﬂuenced, for example, by global changes and world trade) or the implementation of current environmental legislation is expected to impact on economic development and thus on pressures imposed on groundwater resources. In some cases, improvements in groundwater quality might result from the implementation of existing environmental legislation or from changes in economic sector policies favouring less polluting input and production processes. In other cases, urban development and the opening of new markets for industrial products might lead to increased abstraction. And understanding the dynamics of groundwater

Groundwater Management and Planning: How Can Economics Help?

61

systems and connected users is essential in the search for solutions and measures aimed at improving the protection of groundwater resources. The characterisation reports prepared by EU member states in line with the requirements of Article 5 of the WFD provide very diverse illustrations of economic characterisation of water uses. Few of these reports, however, clearly link economic indicators to water use or to groundwater use.1 Investigating today’s situation and plausible socioeconomic futures for given groundwater systems requires collating economic information (often available in statistical systems complemented whenever necessary by surveys of citizens or economic operators), developing links between economic and technical (groundwaterrelated) information or applying foresight methods. Because of the scarcity of information directly available, it builds on exchange and interaction with stakeholders of the groundwater system considered.

2.3.2.2

Methods to Assess Environmental Costs

Recent studies have clearly shown that impacts on water quality and quantity of human activities are becoming signiﬁcant at the European level2,3 as well as elsewhere in the world.4 As a result of increasing nitrates and pesticides concentrations, chlorinated hydrocarbons contamination or salt water intrusion (to quote only selected problems), diﬀerent segments of society are incurring economic losses, referred to as damage costs hereafter. An assessment of the signiﬁcance of these damage costs—which could be avoided through appropriate action—is information policy-makers might use to justify the relevance of engaging into costly groundwater protection actions. Four diﬀerent types of damage costs can be distinguished: (i) cost of disease when the population is exposed to unsafe levels of substances in water; (ii) avoidance costs for water users who have to undertake averting or corrective actions (treatment of water, purchase of bottled water, etc.) in response to groundwater deterioration; (iii) ecological damage of surface ecosystems and subsequent loss of recreational value, when groundwater contamination has an impact on surface ecosystems (rivers, wetlands); and (iv) loss option value (possibility to use groundwater in the future) and non-use value (bequest value). Diﬀerent economic methods have been developed to assess these costs. The cost of illness method aims at assessing costs generated by exposure to contaminants in drinking water, e.g. the cost of foregone wages for victims of contamination and the cost of treating the illness.5,6 This approach may be relevant in cases where health eﬀects are the major cost component, as is the case with arsenic groundwater contamination in Bangladesh for instance.7 It does not seem relevant to pollution in Europe where concentrations of harmful substances found in drinking water very rarely reach a very high level (apart for some rural areas in Romania and Bulgaria, for example). The illness cost due to pollution with pesticide and some selected dangerous organic compounds could be signiﬁcant, but the lack of epidemiological studies does not allow any evaluation of these costs.

62

Chapter 2.3

The avoidance cost method, or averting expenditure approach, consists in assessing the costs of actions undertaken to prevent or mitigate the adverse eﬀects of contamination. Rooted in the household production function model, this approach assumes that consumption of goods or services can substitute for groundwater quality change. The implementation of this method oﬀers a means to generate lower-bound estimates of an important component of the cost of groundwater pollution, namely the use of groundwater as a drinking water source. The averting behaviours reported include purchasing bottled water, monitoring of private borehole water quality and installation of ﬁltering devices. Their estimated cost ranges between US$125 and US$330 per year and per household.8 The contingent valuation method aims at assessing the value that households assign to the preservation of groundwater quality. It builds on household surveys during which respondents have to state their willingness to pay (WTP) for hypothetical groundwater protection (or restoration) scenario. The stated WTP contains both use and non-use values, since households may want to preserve groundwater for their present or future consumption, for the consumption of future generations and/or for the resource as such. The estimated WTP for groundwater protection can be considered as an estimate of the cost of groundwater degradation. Contingent valuation is by far the most widely used method for assessing the beneﬁts of water protection. Household WTP values reported in the literature range from less than h20 to h550 per household per year.9,10

2.3.2.3

Economic Methods for Appraisal of Groundwater Projects and Policies: Cost-eﬀectiveness and Cost–Beneﬁt Analysis

Due to the large number of economic sectors having an impact on groundwater, groundwater overexploitation or pollution problems can almost systematically be solved using diﬀerent technical options or focusing on diﬀerent sectors generating pressures. Consider for instance an overexploited aquifer where total water abstraction exceeds groundwater recharge, resulting in a sustained drop of the water table. The hydrological balance can be restored by reducing water abstraction, through artiﬁcial rechargew or a combination of both. The reduction of water abstraction can be achieved through the development of alternative water resources (inter-basin transfers, creation of reservoir dams, desalination), improvement of irrigation eﬃciency, the development of recycling technologies in industry or wastewater re-use. To identify the cheapest way to restore the hydrological balance, diﬀerent combinations of measures can be compared in terms of their costs and their eﬀects: an approach referred to as ‘‘cost-eﬀectiveness analysis’’. A concrete example of cost eﬀectiveness w

Procedure in which water supplies, sometimes using treated waste water, are pooled over highly permeable aquifers for increased inﬁltration. In some applications, injection wells can be used to recharge water to deeper aquifer levels or into less pervious aquifers.

Groundwater Management and Planning: How Can Economics Help?

63

analysis of water scarcity management can be found elsewhere.11 An additional example of cost-eﬀectiveness analysis applied to a water pollution remediation is further detailed in Section 4 below. The problem can be even more complex when the diﬀerent technical groundwater management options envisaged do not lead to the same environmental outcome. This can be the case, for instance, when diﬀerent levels of contaminated site remediation are envisaged, or when the targeted pollutant concentration is not ﬁxed by law. In such situations, water planners may be interested to assess not only the costs of alternative groundwater remediation options but also their beneﬁts or net beneﬁts (i.e. the diﬀerence between total beneﬁts and total costs of actions). Cost–beneﬁt analysis thus provides a rational and systematic framework for identifying and assessing in monetary terms all positive and negative eﬀects of alternative options. In some cases, it involves the translation into monetary terms of non-marketed environmental, social and other impacts using some of the methods described in Section 2.2. An example of cost–beneﬁt analysis applied to groundwater remediation is presented in Section 5.

2.3.2.4

Economic Behavioural Models

Behavioural models aim at simulating how changes occurring in the economic, regulatory or natural environment aﬀect decisions of economic agents (farmers, households, industrialists) using water (groundwater abstraction) and/or impacting on its quality (pollution). Two broad types of behavioural models can be distinguished: decision-based models and statistical models. Decision-based models aim at representing the decision process that determines water use or pollution emission. Such models have been, for example, extensively applied for simulating farmers’ choice in terms of cropping patterns, water use (irrigation) or fertiliser use. Two types of models can be distinguished: optimisation models and rule-based models. Optimisation models assume that farmers select the combination of crops which maximises their income under a set of technical, regulatory and economic constraints; these models estimate crop choices, input consumption (fertiliser, labour, energy) and farm income for diﬀerent input parameter values (agricultural prices and subsidies, input prices, minimum set aside constraint, etc). Optimisation models have also been developed for households to investigate households’ groundwater use.12 Rule-based models assume that individual decisions result not only from an economic optimisation process but also from interactions between various categories of actors and social objectives which may overrule the proﬁt maximisation objective. Such models are often formulated using agent-based modelling techniques and implemented using object-oriented programming language,13 for an example applied to groundwater management. Statistical models aim at exploring the relationships between (ground) water use behaviour and socioeconomic characteristics of a panel of water users. Such models have been extensively used following the seminal work of Howes and Linaweaver14 for modelling drinking water demand and assessing the

64

Chapter 2.3

sensitivity of water use to price increase.15 More recently, statistical models have also been applied to the farming sector, using positive mathematical programming which investigates relationships between farm crop choices and factors determining production choices.16,17

2.3.2.5

Selected Illustrations

In the following sections, practical case studies are presented to illustrate how the methods and tools described above can be implemented in practice, and which results they deliver. The ﬁrst case study focuses on the assessment of pollution damage costs due to agricultural diﬀuse pollution with nitrates and pesticides (Section 3). The second example, based on a Slovenian case study, illustrates how cost-eﬀectiveness analysis can be used to identify the least costly way to reduce groundwater pollution (Section 4). Section 5 presents the results of assessments of costs and beneﬁts of alternative groundwater protection scenarios from a practical case study conducted in Latvia. In Section 6, we illustrate how economic and groundwater models can be integrated to simulate long-term global change scenarios.

2.3.3

Assessing and Simulating Current and Future Socio-Economic Impact of Groundwater Deterioration

The ﬁrst example focuses on the assessment of socioeconomic impact of groundwater pollution with nitrates and pesticides in the upper Rhine valley aquifer. The demand for this economic assessment emerged in the early 2000s following publication of the results of a trans-boundary groundwater quality survey showing a steady increase of nitrate concentration in very large areas of the aquifer, combined with a drastic increase in the pesticide detection frequency. Reversing this trend would require a drastic reinforcement of groundwater protection measures, in particular for the agriculture sector. But regional policy-makers feared that this would lead to signiﬁcant opposition and protest from the farm lobby. To improve the social acceptance and the legitimacy of groundwater protection measures, they decided to launch a research project to quantify in monetary terms past and future socioeconomic impacts of groundwater pollution. The ﬁrst part of the study consisted in assessing socioeconomic impacts of nitrate and pesticide pollution which had occurred during the last 15 years. The analysis was based on a consultation of experts, review of archives of ﬁnancial actors and interviews with municipalities and industry representatives. The results showed that one-third of drinking water utilities, representing 177 municipalities and 432 000 inhabitants, had been directly concerned by nitrate or pesticide pollution during the 15 years period. And investments made to respond to this pollution were estimated at h26 million resulting in an average increase by h0.2 per cubic metre of the drinking water price (or h30 increase of

Groundwater Management and Planning: How Can Economics Help?

65

the water bill per household per year). This increase would have been twice as high if investments made by drinking water utilities had not been heavily subsidised. Groundwater pollution also contributed to eroding population trust in tap water and to increasing bottled water consumption, generating total additional costs estimated at h165 million over the 15-year period considered.18 The second part of the study consisted in estimating future costs that would occur in the absence of additional groundwater protection measures. It required developing a tool that would enable the simulation of future groundwater quality changes, combined with an economic method for assessing damage costs that would be incurred by water users as a result of groundwater quality changes. The evolution of water quality was simulated using geostatistical methods, assuming trends in water quality observed in the recent past could be extrapolated to the 2015 time horizon.19 The results of the extrapolation (Figure 2.3.1) showed that the estimated average nitrate concentration would increase up to 26.3 mg l1 by 2015 (against 25.7 mg l1 in 2003) with the area where drinking water threshold value for nitrates is exceeded (50 mg l1) being doubled (10% of total area in 2015 versus 5% today). Concerning pesticides, the evolution would diﬀer from one substance to another, with the concentration of atrazine and its metabolites decreasing while metolachlore and alachlore concentrations would increase. Overall, the study showed that 33 drinking water wells belonging to 21 public drinking water utilities would be contaminated by 2015 (mainly by pesticides). Moreover, 68 additional drinking water wells would be at risk as pesticide

2003

2015

Pesticide concentration

Pesticide concentration

2015

Newly affected areas

Sum of all pesticide concentration < 0,05 µg/l Sum of all pesticide concentration > 0,05 µg/l & drinking water threshold values not exceeded Drinking water threshold values exceeded for one individual substance of the sum of substances

Figure 2.3.1

Maps of pesticide concentration for 2003 and 2015.

66

Chapter 2.3

concentration for these wells would be above 0.05 mg l1 in 2015, i.e. 50% of the drinking water threshold value, thus with the possibility that drinking water thresholds are occasionally exceeded. The damage costs associated with the contamination of the 33 wells were estimated assuming that all concerned water utilities would install activated carbon ﬁlter treatment units for eliminating pesticides. In addition, collaborative agreement with farmers would be established to further reduce input of pesticides and nitrates in the nearby aquifer wells. Total costs, including treatment cost (h0.07 per m3) and compensations paid to farmers (h230 per hectare and per year), were estimated at h1.75 million per year in 2015.

2.3.4

Cost Eﬀectiveness Analysis of Groundwater Protection Measures: Finding the Least Costly Way to Reduce Nitrate Pollution to Groundwater

The second example, based on a Slovenian case study, focuses on the identiﬁcation of the least costly way to reduce nitrate pollution in groundwater. The aquifer selected for this case study is the Krsko kotlina aquifer located near the Croatian boarder in Slovenia.20 This aquifer faces increasing nitrate and pesticide pollution originating mainly from agriculture and municipal wastewater collection and treatment. If current pressures remain as they are today, nitrate concentration in groundwater is expected to rise above the 50 mg l1 drinking water threshold value. Pesticides are present at all monitoring locations, with the concentration of some pesticides increasing and being sometimes above threshold values speciﬁed in relevant legislation. Potential measures for reducing groundwater pollution and their cost-eﬀectiveness were investigated in the context of a pilot project aimed at testing methodologies for supporting the implementation of the EU WFD.20 For the agriculture sector, measures considered included adopting agri-environmental measures with lower fertiliser use, the replacement of cropped area by meadows, in particular in water resources protection zones, better management of farm yard manure on farm or the shift to organic farming. For the municipal sector, measures identiﬁed included the installation of new sewage and wastewater treatment facilities, the renewal of leaking sewage, the connection of disconnected households to public sewage networks or the installation and eﬃcient management of sceptic tanks. The expected eﬀects in terms of reduction in nitrate leaching to the groundwater were estimated for individual measures. Costs that were considered included direct investment costs, operation and maintenance costs and in some cases indirect economic costs imposed on economic sectors (e.g. diﬀerences in farm income/value added when shifting from today’s agriculture to organic farming). All costs were annualised based on the expected time life of investments and the distribution of (direct, indirect) costs over time, using a discount rate of 7%. Table 2.3.1 presents estimates obtained for diﬀerent measures and used for building the cost-eﬀective

70 ha

2163 ha

4800 ha 2088 ha

0 ha

1431 ha

2400 ha

696 ha

40 ha 348 ha

0 ha

Shift to meadows in water protection zone 1 for Drnovo and Brege abstraction wells Limits imposed on fertiliser use (170 kg N ha1) in water protection zone 2 (both wells) Best management practice for agriculture Winter green cover

Buﬀer zones Ecological farming

Stricter limitation of fertiliser use in the Brege water protection zone 2 (100 kg ha1) 497 ha

139 ha 1044 ha

Maximum coverage

Potential measures 960 084 SIT ha1

797 SIT ha1

897 SIT ha1 28 997 SIT ha1 1764 SIT ha1 45 397 SIT ha1 1 104 097 SIT ha1

0.00652 mg l1 ha1

0.00056 mg l1 ha1

0.00056 mg l1 ha1 0.00119 mg l1 ha1

0.00596 mg l1 ha1

0.00585 mg l1 ha1 0.00176 mg l1 ha1

Annualised costs

Reduction in nitrates (in absolute terms)

0.0054

3.3172 0.0388

0.0410

0.6243

0.7023

0.0068

Cost eﬀectiveness ratio (mg l1 SIT1)

Coverage, costs and eﬀectiveness of potential measures (PE ¼ population equivalent).

Actual coverage (already implemented)

Table 2.3.1

10

1 6

5

4

3

9

Ranking based on costeﬀectiveness ratio

Groundwater Management and Planning: How Can Economics Help? 67

(continued )

Wastewater treatment and sewage for groups of houses with less than 50 PE Wastewater treatment and sewage for small villages (50 o PE o 2000) Wastewater treatment and sewage for larger settlements (PE 4 2000)

Stricter limitation of fertiliser use in the Drnovo water protection zone 2 (100 kg ha1) Improved septic tanks

Potential measures

Table 2.3.1

50 ha

10 194 PE 297 PE

1019 PE

8115 PE

2548 PE

0 PE

0 PE

0 PE

Maximum coverage

0 ha

Actual coverage (already implemented) 917 SIT ha1

26 042 SIT PE1 19 094 SIT PE1 10 115 SIT PE1 8941 SIT PE1

0.00184 mg l1 ha1

0.00010 mg l1 PE1

0.00017 mg l1 PE1

0.00017 mg l1 PE1

0.00010 mg l1 PE1

Annualised costs

Reduction in nitrates (in absolute terms)

0.0190

0.0168

0.0052

0.0038

0.0067

Cost eﬀectiveness ratio (mg l1 SIT1)

7

8

11

12

2

Ranking based on costeﬀectiveness ratio

68 Chapter 2.3

Groundwater Management and Planning: How Can Economics Help?

69

programme of measures aim at stabilising average nitrate concentration in groundwater at 50 mg l1.20 The activity undertaken as part of the Krka pilot project had a clear methodological focus and was not aimed at delivering results as part of a decision-making process. However, the assessments showed that the most costeﬀective programme for reaching the 50 mg l1 threshold for nitrate concentration would require changes in agriculture for 5400 hectares, with measures for reducing municipal sector pollution targeting around 9100 Population Equivalent.20 Total costs of such a programme were estimated roughly at 340 million SIT (or h1.5 million), equivalent to an average cost of around h200 per hectare for the entire aquifer area. The analysis emphasised that implementing in priority basic measures required for the implementation of existing legislation (Wastewater Treatment Directive, Nitrate Directive), independently of their cost-eﬀectiveness ratio, would lead to signiﬁcantly higher costs for groundwater protection up to h370 per hectare.

2.3.5

Cost–Beneﬁt Analysis of Groundwater Protection: Finding the Economically Optimal Level of Groundwater Protection

The fourth example illustrates how cost–beneﬁt analysis can be applied to assess whether proposed quality objectives or threshold values proposed for groundwater protection can be justiﬁed from an economic point of view. It is based on the results of a case study developed in Latvia as part of the EUfunded BRIDGE research project (see Ref. 21 for more details). The case study investigated the economic consequences of two alternative groundwater quality objectives for petroleum products in the shallow groundwater aquifer underlying the capital city of Riga. There are two main types of pressures on the aquifer that explain pollution by petroleum products: historical pollution sources linked to former military zones, industries, the past operation of fuel ﬁlling stations and tanks or highly contaminated sites that are oﬃcially registered; and current sources of pollution such as fuel ﬁlling stations and tankers, parking places or car disposal places (car cemeteries) that represent potential sources of pollution. The latter has clearly a marginal impact on the existing pollution by petroleum products, with the former representing around 90–92% of the observed pollution. Some pollution in the shallow aquifer also originates from street runoﬀ and unused wells. While the area polluted with petroleum products might appear rather small (97 polluted sites representing 70 hectares out of a total of 30 500 hectares for the entire city of Riga), extremely high concentrations in petroleum products can be found, exceeding the existing (environmental) water quality threshold value of 0.2 mg l1 set in the legislation by 100 to 1000 times. This does not impose threats on human health as shallow groundwater is rarely used for drinking water purposes (drinking water for the city of Riga comes from an upstream dam on the Daugava River). But it creates a permanent ﬂoating layer

70

Chapter 2.3

of petroleum products that, for contaminated sites close to surface waters, imposes a risk to connected surface water ecosystems. Some petroleum product pollution in the Daugava River crossing the city of Riga (close to its outfall into the Gulf of Riga) is, for example, directly coming from the connected shallow aquifer. Highly contaminated sites in the middle of the city also impose constraints on urban and economic development: as new owners of such sites need to clean (groundwater and soil) pollution (from mainly historical sources for which they have no responsibility) whenever new urban developments are proposed. Environmental policies and urban planning strategies are already in place for dealing with petroleum product pollution in shallow aquifers. The review of planning and strategic documents stressed that ﬁnancial resources have already been secured for cleaning four highly polluted sites in the coming years. In addition, good management practices are compulsory for fuel ﬁlling stations, petrol stations and parking places. Clearly, however, these measures will not be suﬃcient for restoring good shallow groundwater quality in the entire aquifer. The analysis investigated in some details the economic implications of two diﬀerent objectives or scenarios. Scenario 1 considers cleaning all sites polluted with petroleum products up to the actual 0.2 mg l1 threshold for 30 sites out of 97 that are highly polluted and (i) connected to surface water ecosystems or (ii) located in a current or future residential area. Scenario 2 is more ambitious and considers cleaning all (97) sites polluted with petroleum products. Diﬀerent actions and measures that would help reducing petroleum product pollution were identiﬁed. These include pumping up existing ﬂoating layers, treating contaminated soils or directly shallow groundwater, building rain water collectors or installing and operating rain water pre-treatment facilities. Based on individual cost ﬁgures (investment, operation and maintenance costs) for each measure, total costs (annualised, using a 4% discount rate) were estimated for each scenario assuming that measures would be implemented by 2015 in both scenarios. Cost ranges were estimated at h26–30.5 million and h50–58.6 million for scenarios 1 and 2, respectively. The beneﬁts accruing to urban and economic development were estimated, assuming these beneﬁts would be equal to the total area cleaned under each scenario multiplied by surrounding land property values existing for new developments. Beneﬁts were estimated by summing up the cadastral value of properties in Riga for each site multiplied by the concerned area of each site. Total beneﬁts (annualised, using a 4% discount rate) accruing to urban and economic development were estimated at h1.9 million and h4.3 million for scenarios 1 and 2, respectively. To estimate total beneﬁts resulting from groundwater quality improvements, a contingent valuation survey was carried out in the case study area. Overall, 510 citizens from the city of Riga and neighbouring areas were interviewed to capture the total value people attach to groundwater quality improvement and to the elimination of petroleum product pollution. Average WTP for removing petroleum products from shallow groundwater was estimated at h24.5 per household per year, with no signiﬁcance diﬀerence between the WTP for

Groundwater Management and Planning: How Can Economics Help?

71

diﬀerent levels of groundwater quality improvement corresponding to scenarios 1 and 2. This WTP value applied to all inhabitants of the city of Riga leads to total beneﬁts at around h68 million (for more details, see Ref. 21), thus signiﬁcantly higher than the beneﬁts accruing from urban development presented above. This stresses the importance of the non-use value people attach to groundwater. The comparison between total (annualised) costs and beneﬁts showed that both scenarios can be justiﬁed from an economic point of view—with the ﬁrst scenario yielding a better economic outcome (total beneﬁts minus total costs) than the second (more ambitious) scenario. This results mainly from the very high population (density) in the area (Riga is the capital city of Latvia and hosts nearly one-third of the entire population of the country) to which WTP values for groundwater quality improvements are applied. Further analyses investigated the time dimension of groundwater improvement programmes. If proposed measures are delayed and implemented for a longer time period to account for the availability of ﬁnancial resources and budgetary constraints, annualised costs are lower. As annualised beneﬁts are also lower (because beneﬁts are obtained after a longer time period), net beneﬁts would be higher than those calculated for a faster implementation time. This could be the basis to justify a longer implementation period for measures: a result which clearly is very site and pollutant speciﬁc and that cannot be extrapolated!

2.3.6

Integrating Economic and Groundwater Models for Simulating Nitrate Pollution in the Upper Rhine Valley Aquifer

The fourth example illustrates how economic and groundwater models can be integrated to investigate the environmental impact of changes in economic sectors/sector policy. The example is based on the results of the MONIT InterReg project conducted by French, German and Swiss partners in the upper Rhine valley aquifer.22 The objective of the project was to develop a modelling tool capable of simulating future evolution of nitrate concentration corresponding to diﬀerent global (economic and environmental) change scenarios. The approach developed recognises that the dynamics of groundwater systems mainly depends on economic drivers inﬂuencing activities generating pollution (e.g. farming), and, subsequently, that the processes governing this economic activity must also be represented and analysed in an integrated model. A farm sector economic model was developed to assess the extent to which levels of pollution emission generated by the farming sector could change depending on global change scenarios. The economic model is developed using linear programming (LP). LP models assume that farmers select the combination of crops which maximises their income under a set of technical, regulatory and economic constraints.23 The models simulate crop choices, input

72

Chapter 2.3

consumption (fertiliser, labour, energy) and farm income for diﬀerent input parameter values (agricultural prices and subsidies, regulatory constraints, changes in the price of input such as energy, fertiliser, labour, minimum setaside constraint, etc.). The models incorporate constraints related to crop rotations, labour availability, production quotas (sugar-beet and milk quotas), manure storage and management (for livestock-oriented farms). Risk is also integrated in LP models to account for the variability of output prices and yields. A crop production function (for corn only) is also introduced in the model to capture the crop yield response to nitrogen input. The LP models representing production choices of twenty selected representative farms were calibrated for sample farms by comparing simulation results with current cropping patterns. The output of the economic model (area under each crop, nitrate use) provides input data to a large-scale nitrate balance model which calculates nitrate leaching for the entire aquifer. This nitrate balance model is developed using the STOFFBILANZ software.23 Some parameters of STOFFBILANZ are adjusted through the use in simulation of a process-oriented nitrate input model developed at a plot scale (soil-plant-model STICS) to determine the nitrate input into the groundwater as a function of crop type, management practice, fertilising practice and climate. The output of the nitrate balance model feeds into the groundwater ﬂow model coupled with a transport model, developed using MODFLOW and MT3D (for more details, see Ref. 23). The economic and soil–groundwater models were then used to simulate the impact of three global change scenarios developed by a group of experts. Given the uncertainty associated with future changes, three contrasted scenario were developed, a baseline scenario and two variants inspired from the scenario developed by the International Panel on Climate Change (IPCC).24 The baseline scenario assumes that the corn root worm detected since 2003 in the region extends over large areas, forcing farmers to increase crop rotation and limiting corn area to less than 50% of the total cultivable area. Energy prices (gazoil) are supposed to increase by 6% per year on average (2015 prices are twice those of 2003). And it is assumed that no ﬁnancial compensation mechanism is implemented by national governments. As a result of energy price growth, fertiliser price increases by 1.5% per year. Due to the European enlargement, temporary labour force increases signiﬁcantly (+66% in twelve years), reducing the proﬁtability of vine, fruit and vegetable crops (in particular in Germany where foreign labour force is signiﬁcantly used for harvesting of these crops). In France, it is assumed that the reform of the Water Act establishes a new water abstraction tax of h0.025 per m3. In Germany, the tax called Wasserpfennig is maintained at its 2003 level (i.e. h0.05 per m3). In both countries, farmers are allowed to produce bio-diesel for their own farm use only. And the bio-fuel industry develops but only for producing ethanol using corn (no bio-diesel industry). The second scenario, called ‘‘A1’’ with reference to IPCC emission scenario A1,24 depicts a more liberal future. Agriculture development aims at maximising competitiveness in markets which tend to function without protectionist

Groundwater Management and Planning: How Can Economics Help?

73

barriers and with minimal environmental constraints: taxes on water abstraction are eliminated in Germany and not introduced in France. Liberalisation favours the full implementation of the decoupling principle promoted by the Common Agriculture Policy reform in France, which result in a drastic change in crop gross margins and relative proﬁtability. Energy price increase is compensated by a ﬁscal stabilisation mechanism and limited to 40 and 68% in Germany and France, respectively. Signiﬁcant technical means are mobilised by government agencies to ﬁght against the corn root worm (pesticides are sprayed by helicopter). And the biofuel industry develops, representing new market opportunities for farmers. The third scenario, inspired from the B2 IPCC emission scenario, corresponds to a vision of the future where agriculture evolved under the double pressure of increasing input prices (energy, fertiliser) and more stringent environmental constraints. Water abstraction taxes are established at the level of the baseline scenario. A tax on fertiliser is introduced at h0.15 per kg in France and h0.26 per kg in Germany. Due to high energy price and an active governmental support to bio-fuel development, crops used for bio-fuel production represent a very attractive market. The proliferation of the corn root worm compels farmers to reduce the area under corn. And ﬁnancial support is granted to fruit and vegetable farms to invest in machinery and compensate for the increase of temporary labour costs. The consequences of these three global change scenarios were assessed using the chain of models described above. The results show that future nitrate concentration in the aquifer will not signiﬁcantly diﬀer from one scenario to the other in the short term (estimated at 19 mg l1 for the baseline and A1 scenarios, and 19.5 mg l1 for the B2 scenario). More signiﬁcant diﬀerences are expected in the longer term (average concentration of 16 mg l1 for baseline and A1 scenarios, 18.2 mg l1 for the B2 scenario). The model shows that the area where nitrate concentration exceeds drinking water thresholds (50 mg l1) will drastically fall from 17 000 hectares in 2005 to around 4000 ha for the baseline and A1 scenarios, and to 6000 ha for the B2 scenario (Figure 2.3.2). Surprisingly, scenario B2 which depict a world with more stringent environmental constraints is also the worst scenario in terms of water pollution due to the increase in areas under industrial crops used for producing bio-fuels.

2.3.7

Designing Economic Instruments for Groundwater Management

As indicated above, recent policy developments have also put emphasis on the potential role economic instruments might play in enhancing the sustainability of water resources. In parallel to requesting EU member states to undertake economic assessments for supporting the selection of measures, the EU WFD also promotes a sounder application of the polluter pays principle, an adequate recovery of the costs of water services and water pricing that provides an incentive to more eﬃcient water use. Furthermore, economic instruments such

74

Chapter 2.3 17000 Baseline

Area where [NO3] > 50 mg/l (ha)

15000

A1 B2

13000

11000

9000

7000

5000

3000 2005

Figure 2.3.2

2015

2025

2035

2045

Simulated evolution of the area (in ha) where nitrate concentration exceeds 50 mg l1 (adapted from Ref. 22).

as water charges/taxes and systems of tradable water rights can be considered when developing programmes of measures. Indeed, economic instruments that provide ﬁnancial incentives to reduce groundwater abstraction might be relevant options when the ecological consequences of groundwater depletion (e.g. sea water intrusion in coastal areas, drying up of wetlands or reduced discharges of river springs) cannot be mitigated by technical solutions alone. The following paragraphs investigate the two main types of economic instruments: environment taxes and tradable groundwater licences/rights.

2.3.7.1

Environment Taxes and Charges

Environmental taxes and charges can be imposed on users who abstract groundwater as an incentive for reducing water use and thus pressures on the aquatic ecosystem. The higher the taxes, the lower abstraction is expected to be. Designing an environment tax system is, however, not trivial. From a theoretical point of view, the tax level should be such that it covers all costs generated by groundwater use, including external costs caused to third parties and to the environmentz. In practice, the design of a tax system consists in ﬁnding a z

Environmental costs can be signiﬁcant where groundwater overexploitation leads to irreversible environmental and economic damages costs (e.g. abandonment of drinking water wells in case of sea water intrusion in coastal aquifer, destruction of natural habitat in wetlands).

Groundwater Management and Planning: How Can Economics Help?

Table 2.3.2

Unitary rates for abstraction taxes and charges for selected member states of the European Union (2004 values: in Ref. 27).

Country France (basic rate of the Seine Normandie river basin)

Estonia

75

Sources of water to which the Tax/charge unitary tax/charge is applied rate (h per m3) Surface water: on volume abstracted Surface water: on volume consumed Groundwater: on volume abstracted Groundwater: on volume consumed Surface water Groundwater

Hungary

All sources of water

Slovenia

All sources of water

0.00071 0.04 0.024 0.04 0.013–0.016 0.016–0.048 depending on use 1.147 for mineral water 0.007–0.02 depending on use 0.03

compromise between economic eﬃciency and social acceptability: the latter criterion being of utmost importance as taxes will generally be resisted by water users who consider water as a free resource.25,26 And the full cost recovery principle promoted by current policies (e.g. the EU WFD of 2000, the 1994 water reform in Australia) is in fact never implemented in practice. As a result, external costs and damages remain borne by economically less sensitive sectors (frequently households). Environmental taxes and charges are more widely used in Europe than in the USA for instancey. Such taxes or charges for water abstraction are rather common as illustrated in Table 2.3.2. When diﬀerent rates are proposed for surface water and groundwater abstraction, those for groundwater are usually higher emphasising the higher status groundwater protection might have in the ﬁeld of water policy. However, a closer look at current tax and charge levels shows that they remain very low as compared to other production costs or value added.28 In some cases, tax exemption is also given to economic sectors to limit its expected negative impact on competitiveness. This situation is illustrated by the Netherlands, where a groundwater tax system was introduced in 1995 as part of an eﬀort to broaden the tax base by shifting the emphasis of revenue generation from conventional taxes to environmental taxes other than energy taxes. The amount of the tax payable is based on the volume of groundwater abstracted with a tax rate of h0.1785 per m3 (2004 ﬁgures). Companies could, however, obtain a signiﬁcant tax rebate (h0.1495 per m3) when inﬁltrating water back y

See Ref. 25 for a discussion of water pollution taxes versus pollution discharge rights.

76

Chapter 2.3

into the aquifer. Furthermore, small groundwater abstractors were exempted. Although small- and medium-sized enterprises (SMEs) and industries faced a water price increase ranging from 40 to 113% when the groundwater tax was ﬁrst introduced, the revenue collected by the groundwater tax amounted to 0.03 and 0.08% of the industry turnover and value added, respectively. This was equivalent to 0.33% of pre-tax proﬁts in the industrial sector,28 thus unlikely to impair the competitiveness of Dutch industry. Overall, environmental taxes and charges are often justiﬁed on environmental grounds, i.e. as an incentive to economic sectors to limit water use or reduce pollution. In practice, however, existing taxes and charges do have limited incentive eﬀects. They mainly play a revenue-raising role, revenues that in some cases (when they are earmarked to environmental protection) can help support ﬁnancially the implementation of groundwater protection programmes and projects.

2.3.7.2

Tradable Groundwater Licences and Rights

The establishment of tradable groundwater abstraction licences or rights is another type of economic instrument that might play a role in groundwater management. Based on hydrogeological knowledge, the total maximum volume that can be abstracted from an aquifer (or sustainable yieldz) can be estimated and converted into a number of individual water abstraction licences or property rights which can be allocated to individual water users. Individual rights can be expressed in volume or in percentage of a sustainable aquifer yield, which can be reassessed every year to account for climatic variability and the eﬀect of past abstractions. Whereas licences are issued for a speciﬁc period and can be revised or cancelled, water (property) rights are granted on a permanent basis. The initial allocation of licence or rights can be based on historical consideration (ﬁrst in time, ﬁrst in right) or on economic consideration (through an auction mechanism, for example, to allocate groundwater to its highest marginal values). Following the initial allocation, water licences or rights can be leased or sold among users. The underlying assumption is that water users will accept to sell part or whole of their licence/right if the ﬁnancial compensation (price) they can obtain for the transfer exceeds the beneﬁts they would derive from using their water directly. In theory, market forces ensure that water is (re-)allocated to the most eﬃcient users, that waste is minimised and that the total economic value of water use is maximised. Water markets have been established in the USA, leading to the reallocation of surface and groundwater resources from one region to another, and from one economic sector to another.29,30 For a detailed description of groundwater markets in Texas and Arizona, see Ref. 31. Similar markets also function in well-regulated river basins in arid area of Chile.32,33 Following the successful z

Sustainable aquifer yield is deﬁned as the groundwater extraction regime, measured over a speciﬁed planning time frame that allows acceptable levels of environmental impact on associated ecosystems and protects the higher value uses that have a dependency on the water.

Groundwater Management and Planning: How Can Economics Help?

77

development of water market in the Murray–Darling basin in the mid1980s,33,34 Australian authorities have also promoted the development of groundwater licence trading as an alternative to the ‘‘command and control’’ approach to moving water use to areas with higher economic value and eﬃciency.35 In Europe, although trading of water use rights has locally been functioning for centuries within irrigation systems in Spain,30 they are rarely considered as measures in water management and policy. As an exception, tradable water permits (quotas) were considered as possible measure for reducing overabstraction for the Beauce aquifer (central France) at the end of the 1990s.36 However, they were never implemented because of the drastic legislation change they would have required. The EU WFD refers also to tradable water rights as possible measure for reaching its environmental objectives. But no EU member state is expected to make use of this type of economic instrument at least in the short and medium term.

2.3.8

Conclusions

This chapter has provided examples of the role economics can play in supporting decisions in the ﬁeld of groundwater management and planning. At the same time, the illustrations presented above clearly show that economics cannot answer policy issues and questions alone. Indeed, supporting groundwater management and planning decision requires multidisciplinary approaches where economics and technical knowledge must be well integrated. Integration between economics with hydrogeology is required for the development of scenarios of plausible futures described for example in Sections 3 and 6. In developing scenarios, economic assessment or modelling can either be used ‘‘upstream’’ or ‘‘downstream’’ from hydrogeological assessment and modelling. An ‘‘upstream’’ use of economics is important when a signiﬁcant evolution of economic activities generating pressures on groundwater (abstraction or pollution) is expected to occur during the time horizon considered. As illustrated in Section 6, an economic model can then help in simulating likely future evolution of economic activities and resulting pressures aﬀecting the state of groundwater resources, under diﬀerent assumption of global economic change (e.g. changes in agriculture policy, demographic growth, changes in energy prices), environmental change (e.g. climate change) or policy action scenario (e.g. implementation of an environmental tax). A ‘‘downstream’’ use of economics is relevant to situations with signiﬁcant economic interests depending on the state of groundwater resources. As illustrated in Section 3, this can help assess the economic consequences (e.g. in terms of economic sector production and value added, employment, competitiveness) of expected future groundwater status changes obtained from hydrogeological model simulation. Multidisciplinary approaches which integrate economics with hydrogeology are also essential to the evaluation of alternative groundwater management

78

Chapter 2.3

options, in line with the requirements of the EU WFD. Whether the evaluation requires a cost-eﬀectiveness analysis (as illustrated in Section 4) or a cost– beneﬁt analysis (presented in Section 5), economic assessments must build on a robust understanding of the impact of possible management actions on groundwater status. And such understanding requires hydrogeological knowledge or modelling. However, studies where speciﬁc groundwater models and economic assessment are conducted together remain rare.37,38 This leads to a relatively high uncertainty associated with cost-eﬀectiveness or cost–beneﬁt estimates: an issue that requires further investigation by the research community. Although the need for multidisciplinary research is increasingly recognised by scientiﬁc and policy-making communities,39 the actual development of multidisciplinary research work in that ﬁeld is somewhat impaired by the fact that academic evaluation (in both the hydrogeology and economics ﬁelds) rarely gives credit to such multidisciplinary approaches. The search for more multidisciplinary work reinforces the issues of scale. On the one hand, it questions the extrapolation to larger scales of results from innovative economic methods developed and applied at local case study level, in particular in the domain of environmental cost and beneﬁt valuation (e.g. contingent valuation). The issue of aggregation of values obtained from contingent valuation studies for estimating total beneﬁts of a given management option and environmental quality change at the river basin scale remains a challenge. Whether the same protocol should apply for building contingent valuation surveys at local or large scale is also a research question that might require further attention. On the other hand, economists are faced with the need to link global economic models representing entire economies (partial equilibrium models or general equilibrium models) to impacts on water and groundwater resources at the macroscale. This implies linking macroeconomic models with groundwater models at regional or national level, a scale at which hydrological functioning is diﬃcult to simplify (and might not have much meaning). Another key challenge for economists working on groundwater management is the need to enhance the appropriation of economic principles, tools and results by policy-makers and stakeholders involved in the preparation of groundwater management plans and programmes. Whereas policy actors might be, by tradition, more familiar with the knowledge mobilised by hydrogeologists, there is still a limited understanding of the role economics can play in supporting their decision—although it is clear that policy-makers are very sensitive to economic arguments. In Europe, this situation is slowly changing with the implementation of the EU WFD gaining momentum and more economic analyses being undertaken and discussed among policy-makers and stakeholders at the national and river basin district scales. Also, one can observe an emerging demand for economic assessments of water management options at the local level (for instance in the Arde`che river basin in France). But we are still at the infancy of more systematic applications of economic assessments. Also, how the results of economic analyses will be used for supporting policy decisions (if used at all) remains to be seen.

Groundwater Management and Planning: How Can Economics Help?

79

A better appropriation of economic concepts and methods by policy-makers and stakeholders will require speciﬁc eﬀorts by the economic community in awareness raising and information, including on the strengths and weaknesses of assessments, on underlying uncertainties and on the demonstration of the added value of the integration between economics and technical expertise. In some cases, it might also require a drastic change in the economists’ thinking and discourse, away from the assumption that individuals are acting purely along the simpliﬁed concepts and principles that are the basis of the microeconomic theory. The real-life application of economic assessments that will be produced in the coming years in the context of the implementation of the WFD and of its daughter groundwater directive will be instrumental in enhancing the shared knowledge between economists, hydrogeologists, policy-makers and stakeholders.

References 1. World Wide Fund for Nature and European Environment Bureau, EU Water Policy: Making Economics Work for the Environment. Survey of the Economic Elements of the Article 5 Report of the EU Water Framework Directive, report prepared by the World Wide Fund for Nature and the European Environment Bureau, Brussels, 2006. 2. A. Scheidleder, J. Grath and G. Winkler, Groundwater Quality and Quantity in Europe, European Environmental Agency, Copenhagen, 1999. 3. S. Nixon, Z. Trent and C. Marcuello, Europe’s Water. An Indicator Based Assessment, Topic Report 1/2003, European Environmental Agency, Copenhagen, 2003. 4. T. Shah and D. Molden D, The Global Groundwater Situation: Overview of Opportunities and Challenges, International Water Management Institute, Colombo, 2000. 5. W. Harrington, J. Krupnick and W. A. Spoﬀord, J. Urban Econ., 1989, 25, 116–137. 6. P. Dasgupta, Environ. Develop. Econ., 2004, 9, 83–106. 7. A. H. Smith, E. O. Lingas and M. Rahman, Bull. World Health Organ., 2000, 78, 9. 8. C. W. Abdalla, Am. J. Agricul. Econ., 1994, 76, 1062–1067. 9. A. Stenger and M. Willinger, J. Environ. Manag., 1998, 53, 177–193. 10. B. Go¨rlach and E. Intervies, Economic Assessment of Groundwater Protection: A Survey of the Literature, Ecologic, Berlin, 2003. 11. A. Gerarsidi, P. Katsiardi et al., Cost-eﬀectiveness analysis for water management in the island of Paros, Greece, International Conference on Environmental Science and Technology, Lemnos Island, Greece, 2003. 12. M. Montginoul, J.-D. Rinaudo, P. Garin, Y. Lunet delajoncquie`re and J. P. Marchal, Water Policy, 2005, 7, 523–541. 13. S. Feuillette and F. Bousquet et al., Environ. Model. Software, 2003, 18(5), 413–427.

80

Chapter 2.3

14. C. W. Howes and F. P. Linaweaver, Water Resour. Res., 1967, 3(1), 13–32. 15. F. Arbue´s and M. A. Garcia-Valinas et al., J. Socio-Econ., 2003, 32(1), 81–102. 16. R. E. Howitt, Am. J. Agricul. Econ., 1995, 77, 329–342. 17. L. Judez and J. M. de Miguel et al., Math. Comput. Modell., 2002, 35(1–2), 77–86. 18. J.-D. Rinaudo, C. Arnal, R. Blanchin, P. Elsass, Meilhac and Loubier, Water Sci. Technol., 2005, 52(9), 153–162. 19. J.-D. Rinaudo, P. Elsass, R. Blanchin, C. Arnal and Meilhac, Assessing the Social and Economic Impact of Nitrate and Pesticide Contamination of the Alsatian Aquifer: Synthesis Report [in French], report Brgm/RP-53172FR.Orle´ans, BRGM, 2006. 20. Selecting Measures to Improve Water Status in the Krka River Sub-Basin, technical report of the Krka Pilot Project, Ljubljana, 2006. 21. K. Pakalniete, H. Bouscasse and P. Strosser, Assessing Socio-Economic Impacts of Diﬀerent Groundwater Protection Regimes. Latvian Case Study Report, research report developed under the EU-funded BRIDGE research project, ACTeon, Orbey, 2006. 22. LUBW, Simulating Nitrate Groundwater Pollution in the Upper Rhine valley [in French and German], Monit Intergreg IIIA project, ﬁnal report, LUBW, Karlsruhe, 2006. 23. P. B. R. Hazell and R. D. Norton, Mathematical Programming for Economic Analysis in Agriculture, Macmillan, New York, 1986. 24. Intergovernmental Panel on Climate Change, Emissions Scenarios: Summary for Policy Makers, IPCC special report, 2000. 25. C. W. Howes, Environ. Res. Econ., 1994, 4, 151–169. 26. A. Dinar, The political economy context of water pricing reforms, in The Economics of Water Management in Developing Countries, ed. P. Koundouri, P. Pashardes, T. Swanson and A. Xepapadeas, Edgard Edwards, Cheltenham, UK, 2003. 27. P. Strosser and S. Speck, Environmental Taxes and Charges in the Water Sector. A Review of Experience In Europe, technical report, Catalan Water Agency, Barcelona, 2004. 28. Ecotec et al., Study on the Economic and Environmental Implications of the Use of Environmental Taxes and Charges in the European Union and Member States, report for the European Union, European Commission, Brussels, 2001 (http://europa.eu.int/comm/environment/enveco/taxation/ environmental_taxes.htm). 29. T. Anderson and P. J. Hill, Water Marketing: The Next Generation, Rowman and Littleﬁel, New York/London, 1997. 30. P. Strosser, Analysing alternative policy instruments for the irrigation sector: an assessment of the potential for water market development in the Chishtian sub-division, Pakistan, PhD thesis, Wageningen Agricultural University, 1997. 31. C. W. Howes, Environ. Develop. Econ., 2002, 7(2), 605–616.

Groundwater Management and Planning: How Can Economics Help?

81

32. C. J. Bauer, Against the Current: Privatization, Water Markets and the State in Chile, Kluwer, Boston MA, 1998. 33. H. Bjornlund and J. McKay, Environ. Develop. Econ., 2002, 7(2), 769–795. 34. S. Thoyer, Int. J. Sust. Develop., 2006, 9, 2. 35. ARMCANZ, Allocation and Use of Groundwater: A National Strategy for Improved Groundwater Management in Australia, policy position paper for advice to states and territories, Agriculture and Resource Management Council of Australia and New Zealand, Task Force on COAG Water Reform, Sustainable Land Water Resources Management Committee, 1996. 36. N. Kosciusko-Morizet, H. Lamotte and V. Richard, What can we expect from the establishment of transferable water entitlements in France: the case of irrigated agriculture [in French], Conference on Irrigation and Collective Water Resource Management in France and in the World, SFER, Montpellier, 19–20 November 1998. 37. J.-D. Rinaudo and S. Loubier, Cost beneﬁt analysis of large scale groundwater remediation in France, in Cost Beneﬁt Analysis and Water Resources Management, ed. R. Brouwer and D. Pearce, Edgard Edwards, Cheltenham, UK/Northampton, MA, 2005, pp. 290–314. 38. R. Brouwer, S. Hess, M. Bevaart and K. Meinardi, The Socio-Economic Costs and Beneﬁts of Environmental Groundwater Threshold Values in the Scheldt Basin in the Netherlands, case study report produced under the BRIDGE research project, IVM report R06-05, Institute for Environmental Studies (IVM), Vrije Universiteit Amsterdam, 2006. 39. Ph. Quevauviller, P. Balabanis, C. Fragakis, M. Weydert, M. Oliver, A. Kaschl, G. Arnold, A. Kroll, L. Galbiati, J. M. Zaldivar and G. Bidoglio, Environ. Sci. Pol., 2005, 8, 203–211.

3. Groundwater Regulatory Framework

CHAPTER 3.1

European Union Groundwater Policyw PHILIPPE QUEVAUVILLER European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium

3.1.1

Introduction

The European Union (EU) regulatory groundwater framework was developed at the end of the 1970s with the adoption of Directive 80/68/EEC on the protection of groundwater against pollution caused by certain dangerous substances. This directive provides a groundwater protection framework by preventing the (direct or indirect) introduction of high-priority pollutants into groundwater and limiting the introduction into groundwater of other pollutants so as to avoid pollution of this water by these substances. In 1982, a major assessment of groundwater resources was carried out in Europe, which consisted of a general survey (Groundwater Resources of the European Community: Synthetical Report) of groundwater quantity. Since it was published, attention has turned in Europe (and the USA) to quality, and not only have groundwater quality monitoring programmes been greatly expanded but many groundwater protection schemes have been put into place. The declaration of the ministerial seminar on groundwater held at The Hague in 1991 recognised the need for further action to avoid long-term deterioration of the quality and quantity of freshwater resources and called for a programme of actions to be implemented by the year 2000, aiming at sustainable management and protection of freshwater resources. This was followed by the elaboration of an action programme for integrated protection and management of groundwater in 1996, in which the Commission pointed to the need to establish procedures for the regulation of abstraction of freshwater and for the monitoring of freshwater quality and quantity. These w

The views expressed in this chapter are purely those of the author and may not in any circumstances be regarded as stating an oﬃcial position of the European Commission.

85

86

Chapter 3.1

considerations coincided with the request made by the European institutions to the Commission to come forward with a proposal for a directive establishing a framework for a European water policy. They were hence naturally embedded into the development of this large policy framework development, which resulted in the adoption of the Water Framework Directive (WFD) 2000/60/ EC on 23 October 2000. This chapter summarises the diﬀerent elements described above, and provides background information about parent legislation which has also to be considered within an enlarged groundwater regulatory framework.

3.1.2

The 1980 Groundwater Policy Framework

In the context of Directive 80/68/EEC on the protection of groundwater against pollution caused by certain dangerous substances,1 groundwater is deﬁned as ‘‘all water which is below the surface of the ground in the saturation zone and in direct contact with the ground or subsoil.’’ This directive provides a groundwater protection framework by preventing the (direct or indirect) introduction of high-priority pollutants (List I) into groundwater and limiting the introduction into groundwater of other pollutants (List II) so as to avoid pollution of this water by these substances (Table 3.1.1). Indirect discharges have to be understood as ‘‘the introduction into groundwater of substances in lists I or II after percolation through the ground or subsoil’’ while direct Table 3.1.1 List I

List II

Lists of substances regulated under Directive 80/68/EEC.

This list contains eight groups of substances, exception being made of substances which are considered inappropriate on the basis of low risk of toxicity, persistence and bioaccumulation: (1) organohalogen compounds and substances which may form such compounds in aquatic environment; (2) organophosphorus compounds; (3) organotin compounds; (4) substances which posses carcinogenic, mutagenic or teratogenic properties in or via the aquatic environment (if this is the case for certain substances of List II, they are included under this category); (5) mercury and its compounds; (6) cadmium and its compounds; (7) mineral oils and hydrocarbons; and (8) cyanides. This list contains individual or groups of substances which could have a harmful eﬀect on groundwater, in particular: (1) metalloids and metals and their compounds such as zinc, copper, nickel, chrome, lead, selenium, arsenic, antimony, molybdenum, titanium, tin, barium, beryllium, boron, uranium, vanadium, cobalt, thallium, tellurium, silver; (2) biocides and their derivatives not appearing in List I; (3) substances which have a deleterious eﬀect on the taste and/or odour of groundwater, and compounds liable to cause formation of such substances so as to render water unﬁt for human consumption; (4) toxic or persistent organic compounds of silicon, and substances which may cause the formation of such compounds in water, excluding those which are biologically harmless or are rapidly converted in water into harmless substances; (5) inorganic compounds of phosphorus and elemental phosphorus; (6) ﬂuorides; and (7) ammonia and nitrites.

European Union Groundwater Policy

87

discharges correspond to ‘‘an introduction without percolation.’’ Pollution is deﬁned as ‘‘The discharge by man, directly or indirectly, of substances or energy into groundwater, the results of which are such as to endanger human health or water supplies, harm living resources and the aquatic ecosystems or interfere with other legitimate use of water.’’ In this framework, consequences of pollution that has already occurred have to be checked or eliminated as far as possible (Article 1). This implies the following. With regard to List I substances, direct discharges are prohibited, whereas indirect discharges (due to disposal or tipping for the purpose of disposal) of these substances are prevented, which is linked to an authorisation procedure preceded by a thorough investigation on a caseby-case basis. In this respect, all appropriate measures have to be taken to prevent any indirect discharges due to either disposal or other activities on or in the ground other than disposal. With regard to List II substances, direct discharges have to be limited and appropriate measures have to be taken to limit any indirect discharges of these substances due to either disposal or other activities on or in the ground other than disposal. An authorisation procedure preceded by a thorough investigation is required in the case of direct discharge or disposal or tipping for the purpose of disposal of these substances. The authorisation is only granted if all the technical precautions for preventing groundwater pollution by these substances are observed. It should be noted that this directive does not apply to discharges of domestic eﬄuents from isolated dwellings not connected to a sewerage system and situated outside areas protected for the abstraction of water for human consumption. In addition, it does not apply to discharges of List I and II substances which are found in a quantity and concentration so small as to obviate any present or future danger of deterioration in the quality of the receiving groundwater, nor does it apply to discharges of matter containing radioactive substances. Another derogation clause concerns the authorisation of discharge of List I substances in groundwater which has been revealed as being permanently unsuitable for other uses (especially domestic or agricultural), providing that their presence does not impede exploitation of ground resources. These authorisations can only be granted if all technical precautions have been taken to ensure that these substances cannot reach other aquatic systems or harm other ecosystems. In addition, authorisation (after prior investigation) may be granted for discharges due to re-injection into the same aquifer of water used for geothermal purposes, water pumped out of mines and quarries or water pumped out for civil engineering works. Finally, artiﬁcial recharges for the purpose of groundwater management are subject to a special authorisation on a case-by-case basis, which may only be granted if there is no risk of groundwater pollution. The directive provides speciﬁc requirements regarding

88

Chapter 3.1

the authorisation procedures, distinguishing direct discharge and indirect discharge. In the above context, monitoring is required only for those speciﬁc cases of authorisation for the purpose of compliance checking and for assessing the eﬀects of discharges on groundwater. Application of measures relevant to this directive may on no account lead, either directly or indirectly, to pollution of groundwater. Finally, where appropriate, one or more member states may individually or jointly take more stringent measures than those provided for under this directive. From the above description, it can be concluded at ﬁrst sight that Directive 80/68/EEC ensures a stringent groundwater protection regime against pollution for all the activities that present a risk of groundwater deterioration through direct or indirect discharges of a wide range of pollutants. The implementation of this directive is, however, sometimes faced with the diﬃculties of a lack of groundwater quality data and objectives. In other words, infringement cases may be diﬃcult to judge in some instances in the absence of clear information on background groundwater quality levels in the zone aﬀected by discharges, and of quality objectives on the basis of which deterioration may unambiguously be identiﬁed. This directive will be repealed in 2013 under the WFD (2000/60/EC),2 after which the protection regime should be continued through the WFD and the new Groundwater Daughter Directive,3 which are further discussed below.

3.1.3

Preliminary Assessment (1982)

In 1982, the Directorate-General for the Environment, Consumer Protection and Nuclear Safety of the European Community carried out a major assessment of groundwater resources within its (then) nine member states. It consisted of a general survey (Groundwater Resources of the European Community: Synthetical Report) and individual reports from each member state.4 This report dealt with four major themes:

the aquifer inventory: location and type; groundwater hydrology: ﬂows within these aquifers; groundwater abstraction; and groundwater availability by area.

The study concluded that the Community had enough groundwater to meet most of its needs but that in most countries abstraction rates were already high: Belgium was abstracting some 70% of its available groundwater resources, Denmark 40%, France 25–50%, Italy 50%, Luxemburg 37%, the Netherlands 62% and the UK at least 25%. Only Ireland, which was abstracting some 3% of its groundwater resources, had major untapped resources. This assessment was mainly concerned with groundwater quantity. Since it was published, attention has turned in Europe (and the USA) to quality, and not only have groundwater quality monitoring programmes been greatly

European Union Groundwater Policy

89

expanded but many groundwater protection schemes have been put into place. These considerations have progressively led authorities to reﬂect on the need to develop a more integrated protection regime for groundwater, both quantitatively and qualitatively. These steps are summarised below.

3.1.4

The Groundwater Action Programme (1996)

The declaration of the ministerial seminar on groundwater held at The Hague in 1991 recognised the need for further action to avoid long-term deterioration of the quality and quantity of freshwater resources and called for a programme of actions to be implemented by the year 2000, aiming at sustainable management and protection of freshwater resources. The ﬁnal declaration stated that: groundwater is a natural resource with both ecological and economic value, which is of vital importance for sustaining life, health, agriculture and the integrity of ecosystems; groundwater resources are limited and should be protected on a sustainable basis; and it is essential to protect groundwater resources from over exploitation, adverse changes in hydrological systems resulting from human activities, and pollution, many forms of which can produce irreversible damage. The declaration stressed that the objective of sustainability should be implemented through an integrated approach, meaning that: surface water and groundwater should be managed as a whole, paying equal attention to both quality and quantity aspects; all interaction with soil and atmosphere should be taken into account; and water management policies should be integrated within the wider environmental framework as well as with other policies dealing with human activities such as agriculture, industry, energy, transport and tourism. Requests made by the Council in 1992 and 1995 in the form of resolutions recommended an action programme and a revision of Directive 80/68/EEC to be undertaken. This was followed up by the presentation by the Commission of a proposal for a Decision of the European Parliament and of the Council on an action programme for Integrated Protection and Management of Groundwater,5 which was adopted on 25 November 1996, and in which the Commission pointed to the need to establish procedures for the regulation of abstraction of freshwater and for the monitoring of freshwater quality and quantity. The action programme was designed along four main lines of action which are summarised below. (i) Action line 1. Planning and management principles: protection and use of groundwater is conceived here in the context of integrated planning and management of fresh water resources, i.e. groundwater is seen as an

90

Chapter 3.1

integral part of the hydrological cycle dynamically interacting with surface water in terms of quantity as well as quality aspects. Protection measures related to (over)exploitation, pollution prevention and integrated planning are described at member state and at Community levels. (ii) Action line 2. Abstraction of freshwater: in this action line, a rational regulatory framework for the abstraction of freshwater in large urban, industrial and agricultural areas, and tourist centres is discussed. Recommendations are expressed on how to secure an appropriate quantity management of groundwater and surface water within each river basin. Here again, actions are deﬁned at member state and Community levels. (iii) Action line 3. Diﬀuse sources of pollution: environmental challenges from diﬀuse sources of pollution are discussed, highlighting the diﬃculty of identifying individual polluters, in particular for groundwater pollution where the time-lag between application or release of polluting substances and the possibility for detection of their presence in the groundwater may span up to several decades. The action distinguishes diﬀuse sources from agricultural and industrial activities, traﬃc and urbanisation either through local impacts or long distance via atmospheric deposition. It provides recommendations to reduce and where possible avoid threats to groundwater from diﬀuse sources in the form of policy recommendations linked to environmental sustainability of agriculture, environmental challenges from nitrates and other mineral emissions and from the use of sewage sludge. (iv) Action line 4. Control of point-source pollution from activities and facilities which may aﬀect groundwater quality: completing action line 3, the recommendations focus on protection from activities and installations producing liquid and solid eﬄuent and/or representing a potential risk of accidental pollution of groundwater resources. Point sources are in principle traceable to speciﬁc activities and may be prone to measures at the source to avert or limit spreading of polluting substances. The action line refers to member states’ responsibilities with regard to authorisation regimes in place in parent legislation (e.g. the Integrated Pollution Prevention and Control Directive) and actions at Community level. The action programme clariﬁes the role of the Commission, and provides a framework for national action programmes, implementation and review. Considerations expressed in this communication coincided with the request made by the European institutions to the Commission to come forward with a proposal for a directive establishing a framework for a European water policy. They were hence naturally taken into account in the development of this large policy framework development, which resulted in the adoption of the WFD 2000/60/EC on 23 October 2000 (see Section 4). Some of them, however, have still not been taken on board, and are still to be considered in the integrated groundwater management under development within the WFD and the new Groundwater Directive.

European Union Groundwater Policy

3.1.5

91

The Groundwater Policy Framework Under the WFD

The WFD (Directive 2000/60/EC) took the 1996 action programme into account and was set to complement the Directive 80/68/EEC by stipulating that member states should implement the measures necessary to prevent or limit the input of pollutants into groundwater and to prevent the deterioration of the status of all bodies of groundwater. In this context, member states have to protect, enhance and restore all bodies of groundwater, ensure a balance between abstraction and recharge, with the aim to achieve good groundwater (chemical and quantitative) status by 2015, following the deﬁnitions given in Table 3.1.2. These requirements include a range of derogation clauses which are summarised in Table 3.1.3. An overview of the diﬀerent articles and annexes of the directive is given in Appendix I. The WFD also requires the implementation of measures necessary to reverse any signiﬁcant and sustained upward trend in the concentration of any

Table 3.1.2

Deﬁnitions of good quantitative and chemical status.

Ref. WFD

Good status

Good quantitative status (Annex V.2.1.2)

The level of groundwater in the groundwater body is such that the available groundwater resource is not exceeded by the long-term annual average rate of abstraction. Accordingly, the level of groundwater is not subject to anthropogenic alteration such as would result in: (a) failure to achieve the WFD environmental objectives for associated surface waters, (b) any signiﬁcant diminution in the status of such waters, and (c) any signiﬁcant damage to terrestrial ecosystems which depend directly on the groundwater body. Alterations to ﬂow direction resulting from level changes may occur temporarily, or continuously in a spatially limited area, but such reversals do not cause saltwater or other intrusion, and do not indicate a sustained and clearly identiﬁed anthropogenically induced trend in ﬂow direction likely to result in such intrusions. The chemical composition of the groundwater body is such that the concentrations of pollutants do not exhibit the eﬀects of saline or other intrusions (as determined by changes in conductivity) into the groundwater body, do not exceed the quality standards applicable under other relevant Community legislation in accordance with Article 17 of the WFD and are not such as would result in failure to achieve the WFD environmental objectives for associated surface waters nor any signiﬁcant diminution of the ecological or chemical quality of such bodies nor in any signiﬁcant damage to terrestrial ecosystems which depend directly on the groundwater body.

Good chemical status (Annex V.2.3.2)

92

Table 3.1.3

Chapter 3.1

Derogation clauses.

Article (WFD)

Derogation

4.4

Extensions may be granted when improvements of status cannot be reasonably achieved within the timescales for reasons of technical feasibility, disproportionate costs or natural conditions. The extension request has to be explained in the river basin management plan under Article 13, as well as a summary of measures required under Article 11 and the reason for the delay in making these measures operational. Extensions are limited to a maximum of two further updates of the RBMP (2027) except where the natural conditions are such that the objectives cannot be achieved within this period. Less stringent environmental objectives may be set out for speciﬁc bodies of water when they are so aﬀected by human activity (as determined by the analysis of pressure and impact under Article 5), or their natural condition is such that the achievement of these objectives would be infeasible or disproportionately expensive and that (a) the environmental and socioeconomic needs served by such human activity cannot be achieved by other means; (b) member states ensure the least possible changes to good groundwater status considering that impacts could not have reasonably been avoided due to the nature of the human activity of pollution; (c) no further deterioration of the aﬀected body of water occurs; (d) the establishment of less stringent objectives and the reasons for it are speciﬁed in the RBMP and this is reviewed every six years. Derogation also concerns temporary deterioration due to natural causes or force majeure which are exceptional or not foreseeable (e.g. extreme ﬂoods or droughts, accidents), providing that (a) all practicable steps to prevent further deterioration are taken, (b) the circumstances are declared in the RBMP, (c) measures are included in the programme of measures, (d) an annual review is undertaken and all practical restoration measures are taken in order to recover the initial status, and (e) a summary of eﬀects of the circumstances and measures are included in the next update of the RBMP. Member states will not be in breach of the directive when failure to achieve good groundwater status or to prevent deterioration in the status of a body of groundwater is the result of alterations to the level of bodies of groundwater, providing that (a) all practical steps are taken to mitigate the adverse impact on the status of the body of water, (b) the reasons for those alterations are set out and explained in the RBMP, (c) the reasons for those alterations are of overriding public interest and/or the beneﬁts of the alterations outweigh those of achieving the WFD environmental objectives, and (d) the beneﬁcial objectives served by these alterations cannot be achieved by other reasons for reasons of technical feasibility or disproportionate costs. The application of the above derogation clauses should not exclude or compromise the achievement of the directive objectives in other bodies of water within the same river basin district, and is consistent with the implementation of other Community environmental legislation.

4.5

4.6

4.7

4.8

European Union Groundwater Policy

93

pollutant resulting from the impact of human activity in order to progressively reduce groundwater pollution. Under this directive, the framework for groundwater protection imposes the following on member states. Delineate groundwater bodies within river basin districts to be designed and reported to the European Commission by member states, and characterise them through an analysis of pressures and impacts of human activity on the status of groundwater in order to identify groundwater bodies presenting a risk of not achieving WFD environmental objectives. This characterisation work had to be carried out in 2004–2005 and reported to the European Commission following requirements summarised in Tables 3.1.4 and 3.1.5. A report giving a synthesis of member states’ reports has been prepared by the European Commission and made available on the europa website in March 2007. Establish registers of protected areas within each river basin districts for those groundwater areas or habitats and species directly depending on water, which had to be carried out in 2004–2005. The registers have to include all bodies of water used for the abstraction of water intended for human consumption6 and all protected areas covered by the Bathing Water Directive 76/160/EEC,7 vulnerable zones under the Nitrates Directive 91/676/EEC8 and sensitive areas under the Urban Wastewater Directive 91/271/EEC,9 as well as areas designated for the protection of habitats and species including relevant Natura 2000 sites designated under Directives 92/43/EEC10 and 79/409/EEC.11 Registers should be reviewed under the River Basin Management Plan (RBMP; see below) updates. In this context, vulnerable zones are deﬁned as ‘‘all known areas of land in member states territories which drain into the waters aﬀected by pollution and waters which could be aﬀected by pollution and which contribute to pollution.’’ For these vulnerable zones, action programmes are required under the Nitrates Directive to reduce pollution caused or induced by nitrates and prevent further pollution. Based on the results of the characterisation phase, establish a groundwater monitoring network providing a comprehensive overview of groundwater chemical and quantitative status, and design a monitoring programme that had to be operational by the end of 2006. Monitoring will have to be reported, following requirements summarised below and detailed in Chapter 6.1. Set up a RBMP for each river basin district which will include a summary of pressures and impact of human activity on the groundwater status, a presentation in map form of monitoring results, a summary of the economic analysis of water use, a summary of the programme(s) of protection, control or remediation measures, etc. The ﬁrst RBMP is scheduled to be published at the end of 2009. A review is then planned by the end of 2015, and every six years thereafter. By 2010, take account of the principle of recovery of costs for water services, including environmental and resource costs, having regard to

94

Chapter 3.1

Table 3.1.4

Characterisation.

Annex II.2 (WFD)

Characterisation

Initial characterisation (par. 2.1)

The initial characterisation concerns all groundwater bodies, assessing their uses and the degree at which they are at risk to meet WFD environmental objectives. This analysis may use existing hydrological, geological, pedological, land use, discharge, abstraction and other data, identifying: the location and boundaries of the groundwater body or groups of bodies, the pressures to which the groundwater is subject to (diﬀuse and point sources of pollution, abstraction, artiﬁcial recharge), the general character of the overlying strata in the catchment area from which the groundwater body receives its recharge, and those groundwater bodies for which there are directly dependent surface water ecosystems or terrestrial ecosystems. It concerns the groundwater (or groups of) bodies which have been identiﬁed as being at risk, and aims to establish a more precise assessment of the signiﬁcance of such risks and the identiﬁcation of any measures top be required under the WFD Article 11. This characterisation has to include relevant information on the impact of human activity and, where relevant, on geological characteristics of the groundwater body (including the extent and type of geological units), hydrogeological characteristics (including hydraulic conductivity, porosity and conﬁnement), characteristics of the superﬁcial deposits and soils in the catchment from which the groundwater body receives its recharge (including the thickness, porosity, hydraulic conductivity, and adsorptive properties of the deposits and soils), stratiﬁcation characteristics of the groundwater, an inventory of associated surface systems (including terrestrial ecosystems and bodies of surface water, with which the groundwater body is dynamically linked), estimates of the direction and rates of exchange of water between the groundwater body and associated water systems, suﬃcient data to calculate the long-term annual average rate of overall recharge, and characterisation of the chemical composition of the groundwater (including speciﬁcation of the contribution from human activity—member states may use typologies for groundwater characterisation when establishing natural background levels for these bodies of groundwater).

Further characterisation (par. 2.2)

the economic analysis conducted under Article 5 of the WFD, and in accordance with the polluter pays principle. Establish a programme of measures for achieving WFD environmental objectives (e.g. abstraction control, prevent or control pollution measures) by the end of 2009, to be operational by the end of 2012. Basic

European Union Groundwater Policy

Table 3.1.5

95

Review of impacts on groundwater and authorisations.

Annex II.2 (WFD)

Reviews of impacts

Impact of human activity (par. 2.3)

For transboundary groundwater bodies (crossing the borders of two or more member states) or bodies identiﬁed at risk following the initial characterisation, additional information, where relevant, will have to be collected and maintained for each groundwater body: (a) location of points in the groundwater body used for the abstraction of water (with the exception of points providing less that 10 m3/day or points for abstraction of water intended for human consumption providing less than 10 m3/day or serving less than 50 persons); (b) the annual average rates of abstraction from such points; (c) the chemical composition of water abstracted from the groundwater body; (d) the location of points in the groundwater body into which water is directly discharged; (e) the rates of discharges at such points; (f) the chemical composition of discharges to the groundwater body; and (g) land use in the catchment (or catchments) from which the groundwater body receives its recharge, including pollutant inputs and anthropogenic alterations to the recharge characteristics such as rainwater and run-oﬀ diversion through land sealing, artiﬁcial recharge, damming or drainage. Bodies for which lower objectives are to be speciﬁed (see Table 3.1.2) have to be identiﬁed by member states, including consideration of the eﬀects of the status of the body on (i) surface water and associated terrestrial ecosystems, (ii) water regulation, ﬂood protection and land drainage, and (iii) human development. Similarly, bodies of groundwater for which lower objectives are to be speciﬁed under Article 4.5 of the WFD (see Table 3.1.2) have to be identiﬁed as a result of the analysis of impact of human activity (Article 5.1).

Impacts of change in groundwater levels (par. 2.4)

Impact of pollution on groundwater quality (par. 2.5)

Article (WFD)

Authorisations

11.3( j)

Authorisations concern: (a) reinjection into the same aquifer of water used for geothermal purposes; (b) injection of water resulting from hydrocarbon extraction or mining activities into geological formations which for natural reasons are permanently unsuitable for other purposes; (c) reinjection of pumped groundwater from mines and quarries or associated with the construction or maintenance of civil engineering works; (d) injection of gas or liqueﬁed petroleum for storage purposes into geological formations which for natural reasons are permanently unsuitable for other purposes, or where there is an overriding need for security of gas supply and where the injection is such as to prevent future deterioration of the receiving groundwater; (e) construction, civil and building works or similar activities on or in the ground which come into contact

96 Table 3.1.5

Chapter 3.1 (continued )

Article (WFD)

Authorisations with groundwater, in accordance with general binding rules developed by the member states; (f) discharges of small quantities of substances for scientiﬁc purposes, providing that such discharges do not compromise the achievement of environmental objectives established for that body of groundwater.

measures include, in particular, controls over the abstraction of groundwater, controls (with prior authorisation) of artiﬁcial recharge or augmentation of groundwater bodies (providing that it does not compromise the achievement of environmental objectives). Point-source discharges and diﬀuse sources liable to cause pollution are also regulated under the basic measures. Direct discharges of pollutants into groundwater are prohibited subject to a range of provisions summarised in Table 3.1.5. The programme of measures has to be reviewed and if necessary updated by 2015 and every six years thereafter. Strategies to prevent and control pollution of groundwater are covered by Article 17 of the WFD, which requires the establishment of criteria for assessing good groundwater chemical status and for the identiﬁcation of signiﬁcant and sustained upward trends and for the deﬁnition of starting points for trend reversals, considering the following. The characterisation of bodies of groundwater as detailed in Annex II.2 of the WFD (see Table 3.1.4). Good status deﬁnitions as detailed in Table 3.1.2, which are based on groundwater level regime (quantitative status) and conductivity and concentrations of pollutants (chemical status). Monitoring requirements to respond to the needs of obtaining a comprehensive overview of groundwater status and to detect the presence of long-term anthropogenically induced upward trends in pollutants. In this respect, surveillance monitoring is aimed at supplementing and validating the impact assessment procedure (carried out under Article 5 of the WFD) and provide information for use in the assessment of long-term trends both as a result of changes in natural conditions and through anthropogenic activity, while operational monitoring should be undertaken in the periods between surveillance monitoring programmes in order to establish the chemical status of all groundwater bodies or groups of bodies determined as being at risk and to establish the presence of any long-term anthropogenically induced upward trend in the concentration of any pollutant. Further details are given in Section 5. Monitoring results should be used to identify long-term anthropogenically induced upward trends in pollutant concentrations and to set up starting points for reversing these trends.

European Union Groundwater Policy

97

Article 17 requests the European Commission to present a proposal based on the above requirements. This proposal of a new groundwater directive was issued in 20033 and resulted in the new Groundwater Directive of which the main orientations are described in the next section.

3.1.6

The New Groundwater Directive 2006/118/EC

While quantitative status requirements are clearly covered by the WFD, it does not include, however, speciﬁc provisions on chemical status, i.e. the diﬀerent conceptual approaches to groundwater protection did not allow achieving an agreement on detailed provisions within the WFD at the conciliation. As mentioned in Section 5, this justiﬁed including a provision, Article 17, requesting the Commission to come forward with a proposal of speciﬁc measures to prevent and control groundwater pollution. This proposal was adopted by the Commission on 19 September 2003 (COM(2003)550 ﬁnal) and was adopted after a conciliation phase among the European Parliament and the Council on 12 December 2006.12 Directive 2006/118/EC is based on the following three main pillars. (i) Criteria linked to good chemical status evaluation, which are based on compliance with EU existing environmental quality standards (nitrates, plant protection products and biocides) and to ‘‘threshold values’’ (playing the same role as EQS) for pollutants representing a risk to groundwater bodies. The latter category of standards has to be established by member states, using common methodological criteria (which were the basis of research developments under the BRIDGE project; see Chapter 9.1), at the most appropriate scale (national, regional or local), taking account hydrogeological conditions, soil vulnerability, types of pressures, etc. They will have to be reported to the Commission by the end of 2008, and will be used as quality objectives for further compliance checking. (ii) Criteria for the identiﬁcation of sustained upward trends of pollutants in groundwater bodies characterised as being at risk. These include measurement principles and requirements regarding trend reversals. (iii) Requirements on the prevention/limitation of pollutant inputs to groundwater, which will ensure a continuity of Directive 80/68/EEC after its repeal in 2013, i.e. the same principle of prevention of hazardous substances introduction and limitation of other pollutants so as to avoid pollution will apply. Other elements concern clariﬁcations about the groundwater use as drinking water (albeit this is well covered by Article 7 of the WFD) and its relation with the present directive, which relates to WFD environmental objectives. Recommendations to undertake research on groundwater ecosystems are also expressed in a recital, which illustrates the awareness for required scientiﬁc integration. Finally, review of technical annexes of the directive (in particular concerning the establishment of groundwater threshold values and methods for

98

Chapter 3.1

identifying and reversing pollution trends) is requested, taking into account scientiﬁc progress, before the end of 2012 and every six years thereafter. This review will have to be carried out following ‘‘comitology’’ rules, i.e. adoption of possible decisions by a regulatory committee composed of member states. Since 2006, these rules imply that the European Parliament will have right of scrutiny on adopted decisions. An evaluation of the functioning of the directive in the light of consistency with parent legislation (see Section 7) is also foreseen. In summary of Sections 2, 5 and 6, the WFD2 (including the new ‘‘daughter’’ groundwater directive3) will complement and ensure a continuity of the Directive 80/68/EEC protection regime.1 This will be achieved through a systematic analysis of pressures and impacts (not done under Directive 80/68/EEC), and requirements related to good chemical status and pollutant trend identiﬁcation/ reversal backed up by surveillance and monitoring programmes. The programme of measures also sets out provisions that are aimed at replacing the existing protection regime. The new groundwater directive aims to provide the necessary common criteria regarding chemical status evaluation, identiﬁcation and reversal of signiﬁcant and upward trends in pollutant concentrations, as well as speciﬁc clauses regarding direct and indirect inputs of pollutants into groundwater to make sure that the existing protection regime will be appropriately strengthened. An overview of the diﬀerent articles and annexes of the directive is given in Appendix II.

3.1.7

Policy Integration

One of the key aspects of environmental integration in the light of the EU Sustainable Development Strategy is linked to policy coordination and integration. With regard to the groundwater policy framework, this policy integration appears to be quite complex since it concerns a range of various directives as illustrated in Figure 3.1.1. This section examines how various relevant directives interact with the groundwater policy under the WFD and Directive 80/68/EEC.

3.1.7.1

Nitrates Directive

The Nitrates Directive8 aims to reduce water pollution caused or induced by nitrates from agricultural sources and to prevent further such pollution. It obliges member states to designate vulnerable zones which correspond to all known areas of land in member state territories which drain into the waters (including groundwater) aﬀected by pollution and waters which could be aﬀected by pollution and which contribute to pollution. A reference is made to action programmes to reduce pollution caused or induced by nitrates and to prevent further pollution, and to requirements for identifying groundwater vulnerable zones as ‘‘those waters which contain more than 50 mg l 1 or could contain more than 50 mg l 1 nitrates if an action programme is not undertaken.’’ The link with groundwater policy is clear in that respect, i.e. nitrate

99

European Union Groundwater Policy Policy integration Seveso, IPPC

Env. Impact assessment Birds, Habitats

URBAN SECTOR

DRINKING WATER

AGRICULTURE

INDUSTRY

Drinking water

Sewage sludge landfill Urban waste water, Construction Products Bathing water

Pesticides, Nitrates, biocides

groundwater

WFD

Figure 3.1.1

The overall groundwater policy framework: integration needs.

contamination levels should not be over the trigger value set at 50 mg l 1 (this argument has been used for proposing this value as an EU groundwater quality standard for groundwater in the new Groundwater Directive (see Section 6). The Nitrates Directive8 requires the implementation of suitable monitoring programmes to assess the eﬀectiveness of action programmes at selected measuring points, making it possible to establish the extent of nitrate pollution in the waters from agricultural sources. The designation and monitoring of vulnerable zones is to be carried out at regular intervals at sampling stations which are representative of groundwater aquifers, taking into account the provisions of the Drinking Water Directive.6 The monitoring has to be repeated at least every four years, except for those sampling stations where the nitrate concentration in all previous samples has been below 25 mg l 1 and no new factor likely to increase the nitrate content has appeared (in which case the monitoring programme needs to be repeated only every eight years). The directive also stipulates that reference methods of measurement have to be used. This, however, concerns freshwaters, coastal waters and marine waters (i.e. no speciﬁc mention is made of groundwater).

3.1.7.2

Urban Wastewater Treatment Directive

The Urban Wastewater Directive9 aims to protect the environment from the adverse eﬀects of discharges of urban wastewater and wastewater from certain industrial sectors. In this context, the identiﬁcation of ‘‘sensitive areas’’ relates essentially to freshwaters, estuaries or coastal waters which are found to be eutrophic, lakes and streams reaching lakes/reservoirs/closed bats with poor water exchange and surface freshwaters intended for the abstraction of drinking water which could contain more than 50 mg l 1 nitrates. This directive is

100

Chapter 3.1

indirectly relevant to groundwater (protection of receiving groundwaters from possibly contaminated wastewaters originating from freshwater sources). The Urban Wastewater Directive9 monitoring obligations are directly related to veriﬁcations of appropriate treatment, prior regulations and/or speciﬁc authorisations of discharges from urban waste treatment plants to freshwaters and estuaries. In the framework of this directive, monitoring will have to focus on discharges from urban wastewater treatment plants to verify compliance with requirements set out in the directive (corresponding to criteria concerning diﬀerent types of discharges) and following control procedures laid down in the annex (reference monitoring methods and evaluation of results). These requirements focus on ﬂow-proportional or time-based 24-hour sample collection at welldeﬁned points in the wastewater treatment plant outlet and if necessary in the inlet in order to monitor compliance with the directive’s requirements for discharged wastewater. They include an obligation to apply good international laboratory practices in order to minimise the degradation of samples between collection and analysis. Note that these monitoring obligations do not concern groundwater.

3.1.7.3

Plant Protection Products Directive

The Plant Protection Products Directive13 concerns the authorisation, placing on the market, use and control within the Community of plant protection products in commercial form. Regarding groundwater, authorisations are only granted when plant protection products have no harmful eﬀect on humans or human health, directly or indirectly, or on groundwater, and they have no unacceptable inﬂuence on the environment, particularly contamination of water including drinking water and groundwater. The ‘‘uniform principles’’ set out in the directive specify that no authorisation shall be granted if the concentration of the active substance or of relevant metabolites, degradation or reaction products in groundwater, may be expected to exceed, as a result of use of the plant protection product under the proposed conditions of use, the lower of (i) the maximum permissible concentration laid down by Directive 80/778/ EEC,7 or (ii) the maximum concentration laid down by the Commission when including the substance listed in the directive, on the basis of appropriate data (in particular toxicological data), or where that concentration has not been laid down, the concentration corresponding to one tenth of the ADI (acceptable daily intake) laid down when the active substance was included in the directive. The monitoring obligations concern the authorisation regime imposed by the member states according to the directive’s provisions. Decision-making provisions are included in the annex to the directive. The granting of authorisations has to take account of the agricultural, plant health or environmental (including climatic) conditions in the areas of envisaged use (this implicitly concerns groundwater, even if this is not speciﬁcally mentioned). These considerations may result in speciﬁc conditions and restrictions of use and, where necessary, in authorisation being granted for some but not other areas within the member state. The control measures are obviously linked to the current analytical knowledge (and authorisation may be limited to a limited period if limitations

European Union Groundwater Policy

101

in analytical science and technology are recognised), with requirements regarding the method’s reproducibility. As mentioned in Section 6, the directive makes a direct reference to groundwater contamination (with drinking water standards not allowed to be exceeded), which therefore requires monitoring. There are no speciﬁc monitoring criteria in this respect other than the mention that analytical methods must reﬂect the state of the art, and analytical criteria on the method’s performance as set out in the annex.

3.1.7.4

Biocides Directive

The Biocides Directive14 concerns the authorisation and the placing on the market for use of biocidal products. Similarly to Directive 91/414/EEC,13 authorisation of biocidal products may only be granted if the products have no harmful eﬀect on humans or human health, directly or indirectly, or on groundwater, and they have no unacceptable inﬂuence on the environment, particularly contamination of water including drinking water and groundwater. Similar principles as the ‘‘uniform principles’’ of Directive 91/414/EEC are set out, which means that the 0.1 mg l 1 quality standard of 80/778/EC6 plays a role of maximum concentration for all groundwater, but that lower standards may be established following the procedure for including the active substance in Annex I of the directive. The decision-making provisions of the annex to the Biocides Directive14 follow the same lines as that described above (related to the Plant Protection Products Directive) with respect to groundwater. Monitoring obligations are closely linked to the authorisation regime which requests a prior risk assessment for which criteria are deﬁned in the evaluation provisions of the same annex. This risk assessment has to take into account any adverse eﬀects arising in any of the three environmental compartments—air, soil and water (including sediment)—and of the biota. The analytical work has, therefore, to focus on the properties and potential adverse eﬀects of the active substances present in the biocidal product for its classiﬁcation. In case this classiﬁcation is not possible, information on bioaccumulation potential, persistence characteristics, information from toxicity studies, etc., have to be taken into account. If appropriate, adequately measured exposure data, likely pathways to environmental compartments, potential for adsoption/desorption and degration, etc., have to be evaluated. This obviously includes eﬀects on groundwater. Speciﬁc monitoring requirements are, however, not included, except the mention that testing should be carried out according to Community guidelines if these are available and applicable. Where appropriate, other methods can be used (e.g. ISO, CEN or other international standard method, national standard method or other methods accepted by the member state) and if relevant ﬁeld data exist, these can also be used.

3.1.7.5

IPPC Directive

The Integrated Pollution Prevention and Control (IPPC) Directive15 concerns integrated pollution prevention and control, which lays down measures designed to prevent or reduce emissions in the air, water and land from a range of

102

Chapter 3.1

activities listed in Annex I of the directive. It establishes provisions for issuing permits for existing and new installations, and makes a speciﬁc reference to groundwater, indicating that the permits shall include appropriate requirements ensuring protection of the soil and groundwater on the basis of emission limit values for pollutants which may be supplemented or replaced by equivalent parameters or technical measures based on best available techniques. The permit procedure under the IPPC Directive15 includes a provision for suitable release monitoring, specifying measurement methodology and frequency, evaluation procedure and obligation to supply data required for checking compliance with the permit. The directive includes a provision for installations that may have signiﬁcant negative eﬀects on the environment of another member state. Monitoring is focused on the releases, the results of which have to be regularly submitted by the operator to the competent authority (and without delay in the case of any incident or accident signiﬁcantly aﬀecting the environment). There are no speciﬁc monitoring requirements for groundwater, but the directive’s provisions obviously imply that risks to groundwater be appropriately monitored.

3.1.7.6

Landﬁll Directive

The Landﬁll Directive16 concerns the landﬁll of waste, which aims to provide for measures, procedures and guidance to prevent or reduce as far as possible negative eﬀects on the environment, including groundwater. Similarly to the IPPC Directive,15 the directive establishes provisions for issuing permits based on a range of conditions including impact assessment studies. Regarding groundwater, site characteristics have to locate groundwater and geological and hydrogeological conditions in the area, prevent groundwater from entering into the landﬁlled waste, take appropriate measures to collect/treat contaminated water and leachate and prevent pollution of the soil, groundwater or surface water using appropriate technical precautions (e.g. combination of geological barrier and bottom liner). The directive establishes criteria for waste testing and acceptance, taking due consideration on the protection of the surrounding environment, including groundwater. The Landﬁll Directive16 imposes control and monitoring procedures with a frequency which is to be deﬁned by the competent authority (and in any event at least once a year) and on the basis of aggregated data, in order to demonstrate compliance with permit conditions. The corresponding article notiﬁes that the quality control of the analytical operations of the control and monitoring procedures are carried out by competent laboratories. Further requirements are provided in the annex. They include reporting obligations for meteorological data (volume of precipitation, temperature, wind, evaporation, atmospheric humidity) to check whether leachate is building up in the landﬁll body or whether the site is leaking. Sampling of leachate and surface water is also required to be collected at representative points (for surface water, no less than two points, i.e. one upstream from the landﬁll and one downstream). A separate section on the protection of groundwater is included, which requests the provision of information on groundwater likely to be aﬀected by the discharging

European Union Groundwater Policy

103

of waste, with at least one measuring point in the groundwater inﬂow region and two in the outﬂow region (this number can be increased on the basis of a speciﬁc hydrogeological survey and the need for an early identiﬁcation of accidental leachate release in the groundwater). Sampling has to be carried out in at least three locations before the ﬁlling operations in order to establish reference values for future sampling (following the requirement of the ISO 5667 standard Sampling Groundwaters, Part 11, 1993). The parameters to be analysed in the samples taken must be derived from the expected composition of the leachate and the groundwater quality in the area, with account of mobility in the groundwater zone and a frequency adapted to the local conditions. Adverse eﬀects are considered to have occurred when an analysis of the groundwater sample shows a signiﬁcant change in water quality as deﬁned by a trigger level which should be determined by the competent authority (taking account of the speciﬁc hydrogeological formations in the location of the landﬁll and groundwater quality) and laid down in the permit whenever possible.

3.1.7.7

Sewage Sludge Directive

The Sewage Sludge Directive17 seeks to encourage the use of sewage sludge in agriculture and to regulate its use in such a way as to prevent harmful eﬀects on soil, vegetation, animals and humans. To this end, it prohibits the use of untreated sludge on agricultural land unless it is injected or incorporated into the soil. Treated sludge is deﬁned as having undergone ‘‘biological, chemical or heat treatment, long-term storage or any other appropriate process so as signiﬁcantly to reduce its fermentability and the health hazards resulting from its use.’’ The directive also requires that sludge should be used in such a way that account is taken of the nutrient requirements of plants and that the quality of the soil and of the surface and groundwater is not impaired. It sets out requirements for the keeping of detailed records of the quantities of sludge produced, the quantities used in agriculture, the composition and properties of the sludge, the type of treatment and the sites where the sludge is used. Limit values for concentrations of heavy metals in sewage sludge intended for agricultural use and in sludge-treated soils are given in annexes to the directive. In the framework of the Sewage Sludge Directive,17 monitoring requirements are focused on speciﬁes rules for the sampling and analysis of sludges and soils, i.e. there are no speciﬁc requirements concerning groundwater.

3.1.7.8

Other Directives

Article 4 of the WFD18 also requires that waste be recovered or disposed of without endangering the environment, which may have an (indirect) eﬀect on protecting groundwater. Finally, the Construction Product Directive19 concerns regulatory provisions for construction products. It indirectly concerns groundwater in that construction products for construction works have to be ﬁt for their intended

104

Chapter 3.1

use and respond to requirements regarding hygiene, health and the environment; in particular there should be no threat to the hygiene or health of occupants or neighbours as a result of pollution or poisoning of the water or soil. The directive focuses on conformity aspects of construction products, taking into account possible risk to water environments (in particular release of dangerous substances to water). As such, the directive does not provide for speciﬁc groundwater monitoring other than the requirement than a veriﬁcation that the construction work is designed and built in such a way that it will not generate pollution of the water or soil. Conformity testing of construction products is generally carried out at the factory or on site from a batch which is ready for delivery. The surveillance concerns the factory production and product testing rather than monitoring possible eﬀects on the environment.

3.1.8

Conclusions: The Need for Worldwide Cooperation

Groundwater protection regimes against pollution and overexploitation are established worldwide either in the form of national laws or regional, national or international conventions, as illustrated by a recent compilation of treaties and other legal instruments produced by FAO/UNESCO.20 These regulatory frameworks should ensure that appropriate actions are being undertaken to protect, enhance and restore the quantitative and quality status of groundwaters. However, much remains to be done to actually eﬀectively implement the diﬀerent legal instruments. The UN International Law Commission (ILC), which is the UN body in charge of the progressive development and the codiﬁcation of international law, adopted at ﬁrst reading in June 2006 a set of draft articles on the law of transboundary aquifers.21 The ILC has sent the draft articles to states, requesting them to provide comments and observations by 1 January 2008. In Europe, eﬀorts linked to the WFD Common Implementation Strategy22 enable exchange of experiences, discussion of diﬃculties and ﬁnding possible solutions in the framework of an international participatory approach (see Chapter 4.1). This certainly constitutes an excellent platform for implementing groundwater protection under the WFD, the new Groundwater Directive and parent legislation at the EU scale, but this will require a constant coordination and accompanying eﬀorts to make this happen. The example could be extended to international cooperation, and initiatives are already under way, e.g. in the Mediterranean basin, leading to shared recommendations.23 Finally, international programmes such as the UNESCO International Hydrological Programme24 oﬀer excellent opportunities to disseminate expertise and experiences, and participate in worldwide harmonisation of groundwater protection regimes and education. UNESCO-IHP has contributed to the work of ILC mentioned above on the draft articles on transboundary aquifers by providing the Special Rapporteur with scientiﬁc and technical support and assistance on the science of hydrogeology (more details can be found at www.isarm.net).

European Union Groundwater Policy

105

References 1. Council Directive 80/68/EEC of 17 December 1979 on the protection of groundwater against pollution, Oﬃcial Journal of the European Communities L 20, 26.1.1980, p. 43. 2. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities L 327, 22.12.2000, p. 1. 3. Proposal for a Directive of the European Parliament and of the Council on the protection of groundwater against pollution, COM(2003)550 ﬁnal. 4. Groundwater Resources of the European Community: Synthetical Report, European Commission, 1982. 5. Proposal for a European Parliament and Council Decision on an action programme for integrated groundwater protection and management, Oﬃcial Journal of the European Communities C 355, 25.11.1996, p. 1. 6. Council Directive 80/778/EEC of 15 July 1980 relating to the quality of water intended for human consumption, Oﬃcial Journal of the European Communities L 229, 5.12.1998, p. 32. 7. Council Directive 76/160/EEC of 8 December 1975 concerning the quality of bathing water, Oﬃcial Journal of the European Communities L 31, 5.2.1976, p. 1. 8. Council Directive 91/676/EEC of 12 December 1991 concerning the protection of waters against pollution caused by nitrates from agricultural sources, Oﬃcial Journal of the European Communities L 375, 31.12.1991, p. 1. 9. Council Directive 91/271/EEC of 21 May 1991 concerning urban waste treatment, Oﬃcial Journal of the European Communities L 135, 30.5.1991, p. 40. 10. Habitats Directive 92/43/EEC, Oﬃcial Journal of the European Communities L 206, 22.7.1992, p. 7. 11. Birds Directive 79/409/EEC, Oﬃcial Journal of the European Communities L 103, 25.4.1979, p. 1. 12. Directive of the European Parliament and of the Council of 12 December 2006 on the protection of groundwater against pollution and deterioration, Oﬃcial Journal of the European Communities, L 372, 12.12.2006, p. 19. 13. Council Directive of 15 July 1991 concerning the placing of plant protection products on the market, Oﬃcial Journal of the European Communities L 230, 19.8.1991, p. 1. 14. Directive 98/8/EC of the European Parliament and of the Council of 16 February 1998 concerning the placing of biocidal products on the market, Oﬃcial Journal of the European Communities L 123, 24.4. 1998, p. 1. 15. Council Directive of 24 September 1996 concerning integrated pollution prevention and control, Oﬃcial Journal of the European Communities L 257, 10.10.1996, p. 26.

106

Chapter 3.1

16. Council Directive 99/31/EC of 26 April 1999 on the landﬁll of waste, Oﬃcial Journal of the European Communities L 182, 16.7.1999, p. 1. 17. Sewage Sludge Directive 86/278/EEC, Oﬃcial Journal of the European Communities L 181, 8.7.1986, p. 6. 18. Waste framework directive 75/442/EEC amended by Directive 91/156/ EEC. 19. Council Directive of 21 December 1988 on the approximation of laws, regulations and administrative provisions of the member states relating to construction products, Oﬃcial Journal of the European Communities L 040, 11.12.1989, p. 12. 20. S. Burchi and K. Mechlem, Groundwater in International Law, FAO Legislative Study, FA0/UNESCO, 86, 2005. 21. Report of the International Law Commission, 58th session, A/61/10, 2006, p. 185 (http://www.un.org/law/ilc/). 22. Common Implementation Strategy for the Water Framework Directive, European Communities, 2003 (ISBN 92-894-2040-5). Final CIS document available at: http://europa.eu.int/comm/environment/water/waterframework/implementation.html. 23. Technical Report on Groundwater Management in the Mediterranean and the Water Framework Directive (http://www.emwis.org/GroundwaterHome. htm). 24. UNESCO, International Hydrological Programme (http://www.unesco. org/water).

CHAPTER 3.2

US Drinking Water Regulation: The Ground Water Rulew CRYSTAL RODGERS-JENKINS US EPA, 1201 Constitution Avenue NW, MC-4607M, Washington, DC 20460, USA

3.2.1

Introduction

The United States Environmental Protection Agency (EPA) ﬁnalised the Ground Water Rule (GWR) in 2006.1 The EPA established the GWR due to concerns related to faecal contamination in groundwater systems (GWSs). The occurrence of faecal contamination in groundwater sources is an indication that viral and bacterial pathogens may also be present. The goal of the GWR is to provide Americans that drink water from public groundwater sources with water that is safe for human consumption. The GWR is the ﬁrst federal drinking water regulation in the USA that requires GWSs, not under the inﬂuence of surface water, to monitor their groundwater sources for faecal indicators. In addition to source water monitoring, the GWR requires periodic sanitary surveys of GWSs to inspect for signiﬁcant deﬁciencies that may lead to contamination. The rule requires GWSs that have an indication of faecal contamination and GWSs with signiﬁcant deﬁciencies to take corrective action to ensure that the drinking water from groundwater sources is safe for human consumption. Prior to the GWR, the only microbial US drinking water regulations that applied to GWSs were the Total Coliform Rule (TCR) and, in part, the Surface Water Treatment Rule (SWTR). The TCR applied to all public water systems and the SWTR applied to public water systems using surface waters or groundwaters under the direct inﬂuence of surface water. The TCR and SWTR were published in 1989.2,3 The TCR requires that each GWS monitor its distribution system for total coliforms and faecal coliforms/E. coli; no monitoring w

The views expressed in this chapter are those of the individual author and do not necessarily reﬂect the views and policies of the US Environmental Protection Agency.

107

108

Chapter 3.2

is required at the groundwater source. All GWSs are required to comply with the TCR provisions, although for some GWSs with pristine groundwater sources—for which the system can demonstrate it has a protected groundwater source and has no history of total coliform presence—the State may allow a reduced distribution system monitoring frequency. Under the TCR, periodic sanitary surveys must also be performed at small GWSs;2 however, the frequency and scope of the TCR sanitary surveys diﬀer from the GWR requirements. The sanitary survey frequency and scope are discussed in Section 3.2.4. The SWTR requires that all GWSs under the direct inﬂuence of surface water to disinfect and achieve at least 3- and 4-log removal and/or inactivation of Giardia and viruses, respectively. Since 1989 the EPA has ﬁnalised ﬁve other microbial drinking water rules for surface water systems: the Information Collection Rule,4 the Interim Enhanced Surface Water Treatment Rule,5 the Filter Backwash Recycling Rule,6 the Long Term 1 Enhanced Surface Water Treatment Rule7 and the Long Term 2 Enhanced Surface Water Treatment Rule.8

3.2.2

Federal Statutory Authority

The GWR was established pursuant to the 1996 amendments to the Safe Drinking Water Act (SDWA). In the 1996 SDWA amendments, US Congress authorised the EPA to require disinfection, if necessary, as a treatment technique for GWSs and to establish criteria for determining when disinfection is needed. The GWR applies to all public water systemsz that use groundwater sources, in whole or in part (including consecutive systems that receive ﬁnished groundwater from another public water system). Public water systems that combine all of their groundwater with surface water prior to treatment at a surface water treatment plant or groundwater under the direct inﬂuence of surface water are not required to comply with GWR. These systems must comply with the Surface Water Treatment Rule.3 Private wells are not regulated under the GWR or by the EPA.

3.2.3

Challenges in Developing the GWR

In 2000, the EPA published the proposed GWR9 for public review. The EPA received several thousand comments on the proposed regulation. The EPA, water industry and States shared a common interest: public health protection. However, the three entities had various perspectives on which requirements would be necessary to protect public health. In general, the water industry focused on minimal capital and operational and maintenance costs which translates to fewer rule requirements; and States—oftentimes struggling with implementing existing regulations due to limited personnel and ﬁnancial z

Public water systems are systems that provide drinking water to the public through pipes or other constructed conveyances, if such system has at least 15 service connections or regularly serves an average of at least 25 people per day at least 60 days out of the year.

US Drinking Water Regulation: The Ground Water Rule

109

constraints—emphasised the need for greater ﬂexibility. The EPA’s goal in developing the GWR was to meet the federal statutory requirement through a scientiﬁcally sound, ﬂexible and cost-eﬀective regulation. The EPA met numerous challenges in achieving this goal. Some of the main challenges were: the number of GWSs: approximately 150 000 systems; the number of very small GWSs: more than 60% of the GWSs serve r500 people; the lack of a single, comprehensive national microbial occurrence study of GWSs in the USA; and limited data on public health risks related to exposure to microbial pathogens in groundwater.

3.2.3.1

US Groundwater System Demographics

A total of approximately 150 000 GWSs are unevenly spread across the country’s extensive (over 3.7 million square miles (over 9.6 million km2))10 and extremely varied landscape. The water quality of the public wells that provide groundwater to GWSs also diﬀers throughout the US regions. In some geographical areas in the USA, the raw groundwaters from public wells are clean and safe to drink or the GWSs may need to provide very little treatment prior to human consumption. US groundwater occurrence data and information on the waterborne disease outbreak in groundwater systems suggest this is not the case for other areas in the USA. The nearly 150 000 GWSs supply drinking water to approximately 114 million people. The EPA divides public water systems into three categories: community water systems (CWS), non-transient non-community water systems (NTNCWS) and transient non-community water systems (TNCWS).y Nontransient non-community water systems and transient non-community water systems account for about 70% of the total number of GWSs in the USA. A relatively small percentage of the population (13%) consumes groundwater supplied by NTNCWSs or TNCWSs. About 70% of NTNCWS and TNCWSs serve o100 people and over 60% of CWSs serve o500 people. During implementation of the GWR, GWSs that serve very small populations (e.g. o500 people) will have the greatest economic impact compared to larger GWSs due to economies of scale. The GWR requires systems of all sizes to adhere to rule requirements intended to ensure the same level of public health protection. The EPA addressed the challenges of the large number of GWSs and the substantial percentage of very small GWSs through GWR’s risk-targeted strategy. This strategy targets the subset of the total number of GWSs that are susceptible to y

Community water systems serve at least 25 people year round. Non-transient non-community water systems provide water to the same 25 people for at least six months per year (e.g. schools). Transient non-community water systems serve 25 persons per day for 60 days per year (e.g. restaurants).

110

Chapter 3.2

Table 3.2.1

Ground Water Rule population and system baselines.

System size (population served) CWS r100 101–500 501–1000 1001–3300 3301–10 000 10 001–50 000 50 001–100 000 100 001–1 million 41 million National total NTNCWS r100 101–500 501–1000 1001–3300 3301–10 000 10 001–50 000 50 001–100 000 100 001–1 million 41 million National total TNCWSs r100 101–500 501–1000 1001–3300 3301–10 000 10 001–50 000 50 001–100 000 100 001–1 million 41 million National total Grand total

Total population

Total number of ground water systems

749 084 3 377 075 3 310 229 10 867 911 16 242 873 29 704 225 10 889 872 21 334 033 3 933 533 100 408 836

12 843 14 358 4649 5910 2884 1444 167 103 3 42 361

476 998 1 527 684 1 274 145 1 109 731 376 195 207 644 66 000 110 000 – 5 148 396

9456 6758 1894 715 73 10 1 1 – 18 908

2 461 310 3 360 731 1 242 779 828 502 371 291 299 629 51 850 125 000 – 8 741 092 114 298 324

64 448 18 993 1940 585 74 19 1 1 – 86 061 147 330

Note: detail may not add to totals due to independent rounding. Source: Ref. 14.

faecal contamination and requires these at-risk systems to take corrective action. The strategy is discussed in greater detail later in Section 3.2.4. The population and system baselines of the GWR are summarised in Table 3.2.1.

3.2.3.2

Occurrence Data

Occurrence data play a major role in risk assessments conducted for drinking water regulations. Some regulations, for example the Long Term 2 Enhanced Surface Water Treatment Rule, are based on comprehensive occurrence

US Drinking Water Regulation: The Ground Water Rule

111

studies—the Information Collection Rule (ICR) and the ICR Supplemental Surveys—conducted speciﬁcally for that regulation.11 Groundwater is a vital natural resource in the USA; however, there are a limited number of studies that assess the groundwater quality. The ﬁnal GWR risk assessment relied on occurrence data from multiple studies due to lack of a single study on groundwater systems across the USA. The EPA’s GWR risk assessment analyses data from 15 occurrence studies to obtain an estimate of the national occurrence of viral pathogens and faecal indicators, speciﬁcally evaluating the occurrence of enteroviruses based on culture methods, and E. coli.12 When the EPA proposed the GWR, it based the estimate of the national occurrence of pathogenic viruses and faecal indicators on two occurrence studies although 16 studies were evaluated. The two studies were more comprehensive than the other 14 studies and were selected because they provided a relatively large amount of occurrence data among diﬀerent geographical areas: one of the studies included data from 17 States and 2 US territories, the other study analysed groundwater sources in 35 States and 2 US territories. The remaining 14 studies were more limited in scope and mostly individual State occurrence studies. The EPA intended to use the two datasets selected for the proposed GWR risk analysis to estimate the national occurrence of pathogenic viruses and faecal indicators. The EPA did not include bacterial pathogen data in the risk analysis due to the lack of available occurrence data. Based on the objectives and scope of the studies, the EPA preliminarily concluded that the two sets of data represent the national pathogenic virus and faecal indicator occurrence and concentrations in US groundwater sources.9 The EPA received adverse comments during the public comment period for the proposed GWR related to the EPA’s reliance on two studies to represent national pathogenic virus and faecal indicator occurrence in US groundwater sources. Therefore, the EPA re-evaluated the 16 occurrence studies cited in the proposed GWR and several other studies that became available after the GWR was proposed.13 The EPA faced the challenge of using multiple studies with varying objectives and scopes to estimate that national pathogenic virus and faecal indicator occurrence. To address this challenge, the EPA convened a two-day statistical workshop in 2005 to obtain advice from expert statisticians on a scientiﬁcally sound and meaningful way to select relevant studies and combine data from multiple studies to model the occurrence of pathogenic viruses and faecal indicators in public water wells.12 The statisticians recommended that EPA use all available data unless it was aware of data quality assurance problems or the well contamination situation was outside the realm of normal operation of public wells in the USA. The EPA followed the statisticians’ recommendations and used all available and relevant US groundwater studies that had enterovirus cell culture and E. coli data. The modelling approach used to combine data from the studies used in the GWR risk analysis is outside the scope of this chapter; however, detailed information about the modelling approach can be found in Economic Analysis for the Final Ground Water Rule.14

112

3.2.3.3

Chapter 3.2

Public Health Risks

Some individuals may be at risk due to exposure to viral and bacterial pathogens because the pathogens and/or faecal indicators may occur in GWSs that provide no treatment, provide less than optimal treatment (o4-log treatment of viruses) or have treatment deﬁciencies/failures. EPA estimates that GWSs serve untreated groundwater to 20 million people. About half of the 20 million people receive their water from CWSs and the other 50% from NTNCWSs and TNCWSs. Waterborne disease outbreak data (Table 3.2.2) indicate that GWSs that do not disinfect may be susceptible to faecal contamination. The data also show that waterborne disease outbreaks have occurred at disinfecting GWSs due to treatment deﬁciencies and failures. Waterborne disease outbreaks in GWSs demonstrate that individuals have been exposed to pathogenic viruses and bacteria,1 although the EPA believes that the data underestimates a signiﬁcant number of waterborne disease outbreaks and cases of illness due to underreporting. Gastrointestinal illness is the commonly reported illness in waterborne disease outbreaks. In healthy individuals the illness may cause mild symptoms like mild diarrhoea, fever and vomiting which is usually treated at home without seeking medical attention. In infants the illness may be more severe and, in some cases, life-threatening.

Table 3.2.2

Sources of waterborne disease outbreaks in groundwater systems, 1991–2000.

Cause of contamination Community water systems Untreated groundwater Treatment deﬁciency Distribution system deﬁciency Miscellaneous/unknown Total Non-community water systems Untreated groundwater Treatment deﬁciency Distribution system deﬁciency Miscellaneous/unknown Total Combined Untreated groundwater Treatment deﬁciency Distribution system deﬁciency Miscellaneous/unknown Total Sources: Refs. 15–19.

Number of outbreaks

Outbreaks (%)

Cases of illness

Illness (%)

Cases per outbreak

5 7 5

26 37 26

167 1,624 803

6 58 29

33 232 161

2 19

11 100

183 2777

7 100

92 146

23 19 6

47 39 12

4057 3264 442

50 40 5

176 172 74

1 49

2 100

386 8149

5 100

386 166

28 26 11

41 38 16

4224 4888 1245

39 45 11

151 188 113

3 68

4 100

569 10 926

5 100

190 161

US Drinking Water Regulation: The Ground Water Rule

113

Prospective epidemiological studies and pathogen dose–response information are needed to assess the endemic ‘‘baseline’’ level of health risk related to human consumption of contaminated water from GWSs. The actual endemic risk level associated with consumption of contaminated groundwaters is diﬃcult to ascertain because in many situations the illnesses that may result do not rise to the level of medical attention. In addition, there are very few epidemiological studies conducted in communities where the people drink water solely from GWSs. In addition, dose–response data for pathogenic viruses are available but are also limited. Pathogenic bacteria dose–response data are not considered due to lack of bacterial pathogen occurrence information. Therefore, the quantitative risk analysis is underestimated because it is based solely on viruses. The reader is referred to the Economic Analysis for the Final Ground Water Rule14 for more information on the health eﬀect predictions for the GWR. Because of the limited epidemiological studies and dose–response information, the EPA believes that the GWR risk analysis understates the viral illnesses and deaths that are related to GWSs. The estimated endemic level is also underestimated because illness and deaths due to exposure to bacterial pathogens, e.g. E. coli O157:H7, in groundwater are not quantiﬁed in the GWR risk analysis due to data limitations. E. coli O157:H7 causes kidney failure and possibly other severe illnesses. The EPA dealt with the challenges related to estimating the viral baseline illness and risk reductions (beneﬁts due to the number of avoided illness and deaths that would result from the GWR) by quantifying the risk using dose– response information for rotavirus and echovirus in the GWR quantitative risk analysis. Also, the Economic Analysis for the Final Ground Water Rule14 includes a detailed discussion on unquantiﬁed beneﬁts, for example beneﬁts due to avoided bacterial illness and deaths.

3.2.4

The Risk-Targeted Approach

In the ﬁnal GWR, EPA established the risk-targeted approach to focus the GWR requirements on GWSs that are susceptible for faecal contamination. The intent of the risk-targeted approach is to target a subset of the 150 000 GWSs that have groundwater sources that may be vulnerable to faecal contamination and to require those GWSs to take corrective action. The proposed GWR established a multiple barrier approach. The multiple barrier approach and the ﬁnal GWR risk-targeted strategy involve the same concept: identiﬁcation and corrective measures for GWSs at risk of faecal contamination. The multiple barrier approach included ﬁve provisions: sanitary surveys, triggered source water monitoring, hydrogeological sensitivity assessments/routine source water monitoring, treatment technique requirements and compliance monitoring. The main diﬀerence between the two strategies is that the multiple barrier approach included a hydrogeological sensitivity assessment (HSA) requirement

114

Chapter 3.2

to determine if the GWSs obtain water from sensitive aquifers (e.g. karst, gravel or fractured bedrock aquifers). The multiple barrier approach also included a routine source water monitoring requirement for GWSs that obtain water from sensitive aquifers. In developing the ﬁnal GWR, the EPA removed the HSA requirement and the accompanying routine source water monitoring requirement. The EPA believes these changes alone resulted in a more ﬂexible and more targeted approach since it provides the State with the option to conduct more comprehensive monitoring (assessment source water monitoring) at GWSs that may have a higher likelihood of faecal contamination. The State may choose to use HSA along with other information to determine GWSs most vulnerable to faecal contamination. Below is a summary of the ﬁnal GWR provisions. The reader is referred to the GWR regulatory text and preamble1 for the detailed rule requirements and a comprehensive discussion of the rationale for the requirements. Sanitary surveys. Sanitary surveys must be conducted at all GWSs. The purpose of the sanitary survey is to identify signiﬁcant deﬁciencies that may result in contamination of the public water supply. The State is responsible for conducting the on-site assessment every 3 years for CWSs and every 5 years for non-community water systems (NCWSs) after the initial sanitary survey cycle. The rule includes conditions for the State to conduct sanitary surveys every 5 years (in lieu of every 3 years) for CWSs. Under the TCR, sanitary surveys are required only for small systems that routinely collect less than 5 TCR distribution system samples per month. Sanitary surveys must be conducted every 5 years for small CWSs and NCWSs (serving populations of 4100 or less) after the initial sanitary survey cycle. For NCWSs using only protected and disinfected groundwater, the system must undergo subsequent sanitary surveys at least every 10 years. The scope of the on-site assessments required under the GWR includes evaluations of the following 8 critical elements: (1) source; (2) treatment; (3) distribution system; (4) ﬁnished water storage; (5) pumps, pump facilities and controls; (6) monitoring, reporting and data veriﬁcation; (7) system management and operation; and (8) operator compliance with state requirements. The TCR does not specify the scope of the sanitary surveys. Under the GWR, if a signiﬁcant deﬁciencyz is identiﬁed during the sanitary survey then the GWS must take corrective action. The GWR requires the GWS to inform its customers of an uncorrected signiﬁcant deﬁciency. The GWS must continue to inform the public annually until the signiﬁcant deﬁciency is corrected. The ﬂexibility of the rule provides the State with the option of requiring GWSs to inform customers of corrected signiﬁcant deﬁciencies. z

In the GWR, signiﬁcant deﬁciencies include, but are not limited to, defects in design, operation or maintenance, or a failure or malfunction of the sources, treatment, storage or distribution system that the State determines to be causing, or have potential for causing, the introduction of contamination into the water delivered to consumers.

US Drinking Water Regulation: The Ground Water Rule

115

Triggered source water monitoring. The triggered source water monitoring requirement builds on the existing TCR. The monitoring requirement applies to GWSs that do not provide at least 4-log treatment of viruses and have a TCR distribution system sample that has tested total coliformpositive8. A GWS that is subject to the triggered monitoring provision must monitor its groundwater source(s) for one of three State-speciﬁed faecal indicators (E. coli, enterococci, coliphage). If the groundwater source sample is faecal indicator-positive, the GWR requires the GWS to notify the State and the public within 24 hours. Unless directed by the State to take immediate corrective action following the initial faecal indicator-positive sample, the GWS must collect and test 5 additional groundwater source samples for the presence of the same Statespeciﬁed faecal indicator within 24 hours. If any one of the 5 additional groundwater source samples tests positive for the State-speciﬁed faecal indicator, the GWR requires the GWS to notify the State and public and to take corrective action. As a complement to the triggered source water monitoring provision, States have the discretion to require GWSs to conduct assessment source water monitoring on a case-by-case basis. This optional provision provides States with ﬂexibility and the opportunity to require additional source water monitoring and to further evaluate GWSs that may have a higher likelihood of faecal contamination. Treatment technique requirements. The treatment technique requirements, also called corrective action requirements, apply to GWSs that have received written notice from the State of a signiﬁcant deﬁciency and GWSs that have received written notice from the laboratory of a faecal indicator-positive sample. The GWR requires that within 120 days of receiving the notiﬁcation from the State of a signiﬁcant deﬁciency the GWSs must take corrective action. Under the GWR, GWSs that have received notice from the laboratory that one of the 5 additional groundwater source samples has tested faecal indicatorpositive must take corrective action within 120 days. In some cases, the State may require corrective action to be taken within 120 days following the initial faecal indicator-positive sample. The GWR treatment technique provision requires that GWSs implement at least one of the following corrective action options when taking corrective measures: correct all signiﬁcant deﬁciencies; provide an alternative source of water; 8

A GWS is not required to collect a faecal indicator sample following a TCR total coliform-positive distribution system if (a) the State determines that the total coliform-positive sample is caused by a distribution systems deﬁciency, or (b) the TCR total coliform-positive distribution system sample is collected at a location that meets State criteria for distribution system conditions that will cause total coliform-positive samples.

116

Chapter 3.2

eliminate the source of contamination; and provide treatment that reliably achieves 4-log treatment of viruses. Compliance monitoring. The compliance monitoring requirement applies to GWSs that at anytime during implementation of the GWR provide at least 4-log treatment of viruses using chemical disinfection, membrane ﬁltration or a State-approved alternative treatment technology. The purpose of the provision is to ensure that treatment eﬀectiveness is maintained.

3.2.5

Conclusions

Federal drinking water regulations developed to protect consumers from microbial pathogens that may occur in surface water systems have existed for many years. The ﬁnalisation of the GWR has made a signiﬁcant contribution to US public health protection objectives aimed at providing adequate protection against pathogenic viruses and bacteria for many Americans that obtain drinking water from GWSs. The EPA faced many changes in developing the regulation. However, through the EPA’s careful consideration of the public comments received on the proposed GWR rule requirements, incorporation of expert recommendations on meaningful ways to use multiple occurrence studies and recognition of the limitations of the GWR risk assessment, the EPA has met the challenges and believes that the GWR is a scientiﬁcally sound, ﬂexible and cost-eﬀective regulation.

Acknowledgements I would like to thank the following individuals for their review and contributions to this chapter: Philip Berger, Michael Finn, Jennifer McLain, Michael Messner and Stig Regli.

References 1. EPA, National Primary Drinking Water Regulations; Ground Water Rule; Final Rule, Federal Register, 8 November 2006, vol. 71, p. 65574 (http:// www.epa.gov/fedrgstr/EPA-WATER/2006/November/Day-08/w8763.htm). 2. EPA, Drinking Water; National Primary Drinking Water Regulations: Total Coliforms (Including Fecal Coliforms and E. coli); Final Rule, Federal Register, 54(124): 27544-27568, 29 June 1989 (http://www. epa.gov/safewater/disinfection/tcr/regulation.html#1989rule). 3. EPA, National Primary Drinking Water Regulations; Filtration, Disinfection; Turbidity, Giardia lamblia, Viruses, Legionella, and Heterotrophic Bacteria; Final Rule, Federal Register 54(124): 27486, 29 June 1989. 4. EPA, National Primary Drinking Water Regulations: Monitoring Requirements for Public Drinking Water Supplies; Final Rule, Federal Register

US Drinking Water Regulation: The Ground Water Rule

5.

6.

7.

8.

9.

10. 11.

12.

13.

14.

15.

16.

17.

117

61(94): 24353-24388, 14 May 1996 (http://www.epa.gov/fedrgstr/EPAWATER/1996/May/Day-14/pr-20972DIR/pr-20972.txt.html). EPA, Interim Enhanced Surface Water Treatment Rule; Final. Federal Register 63(241): 69477-69521, 16 December 1998 (http://www.epa.gov/ fedrgstr/EPA-WATER/1998/December/Day-16/w32888.htm). EPA, National Primary Drinking Water; Filter Backwash Recycling Rule; Final Rule; Federal Register 66(111): 31085-31105, 8 June 2001 (http:// www.epa.gov/fedrgstr/EPA-GENERAL/2001/June/Day-08/g13776.htm). EPA, Long Term 1 Enhanced Surface Water Treatment Rule; Final Rule, Federal Register 67(8): 1812-1844, 14 January 2002 (http://www.epa.gov/ fedrgstr/EPA-WATER/2002/January/Day-14/w409.htm). EPA, Long Term 2 Enhanced Surface Water Treatment Rule; Final Rule, Federal Register 71(3): 653-702, 5 January 2006 (http://www.epa.gov/ fedrgstr/EPA-WATER/2006/January/Day-05/w04a.htm). EPA, National Primary Drinking Water Regulations: Ground Water Rule; Proposed Rule, Federal Register, 10 May 2000, vol. 65, p. 30194 (http:// www.epa.gov/fedrgstr/EPA-WATER/2000/May/Day-10/w10763.htm). US Geological Survey, Materials in Use in U.S. Interstate Highways, fact sheet 2006-3127, US Department of Interior, October 2006. EPA, Occurrence and Exposure Assessment for the Long Term 2 Enhanced Surface Water Treatment Rule, EPA–821–R–06–002, US Environmental Protection Agency, Oﬃce of Water, Washington, DC, 2005 (http:// www.epa.gov/safewater/disinfection/gwr/regulation.html). EPA, Occurrence and Monitoring Document for the Final Ground Water Rule, EPA–815–R–06–012, US Environmental Protection Agency, Oﬃce of Water, Washington, DC, 2006 (http://www.epa.gov/safewater/disinfection/ gwr/regulation.html). EPA, National Primary Drinking Water Regulations; Ground Water Rule; Notice of Data Availability, Federal Register, 27 March 2006, vol. 71, p. 15105 (http://www.epa.gov/fedrgstr/EPA-WATER/2006/March/Day-27/ w2931.htm). EPA, Economic Analysis for the Final Ground Water Rule, EPA–815–R– 06–014, US Environmental Protection Agency, Oﬃce of Water, Washington, DC, 2006 (http://www.epa.gov/safewater/disinfection/gwr/regulation.html). R. S. Barwick, D. A. Levy, G. F. Craun, M. J. Beach and R. L. Calderson, Surveillance for waterborne-disease outbreaks: United States, 1997–1998, Morbid. Mortal. Weekly Rep., 2000, 49(SS-04), 1–35 (http://www.cdc.gov/ mmwr/preview/mmwrhtml/ss4904a1.htm). M. H. Kramer, B. L. Herwaldt, R. L. Calderson and D. D. Juranek, Surveillance for waterborne-disease outbreaks: United States 1993–1994, Morbid. Mortal. Weekly Rep., 1996, 45(SS-1), 1–33 (http://www.cdc.gov/ mmwr/preview/mmwrhtml/00040818.htm). S H. Lee, D. A. Levy, G. F. Craun, M. J. Beach and R. L. Calderon, Surveillance for waterborne disease outbreaks: United States, 1999–2000, Morbid. Mortal. Weekly Rep., 2002, 51(SS-08), 1–28 (http://www.cdc.gov/ mmwr/preview/mmwrhtml/ss5108a1.htm).

118

Chapter 3.2

18. D. A. Levy, M. S. Bens, G. F. Craun, R. L. Calderon and B. L. Herwaldt, Surveillance for waterborne disease outbreaks: United States, 1995–1996, Morbid. Mortal. Weekly Rep., 1998, 47(SS-05), 1–34 (http://www.cdc.gov/ mmwr/preview/mmwrhtml/00055820.htm). 19. A. C. Moore, B. L. Herwaldt, G. F. Craun, R. L. Calderon, A. K. Highsmith and D. D. Juranek, Surveillance for waterborne disease outbreaks: United States, 1991–1992, Morbid. Mortal. Weekly Rep., 2003, 42(SS-05), 1–22 (http://www.cdc.gov/mmwr/preview/mmwrhtml/00025893.htm).

4. Stakeholder Interactions

CHAPTER 4.1

Principles of the Common Implementation Strategy of the WFD: The Groundwater Working Groupw PHILIPPE QUEVAUVILLER,a JOHANNES GRATHb AND ANDREAS SCHEIDLEDERb a

European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium; b Umweltbundesamt GmbH, Spittelauer Laende 5, AT-1090 Wien, Austria

4.1.1

The Need for Multi-stakeholder Involvement in the Environmental Policy Development and Implementation Process

The development of environmental policies is a complex process, which mixes legal requirements with issues of technical feasibility, scientiﬁc knowledge and socioeconomic aspects, and which requires intensive multi-stakeholder consultations. In this context, the consideration of scientiﬁc progress and access to technical information represent key aspects for the design of new policies and the review of existing ones.1 This is discussed in detail in Chapter 2.1. Within the European Union (EU), this consideration is fully embedded in the Sixth Environment Action Programme which stipulates that ‘‘sound scientiﬁc knowledge and economic assessments, reliable and up-to-date environmental data and information, and the use of indicators will underpin the drawing-up, implementation and evaluation of environmental policy.’’2 This requires, therefore, that scientiﬁc inputs should constantly feed the environmental policy process. This integration also involves various players, namely the scientiﬁc and w

The views expressed in this chapter are purely those of the authors and may not in any circumstances be regarded as stating an oﬃcial position of the European Commission.

121

122

Chapter 4.1

Figure 4.1.1

Consultation process involving stakeholders.

policy-making communities, but also representatives from industry, agriculture, non-governmental organisations (NGOs), etc. (Figure 4.1.1). This aspect is described in depth in Chapter 4.4. An example of successful multi-stakeholder consultation and participation concerns the implementation of the Water Framework Directive (WFD; 2000/ 60/EC). In this context, a Common Implementation Strategy (CIS) has been agreed with the EU member states, candidate and associate countries and stakeholder organisations and has been operational since 2001.3 In this framework, various topics are under discussion by experts from EU member states, industry, agriculture, scientists, etc., with the aim to gather and share knowledge and concern on WFD relevant issues, as examined from diﬀerent perspectives. This approach, albeit time-consuming, has considerably enhanced the knowledge and common interpretation of the key provisions of the WFD, and it has been considered as a very powerful tool for sharing good practices and an example of good governance.

4.1.2

The WFD Common Implementation Strategy

4.1.2.1

General Principles

As already mentioned above, it has become clear, soon after the WFD adoption, that the successful implementation of the directive will be, at the least, equally as challenging and ambitious for all countries, institutions and stakeholders involved. Therefore, a strategic document establishing a CIS for the WFD has been developed and ﬁnally agreed by the EU’s Water Directors under the Swedish presidency in 2001.3 Despite the fact that the full responsibility of the individual member states for implementing the WFD was recognised, a broad consensus existed among the Water Directors of the

Principles of the Common Implementation Strategy of the WFD

123

member states, Norway and the Commission that a European joint partnership was necessary in order to:

develop a common understanding and approaches; elaborate informal technical guidance including best practice examples; share experiences and resources; avoid duplication of eﬀorts; and limit the risk of bad application.

Furthermore, the Water Directors stressed the necessity to involve stake holders, NGOs and the research community in this joint process as well as to enable the participation of candidate countries in order to facilitate the cohesion process. Following the decision of the Water Directors, a comprehensive and ambitious work programme was started of which the ﬁrst phase, including ten Working Groups and three Expert Advisory Forum (EAF) groups, was completed at the end of 2003 and led to the availability of fourteen guidance documents which are publicly available (in the form of CD-ROM and on the internet on the WFD europa website). The second phase of the CIS (2003–2004) involved four Working Groups, namely on Ecological Status (WG A), Economics and Pilot River Basins (WG B), Groundwater Body Characterisation and Monitoring (WG C) and Reporting (WG D), as well as two EAF groups, of which the discussions focused on developing policies linked to the WFD (i.e. Priority Substances Directive, and revision of the Reporting Directive). These groups were re-conducted in the third phase (2005–2006), and this process is now continued under new mandates for the period 2007–2009 (see Figure 4.1.2), which is detailed with regard to groundwater in the section below.

4.1.2.2

Supporting Activity: The Pilot River Basin Network

In the context of the CIS, a network of Pilot River Basins (PRBs) has been established to test and validate guidance documents developed under the CIS of the WFD. The network covers 15 PRBs in 12 countries. Activity reports are published on a yearly basis. Besides the CIS process, a range of PRBs are linked to research and demonstration projects. They indeed represent an opportunity for researchers to test new developed techniques or methodologies (e.g. risk assessment methods, monitoring tools, modelling) in well-characterised areas which have direct links to WFD implementers. This network is described in detail in Chapter 4.2.

4.1.3

The CIS Working Group on Groundwater

4.1.3.1

Objectives

The CIS Groundwater Working Group (WG C) aims both to clarify groundwater issues that are covered by the WFD and to prepare the development of technical guidance documents and exchange best practices on several issues in

124

Chapter 4.1

Figure 4.1.2

CIS operational diagram for the period 2007–2009.

the light of the orientations of the newly adopted Groundwater Directive (see Chapter 3.1).

4.1.3.2

Leadership and Network

The Commission/DG ENV chairs the WG C which is co-chaired by Austria. The Working Group is composed of representatives of EU member states, associated and candidate countries, industrial and scientiﬁc stakeholders and NGO representatives (around 80 members in total). Plenary meetings are open to all participants, while ad hoc activities are operated by groups of a maximum of 15–20 participants which develop documents that are scrutinised by the plenary group.

4.1.3.3

Achievements from 2003 to 2006

The focus in the period 2003–2006 was on the development of technical reports and guidance documents primarily focusing on the issues covered by the WFD, namely monitoring, prevent/limit measures and groundwater protected areas. In addition, a speciﬁc activity concerned exchange of views on groundwater management in the Mediterranean area (linked to the EU Water Initiative). Activities of the WG were conceived with the view of collecting targeted data

Principles of the Common Implementation Strategy of the WFD

125

and information, avoiding duplication with existing guidance documents and ensuring an eﬃcient use of available data and information. Series of workshops were held in 2003–2004, which led to three technical reports gathering member states’ practices in the ﬁeld of groundwater risk assessment, monitoring and programmes of measures.4–6 The orientations in 2005–2006 were concerned with the drafting of guidance documents on groundwater monitoring, protected areas and measures to prevent/limit pollutant introduction into groundwater. The Monitoring Guidance for Groundwater document was ﬁnalised and endorsed by the EU Water Directors on 30 November 2006 at the meeting under the Finnish presidency in Inari (Lapland).7 The two other guidance documents were slightly delayed owing to the negotiation of the new Groundwater Directive, and were rescheduled for the 2007–2009 work programme of WG C (see Section 3.4).

4.1.3.4

Perspectives for 2007–2009

The main orientations of the 2007–2009 mandate of WG C were discussed at the occasion of the Groundwater Conference in Vienna on 22–23 June 20068 and through an e-mail consultation of all WG C members. The main aims and objectives of WG C for the period 2007–2009 are to pursue exchanges in support of the implementation of the new Groundwater Directive along the CIS principles,9 focusing in particular on: best practices related to groundwater programmes of measures, including measures related to diﬀuse sources of pollution and megasites; common methodology for the establishment of groundwater threshold values; compliance, status and trend assessment; and recommendations for integrated risk assessment, including conceptual modelling. The WG C work programme 2007–2009 consists of three core activities led or co-led by member states or stakeholder organisations, which will develop their work programme as described in activity sheets. The activities (drafting or exchanges of good practices) will be undertaken with selected WG participants (groups of ideally 15–20 participants) willing to actively contribute to the drafting of documents and to participate in ad hoc meetings (possibly organised by the activity leaders). The progress of the activities will be reported and discussed at plenary meetings of WG C held twice a year and organised under the EU presidency umbrella. The objectives of WG C for the period 2007–2009 are separated into three core activities. 1. Activity 1: Programmes of Measures (PoM). Discussions will focus on exchanges on best practices and recommendations needed by member states in the context of the identiﬁcation of measures related to

126

Chapter 4.1

groundwater that will have to be included in the First River Basin Management Plan. The activity will cover, in particular, the following. The ﬁnalisation of the ‘‘prevent/limit’’ guidance initiated in 2006, which aims to support the implementation of Article 6 of the new Groundwater Directive. Exchanges on best available technologies related to groundwater measures, taking into account programmes of measures required under other EU directives. These will include guidance on both point sources of pollution (including historical contaminated sites) and on diﬀuse sources, including agricultural diﬀuse pollution and megasites (large polluted areas, e.g. harbour areas and industrially contaminated sites). This item will be closely linked to a CIS expert group on ‘‘WFD & Agriculture’’ with regard to diﬀuse agricultural sources. 2. Activity 2: Compliance and Trends. The work will be directed toward the development of a guidance document on compliance and trends, as well as on recommendations establishing on groundwater threshold values, considering the following. Adoption of a common methodology for the establishment of groundwater threshold values based on the outcome of the methodology developed by the BRIDGE project (see Chapter 9.1), and exchanges of experiences among the member states in support of the new Groundwater Directive (see Chapter 3.1). Development of the status compliance and trends guidance document, concerning both quantitative and chemical status issues. This had been planned in the former work programme but could not be initiated owing to delays in the adoption of the new Groundwater Directive. With respect to trend assessment, the document will be largely based on a technical report developed in 2002,10 and will provide recommendations to member states on how to undertake and interpret trend studies (including considerations on lag time of groundwater systems and how to integrate this in trend assessment). 3. Activity 3: Integrated Risk Assessment. Discussions will focus on recommendations for improving risk assessment for groundwater at river basin level in an integrated way, in view of the preparation of the First River Basin Management Plan. The activity will cover, in particular, the following. Discussions on how to improve risk assessment and recommendations on conceptual modelling for water systems, including (publicly available) databases, mapping (e.g. vulnerability, hydrogeology) and visualisation of subsurface processes. Good management practices, including issues such as artiﬁcial recharge and transboundary aquifer management. This item will be operated in close connection with a CIS expert group on water scarcity and drought.

Principles of the Common Implementation Strategy of the WFD

4.1.4

127

Perspectives

The successful implementation of the new Groundwater Directive will closely depend upon an eﬃcient participatory approach and harmonised groundwater risk assessment, monitoring and programmes of measures throughout the EU. The CIS Working Group on Groundwater will be an indispensable element supporting this implementation, in particular in view of the preparation of the First River Basin Management Plan expected for publication at the end of 2009. The working group will also help in gathering the necessary scientiﬁc and technical knowledge which will enable a scientiﬁcally sound review of the new Groundwater Directive required under Article 7, i.e. by the end of 2012, which will coincide with the operational start of the WFD Programme of Measures.

References 1. Ph. Quevauviller, P. Balabanis, C. Fragakis, M. Weydert, M. Oliver, A. Kaschl, G. Arnold, A. Kroll, L. Galbiati, J.M. Zaldivar and G. Bidoglio, Environ. Sci. Pol., 2005, 8, 203. 2. European Commission, 6th Environment Action Programme 2002–2012, 2002. 3. Common Implementation Strategy for the Water Framework Directive, European Communities, 2003 (ISBN 92-894-2040-5). Final CIS document available at: http://cc.europa.eu/environment/water/water-framework/ objectives/implementation-en.htm. 4. European Commission, Groundwater Body Characterisation, technical report, 2004. 5. European Commission, Groundwater Risk Assessment, technical report, 2004. 6. European Commission, Groundwater Monitoring, technical report, 2004. 7. European Commission Monitoring Guidance for Groundwater, CIS Guidance no. 15, Common Implementation Strategy of the WFD, European Commission, 2007. 8. Proceedings of the European Groundwater Conference, Vienna, 22–23 June 2006. 9. Mandate of Working Group C ‘‘Groundwater,’’ Common Implementation Strategy of the WFD, European Commission, 2007. 10. Technical report on Statistical aspects of the identiﬁcation of groundwater pollution trends, and aggregation of monitoring results, Common Implementation Strategy, 2001.

CHAPTER 4.2

The Pilot River Basin Network: Examples of Groundwaterrelated Activitiesw LORENZO GALBIATIa AND GIOVANNI BIDOGLIOb a

Age`ncia Catalana de l’Aigua, Provenc¸a 204-208, ES-08036 Barcelona, Spain; b European Commission, Joint Research Centre, Via E. Fermi 1, TP 460, IT-21020 Ispra (VA), Italy

4.2.1

Introduction

The Water Framework Directive1 (WFD) sets the guidelines for sustainable water management in Europe. Its aim is to achieve ‘‘good status’’ for all environmental waters by 2015 at the latest. This is to be achieved through programmes of measures in which each river basin is treated as a coherent unit (see Chapter 3.1). To help the compliance of this ambitious objective the European Union (EU) water directors have agreed to establish a Common Implementation Strategy (CIS) of the WFD (see Chapter 4.1). During the 2001–2002 CIS period, a series of guidance documents (GDs) concerning all major aspects the WFD implementation were developed by working groups (WGs) including representatives of EU member states, accession countries, national experts and the European Commission. In order to test and cross-validate these GDs, a Pilot River Basin (PRB) network has been established. It was foreseen that such a network would act as an interface between the Commission and member state authorities and thus contribute to the implementation of the WFD. The PRB exercise was structured into two phases. Phase I focused on testing and reporting on coherence amongst the diﬀerent GDs, leading to the longterm development of River Basin Management Plans and preparation of w

The views expressed in this chapter are purely those of the authors and may not in any circumstances be regarded as stating an oﬃcial position of the Catalan Water Agency or the European Commission.

128

The Pilot River Basin Network: Examples of Groundwater-related Activities

129

programmes of measures. This ﬁrst phase was ﬁnalised at the end of 2004, the main deliverables of which were: the PRB outcome report on the testing of WFD GDs;2 the Article 5 reports produced by some PRBs (available on the PRB’s website and on the PRB section of CIRCA); and the PRB thematic workshops, which were organised on the topics of groundwater, water body delineation, economy, Mediterranean dimension, and research and technology integration in support of the WFD. Phase II mainly consisted of gaining experience and developing methods necessary to draw up the monitoring programmes and programmes of measures for the PRBs according to the WFD deadlines. In this phase, PRBs have been used as tools to reach the goals of the WGs in the CIS project. Phase II of the exercise started at the beginning of 2005 and was ﬁnalised at the end of the 2006.

4.2.2

Science–Policy Integration in the PRB Exercise Linked to Groundwater Management

Under Article 5 of the WFD, member states had to identify water bodies by end of December 2004. In this context, member states carried out an initial characterisation of all groundwater bodies including their location and boundaries as well as the identiﬁcation of pressures and groundwater bodies at risk of failing to meet the objectives of the WFD. In a forefront initiative to support this speciﬁc aspect of the implementation process, the PRBs tested the applicability of the GD on groundwater. Generally in this phase the PRBs perceived that the deﬁnition of the river basin district boundaries is the most important and complex issue which is not only aﬀecting the activities underlined by the Planning GD but is strongly aﬀecting the activities related with the Groundwater GD, especially in those cases where the groundwater body deﬁnition concerns shared aquifers. These criteria were chosen in most countries at national level, and then adopted with the necessary adaptation by all basin districts. Criteria were also set for assignment of each shared groundwater body to all the pertaining river basins, based on available information about hydrogeology (bedrock geology, tracing study results, groundwater ﬂow regime and direction) and the presence of dependent ecosystems (groundwater-fed lakes, rising from underground streams, groundwater-dependent terrestrial ecosystems). In the case of shared river basins or groundwater bodies, the issue of cooperation becomes essential, particularly in the development of the programme of measures and river basin management plans, to ensure that such interconnected water bodies and associated ecosystems are adequately protected. Most of the activities related to groundwater management undertaken by the PRBs have been focused on the diﬀerent criteria applied to deﬁne the groundwater bodies. Section 3 presents a methodology proposed by the Shannon PRB for the managing of groundwater bodies.

130

Chapter 4.2

During phase II of the exercise, the PRBs developed their activities under the umbrella of diﬀerent CIS WGs contributing to the testing of diﬀerent methodologies to tackle and solve problems related with the WFD implementation. For example, Article 17 of the directive stipulates that criteria for establishment of the groundwater chemical status should be developed in a proposal made by the European Commission, i.e. the ‘‘Groundwater Daughter Directive’’ adopted in December 2006 (see Chapter 3.1). In this directive, the Commission speciﬁes that the good chemical status of groundwater should be partly deﬁned by the establishment of groundwater standards (‘‘threshold values’’) by member states themselves. The idea is that the chemical status of groundwater will be based on existing community quality standards and on the requirements for member states to identify pollutants and related threshold values that are representative of groundwater bodies found being at risk, in accordance with the analysis of pressures and impacts carried out under the WFD. The interface role of the PRBs qualiﬁed them as appropriate places for researchers to test new methodologies (e.g. risk assessment and monitoring approaches, monitoring tools).3 An example is the role of the PRBs in the development of pollutant threshold values in the context of the BRIDGE project,4 which involved the following PRBs: Tevere (Italy), Pinios (Greece), Scheldt (Belgium) and Odense (Denmark). Section 4 gives a brief overview of the case study developed by the Tevere PRBs for the identiﬁcation of the groundwater natural background levels and the deﬁnition of threshold for an aquifer in a Mediterranean catchment.5

4.2.3

Managing Groundwater Bodies in the Shannon PRB for the Implementation of the WFD

The Shannon PRB is the largest river basin in Ireland draining a land area of some 18 000 km2 in central Ireland. It includes part of 18 local authorities in the Republic of Ireland and has a small transboundary component of approximately 6 km2 in County Fermanagh, Northern Ireland. Carboniferous rocks dominate the bedrock geology of the Shannon PRB. Of these, highly karstiﬁed pure bedded limestones predominate in the upper reaches of the basin. Groundwater ﬂow in these rocks is dominated by conduit ﬂow. In contrast, in most of the rest of the basin, groundwater ﬂows through ﬁssures and faults in relatively low transmissivity aquifers. In the west, on either side of the Shannon estuary, bedded shales and sandstones of Namurian age dominate. Between the upper and lower reaches of the basin, embedded pure limestone and impure limestone are folded around cores of older rocks. Agriculture is the principal activity in the river basin (73% of total area); the dominant land use being pasture. There are some signiﬁcant areas of wetland (12%), mainly peatland. The catchment is not notably industrialised and agri-industries, such as milk and meat processing, are the most prominent.

The Pilot River Basin Network: Examples of Groundwater-related Activities

4.2.3.1

131

Groundwater Body Delineation

The Geological Survey of Ireland (GSI) has carried out the delineation of groundwater bodies in Ireland, including the Shannon PRB. The delineation process involved several stages. Mapped rock units were assigned an aquifer class based on the existing GSI aquifer classiﬁcation system. These aquifer classes were then grouped into four aquifer types based on groundwater ﬂow regime, i.e. karst aquifers, gravel aquifers, productive ﬁssured bedrock aquifers and poorly productive bedrock aquifers. Preliminary groundwater bodies were then delineated using no-ﬂow geological boundaries, as well as boundaries based on groundwater highs, diﬀering ﬂows and ﬂow lines. Final delineation incorporated major surface water catchment boundaries except in areas where the inﬂuence of topography is diminished (e.g. karstic or conﬁned aquifers). This process resulted in the delineation of 97 bedrock groundwater bodies with a median size of 53 km2.

4.2.3.2

Groundwater Management for the WFD

The ﬁrst requirement of the WFD is to identify groundwater bodies at risk of failing to meet the environmental objectives set out in Article 4. To achieve these objectives requires making operational the programme of measures speciﬁed in the River Basin Management Plan. A proposed risk assessment methodology to identify groundwater bodies (GWBs) at risk is presented. This process will allow for the prioritisation of resources in the River Basin Management Plan. The focus of the programme of measures should be on the high impact potential areas of ‘‘at risk’’ GWBs. Diﬀerent aquifer types will require diﬀerent management responses appropriate to their spatial extent, ﬂow regime, degree of groundwater–surface water interaction and connectivity with groundwater-dependent terrestrial ecosystems. This approach will require a detailed conceptual understanding of each GWB to ensure that the most suitable programmes of measures are applied and the use of limited resources is optimised.

4.2.3.3

Example of Risk Assessment Methodology for Diﬀuse Groundwater Pollution in the Shannon PRB

The following approach is a screening exercise using available GIS layers and follows the ‘‘source-pathway-receptor’’ model. The objective is to identify groundwater bodies at risk and allow for prioritisation in the programme of measures and river basin management plan. To reach this goal a ﬁve-step procedure has been applied. The ﬁrst and second steps are to develop of a good conceptual understanding of each groundwater body and combine information on groundwater vulnerability with aquifer ﬂow regime characteristics using risk matrices to identify the degree of pathway susceptibility to diﬀuse pollution (Figure 4.2.1). Then pressure magnitude thresholds, e.g. for stocking density, are set up (Figure 4.2.2). Thresholds will

132

Chapter 4.2

Figure 4.2.1

Step 1 and step 2: develop a good conceptual understanding and combine information on groundwater vulnerability.

Figure 4.2.2

Step 3: set pressure magnitude thresholds.

need to be developed for all pollutant types. The next step is the combination of the pathway susceptibility and pressure magnitude using risk matrices to produce an impact potential map (Figure 4.2.3). As a last step, the combination of all these steps produces the ﬁnal risk designation (Figure 4.2.4). The percentage area aﬀected by pollution combined with a veriﬁcation using monitoring data will determine the identiﬁcation of whether a groundwater body is ‘‘at risk’’ or not. Lack of monitoring data and pressure layer information will aﬀect the conﬁdence in the risk designation. Further assessment may be required to determine whether associated surface waters or groundwaterdependent terrestrial ecosystems are adversely impacted.

The Pilot River Basin Network: Examples of Groundwater-related Activities

Figure 4.2.3

Step 4: pathway susceptibility and pressure magnitude.

Figure 4.2.4

Step 5: ﬁnal risk designation.

4.2.4

133

Groundwater Natural Background Levels and Threshold Deﬁnition in the Tevere PRBs Under the BRIDGE Project

The BRIDGE project has been structured with diﬀerent working packages (WPs), each of them being a diﬀerent step in the deﬁnition of the background criteria for the identiﬁcation of groundwater threshold (see further details in Chapter 9.1): work package 1: survey of representative groundwater pollutants project launching and co-ordination; work package 2: study of groundwater characteristics; work package 3: criteria for environmental thresholds and methodology to deﬁne a good status; work package 4: representative sites/water body studies and compliance testing; work package 5: economic and social costs linked to the establishment of groundwater threshold values; and work package 6: information and dissemination.

134

Chapter 4.2

The example of the Tevere and the PRBs presented in this section are related to the developed in the framework of the WP4. The objective of WP4 was to evaluate the developed approach and application of environmental thresholds recommended in BRIDGE WP3 at selected European representative sites. Results of this phase were then transferred to WP5 and WP6 for economic assessment and communication, respectively. The aim of the BRIDGE WP3 was to propose a practical approach for a methodology to deﬁne the good status for groundwater bodies and a general structure for developing criteria related to this status deﬁnition, providing guidance on the derivation process for environmental thresholds. Following the structure of the project, the activities developed in this WP referred to data and information coming from WP1 and WP2. In this section a brief summary of the case study developed by the Tevere in the framework of the BRIDGE project is given. A complete description of the area and the groundwater characterisation are available as a full report at the BRIDGE web page (www.wfd-bridge.net). The section focuses on the description of a methodology applied to obtain the groundwater status evaluation by using threshold values.

4.2.4.1

Tevere PRB: The Colli Albani Case Study

The Colli Albani volcanic area has a surface of about 1950 km2. The area is characterised by four WBs for the surface aquifers on the basis of groundwater watersheds and ﬂow direction. Overexploitation is the main problem in all the four WBs. The impacts produced by this pressure on the aquifer aﬀect diﬀerent receptors, among them groundwater itself, terrestrial ecosystems, aquatic ecosystems and drinking water. All of the receptors considered in BRIDGE are present in the Colli Albani aquifer, the following in particular: watercourses extending in a radial direction along the slopes of the volcanic structure fed by point and linear sources; lakes in hydraulic continuity with groundwater; wetlands associated with terrestrial ecosystems dependent on groundwater; coastal areas subject to saline intrusion; areas destined for water abstraction for drinking water use; and groundwater itself. The main pressure in the area is groundwater abstraction, coming from about 33 000 points legally authorised and from an estimated equivalent number of illegal wells. Diﬀuse and point sources of pollution, coming from agricultural areas (38% of the surface) is also an important pressure in the area. Concerning the determination of a natural background level (NBL) of pollutants no existing methodologies are available neither at a national nor regional level. However, many studies on the characterisation of the hydrochemical facies of the Colli Albani aquifers are available, based on a monitoring network developed in the area since the 1970s. It is possible to observe several important eﬀects caused by the main pressure on groundwater. With respect to the quantitative aspect these are: substantial variation of the piezometric level (especially in the intensely exploited areas), lowering of the level of the lakes, 60% reduction of the total base ﬂow in the

The Pilot River Basin Network: Examples of Groundwater-related Activities

135

watercourses and salty water intrusion in the coastal areas. Looking at the groundwater quality as aﬀected by groundwater abstraction, no signiﬁcant values for anthropogenic hazardous substances are visible. The abstraction is also aﬀecting the water-dependent ecosystems. In this case the main impact is related to the variations in the extension of wetlands and disappearance of springs, which also led to a reduction of the base ﬂow in surface watercourses. As a consequence, the dilution capacity of watercourses is substantially reduced and under certain conditions discharges completely replace base ﬂow. Consequently, groundwater itself becomes the receptor of the water recharge coming from the watercourses.

4.2.4.2

Groundwater Status Evaluation by Threshold Values

The application of WP3 methodologies was preceded by a study of the hydrogeochemical characteristics of the Colli Albani aquifer. For this purpose chemical analyses carried out on 15 springs characterised by marked hydrogeochemical anomalies were used. About 100 samples from the period 1970– 1980 were selected. They were characterised by low anthropogenic pressure. Furthermore, chemical analyses from a measurement campaign carried out in 2005 by the Lazio Regional Environment Agency on about 50 wells used for drinking water abstraction were considered, especially focusing on natural elements such as arsenic, vanadium and ﬂuorides. In some cases these natural elements exceeded drinking water standard values. For example, high values of arsenic, vanadium and ﬂuorides seem to be located in measuring points in the southeastern sector of the volcanic structure. The high variability of the chemical characteristics of water and consequently of the natural background levels of volcanic aquifers suggested that WP3 methodologies can be applied successfully only if the water families that interact with receptors are identiﬁed ﬁrst. The following steps were carried out in this case study: mapping of the hydrochemical characteristics of water in the entire aquifer; mapping of the receptors; description of receptor–groundwater body interaction; selection of Environmental Quality Standards (EQS) values that can be associated with the receptors; identiﬁcation of NBL values of groundwater that interact with the receptor; identiﬁcation of signiﬁcant parameters that may cause variations in the receptor; and application of WP3 methodologies for threshold identiﬁcation. This approach requires that for each receptor the type of interaction with groundwater, the EQS values to take into account and the natural background levels of local groundwater that interact with the receptor are known. Threshold values were calculated only for parameters that cause negative impacts on

136

Chapter 4.2

the receptor and show growing trends in respect to the natural background values. At present, there are not suﬃcient data to produce a detailed map of the chemical characteristics of the entire Colli Albani aquifer. Therefore, the identiﬁcation of threshold values was carried out starting from each receptor in relation to available data. In particular, WP3 methodologies were applied to two speciﬁc areas: groundwater abstraction area destined for drinking water use (SaloneAcqua Vergine area); and salty water intrusion near the protected area of Castelporziano.

4.2.4.3

Case Study of Salone-Acque Vergini System

The study area is a protected zone known as the Salone-Acqua Vergine system. The main pressure on the area is groundwater abstraction for drinking water use, with four main abstraction points, which supply a total of 600 l s 1. Monitoring activities, are carried out mainly by water supply agencies and consist of about 2000 samples collected from 1992 to 2005. In 2001 a ﬁeld campaign was undertaken to analyse the hydrogeochemical composition of the water which is feeding the springs and wells in water abstraction points for drinking water supply. Eighteen regularly sampled monitoring points were identiﬁed. The considered EQS values were drinking water standards according to those proposed by Directive 1998/83/EC.6 Time-series analyses of parameters such as Ca, Mg, SO4, Na, Cl, Si, Fe, V, Li, B and NH41 have shown ﬂuctuations around a central value that remains constant. This can be ascribed to seasonal ﬂuctuations or to analytical variability. The NO3 parameter was considered signiﬁcant in respect to the impact on the receptor due to its signiﬁcant upward trend, probably due to anthropogenic pressures (Figure 4.2.5). The trend analysis from 1997 showed an increase in some wells of the nitrates value, which reached 30 mg l 1 in this speciﬁc area, probably due to anthropogenic factors related to agricultural activities and/or wastewater point sources coming from urban areas. NBL values of the chemical compounds were derived from 90th percentile calculations of analyses carried out on six monitoring points in 1997. Arsenic is another signiﬁcant parameter that characterises the receptor. It varied between 5 and 10 mg l 1 and did not show signiﬁcant trends. The background levels derived from the statistical analysis of NO3 and as are presented in Table 4.2.1.

4.2.4.4

Case Study of the Protected Area of Castelporziano

The national protected area of Castelporziano was included in the ‘‘Natura 2000’’ ecological network as a Special Conservation Zone. It is a strip of Mediterranean forest, which has remained unaltered for many centuries, with terrestrial ecosystems depending on water and wetlands. Reclamation works carried out in the surrounding marshland during the 1930s as well as groundwater overexploitation for agricultural, household and industrial uses modiﬁed

The Pilot River Basin Network: Examples of Groundwater-related Activities

137

Figure 4.2.5

Nitrates concentration trend in two wells located in protected area for drinking water uses.

Table 4.2.1

Nitrates and arsenic threshold values in the drinking water of the Salone-Acqua Vergine area.

Parameter 1

Nitrate (mg l ) Arsenic (mg l 1)

NBL BRIDGE

TV

Drinking water standard

24 10

37 10

50 10

Case 2 Case 3

aquifer recharge causing progressive inﬁltration of saline water from the sea and the ﬁnal stretch of the Tevere River. Phenomena linked to mainly gaseous endogenous ﬂuids take place in the inland area of Castelporziano, in the Malafede valley, characterised by the presence of mineral and thermomineral water. Since salinity variations may cause adaptive changes in the ﬂora and fauna of the protected area’s terrestrial ecosystem, conductivity, salinity and piezometric levels are monitored in the area. Methodologies for the identiﬁcation of salinity levels that may induce these adaptive changes still need to be consolidated. Currently, EQS values to use for salinity and conductivity are being assessed by expert judgement, mainly considering trend analysis rather than threshold values. Tables 4.2.2 and 4.2.3 show the results of two measurement and chemical ﬁeld analyses, carried out in the area during 1999 and 2004.7 A more detailed reconstruction of the phenomenon is shown in Figure 4.2.6, where also sampling points external to the protected area were taken into account. The map shows that higher conductivity levels are located in the most intensively urbanised areas, while lowest conductivity levels are located at the centre of the protected area.

T (1C)

18.0 17.7 17.6 16.6 18.1 15.9 16.8 17.2 16.0 15.5 17.2 17.9 15.5 18.1 17.0 15.9 17.2 18.0 18.1

F2 F3 E1 E2 E3 E4 E7 E8 E11 E26 E27 C1 min max med P10 P50 P90 P97.7

6.7 6.6 7.3 7.2 7 6.9 6.6 6.1 6.8 6.6 7.7 6.9 6.1 7.7 6.9 6.6 6.9 7.3 7.6

pH

1.95 0.87 1.14 0.79 1.04 0.99 0.34 1.26 0.98 1.44 1.05 1.12 0.34 1.95 1.08 0.80 1.05 1.28 1.40

Cond. (mS cm 1) 144.0 87.0 54.0 40.0 108.0 40.0 38.0 67.0 54.0 76.0 75.0 56.0 38.0 144.0 69.9 40.0 61.5 79.2 101.4

Ca (mg l 1) 14.8 9.2 3.7 5.4 4.3 4.4 8.8 5.6 2.0 6.6 32.4 3.4 2.0 32.4 8.4 3.4 5.5 11.2 27.5

K (mg l 1) 194.0 39.0 123.0 92.0 90.0 113.0 16.0 135.0 124.0 149.0 69.0 100.0 16.0 194.0 103.7 42.0 106.5 136.4 146.1

Mg (mg l 1) 54.5 24.8 36.5 12.3 10.8 25.0 3.7 31.0 22.7 35.0 39.3 34.7 3.7 54.5 27.5 11.0 28.0 36.8 38.7

Na (mg l 1) 888.0 403.0 481.0 422.0 407.0 505.0 169.0 533.0 565.0 537.0 470.0 497.0 169.0 888.0 489.8 403.4 489.0 539.8 559.2

HCO3 (mg l 1) 204.0 85.1 76.4 45.1 114.0 66.7 17.9 146.0 48.0 186.4 86.0 86.3 17.9 204.0 96.8 45.4 85.6 150.0 178.0

Cl (mg l 1)

40.7 0.2 64.5 29.2 40.1 34.0 13.2 37.0 30.4 33.6 54.0 52.4 0.2 64.5 35.8 14.8 35.5 55.1 62.3

SO4 (mg l 1)

Chemical analyses and physical parameters of 12 samples in the protected area of Castelporziano, 1999.

Sample

Table 4.2.2

1540.0 648.0 839.0 646.0 775.0 788.0 267.0 955.0 846.0 1023.0 825.0 829.0 267.0 1540.0 831.8 646.2 827.0 961.8 1008.9

TDS (mg l 1)

138 Chapter 4.2

T (1C)

17.6 18.5 18.1 18.9 17.4 19.3 17.8 18.7 16.9 18.5 18.6 18.6 16.9 19.3 18.2 17.4 18.5 18.9 19.2

F2 F3 E1 E2 E3 E4 E7 E8 E11 E26 E27 C1 min max med P10 P50 P90 P97.7

7.0 7.2 6.9 6.8 6.8 7.1 6.5 6.8 6.8 6.9 7.7 7.1 6.5 7.7 7.0 6.8 6.9 7.2 7.6

pH

1.34 0.91 1.74 0.91 1.25 0.65 0.78 1.55 0.41 1.40 0.99 1.11 0.41 1.74 1.09 0.66 1.05 1.57 1.70

Cond. (mS cm 1) 110.0 40.0 184.0 104.0 119.0 69.2 69.3 92.4 28.7 159.0 52.0 131.0 28.7 184.0 96.6 41.2 98.2 161.5 178.8

Ca (mg l 1) 10.9 10.8 6.8 7.3 6.5 5.4 6.8 9.1 4.0 7.7 19.2 6.1 4.0 19.2 8.4 5.5 7.1 10.1 17.1

K (mg l 1) 27.8 26.4 46.4 14.4 16.4 11.2 13.0 42.7 5.9 34.5 24.5 35.8 5.9 46.4 24.9 11.4 25.5 43.1 45.6

Mg (mg l 1) 140.0 101.0 94.2 69.5 113.0 46.0 48.3 168.0 35.3 72.5 93.0 54.4 35.3 168.0 86.3 46.2 82.8 118.5 156.6

Na (mg l 1) 713.7 433.1 701.5 420.9 390.4 298.9 311.1 488.0 164.7 610.0 359.9 481.9 164.7 713.7 447.8 300.1 427.0 619.2 682.6

HCO3 (mg l 1) 137.0 89.0 182.0 96.0 183.0 58.0 60.0 238.0 41.0 172.0 76.0 92.0 41.0 238.0 118.7 58.2 94.0 188.5 226.6

Cl (mg l 1)

35.0 0.0 105.0 36.0 81.0 21.0 21.0 86.0 16.0 26.0 22.0 48.0 0.0 105.0 41.4 16.5 30.5 87.9 101.1

SO4 (mg l 1)

Chemical analyses and physical parameters of 12 samples in the protected area of Castelporziano, 2004.

Parameter Sample

Table 4.2.3

1174.0 700.0 1320.0 748.0 909.0 510.0 529.0 1124.0 295.0 1082.0 647.0 849.0 295.0 1320.0 823.9 511.9 798.5 1143.6 1279.4

TDS (mg l 1)

The Pilot River Basin Network: Examples of Groundwater-related Activities 139

140

Chapter 4.2

Figure 4.2.6

Delimitation of conductivity levels in the protected area of Castelporziano as measured in 2004.

Historical data on groundwater conductivity before that saline intrusion occurred are not available; therefore sampling points located in the protected area of Castelporziano characterised by the lowest salinity values were selected as NBL value identiﬁcation. In this case the NBL value is about 1000 mS cm 1 lower than the P90 values in the areas of about 1300 mS cm 1, as shown in Tables 4.2.2 and 4.2.3. The threshold value was calculated as 2000 mS cm 1, since an EQS value derived from a consolidated methodology is not given. The EQS value will be set by means of expert judgement and probably it will be inferior to the threshold value, considering the particular naturalistic value of the area. The application of the methodology in the protected area of Castelporziano showed that the threshold value for parameters measuring marine water intrusion were exceeded in sample coastal areas, showing a general upward trend of the values from 1999 to 2004.

4.2.5

Conclusions

The chemical composition of the water in the Colli Albani aquifer system is very variable due to the nature of the rocks and the interaction with gasses and

The Pilot River Basin Network: Examples of Groundwater-related Activities

141

ﬂuids originating at depth. The statistical treatment (P90, P97.7) of a large number of samples from diﬀerent parts of the aquifer can lead to assessment errors of NBL values even though they are not subject to anthropogenic pressure. The application of WP3-BRIDGE methodologies under these conditions is useful for the purpose of threshold identiﬁcation and trend reversal measures only if the geochemical groundwater families that interact with the receptors are known. WP3 methodologies were applied in two areas of the hydrostructure where there were suﬃcient data to characterise groundwater and to identify signiﬁcant parameters of receptor interaction. A general lowering of the piezometric level is visible due to intense water abstraction. Therefore, quantitative recovery should be the ﬁrst objective of the programme of measures to be adopted by the management plan of the WFD. The Colli Albani aquifer is currently subject to safeguard measures aimed at limiting abstraction permits in critical areas where there are concentrated withdrawals.

References 1. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities L 327, 22.12.2000. 2. L. Galbiati, J. M. Zaldivar, F. Somma, F. Bouraoui, M. C. Moren Abat, G. Bidoglio and J. D’Eugenio, Pilot River Basin Outcome Report: Testing of the WFD Guidance Documents, EUR Report 21518, 2005. 3. Ph. Quevauviller, P. Balabanis, C. Fragakis, M. Weydert, M. Oliver, A. Kaschl, G. Arnold, A. Kroll, L. Galbiati, J. M. Zaldivar and G. Bidoglio, Environ. Sci. Pol., 2005, 8, 203. 4. Bridge general application and evaluation of a proposed methodology for derivation of groundwater threshold values: a case study summary report. 5. A. Di Domenicantonio, M. Ruisi and P. Traversa, Groundwater natural background levels and threshold deﬁnition in the Colli Albani volcanic aquifers in central Italy, Autorita` di Bacino del Fiume Tevere BRIDGE Project, 2007. 6. Directive 1998/83/EC of the Council of 3 November 1998 on the quality of water intended for human consumption, Oﬃcial Journal of the European Communities L 330, 05.12.1998. 7. M. Bucci, Stato delle risorse idriche, Il sistema ambientale della tenuta presidenziale di Castelporziano, Accademia Nazionale delle Scienze detta dei Quaranta, Scritti e documenti XXVII, 2006, pp. 327–387.

CHAPTER 4.3

The Harmoni-CA Initiative GEO E. ARNOLD, WIM J. DE LANGE AND MICHIEL W. BLIND RIZA, PO Box 17, NL-8200 AA Lelystad, The Netherlands

4.3.1

Introduction

The need to improve the integration of research into the policy-making process is one of the major challenges in managing complex environmental problems like water resources management. For many years, research and technology development (RTD) activities have paid more and more attention to incorporating policy-relevant topics in their research agendas. Current RTD projects have established operational links with practitioners. However, the objective of transferring newly developed tools from the research community to operational use by water managers has not been achieved and the eﬀorts of diﬀerent projects have not been co-ordinated. One of the actions for enhancing the use of tools and closer cooperation between the RTD and the European Union (EU) Water Framework Directive (WFD) worlds was the establishment of the Harmoni-CA (Harmonised Modelling Tools for Integrated Basin Management, EVK1-2001-00192) project in October 2002. Harmoni-CA is a concerted action, supported by the European Commission (DG RTD) under the 5th Framework Programme, which aims to facilitate the speciﬁc clustering activities. Harmoni-CA aims to facilitate the dialogue and help bridge the gap between research and policy, by synthesising the available knowledge produced by the various RTD (CatchModw) projects and facilitating the development and the use of these methodologies and tools to support the use of information, communication and technology (ICT) tools in implementing the WFD. Harmoni-CA started in October 2002 and will ﬁnalise in September 2007. This chapter gives a brief description of the way Harmoni-CA started the process of bridging the gap between research and policy. The process as initiated by the Harmoni-CA project and the lessons learned will be of interest w

CatchMod is a group of EC FP5- and FP6-funded projects aiming at development of ICT tools and supporting methodologies for integrated river basin management (IRBM).

142

143

The Harmoni-CA Initiative

for the implementation of the new Groundwater Directive, adopted by the European Parliament in December 2006.

4.3.2

The Harmoni-CA Initiative

The concerted action Harmoni-CA, supported by the European Commission (EC) under the 5th Framework Programme,1 aims to facilitate the dialogue and to help bridge the gap between research, consultancy, operational management and policy. Harmoni-CA therefore synthesises the available knowledge produced by the various RTD (CatchMod) projects and facilitates the development and the use of these methodologies and tools to support the use of ICT tools in implementing the WFD (Figure 4.3.1). The long-term objective of Harmoni-CA is to set up a forum for communication, information exchange and the harmonisation of the use and development of ICT tools that goes beyond the duration of Harmoni-CA. Harmoni-CA therefore formulates a set of speciﬁcations for the continuation and extension of the communication forum.

4.3.3

Process of Bridging the Gap Between Research and Policy/Water Management

As described above, Harmoni-CA acts along two tracks. Firstly, Harmoni-CA facilitates activities within the CatchMod cluster, like the identiﬁcation and enhancement of complementarities between diﬀerent research projects, and disseminates research results focusing on the projects in the EC-supported CatchMod modelling cluster. Secondly, Harmoni-CA brings together the demand and support for ICT tools and methodologies for the implementation of the WFD. The activities can be divided into a process-related activity and products/tools that support this process. Establishing a dialogue among the scientiﬁc and policy-making communities is one of the most important achievements of Harmoni-CA. For establishing a dialogue among the scientiﬁc and policy-making communities Harmoni-CA organised yearly forums and conferences. The forums and conferences were supported by workshops on speciﬁc topics.

Harmoni-CA

Fundamental Research

Figure 4.3.1

Applied research (e.g. Catch Mod)

mediation

WFD implementation

The role of Harmoni-CA between WFD and applied research.

144

4.3.3.1

Chapter 4.3

Harmoni-CA Forums and Conferences

As stated above, the aim of organising forums and conferences was to improve the relation between research and policy. The ﬁrst step in bringing policymakers and researchers together was to understand the demands and needs from the policy and water management side for the implementation of the WFD and hence the support to be given from the research and development communities. The ‘‘demand ’’ from the policy-makers and the operational managers was based on the experiences in the WFD pilots and in the implementation of the WFD. The ‘‘support’’ from the methodology and technology providers was given by the EC-supported research (mainly CatchMod) projects, national initiatives and other sources. The main objective of the forums and conferences was to facilitate the dialogue between these groups (Figure 4.3.2). The forums and conferences were targeting activities following the time schedule of the WFD implementation, such as the characterisation reports (WFD, Article 5), the monitoring activities and the preparation of integrated river basin management plans. The focus of the conferences was on stimulating the networking between the four target groups involved. During the ﬁrst conference (February 2004) it became clear that a large gap existed between the ‘‘demand’’ of the policy-makers and the operational managers and the ‘‘support’’ oﬀered by the scientists. This gap is for diﬀerent reasons, among others: scientists and policy-makers speak a diﬀerent language and they have diﬀerent interests and targets, diﬀerent agendas and diﬀerent timetables. Discussions between scientists, policy-makers and stakeholders showed that knowledge generated by many research and demonstration

Policy makers

Harmoni-CA Operational managers

Forums and Conferences

Methodology providers

Technology providers

Figure 4.3.2

Target groups for Harmoni-CA (forums and conferences).

The Harmoni-CA Initiative

145

projects does not reach policy-makers in an eﬃcient way. On the other hand, the policy-making community does not consider research results, as it should do, mainly for political reasons and diﬃculties in integrating research developments in legislation. This is further discussed in Chapter 11.3. An important conclusion of this conference was the need to improve and to make operational a ‘‘science–policy interface’’ linked to the implementation of the WFD. As a result a common ‘‘scope paper’’2 has been drafted to strengthen the cooperation between CatchMod/Harmoni-CA (under the umbrella of DG RTD) and those responsible for implementing the WFD (under the umbrella of DG Environment). In this scope paper the three following activities were deﬁned: linking WFD requirements and RTD products; building of a web portal; and close co-operation within Pilot River Basins (PRBs). Since 2004 these actions have played a central role in the activities of HarmoniCA.3 At the same time the EC started a discussion on the integration of scientiﬁc and technological progress into the policy-making and implementation process.4 A need was felt for science–policy integration in the implementation process of the EU WFD.

4.3.3.2

CatchMod/Harmoni-CA Workshops

In combination with and additional to the conferences Harmoni-CA supported the organisation of CatchMod/Harmoni-CA workshops aiming to facilitate activities within the CatchMod cluster, like the identiﬁcation and enhancement of complementarities between diﬀerent research projects and to strengthen the discussion and contacts between researchers and operational managers. The workshops were organised around speciﬁc topics, in relation to ICT tools and focused on the tasks and activities of the diﬀerent Harmoni-CA work packages (toolboxes, planning methodology, joint use of monitoring and modelling and public participation). Since the end of 2003, but in particular since the 1st Harmoni-CA Forum and Conference in April 2004, the CatchMod/Harmoni-CA consortium started to link its activities with some Common Implementation Strategy (CIS) activities initiated by DG Environment, policymakers and water managers.

4.3.3.3

Conclusions and Lessons Learned from the Conferences and Workshops

Some conclusions and lessons from the conferences and workshops organised by Harmoni-CA/CatchMod are the following. Conferences, workshops, action plans, reports, websites and newsletters are good communication tools but should be ‘‘tailor-made’’ for each user category (policy-makers, operational managers, stakeholders, etc.).

146

Chapter 4.3

New research is too late for actual use. This is often the case for new research projects under the 6th EU Framework Programme and certainly for new research programmes under the 7th Framework Programme. Exceptions are those projects that are truly linked to a WFD activity (e.g. projects for policy support, such as REBECCA (http://www.environment.ﬁ/syke/rebecca) and BRIDGE (see Chapter 9.1)). Many research projects provide results and tools that do not yet meet the requirements for broad application by water managers. Some additional investments are required to make operational tools and create supporting organisations. In addition to the yearly conferences and workshops, it is of utmost importance to participate in regional meetings and river basin meetings to listen to and to discuss the experiences of water managers and to demonstrate and to bring locally relevant research to water authorities. There is a need for a closer co-operation between DG ENV and DG RTD, including the development of funding mechanisms for joint research projects in which the role of water managers is strengthened.

4.3.4

Products/Tools of Harmoni-CA

Beside the process as described above, Harmoni-CA is delivering products and tools to support the communication process. Important products are the WISE-RTD web portal, guidance documents, synthesis reports and summaries.

4.3.4.1

WISE-RTD Web Portal

In close co-operation with Commission services Harmoni-CA developed a web portal which aims to provide direct access to scientiﬁc information supporting water policy implementation (see also Chapter 11.3). This portal is under development and will be linked to the oﬃcial launch of WISE (March 2007). The WISE-RTD portal provides an intelligent search of information on tools and experiences issued from RTD projects, as well as guidance documents in support of the implementation of the WFD. The search is based on WFDspeciﬁc issues (e.g. WFD milestones, WFD terminology) and serves multiple user groups (policy-makers, water managers, stakeholders, modellers, etc.). Hence the portal links ‘‘demands and support/oﬀers’’, while taking the diﬀerent languages/terminologies of the target groups into account. The users can get support from a ‘‘Communication Service Center’’ (CSC). Figure 4.3.3 shows the main idea behind the web portal. WISE-RTD links to websites that contain a wide range of information such as CIS guidance documents, reports from PRBs, reviews and selections of ICT tools, or results of national and EC-funded projects (e.g. the CatchMod cluster). The system already links to more than 60 research and application projects, over 100 tools and provides direct entry inside CIS guidance and

The Harmoni-CA Initiative

Figure 4.3.3

147

Information ﬂow.

technical guidance documents (January 2007 status). The portal prototype can be viewed at www.wise-rtd.info.

4.3.4.2

Guidance Documents, Synthesis Reports and Summaries

Another activity to support the process of matching the demand and support for knowledge and ICT tools for implementing the WFD is the preparation of guidance documents and synthesis reports. Guidance documents are meant to give guidance to the implementation of activities or tasks. Target groups are operational managers or more general, implementers of a certain task within the WFD implementation. Examples of guidance documents in preparation are:

Uncertainty Calibration Sensitivity analysis Environmental economics Quality assurance in modelling Public participation Monitoring network design Planning methodology for the WFD.

148

Chapter 4.3

Besides guidance documents activities (synthesis and summaries) are carried out within Harmoni-CA, which bring results of projects forward, provide support to new research, etc. Synthesis reports give an evaluation of results on a study or a methodology, etc. Synthesis reports describe amongst others the area of application, gaps, etc. Target groups can be researchers or at a more strategic level policy-makers, water managers and also research funding boards. Examples of these activities are: improving the toolbox on nutrient emission tools to include additional tools; developing a portal for useful standards relating to software in water management; improving quality assurance support tools to make them better adaptable to new (modelling) processes; analysis of decisions support systems, leading to advice on how to develop such systems successfully; and analysis of end-user involvement in European-funded projects leading to advice as to how this can be done more eﬀectively, and thus should lead to improved outputs of research with respect to operational water management. The guidance documents and synthesis reports will also be incorporated in the web portal.

4.3.5

SPI-Water

Four years into the Harmoni-CA project it can be stated that a fruitful communication between research and policy has been established. However, the duration of the concerted action Harmoni-CA is limited. The project started in October 2002 and will end in October 2007. As stated before, the long-term objective of Harmoni-CA is setting up a forum for communication, information exchange and the harmonisation of the use and development of ICT tools that will go beyond the duration of Harmoni-CA. This objective will partly be carried out by SPI-Waterz, a new project that is funded by the 6th Framework Programme, as part of Priority 8.1: policy-oriented researchscientiﬁc support to policies. SPI-Water elaborates on a theme already started by Harmoni-CA and proposes a number of concrete actions to bridge the gap in communication by developing and implementing a ‘‘science–policy interface,’’ focussing on setting up a mechanism enhancing the use of RTD results in the WFD implementation. As a ﬁrst action, existing science–policy links will be investigated. RTD and LIFE projects with direct relevance for the implementation of the WFD will be identiﬁed and analysed and their results will be extracted, ‘‘translated’’ and z

Science–Policy Interfacing in support of the WFD implementation.

The Harmoni-CA Initiative

149

synthesised in a way that can eﬃciently feed the WFD implementation. Secondly, the information system (WISE-RTD web portal) will be further developed to cater for an eﬃcient and easy-to-use tool for dissemination as well as retrieval of RTD results.5 The web portal will be tested in four selected river basins to better tune the ‘‘product’’ to the needs of WFD stakeholders, policymakers and scientists. In parallel, the web portal will be disseminated to WFD stakeholders. As a third and last action, this science–policy interfacing of WFD-related topics will be extended to non-EU countries, taking their speciﬁc needs into account. An assessment of recent practices and needs of non-EU countries, together with an in-depth analysis of the operational needs in two Mediterranean pilot river basins, will allow the preparation of recommendations for an eﬃcient transfer of knowledge.

4.3.6

Groundwater Directive

The initiatives started by Harmoni-CA and the lessons learned during the Harmoni-CA project will be very important for the implementation of the Groundwater Directive (see Chapter 3.1). Indeed, groundwater and surface waters are strongly connected and a lot of the tools already developed and experiences gained can be used for groundwater. Communication between water managers, policy-makers and surface water specialists is not easy. For groundwater, an invisible and ‘‘hidden’’ water resource with quite diﬀerent characteristics, this is even more diﬃcult. In this case public participation, one of the topics within the Harmoni-CA project, can play an important role. The WISE-RTD web portal, already developed by Harmoni-CA and continued and extended by the new SPI-Water project, will be a challenge for the implementation of the water policies in general, and the new Groundwater Directive in particular. All groundwater activities and projects are invited to make use of this tool.

References 1. Description of Work HarmoniCA (Harmonised Modelling Tools for Integrated Basin Management), EVK1-2001-00192, 2002. 2. G. Arnold and Drafting Group, Research supporting the WFD implementation, Mutual gains from cooperation (scope paper), Lelystad, The Netherlands, 2004. 3. G. E. Arnold, W. J. De Lange and M. W. Blind, Environ. Sci. Pol., 2005, 8, 213–218. 4. Ph. Quevauviller, P. Balabanis, C. Fragakis, M. Weydert, M. Oliver, A. Kaschl, G. Arnold, A. Kroll, L. Galbiati, J. M. Zaldivar and G. Bidoglio, Environ. Sci. Pol., 2005, 8, 203–211. 5. P. Willems and W. J. De Lange, Environ. Sci. Pol., in press.

CHAPTER 4.4

Linking Public Participation to Adaptive Management CLAUDIA PAHL-WOSTL, JENS NEWIG AND DAGMAR RIDDER Institute of Environmental Systems Research, University of Osnabru¨ck, Barbarastrasse 12, DE-49069 Osnabru¨ck, Germany

4.4.1

Introduction

In recent years the awareness that sustainable and integrated water resource management cannot be realised based on expert knowledge and technical solutions alone has increased. Participatory approaches are required in which stakeholders are involved in developing, implementing and monitoring management plans to cope with the complexity of issues to be tackled and the ensuing conﬂicts of interest.1 Such is also the spirit of the novel European Water Policy aiming at the integration of the hitherto existing fragmented regulatory framework. The European Water Framework Directive (WFD; Directive 2000/60/EC) can be labelled as an example of a new generation of European Union (EU) directives with which the EU seeks to partly overcome the established technocratic, top-down method of European policy-making. The EU member states have more freedom to develop an implementation plan targeted towards their needs, taking into account national and regional conditions.2 Organised interest groups and the public at large are to be involved in the process. The upcoming Groundwater Directive should be implemented in the same spirit and public involvement can be expected to encounter similar challenges in implementing participatory approaches as experienced during implementation of the WFD. The recognition of uncertainties such as climate change or changes in administrative systems underline the need for new, more adaptive management styles in water management in general and groundwater management in particular. Participatory processes are expected to foster the learning of individuals and groups. Eventually, social learning processes are expected to support adaptive management. One precondition for achieving this is 150

Linking Public Participation to Adaptive Management

151

participatory processes that go beyond informing or consulting the public by trying to actively involve the relevant stakeholders in decision-making.

4.4.2

Adaptive Management and Public Participation

Water management in Europe largely lacks experience and expertise in involving the public. At the operational level of water management, public participation is often perceived as being too resource intensive and is seen as a barrier to eﬃcient management rather than an opportunity for developing and implementing innovative management approaches. However, innovation towards more ﬂexible governance systems and management strategies that take diﬀerent kinds of uncertainties into account are urgently neededw. We argue that stakeholder participation is required to develop, implement and sustain such management approaches, taking into account that: ambiguity exists when deﬁning operational targets for the diﬀerent management goals to be achieved, and that conﬂicts of interest require participatory goal setting (not by experts alone) and a clear recognition of uncertainties in this process; the outcomes of management measures are uncertain due to the complexity of the system to be managed and to uncertainties in environmental and socioeconomic developments inﬂuencing the performance of implemented management strategies; new knowledge about system behaviour may suggest options for change in management strategies; and changes in environmental and/or socioeconomic conditions may demand changes in management strategies. Given current water management practice, one can identify a clear need for a more coherent and comprehensive approach, an approach based on sound conceptual foundations to deal with uncertainties in water management in general and groundwater management in particular. The idea of adaptive management has already been the subject of ecosystem management discussions for some time.3–6 It is based on the insight that the ability to predict future key drivers inﬂuencing an ecosystem, as well as a system’s behaviour and its responses, is inherently limited. Hence management must be adaptive and must include the ability to change managerial practices based on new insights. One form of adaptive management uses management programmes that are designed to experimentally compare selected policies or practices by evaluating alternative hypotheses about the system being managed (e.g. Refs. 7–9). However, adaptive management is not limited to an experimental approach but can more generally be deﬁned as a systematic process for improving management policies and practices by learning from the outcomes of w

The development of such strategies is the main objective of the EU-funded project NeWater ‘‘New approaches to adaptive water management under uncertainty’’ (www.newater.info).

152

Chapter 4.4

implemented management strategies. As Bormann et al.10 pointed out, ‘‘Adaptive management is learning to manage by managing to learn.’’ Learning may encompass a wide range of processes that span the ecological, economic and sociopolitical domains in the testing of hard and soft approaches.11,12 Adaptive management emphasises the importance of the process nature of management without claiming that the process is an end in itself but by explicitly recognising that management strategies and even goals may have to be adapted during the process. We argue that this can be carried out most eﬀectively if a learning cycle unites all the relevant actors in the diﬀerent phases of policy development, implementation and monitoring. The actors to be involved can be represented by the public at large. Nonetheless, the more common and realistic form of participation in adaptive management is stakeholder participation: organised groups are represented by one or more persons. Organised groups can be very diverse, e.g. environmental NGOs, groups representing social classes, age or gender groups, lobby groups of industrialists, or they can even be another state or regional authority that has traditionally not been involved in the decisionmaking process. The whole adaptive management process (see Figure 4.4.1) requires a number of steps that are part of an iterative cycle. All steps should be participatory (0, 1, 2, 3, 4). In a participatory process diﬀerent perspectives need to be taken into account in the deﬁnition of the problem (0). The design of policies should include scenario analyses to identify key uncertainties and to ﬁnd strategies that perform well under diﬀerent possible but initially uncertain future developments rather than searching for a strategy that performs optimally under very speciﬁc

Figure 4.4.1

Iterative cycle of policy development and implementation in adaptive management.

Linking Public Participation to Adaptive Management

153

conditions (e.g. climate) but performs purely if these conditions are not met (1). Policies must be understood as semi-open experiments that require the careful evaluation of potential positive or negative feedback mechanisms by planning and implementing other related policies (1, 2). Decisions should be evaluated by the costs of reversing them. Large-scale infrastructure or rigid regulatory frameworks increase the costs of change. But costs may also be related to a loss of trust and credibility if uncertainties and the possible need for changes are not addressed by the competent authority during policy development (3). Monitoring programmes should include processes to highlight undesirable developments at an early stage. This might imply diﬀerent kinds of knowledge, including community-based monitoring systems13 (3). The policy cycle must include institutional settings in which actors assess the performance of management strategies and implement change if needed (4). Continuous replanning and reprogramming based on the results of monitoring and evaluation should be institutionalised (4).

In principle, the WFD is compatible with such an iterative and adaptive approach. However, implementation seems to follow largely established management practice based on a more linear and optimisational approach. Uncertainties are taken into account only to a limited extent.14 The implementation of an adaptive management approach is only possible if certain structural conditions are fulﬁlled. Hence the implementation of adaptive management needs an integrated system linked to a new understanding of management as learning rather than a control process. Response to new insights is only possible if the measures implemented can be changed. The transition to adaptive management relies on increasing the adaptive capacity of the (water) system. It aims at an integrated system based on the understanding of the interdependence between technologies, economic and environmental factors and the formal and informal institutional context. The aim is to increase the ability of the entire system to respond to change rather than to simply react to undesirable impacts of change. Institutionalising this learning capability in the long run will secure the adaptive foundation of management. What are the current requirements for the adaptive management of groundwater resources? New information must be made available and/or consciously collected (e.g. indicators of the performance of management regimes like the relation of groundwater retrieval to groundwater recharge) and monitored over appropriate time scales (given the often very slow percolation of pollutants through the soil and the long groundwater travel times, monitoring time scales need to be considerably longer than those mandated by short-term political objectives).

154

Chapter 4.4

The actors within the management system must be able to process this information and draw meaningful conclusions from it. This can be best achieved if the learning process unites the actors in all the phases of assessment, policy implementation and monitoring. If the information collected provides answers to the questions (hypotheses) they posed, then management is transparent to all those involved. Change must be possible in ways that are open and understandable to all actors. Management must have the ability to implement change, based on processing new information in a learning process where it is clear as to who decides how and when to change management practices, based on which evidence, and why.

Such conditions are not met in many countries in Europe, in particular in groundwater management where long-lasting conﬂicts may prevail. In the upper Guadiana basin, for example, groundwater use exceeds capacity and the groundwater table drops due to the existence of numerous illegal wells. Representatives from the competent authority argue for the need to enforce legal regulations and for the wells to be closed down by governmental intervention (personal communication). However, given the failure of such attempts in the past one may need to reconsider whether participatory approaches based on learning rather than on control may have a greater chance of triggering change. It can be concluded that adaptive management requires the incorporation of a learning element in its management process. This learning component encompasses social learning: in which the actors involved learn about their diﬀerent perspectives with regard to groundwater management and their motives behind opting for certain management strategies and actions. Only then can it be guaranteed that the planned activities respect the process nature of groundwater management and that management strategies leave suﬃcient room for reaction to these changes. Active participation is the foundation for achieving social learning and adaptive management. In order to guarantee that new information is freely distributed and equally understood, the learning cycle must unite all the relevant actors in the diﬀerent phases of policy development, implementation and monitoring. The actors to be involved can be represented by the public at large or by stakeholders. The more common and realistic form of participation in adaptive management at this stage is stakeholder participation: organised groups are represented by one or more persons. Organised groups can be very diverse including, e.g. environmental NGOs, groups representing social classes, age or gender groups, lobby groups of industrialists, or they can even be another state or regional authority that was traditionally not involved in the decision-making process. In addition to the current forms of participation such as informing and consulting other people or groups, new, more active forms of participation will be required to optimise the learning process and eventually the adaptability of management.

155

Linking Public Participation to Adaptive Management

4.4.3

Rationales and Requirements for Eﬀective Participation in Groundwater Management

4.4.3.1

Rationales and Goals for Public Participation

‘‘Public participation is not an end in itself but a tool to achieve the environmental objectives of the Water Framework Directive.’’15 This quote from the CIS guidance document on public participation according to the WFD perhaps best describes the dominant rationale for public participation in the current EU water policy. Thus, the current emphasis on participation seems to be predominantly rooted in a certain disillusionment in the eﬀectiveness of governmental steering eﬀorts in the face of the continuing implementation deﬁcits of state environmental policy.16–18 It expresses both a hope and an expectation that participatory processes will lead to an improved compliance and implementation (measured by the agreed environmental goals) due to a more sound knowledge base and an improved acceptance of decisions: in short, an enhanced eﬀectiveness of the pursued policy.19,20 Moreover, some observers generally expect that an increasing societal complexity requires poly-centric and participatory modes of governance.1,16,21,22 Taking a closer look at the CIS guidance document on public participation, two main strands of arguments become apparent (see Table 4.4.1). One is that the quality of decisions is expected to be better: In the course of the participatory process, information is generated or made available that would not have been so otherwise; furthermore, the decision beneﬁts from this information, i.e.

Table 4.4.1

Diﬀerent rationales for public participation as they appear in the CIS Public Participation Guidance Document relative to the Water Framework Directive.15 (Table after Ref. 60.)

Rationales for public participation

CIS PP Guidance

Quality of decision

pp. 24, 26, 41

Quality of implementation

Make available lay local knowledge to the authority Make available knowledge regarding attitudes and acceptance on the part of the non-state actors to the authority Improve environmental quality, reach environmental goals Increase environmental awareness, education, information on the part of the non-state actors Build acceptance of and identiﬁcation with a decision on the part of the nonstate actors Build trust among non-state actors and between these and the authority Alleviate conﬂicts by mediation of interests

p. 24 pp. 7, 26 pp. 4, 26 pp. 4, 26, 41 pp. 26, 41 pp. 26, 41

156

Chapter 4.4

the information is actually incorporated into the decision. Thus, water policy and management decisions can beneﬁt from the factual knowledge of involved actors about their (local) conditions,23–25 assuming that those who are closest to a problem develop the best understanding of it.26,27 Other authors, however, contest this claim and hold that it is rather the authorities who have diﬀerent and usually more reliable means of information provision at their disposal,28 especially regarding highly technical issues, and the corresponding need for specialised expert knowledge.29 Then again, there may be information that ‘‘emerges’’ from the close interaction of actors in a group process. Many authors stress the positive eﬀects of social learning, the plurality of perspectives and thus the more creative decision-making as characteristics of participatory decision-making.30 Yet group processes also have the potential to create adverse eﬀects. For instance, Cooke31 points out problematic ﬁndings from social psychology regarding consensus-oriented group processes, such as the tendency towards taking risky decisions or an immunisation towards independent and critical arguments. Another type of information from which decisions could proﬁt is information regarding the extent to which planned measures will be accepted by the addressees. In this respect, participation becomes an ‘‘instrument for the anticipation of resistance to planning and implementation.’’32 Generally, participation is expected to prevent implementation problems from occurring.33 Quite plausibly, the addressees of a decision must know of it in order to be able to implement it: obey rules, comply with requirements. If future addressees are involved in decision-making, they can be assumed to be thoroughly informed about these decisions, and a higher rate of compliance can reasonably be expected, as the possibly necessary measures of reorganisation and adaptation to new (regulatory) conditions, which usually take some time, can duly be taken. Furthermore, compliance with a decision is expected to depend positively on the degree of acceptance, or even identiﬁcation, on the part of the addressees (e.g. Refs. 33, 34). Acceptance may, ﬁrstly, be supported by providing the interested actors with early and comprehensive information. This may prevent actors from feeling left out or ignored, and create a sense of involvement and belonging. Also, certain educational eﬀects, e.g. in the sense of an improved environmental awareness, can play a role.35 Moreover, an intensive involvement of the concerned actors in a decision process that is perceived as fair and based on mutual communication is expected to enhance the acceptance of the decision. This even holds when the result does not correspond to the actors’ expectations,27,36 as procedural justice research has found that the acceptance of a decision crucially depends on aspects of fairness of the decision procedure.33,37–39 Furthermore, a decision that involves conﬂicting interests is more likely to be accepted by the diﬀerent parties if it is based on either a consensus or at least a compromise to which most of the parties agree. This in turn most likely requires an intensive participatory process that allows the concerned actors to eﬀectively claim their stakes, but also a spectrum of interests that does not fundamentally rule out any consensual solutions.

Linking Public Participation to Adaptive Management

157

Furthermore, in the medium and long term, the building of trust relationships among the non-state actors, the state actors and also between non-state and state actors through participation33 can lead to an increased regional collective social capital, and can thus inﬂuence the context of future decision processes. In particular, the building of trust can improve acceptance of and thus the willingness to comply with measures, as empirical studies in other contexts have shown.39

4.4.3.2

Public Participation Provisions in the WFD

The WFD demands diﬀerent forms of involving the public in decision-making, which are explicated in further detail in the PP Guidance Document, although of course not in a legally binding manner.

4.4.3.2.1

Legally Binding Three-stage Consultation Process for the River Basin Management Plans

Most stringently regulated is the consultation for the river basin management plans (RBMPs). RBMPs form the principal instrument of transparency and communication of the current status of waters and planned measures.40 From the end of 2006 onwards, the public must be informed, and can voice its concerns, at three annual intervals regarding the working programme (2006), the most important water management issues (2007) and the draft management plans (2008), leaving the public 6 months each time to produce written statements (Art. 14 (1.2) and (2) WFD). Since this formal consultation procedure only has to be implemented at the level of the rather large river basin districts, the impact of public participation on decision-making can be expected to remain rather poor. This is even more the case since the current cooperation in the river basin districts in federative EU member states such as Germany is restricted to exchange between authorities, while the actual planning process takes place at the level of the federal states. While consultation is only required for the RBMP and not for the more important programmes of measures, the latter is required by the SEA Directive. In practice, both consultations will most probably be combined into one single procedure (see Ref. 2).

4.4.3.2.2

Free Access to Background Information

According to Art. 14 (1.3) WFD, the public is to be granted free access to the documents used for preparing the RBMP on request.

4.4.3.2.3

Encouragement of ‘‘Active Involvement’’

Probably the most important provision regarding public participation is the required encouragement of the ‘‘active involvement’’ of all interested parties

158

Chapter 4.4

(Art. 14 (1.1) WFD). Active involvement diﬀers from the consultation process in three respects. First, relevant involvement relates to the ‘‘interested parties’’ and thus to a smaller circle of the public than in consultation. Second, ‘‘active involvement’’ implies a much stronger involvement of actors than in consultation. Third, ‘‘active involvement’’ relates to the implementation of the whole directive and not exclusively to the RBMP. Although this requirement is less legally binding than the consultation process and, moreover, not at all operationalised by the directive, regional authorities in the member states have already taken great pains to do this requirement justice (see the examples below).

4.4.3.2.4

Adaptive Regulatory Impact Assessment

As a learning instrument and to systematically monitor its success, the WFD requires reports to be drawn up about the public participation conducted. According to Art. 13 (4), RBMPs must also include information on how public participation has aﬀected or changed the plan. The Guidance Document stresses that this instrument serves not only ex-post control by the Commission but, predominantly, to improve public participation in the following planning cycle. While this instrument appears to have been underestimated until now, it allows the successive improvement of public participation from one planning cycle to another. The collection and systematic analysis of experiences enables the adaptation of public participation in the following cycle. Thus, for the ﬁrst time, legal evaluation is not only institutionalised as a retrospective regulatory impact assessment41 but, moreover, as an adaptive management procedure.

4.4.3.2.5

Imperative to Conduct Actor Analyses?

When the competent authorities determine the relevant public to be involved, diﬀerentiating according to the diﬀerent phases and demands of implementation, and also distinguish diﬀerent forms and degrees of participation—in short, when they tailor participation instruments to their target groups—this will in many cases require a systematic identiﬁcation, analysis and classiﬁcation of (potentially) relevant actors.15,42–44

4.4.3.3

German Experiences with Public Participation

Although the WFD commands implementation at the level of the river basin districts, the federal organisation in Germany has led to implementation structures at the level of the federal states. Thus, the most important public participation instruments are organised at state level. These diﬀer considerably from state to state. As Figure 4.4.2 illustrates, diﬀerent forms of participation have been institutionalised. Some are located at the state level (such as councils or a steering group) but most are situated at more local scales. And while some instruments are targeted at the public at large, others include only selected

159

Linking Public Participation to Adaptive Management

state level

Steering Group (1 state)

Council (7 states)

Road Shows (3 states) Regional Conferences (4 states) working level the public at large

Councils (1 state)

Area Cooperations (1 state)

Working Groups (2 states)

selected stakeholders

degree of participation and of decreasing publicness

Figure 4.4.2

Overview of public participation institutions (‘‘active participation’’) in German federal states (adapted from Ref. 2).

stakeholders, thus allowing for more intensive cooperation in the WFD’s implementation. On the whole, a surprising multitude of participation activities is being undertaken, covering information, consultation and active participation, even before the oﬃcial consultation process regarding the RBMPs has even started. The Groundwater Directive is not new in terms of substantive provisions but rather in that it expands the procedural law in terms of instruments in order to attain the water quality goals.45

4.4.3.4

Example: Regional Participation in Groundwater Protection from Agricultural Nitrate

This example focuses on the participation activities in Lower Saxony, Germany’s ‘‘principal agricultural state.’’ Agriculture is one main, if not the main, addressee of groundwater regulations because of its high potential to aﬀect groundwater quality due to its area-wide operation. Diﬀuse pollution due to nitrate is still one of the most important unresolved problems of groundwater quality. Experiences made during the implementation process of the WFD provide valuable insights for the groundwater directive, too. Information is provided by the authorities through a leaﬂet and two internet pages. Roadshows in diﬀerent places and for diﬀerent catchment areas (‘‘regional’’ and ‘‘area conferences’’) have been organised to inform

160

Chapter 4.4

stakeholders and the interested public about the ﬁrst inventories that were developed for the sub-basins within the administrative boundaries of Lower Saxony. The number of participants ranged between 70 and 200 persons. In January 2003, the Lower Saxon Ministry of Environment (MELS) established a council for the implementation of the WFD in Lower Saxony, aiming to inform the most important stakeholders at the state level, but also to allow some room for discussion as regards the current and future implementation steps. The council meets once or twice a year and is made up of about 50 representatives from diﬀerent sectors. At the ministerial level, a technical expert group involving about a dozen stakeholders has been established to support the development of the methodological basis for the implementation of the WFD, with a speciﬁc emphasis on the objectives for groundwater bodies. Although the oﬃcial process for informing and consulting the public has not yet begun, the variety of the more or less institutionalised participation eﬀorts illustrated above show that some of the provisions of the WFD have already been surpassed by the water authorities in Lower Saxony. On the other hand, these institutions do not allow non-state actors to participate in planning concrete measures at the regional and local scales and therefore hardly meet the needs of regional stakeholders, since the meetings are organised at higher levels, making the translation of partly abstract and technical decisions into the lower levels diﬃcult. Moreover, the number of participants, e.g. in the council, is too large for a constructive working atmosphere. Consequently, the MELS established a more local and direct form of active involvement. In autumn 2005, 30 so-called ‘‘areas of cooperation’’ were initiated at the sub-sub-basin level covering the whole of Lower Saxony. They were designed as long-term institutions that would contribute to the formulation of the RBMPs, while leaving the ﬁnal decision competence with the state authorities (MU Niedersachsen 2005, p. 2). Although the oﬃcial consultation process at the level of the whole river basin districts, starting by the end of 2006, could also inﬂuence the implementation of the WFD, the most important discussions, and perhaps decisions, will take place within these areas of cooperation. They hold a great potential to break up the existing alliances of agriculture and state actors and allow for a true consideration of the environmental goals provided by the WFD. Whether or not, and to what extent, the area of cooperation will succeed in terms of the stringent implementation of WFD demands and whether the measures decided upon by the area of cooperation will in fact lead to substantially reduced nutrient intakes into the region’s waters depends on a series of factors that will be analysed in the next section. The successful implementation of the WFD in a region with intensive agriculture surely depends on many aspects. First of all, the historical and economic role of agriculture, the regional experiences in water management and the diﬀerent resources and interests of actors are decisive. A crucial aspect for Lower Saxony in general will be the speciﬁc motivation of stakeholders as to why and how the WFD should be implemented. Is the WFD only seen in terms of fulﬁlling European legislation or is there a true desire to improve the status of water bodies? Where the latter is the case—which will be true if a clear

Linking Public Participation to Adaptive Management

161

beneﬁt for the person or group in question becomes obvious—the area of cooperation can be of great assistance in achieving ecological targets, without completely dismissing economic interests. Exploring all possible measures and instruments that can contribute to good water status and negotiating the opportunities is a great chance for stakeholder involvement and also gives weak actors the chance to state their interests.46 Integrating new actors and perspectives and developing the capacity for eﬀective and eﬃcient communication require processes of social learning.

4.4.4

Social Learning in Public Participation: Support for Adaptive Management

What is social learning? As already explained above, adaptive management places an emphasis on the improvement of management processes by learning. Here, it becomes obvious how the concept of social learning has greatly inﬂuenced the development of the meaning of ‘‘adaptive management.’’ Learning at its best should be an active process, and social learning with respect to sustainable development is based on the participatory processes of social change and societal transformations.16 The necessity of participatory approaches is also considered as crucial in adaptive management processes. Therefore both concepts, social learning and adaptive management, cannot be applied without active stakeholder involvement in planning and decisionmaking. In a very broad sense, social learning is referred to as building knowledge within groups, organisations or societies. Mostert47 explains it as ‘‘the growing capacity of social entities to perform common tasks, such as the management of a water resource.’’ But it should not be forgotten that it is often also an individual’s gain in knowledge in a well-managed stakeholder process that can make the diﬀerence. The concept of social learning can therefore very generally support the traditional use of economic and hydrological information as well as general expert knowledge in making water management decisions. In bringing together a large spectrum of relevant actors for groundwater management not only communication will be improved and thus information better exchanged, but also new information will be gained. A better understanding of feedback mechanisms and actors’ dependencies will evolve. This newly acquired knowledge facilitates better cooperation and eventually solution-ﬁnding in a team. The HarmoniCOP projectz resulted in a handbook on improving social learning and participation in water management. This process was brieﬂy summarised as ‘‘learning together to manage together.’’48 The major diﬀerence of this interpretation of social learning from Bormann’s13 ‘‘adaptive management is learning to manage by managing to learn’’ is its focus on learning as a collective experience. z

Harmonising Collaborative Planning (HarmoniCOP), an EU project funded within the 5th framework programme (www.harmonicop.info).

162

Chapter 4.4

The following aspects should be distinguished in order to create conditions conducive to achieving social learning:

appropriate framing conditions (legal framework/authorities); well-designed process management; well-selected methods and tools; and leadership issues.

If water managers manage to use active participation and social learning to result in collective decision-making, it feeds back to the water managers and ﬁnally to policy-makers. Successful social learning therefore ampliﬁes its impact by continuously improving its positive policy environment, and thereby supports the idea of the iterative policy cycle (see Figure 4.4.1).

4.4.4.1

Appropriate Framing Conditions

Social learning not only beneﬁts the creation of new joint ideas and common visions to realise bottom-up planning of societal relevance, but it can and should also support institutional change. In practise, this could be realised, for example, by the improved cooperation of agencies of agriculture with agencies of nature and water protection. During the discussions on how to implement the participatory aspects of the European WFD it was requested at an early stage that the environmental measures in agriculture should become an integral part of implementing the WFD.49 Corresponding to the requests of the WFD to implement water protection measures in a cross-sectoral manner, it will also require a change in behaviour of the relevant actors coming from diﬀerent ﬁelds of policy.50 Here, the above-mentioned change of relations and the knowledge built between individuals, groups and organisations helps change the common practise by increasing the understanding of the complexity of the problem. Compared to merely having information, social learning is therefore the more sustainable form of participation: facilitating adaptive management and supporting institutional change into a desired direction determined in a transparent process of negotiation. Social learning therefore accordingly supports the demand of the WFD (‘‘active involvement of all interested parties’’ according to Art. 14 WFD and of the guidance document on ‘‘Public participation in relation to the WFD,’’ Section 3: ‘‘Active involvement of all interested parties in the planning process of the directive’’15 to involve a large spectrum of actors. Aspects such as trust, social learning and the building of networks are nowadays considered to be the key elements of sustainable water management.12,14 Conducive framing conditions for social learning and adaptive management are characterised by a policy framework that, among other things, allows the following conditions to evolve: decentralisation of decision-making to make it as local as appropriate (e.g. if the topics to be discussed and decided upon become very abstract for stakeholders, their motivation to participate decreases because it

Linking Public Participation to Adaptive Management

163

becomes diﬃcult for them to integrate their knowledge into the overall context; but the right level of decentralised decision-making implies also that factors like the NIMBY eﬀect are taken into consideration which, if it occurs, would call for a higher level of decision-making); high ﬂexibility in planning groundwater management as well as ﬂexible and adaptive working structures to better react to uncertainties; and eﬃcient management of relations among relevant actors as well as their roles in the network to guarantee long-term cooperation with clear-cut deﬁned roles (including requirements and beneﬁts) for the stakeholders. Stakeholder participation and social learning in groundwater management must develop within a solid legal and institutional framework. Otherwise there is a great risk that the stakeholders are reluctant to participate. It is especially important to deﬁne:51 the rights and duties of representatives; and procedures for those who are continually reluctant to participate (this may imply personal visits and additional information for the stakeholders concerned or even special incentives to participate but also a clear working modus with deadlines for continuation). Although social learning may initiate institutional change, institutional factorsy can also become a barrier to social learning.52 Bureaucratic working procedures, centralised structures of organisations and corresponding centralised decision-making reduce the necessary ﬂexibility of participatory processes. Moreover, it increases the risk of ‘‘solving’’ problems that are not the actors’ real problems. Centralised structures and rigid, over-regulated bureaucracy hinder the creation of a conducive, consensus-oriented culture of discussion from the very beginning.

4.4.4.2

Well-designed Process Management

At the beginning of a stakeholder process in groundwater management it is important to deﬁne a common problem. For example, a water supplier is interested in abstracting groundwater of a suﬃcient quality and quantity to reduce production costs. Agricultural associations, for example, will defend the interests of farmers to continue their current practise of fertilisation if compensation for losses in productivity is not guaranteed. In such a case, starting out with a problematic statement like ‘‘reducing the nitrate levels in groundwater to potable water level’’ would be too one-sided. A more holistic statement allowing give-and-take in terms of measures and beneﬁts should be preferred. y

The term ‘‘institutions’’ is used in a sociological sense where formal institutions such as laws, acts and policies are distinguished from informal institutions such as general unwritten rules and norms. Institutions in general guarantee the organisation of individuals and groups in communities (cf. Ref. 59).

164

Chapter 4.4

One simple example could be ‘‘improving the groundwater status,’’ rather than already focusing on one pollutant and giving indirect thresholds. Examples of important aspects in successful process management are: making use of external facilitation for setting the rules of the process as well as having a ‘‘neutral mirror’’ for reﬂecting upon the process; making clear how the process inﬂuences decisions and results; and the rather informally designed participatory process should increasingly result in formal agreements and responsibilities. Well-designed process management will lead to social learning among participating actors by sensitising to for the problems and needs of others. This new perspective for individuals and groups is based on their interaction leading to not only improved but also new relations. The following paragraph will demonstrate how selected methods and instruments support this learning process over a required time span. The results of learning are measurable in terms of the quality of the relations among actors and in terms of the quality of the technical outcomes of measures and interventions. Unfortunately in water management, many results only become measurable after a longer time span. This is especially true for groundwater. In the preceding paragraph the importance of an eﬃcient management of network relations was already emphasised. This is partly guaranteed by the right framework. But the process management in itself can also contribute to the good management of actor relations within the network. Within the process of participation it is ﬁrst necessary to sensitise actors to the importance of such functional network management; help set up or improve the network; and train the relevant actor(s) in managing and maintaining the network. To improve communication and relations among stakeholders, it is even recommended to engage social scientists to map the existing communication networks amongst the various ‘‘message senders’’ and ‘‘message receivers’’ involved in the management and to use a speciﬁc aquifer.51 The organisation in charge of the participation process must take these results into account for their process design.

4.4.4.3

Well-selected Methods and Tools

Selecting the appropriate method or the right information and communication (IC) support tool can determine the success of participation. IC methods and tools are deﬁned as ‘‘material artefacts, devices or software, that can be seen and/or touched, and which are used in a participatory process to support the interaction between stakeholders (including scientists) and with the public through two-way communication processes.’’53 Examples include maps and Geographic Information Systems (GIS), as well as group model building and

Linking Public Participation to Adaptive Management

165

role playing games. The precondition is that the method or tool supports the dialogue: information should be dissipated equally in all directions. Since social learning requires a multitude of well-set conditions to evolve and also deﬁnes a multitude of goals in itself, it is unrealistic to believe that one stakeholder meeting, one workshop or one method will suﬃce to guarantee its development. For this reason, social learning is explicitly deﬁned as a process. This means that the IC tools are not always appropriate in each phase of the participation process (in the starting, managing and improving phases; compare Table 4.4.2) and that not all conditions and goals of social learning are equally supported. Table 4.4.2 shows the assessment of tools and methods with regard to their applicability in diﬀerent participation phases (good to low applicability), depicting their basic eﬀect with regard to selected conditions and/or objectives of social learning.54 The applicability and beneﬁts of IC methods and tools in participatory processes can be summarised as follows: diﬀerent methods support the various phases of participation with diﬀerent intensity; diﬀerent goals of the process require diﬀerent methods; trust-building methods are a precondition for creating the transparency of the process; methods must be culturally adapted so that participants are encouraged to formulate their interests; interest groups must be integrated into the development of new methods of participation; and additional information and expertise will ﬁnally be gathered.

4.4.4.4

Leadership Issues

The importance of good and neutral facilitation has already been highlighted under the issue of process management. In many cases governmental authorities will not only take the lead in the participation process but will also, to a certain extent, facilitate working sessions of stakeholder groups in which they represent a stakeholder themselves. During the implementation of the WFD it is and has often been the authority that maintains the ﬁnal power to decide which other stakeholders to invite to a meeting in which all were supposed to have a say. The fact that one authority or representative may invite, moderate and contribute to a participatory process can certainly raise diﬃculties in creating an egalitarian working atmosphere. It is an important topic when planning the overall moderation and process management of participation. Additionally, it may be the case that other technical authorities are invited to these meetings that may compete with each other for governmental budgets. This would be an additional reason to call for professional moderation, adding to the credibility of the process.

’

’ ’

’

K

K

K

’, Good applicability; K, medium applicability; m, low applicability.

Website

Maps Spatial mental models and maps

Facilitated session in which participants build a model to improve their understanding of the issue Game situation in which players act out roles in a real or imaginary context Facilitated and reported open discussion between participants System used for the storage, mapping and analysis of geographic data Graphic scale models Geographic representation and structuring of perceptions about issues Computer-based collection of information accessible on the internet, sometimes including a forum

Short description

Phase: starting

’

’ ’

’

’

’

’

Phase: managing

’

m

m

’

K K

K

’ ’

’

’

Phase: improving

m

Knowing about system complexity

’

’

Fairness of the process

’

’

Learning about other perspectives

’

’

’

Distribution of information

’

’

Common problem deﬁnition

Tools and methods and their applicability in diﬀerent participation phases, and their eﬀect with regard to selected conditions and/or objectives of social learning.

Geographic Information System (GIS)

Round table conference

Role playing game

Group model building

Name of tool or method

Table 4.4.2

166 Chapter 4.4

Linking Public Participation to Adaptive Management

167

Experiences with participation and social learning are currently largely based on the interaction of the representatives of interest or stakeholder groups in the form of informal to formal working groups, forums and workshops. Unfortunately, there is a large knowledge gap on the mechanisms that guarantee that individual stakeholders correctly represent the interest of their group in the participation process and that information gathered during the participation process is correctly fed back to their group. One precondition for a successful participation process is certainly that representatives should have the clear mandate of their group. Here, it is also the structure of the organisation they represent that inﬂuences the kind of mandate provided and the way in which information and decisions are fed back into the organisation. In additionally to the provided mandate, the ‘‘quality’’ of a stakeholder in a process depends on his or her leadership qualities. On one hand the individual qualities and skills of the representatives largely determine the satisfaction of their represented group with the outcomes of participation. On the other hand these qualities and skills determine the acceptance and individual success of the representative in the participation process. It therefore becomes clear that the two concepts of social learning and leadership inﬂuence the success or failure of the participation process. Examples of individuals’ high leadership qualities have already been described by Maslow.55 Among the characteristics of people with high leadership qualities Maslow mentioned ‘‘they focus on problems outside themselves.’’ This attribute is at the same time one inherent goal of social learning: becoming aware of other perspectives of a problem and better understanding why other people hold a contrasting position. The importance of leadership issues was underlined by the outcomes of the HarmoniCOP project, where it was recognised that ‘‘the success of social learning is dependent on the participation of key individuals and their attitude.’’56 Not only organisations and facilitators should become increasingly aware of the required leadership skills but also research should be conducted to investigate which kinds of leadership styles and competencies may be most needed at diﬀerent scales of participation in water management. Finally, another risk to the often used stakeholder participation must be mentioned. Groups that are composed of diﬀerent representatives are often not really representative, despite all eﬀorts to involve all groups or people concerned. Groups that meet on a regular basis may tend towards corporatism, which involves the risk that decisions are taken that may have a negative impact on those actors who did not participate. This risk underlines the necessity of a careful process management that may include a stakeholder or actor analysis at the beginning of the process (cf. Section 3.2). All steps of the process require an intermediary evaluation to notice when corporatism with its negative eﬀects is developing.

4.4.4.5

An Example of Participation-based Measures in Groundwater Management

Experiences from an EU-funded project on sustainable groundwater management show good results with voluntary agreements (in the form of contracts)

168

Chapter 4.4

between land managers and the water supply companies and/or the federal state of Lower Saxony in Germany.57 Under the terms of the contracts, the enterprises (principally farmers, but also forestry and horticultural enterprises) are obliged to observe particular restrictions and conditions which go beyond good practice. Farmers are paid compensation for economic losses that may arise. Thus, the farmers make a contribution towards groundwater protection which they would otherwise not be required to do, in return for payment. The money required comes from water abstraction charges and, in some cases, from the EU. The basis for developing these voluntary agreements and implementing a variety of measures therein was the development of the cooperation model in Lower Saxony. Cooperation committees were formed for all of the waterworks where representatives from agriculture and forestry operations, the water companies and the authorities involved, namely the Chamber of Agriculture and the Water Authority, sit at one table. The objective of this cooperation was to ﬁnd a common solution to lower the nitrate emissions into groundwater for drinking water protection. The preconditions for the successful implementation of these voluntary agreements are the following. Local water advisors work in the ﬁeld to supervise/aid farmers. Measures are jointly developed by local and regional water management authorities, water suppliers and farmers. Economic disadvantages of farmers are compensated. Random control of measures: farmers are directly approached if they do not comply with contractual agreements. In worst cases, farmers can be excluded from further voluntary agreements after the payment of compensation has stopped. Since money for compensation is limited, this instrument has only been implemented for water protection areas until now. Available measures are restricted to changes in agricultural production. But restrictions of this instrument exist, too: e.g. even a high economic incentive cannot work for farmers with high production levels of liquid manure because they are physically forced to get rid of the manure if they do not completely change their farming system. What makes it diﬃcult to judge success is that the expected beneﬁt of reduced nitrate levels in groundwater—depending on the measure in detail and the natural factors—are only visible after several years. One advantage of voluntary agreements lies in their self-control mechanism. Groups that join for a common goal or common intervention tend to socially control each other so that no ‘‘free-riders’’ beneﬁt from the system. The given example shows what kind of steps can be taken towards achieving more adaptive groundwater management in practise. If this new instrument in the form of voluntary agreements becomes more eﬀective in the sense of adaptive groundwater management, special attention should be paid to monitoring and evaluating the process and to the learning that has taken place among the actors involved. Naturally, more adaptive groundwater management requires more

Linking Public Participation to Adaptive Management

169

than just new instruments. The requested new management approaches may include diﬀerent working attitudes, new institutions and the previously mentioned innovative instruments for implementing the management strategies. In the case of Germany, groundwater protection presents the most important drinking water resource. Accordingly, water suppliers and relevant authorities are urgently awaiting a rigid daughter directive of the WFD on groundwater.58 But since large implementation gaps already exist for policies, laws and ordinances in water protection, it is unlikely that augmented regulation will necessarily aid groundwater protection. Cross-sectoral collaboration and stakeholder participation become increasingly important because of the currently acknowledged complexity of problems. Also, in an era of job cuts in the relevant authorities, it is also an economic necessity. This development underlines the necessity of making use of new instruments. This is because these participatory-based voluntary agreements may demand less eﬀort in controlling their successful implementation than ‘‘traditional’’ instruments in water protection based on mechanisms of command and control. Besides the economic needs that become drivers of change in water management, the previously mentioned factors such as climatic change, including the increasing uncertainty in water scenarios, also make adaptive management an appropriate solution. Adaptive management will consist of many small experiments trying to come up with new ideas and instruments like the one described. All of these steps, whether successful or not, should be planned, directed and certainly documented and well reﬂected upon to prepare the ﬂoor towards a more open, transparent and participatory management style.

4.4.5

Conclusions

Adaptive water management and participatory approaches are not advocated as a panacea to solve all kinds of water problems. Diﬀerent management traditions, cultures and styles make it necessary to carefully explore what is possible and to develop approaches adapted to the socioeconomic and environmental context. The acceptance of this new approach by water managers and other actors is a precondition for its success. If adaptive management cannot be meaningfully embedded in its policy environment, it would also be useless to force its implementation. When can adaptive groundwater management been considered especially useful? new, more creative solutions are required due to new, unpredicted problems; major uncertainties about the eﬀect of measures prevail; slow processes of change with unreliable data and information; problems and their solutions have and will have an impact on many actors; and there is a general need to adapt traditional instruments of groundwater management.

170

Chapter 4.4

When can adaptive groundwater management been considered too resourceintensive to be implemented on a routine basis? Decisions are routine and very clear: the processes are technical; and there is an agreement among actors in groundwater management that techniques, methods and instruments are most appropriate and functional to solve a problem. Eventually, it is necessary for water managers to understand that adaptive water and adaptive groundwater management are composed of many measures and activities that can be placed on a continuum from ‘‘hardly’’ adaptive to ‘‘very’’ adaptive. It is the ﬁnal selection in a participatory process carefully evaluating a whole range of possible management approaches that makes it the best option for a particular problem in a speciﬁc area.

References 1. C. Pahl-Wostl, Aquat. Sci., 2002, 64, 394–411. 2. J. Newig, Z. Umweltpolitik Umweltrecht, 2005, 28(4), 469–512. 3. C. S. Holling (ed). Adaptive Environmental Assessment and Management, John Wiley, New York, 1978. 4. C. J. Walters, Adaptive Management of Renewable Resources, McGraw Hill, New York, 1986. 5. C. Pahl-Wostl, The Dynamic Nature of Ecosystems: Chaos and Order Entwined, Wiley, Chichester, 1995. 6. K. N. Lee, Conserv. Ecol. 1999, 3(2) (online at: http://www.consecol.org/ vol3/iss2/art3). 7. L. Gunderson, Conserv. Ecol. 1999, 3(1) (online at: http://www.consecol.org/vol3/iss1/art7). 8. C. F. Kiker, J. W. Milon and A. W. Hodges, Ecol. Econ., 2001, 37, 403–416. 9. B. D. Richter, R. Mathews, D. L. Harrison and R. Wigington, Ecol. Appl., 2003, 13, 206–224. 10. B. T. Bormann, P. G. Cunningham, M. H. Brookes, V. W. Manning and M. W. Collopy, Adaptive Ecosystem Management in the Paciﬁc Northwest, USDA Forest Service General Technical Report PNWGTR-341, 1994. 11. P. Gleick, Science, 2003, 302, 1524–1528. 12. C. Pahl-Wostl, Transitions towards adaptive management of water facing climate and global change, Water Resour. Manag., 2007, 21(1), 49–62. 13. D. Ridder, E. Mostert and H. A. Wolters (eds), Learning Together to Manage Together: Improving Participation in Water Management, Handbook of the HarmoniCOP project, 2005 (ISBN 3-00-016970-9). 14. V. Galaz, Does the Water Framework Build Resilience? Harnessing SocioEcological Complexity in European Water Management, policy paper 1, Swedish Water House/The Resilience and Freshwater Initiative, 2005

Linking Public Participation to Adaptive Management

15.

16.

17. 18. 19. 20.

21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

31.

32.

33. 34.

171

(http://www.swedishwaterhouse.se/images/partners/20050810093304WFD BuildResiliencReport.pdf/). European Union, Guidance on Public Participation in Relation to the Water Framework Directive. Active Involvement, Consultation, and Public Access to Information, Common Implementation Strategy, Guidance Document no. 8, Luxemburg, 2002. J. Minsch, P. H. Feindt, H.-P. Meister, U. Schneidewind, T. Schulz and J. Tscheulin, Institutionelle Reformen fu¨r eine Politik der Nachhaltigkeit. Studie im Auftrag der Enquete-Kommission ‘‘Schutz des Menschen und der Umwelt’’ des Deutschen Bundestages, Springer, Heidelberg/New York/Berlin, 1998. A. Voßkuhle, Verwaltungsarchiv, 2001, 92, 184–215. M. Lee and C. Abbot, Modern Law Rev., 2003, 66(1), 80–108. C. Coglianese, Duke Law J., 1997, 46, 1255–1346. F. H. J. M. Coenen, D. Huitema and L. J. O’Toole, Participation and environmental decision quality: an assessment, in Participation and the Quality of Environmental Decision Making, ed. F. H. J. M. Coenen, D. Huitema and L. J. O’Toole, Kluwer Academic, Dordrecht/Boston/ London, 1998, pp. 307–324. V. Ostrom, C. M. Tiebout and R. Warren, Am. Polit. Sci. Rev., 1961, 55(4), 831–842. C. Pahl-Wostl, Environ. Modell. Software, 2007, 22(5), 561–569. J. A. Lo´pez Cerezo and M. Gonza´lez Garcı´ a, Phil. Tech., 1996, 2, 1. L. Pellizzoni, Environ. Values, 2003, 12(2), 195–224. S. Yearley, S. Cinderby, J. Forrester, P. Bailey and P. Rosen, Environ. Values, 2003, 12(2), 247–262. J. Steele, Oxford J. Legal Studies, 2001, 21(3), 415–442. J. C. Thomas, Public Participation in Public Decisions. New Skills and Strategies for Public Managers, Jossey-Bass, San Francisco, CA, 1995. A. Fisahn, Demokratie und O¨ﬀentlichkeitsbeteiligung, Mohr Siebeck, Tu¨bingen, 2002. G. Munnichs, J. Agric. Environ. Ethics, 2004, 17, 113–130. J. Doak, Changing the world through participative action: the dynamics and potential of local agenda 21, in Participation and the Quality of Environmental Decision Making, ed. Frans H. J. M. Coenen, Dave Huitema and Laurence J. O’Toole, Kluwer, London, 1998, pp. 77–91. B. Cooke, The social psychological limits of participation? in Participation: The New Tyranny? ed. B. Cooke and U. Kothari, Zed Books, London/ New York, 2001, pp. 102–121. W. Linder and A. Vatter, Kriterien zur Evaluation von Partizipationsverfahren, in Planung und Kommunikation, ed. K. Selle, Bauverlag, Wiesbaden/Berlin, 1996, pp. 181–188. H. Bulkeley and A. P. J. Mol, Environ. Values, 2003, 12(2), 143–154. T. Webler and O. Renn, A brief primer on participation: philosophy and practice, in Fairness and Competence in Citizen Participation: Evaluating

172

35.

36. 37. 38. 39. 40.

41.

42. 43.

44. 45. 46.

47.

48. 49.

50.

51.

Chapter 4.4

Models for Environmental Discourse, ed. O. Renn, T. Webler and P. Wiedemann, Kluwer Academic, Dordrecht, 1995, pp. 17–33. H. Ryﬀel, Bedingende Faktoren der Eﬀektivita¨t des Rechts, in Zur Eﬀektivita¨t des Rechts, ed. M. Rehbinder and H. Schelsky, Bertelsmann Universita¨tsverlag, Du¨sseldorf, 1972, pp. 225–246. J. L. Creighton, The Public Involvement Manual, Abt Books, Cambridge, MA, 1981. E. A. Lind and T. R. Tyler, The Social Psychology of Procedural Justice, Plenum, New York/London, 1988. T. R. Tyler, Why People Obey the Law, Yale University Press, New Haven, CT, 1990. K. Murphy, Law Human Behav., 2004, 28(2), 187–209. European Union, Common Implementation Strategy for the Water Framework Directive (2000/60/EC), Guidance Document no. 11 on the Planning Processes, Luxemburg. 2003. C. Bo¨hret and G. Konzendorf, Guidelines on Regulatory Impact Assessment (RAI), prepared for the Federal Ministry of the Interior and for the Ministry of the Interior of Baden-Wuerttemberg, Speyerer Forschungsberichte, Speyer, 2004, vol. 234. W. M. Babiuch and B. C. Farhar, Stakeholder Analysis Methodologies Resource Book, National Renewable Energy Laboratory, Colorado, 1994. A. Harrison, G. Schmidt, C. Avis and R. Hauser, WWF’s preliminary comments on public participation in the context of the Water Framework Directive and Integrated River Basin Management, Copenhagen, 2001. E. Mostert, Water Policy, 2003, 5(2), 179–197. V. Nies, B. Rocha and A. Skupien, Agrar- und Umweltrecht, 2006, 4, 120–126. B. Kastens and J. Newig, Will participation foster the successful implementation of the WFD? The case of agricultural groundwater protection in North-West Germany, Local Environ., accepted for publication. E. Mostert, Participation for sustainable water management, in Sustainable Management of Water Resources: An Integrated Approach, ed. G. Giupponi, A. J. Jakeman, D. Karssenberg and M. Hare, The Fondazione Eni Enrico Mattei Series on Economics, the Environment and Sustainable Development, Cheltenham, UK, 2006, p. 171. D. Ridder and C. Pahl-Wostl, Regional Environ. Change, 2005, 5(4), 188–196. A. Ro¨hring, Die Umsetzung der EU-Wasserrahmenrichtlinie: ‘‘problems of interplay’’, in Das Flussgebiet als Handlungsraum. Institutionenwandel durch die EU-Wasserrahmenrichtlinie aus raumwissenschaftlichen Perspektiven, ed. T. Moss, Mu¨nster, 2003, p. 256. T. Moss, Induzierter Institutionenwandel ‘‘von oben’’ und die Anpassungsfa¨higkeit regionaler Institutionen. Zur Umsetzung der EUWasserrahmenrichtlinie in Deutschland, in Das Flussgebiet als Handlungsraum. Institutionenwandel durch die EU-Wasserrahmenrichtlinie aus raumwissenschaftlichen Perspektiven, ed. T. Moss, Mu¨nster, 2003, p. 158. H. Gardun˜o, M. Nanni, A. Tuinhof, K. Kemper and C. Dumars, Stakeholder participation in groundwater management: mobilizing and

Linking Public Participation to Adaptive Management

52. 53.

54.

55. 56.

57.

58. 59. 60.

173

sustaining aquifer management organization, in Sustainable Groundwater Management: Concepts and Tool, Brieﬁng Note 6, World Bank, Washington, DC, 2003. J. Tippet, B. Searle, C. Pahl-Wostl and Y. Rees, Environ. Sci. Pol., 2005, 8(3), 287–299. P. Maurel (ed.), Public Participation and the European Water Framework Directive. Role of Information and Communication Tools, Work Package 3, Report of the HarmoniCOP Project, 2003 (www.harmonicop.info). N. Kranz, M. Patel and D. Ridder, Public Participation in European River Basin Management. Lessons from the HarmoniCOP Project, Ecologic, Briefs Ecologic, Berlin, 2006. A. Maslow, Motivation and Personality, Harper & Row, New York, 1954. Y. Rees, B. Searle, T. Tippett and A. Johannessen, Good European Practices for Stakeholder Involvement: Lessons from Real Planning Processes, Work Package 5, report of the HarmoniCOP project, 2005 (www.harmonicop.info). C. Aue and K. Klaassen (coord.), Sustainable Groundwater Management: Handbook of Best Practice to Reduce Agricultural Impacts on Groundwater Quality (Water4All-Project), Aalborg, Denmark, 2005. F. Haakh, Anforderungen an den Gewa¨sserschutz aus Sicht der Wasserversorgung, Wat-Berlin Tagungsband, DVGW, 2006, pp. 18–19. H. Esser, Soziologie. Spezielle Grundlagen, Bd 5, Institutionen, Frankfurt aM/New York, 2000. J. Newig, Does public participation in environmental decisions lead to improved environmental quality? Towards an analytical framework, Communication, Cooperation, Participation. Research and Practice for a Sustainable Future, 2007, 1(1), 51–71.

5. Groundwater Characterization and Risk Assessment

CHAPTER 5.1

Groundwater Characterisation and Risk Assessment in the Context of the EU Water Framework Directivew ANDREAS SCHEIDLEDER,a JOHANNES GRATHa AND PHILIPPE QUEVAUVILLERb a

Umweltbundesamt GmbH, Spittelauer Laende 5, AT-1090 Wien., Austria; European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium

b

5.1.1

Legal Background

One of the primary and key steps within the implementation of the European Union (EU) Water Framework Directive (WFD) is the analysis of the characteristics of groundwater bodies, the review of the environmental impact of human activity and the economic analysis of water use as it is laid down in Article 5 and speciﬁed in Annex II in order to identify groundwater bodies presenting a risk of not achieving WFD environmental objectives laid down in Article 4 of the WFD. The ﬁrst analysis and review was due on 22 December 2004 and had to be reported by EU member states to the European Commission in March 2005. These analyses have to be reviewed, and, if necessary, updated at the latest in 2013 and every six years thereafter. The content of this chapter is very much inspired by the discussion within and the outcome of technical workshops of WG C ‘‘Groundwater’’ (see Chapter 4.1) on groundwater body characterisation and risk assessment1–3 during 2004–2005 which were already based on guidance documents elaborated during the ﬁrst phase of the Common Implementation Strategy.4–7 w

The views expressed in this chapter are purely those of the authors and may not in any circumstances be regarded as stating an oﬃcial position of the European Commission.

177

178

Chapter 5.1

The identiﬁcation and delineation of groundwater bodies is the ﬁrst part of the analysis of the characteristics of the river basin districts under the WFD (Article 5). Each groundwater body has to be characterised in order to assess the uses, the degree to which they are at risk of failing to meet the environmental objectives and to identify any measures to be required under Article 11. The speciﬁcation for this impact review for groundwater is laid down in WFD Annex II and includes ﬁve parts (see Figure 5.1.1): initial characterisation, including identiﬁcation of use, pressures and risk of failing to achieve objectives; further characterisation of groundwater bodies identiﬁed as being at risk; review of the impact of changes in groundwater levels for groundwater bodies for which lower objectives are to be set according to Article 4(5); review of the impact of human activity on groundwaters for transboundary and at risk groundwater bodies; and review of the impact of pollution on groundwater quality for which lower objectives are to be set. The most important goal of this ﬁrst review was to understand the signiﬁcant water management issues within each river basin and how they aﬀect each individual water body. The timetable for completing the ﬁrst pressures and impacts analyses and reporting their results was very short. The ﬁrst analyses therefore relied heavily on existing information on pressures and impacts and existing assessment methods.

Figure 5.1.1

The WFD speciﬁes requirements for impact analysis.6

Groundwater Characterisation and Risk Assessment

179

The groundwater risk assessment is part of the characterisation and the review of the environmental impacts of human activity already introduced. For each groundwater body the degree to which it is at risk of failing to meet the objectives under Article 4 has to be assessed. Ideally, a pressures and impacts assessment will be a four-step process. identifying the driving forces (especially land use, urban development, industry, agriculture and other activities which lead to pressures) without regard to their actual impacts and identifying pressures with possible impacts on the water body and on water uses; identifying the signiﬁcant pressures, by considering the magnitude of the pressures and the susceptibility of the water body; assessing the impacts resulting from the pressure; and evaluating the likelihood of failing to meet the objective. To undertake the four key stages, three supporting elements must be considered (shown on the left of Figure 5.1.2). The description of a water body and its catchment area will underpin the pressures and impacts analysis. During the process, monitoring data relevant to the water body may be introduced. A comparison of monitoring data with driving forces may help to screen where pressures are likely to cause a failure in meeting objectives. It is also necessary to understand the objectives against which the actual state is compared. In many cases these key stages need not be undertaken as a linear sequence, but in general all key stages are to be addressed.

Figure 5.1.2

Key components in the analysis of pressures and impacts.6

180

5.1.2

Chapter 5.1

Groundwater Body Identiﬁcation and Delineation

As the environmental objectives of the WFD must be applied to ‘‘water bodies’’ it is important to have a closer look at the identiﬁcation and delineation of groundwater bodies. The application of the term ‘‘body of groundwater’’ must be understood in the context of the hierarchy of relevant deﬁnitions provided under Article 2 of the WFD wherein a ‘‘body of groundwater’’ means a distinct volume of groundwater within an aquifer or aquifers; ‘‘groundwater’’ means all water that is below the surface of the ground in the saturated zone and in direct contact with the ground or subsoil; and ‘‘aquifer’’ means a subsurface layer or layers of rock or other geological strata of suﬃcient porosity and permeability to allow either a signiﬁcant ﬂow of groundwater or the abstraction of signiﬁcant quantities of groundwater. Based on the deﬁnitions a body of groundwater must be within an aquifer or aquifers. However, not all groundwater is necessarily within an aquifer. The WFD’s deﬁnition of aquifer requires two criteria to be considered in determining whether geological strata qualify as aquifers: a signiﬁcant ﬂow of groundwater or the abstraction of signiﬁcant quantities of groundwater. If either of the criteria is met, the strata will constitute an aquifer or aquifers. The signiﬁcance of groundwater ﬂow should be understood in the context of the purpose and provisions of the WFD. A key purpose of the WFD is to prevent further deterioration of and protect and enhance the status of aquatic ecosystems, and with regard to their water needs, terrestrial ecosystems and wetlands directly depending on groundwater. The objective of protecting and restoring good groundwater status is designed to help achieve this purpose. It applies to all bodies of groundwater. Consequently, to ensure that the purpose of the directive can be achieved, the deﬁnition of signiﬁcant ﬂow must encompass all groundwater ﬂow that is important to aquatic and terrestrial ecosystems. Geological strata that permit such ﬂow should therefore qualify as aquifers. In practice, the criteria mean that nearly all groundwater in the Community would be expected to be within aquifers (Figure 5.1.3). The WFD leaves ﬂexibility to the member states for the delineation of groundwater bodies to adopt the most eﬀective means of achieving the directive’s objectives. The delineation should take into account that the groundwater bodies can be accurately described and take regard of major diﬀerences in the status of the groundwater at diﬀerent depths. This does not mean that a body of groundwater must be delineated so that it is homogeneous in terms of its natural characteristics, or the concentrations of pollutants or level alterations within it. Groundwater bodies should be delineated in three dimensions and the depth of groundwater considered should depend on the risk to fail the directive’s objectives. It seems appropriate to diﬀerentiate between shallow areas which

Groundwater Characterisation and Risk Assessment

Figure 5.1.3

181

The WFD’s deﬁnition of aquifers.5

are more immediately aﬀected by pressures on the surface and deeper groundwater bodies which might be aﬀected as well but react after a certain time-lag. Nearly all member states started with the identiﬁcation of geological and hydrogeological boundaries and applied a comprehensive, additional set of further criteria like vulnerability maps, subsoil properties, risk potential, utilisation and protection need, economic importance and water management aspects. The most important aim of the member states was to achieve eﬃcient and practical inventory and management units and to keep the administrative burden and the ﬁnancial eﬀorts within practicable dimensions. Each groundwater body has to be assigned to a river basin district. The identiﬁcation of groundwater bodies must be consistent and coordinated within a river basin district. In particular, the international river basin districts need to develop common approaches for the whole river basin. Groundwater bodies may be grouped for the purposes of the risk assessment, for monitoring, reporting and management purposes where monitoring of suﬃcient indicative or representative water bodies in the subgroups of groundwater bodies provides for an acceptable level of conﬁdence and precision in the results of monitoring, and in particular the classiﬁcation of water body status. The ability to group bodies will depend on the characteristics of the river basin district and the type and extent of pressures on it and can contribute to a cost eﬃcient and pragmatic implementation of the directive and management of groundwater. Especially the Nordic countries Finland and Norway are

182

Chapter 5.1

confronted with a huge number of groundwater bodies due to the speciﬁc geological characteristics. However, such grouping must be undertaken on a scientiﬁc basis so that monitoring information obtained for the group provides for a suitably reliable assessment that is valid for each body in the group and for an acceptable level of conﬁdence and precision in the results of monitoring. It should be emphasised that the identiﬁcation of ‘‘groundwater bodies’’ is a tool and not an objective in itself. Groundwater bodies are the units which are used for reporting and assessing compliance with the directive’s principal environmental objectives. Member states had to identify such bodies by 22 December 2009 and where necessary verify and reﬁne the body identiﬁcation in the period before the publication of each river basin management plan (RBMP) by 22 December 2009 and then every six years. As commonly information from the characterisation process and monitoring was not available before 2004 it is most likely that member states will need to update the delineation of groundwater bodies, and therefore veriﬁcation and reﬁnement steps of groundwater body identiﬁcation should be foreseen in the implementation process. However, all groundwater bodies must at least be ﬁxed for each planning period. A key descriptor in this context is the ‘‘status’’ of those bodies. If water bodies are identiﬁed that do not permit an accurate description of their status, member states will be unable to apply the directive’s objectives correctly. At the same time, an endless subdivision of water bodies should be avoided in order to reduce administrative burden if it does not fulﬁl any purpose as regards the proper implementation of the directive. In addition, the aggregation of water bodies may, under certain circumstances, also help to reduce meaningless administrative burden, in particular for smaller water bodies. Finally, it has to be recognised that the objectives of preventing or limiting inputs of pollutants (Article 4.1 of WFD) and reversing any signiﬁcant and sustained upward trend in the concentration of any pollutant are subject of all groundwater and not only groundwater bodies.

5.1.3

Initial Characterisation

The need for an initial characterisation of all groundwater bodies is laid down in Article 5 of the WFD and speciﬁed in Annex II. It covers the analysis of the river basin district characteristics, the identiﬁcation of uses and pressures and a review of the environmental impact of human activity for assessing the degree to which the groundwater bodies are at risk of failing to meet the objectives of Article 4 of the WFD, namely the achievement of good (quantitative and chemical) status of groundwater at the latest by the end of 2015. Groundwater bodies may be grouped for the purposes of this initial characterisation. Based on the delineation of groundwater bodies, pressures to which the groundwater bodies or groups of bodies are liable to be subjected need to be identiﬁed (including diﬀuse and point sources of pollution, abstraction and

Groundwater Characterisation and Risk Assessment

183

artiﬁcial recharge). In addition, the general character of the overlying strata in the catchment from which the groundwater body receives its recharge shall be described, as well as the groundwater bodies for which there are directly dependent surface water ecosystems or terrestrial ecosystems. The initial characterisation requires a general analysis of pressures corresponding to that described above, but set in the context of evaluating the risk of failing to meet the objectives. This requires an understanding of the nature of the impact that may result from a pressure, and appropriate methods to monitor or assess the relationship between impact and pressure as it is described by the conceptual model. Where no monitoring data for a groundwater body are available, the likely presence or absence of pressures and impacts should be considered when making a decision of the likely status of the groundwater body. Where it is clear from monitoring data that the groundwater body is ‘‘at risk’’, or where there are inadequate data to make a decision with reasonable conﬁdence that a groundwater body is ‘‘at risk’’, the process should continue to further characterisation.

5.1.4

Further Characterisation

Following the initial characterisation, a further characterisation has to be carried out for those groundwater bodies or groups of bodies which have been identiﬁed as being at risk in order to establish a more precise assessment of the signiﬁcance of such risk and identify any measures to be required under Article 11 of the WFD. The approach recommended follows that outlined for the initial characterisation, but requires the collection of more detailed information and data, such as that detailed in Annex II 2.3, e.g. geological and hydrogeological characteristics, the characteristics of the superﬁcial deposits and soils, stratiﬁcation characteristics in the groundwater, an inventory of associated surface systems including terrestrial ecosystems and bodies of surface water, with which the groundwater body is dynamically linked, estimates of the directions and rates of exchanges of water between the groundwater body and associated surface systems, long-term annual average rate of overall recharge and characterisation of the chemical composition of the groundwater, including speciﬁcation of the contributions from human activity. The wording of Annex II suggests that the information speciﬁed shall be included ‘‘where relevant’’. In this context ‘‘relevant’’ is taken to mean relevant to the assessment of risk of failure to meet the environmental objectives of the WFD. This does not give license to neglect collecting information. ‘‘Relevance’’ also involves questions of the level of detail that should be sought and, for human activities, the timescale over which the eﬀects of the activity may be deemed relevant. In deciding these matters it is important to refer back to the purpose of further characterisation: to improve the assessment of risk and identify any measures to be required under Article 11.

184

Chapter 5.1

5.1.5

Additional Requirements of the WFD

5.1.5.1

Transboundary Groundwater Bodies

Within the characterisation process speciﬁc provisions concern those groundwater bodies that are transboundary and cross the boundary between two or more member states, focusing mainly on quantitative aspects. Information is requested on, for example, the location of groundwater abstraction points serving more than 10 m3 a day or more than 50 persons, the abstraction rates and direct discharges to groundwater and the chemical composition of water abstracted and recharged. First experiences show that the way to deal with transboundary groundwater bodies is not yet fully clariﬁed. In many cases considerable, time-consuming discussion, cooperation and commitment between member states sharing a common transboundary groundwater body are still needed in order to characterise the overall groundwater body, review the impacts and assess the risk of failing to achieve the objectives of the WFD in a harmonised way.

5.1.5.2

Groundwater Bodies with Lower Objectives

Connected to the further characterisation, the WFD also requires the identiﬁcation of those bodies of groundwater for which lower objectives are to be speciﬁed under Article 4 where, as a result of the impact of human activity, and as determined in accordance with the analysis of pressures and impacts, the body of groundwater is so polluted that achieving good groundwater chemical status is infeasible or disproportionately expensive.

5.1.5.3

Interaction with Aquatic and Terrestrial Ecosystems

Important aspects to be considered for the characterisation of groundwater bodies are the interactions with associated surface waters and terrestrial ecosystems. Indeed, the deﬁnition of good groundwater status implies that the concentrations of pollutants and directions and rates of water exchange in a deﬁned groundwater body should not result in failure to achieve the environmental objectives under Article 4 of the WFD for associated surface waters nor any signiﬁcant diminution of the ecological or chemical quality of such bodies nor in any signiﬁcant damage to terrestrial ecosystems which depend directly on the groundwater body.

5.1.6

Conceptual Model/Understanding

Assessing the impacts on a water body requires some quantitative information to describe the state of the water body itself and/or the pressures acting on it. This assessment requires a conceptual understanding of what happens in a groundwater body and what causes impacts.

Groundwater Characterisation and Risk Assessment

185

Conceptual understandings or conceptual models are simpliﬁed representations, or working descriptions, of the hydrogeological system being investigated and reﬂect the understanding of how the real hydrogeological system of a groundwater body, its processes and reactions are believed to behave. It is very much a set of working hypotheses and assumptions, based on evidence and should be written down so that it can be tested. The development of a conceptual understanding is mainly based on the work carried out as part of the characterisation process and reﬁned by additional information and monitoring data gathered during the following implementation procedure of the WFD. As the amount of, and conﬁdence in, available environmental information increases, the accuracy and complexity of the model/understanding improves, so that it becomes a more eﬀective and reliable description of the system. The level of reﬁnement needed in a model is proportionate to (a) the diﬃculty in making the assessments or predictions required, and (b) the potential consequences of errors in those assessments. A conceptual model/understanding is furthermore necessary to design monitoring programmes and it is necessary to interpret the data provided by those programmes, and hence assess the achievement of the directive’s objectives. The testing of conceptual models/understandings is important to ensure they provide for acceptable levels of conﬁdence in the assessments they enable. A successful pressure and impact assessment will not be one that follows prescriptive guidance. It will be a procedure in which there is proper understanding of the objectives, a good description of the water body and its catchment area (including monitoring data) and a knowledge of how the catchment system functions. A conceptual understanding/model is dynamic, evolving with time as new data and information are obtained and as the model is tested. Its development and reﬁnement should adopt an iterative approach (see Figure 5.1.4). The approach therefore ﬁts in well with the various levels of knowledge required at diﬀerent stages of the WFD. For example a basic model will be appropriate for initial characterisation; this (if appropriate) will be reﬁned and improved during further characterisation, and again during the review cycle of the RBMP. The

Figure 5.1.4

Reﬁnement of conceptual model/understanding.7

186

Chapter 5.1

drafting of basic conceptual models of groundwater ﬂow and chemical systems and then of groundwater bodies should be undertaken early in the process of initial groundwater characterisation. This will include the delineation of the groundwater body boundaries and an initial understanding of the nature of the ﬂow and geochemical system, interaction with surface water bodies and terrestrial ecosystems and an early assessment of pressures.

5.1.7

Identiﬁcation of Driving Forces and Pressures

A common understanding of terms and the most eﬀective approach for groundwater risk assessment was developed within the IMPRESS working group during the ﬁrst phase of the Common Implementation Strategy.6 The widely used driver, pressure, state, impact, response (DPSIR) analytical framework had been adopted with deﬁnitions as in Table 5.1.1. It is worth noting in the context of the DPSIR framework as described above that objectives deﬁned by the WFD relate to both the state and the impact, since standards from other European water quality objective legislation relate to the concentration of pollutants in the water body (i.e. its state), while the biological elements of the WFD clearly indicate impacts. Driving forces are sectors of activities that may produce a series of pressures. A pressure results from an activity that may directly cause deterioration in the status of a water body. In most cases a pollution pressure relates to the addition or release of substances into the environment. This can be the discharge of a waste product, but may also be the side eﬀect or by-product of other activities, such as leaching of nutrients from agricultural land. A pollution pressure may also be caused by an action such as a change in land use. The most usual categorisation of pollution pressures is to distinguish between diﬀuse and point sources. However, the distinction between point and diﬀuse sources is not always clear, and may again relate to spatial scale. For example, areas of contaminated land might be considered as either diﬀuse or point sources of pollution. A quantitative pressure relates to the change of Table 5.1.1

The DPSIR framework as used in the pressures and impacts analysis.

Term

Deﬁnition

Driver

An anthropogenic activity that may have an environmental eﬀect (e.g. agriculture, industry) The direct eﬀect of the driver (e.g. an eﬀect that causes a change in ﬂow or a change in the water chemistry) The condition of the water body resulting from both natural and anthropogenic factors (i.e. physical and chemical characteristics) The environmental eﬀect of the pressure (e.g. ecosystem modiﬁed) The measures taken to improve the state of the water body (e.g. restricting abstraction, limiting point-source discharges, developing best practice guidance for agriculture)

Pressure State Impact Response

Groundwater Characterisation and Risk Assessment

187

groundwater levels or the modiﬁcation of ﬂow directions but also to the intrusion of salinity, the reduced dilution of chemical ﬂuxes or the modiﬁcation of dependent aquatic or terrestrial ecosystems. Such pressures can be changes in land use like land sealing, water abstraction or artiﬁcial recharge. The pressures and impacts analyses should be focused in such a way that the eﬀort involved in assessing whether any groundwater body, or group of bodies, is at risk of failing to achieve its environmental objectives is proportionate to the diﬃculties involved in making that judgement. A screening approach helps simplifying the tasks prior to additional description and analysis at a later stage, as it means focusing on the search for pressures on those areas and pressure types that are likely to prevent meeting the objectives and pointing out with simple assessments those water bodies that are clearly ‘‘at risk’’ or not at risk in failing to achieve good status in 2015. The screening approach may be carried out using driving force assessment as substitute of pressures. Driving forces are quantiﬁed by aggregated data, simple to obtain, e.g. hectares of arable land, population density per area. This screening should identify issues to be addressed in the drawing up of the RBMP, and it may also reveal a number of gaps in data or knowledge that should be ﬁlled during the process of drawing up the RBMP and the monitoring programme. A list of pressures and the assessment of impacts on a water body shall ensure the identiﬁcation of all of the potentially important problems. Assessing the likely impacts arising from each of the pressures will produce a list that can be used to identify points where monitoring is necessary to better understand if the water body is at risk of failing to achieve good status. The screening procedure is not only a way to accelerate data collection by focusing on those pressures that are reasonably expected, it provides an independent assessment of pressures and impact relationships, which is valuable especially if emission and abstraction registers are poorly populated. Clearly the use of GIS facilitates this process.

5.1.8

Identiﬁcation of Signiﬁcant Pressures

The inventory of pressures is likely to contain many that have no or little impact on the groundwater body. The assessment of whether a pressure on a water body is signiﬁcant must be based on the knowledge of the pressures within the catchment area together with a conceptual understanding of the groundwater bodies in the context of the receptors. This understanding coupled with the list of all pressures and the particular characteristics of the catchment makes it possible to identify the signiﬁcant pressures. However this approach often requires two steps. In the ﬁrst step, correlation assessment can be carried out. This has the advantage of using monitored data and does not require complex hypotheses. In the second step, the conceptual understanding is embodied in a set of simple rules indicating directly whether a pressure is signiﬁcant or not.

188

Chapter 5.1

One approach is to compare the magnitude of the pressure with a criterion or threshold relevant to the water body type but this must take account of the particular characteristics of each water body and its vulnerability to the pressure. This approach eﬀectively combines the pressure identiﬁcation with the impact analysis, since, if any threshold is exceeded, the water body is assessed as likely to fail its objectives. To improve conﬁdence, the estimates of the type and magnitude of pressures should be crosschecked, where possible, with monitoring data and with information on the key drivers for the pressures.

5.1.9

Assessing the Impacts of Pressures

Within the initial characterisation the concept of ‘‘potential impact’’ could be introduced to describe the eﬀects a pressure is likely to have on a groundwater body, and that ‘‘potential impact’’ is used in the evaluation of whether the body is ‘‘at risk’’. This concept recognises that, with the constraints on the characterisation process, it will not always be possible to accurately measure the impact by monitoring groundwater levels and quality. For pollution pressures the potential impact is judged by considering the pollution pressure (where this occurs at the surface) in combination with the vulnerability of the groundwater body to pollution. Thus, for example, a high pollution pressure caused by anthropogenic activities at the surface may have little impact on a groundwater body if that body is protected by a signiﬁcant thickness of low permeable layers. Within the further characterisation a review of the impact of human activity for groundwater bodies characterised to be ‘‘at risk’’ and for those crossing member state boundaries is required. Assessing the impacts on a water body requires some quantitative information to describe the risk status of the groundwater body and/or the pressures acting on it. The type of analysis will depend on what data are available. Regardless of the particular process to be adopted, and as with the identiﬁcation of signiﬁcant pressures described above, the assessment requires a conceptual understanding of what causes impacts. In many cases a simple approach might be absolutely suitable for assessing the impact of a pressure. However, there will be a vast range of catchment types, aquifer types, interacting pressures, process conceptualisations, data requirements and possible impacts, and adopting such a simple model for all cases might not be appropriate. Tools which might assist assessing the impacts comprise the use of observed data to assess and to reﬁne the assessment, a conceptual model, the use of analogue water bodies and the use of numeric models.

5.1.9.1

Tools to Assist

A pressure checklist can be considered as a reminder of the driving forces and the pressures that should be considered and therefore represents a precursor to the actual pressures and impacts analysis.

Groundwater Characterisation and Risk Assessment

189

If monitoring data are available for the groundwater body, it might be possible to perform a direct assessment of the impact. However, it must be kept in mind that most pressures do not create a clear-cut impact. Monitoring data may indicate that there are no current impacts. This information itself reveals that none of the pressures identiﬁed in the initial screening process is signiﬁcant, or that the time lag required for a pressure to give rise to an impact has not yet passed. The latter is likely to be of particular importance when assessing groundwater bodies in which pollutants travel very slowly. A conceptual model/understanding of the ﬂow system, chemical variations and the interaction between groundwater and surface ecosystems is essential for characterisation and for assessing the impacts of pressures. A signiﬁcant strength of the approach is that it allows a wide variety of data types (including, for example, physical, biological and chemical data) to be integrated into a coherent understanding of the system. As new data are obtained they help to reﬁne or adjust the model; conversely the model may indicate errors and inadequacies in the data. A further step could be a shift to mathematical models. In general the more complex the model, the greater the data requirements and the greater the time and costs needed to improve it; therefore such an additional eﬀort seems to be appropriate where water bodies appear to be at risk, or where a detailed programme of measures needs to be developed only. However, in the context of groundwater body characterisation under the WFD there are many questions that may be answered adequately with a simple model. In situations with no observed data, a possible tool to evaluate status is to compare with a similar analogous groundwater body for which data are available, and to assume that the assessment made from the observed data can be applied validly to both bodies. Furthermore, the WFD oﬀers the possibility of grouping water bodies for the purpose of pressure and impact analysis and monitoring. Groundwater vulnerability maps or indices are useful tools for assessing the likely impact of pollution pressures during the characterisation process. By taking account of a range of factors the susceptibility or vulnerability of groundwater to pollution from pollution pressure on the land surface can be ranked. Groundwater vulnerability maps can be used as a screening tool to rapidly assess the relative scale of impacts arising from pressures. They may be useful for assessing whether groundwater bodies are ‘‘at risk’’ from pollution sources during the initial characterisation. Groundwater vulnerability assessments may be combined with models of diﬀuse pollution source behaviour, to consider the overall risks to water quality on a groundwater body scale.

5.1.9.2

Scaling Issues

Diﬀerent kinds of pressures do not impact the diﬀerent water bodies at the same time and space scales. It is important to adopt appropriate temporal scales in the pressures and impacts analysis since some pressures may result in impacts many years in the future, and some future impacts will relate to past

190

Chapter 5.1

pressures that no longer exist. For example, pesticide application may lead to increased concentrations of the pesticide in the groundwater many years after it was released. When a groundwater body currently has good status but it is thought that pressures may cause its status to be rendered poor by 2015, then the body is ‘‘at risk’’ and will require further characterisation. It should be noted that a body currently determined to have poor status will automatically be ‘‘at risk’’. Regarding spatial scales, it is important to consider the location of where the data are gathered from and where the pressures occur, especially if the groundwater body consists of, for example, a separate recharge area and an area of conﬁned groundwater that responds diﬀerently to the pressure. For example, considering conﬁned groundwater, the relevant pressure data are those on the recharge area only, not over the total extent of the water body. The WFD does not diﬀerentiate between groundwater in diﬀerent strata: all groundwater requires the same degree of protection from pollution. However, the impact that a pollution pressure is likely to have on groundwater varies from site to site, depending on the hydrogeological properties of the aquifer and geological strata. Consequently, for a given pollution pressure, the impact on the status of a groundwater body, and the potential programme of measures will vary in diﬀerent aquifers.

5.1.10

Evaluating the Likelihood of Failing to Meet the Objectives

Evaluating the risk of failing to meet the objectives should have been a straightforward comparison of monitored data with the provisions of good status as laid down in the WFD. The ﬁrst pressure and impact assessment had to be completed by the end of 2004. However, speciﬁcations required to meet most of the objectives of the WFD had not been ﬁrmly deﬁned by this date as Article 17 requested the European Commission to come up with a ‘‘daughter’’ directive laying down a criteria for assessing good groundwater chemical status and criteria for identifying signiﬁcant sustained upward trends and for the deﬁnition of starting points of trend reversal. After four years of discussion, the new Groundwater Daughter Directive was ﬁnally adopted in December 2006 (see Chapter 3.1). The criteria for good chemical status are based on EU-wide quality standards, groundwater threshold values and WFD criteria, wherein the groundwater threshold values have to be developed by member states on a national, regional or local level for those substances which are causing risk of failing to meet the environmental objectives of the WFD. The groundwater threshold values have to be reported together with the draft RBMPs in 2008. The conﬁdence and precision of the estimated environmental eﬀects of diﬀerent pressure types will also be very variable, depending to a great extent on the quality of national and local information and assessment expertise. This is because consideration of many of the pressures and impacts relevant under

191

Groundwater Characterisation and Risk Assessment

Annex V and GWDD

Annex II requirements

Evaluating risk of

requirements

failing to meet Characterisation,

Figure 5.1.5

Defining thresholds for

The iterative evaluation of the risk of failing objectives (based on Ref. 6).

the WFD has not previously been required by other Community water legislation. Member states completed the ﬁrst analyses using appropriate estimates for pressures and impacts but they had to be aware of, and had to take account of, the uncertainties in the environmental conditions required to meet the WFD’s objectives and the uncertainties in the estimated impacts. The consequence of these uncertainties is that member states’ judgements on which bodies are at risk, and which are not, are likely to contain more errors in the ﬁrst report than will be the case in subsequent planning cycles. It is clear that the process of evaluating the risk of failure is to some degree an iterative collaboration between those undertaking the pressures and impact analysis, and those deﬁning thresholds for the as yet undeﬁned elements of status (Figure 5.1.5). Furthermore, it will be important for member states to be aware of the uncertainties so that their monitoring programmes can be designed and targeted properly in order to provide the information needed to improve the conﬁdence in the assessments. Where the assessment contains signiﬁcant uncertainty, those water bodies should be categorised as at risk of failing to meet their objectives. Obvious failing of pressures is not an uncertainty.

5.1.11

Reporting on the Characterisation and Risk Assessment

Article 15 requires member states to submit a summary report of the pressures and impact analyses to the Commission within three months of their completion (i.e. the ﬁrst report had to be submitted by March 2005). This analysis has to be reviewed, and if necessary updated at the latest in 2013 and every six years thereafter. The summary reports sent to the Commission should be concise and give an overview of relevant characteristics and main water bodies within a river basin district, tables and maps showing signiﬁcant pressures and water bodies at risk. Furthermore, the summary report should include methodologies, tools, threshold values, environmental quality objectives, classiﬁcation schemes, etc., used

192

Chapter 5.1

within the risk assessment and an indication of the amount of uncertainty of the pressure analysis and results. All these reports provided to the Commission are publicly available on the internet. More detailed information should be available on demand for public and stakeholder consultation. The European Commission is currently developing guidance on reporting to the Commission in the form of reporting sheets for the single reporting obligations of the WFD contributing to a comparable basis for harmonisation of water management on a river basin scale between member states within international river basin districts and to provide a transparent overview of the analysis and results to communicate with government, stakeholders and the public. Moreover, the development of WISE (Water Information System for Europe) will enable electronic reporting and visualisation of reported data. The system has been publicly opened on 22 March 2007.

References 1. European Commission, Groundwater Summary Report, Technical Report on Groundwater Body Characterisation, Monitoring and Risk Assessment, 2005. 2. European Commission, Groundwater Body Characterisation, technical report, 2004. 3. European Commission, Groundwater Risk Assessment, technical report, 2004. 4. Common Implementation Strategy for the Water Framework Directive, European Communities, 2003 (ISBN 92-894-2040-5). 5. European Commission, Guidance Document no 2, Identiﬁcation of Water Bodies, 2003 (ISBN 92-894-5122-X). 6. European Commission, Guidance Document no 3, Analysis of Pressures and Impacts, 2003 (ISBN 92- 894-5123-8). 7. European Commission, Guidance Document no 7, Monitoring under the Water Framework Directive, 2003 (ISBN 92-894-5127-0).

CHAPTER 5.2

Groundwater Quality Background Levels EMILIO CUSTODIOa AND MARISOL MANZANOb a

Technical University of Catalonia, Department of Geotechnical Engineering, Gran Capita`, s/n Ed D-2, ES-08034 Barcelona, Spain; b Technical University of Cartagena, Paseo Alfonso XIII, 52, ES-30203 Cartagena, Spain

5.2.1

Introduction

Groundwater forms complex, three-dimensional bodies in which recharge, ﬂow conditions and interaction with the solid matrix are point dependent. This means that, in a given groundwater body, the chemical, radiochemical and biochemical characteristics of water vary both in space (horizontally and vertically) and slowly with time. When anthropogenic eﬀects are added, variations may be intensiﬁed with respect to pristine conditions. Thus, the water quality of a given groundwater body cannot be represented by any set of single analytical values, and the degree of human inﬂuence cannot be established by a simple comparison to a reference list. The terms background, threshold and baseline quality values have been classically used in many scientiﬁc disciplines to try to identify anomalous concentrations with respect to what are considered as ‘‘typical’’ values. These values are critical to deﬁne water quality for a given use, and have to be deﬁned for groundwater as a guide for protection and remediation programmes. The baseline chemical composition or baseline quality of a groundwater body may be deﬁned, quite instinctively, as the physicochemical conditions due only to natural processes during recharge, ﬂow and water–rock interaction. Should this be possible, any impact on groundwater quality could easily be shown by comparing actual values to baseline values. Nevertheless in practice, problems appear when deﬁning baseline due to the common variability of the signiﬁcant diﬀerent chemical parameters. It is necessary to know whether a given concentration is natural, the result of hydrologic changes due to human activities or the result of introducing substances from outside into the groundwater body. 193

194

Chapter 5.2

Any adverse change in physicochemical properties is called contamination, and the substances producing this change are called contaminants. If these substances are artiﬁcially introduced in some way they can be called pollutants, and the result is pollution (see Chapter 1). These are the deﬁnitions that will be used in this chapter, although they are not universally admitted. Natural variability can be taken into account by describing the statistical distribution of values for any given parameter of interest reﬂecting the water’s natural quality of a groundwater body. This implies that a large enough number of unbiased measurements were made, which is not an easy task in a large, threedimensional groundwater body. For a given chemical or physical parameter, the range and distribution of values can be described by a set of statistical magnitudes like the mean, median, standard deviation, percentiles, maximum, minimum, etc., or by its full statistical distribution. In practice, this can be done only for a limited number of samples and for a small number of characteristic parameters, which may vary from case to case. Although this is not a new problem, existing experience is currently limited. However, statistical values are badly needed to correctly and eﬀectively apply the European Water Framework Directive (WFD),1 and especially the recently adopted Groundwater Directive.2 The situation explained above was in mind when the European Union BaSeLiNe (Natural Baseline Quality in European Aquifers, EVK1-CT199900032 and EVK1-CT-2002-00527) project was elaborated. The project started in 1999 and closed in 2003.3 In this project the following deﬁnition of baseline was adopted: ‘‘groundwater quality baseline is the concentration range in water of a given present element, species or substance, derived from natural geological, biogenic or atmospheric sources.’’ Thus, chemical concentrations are considered, taking into account water– rock interaction and the natural behaviour of chemical compounds along groundwater ﬂow lines. Both atmospheric contributions and chemical reactions are time dependent, and not all components have the same residence time in the system. As a consequence, the quality baseline of a given aquifer shows a range—sometimes a wide one—of values that varies in space and evolves slowly with time. For a given groundwater body, lithological heterogeneity and the fact that groundwater moves following more or less well-deﬁned ﬂow lines are the main contributors to baseline spatial variability. The main actions controlling baseline temporal variability are chemical reactions (redox, mineral solubility and surface processes such as adsorption and ionic exchange) and recharge conditions. This one involves evapo-concentration of airborne and soil-released salts, as well as solutes contributed by surface water. Groundwater pollution may clearly appear when looking at speciﬁc substances introduced as pollutants (tracers) and non-existing (to some extent) from baseline, especially those of fully artiﬁcial origin. Substances such as NO3, NH4, F, As, heavy metals, some radioisotopes (and isotope changes) may not be suitable tracers for pollution, since they may also appear naturally in groundwater under some hydrogeological conditions. The objective of the BaSeLiNe project was twofold: (i) setting scientiﬁc criteria to deﬁne groundwater quality baseline, and (ii) developing standard methods to

Groundwater Quality Background Levels

195

be applied in the European Union territory in order to guide in the fulﬁlment of the WFD requirements. The project, coordinated by researchers of the British Geological Survey, initially consisted of a consortium of 11 research groups of 9 European countries: Denmark, Spain, Portugal, Belgium, France, Estonia and Poland, United Kingdom and Switzerland as an associate country; later on, in 2002, they were joined by three other countries: Malta, Bulgaria and Czech Republic, which at the time were in the process of joining the European Union. The main objectives of the project were attained by means of theoretical and practical conceptual approaches, based on experience and real data from 21 aquifers of the consortium countries. It is clear that each aquifer is unique and its baseline quality depends on a particular combination of geological, hydrodynamical, climatic and soil cover characteristics. This means that results from a given aquifer cannot be used to accurately characterise any other aquifer. Nevertheless when climatic conditions, including distance to the coast, are not too diﬀerent and of secondary importance, it is possible to establish some typologies, based essentially on the dominant lithology, in order to help deﬁning baseline quality and planning analysis sampling, monitoring and study. But lithology is only one of the many factors inﬂuencing baseline quality. It becomes dominant in well-recharged aquifers by continental rainfall, but climatic conditions may dominate in semi-arid and arid areas, especially near the seashore, reducing and even overcoming the importance of lithological inﬂuence. This is a common situation in many peri-Mediterranean areas. Furthermore the residual inﬂuence of the sea on sediments may exert an important inﬂuence on groundwater quality, as happens in numerous coastal areas. Figure 5.2.1 shows how the airborne chloride deposition rate varies on continental Spain and how this is translated into groundwater chloride content due to water evapo-concentration in the soil. In this chapter some of the main methodological and conceptual results of the BaSeLiNe project are mentioned to illustrate how to determine the natural quality baseline of groundwater. This includes both the items to be considered and the preventive measures to be taken into account. The ﬁnal report of the project was made available via the internet.3

5.2.2

Rationale to Establish the Groundwater Quality Baseline

The need to deﬁne and establish baseline values of groundwater quality, and the criteria to set them, are a consequence of the enforcement, over the years, of diﬀerent acts and laws. In the USA, the Environmental Protection Agency (EPA) relies on the Federal Water Pollution Control Act Amendments of 1972, amended in 1977 and revised several time through the years. It is known as the Clean Water Act. Although dealing mostly with surface water, this act considers cleaning polluted groundwater in disposal sites and recognises that planning is needed to address the critical problems posed by non-point-source pollution. In the European Union, the WFD (2000) has been enacted, the main objective of which is setting the framework to protect continental, surface,

196

Chapter 5.2 10

France

3

3

1

1

0.5 10

3

1

1

1 0.5

Portugal

3

0.5

3 Airborne salts -2 -1 g·m ·yr Cl

1

3

0.5

N

1

1

3

3

3

3

0

100

200 km

France 20

10

10

5

5

10 5

20

Portugal

5

10 10

10 5

10 10

50 20

Figure 5.2.1

20

5

20 50 100

5

5

10

100

5 10

Recharge, mg/LCl-

20 20 50

N

5 100 0

100

200 km

Change of chloride content in groundwater in continental Spain. The upper panel shows atmospheric airborne chloride contribution in mg m2 yr1 and the lower panel shows the result of climatic/pedologic evapo-concentration, once discounted runoﬀ, as reﬂected in chloride content (mg l1) in the upper part of the groundwater table. Chloride baseline changes conspicuously throughout the territory. (Modiﬁed from Ref. 4.)

Groundwater Quality Background Levels

197

transitional, coastal and ground water, in order to prevent any further environmental deterioration and to improve current status. The Groundwater Directive,2 which was adopted at the end of 2006, is aimed at establishing strategies to prevent, control and correct groundwater contamination, as stated in Article 17 of the WFD. It neither modiﬁes nor enlarges the WFD objectives, but complements them in a way that presents some diﬃculties, since it is still insuﬃciently known by many of the oﬃcials that have to apply the provisions. The European directives are obligatory in the European Union territory, and have to be incorporated into national legislations in accordance with the subsidiarity principle. In fact, the WFD is now incorporated into countries’ national laws and water acts, and related legislation and rules have been or will be correspondingly adapted (see Chapter 3.1 for further details). In North America (USA and Canada), the term baseline applied to natural groundwater quality appears often in documents of the EPA and the United States Geological Survey (USGS), since at least the mid-1990s. But it seems that there is neither an oﬃcial document giving a deﬁnition nor the criteria to establish it. In Europe, the WFD does not use explicitly the term baseline, but the expression ‘‘background levels.’’ Moreover, it mentions repeatedly what is called ‘‘quality of surface water and groundwater bodies of the diﬀerent countries.’’ A ‘‘groundwater body’’ is deﬁned as a clearly diﬀerentiated volume of groundwater inside a given aquifer or aquifer system. Furthermore, the new Groundwater Directive2 incorporates the term ‘‘threshold values,’’ meaning ‘‘a concentration limit for a pollutant in groundwater, the exceedance of which would cause a body of groundwater to be characterised as having poor chemical status.’’2 The baseline or ‘‘threshold’’ should be established in waters participating actively in the hydrological cycle, except if they are already modiﬁed by contamination. This participation may be due to natural conditions or created by human intervention, such as pumping or deep drainage. Taking into account the spatial and temporal variability of natural quality baseline composition and the current scarcity of monitoring data and knowledge on the functioning of many aquifers, the Groundwater Directive does not provide a list of quality standards to be uniformly applied in the whole European Union, though it prescribes the application of existing nitrate and pesticide norms. In this respect, norms on drinking water, which are useful to protect human health, are not necessarily adequate as environmental guidelines. However, from this derives that groundwater bodies in which the limits are overcome must be classiﬁed as water bodies in ‘‘poor chemical status.’’ Thus, a body or group of bodies of groundwater shall be considered as having good groundwater chemical status when, according to the WFD and new Groundwater Directive,1,2 the nitrate concentration does not exceed 50 mg l1 NO3 (or a lower one if established for a nitrate vulnerable zone, following Directive 91/676/EEC), and the total content of active ingredients in pesticides, including their relevant metabolites (degradation and reaction products), does not exceed 0.1 mg l1. With regard to any other polluting substances, and especially for NH4, As, Cd, Cl, Pb, Hg, SO4, trichloroethylene and

198

Chapter 5.2

tetrachloroethylene, groundwater status has to be below the threshold value set by each country.2 The member states of the European Union are encouraged to enlarge the list according to identiﬁed risks to groundwater. The rationale for the new Groundwater Directive2 mentions that the 2003 meeting of the BaSeLiNe project in Funchal, Madeira Island, stressed the diﬃculty of setting uniform quality standards for groundwater, and emphasised the need to consider aquifer characteristics and actual pressures from human activity. Moreover, the document establishes criteria to identify signiﬁcant and sustained upward trends in pollution from human activity and to determine if there is a reversal, calling for a common methodology to test the statistical signiﬁcance of these trends (Article 1 of the directive). In order to correctly monitor the possible natural baseline deterioration of the diﬀerent water bodies, following the conclusions of the BaSeLiNe project, some groundwater ‘‘types’’ may be considered, according to the kind of aquifer containing the water and the concentration range of speciﬁc baseline indicators. Nevertheless, as already stated, baseline is a complex result and consequently the aquifer type is only one of the factors.

5.2.3

Methods to Establish the Natural Baseline Quality of Groundwater

To establish the baseline quality of an aquifer or groundwater body, the ideal situation is when available chemical data correspond reasonably well to areas unaﬀected by human activity. Generally this means pre-industrial age water. But this is not always easy or possible to get. Shallow levels of water table aquifers often contain anthropogenic components of diverse origin (acid rain, airborne pollutants, agrochemicals) and must be discarded. Multilayer aquifers present, in theory, some ideal conditions to obtain non-impacted waters from their deepest levels, provided they are not stagnant or contain saline water. However, often wells and boreholes are poorly constructed and grouted, and may produce a by-pass between exploited deep and shallow contaminated levels. This means that, sooner or later, young contaminated water may penetrate pre-industrial age water levels. Nevertheless, in some cases, there are aquifers containing pollution-free, young water, which are fully acceptable to characterise the reference natural baseline quality. In order to distinguish water of natural origin from anthropogenically impacted water, the BaSeLiNe project3 recommended the following approaches to be adopted, if applicable: (1) looking for the evidence that water age (or mean residence time) exceeds 50 to 100 years; (2) extrapolating available chemical data time series backwards, until it reaches a (theoretical) initial time in which there was no anthropogenic activity in the area; and (3) looking for substances that are clear indicators of human activity. These substances may be agrochemicals and their degradation compounds (including metabolites of pesticides), industrial products or an increase of dissolved nitrogen species or of total dissolved organic carbon. In order to identify the existence of a fraction

Groundwater Quality Background Levels

199

of young water with an anthropogenic inﬂuence, fully artiﬁcial substances are especially useful, such as many organic solvents, SF6, CFSs, etc. Even though the presence and the impact intensity of anthropogenic contaminants in water can be currently easily identiﬁed through modelling, this approach allows taking reasonable initial decisions on the way, provided that the signiﬁcant compounds are analysed with the needed analytical sensitivity, at least in the survey stages, independently of being part of the monitoring program. One of the more frequent contamination forms is the arrival to the recharge area of a given aquifer or water body of airborne contaminants external to the zone. This complicates the determination of natural quality baseline in small, intensively exploited aquifers, which are often an essential local water resource characterised by the short turnover time. The BaSeLiNe project suggests using, as a reference for these aquifers, the natural baseline established for other aquifers under similar geological, hydrogeological, climatic, etc., characteristics, reinforced when needed with the help of hydrogeochemical modelling, and the drilling of new monitoring boreholes for sampling, when some parts under natural conditions can be expected to be found. Since the same non-impacted aquifer may contain groundwater bodies of diﬀerent chemical composition (e.g. due to the presence of redox fronts, ion exchange gradients, waters of diverse marine continental origin), in practice the natural baseline quality of speciﬁc water bodies and their characteristic values should be explained by means of some main geochemical processes and the heterogeneities existing in the aquifer. In order to explain correctly a given baseline quality composition of a water body inside an aquifer, it is convenient to use ambient descriptor properties (and terms) such as ‘‘conﬁned,’’ ‘‘water table,’’ ‘‘oxidant,’’ ‘‘reducing.’’ The diﬀerent integrated tools applied in the BaSeLiNe project are commented on below. They constitute a proposal of methodology to establish the natural groundwater baseline quality, and contain the main conceptual and applied results of the project. These tools are: to study the major and trace inorganic components chemical data, in order to establish the variation range of natural baseline quality; to study the organic carbon data in order to establish the variability range of baseline quality and its usefulness as a contamination indicator; to carry out hydrogeochemical modelling, in order to identify and establish the types and characteristic times of the basic reactions controlling the baseline quality of the diﬀerent aquifer types; to use tracers and dating techniques to know the time scales that control the variation ranges of the diﬀerent components under consideration; to study baseline trends to know their causes and how to discriminate between those due to natural processes and those due to contamination. Adequate sampling, which is a key issue to know aquifer water quality, has to address the three-dimensional character of groundwater ﬂow and the importance of surface processes.5,6

200

5.2.3.1

Chapter 5.2

Study of Major and Trace Inorganic Component Chemistry

Natural water quality depends on characteristic concentrations of diﬀerent components, represented by the statistical mean or median of a set of values and their distribution (dispersion) around these central values. If the distribution is normal or log–normal, it is possible to deﬁne their dispersion by means of the standard deviation. However often this is not the case due to the simultaneous or correlative presence of more than one physicochemical process. In the BaSeLiNe project, in order to deﬁne the baseline quality, and after a ﬁrst evaluation of available data, it was decided to adopt the median as the most characteristic value for a parameter or component, and the 2.3% and 97.7% percentiles to show the variation range. Thus, most of the studied population (95.4% for a normal distribution) is inside the range. In order to reasonably describe the possible spatial variation of baseline quality, the chemical study of an aquifer or groundwater body should be carried out by using a large enough number of groundwater samples. This implies that new samplings may be needed. Moreover, this means that historical data prior to 1985 should be evaluated before being integrated to the younger ones, since many analytical techniques were less accurate than the current ones, and detection limits were too high for some components. An approach based on simple statistical, univariant techniques is proposed to establish characteristic values and variation ranges. In the BaSeLiNe project the cumulative frequency representation of data was selected in order to identify the main processes controlling the observed distributions (Figure 5.2.2). This type of representation was already used by Davis and de Wiest7 to show the distribution of elements in fresh groundwater. Figure 5.2.3 is a simple case. Other graphical representation types, which were considered adequate to show the natural baseline, are the box or whiskey plots (Figures 5.2.4 and 5.2.5) and that of time/space evolution of concentrations. Some of the main methodological remarks are the following. For populations not following the normal or log–normal distribution, outliers may be part of groundwater composition natural baseline quality and not necessarily the result of contamination. Such are the appearance of dissolved Fe21 and Mn21 in reducing groundwater ambients, the disappearance of NO3 and SO42 in similar environments or the high concentrations of sulfate when sulﬁde-rich sediments are supplied with dissolved oxygen or become desaturated. Thus, a careful study of these data is needed. The baseline quality variability may be of the same order of magnitude or even larger than that produced by contamination. The natural baseline quality being established depends on available data, and it is often biased due to diverse circumstances, such as taking samples by pumping (selection of the most permeable layers) or carrying out samplings with preference for some depths, or by mixing diﬀerent ﬂow lines (Figure 5.2.6).

201

Groundwater Quality Background Levels

6- Part of values are below analytical sensitivity (no data) 5- Small contents are discarded (less than limit) or have disappeared 1- Normal distribution 2- Concentrations are close to chemical equilibrium with minerals 3- Upper values are solubility controlled 4- Contamination or saline water admixture 1’- Bimodal distribution

Cumulative frequency, %

90 6

2 4 5

1

3 1’

50

Median

10

Log concentration

Figure 5.2.2

Plot of cumulative frequencies of chemical parameters, showing some typical circumstances. The median is used as the regional reference level, or the value to compare diﬀerent parameters. Type 1, normal distribution; type 1 0 , multimodal (bimodal as shown) distribution, both reﬂecting variability of recharge, water–rock interaction and turnover (residence) time in natural ﬂow systems. Type 2, small variability due to closeness to chemical equilibria with relevant minerals (for Si, Ca, Mg, etc.). Type 3, small variability at high concentrations reﬂecting that mineral solubility exerts control (e.g., ﬂuorine content due to ﬂuorite dissolution). Type 4, large variability of high concentrations resulting from addition of contaminated or saline water to a small fraction of samples. Type 5, fast decrease of low values pointing to the preferential reduction or elimination of a component by a geochemical reaction (e.g. nitrate reduction or sulfate reduction). Type 6, low values are below a threshold due to analytical conditions. (After Ref. 3.)

As a consequence, the following guidelines are proposed to determine the baseline quality of an aquifer or water body. To exclude samples known to be contaminated (information provided by some components). To carry out samplings along a ﬂow line and normal to it. The data used must take into account the three-dimensional distribution of water characteristics in the aquifers. This means considering the sample position with respect to the groundwater ﬂow network. In many cases what matters is the ﬂow conﬁguration existing under prevalent aquifer conditions, due to the slow movement and replacement of groundwater, and not the current ﬂow regime under disturbed conditions. This may be a major handicap for studies lacking scientiﬁc support.

202

Chapter 5.2

Figure 5.2.3

Plot of cumulative frequencies for major ions in Madrid’s basin aquifer. Fresh recharge water from the basin sides mixes with saline remnants in the central basin sediments, derived from old playa lake situations. The higher values are controlled by reactions with sedimentary silicates. (After Ref. 8.)

To compare local data to information from other areas that are surely not aﬀected by human activity. To use times series, when existing, to detect early time conditions. To use hydrogeochemical modelling as a tool to know if the assumed and deduced processes are natural or need artiﬁcial conditions to be active. To make a limited use of statistical techniques, since hydrologic and hydrogeochemical considerations are generally not taken into account and thus not enabling key processes to be clearly identiﬁed.

5.2.3.2

Organic Component Chemistry

Organic carbon dissolved in groundwater is an important reactant in natural geochemical processes. Furthermore, it may be useful to determine the natural or uncontaminated status of a given water as a contamination indicator, e.g. of disposal sites and used waters, and may be a potential contaminant as well. It is a source of energy and food to bacterial populations, both in aquifers and in distribution networks, and also plays an important role in the mobilisation of trace metals, radionuclides and biogenic components coming from outside.

Groundwater Quality Background Levels

Figure 5.2.4

203

Box plot (whiskey plot) of major ion concentrations in groundwater samples from Madrid Tertiary Detrital aquifer. Groundwater is predominantly of the sodium bicarbonate type, in agreement with the arcosic nature of sediments, but there is also some saline groundwater formed in old evaporating playa lakes, of mostly the sodium–calcium sulfate type. Excess sodium is part of baseline. (After Ref. 8.)

Additionally, its presence and concentration is important for the evolution of underground redox fronts. Organic carbon is part of the organic matter present in groundwater, which is of two types: humic and non-humic substances. The latter group includes decomposable plant material, living biomass and woody plant material. Both types can be present as solved species or as particles. All natural waters contain some dissolved organic carbon (DOC), here considered—from a practical approach—as that remaining in water ﬁltered through a 0.45 mm sieve. The total organic carbon (TOC) is that measured in unﬁltered waters and includes both the dissolved and the particulate fraction (larger that 0.45 mm), such as bacteria and phytoplankton. For correctly sampled groundwater, in most cases DOC is about the same as TOC since the sieving eﬀect takes place in the aquifer, except for coarse formations, in which even some micro-organisms could be transported. Work carried out mostly in the last decade recognises the signiﬁcant role of TOC in some important hydrogeochemical processes, such as natural weathering or redox reactions, which control the evolution of some hydrogeochemical environments in aquifers. Furthermore, the particulate matter, especially that of colloidal size (most of it smaller than 0.45 mm),

Figure 5.2.5

Whiskey plots (box plots) to show median and range variation of diﬀerent major and trace components and physicochemical parameters of interest for baseline quality in fresh water (A) and saline water (B) of the same aquifer (Don˜ana, southwestern Spain). The aquifer system has at least two hydrogeologically diﬀerent water bodies. (After Ref. 9.)

204 Chapter 5.2

Groundwater Quality Background Levels

Figure 5.2.6

205

Schematic representation of aquifer behaviour in El Abalario, Don˜ana Natural Park (Huelva, Spain), showing a cross-section between El Abalario dome and La Rocina creek. The water table aquifer consists of ﬁne–medium sands with a thin coarse layer of a much more permeable gravely formation, where wells have their screens. The ﬂow lines (lines with arrows) have an important vertical component, which is downwards in most of the area. Nitrate contamination in irrigated ﬁelds (mostly fruit trees and strawberries under plastic cover) moves downwards in the sands, at about 0.5 m per year. Currently a large part of the sands is contaminated by high concentration of nitrate, but this still does not appear in many wells that are deep screened nor in deep discharges into the main water course, though it does in the shallow local creeks. Well water still shows baseline values but the water body is seriously damaged, and it will progressively worsen. The medium is oxidant, so nitrate is not reduced. In other areas, in which the aquifer is thinner, nitrate pollution attains the full thickness.

plays an important role in the transport, mobilisation and degradation of contaminants. Natural groundwater often has TOC concentrations less than a few milligrams per litre, although values higher than 50 mg l1 can be occasionally found. Data studies carried out within the BaSeLiNe project show that TOC concentrations decrease with increasing depth, and that diﬀerences between median values of natural water from siliceous and carbonate aquifers containing young and old water are small. The measured values vary between 0.7 and 1.8 mg l1, which are signiﬁcantly smaller than those measured in clearly

206

Chapter 5.2

contaminated aquifers, in which the values may be of the order of several tens to some hundred milligrams per litre. A clear relationship between TOC and other similar parameters such as assimilable organic carbon, halogenated organic compounds or the bacterial counting has not yet been found, and the scientiﬁc knowledge about the diﬀerent organic molecules and their reactivity, toxicity and ability to mobilise various contaminants is still being developed. Though early work was done more than 40 years ago,10 research intensiﬁcation has occurred during the last ten years, producing excellent manuals11 and a good number of scientiﬁc papers. Published work is mostly related to biodegradation in natural remediation, paying attention to some compounds (plaguicides, halogenated organic carbons, hydrocarbons) and their metabolites, as well as selective behaviour through 13C isotope evolution.12,13 The knowledge gained in upcoming years could be relevant.

5.2.3.3

Hydrogeochemical Modelling

As already stated, groundwater natural quality is the result of complex interactions between the solid, gaseous and liquid phases. The resulting composition may be in some places of the same or of a higher order of magnitude than that due to contamination in other places. Hydrogeochemical modelling coupled to water ﬂow is a necessary tool to get a qualitative and quantitative knowledge of the main and more frequent processes controlling groundwater quality (dissolution/precipitation, ion exchange, redox reactions, adsorption; Figures 5.2.7 and 5.2.8). Once the signiﬁcant processes are known and quantiﬁed, modelling can be used to predict future water quality changes, both due to ambient

Figure 5.2.7

Interpretation of chemical logs along boreholes to show hydrogeochemical changes when going through a redox front. Data points are experimental values and the continuous curves are the result of ﬂow and transport modelling. Values are in mmole l1 (mM). (After Ref. 3.)

207

Groundwater Quality Background Levels

Interpretation of measures along a ﬂow line to show hydrogeochemical changes when there is water–rock interaction. Only Ca21 evolution is shown. Data points are experimental values and the curves are the result of ﬂow and transport modelling, with only one process (insuﬃcient to explain results) and with two processes. (After Ref. 3.)

Concentration

Figure 5.2.8

Extrapolated trends Modelled trend

Baseline data period

First available data

Figure 5.2.9

Today

Time

Schematic representation of trend modelling using the data period. Extrapolation backwards until anthropogenic processes are not present or a steady state is observed allows the approximate knowledge of baseline quality for the chemical parameter being considered. Extrapolation forwards allows visualisation of trends under given scenarios or hypothesis. (After Ref. 3.)

natural variations (e.g. natural climatic changes, tectonically induced ﬂow modiﬁcation, subsidence, sedimentation, erosion) and to the impact of diﬀerent human activities, including the future removal of some of the current ones (Figure 5.2.9). Modelling facilitates the study of spatial and temporal trends, and helps in the correct design of expensive monitoring programmes.

208

Chapter 5.2

In the BaSeLiNe project, programme PHREEQC was used to study the processes responsible for the natural composition observed and its evolution, especially the changes occurring along a ﬂow line in three aquifer types that illustrate frequent situations in Europe: silica-dominated sedimentary aquifers with rare carbonates; carbonate aquifers or aquifers containing abundant carbonate; and aquifers with old saline water being displaced by younger freshwater. More sophisticated approaches, although not necessarily more eﬀective, are those simulating ﬂow and reactive mass transport. Nevertheless a detailed insight to processes allows one to deﬁne more accurately what is happening14 or to identify controlling facts, e.g. in seawater intrusion into aquifers.15 The most signiﬁcant conclusions related to modelling are the following. Many patterns and trends appearing often in natural waters, such as ion exchange gradients, are the result of processes occurring at geological time scale and are generally due to ﬂow conditions that preceded those existing today, even if equilibrium is quickly attained. Under such conditions the supply of reactants may be the limiting factor. This means that the ‘‘time’’ parameter must be carefully used in simulation works. Independent dating measures using diﬀerent tracers are desirable to limit uncertainty. When this information is not available and time is derived from a modelling exercise, the uncertainty associated must be clearly shown. Groundwater development distorts natural chemical gradients, and this is very diﬃcult to correct and simulate. Mixing processes between old and young waters also change the original composition. Therefore, some discontinuities in the natural composition of some aquifers may be due to ‘‘age gradients,’’ while smooth changes following the groundwater head gradient show the existence of continuous processes. When an aquifer does not contain water of natural origin, modelling may yield a realistic estimation of the original concentration of some elements.

5.2.3.4

Tracers and Temporal Scales

In order to interpret water quality changes referring to variation of the natural baseline quality, the knowledge of the age – the turnover time – of groundwater in the ﬂow system is needed, as well as the temporal scale at which the diﬀerent hydrogeochemical processes explaining natural baseline chemistry occur. Under favourable conditions water age may be reasonably estimated with suﬃcient accuracy, but this may be often impossible, due to the unavailability of suitable ideal tracers. In order to solve these diﬃculties, the use of several tracers and the numerical simulation of water ﬂow and solute transport are necessary. The dating principles are the following. It is interesting to measure the components whose concentration in water varies along time, due to known causes distinct from water–rock interaction.

209

Groundwater Quality Background Levels

These components are like ‘‘clocks.’’ The diﬀerent known types are: – cumulative processes (3He, 4He, some chemical components, etc.) – radioactive decay processes (39Ar, 14C, 32Si, 3H, 85Kr, etc.) – variable but known incorporation to groundwater (3He, CFCs, SF6, etc.). Within the BaSeLiNe project diﬀerent tracers (13C, 14C, 3He/3H, 85Kr, 39Kr, CFSs, etc.) and water molecule isotopes (18O, 2H, 3H) are used, mostly to know the groundwater age structure in the studied aquifers (Table 5.2.1). Figure 5.2.10 is an example. Age structure is the result of integrating values derived from diﬀerent techniques into the hydrogeochemical and hydrodynamical processes responsible for water characteristics, age being one of them. Water age determination implies that processes other than time-dependent ones (e.g. radioactivity) have been adequately corrected to remove other eﬀects such as dilution by dead (non-radioactive) matter or exchange with solid. This is not an easy task. Ages obtained without corrections (or with only partial corrections), namely apparent ages, are only an approximation, and sometimes a crude one. One of the main contributions of time tracers to natural baseline quality determination is the estimation of the residence (turnover) time scale of groundwater in the aquifers being studied. Before carrying out long, extensive and expensive hydrogeochemical studies, this information will help to predict if water will keep in the future its natural composition. This aspect may be shown by determining whether a young water component (3H, 85Kr, 39Ar, CFSs, SF6, etc.) is present or not, taking into account the following. If the presence of some (generally various) of these components shows that water is younger than 50 years, this means that it is potentially impacted by human activity. Therefore, it is probable that the natural background has been changed, and consequently the baseline quality. The lack of any young component guarantees that original natural conditions prevail.

Table 5.2.1

Substances (asterisk indicates radioactive) potentially useful for groundwater to date under favourable circumstances and range of years that can be dated.

Substance

Origin

Range (years)

Application

3

Natural, nuclear bombs, nuclear reactors Nuclear reactors Industrial, domestic Natural Natural, (nuclear bombs) Natural

5–50 (200)

Easy

10–50 10–50 50–2000 1000–20 000

Diﬃcult Medium Diﬃcult Easy

200–10 000 000

Medium

H* (tritium); 3 H*/3He 85 Kr* CFCs, SF6 39 Ar* 14 C* (radiocarbon) 4

He

210

Figure 5.2.10

Chapter 5.2

Simultaneous measurement of three isotopes to date in a north–south cross-section in Don˜ana National and Natural Parks, southwestern Spain. The ﬁgure shows tritium (half-life of 12.43 years) in tritium units (TU), radiocarbon (14C, half-life of 5730 years) in percent modern organic carbon (pmc), and 13C content (stable, indicating origin and behaviour of dissolved carbon in the aquifer) in deviation per mile from the PDB standard. Most samples are mixtures of waters from diﬀerent depths, where relatively high tritium (young water) may coexist with relatively low radiocarbon (old water). This helps in interpreting chemical data. The measurement of 85Kr and 39Ar at some points allow improved understanding. (After Ref. 16.)

If the lack of young components indicates that water age is greater than 50 years, groundwater quality corresponds to water–rock interactions evolving through time. The knowledge of these reactions needs both hydrogeochemical modelling and dating with adequate tracers. What has been stated allows for the preliminary classiﬁcation of any chemical parameter, independently of its variation range, into two groups with diﬀerent residence times: (1) parameters that are measured in water with young components, and (2) parameters that are measured in water free of young components. This simpliﬁes sampling and data treatment, in order to establish the natural baseline quality. The main result of the study of water residence time in the diﬀerent aquifers considered in the BaSeLiNe project is that a unique universal technique or set of techniques does not exist, but each aquifer needs the application of a series of speciﬁc techniques, which mainly depend on the time scale of water residence time, the geochemical ambient (e.g. in redox media many of the available

Groundwater Quality Background Levels

211

tracers may be not useful), the economic resources and the analytical tools that can be used.

5.2.3.5

Study of Natural Baseline Trends

This activity is a result of what is required in the Groundwater Directive:2 the use of time series in order to observe the possible presence of increasing, sustained and statistically signiﬁcant trends in the concentration of some chemical components, due to contamination; and monitor and detect the reversal in these trends after implementing the corresponding remediation actions. The use of statistical techniques as a principal indicator is still much discussed in hydrogeological forums. If hydrogeochemical and hydraulic techniques are not considered, there is the risk of mistaking natural increasing trends for contamination. In the framework of the BaSeLiNe project the study of historical data sets has been addressed: to know which type of time series can be expected to be found in diﬀerent European Union countries; to observe and deﬁne natural baseline trends in order to obtain chemical support to understand the natural functioning of aquifers; and based on what has been said above, to discriminate natural changes from those due to anthropogenic activity. Only a few countries have good time quality series on groundwater quality, except for some special aquifers. Length is often less than 15 years, but some Eastern Europe countries have the longest series, up to 70 years. The study of these series has allowed one to distinguish two types of spatial and time trends. 1. Trends of natural origin: Due to processes that cause changes at the aquifer scale. They depend on solute transport velocities through the medium, and therefore are very slow, such as the replacement of saline water by recently recharged freshwater. Due to small-scale space variability caused by aquifer heterogeneity, also of small scale. This causes ﬂuctuations around some level that may be erroneously interpreted, if only a statistical approach is used, as the result of contamination processes (increasing trends) or as reversals of them (decreasing trends following other increasing tendencies). 2. Trends due to aquifer development, which have their only cause in natural processes and not in human contamination. The cause is the ﬂow

212

Chapter 5.2

velocity increase due to groundwater development, thus accelerating the appearance of natural trends that otherwise would require much longer times to be observed. Groundwater development may also induce natural chemical reactions that otherwise would not be produced if abstraction did not exist. This is the case of water mixing by upconing or by changes in the redox state, pH, etc., related to water table oscillation or the displacement of water bodies of diﬀerent characteristics. The former trends are related to hydrodynamic and hydrogeochemical conditions, and have to be interpreted and studied before trying to interpret the latter ones. Figure 5.2.11 shows a simple example and Figure 5.2.12 shows how diﬀerent chemical parameters correlate with chloride, indicating processes that may be present in trends. The main methodological and conceptual conclusions of the study of chemical trends are the following. In order to diﬀerentiate a natural trend from an evolution due to contamination, the origin of the observed trends has to be understood.

Figure 5.2.11

Chemical trends. The hydrodynamic and hydrogeochemical study is the key to know if there is a trend and if its causes are natural or due to contamination. Time evolution of Ca21 in a pumping well with trend and time series of Mn21 in a pumping well without trend. Central trends are also shown. (After Ref. 3.)

Groundwater Quality Background Levels

Figure 5.2.12

213

Plot showing correlations of some ions versus chloride content. It shows the existence of geochemical processes that may not be obvious from the time series plot.

The main processes producing changes in groundwater quality trends are – changes in the mixing proportions between diﬀerent waters; this happens in the salinisation by lateral ﬂow or saline upconing or when substituting connate saline water with freshwater; – changes of redox conditions and fronts, such as the occurrence of pyrite oxidation and the displacement of nitrate, sulfate or iron reduction fronts; and – changes in recharge water composition, such as those due to variations of recharge rate or rainfall composition. In order to conﬁrm the presence of water quality trends, be they natural or of anthropogenic origin, historical series are needed. When they exist, trends have to be interpreted. Going backwards, the original composition of groundwater can sometimes be determined, which is very important for chemical data interpretation. It is recommended that all new drilled wells or boreholes in a given aquifer be adequately sampled, in order to keep this information as an initial reference in databases. In the same way, the measurement of initial chemical characteristics will allow future studies to beneﬁt from a highly valued initial reference

214

Chapter 5.2

information to study the natural trend, and also for numerical modelling and for aquifer management. In order to monitor groundwater quality trends in a given aquifer or groundwater body, ﬁrstly the diﬀerent natural trends have to be identiﬁed.

5.2.4

Conclusions

The relatively recent enforcement of the European Union Water Framework Directive,1 and the enforcement of the newly adopted Groundwater Directive,2 with their strict requirements of member states with respect to the quality status characterisation of diﬀerent groundwater bodies and the implementation of remediation programmes, need a deﬁnition of baseline quality with two clear objectives: to distinguish between natural quality and quality modiﬁed by the presence of anthropogenic components; and to establish the characteristic natural composition of the diﬀerent aquifers or groundwater bodies to serve as a reference, in case of implementing remediation activities. In the European Union project BaSeLiNe,3 carried out between 1999 and 2003, a deﬁnition of natural baseline was established, with scientiﬁc support, being at the same time usable by managers and policy-makers. Furthermore, a methodology to establish the natural baseline quality and its origin was proposed and tested, appearing to be adequate for application to all Europe. This methodology has been summarily presented above, as well as the main applied and conceptual conclusions derived from the application to 21 European aquifers. As a summary, the main conclusions are the following. The interpretation of observed groundwater chemical composition should always be made taking into account the aquifer hydrodynamical functioning. This is the basis to establish natural baseline quality or to deduce which natural or contamination-induced geochemical processes are acting, as well as to detect trends and their possible initial and ﬁnal time. Since the functioning of most aquifers is poorly known, this introduces the need, supported by water law requirements, of carrying out the studies and undertakings adequate to get a basic knowledge in the upcoming 10–15 years. Among these works are (a) the drilling of some new boreholes and their adequate and complete study by means of geological, geophysical, geochemical and hydrodynamical logs, (b) the start of time quality series to monitor baseline quality and to detect any change since the beginning and (c) the use of numerical models, not necessarily sophisticated ones, in order to (i) complete and reinforce

Groundwater Quality Background Levels

215

aquifer functioning conceptual models, including water quality origin, (ii) know what is naturally or artiﬁcially producing the modiﬁcations, (iii) forecast the foreseeable future evolution and (iv) establish the main characteristics of natural baseline quality when it is not reﬂected in the water currently in the aquifer.

Acknowledgements Most ideas and results reﬂected in this chapter derive from the joint work of the European Union research project BaSeLiNe (natural baseline quality in European aquifers: a basis for aquifer management, EVK–1–CT1999–0006), in which more than 30 scientists and experts—among them the authors of this chapter—worked during the period 1999–2003. The project results were considered in drafting the text of the European Daughter Directive on Groundwater. A synthesis of the work is available at the project site.3 Due credit should be given to the team responsible for each of the work packages, from whom many of the ideas expressed in this chapter have been borrowed. Some ﬁgures come also from the project papers. Besides the documents available through the web, a book is being prepared, and is well advanced, co-edited by Mike Edmunds and Paul Shand, managers of the BaSeLiNe project.17

References 1. Directive 2000/60/EC of the European Parliament and of the Council establishing a framework for the Community action in the ﬁeld of water policy (EU Water Framework Directive), Oﬃcial Journal, OJ.L.327, 22-12-2000, 2000. 2. Directive of the European Parliament and of the Council on the protection of groundwater against pollution and deterioration, OJL 372, 12.12.2006. 3. Natural BaSeLiNe quality in European aquifers: a basis for aquifer management, 2003 (www.bgs.ac.uk/hydrogeolgy/baseline/europe/EU_Baseline.pdf ). 4. F. G. Alcala´, Recarga a los acuı´ feros espan˜oles mediante balance hidrogeoquı´ mico [Recharge to Spanish aquifers by means of hydrogeochemical balance], doctoral thesis, Technical University of Catalonia, Barcelona, 2005. 5. L. G. Everett, Groundwater Monitoring, Genium, Schnectady, NY, 1987. 6. J. R. Boulding, Practical Handbook of Soil Vadose Zone and Ground-water Contamination, Boca Raton FL, Lewis, 1995. 7. S. N. Davis and R. J. M. de Wiest, Hydrogeology, John Wiley, New York, 1966. 8. M. A. Herna´ndez-Garcı´ a and E. Custodio, Environ. Geol., 2004, 46, 173–188. 9. M. Manzano and E. Custodio, Groundwater baseline chemistry in the Don˜ana aquifer (SW Spain) and geochemical controls, 4th Assembleia Luso-Espanhola de Geodesia e Geofı´ sica, Figueira de Foz, Resumos, 2004, S13.7, pp. 729–730.

216

Chapter 5.2

10. M. A. Messineva, in: Geologic Activity of Microorganisms, ed. S. I. Kuznetsov, Trans. Inst. Microbiol. IX. Consultants Bureau, New York, 1962. pp. 6–24. 11. F. H. Chapelle, Groundwater Microbiology and Geochemistry, John Wiley, New York, 2001. 12. J. T. Wilson, R. Kolhatkar, T. Kuder, P. Philp and S. J. Daugherty, Ground Water Monit. Remed., 2005, 25, 108–116. 13. L. G. Kennedy, J. W. Everett and J. Gonzales, J. Cont. Hydrol., 2006, 83, 221–236. 14. T. Xu, J. Samper, C. Ayora, M. Manzano and E. Custodio, J. Hydrol., 1999, 214, 144–164. 15. M. Rezaei, E. Sanz, E. Raeisi, C. Ayora, E. Va´zquez-Sun˜e´ and J. Carrera, J. Hydrol., 2005, 311, 282–298. 16. M. Manzano, E. Custodio and M. Colomines, El fondo hidroquı´ mico natural del acuı´ fero de Don˜ana (SO Espan˜a) [Natural hydrochemical baseline of the Don˜ana aquifer, southwestern Spain], 5th Congreso Ibe´rico de Geoquı´ mica/9th Congreso de Geoquı´ mica de Espan˜a, Soria, 2005, pp. 1–13. 17. W. M. Edmunds and P. Shand, The Natural Baseline Quality of Groundwater, Blackwell, Oxford, in press.

CHAPTER 5.3

Groundwater Age and Quality KLAUS HINSBY,a ROLAND PURTSCHERTb AND W. MIKE EDMUNDSc a

Geological Survey of Denmark and Greenland, GEUS, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark; b Climate and Environmental Physics, Physics Institute, University of Bern, Sidlerstrasse 5, CH-3012 Bern, Switzerland; c Oxford Centre for Water Research, Oxford University Centre for the Environment, South Parks Road, Oxford OX1 3QY, UK

5.3.1

Introduction

The pressures on groundwater quality and quantity have increased dramatically during the past 50 years due to increasing demands for freshwater and contamination from a wide range of human activities. Before 1950 the human impact on groundwater quality was insigniﬁcant or limited, and the groundwater composition was in practice close to the natural background in most aquifers,1,2 Since about 1950 increasing contents of contaminants such as nitrate and pesticides have been found in groundwater and ecosystems globally.2–4 In the same period an increasing number of pollution plumes from point sources appear below urban and industrial areas in most parts of the world. The increasing pressure on groundwater quality and quantity leads to an increasing pressure on both water resources and dependent ecosystems and an increasing need for eﬃcient tools for the development of a sustainable integrated water management and policy. The European water framework and groundwater directives (see Chapter 3.1) provide the framework for developing such tools, e.g. as exempliﬁed by the eﬀorts of estimating the natural background quality of European groundwater and nutrient mobility within European river basins.1,5 Groundwater dating by environmental tracers and numerical groundwater ﬂow models are important tools for understanding the temporal and spatial hydrochemical evolution and the pressures of pollutants on groundwater and dependent ecosystems. Figure 5.3.1 illustrates trends of important environmental tracers in the atmosphere, which can be used to estimate groundwater ages, evaluate the temporal hydrochemical evolution of groundwater and indicate human impact. 217

218

Chapter 5.3 CFC, SF6 (pptv), 14C (pmc)

85

85

600

Kr (mBq m-3)

3

H (TU)

6000

Kr CFC-12 1200

500

5000 3

400

1000

SF6 x100

H

4000 800

300

3000 CFC-11

200

600 400

2000

14

C

100

200

1000

CFC-113 0 1940

Figure 5.3.1

1950

1960

1970

1980 Year

1990

2000

0 2010

Concentration trends of selected environmental tracers in the atmosphere applicable for groundwater dating (modiﬁed after Hinsby et al.2).

Important factors controlling both groundwater quality and quantity issues and the eﬀect on dependent ecosystems include how fast groundwater is recharged, how fast it ﬂows and how it interacts with the aquifer sediments and rocks. The computation of these parameters requires detailed geological information, which is most commonly not available for the subsurface. The content of environmental tracers in groundwater, and groundwater age estimation of water sampled from wells or monitoring points provide valuable information on travel times and the risk of contamination at these points. A simple evaluation of the existence of a modern water component in a water sample from a monitoring point or well by measuring the contents of one of more of the tracers shown in Figure 5.3.1 gives a ﬁrst indication of possible contamination at this point.2,6 Considerable research eﬀorts during the past decade have demonstrated more advanced applications of new environmental tracers and groundwater modelling techniques. The new techniques provide more detailed understanding of subsurface ﬂow systems including the possibility of estimating absolute groundwater ages, history and fate of groundwater contaminants and the interaction with dependent ecosystems (see Chapter 10.2).7–10 The quality and the quantity of the subsurface water resources and their impact on dependent ecosystems are closely linked to the age of groundwater.11,12 The aquifers have a considerable attenuation potential for most contaminants,13,14 and the risk of pollution therefore decreases with increasing groundwater age along ﬂow paths. Sustainable management of the water resources and the dependent ecosystems therefore requires a solid understanding of the groundwater age

Groundwater Age and Quality

219

distribution, the temporal and spatial evolution of the subsurface hydrochemistry and the groundwater/surface water interaction.10–19

5.3.2

Groundwater Age Estimation

There are basically two diﬀerent ways of estimating groundwater age at a groundwater well or monitoring point: (1) by environmental tracers and geoindicators and (2) by groundwater ﬂow modelling. This chapter focuses on the application of environmental tracers and geoindicators as groundwater dating tools, while groundwater ﬂow modelling is only brieﬂy discussed.

5.3.2.1

The Deﬁnition of Groundwater Age

Groundwater age is generally considered as the average travel time for a water parcel from either the surface or from the water table (point of recharge) to a given point in the aquifer. In humid sandy areas with thin (o5–10 m) unsaturated zones the diﬀerence between these is generally negligible. In arid areas with thick unsaturated zones the diﬀerence may be considerable. For European conditions the diﬀerence is generally believed to be of less importance and in this chapter we therefore use the term groundwater age to cover both situations. The term ‘‘residence time’’ was originally deﬁned diﬀerently, but we prefer to use it synonymously with ‘‘age’’ as this is commonly done at present, and as this concept is of more use in groundwater studies. Hence, the residence time of groundwater is here deﬁned as the average travel time between the point of recharge and the point of discharge, e.g. to a river or a lake or to any monitoring point in the groundwater zone. The tracer age estimate is normally considered and described as the average age of the water sample. This is a good approximation in cases where the ﬂow system is simple, and can be approximated by a piston ﬂow model (insigniﬁcant mixing and dispersion19). However, where signiﬁcant mixing and dispersion occur, e.g. in long screens in water supply wells, in fractured dual porosity aquifers and in groundwater bodies with signiﬁcant aquitards, the estimated tracer model age may either underestimate or overestimate the actual mean age of the water parcel.20,21 A sound knowledge of the geological setting and both physical and chemical processes in the aquifers is therefore important for the right interpretation and application of the environmental tracers and computed groundwater ages.

5.3.2.2

Environmental Tracers for Absolute Age Estimation

All environmental tracer dating methods are based on chemical or isotopic concentration variations as function of time. The time resolution of a groundwater dating method depends on two fundamental factors: the accuracy of the functional relation between a measured concentration and the groundwater age; and

220

Chapter 5.3

the temporal gradient of the tracer concentration in relation to the analytical precision of the measurements. Varying tracer concentrations are groundwater is either the result of a changing input concentration or sinks and sources of tracer in the subsurface. For dating the input history and the accumulation or decay rates have to be known. Many natural tracers are produced in the atmosphere due to nuclear reactions with cosmic rays (examples are 39Ar, 14C or 81Kr). A constant atmospheric equilibrium concentration will be achieved if the production rate is not changing. Dissolved in the precipitation these isotopes enter the subsurface and are subject to radioactive decay. The radioactive ‘‘clock’’ is in the ideal case the only process changing the tracer concentration in the subsurface. This is very often true for noble gases and why they are preferred tracers despite the high analytical requirements. On the other hand, chemically reactive tracers are easier to measure, but they can be aﬀected by dissolution or degradation processes, which are diﬃcult to quantify. The use of 14C for example is therefore complicated for groundwater dating in some settings. During the past 50 years, human activities have released a number of chemical and isotopic substances into the atmosphere. In the atmosphere they have mixed and spread worldwide. The monitoring of the atmospheric concentrations of these substances provides an ideal input function for groundwater dating (Figures 5.3.1 and 5.3.2). 3H was introduced in the atmosphere as a result of nuclear bomb tests, chloroﬂuorocarbons (CFCs) were intensively used in refrigerators and air-conditioning, 85Kr is released during reprocessing of fuel rods from nuclear reactors and SF6 is mainly used as an electrical insulator in high-voltage switches.23 All these tracers are gases, except 3H, which is usually part of a water molecule. The initial amount of tracer dissolved in the water depends therefore also on the recharge conditions namely the recharge temperature and pressure as well as the entrapment of excess air.24 3H and 85Kr decay with half-lives of 12.32 and 10.76 years respectively. CFCs and SF6 are non-radioactive and chemically stable under oxic conditions. However, CFCs can be degraded in anaerobic environments.25–27 resulting in an overestimation of groundwater residence times deduced from CFC measurements. Mixing of younger and older water due to dispersion or extended screen intervals is a factor that has to be considered for all dating tracers. The combination of diﬀerent tracers or time series can help to identify and to quantify such processes (Figure 5.3.2) The input function and the decay rate in the subsurface are the crucial parameters deﬁning the most sensitive dating range of a tracer. Normalised functional relations of tracer concentrations and groundwater age are shown in Figure 5.3.3(a). Compared to the atmospheric input these concentrations are smoothed out due to dispersion. The analytical uncertainties of the methods divided by the concentration gradient gives an estimate of the time resolution of a dating method (Figure 5.3.3(b)). Other factors aﬀecting the tracer age are not considered here. The high precision of mass spectrometric (MS) measurements result in very accurate dating results for the 3H/3He method28 and age

Groundwater Age and Quality

Figure 5.3.2

221

Dating principle using environmental tracer methods. The tracer concentration in the recharge water is given by the input concentration (1) in the atmosphere and the recharge conditions (2) (temperature, pressure, excess air, etc.). In the aquifer diﬀerent ﬂow lines with diﬀerent ﬂow times mix to a whole age distribution, e.g. in a screened borehole (3). The correct age distribution has either to be assumed based on the hydrogeological situation or can be constrained, e.g. by measured time series (4). The admixture of old water (5) (450 years) which is free of 3 H, 85Kr, etc., can be quantiﬁed with a two-tracer approach (5). The old water shifts both concentrations towards zero. The dilution ratio indicates the amount of old water whereas the isotope ratio deﬁnes the age of the young water (examples: (a) 100% young water with an age of 20 years; (b) 60% young water with an age of 10 years and 40% water older than 50 years). The age of the old water can further be constrained with 39 Ar and 14C measurements.

diﬀerences of a few weeks can be measured.29 The uncertainties of 85Kr, SF6 and CFC ages are in the ideal case of the order of months to a few years. Decreasing CFC concentrations since the 1990s reduce the dating resolution for young waters. A similar problem arises for 3H with almost no concentration gradient over the last 10 years. 39Ar dating provides the most reliable dating results in the age range 60–900 years with a dating error of approximately

222

Figure 5.3.3

Chapter 5.3

(a) Normalised tracer concentrations in groundwater in a dispersive aquifer as a function of groundwater age. (b) Dating resolution calculated based on the analytical error of tracer measurements and the concentration gradients shown in the ﬁgure above. These errors have to be interpreted as lower limits because of other limiting factors like degassing, unknown recharge conditions, degradation or subsurface production, etc. (c) Most advantageous dating ranges of the diﬀerent methods.

30–50 years. For older waters 14C is the method of choice. However, because groundwaters are commonly mixtures of waters with diﬀerent age multi-tracer measurements are required for the characterisation of complex age structures. An example is given in Figure 5.3.2. With the combination of two tracers it is possible to determine, for example, the age and mixing portion of young water components. In the context of this volume it is very often suﬃcient to investigate whether or not recent water components are present in a sample. For this purpose any of the tracer methods 3H/3He, CFC, SF6 or 85Kr is in principle

Groundwater Age and Quality

223

suitable. The most favourable tracer has to be investigated in the individual case based on the objectives and the hydrological and geochemical conditions.

5.3.2.3

Geoindicators: Estimating Environmental Change and Relative Ages

Indicators of environmental change have been extensively developed, mainly using biological parameters. A series of geoindicators has been proposed,30 which can detect environmental change over a timescale of 100 years. Among these groundwater quality has been proposed as a sensitive indicator of overall change since human impacts are recorded in the groundwater body.31 Two levels of indicators are suggested which monitor physical change: changes in the natural hydrogeochemistry and the main anthropogenic inﬂuences. The primary indicators (water level, HCO3, DO, Cl, NO3, SO4, DOC) may be supported where possible by various secondary indicators, which help to characterise the sources of contaminant or the geochemical processes involved. The unsaturated zone in unconsolidated lithologies is also proposed as a target for monitoring. Under favourable circumstances in porous media a decadal record of the recharge rate, recharge history, products of geochemical reaction and records of pollution may be observed since downward rates of transport amount to 0.5–1.0 m yr 1. Chemical information in groundwater may also be used as an indicator of residence time since many processes and reactions are time dependent and these have been used to supplement information obtained from the quantitative radiometric tools.32,33 Chemical tracers are best investigated in downgradient proﬁles in conﬁned groundwaters where sequential hydrogeochemical changes (including changes in cation ratios, salinity (Cl) changes, build-up of trace elements) may occur along ﬂow lines. Records obtained may represent timescales extending over many millennia (103–105 yr) and give an indication of former climatic or environmental conditions as well as extending age ranges, e.g. above the radiocarbon dating limit.32 Information of relative age may also be obtained from depth proﬁles, distinguishing modern water (recognisable by various contaminants) from pristine sources at depth. Glynn and Plummer34 provide a comprehensive review of the present knowledge about subsurface geochemistry and the understanding of groundwater systems. Freshwater diagenesis by groundwater moving into sediments formerly occupied by saline waters of marine or continental origins produces sequential changes in the groundwater with distance and depth. These reactions are of two types: gradual dilution of the marine pore water of clay-rich sediments with associated changes in groundwater cation ratios and incongruent reactions between groundwater and carbonate minerals also leading to distinctive cation ratios. In the ﬁrst case, often termed freshening aquifers,35–37 two end members are concerned: a saline NaCl-type marine water initially ﬁlling the aquifer, and a fresh CaHCO3-type recharge water. Cation exchange processes take place after the marine pore water has been replaced by fresh recharge water, causing

224

Chapter 5.3

a disequilibrium between the new pore water and the marine cations that are still adsorbed on the exchange sites of (mainly) the clay minerals. The marine cations Na1, K1 and Mg21 are desorbed and exchanged for Ca21 in the new pore water according to the aﬃnity sequence, leading sequentially to an increase in Na1, followed by a K1 peak, and ﬁnally a Mg21 peak, in the upﬂow direction. Marine carbonates (low- or high-Mg calcites as well as aragonite) contain Mg, Sr and other impurities which help stabilise the mineral lattices under the biogenic conditions of formation. Carbonates under subaerial conditions then form aquifers and undergo freshwater diagenesis, the marine carbonate minerals releasing their impurities to groundwaters, leaving a purer calcite in the process.38 This process of incongruent dissolution is time dependent and Mg and Sr increases (up to limits of dolomite or celestite solubility) can be used to distinguish younger waters from older more evolved waters. This is illustrated for the chalk aquifer of Wessex in southern England in Figure 5.3.4. In the freshwater aquifer, where it can be demonstrated that inert tracers are derived predominantly from atmospheric inputs, it may be possible to infer

Figure 5.3.4

Trilinear diagram showing the evolution of groundwater in the chalk of Wessex, UK. Initial reaction of low-Mg calcite in the unsaturated zone and shallow aquifer produces low Mg/Ca ratios, but increasing Mg/Ca ratios (up to the limiting 1:1 ratio of dolomite saturation) indicate increasing residence time as the marine calcite undergoes recrystallisation. Similar trajectories are shown for the Berkshire area (London Basin) and in both areas the inﬂuence of residual connate water is shown by Na+K increase.

225

Groundwater Age and Quality

°

Figure 5.3.5

°

Lithium, strontium and 13C as indicators of groundwater residence time in the East Midlands aquifer, UK. Groundwater temperature has been used as a proxy for distance downgradient.

palaeoclimatic information and, therefore, independently, another indicator of age. This approach has been used in the East Midlands aquifer,32,39 where the Cl concentrations are atmospherically derived and the Br/Cl and the 36Cl/Cl atomic ratios may provide information on past recharge rates and the changes in source areas of precipitation. In turn the chemical data may be used to ﬁngerprint modern and older, pristine water as in the case study below. In a further example from the East Midlands aquifer, reactive tracers such as 13 C, Li and Sr may be used as proxies for residence time.32 In Figure 5.3.5 the groundwater radiocarbon ages are indicated and show a near linear increase with distance from outcrop (here represented by temperature as proxy). There is a gap in ages from approx. 10–20 kyr which is thought to represent an absence of recharge during the LGM (Last Glacial Maximum). Strontium increases and the enrichment in carbon-13 are the result of two processes: the incongruent dissolution of carbonate minerals and the dissolution of trace gypsum with Sr as impurity. Lithium shows a strong linearity and is derived from slow release from silicate minerals (probably K-feldspar) in the aquifer framework. A lithium timescale has been suggested for this aquifer.32 Lithium is unrelated to modern contamination and at outcrop the lithium in young groundwaters is below 10 mg l 1 but concentrations in excess of 20 mg l 1 indicate residence times in excess of 20 kyr.

5.3.2.4

Numerical Modelling of Groundwater Age

Groundwater ﬂow models and environmental tracers have been used to assess travel times and groundwater ages for some decades. The studies have

226

Chapter 5.3

investigated both very old groundwaters in large sedimentary basins40 and young groundwater in shallow aquifers.41,42 Presently the studies are increasingly focusing on young contaminated groundwaters, and the interaction with surface waters (Chapters 9.3 and 10.2),10,42,43 due to increasing environmental impact and the development of more advanced numerical modelling and environmental tracer tools for dating young groundwater. Groundwater ages can be computed by groundwater ﬂow models or integrated hydrological models by various techniques.11,44–46 Many studies have demonstrated that the computation and comparison of groundwater ages estimated by groundwater ﬂow models and environmental tracers are not trivial.21,46–49 Modelling studies have, for example, illustrated that tracer-derived groundwater ages may both underestimate or overestimate the actual average age of groundwater in settings where young waters with human impact are mixed with old waters without human impact20,21,45,50 or where low permeability aquitards or units occur along the ﬂow paths,20,51–53 respectively. While this is a drawback to the application of tracer ages as a calibration tool for groundwater models, unless dispersive processes and mixing in the well can be accounted for in the groundwater ﬂow and transport model, underestimation of groundwater age may actually be considered a beneﬁt when evaluating groundwater vulnerability at water supply wells,6 since soluble pollutants generally behave similar to the environmental tracers. Similarly there may be a considerable uncertainty on results from the numerical modelling as these reﬂect the uncertainty of the geological model and the physical parameters of the aquifers and aquitards in the subsurface and uncertainty in climatological parameters.45,54,55 However, it is beyond the scope of this chapter to discuss strengths and drawbacks for the diﬀerent methods. For the purpose of this chapter it suﬃces to say that both numerical hydrological models and environmental tracers are very strong tools in groundwater research and management, and that the best and most detailed information of the investigated system is obtained by the combined use of both methods. Generally, extensive geological knowledge and quantiﬁcation of both geochemical and physical subsurface processes are needed to obtain sound estimates of groundwater ages.34

5.3.3

Groundwater Age and Water Quality and Quantity Issues

5.3.3.1

Groundwater Quality as a Function of Age

As mentioned earlier, groundwater quality changes with time due to both changes in anthropogenic activities and water/rock interaction in the surface. Generally the risk of pollution decreases with increasing groundwater age, while the risk of saltwater intrusion and increase in dissolved minerals, trace metals and radionuclides increases. No general rules can be given for the advance of modern contaminated groundwater in European aquifers as it varies considerably with the geological and climatological setting.2 Figure 5.3.6 illustrates the temporal evolution of nitrate concentrations in groundwater in

227

Groundwater Age and Quality 50

30 40

NL g 30

20

20 USf

N in groundwater (mg/L)

N in fertilizer (g/m2/y)

NL f

DK f

10

10 DKg US g 0

0 1940

Figure 5.3.6

1950

1960

1970 1980 Recharge date

1990

2000

Average (smoothed) nitrate concentrations in groundwater in Denmark (for all samples in monitoring database with O2 4 1 mg l 1),68 the Netherlands (Noord-Brabant)12,57 and the USA (watershed in Maryland)7 as a function of recharge year estimated by CFCs (DK and US) and 3H/3He (NL). The NO3-N concentration evolution in groundwater is compared to the fertiliser consumption for the relevant regions in the same period (subscripts g and f indicate groundwater and fertilizer curves, respectively).

Denmark, the Netherlands and the USA compared to the amount of applied fertilisers in the investigated regions. Environmental tracers have been applied widely in evaluation of the transport of nitrate in the subsurface and to dependent ecosystems.7,10,16,56–58

5.3.3.2

Groundwater Age and Monitoring

The location of groundwater monitoring points in relation to the spatial distribution of groundwater age is of major importance. Ideally groundwater monitoring programmes should include also analyses of environmental tracers such as tritium or CFC gases as these provide valuable information on the vulnerability of the monitoring wells and the temporal evolution of the groundwater around them. An example is shown in Figure 5.3.6 were the nitrate contents of oxic groundwater in Denmark and the USA are shown as a function of groundwater ages estimated by CFC gases compared to nitrate contents in groundwater in the Netherlands dated by the 3H/3He method.

228

5.3.3.3

Chapter 5.3

Groundwater Age and Water (Over)exploitation

Generally as the groundwater age and depth increases the risk of overexploitation and mining also increases as the recharge of the groundwater resource may be slow. However, if too high abstraction rates lead to severely lowered water tables and strongly increased hydraulic gradients, the risk of both saline water intrusion and pollution from the surface increases. Furthermore, if groundwater tables are lowered below reduced sediments or minerals this will induce natural geochemical processes, which may result in a groundwater quality breaching quality standards for several elements or substances. Lowered water tables will result in increased oxidation of reduced organic carbon and minerals potentially present in the sediment, and this may lead to increased concentrations of dissolved carbon, bicarbonate, sulfate, arsenic, nickel, etc.,59 Furthermore the high abstraction may aﬀect the quality of the dependent surface waters primarily due to reduced discharge (Chapter 9.3).

5.3.4

Groundwater Age and the Water Framework and Groundwater Directives

The spatial distribution of groundwater ages in the subsurface is an import indicator of vulnerable and less vulnerable parts of the groundwater resources.41,60 and the potential impact from the diﬀerent parts on dependent aquatic and terrestrial ecosystems.9–11 Hence it is an important parameter for the temporal and spatial evolution of the status of both groundwater itself and the dependent ecosystems.

5.3.4.1

Groundwater Age and Derivation of Natural Background Levels and Threshold Values

It is stipulated in the Water Framework Directive and the daughter Groundwater Directive (see Chapter 3.1) that the water bodies in Europe have to reach good status in 2015. The assessment of the qualitative status for groundwater depends on the natural background of the groundwater quality and the relevant environmental quality standards for groundwater itself and for the dependent ecosystems. The environmental tracers and groundwater dating are the most obvious and important tools for identifying natural background quality groundwater without human impact (see Chapter 5.2).61

5.3.4.2

Groundwater Interaction with Dependent Ecosystems

Shallow young groundwaters aﬀect the dependent ecosystems more than deep groundwaters, quantitatively and qualitatively (Chapters 9.3 and 10.2).10 Hence a sound understanding of the relation between the spatial hydrochemical evolution of groundwater quality and the groundwater age distribution provides important information on the possible impact of groundwater on the

229

Groundwater Age and Quality

dependent ecosystems.7,9 Groundwater/surface water interaction may occur in both directions and may change downstream as some parts of the streams may be gaining, while others may be loosing. Hence a bad status groundwater may aﬀect streams negatively in some parts of a stream, while in other parts the stream water quality may lead to deterioration of the groundwater quality (see Swiss case studies in Section 5.2).

5.3.5

Case Studies

Environmental tracers and groundwater ﬂow modelling are applied for evaluation of groundwater ages in an increasing number of studies as important tools for evaluation of the hydrochemical evolution in both deep and shallow groundwater systems. Both tracers and groundwater ﬂow modelling are generally necessary tools for unravelling groundwater ﬂow and hydrochemical evolution in the usually complex geological structures of the subsurface. Below we brieﬂy describe selected case studies from Denmark, Switzerland, Germany and the UK, in which environmental tracers and groundwater modelling have been cardinal tools.

5.3.5.1

Examples of Danish Case Studies

5.3.5.1.1

Nitrate Reduction in a Pyritic Sandy Aquifer at Rabis Creek

In a classic study, Postma et al. investigate transport and degradation of nitrate in a pyritic sand aquifer.19,56 The study identiﬁes the tritium bomb peak and uses this together with earlier evaluations of the migration of bomb tritium in unsatured and saturated zones at a location close to the Rabis Creek site62 to estimate vertical groundwater ﬂow velocity across the redox boundary, and the progression of the redox front (Figure 5.3.7). Postma et al. demonstrate that the redox front progression is accelerated by a factor of 5 by nitrate pollution to about 2 cm a year as nitrate is reduced and use up pyrite in the sediment.56 Later studies at the Rabis Creek test site investigate the subsurface geology63 Agricultural areas

Forest and Heath T2 T3 T4

T1

60 100 50

T5

T6

T10

100 50

Elevation, m

Water Table 40

? 20

0

Figure 5.3.7

NO3 > 0.1 mM

O3 < 0.05 mM

0.5 km

The advance of the nitrate plume at the Rabis Creek test site. Modiﬁed after Postma et al.56 The curves to the left of the multisamplers T1 and T2 indicate the location of the 3H bomb peak (Figure 5.3.1) in the Rabis Creek aquifer in 1988.

230

Chapter 5.3

and model nitrate removal,64 transport of tritium,65 groundwater ages66 and the transport and degradation of CFCs in the saturated25 and unsaturated zone.67 The site is included in the National Groundwater Monitoring Programme and more than 100 monitoring points at the site are sampled and analysed once a year.68

5.3.5.1.2

Natural Background Levels and Hydrochemical Evolution

The hydrochemistry of the deep-lying Ribe Formation aquifer in SW-Jylland, Denmark, was investigated in a European research project ‘‘PALAEAUX’’ on the hydrochemical evolution of European aquifers since the end of the last ice age.51 In this study it was mainly the radioactive isotopes 3H and 14C that were used and proved to be of great value in identifying natural background composition or modern water impacts. After correction for the eﬀects of geochemical and physical (diﬀusion) processes in the subsurface, the groundwater ages obtained by 14C dating and groundwater ﬂow models compared quite well and showed that the groundwater in the Ribe Formation is a few thousand years old, totally pristine and has a very high quality. The study showed that the Ribe Formation, which is mainly a freshwater sediment deposited in the Miocene period about 20 million years ago, was ﬁrst salinised during later sea-level high stands and then again (re)freshened during the Pleistocene and Holocene geological periods. The Ribe Formation was studied further in another European Union research project on the estimation of natural baseline (background) levels of hydrochemical parameters in European aquifers.69 The study showed that the aquifer is an excellent example of a groundwater body with a pristine natural background quality that also to some extent can be used to approximate the natural background quality in anoxic parts of carbonaceous Pleistocene sands above.

5.3.5.1.3

Vulnerability of Groundwaters in Buried Quaternary Valleys

Deep buried valleys may contain pristine groundwater of very high quality, but may also create shortcuts or pollution ‘‘highways’’ between the surface and adjacent deep aquifers like the Miocene Ribe Formation described above. Very diﬀerent scenarios may exist depending on the valley inﬁll, surrounding aquifers and aquitards, and regional climate and exploitation. The environmental tracers are excellent tools for evaluating the vulnerability of the valley aquifer and adjacent aquifers. Selected environmental tracers were applied in a regional European research project on mapping of water resources in buried valleys for evaluating recharge, groundwater age and vulnerability in selected Danish and German Quaternary buried valleys.70 Results demonstrate the value of the environmental tracers in the understanding of groundwater ﬂow systems and in the identiﬁcation of leakage in some monitoring wells.

231

Groundwater Age and Quality

5.3.5.1.4

Pollutant Breakthrough and Tracer Transport Modelling in the Odense River Basin

In the ﬁnal Danish example multiple environmental tracers (3H/3He, CFCs, 85 Kr and SF6) were used to constrain and partly calibrate an integrated hydrological model including groundwater/surface water interaction, set up to estimate the breakthrough and future evolution of the pollutant 2,6dichlorobenzamide (BAM) concentrations at water supply wells of the Odense Water Company.45,55,71,72 Figure 5.3.8 shows the simulated evolution of the pollutant concentrations at the well site for the next 50 years.

5.3.5.2

Examples of Swiss and German Case Studies

5.3.5.2.1

Quantiﬁcation of River Water Inﬁltration in a Shallow Aquifer in Switzerland

Pollution of surface water can cause degradation of groundwater quality and conversely pollution of groundwater can degrade surface water. Hence, the exchange rate of groundwater with surface water in particular with rivers determines the risk potential for cross contamination between these two systems which have to be regarded as a single resource. In this example the environmental tracers 3H and 85Kr were used in order to estimate the mixing portion of river water in a shallow groundwater body.73 In the area of investigation the river water is enriched in 3H due to emissions from the local BAM root zone (µg/l) 1.50

BAM well field (µg/l) 0.50

0.45 1.25

0.40 0.35

1.00

0.30 0.75

0.25 0.20

0.50

0.15 0.10

0.25

0.05 0.00 0.00 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060

BAM concentration below root zone (input) BAM concentration at well field (fully mixed)

Figure 5.3.8

Simulated future evolution of the pollutant 2,6-dichlorobenzamide (BAM) at a well ﬁeld.72

232

Chapter 5.3

watch industry compared to 3H in recharge from precipitation. 85Kr on the other hand is exclusively of atmospheric origin. The elevated 3H values found in two groundwater samples (Figure 5.3.9) clearly indicate the presence and the mixing portion of river water in the groundwater. In contrast, 85Kr is a sensitive indicator for the mean residence time of the groundwater.

5.3.5.2.2

Constraining the Age Distribution of a Mineral Water with Multi-Tracer and Time Series Tracer Measurements

The age structure of groundwater is a key parameter for the sustainable exploitation of a resource in terms of water quality and quantity. The accuracy of the dating depends not only on the intrinsic properties of the applied tracer methods (Figure 5.3.3(a)) but also on the understanding of the subsurface ﬂow pattern and the shape of the transit time distribution.74 This information has to be based on the hydrogeological environment but can also be constrained by the tracer data. A sample from a mineral water production well revealed 3H and 85 Kr values of 22.1 TU and 22.3 dpm cm 3 Kr respectively.75,76 The relation between the 3H/85Kr ratio and the mean groundwater residence time depends on the assumed age distribution (Figure 5.3.10). The diﬀerence between a piston ﬂow ( pfm: no mixing) and the exponential model (EM: strong mixing),

Figure 5.3.9

Estimation of river water portion and residence time of a shallow aquifer in eastern Switzerland using 3H and 85Kr data. The proportion of river water is about 80% and 30% in samples (a) and (b) respectively. The residence time of groundwater is around 2 years in both cases.

233

Groundwater Age and Quality 140

0.2 a)

PF (Tm = 12 years)

100

EM

0.0

Tritium [TU]

log( 3 H/ 85 Kr)

0.1

b)

120

PM

measured (1996)

-0.1 -0.2 t

-0.3

80 60 40 EM (T =15 years) m 20

-0.4 0 -0.5 0

5

10

15

20

25

mean residence time

Figure 5.3.10

1975 1980 1985 1990 1995 2000

sampling year

3

H and 85Kr concentrations of a mineral water. (a) Tracer ratio H/85Kr as function of mean residence time plotted for two assumed age distributions (PM: piston ﬂow; EM: exponential model); (b) 3H time series. 3

which can be regarded as two extreme scenarios, increases signiﬁcantly for residence times over 10 years. This ambiguity can be reduced if the 3H time series is taken into account (Figure 5.3.10(b)), which evidently favours the EM age distribution before the PM. Both the 85Kr and 3H tracer ratio and the 3H time series point to an EM age distribution with a mean residence time of 15 years for this sample.

5.3.5.2.3

Inter-Aquifer Leakage

Shallow unconﬁned Quaternary basins provide the main groundwater resource in many areas in Europe. However, as in southwestern Germany, these basins may be hydraulically interconnected with deeper groundwater systems, e.g. in the Malm-Karst.77 The direction of the hydraulic gradient between these two resources determines whether the shallow systems, which are vulnerable to pollution, are ‘‘fed’’ by older high-quality water from the deep aquifer or whether the deep aquifer is under risk to become polluted from the shallow system. The deep water is often older than 50 years,78 and is therefore free of young residence time indicators (3H, SF6, 85Kr, etc.). Hence, the admixture of deep and old water is manifested by a dilution of the tracer concentrations (Figure 5.3.11). Three groups of waters can be distinguished in Figure 5.3.11. Samples plotting near the suspected model curve (dotted line) represent waters from local recharge. The residence times of these waters range up to 20 years. A group of samples, mainly in the age range 20–25 years and originating essentially from greater depths of the Quaternary basins, become increasingly inﬂuenced by the deep water (group 2). The third group of samples is dominated by old water from the deep karst aquifer.

234

Chapter 5.3 45.0

40a

40.0 35.0

Tritium [TU]

30.0

30a 51a

young water age 25a

25.0 20.0

20a 10a

15a

15.0

5a

10.0 80%

60%

5.0

40%

young water portion

20%

0.0 0.0

10.0

20.0

30.0

40.0

50.0

60.0

85Kr[dpm/ccKr]

Figure 5.3.11

5.3.5.3 5.3.5.3.1

Age dating of young water components and estimation of the portion of old water from deeper underlying aquifers using a two-tracer plot.

Example of a British Case Study Chloride as Indicator of Modern and Pristine (Older) Water: East Midlands Aquifer

Chemical tracers can provide diagnostic information of age in many types of aquifer as outlined above: either alone or in conjunction with some other indicator. The East Midlands Triassic sandstone aquifer is well studied and presents a clear example of how tracers32,33,79 behave in groundwater which has no residual salinity over most of its developed section. Chloride acts as an inert tracer and its inputs relate to climatic or anthropogenic factors and geogenic inﬂuences are negligible.39 The timescale for groundwaters in the aquifer, obtained from corrected 14C data,80 extends beyond the radiocarbon dating range and several other absolute indicators have now been applied to this aquifer which demonstrate the age proﬁle downgradient, albeit with a degree of mixing. In the example shown (Figure 5.3.12), where groundwater temperature is used as a proxy for distance from outcrop, the timescale is represented qualitatively by d18O. There is a clear separation (about 1.4%) between lighter isotopic signatures representing recharge during the late Pleistocene and recharge from the modern era; the scatter can be related to several factors such as abstraction rates and gives an idea of the amount of mixing in this aquifer which also shows strong vertical age stratiﬁcation. The extent of invasion by modern groundwater is shown clearly by high Cl (above 20 mg l 1) derived from modern aerosol input with Cl enhanced by industrial emissions, but also locally by agrochemicals and from industrial sources. This ‘‘front’’ is also shown by other chemical indicators such as nitrate.33 The pristine waters are

Groundwater Age and Quality

Figure 5.3.12

235

Modern water with high Cl entering the East Midlands (UK) aquifer in contrast to the low Cl water of Holocene and late Pleistocene age.

remarkably low in Cl and represent the rainfall signature of the early Holocene and Pleistocene; the absence of an increase in Cl with depth over this section contrasts with many aquifers, where residual salinity is still found. In this aquifer salinity does increase in the very deepest groundwater, however, related to formation waters associated with continental evaporites. The main conclusion is that if the aquifer is well understood using a multi-tracer approach, then Cl alone may be used for management purposes to monitor the extent of overprinting of the pollution front.

5.3.6

Conclusions

A sound understanding of groundwater ages and travel times is a prerequisite for the evaluation of the history and fate of contaminants in the subsurface, and hence for evaluating and protecting the quality and quantity of both groundwater itself and its dependent ecosystems. Environmental tracers as well as groundwater or integrated hydrological models are the important and only tools for evaluating travel times in the subsurface, and these should be

236

Chapter 5.3

combined whenever possible to get the best description of groundwater ages and residence times in the hydrological system. There is a relatively large number of environmental tracers that can be applied for evaluating travel times and groundwater ages, as described in this chapter; which ones to use will depend on the hydrogeological setting, etc. Generally, it is recommended to use a combination of multiple tracers as this, for example, provides a possibility of evaluating the relative contribution of young and old waters in mixed water samples. Groundwater age or residence times from properly designed monitoring networks provide valuable information to all assessments of the evolution of groundwater quantity and quality and its eﬀects on dependent ecosystems. Hence, they are important tools for developing a sustainable management policy for the protection of water resources and the aquatic environment.

References 1. W. M. Edmunds and P. Shand, in The Natural Baseline Quality of Groundwater, Blackwell, 1st edn, 2007, in press. 2. K. Hinsby, W. M. Edmunds, H. H. Loosli, M. Manzano, T. Melo and F. Barbecot, Geological Society Special Publications, 2001, 189, 271. 3. W. Alley, in Regional Ground-Water Quality, Van Nostrand Reinhold, 1st edn, 1993. 4. J. N. Galloway, F. J. Dentener, D. G. Capone, E. W. Boyer, R. W. Howarth, S. P. Seitzinger, G. P. Asner, C. C. Cleveland, P. A. Green, E. A. Holland, D. M. Karl, A. F. Michaels, J. H. Porter, A. R. Townsend and C. J. Vorosmarty, Biogeochemistry, 2004, 70, 153. 5. A. L. Heathwaite, G. Billen, C. Gibson, C. Neal, P. Withers and L. Bolton, J. Hydrology, 2005, 304, 1. 6. A. H. Manning, D. K. Solomon and S. A. Thiros, Ground Water, 2005, 43, 353. 7. J. K. Bohlke and J. M. Denver, Water Resour. Res., 1995, 31, 2319. 8. Mattle, W. Kinzelbach, U. Beyerle, P. Huggenberger and H. H. Loosli, J. Hydrology, 2001, 242, 183. 9. P. B. McMahon and J. K. Bohlke, J. Hydrology, 1996, 186, 105. 10. E. Modica, H. T. Burton and L. N. Plummer, Water Resour. Res., 1998, 34, 2797. 11. H. P. Broers, J. Hydrology, 2004, 299, 84. 12. H. P. Broers and B. van der Grift, J. Hydrology, 2004, 296, 192. 13. T. H. Christensen, P. Kjeldsen, P. L. Bjerg, D. L. Jensen, J. B. Christensen, A. Baun, H. J. Albrechtsen and C. Heron, Appl. Geochem., 16, 2001, 659. 14. T. H. Christensen, P. L. Bjerg and P. Kjeldsen, Ground Water Monitor. Remed., 2000, 20, 69. 15. J. K. Bohlke and Denver, Water Resour. Res., 1995, 31, 2319. 16. A. J. Tesoriero, T. B. Spruill, H. E. Mew, K. M. Farrell and S. L. Harden, Water Resour. Res., 2005, 41, 2005.

Groundwater Age and Quality

237

17. R. Kunkel and F. Wendland, J. Hydrology, 2002, 259, 152. 18. F. Wendland, H. Bogena, H. Goemann, J. F. Hake, P. Kreins and R. Kunkel, Phys. Chem. Earth, 2005, 30, 527. 19. C. A. J. Appelo and D. Postma, in Geochemistry, Groundwater and Pollution, A. A. Balkema, Leiden, 2nd edn, 2005, p. 1. 20. P. Maloszewski, W. Stichler and A. Zuber, Isotopes Environ. Health Stud., 2004, 40, 21. 21. E. M. LaBolle, G. E. Fogg and J. B. Eweis, Water Resour. Res., 2006, 42(7). 22. P. G. Cook and A. L. Herczeg, in Environmental Tracers in Subsurface Hydrology, Kluwer Academic, Boston, MA, 1st edn, 2000, p. 1. 23. M. Maiss and I. Levin, Geophys. Res. Lett., 1994, 21, 569. 24. E. Mazor, Geochim. Cosmochim. Acta, 1972, 36, 1321. 25. K. Hinsby, A. L. Højberg, P. Engesgaard, K. H. Jensen, F. Larsen, L. N. Plummer and E. Busenberg, Water Resour. Res., 2007, in press. 26. IAEA, in Use of Chloroﬂuorocarbons in Hydrology: A Guidebook, International Atomic Energy Agency, Vienna, 1st edn, 2006, p. 1. 27. L. N. Plummer and E. Busenberg, in Environmental Tracers in Subsurface Hydrology, ed., P. G. Cook and A. L. Herczeg, Kluver Academic, Boston, MA, 2000, pp. 441–478. 28. U. Beyerle, W. Eschbach-Hertig, D. M. Imboden, H. Baur, T. Graf and R. Kipfer, Environ. Sci. Technol., 2000, 34, 2042. 29. D. K. Solomon, S. L. Schiﬀ, R. J. Poreda and W. B. Clarke, Water Resour. Res., 1993, 29, 2951. 30. A. R. Berger and W. J. Iams, in Geoindicators: Assessing Rapid Environmental Change in Earth Systems, A. A. Balkema, Rotterdam, 1st edn, 1996. 31. W. M. Edmunds, in Geoindicators: Assessing Rapid Environmental Change in Earth Systems, ed., A. R. Berger and W. J. Iams, A. A. Balkema, Rotterdam, 1st edn, 1996, pp. 135–150. 32. W. M. Edmunds and P. L. Smedley, Appl. Geochem., 2000, 15, 737. 33. P. L. Smedley and W. M. Edmunds, Ground Water, 2002, 40, 44. 34. P. D. Glynn and L. N. Plummer, Hydrogeol. J., 2005, 13, 263. 35. C. A. J. Appelo, Water Resour. Res., 1994, 30, 2793. 36. M. Coetsiers and K. Walraevens, Hydrogeol. J., 2006, 14, 1556. 37. K. Walraevens, M. Van Kamp, J. Lermytte, W. J. M. Van der Kemp and H. J. Loosli, Geological Society Spec. Publ., 2001, 189, 49. 38. W. M. Edmunds, J. M. Cook, W. G. Darling, D. G. Kinniburgh, D. L. Miles, A. H. Bath, M. Morgan-Jones and J. N. Andrews, Appl. Geochem., 1987, 2, 251. 39. J. N. Andrews, W. M. Edmunds, P. L. Smedley, J. C. Fontes, L. K. Fiﬁeld and G. L. Allan, Earth Planet. Sci. Lett., 1994, 122, 159. 40. M. C. Castro, P. Goblet, E. Ledoux, S. Violette and G. de Marsily, Water Resour. Res., 1998, 34, 2467. 41. A. Zuber, S. Witczak, K. Rozanski, I. Sliwka, M. Opoka, P. Mochalski, T. Kuc, J. Karlikowska, J. Kania, M. Jackowicz-Korczynski and M. Dulinski, Hydrolog. Processes, 2005, 19, 2247.

238

Chapter 5.3

42. P. G. Cook, S. Lamontagne, D. Berhane and J. F. Clark, Water Resour. Res., 2006, 42, 2006. 43. D. C. Gooddy, W. G. Darling, C. Abesser and D. J. Lapworth, J. Hydrology, 2006, 330, 44. 44. D. J. Goode, Water Resour. Res., 1996, 32, 289. 45. L. Troldborg, Technical University of Denmark/Geological Survey of Denmark and Greenland, 2004. 46. M. Varni and J. Carrera, Water Resour. Res., 1998, 34, 3271. 47. P. Maloszewski, W. Stichler and A. Zuber, Isotopes Environ. Health Stud., 2004, 40, 21. 48. P. Maloszewski and A. Zuber, Water Resour. Res., 1991, 27, 1937. 49. P. Maloszewski and A. Zuber, J. Hydrology, 1982, 57, 207. 50. C. M. Bethke and T. M. Johnson, Geology, 2002, 30, 107. 51. K. Hinsby, W. G. Harrar, P. Nyegaard, P. Konradi, E. S. Rasmussen, T. Bidstrup, U. Gregersen and E. Boaretto, Geological Society, Spec. Publ., 2001, 189, 29. 52. E. M. LaBolle, G. E. Fogg and J. B. Eweis, Water Resour. Res., 2007, 42, 2007. 53. W. E. Sanford, Ground Water, 1997, 35, 357. 54. J. C. Refsgaard, P. van der Keur, B. Nilsson, D. I. Mu¨ller-Wohlfeil and J. Brown, J. Hydrol. Earth Syst. Sci., in press. 55. L. Troldborg, J. C. Refsgaard, K. H. Jensen and P. Engesgaard, Hydrogeol. J., 2007, doi: 10.1007/j.envpol.2007.01.027. 56. D. Postma, C. Boesen, H. Kristiansen and F. Larsen, Water Resour. Res., 1991, 27, 2027. 57. A. Visser, H. P. Broers, B. van der Grift and M. F. P. Bierkens, Environ. Pollut., 2007, doi: 10.1016/j.envpol.2007.01.027. 58. K. Zoellmann, W. Kinzelbach and C. Fulda, J. Hydrology, 2001, 240, 187. 59. F. Larsen and D. Postma, Environ. Sci. Technol., 1997, 31, 2589. 60. A. Zuber, S. M. Weise, J. Motyka, K. Osenbruck and K. Rozanski, J. Hydrology, 2004, 286, 87. 61. W. M. Edmunds, P. Shand, P. Hart and R. S. Ward, Sci. Total Environ., 2003, 310, 25. 62. L. J. Andersen and T. Sevel, in Isotope Techniques in Groundwater Hydrology, IAEA, Vienna, 1974. 63. H. Olsen, C. Ploug, U. Nielsen and K. Sorensen, Ground Water, 1993, 31, 84. 64. P. Engesgaard and K. L. Kipp, Water Resour. Res., 1992, 28, 2829. 65. P. Engesgaard, K. H. Jensen, J. Molson, E. O. Frind and H. Olsen, Water Resour. Res., 1996, 32, 3253. 66. P. Engesgaard and J. Molson, Ground Water, 1998, 36, 577. 67. P. Engesgaard, A. L. Hojberg, K. Hinsby, K. H. Jensen, T. Laier, F. Larsen, E. Busenberg and L. N. Plummer, Vadose Zone J., 2004, 3, 1249. 68. GEUS, Groundwater Monitoring 2005, Geological Survey of Denmark and Greenland (GEUS), 2005 (www.geus.dk). 69. K. Hinsby and E. S. Rasmussen, in The Natural Baseline Quality of Groundwater, ed. W. M. Edmunds and P. Shand, Blackwell, 1st edn, 2007.

Groundwater Age and Quality

239

70. K. Hinsby, in Groundwater Resources in Buried Valleys: A Challenge for Geosciences, ed. BurVal Working Group, Leibniz Institute for Applied Geosciences (GGA-Institute), Hannover, 2006, pp. 141–148. 71. J. A. Corcho Alvarado, R. Purtschert, K. Hinsby, L. Troldborg, M. Hofer, R. Kipfer, W. Aeschbach-Hertig and H.-A. Synal, Appl. Geochem., 2005, 20, 599. 72. K. Hinsby, L. Troldborg, R. Purtschert and J. A. Corcho Alvarado, in Isotopic Assessment of Long Term Groundwater Exploitation, IAEA-TECDOC-CD-1507, 2006, 73–95. 73. K. Osenbruck, Pilot-Isotopen-Studie, Amt fu¨r Umwelt des Kanton Thurgau, 2001. 74. D. W. Waugh, T. M. Hall and T. W. N. Haine, J. Geophys. Res. Oceans, 2003, 108, 2003. 75. P. Hartmann, ETH Zu¨rich, 1998. 76. R. Purtschert, Universita¨t Bern, 1997. 77. B. W. Bertleﬀ, Abh. Geol. Landesamt Baden-Wu¨rttemberg, 1986, 12. 78. M. Heidinger, Grundwasser Bewirtschaftungskonzept Singen, Stadtwerke Singen, Singen, 1996. 79. W. M. Edmunds, A. H. Bath and D. L. Miles, Geochim. Cosmochim. Acta, 1982, 46, 2069. 80. A. H. Bath, W. M. Edmunds and J. N. Andrews, in International Symposium on Isotope Hydrology, IAEA-SM-228/27, Vienna, 1979, vol. II, pp. 545–568.

CHAPTER 5.4

Characterisation of Groundwater Contamination and Natural Attenuation Potential at Multiple Scales THOMAS PTAKa AND JERKER JARSJO¨b a

University of Go¨ttingen, Geosciences Center, Goldschmidtstrasse 3 DE-37077 Go¨ttingen, Germany; b Stockholm University, Department of Physical Geography and Quaternary Geology, SE-106 91 Stockholm, Sweden

5.4.1

Introduction

Groundwater pollution is an important problem at many locations all over Europe. Sources for contaminants in aquifers such as chlorinated compounds, petroleum hydrocarbons, etc., are, for example, leaking underground storage tanks and pipelines, petrol stations, gasworks sites and all types of industries. At these locations, rapid industrial development, missing regulations and/or safety measures, changes in land use and ownership as well as the hydraulic and hydrogeochemical aquifer heterogeneity cause complex and irregular contamination patterns with very often unknown locations of pollutant hot spots. It is generally accepted that present approaches for site investigation and assessment are either not reliable enough or not cost eﬀective, making the development of new approaches for the characterisation of groundwater contamination and natural attenuation potential at multiple scales necessary. The European Union (EU) FP 5 project INCORE (Integrated Concept for Groundwater Remediation, EVK1-1999-00080) and the German BMBFfunded project SAFIRA C2.1 (Erkundung der Schadstoﬀfracht in kontaminierten Aquiferen zur Dimensionierung von in-situ-Sanierungsreaktoren, BMBF 02WT9948/0) are aimed at the development and implementation of a new approaches and methods for contaminated land assessment and revitalisation in urban industrial areas, focusing on groundwater quality and complex 240

Characterisation of Groundwater Contamination

241

contamination patterns at megasites which are typical for many European cities. The principal idea is to start at large scale, e.g. at the scale of an entire industrial site, and, in a ﬁrst step, to assess groundwater contamination, using an innovative integral investigation method to estimate contaminant concentrations and mass ﬂow rates across control planes as well as the natural attenuation potential. Using backtracking methods, potential contamination source zones can then be delimited, considering parameter uncertainty. Finally, detailed high-resolution investigation methods such as direct push-based proﬁling or multilevel sampling can be applied for small-scale investigations at the identiﬁed hot spot zones. In this approach, a large potentially contaminated area is screened initially at a high level of certainty, but only a small area may be ﬁnally considered for further investigations and remediation measures. Consequently, this large- to small-scale screening procedure yields a signiﬁcant reduction of costs needed for land revitalisation. The new tools were developed, implemented and tested under real-world conditions, considering administrative aspects also. In this way, the project results provide an important basis for the development and implementation of EU directives on contaminated land assessment and revitalisation. This chapter summarises some of the new developed approaches and tools. The principle of the integral groundwater investigation method is described at ﬁrst. Then a general, integral investigation-based methodology for assessing the eﬀects of aquifer parameter uncertainty on the estimates of mass ﬂow rates and concentrations as well as on delimiting both contaminant source zones and zones absent of source is presented. It is also shown how the integral investigation method can be applied to estimate natural attenuation rates at ﬁeld scale. Finally, a multilevel version of the integral investigation approach for local-scale investigations is described. In addition, examples of application are given.

5.4.2

The Integral Groundwater Investigation Method

A reliable characterisation of contaminant plumes in groundwater and source zone locations is essential for decisions about future land use at contaminated sites, and for choosing appropriate remediation measures. The basic problem of the characterisation is that at many contaminated sites pollutant hot spots with positions not exactly known, and preferential transport paths and low conductivity zones within the aquifer cause an irregular distribution of contaminants in groundwater (Figure 5.4.1). In such situations, standard subsurface investigation procedures based on interpolation of point-scale concentration measurements are very likely to yield poor results. An eﬀective way of obtaining investigation results with a high level of certainty at large scale is to apply the integral groundwater investigation method,1–4 in which relatively large water volumes are sampled within so-called integral pumping tests (IPTs; Figure 5.4.1). Due to the large sampling volumes the small-scale concentration variance is averaged out, which may otherwise inﬂuence and bias point-scale concentration

242

Chapter 5.4 Plot of concentration vs. time during pumping tests (compound specific)

Pumping tests with concentration time series measurements Contaminated Mean groundwater flow direction site Isochrones Source Well 1 Control cross-section of Well 2

C

t1

pollutant Well 3

t2 t 1

Well 2

C

Well 3

t2 t1

t2

Contaminant plume

Contaminant mass fluxes and concentrations at control cross-section

Figure 5.4.1

C Well 1

Transient inversion algorithm based on a numerical flow and transport model of the field site

Concept of the integral investigation method for the quantiﬁcation of groundwater contamination.3

measurements, and there is no need to interpolate point-scale concentration measurements. Therefore a high level of certainty can be expected for the investigation results.

5.4.2.1

Concepts and Principles

The basic idea of the integral groundwater investigation method1–3 is to cover a whole cross-section of a contaminant plume downstream of a pollutant source, employing pumping tests with multiple contaminant concentration measurements at the pumping wells. Due to the spatial integration of a pumping test, and due to the increasing capture zone with pumping time, both the spatial distribution of the contaminants as well as the total mass ﬂow rate within a contaminant plume can be estimated. To apply the integral investigation method, one or more pumping wells are placed along a control plane (control cross-section) perpendicular to the groundwater ﬂow direction and operated simultaneously, or in subsequent pumping campaigns, downstream of a suspected pollutant source zone. The positions, pumping rates and pumping times are designed in a way so as to allow the well capture zones to cover the overall width of the potentially polluted area (Figure 5.4.1). Typically, the well is operated for a time period of some days, in order to obtain a large enough well capture zone. During pumping, as the capture zones increase, the concentration of groundwater contaminants and/or other groundwater quality parameter values is measured as a function of time at each of the pumping wells. The concentration time series yield information on the position and extent of the contaminant plume(s) as well as on the concentrations of the target substances in the plume(s). In Figure 5.4.2, four typical scenarios of concentration time series are shown as usually observed during IPTs, together with a possible interpretation of the subsurface contaminant plume size and location.4

243

1

Idealized contaminant plume maximum isochrone at end of pumping period pumping well

concentration

concentration

Characterisation of Groundwater Contamination

3

time

concentration

concentration

time

2 time

Figure 5.4.2

4

time

Typical scenarios of concentration time series recorded during an integral pumping test and possible interpretation with respect to position and size of the contaminant plume.4

The type 1 scenario represents a site where the pumping well is located outside of a relatively narrow contaminant plume. At the beginning of the pumping, the well capture zone initially covers the uncontaminated aquifer volume, and the concentrations in the well are below the detection limit. As the well capture zone (isochrone) grows into the area of the plume, the concentrations in the well discharge increase. At a later time, when the well capture zone has reached the outer fringe of the plume, the concentrations decrease again, due to dilution with clean water from outside the plume. Note that the measured concentration does not decrease to zero, as always the contaminant plume is within the isochrone and therefore some contaminated water is pumped at the well. It becomes clear from this interpretation, that the shape of the concentration time series is determined by the well location relative to the location of the plume and the plume width. The scenario of type 2 is observed if the pumping well is located within the contaminant plume, which itself is limited in width. In such a case, concentrations in the pumping well are high at the beginning of the pumping, but they decrease as the size of the capture zone and hence the dilution increases with time. This type can also be observed if the fringe of the plume is not sharp but given by decreasing concentrations. In case of the type 3 scenario, the pumping well is located outside of the plume, and the plume width reaches beyond the maximum isochrone extent. In this case, the width of the plume cannot be assessed. Finally, the type 4 scenario is characterised by a more or less constant concentration detected at the pumping well. In this case, the well is located within a wide plume with only slightly varying concentrations. Again, the plume width cannot be assessed. Only for a concentration time series of type 4, the value of the contaminant concentrations measured in the well actually represents the real concentration of the plume within the aquifer. The mass ﬂow rates and mean concentrations within the well capture zones as well as possible contaminant concentration distributions across the control plane can be determined by an inversion procedure.

244

Chapter 5.4

5.4.2.2

The Inversion Problem

5.4.2.2.1

Numerical Solution

The algorithm used for the numerical inversion of the measured concentration time series at the wells is implemented in the newly developed program code CSTREAM,5,6 an improved version of the original code by Schwarz,7 focusing on applications in highly non-uniform groundwater ﬂow systems due to aquifer heterogeneity (hydraulic conductivity distribution) and boundary conditions, and allowing investigations in a broad range of hydrogeological conditions. CSTREAM requires a transient ﬂow model of the ﬁeld site, which incorporates the pumping rates and pumping times at the abstraction wells, simulates all pumping tests at one control plane in one model run, and which provides the thickness, hydraulic conductivity and porosity of the aquifer as well as the local hydraulic gradient. Using CSTREAM, irregular well capture zones due to aquifer heterogeneity and the eﬀects of the local hydraulic gradient on the development of capture zones are accounted for. Mass ﬂow rates are thus estimated using local groundwater ﬂow terms at the wells. The program operation is brieﬂy described in the following using a two-dimensional, depth-integrated formulation.5 The setting is outlined in Figure 5.4.3, where the mean groundwater ﬂow under natural conditions is along the y-axis. A contaminant plume is assumed to pass a control plane (CP) located at y ¼ 0. The control plane has an area ACP [L2] deﬁned by the thickness of the water body and the maximum capture width LCP [L] during a pumping test performed at a pumping well with the position (0, 0). The distribution of contaminant concentration in the aquifer is denoted by C(x,y,t) [M L3]. C(x,y,0), i.e. the concentration distribution at time t ¼ 0 before start of the pumping, is assumed to be constant in ﬂow direction along the streamlines within the well capture zone. The well is operated for times t 4 0, and the contaminant concentrations at the pumping well are recorded as concentration y pumping well (0,0)

a)

QW(t) CW(t)

b)

b A CP LCP q

q

x

(p)

(n)

LI(t) plume

Figure 5.4.3

b(x,y)

Concept for the inversion of measured concentration time series. (a) The plume and the control plane are located at the well position perpendicular to the mean groundwater ﬂow direction. (b) Convergent ﬂow ﬁeld with isochrone LI(t) at time t (see text for an explanation of the symbols8).

245

Characterisation of Groundwater Contamination

time series CW(t). This measured concentration time series is used in order to estimate the total mass ﬂow rate. The concentration distribution along the CP before pumping, C(x,0,0), can be expressed as a function of the concentration time series CW(t) measured at the pumping well through an integral equation describing the mass balance within the volume of pumped water:5 CW ðtÞ ¼

1 QW

I

Cðx; y; 0Þj~ qðpÞ ðx; yÞjbðx; yÞ dl

ð1Þ

LI ðtÞ !

where jqðpÞ ðx; yÞj ½L T1 denotes the DarcyH velocity during the pumping test, ! b(x, y) [L] is the saturated thickness, QW ¼ LI ðtÞ jqðpÞ ðx; yÞjbðx; yÞ dl ½L3 T1 is the pumping rate at the well and LI(t) [L] is the isochrone corresponding to time t. Equation (1) is solved for each value of the measured concentration time series, taking into account the concentration distribution in the aquifer obtained by the previous CW(t) values. In this way a concentration distribution within the aquifer is obtained, which implicitly yields a concentration distribution across the control plane before start of pumping, C(x,0,0) [M L3]. Figure 5.4.4 summarises the basic principle of the numerical inversion solution of an IPT. The left side of Figure 5.4.4 shows a plan view of the isochrones and the streamlines of the undisturbed groundwater ﬂow ﬁeld corresponding to their maximum extent. The ﬁrst concentration measurement shown on the right side of Figure 5.4.4 corresponds to the ﬁrst, i.e. smallest, isochrone. The sample concentration is the mean (mixed) concentration along the isochrone. Using the assumption that the concentration is practically constant along a streamline at

concentration [c/cmax]

1.0

0.5

0 0

Figure 5.4.4

1

2 3 time [d]

4

Plan view of the isochrones, streamtubes of the undisturbed groundwater ﬂow ﬁeld and the pumping well (left) as well as measured concentrations at the well (right) for a demonstration run of an integral pumping test.9

246

Chapter 5.4

the scale of the isochrones, the area covered by the streamtube corresponding to the ﬁrst isochrone can be assigned the mean concentration of the ﬁrst sample (i.e. zero, in this example). The groundwater of the second sampling is representative of the second isochrone. The concentration of a part of the second isochrone, i.e. the part crossing the streamtube of the ﬁrst sample, is already known from the ﬁrst sample. The average concentration of the remaining parts of the second isochrone can then be calculated, yielding a mean concentration along the second isochrone corresponding to the second measured concentration. This procedure is then successively repeated for all measured concentrations. The resulting concentration distribution can then be integrated along the control plane in order to obtain the total mass ﬂow rate under steady state ﬂow conditions across the control plane, MCP [M T1]:5 Z MCP ¼ Cðx; 0; 0Þj~ qðnÞ ðx; 0Þjbðx; 0Þ dx ð2Þ lCP !

wherejqðnÞ ðx; yÞj ½L T1 is the Darcy velocity perpendicular to the CP under natural ﬂow conditions. The numerical program CSTREAM enables the estimation of the total mass ﬂow rate at a CP, MCP, for (known) heterogeneous conditions. The inversion of the measured concentration time series at the pumping well CW(0, 0, t), yielding the concentration distribution along the CP before pumping, is performed based on a transient and heterogeneous numerical ﬂow model of the ﬁeld site. MODFLOW is used here.10 The code MODPATH11 is used for the deﬁnition of the isochrones (transient backtracking during pumping) and of the streamtubes (steady-state forward tracking before pumping, i.e. at time t ¼ 0). Finally, the integral eqn. (1) is solved numerically within the code CSTREAM, and mass ﬂow rates are obtained applying eqn. (2). Veriﬁcation examples are provided by Bayer-Raich et al.5

5.4.2.2.2

Analytical Solution

If a groundwater ﬂow model of the investigated site is not available, for example in an initial stage of site investigation, and/or if no information on aquifer heterogeneity can be obtained, a simpliﬁed analytical solution for eqn. (1) and (2) can be applied for the inversion of the measured concentration time series. Under the assumption of homogeneous aquifer parameters around the pumping well and radially symmetrical ﬂow towards the pumping well, i.e. neglecting the inﬂuence of the natural groundwater ﬂow during the pumping period, the following analytical equation may be used for the estimation of mass ﬂow rates:7,12

MCP ¼ 2

n X i¼1

c^i Qi

with

1 p iP rk1 rk ci arccos c^k arccos 2 k¼1 r ri i c^i ¼ ri1 arccos ri

ð3Þ

Characterisation of Groundwater Contamination

247

where MCP [M T1] represents the mass ﬂow rate perpendicular to the control plane, ci [M L3] the concentration measured at the pumping well at time ti, i.e. ci ¼ c(ti), and c^i the average of the concentrations of the two streamtubes of the natural groundwater ﬂow ﬁeld positioned left and right from the pumping well at a distance r [L] (with ri1 o r o ri ). Qi ¼ kjrhjbðri1 ri Þ ½L3 T1 is the discharge under natural, i.e. undisturbed, conditions passing the control plane at both left and right streamtubes. The radius of the isochrone at time t is given by pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ri ¼ QW ti =pbf ð4Þ where QW [L3 T1] is the pumping rate at the well, b [L] the aquifer thickness and f the porosity. For the ﬁrst time step, c^1 ¼ c1 , and n is the total number of samples. This solution has been applied successfully at several sites (e.g. Refs. 4 and 13). Criteria for a site-speciﬁc decision on the application of the simpliﬁed analytical solution are discussed elsewhere.2 An expansion of the analytical inversion solution considering non-radial isochrones due to natural groundwater ﬂow is given in Ref. 12.

5.4.2.2.3

Non-uniqueness of Inversion

The inversion of the measured concentration time series is not unique, as many possibilities exist for the distribution of the contaminant in the aquifer to generate the measured concentration time series in the pumping well. Therefore, a symmetrical solution is usually used, where the contaminant mass is distributed equally on both sides of the pumping well, which yields a symmetrical concentration distribution in the aquifer along the control plane. If additional information on concentration in the aquifer is present, i.e. a pointscale measurement within the well capture zone, conditioning of the inversion algorithm is possible to include this a priori information. For aquifers with medium heterogeneity the uncertainty introduced by the non-uniqueness of the concentration distribution can be expected to be smaller than 50% of the estimated mass ﬂow rate.5 If more than one pumping well is used in subsequent pumping campaigns to measure the concentration time series, subsequent overlapping of the individual capture zones allows one to reduce the non-uniqueness of the solution.2 The shifting of the plume position due to pumping in subsequent campaigns is of course considered in the numerical solution.

5.4.2.3

Application of the Integral Investigation Method

Up to now the integral investigation method has been applied under real-world conditions at numerous locations in Europe and North America, covering a large variety of contaminant types and hydrogeological conditions. As an example, some of the results obtained from applications at an industrial site in southwest Germany will be summarised in the following.

248

5.4.2.3.1

Chapter 5.4

Application of the Integral Investigation Method at an Industrial Site in Southwest Germany

5.4.2.3.1.1 Site Description. The investigated site in southwest Germany is a more than 100-year-old industrial area with multiple potential sources of groundwater contamination. Abandoned landﬁlls (with municipal waste, demolition debris and industrial waste) are overlain by former and recent industrial sites. Figure 5.4.5 shows a plan view of the study area, which has an extent of 4.5 km2. The area is contaminated with mineral oil, BTEX compounds, chlorinated hydrocarbons as well as PAHs. The aquifer system under investigation is a Quaternary porous aquifer consisting of poorly sorted sand and gravel deposits. The aquifer thickness shows a high variability between 1 and 10 m mainly due to the irregular surface of the aquifer bottom. The average thickness of the aquifer is 3.5 m, and the mean transmissivity amounts to 6.85 103 m2 s1. The ln K values determined by slug tests and short duration pumping tests range from 9.8 to 3.1 with a variance of 1.9. This indicates the considerable heterogeneity of the aquifer. 5.4.2.3.1.2 Deﬁnition of Control Plane Locations, Field Investigations, Design and Performing of Integral Pumping Tests. An extensive compilation of the existing data provided the basis for the planning of the ﬁeld investigations.13 The relevant data included the characterisation of about 50 potential source zones with groundwater ﬂow directions, aquifer parameters (where available) and the analysis of the existing network of about 300 monitoring wells. Within

source characterization pumping test abandoned landfill

pot. contaminated site

Figure 5.4.5

Plan view of the study area in southwest Germany and position of integral pumping tests.13

Characterisation of Groundwater Contamination

249

the ﬁrst phase of the ﬁeld investigations starting in April 1997, 59 new monitoring wells were drilled within the study area. The wells for the (source zone characterisation) IPTs were placed along several control planes transverse to the main groundwater ﬂow direction (Figure 5.4.5). The cross-sections are completed by wells which are located downstream of sites with major contaminant sources. In addition, a number of wells were drilled to complete the existing network of monitoring wells. In each of the new wells and in some of the existing wells a short-duration (3–4 h) step drawdown test was conducted providing information about the well capacity and the aquifer parameters. Groundwater samples, obtained at the end of the step drawdown tests were used for some ﬁrst screening of the kind of contaminants and the range of concentrations to be expected. This information provided the database for the detailed planning of the pumping tests (discharge rates, pumping periods, sampling intervals, type of discharge treatment). The most critical cost factor was how to obtain the maximum possible size of the capture zone within a minimum of pumping time. This in general is achieved through maximum discharge rates. The sampling intervals for concentration measurements were chosen such that every concentration measurement represents equal spatial averaging with respect to the capture zone development. From October 1997 until June 1998, 34 full-scale IPTs were conducted in the study area.13 Due to the high hydraulic conductivity of the aquifer the pumping rates were 5.3 l s1 on average. With an average pumping period of 5.3 days, the resulting capture zone widths range from 30 to 120 m. In general about 10 groundwater samples were obtained during each pumping test and analyzed for the major organic compounds. The numerical model and the numerical evaluation of the IPTs is described in sections below dealing with parameter uncertainty and delimiting of contaminant source zones.

5.4.3

Methodology to Consider Aquifer Parameter Uncertainty and to Delimit Contaminant Source Zones Using Integral Measurements

5.4.3.1

Principles

Although estimations of plume locations form a part of the inversion problem, as outlined in the previous sections, it should be recognised that robust integral estimations of average concentrations and total mass ﬂows may not necessarily require an exact answer regarding the plume location. This is because diﬀerent plausible plume locations and ﬂow ﬁeld conceptualisations may yield similar results in terms of concentration and mass ﬂow. Formally, this view is supported by the analytical analyses of inverse problems of Bayer-Raich et al.,12 where it is shown that whereas solutions for exact plume locations are mathematically ill-posed, the corresponding solutions for average concentrations are relatively robust. In the following, we will explore this topic further

250

Chapter 5.4

and systematically review diﬀerent sources of uncertainty, summarising and discussing current knowledge regarding their inﬂuence on integral investigation results. We will also relate this to practical aspects related to IPT investigations of contaminated sites, including discussions and suggestions on how measurements should be performed in the light of existing uncertainties.

5.4.3.1.1

Sources of Uncertainty

One can distinguish between three principal sources of uncertainties in IPT investigations: (i) ﬂow model uncertainties, (ii) transport/reaction model uncertainties and (iii) the non-uniqueness of inversion results, leading to uncertainties of the original contaminant distribution in the aquifer. Regarding (i), we note that ﬂow model uncertainties can be a result of uncertain aquifer properties or uncertain initial or boundary conditions. For instance, considering IPTs performed such that the pumping rate is large relative to the natural ﬂow in the aquifer (a preferable condition for obtaining dependable IPT results), isochrones will be approximately circular around the pumping well in homogeneous aquifers. This is formally the case if the pumping rate Q is greater than or equal to 2pbq02t/ne, where b is the aquifer depth, q0 is the speciﬁc discharge, t is the pumping time and ne is the eﬀective porosity, as further outlined in Bayer-Raich et al.12 Then, in homogeneous aquifers, even if the hydraulic conductivity value K is uncertain, the inversion procedure implies that exact estimates of average concentrations can be obtained. The reason is that the circular isochronous positions and linear streamline positions are independent of the uncertain K estimate. However, associated mass ﬂows are in this case associated with uncertainty since they are dependent on K. In contrast, boundary condition uncertainties and uncertain K values may in heterogeneous aquifers result in both uncertain average concentration and mass ﬂow estimates. This is because the isochronous positions and streamlines generally depend on the assumed K distribution and boundary conditions (BCs), which inﬂuence inversion results. Jarsjo¨ et al.14 used a MODFLOW site model and a numerical inversion procedure for quantiﬁcation of the eﬀects on IPT results of a BC uncertainty (that, in turn, was caused by uncertainties in an underlying water balance study). Results showed that the eﬀects were relatively small: a factor two diﬀerence in constraining groundwater recharges resulted in relative errors in average concentrations and mass ﬂows that were commonly less than 10%. Additional uncertainties related to spatially variable K values may be quantiﬁed using Monte Carlo simulations of the inversion, in which a relatively large number of equally likely aquifer realisations are used for obtaining statistics on the associated variability in IPT results (e.g. Ref. 15). However, for most practical applications, it may be suﬃcient to consider only some limiting cases of aquifer heterogeneities, such as the fully stratiﬁed case (with K variability in the vertical direction only), for which the uncertainty analyses can be considerably simpliﬁed by application of numerical or analytical methods to each aquifer layer independently (e.g. Ref. 6).

Characterisation of Groundwater Contamination

251

Regarding transport/reaction parameter uncertainties, case (ii) above, Bayer-Raich et al.16 recently investigated the inﬂuence of linear sorption on IPT interpretations (during the pumping phase). Results showed that interpretation results are independent of the retardation factor under equilibrium conditions, although linear retardation due to sorption decreases the investigation volume in comparison with the non-retarded case. In other words, whereas this has implications for the technical dimensioning of the tests (e.g. required number of wells, pumping rates and pumping times for investigating a cross-section of certain dimensions), uncertainties in the retardation factor do not inﬂuence subsequent average concentration or mass ﬂow results. Furthermore, for obtaining a solution to the inversion problem, it is necessary to make some assumption regarding the contaminant attenuation in the direction of ﬂow. So far, all existing IPT studies have, for simplicity, assumed that concentrations are approximately constant over the well capture zone (in the ﬂow direction). A basic and relevant question posed by Zeru and Scha¨fer17 is how much various violations of this simplifying assumption can bias IPT results, particularly if strong concentration gradients exist within the well capture zone (for instance as a result of dispersion, (bio)degradation and ﬂuctuations of the ﬂow ﬁeld). In response, Bayer-Raich et al.18 showed that IPT results are unbiased as long as the concentration attenuation along the ﬂow direction is linear, regardless of the concentration gradient. This hence implies that biases in IPT investigations due to the ‘‘constant concentration’’ assumption are only introduced if the concentration attenuation is nonlinear. Jarsjo¨ and Bayer-Raich19 further considered wide contaminant plumes and provided an analytical expression from which the inﬂuence of concentration attenuation on IPT results can be quantiﬁed. Speciﬁcally, eﬀects of exponential ﬁrst-order decay were investigated, being the most common model for natural attenuation (NA) processes. Results show that the ‘‘constant concentration’’ assumption does not considerably bias predictions even if the investigated contaminant undergoes ﬁrst-order decay, unless the resulting attenuation is very large. For instance, a 60% concentration decrease over the capture zone extent caused by ﬁrst-order decay yields a prediction error of 4%. Finally, the uncertainty related to the non-uniqueness of the inversion results, case (iii) above, is illustrated in Figure 5.4.6. The upper part of the ﬁgure illustrates a C(t) curve, i.e. a contaminant concentration–time series measured in a pumping well, which is used as input for the IPT inversion analysis. The curve shows that the concentration is zero, or close to zero, in the beginning of the pumping test, implying that the well is located in a relatively clean part of the aquifer. After some time, concentrations in the well increase as one or several contaminant plumes are drawn towards the well due to the pumping. However, the original location of the contaminant plume relative to the well cannot be judged on the basis of the measured C(t) curve, which means that several diﬀerent interpretations (or conceptual models) regarding the spatial contamination distribution are possible on the basis of the same curve. In analogy with Jarsjo¨ et al.,14 this uncertainty will in the following be denoted contamination model uncertainty.

252

Chapter 5.4

?

concentration in pumping well

Observation:

? time

? Interpretation I: Plume at left hand side

Interpretation III: Plume at right hand side

Interpretation II: Plume at both sides

280

280

280

270

270

270

260

260

260

250

250

250

240

240

240

230

230

230

220 220 220 220 230 240 250 260 270 280 220 230 240 250 260 270 280 220 230 240 250 260 270 280

Figure 5.4.6

Illustration of the non-uniqueness in the inversion results, regarding the plume position relative to the well. The lower part of the ﬁgure shows possible hydraulic and contaminant situations in the nearest well vicinity (plan views), with the shaded region representing the plume location, the thin black lines representing isolines of hydraulic head and the closed curves representing isochrones. The well is located within the inner isochrone. The numbers on the axes are spatial coordinates in the x and y directions, in metres.

Figure 5.4.6 illustrates three possible interpretations that can be made regarding the contaminant distribution: the contaminant plume may be located at the left-hand side of the well only, at the right-hand side of the well only, or at both sides. Jarsjo¨ et al.14 quantiﬁed this left–right uncertainty considering 19 IPTs performed in a strongly heterogeneous aquifer at an industrial site in southwest Germany (see above). Speciﬁcally, for each well, the contaminant mass ﬂow (MF) was quantiﬁed for each considered plume position (left, right or both, i.e. symmetrical; see Figure 5.4.6), resulting in three diﬀerent MF estimations per well and contaminant. The uncertainty in the result was then for each well and contaminant quantiﬁed through the relative MF arising from the three considered positions, deﬁned as RMF ¼ (max MF min MF)/(average MF). Results showed that large uncertainties mainly occurred if the aquifer properties were strongly heterogeneous in the nearest vicinity of the pumping well. A regression analysis showed that RMF was correlated to the standard deviation of the log-transmissivity slnT within the capture zone volume of the qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ aquifer model, quantiﬁed as sln T 1=ðnCZ 1Þ Sðln T mln T Þ2 , where

Characterisation of Groundwater Contamination

253

mlnT is the mean value of ln T, the summation is performed over all the model cells belonging to the well capture zone and nCZ is the total number of model grid cells belonging to the well capture zone. Speciﬁcally the correlation was RMF E 0.5slnT + 0.1, with R2 ¼ 0.7. The constant 0.1 in this relation illustrates that as long as the aquifer variability is low (with slnT close to zero) in the nearest well vicinity, the left–right uncertainty is of the order of 10%, even though the considered aquifer as a whole is strongly heterogeneous (in this case, slnT,Tot is 1.6 for the whole aquifer). The relation further shows that the uncertainty increases with a proportionality constant of approximately 0.5 as slnT of the well vicinity increases. In summary, we have here reviewed the main sources of uncertainties in IPT investigations. Some of the uncertainty sources are the same as, or similar to, uncertainties in traditional point measurement interpretations. This is for instance the case for estimations of mass ﬂows on the basis of measured concentrations. However, a main diﬀerence between the IPT measurements and the traditional point measurements is that, in the point measurement case, there will for most practical applications always remain an uncertainty as to the contamination situation in between the measurement points. This interpolation uncertainty cannot be quantiﬁed unless new measurement points are established. For the case of IPTs, the investigation volume is large and the traditional interpolation uncertainty is transformed into a quantiﬁable uncertainty related to the (unknown) position of the plume relative to the pumping well. The uncertainty can be quantiﬁed if the aquifer property statistics are known or can be estimated at a reasonable level of conﬁdence, as illustrated through the above example. Alternatively, the uncertainties can be reduced by constraining the plume position interpretations (Figure 5.4.6), e.g. on the basis of point concentration data.

5.4.3.1.2

Source Zone Delimiting

The primary outputs of IPT investigations are average concentration values and mass ﬂows of dissolved contaminants over CPs. However, the location of (free phase) sources that give rise to the dissolved contaminant plumes that can be detected at CPs is in many cases unknown. IPT results can aid in delimiting the location of such source zones. In addition, a negative IPT result, i.e. the case that no contaminants are detected in the pumping well, implies that the whole length of the CP is essentially free from contamination. Such a result can be used for delimiting larger upstream areas absent of contaminant sources, which means that these areas can be excluded from further investigation. In the following, we outline and discuss the principles of a technique described in Jarsjo¨ et al.,14 through which source zones and source absence zones can be delimited at a chosen level of conﬁdence, depending on individual or administrative rules and concepts. The source zone delimiting procedure can also be applied to sections of a control plane, deﬁned, for example, by the isochrones corresponding to sampling times within a concentration time series.

254

Chapter 5.4

With respect to delimiting of the position of the source zone, the approach is to use particle backtracking techniques considering diﬀerent alternative aquifer/boundary condition realisations, thereby accounting for ﬂow model uncertainties (see also the discussion of the previous section and the ﬂow charts below). The principle is illustrated in Figure 5.4.7, in which ‘‘Model 1’’ and ‘‘Model 2’’ represent identiﬁed worst-case scenarios with regard to ﬂow directions. Figure 5.4.7(a) shows how particle backtracking then is performed by placing particles along the length of the considered CP and backtracking them to the upstream boundary, hence calculating the upgradient pathway of particles detected at the CP. This yields a distribution of spatial limits in the transverse direction, which is relevant for inert compounds, for which the concentration along streamlines is constant over long distances (kilometres). The resulting outer limits, including streamlines from the union of the Model 1 and Model 2 results, are relevant for delimiting the possible source zone location of inert compounds (Figure 5.4.7(b); thick black line). The inner limits, including streamlines from the intersection of the Model 1 and Model 2 results (Figure 5.4.7(b); thick grey line), are relevant for delimiting the source absence zone for inert compounds at the same level of conﬁdence as for the previous case, as further explained in Jarsjo¨ et al.14 However, most organic compounds are subject to, for example, (bio-) degradation and sorption, which may lead to the development of relatively

Figure 5.4.7

Principle of (a) using particle backtracking from diﬀerent, equally plausible ﬂow models, to delimit the source zone location (or a source absence zone) at the same signiﬁcance level, for (b) inert and (c) degrading (reactive) compounds. (After Ref. 14.)

Characterisation of Groundwater Contamination

255

short (down to 10 m or so in length) and stable plumes. Since degradable contaminants cannot be detected in the ﬂowing groundwater after a certain downgradient distance from the source (the plume length), predictions about upgradient source absence (for the result that no contaminant was found in the pumping well) are invalid for distances larger than this plume length. Further (for the result that contaminant was found in the pumping well), the source cannot be located further upgradient than this critical length. We here deﬁne the plume length as the longest distance from the source, among the bundle of streamtubes leaving the source, to a downgradient point in which the concentration c is lower than some ﬁxed value clow. Recognising that site- and contaminant-speciﬁc prediction of attenuation functions is essentially beyond state-of-the-art knowledge, we further make the assumption that it at least is possible to estimate (e.g. based on empirical, historical data) the total plume lengths LMAX and LMIN that will not be exceeded, and will be exceeded, respectively (at some chosen conﬁdence level a), when allowing the plume to develop fully in a given setting (i.e. considering type of contaminant, the aquifer type, ambient physical and chemical conditions, etc.; see Section 5.4.3.3 for an example). For example, Ru¨gner and Teutsch20 provide empirical plume length statistics for diﬀerent compounds (e.g. benzene, chlorinated hydrocarbons and polyaromatic hydrocarbons), which is useful in estimations of LMAX and LMIN. Alternatively, ﬁrst-order decay rate models or sophisticated multispecies– multiprocess reactive transport models can, in principle, be used to estimate the required lengths. Using a similar reasoning as for inert compounds, the possible source zone for degrading compounds can now be spatially delimited by considering LMAX and the outer streamline limits (Figure 5.4.7(c); thick black line). Furthermore, the source absence zone for degrading compounds can be delimited by considering LMIN and the inner streamline limits (Figure 5.4.7(c); thick grey line).

5.4.3.2

Decision Tree Approach

Regarding IPT evaluations of mass ﬂows and average concentrations, Figure 5.4.8 shows a general scheme as to how the associated main sources of uncertainty, identiﬁed in the previous sections, in practice can be quantiﬁed and addressed, using various models and methods, considering a speciﬁc site. Also, Figure 5.4.8 indicates how the inversion results can be used for comparison with, for example, regulatory limits, providing a basis for clean-up decisions or related measures. In particular, the uncertainty analyses imply that a set of diﬀerent, equally plausible, mass ﬂow and average concentration values are obtained, corresponding to diﬀerent groundwater and contamination models (boxes I to VI in Figure 5.4.8). Relevant statistics describing the set of values are then compiled (box VII in Figure 5.4.8) and compared with regulatory limits (box VIII in Figure 5.4.8), which implies that decisions can be taken at diﬀerent (chosen) levels of conﬁdence. Furthermore, considering delineations of source zones and source absence zones, Figure 5.4.9 shows that, depending on site-speciﬁc conditions and

256

Chapter 5.4 I. For each groundwater model i

II. Run hydraulic model→ Flow velocities vi ; streamlines si III. For each contamination model IV. Perform inversion → mass flow, mi,j ; concentrations ci,j V. Next contamination model j VI. Next groundwater model i VII. Compile statistics of mi,j & ci,j (e.g., mean value and standard deviation of box IV results) VIII. Compare m and c statistics with regulatory limit(s), L. (Example outcome: m < L with 95% confidence.) DECISION

Figure 5.4.8

Flow chart for assessing uncertainty in IPT estimates of mass ﬂows and average concentrations.

general knowledge about the contaminant under investigation, one of the following end results can be expected (grey boxes in Figure 5.4.9). (I) Source zone location can be delimited both in transverse and longitudinal directions. (II) Source zone location can be delimited only in the direction transverse to ﬂow. (III) Source absence zone can be delimited both in transverse and longitudinal directions. (IV) Source zones and source absence zones cannot be delineated for the considered contaminant. Type IV results occur if the considered contaminant is not detected and at the same time the state-of-the-art knowledge is too limited for predicting plume lengths under the ambient conditions. Alternatively, the same result is obtained if the contaminant is not detected and at the same time the present analytical

257

Characterisation of Groundwater Contamination Previous studies of plume lengths under similar conditions?

For each considered compound

Enough data to perform statistical analysis?

yes

yes

no no Contaminant detected?

no yes

no RESULT – type IV: Cannot delimit the source zone, or draw conclusions regarding the absence in groundwater of upgradient contaminant.

Contaminant detected?

yes

Fewer than a handful studies?

yes

RESULT – type II : Source zone can be delimited, but only in the direction transverse to the flow.

Measured concentration above regulatory concentration limit?

LMAX = documented longest plume length LMIN = documented shortest plume length

LMAX = 75 to 99 percentile length LMIN = 1to 25 percentile length

yes

no

RESULT – type I: Source not further upgradient than LMAX. Source zone delimited by LMAX. in combination with particle tracking results.

no

Analytical detection limit considerably below the regulatory concentration limit?

yes

Measured concentration considerably below the regulatory concentration limit?

yes

(Historical) Data shows that contaminant is old enough for the development of a stable or shrinking plume?

yes

no no

RESULT – type IV: Cannot delimit the source zone, or draw conclusions regarding the absence in groundwater of upgradient contaminant.

Figure 5.4.9

no/ no data

RESULT – type III: No source within upgradient distance LMIN. Clean zone delimited by LMIN in combination with particle tracking results.

Flow chart for assessing uncertainty in source zone, or a source absence zone, delineations.

procedures are insuﬃcient for the purpose, or if the plume is young/not at steady state. Note, however, that a no-detect result otherwise (i.e. if analytical procedures, state-of-the art process knowledge and historical records are appropriate) implies that the source can be excluded with high certainty from a speciﬁc area (a type III result). A detection of contaminant at a CP implies that the source causing the contamination (of the speciﬁc CP) can be delimited

258

Chapter 5.4

in the direction transverse to ﬂow (at diﬀerent levels of certainty), through particle tracking (a type II result). Furthermore, if the state-of-the-art knowledge of plume lengths under ambient conditions is appropriate, a detection of contaminant implies that the location of the source can be delimited both along and transverse to the direction of ﬂow, at a chosen level of certainty (a type I result).

5.4.3.3

Example of Application at an Industrial Site

Figure 5.4.10 shows a comparison of regulatory concentration limits (State of Baden-Wu¨rttemberg, Germany) and IPT estimations of average benzene concentrations across 19 considered CPs at the application site in southwest Germany (see above). If the contaminant concentration in the groundwater at the CP is above the limit (dark grey CPs in Figure 5.4.10), or even up to one order of magnitude (OM) below the limit (lighter grey CPs in Figure 5.4.10), we consider it contaminated and use the methodology described in previous sections to delimit the location of the contamination source. If, on the other hand, the concentration at the CP is more than one OM below the limit, or if the contaminant was not detected analytically at all, we consider the groundwater clean (corresponding CPs are white in Figure 5.4.10) and use the methodology to delimit zones absent of source.

Figure 5.4.10

Predicted zones delimiting benzene source locations, and zones absent of benzene source, at a ﬁeld application site in southwest Germany. (After Ref. 14.)

Characterisation of Groundwater Contamination

259

In Figure 5.4.10, the borders of the predicted zones delimiting contaminant sources are shown with black lines (thick if the concentration at the CP is above the regulatory limit). Furthermore, the zones predicted to be absent of contaminant source are dashed. Since observations of benzene plumes (which are generally biodegradable) are relatively frequent and well-documented, and statistical analyses of benzene plume lengths are performed and summarised in Ru¨gner and Teutsch,20 we can here present results of type I and III, using the notation of the ﬂow chart in Figure 5.4.9. Figure 5.4.10 shows, as an example, results for a conﬁdence level of 75%. Considering the Ru¨gner and Teutsch20 database and referring to the notation introduced in Section 5.4.3.1.2, the parameter LMAX is taken as an upper threshold value, chosen such that 75% of the reported plume lengths will fall below the threshold. Furthermore, LMIN is taken as a lower threshold value, chosen such that 75% of the reported plume lengths will fall above the threshold. For benzene, the analysis resulted in LMAX ¼ 420 m and LMIN ¼ 60 m. As a consequence, at our ﬁeld site the benzene source could be delimited to within 420 m in the upstream direction from the measurement well (see wells 5, 9, 10, 20, 32, 62, 64, 2202 and 2058 of Figure 5.4.10), whereas source absence of benzene (for the wells where no benzene was found) could be delimited to within 60 m at the same level of conﬁdence (see wells 2, 3, 7, 15, 16, 17, 66 and 359 of Figure 5.4.10). Performing the analysis at a higher conﬁdence level would imply larger extents (in the ﬂow direction) of the source presence zones and smaller extents of the source absence zones. This methodology for zone delineations at given levels of conﬁdence can be considered a basis for planning of future land use and/or in necessary additional investigation and remediation activities.

5.4.4

Quantiﬁcation of Natural Attenuation Rates Using Integral Measurements

5.4.4.1

Principles

After screening an area at large scale as described above, if compound-speciﬁc total mass ﬂow rates (MF) have been quantiﬁed within selected contaminated smaller scale sub-areas at diﬀerent distances from a contaminant source zone using the integral groundwater investigation method, and if the average travel time Dt in groundwater between the two operated control planes CP(I) and CP(II) is known, it is possible to quantify compound-speciﬁc eﬀective ﬁrstorder natural attenuation (NA) rates using the following equation (assuming a retardation factor equal to one for the eﬀective values): MFCPðIIÞ ¼ MFCPðIÞ elDt

ð5Þ

MFCPðIIÞ 1 l ¼ ln MFCPðIÞ Dt

ð6Þ

leading to

260

Chapter 5.4

with l [T1], MFCP(I) and MFCP(II) representing the eﬀective natural attenuation rate constant and the measured compound-speciﬁc mass ﬂow rates at control planes CP(I) and CP(II), respectively.4 Total mass ﬂow rates and average concentrations can be simultaneously estimated for a number of target compounds at each abstraction well. This may include not only the original contaminants, but also potential degradation products or hydrogeochemical indicators for natural attenuation processes, e.g. pH, EH, sulfate, nitrate, dissolved iron. It should be noted that the diﬀerences in mass ﬂuxes between any two control planes can be due to degradation, sorption and volatilisation of the target compound. In general, a calculation of NA rates with eqn. (6) does not allow one to diﬀerentiate between these processes. NA rates estimated in this way consequently incorporate all mass ﬂow reducing factors such as sorption and degradation. Recharge and dispersion, however, do not aﬀect the results because of the spatial integration inherent to the mass ﬂow rate estimation. To be able to estimate the relative contribution of the diﬀerent mass ﬂow reducing processes to the measured eﬀective contaminant mass ﬂux reduction between two control planes, and hence to obtain evidence of degradation, the ﬁeld-scale measurements must be accompanied by reactive transport modelling.

5.4.4.2

Example of Application at a Former Gasworks Site

The studied former urban gasworks site is situated in a river valley in southwest Germany4 (Figure 5.4.11). The contaminated aquifer is composed of shallow Quaternary gravels with locally embedded sand, silt and loamy clay. Based on pumping tests, the arithmetic average of the hydraulic conductivity at the site was estimated as 2.5 103 m s1.22 Hydraulic heads were monitored over a 3-year period with no indication of signiﬁcant seasonal changes or temporally variable groundwater ﬂow directions.23 The steady state of the local groundwater ﬂow ﬁeld is due to the artiﬁcial regulation of the water level of the river that runs parallel to the eastern border of the ﬁeld site (Figure 5.4.11). Transport parameters have been determined in the ﬁeld employing a natural gradient multi-tracer test.24 At this site, two control planes were installed4 (Figure 5.4.11). The contaminant source is formed by NAPL phase covering an area of approximately 20 000 m2. The NAPL originates from a number of point sources. In the source zone, total BTEX concentrations in groundwater range up to 12 mg l1, whereas PAH concentrations are up to 3.2 mg l1. The resulting PAH plume has an approximate width of 120 m. Of the 16 EPA-PAHs, only acenaphthene shows high concentrations of 190 mg l1 at distances of about 280 m downstream of the source zone. The overall length of this plume is unknown yet, as no monitoring wells are available further downstream. The overall length of the BTEX plume is assumed to be less than 280 m, as only p-xylene showed concentrations exceeding 0.2 mg l1 at control plane 2 (Figure 5.4.11). Figure 5.4.12 illustrates an example of estimated mass ﬂow rates of BTEX and other hydrocarbons based on the analytical inversion of the concentration

261

Characterisation of Groundwater Contamination Estimated PAH-plume extension

?

2

e1

l

ro

nt

Co

Control plane Employed well Existing well NAPL phase Flow direction from tracer tests

an pl

2069 B72

B73

NT01

ne1

Va

l tro

n Co

y lle bo

P1

B41 P2

un

ree St

r ve

ry

Ri

B42

da

N

pla

t

01

50

00m

NAPL free phase

Figure 5.4.11

Site overview with the location of the employed monitoring wells and the installed control planes for the integral groundwater investigation.4

10.0000

Flux [g/d]

1.0000 0.1000 0.0100 0.0010

Figure 5.4.12

ne

ne

de In

B

da In

TM

M B

2, 31,

TM B

4T

2, 1,

PB 1,

3,

5-

PB

yl

yl

X

oIs

o-

nz

X p-

l

Be

To

E-

Be

nz

0.0001

Mass ﬂow rates of BTEX and other aromatic hydrocarbons at control plane 1 (solid bars) and control plane 2.21

262

Chapter 5.4

Effective NA rate constant [1/d]

0.15

0.10

0.05

Figure 5.4.13

e en

an

B

e In d

In d

B

3TM

1,

2,

B

4TM

1,

2,

TM

PB 1,

3,

5-

PB

yl

Is o-

yl

oX

nz

pX

l

Be

To

E-

Be

nz

0.00

Eﬀective natural attenuation rate constants for BTEX and other aromatic hydrocarbons.4

time series measured at the abstraction wells situated at control plane 1 and control plane 2 during the integral site investigation. Based on the measured compound-speciﬁc mass ﬂow rates at the two control planes, eﬀective ﬁrst-order natural attenuation rate constants could be estimated using eqn (6). The results are shown in Figure 5.4.13. The estimated NA rate constants for the BTEX compounds agree well with biodegradation rate constants described in the literature (e.g. Ref. 25). It should be mentioned that it is only the irreversible (bio-)degradation process that really removes mass from the investigated aquifer system. In order to estimate the individual contributions of reactive transport processes such as sorption, (bio-)degradation, etc., to the reduction of the total mass ﬂow rate between the two (or more) control cross-sections, a process-based reactive transport model has to be applied, such as PHT3D26 or MT3D-IPD.27 The numerical models can account for sorption, (bio-)degradation, etc., as well as for the heterogeneity of both the hydrodynamic and reactive transport (e.g. sorption) parameters. The individual contributions of contaminant mass ﬂow reducing reactive transport processes such as sorption and (bio-)degradation to the total mass ﬂow reduction can be quantiﬁed by comparing the measured contaminant mass ﬂow reduction with the results from the numerical model within a combined forward-inverse modelling framework.

5.4.5

Multilevel Integral Investigation of Contamination

5.4.5.1

The Multilevel Integral Investigation Method

At large scale, the IPT method is usually applied in a two-dimensional depthaveraged approach, i.e. assuming the concentrations and the IPT capture zone extent to be constant over the aquifer thickness. However, in many cases

Characterisation of Groundwater Contamination

263

contamination may be limited to distinct aquifer levels, and, due to aquifer layering, the capture zone width may vary with depth. Therefore, after screening an area at large scale as described above, for investigations at selected contaminated smaller scale sub-areas the IPT method was improved to allow multilevel measurements and to account for aquifer layering.28,29 Employing the improved investigation method, site assessment and the total contaminant mass ﬂow rate to be removed by a remediation measure can now also be obtained with a vertical resolution. The concept of the improved integral investigation method for multilevel measurements is shown in Figure 5.4.14. To obtain the required multilevel concentration time series, a ﬂow separation technique was developed, allowing multilevel groundwater sampling also within fully screened pumping wells.28 The algorithm used for the numerical inversion of the measured multilevel concentration time series at the pumping wells is implemented in the program code CSTREAM.5 In case of multilevel concentration time series, the twodimensional, depth-integrated formulation5 (see also above) is applied to distinct aquifer layers, which are deﬁned by the vertical extent of the multilevel sampling sections. This layered approach is applicable to situations with no signiﬁcant vertical mass transport.

5.4.5.2

Example of Application

The multilevel integral investigation method was ﬁrst applied at a former chlorobenzene manufacturing site in Germany. At this site chlorobenzene concentrations in groundwater of up to 55 mg l1 could be measured. The porous aquifer has a thickness of about 15 m with a mean hydraulic conductivity of 5 1041 103 m s1. A total of three pumping wells, each with four multilevel sampling sections and a maximum pumping rate of 7.5 l s1, were positioned along a control cross-section of about 140 m length. The pumping time per well was up to about 8 days. During pumping, groundwater concentration time series of BTEX, chlorobenzene and chlorotoluene compounds were measured at each of the pumping wells and at each multilevel sampling section with variable sampling intervals, allowing a spatial resolution of about 1–2 m along the control plane. Further details are given in Ptak et al.[28] Following the measurements, the multilevel concentration time series were evaluated using the CSTREAM code, and a local-scale three-dimensional ﬂow and transport model of the site. As an example of the results, Figure 5.4.15 shows the relative distribution of mean concentrations in the vertical direction across the control plane for benzene and 1,2-dichlorobenzene. The multilevel capture zones, the concentration values and the mass ﬂow rates obtained from the multilevel inversion are signiﬁcantly variable in space, indicating an irregular distribution of contaminant mass within the plume. Distinct zones with relatively high concentrations can be identiﬁed. It should be noted that the maximum concentrations may be found in aquifer zones between the pumping wells, and that the high concentration and the high mass ﬂux zones are not necessarily coincident.

264

Chapter 5.4 Pumping tests with measurement of multilevel concentration time series Well i+1

Well i

Position z j Position z j+1

Mass flux

Contaminant source

Control plane

Well i+1

Well i

Position z j+1

C

C Position z j+1

Position z j

t0

t1

Position z j

t1

t2

Multilevel concentration time series (compound specific)

Multilevel inversion algorithm based on a 3D numerical flow and transport model

Level-oriented contaminant mass fluxes and concentrations at control plane

Figure 5.4.14

Multilevel integral investigation method.29

Employing the multilevel integral investigation method, the spatial distribution of contaminant concentrations and mass ﬂow rates across control planes can be estimated without the need for a dense sampling network. This information can then be used for example to obtain an optimal design of remediation measures by focusing on the highly contaminated aquifer zones. Backtracking starting from the identiﬁed high mass ﬂux sections at the control plane allows a more detailed local-scale delimiting of the contaminant source

265

Characterisation of Groundwater Contamination 1,2-DCB (BLACK=2 mg/L)

Benzene (BLACK=14 mg/L) 41

40

38

75

75

70

70

65

65

60

60

55

55

50 5720100

5720150

Figure 5.4.15

5720200

5720250

5720300

50 5720100

41

40

38

5720150

5720200

5720250

5720300

Relative distribution of mean concentrations in the vertical direction across the control plane for benzene and 1,2-dichlorobenzene (1,2-DCB) (greyscale related to the maximum concentration shown in black).29

zone, compared to the two-dimensional approach described above. When applied upstream and downstream of a remediation measure, the eﬃciency of the remediation can be quantiﬁed by comparing the total upstream and downstream mass ﬂow rates. Multilevel integral measurements at two or more control planes positioned in the downstream direction allow the quantiﬁcation of the natural attenuation potential at a high spatial resolution. In this way the new multilevel integral investigation approach may signiﬁcantly contribute to an improvement of contaminated land assessment and revitalisation.

5.4.6

Conclusions

Employing the control plane-based integral investigation method, the compound-speciﬁc average contaminant concentration, the spatial distribution of concentration values and mass ﬂow rates along a control plane, as well as the total contaminant mass ﬂow rates downstream of an area under investigation can be estimated quickly and with a high level of certainty. The information obtained from this analysis can be considered a basis for planning of future land use. The results from the integral investigation can be used for risk assessment purposes, for the quantiﬁcation of the natural attenuation potential and for the design of remediation measures. In addition, a consistent quantiﬁcation of uncertainties in the results from the application of the integral groundwater investigation method is possible, considering uncertainty in the boundary conditions and uncertainty in the hydraulic property values of the aquifer. Finally, the delimiting of the source zone extent and its uncertainty allows one to deﬁne priorities for further investigation measures at a smaller scale and to

266

Chapter 5.4

develop cost-optimised clean-up strategies at sites with complex contamination patterns. In this way, employing the large- to small-scale screening procedure it is possible to obtain optimal results for the eﬀorts spent. Therefore, the new approaches may become a basis for the development and implementation of EU directives on contaminated land assessment and revitalisation in urban industrial areas.

Acknowledgements Financial support for the presented work was provided by the European Union, by the Landesanstalt fu¨r Umweltschutz Baden-Wu¨rttemberg, by the Projekt Wasser-Abfall-Boden Baden-Wu¨rttemberg, by the BMBF, by the Deutsche Forschungsgemeinschaft and by the UFZ Leipzig-Halle GmbH. The authors gratefully acknowledge the contributions of Luca Alberti, Marti Bayer-Raich, Sebastian Bauer, Diego Bianchi, Sara Ceccon, Philippe Elsass, Thomas Ertel, Jadwiga Gzyl, Thomas Holder, Hermann J. Kirchholtes, Christian Kolesar, Dietmar Mu¨ller, Caterina Padovani, Georgia Spausta, Gilles Rinck, Gerhard Scha¨fer, Maria Giovanna Tanda, Georg Teutsch and Andrea Zanini within the INCORE project.

References 1. G. Teutsch, T. Ptak, R. Schwarz and T. Holder, Grundwasser, 2000, 4, 170–175. 2. T. Ptak, R. Schwarz, T. Holder and G. Teutsch, Grundwasser, 2000, 4, 176–183. 3. T. Ptak and G. Teutsch, Development and application of an integral investigation method for the characterization of groundwater contamination, in Contaminated Soil 2000, Thomas Telford, London, 2000, pp. 198–205. 4. A. Bockelmann, T. Ptak and G. Teutsch, J. Contam. Hydrol., 2001, 53, 429–453. 5. M. Bayer-Raich, J. Jarsjo¨, T. Holder and T. Ptak, Numerical estimations of contaminant mass ﬂow rate based on concentration measurements in pumping wells, ModelCare 2002: A Few Steps Closer to Reality, IAHS Publication no. 277, 2003, pp. 10–16 (ISBN 1-901502-07-4). 6. M. Bayer-Raich, Integral pumping tests for characterization of groundwater contamination, PhD thesis, Center for Applied Geoscience, University of Tu¨bingen, 2004. 7. R. Schwarz, Grundwasser-Gefa¨hrdungsabscha¨tzung durch Emissions- und Immissionsmessungen an Deponien und Altlasten, PhD thesis, Center for Applied Geoscience, University of Tu¨bingen, 2002. 8. S. Bauer, T. Holder, M. Bayer-Raich, T. Ptak, Ch. Kolesar and D. Mu¨ller, J. Contam. Hydrol., 2004, 75, 183–214. 9. S. Bauer, M. Bayer-Raich, T. Holder, J. Jarsjo¨, T. Ptak and G. Teutsch, The integral groundwater investigation method: inversion of

Characterisation of Groundwater Contamination

10.

11.

12. 13.

14. 15.

16. 17. 18. 19. 20.

21.

22. 23. 24.

267

concentration-time series and ﬁeld application at the INCORE Strasbourg site, in IMAGE-TRAIN, Innovative Management of Groundwater Resources in Europe, ed. G. Prokop, Training and RTD Coordination Project, Proceedings of the 1st IMAGE-TRAIN Cluster Meeting, Karlsruhe, Germany, Federal Environment Agency, Vienna, 2002, pp. 75–79. M. G. McDonald and A. W. Harbaugh, MODFLOW. A modular threedimensional ﬁnite diﬀerence groundwater ﬂow model, investigation of United States Geological Survey, Washington, DC, 1988. D. W. Pollock, User’s Guide for MODPATH/MODPATH-PLOT, Version 3: a particle tracking post-processing package for MODFLOW, US Geological Survey ﬁnite diﬀerence ground-water ﬂow model, US Geological Survey, 1994. M. Bayer-Raich, J. Jarsjo¨, R. Liedl, T. Ptak and G. Teutsch, Water Resour. Res., 2004, 40, W08303. T. Holder, T. Ptak, R. Schwarz and G. Teutsch, Groundwater risk assessment at contaminant sites. A new approach for source zone characterization: the Neckar valley study. Groundwater ouality: remediation and protection, IAHS Publication no. 250, Tu¨bingen, 1998. J. Jarsjo¨, M. Bayer-Raich and T. Ptak, J. Contam. Hydrol., 2005, 79(3–4), 107–134. A. Peter, Assessing natural attenuation at ﬁeld scale by stochastic reactive transport modelling, PhD thesis, Center for Applied Geoscience, University of Tu¨bingen, 2002. M. Bayer-Raich, J. Jarsjo¨, R. Liedl, T. Ptak and G. Teutsch, Water Resour. Res., 2006, 42, W08411. A. Zeru and G. Scha¨fer, J. Contam. Hydrol., 2005, 81, 106–124. M. Bayer-Raich, J. Jarsjo¨ and G. Teutsch, J. Contam. Hydrol., 2007, 90, 240–251. J. Jarsjo¨ and M. Bayer-Raich, Water Resour. Res., in press. H. Ru¨gner and G. Teutsch, Literature study, Natural attenuation of organic pollutants in groundwater, Final Report for EU-FP5 project INCORE, 2001. A. Bockelmann, T. Ptak, R. Liedl and G. Teutsch, Mass ﬂux, transport and natural attenuation of organic contaminants at a former urban gasworks site, in Prospects and Limits of Natural Attenuation at Tar Oil Contaminated Sites, Dechema eV Texte, Frankfurt am Main, 2001, pp. 325–336. M. Herfort, T. Ptak, O. Hu¨mmer, G. Teutsch and A. Dahmke, Grundwasser, 1998, 3(4), 159–166. M. Herfort, Reactive transport of organic compounds within a heterogeneous porous aquifer, PhD thesis, Universita¨t Tu¨bingen, 2000. D. Bo¨sel, M. Herfort, T. Ptak and G. Teutsch, Design, performance, evaluation and modelling of a natural gradient multitracer transport experiment in a contaminated heterogeneous porous aquifer, in Tracers and Modelling in Hydrogeology, ed. A. Dassargues , IAHS, Liege, 2000, pp. 45–51.

268

Chapter 5.4

25. T. H. Wiedemeier, H. S. Rifai, C. J. Newell and J. T. Wilson, Natural Attenuation of Fuels and Chlorinated Solvents in the Subsurface, Wiley, New York, 1999. 26. H. Prommer, D. A. Barry and C. Zheng, Ground Water, 2003, 42(2), 247–257. 27. R. Liedl and T. Ptak, J. Contam. Hydrol., 2003, 66(3–4), 239–259. 28. T. Ptak, M. Bayer-Raich and S. Bauer, Grundwasser, 2004, 4(9), 235–247. 29. T. Ptak, M. Bayer-Raich and S. Bauer, Multilevel integral investigation of contamination in large polluted aquifers, in MODFLOW and MORE 2006: Managing Ground Water Systems, Conference Proceedings, International Ground Water Modeling Center (IGWMC), Colorado School of Mines, 2006, pp. 574–578 (www.mines.edu/igwmc/).

CHAPTER 5.5

Improved Risk Assessment of Contaminant Spreading in Fractured Underground Reservoirs CHRISTOS D. TSAKIROGLOU FORTH/ICE-HT, Stadiou Street, Platani, PO Box 1414, GR-26504 Patras, Greece

5.5.1

Introduction

Soil and groundwater contamination by hazardous substances is tending to become one of the most signiﬁcant problems for the environmental and economic policies of the European Union (EU). Fractures are widespread in various types of soils and rocks (e.g. clay till, sandstone, chalk, granite, limestone) and inﬂuence drastically the pathways of liquid pollutant migration in the subsurface with the potential contamination of underground aquifers. In the past, numerical simulators were developed to forecast the various contaminant transport (e.g. gravity ﬂow, sorption, dissolution/dispersion) and reaction (e.g. biodegradation) processes taking place in the fractured zones of the subsurface. Such simulators may be used as tools in the risk assessment as well as in the design of remedial actions on fractured sites contaminated by non-aqueous phase liquids (NAPLs). Nevertheless, in spite of the progress that has been accomplished in the development of intelligent numerical solvers of the equations modelling the contaminant transport/biodegradation in fractured media, there is still a lack of fundamental knowledge concerning: the quantitative description of fractured media at diﬀerent scales (e.g. single fractures, fracture network); the correlation of the NAPL transport pathways with fracture morphology and NAPL rheology (mobility); 269

270

Chapter 5.5

the quantiﬁcation of the up-scaled transport coeﬃcients of highly heterogeneous fractured media in terms of reliable phenomenological models; and the introduction of the foregoing information into macroscopic numerical simulators to produce reliable data for the risk analysis of contaminated fractured sites.

5.5.1.1

Literature Review

The design and implementation of remediation strategies for fractured soils and aquifers contaminated by NAPLs is one of the most intractable problems. In most cases, excavation down to the fractured rock, soil or sediment and the removal of the contaminated material are very expensive tasks. Potential remediation alternatives are dewatering of the contaminated zone at high pumping rates and removal of the volatile NAPLs through soil vapour extraction,1 surfactant-enhanced NAPL dissolution and mobilisation,2 in situ bioremediation,3 steam injection,4 etc. The design and installation of the most suitable remediation scheme on a contaminated fractured site requires information about the spatial and temporal distribution of pollutants throughout the subsurface. Macroscopic simulators of NAPL transport in the subsurface oﬀer a cost-eﬀective method for the long-term mapping of the distribution of pollutants in fractured sites.5,6 The characterisation of a contaminated fractured site at the scale of single fractures and fracture networks is a prerequisite for the determination of the transport properties of such media.7–9 Contaminant transport in fractured permeable formations is typiﬁed by the interaction of the hydraulic properties of fractures and matrix. In most cases, the permeability of the fracture network is much higher than that of the host rock (matrix), but most of the capacity for storing a pollutant is provided by the matrix porosity.10 The simulation of the contaminant transport in such formations is based on dual continuum models which separate the heterogeneous formation into two homogeneous media: one representing the fracture system and one representing the matrix.11 Depending on whether the permeability of the matrix is neglected or taken into account, the model is characterised as dual porosity/single permeability (DPSP) or dual porosity/dual permeability (DPDP).12–15 The abovementioned approach minimises the computational requirements for ﬁeld-scale simulations, with the accuracy of the numerical predictions depending strongly on the use of representative up-scaled transport properties8,16 and reliability of the mathematical models.17 Much attention has been focused on the modelling of NAPL migration through fractured and low-permeability media with signiﬁcant matrix porosity, such as clay till. Simulating a scenario with a sandy aquifer situated beneath a fractured clay aquitard showed that the lower aquifer was vulnerable to contamination from DNAPLs leaking on the ground surface; the DNAPL migration pathways towards the aquifer are mainly governed by the fracture aperture, matrix porosity and form of relative permeability curves. Whether or

Improved Risk Assessment of Contaminant Spreading

271

not the DNAPL can enter and persist within the underlying aquifer depends on a variety of factors such as the DNAPL release rate and its composition at the source, the thickness of the fractured clay layer and the diﬀusion/sorption properties of clay.18 The presence of sand lenses intersecting the fractured clay has been found to increase the time required for the non-wetting phase (DNAPL) to migrate through the vertical extent of a clay sequence from a few days to several years.19 Parametric analysis of the inﬂuence of the main fracture characteristics (mean aperture, roughness, and correlation length) on the hydraulic properties of single fractures has shown that the mean fracture aperture is the most important parameter aﬀecting strongly the permeability and entry pressure.20 The drainage/imbibition hysteresis eﬀect increases with increasing capillary number.20

5.5.2

Objectives and Approach of the TRACE-Fracture Project

The overall objective of the EU-funded TRACE-Fracture project (1/2/2000-31/ 1/2003, contract no. EVK1-CT-1999-00013)21–24 was (1) to develop a novel method for the characterisation of fractured media at the scales of single fractures and fracture networks, (2) to develop reliable and predictive phenomenological models that provide the single-phase ﬂow, two-phase ﬂow and solute dispersion eﬀective coeﬃcients of fractured porous media as functions of fracture morphology, ﬂuid rheology and hydrodynamics, (3) to integrate the new phenomenological models into a novel numerical simulator of the macroscopic contaminant transport in fractured underground reservoirs, (4) to integrate the new numerical tool into a generalised methodology of risk assessment and rational design of remedial strategies for contaminated fractured aquifers and (5) to implement the results in two geologically diﬀerent fractured sites contaminated by NAPLs. The general concept of the approach used in the TRACE-Fracture project is illustrated in Figure 5.5.1. Accurate geostatistical properties and reliable eﬀective transport coeﬃcients of fractured soils and rocks are determined at multiple scales ranging from single fractures to fracture networks, by combining properly data from ﬁeld-work and photo-geological analysis with laboratory-scale experiments and computational methods. This information is integrated into an updated numerical simulator of the organic pollutant transport in fractured media (SIMUSCOPP). The simulator is used as a tool for the cost-eﬀective calculation of the spatial and temporal distribution of the saturation of the bulk NAPL in unsaturated/saturated zones, and concentration of NAPL compounds in groundwater. Long-term numerical predictions of the chemical status of groundwater under various scenarios of pollution that simulate the site contamination history are coupled with additional information (e.g. exposure pathways, potential receptors, toxicity of substances) for the risk assessment of fractured sites contaminated by organic pollutants. Two very diﬀerent highly heterogeneous fractured sites which have been contaminated by

272

Chapter 5.5 Fi Fieldwork on site & llab-scale experiments multig eological multi- scale geological informat ion

History matching of ma experiments expe

Risk Ri assessment of of f fractured sites

Por Pore-network si simulators

Dual Dual-porosity macroscopic simulator

Figure 5.5.1

Concept of TRACE-Fracture project.

the waste oils of industrial facilities were investigated to assess the risks threatening human health and the ecosystem: one site overlying clay till sediment and situated in an urban area of Denmark (Ringe site) and one site overlying granite rock and situated in northern Spain (Spanish site). The results of the project might be helpful in formulating protocols for the risk assessment of fractured contaminated sites, designing cleanup strategies and setting the boundaries of the protection zones for aquifers underlying fractured areas. In the following, our attention is focused on the Ringe site, whereas the application of the methodology to the Spanish site has been published elsewhere.15

5.5.3

Description of Ringe Site

The investigated Ringe site is situated in an abandoned asphalt and creosote factory at Ringe, on Funen island, Denmark. The subsurface consists of clay till overlaying a primary sandy aquifer. The clay till is approximately 8–12 m thick and is dissected by multiple fractures, which are responsible for the transport of pollutants through a normally tight layer. The site was contaminated by leaking storage tanks of a creosote and asphalt factory (1929–1962) as well as by waste oils of several companies and automobile workshops (1962–1988). In 1988, it was discovered that the subsurface of the site was strongly contaminated by creosote, various compounds of which were traced in the aquifer, 22 m below ground, and a long-term site remediation programme was adopted by the Danish Environmental Protection Agency. The site was used for ﬁeld studies in a number of research projects (1994–2003): a great number of wells were established and four open pits were excavated. A geological model was developed (Figure 5.5.2), and numerous investigations were performed to collect data concerning the hydraulic and solute transport

Improved Risk Assessment of Contaminant Spreading

Figure 5.5.2

273

Macropore distribution on a representative region of the Ringe site.

characteristics of clayey fractured sediments.25,26 It is worth mentioning that clay till is a sediment that is widespread in the subsurface of regions covered by ice during earlier glacial times, such as Canada, the USA, northern Europe, as well as all Alpine and Alpine marginal areas.27

5.5.3.1

Conceptual Fracture Network Model

The upper 5 m of till may be separated into three zones diﬀering with respect to the characteristic distribution of pores and fractures (Figure 5.5.2): (i) an upper zone (0.5–2.5 m below ground surface) dominated by bio-pores (burrows), desiccation fractures and a highly porous matrix; (ii) a central zone (2.5–4 m

274

Chapter 5.5

below ground surface) dominated by well-connected desiccation and glaciotectonic shear fractures; and (iii) a lower zone (44 m below ground surface) dominated by glaciotectonic shear fractures. Three distinct fracture systems as well as a number of randomly oriented fractures were recognised in the upper layer (0–5 m below ground surface) (Figure 5.5.2). Systematic measurement of the aperture of desiccation and glaciotectonic fractures was done by analysing 2D BSEM images of resin-impregnated samples (Figure 5.5.3). It was revealed that the fracture aperture resembles a 2D network of elliptical channels and such a model was employed in experimental and theoretical approaches to estimate the single- and multiphase eﬀective transport coeﬃcients at the scale of a single fracture. Based on the foregoing information and statistics of fracture intensity/ spacing, a conceptual model of the fracture network/porous matrix system was established (Figure 5.5.4). The weathered microporous matrix is dominant

Sample: SF-2D

1000 µm

Figure 5.5.3

BSEM image of resin-impregnated single fracture. The 2D aperture area was digitized with ScanPro 5.0 and appears dark. Matrix Matr zones

Clay Till High porosity CaCO3-poor

Clay Till Medium porosity CaCO3-rich

Macro-pore zones Borrows and rootholes

Desiccation fractures and tectonic fractures

Tectonic fractures

Sand

Figure 5.5.4

Fracture network/porous matrix model of Ringe site.

Improved Risk Assessment of Contaminant Spreading

275

in the upper zone (o2.5 m below ground surface). In this zone, the limestone was dissolved thus leaving a highly porous matrix with a porosity ranging from 0.3 to 0.45 (Figure 5.5.4). The unweathered matrix is dominant in the lower layers, and consists of ﬁrmly consolidated clay till with a weight fraction of limestone close to 0.25–0.3 and porosity ranging from 0.25 to 0.30 (Figure 5.5.4). In depths between 3.5 m and 15 m sand-lenses of higher permeability appear.

5.5.4

Hierarchical Methods for the Determination of Transport Properties

Experimental techniques and numerical methods were developed for the determination of the single-phase transport properties of fractured media. With the aid of the critical path analysis (CPA) of percolation theory and eﬀective medium approximation (EMA), accurate phenomenological models were developed to relate explicitly the absolute permeability and electrical formation factor of single fractures with microscopic properties of their aperture.28 The results of the geological characterisation (Figure 5.5.4) were employed for the computer-aided construction of networks of glaciotectonic and desiccation fractures intersecting the clay till29 (Figures 5.5.5(a)–(d)). The up-scaled transport properties of fractured clay till (Figures 5.5.5(a)–(d)) were calculated by using hierarchical pore network simulations (Table 5.5.1) and were found to be comparable to results of ﬁeld- and laboratory-scale hydraulic tests (Table 5.5.2).25,26 Phenomenological models were developed to relate the ﬂow velocity of inelastic shear-thinning NAPLs (e.g. asphalt, emulsions of creosote with water, crude oil) in single fractures and fractures embedded into porous matrices with NAPL rheology (e.g. parameters that specify the viscosity as a function of shear rate) and fracture aperture properties.30,31

5.5.4.1

Multiphase Transport Coeﬃcients of Single Fractures and Fracture Networks

Experimental procedures and numerical methods were developed to determine the two-phase ﬂow properties (capillary pressure and relative permeability curves) of fractured media. Network-type hierarchical simulators of the immiscible displacement of an aqueous phase by a NAPL in single fractures, single fracture embedded into porous matrix and fracture networks were developed to compute the up-scaled relative permeability and capillary pressure curves29 (Figures 5.5.5(e) and (f)). Visualisation experiments of drainage and imbibition performed on glass micromodels revealed that the transient immiscible displacement growth patterns in fractures are aﬀected strongly by NAPL rheology, ﬂow rates and fracture surface wettability32–34 (Figure 5.5.6). With the aid of numerical algorithms of history matching, the capillary pressure (Figures 5.5.7(a) and (d)) and relative permeability (Figures 5.5.7(b) and (e)) curves of porous and fractured media were estimated simultaneously from datasets of unsteady-state displacement experiments and were found strongly

276

Chapter 5.5

–2

–2

–2.5 –3

–2.5 –3

–3.5 –4

–3.5

–4.5 –5 10

–4.5

–4

–5 10

8 6 4

(a)

8

6 2 0 0

2

8

10

6 4

4

2

2

0 0

(b)

10

8

6

4

–2 –2.5 –3 –3.5 –4 –4.5 –5

(c)

0

(d)

regions 1-2-3 region 1 region 2 region 3 region 4

0.01

1E-3 0.0

(e)

Figure 5.5.5

10

9

8

7

6

5

4

3

2

1

0

1.0

Relative Permeability, krw, k rnw

Capillary pressure, Pc (bar)

0.1

5

0.2

0.4

0.6

Water Saturation, Sw

0.8

0.8

0.6

0.4

(f)

regions 1-2-3 region 1 region 2 region 3 region 4

0.2

0.0 0.0

1.0

k rw

k rnw

0.2

0.4

0.6

0.8

1.0

Water Saturation, Sw

Generation of the two fracture systems. (a) Fracture system 1: desiccation fractures. (b) Fracture system 2: tectonic fractures. (c) Upper view and (d) side view of the fracture network; region 1 consists of tectonic (light grey) and desiccation (mid grey) fractures (number ratio ¼ 1); region 2 consists of tectonic (light grey) and desiccation (dark grey) fractures (number ratio ¼ 1/3); region 3 consists of tectonic (light grey) and desiccation (mid grey) fractures (number ratio ¼ 1/6); region 4 consists only of tectonic fractures. (e) Capillary pressure and (f) relative permeability curve for the entire fracture network, and sub-networks corresponding to diﬀerent depth intervals (Table 5.5.1).

277

Improved Risk Assessment of Contaminant Spreading

Table 5.5.1

Calculated fracture network permeability at various depths (Figure 5.5.5).

Permeability (mD)

Three regions

Region 1

Region 2

Region 3

Region 4

Kvertical//gravity Khorizontal//tectonic Khorizontal//desiccation

1073

1372 836 230

775 594 76

538 514 29

372 372 0

Table 5.5.2 Depth (m) 0–2 2–4 4–6 6–13

Figure 5.5.6

Experimental values of the permeability over the various zones of Ringe site. Test Inﬁltration tests Column tests Inﬁltration slug tests Falling head tests Inﬁltration tests

Hydraulic conductivity (m s1) 5

4

1.5 10 1.5 10 3 106 106–1.2 105 3 107 1.2 107

Permeability 1.5–15 Da 300 mD 100 mD–1.2 Da 30 mD 12 mD

Final ﬂuid distribution after the displacement of NAPL by water in a dual-pore network (artiﬁcial glass-etched model of a single fracture surrounded by the clay till porous matrix; the displacement is from left to right): (a) the NAPL is paraﬃn oil (Newtonian ﬂuid); (b) the NAPL is a suspension of ozokerite (natural wax) of concentration 2% in paraﬃn oil (shear-thinning ﬂuid).

sensitive to the ratio of viscous to capillary forces (capillary number) for Newtonian (Figure 5.5.7(c)) and shear-thinning (Figure 5.5.7(f)) NAPLs.33–36 A fully automated technique, based on the high sensitivity of the colour intensity of an aqueous solution to pH, was devised35,37 for performing solute dispersion visualisation experiments on transparent porous media models, and measuring precisely the transient solute concentration proﬁles with image analysis (Figure 5.5.8(a)). Chaotic, buoyancy-driven solute dispersion regimes were identiﬁed38 at the scale of pore network/single fracture (Figures 5.5.8(b)–(d)). An inverse modelling method was developed for the simultaneous estimation of the longitudinal and transverse dispersion coeﬃcients of fractures from transient solute concentration proﬁles measured with miscible displacement and single-

278

Chapter 5.5 Relative permeability, krw krnw

Capillary pressure, Pc (Pa)

3000 2000

1000 900 800 700 600 500

(a)

400 300 0.0

0.2

0.4

0.6

0.8

1.0

Non-wetting phase saturation, Snw Relative permeability, krw krnw

Capillary pressure, Pc (Pa)

1000 900 800 700 600 500 400 300 0.0

(d) 0.2

0.4

0.6

0.8

1.0

Non-wetting phase saturation, Snw

Figure 5.5.7

0.9 0.8 0.7 0.6

krnw

krw

0.5 0.4 0.3 0.2

(b)

0.1 0.0 0.0

0.2

0.4

0.6

0.8

(c) 1.0

Non-wetting phase saturation, Snw

3000 2000

1.0

1.0 0.9 0.8 0.7

k rw

0.6

k rnw

0.5 0.4 0.3 0.2

(e)

0.1 0.0 0.0

0.2

0.4

0.6

0.8

(f)

1.0

Non-wetting phase saturation, Snw

Estimated two-phase (a) capillary pressure and (b) relative permeability curves from (c) transient data of the displacement of an aqueous (dark grey) phase by a Newtonian (light grey) NAPL (paraﬃn oil), at various values of the capillary number. Estimated two-phase (d) capillary pressure and (e) relative permeability curves from (f) transient data of the displacement of an aqueous (dark grey) phase from a shear-thinning (light grey) NAPL (ozokerite 1.5% in paraﬃn oil), at various values of the capillary number.

source solute transport experiments.37 The longitudinal dispersion coeﬃcient increases and tends to be dominated by macrodispersion as the variability of the fracture aperture is enhanced (Figures 5.5.9(a) and (b)) and is correlated with fracture morphology and Peclet (ratio of convective to diﬀusive ﬂux) number.35 Conclusively there is a variety of parameters that should be taken into account in the deternination of the eﬀective transport coeﬃcients of fractured media: (1) the geometry and topology of the aperature of single fractures; (2) the morphology of fracture networks; (3) the rheology of NAPL; (4) the ratio of viscous to capillary forces (capillary number); and (5) the ratio of convective to diﬀusive ﬂux (Peclet number).

5.5.5

Numerical Modelling of NAPL Fate in Unsaturated and Saturated Zones

A macroscopic numerical tool was developed for quantifying the NAPL spreading in fractured clay till sediments and underlying aquifers. The macroscopic

Improved Risk Assessment of Contaminant Spreading

Figure 5.5.8

279

Visualization single source-solute transport experiments performed on a single fracture. A low solute concentration aqueous solution (groundwater/dark colour) ﬂows steadily from the left to the right, whereas a high solute concentration aqueous solution (pollutant/light colour) is injected at a low ﬂow rate through a hole (single source). (a) Steady-state solute dispersion regimes across the single fracture at various values of Peclet (Pe) number (Pe ¼ u0lp/Dm, u0 ¼ pore velocity, lp ¼pore length, Dm¼solute/solvent diﬀusion coeﬃcient) without the action of gravity. (b-d) Buoyancy-driven (chaotic) successive steady-state solute dispersion regimes across a single fracture at various Pe values (the gravity acts vertically to the main ﬂow direction and lobe-shaped instabilities are created by the downward ﬂow of the heavier liquid and the upward ﬂow of the lighter liquid).

simulator SIMUSCOPP (owned by Institut Francais du Petrole) was updated to simulate the contaminant transport in fractured porous media.39 The eﬀective transport coeﬃcients of the clay till fracture systems were up-scaled to the block size of the numerical grid.29 The geological model was transformed to an equivalent macroscopic dual porosity–dual permeability numerical model of the fractured clay till site.29 The spatial and temporal distribution of NAPL saturation in the unsaturated zone of clay till was determined with the aid of SIMUSCOPP for the following pollution scenario: creosote was leaking from tanks for a period of 30 years on a surface of 25 m2, at a constant ﬂow rate of 3 105 m3 per day. The migration pathways of pollutants over the period of 60 years were simulated29 (Figure 5.5.10(a)). The matrix is highly saturated by

280

Chapter 5.5 x

x y

y Solute injection

1

Solute injection

2

1000

(a)

1000

3

DL/Dm=0.04 Pe1.48 100

DL/Dm=0.087 Pe0.9

DL/ Dm

DL/ Dm

2

(b)

DL/Dm=7.4e-4 Pe1.78

100

10

1

DL/Dm=0.21 Pe1.06 10 Dual pore network

1

Single pore network

1

Estimated values Taylor dispersion in large pore Taylor dispersion in small pore

Estimated values Taylor dispersion in a pore

0.1 1

10

100

1000

Pe=up0Lp /Dm

Figure 5.5.9

10000

0.1 1

10

100

1000

Pe=up0Lp /Dm

Longitudinal dispersion coeﬃcient estimated from miscible displacement experiments performed on (a) simple (single fracture) and (b) dual (single fracture surrounded by matrix) pore networks, as function of Peclet number.

water and is bypassed by the NAPL, most of which ﬂows downwards through the fracture network pathways (Figure 5.5.10(a)). Moreover, in the fractured zone, the NAPL is retained in the sand lenses rather than in the clay till, which is fully saturated by water (Figure 5.5.10(a)). In the simple medium simulation, contrary to the dual medium simulation results, NAPL does not reach the aquifer (Figure 5.5.10(b)). Without the presence of the fracture network, an important back ﬂow of NAPL and water occurs at the surface (soil/air interface). The water-saturated and low-permeability matrix acts as a barrier that prevents the downward NAPL ﬂow (Figure 5.5.10(b)). The dual-porosity simulations performed on the cross-section (Figure 5.5.10(a)) provided the NAPL ﬂux towards the aquifer. This ﬂux was used as a boundary condition in the single-porosity SIMUSCOPP simulator to predict the transport of the dissolved compounds in the aquifer. The groundwater pollution at a distance 500 m downstream from the source was examined. The aquifer was modelled as a 2D cross-section 770 m long (500 m downstream and 250 m upstream from the source), and 10 m deep. The spatial and temporal distribution of naphthalene and phenol concentration in the homogeneous sandy aquifer (saturated zone) underlying the fractured clay till zone was determined with the aid of SIMUSCOPP (Figure 5.5.11). The predicted concentrations of naphthalene and phenol

Improved Risk Assessment of Contaminant Spreading

Figure 5.5.10

281

Simulation of the spatiotemporal evolution of NAPL saturation within the unsaturated zone of wet clay till (a rainfall rate equal to 5.5 m3 per year or equivalently 73 mm annual recharge was assumed for the entire surface of 25 m2): (a) simulation by accounting for the presence of fractures; (b) simulation by ignoring the existence of fractures.

in groundwater exceed signiﬁcantly the maximum ones measured on the ﬁeld, because some parameters introduced in the simulator, such as the weight fractions of the compounds in creosote and the rate of NAPL ﬂux on the groundwater table, might be overestimated.

5.5.6

Risk Assessment of Contaminated Sites

The RAGS (Risk Assessment Guidance for Superfund) approach of the US Environmental Protection Agency (EPA),40 focused on human health risk assessment, was selected for implementation in the TRACE-Fracture project. The risk analysis included the following steps: (1) a hydrological model of Ringe site was established; (2) the database containing ﬁeld measurements of the spatial and temporal distribution of the concentration of oil pollutants (PAH, phenols, BTEX) in the subsurface of Ringe site was revised and updated (Table 5.5.3; Figure 5.5.12); (3) the current practices were combined with the numerical predictions of the updated SIMUSCOPP model (Figure 5.5.11); and (4) risk analysis of the threats posed to potential receptors of the contaminated groundwater of clay till site was done. By using ﬁeld measurements (Table 5.5.3) and simulation results (Figure 5.5.11), along with RBCA software, there were found risks associated with the high levels (exceeding the Dutch intervention limits) of several species (BTEX, naphthalene, phenols) concentration in soil and groundwater, at long distances from the source of

282

Chapter 5.5

Figure 5.5.11

Simulation of naphthalene and phenol dissolution/dispersion in the primary sandy aquifer underlying the fractured clay till site.

pollution. The latest available data (Table 5.5.3) indicate that the contaminants have travelled long distances (hundreds of metres) away from the site (Figure 5.5.12). However, the groundwater is used by the local community as a source of drinking water and hence eﬀective methods are to be selected to clean up a volume of about 5000 m3 of contaminated soil.

5.5.6.1

Site Remediation

The following technologies were screened as alternatives to decontaminate the Ringe site: (1) bioremediation, (2) bioventilation (3) enhanced bioremediation, (4) chemical oxidation, (5) soil ﬂushing, (6) thermal treatment, (7) electrical resistance heating, (8) radio frequency/electromagnetic heating and (9) hot air/ steam injection. Moreover, the following technologies were examined as alternatives for groundwater remediation at the Ringe site: (1) enhanced bioremediation, (2) slurry wells, (3) pump-and-treat, (4) bioreactors, (5) sprinkler irrigation, (6) granulated active carbon/liquid-phase carbon adsorption and (7) separation.

0 0 0 0 0 0 0 0 0

0 0.06 0.006 0 0 0.01 0 0 0

0 0 0 0 0 0 0 0 0

0.1

1003 1/1/ 98

0.8 0 0 0

0 0 0 0 0 2 0.06

1/7/ 98

0.9 0 0 0

0.08

1001 1/1/ 98

0 0 0 0 0 0 0 0 0

0 0 0 0 0 3 0

1/7/ 98

0 0 0 0 0 0 0 0 0

0 0.29 0.19 0 0 0 0 0 0 0 0 0 0

1999

0 0 0 0 0 0 0 0 0

0 0 0 0

0

1004 1/1/ 98

0 0 0 0 0 0.008 0 0 0

0 0 0 0 0 5 0

1/7/ 98

0 0 0 0 0 0 0 0 0

0 0 0 0

0

1009 1/1/ 98

0 0.07 0.01 0.08 0 10 0 0 0

0 0 0.8 0 0 7 0.08

1/7/ 98

0 0.07 0.01 0.08 0 10 0 0 0

0 0 0.8 0 0 7 0.08

1/7/ 98

0.5 0.7 0.6 0.2 0.4 0.3 0.3 0.04 0.03

0.5 0 0 8

8

1010 1/1/ 98

Chemical species concentration (mg l1) in groundwater of Ringe site.

Benzene Toluene Xylenes 1,2,3-Trimethylbenzene 1,2,4-Trimethylbenzene 1,3,5-Trimethylbenzene Naphthalene 1-Methylnaphthalene Phenol o-Cresol m-/p-Cresol 2,6-Dimethylphenol 2,4- and 2,5Dimethylphenol 2,3-Dimethylphenol 3,4-Dimethylphenol 3,5-Dimethylphenol Carbazol Quinolin Thiophen Benzothiophen Benzofuran Dibenzofuran Acenaphthylene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz[a]anthracene Chrysene Benzo[b]ﬂuoranthene Benzo[k]ﬂuoranthene Benzo[a]pyrene Dibenzo[a,h]anthracene

Date

Well no.

Table 5.5.3

0.7 0.8 0.75 0.3 0.5 0.5 0.4 0.04 0.05

1100 3 60 3 6 7 3

1/7/ 98

0.8 1 0.1 0 0.07 0 0 0 0

0.15 0 0 1

12

1021 1/1/ 98

0.008 0.01 0 0 0 0.007 0 0 0

200 0 10 2 4 4.5 0.15

1/7/ 98

0.28 0.53 0.33 0.46 0 0 8.1 0.78 0.34 0 4233 0 0 0 0 0 0 0 0 0 0 0

0 0.4 17 0.83 6.5 0.76 2.5 0 0 0 0 17 0.55

1999

0 4152 0 0 0 0 0 0 0 0 0 0 0

21

2000

1100

0 1.4 0.29 0 0 4.3 11 0.56 0

244 0.74 18 0.87 4.7 0 12 0.12 7.9 0 0 150 0.38

1999

9811

7.1 12 4.3 0 0 73 173 3.7 0

1800 12 365 16 88 0 254 0.75 34 0.85 1 450 39

1999

9911

430 150 62 3.4 0 52 128 168 0.58

1800 43 296 10 58 13 1000 43 44 1.7 6.7 430 1200

1999

9912

Improved Risk Assessment of Contaminant Spreading 283

284

Figure 5.5.12

Chapter 5.5

Locations of supply and monitoring wells around Ringe site.

The abovementioned methods of remediation were the output of the screening process included in the RAGS methodology. Once diﬀerent remediation alternatives were screened, they should be evaluated in detail with respect to the speciﬁc characteristics of the Ringe site. The RAGS methodology utilises nine criteria with equal weight fractions to evaluate and compare diﬀerent remediation alternatives (Table 5.5.4). It seems that the most suitable method of soil decontamination might be a thermal treatment combined with an extraction system (Table 5.5.4). Abstraction of groundwater in conjunction with adequate treatment processes (pump-and-treat) is likely the best option to decontaminate the groundwater at the Ringe site (Table 5.5.4). On the basis of the foregoing analysis, the suggested strategy for remediation of the Ringe site is a combination of diﬀerent systems and technologies, including: steam injection or electrical heating of soil; dual-phase vacuum extraction of near surface liquids and vapours; groundwater abstraction to create a hydraulic barrier to prevent additional contaminant migration; and

Bioremediation

+ + + + 2

Overall protection Long-term eﬀectiveness Reduction of toxicity Short-term eﬀectiveness Implementation Cost Score

Soil

Soil ﬂushing + + 2

Chemical oxidation + + +/ 0.5

Evaluation of potential remediation technologies.

Criteria

Table 5.5.4

+ + + + + 4

Thermal treatment and bioslurping + + + 0

Bioremediation

+ + 2

Slurry walls

Underground water

+ + + + + 5

Pumping and treatment

Improved Risk Assessment of Contaminant Spreading 285

286

Chapter 5.5

water treatment plant to treat the abstracted groundwater before reinjecting it to the upstream.

5.5.6.2.

In Situ Stimulation/Remediation of Contaminated Sites

Ringe site is representative of numerous contaminated fractured and lowpermeability sediments/soils covering a major part of the surface, especially in northern and central Europe. Such soil types possess a special problem in relation to the spreading of contaminants into groundwater. The fractures form hydraulic avenues through the otherwise low-permeability clayey sediments/ soils. Traditional remediation technologies used in high-permeability soils (extraction, ventilation, etc.) are primarily based on vertical wells that are installed on the subsurface. Regarding the fractured low-permeability sediments, the problem is that the transport takes place predominantly in the vertical fractures whereas NAPL is accumulated in the impermeable matrix. Therefore any remediation method based on vertical wells is expected to be very ineﬃcient. In order to perform eﬀective in situ remediation of fractured sediments a large number of fractures have to be connected to a well or to a highly permeable sediment layer. The bulk hydraulic conductivity in the fractured sediments may be stimulated either by increasing the fracture aperture and/or the connectivity between fractures and/or the density of the fractures. During hydraulic fracturing, new fractures are introduced into the system and the aperture of the existing fractures is increased due to the uplift of the soil above the fracture. Fracturing is a method whereby a gas (pneumatic fracturing) or water/slurry (hydraulic fracturing) is injected into the subsurface at pressures exceeding the in situ pressure at ﬂow rates exceeding the ﬂow rates corresponding to the natural in situ permeability. The induced fracture itself is commonly a sheet-like feature with maximum dimensions of roughly 20 m and a thickness of 1 to 20 mm depending on the type of injected ﬂuid. Hydraulic fractures are commonly ﬁlled with granular material, which keep the fractures open. Pneumatic fractures are not ﬁlled with granular material and are kept open due to irregularities along the fracture walls. Investigations over the past 15 years in North America have shown that fractures can be created in contaminated, ﬁne-grained sediments, where they increase the ﬂow rates to and from wells by one or two orders of magnitude.41 The technique appears to oﬀer the possibility of signiﬁcantly reducing the costs of remediation of contaminated sites underlain by clay till by increasing the rate at which remediating agents can be introduced into the subsurface and the rate at which contaminated ﬂuids can be extracted. Induced fractures can be established either from vertical wells (most common in groundwater) or from angled/horizontal wells. Hydraulic fracturing is widely used in the petroleum industry where the fractures are created at great depth in rock to improve the productivity of oil

Improved Risk Assessment of Contaminant Spreading

287

wells. It has been shown that hydraulic fractures may be created at shallow depths in sediments to increase their hydraulic conductivity and improve the remediation of contaminated sites.42,43 Most of the environmental applications have been developed by researchers in Cincinnati, Ohio, with the applications conducted in silty and clayey glacial drift similar to the deposits found throughout Scandinavia, the Baltic countries and large parts of Germany, the Netherlands, the UK, Poland and other areas that were transgressed by glaciers during former ice ages. The technique is applicable to the remediation of a wide range of contaminant types, including petroleum hydrocarbons, chlorinated solvents, pesticides and other compounds.44 The properties of hydraulic fractures vary considerably, but many demonstrations have shown that the rate of remediation can be increased by one to two orders of magnitude.42 The technique appears to oﬀer the possibility of signiﬁcantly reducing the remediation costs of contaminated sites underlain by especially silty clay till.

5.5.7

Socioeconomic Relevance and Policy Implications

Liquid pollutants penetrate into the subsurface from industrial waste disposal ponds or municipal waste landﬁlls, leaking underground storage tanks, agricultural chemicals, oil spills, etc. Unavoidably, pollutant inﬁltration through soils leads to the contamination of underground aquifers and the introduction of hazardous chemicals in plant, animal and human tissues. The European Inventory of Existing Chemical Substances lists over 100 000 compounds.45–47 The threat posed by many of these chemical remains uncertain because of the lack of knowledge about their concentrations and the ways in which they move through and accumulate in the environment and then impact on humans and other life forms. For most European countries, underground aquifers remain the main sources of water supply to urban and rural areas. The pollution of water by organic and inorganic wastes of domestic, industrial and agricultural activities in combination with climatic changes, aﬀecting the aquatic cycle, has led to a dramatic reduction of the reserves of ‘‘drinking’’ groundwater. All water, polluted by households, industry or agriculture, returns back one way or another and may cause environmental damage. Over 300 000 potentially contaminated sites have been identiﬁed in Western Europe, and the estimated total number in Europe adds up to 1 500 000.45–47 In Eastern Europe, soil contamination by fuels (e.g. gasoline, diesel, kerosene, crude oil) around abandoned military bases, airports and oil transportation pipes poses the most serious risk. Current practices used by environmental companies for the risk assessment of contaminated sites follow some protocols suggested by environmental agencies. Commonly, the heterogeneous/fractured nature of soils is overlooked in these protocols. The new methodologies developed in the course of the TRACE-Fracture project might be integrated as options into such protocols. The suggested methodology of site characterisation reduces substantially the number of excavations/monitoring wells and the number of ﬁeld-scale/

288

Chapter 5.5

laboratory-scale experiments and tests that are commonly required for characterising highly heterogeneous formations. The methods and approaches developed for fractured clay till are generic and can be extended to (1) any other fractured site with similar geological characteristics, (2) any other numerical simulator that includes the module of dual porosity/dual permeability medium and (3) any type of organic pollutant. The market is full of commercial numerical codes for the simulation of contaminant transport in the unsaturated and saturated zones of the subsurface. However, the main shortcoming of all these simulators is the lack of eﬃcient up-scaling procedures that transform the actual heterogeneous geological models to equivalent numerical models. For this reason, the reliability of the predictions of most macroscopic simulators is questionable. The generalised procedures, developed in the context of the TRACE-Fracture project for the accurate transformation of conceptual geological models into numerical grids, could be useful not only for updating the capabilities of the SIMUSCOPP, but also for improving the predictability of any numerical code used in risk assessment of such contaminated lands (e.g. QUMPFS, TOUGH2, COMPFLOW). The economic impact of the project within the EU is hard to quantify with any reasonable precision. In the USA, the various sources place the clean-up costs of contaminated land, water and structures between $500 billion and $1 trillion over the next 50 years. In Europe, the corresponding ﬁgures are deemed to climb to the same order of magnitude or even higher, given the increased drilling and excavation costs compared to those in the USA. The accurate characterisation of the fractured polluted sites coupled with reliable estimates of pollutant migration can contribute substantially to the design and application of restoration strategies that can deliver safe, reconstituted fractured soils at reduced cost and eﬀort compared to current practices. The number of necessary laboratory-scale experiments and ﬁeld tests will decrease, companies will save money and time and the cost of the decision-making procedures will be reduced substantially. Roughly, with the use of the new methodologies of site characterisation, the cost may be reduced from h1 000 000 to h300 000.

Acknowledgements The TRACE-Fracture project (1/2/2000-31/1/2003) was supported by the European Union under the Energy Environment and Sustainable Development (EESD) sub-programme of the 5th FP (contract number EVK1-CT199900013). I would like to express my thanks to many people who contributed to the TRACE-Fracture project by mentioning just a few of them: Dr M. Theodoropoulou (TEI of Patras and University of Patras, Greece); Dr V. Karoutsos (University of Patras, Greece); Dr K. E. Klint (GEUS, Denmark); Dr P. Gravesen (GEUS, Denmark); Dr C. Laroche (IFP, France); Dr P. LeThiez (IFP, France); Mr L. Molineli (CH2M-Hill, Spain); and Mr F. Sanchez (CH2M-Hill, Spain).

Improved Risk Assessment of Contaminant Spreading

289

References 1. A. S. Goldford, G. A. Vogel and D. E. Lundquist, Waste Manag., 1994, 14, 153. 2. K. D. Pennell, M. Jin, L. M. Abriola and G. A. Pope, J. Contam. Hydrol., 1994, 16, 35. 3. E. L. Madsen, J. L. Sinclair and W. C. Ghiorse, Science, 193, 252, 830. 4. B. R. Keyes and G. D. Silcox, Environ. Sci. Technol., 1994, 28, 840. 5. Y.-S. Wu, K. Zhang, C. Ding, K. Pruess, E. Elmorth and G. S. Bodvarsson, Adv. Water Resour., 2002, 25, 243. 6. K. Zhang and A. D. Woodbury, Adv. Water Resour., 2002, 25, 705. 7. X.-H. Wen and J. J. Gomez-Hernandez, J. Hydrol., 1996, 183, ix. 8. C. T. Miller, G. Christakos, P. T. Imhoﬀ, J. F. McBride and J. A. Pedit, Adv. Water Resour., 1998, 21, 77. 9. G. S. Bodvarsson, Y.-S. Wu and K. Zhang, J. Contam. Hydrol., 2003, 62–63, 23. 10. J. Birkholzer, H. Rubin, H. Daniels and G. Rouve, J. Hydrol., 1993, 144, 1. 11. J. E. Warren and P. J. Root, SPE J., 1963, Sept., 245. 12. F. W. Schwartz and L. Smith, Water Resour. Res., 1988, 24, 1360. 13. R. H. Dean and L. L. Lo, SPE Res. Eng., 1998, May, 638. 14. H. H. Gerke and M. T. van Genuchten, Water Resour. Res., 1993, 29, 305. 15. K. E. Klint, P. Gravesen, A. Rosenbom, C. Laroche, L. Trenty, P. LeThiez, F. Sanchez, L. Molinelli and C. D. Tsakiroglou, Water, Air and Soil Pollution: FOCUS, 2004, 4, 201. 16. B. Bourbiaux, M. C. Cacas, S. Sarda and J. C. Sabathier, Revue de l’IFP, 1998, 53, 785. 17. S. M. Hassanizadeh and W. G. Gray, Adv. Water Resour., 1993, 16, 53. 18. K. J. Slough, E. A. Sudicky and P. A. Forsyth, J. Contam. Hydrol., 1999, 40, 107. 19. D. A. Reynolds and B. H. Kueper, J. Contam. Hydrol., 2001, 51, 41. 20. K. Vandersteen, J. Carmeliet and J. Feyen, Transp. Porous Media, 2003, 50, 197. 21. C. D. Tsakiroglou, TRACE-Fracture, 1st Annual Progress Report, March 2001. 22. C. D. Tsakiroglou, TRACE-Fracture, 2nd Annual Progress Report, March 2002. 23. C. D. Tsakiroglou, TRACE-Fracture, 3rd Annual Progress Report, March 2003. 24. C. D. Tsakiroglou, TRACE-Fracture, Final Report, March 2003. 25. R. C. Sidle, B. Nilsson, M. Hansen and J. Fredericia, Water Resour. Res., 1998, 34, 2515. 26. B. Nilsson, R. C. Sidle, K. E. Klint, C. E. Boggild and K. Broholm, J. Hydrol., 2001, 243, 162. 27. M. Houmark-Nielsen, Bull. Geol. Soc. Denmark, 1987, 36, 1. 28. C. D. Tsakiroglou, Ind. Eng. Chem. Res., 2002, 41, 3462.

290

Chapter 5.5

29. C. Laroche, C. Henrique, S. Bekri, L. Trenty, P. LeThiez, Proceedings of 8th International Conference on Contaminated Soil CONSOIL 2003, Gent, Belgium, 12–16 May 2003. 30. M. Theodoropoulou, V. Karoutsos and C. Tsakiroglou, Environ. Forensics, 2001, 2, 321. 31. C. D. Tsakiroglou, J. Non-Newtonian Fluid Mech., 2002, 105, 79. 32. C. D. Tsakiroglou, M. Theodoropoulou, V. Karoutsos, D. Papanicolaou and V. Sygouni, J. Colloid Interf. Sci., 2003, 267, 217. 33. C. D. Tsakiroglou, M. Theodoropoulou and V. Karoutsos, AIChE J., 2003, 49, 2472. 34. C. D. Tsakiroglou, J. Non-Newtonian Fluid Mech., 2004, 117, 1. 35. C. D. Tsakiroglou, M. A. Theodoropoulou, V. Karoutsos and D. Papanicolaou, Water Res. Resour., 2005, 41, W02014. 36. M. A. Theodoropoulou, V. Sygouni, V. Karoutsos and C. D. Tsakiroglou, Int. J. Multiphase Flow, 2005, 31, 1155. 37. M. Theodoropoulou, V. Karoutsos, C. Kaspiris and C. D. Tsakiroglou, J. Hydrol., 2003, 274, 176. 38. C. D. Tsakiroglou, M. Theodoropoulou and V. Karoutsos, Oil Gas Sci. Technol. La Revue de l’IFP, 2005, 60, 141. 39. L. Trenty, C. Henrique, C. Laroche, P. LeThiez, Proceedings of 8th International Conference on Contaminated Soil CONSOIL 2003, Gent, Belgium, 12–16 May 2003. 40. US EPA, Risk Assessment Guidance for Superfund, Vol. I, Human Health Evaluation Manual (Part A), Interim Final, USEPA/540/1-89/002, 1989. 41. US EPA, Risk Reduction Laboratory and the University of Cincinnati, Hydraulic Fracturing Technology: Applications Analysis and Technology Report, USEPA/540/R-93/505, 1993. 42. L. C. Murdoch, W. Slack, B. Siegrist, S. Vesperm and T. Meiggs, Am. Soc. Civ. Eng., 1997, 10A. 43. L. C. Murdoch, M. C. Kemper, M. Narayanaswamy, A. J. Wolf, Demonstration of Hydraulic Fracturing to Facilitate Remediation, HWRIC RR-068, Illinois DNR Hazardous Waste Research & Information Center, Urbana-Champaign, 1997. 44. US EPA, Alternative methods for ﬂuid delivery and recovery, EPA/625/ R-94/003, 1994. 45. EEA, Europe’s Environment: The Second Assessment, European Environment Agency, Luxembourg, 1998. 46. EEA, Environment in the European Union at the Turn of the Century: Summary, European Environment Agency, Copenhagen, 1999. 47. EEA, Environment in the European Union at the Turn of the Century: Appendix to the Summary. Facts and Findings per Environmental Issue, European Environment Agency, Copenhagen, 1999.

CHAPTER 5.6

Groundwater Risk Assessment at Contaminated Sites (GRACOS): Test Methods and Modelling Approaches PETER GRATHWOHLa AND HANS VAN DER SLOOTb a

Centre for Applied Geoscience, Universita¨t Tu¨bingen, Sigwartstrasse 10, DE-72076 Tu¨bingen, Germany; b Energy Research Centre (ECN), P.O. Box 1, NL-1755 Petten ZG, The Netherlands

5.6.1

Introduction

All over Europe there are numerous contaminated sites where top layers of soil or other materials are contaminated. Remediation of all these sites would be economically impossible and therefore site prioritisation is needed based on an evaluation of the risk to subsoil and groundwater contamination. The deﬁnition of ‘‘groundwater risk assessment’’ as used here is the risk of a compound migrating from a contaminated source in the unsaturated zone to the groundwater. The goal is usually to predict the contaminant concentration at diﬀerent points of compliance, which include (1) the boundary between disposed materials and the natural subsurface environment, (2) the transition zone between groundwater and vadose zone below disposed or contaminated materials and (3) the groundwater downstream of a contaminated zone (often the property boundary). The European Union (EU) project GRACOS and similar projects have addressed these issues and the relevant processes as depicted in Figure 5.6.1. For the release of inorganic (major, minor and trace elements) and organic constituents from contaminated materials into water, it is important to identify a few key issues: the nature of the constituents of concern, which may have very diﬀerent release behaviours based on their chemistry and the local conditions; 291

292

Chapter 5.6 Groundwater Risk Assessment in Case of Contaminated Soil / Contaminated Soil Air Source zone: The contaminant concentration in a soil leaching test or in soil air is above the legal limit !!! → Is the concentration in the groundwater above the legal limit ???

Contaminant Concentration

Depth (unsaturated zone)

Groundwater recharge (A)

Soil air sampling CO , CH

Column test

Possible scenarios: (A )

Residual oil phase petroleum hydrocarbons

(B)

B) No recharge (immobile Contaminants)

(C)

D)

(D)

datio

n (A+

A) Contaminant transport by seepage water C) no recharge: vapor diffusion only

egra

O -Diffusion

(E)

Groundwater recharge: Persistent contaminants can reach the groundwater No recharge: Nonvolatile compounds are immobile Volatile compounds, no recharge: Peristent compounds can reach the groundwater by vapor diffusion Biodegradable compounds may reach the groundwater in very low concentrations (below legal limit ?) Further attenuation in the shallow plume due to volatilization and biodegradation ???

Biod

D) Biodegradation ??? CO

With

E) Volatilization

from

biodegradation

O -Supply Capillary fringe

Shallow plume E) Biodegradation ???

Aqueous concentration

Contaminant transport across the capillary fringe by diffusion and transverse dispersion

Contaminant volatilization across the capillary fringe downgradient from the source zone

Contaminant source zone

Groundwater

Oxygen transport across the capillary fringe aerobic biodegradation of hydrocarbons

Downgradient attenuation zone Distance from source zone

Figure 5.6.1

The GRACOS scenario.

the major release mechanisms in soil and soil-like materials, which will usually be dominated by percolation or (vapour) diﬀusion; the hydrology of the site under consideration, i.e. how much inﬁltration will occur and what are the relevant/preferential ﬂow paths; the nature of the contamination source in terms of release controlling parameters such as pH, EC, redox conditions and dissolved organic carbon (DOC); the changes in release conditions with time due to processes such as depletion of source material and shifts in the main release controlling parameters; the biodegradation/natural attenuation of released constituents in the subsoil and in the groundwater; and the targets to be evaluated for the consideration of proper mitigating measures, including obtaining information on the allowable release rates post-remediation and the applicable monitoring strategies (whether active remediation took place or not).

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

293

Almost all of these questions can be traced back to a quantiﬁcation of the release rates and/or aqueous concentrations of inorganic and organic contaminants at a speciﬁed point of compliance. Obviously, judgment based on potential leaching is more appropriate for assessing long-term impacts of a contaminant release than evaluations based on total composition (i.e. contaminant concentration in solids). Consequently, in material testing and soil contamination analysis, a paradigm shift from concentrations in solids towards concentrations in the aqueous phase has occurred during the last decade. For these tests, the main question is how to achieve a reliable groundwater risk assessment without spending too much money, so that eﬀorts are deemed economically viable. The economics behind these studies are very relevant because of the large material streams involved. These streams include releases from contaminated soils, demolition wastes, dredged sediments and waste materials from mining and industry. Since test applications depend on compound properties, two types of contaminant classes have been distinguished in GRACOS: volatile contaminants which travel in the gas phase in the unsaturated zone (e.g. volatile solvents and gasoline constituents); and non-volatile compounds which more or less exclusively migrate with the seepage water (e.g. most inorganic contaminants, heavy metals and high molecular weight organic compounds). For the assessment of volatile compounds, vapour-phase monitoring and Henry’s law constant allow for the derivations of in situ aqueous concentrations. For non-volatile and ionic compounds, aqueous leaching tests are needed to determine the aqueous concentrations. The subsequent sections will specify and present key conclusions based on selected datasets form the GRACOS project. More detailed information is available from www.uni-tuebingen. gracos.de and from the numerous GRACOS publications, of which the most important ones are listed as Refs. 1–4.

5.6.2

Leaching Tests (Heavy Metals, Low-Volatility Organic Compounds)

5.6.2.1

Total Composition vs. Aqueous Concentrations

Total composition (i.e. concentrations based on dry solids) is inadequate for most environmental assessment purposes because for many of the constituents a signiﬁcant fraction of their total content is essentially non-leachable by water. This potential contaminant release into water can be subdivided into (1) potentially leachable, which is a maximum amount leached under predeﬁned worst-case conditions, and (2) actual leached amount, which is the amount leached under the conditions imposed by the material itself. In Figure 5.6.2, the leaching behaviour of cadmium from an agricultural soil heavily contaminated from a sewage sludge application is shown for illustration. It

294

Chapter 5.6 1000 ANC mode Duplicate INGESTION INHALATION

Leached at L/S=10 (mg/ kg)

100

SCE

ACIDIC ENVIRONMENTS PLANT UPTAKE

NaNO3 Hac CaCl2 CEN

NATURAL SOIL

10

1

SCE2 Total

SOIL LIMING

Cd

0.1

CEMENT STABILIZATION OF CONTAMINATED SOIL

0.01 1

3

5

7

9

11

13

pH

Figure 5.6.2

Relevant pH domains for assessing diﬀerent questions in relation to diﬀerent types of impact (L/S, liquid/solid ratio).

demonstrates that diﬀerent test methods can be placed with respect to one another by plotting such test data as a function of pH. This even applies to the test data obtained from sequential chemical extraction schemes, provided the data are calculated as cumulative leached amounts in subsequent leaching steps. It is recommended that site evaluations be based on leaching rather than on total composition analysis, which is currently being employed by regulations. The main advantage of judgment based on leaching vs. total composition is that the assessment is taking place on the true aspects causing environmental impact. This is because the non-leachable constituents will not cause confusion or controversy, since they would not show up in a leaching test.3–6

5.6.2.2

Percolation vs. Batch or Shaking Tests: Comparison to Field

Although batch shaking tests are very popular in environmental analysis of contaminated materials, it has become obvious that for the assessment of long-term contaminant behaviour, dynamic tests closer to natural conditions

295

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

are needed. Column leaching tests are important laboratory techniques commonly used for the determination of desorption or dissolution rates of (mobile) contaminants from various materials (soils and sediments, mining wastes, recycling and construction materials, demolition waste, etc.). Critical parameters in leaching tests are the contact times and the mass transfer rates into the aqueous phase. Extended contact times result ultimately in equilibrium conditions between the solid phase and the water. Equilibrium concentration levels may depend on dissolved organic matter and colloids and for heavy metals additionally on pH and redox conditions. At equilibrium, results from diﬀerent tests are interrelated as shown in Figure 5.6.3, provided that diﬀerent test conditions have no inﬂuence on other parameters that determine the aqueous concentrations of the target compound. For example, pH, DOC and turbidity (the amount of suspended particles) typically depend on the liquid/solid (L/S) ratio used in the test (at high L/S ratios in shaking tests DOC may be diluted whereas turbidity may increase). The equilibrium concentration in the aqueous phase (Cw) depends on the L/S ratio and the sorption capacity of the sample for a speciﬁc compound: Cw ¼

Cs;ini Kd þ LS

ð1Þ

where Kd is the distribution coeﬃcient (l kg1) deﬁned as the ratio of the concentration in the solids to the concentration in water (Kd ¼ Cs/Cw). Cs,ini 10 LS = 0.25

concentration Cw

1 LS = 2 0.1 LS = 10 0.01

1⋅10

3

0.1

1

10

100

1⋅103

distribution coefficient Kd

Figure 5.6.3

Decrease in water concentration with increasing distribution coeﬃcient Kd at diﬀerent liquid/solids ratios (LS) starting at an initial concentration in the solids (Cs,ini) of 1 mg kg1, mg kg1, etc. The least dilution, meaning almost equal concentrations independent of LS is observed for Kd 4 100 (inverse linear relation between concentration and Kd). The highest concentrations in water are observed for small LS ratios and low Kd values (LS ¼ 0.25 represents approximately the conditions in a column experiment or a natural porous medium; LS then corresponds to the ratio of porosity and bulk density ¼ n/rbulk).

296

Chapter 5.6

denotes the initial concentration in the solids. If Kd is much larger than L/S then the aqueous concentration is independent of L/S. For small values of Kd, increasing L/S ratios cause simply a dilution. When pH changes occur, Kd can no longer be considered constant and other approaches are needed. For inorganic compounds, adsorption phenomena are often strongly nonlinear, and hence their solid/solution distribution cannot be described by a linear, i.e. concentration independent, Kd relationship. In addition, constituents cannot be considered to be transported through soil fully independent of each other. To account for multicomponent eﬀects typical for inorganic constituents, such as competition for adsorption sites, a mechanistic modelling approach is preferable over empirical (Kd) models. Several models are available that take into account relevant processes that inﬂuence the solid/solution partitioning and subsequently the transport rates. These processes include, among others, mineral precipitation, incorporation into solid solutions, sorption onto iron, aluminium and manganese (hydr)oxide surfaces and interaction with particulate (POM) and dissolved organic matter (DOC). Depending on the choice of input parameters, such mechanistic geochemical models allow site-speciﬁc as well as generic model predictions. In addition, the calculation of element partitioning between dissolved, complexed and diﬀerent particulate phases is relevant with respect to potential bioavailability of elements, as it is generally believed that complexed forms are less likely to be taken up by organisms than elements in free ionic form. DOC bound constituents are likely to be transported over larger distances than free forms due to facilitated transport. This may also be true for very poorly water soluble organic micro-pollutants. In Figure 5.6.4, the element Zinc Contaminated soil NL 0.001

Model

0.0001 0.00001 0.000001

Partitioning liquid and solid phase, [Cu+2]

0.001 0.0001

Tenorite

0.00001

Clay

0.000001

FeOxide POM-bound

0.0000001

DOC-bound

0.00000001

Free

0.0000001 1

2

3

4

5

6

7

8

9

10 11 12 13 14

0.000000001 1

Cu+2 fractionation in solution

100%

2

3

4

5

6

7

8

9 10 11 12 13 14

Cu+2 fractionation in the solid phase 100%

Fraction of total concentration (%)

Fraction of total concentration (%)

0.01

Concentration (mol/l)

Concentration (mol/l)

[Cu+2] as function of pH 0.01

80% 60% 40% DOC-bound

20%

80%

Tenorite Clay

60%

Fe Oxide 40%

POM-bound

20%

Free 0%

0% 1

2

3

4

5

6

7

8

pH

Figure 5.6.4

9 10 11 12 13 14

1

2

3

4

5

6

7

8

9

10 11 12 13 14

pH

Prediction of copper release from zinc-contaminated soil illustrating the adequate match between test data (ﬁlled circles; TS14429) and model results (solid curve). Partitioning between dissolved and particulate phases in concentration (top right). Fractionation of copper in dissolved (free and DOC associated) and solid phases (particulate organic matter (POM), iron oxide, clay and speciﬁc minerals).

297

Groundwater Risk Assessment at Contaminated Sites (GRACOS) 100

0.1 0.01 0.001 0.0001 0.00001

Ni

0.000001 0.0000001

10

Concentration (mg / l)

Cumulative release (mg / kg)

1

Leached (mg / kg)

1

100

10

1 0.1 0.01 0.001 0.0001 0.00001

3

5

7

9

11

13

0.00001

10

0.00001

100 10 1 0.1

DOC

0.001 5

7

9

11

13

1000 100 10 1 0.1

1000

100

10

0.01

0.00001

0.001

pH

Figure 5.6.5

10

10000

0.001 3

0.1

10000

Concentration (mg/l)

Cumulative release (mg/kg)

1000

0.001

L/S (l/ kg)

100000

10000

Leached (mg/kg)

0.1

L/S (l/kg)

100000

1

0.001

0.0001 0.001

pH

0.01

0.01

0.000001

0.0000001 1

0.1

0.1

L/S (l/kg)

10

1 0.00001

0.001

0.1

10

L/S (l/ kg)

Relationships between pH dependence and percolation tests (laboratory) for a mixture of wastes (integral mix largely consisting of contaminated soil, sediments and soil cleaning residues; ﬁlled diamonds) disposed in a 12 000 m3 pilot cell with leachate data from 1.5 m3 lysimeters (triangles) and the full-scale pilot (ﬁlled circles).

distribution between dissolved and particulate phases obtained with a mechanistic geochemical modelling approach is shown for a zinc-contaminated soil. A detailed description of this modelling approach is referred to in Dijkstra et al.7 Figure 5.6.5 shows that results from column percolation tests and ﬁeld lysimeter measurements can be related to ﬁeld-scale data, if the sampling and testing are done in an appropriate manner. Under ﬁeld conditions, the relationship between time, inﬁltration and L/S can be calculated simply based on the dry bulk density (rbulk [kg m3]), the height of the percolated material of interest (h [m]) and the net inﬁltration rate (N [mm a3]):

t ½a ¼

LSrbulk h N

ð2Þ

The consistency between the data from diﬀerent scales of testing (laboratory to pilot to ﬁeld scale), as shown in Figure 5.6.5, demonstrates that the underlying processes are suﬃciently understood and can be quantiﬁed. In addition, this observation indicates that appropriate sampling procedures and mixing of samples to constitute a representative leachate sample lead to more meaningful

298

Chapter 5.6

results than the often reported scattered data from individual sub-samples obtained in the ﬁeld or laboratory. Using the new chemical speciation/transport tool LeachXS-Orchestra,8,9 the percolation test data for the integral waste mix consisting largely of contaminated soil, sediments and soil cleaning residues were modelled (Figure 5.6.6). The mineral assemblage and the iron and aluminium oxide sorption parameters were determined from modelling the pH dependence leaching test results of the same mix. The DOC release is assumed to be a continuous decay function. In the case of the lysimeter and the ﬁeld data graph showing preferential ﬂow, the 25% of total mass involved in leaching accounts for the diﬀerence from the percolation test. The simultaneous modelling of some 25 major, minor and trace elements is ambitious and the modelling results are quite promising. [Na+] as function of L/S 1.0E+00

Concentration (mol/l)

pH

pH as function of L/S 8.5 8.3 8.1 7.9 7.7 7.5 7.3 7.1 6.9 6.7 6.5 0.0001

0.001

0.01

0.1

1

10

1.0E-01

1.0E-02

1.0E-03

1.0E-04 0.0001

100

0.001

0.01

L/S (l/kg) 1.0E-01

Concentration (mol/l)

Concentration (mol/l)

1.0E-02

0.001

0.01

0.1

1

10

1.0E-03

0.001

0.01

Concentration (mol/l)

Concentration (mol/l)

L/S

Figure 5.6.6

10

100

[Zn2+] as function of L/S

1.0E-07

0.1

1

1.0E-03

1.0E-06

0.01

0.1

L/S

[Cu2+] as function of L/S

0.001

100

1.0E-02

1.0E-04 0.0001

100

1.0E-05

1.0E-08 0.0001

10

[SO4 ] as function of L/S

L/S 1.0E-04

1

-2

[Ca2+] as function of L/S 1.0E-01

1.0E-03 0.0001

0.1

L/S

1

10

100

1.0E-04 1.0E-05 1.0E-06 1.0E-07 1.0E-08 1.0E-09 0.0001

0.001

0.01

0.1

1

10

100

L/S

Full mechanistic modelling (open diamonds) of percolation tests (laboratory) for a mixture of wastes (integral mix largely consisting of contaminated soil, sediments and soil cleaning residues; ﬁlled circles) disposed in a 12 000 m3 pilot cell in comparison with leachate data from a 1.5 m3 lysimeter (triangles) and the full-scale pilot (squares).

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

299

Apart from the prediction of the release as a function of L/S (or time), information on the partitioning within the column as a function of time at a speciﬁed depth or as a function of depth at a speciﬁed time can be obtained without any extra eﬀort. This provides further insight into the processes occurring within the column, which may not yet be apparent in the outﬂow. In Figure 5.6.7, results are shown for Ca, Cu and Mo. In the case of Ca, several minerals are of relevance; for Cu, organic matter dominates the release; for Mo, PbMoO4 and Fe are controlling phases. In both cases of Ca and Mo, dissolved concentrations locally increase within the column as a result of local conditions. This illustrates the kinetics between all mutually interacting inorganic constituents.

5.6.2.3

Boundary Conditions Leading to Changing Release Rates

Concentrations of many compounds, especially heavy metals, often depend strongly on pH and DOC, as shown in Figure 5.6.4 for an example of an inorganic constituent. Besides the complexation of heavy metals, DOC also causes solubilisation of organic compounds which decreases the sorption coeﬃcients and increases leaching. The solubilisation factor (corresponding to the decrease in sorption or Kd) is given by S 0 ¼ 1 þ fDOC KDOC

ð3Þ

where fDOC and KDOC denote the fraction of dissolved organic carbon in the aqueous phase and the organic carbon normalised distribution coeﬃcient (l kg1). Figure 5.6.8 shows the inﬂuence of increasing DOC values on distribution patterns of polycyclic aromatic compounds (PAHs) leaching from demolition waste. Low-solubility compounds are aﬀected most but they do not contribute much to the sum of the 16 EPA PAHs, if the DOC concentration stays below 30 mg l1. It should be noted that in both cases—heavy metals and organic compounds—DOC can cause signiﬁcantly enhanced leaching and in many ﬁeld cases is probably more important than particle facilitated transport. In contrast to suspended particles, DOC cannot be ﬁltered (in the laboratory or in the ﬁeld) and therefore is an important parameter if present in suﬃcient concentrations (430 mg l1).

5.6.2.4

Release Kinetics

So far, the considerations have been concerned with contaminant release under equilibrium conditions, which may not always be the case. There are generally two types of non-equilibrium: chemical non-equilibrium and physical nonequilibrium. Chemical non-equilibrium in a leaching test may occur when the kinetics of mineral dissolution/precipitation and/or sorption processes are slow relative to the contact time with the water. Physical non-equilibrium indicates the mass transfer of contaminants from a sorbed or solid state to mobile water by ﬁlm diﬀusion and intraparticle diﬀusion.

Figure 5.6.7

0.001

0.01

6.67 DOC-bound

depth (m)

10.00

13.33

3.33

POM-bound PbMoO4[c]

6.67

depth (m)

10.00

13.33

Concentration profile for Ca+2 at 3 cm

POM-bound Calcite time (days)

DOC-bound time (days)

Free FeOxide

POM-bound PbMoO4[c]

time (days)

0.000000001 0.00 2.00 4.06 6.15 8.23 10.2912.3814.4616.54

0.00000001

0.0000001

0.000001

0.00001

0.0001

Concentration profile for MoO4-2 at 3.3 cm 0.001

Free POM-bound

Concentration profile for Cu+2 at 3.3 cm 0.01 0.001 0.0001 1E-05 1E-06 1E-07 1E-08 1E-09 1E-10 1E-11 0.00 2.00 4.06 6.15 8.23 10.29 12.38 14.46 16.54

Free Anhydrite

0.0001 0.00 2.00 4.06 6.15 8.23 10.29 12.38 14.46 16.54

0.001

0.01

0.1

1

10

Concentration profile for Ca+2 at 10 cm

POM-bound Calcite

time (days)

Concentration profile for MoO4-2 at 10cm

Free FeOxide

POM-bound PbMoO4[c]

time (days)

0.0000001 0.00 2.00 4.06 6.15 8.23 10.2912.3814.4616.54

0.000001

0.00001

0.0001

0.001

Concentration profile for Cu+2 at 10cm 0.01 0.001 0.0001 1E-05 1E-06 1E-07 1E-08 1E-09 1E-10 1E-11 0.00 2.00 4.06 6.15 8.23 10.29 12.38 14.46 16.54 Free DOC-bound POM-bound time (days)

Free Anhydrite

0.01 0.00 2.00 4.06 6.15 8.23 10.29 12.38 14.46 16.54

0.1

1

10

Predicted partitioning between dissolved and particulate phases in the laboratory column of integral waste consisting of contaminated soil, sediments and soil cleaning residues as a function of depth at a speciﬁed time (15.5 days) and as a function of time at a speciﬁed depth (3 and 10 cm from the inﬂow into the column) for Ca, Cu and Mo as representatives for the behaviour of a major element, a metal and an oxyanion.

Free FeOxide

0.000000001 0.00

0.00000001

0.0000001

0.000001

0.00001

0.0001

Concentration profile for Mo after 15.5 days 0.001

1E-11 0.00 3.33 Free POM-bound

1E-10

1E-09

1E-08

1E-07

1E-06

1E-05

6.67 10.00 13.33 alpha-TCP depth (m)

Concentration profile for Cu after 15.5 days

3.33 POM-bound Calcite

Concentration profile for Ca after 15.5 days

0.0001 0.00 Free Anhydrite

0.001

0.01

0.1

1

0.0001

Concentration (mol/l)

Concentration (mol/l)

Concentration (mol/l)

Concentration (mol/l)

Concentration (mol/l) Concentration (mol/l)

Concentration (mol/l) Concentration (mol/l) Concentration (mol/l)

10

300 Chapter 5.6

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

301

100 0

3

10

20

50

Solids

concentration µg l-1

10

1

0. 1

Figure 5.6.8

C hr Bb F Bk F Ba In P de no D ah A Bg hi P Su m

Py Ba A

n Ph e An t Ft h

Fl

A

ny Ac e

0. 01

Distribution pattern of 16 EPA PAHs in aqueous leachate concentrations of soils. With increasing DOC (0–50 mg l1) the aqueous phase concentrations shift towards the distribution pattern of the solids (righthand bar for each PAH). Note that low molecular weight PAHs up to anthracene (Ant) are not much aﬀected and that more than 20 mg l1 is needed to cause a signiﬁcant increase in the sum of the 16 EPA PAHs.

Small grains in both cases cause higher release rates because of the higher surface area present per unit volume of porous media. Intraparticle diﬀusion is relevant for porous materials such as clay aggregates, rock fragments, concrete in demolition waste and construction products, slag, etc., i.e. the most abundant soil and waste materials of interest. Intraparticle diﬀusion limits mass transfer increasingly with time because of decreasing concentration gradients inside the particle during leaching. Initially, concentration gradients are steep leading to high release rates and equilibrium is established after short ﬂow distances in the contaminated material. Therefore, most column experiments (and percolation in the ﬁeld) start with equilibrium conditions which sooner or later shift into nonequilibrium. The lag time towards non-equilibrium depends on grain sizes (longer for ﬁne-grained materials or for a high fraction of ﬁne-grained material in a heterogeneous, coarse sample) and the sorption capacity (expressed as Kd). High sorption capacity causes extended periods of equilibrium contaminant leaching and therefore higher solubility (low sorption) compounds leach faster than low solubility compounds in a chromatographic manner (shown for PAHs in Figure 5.6.9). Diﬀusion-limited release causes concentrations and release rates to decrease with the square root of time: in this domain the release rates are independent of the ﬂow rate. If intraparticle porosities and tortuosities are known, the leaching of a series of compounds can be predicted based on the relative water solubilities or relative Kd values of the compounds, as shown in Figure 5.6.9.

302

Chapter 5.6 100

Concentration in water [µg l-1]

Demolition waste 10

1

0.1 Nap Ace 0.01 Phe Fth

Comparison of measured and numerical modelled datas from column test

0.001 0.1

1

10

100

1000

time [d]

Figure 5.6.9

Long-term release of PAHs (Nap, naphthalene; Ace, acenaphthene; Phe, phenanthrene; Fth, ﬂuoroanthene) from demolition waste. Equilibrium concentration plateau followed by diﬀusion limited release (tailing); symbols denote data and curves are numerical simulations based on the spherical intraparticle diﬀusion model.

The time needed for equilibration in a leaching test increases with increasing L/S ratio (at high L/S more compound has to diﬀuse from the solid to the aqueous phase until the equilibration concentration is reached), therefore column tests (L/S o 0.25) more likely yield equilibrium concentrations initially than typical batch shaking tests (i.e. L/S¼10). Figure 5.6.9 also shows that, in principle, only a short-term column test (one day) is needed to determine the equilibrium concentration. This may prove of relevance for the concentration which is to be found under ﬁeld conditions shortly after placement. If information about the long-term dynamics of leaching is required, than an ongoing short-term column test can be simply continued beyond one day. Experience shows that the eﬀort for a short-term column test actually is not greater than for shaking tests, and at the same time the results are more precise and of better reproducibility.

5.6.2.5

Column Tests: Standardisation and Design

In CEN TC 292 ‘‘Characterisation of waste’’ and in ISO TC 190 ‘‘Soil’’ there have been developed standardised protocols for waste (CEN TS TS 1442910 and 1440511), soil and soil-related materials, such as sediments (ISO 21268-312 and 413). Currently, the same methods are proposed for characterisation of granular construction materials (CEN TC 351). This constitutes a horizontal (across diﬀerent sector) approach to standardisation as opposed to a materialspeciﬁc test development. For a broad range of inorganic constituents, leaching

303

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

kinetics have been evaluated recently in pH-dependence leaching tests for MSWI bottom ash14 and percolation tests.15 Column tests can in principle be ﬂexible in their dimensions. For practical reasons largely related to the particle size of the granular material to be tested, column dimensions of 5–10 cm diameter and a length of 15–30 cm have been proposed (resulting in 0.5 kg to several kilograms solids in the column depending on density). Generally, the percolation is run to reach a certain liquidto-solid (L/S) ratio. For quality control purposes (compliance testing) there is no need to go to high L/S. Under such conditions suﬃcient eluate for analysis can be obtained already after L/S ¼ 2:1 within one or two days. These dimensions allow for investigation of many materials that are received in the laboratory without too much pretreatment (grinding, etc.). This also has a positive eﬀect on sampling requirements, which can be less severe than in the case of analysis of composition on relatively small laboratory samples. Figure 5.6.10 shows the typical set-up of a column leaching test. One limitation of column tests is given for ﬁne-grained clayey samples which in compacted form have very low permeabilities: such materials can be percolated in a mix with coarse sand, as aggregates or as compacted bodies with ﬂow around them, as this is how they often occur in the ﬁeld. Note that the packing density does not inﬂuence the equilibrium concentration. Prerequisites for reproducible results are representative samples and a reasonably homogeneous contaminant distribution in the sample. Recently, very reproducible results were obtained for the leaching of many inorganic constituents from MSWI bottom ash, a relatively heterogeneous material.15 Usually columns have a suﬃcient self-ﬁltration capacity yielding turbidities in the eﬄuent that are too low to inﬂuence the leaching of organic compounds. It is important to note that since artefacts are very likely to occur with custom ﬁlters (such as cellulose or glass ﬁbre ﬁlters) and since DOC cannot easily be removed, it is recommended for the release of organic compounds that only centrifugation methods be applied for when turbidity in the eﬄuent is too high (i.e. 410 FNU). Filtration is only required for inorganic compounds if suspended particles are comprised of target metals such as iron-containing

storage tank ss tubing

glass column

(deionized pure water)

(diameter 6cm, length 16cm, volume 450ml)

PVC-tube soil sample filter (quartz sand)

peristalticpump

solvent e.g. cyclohexane for extraction of organic compounds

column effluent

Figure 5.6.10

Typical set-up of a column leaching test (see also CEN TC 29216 ‘‘Characterisation of waste’’ and ISO TC 19012 ‘‘Soil percolation tests’’).

304

Chapter 5.6

mica or clay minerals. For percolation, water free of the target compound is needed. Changes in the redox conditions or biodegradation need to be avoided by removing major electron acceptors such as oxygen, nitrate or sulfate by stripping and deionising of the water. Light-induced bioactivity can be avoided relatively simply by wrapping a column set-up (e.g. in aluminium foil). Measurements of parameters besides the contaminants, such as pH, EC, DOC and turbidity (and Eh), are necessary when further data interpretation, modelling and data comparison is foreseen.

5.6.2.6

Concluding Remarks for Leaching Tests

Most factors that inﬂuence the release of contaminants from a contaminated site can be assessed through aqueous leaching tests. Although a wide variety of leaching tests is available in the literature, only a limited set of tests suﬃce to address key processes and inﬂuencing factors. For a largely percolationdominated situation, as in most real-world cases, a column leaching test is the most appropriate method to evaluate contaminant release. In the case of rather impermeable clay or clayey loam samples, aggregated material can be ﬁlled into the columns, or depending on the scenario compacted material in a ‘‘ﬂow-around-monolith’’ mode can be used (according to NEN 7347). Changing conditions of pH, EC, redox and DOC should be monitored in the leachate. By carrying out a pH dependence test (i.e. CEN/TS1442910), the importance of the chemical speciation and partitioning can be addressed, which are relevant for assessing bioavailability of contaminants and mobility in soil and groundwater. For the prediction of long-term behaviour, modelling is an important tool (bench-scale and even full-scale testing can cover only a limited time span relative to the timeframe over which answers are sought). In general, the ﬁndings of leaching of pollutants support the tiered approaches which, for example, consist of basic, i.e. comprehensive, characterisation tests for solid materials, which are not well known, and later a simple, short-term compliance test for already well-characterised materials. With compliance tests the eﬀort and costs in assessing large material streams such as demolition wastes, incineration ash, slag, dredged sediments, etc., can be minimised.4 Basic characterisation allows for the understanding of the longterm dynamic behaviour of a speciﬁc material (including changes in redox, pH, etc.) whereas a compliance test only conﬁrms that the sample belongs to a wellknown (i.e. characterised in a basic test) group of materials. Because of that, compliance tests can be very simple ranging from visual appearance to batch and short-term percolation tests. On the other hand, basic characterisation tests have to be designed in a way which allows for the understanding of the longterm leaching behaviour of a contaminant. Results from such tests can be used in mechanistic models which then allow for predictions of the long-term behaviour under ﬁeld conditions. According to the experiences gained in GRACOS, short-term column tests are less diﬃcult to perform and yield more robust results than the widely used shaking or batch tests. It is well known in the scientiﬁc community that the shaking tests originally developed for

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

305

activated sludge from waste water treatment are not suitable for materials such as demolition waste, dredged sediments, incineration ash, etc. For inorganic contaminants, column tests in combination with batch testing as a function of pH provide the most robust results, as the information covers a wide range of possible conditions relevant for judging releases from granular materials.

5.6.3

Groundwater Risk Assessment for Volatile Compounds

5.6.3.1

Vapour-phase Diﬀusion

Contrary to ionic or highly water soluble compounds, which are predominantly transported by seepage water, volatile compounds can reach the groundwater table via the vapour phase. Some of the most frequent groundwater pollutants such as fuel constituents (benzene, toluene, ethylbenzene, xylenes (BTEX)) and chlorinated solvents (trichloroethene (TCE), perchloroethene (PCE), etc.) are volatile and migrate in the subsurface environment by vapour-phase diﬀusion. Diﬀusive spreading follows Fick’s second law: @C @2C ¼ Da 2 @t @x

ð4Þ

where Da denotes the apparent diﬀusion coeﬃcient [m2 s1] in the three-phase system soil solids/water/air, which is deﬁned as Da ¼

Dg ðn2:5 De g =nÞ ¼ ng þ ðnw =HÞ þ ðKd r=HÞ a

ð5Þ

where n, ng and nw denote the total, gas-ﬁlled and water-ﬁlled porosities, respectively. Dg is the molecular diﬀusion coeﬃcient in the gas phase and H is Henry’s law constant, deﬁned as the concentration ratio of gas to water. The term ng2.5/n denotes an empirical relationship for calculation of the eﬀective diﬀusion coeﬃcient (De) for vapour-phase diﬀusion in the unsaturated zone.17 Kd is the sorption coeﬃcient which can be estimated based on empirical relationships in terms of the organic carbon content of the soil.18 Note that partitioning into the water is often more important for retarded diﬀusion of volatile contaminants in the soil air than sorption onto the soil solids. With equation (4) and (5), the transient spreading of contaminants in the unsaturated zone can be assessed (the simplest case is of a concentration front, i.e. C/Co D 0.5, spreading from a source of ﬃ constant concentration where the pﬃﬃﬃﬃﬃﬃﬃ distance travelled is proportional to Da t). For the volatilisation of constituents from mixtures of organic compounds (e.g. benzene from gasoline) the change in composition over time causes changing vapour-phase concentrations in the source, which has to be accounted for by Raoult’s law: Ci ¼ wi

p0i mi g ¼ wi Cgsat gi RT i

ð6Þ

306

Chapter 5.6 2.0 MTBE

(b) Mixture 2

1.5 Flux (g m -2 h-1)

methylcyclohexane

toluene

1.0 TCE

1,1,1-TCA

ethylbenzene

0.5 PCE

0.0 10

0

20

30

40

50

60

70

80

90

100

Time (h)

Figure 5.6.11

Pure forward prediction (curves) of measured (symbols) diﬀusive ﬂuxes of gasoline constituents though a sand column of 30 cm length.19

where Cgsat, pi0, mi, R and T are the saturation concentrations in the gas phase of the pure compound [g l1], the saturation vapour pressure [kPa] and the molecular weight [g mol1] of compound i, the gas constant [l kPa mol1 K1] and the temperature [K], respectively. Ci, wi and gi denote the equilibrium vapour concentration [g l1], the molar fraction and the activity coeﬃcient of compound i in the mixture, which equals 1 in an ideal mixture (this is often assumed). The volatilisation rates are proportional to the vapour pressure and therefore depend on the molar fraction in the mixture. Figure 5.6.11 shows that with equation (4)–(6) the ﬂuxes of gasoline constituents diﬀusing from a non-aqueous gasoline source in a sand column over 30 cm distance can be predicted in pure forward fashion (i.e. no ﬁtting) reasonably well by a simple numerical model.19

5.6.3.2

Coupled Models for Simulation of Field Sites

Numerical models proved successful in predicting the behaviour of organic vapours in the unsaturated zone at the ﬁeld scale in well-controlled lysimeter and ﬁeld experiments.20–22 Such numerical models couple vapour-phase diﬀusion, biogeochemical reaction and ﬂow of seepage water allowing for the simulation of the spatiotemporal contaminant concentrations in the unsaturated zone and the capillary fringe as shown in Figure 5.6.12. Furthermore, these models allow for simulating the aging of complex organic mixtures in the ﬁeld by considering volatilisation of constituents from a source zone in the soil, the transport of pollutants to the atmosphere and the increased concentration gradients due to biodegradation (which accelerates the volatilisation of degradable compounds).22 Initially, the high-volatility compounds volatilise leading to an increase of the mole fraction of the residual compounds in the organic mixture which subsequently leads to an increase of the equilibrium vapour-phase

307

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

gaseous concentration [mg/l air] cyclo-pentane 50 20 2 1 0.2 0.02 0.002 0.0002 30

G_i, T = 3.000E+ 00 days

1

2 2

2

0

5

0.2

0.0

1

4

0.2 0.02 0.00.002 00 2 0.0002

20

0.2

2

0.02 0.2

3

0.002

1

2

0.02

1

2

0.000 2 0.002

1

0.002 0.0002

depth [m]

0

10

15 X [m]

20

25

G_i, T = 9.000E+ 01 days 0.2

2 1

2

2 1

0.02 0.0002

0.2

2

0

5

0.002

1

3

4

10

15

20

02

0.0

0.2

2

gaseous concentration [mg/l air] cyclo-pentane 50 20 2 1 0.2 0.02 02 0. 0.002 0.0002 30 0.0

1

1

2

depth [m]

0

25

X [m] G_i, T = 3.500E+ 02 days 0.002

0.000

2

1 2

0.0

0.00

2

02

depth [m]

0

3 0.002

4 0

5

Figure 5.6.12

10

15 X [m]

0.02

20

25

gaseous concentration [mg/l air] cyclo-pentane 50 20 2 1 0.00 0.2 0.02 0.002 0.002 0.0002 30

Development of an organic vapour-phase plume (cyclopentane) from a kerosene source in the unsaturated zone after 3 (initial spreading), 90 (maximum spreading) and 350 days (shrinking and detached plume in the groundwater), two-dimensional simulation, groundwater ﬂows from left to right.21

concentrations as shown in Figure 5.6.11 and for the ﬁeld case in Figure 5.6.13. Since vapour-phase pressures are dependent on temperature this has to be considered for obtaining an improved ﬁt between observation and prediction. However, already a simple isothermic model can predict the general behaviour reasonably well as shown in Figure 5.6.14. Also important is the reproduction of the CO2 produced by biodegradation of the organic vapour-phase compounds inﬂuenced by changing water contents in the soil because this strongly inﬂuences the eﬀective diﬀusion coeﬃcients (equation (5)) as shown in Figure 5.6.15. The only ﬁtting parameter needed in this case is the biodegradation rate constant.

5.6.3.3

Concluding Remarks for Risk Assessment for Volatile Compounds

Concentrations of volatile compounds in seepage water and the capillary fringe can be calculated from concentrations in the soil air based on Henry’s law constant. This also applies to the saturation vapour concentrations at the boundaries to non-aqueous phase liquids such as fuels which can be calculated from Raoult’s law (in most cases an activity coeﬃcient of unity is appropriate). Both laws

308

Chapter 5.6 40 35

molefraction [%]

30 25

hexane 3D model decane 3D model iso-octane 3D model Værløse field data

20 15 10 5 0

0

100

200

300

time [days]

Figure 5.6.13

Development of mole fractions of kerosene constituents during the ﬁrst year of the GRACOS ﬁeld experiment; ‘‘aging’’ of the source in the three-dimensional model (curves) agrees very well with ﬁeld data for hexane, isooctane and decane.21

require the ‘‘local equilibrium assumption’’ in models which can be conﬁrmed in bench-scale laboratory tests19 as well as in lysimeters22 and a well-controlled ﬁeld experiment run over a time span of one year.20,23,24 Diﬀusion coeﬃcients can be estimated based on empirical relationships17,19 or tracer tests.25–27 For ‘‘real’’ ﬁeld scenarios, biodegradation of organic vapours and the subsequent geochemical changes such as O2 depletion and CO2 production have to be considered. Coupled (seepage water ﬂow–vapour-phase diﬀusion– biogeochemical reaction) models such as MIN3P24 could be validated with the well-controlled ﬁeld experiment of a kerosene spill at the GRACOS ﬁeld test site. Volatilisation to the atmosphere has been proved to be the most important natural attenuation process in the unsaturated zone for volatile organic pollutants from shallow spills when the surface is not sealed, followed by biodegradation (Figure 5.6.16). Since biodegradation causes steep concentration gradients in the vadose zone, spreading is limited but aging or depletion of compounds from the source is accelerated. Modelling results illustrate that the overall biodegradation rates depend mainly on distribution parameters such as Henry’s law constant of the fuel constituents (because degradation takes place in the aqueous phase exclusively), on the biological degradation rate constant, on the soil water content and on the temperature. Temporal changes of temperature and inﬁltration rates aﬀect volatile organic compound behaviour signiﬁcantly (highly dynamic system): for a ﬁrst assessment, isothermal and stationary boundary conditions can be assumed. Transport to groundwater

309

Groundwater Risk Assessment at Contaminated Sites (GRACOS) x=0.0, z=-2.3m, located 1 m below center of NAPL source component gaseous concentrations in mg/l, 3D model

1.4

toluol(g) Værløse field data toluol(g) 3D model steady state toluol(g) 3D model transient

gaseous concentration [mg/l]

1.2

1

0.8

0.6

0.4

0.2

0

0

100

200

300

time [days] 20 15

5 0 air T NAPL source 2.25mdepth

Figure 5.6.14

T [˚C]

10

-5 -10

Fit of predicted toluene concentrations in the unsaturated zone below the kerosene source in the GRACOS ﬁeld experiment with consideration of temperature ﬂuctuations (dash–dot line) and under isothermal conditions (solid curve). The lower diagram shows the temperature in the air at the surface and at 1 m and 2.25 m depth.21

was found to be controlled by dispersion–diﬀusion processes: although only a small fraction of contaminants may reach the groundwater, legal limits in the capillary fringe are locally likely to be exceeded.4,22,28–30

5.6.4

Modelling for Groundwater Risk Assessment of Inorganic Constituents

As already explained above, for inorganic substances, mechanistic geochemical modelling approaches have more perspective over simple empirical Kd models,

310

Chapter 5.6 2D model location 1.3 m below NAPL source

CO2(g) CO2(g)transientflow

0.012 0.01

precipitation

-0.03 -0.025 -0.02

0.006

-0.015 -0.01

0.004

-0.005 0

0.002 evapot ranspiration

0

Figure 5.6.15

groundwater recharge

100

0.005

200 time [days]

0.01

300

3

total inflow total outflow

0.008

tota l outflow [m /day]

partial gas pressure [atm]

0.014

Inﬂuence of precipitation and evaporation, i.e. soil water contents, on the development of CO2 partial pressure due to biodegradation of organic vapours in the unsaturated zone: solid curve, steady state; dash–dot curve, transient inﬁltration conditions.16

mass in % of initial

100 biodegradation

80 degassing to atmosphere

60

1,2,4-TMB

40 20

mass still present

0 −6 10

Figure 5.6.16

−5

−4

10 10 degradation rate constant [s-1]

10

−3

Competition of biodegradation and degassing to the atmosphere as a function of the biodegradation rate constant (pseudo ﬁrst order) for the residual mass of 1,2,4-trimethylbenzene (1,2,4-TMB) after one year (GRACOS two-dimensional ﬁeld scenario).21,28 With increasing biodegradation, degassing to atmosphere is diminished and the volatilisation from the source is accelerated.

as their outcome has a much more predictive value.7,14,15 Unlike the apparent complexity of these models, there is a great perspective in limiting their input to obtain results of routine tests. Similarly, the choice of binding parameters for these models allow only a few degrees of freedom, as for all model components, ‘‘generic’’ parameter sets for a broad range of elements have been derived. This means that this type of modelling is characterised by a general applicability with respect to a wide variety of contaminants and a variety of soils, sediments

311

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

and wastes. Models that can couple chemistry with transport include, among others, PHREEQC,31 ECOSAT32 and ORCHESTRA.8 The latter has the most ﬂexibility for incorporating new adsorption models, solid solution descriptions, surface precipitation and organic matter interaction such as Nica Donnan. PHREEQC at present (2006) does not contain a module for the binding of metals to organic matter, which is needed to predict metal mobility in the upper soil layers with high contents of dissolved and particulate organic matter. Although not always necessary for risk assessment, the cited model codes rarely have detailed hydrological capacity standards available (only onedimensional transport; two- or three-dimensional in principle is possible, but not standard). At present, there are already some models such as PHAST,33 MIN3P34 and PHT3D35 that combine detailed hydrology with detailed chemistry. A mechanistic modelling approach was used in 2006 to derive new limit values for building materials in the revised Dutch Building Materials Decree. The methodology followed to derive these limit values is similar to that described in EN1292016 and is shown in Figure 5.6.17. The source term was based on contaminant-speciﬁc decay functions derived from percolation tests (TS1440511) performed on a variety of construction and waste materials. A percolation rate was assumed of 300 mm per year (average net rainfall in the Netherlands) and diﬀerent heights of application were assumed (i.e. 0.5 m to 5 m) that establishes the relationship between time and L/S ratio. Existing data from typical Dutch soil proﬁles were used to deﬁne the properties of the underlying soil proﬁle (i.e. in terms of water saturation and sorbent surfaces present in the soil such as clays and organic matter). The

Establish regulatory limits (iterative process)

Percolation through waste

Transport through soil

Figure 5.6.17

Concentration. at POC

Model simulation of contaminant leaching from a waste application to soil and groundwater rain

limit value

Time

“Point of compliance” with Environmental limit value

Simpliﬁed ﬁgure showing the process of modelling groundwater contamination and establishment of regulatory limit values.

312

Chapter 5.6

calculations were performed with ORCHESTRA8 using one-dimensional ﬂow of water along a streamline. Next, on a given ‘‘point of compliance’’, breakthrough curves were established and concentrations were compared to limit values (i.e. ecotoxicological values). Using an iterative process, limit values were then derived for building materials for each contaminant (expressed in mg kg1 leached at L/S ¼ 10) in such a way that concentrations at the point of compliance just fulﬁl the criteria at the point of compliance (17). The advantage of this semi-mechanistic approach over the often used Kd approach is obviously that important interactions between contaminants (and major elements such as iron, aluminium and calcium in the soil and groundwater) are captured by the model. In addition, this way of modelling is fully transparent as all thermodynamic constants used in the modelling are justiﬁed in the scientiﬁc literature. The next step in modelling is to describe both the source and the transport in the unsaturated and saturated zones by the full mechanistic modelling approach.

5.6.5

Conclusions/Recommendations

In general, for a contaminant of interest, the groundwater risk assessment should consider (1) the amount that is available for release or leaching from the material, (2) the multiphase local equilibrium distribution (i.e. aqueous/solid for non-volatile contaminants; vapour/ water/solid for volatile contaminants) and (3) any factors that may signiﬁcantly alter the multiphase local equilibrium (e.g. pH, DOC, colloids and suspended particles). For assessment of volatile contaminants, identifying the extent of the contaminant source area, whether or not a separate organic contaminant phase (dispersed or continuous) is present, and the extent of the vadose zone or groundwater migration plume are important. Monitoring soil vapour concentrations and the assumption of the local vapour/water equilibrium are generally suﬃcient (Henry’s law). When a separate organic phase (i.e. a complex organic mixture such as fuels, lubrication oils or solvents) is present, then the constituent’s vapour pressure can be calculated based on Raoult’s law. For assessment of non-volatile contaminants, leaching is best characterised using up-ﬂow column percolation tests in conjunction with pH controlled batch testing for pH-sensitive compounds and for when long-term release is of interest. Column testing provides initial equilibrium concentrations in the leachate and information on the long-term leaching dynamics as a function of elution liquid-to-solid ratio (L/S) or pore volumes. Results from the testing methods discussed above allow for the parameterisation and use of a contaminant release model for the source, which can then be coupled with a fate and transport model for considering dilution and attenuation from the source location to the point of compliance.36 Such models require a detailed description of hydrology at the ﬁeld site, including possible preferential ﬂow patterns. In response to the speciﬁc evaluation needs, the source term model may be either a simplistic, semi-analytical (i.e. spreadsheet)

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

313

model that provides an overestimate of release in order to be intentionally conservative, or it may be a more complex coupled local equilibrium model (i.e. including geochemical speciation for inorganic species) or a kinetic model that is coupled with representative mass transfer properties. Future tasks aim at the EU-wide validation of the characterisation leaching tests so that for these methods performance characteristics can be obtained. Additionally, based on many years of experience, a database/expert system has been developed which couples laboratory test data, physical aspects of release (hydrology), chemical changes in material properties with time and chemical reaction/transport modelling.9

Acknowledgement This work was supported by the EU 5th framework programme project GRACOS (Groundwater Risk Assessment at Contaminated Sites, EVK1CT1999-00029).

References 1. D. Halm and P. Grathwohl (eds), Proceedings of the 1st International Workshop on Groundwater Risk Assessment at Contaminated Sites (GRACOS), Tu¨bingen, Germany, 21–22 February 2002, Tu¨binger Geowissenschaftliche Arbeiten (TGA) C 61, 2002. 2. D. Halm and P. Grathwohl (eds), Proceedings of the 2nd International Workshop on Groundwater Risk Assessment at Contaminated Sites (GRACOS) and Integrated Soil and water Protection (SOWA), Tu¨bingen, Germany, 20–21 March 2003, Tu¨binger Geowissenschaftliche Arbeiten (TGA) C 69, 260, 2003. 3. H. A. van der Sloot, Waste Manag., 2002, 22, 693–694. 4. Guideline for Groundwater Risk Assessment at Contaminated Sites (GRACOS), www.uni-tuebingen.gracos.de. 5. D. S. Kosson, H. A. van der Sloot, F. Sanchez and A. C. Garrabrants, Environ. Eng. Sci., 2002, 19(3), 159–204. 6. H. A. van der Sloot, L. Heasman and Ph. Quevauviller (eds), Harmonization of Leaching/Extraction Tests, Studies in Environmental Science, Elsevier Science, Amsterdam, 1997, vol. 70. 7. J. J. Dijkstra, J. C. L. Meeussen and R. N. J. Comans, Environ. Sci. Technol., 2004, 38, 4390–4395. 8. J. C. L. Meeussen, Environ. Sci. Technol., 2003, 37, 1175–1182. 9. LeachXS: a database–expert system for leaching and environmental impact assessment, 2005 (http://www.leachxs.com). 10. CEN/TC292 (2005), Characterization of waste—Leaching behaviour tests—Inﬂuence of pH on leaching with initial acid/base addition, CEN/ TS 14429.

314

Chapter 5.6

11. CEN/TC292 (2004), Characterization of waste—Leaching behaviour tests— Up-ﬂow percolation test (under speciﬁed conditions), CEN/TS 14405. 12. ISO 21268-3, Soil quality—Leaching procedures for subsequent chemical and ecotoxicological testing of soil and soil materials—Part 3: Up-ﬂow percolation test. 13. ISO 21268-4, Soil quality—Leaching procedures for subsequent chemical and ecotoxicological testing of soil and soil materials—Part 4: Inﬂuence of pH on leaching with initial acid/base addition. 14. J. J. Dijkstra, H. A. Van der Sloot and R. N. J. Comans, Appl. Geochem., 2006, 21, 335–351. 15. J. J. Dijkstra, H. A. Van der Sloot and R. N. J. Comans, A consistent geochemical modeling approach for the leaching and reactive transport of major and trace elements in MSWI bottom ash, submitted. 16. CEN/TC292 (2005), Characterisation of waste—Methodology guideline for the determination of the leaching behaviour of waste under speciﬁed conditions, EN 12920. 17. P. Moldrup, T. Olesen, J. Gamst, P. Schjønning, T. Yamaguchi and D. E. Rolston, Soil Sci. Soc. Am. J., 2000, 64(5), 1588–1594. 18. R. Allen-King, P. Grathwohl and W. P. Ball, Adv. Water Res., 2002, 25(8–12), 985–1016. 19. G. Wang, S. B. F. Reckhorn and P. Grathwohl, Vadose Zone J., 2003, 692, 701. 20. M. Broholm, M. Christophersen, U. Maier, E. Stenby, P. Hoehener and P. Kjeldsen, Environ. Sci. Technol., 2005, 39(21), 8251–8263. 21. U. Maier and P. Grathwohl, in Reactive Transport in Soil and Groundwater, ed. G. Nu¨tzmann, P. Viotti and P. Aagard, Springer, 2005, pp. 141–155. 22. G. Pasteris, D. Werner, K. Kaufmann and P. Ho¨hener, Environ. Sci. Technol., 2002, 36, 30–39. 23. P. Gaganis, P. Kjeldsen and V. N. Burganos, J. Vadose Zone Res., 2004, 3, 1262–1275. 24. P. Gaganis, H. K. Karapanagioti and V. P. Burganos, Adv. Water Res., 2002, 25, 723–732. 25. D. Werner and P. Ho¨hener, Environ. Sci. Technol., 2002, 36, 1592–1599. 26. D. Werner and P. Ho¨hener, Environ. Sci. Technol., 2003, 37(11). 27. D. Werner, P. Grathwohl and P. Ho¨hener, Vadose Zone J., 2004, 3, 1240– 1248. 28. P. Grathwohl, I. D. Klenk, U. Maier and S. B. F. Reckhorn, IAHS Publ., 2002, 275, 141–146. 29. I. D. Klenk and P. Grathwohl, J. Contam. Hydrol., 2002, 58(1–2), 111–128. 30. I. D. Klenk, Transport of volatile organic compounds (VOCs) from soilgas to groundwater, PhD dissertation, TGA C55, Center for Applied Geosciences, Tu¨bingen, 2000. 31. D. L. Parkhurst and C. A. J. Appelo, User’s guide to PHREEQC (version 2): a computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations, Water Resource Inv. Report 99-4259, US Geological Survey, Denver, CO, 1999.

Groundwater Risk Assessment at Contaminated Sites (GRACOS)

315

32. M. Keizer and W. H. Van Riemsdijk, ECOSAT: a computer program for the calculation of speciation and transport, Department of Soil Quality, Wageningen University, 1996. 33. D. L. Parkhurst, K. L. Kipp, P. Engesgaard and S. R. Charlton, PHAST: a computer program for simulating ground-water ﬂow, solute transport, and multicomponent geochemical reactions, 2005. 34. K. U. Mayer, E. O. Frind and D. W. Blowes, Water Resour. Res., 2002, 38(9), 1174–1195. 35. H. Prommer, D. A. Barry and C. Zheng, Ground Water, 2003, 42(2), 247–257. 36. P. Grathwohl, in Boden und Altlasten, ed. Franzius, Lu¨hr and Bachmann, Erich Schmidt Verlag, 2000, vol. 9, pp. 41–60.

CHAPTER 5.7

INCORE: Integrated Concept for Groundwater Remediation THOMAS ERTELa AND HERMANN J. KIRCHHOLTESb a

Sachversta¨ndigen-Bu¨ro, Boschstr. 10, DE-73734 Esslingen, Germany; Landeshauptstadt Stuttgart, Amt fu¨r Umweltschutz, Hermann Josef Kirchholtes, 36-3.51, Gaisburgstr. 4, DE-70182, Stuttgart, Germany

b

5.7.1

Motivation and Basic Concept

European cities located in river basins are using groundwater from local shallow aquifer systems. Industrial development in the 20th century was rapid and caused urban groundwater pollution, often exceeding the legal limits. Changes in land use during this period have created complex contamination patterns, such as heterogeneous distribution of contaminants, the presence of diﬀerent contaminants and large landﬁll areas (Figure 5.7.1). Besides the threat to the wider environment, existing soil and groundwater contamination has resulted in incalculable costs for long-term groundwater remediation. Healthy residential and working conditions can be guaranteed only on uncontaminated ground. The presence of environmental pollution can limit investment in urban development on brownﬁeld sites. However, structural changes now in place oﬀer an opportunity to improve soil and groundwater quality. Considerable thought must be made in order to reach sustainable improvements: this is exempliﬁed by the integral procedures developed by INCORE. The current legal approach for the treatment of soil and groundwater pollution is focused on a particular set of problems caused by a speciﬁc polluter. All measures aim at a rapid reduction of environmental damage so that risk to the public associated with a particular property is removed. However, this approach fails in heavily polluted areas with diﬀerent property owners and complex pollutant patterns (Figure 5.7.2). Large amounts of private and public money are being spent to identify and assess point sources of contamination without being able to quantify reliably their impact on groundwater quality; numerous remediation schemes are undertaken without an economic evaluation of their long-term performance. 316

INCORE: Integrated Concept for Groundwater Remediation

Figure 5.7.1

Aerial view of Stuttgart Neckar valley.

Figure 5.7.2

Complex pollution pattern schema.

317

Five European cities, Stuttgart, Linz, Strasbourg, Milan and Bydgoszcz, which share similar groundwater problems in their industrialised urban areas, committed themselves to develop jointly suitable solutions. Speciﬁc local conditions vary in these ﬁve INCORE project areas; they vary with respect

318

Chapter 5.7

to groundwater conditions, existence of public and private monitoring wells, type of pollutants, size of problem areas, rising groundwater problems, etc. Therefore they provide a representative range of the conditions to be expected across Europe. In order to achieve the INCORE project goals in a cost-eﬀective way, diﬀerent parts of the anticipated tool set were applied and evaluated at diﬀerent levels of detail in the ﬁve selected areas. The proposed INCORE strategy for the investigation, remediation and revitalisation of industrial areas is based on an integrated quantiﬁcation of total contaminant emissions. It considers entire industrial areas instead of particular single sites, in order to achieve a high level of conﬁdence in the investigation results. A cyclic approach is proposed, beginning with the screening of groundwater plumes at the scale of entire industrial areas, and ending with the remediation of individual source areas or the containment of plumes. The major advantage of this approach is that the number of local-scale sites, or the size of the area to be considered, is reduced stepwise from one cycle to the next. Thus, a large potentially contaminated area would be screened but ultimately only a small area may need remedial actions. Figure 5.7.3 presents a schematic of this new approach. This new approach repeats an investigation/assessment/revitalisation cycle three times at diﬀerent scales. Cycle I. The groundwater quality is screened downstream of the potential source areas. Cycle II. Only those sites where groundwater quality is not acceptable are considered further. In these cases analytical methods are used to backtrack and identify sources of contamination.

Figure 5.7.3

Cyclic approach.

INCORE: Integrated Concept for Groundwater Remediation

319

Cycle III. The characteristics of the source zones are considered for remediation to control emissions, or implementation of monitored natural attenuation. The most appropriate technology is selected by establishing what level of contamination reduction is required, considering the proposed future use of the site. This ‘‘no further action’’ approach provides a cost eﬀective set of tools for the optimised investigation, evaluation and management of contaminated groundwater and land in industrialised urban communities.

5.7.2

Cycle I: Plume Screening

At the beginning of the project data inventories compiling historical information and present land use data are set up, combined with the development of a conceptual site model. Based on this, ﬁrst technical investigations are emission oriented and are focused on groundwater contamination. Quantiﬁcation of contaminant emission is obtained from the application of a novel integral groundwater investigation method (see Chapter 5.5), which yields the total pollutant mass ﬂux and the mean and maximum pollutant concentrations originating from contaminant source zones. The basic idea of the new integral groundwater investigation technique is that the total contaminant mass ﬂux downstream of potentially contaminated sites is covered by the capture zones of one or more pumping wells, which are positioned along control planes perpendicular to the mean groundwater ﬂow direction. Analyses of multiple groundwater samples obtained at the wells during pumping yield concentration time series. Results of this integral groundwater investigation method are the total contaminant mass ﬂux (emission) and the mean and maximum concentrations within the undisturbed groundwater ﬂow ﬁeld as well as determination and the delimitation of boundaries of potential polluted source zones. Further modelling leads to estimations of probabilities of contaminant concentrations exceeding regulation limits within large areas. The size of these areas under consideration depends on the transport and degradation behaviour of diﬀerent contaminants. The ﬁnal result is a map (see Figure 5.7.4) of the investigation area distinguishing between areas with diﬀerent levels of groundwater impact. This allows a ranking of these areas with a distinct level of conﬁdence. This map enables the administration to set priorities for further activities. Focusing the eﬀorts on the areas of greatest impact helps to concentrate personnel resources on those sites which cause major impacts on groundwater pollution. This leads to maximum eﬀectiveness in administrative activities. The mapping of areas with diﬀerent groundwater impact further identiﬁes areas in which urban and economic development can take place without any hindrance by groundwater pollution. This ensures development and structural change with lower risks on investment.

320

Chapter 5.7

source area with high gw impact source area with low gw impact no gw impact gw control plane investigation area direction of gw flow

Figure 5.7.4

5.7.3

Results of plume screening.

Cycle II: Source Identiﬁcation

The results of the integrated investigation provide a rough localisation of suspected source areas. However, more precise identiﬁcation of the contamination source area is needed in order to apply the polluter pays principle. With the means of cost-eﬀective laboratory and on-site analytical systems as well as isotopic ﬁngerprinting techniques the backtracking from the control plane along the path line of the plume yields a precise localisation of the source of contamination (see Figure 5.7.5). The results of cycle II verify the sources of groundwater pollution and identify the polluter with a very high reliability. This secures the application of the ‘‘polluter pays principle.’’ These results also help to avoid law-suits which leads to an acceleration of investments and to faster administrative procedures.

5.7.3.1 5.7.3.1.1

GC-MS Fingerprinting for Petroleum Hydrocarbons General Approach

Fingerprinting of hydrocarbons was developed in the petroleum industry, so as to understand the source of crude oil and natural gases. The objectives of ﬁngerprinting investigations (environmental forensics) pertaining to history and source of environmental pollutants are overall very similar to those of petroleum geochemistry. Fingerprinting is deﬁned as a scientiﬁc methodology

INCORE: Integrated Concept for Groundwater Remediation

321

source area with high gw impac source control planearea source area with low g A contributes 80 % to detected plume no gw impact 20 % to detected plume B contributes gw control plane to detected plume C no contribution investigation area direction of gw flow

C C

C

B

C

C C A A

Figure 5.7.5

Results of source screening.

developed to be used in the environmental assessment of a fuel pollutant (or fuel contaminant) to: characterise the type of the fuel contaminant; quantify the concentration of potentially environmental hazardous compounds; and identify the composition of all compounds within the fuel contaminant that can be used to reliably determine the source, and understand the rate and transport of the fuel pollutants. Hydrocarbons released into the environment are subjected to biotic and abiotic transformation reactions in the soil and groundwater media. The rate and behaviour of these hydrocarbons depends on a number of physicochemical and biological processes including (1) evaporation, (2) dissolution, (3) microbial degradation, (4) photooxidation and (5) interaction between oil and sediments. The combination of these processes known as weathering reduces the concentration of released hydrocarbons in soil and groundwater and alters their chemical composition. For a speciﬁc site investigation and oil spill identiﬁcation, analytical approaches have to be considered which provide detailed compositional information, and are sensitive enough for measuring spilled oil and petroleum products in soils and groundwater. The following list provides major target analytes which are suggested for the characterization and source identiﬁcation of the contaminants. Aliphatic hydrocarbons, including n-alkanes in the C8–C40 range, and selected isoprenoids like pristane (C19) and phytane (C20).

322

Table 5.7.1

Chapter 5.7

Database-base input and its variety.

Database section

Data key

n

Database section

Data key

n

Contaminant type

1 gasoline

22

Age of contamination

1o10 years

25

2 diesel/heating oil 3 bunker C oil 4 lubricating oil 5 creosote

47

2 10–20 years

18

7 16

3 20–50 years 4450 years

44 8

See list above

50

3

Chemical parameter

Number of considered samples ¼ 95.

Single-ring volatile aromatic hydrocarbons, so called BTEX compounds (benzene, toluene, ethylbenzene and m-, o- and p-xylene), and alkylated benzenes (C3–CS benzenes). Polynuclear aromatic hydrocarbon compounds (PAH; Table 5.7.1), particularly the two- to four-ring compounds (including dibenzothiophenes) can be used to identify both source and extent of degradation. These compounds include the 16 EPA priority pollutant PAHs, and their associated alkylated homologues. Biological markers (or biomarkers) are molecular fossils, meaning that these compounds are derived from formerly living organisms.1 Biomarkers are complex organic compounds composed of carbon, hydrogen and other elements which are found in crude oil and its heavy reﬁned products. The structural characteristics of these compounds are chemically stable during degradation. Terpanes and steranes are the most common biomarkers in crude oils and their medium and heavy reﬁned products. Besides identiﬁcation of discrete fuel types, if suﬃcient release information is available, this technique can often discriminate whether a contaminant release was a single event, a series of events or a continuous release of a single or multiple products. By considering the results of other workers,2–5 Kaplan et al.6 have elaborated biodegradation which is presented in Figure 5.7.6, illustrating the relative level of diﬀerent hydrocarbon types in fuels with a volatility range from gasoline to bunker C oil. However, this degradation chart should be used cautiously because biodegradation is a complex ‘‘quasi-stepwise’’ process that cannot be described as truly sequential alteration of compound classes.3 Weathered oils from diﬀerent spills (sources) can be distinguished by using the source ratio (C3-dibenzothiophenes/C3-phenanthrenes or D3/P3) and weathering ratio (D3/C3-chrysenes or D3/C3) (Figure 5.7.7). Compounds in source ratios degrade relatively at the same ratio whereas those in weathering ratios change substantially with weathering and biodegradation.

323

INCORE: Integrated Concept for Groundwater Remediation

Figure 5.7.6

Degradation scheme.

12

8

6

4

Increasing Oil Weathering

D3/C3 Weathering Ratio

10

North Sea, mean = 0.80 ∓ 2 SD

Alaska North Slope l crude, mean = 1.190 ∓ 2 SD

Iranian Crude, mean = 2.48 ∓ 2 SD

2 0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

D3/P3 Souce Ratio

Figure 5.7.7

5.7.3.1.2

Plot of D3/C3 vs. D3/P3.10

Contaminant Type Identiﬁcation by Statistical Analysis

5.7.3.1.2.1 Objectives. The application of the ﬁngerprinting method in practice demands much expert knowledge. The interpretation of the chromatograms

324

Chapter 5.7

and/or mass spectra needs long lasting experience and a large extent of know-how. Statistical analysis transfers the expertise-based method of ﬁngerprinting interpretation to a tool comprehensible to each party involved in the investigation or evaluation process. This transfer was done using several statistical methods so the results of the ﬁngerprinting tool will be both comprehensible and objective. The main goal of statistical analyses in the INCORE strategy is to develop a tool which could be used to identify the type of contaminant found in the plume and to determine source–plume relationships to potential source areas. 5.7.3.1.2.2 Database. The most detailed analytical data available from the INCORE test ﬁelds are those from samples taken at Stuttgart investigation site, which show the typical picture of a multiple used and therefore multiple contaminated industrial site. However, since the database should contain results from as many real samples as possible, the database was enlarged with other data derived from projects mainly in Germany, covering a wide range of diﬀerent types of contaminants having undergone diﬀerent levels of weathering (biodegradation). A set of 50 diagnostic parameters consisting of alkylbenzenes, alkylcyclohexanes, isoprenoides, PAHs and biomarkers such as steranes and terpanes were selected for further statistical consideration. For pattern identiﬁcation among the parameters the original measured values were transformed in respect to the sum of the total analysed peak areas. Within the INCORE investigation and other related projects, the compiled database shows a broad range of real contaminated samples. The enlargement of the data with reference samples of ‘‘pure’’ contaminant types improves further the database. These reference samples were collected as fresh directly from service stations (diesel, gasoline) or heating oil storage tanks. An index of the database input and its variety is given in Table 5.7.1. 5.7.3.1.2.3 Discriminant Analysis. In a ﬁrst step the database was ﬁltered for the reference samples and those samples which were deﬁnitely contaminated with only one product type. This data subset was used for the model development. From the samples with known group membership (contaminant type), a set of linear discriminant functions is generated. These functions are based on a linear combination of the ‘‘predictor variables,’’ those parameters which in a stepwise procedure were identiﬁed to provide the best discrimination between the groups. In processing the stepwise method at least 17 parameters were selected which are signiﬁcant for the classiﬁcation of the ﬁve contaminant types. The so-called canonical correlation coeﬃcients (CCC) reach values between 0.999 and 0.913 which testify to the very good classiﬁcation of the model for the development data set. Using function 1 and function 2, the discriminant scores were calculated for each sample. The scatterplot shows that the ﬁve diﬀerent contaminant types can be divided signiﬁcantly. In practice most contamination belongs to multiple product types with probably diﬀerent ages and diﬀerent stages of degradation. Taking this information into account in a further step,

325

INCORE: Integrated Concept for Groundwater Remediation 10 discrimi. scores function 2

discrimi. scores function 2

10 5 0 gasoline middle distillates bunker C oil lubricating oil creosote

-5 -10 -15 -20 -20

gasoline middel distillates bunker C oil lubricating oil creosote

8 6 4 2 0 -2 -4 -6 -8 -10

0

20

40

discrimi. scores function 1

Figure 5.7.8

60

-5

0 5 10 15 discrimi. scores function 1

20

Calculated discriminant scores for development data set (left) and all samples (right).

all available data of the database were used to develop a new model in order to provide wider regions for the identiﬁcation of real samples. As expected the classiﬁcation of the enlarged sample set shows less sharp results. Regarding the CCC of function 1 which reached a value of 0.999, for the reference samples it is now reduced to 0.969. Nevertheless this result indicates a signiﬁcant classiﬁcation of the ﬁve contaminant types. Calculating the discriminant scores from function 1 and function 2 and plotting them together as x- and y- axis the regions of the diﬀerent contaminant types can be visualised (Figure 5.7.8). It can be seen that the heavy distillates such as bunker C oil and creosote can be divided easily from the remaining samples. Also the discrimination of gasoline and lubricating oil is signiﬁcant. Only the discrimination of some samples contaminated with middle distillates is not clear. For that reason a special discriminant function was estimated for separating gasoline, middle distillates and lubricating oil. The result of this approach in general shows a better discrimination between the three groups, but some samples still cannot clearly be identiﬁed. The reason for this is the sample composition and not the statistical model, because those samples show an untypical pattern of contamination, caused most likely by mixed contamination occurring at gas stations and the somehow intermediate character of so-called ‘‘winter diesel.’’ 5.7.3.1.2.4 Field Site in Stuttgart. Based on the results of the investigation performed in cycle I, this site was selected as a potential source of BTEX and PAH observed in well NT 9. For further identiﬁcation of groundwater contamination and related plumes, a series of integral pumping tests (IPTs) was performed in newly installed wells given in Figure 5.7.9. The wells were placed according to potential sources known from the historical survey and previous investigations. Figure 5.7.10 gives an example of the IPT results in NT 108 showing a rapidly increasing PAH concentration, which leads to a PAH plume with its centre more than 20 m

326

Chapter 5.7

Figure 5.7.9

Results of groundwater sampling.

16

concentration [µg / l]

14 12 10 8 6 4 2 0 0

5

10

15

20

25

30

35

40

45

50

width of capture zone [m]

Figure 5.7.10

Example result of IPT.

besides NT 108. Figure 5.7.11 summarises the results of the IPT campaign giving a rough estimation of the shape of the identiﬁed plumes. According to the dynamic workplan an adaptive sampling and analysis campaign was performed to localise deﬁnitively the contaminant sources. The results are also given in Figure 5.7.11.

INCORE: Integrated Concept for Groundwater Remediation

Figure 5.7.11

327

Results of the IPT campaign with estimated shape of the identiﬁed plumes.

A total of 10 soil samples collected from hot spots of the contaminated site, as well as 14 groundwater samples from IPTs have been studied using ﬁngerprinting methods/parameters presented in Table 5.7.2. The correlation of the fuel contaminants identiﬁed in the monitoring wells has been performed mainly on the basis of their PAH distribution patterns. Results of this correlation lead to an interpretation of source–plume relationship as shown in Figure 5.7.11. 5.7.3.1.2.5 Hierarchical Cluster Analysis. Figure 5.7.12 shows the result of the hierarchical cluster analysis with a database subgroup consisting of all samples from Ulmer Straße. On the left a dendrogram is depicted. It shows the diﬀerent steps of the cluster analysis, starting with every sample in one cluster proceeding to only one cluster left. The clusters were computed with samples which are relatively similar among the cluster group but relatively diﬀerent to the other groups. From the distances, conclusions can be drawn if for example a particular test sample is nearer to one or another sample. On the right the clustering process is shown as a table. The ﬁrst column indicates the cluster aﬃliation of the samples when eight clusters were computed. The second column indicates the cluster aﬃliation when only seven clusters were computed and so on.

328

Table 5.7.2

Chapter 5.7

Parameter and related laboratory methods used for ﬁngerprinting at Ulmer Strasse. Used method H-18 GC-MS GC-MS method GC-MS method Full scan GC-MS of an extract from which aliphatic compounds are removed

Figure 5.7.12

cluster 7

cluster 6

cluster 5

cluster 4

cluster 3

cluster 2

Ulmer Straße 105/1 GW Ulmer Straße 105/2 GW Ulmer Straße 105/5 GW Ulm Str. W. NT13/6 Ulmer Straße 74 (2 m) Ulmer Straße 79 (3,8 m) Ulmer Straße 67 (5 m) Ulmer Straße 62 (4,2 m) Ulmer Straße 62 (2,2 m) Ulm Str. W. 108/1 Ulm Str. W. 108/5 Ulm Str. W. NT16 Ulmer Straße 71 (3 m) Ulmer Straße 71 (4-6 m) UlmS tr.W .N T10 Ulm Str. W. NT109/4 Ulm Str. W. NT109/8 Ulm Str. W. NT13/1 Ulm Str. W. NT9 Ulmer Straße 67 (3 m)

cluster 8

GC-MS method

Sample ID

Parameter Total petroleum hydrocarbon content (TPH) Volatile aromatic hydrocarbons (BTEX) MTBE (oxygenate) EDB and EDC (lead scavengers) Parameter pattern: Total ion chromatogram (TIC) Normal alkanes ( using m/z 85 mass chromatogram) Isoprenoids (using m/z 113 mass chromatogram) C4 alkylbenzenes (using m/z 134 mass chromatograms) Steranes and triterpanes (using m/z 217 and m/z 191 mass chromatograms, respectively) Polynuclear aromatic hydrocarbons (PAHS)

1 1 1 1 2 3 3 3 4 4 4 4 8 8 5 5 5 5 5 7

1 1 1 1 2 3 3 3 4 4 4 4 4 4 5 5 5 5 5 7

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

1 1 1 1 2 3 3 3 3 3 3 3 3 3 4 4 4 4 4 5

1 1 1 1 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4

1 1 1 1 1 2 2 2 2 2 2 2 2 2 3 2 2 2 2 3

1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

Cluster analysis.

In contrast to the results of the discriminant analysis, the cluster analysis shows the diﬀerences between the samples in more detail (Figure 5.7.12). Not all gasoline or middle distillates samples were grouped in one class, as under real ﬁeld conditions every individual sample is diﬀerent in composition, age and degradation level. Nevertheless following the clustering process information can be extracted as to which sample is comparable with another and which samples do not ﬁt with any other. For example, the samples from diﬀerent depths of GW 105 and the sample from NT 13 grouped together indicate that their main contaminant is gasoline. During the clustering process the sample from 74 (2m) is grouped together with

INCORE: Integrated Concept for Groundwater Remediation

329

these gasoline-contaminated samples which means that the contaminant pattern of this samples is most comparable to these others. This ﬁts quite well to the real situation because 74 (2m) consists of gasoline and of minor amounts of diesel. Another group of samples which are relatively homogeneous are the samples from NT 109, NT 9, NT 13/1 and NT 10. This group is mainly contaminated with weathered gasoline and traces of moderately altered diesel. All other samples are relatively not comparable to this group which can be seen in Figure 5.7.12. The case study shows that statistical correlation can be a powerful tool to support the interpretative work of the expert and help to visualise results.

5.7.3.2 5.7.3.2.1

Isotopic Fingerprinting for Chlorinated Hydrocarbons General Approach

The isotopic composition of chlorinated hydrocarbon compounds released into the environment is in fact inﬂuenced by the diﬀerent raw material sources and manufacturing processes associated with the synthesis.7 In this way, if the isotopic signature of the dissolved contaminant has not undergone signiﬁcant fractionation, it still reﬂects the isotopic composition of the source, which can thus be identiﬁed. Nevertheless, due to extensive degradation the correlation of the pollutants with their suspected sources could be diﬃcult, because the bulk isotopic composition can be aﬀected. Preferential microbial degradation of the light compounds causes isotopic fractionation. This leads to an enrichment of the heavy isotopes in the remaining source compound and respectively to a depleted isotopic signature in the degradation product. Recent publications on PCE and TCE as the main pollution species in groundwater show a signiﬁcant diﬀerence between initial product d13C and their degradation product counterparts. The isotopic fractionations reported are as follows: 2%

4%

12%

26%

PCE ! TCE ! DCE ! VC ! Eth: As the initial molecule is degraded, the associated carbon isotope fractionation is more and more important, reaching values up to 26% (Figure 5.7.13). The degradation process can be mathematically described by Rayleigh models. By means of these models a clear distinction between degradation and a mixture of diﬀerent sources as a reason of diﬀering isotopic signatures can be achieved.

5.7.3.2.2

Source–plume Relationship by Isotopic Fingerprinting at Nesenbach Site

5.7.3.2.2.1 Initial Situation. The site is located in the Nesenbach valley in the city centre of Stuttgart (see site map in Figure 5.7.14). From the beginning of the 20th century to 1976 a dry-cleaning facility was situated on the site. The

330

Chapter 5.7 0 -10

-30 Total

-40

PCE TCE

-50

cDCE

δ

13

C [‰ vs. PDB]

-20

-60

VC Ethene

-70 0

10

20

Time (j)

Figure 5.7.13

Experimental carbon isotope fractionation associated with biodegradation.8

cleaning agent ﬁrst used was BTEX and later CHC, predominantly tetrachloroethene (PCE). In 1989 soil and groundwater contamination was detected. Initial investigations revealed concentrations up to 435 000 mg l1 CHC and 200 mg l1 BTEX in the groundwater. In 1990 the contaminated soil of the unsaturated zone was almost completely excavated. A probably remaining pool in the saturated zone still seems to feed a signiﬁcant plume. The prerequisite for any further consideration on remediation technology is to prove whether this known source really is responsible for the extended plume and if there are other sources further downstream contributing to this plume. Besides classic geochemical characterisation methods, this identiﬁcation of source– plume relationship was done by isotopic ﬁngerprinting using d13C on speciﬁc CHC compounds. Beneath a covering of anthropogenic ﬁll and quaternary sediments the subsurface consists of clay stones, weathered gypsum and dolomite layers of the triassic Middle and Lower Keuper. The diﬀerent permeabilities of the clay stones and the fractured dolomite stones cause a vertical subdivision into several groundwater bearing zones: a low permeable ﬁrst aquifer in the Middle Keuper and a more yielding second aquifer in the Lower Keuper. The area is tectonically disrupted by two major faults. Further consideration of the plume therefore is restricted to this second aquifer. 5.7.3.2.2.2 Plume Localisation by Local-scale IPTs. Based on the results of a sampling and analysis campaign in the wells shown in Figure 5.7.14, a rough localisation of the central area of the plume was performed by a series of localscale IPTs. Figure 5.7.15 shows the concentration time series for CHC of the IPTs in the downgradient monitoring wells 5, 7, 8, 9 and 12. The increasing CHC

INCORE: Integrated Concept for Groundwater Remediation

Figure 5.7.14

331

Map of investigation area.

concentration during the IPTs is typical of wells located randomly to an existing plume. The rather straight lines in GWM 8 and GWM 12 indicate a central position in a plume. The lateral borders of the plume were not reached during the pumping phase.

5.7.3.2.3

Isotopic Fingerprinting

To prove the source–plume relationship and identify additional polluters which might contribute to the CHC plume in this area with many other potential polluters, several wells have been selected for isotopic ﬁngerprinting by 13C isotopes of the diﬀerent CHC compounds. Figure 5.7.16 summarises the results by giving both total CHC contents and d13C values.

332

Chapter 5.7

Figure 5.7.15

Concentration vs. time series during IPTs. δ13C [‰V-PDB]

-15

-17

-19

-21

-23

-25

-27

-29

-31

-33

-27,1

GWM 1 PCE: 37630 µg/l

-28,3

TCE: 924 µg/l cDCE: 543 µg/l

-30,1 -27,1

GWM 2 PCE: 6230 µg/l

-24,3

TCE: 99,6 µg/l

-25,9

cDCE: 87 µg/l

-27,5 -27,5

GWM 8 PCE: 744 µg/l TCE: 56,1 µg/l

-30,1

cDCE: 67,5 µg/l

-25,4

GWM 7ku PCE: 150 µg/l

-30,6 -30,2

TCE: 26,2 µg/l cDCE: 21 µg/l

-27,4

GWM 9 PCE: 250 µg/l

-23,0

TCE: 11,1 µg/l cDCE: <5 µg/l

-26,6

GWM 12W PCE: 137 µg/l TCE: 4,28 µg/l

-17,5

cDCE: <5 µg/l

Figure 5.7.16

Documentation of d13C values on CHC.

GWM 1 and GWM 2 characterise the source area. GWM 1 in the ﬁrst aquifer gives the initial source values with about 27% V-PDB for PCE. The values for TCE and CIS indicate a slight degradation occurring according to an established degradation model described in Chapter 3.2. According to this

INCORE: Integrated Concept for Groundwater Remediation

Figure 5.7.17

333

Fractionation described by a Rayleigh model.

model the development of the resulting fractionation can be described by a Rayleigh model given in Figure 5.7.17. This model indicates: for PCE, all wells are dominated by the contamination originating most probably from the identiﬁed source; TCE in GWM 2 (second aquifer close to source area) indicates an already existing TCE contamination upstream of the source area; TCE in GWM 7 Ku is inﬂuenced by another TCE contamination which diﬀers signiﬁcantly from the one upstream of the main source area; and TCE in GWM 9 and GWM 12 shows very heavy isotopic values indicating another TCE contribution not related to those located upstream of the source or west of the centre of the plume. Considering the total contents of TCE and CIS this additional contribution of TCE in 9 and 12 can be disregarded. A rough estimation based on mixing of contaminations in GWM 7 and GWM 2 assuming the second component following in the most common range of 16–32 d13C values for initial TCE leads to TCE and CIS contents of about 20–100 mg l1 for the additional components.

5.7.4

Cycle III: Remediation Strategy

5.7.4.1

Remediation Scenarios

Monitored natural attenuation as well as emission-focused in situ remediation combined with passive remediation techniques are major options which have to be considered in the management of contaminated groundwater and land in

334

Chapter 5.7

scenario 1: joint remediation of whole area using one reactive barrier

scenario 2: remediation of each hot spot / source separately

scenario 3: combined remediation of clusters of related spots

Figure 5.7.18

Remediation scenarios.

urban industrial areas. The work is focused on dual solutions of source and plume remediation, taking into account natural attenuation as a part of a comprehensive remediation approach, depending on the remaining/tolerable contamination levels and extent. The basic idea is to ﬁnd the most eﬃcient hot spot treatment technology for a given hydrogeological and contamination situation, and to treat the remaining plumes with passive remediation technologies. These alternative or partly combined remediation scenarios were evaluated in feasibility studies (see Figure 5.7.18): scenario 1: joint remediation of the whole area using one reactive barrier; scenario 2: remediation of each hot spot/source separately; and scenario 3: combined remediation of clusters of related hot spots.

INCORE: Integrated Concept for Groundwater Remediation

335

Remediation schemes lead to joint remediation strategies for several sources of contamination. The application of ‘‘private–public partnership’’ and innovative technologies will result in cost-eﬃcient and faster treatment of polluted sites. By this means the costs for industrial settlement as an important locational factor for industry will be lowered.

5.7.4.2

ISIRE: In Situ Remediation Technologies—Decision Support

In order to remediate the contaminated areas, it is important to analyse and compare the best available remediation technologies, taking into consideration site and contaminant characteristics. Preferably those remediation techniques are chosen that treat and utilize the soil within the site, reducing costs and risks caused by transport and disposal. The concept of ‘‘best available technology at tolerable costs’’ is pursued, but it is not explained how to evaluate it. ISIRE contributes to this evaluation collecting and updating information about innovative techniques just tested at pilot scale or only demonstrated at full scale to a limited extent. ISIRE 1.0 is developed by the Polytecnico of Milan. The software supports the preliminary choice of suitable selection of in situ remediation techniques. ISIRE allows a fast screening among available techniques, considering at the same time the contaminant- and site-speciﬁc characteristics. The software can help to speed up decision-making processes providing basis information and references to start a feasibility study and to choose the best in situ remediation technology (Figure 5.7.19). ISIRE has been compiled in Visual Studio 6.0 in order to have a user-friendly interface. It is possible to download the software at the following web site: http://geologica.dstm.polimi.it. The main bibliographic source, on which the software contents are based, is the literature study ‘‘In situ remediation techniques’’ developed during the INCORE project.9 During the European project, the program has been tested to evaluate which were the best available in situ technologies for a site situated in the southeast zone of Milan, Italy, contaminated by chlorinated solvents. ISIRE gives the user an immediate answer about the applicability of several in situ technologies, considering simultaneously the type of contaminant and the site-speciﬁc characteristics. The program checks the availability of innovative technologies and the combinations of more remediation technologies; nevertheless it is not possible to have an ordered list according to their eﬀectiveness. Once the solutions are found, the best one has to be chosen according to a detailed feasibility study. With the ﬁrst steps it is possible to reject all those technologies which anyway cannot be applied to the studied area, because of the depth of contamination and/or of the concentration of the contaminant and/or for several other reasons. In the second phase, it is possible to check if some

336

Chapter 5.7 All considered in situ technologies (N) Contaminants

Contaminant characteristics (depth to restore, concentration, dimension of area)

Soil characteristics (conductivity, subsoil homogeneity, saturation)

Technology characteristics (remediation time, costs)

Figure 5.7.19

Applicable technologies (N-y1)

Not applicable technologies (y1)

Applicable technologies (N-y1-y2)

Not applicable technologies (y2)

Applicable technologies (N-y1-y2-y3)

Not applicable technologies (y3)

Applicable technologies (N-y1-y2-y3-y4)

Not applicable technologies (y4)

ISIRE program.

technical solutions are available to solve speciﬁc problems or whether it is possible to study in depth only the eﬃcient technologies. ISIRE gives also a qualitative indication about costs and remediation time. The program has four main dialog windows, as follows. A research table invites the user to select the contaminants, the depth to restore, the concentration, the conductivity, the subsoil homogeneity and the saturation. In the dialog window, which contains the results, is pointed out two categories of available technologies for the area described by the user: – Characteristics table: the user selects the technology, then the ﬁelds in which the technology has been applied with success and the ones in which it has been less eﬃcient are shown schematically through check boxes. Nevertheless, all the outlines shown could be more or less complete according to the technology development status. In the same Characteristics table there are links to other windows dialogue with more information about the selected technology: enhancements, ﬁgure, pre-requirements, valuations, technology development status and references. – Combination: this checks the possible combination of more technologies.

INCORE: Integrated Concept for Groundwater Remediation

337

The creation of ISIRE has reached the aim of developing a tool helpful in the research of the best available in situ technology. In this way, the time to search and choose the best available in situ technologies is shortened. Moreover it is also easy to consult for non-experts. The free download invites potential users to try it so that the ﬁrst version can be further developed by the input of users.

5.7.5

Implementation

5.7.5.1

Administrative Aspects

Management of contaminated sites aims at the assessment, safeguarding and restoration of healthy environmental conditions. Special emphasis is placed on the defence of risk or damages for human health. Management of contaminated sites is an important public task; the measures required must be well deﬁned and enforced by public authorities. The INCORE approach for management of contaminated sites is based on the strategy of the European Union Water Framework Directive. The legal requirements basically lead to particular solutions for single sites. Very few regulations for the management of complex, overlapping groundwater contamination exist under prevailing law in the countries considered. On the other hand, there are no signiﬁcant conﬂicts between the integral INCORE approach and the prevailing national laws.

5.7.5.2

Cost Considerations

The costs for investigations in cycle I and II can roughly be compared to the conventional approach. To determine the location of the contaminant plumes, their extent and the sources of the groundwater contamination, the INCORE methodology proposes to establish a number of control planes and to perform IPTs and on-site analysis. The level of knowledge achieved by this approach is represented by the ‘‘level of conﬁdence,’’ which can be estimated by calculating the probability of the plume length for each contaminant based on plume length statistics. The level of conﬁdence achieved and the size of the area investigated using the INCORE methodology diﬀer from conventional investigation, where groundwater contamination is detected point by point in the source zone, usually without considering the plume. This allows only a limited comparison of the results and the costs of conventional and integrated methodologies. Figure 5.7.20 gives the cost structures for IPTs performed in four project areas broken down by the diﬀerent tasks. This reveals signiﬁcant diﬀerences between the project areas depending on the diﬀerent contaminants being considered, the possibility to use or partly use existing wells, as well as varying hydrogeological conditions. Table 5.7.3 summarises the costs for integral pumping tests. Although the costs vary quite widely, it is obvious that the most inﬂuencing factors are the length of the control planes and the number of wells. The overall costs were

338

Figure 5.7.20 Table 5.7.3

Chapter 5.7

Costs structures for integrated preliminary investigation (%).

Integrated preliminary investigation costs.

Integral pumping test costs Number of wells Total costs/length of control plane Total costs/squaremeter of costs section Total costs/pumped volume Total costs/well

Unit

Stuttgart

Strasbourg Linz

Milan

h

751 400

124 000

162 000

135 260

h per m

34 289.00

2 354.29

10 294.55

7 294.04

h per m2

96.33

47.80

32.73

19.60

h per m3

9.01

3.61

4.42

1.05

h per well

22 100

62 000

16 200

19 300

highest in Germany where an investigation comprising 34 wells was conducted. The total costs were lowest in France, because no drilling was carried out. However, the costs per well in France are about three times higher than in the other cities, because of relatively high ﬁxed costs for modelling and pumping (because of the high transmissivity of the aquifer). It can be concluded that on average the costs per well can be approximated as about h20 000 per well and about h300 per metre of the length of a control plane. The comparison with the conventional approach, which is assumed with h20 000 per site as average costs, concludes that the decision between conventional and integral investigation should be made on a case-by-case basis. Considering

339

INCORE: Integrated Concept for Groundwater Remediation

Cycle I: Plume Screening definition: objectives, project area; present and future land/gw use data inventory, conceptual site model integral pumping test; pollution analysis criteria A mean / max concentration baseline health risk ass., risk based concentration

criteria C

criteria B contaminant load, total mass flux

c riteria D delimitation of area under consideration

delimitation of boundaries of contamination

strength of groundwater impact

mapping of no-information-area

pollution not relevant

evaluation of

no further action

criteria A-C pollution relevant

Cycle II: Source Screening dynamic investigation using on-site analysis finger printing soil and groundwater

matching fingerprints

source intensity fits to load in plume

yes

updated conceptual/geological model

yes

further "remediation" is needed

no

yes

no

yes

no further action

monitored natural attenuation

periodical monitoring

Cycle III: Sourceand / or Plume Remediation feasibility study - site specific areal/combined solutions technical evaluation using the computer program ISIRE

economical evaluation risk assessment according to riskbased land management concept

implementation of remedition technique

Figure 5.7.21

The ﬂow chart.

large areas with many potential sources of contamination, the integral INCORE approach leads to signiﬁcantly reduced costs. This INCORE phased approach is illustrated in a strategy ﬂowchart in Figure 5.7.21. Following this chart guides the application of the approach and,

340

Chapter 5.7

where required, enables the user to go straight to more detailed explanations in the extended report (enclosed on the CD).

5.7.5.3

Implementation in Projects

INCORE has provided guidance for a structured management approach for dealing with groundwater contamination in large industrialised areas. This approach consists of several technical tools embedded with a concept for related administrative actions. The following projects are already on the way in partner cities. The municipality of Stuttgart acquired the ownership of an area around the present central railway station from the Deutsche Bahn AG. The station will be relocated underground and the existing infrastructure will be phased out between now and 2013. This area, ‘‘Stuttgart 21,’’ covers about 120 ha and is located in the heart of Stuttgart city centre. Over the last 100 years land use closely connected to railway activities has caused a complex contamination situation in soil and groundwater. The INCORE approach was applied to this area in 2002–2003. In June 2005, the city started an integral groundwater investigation in the Feuerbach district. This district has had intensive industrial use over many years. Nowadays the integral approach is the standard procedure for complex cares in Stuttgart and Baden-Wu¨rttemberg. The public administration of Milan has to manage and investigate contamination of soil and groundwater on a large scale, in order to protect the public drinking water supplies (from about 600 wells). By applying the INCORE strategy over the whole city, the municipality intends to be able to prioritise its remediation work, starting with the areas that are a real risk for the health of citizens.

References 1. G. Eglington and M. Calvin, Chemical fossils, Sci. Am., 1967, 261, 32–43. 2. J. M. Moldowan, P. Sundararaman, T. Salvatori, A. Alajbeg, B. Gjukic, C. Y. Lee and G. J. Demaison, Source correlation and maturity assessment of select oils and rocks from the Central Adriatic Basin (Italy and Yugoslavia), in Biological Markers in Sediments and Petroleum, ed. J. M. Moldowan, P. Albrecht and R. P. Philp, Prentice Hall, Englewood Cliﬀs, NJ, 1992, pp. 370–401. 3. K. E. Peters, J. M. Moldowan, The Biomarker Guide, Prentice Hall, Englewood Cliﬀs, NJ, 1993, p. 363. 4. L. K. Volkman, R. Alexander, R. I. Kagi, R. A. Noble and G. W. Woodhouse, Geochim. Cosmochim. Acta, 1983, 47, 2091–2106. 5. L. K. Volkman, R. Alexander, R. I. Kagi and G. W. Woodhouse, Geochim. Cosmochim. Acta, 1983, 47, 785–794.

INCORE: Integrated Concept for Groundwater Remediation

341

6. I. R. Kaplan, Y. Galperin, H. Alimi, R. P. Lee and S. T. Lu, Patterns of chemical changes during environmental alteration of hydrocarbon fuels, Groundwater Monit. Remed., 1996, Fall issue, 113–125. 7. H. Dempster, B. Lollar and S. Feenstra, Environ. Sci. Technol., 1997, 31(11), 3193–3197. 8. D. Hunkeler, P. Ho¨hener, S. Bernasconi and J. Zeyer, J. Contam. Hydrol., 1999, 37, 201–223. 9. INCORE project, ﬁnal report: UW Umwelt Wirtschaft GmbH, Stuttgart, Germany, 2003 (http://www.umweltwirtschaft-uw.de/incore/). 10. G. S. Douglas, A. E. Bence, R. C. Prince, S. J. McMillan and E. L. Butler, Environmental stability of selected petroleum hydrocarbon source and weathering ratios, Environ. Sci. Technol., 1996, 30, 2332–2339.

6. Groundwater Monitoring

CHAPTER 6.1

Groundwater Monitoring in the Policy Context JOHANNES GRATH,a ROB WARDb AND ANDREAS SCHEIDLEDERa a

Umweltbundesamt GmbH, Spittelauer Laende 5, AT-1090 Wien, Austria; Environment Agency—England and Wales, Olton Court, Solihull, West Midlands, B92 7HX, UK

b

6.1.1

Introduction

Groundwater is an important resource across Europe. It supplies a high proportion of Europe’s drinking water needs, this proportion being as high as 100% in Denmark, 99% in Austria, 84% in Iceland, 83% in Switzerland and 80% in Italy.1 It also supports many industries and is of vital importance for supporting the base ﬂow to rivers and groundwater-dependent terrestrial and aquatic ecosystems (wetlands). In 1991 the environment ministers of the European Union (EU) initiated a process for the protection of groundwater in Europe. As a result of this, in 1996, a ‘‘Groundwater Action Programme’’ was proposed (see Chapter 3.1), the recommendations of which have been considered in the development of the legal framework of Directive 2000/60/EC of the European Parliament and Council establishing a framework for Community action in the ﬁeld of water policy: the so-called Water Framework Directive (WFD).2 Complementary regulations to support the WFD, and in particular Article 17, have also been developed, leading in particular to the adoption of the so-called Groundwater Daughter Directive setting criteria for establishing environmental quality standards, trend and trend reversal, and establishing provisions for preventing and limiting direct input of pollutants into groundwater (see details in Chapter 3.1). The WFD provides a framework for the protection of the water environment including groundwater. It sets out a series of environmental objectives that member states of the EU are required to achieve by implementing programmes of measures. The achievement of the objectives for groundwater means that 345

346

Chapter 6.1

groundwater bodies will be at good status by the end of 2015. It is recognised that there are diverse conditions and needs within the EU which require diﬀerent speciﬁc solutions and that this diversity needs to be taken into account when planning and executing the programmes of measures. It is therefore necessary to undertake an analysis of the characteristics of river basins and the impacts of human activity as well as an economic analysis of water use. This should be supported by monitoring in a systematic and comparable way throughout the Community. This information is needed to provide a sound basis for identifying where programmes of measures are needed and assessing their success, or otherwise, in achieving the relevant environmental objectives. The WFD requires that monitoring programmes are designed and implemented. It also requires the use of data from these programmes to support diﬀerent elements of the WFD, e.g. characterisation and risk assessment, status and trend assessment, assessing eﬀectiveness of programmes of measures and supporting the achievement of drinking water protected area objectives. For groundwater both quantitative (level) and chemical (quality) monitoring is required. To provide support to the member states of the EU and to harmonise the implementation, the Common Implementation Strategy (CIS) was initiated. This activity is based on the contributions of member state experts, the scientiﬁc community and stakeholders. It allows for broad involvement of all relevant groups. Further details can be found in Chapter 4.1. One output of the CIS is the ‘‘Monitoring Guidance for Groundwater.’’3 This resulted from the establishment of a Groundwater Working Group and a smaller drafting group made up of 28 groundwater monitoring experts within a larger expert network (the CIS Working Group C on Groundwater; see Chapter 4.1). This allowed a range of expertise and viewpoints to be considered when developing the guidance. The EU ‘‘Monitoring Guidance for Groundwater’’3 will support the implementation of national WFD-compliant groundwater monitoring programmes. The establishment of high-quality, long-term monitoring programmes is essential if the implementation of the WFD is to be eﬀective and achievement of its objectives realised. They will provide the information needed for planning, decision-making and assessing the achievement of environmental objectives. The groundwater monitoring and management activities required for the WFD are outlined in Figure 6.1.1. The WFD provisions and the groundwater management objectives are at the centre of the monitoring cycle. This means that the overall process is driven by policy and environmental needs but there must also be continuous interaction between these, monitoring programme management activities and technical implementation. As stressed above, the overall goal in the WFD for groundwater is the achievement of good groundwater status by 2015. The assessment of groundwater status is based on an assessment of both chemical status and quantitative status. The assessment unit is the groundwater body and this is deﬁned in the

347

Groundwater Monitoring in the Policy Context Reporting

conceptual model

modelling & assessment WFD and Management Objectives

monitoring design

sampling data management laboratory analyses

Figure 6.1.1

Monitoring: cyclic procedure according to the Water Framework Directive.3

WFD Article 2 (12) as: ‘‘A distinct volume of groundwater within an aquifer or aquifers.’’2 Further details for groundwater body delineation and characterisation are given in Ref. 4. The results from chemical and quantity monitoring are the basis for status assessment. As indicated in Figure 6.1.1, the conceptual model of a groundwater body is the foundation for the monitoring design. The general principles for the conceptual model of a groundwater body and network design, covering monitoring principles, site selection, parameter selection and monitoring frequency, are the subject of the following sections. Other important issues concerning groundwater monitoring, like sampling, quality assurance, laboratory analysis/screening methods and assessment, are dealt with in other chapters of this book, respectively Chapters 6.2, 6.3 and 9.1. While monitoring obligations are mainly embedded into the WFD regulatory framework (Article 8, Annex V), they are also closely related to the new Groundwater Directive, in particular regarding compliance to groundwater quality standards and threshold values, trend identiﬁcation and reversal, and measures to prevent/limit inputs of pollutants into groundwater. Details on this directive are provided in Chapter 3.1.

6.1.2

Monitoring Requirements: Legal Background and Objectives

Article 8 of the WFD requires the establishment of programmes for the monitoring of groundwater. WFD groundwater monitoring is focused primarily on the groundwater body but it also supports the overall management of the river basin district and the achievement of its environmental objectives. The groundwater monitoring programmes must provide the information necessary to assess and demonstrate whether relevant environmental objectives according to Article 4 of the WFD are being met (see Chapter 3.1).

348

Chapter 6.1

The WFD sets out the requirements for a stepwise approach which comprise: delineation of groundwater bodies and groups of groundwater bodies (see Annex II, WFD2 and the CIS guidance no. 2 ‘‘Identiﬁcation of Water bodies’’4); characterisation of groundwater bodies, the hydro(geo)logical conditions and the pressure situation (Annex II, WFD); assessment as to whether there is risk of failing to meet the environmental objectives for each groundwater body; and establishment of monitoring programmes. The monitoring programmes comprise the following. Groundwater quantity monitoring, to supplement and validate the Article 5 characterisation and risk assessment procedure with respect to risks of failing to achieve good groundwater quantitative status in all groundwater bodies, or groups of bodies. Its principal purpose, however, is to enable assessment of the quantitative status of a groundwater body. Groundwater quality (chemical) monitoring. This is subdivided into: – surveillance monitoring to (a) supplement and validate the Article 5 characterisation and risk assessment procedure with respect to the risks of failing to achieve good groundwater chemical status, (b) provide information for use in the assessment of long-term trends in natural conditions and in pollutant concentrations resulting from human activity and (c) to identify, in conjunction with the risk assessment, the need for operational monitoring; – operational monitoring to (a) determine the chemical status of each groundwater body identiﬁed as being ‘‘at risk’’ and (b) identify the presence of signiﬁcant and sustained upward trends in the concentration of pollutants and their reversal; and – appropriate monitoring to support the achievement of Drinking Water Protected Area (DWPA) objectives. The results of the monitoring should also be used to: assist in further characterisation of groundwater bodies; estimate the direction and rate of ﬂow in groundwater bodies that cross member states’ boundaries; assist in the design of programmes of measures; evaluate the eﬀectiveness of programmes of measures; and characterise the natural quality of groundwater including natural trends (baseline). Speciﬁc provisions concern those bodies of groundwater that cross the boundary between two or more member states. Bilateral agreement should be reached on monitoring strategies, which requires coordination of conceptual model

Groundwater Monitoring in the Policy Context

349

development and the exchange of quality assured data and information. Surveillance monitoring in all transboundary groundwater bodies must include monitoring of those parameters which are relevant for the protection of all uses supported by the groundwater ﬂow.

6.1.3

General Principles

The monitoring programmes must provide the information necessary to assess whether the WFD environmental objectives are being achieved. This means that a clear understanding of the environmental conditions required for the achievement of the objectives, and of how these could be aﬀected by human activities, is essential for the design of eﬀective monitoring programmes. The monitoring programmes should therefore be designed on the basis of the results of the groundwater body characterisation and risk assessment procedure and the conceptual model/understanding of the groundwater system in which the general scheme of ‘‘recharge–pathway–discharge’’ is known. It is recognised that a very wide variability in groundwater systems exists across Europe and this is considered within the published guidance. A prescriptive approach is not appropriate or possible. Instead the guidance advocates a risk-based approach that considers the overarching objectives and the considerable variation in environmental factors that will inﬂuence monitoring programme design and operation. In general this will mean that monitoring will be targeted in bodies that are at risk and the amount of monitoring required (number of points and sampling frequency) will generally be proportional to (a) the diﬃculty in judging the status of the groundwater body, (b) identifying the presence of adverse trends and (c) the implications of errors in such judgements, in particular with regard to setting up programmes of measures.

6.1.3.1

Conceptual Model

A fundamental requirement is an understanding of the groundwater body. This is supported by developing a conceptual model that draws together all the available information on the body: physical, chemical and biological. This conceptual model is a simpliﬁed representation, or working description, of the hydrogeological system being studied. It provides the link between the characterisation and risk assessment process and monitoring. Figure 6.1.2 outlines the principles and relationship of the conceptual model to the monitoring programme.

6.1.3.2

Three-dimensional Characteristics and Variability

Groundwater systems are three-dimensional in nature and considerable variation in water chemistry and level can exist both spatially and vertically. This variation must be considered as part of the conceptual model development and in the design of the monitoring programmes.

350

Figure 6.1.2

Chapter 6.1

Link between the conceptual model/understanding and monitoring.4

The monitoring network needs to be able take into account this variability along with the pressures and objectives of monitoring. The selection of appropriate monitoring site locations, site types and spatial coverage should be informed by considering: vertical variation in hydraulic properties; existing quality and/or quantity data (length, frequency, range of parameters and evidence of vertical stratiﬁcation in groundwater quality); construction characteristics of existing sites and the abstraction regime; the spatial distribution of existing sites compared to the scale of the groundwater body; practical considerations relating to easy and long-term access, security, health and safety; travel times and/or groundwater age data; and monitoring site types. The selection of appropriate monitoring site types within a monitoring network at groundwater body level should again be based on an understanding of the objectives of monitoring. An understanding of the groundwater system and its behaviour is also important to ensure that the most appropriate monitoring point type is selected. Where appropriate, integrated monitoring—where a site is used for more than one purpose—should be considered. An example of this is the use of a spring for indicating the chemistry of groundwater and surface water and also ﬂows and discharges. This will not only contribute signiﬁcantly to cost

351

Groundwater Monitoring in the Policy Context

eﬃciency but also enable the interactions between programmes, environmental compartments (e.g. groundwater and surface water) and water body types to be investigated and better understood. Detailed information about each monitoring site should be recorded and the information frequently reviewed to ensure that it is still valid and the monitoring point making a valid contribution to the programme. The information needed to characterise a sampling site is summarised in Table 6.1.1.

6.1.3.3

Aquifer Types

A diverse range of hydrogeological settings and aquifer types is found across Europe. This broad variation has major implications for the number of monitoring sites required, the suitability of diﬀerent types of sampling Table 6.1.1

Monitoring point information: essential and desirable factors. Chemical monitoring points

Quantitative monitoring points

E D E

E D E

E E E

E E E

E E

D E

D E

D D

E

D

E E D

E E E

D D D

E D D

Static or rest water level Datum elevation and description of datum

D D

E E

Artesian/overﬂowing Borehole log (geological) Aquifer properties (transmissivity, hydraulic conductivity, etc.)

E D D

E D D

Factor Aquifer(s) monitored Location map and photographs Location (grid reference), name of monitoring point and unique identiﬁer Groundwater body that monitoring point is within Purpose(s) of monitoring site Type of monitoring point: farm borehole, industrial borehole, spring, etc. Depth and diameter(s) of boreholes/wells Description of headworks: grouting integrity, slope of ground around borehole Depth of screened/open sections of boreholes/wells Vulnerability or indication of subsoil thickness and type at monitoring point Visual appraisal of recharge area (including land use and pressures, potential sources of point pressures) Construction details Amount abstracted or total discharge (at springs) Pumping regime (qualitative description: intermittent, continuous, overnight, etc.) Drawdown (pumped water level) Zone of contribution/recharge area Pump depth

E. . .Essential, D. . .Desirable.

352

Chapter 6.1

installation and how eﬀectively they represent changes in groundwater systems. Monitoring design needs to be tailored accordingly. For all groundwater bodies, there is a need to consider the characteristics of the strata forming the aquifers with regard to ﬂow paths and ﬂow mechanisms, storage, unsaturated zone thickness, groundwater recharge and discharge before determining the most appropriate means of monitoring. The scale of the groundwater body, i.e. whether there are local and rapid ﬂow paths or much longer and slower regional ones, and the geological controls on ﬂow, e.g. intergranular or fracture ﬂow and the impact on response times, are key factors that must be considered in site selection and operation.

6.1.4

Chemical Status and Trend Monitoring

6.1.4.1

Overall Objectives

Groundwater monitoring programmes are required to provide a coherent and comprehensive overview of water status within each river basin, to detect the presence of long-term anthropogenically induced trends in pollutant concentrations and to ensure compliance with protected area objectives. A groundwater body will be at good chemical status if the following criteria are satisﬁed. General water quality. The concentrations of pollutants should not exceed the quality standards applicable under other relevant Community legislation in accordance with Artile 17, WFD. Impacts on ecosystems. The concentration of pollutants should not be such as would result in failure to achieve the environmental objectives speciﬁed under Article 4 for associated surface waters nor any signiﬁcant diminution of the ecological or chemical quality of such bodies nor any signiﬁcant damage to terrestrial ecosystems which depend directly on the groundwater body. Saline intrusion. The concentrations of pollutants should not exhibit the eﬀects of saline or other intrusions as measured by changes in conductivity. The WFD requires surveillance and operational monitoring programmes to be established to provide the information needed to support the assessment of chemical status and the identiﬁcation and monitoring of pollutant trends.

6.1.4.2

Surveillance Monitoring Programme

Surveillance monitoring focuses on the groundwater body as a whole. A ‘‘surveillance monitoring’’ programme is required to: validate risk assessments: supplement and validate the characterisation and risk assessment procedure with respect to risks of failing to achieve good groundwater chemical status;

Groundwater Monitoring in the Policy Context

353

classify groundwater bodies: conﬁrm the status of all groundwater bodies, or groups of bodies, determined as not being at risk on the basis of the risk assessments; and assess trends: provide information for use in the assessment of long-term trends in natural conditions and in pollutant concentrations resulting from human activity. Surveillance monitoring is required in bodies or groups of bodies both at risk and not at risk of failing WFD objectives. The programme must be carried out during each river basin management cycle, irrespective of whether the groundwater body (or group of bodies) is at risk. Surveillance monitoring should be undertaken in each plan period and to the extent necessary to adequately supplement and validate the risk assessment procedure for each body or group of bodies of groundwater. The surveillance monitoring programme is also needed for deﬁning natural background quality within the groundwater body. This will enable future changes in conditions to be assessed, reference data to be acquired and typologies to be investigated. This information will be useful for characterising transboundary water bodies and as a basis for European-wide reporting.

6.1.4.3

Operational Monitoring Programme

Operational monitoring is focused on the groundwater body as a whole. An ‘‘operational monitoring’’ programme is required to determine: the chemical status of all groundwater bodies, or groups of bodies, determined as being ‘‘at risk’’ the presence of any long-term anthropogenically induced upward trends in the concentration of any pollutant; and it can also be used to assess the eﬀectiveness of programmes of measures implemented to restore a body to good status or reverse upwards trends in pollutant concentrations. Operational monitoring is required only in bodies ‘‘at risk’’ of failing to meet WFD objectives. It should be carried out during the periods between surveillance monitoring. In contrast to surveillance monitoring, operational monitoring focuses on assessing the speciﬁc, identiﬁed risks that may compromise the achievement of the WFD’s objectives. In the following sections general criteria for selection of monitoring sites, determinands and monitoring frequency—valid for both surveillance and operational monitoring—are further described.

6.1.4.4

Selection of Representative Monitoring Sites

The selection of sampling sites and their operation is of major importance for the results of the later assessment procedure especially as contaminants are

354

Chapter 6.1

often unevenly distributed across a body of groundwater. The spatial distribution of contaminants is related to the location of diﬀerent pressures, e.g. point and diﬀuse sources (diﬀerent types of land use) and their subsequent migration within the groundwater system. A body of groundwater is threedimensional and the concentration of contaminants may also vary signiﬁcantly with depth and over short timescales along with other physicochemical parameters (e.g. electrical conductivity, temperature and contaminant concentrations). The site selection process should be based on three main factors: the conceptual model(s) including assessment of the hydrological, hydrogeological and hydrochemical characteristics of the body of groundwater including characteristic travel times, distribution of diﬀerent types of land uses (e.g. settlement, industry, forest, pasture/farm land), pathway susceptibility, receptor sensitivity and existing quality data; assessment of risk and the level of conﬁdence in the assessment, including the distribution of key pressures; and practical considerations relating to the suitability of individual sampling points, e.g. sites need to be easily accessed, secure and be able to provide long-term access agreements. An eﬀective monitoring network will be one in which the sites are able to monitor for the potential impacts of identiﬁed pressures and the evolution of groundwater quality along the ﬂow paths within the body. Where risk issues relate to speciﬁc receptors such as ecosystems, additional sampling points may need to be focussed in areas that are close to these receptors. In these cases, where the location of pressures (point sources) is well known, sampling points will often be used to help isolate impacts from diﬀerent pressure types, assess the areal extent of impacts and determine contaminant fate and transport between the pressure and the receptor. Additional aspects to be considered for site selection are as follows. Large abstractions and springs may provide suitable sampling sites, as they draw water from a large area and volume of aquifer particularly in homogeneous systems—preferably for surveillance monitoring, since the objective is not to isolate the impact of individual pressures and the eﬀectiveness of a programme of measures. Springs are particularly recommended in karstic or shallow fracture ﬂow dominated aquifers. However, a representative monitoring network should ideally be based on a balanced mixture of diﬀerent sampling site types as well as sampling site uses (e.g. abstraction, monitoring). In some hydrogeological systems where the groundwater contributes signiﬁcantly to the (base)ﬂow of the surface water course, then sampling of the surface water may provide a representative groundwater sample. Representativity. In some aquifer systems, stratiﬁcation may occur. In this case the location of monitoring points must be focused on those parts

Groundwater Monitoring in the Policy Context

355

of the groundwater body that are most susceptible to pollution. This will often be the upper parts. However, to provide a representative assessment of the distribution of contaminants for the groundwater as a whole additional monitoring in other parts of the groundwater body is also required. ‘‘At risk’’ bodies. Surveillance monitoring sites will provide the basis for the operational monitoring, i.e. based on the results the network can be adapted accordingly. Sites could be used for both programmes. ‘‘Not at risk’’ bodies where conﬁdence in the risk assessment is low. The number of monitoring points should be suﬃcient to be representative of the range of pressure and pathway conditions in the groundwater body (or group of bodies) with the aim of providing the data necessary to supplement the risk assessment, i.e. increase conﬁdence. The location of sampling points may therefore be focused on the most susceptible areas of the groundwater body or bodies for each pressure/pathway combination. The ﬁnal distribution per grouping will depend on availability of suitable surveillance sites and the distribution of pressures. As a general guide of WG 2.8, a minimum of three points in a groundwater body or group of bodies is recommended.4 However where groundwater bodies are large and heterogeneous, it is likely that signiﬁcantly more monitoring points will be needed to meet the monitoring objectives. Potential linkages with existing/planned surface water monitoring sites and other monitoring programmes that have complementary needs. The cost of installing the infrastructure needed at a monitoring site is an important factor. Hence many programmes depend, to a large extent, on existing wells or boreholes. It is important therefore that the characteristics of the site are well known and there is conﬁdence in the site’s suitability. To support the assessment of the suitability of monitoring sites and aid the interpretation of monitoring results, detailed site information should be available and documented for each site. Table 6.1.1 gives an example of the type of information needed. At quantitative monitoring sites the objective is to monitor the groundwater level in order to assess the impacts of abstractions at the groundwater body scale. Therefore: monitoring points should not be pumped or should only be pumped for very short periods so that measured water levels are not aﬀected by localised abstraction; and the locations of monitoring points should be outside the immediate hydraulic inﬂuence of abstraction pressures such that day-to-day variations in pumping will not be evident in the data. Note that data from stations which function as continuous abstraction wells may be acceptable if accompanied by detailed (e.g. hourly) pumping records.

356

6.1.4.5

Chapter 6.1

Selection of Monitoring Determinands

The recommended core set of determinands comprises dissolved oxygen, pH value, electrical conductivity, nitrate, ammonium, temperature and a set of major and trace ions. Parameters such as temperature and a set of major and trace ions are not formally required by the WFD but may be helpful to validate the risk assessment and the conceptual models. Selective determinands (e.g. heavy metals and relevant basic radio nuclides) will be needed for assessing natural background levels. Additional indicators of anthropogenic contaminants typical of land use activities in the area and with the potential to impact on groundwater will also be required on an infrequent basis to provide additional validation of WFD risk assessments and to check for impact from any new pressures. In addition at all sites monitoring of the water level is recommended in order to describe (and interpret) the ‘‘physical status of the site’’ and to interpret (seasonal) variations or trends in chemical composition of groundwater. For operational monitoring in most cases, both core and selected determinands will be required at each sampling station. For the selection of additional parameters analysis of pressures identiﬁed in the risk characterisation should be considered. These could include agricultural, industrial and municipal activities, waste disposal activities, transport and groundwater (over)exploitation (which may lead to intrusion of pollutants). Further, the characteristics of the pollutants, their physicochemical and (eco)toxicological properties also need to be considered in the selection process.

6.1.4.6

Monitoring Frequency

The selection of appropriate monitoring frequency will generally be based on the conceptual model and knowledge of groundwater quality/level behaviour from existing monitoring data. Where there is adequate knowledge of the groundwater system and a long-term monitoring programme is already established this should be used to determine an appropriate frequency for surveillance monitoring. Where knowledge is inadequate and data are not available a schema has been proposed for initial monitoring frequencies that takes into account diﬀerent aquifer characteristics and responses. The suggested initial frequencies for surveillance monitoring for diﬀerent aquifer types are shown in Table 6.1.2. The results of surveillance monitoring should be reviewed on a regular basis and frequencies adjusted accordingly to ensure that the information requirements, including those for trend assessment, are fully met and a cost-eﬀective programme maintained. In a similar way to surveillance monitoring a schema for operational monitoring has also been proposed (Table 6.1.3). Where there is a good understanding of groundwater quality and the behaviour of the hydrogeological system, alternative monitoring frequencies can be adopted as necessary.

Twice per year Annual Every 6 years

Annual Every 6 years

Quarterly

Annual

Quarterly

Shallow ﬂow

Every 6 years

Annual

Twice per year

Quarterly

Fracture ﬂow only

–

Twice per year

Twice per year

Quarterly

Karst ﬂow

Note: This table proposes monitoring frequencies that can be used as a guide where the conceptual understanding is limited and existing data are not available. Where there is a good understanding of groundwater quality and the behaviour of the hydrogeological system, alternative monitoring frequencies can be adopted as necessary.

Generally high-mod transmissivity Generally low transmissivity Additional parameters (on-going validation)

Long term frequency – core parameters

Twice per year Every 2 years Every 6 years Every 6 years

Conﬁned

Signiﬁcant deep ﬂows common

Intergranular ﬂow signiﬁcant

Unconﬁned

Aquifer ﬂow type

Proposed monitoring frequencies for surveillance monitoring (where understanding of aquifer systems is inadequate).

Initial frequency – core & additional parameters

Table 6.1.2

Groundwater Monitoring in the Policy Context 357

Trend assessments

Lower vulnerability groundwater

Continuous pressures Seasonal/intermittent pressures Continuous pressures Seasonal/intermittent pressures Annual

Annual Annual

Annual Annual

Conﬁned

Twice per year

Annual Annual

Twice per year Annual

Signiﬁcant deep ﬂows common

Twice per year

Twice per year As appropriate

Twice per year As appropriate

Shallow ﬂow

Twice per year

Twice per year As appropriate

Quarterly As appropriate

Fracture ﬂow only

Unconﬁned

Aquifer ﬂow type Intergranular ﬂow signiﬁcant

Proposed frequencies for operational monitoring.

Higher vulnerability groundwater

Table 6.1.3

–

Quarterly As appropriate

Quarterly As appropriate

Karst ﬂow

358 Chapter 6.1

Groundwater Monitoring in the Policy Context

359

Sampling frequency and sample timing at each monitoring location should furthermore consider the following. Requirements for trend assessment. The proximity to pressure(s). The level of conﬁdence in risk assessments, and changes in the assessments over time. Short-term ﬂuctuations in pollutant concentrations, e.g. seasonal eﬀects. Where seasonal and other short-term eﬀects are likely to be encountered, it is essential that sampling frequencies and timings are adjusted (increased) accordingly and that sampling takes place at the same time(s) each year, or under the same conditions, to enable comparable data. Land use management patterns, e.g. the period of pesticides or nitrate application. This is especially important for rapid ﬂow system like karstic aquifers and/or shallow groundwater bodies.

6.1.5

Quantity (Water Level) Monitoring

6.1.5.1

Overall Objective

A groundwater level monitoring network is required to assist in characterisation, to determine the quantitative status of groundwater bodies, to support the chemical status assessment and trend analysis and to support the design and evaluation of the programme of measures. A groundwater body will be at good quantitative status if: the available groundwater resource is not exceeded by the long-term annual average rate of abstraction; the groundwater levels and ﬂows are suﬃcient to meet environmental objectives for associated surface waters and groundwater-dependent terrestrial ecosystems; and anthropogenic alterations to ﬂow direction resulting from level change do not cause saline or other intrusion. As with other networks, the monitoring design should be based on a conceptual understanding of the groundwater system and the pressures. The key elements of the quantitative conceptual understanding will be: assessments of recharge and water balance; existing groundwater level or discharge assessments and relevant information on the risks for groundwater-dependent surface waters and groundwater-dependent terrestrial ecosystems; and the degree of interaction between groundwater and related surface and terrestrial ecosystems where this interaction is important and could potentially cause the surface water body status to be aﬀected.

360

Chapter 6.1

The development of a quantitative monitoring network can be iterative; data collected from new monitoring points being used to enhance and reﬁne the conceptual model used to locate each monitoring point in the groundwater body as a whole and the operation of the quantitative monitoring programme. Implementation of a numerical groundwater model or a hydrological model integrating groundwater and surface water are useful tools in compiling and interpreting quantitative monitoring data and identifying resources and ecosystems at risk. Furthermore, the uncertainty estimates that can be obtained with a numerical model can help identify parts of a groundwater body where additional data points will add most to the description of groundwater quantity and ﬂow.

6.1.5.2

Monitoring Parameters

Although the WFD identiﬁes groundwater level as the metric for determining quantitative status, in practice, the requirements of status assessment mean that additional supporting information will be required. Recommended supporting parameters for the purposes of quantitative assessment of groundwater include: groundwater levels in boreholes or wells; spring ﬂows; ﬂow characteristics and/or stage levels of surface water courses during drought periods (i.e. when the ﬂow component directly related to rainfall can be neglected and discharge is sustained substantially by groundwater); stage levels in signiﬁcant groundwater-dependent wetlands and lakes; chemical and indicator parameters (e.g. temperature, electrical conductivity) monitoring for saline or other intrusions (for island aquifers it may also be appropriate to monitor the fresh/saline water transition zone) rainfall and the components required to calculate evapo-transpiration (to calculate groundwater recharge); ecological monitoring of groundwater-dependent terrestrial ecosystems (including ecological indicators); and groundwater abstraction (and artiﬁcial recharge). Speciﬁc requirements for the supportive monitoring data, to supplement the knowledge gained from groundwater level monitoring will largely be determined by the tools/methods employed to support the assessment of risk or status and the conﬁdence required in this assessment. Key to parameter selection is how representative the parameter is of the hydrogeological setting being monitored and the signiﬁcance of its role in determining risk or status. In some hydrogeological settings monitoring groundwater levels in a borehole may be inappropriate for the purposes of the WFD and in some cases highly be misleading. In these circumstances the ﬂow characteristics of associated watercourses or springs may provide better data with which to undertake

Groundwater Monitoring in the Policy Context

361

an assessment. This is most likely to be the case in low permeability/fractured aquifers. There are cases when the water level remains more or less stable but water from other aquifers, surface waters or even seawater is intruding. Speciﬁc conditions should be considered for groundwater bodies on islands. If there is the risk of waters intruding, then appropriate water quality indicators should be monitored, e.g. electrical conductivity and water temperature.

6.1.5.3

Selection of Monitoring Density

Monitoring may be required at two diﬀerent scales to meet the various requirements of the environmental objectives (Article 4, WFD). Firstly, where possible, groundwater levels and ﬂows across a groundwater body should be assessed. These may be related to the water balance assessment for the body as a whole. Secondly, more focused ‘‘local’’ monitoring of levels and ﬂows that relate to relevant local groundwater supported receptors, i.e. surface water bodies (rivers, lakes, estuaries) and groundwater-dependent terrestrial ecosystems, may be needed. The latter may include supporting information, e.g. salinity monitoring (with respect to saline intrusions), or supporting information from ecological monitoring as already performed under other relevant community legislation (as evidence of impact on ecosystems from groundwater abstractions). In groundwater bodies or groups of groundwater bodies assessed as being ‘‘not at risk,’’ the monitoring can be minimised. Indeed, monitoring need not be located in each body within a group, provided that the groups are hydrogeologically comparable. In groundwater bodies or groups of groundwater bodies assessed as being ‘‘at risk,’’ the distribution of monitoring points will reﬂect the need to understand the hydrogeological conditions that relate to the receptors identiﬁed as being ‘‘at risk’’ and to their perceived importance. Monitoring density must be suﬃcient to ensure proper assessment of impacts due to abstractions and discharges on groundwater level. Speciﬁc provisions concern those bodies of groundwater which cross the boundary between two or more member states, such as the location of groundwater abstraction points providing more than 10 m3 a day or serving more than 50 persons, the abstraction rates, direct discharges to groundwater, etc. The number of sampling sites should be suﬃcient to be able to estimate the direction and rate of groundwater ﬂow across the member state boundary.

6.1.5.4

Monitoring Frequency

The amount and frequency of monitoring will be determined by the data needed to determine risk and status, and where necessary to support the design and assessment of a programme of measures. Frequency of monitoring predominantly depends of the characteristics of the water body and the monitoring site. Sites with signiﬁcant annual variability should be monitored more frequently than sites with only minor variability. In

362

Chapter 6.1

general monthly monitoring will be suﬃcient for quantity monitoring where variability is low but daily monitoring would be preferred (particularly when measuring ﬂows). The frequency should be revised as knowledge of the aquifer response and behaviour improves and in relation to the signiﬁcance of any changes in pressures on the groundwater body. This will ensure that a costeﬀective programme is maintained.

6.1.6

Review and Update

To ensure that the monitoring programmes continue to be ‘‘ﬁt for purpose,’’ the conceptual model should be reviewed and reﬁned as necessary. As the understanding of the hydrogeology and hydrochemistry of the groundwater system improves, the network design should also be reviewed and adapted as required. To support this process the monitoring results obtained from the networks should be interpreted regularly and the monitoring networks and their operation reviewed at least once every six years, but ideally more frequently.

References 1. A. Scheidleder, J. Grath, G. Winkler, U. Sta¨rk, C. Koreimann and C. Gmeiner, Austrian Working Group on Water; S. Nixon and J. Casillas, Water Research Centre; P. Gravesen, Geological Survey of Denmark and Greenland; J. Leonard, International Oﬃce for Water; M. Elvira, Centro de Estudios y Experimentacio´n de Obras Pu´blicas; S. Nixon and ETC-IW Leader T. J. Lack, Environmental Assessment Report No 3: Groundwater Quality and Quantity in Europe, OPOCE (Oﬃce for Oﬃcial Publications of The European Communities), 2000 (http://reports.eea.europa.eu/ groundwater07012000/en). 2. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities, L 327, 22.12.2000. 3. European Commission, Guidance document No. 15, Groundwater monitoring, Common Implementation Strategy, 2007. 4. Common Implementation Strategy for the Water Framework Directive, Communities, No. 2 Identiﬁcation of Water Bodies, No. 7 Monitoring under the Water Framework Directive, Technical Report 1, EU Water Framework Directive: Statistical aspects of the identiﬁcation of groundwater pollution trends and aggregation of monitoring results, 2003 (ISBN 92-894-2040-5). Final CIS document available at: http://ec.europa.eu/environment/water/ water-framework/guidance_documents.html.

CHAPTER 6.2

Screening Methods for Groundwater Monitoring CATHERINE GONZALEZ,a ANNE-MARIE FOUILLACb AND RICHARD GREENWOODc a

Ecole des Mines d’Ale`s, 6 avenue de Clavie`res, FR-30319 Ale`s Cedex, France; b Bureau de Recherches Ge´ologiques et Minie`res (BRGM), 3 avenue Claude Guillemin, FR-45060 Orle´ans ce´dex 2, France; c School of Biological Sciences, University of Portsmouth, King Henri Building, King Henri I street, UK-Portsmouth, PO1 2DY, United Kingdom

6.2.1

Groundwater Monitoring Requirements and Speciﬁc Issues

Article 17 of the European Union (EU) Water Framework Directive (WFD; 2000/60/EC) stipulates the strategies that have to be developed under the daughter directive to prevent and control pollution of groundwater. This new directive (2006/118/EC) on the protection of groundwater against pollution (see Chapter 3.1) aims to improve and to protect the quality of the groundwater and notably stipulates that: In order to protect the environment as a whole, and human health in particular, detrimental concentrations of harmful pollutants in groundwater should be avoided, prevented or reduced. Measures to prevent and control groundwater pollution should be adopted, including criteria for assessing good groundwater chemical status and criteria for the identiﬁcation of signiﬁcant and sustained upward trends in pollutant concentrations and for the deﬁnition of starting points for trend reversals taking into account the likelihood of adverse eﬀects on associated aquatic ecosystems or dependent terrestrial ecosystems. Moreover, it is necessary to ensure the continuity of the protection provided by Directive 80/68/EEC with regard to measures aimed at preventing or limiting both direct and indirect inputs of pollutants into groundwater. 363

364

Chapter 6.2

Groundwater chemical status provisions do not apply to high naturally occurring concentrations of substances or ions, or their indicators, that are not covered by the deﬁnition of pollution and are contained either in a body of groundwater or in associated bodies of surface water due to speciﬁc hydrogeological conditions. Equally, they do not apply to temporary, spatially limited changes in ﬂow direction and chemical composition that are not regarded as intrusions. However, where there are existing plumes of pollution in bodies of groundwater that may threaten the achievement of the objectives of the directive, and particularly where these result from point sources and contaminated land, EU member states should carry out additional trend assessments. The latter are necessary in order to verify that plumes of identiﬁed pollutants from contaminated sites do not expand to reduce the chemical status of the body or group of bodies of groundwater, and do not present a risk to human health and the environment. It is necessary to establish groundwater monitoring programmes in order to meet the requirements of the WFD and the new Groundwater Directive. Those programmes should include quantitative measurements of chemical quality to establish the current status and enable the assessment of trends in quality, and to underpin the characterisation of groundwater in the context of legislation safeguarding the quality of drinking water. It is essential in the implementation of these programmes that there should be conﬁdence in the results of any risk assessment based on them, and that the methods employed be cost-eﬀective and ﬁt for purpose on the regional, national and river basin scales across Europe. The monitoring programmes focus on phenomena aﬀecting the overall state of the groundwater body and should be designed on the basis of the conceptual model/understanding of the groundwater system in which the general scheme of ‘‘recharge–pathway–discharge’’ is known. The conceptual model will represent the current understanding of the groundwater system based on the knowledge of its natural characteristics (e.g. the aquifer type, three-dimensional structure, dynamics and boundary conditions), as well as perceived pressures and knowledge of impacts. A consideration of the diﬀerent types of aquifers is an essential part of the conceptual model/understanding.1,2 A diverse range of hydrogeological backgrounds and aquifer types is found across Europe. This broad variation has major implications for the suitability of diﬀerent types of sampling installation and how eﬀectively they represent changes in groundwater systems. The design of monitoring programmes needs to be tailored accordingly. For all groundwater bodies, there is a need to consider the characteristics of the strata forming the aquifers with regard to ﬂow paths and ﬂow mechanisms, storage, unsaturated zone thickness, groundwater recharge and discharge, before determining the most appropriate means of monitoring. Given the large scale of the necessary monitoring activities under the new legislation, there is a need for new analytical methods and parameters that can improve the quality of the information provided and at the same time deliver eﬃciency. Screening methods can be used for a rapid estimation of the quality of a groundwater body and thus might help to facilitate the ﬁeld monitoring

Screening Methods for Groundwater Monitoring

365

required for supporting and reﬁning such conceptual models. However, standard methods are not yet generally available for those emerging sampling and analytical methods, and under these circumstances there is a necessity for ‘‘inhouse’’ validated methods, and for the publication of regular evaluations of their performances. In England and Wales, and in the Netherlands, short-term screening tests are being evaluated in order to demonstrate whether a combination of screening methods and traditional spot sampling procedures can be cost eﬀective. One advantage of such an approach could be a reduction in the frequency of monitoring and a reduction in the number of monitoring stations necessary to ensure a representative picture of water quality, and a concomitant decrease in the number of chemical analyses. This would allow a pragmatic approach to the design of a monitoring strategy, whilst increasing its cost/ eﬃciency. Indeed screening tools might be able to contribute to the surveillance and operational monitoring of the quality of a groundwater body as a whole in a number of ways. Understanding of conceptual models and of the hydrogeology and hydrochemistry of the groundwater system can be improved, enabling review and adaptation, where appropriate, of the network design. The design of a monitoring network should take into account the threedimensional nature of the groundwater system and both spatial and temporal variability. This is especially important when selecting the location and nature of monitoring sites. The network should have a spatial and temporal density that matches the natural characteristics of the groundwater body (conceptual understanding) and the pollution risks. This will focus monitoring activities in areas of signiﬁcant pressures and high vulnerability. The recommended core set of determinands comprises dissolved oxygen, pH value, electrical conductivity, nitrate, ammonium, temperature and a set of major and trace ions. Parameters such as temperature and the set of major and trace ions are very helpful in the validation of risk assessments and of the underlying conceptual models, as well as in the assessment of natural background levels. The use of multi-parameter probes that can provide continuous measurements of transient physical/chemical ﬁeld parameters (pH, temperature, conductivity, redox potential and dissolved oxygen) over a range of depths can facilitate chemical logging of the water table, and provide an instantaneous ‘‘chemical photograph’’ of the water body where there is a vertical variation in the aquifer characteristics and stratiﬁcation of groundwater quality. Such multiparameter probes can also be very eﬃcient in the identiﬁcation of saline or other intrusions resulting from alterations in ﬂow within the groundwater body, and hence in the quantiﬁcation of water volumes. This can be combined with the monitoring of water levels at all the relevant sites in order to describe the ‘‘physical status of the site’’ and to interpret (seasonal) variations or trends in the chemical composition of the groundwater.

366

Chapter 6.2

The monitoring that needs to be focused on point sources can be limited where risks are related to speciﬁc sites, such as particular ecosystems, that receive water from an underground source. Here sampling points can be focused in areas that are close to the aﬀected system. In these cases, where the location of pressures (point sources) is known, sampling points will often be used to help identify impacts from diﬀerent pressure types, to assess the extents of these and to determine contaminant fate and transport between the pressure and the receiving system. In some cases this may require the use of physicochemical probes and of multilevel samplers. An appropriate selection of bioassays may be useful in identifying contamination hotspots. – Such assays enable mapping at a local scale to identify sources of pollutants; assessment of the pressures in the case of multiple sources of pollutants; and evaluation of temporal changes in concentration in a plume of a pollutant. – Sensors and immunoassay test kits may enable on-site mapping of a water body, for example following an accidental pollution incident. These tools can help in the on-site selection of samples to be brought back to the laboratory for further conﬁrmation or more accurate analysis. Thus screening methods can be used for a rapid assessment of the chemical status of a water body in order to identify the most relevant sites for surveillance or operational monitoring. Where relevant historical data are available, the utility of routinely measured parameters may be known. However, in the absence of such information, it may not be possible to establish useful indicators without extensive research and this may be prohibitively expensive.3

6.2.2

Environmental Variability: Spatial and Temporal Groundwater Quality Variability

As already highlighted in this chapter, the design of a monitoring network should take into account the three-dimensional nature of the groundwater system and both spatial and temporal variability, especially when determining the location of monitoring sites and selecting appropriate monitoring site types. The variability in time and space displayed by the water body being monitored will determine the frequency and number of sites necessary for representative monitoring. This observed variation may be, for example, the result of environmental conditions such as tidal cycles in coastal areas and estuaries, or ﬂood events in riverine environments. Other sources of variation include seasonal and spatial trends in pesticide applications, or the impact of sewage, storm water and industrial eﬄuents.3 Where there is high variability in environmental quality of a water body, then it will be necessary to use a larger number of

Screening Methods for Groundwater Monitoring

367

sampling sites and/or more frequent sampling in order to obtain a representative picture of the overall quality. In these circumstances it may not be possible because of the high cost of employing spot sampling and classic (laboratory) analysis. Commonly accepted sampling frequencies and spatial distributions of sampling sites may provide misleading information. They may either overestimate a problem (where samples are taken at peak levels of ﬂuctuating concentrations, or in localised ‘‘hot spots’’) or underestimate it (where samples are taken when concentrations are temporarily below normally high levels, or in localised clean areas). Here, despite the well-deﬁned, relatively low uncertainties associated with the classic sampling and analytical procedures, the wrong conclusions will be drawn, and this could be costly in terms of an assessment of risk based on such data. Under these circumstances it may be better to use more rapid, less expensive tools that will enable more comprehensive monitoring of a water body, even if the uncertainties associated with them are relatively high, just so long as the uncertainties are well deﬁned. It would be safer to be roughly right than precisely wrong. Such tools would give greater conﬁdence in the use of standard monitoring programmes based on infrequent spot sampling, and also in risk assessments. Evaluation of the magnitude of environmental variability of water bodies compared with both sampling and analytical (classic in-laboratory) uncertainties is important for the interpretation of the monitoring data. Currently it is assumed that information based on sparsely distributed, infrequent spot sampling (where levels of key pollutants are compared with environmental quality standards) is representative of overall water quality. Whilst laboratories are aware that sampling can contribute to the overall uncertainty of measurement, its contribution is diﬃcult to quantify and the calculation of measurement uncertainty does not usually take into account the complete analytical chain. Indeed, in most cases, there is no information available concerning the contributions of the uncertainty of the sampling and the uncertainty of the analytical measurement to the overall uncertainty in estimates of environmental levels over space and time. However, awareness of this situation is increasing. When the most suitable methods or techniques are not implemented in the ﬁeld, there is a risk that the chemical data derived from them may be wrongly interpreted, not only with regard to statutory regulations but also in terms of our understanding of natural phenomena. The sum of external factors that may inﬂuence or interfere with a sample therefore needs to be quantiﬁed precisely in order to ensure that they do not mask real natural variations in the environment, which may be spatial or temporal. It has been investigated in the SWIFTWFD project during a ﬁeld trial on the Meuse river where little variability of the medium occurred during the experiment. In this example, experiments were carried out to assess the contributions of (i) the sampling procedure throughout the whole measurement process and (ii) the variability of the medium. During the case study, the various steps in sampling process and sample preparation (e.g. ﬁltration, ﬂask/tube, HNO3 acidiﬁcation, transport, blank eﬀect) that could aﬀect the uncertainty were investigated. These tests were conducted on natural samples, which imply matrix eﬀects, and possibly temporal

368

Chapter 6.2

ﬂuctuations in the composition of the medium. Complementary tests were carried out in order to estimate temporal and spatial variability in the concentrations of analytes of interest in the water body during the sampling period. Diﬀerent sampling sites were selected, according to the geographical constraints, and then spot sampling was performed at a range of depths at each site to determine depth proﬁles and spatial variability. The depth proﬁles were measured several times each day for three days by the same operator to estimate temporal variability. All depth proﬁles were measured with the same YSI multiparameter probe, in order to avoid a probe eﬀect. The parameters recorded were conductivity, temperature, pH, turbidity, dissolved O2 and redox potential. The complementary assessment of the (spatial or temporal) variability of the medium was based on analytical measurements methods of spot samples by classic methods in the laboratory.

6.2.3

Screening Methods Towards Groundwater Monitoring Needs

A technical report, recently published under the EU’s 6th Framework Project, Screening Methods for Water Data Information in Support of the Implementation of the Water Framework Directive (SWIFT-WFD project; www. swift-wfd.com), lists the commercially available and prototype techniques or tools that can be used in the water quality monitoring programmes as required by the WFD.4 On the basis of this report, this section aims to highlight the most promising methods/tools that could be suitable for groundwater monitoring. The parameters monitored in groundwater are classiﬁed in several groups: inorganic parameters (nitrate, nitrite, ammonium, phosphate, alkali metal, alkaline earth metal (potassium, calcium, magnesium)); heavy metals (arsenic, mercury, cadmium); volatile organic compounds (VOCs; generally halogenated hydrocarbons); and pesticides (triazine, phenoxy alkane carbone acids) and polyaromatic hydrocarbons. When groundwater is expected to contain high naturally occurring concentrations of substances or ions, or their indicators, due to a speciﬁc hydrogeological context, it may be useful to identify a cost-eﬀective and rapid way to test such an assumption by determining natural background levels of inorganics or environmental threshold values.

6.2.3.1

Common Physicochemical Methods Used for Groundwater Assessment

Some physicochemical methods are commonly used for the assessment of groundwater quality5 depending on the compounds to be monitored (organic

Screening Methods for Groundwater Monitoring

369

and inorganic compounds, volatile compounds and heavy metals). These technologies are based on laboratory methods, including, for example, speciﬁc gas chromatographic techniques developed for the analysis of VOCs extracted from groundwater or soils. Some research6 has focused on the development of new analytical instrumentation based on spectroscopic methods (e.g. laser-induced ﬂuorescence methods, and laser-ﬁbre-optical instruments) that are suitable for in situ detection of groundwater contaminants (e.g. aromatic hydrocarbon pollutants). An optical reﬂectance sensor has been proposed7 for the in situ monitoring of BTEX in groundwater. Some automatic analysis systems have been also been developed and include several sensors (e.g. ion-selective potentiometric sensors, optoﬂuidic measuring systems, conductivity cells, pH electrodes).

6.2.3.2

Emerging Screening Tools

Some emerging screening tools can be used in groundwater sampling to give valuable information that provides a basis for the design of groundwater monitoring programmes. For instance, they can help in the selection of an appropriate number of test sites (wells) and the identiﬁcation of representative sites. Moreover, the potential impact of groundwater on surface water might be eﬃciently monitored in the relevant selected sites by an appropriate choice of bioassays. These may be useful for linking biological/ecological monitoring to chemical or quantitative information, allowing the early detection of biological imbalances.

6.2.3.2.1

Electrochemical Sensors

Voltamperometric probes and selective, or ion-speciﬁc, electrodes are particularly suited to in situ measurements because they require little or no sample treatment and enable continuous measurements to be performed. Electrochemical measurement systems have been miniaturised into screen-printed electrodes (SPEs) that are incorporated in hand-held equipment for on site monitoring of many heavy metals and certain pesticides. The University of Florence8 produces SPEs which have either a graphite working electrode or a gold working electrode and counter electrode and a silver reference. These sensors can be obtained from Palm Instruments. Recently, Palm Instruments and the company EcoBioServices & Researches have developed ready-to-use strips for the detection of photosynthetic inhibitors (herbicides and heavy metals).

6.2.3.2.2

Immunoassay Test Kits as Screening Tools

Immunoassays have been developed and proposed as screening tools in order to assess groundwater quality and monitor some targeted pesticides.9 The latter is useful since the majority of surveyed groundwater sources, sampling both shallow wells and deeper aquifers, contained pesticide residues. Atrazine and simazine (triazine pesticides) are the pesticides most frequently detected in wells in both agricultural and urban areas. Immunoassay (IA) tests have been

370

Chapter 6.2

introduced in order to reduce the cost of water quality monitoring and enhance ﬁeld measurements. Good selectivity, sensitivity, precision and portability make immunoassays a cost-eﬀective method for groundwater monitoring. Due to their portability, immunoassay test kits show promise as screening methods in particular for mapping of pollutants and the characterisation of contaminated sites. These immunoassays are also very useful in the case of known contaminant monitoring.10–12 Enzyme-linked immunosorbent assays (ELISAs) based on the use of labelled enzyme conjugates are widely available in a range of formats such as coated tubes, magnetic particles or 96-well plates, enabling the simultaneous processing of a large number of samples. Enzyme conjugates are competitively displaced from binding sites by the free analytes. The tubes, magnetic particles or well-plates are rinsed and a chromogen is added to react with enzyme conjugates producing a coloured chemical. After a period of time the reaction is stopped, enabling spectrophotometric quantiﬁcation of immobilised enzyme conjugates, and thus, from the diﬀerence, the concentration of analytes initially present in the sample.13,14

6.2.3.2.3

Biosensors

Biosensors can be used for long-term monitoring of groundwater. The main advantages are: quick analytical turnaround time, cost-eﬀectiveness, portability, in situ measurement, high sensitivity and speciﬁcity. In general, biosensor devices combine a biological recognition element in contact with a transduction element, and have the potential to assess speciﬁc biological eﬀects (e.g. toxicity, cytotoxicity, genotoxicity, endocrine-disrupting eﬀects) due to contaminants. A very useful overview has been published recently describing an environmental application15 involving monitoring both organic compounds (e.g. pesticides, PAHs, PCBs and phenols) and inorganic compounds (e.g. heavy metals, nitrate, and phosphate).

6.2.3.2.4

Whole-organism Bioassays

A whole-organism bioassay relies on the measurement of the biological response (acute or chronic toxicity) of a test organism to a mixture of contaminants present in water. Organisms commonly used include microorganisms, algae, amphipods, daphnids, oysters and chironomid larvae. The test parameters usually measured include mortality, bioluminescence, metabolic status and growth rate inhibition. Inhibition of bioluminescence in the bacterium Vibrio ﬁscheri is the basis of the most common test. This type of assay is relatively simple to implement, and many commercial devices are available. A large toxicity database including many chemicals has been established for this assay, for which standard protocols exist (ISO 11348). These toxicity assays can be used for the rapid screening or mapping of contaminant levels (general toxicity, genotoxicity) in a water system.

Screening Methods for Groundwater Monitoring

6.2.3.2.5

371

Passive Samplers

The tools described earlier in this section provide alternatives to classic laboratory-based analytical methods, and in some cases still depend on the collection of grab (spot) samples. In contrast, passive samplers replace spot sampling by accumulating pollutants over a prolonged deployment period to provide measurements of the average concentrations to which they have been exposed. The samplers are returned to the laboratory for extraction and analysis, and the latter is usually carried out using classical methods. However, the analysis could be performed using a wide range of alternative methods including those described earlier in this section. The extracts from the sampler receiving phases can be tested in bioassays. A range of integrative passive sampling devices has been developed and used in recent years. A comprehensive review of the currently available passive sampling devices has been published.16 Among the most widely used samplers are the semi-permeable membrane devices (SPMDs) for hydrophobic organic pollutants17 and the diﬀusive gradients in thin ﬁlms (DGTs) for metals and inorganic ions.18 Several novel passive sampling devices suitable for monitoring a range of polar organic chemicals, including pesticides pharmaceutical/veterinary drugs and other emerging pollutants of concern have recently been developed.19,20 Passive samplers can be applied to investigate long-term temporal trends in water contaminants and to evaluate the location of point and diﬀusive contaminant sources.21,22 These passive samplers are particularly relevant when the objective is to screen for the presence or absence of targeted pollutants (metals, pesticides). Due to their integrative function it is possible to quantify pollutants that are below the level of detection in bottle samples of water. Furthermore, they can provide a more representative picture of the level of freely dissolved contamination (but not material bound to particulate matter) since the mass accumulated will be indicated by the area under the concentration time proﬁle to which they have been exposed. This is in contrast with spot sampling, which only provides an instantaneous measurement of concentrations of pollutants at the moment that the sample was taken. Passive samplers can thus reduce the uncertainties associated with the use of infrequent spot samples where contaminant concentrations ﬂuctuate. Most passive samplers are suitable for deployment in surface waters, but only a few designs (e.g. the ceramic dosimeter that has been used to monitor a range of organic pollutants including volatile organic compounds23) have been developed for long-term deployment in bore holes to monitor groundwater.

6.2.3.3

Potential Uses of Emerging Tools

Some examples/case studies performed in the framework of SWIFT-WFD European project SSPI-CT-2003-502492 illustrate the potential of screening methods for use in water monitoring programmes.

372

6.2.3.3.1

Chapter 6.2

River Daugava (Latvia)

The main objective was to assess whether a number of sampling sites gave a representative measure of the impact of urban discharge on water quality of the river. For that purpose, seven sampling sites were selected upstream and downstream of the city of Rezekne,24 and biogens (NO3, NH41, PO43 ) were measured either in situ or on site, using a multi-parameter probe (YSI), colorimetric kits (Merck) and a portable UV instrument (Pastel UV). The results obtained using the emerging tools tested were consistent with the data obtained using routine standard laboratory methods, and demonstrated the impact on the river of discharges from the city and its waste water treatment plant. The use of Pastel UV in chemical monitoring proved valuable for the evaluation of a point source pollution plume, the selection of representative sites and for investigative monitoring. The emerging tools tested are highly suitable for the rapid assessment of the concentration of biogens in relatively contaminated waters.

6.2.3.3.2

Orlice River (Czech Republic)

The main tools deployed were portable laboratory-based instruments and test kits that could be used on site. The measurements focused on ammonium, nitrate, total organic carbon, BOD and alachlor, copper, aluminium and zinc. The on-site systems (Pastel UV, chemical test kit) are useful tools for assessing water treatment plant eﬃciency, in particular to monitor global parameters (TSS, TOC, COD, BOD) and speciﬁc species (nitrate, phosphate). Nitrate concentrations were measured upstream (input channel) and downstream (output channel) of a treatment plant. Speciﬁc discharge problems were detected as shown by the increase in nitrate concentrations caused by treatment (Figure 6.2.1). Screening methods (chemical test kit and Pastel UV, for example) are very powerful tools for assessing the variability of downstream water and troubleshooting the WWTP.

6.2.3.3.3

Hart-catchments (Alsace, France)

In the framework of SWIFT-WFD, the quality of the groundwater in drinking water catchments was investigated in the Alsace region of France. There is a problem with pesticide contamination of groundwater in the area studied. Two compounds (atrazine and alachlor that are used for weed control in ﬁeld corn, and are sometimes mixed in some commercial herbicide formulations) are notable pollutants. Neither the sources of these pesticides nor their relative importance as inputs in groundwater has yet been fully identiﬁed. The hydrology of the catchments is such that surface waters inﬁltrate the groundwater through several gravel pits, and this may be the entry route through which pesticides from agricultural activities enter. The aim of this study was to investigate the spatial variability of nitrate and pesticides (atrazine, alachlore) in wells in order to assess groundwater quality. In addition, the YSI probe was

373

Screening Methods for Groundwater Monitoring 1000 Upstream Downstream

900

Concentration mg/L

800 700 600 500 400 300 200 100 0 TSS

Figure 6.2.1

TOC

NO3

COD

BOD

Wastewater treatment monitoring by portable Pastel UV.

Wells 1200 pH unit

7.30

1000 800

7.20

600 400

7.10

200 7.00

0 PZ1

Figure 6.2.2

Conductivity µS/cm

1400

7.40

PZ2

PZ5 Samples

PZ6

PZ30 pH Conductivity 25˚C

YSI probe in situ measurement of pH and conductivity values (Hart-catchments, Alsace, France).

used for pH and conductivity measurements in the ﬁve wells studied (Figure 6.2.2). Four of the wells (PZ1, PZ2, PZ3, PZ30) showed similar conductivity values (about 800 mS cm 1) and the conductivity for PZ6 was in the region of 1200 mS cm 1. Both pH and conductivity values were compliant with the quality standards deﬁned by drinking water standard directive 98/83/CE. Global parameters (TOC, COD, BOD, suspended matter) and speciﬁc parameters (nitrates, surfactants as DBS) were estimated using portable instruments (portable spectrophotometer Pastel UV), and pesticides (atrazine and alachlore) by immunoassays. These parameters were selected on the basis of the history of these wells, and their values were mapped to identify critical

374

Chapter 6.2 Wells 80.0

250.0

Concentration mg/L

70.0

NO3-

60.0

DBS

50.0

COD BOD

40.0

200.0 150.0

SPM 100.0

30.0 20.0

50.0

Concentration SPM (mg/L)

TOC

10.0 0.0

0.0 PZ1

PZ2

PZ5

PZ6

PZ30

Samples

Figure 6.2.3

Variability of global and speciﬁc parameters assessment by Pastel UV values (Hart-catchments, Alsace, France).

points in terms of pollution by nitrate and atrazine in order to assess the potential for using some of the wells for feeding into drinking water sources. In two wells (PZ5 and PZ6) the concentrations of nitrates (Figure 6.2.3) exceeded the quality standard (50 mg/L) deﬁned by drinking water standard directive 98/83/CE and nitrate directive 91/676/EC. The global parameters measured (Figure 6.2.3) provide a useful proﬁle of groundwater quality and variability between the wells. In two of the wells (PZ1 and PZ2), atrazine concentrations exceeded the quality standard (0.100 mg l 1) deﬁned in the drinking water directive, whereas the level of alachlore contamination was below quality standard in all ﬁve wells (Figure 6.2.4). This example demonstrates that the application of screening tools (YSI probes, portable instruments and immunoassays) is very useful for mapping the spatial variability of groundwater quality, particularly for investigative monitoring.

6.2.4

Screening Methods and Priority Substances

As part of the SWIFT-WFD project, tools and techniques either available commercially or as prototypes were identiﬁed in a literature search, and on the basis of this a toolbox was constructed to provide a set of alternative methods that could potentially be used to provide improved information on a water body in order to meet the challenges of chemical and biological monitoring.3,4 The aim was to evaluate a subset of the available tools which could be used to meet monitoring requirements that would be diﬃcult or expensive to fulﬁl using the established and accepted spot sampling methods linked with classical laboratory analysis. Well-deﬁned roles or functions have been identiﬁed for

375

Screening Methods for Groundwater Monitoring Wells 0.400

Concentration (µg)

0.350

Atrazine Alachlor

0.300 0.250 0.200 0.150 0.100 0.050 0.000 PZ1

Figure 6.2.4

PZ2

PZ5 Samples

PZ6

PZ30

Variability of atrazine and alachlor concentrations values (Hartcatchments, Alsace, France.)

the various types of tools to facilitate their introduction into, and optimise their impact on, monitoring programmes. Finally, selected results from a series of ﬁeld trials conducted within the SWIFT-WFD project are presented as case studies to demonstrate the tasks embedded in monitoring programmes. Examples include (i) time-weighted measurement of labile concentrations of heavy metals by Chemcatcher for the long-term monitoring of natural changes or changes due to anthropogenic activity, (ii) the tracing of point or diﬀuse sources of contaminants using the Ecoscope sampler and (iii) rapid and onsite screening for selected metals using the Palmsens and screen-printed electrodes.

6.2.5

New Trends and Perspectives

Passive sampling techniques that measure time-weighted average concentrations may be used for monitoring long-term trends, the screening of a large range of contaminants at very low concentrations, the measurement of metal speciation or the identiﬁcation of sources of pollution.3 In addition it is possible to use passive sampling techniques to conﬁrm the risk assessment elaborated for river basins across Europe.3 A development that will extend the utility of passive samplers in investigative monitoring is the linking of the accumulation over a prolonged period with toxicological assays. The Toximeter is a passive sampler based on the ceramic dosimeter developed for monitoring organic compounds in groundwater. The beads comprising the receiving phase can be used directly in cell-based bioassay systems without solvent extraction.25 This shows great promise for a wide range of applications within investigative monitoring. In common with many of the emerging techniques, work is required to demonstrate their validity and to develop acceptable quality assurance

376

Chapter 6.2

procedures. Without these prerequisites they will not be used to underpin legislation. Work on the development of national and international standards for the use of passive samplers has started, with the publication of a national norm (BSI PAS61), but further work is necessary to develop European or international norms. Field demonstrations of the performance of these technologies alongside routine spot sampling carried out by regulatory agencies, together with an evaluation of relative costs, would further increase conﬁdence in the emerging tools. Groundwater monitoring programmes are required to provide a coherent and comprehensive overview of water status within each river basin, to detect the presence of long-term anthropogenically induced trends in pollutant concentrations and ensure compliance with protected area objectives. The reliability of currently used spot sampling methods for measuring water quality is questionable where concentrations of pollutants are low, and hence analytical uncertainties are high, and where there are short-term temporal and spatial variations in pollutant levels. This makes it diﬃcult to detect trends in order to monitor the impact of remedial actions on pollutant concentrations. Under these circumstances the deployment of time-integrated sampling systems might be a helpful alternative since masses of pollutants accumulated over several weeks are large enough to decrease the uncertainties associated with the chemical analysis. Furthermore, they provide measures of the average concentrations to which they have been exposed rather than the instantaneous estimates provided by spot sampling. Thus, even though there are additional sources of uncertainty (calibration procedures, and the inﬂuence of environmental variables such as temperature and turbulence) associated with some of the emerging tools, they may provide more representative and accurate measures of water quality. There are great potential beneﬁts to be gained from the availability of a set of wellcharacterised tools from which those responsible for monitoring programmes can select the most appropriate for a particular application. This would increase conﬁdence in the risk assessments used in the management of water quality, and in the measurement of trends in water quality.

References 1. F. Wendland, A. Blum, M. Coetsiers, J. Griﬀoen, J. Grima, K. Hinsby, R. Kunkel, A. Marandi, S. Vermooten, K. Walraevens, Aquifer typologies: a practical framework for an overview about major groundwater composition on a European scale, European Conference on Groundwater, Vienna, 22–23 June 2006. 2. F. Wendland, A. Blum and R. Kunkel, Approach to assess natural background levels as well threshold values for the groundwater, European Conference on Groundwater, Vienna, 22–23 June 2006. 3. I. J. Allan, G. A. Mills, B. Vrana, J. Knutsson, A. Holmberg, N. Guigues, S. Laschi, A. -M. Fouillac and R. Greenwood, Trends Anal. Chem., 2006, 25, 7.

Screening Methods for Groundwater Monitoring

377

4. I. J. Allan, B. Vrana, R. Greenwood, G. A. Mills, B. Roig and C. Gonzalez, Talanta, 2006, 69(2), 302–322. 5. US EPA, Compendium of ERT Field Analytical Procedures, OSWER Directive 9360.4-04, and Compendium of ERT Ground-Water Sampling Procedures, OSWER Directive 9360.4-06. 6. S. J. Hart, Y. M. Chen, J. E. Kenny, B. K. Lien and T. W. Best, Field Anal. Chem. Technol., 1997, 1, 343–355. 7. S. L. Huntley, J. L. Ritchie, S. J. Setford and S. Saini, Sci. World J., 2002, 2, 1101–1107. 8. S. Laschi, I. Palchetti, G. Marrazza and M. J. Mascini, Electroanal. Chem., 2006, 593, 211–218. 9. L. Candella, J. Caballero, T. Melo and E. Torres, J. Environ. Technol., 1998, 6, 1–9. 10. B. Lesnik, Immunoassays techniques in environmental analysis, in Encyclopedia of Analytical Chemistry, ed. R. A Meyers, John Wiley, Chichester, 2000, pp. 2653–2672. 11. M. -C. Hennion and D. Barcelo, Anal. Chim. Acta, 1998, 362, 3. 12. M. A. Bacigalupo, G. Meroni, M. Mirasoli, D. Parisi and R. Longhi, J. Agric. Food Chem., 2005, 53, 216. 13. U. Pfeifer-Fukumura, I. Hartmann, H. Holthues and W. Baumann, Talanta, 1999, 48, 803. 14. B. Ballesteros, D. Barcelo, A. Dankwardt, P. Schneider and M. -P. Marco, Anal. Chim. Acta, 2003, 475, 105. 15. S. Rodriguez-Mozaz, M. J. Lopez de Alda and D. Barcelo, Anal. Bioanal. Chem., in press. 16. J. Namiesnik, B. Zabiega1a, A. Kot-Wasik, M. Partyka and A. Wasik, Anal. Bioanal. Chem., 2005, 381, 279–301. 17. J. N. Huckins, G. K. Manuweera, J. D. Petty, D. Mackay and J. A. Lebo, Environ. Sci. Technol., 1993, 27, 2489. 18. H. Zhang, W. Davison, B. Knight and S. McGrath, Environ. Sci. Technol., 1998, 32, 704. 19. J. K. Kingston, R. Greenwood, G. A. Mills, G. M. Morrison and B. L. Persson, J. Environ. Monit., 2000, 2, 487. 20. D. A. Alvarez, J. D. Petty, J. N. Huckins, T. L. Jones-Lepp, D. T. Getting, J. P. Goddard and S. E. Manahan, Environ. Toxicol. Chem., 2004, 23, 1640. 21. B. Vrana, A. Paschke, P. Popp and G. Schu¨u¨rmann, Environ. Sci. Pollut. Res., 2001, 8, 27. 22. L. B. Blom, G. M. Morrison, J. Kingston, G. A. Mills, R. Greenwood, T. J. R. Petersson and S. Rauch, J. Environ. Monit., 2002, 4, 258. 23. S. K. Bopp, H. Weiß and K. Schirmer, J. Chromatogr. A., 2005, 1072, 137. 24. N. Guigues, J. -C. Foucher, R. Scha¨ﬀer, M. Motelica, J. Bruveris, R. Rimsa, P. Petrova, P. Fordor, A. -M. Fouillac, Testing emerging tools for biogens: the Latvian case study, SWIFT-WFD Workshop, Berlin, 2005. 25. S. K. Bopp, N. C. Bols and K. Schirmer, Environ. Toxicol. Chem., 2006, 25, 1390.

CHAPTER 6.3

Quality Assurance for Groundwater Monitoring PHILIPPE QUEVAUVILLERa AND STE´PHANE ROYb a

European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium; b Bureau de Recherches Ge´ologiques et Minie`res (BRGM), 3 avenue Claude Guillemin, FR-45060 Orle´ans ce´dex 2, France

6.3.1

Need for Quality Assurance for Groundwater Analysis

Groundwater analyses are currently performed in routine and research laboratories for the control of quality of water used for human consumption and of environmental contamination levels according to EC legislation (see Chapter 3.1). A good accuracy is a prerequisite for analysis for achieving a good comparability of data and hence allowing sound decisions to be taken by authorities. The lack of analytical quality control renders useless many of the results obtained in monitoring campaigns and it is obvious that many studies performed in the past have not been performed with a suﬃcient quality control mechanism and that many of their conclusions were biased. Poor performance by analytical laboratories and consequent incomparability of data create economic losses due to extra-analyses, court actions, etc. The costs in terms of quality of life can hardly be estimated but is certainly very high. Various measures are necessary to ensure not only a good reproducibility but also a good accuracy. This chapter summarises various aspects of quality assurance for water analysis on the basis of previous publications.1 Most attention will be paid to interlaboratory studies and use of certiﬁed reference materials (CRMs) according to experiences gained in EC-funded projects.2–4 A particular focus will be given to groundwater, which is currently monitored by European Union (EU) laboratories to control the level of contamination by trace or major elements, in particular in support of EC directives (80/68/EEC, 80/778/EEC and 91/676/EEC), and which will have to be monitored on a systematic basis in the context of the Water Framework Directive (2000/60/EC) 378

Quality Assurance for Groundwater Monitoring

379

and its related daughter Groundwater Directive (adopted in early 2007) from 2007 onward. Furthermore, speciﬁc emphasis is placed on the sampling step, which needs to be taken into account to ensure a good traceability of the analytical chain.

6.3.2

Within-laboratory Quality Measures

The accuracy and uncertainty of a measurement are two basic parameters that should be considered when discussing results of any analysis. Accuracy is of primary importance but if the uncertainty in a result is too high, that result cannot be used to reach any conclusions about the outcome of an experiment or to judge the quality of the water analysed. In water analysis, all basic principles of calibration, elimination of sources of contamination and losses, correction for interferences, etc., should be followed,5,6 although calibration seems to be so obvious to many analytical chemists, experiences gained in interlaboratory programmes have shown that many systematic errors could be related to calibration errors, showing that insuﬃcient attention had been paid to this part of the analytical process.6 A typical example concerned estuarine water analysis for which the need to use standard addition procedures was highlighted due to its complex nature, i.e. diﬀerences of up to 30% could be observed between results obtained by external calibration and standard additions.7 The increasing awareness for quality assurance has led to the establishment of series of guidelines, the most comprehensive one being the ISO 17025 standard (now adopted as European Norm) which describes the way in which the laboratory should work, its organisation and the way to produce valid results, i.e. involving managerial aspects, quality assessment (inter- and intralaboratory programmes, use of reference materials), statistical quality control (control charts), maintenance of apparatus, chemicals and reagents, sampling and storage, laboratory analysis, documentation and reporting, and archiving. In the framework of this chapter, most attention is paid to interlaboratory studies, materials for use in control charts and certiﬁcation of reference materials.

6.3.3

Statistical Control

When a laboratory works at a constant level of high quality, ﬂuctuations in the results become random and can be predicted statistically.8 This implies that the limits of determination and detection should be constant and well known. Rules for rounding-oﬀ ﬁnal results should be based on the performance of the method. Furthermore, in the absence of such systematic ﬂuctuations, normal statistics should be applied to study the results wherever necessary (e.g. regression analysis, t- and F-tests, analysis of variance).9 Control checks should be carried out as soon as the method is in control in the laboratory. These can be done by setting up control charts which provide a

380

Chapter 6.3

graphical way of interpreting the method’s output in time in order to evaluate the reproducibility of the results over a period of time. To do so, one reference material of good homogeneity and stability should be analysed with, for example, 10–20 unknown samples. In order to be able to detect non-random ﬂuctuations in the real analysis, the RM used should pose the same or similar problems to the analytical chemist as the unknown samples do (analytical similarity) and their composition should be homogeneous and stable in time. A detailed description of the diﬀerent types of control charts (e.g. Shewhart or cusum control charts) is given elsewhere.9

6.3.4

Comparison of Analytical Methods

The use of control charts enables one to detect whether or not a method is still under control. However, it does not allow detection of a systematic error which could be present from the moment of the introduction of the method in the laboratory. To overcome this risk of undetected bias, results should be veriﬁed by using other methods. All analytical methods have their own particular source of error, e.g. diﬀerent interferences occurring in spectrometric or voltametric techniques. Independent methods displaying diﬀerent sources of error may therefore be used to verify the validity of results of chemical analysis. If results of two independent methods are in good agreement, it may be concluded that the risk of error of systematic nature is unlikely; this conclusion is strongest when the two methods diﬀer widely. If the methods have similarities, such as a digestion step in the case of voltametry and spectrometry, a comparison of the results will only allow conclusions to be drawn on the accuracy of the ﬁnal determination and not the whole analytical procedures. In some cases, the method of comparison is a technique of which the sources of error are well known and controlled and which is relatively insensitive with respect to human failures; such techniques are often referred to as ‘‘reference method’’, e.g. isotope dilution mass spectrometry. When possible, the application of such a reference method is used in a good quality control system. However, it must be stressed that good independent or reference methods do not always exist (e.g. there are few reference methods for organic analysis). Furthermore, these techniques require a good experience of the technician who might not have a suﬃcient skill to produce reliable results even with a so-called reference method. Further details are given in the literature.5,6

6.3.5

Interlaboratory Studies

6.3.5.1

Introduction

Interlaboratory studies are useful tools to detect systematic errors linked to a speciﬁc method or incorrect application of a method in a laboratory. In general, besides the sampling error, three sources of error can be detected in all analyses: (i) sample pre-treatment (e.g. digestion, extraction, separation,

Quality Assurance for Groundwater Monitoring

381

pre-concentration), (ii) ﬁnal measurement (e.g. calibration errors, spectral interferences, peak overlap) and (iii) insuﬃcient experience (e.g. lack of training, care applied to the work, awareness of pitfalls, management, clean bench facilities). A relatively large laboratory where more techniques are applied by diﬀerent experienced technicians may be in a position to eliminate most errors related to (i) and (ii) above in a continuously and carefully applied quality control scheme. Smaller laboratories or laboratories having one or two techniques only are not always able to assess the quality of their measurements with respect to (i) and (ii) in the absence of other tools (e.g. use of CRMs as described below). For both types of laboratories, however, interlaboratory studies to evaluate step (iii) are necessary. Interlaboratory studies can be conducted with several aims: (i) to detect the pitfalls of analytical techniques and ascertain their performance; (ii) to measure the quality of a laboratory or part of a laboratory (e.g. audits for accredited laboratories); (iii) to improve the quality of a laboratory in collaborative work in a mutual learning process; and (iv) to certify reference materials. When laboratories are working under control (of a recognised quality control scheme), interlaboratory studies of types (ii) and (iv) are conducted only. However, studies of types (i) and (iii) also play an important role. The organisation of interlaboratory studies, the types of samples, laboratories invited, management of the trials, etc., have been described in detail elsewhere.6,10 Interlaboratory studies are a useful and important means to assess the quality of the work done in the laboratory, to motivate laboratory workers and to demonstrate the quality of a laboratory’s result to a ‘‘customer.’’ The following paragraphs describe experiences obtained in a programme concerning groundwater analysis.2–4 In this context, participants of diﬀerent EU member states discussed thoroughly the reasons for discrepancies in collaborative trials and were able to improve their methods in the light of these discussions. Many exercises followed a step-by-step approach, i.e. series of studies with samples of increasing diﬃculty (e.g. synthetic samples, natural samples spiked with the analyte(s) of concern, natural samples or matrixmatched synthetic samples).

6.3.5.2

Scope of the Groundwater Interlaboratory Programme

One of the most powerful tools to detect and remove sources of error due to a particular technique or a lack of quality control within a laboratory is to participate in interlaboratory studies.10 In general, besides the sampling error, a series of sources of error can be detected in trace element determinations and may be related, for example, to sample pretreatment (e.g. storage, ﬁltration, digestion, preconcentration, dilution), ﬁnal measurement (e.g. calibration

382

Chapter 6.3

errors, spectral interferences, background corrections), contents too close to the quantiﬁcation limit and/or laboratory organisation (e.g. training and educational level of workers, care applied to the work, awareness of pitfalls, management, clean bench facilities). When diﬀerent laboratories participate in an interlaboratory study, diﬀerent sample pretreatment methods and diﬀerent techniques of ﬁnal determination may be compared and discussed. An interlaboratory study can be organised (i) to detect the pitfalls of a commonly applied method and to ascertain its performance in practice, (ii) to measure the quality of a laboratory or a part of a laboratory (e.g. audits for accreditation of laboratories), (iii) to improve the quality of a laboratory in collaborative work in a mutual learning process and (iv) to certify the contents of a reference material.6 The studies described in the following paragraphs are of type (iii); they involved 15–20 laboratories from EU member states.2,3

6.3.5.2.1

Major Elements in Groundwater

Artiﬁcial reference materials representative of groundwater matrices were considered to be the best compromise to achieve the stability of major elements.2 The parameters selected were based on a literature search,11 which enabled the deﬁnition of a mean composition for two types of groundwater reference materials with, respectively, low and high carbonate contents (Table 6.3.1). The optimal conditions for the preparation of groundwater candidate CRMs were tested in a feasibility study.2 Two batches of solutions, matching respectively typical carbonate and sandstone media, were prepared from ultrapure water by adding pro-analysis grade chemicals and their homogeneity was veriﬁed to evaluate possible eﬀects of the preparation procedure on the sample Table 6.3.1

List of parameters selected for the preparation of synthetic groundwater samples and mean concentrations in EC groundwaters (adapted from Ref. 2).

Element/ compound

Directive 80/778/CEE (mg l1)

Typical carbonate (mg l1)

Typical sandstone (mg l1)

Na K Ca Mg Cl SO4 NO3 H3O1 HCO3 Fe Mn NH41 PO4

20–150 10–12 100 30–50 25 25–250 25–50 6.5–9.5

2–20 o1 40–90 10–50 5–15 5–50 1–20

3–30 0.2–5 5–40 0–30 5–20 10–30 0.5–10

150–300 o0.4 o0.1

2–25 0.1–5 o0.1

0.05–0.2 0.02–0.05 0.05–0.5 0.27–3.35

Quality Assurance for Groundwater Monitoring

383

composition. The stability was also checked at +4 1C and +20 1C over a period of three months. The results of this study are described in detail elsewhere.2 The addition of lauryl sulfate to match the presence of humic acids was carried out, as was successfully performed for the certiﬁcation of nitrate in artiﬁcial freshwater.12 Autoclave sterilisation was applied without signiﬁcantly changing the nutrient pattern. The short-term stability study demonstrated that the samples remained stable under the conditions tested2 which encouraged the preparation of a large batch of candidate reference materials for the interlaboratory programme.

6.3.5.2.2

Trace Elements in Groundwater

Natural groundwater samples were selected on the basis of monitoring data obtained by the Danish Geological Survey over the period 1993–1998. Samples for trace elements were ﬁltered on-line during collection to avoid oxidation of the groundwater and were immediately acidiﬁed with nitric acid at pH ¼ 2; they were placed in polyethylene bottles previously cleaned with hot HCl and rinsed with Milli-Q water. Samples for bromide were ﬁltered at 0.45 mm (Sartorius ﬁlter) and diluted with Milli-Q water; they were placed in preconditioned brown 25 ml ampoules and packed in boxes each containing 5 ampoules. The homogeneity and stability were veriﬁed and found to be suﬃcient for the purpose of an interlaboratory study subsequently carried out.3 Most of the errors detected in this trial were due to calibration errors, in particular for Al, As and Pb. Recommendations and observations were drawn from this interlaboratory study, e.g. precautions to be taken to avoid incomplete extraction due to complexing agents, highlights of interference problems in inductively coupled plasma mass spectrometry (ICPMS) and the need to use pre-concentration. Spectral interferences were already observed for copper in a certiﬁcation campaign of trace elements in estuarine water.13 The state of the art was, however, considered to be suitable to contemplate certiﬁcation for the trace elements of concern, as well as for bromide; for iodide, however, it was considered that further improvements would be needed before proposing this element for certiﬁcation in groundwater.3

6.3.6

Certiﬁed Reference Materials

6.3.6.1

General Principles

The use of CRMs is a recognised means to verify the accuracy of an analytical procedure in a laboratory. The laboratory which measures such a reference material by its own procedure and ﬁnds a value in disagreement with the certiﬁed value is warned that its measurement includes an error of which the source must be identiﬁed. CRMs can also be used in some cases to calibrate equipment which requires a calibrant similar to the matrix (e.g. X-ray ﬂuorescence (XRF) spectrometry) and to demonstrate equivalence between methods. Other materials, known as laboratory reference materials (LRMs), are used to

384

Chapter 6.3

monitor the performance of the method (control charts); CRMs should not be used for this purpose but can be used as a basis of comparison to prepare LRMs. Diﬀerent categories of CRMs can therefore be mentioned:

pure substances; matrix materials; CRMs for calibration of relative methods (e.g. XRF); and CRMs certiﬁed for operationally deﬁned parameters.

Several requirements have to be followed in the preparation of CRMs which often correspond to compromises, e.g. stabilisation procedures slightly changing the physical or chemical state of a sample. The homogeneity and stability of the CRMs are indeed the strongest requirements that are to be fulﬁlled along with the representativeness of the materials (i.e. similarity with real samples). These parameters are of course thoroughly studied for each material starting with the stabilisation procedures (e.g. gamma-irradiation or simple acidiﬁcation), the between-bottle homogeneity and the long-term stability. These requirements are described elsewhere.6,10 Experience gathered within the EU BCR programme has shown that the most reliable way to obtain a certiﬁed value in a reference material is to compare results obtained by diﬀerent analytical techniques of proven performance which are applied by diﬀerent independent laboratories. If possible, the use of a ‘‘reference method’’ is recommended. The laboratories participating in certiﬁcation campaigns are requested to follow high standards with regard to calibration procedures, care in avoiding contamination or losses (e.g. due to incomplete digestion), application of good quality control principles, etc. As mentioned in Section 5.1, the certiﬁcation of BCR reference materials has usually been carried out via the organisation of interlaboratory studies. This has been the case for all the water reference materials that have been produced within the years 1990–2000,6,10,14 including groundwater reference materials certiﬁed for their major and trace element contents, of which the production is broadly described below (further details are given in Refs. 2–5).

6.3.6.2 6.3.6.2.1

CRMs for Major Elements in Groundwater Preparation of the Materials

The preparation and homogenisation of each candidate reference material was carried out in a high-density polyethylene container (150 l) designed to avoid microbiological and dust contamination. A magnetic drive pump (with no metal parts in contact with the solution) was used for homogenisation. All rings and piping devices were made of polypropylene.15 The container and connections were rinsed with ultrapure water and cleaned with nitric acid solution and hydrogen peroxide (3% v/v) for disinfection, and further rinsed with ultrapure water. Bacterial examination of the rinsing water did not show any microbiological contamination.15 Groundwater samples were stored in

385

Quality Assurance for Groundwater Monitoring

Table 6.3.2

Substances added for the preparation of candidate CRMs and concentrations expected upon spiking.15

Element

CRM 616 (mg kg1)

CRM 617 (mg kg1)

Ca Mg Na K Fe Mn Cl SO4 NO3 NH4 PO4

39.9 23.8 61.6 0.50 0.050 0.018 50.2 57.9 50.7 0.50 3.35 pH ¼ 6.0

14.3 7.32 14.3 9.67 0.206 0.050 26.4 27.0 25.8 0.05 0.275 PH ¼ 6.3

previously cleaned and sterilised 100 ml borosilicate glass ampoules which were immediately heat-sealed after ﬁlling. Additional details on the cleaning and preparation procedures are given elsewhere.15,16 The materials were stored at ambient temperature (ca. 20 1C) in the dark after autoclave sterilisation. Two batches of artiﬁcial groundwater samples, corresponding to typical carbonate (BCR-616) and sandstone (BCR-617) media, were prepared from utlrapure water to which freshly prepared solutions of ammonium chloride (both CRMs), calcium chloride (both CRMs), calcium hydroxide (BCR-616), calcium nitrate (BCR-616), magnesium chloride (both CRMs), magnesium nitrate (BCR-616), magnesium sulfate (both CRMs), maganese(II) sulfate (both CRMs), potassium nitrate (both CRMs), iron citrate (both CRMs), sodium hydrogen phosphate (both CRMs), sodium carbonate (BCR-616), sodium hydrogen carbonate (BCR-617), sodium sulfate (BCR-617) and lauryl sulfate were added (Table 6.3.2). All reagents were of pro-analysis grade quality. The salt solutions were previously ﬁltered through a 0.2 mm sterilised acetate cellulose membrane to reduce microbiological contamination risks. Additional details on this preparation are given elsewhere.15,16 The homogeneity and stability were veriﬁed by potentiometry (pH), ﬂame atomic absorption spectrometry (Ca, K, Mg and Na), colorimetry (NH4, NO3, PO4 and SO4) and electrothermal atomic absorption spectrometry (ETAAS; Fe and Mn). The stability tests were performed over 52 weeks’ storage at +4 1C and +20 1C. A short-term stability study was also carried out at +40 1C to test worst-case transport conditions. No instability could be detected for any of the conditions tested.15,16

6.3.6.2.2

Certiﬁcation

Various analytical techniques were used in the certiﬁcation, involving diﬀerent types of pre-treatment. Calibration was generally based on the use of

386

Chapter 6.3

calibration graph. The techniques used in the certiﬁcation of the CRMs BCR616 and 617 are summarised in Table 6.3.3. Detected sources of errors were mainly due to calibration errors, high blanks explaining high results and uncontrolled interferences (e.g. in ICPMS).15,16 Samples which were left opened and left for three days displayed precipitation Table 6.3.3

Summary of techniques applied in element determination.

Element

Technique

Ca

FAAS: deuterium background correction; air/ acetylene; addition of HNO3 and La(III); calibrant: calcium nitrate ZETAAS: Pd matrix modiﬁer; addition of HNO3 and La(III); calibrant: iron nitrate FAAS: deuterium background correction; air/ acetylene; addition of HNO3 and La(III); calibrant: magnesium nitrate ZETAAS: Pd matrix modiﬁer; addition of HNO3 and La(III); calibrant: manganese nitrate FAAS: air/acetylene; addition of HNO3 and La(III); calibrant: potassium nitrate FAAS: air/acetylene; addition of HNO3 and La(III); calibrant: sodium nitrate Colorimetric method with segmented continuous ﬂow system (660 nm); addition of potassium sodium tartrate, sodium citrate, sodium salicylate/sodium hydroxide, sodium nitroprusside and sodium dichloroisocyanurate; calibrant: ammonium chloride Colorimetric method with segmented continuous ﬂow system (490 nm); addition of Hg(II) thiocyanate and Fe(III) nitrate; calibrant: sodium chloride Colorimetric method with segmented continuous ﬂow system (540 nm); reduction on Cd/Cu-column; addition of phosphoric acid, ammonium chloride, sulfanilamide and N-(1-naphthyl)ethylenediamine; calibrant: sodium nitrate Colorimetric method with segmented continuous ﬂow system (880 nm); addition of ammonium molybdate and ascorbic acid; calibrant: potassium dihydrogen orthophosphate Colorimetric method with segmented continuous ﬂow system (460 nm); cation exchange; addition of Ba(II) methylthymol blue solution; calibrant: sodium sulfate

Fe Mg Mn K Na NH41

Cl

NO3

PO4

SO4

LQ (mg l1)

RSD (%) (no. of replicates)

0.6

1.4. (12)

0.02

5.6 (16) 1.0 (12)

1.5 0.01

6.0 (10)

0.9

4.0 (12) 2.1 (12) 2.8 (12)

4.7 0.015

2.5

2.5 (12)

0.4

2.3 (12)

0.003

1.2 (12)

2.1

1.9 (12)

Quality Assurance for Groundwater Monitoring

387

phenomena which explained low results for chloride. It was hence recommended to analyse the materials immediately after opening the ampoules; this recommendation was supported by similar observations made for sulfate results. For ammonium, some colorimetric results were too close to the detection limits which explained the apparent spread of results. In addition, interferences in ion chromatography were observed with sodium. Considering the spread of results, the participants recommended that the ammonium values be given as indicative only in the certiﬁcation report.15 It was stressed that ICP atomic emission spectrometry (ICPAES) and ICPMS enabled total phosphorus to be determined whereas ion chromatography determined orthophosphate and colorimetry determined reactive phosphorus (both ‘‘inorganic phosphate’’). It was hence decided to certify the content of inorganic phosphate and to use the ICPAES and ICPMS results as supporting results. Argon interferences with 39 K and 64Fe were likely the cause for high standard deviations in ICPMS for potassium and iron, respectively. In the case of BCR-617, losses of Fe were suspected to have occurred during the preparation, considering the signiﬁcant diﬀerence of the mean of laboratory means ((0.191 0.012) mg kg1 in comparison to 0.206 mg kg1). Precipitation of Fe(III) could be one cause of loss during ﬁltration; participants rather suspected that adsorption occurred on the ampoule since two sets of results corresponding to solutions acidiﬁed in the ampoule yielded higher results. Since risks of release could occur in the case of adsorption, participants recommended that the Fe value in BCR-617 be considered as indicative only. The certiﬁed values are given in Table 6.3.4.

6.3.6.3 6.3.6.3.1

CRMs for Trace Elements in Groundwater Preparation of the Materials

Pre-cleaned polyethylene containers were used for the sampling, preparation and homogenisation of the candidate CRMs. The ﬁlters used for ﬁltration were Sartorius depth ﬁlter (particle retention size 0.7 mm) and Sartorius PH membrane ﬁlter capsule (particle retention size 0.45 mm). Polyethylene bottles produced by Nalgene were used for each of the reference material. Some 2600 pieces of 500 ml bottles were cleaned, using the following procedure. Detailed descriptions of the cleaning and rinsing procedures used for the containers, tubings used for ﬁltration and homogenisation, and bottles are given elsewhere.17 The sampling procedure was similar to that described in Section 5.2.2. The groundwater samples were collected from boreholes in Danish waterworks and subsequently ﬁltered on-line through 0.7 mm Sartorius PE ﬁlter by use of a peristaltic pump system. Both groundwater samples (CRMs BCR-609 and BCR-610) were acidiﬁed with suprapure nitric acid to pH o 2 and ﬁltered with a Sartobran PH membrane capsule ﬁlter (0.65 and 0.45 mm pore size). Freshly prepared solutions of trace elements were added to the samples to achieve concentration levels given in Table 6.3.5. The homogenisation of the samples was achieved by constant circulation for more than 24 h using a peristaltic pump system. Bottle ﬁlling took

388

Table 6.3.4

Chapter 6.3

Certiﬁed contents of major elements in BCR-616 and BCR-617.15 Certiﬁed value (mg kg1)

Uncertainty (mg kg1)

p (no. of datasets)

Ca Cl Fe K Mg

38.5 49.8 0.052 0.58 23.9

0.9 1.0 0.003 0.03 0.3

17 11 10 11 17

Mn Na NO3 PO4 SO4

0.0197 61.5 50.4 3.36 57.3

0.0007 0.7 0.9 0.13 1.1

13 17 10 11 10

Ca Cl K Mg

14.6 26.4 9.93 7.32

0.4 0.4 0.26 0.15

18 11 14 18

Mn Na NO3 PO4

0.050 14.6 25.6 0.272

0.002 0.3 0.5 0.009

16 17 10 7

SO4

26.3

0.5

10

BCR-616

BCR-617

p number of data sets.

Table 6.3.5

Amounts of elements added for the preparation of the candidate CRMs BCR-609 and BCR-610 and concentrations present before spiking (adapted from Ref. 17). CRM 609 (mg kg1)

CRM 610 (mg kg1)

Element

Before spiking

Amount added

Before spiking

Amount added

Al As Cd Cu Ni Pb

0.2 o0.5 0.05 2.0 3.0 o0.05

47 0.940 0.118 2.0 6.27 1.57

0.3 o0.5 o0.005 o0.5 o0.1 o0.005

155 10.1 3.10 46.6 23.3 7.76

place in a clean room. The (conditioned) 500 ml polyethylene bottles were ﬁlled with the CRM solutions by means of a peristaltic pump. All precautions were taken to avoid contamination during the sample handling. Additional details on the bottle conditioning and transport are given elsewhere.17

Quality Assurance for Groundwater Monitoring

389

The analytical methods used to verify the homogeneity and stability of the CRMs were ETAAS (Al, Cd, Cu, Pb and Ni) and hydride generation AAS (HGAAS) (As). The stability tests were performed over 52 weeks’ storage at +4 1C and +20 1C. A short-term stability study was also carried out at +40 1C to test worst-case transport conditions. No inhomogeneity or instability could be detected for any of the conditions tested in both materials.17

6.3.6.3.2

Certiﬁcation

Various analytical techniques were used in the certiﬁcation, involving diﬀerent types of pre-treatment. Calibration was generally based on the use of calibration graphs or standard additions. The techniques used in the certiﬁcation of the CRMs BCR-609 to BCR-612 are summarised in Table 6.3.6. The most common sources of systematic errors which were detected during the technical discussions were contamination problems, calibration errors or uncontrolled matrix eﬀects.17 Results obtained by total reﬂection XRF spectrometry could not be retained for certiﬁcation since this technique does not comply with traceability requirements requested for certiﬁcation; however, no doubts were expressed on its suitability for routine measurements and the results were used as conﬁrmation for As, Cu and Pb. For Al, the lack of use of modiﬁer for Al determination by ETAAS explained the low results which were consequently withdrawn. High results obtained by one laboratory for As by ICPMS were likely due to ArCl interferences; the results were withdrawn. This interference was conﬁrmed by another laboratory which tested the use of a silver nitrate cartridge to remove chloride: while no diﬀerence was observed for BCR-609 between ﬁltered and not ﬁltered samples, signiﬁcant diﬀerences were observed for BCR-610 which contained higher chlorine content (i.e. higher results were obtained for non-ﬁltered samples). A precipitate was observed by Table 6.3.6

Summary of techniques of ﬁnal determination for the certiﬁcation of BCR-609 to BCR-612.

Element

Techniques

Al As Cd Cu Pb Ni I Br

ETAAS, ICPAES, ICPMS, INAA ETAAS, HGAAS, ICPMS, INAA, TXRF DPASV, ETAAS, ICPMS DPASV, ETAAS, ICPAES, ICPMS, TXRF DPASV, ETAAS, ICPAES, ICPMS, TXRF DPASV, ETAAS, ICPAES, ICPMS, TXRF IC, ICPMS, INAA, SPEC HR-ICPMS, IC, ICPMS, INAA, SPEC

DPASV, diﬀerential pulse anodic stripping voltammetry; ETAAS, electrothermal atomic absorption spectrometry; HGAAS, hydride generation atomic absorption spectrometry; IC, ion chromatography; ICPAES, inductively coupled plasma emission spectrometry; ICPMS, inductively coupled plasma mass spectrometry; ID-ICPMS, isotope dilution inductively coupled plasma mass spectrometry; INAA, instrumental neutron activation analysis; SPEC, spectrophotometry; TXRF, total reﬂection X-ray ﬂuorescence spectrometry.

390

Chapter 6.3

Table 6.3.7

BCR-609 Al As Cd Cu Pb

Certiﬁed contents of trace elements in BCR-609 and BCR-610.17 Certiﬁed value (mg kg1)

Uncertainty (mg kg1)

p (no. of datasets)

47.7 1.20 0.164 2.48 1.63

1.6 0.12 0.012 0.09 0.04

14 8 16 9 10

159 10.8 2.94 45.7 7.78

4 0.4 0.08 1.5 0.13

12 10 13 7 9

BCR-610 Al As Cd Cu Pb

another laboratory in the BCR-610 which was attributed to the presence of higher Fe content in comparison to BCR-609. Digestion by high-pressure ashing was necessary to ensure a good recovery for As by diﬀerential pulse anodic stripping voltammetry (DPASV). In the case of Pb, it was recommended that calibration be performed by standard additions to take matrix eﬀects into account. Similar problems of precipitation in BCR-610 were experienced by one laboratory which withdrew its results obtained by HR-ICPMS, DPASV and ETAAS. A wide spread of results was observed for Ni in both CRMs. After technical scrutiny, it was found that most of the outlying results were due to either calibration errors (using external calibration) or blank problems. A strong recommendation was given to use standard additions for the determination of Ni by ETAAS in these materials; this recommendation was supported by the fact that sets of data of laboratories which applied standard additions were in good agreement (using the following techniques: ICPMS (two laboratories), ETAAS and isotope dilution ICPMS). In the case of BCR-610, high results obtained by ICPMS by one laboratory could be due to interferences on 58 Ni from ArO and NaCl which were only observed to a lesser extent for BCR609. Considering the spread of results, it was decided to consider Ni as indicative only on the basis of the data obtained by standard additions. The certiﬁed values are listed in Table 6.3.7.

6.3.6.4 6.3.6.4.1

CRMs for Bromide in Groundwater Preparation of the Materials

Pre-cleaned polyethylene containers were used to collect aerated groundwater from Danish waterworks.17 For the BCR-611 (low level), the sample was aerated for one day, then ﬁltered and pumped into a clean container; the

391

Quality Assurance for Groundwater Monitoring

Table 6.3.8

Certiﬁed contents of bromide in CRMs BCR-611 and BCR-612.17

Bromide

Certiﬁed value (mg kg1)

Uncertainty (mg kg1)

p (no. of datasets)

CRM 611 CRM 612

93 252

4 10

10 8

sample was subsequently mixed with freshly distilled water with a mechanical stirrer unit, after which a sample was taken for a pre-contamination microbiological test. No microbiological growth was found in the sample. The sample was then re-ﬁltered through a Sartobran ﬁlter (0.45/0.2 mm pore size). Samples were placed in 25 ml ampoules using an ampoule-ﬁlling machine. The ﬁrst and last ampoules ﬁlled were tested for microbiological contamination: no signs of microbiological contamination were detected. Ampoules were autoclaved at 120 1C for 20 min; they were then labelled individually and sets of four ampoules were packed in specially designed boxes. For the BCR-612 (high level), similar preparation procedures were used as for comparison with BCR611, with the exception that this material was not diluted with distilled water. Additional information on the collection and sample pre-treatment is given elsewhere.17 Bromide determinations were carried out by ion chromatography for homogeneity and stability studies. Both materials were found to be homogenous and stable over 52 weeks’ storage at +4 1C and +20 1C, and no instability could be detected at +40 1C over 15 days (worst-case transport test).17

6.3.6.4.2

Certiﬁcation

Bromide was determined solely by ion chromatography, whereas total bromine was determined by ICPMS and instrumental neutron activation analysis. Bromide was hence certiﬁed on the basis of ion chromatography measurements, whereas total Br was given as certiﬁed value.17 The large spread of results did not allow the certiﬁcation of iodide in any of the CRMs; consequently, indicative values were proposed only.17 The certiﬁed values, along with their uncertainties, are given in Table 6.3.8.

6.3.6.5

Conclusions: Availability of Water CRMs

The six CRMs described in this chapter clearly fulﬁl a demand for quality control tools for groundwater monitoring. They represent a compromise, both in terms of chemical composition (artiﬁcial materials instead of natural ones for major elements) and origin (natural ground waters collected in Denmark), which is the ﬁrst step for helping EU laboratories to achieve comparability of data in this ﬁeld of analysis. These materials also oﬀer reference tools to laboratories willing to prepare their own LRMs (e.g. better matching the composition of ground waters currently monitored in their region). Their

392

Chapter 6.3

availability undoubtedly enabled considerable improvement in the state-of-theart of groundwater monitoring. There are a number of suppliers of water CRMs, some of them being specialised in a particular ﬁeld of interest (e.g. the National Research Council of Canada for marine monitoring). Two main bodies, NIST (USA, formerly NBS) and the European Commission (Institute for Reference Materials and Measurements, BCR), cover several ﬁelds and ensure long-term availability of CRMs. Surveys are regularly undertaken by international organisations. A recent compilation of existing water CRMs is given in the scientiﬁc literature.10,14 It should be noted that CRMs are products of very high value. Typically, the production of a water CRM costs some hundred thousand euros. Therefore, these materials should be reserved for the ﬁnal veriﬁcation of analytical procedures and not as materials for routine measurements (e.g. in control charts). The CRMs described in this chapter are available at the Institute for Reference Materials and Measurements (IRMM), Retieseweg, B-2440 Geel, Belgium. They are delivered along with certiﬁcation reports containing a full description of the material preparation, the homogeneity and stability studies, the techniques used in the certiﬁcation, the technical and statistical evaluation of the results, and all the individual results provided by the certifying laboratories. Direct information can also be obtained by e-mail at [email protected] and through the IRMM website at http://www.irmm.jrc.be/mrm.html.

6.3.7

Assessment of Uncertainty Linked to Groundwater Sampling. A Case Study: The METREAU Project

6.3.7.1

Introduction

The uncertainties associated with results of chemical analyses, as deﬁned by international terminology, are widely recognised as being indispensable in terms of traceability and reliability of data transmitted to the client.19 As previously described, the consideration of uncertainties is now explicitly stated in the international normative reference ISO/CEI 17025. In addition, several guidelines describe the speciﬁcs of these obligations and/or propose calculation methodologies.20 The uncertainties of measurement are, therefore, intrinsically linked to every single analytical result. The determination takes into account every aspect of laboratory measurements and analytical procedures, from the instant the samples are received by the laboratory.21 These calculations, however, do not usually take into account the complete analytical chain. In particular, uncertainties related to sampling steps, including sampling itself, packaging, sample conditioning, transportation, etc., are often ignored or poorly quantiﬁed.22 Progress has been made with regard to sampling methodologies for solid phases, in particular in the ﬁeld of mining research,23 in which sampling representativeness is of key importance due to the well-known ‘‘nugget’’ eﬀect. In contrast, in the case of chemical analysis of aqueous phases in water, no studies speciﬁcally dealing with the notion of representativeness at the sampling stages have been performed to the authors’ knowledge.

Quality Assurance for Groundwater Monitoring

393

Awareness about this fact, regarding chemical analysis of water, is however increasing. Sampling may be deﬁned as ‘‘a process of selecting a portion of material small enough in volume to be transported conveniently and handled in the laboratory, while still accurately representing the part of the environment sampled’’.6,24 The diﬃculties mainly concern the representativeness of the medium and the conservation of samples without any signiﬁcant perturbation. These principles, and the questions and doubts they raise, are not new and have been widely discussed by a number of authors.25,26 However, solutions to these diﬃculties, particularly in terms of consensual methodologies, are still to be found. In the absence of an overall agreement on the most suitable methods/ techniques to be implemented in the ﬁeld, there is a risk that derived chemical data may be wrongly interpreted (see Section 1), not only with regard to statutory regulations but also in terms of our understanding of natural phenomena. Indeed, the sum of external factors, which may inﬂuence or interfere with a sample, needs therefore to be precisely quantiﬁed in order to ensure they do not mask real natural variations in the environment, which may be chemical, spatial or temporal. So, under the acronym of METREAU, a research project supported by the French ministry of research was carried out between 2002 and 2003. This project was conducted in the framework of RITEAU, a network of innovative technological developments in the ﬁeld of water. Partners of the METREAU project were the Laboratoire National d’Essais (LNE), Merck-Eurolab, Bipea, the Laboratoire De´partemental de Pe´rigueux (LDAR) and BRGM, the French Geological Survey.27,28 The main objectives of this project were to deﬁne and create a reliable metrological structure that would allow an improvement in the quality of analytical determinations on water samples performed by laboratories. As a part of this project, BRGM was in charge of conducting on-site tests and measurements, and implementing these factors in the evaluation of total uncertainty over the whole analytical chain, from the ﬁeld to the ﬁnal results, including any laboratory eﬀects.28 In this respect, a new methodological approach to sampling was developed in order to make an estimation of the uncertainties associated with the prelaboratory stages. This approach was applied, by way of an example, to the case of chemical analysis of heavy metals in ground waters of an industrial site.

6.3.7.2

Groundwater Sampling: New Devices

Methodologies for sampling natural water diﬀer signiﬁcantly for surface water and for groundwater. The diﬀerences are, for the most part, related to the lack of visibility inherent in groundwater run-oﬀ and the diﬃculties in accessing groundwater. Indeed, the means which are most widely used for accessing such large underground reservoirs of water for monitoring purposes are boreholes and, more speciﬁcally, those required by statutory regulations concerning the quality of groundwater.

394

Chapter 6.3

In order to overcome these diﬃculties, and following approaches developed in oilﬁeld exploration based on geophysics, a chemical logging was used to obtain an initial ‘‘chemical photograph’’ of the water table aﬀected by pollutants.29,30 The apparatus used was a multi-parameter probe which allows continuous measurements of transient physical/chemical ﬁeld parameters (pH, temperature, conductivity, redox potential and dissolved oxygen) as a function of depth. Groundwater samples were collected by means of a pressurised nitrogen system. This device, developed by BRGM, allows sampling to be done at a speciﬁcally selected depth, without any mixing with adjacent layers in the water column. Chemical proﬁles as a function of depth, called chemical logs (Figure 6.3.1), allow one to determine in a matter of minutes (descent speed around 3–5 m min1) the type of water present and to identify any chemical stratiﬁcation within the water column.30–32 The depth-dependent eﬀects, well known to hydrogeologists, are related to the diﬀerent lithologic layers constituting the soil which may generate distinct ﬂow patterns and, consequently, variable spatial and temporal diﬀusion of pollutants.

6.3.7.2.1

Example of Heavy Metals in Groundwaters of an Industrial Site

A small industrial ore processing site, located in eastern France, was chosen according to chemical data previously obtained. The site covers an area of around 15 000 m2 (about 300 m long and 50 m wide). The groundwater directly below the site is contaminated with a variety of heavy metals (Pb, Zn, As, etc.). For this purpose, analyses focused on several key metals and, in particular, Pb. Lead levels were determined at concentrations of around 300 mg l1. Three groundwater monitoring boreholes are available on this site. The initial chemical logs acquired in boreholes 1 and 2 are shown in Figure 6.3.1. Despite the proximity of the two boreholes (200 m apart), the depth proﬁles of the selected parameters exhibited signiﬁcant diﬀerences. Borehole 2 seemed to be chemically homogeneous, except for dissolved O2 levels which displayed the traditional decrease with depth. In contrast, borehole 1, located upstream of borehole 2 in hydraulic terms, exhibited signiﬁcant heterogeneity. Borehole 1 showed a distinct stratiﬁcation in the water column, marked by a shift of all physicochemical parameters at a depth of 2 m deﬁning an upper and a lower layer with diﬀerent compositions. This observation was conﬁrmed during the whole duration of the project, at each sampling period,28 even after pumping tests at diﬀerent ﬂows. These observations, acquired by logging, were fully corroborated by chemical assays of major ions and Pb (Table 6.3.9), carried out in the laboratory by ion chromatography (Cl, NO3, SO42), ICP emission spectrometry (Ca21, Mg21, Na1, K1), potentiometry (HCO3) and ICPMS (Pb21). Water samples were taken at three diﬀerent levels chosen arbitrarily (top, medium and bottom) in each of the two boreholes. In borehole 1, the results

Quality Assurance for Groundwater Monitoring

Figure 6.3.1

395

Chemical logs of boreholes 1 and 2 at the ore processing site (C ¼ conductivity at 25 1C (mS cm1), T ¼ temperature (1C), O2 ¼ dissolved oxygen (ppm and %), Eh ¼ oxido-reduction potential (mV); pH expressed as pH units).28

396

Table 6.3.9

Chapter 6.3

Chemical element content in boreholes 1 and 2 as a function of depth, in February 2002.28 Borehole 1 (mg l1)

Borehole 2 (mg l1)

Element

Top

Middle

Bottom

Top

Middle

Bottom

Na K Ca Mg Cl SO4 NO3 HCO3 Pba

65 30 19 7 70 53 12 127 0.6

26 3 25 11 50 53 12 42 0.1

25 3 24 11 51 54 11 42 0.1

35 262 44 15 268 54 5 244 0.4

36 241 49 17 254 53 4 254 0.3

36 240 48 17 254 48 4 252 0.1

a

Values in mg l1.

showed signiﬁcant diﬀerences for ions such as Na, K, HCO3 and Pb between the upper level and the two lower levels, and moderate diﬀerences for Ca, Mg and Cl. In contrast, NO3 and SO4 contents were essentially identical in the three levels. In borehole 2, chemical assays also corroborated the log proﬁle with no major changes, since the diﬀerences observed as a function of depth were minimal. A slightly lower Pb content was observable in the bottom level compared to the two upper ones, but the diﬀerence was not signiﬁcant considering the very low Pb contents measured and the large associated analytical uncertainty (see below). The observations above imply that borehole 1 is really composed of two chemically distinct layers containing diﬀerent amounts of pollutants, possibly because factors controlling pollutant concentration relate to diﬀerent processes. As a consequence several depth levels must be sampled in such situation if one is to characterise precisely the chemical composition of groundwater. For borehole 2, laboratory measurements conﬁrmed the homogeneity of water composition with depth documented on site (logging), a somewhat surprising result considering the small distance separating the two boreholes and the downstream location of borehole 2. This notion of stratiﬁcation or non-stratiﬁcation of groundwater has important implications for the strategy of sampling and, in particular, the representativeness of selected samples. This is especially important when it comes to monitoring a site for the purpose of studying transfer of contaminants in compliance with any applicable regulations.

6.3.7.2.2

Representativeness of Groundwater Sampling: The Advantage of Using a Logging Probe

This example demonstrates that groundwater should not be assumed a priori to be homogeneous from a physicochemical standpoint. Marked variations in the

Quality Assurance for Groundwater Monitoring

397

contents of dissolved chemical compounds are documented to occur at industrial sites. Such heterogeneity must therefore be taken into account prior to deﬁning a sampling strategy for groundwater in order to avoid later making decisions based on unreliable analytical values due to non-representativeness of initial sampling. More generally, the use of in situ measuring tools enables the sampling strategy to be ﬁnely tuned, not only in terms of selecting appropriate sampling levels, but also regarding the minimum number of samples required. On-site logging has the advantage of providing quick and reliable measurements. Depending on the pursued objective, it is necessary to deﬁne a proper sampling strategy when carrying out a pollution study on an unknown site. In contrast, when monitoring an industrial site in which the emission of pollutants is well constrained historically, or in the case of water pumping operations for a speciﬁc use, a single representative sample of the water column may be suﬃcient.

6.3.7.3 6.3.7.3.1

Uncertainties Associated with the Sampling Stage Case of Heavy Metals Using Speciﬁc Methodology

The uncertainties associated with the sampling stage were estimated using a methodology speciﬁcally developed for this project. For this purpose, a series of careful repetitive ﬁeld tests was carried out. However, considering the natural environment is under continuous evolution, the methodology groups together, in a single step, the repeatability and reproducibility tests of the entire analytical chain. In borehole 2, which shows the highest Pb content, water samples were collected from three diﬀerent well-deﬁned depths. Because chemical logging of this borehole did not reveal any major physical or chemical diﬀerences with depth in the water column, the three levels for sampling were arbitrarily chosen at 4 m, 8 m and 12 m. Ten water samples were collected at each of these three levels, in order to have suﬃcient data for statistical processing. Therefore, 30 independent tests were carried out at three diﬀerent depths in the borehole, covering the entire analytical chain and including collection of ten independent water samples at each level, conditioning and packaging, transportation, conservation and analysis by ICPMS. The analytical uncertainties associated with Pb present in each of the three depth levels were calculated using the ANOVA (analysis of variance) method.33 Average Pb contents measured for samples collected are shown in Table 6.3.10, along with the 95% conﬁdence interval. It appears that values are diﬀerent from those measured earlier (see Table 6.3.9). Since the data in Tables 6.3.9 and 6.3.10 are related to the same borehole, the same depths and the same sampling methodology, the fact that measured Pb contents were diﬀerent illustrates that Pb concentrations in the water may vary as a function of time in a given borehole.

398

Chapter 6.3

Table 6.3.10

1

[Pb] (mg l ) 2s (mg l1) 2r (%)

Average Pb concentrations and calculated uncertainties in each level of borehole 2 (July 2002). Measured values corrected from background contribution were used in the calculations.28 Top level, n ¼ 10

Medium level, n ¼ 10

Bottom level, n ¼ 10

0.96 0.16 18

1.20 0.33 28

0.51 0.12 24

The results of statistical calculations34 also showed that the three depth levels in borehole 2 were not homogeneous. The upper two levels were statistically equivalent, but diﬀerent from the bottom level, in which measured concentrations were signiﬁcantly lower, by a factor of nearly 2. In the concentration range of the samples studied, the analytical repeatability and reproducibility tests carried out using ICPMS (Plasma Quad 3) for Pb indicated that analytical uncertainty is no more than 3%. Therefore, the total calculated uncertainty, which combines ﬁeld uncertainties and analytical uncertainties, also included the eﬀect of non-homogeneity of Pb in the water column within borehole 2. Unfortunately, the exact contribution of this eﬀect cannot be quantiﬁed directly, since the uncertainty speciﬁc to the ﬁeld tests is not known. In the light of these results, it is postulated, however, that true total (‘‘ﬁeld+analytical’’) uncertainty did not exceed the lowest calculated value of 18% for the top level (Table 6.3.10). A corollary to this is that the excess uncertainty factor obtained in the two other depth levels is essentially due to the non-homogeneity eﬀect as a function of depth. The results above illustrate that if a ﬁeld protocol is properly deﬁned and applied in a meticulous manner, the magnitude of uncertainties associated with groundwater sampling stages is overall comparable to that of analytical uncertainties. This is an important observation, which was not intuitive prior to this case study. In fact, one might have intuitively assumed that the upstream sampling stages would induce much higher uncertainties than purely analytical uncertainties.

6.3.7.3.2

Case of Heavy Metals without any Speciﬁc Methodology

The calculations described above were based on repetitive tests that were reproduced ten times in each water level. Let us now consider a hypothetical sampling scenario closer to traditional ﬁeld practices, i.e. a sampling strategy limited to the collection of a single sample of water, taken at random in the borehole. In such a case, one may assume that the laboratory measurement of Pb concentration for this single sample would fall on or close to the average of the

Quality Assurance for Groundwater Monitoring

399

30 measurements performed in our tests. This average (Pbmean ¼ 0.88 mg l1) is the closest estimate of the true value. In this case, the associated uncertainty may be calculated as the blind prediction of an unknown value, using the following relationship: pﬃﬃﬃ Blind prediction ðANOVA methodÞ 1:96s= n where s is the standard deviation for the method and n is the number of samples (¼ 1). Hence ½Pbmean ¼ 0:88 0:82 mg 11 The uncertainty calculated in this way is 0.82 mg l1 (95% conﬁdence), in this example almost identical to the Pb concentration hypothetically measured for the single sample. This means that the uncertainty associated with a single data point would represent nearly 100% of Pb concentration present in the water. This is undoubtedly a situation which would lead to serious diﬃculties in terms of interpretation.

6.3.7.4

Conclusions

Based on this case study of an industrial site exhibiting groundwater contaminated with heavy metals, especially with Pb, the advantage and necessity of considering ﬁeld stages in order to have a full understanding of environmental phenomena have been demonstrated. In addition, a new approach to groundwater sampling has been tested and validated, utilising in situ chemical logging. This methodology has signiﬁcant potential. It is technically simple and rapid, and the cost is in proportion with the information provided. This work enables an estimation of the representativeness of groundwater sampling and calculation of the associated sampling uncertainties in order to establish an overall evaluation of the impact of ﬁeld activities compared to analytical measurements. As an example, for Pb concentrations in the area of 1 mg l1, sampling uncertainty did not exceed 18% of the measured value, if a reﬁned sampling strategy and procedure are applied. In contrast, if sampling is limited to a single sample collected at random (‘‘blind’’) in the water column, the uncertainty at such low concentration approaches 100%. The chemical logging-based approach proposed here for the so-called ‘‘ﬁeld’’ stages constitutes a signiﬁcant improvement in data quality and representativeness. It is a very critical step in building up awareness of the fact that reliable and eﬃcient decisions cannot be made on the basis of a single analytical value. This particular study is a good illustration that natural changes in the medium can generate variations in the data much more signiﬁcant than the eﬀects of human actions (such as sampling, packaging, transportation, analysis).

400

Chapter 6.3

This study provides a good basis for an approach that could be adopted and implemented by investigators worldwide. It is particularly relevant and appropriate for natural groundwater monitoring programmes conducted at high-risk sites. Furthermore, it highlights the needs for specifying, as soon as possible and within a regulatory context, best-suited sampling methodologies for groundwater, particularly in the context of the implementation of the new Groundwater Directive and related guidelines.35

Acknowledgements The preparation of CRMs BCR-616 and BCR-617, as well as the homogeneity and stability studies and the coordination of the certiﬁcation campaign, were performed by the Empresa Portuguesa das A´guas Livres in Lisbon (Portugal) under EC contract MAT1-CT93-0012. The analyses for certiﬁcation were carried out by the following laboratories: Aristotle University, Lab. Analytical Chemistry, Thessaloniki (Greece); Compagnie Ge´ne´rale des Eaux, Anjou Recherche, Maisons-Laﬁtte (France); CSIC, Instituto J. Almera, Barcelona (Spain); Empresa Portuguesa das A´guas Livres, Lisbon (Portugal); GSFForschungszentrum fu¨r Umwelt und Gesundheit, Oberschleißheim (Germany); Institut Scientiﬁque de Service Public, Lie`ge (Belgium); Istituto Italiano di Idrobiologia, CNR, Pallanza (Italy); Istituto di Ricerca sulle Acque, CNR, Brugherio (Italy); KIWA NV, Nieuwegein (The Netherlands); Lyonnaise des Eaux, CIRSEE, Le Pecq (France); Service Central d’Analyse, CNRS, Vernaison (France); Swedish University of Agricultural Sciences (Sweden); Technologiezentrum Wasser, Karlsruhe (Germany); Universidad Complutense, Depto. de Quı´ mica Analı´ tica, Madrid (Spain); Universiteit Gent, Instituut voor Nucleaire Wetenschappen, Gent (Belgium); Water Quality Institute, Hørsholm (Denmark); Yorkshire Water, Alcontrol, Rotherham (UK). The preparation of CRMs BCR-609-610 and BCR-611-612, the veriﬁcation of their homogeneity and stability and the management of the certiﬁcation campaign were carried out by the Water Quality Institute (VKI) in Hørsholm (Denmark) and KIWA NV in Nieuwegein (The Netherlands) under EC contract MAT1-CT94-0002. The following laboratories participated in the certiﬁcation campaign: Anglian Water Services Ltd, Cambridge (UK); Aristotle University, Lab. of Analytical Chemistry, Thessaloniki (Greece); A/S AnalyCen, Fredericia (Denmark); Bundesamt fu¨r Seeschiﬀahrt und Hydrographie, Hamburg (Germany); Canal Isabel II, Madrid (Spain); CNRS, Service Central d’Analyse, Vernaison (France); Empresa Portuguesa das A´guas Livres, Lisbon (Portugal); Energieonderzoek Centrum Nederland, Petten (The Netherlands); Finland Miljo¨central, Helsinki (Finland); GSF-Forschungszentrum fu¨r Umwelt und Gesundheit, Oberschleißheim (Germany); Istituto Superiore di Sanita`, Roma (Italy); Institut Scientiﬁque de Service Public, Lie`ge (Belgium); KIWA NV, Nieuwegein (The Netherlands); Ministe`re de l’Inte´rieur, DGCCRF, Talence (France); NIVA, Oslo (Norway); NV-PWN Waterleidingbedrijf Noord-Holland (The Netherlands); State Laboratory, Dublin

Quality Assurance for Groundwater Monitoring

401

(Ireland); Stockholm Universitet, ITM, Solna (Sweden); Sociedad General das Aguas de Barcelona (Spain); Universiteit Gent, INW, Gent (Belgium); Water Quality Institute, VKI, Hørsholm (Denmark); Water Research Centre, WRC, Medmenham (UK). With regard to the METREAU project, the French Ministry of Research (Riteau network) is acknowledged for providing ﬁnancial support. We also acknowledge the Metreau project partners: the Laboratoire National d’Essais, Merck-Eurolab, Bipea, the Laboratoire De´partemental d’Analyses et de Recherche de Pe´rigueux (France). We would also like to acknowledge the in situ measurement team of BRGM (M. Brach, G. Braibant, J. C. Foucher and F. Jouin). Special thanks are also expressed to the industrial company for permission to work on site and for cooperation.

References 1. Ph. Quevauviller, Trends Anal. Chem., 1994, 13, 404. 2. M. J. Benoliel, Ph. Quevauviller, E. Rodrigues, M. E. Andrade, M. A. Cavaco and L. Cortez, Fresenius’ J. Anal. Chem., 1997, 358, 574. 3. Ph. Quevauviller, K. Andersen, J. Merry and H. van der Jagt, Analyst, 1998, 123, 955. 4. Ph. Quevauviller, M. J. Benoliel, K. Andersen and J. Merry, Trends Anal. Chem., 1999, 18, 376. 5. E. Prichard, Quality in the Analytical Laboratory, John Wiley, Chichester, 1995. 6. Ph. Quevauviller, Quality Assurance for Water Analysis, John Wiley, Chichester, 2002. 7. Ph. Quevauviller, K. Kramer and T. Vinhas, Mar. Pollut. Bull., 1994, 28, 506. 8. J. K. Taylor, Handbook for SRM-Users, NBS Special Publication 260-100, 1985. 9. J. Vogelsang, Fresenius’ Z. Anal. Chem., 1987, 328, 213. 10. Ph. Quevauviller and E. A. Maier, Certiﬁed Reference Materials and Interlaboratory Studies for Environmental Analysis: The BCR Approach, Elsevier, Amsterdam, 1996. 11. M. J. Benoliel and L. Cortez, Groundwaters of the European Community, Report of the Contract 5528/1/9/297/91/08/BCR-P, European Commission, 1992. 12. Ph. Quevauviller, M. Va´lcarcel, M. D. Luque de Castro, J. Cosano and R. Mosello, Analyst, 1996, 121, 83. 13. Ph. Quevauviller, K. J. M. Kramer and T. Vinhas, Fresenius’ J. Anal. Chem., 1996, 354, 397. 14. K. J. M. Kramer, Analyst, 1998, 123, 991. 15. Ph. Quevauviller, M. J. Benoliel, R. Neves Carneiro and L. Cortez, EUR Report, 18286 EN, European Commission, Brussels, 1998.

402

Chapter 6.3

16. Ph. Quevauviller, M. J. Benoliel, R. Neves Carneiro and L. Cortez, Fresenius’ J. Anal. Chem., 1999, 363, 23–27. 17. P. Quevauviller, K. Andersen, J. Merry and H. van der Jagt, EUR Report, 18317 EN, European Commission, Brussels, 1998. 18. P. Quevauviller, K. Andersen, J. Merry and H. van der Jagt, Sci. Total Environ., 1998, 220, 223. 19. NF X 07-001, Vocabulaire international des termes fondamentaux et ge´ne´raux de me´trologie, 1994. 20. Eurachem/CITAC Guide, Quantifying Uncertainty in Analytical Measurements, 2nd edn, 2000. 21. M. Feinberg, La validation des me´thodes d’analyses, une approche chimiome´trique de l’assurance qualite´ au laboratoire, ed. Masson, Paris, 1996. 22. M. Deleuil, Introduction au proce´de´ d’e´chantillonnage, Analusis, 1999, 27(6), 504. 23. P. Gy, Sampling and Particulate Materials. Theory and Practice, Elsevier, 2nd edn, 1982. 24. Ph. Quevauviller, Sampling and sample handling, in Quality Assurance for Water Analysis, Water quality measurements series, John Wiley, 2002, ch. 3. 25. L. H. Keith, Environmental Sampling and Analysis: A Practical Guide, Lewis Publishers, 1991. 26. J. Krajca, Water Sampling, Ellis Horwood books on water and wastewater technology, 1989. 27. P. Charlet and A. Marschal, Trends Anal. Chem., 2004, 23, 178. 28. S. Roy and A.-M. Fouillac, Trends Anal. Chem., 2004, 23, 185. 29. N. Guigues, S. Roy, J. C. Foucher, J. L. Pinault and F. Lenain, International Meeting on Chemical Sensors, Boston, MA, July 2002. 30. S. Roy, A.-M. Fouillac and M. Coroller, 3rd Senspol Workshop on Monitoring in Polluted Environments for Integrated Water-Soil Management, Krakow, June 2003. 31. A.-M. Fouillac and S. Roy, ChemRawn XV: Quality Assessment and Metrology, Paris, June 2004. 32. A. Strugeon, N. Guigues, T. Despas, S. Roy and A. M. Fouillac, International Metrology Congress, Lyon, June 2005. 33. A. M. H. Van Der Veen and J. Pauwels, Accred. Qual. Assur., 2000, 5, 464. 34. A. M. H. Van Der Veen and J. Pauwels, Accred. Qual. Assur., 2001, 6, 26. 35. European Commission, Groundwater Monitoring Guidance, WFD Common Implementation Strategy, in press.

7. Groundwater Pollution Prevention and Remediation

CHAPTER 7.1

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites: Integrated Management Strategy and its Application on Megasite Cases JEROEN TER MEER, HANS VAN DUIJNE, ROB NIEUWENHUIS AND HUUB RIJNAARTS TNO Netherlands Institute of Applied Geoscience – National Geological Survey, Princetonlaan 6, PO Box 80015, NL-3508 TA Utrecht, The Netherlands

7.1.1

Addressing Groundwater on Large (Former) Industrial Sites

The European Union (EU) Water Framework Directive (WFD)2 and the Groundwater Directive (GWD)3 consider groundwater as a valuable natural resource which should be protected for future generations. There are several human activities that cause a deterioration of groundwater quality, and one of these is pollution from industries. As a result of industrial activities, chemicals are accidentally spilled and result in soil contamination. Some contaminants have a tendency to adsorb to the soil and remain in the top layer, whereas others are likely to move to the deeper layers and disperse into the groundwater. In the groundwater the contaminants might dissolve and form a plume in the direction of the groundwater ﬂow. The site-speciﬁc situation and conditions determine the extent of this plume and its impact on the groundwater and nearby receptors (e.g. surface waters, drinking water abstractions and aquatic and terrestrial ecosystems). 405

406

Chapter 7.1

For individual small-scale industrial sites with single point sources, the impact on the groundwater system will be limited, and due to the limited extent of the contamination it will be possible to take adequate manageable measures to protect the groundwater and nearby receptors. Megasites are large-scale sites or areas where a large number of industrial activities have taken place for long periods (i.e. decades or centuries) and in some cases still continue. These activities have resulted in a contamination from diﬀuse and/or multiple point sources. At megasites the land is historically contaminated and hazardous substances have often already entered the groundwater and have formed plumes. Typically, the extent and/or number of plumes is relatively large compared to individual small-scale industrial sites. From a technical and economical point of view it will not be feasible within the megasite area to remove the contamination from the groundwater and to comply with threshold values that are intended for pristine groundwater systems (disproportionality principle). Instead, a risk-based approach is the only feasible way to prevent the future impact on the surrounding groundwater system and receptors. It has been estimated that in Europe there are tens of thousands of megasites, representing environmental redevelopment costs of up to h100 billion. Hence due to the large number of megasites and the huge environmental and ﬁnancial consequences, a distinct approach for managing the groundwater and surface water quality at megasites is of considerable importance to comply with the objectives of the EU WFD and GWD.

7.1.2

Risk-based Approach for Contaminated Megasites

7.1.2.1

Integrated Management Strategy

An integrated management strategy (IMS) has been developed in the EU 5th framework project Welcome.1 The IMS assists megasite owners and their consultants in establishing a risk-based management plan that complies with the WFD/GWD and protects the water resources at and around the megasite through a cost-eﬃcient programme of measures. In essence, the IMS involves the identiﬁcation and application of measures to control and reduce emissions from polluted sites. Pollutants travel from areas with large amounts of contaminants in soil and groundwater towards other (ground)water and soil systems (receptors), which have to be protected because good quality status needs to be achieved or maintained. The source–pathway–receptor concept is the basis of the risk-based management approach and focuses primarily on the protection of the receptors, rather than on the removal of sources. Taking such an approach means: evaluating the impact of contaminants to diﬀerent receptors, considering all possible exposure pathways; identifying potential measures to reduce or eliminate the impact; and setting priorities with respect to timing and the allocation of ﬁnancial resources, taking the ﬁnancial situation of the local economy into account.

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites

7.1.2.2

407

Megasite Objectives

In accordance with the WFD and the GWD three objectives can be deﬁned for managing the groundwater quality at megasites: 1. The main objective for groundwater in the WFD is to control the chemical status of the entire groundwater body. Megasites interact with (a part of) a certain groundwater body, and therefore eﬀects on the chemical status of the groundwater body as a whole must be prevented. 2. For megasites where industrial activities still take place, suﬃcient protective measures should be taken to prevent the occurrence of new contamination in soil and groundwater. 3. For megasites where the groundwater is historically contaminated, measures must be taken to reduce the risks for receptors and further pollution. The term ‘‘receptor’’ must be taken in its widest context to include not only the existing uses of groundwater but all possible future uses to which the groundwater might be subjected. (Note that in most EU member states the groundwater itself is considered as receptor, independent from its use.)

7.1.2.3

Concept for Megasites

A generic concept has been developed that is applicable to all types of megasites and provides suﬃcient freedom to accommodate case-speciﬁc aspects. The concept was developed in the Welcome project and is shown in Figure 7.1.1. The point of departure in the concept is the protection of the receptor. The quality of the groundwater at the receptor has to comply with certain targets. To control these targets, points (or lines, or planes) are deﬁned: the so-called points of compliance (POC). At or close to the receptor the third POC is situated. The targets are threshold values that have to be achieved and are further referred to as values of compliance (VOC). The VOCs depend on the functional use of the groundwater and the ecological importance. In some EU member states, the groundwater is also considered to have an intrinsic value which makes it necessary to deﬁne VOCs, even in situations where the groundwater is not used nor aﬀects areas with a high ecological importance.

Figure 7.1.1

Concept for megasites.

408

Chapter 7.1

In practice these VOCs are based on or equal to risk-based values and/or legislative standards, which are deﬁned by the members states. As an early warning for a possible impact to the receptor, upstream (in the direction of the source area) another POC is situated (second POC). The position of the second POC and the corresponding VOC must be suﬃciently protective for the third POC, and has to make sure that suﬃcient time and space is available to take measures. The preferred position and values are casespeciﬁc and depend on the fate and transport of the contamination, the legislative boundary conditions of member states and local considerations (liability, costs, technical feasibility, accessibility and restrictions for monitoring network). For some megasites the second POC is equal to the boundaries of the megasite itself to control the impact of the contamination on the surrounding groundwater, while for other megasites the second POC is situated closer to the receptor to create more time and space for management and measures. The VOC can vary from a strict legislative standard to a purely risk-based value which considers long travelling times towards the receptor and natural attenuation mechanisms that reduce the load of contaminants. In this latter case the corresponding VOCs at POC 2 can therefore be less stringent than the VOCs at the receptor itself (POC 3). Further upstream, the ﬁrst POC is located at the top of the groundwater table, or close to the source area in the aquifer. At this POC it is checked to what extent the contaminants reach the groundwater zone from which further transport takes place. At some megasites this takes place in shallow groundwater layers; for others it takes place primarily in the deeper groundwater layers. Since megasites are historically contaminated, the groundwater has already been contaminated to some extent. The primary aim of this plane of compliance is to determine concentration trends and processes that aﬀect contaminant transport and aﬀect the receptor in the future. This information can be used to take measures at an early stage. In case the groundwater below this POC 1 is considered as receptor because the groundwater is abstracted and used, or considered to have an intrinsic value, VOCs have to be deﬁned to obtain the required quality level. POC 0 is located at the bottom of the source in the unsaturated zone (for point sources as well as for diﬀuse sources). It can be described as the point at which the input enters the soil–groundwater environment. The purpose of POC 0 is to assess whether the groundwater could be aﬀected in the future. For megasites this POC 0 might be useful to characterise the contaminant load and the potential ecological and human risks at the surface. The POC 0 is only useful as compliance point for groundwater quality when a risk assessment is carried out with leaching models. For megasites the application of VOCs at this POC 0 is not considered appropriate.

7.1.2.4

Megasite Categories

Diﬀerent (types of) megasites result in diﬀerences in the impact on the groundwater system. At some megasites the situation is critical due to the

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites

409

extent of the combined plumes and the sensitivity of the receptor, while at other megasites the plumes are unlikely to reach the receptors. Based on the level of impact on the groundwater system, diﬀerent categories can be distinguished. (i) Contamination present in unsaturated zone. In this category the contaminants are not expected to reach the groundwater (Figure 7.1.2). The physical and chemical properties of the contaminant prevent it becoming mobile and/or the hydrogeological situation hampers the downward transport to the groundwater. This is the case in situations where seepage of the groundwater takes place or impermeable clay layers are present. Because the groundwater system is not impacted, this category is not relevant in the context of groundwater management and therefore not further elaborated. (ii) Contamination present in (phreatic) groundwater zone. The contamination has locally reached the phreatic groundwater, but due to the limited extent will not further aﬀect the surrounding and deeper groundwater systems (Figure 7.1.3). This category will not require a risk-based approach at megasite level to manage the groundwater contamination. A site-speciﬁc approach will be feasible. (iii) Contamination present in aquifer. As a result of historical contamination, the contaminants have already reached the groundwater and potentially lead to a further impact to the surrounding groundwater system (Figure 7.1.4). Four diﬀerent stages are distinguished to indicate the (expected) level of impact and the required action. a. Alert stage: time and space are available to control the plume. The VOCs are not exceeded now, nor expected to be exceeded at POC 2 in the future. The situation can be managed by monitoring and no active measures are required. b. Active stage: the VOCs at POC 2 are exceeded. This indicates that the plume is not under control and forms a future risk for the receptor(s). In addition to monitoring the plume, active (remediation) measures

Figure 7.1.2

Megasite category I.

410

Chapter 7.1

Figure 7.1.3

Megasite category II.

have to be taken to protect the receptor(s). These measures can be part of a long-term risk management strategy. c. Critical stage: the VOCs at POC 3 have been exceeded at POC 3 and form a direct risk for the receptor(s). Active measures have to be taken immediately to protect the receptor(s).

7.1.2.5

Modelling and Monitoring

To be able to categorise a megasite, the impact of groundwater contamination needs to be predicted for the current as well as the future situation. Based on the predicted level of impact the megasite can be divided into clusters. Having a diﬀerent level of impact, the clusters obtain a certain priority and requirement for risk management scenarios. Due to the complexity and extent of contamination at megasites, a reactive transport model is required to make such predictions. The POCs should be included in the model to predict if the targets (VOCs) will be achieved now and in the future. Modelling alone is not suﬃcient and always has to be combined with monitoring. Modelling results give a prediction of the situation with a relatively large degree of uncertainty. Monitoring data are needed to determine whether or not the VOCs are actually achieved and to indicate contaminant concentration trends. Modelling and monitoring go hand in hand. The model indicates where a potential (future) impact to the groundwater takes place and where the monitoring should be carried out. The monitoring data can be used to verify and improve the model.

7.1.2.6

Risk Management Scenarios

In the ﬁrst place, monitoring is intended to conﬁrm that the targets are met at POC 3 and POC 2. Active measures must be taken as soon as the VOCs are exceeded at these corresponding POCs. In the case of POC 3 immediate action is required to protect the receptor. If POC 3 is not yet impacted, more time and

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites

Figure 7.1.4

411

Megasite categories (a) IIIa, (b) IIIb and (c) IIIc.

space is available to implement the optimal measures. Additional monitoring data at POC 1 and POC 0 will increase the understanding of the long-term plume behaviour, which makes it possible to rely on natural attenuation to control the groundwater contamination. Together with active measures the natural attenuation processes will lead to a reduction of the impact at POC 3 and POC 2. By providing space and time for natural attenuation to take place in parts where contamination causes a limited impact to the groundwater system and focusing the active measures primarily on those parts where the impact to the receptor is urgent, an optimal and cost-eﬃcient risk management scenario is established.

412

Chapter 7.1

Besides controlling the historical contamination, measures are needed for preventing new contamination entering the groundwater. These measures are typically taken at the industrial facilities and include impermeable ﬂoors and early warning systems for leakages. Also remediation measures in the unsaturated zone can be implemented to prevent the contamination reaching the groundwater.

7.1.3

Relevance of Risk-based Management of Megasites for the Groundwater Directive

Many countries are developing risk-based strategies for the management of megasites such as: large industrial sites; soil and groundwater under urban regions; and diﬀusely contaminated rural areas. For the implementation of the GWD, megasites can be seen as a combination of plumes resulting from historically contaminated land. To implement the GWD, member states shall carry out additional trend assessment for identiﬁed pollutants in order to verify that plumes resulting from megasites: do not expand; do not degrade the chemical status of the body of groundwater; and do not present a risk for human health and the environment. Objectives for managementw of contaminated groundwater plumes within the megasite are based on the risks for important receptors: e.g. the surrounding groundwater, protected areas, the environment and human health. The type of management of megasites depends on factors that are listed in Table 7.1.1. The management of megasites should be summarised in a risk-based management plan, which can be part of a river basin management plan.

7.1.3.1

Case Studies

A number of cases have been collected in Poland (Tarnowsky Go´ry), France (eastern France and Alsace), Belgium (Kempen, Antwerp), Germany (Bitterfeld) and The Netherlands (Kempen, Rotterdam). All these cases are archetype 3 (see Table 7.1.1). According to the approach that has been indicated above, these cases have been elaborated and combined in Table 7.1.2. These cases include examples of industrial megasites and urban areas with a large-scale groundwater contamination. w

We focus in this section on the risk-based management of existing historical contamination. Another important aspect of management of megasites is prevention and remediation of new pollution.

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites

Table 7.1.1

413

Factors inﬂuencing megasite management.

Archetype

1

2

3

Type of pollution Hydrogeology

Immobile Not relevant

Receptors (aquifer, surface water, terrestrial ecosystems, protected area) Objectives and measures (risk-based management, prevention, remediation measures within socioeconomic and legislative context) Monitoring

Not relevant

Mobile Prevents spreading of pollution Not relevant

Mobile Favours spreading of pollution Identify receptors that may be inﬂuenced by plume

Focused on preventing new pollution and managing human and ecological risks in the polluted area

Focused on preventing new pollution and managing human and ecological risks in the polluted area

Focused on new pollution and human and ecological risks in the polluted area

Focused on new pollution and human and ecological risks in the polluted area; and controlling so that spreading outside the area does not take place

Low

Moderate

Identify objectives for the quality of receptors and the quality of the plume and measures to obtain these objectives. In cooperation with legislation and stakeholders Focused on new pollution and human and ecological risks in the polluted area; and controlling so that spreading outside the area does not exceed values of compliance (at points of compliance) and on the eﬀect on identiﬁed receptors High

Management intensity

For each of these cases the steps indicated in Figure 7.1.5 have been followed and have been checked for a clear transparent description of the megasite from ‘‘actions to measures’’ that were already made or need to be taken to comply with the requirements from the WFD and GWD for good chemical status and trend reversal. If the requirements cannot be met in the established timeframe a risk-based management plan needs to be elaborated. This management plan develops in several loops within certain periods. The ﬁrst loop can, for instance,

Antwerp (B)

Kempen (NL/B)

Bitterfeld (D)

Questions addressed in the examples. Tarnowsky Gory (PL)

France site 1 (F)

Alsace site 2 (F)

X X

X

X

X

What is the location of points of

The POCs are those borders on which active

Various pollution paths are investigated

Surface water, aquifer, industrial

Source, pathway and receptor clusters

3. Area deﬁnition: what are the criteria rules that we can use to deﬁne the boundaries of the area (contaminated land/plume)? Approx. 2600 km2 Whole port area, 14 What is the size of the A good number of sites 5–6 km2 delineated by 000 ha. The contaminated area/ of a few km2. Total plume maximum pilot smaller in plume in relation to extent area of megasite: size groundwater body? approx. 15 km2

2. Describe the situation Situation 1 Situation 2 Situation 3a Situation 3b Situation 3c Situation 3d

Drinking water wells, surface water

40 km2

X

Aquifer, urban areas, arable land

200 km2

X

1. What should be reported to the EU about megasites (contaminated land and plumes resulting from contaminated land) in the river basin management plan? Result of monitoring Conclusions from Several surveys have Monitoring combined Both (conceptual) Survey for smaller Results of monitoring models been completed with modelling models and part harbour or only monitoring area conclusions? Contamination very IMS applied and phases Complex situation Long-term remediation Description of the Sources are Management / large and no approached with with short term options for megasite and the identiﬁed remediation measures indicated conceptual model applied and longcontaminated consequences in according to measures? in structural way and afterwards term measures groundwater to near and far future activity map of monitored considered (60–100 meet WFD criteria the port years). Further development for removal of waste necessary

Type

Table 7.1.2

Contamination largely restricted to groundwater within port area. The impact on the groundwater body Rhine-West is expected to be limited POCs are the border of the port area and the bottom of the

X (expected)

Treatment of sources according to agreement with port of Rotterdam (owner) and industrial companies (renter). Long-term management of contaminated groundwater plumes according to risk-based approach

Monitoring combined with modelling

Port of Rotterdam

414 Chapter 7.1

monitoring takes place¼cluster border Surface water, aquifer (shallow), Schelde river. Water does not have an intrinsic value

Measures are taken to prevent the pollutants ﬂowing into the surface water

Close the tap policy. Reduction of source input; removal of zinc ashes and other contaminants. Ongoing!

Applied. At risk are two separate Triassic aquifers

Complex situation in conceptual model presented

Yes with simpliﬁed assumptions

Reverse source–path– receptor approach applied

The waste stocks still are present and form a continuous active source

Containment of plume

Essential removal of sources (wastes) was done

Eﬀorts are made to close all the taps

5. What are appropriate measures, best available technique or economically feasible? Being investigated Applied Source–path–receptor IMS applied based on receptor protection

New contamination: close the tap!

Risk-based management at megasite scale is the only way to manage the groundwater contamination Immediate removal of the entire contamination Diﬀerent measures have been taken. It seems no riskbased management or IMS has been applied

IMS applied

IMS is applied. Contaminations were successfully contained

Risk-based management applied

The receptor is the groundwater outside the port area. The groundwater zone within the port area provides time and space for NA

Threshold values are a criterion at the planes of compliance (and a protective for the surrounding groundwater body). Trend reversal is the criterion for plumes within the port area

Aquifer, urban areas, arable land

ﬁrst aquifer (25 mbgl)

2015 will not be reached to obtain good status water quality

Surface water, arable land, regional aquifer

Trend reversal for water bodies by 2015

Surface water, two groundwater aquifers, urban areas, arable land

Combined remediation policy to obtain trend reversal; this could take a time span of 100 years

Groundwater does not have an intrinsic value

Groundwater extraction wells, surface water (brooklets), nature reserve and agricultural land

are identiﬁed and monitored

Trend reversal is being implemented; however, the results can only be expected after many years

complexes and urban areas

and need to be identiﬁed

4. What are the objectives for the quality of the plume? The total inventory has Inventory has only Standstill/trend not been completed been done for reversal/threshold yet. Trend reversal smaller part of values/other is main objective the harbour compliance values area (4th dock). (see guidance Limited sourcedocument on path-receptor prevent and limit) approach was elaborated. Reduction of ﬂux of contaminants by 80% IMS under construction Risk-based Historical as part of the riskmanagement contamination: based management can be used and risk-based plan is coherent with management Flemish regulations

What are the receptors (does groundwater have an intrinsic value and should it therefore be considered as a receptor itself)?

compliance (where and how deep)?

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites 415

Risk-based approach and with economically and technically sound measures

Yes

Proportionality/ feasibility

Cost eﬃciency

Measures. for monitoring programme ﬂux calculations are done

In relation to measures

Consequences

Relation monitoring with measures

Measures are being studied. Focus on active sources removal; socially acceptable management, which is risk controlled, receptor oriented and cost eﬃcient Active groundwater monitoring in combination with modelling and management. Defensive monitoring

Several measures are taken: hydraulic ﬂow measures, source removal, ENA, reactive barriers. Necessary to create an organisation that manages overall megasite Monitoring systems applied

Yes

Measures are designed according to measurements

100 monitoring wells

Yes

No measures taken yet

Jointly monitoring and modelling

Great deal has been done with modelling

No deﬁnite measures are proposed. CBA are only directing in the application of area of derogation

Not clear

Yes, but disproportionate

Yes

Yes, aiming at cost eﬃcient remediation

Disproportionate cost arguments; CBA are c1

Including NA

Risk-based approach applied for groundwater contamination and NA is considered because of high groundwater remediation costs Yes

Due to the improved approach initiated an economic development during the past 10 years

Extend is very big, shallow groundwater extractions are stopped to comply with environmental standards for the region Yes; isolated sitespeciﬁc approaches not cost eﬃcient

Alsace site 2 (F)

France site 1 (F)

Tarnowsky Gory (PL)

Bitterfeld (D)

Kempen (NL/B)

6. How do we deﬁne our monitoring programme (do we have to deﬁne a monitoring programme)? Yes Yes, for smaller part Yes, POCs are deﬁned Spatial and temporal of the harbour to further envisage scale/point of contamination compliance trends

Antwerp (B)

(Continued ).

Type

Table 7.1.2

Yes

Compliance monitoring at POC at border of port area; trend/ process monitoring at highly impacted areas within port area Violating the monitoring objectives yields an intensiﬁcation of the monitoring eﬀort and/or appropriate remediation measures

Cost reduction of factor 3 expected compared to conventional approach

Risk-based approach with economically and technically sound measures

Port of Rotterdam

416 Chapter 7.1

Pollution Size (km2) POC exceeding IMS application cycle Status for WFD reporting, trend analyses (IMS step 1-2; GWD Art. 5.5) Status for WFD reporting, measures (IMS step 3-4; GWD Art. 6.a) Cost eﬃciency regional/ megasite approach

‘‘Real’’ or virtual (modelling) POCs. Situation 3b

Real POCs

Kempen (NL/B) Heavy metals 2600 3, strongly 1–2 Several site surveys, modelling

Measures in preparation

Yes; isolated sitespeciﬁc approaches not cost eﬃcient

Real POCs

Antwerp (B) Petrochemicals 140 2 0–1 Monitoring and modelling; part of site

No active measures prepared yet

Yes

Measures are designed

Yes

Yes

Yes, aiming at cost eﬃcient remediation

France site 1 (F) Trichloro chemicals 40 3 0–1 Monitoring

Real POCs

No measures taken yet.

Tarnowsky Gory (PL) Boron, TCE 5–6 3, strongly 1 Monitoring and modelling

Real POCs

Measures are taken and further developed

Bitterfeld (D) Chloro chemicals 15–20 3, strongly 2–3 Detailed monitoring and modelling

Real planes of compliance

No deﬁnite measures proposed. CBA directing to derogation Yes, but disproportionate

Alsace site 2 (F) Potash mining 200 3, strongly 0 Modelling

Virtual modelling is done, reﬁning of models, etc.

Intensiﬁed monitoring, appropriate remediation measures prepared Cost reduction of factor 3 compared to conventional approach

Port of Rotterdam (NL) Petrochemicals 800 2 2 Site surveys, modelling, partial detailed modelling

Both

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites 417

418

Chapter 7.1 Action Plan: Medium-term results 3 – 7 years

Action Plan: Long-term results 7 – 10 years

Action Plan: Short term results 1 – 3 years

Implementation Management plan Monitoring program Risk management scenario’s New MegaSite

Basic scenario’s

- no inventory Risk assessment

- no actionplan

Megasite characterisation Problem definition Boundery conditions Inventory of information Building conceptual model

Figure 7.1.5

Review process

Potential & preferred scenario’s Final scenario

Clustering Modelling

Evolving Risk Based Management Plans using a Integrated Management Strategy delivering to River Basin Management Plans

Flow chart of integrated management strategy.

be made in a relatively short period (0–3 years) to characterise the situation, assess the risks and deﬁne measures to prevent and limit inputs of pollutants. The cycle needs to be further elaborated to acquire more knowledge and data for improving the megasite characterisation and to formulate a more profound plan to take active measures. However, in the meantime the management plan can be approved and used for an established period. In some cases the competent authority can issue a (partial) derogation for a megasite or a cluster within the megasite. This megasite or cluster can be approached as a separate groundwater body, for which derogation is applied. Alternatively, the megasite or cluster remains part of the groundwater body, but has distinct (less stringent) risk-based objectives. A prerequisite is that these objectives are suﬃciently protective for the rest of the groundwater body. The next loop will be initiated in case more information and results from measures become available. This can be achieved in a moderate time span of 4– 7 years. The ﬁnal loop takes some 7–10 years and goes through the management system again. It focuses on even more additional measures that are required to comply with the criteria of the WFD and GWD. In this loop it will be demonstrated that the criteria are met and that the improvements have resulted in a trend reversal. In all these loops the national, regional and local authorities keep a close watch on the progress (surveillance and operational monitoring), verify the WFD/GWD objectives and decide on the approval of the management plan. The loops are shown in the ﬂowchart in Figure 7.1.5. The above approach has been used in the example cases that are mentioned in Table 7.1.2, which indicates the summary of the cases that were analysed according to the following questions.

Prevention and Reduction of Groundwater Pollution at Contaminated Megasites

419

1. What should be reported to the EU about megasites (contaminated land and plumes resulting from contaminated land) in the river basin management plan? Results of monitoring or only conclusions? Management/remediation measures? 2. Describe the situation (situation 1, 2, 3a–d). 3. Area deﬁnition: what are the criteria/rules that we can use to deﬁne the boundaries of the area (contaminated/plume)? What is the size of the contaminated area/plume in relation to the groundwater body? What is the location of the points of compliance (where and how deep)? What are the receptors (does groundwater have an intrinsic value and should it therefore be considered as receptor)? 4. What are the objectives for the quality of the plume? Standstill/trend reversal/threshold values/other compliance values (see guidance document on prevent and limit). Historical contamination: risk-based management. New contamination: close the tap! 5. What are appropriate measures, best available techniques or economically feasible? Source/path/receptor. Proportionality/feasibility. Cost eﬃciency. 6. How do we deﬁne our monitoring programme (do we have to deﬁne a monitoring programme)?

Spatial and temporal scale/points of compliance. Consequences. Relation monitoring with measures. Real or virtual (modelling) points of compliance.

As a result of these questions the outcomes vary considerably. The results of analyses were done from a range of real monitoring data to only modelled data to design a conceptual model. Diﬀerent receptors were identiﬁed: surface water, drinking water reserves, acquifers, urban areas and agricultural land. All these are exposed to diﬀerent types of pollutants in the diﬀerent cases. The size of the megasites are diﬀerent, ranging from 10–20 ha to hundreds of square kilometres. In those cases where no IMS is used no feasible action plan is available and the size and impact of the contamination is estimated quite large with no speciﬁcations for smaller areas (clusters) or with detailed focus to reverse trends. The megasites that have not been analysed with an IMS do not obtain the overview and one is hesitant to set priorities.

420

Chapter 7.1

The main conclusion is that for those megasites that were analysed with an IMS an action plan has already started to deal with the contamination according to the priorities that were set for clustered areas with a similar approach. In some cases threats of possible major pollution were identiﬁed. After the results of the application of this IMS some of the supposed threats were taken away and framed to a feasible size, which could be handled with more rigour. This was a real eye-opener for the competent authority.

7.1.4

Conclusions

This chapter presents approaches and diﬀerent project examples of groundwater management plans for megasites and urban areas that deal with historically contaminated sites in relation to the WFD and GWD. By categorising megasites in diﬀerent archetypes, a systematic approach can be followed to assess the risks to the groundwater system (as a receptor) and the measures to monitor and manage the risks. These measures depend on the current and future impact at points at diﬀerent distances within the groundwater system to the receptor(s), the so-called points of compliance. Active measures have to be taken in situations where the contamination forms an immediate risk for a receptor. In cases where the contamination does not form a direct risk for receptors, the situation is less critical and groundwater contamination can be controlled by long-term measures. It can be concluded that after analysing the diﬀerent megasite cases, the WFD and GWD give suﬃcient options to develop and implement a risk-based approach for megasites and to deal with historically contaminated groundwater in an adequate way.

References 1. Water, Environment, Landscape management of Contaminated Megasites (WELCOME) project, EVK1-CT-2001-00103 (www.euwelcome.nl/kims). 2. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy (Water Framework Directive). 3. Directive 2006/118/EC of the European Parliament and the Council in the ﬁeld of the protection of groundwater against pollution and deterioration (Groundwater Directive).

CHAPTER 7.2

Forecasting Natural Attenuation as a Risk-based Groundwater Remediation Strategy RYAN D. WILSON, STEVEN F. THORNTON AND DAVID N. LERNER Groundwater Protection and Restoration Group, University of Sheﬃeld, Kroto Research Institute, Broad Lane, Sheﬃeld S3 7HQ, UK

7.2.1

The Nature of Groundwater Pollution from Point Sources

7.2.1.1

Sources and Plumes

One of the major environmental problems caused by contaminated land is the pollution of the underlying groundwater. Groundwater can carry the pollutants oﬀsite and so present a hazard to the wider environment and human health wherever the groundwater travels. Understanding how such pollution plumes develop, having tools to estimate the associated risks and how the risk can be reduced are clearly important for the management of contaminated sites. Groundwater spreads pollutants that are dissolved or suspended in it. Sometimes the pollutants are already dissolved in the water that enters the subsurface, for example a leak from a sewer or wastewater lagoon. A more common and serious source is water ﬂowing through a contaminated zone and dissolving the pollutants, to form a plume in the groundwater, downgradient of the source. These zones often contain non-aqueous phase liquids (NAPLs). Sources of this type are serious problems because they are long-term point sources of groundwater pollution and are very common. The contaminant source zone presents a local hazard, whereas the resulting plume can be a hazard to both human health and groundwater resources a signiﬁcant distance downgradient from the source. Example point sources of groundwater pollution associated with contaminated land include industrial premises (petrochemical manufacturing facilities, gasworks, coal tar distillation 421

422

Chapter 7.2

and coking plants, oil reﬁneries, petrol stations), waste disposal (landﬁlls and storage lagoons) and chemical spillages in transport (storage and distribution depots). The pollutants released include organic chemicals, inorganic compounds, metals and colloids. They may be present singly or as mixtures, depending on the composition of the contaminant source, the groundwater chemistry, aquifer setting and biogeochemical processes.

7.2.1.2

Risk Assessment: Is Clean-up Required?

Where contamination of groundwater has occurred, government agencies and site owners must decide whether action is needed to remediate (clean up) the problem. However, remediation of contaminated land and groundwater is often technically diﬃcult and expensive. Any action should be carefully designed and targeted to make the wisest use of the planet’s resources, including the limited resources of the government and site owner. The basic decisions to be made are: (a) is any action required at all, (b) what standard of remediation is required and (c) should the source zone or the plume be targeted? Remedial action may be required by statute, as a requirement of insurance policies or by a prospective purchaser of the land. These undiscriminating requirements may lead to expensive remediation for sites where there is in fact no signiﬁcant threat to health or the environment. A more rational approach is to estimate the risks that the contaminated sites pose, and then assess if these risks are signiﬁcant enough to justify remedial action. Such risk-based decisionmaking is becoming more accepted and widespread as a philosophy for several reasons. It is a rational approach, which operates in a transparent way and provides comparability between sites. This leads to cost eﬀectiveness, as resources are targeted on those sites which present the highest risks, not just those with high public proﬁle; it also makes it harder for vested interests to resist requests to carry out remedial work. Risk assessment usually adopts the source–pathway–target conceptual model to identify pollutant linkages. This requires answers to the following questions. Is there a source of pollution, and what are its characteristics and signiﬁcance? In other words, is there a hazard? Is there a pathway which could conduct the pollutants to a receptor, and how will concentrations change as the pollutants move along the pathway? Are there targets (or receptors) which are vulnerable to these pollutants, and what is the dose–eﬀect relationship? Quantifying the answers to the risk assessment questions will enable sensible decisions to be made about whether to clean up individual sites. Understanding the linkages between source and target will help to decide where action should be targeted: is it better to remediate the source, block the pathway or protect the receptor? Some of the options for achieving remediation are discussed in the next section.

Forecasting Natural Attenuation

423

Of particular interest for this chapter is the question of how pollutant concentrations change along the pathway between source and target. Clearly, if the natural processes occurring in the pollution plume are reducing concentrations (or toxicity), then risk will be reduced with time or travel distance. If such natural attenuation can be reliably described and quantiﬁed, it can be included in the assessment of risk from the site. This is likely to reduce the need for engineered remediation: in some cases no action beyond monitoring will be required. However, natural attenuation processes are varied and can be complex. Good scientiﬁc understanding and site investigation are needed to reliably quantify them. The bulk of this chapter explains the science of natural attenuation, how it can be quantiﬁed and then presents a new screening tool which takes advantage of recent theoretical developments in the understanding and modelling of natural attenuation.

7.2.1.3

Options for Remediation

If remediation is required, then remediation of the source is almost always required. The source zone is a long-term reservoir of pollutants which will be continually released into groundwater. Numerous studies have shown that NAPL sources will have lifetimes of decades and centuries. If they are not removed, the plume will continue to exist for as long as the source does, and remediation of the plume will become an unsustainably long-term commitment. The options for source remediation are broadly containment and removal. Containment prevents water ﬂowing through the source and transporting the pollutants away, and may take the form of a recharge excluder for sources above the water table, barrier walls of various types, encapsulation or immobilisation with for example cement additives, or hydraulic containment with pumping wells. There is a variety of removal technologies, such as hydraulic or chemical ﬂushing, biodegradation, heating and vapour or ﬂuid extraction. The choice of appropriate technology for each site is a specialist design problem, depending on the geology of the site and nature of the pollutants, and is not dealt with in this book; however, specialist texts are available.1,2 There are a variety of technical approaches to remediating pollution plumes in groundwater. The simplest is pump-and-treat, in which polluted groundwater is removed by pumping; this is a barrier or containment method. In situ chemical treatment and bio-remediation require chemicals (and sometimes bacteria) to be injected and mixed into the plume; mixing in groundwater is notoriously slow, as discussed below. Permeable reactive barriers (PRBs) have been very successful for certain types of pollutants. In a PRB, some kind of reactive ﬁlter is introduced into the path of the plume in the subsurface. As the plume ﬂows through it, usually with natural groundwater ﬂow, the pollutants are destroyed or removed by the ﬁlter material. Finally, natural attenuation is accepted as a remedial strategy in many jurisdictions, provided that it can be reliably quantiﬁed and shown to reduce risk to acceptable levels within an acceptable time scale. The duration of any plume remediation will be partly determined by the nature and longevity of the source zone.

424

Chapter 7.2

7.2.2

Natural Attenuation of Contaminants

7.2.2.1

Why Natural Attenuation is Important

Dissolved contaminants in groundwater are a problem because they may (1) pose a risk to people due to human toxicity, (2) compromise the quality of surface and groundwater resources, (3) cause aesthetic degradation of drinking water, (4) persist for many decades, (5) arise from sources that are very diﬃcult to remove and (6) are costly and disruptive to treat either at the point of use or in situ. Natural attenuation (NA) represents the combination of naturally occurring physical, chemical and biological processes which occur in the subsurface to reduce the concentration, mass or toxicity of contaminants, ultimately slowing the rate at which dissolved contaminants migrate in groundwater and potentially contributing to a reduction in the exposure risk to receptors. The fact these processes occur is important because they can form all or part of a strategy to manage contaminant plumes in groundwater. Where NA processes occur at an acceptable rate, they may be used as a sole management approach known as monitored natural attenuation (MNA). Where the rate is not entirely protective of risk, MNA may still serve as a polishing step for more aggressive in situ treatment approaches. Thus it is of considerable importance to be able to make robust predictions on the magnitude of individual NA processes. It has been shown that the cost to apply MNA as a sole plume management strategy is far less than installing and maintaining engineered in situ remedial solutions. It therefore follows that credible NA assessments made early on in site investigation (using tools requiring a minimum of input data) both helps make informed management decisions and costeﬀective allocation of resources.

7.2.2.2

Contributing Processes

Natural attenuation produces a reduction in contaminant concentration along a ﬂowpath as a result of volatilisation, dilution, sorption, dispersion and biodegradation (see Figure 7.2.1 for summaries of these processes). The contribution these ﬁve processes make to natural attenuation depends on the nature of the contaminants, the characteristics of the aquifer and the mode of contaminant release. Weakly attenuating contaminant species typically have low volatility, low sorption and low biodegradation potential. One example is the petroleum additive MTBE, which is attenuated primarily by dispersive mixing with uncontaminated water and dilution at receptors. A converse example is benzene, another component of fuels, which is relatively easily biodegraded. Volatilisation (partitioning of contaminants from the water phase to the gas phase) will make a more signiﬁcant contribution to the attenuation of highly volatile contaminants (larger Henry’s constant) relative to low-volatility contaminants (Figure 7.2.2). Thus, it could be expected that volatilisation will be a more signiﬁcant NA process for benzene (Hcc ¼ 0.24) than MTBE

425

Forecasting Natural Attenuation Volatilisation: partitioning of a contaminant from a separate organic liquid phase to soil gas. Dilution: mixing of plume with uncontaminated water at e.g a river discharge or abstraction well Retardation: reduction incontaminant velocity due to partitioning to aquifer matrix Dispersion: mechanical and diffusive mixing of contaminant and background water during transport Biodegradation: transformation of contaminants by indigenous microbes

Figure 7.2.1

Natural attenuation processes.

Contaminants partition from the dissolved phase to air (soil gas) according to Henry’s Law: Hcc =

Ca

Cw

where Ca is air concentration and Cw is water concentration (both in mass/volume, e.g. mg/L). In this form,H is dimensionless – beware of units in other forms. A more volatile contaminant will have a larger Hcc compared to a less volatile contaminant.

Figure 7.2.2

Volatilisation: air–water partitioning.

(Hcc ¼ 0.023), even though both may be found in the same plume dissolving from an unleaded petroleum fuel hydrocarbon spill. Also, signiﬁcant volatilisation can only occur at or above the water table, and therefore this process will contribute more to the attenuation of plumes emanating from spilled petroleum fuel hydrocarbons (light NAPLs that ﬂoat on the water table) than those emanating from a chlorinated solvent spill (a dense NAPL distributed to depth), even though both spills may have constituents with similar Henry’s constants. Dilution occurs where a plume intersects a receptor (water supply well, river, etc.). Since the plume represents only a fraction of the groundwater captured by the receptor, plume contaminant concentrations are diluted by contaminantfree water. The magnitude of dilution depends on the proportion of plume to clean water ﬂux into the receptor. Thus, for a given receptor, plumes with a small cross-sectional area of ﬂow will be diluted more than larger plumes.3 However, dilution in a receptor, such as a public supply well, is not to be considered as NA, although it can be taken into account in risk assessment.4

426

Chapter 7.2 Contaminants partition from the dissolved phase to the solid phase (i.e. aquifer media) according to a distribution coefficient (Kd):

Kd =

Cs Cw

where Cs is the concentration of contaminant on the solids. There are published Kd values and a number of semi-empirical methods to estimate it for most common contaminants. Solid-water partitioning slows contaminant transport velocity by a retardation factor: R=

w

c = 1 +

b

Kd

where νc and νw are contaminant and water velocity, respectively, ρb is aquifer bulk density and θ is aquifer porosity.

Figure 7.2.3

Sorption: solid–water partitioning.

Sorption attenuates contaminants by reducing dissolved concentrations and slowing (retarding) their transport velocity through the aquifer (Figure 7.2.3). Contaminants that readily partition to organic carbon or ion exchange sites on geological materials will migrate more slowly than those that weakly partition. For most of the common organic contaminants, the magnitude of sorption does not vary with contaminant concentration, meaning that retardation can be estimated using a simple partitioning relationship. If local contaminant concentrations are constant in space (i.e. the plume is at steady state, neither growing nor shrinking), then retardation does not contribute to overall contaminant attenuation because solid–water partitioning is in equilibrium. Only when concentrations are not in equilibrium (i.e. during plume growth or decay) does retardation contribute to attenuation. Dispersion is the process responsible for mixing waters during transport. Dispersion has two additive components: advection and diﬀusion. The magnitude of the advective component of mixing is deﬁned by groundwater velocity and an aquifer characteristic called dispersivity. Aquifers with a wide range of pore sizes (greater heterogeneity) typically have greater dispersivity. The diﬀusive component is driven by the chemical gradient between higher and lower concentration areas of the plume. In most aquifers groundwater velocity is high enough that diﬀusion is the lesser component of dispersion. Biodegradation by indigenous microorganisms is the only NA process that results in the destruction of contaminants, either via complete mineralisation (conversion to carbon dioxide and water) or transformation to another species. It is the most important process for the NA of organic pollutants. Because of this, NA assessments focus on collecting data to establish the rate of biodegradation. While biodegradation is often a complex process involving

Forecasting Natural Attenuation

427

interacting microbial consortia operating over a range of geochemical conditions, at the early stages of NA assessment it can be approximated by pseudo ﬁrst-order decay. In the case of contamination by hydrocarbons or their derivatives, decay is often rapid enough to be treated as eﬀectively instantaneous (since groundwater velocities are typically small). This allows the complex steps involved in the biodegradation reactions to be treated and interpreted in a relatively simple way.

7.2.2.3

Plume Development and Life Phases

If a suﬃcient mass of contaminant is released into the subsurface, some of that mass may be transported to the water table. The result is a groundwater contaminant plume. The manner in which the plume forms depends on the nature of the contaminant release. Some contaminants (e.g. ammonium and nitrate) are released as aqueous solutions, while others (petroleum, chlorinated solvents, coal tar) enter the subsurface as NAPLs. Aqueous contaminants can be released below the water table or be ﬂushed down to the water table by recharging water. Constituents dissolving from a NAPL located above the water table may also be ﬂushed down to the water table to form a plume, or the NAPL itself may reach the water table and then generate a plume. In either case, a plume of contamination will emanate from what is called a localised or point source area. Non-point sources of contamination are those that originate from more diﬀuse sources (e.g. from agricultural activity). Natural attenuation assessment of plumes derived from such sources can be performed, but because the source area may be many hectares in size and spatially complex, they must be treated at a very large scale. The various assumptions that must be made to simplify the source means any estimate of plume length will have a high degree of uncertainty.

7.2.2.3.1

Growth Phase

Immediately after a contaminant source is introduced to the subsurface, a plume will begin to form in groundwater and be transported downgradient if contaminant concentrations are stoichiometrically greater than the available electron acceptors (Figure 7.2.4). The extent to which a plume grows thus depends on concentrations, groundwater velocity, sorption and biodegradation: by estimating or measuring these parameters, it is possible to estimate the time required for a given plume to reach the end of the growth phase or attain its maximum plume length.

7.2.2.3.2

Steady-state Phase

Once the contaminant mass ﬂux (cross-sectional plume area groundwater ﬂux contaminant concentration) is oﬀset by the combined rate of mass attenuated through natural processes, the plume has reached steady state. As long as contaminant ﬂux from the source area continues and the rate of

428

Chapter 7.2

a

>

Σ

plume mass flux attenuation

source

plume

growth

b

=

Σ

steady state flow

c

<

Σ

decay

d

<

Σ

or

decay

Figure 7.2.4

Plume life phases deﬁned by the balance between plume mass ﬂux and plume attenuation due to natural processes.

attenuation does not change, the plume will remain at steady state. In the case of a dissolved source, it is rare that the associated plume will reach a steadystate condition because the source changes with time. NAPL sources, on the other hand, may last for decades or centuries due to the low eﬀective solubility of the constituents dissolving into the plume. Thus, the steady-state phase of plumes emanating from NAPL sources may be of similar duration. The plume length under this condition will be a function of the contaminant loading in the plume and availability of electron acceptors in the aquifer to support degradation. Hence, it is expected that the steady-state plume length will be diﬀerent for the same group of contaminants in diﬀerent aquifers. Predicting this steadystate plume length and the time scale to achieve it is a key objective of NA assessments.

7.2.2.3.3

Decay Phase

When contaminant mass ﬂux from the source becomes smaller than the rate of attenuation (the source is exhausted or the rate of attenuation increases),

Forecasting Natural Attenuation

429

the plume will shrink or decay. If steady state is disturbed because the attenuation rate has increased, the plume may shrink until it achieves a new steady-state length. Conversely, if the attenuation rate decreases for some reason (e.g. renewed release of contaminant mass from the source or accumulation of microbial by-products which inhibit degradation), the plume may grow until it again reaches a new steady state. If the source is exhausted, the plume will decay until it disappears. Note that the plume may appear to shrink in length for a period of time, but since attenuation occurs everywhere in the plume (at various rates), the plume will in fact eﬀectively ‘‘dissolve’’ away.

7.2.2.4

Biodegradation in Detail

Fortunately many, if not most, of the contaminants found in groundwater are amenable to microbially mediated degradation. There are two fundamental types of reaction that involve biodegradation of contaminants: oxidation and reduction. Contaminants that are oxidised lose electrons, while those that are reduced gain electrons. The compound losing the electron is termed an electron donor, whereas the compound gaining the electron is termed the electron acceptor. The complete degradation reaction is created by the combination of two redox half reactions, where electrons are transferred from the electron donor to the electron acceptor. This electron transfer is regulated by microorganisms, which gain energy for growth by facilitating the process. These fundamental reaction types should not be confused with oxidising and reducing geochemical conditions, commonly expressed as redox (Eh), or oxidising– reducing potential (ORP). This is a measure of the tendency for the system to transfer electrons. As electron acceptors are consumed, the groundwater becomes more reducing and the redox potential of the groundwater decreases.

7.2.2.4.1

Oxidation of Contaminants

The vast majority of petroleum hydrocarbons (petroleum, diesel, fuel oil, coal tar, etc.) consist of compounds that contain carbon in a reduced state. Microorganisms that degrade these compounds do so by facilitating the exchange of electrons from the compound (e.g. organic contaminant) to an electron acceptor, which is itself reduced in the reaction. Note that the oxidation of these compounds can occur under oxidising or reducing geochemical conditions, which is often a source of some confusion. For example, benzene can be biodegraded by organisms using oxygen as an electron acceptor (under aerobic or oxidising geochemical conditions). The relevant half reactions are: O2 + 4H1 + 4e - 2H2O C6H6 + 12H2O - 6CO2 + 30H1 + 30e

430

Chapter 7.2

Balancing and adding these half reactions yields the coupled reaction: C6H6 + 7.5O2 - 6CO2 + 3H2O So, 7.5 mol of oxygen are needed to mineralise 1 mol of benzene. This can be used to estimate the mass of oxygen (molecular weight 32 g mol1) needed to convert a given mass of benzene (78 g mol1). An amount of 240 g of oxygen (¼ 7.5 mol 32 g mol1) is required to degrade 78 g of benzene (¼ 1 mol 78 g mol1), or 3.1 g oxygen is required per gram of benzene (¼ 240/78). Benzene can also be oxidised by organisms that use sulfate as an electron acceptor (under moderately to strongly reducing geochemical conditions). The sulfate reduction half reaction is: 8e + 10H1 + SO42– - H2S + 4H2O Balancing and adding this to the benzene half reaction above yields the coupled reaction: 7.5H1 + 3.75SO42 + C6H6 - 6CO2 + 3.75H2S + 3H2O In both cases, from the half reactions it can be seen that benzene loses electrons (is oxidised) while the electron acceptor gains electrons (is reduced). It is important to note that the geochemical conditions under which each reaction occurs are mutually exclusive: aerobic reactions cannot occur under sulfate reducing conditions and sulfate reduction cannot occur under aerobic conditions.

7.2.2.4.2

Reduction of Contaminants

Halogenated aliphatic compounds such as trichloroethylene (TCE), perchloroethylene or tetrachloroethene (PCE), 1,1,1-trichloroethane (1,1,1-TCA) and dibromoethene (DBE) are common contaminants in groundwater and are examples of compounds where carbon exists in an oxidised state. These compounds may be degraded by microorganisms via reductive dehalogenation, which is the transfer of electrons from a donor (typically molecular hydrogen) to the contaminant (functioning as the electron acceptor). The half reactions for the reductive dechlorination of TCE to cis-1,2-dichloroethylene (cis-1,2DCE) are: H2 - 2H1 + 2e and C2HCl3 + 2H1 + 2 - C2H2Cl2 + H1 + Cl yielding the coupled reaction: C2HCl3 + H2 - C2H2Cl2 + H1 + Cl Thus TCE gains electrons (is reduced) while hydrogen loses electrons (is oxidised). The organisms that perform these reductive dehalogenation

431

Forecasting Natural Attenuation

reactions require very strongly reducing conditions. Thus, the greatest rate of contaminant dehalogenation will occur near the core of the plume where the necessary anaerobic conditions exist.

7.2.2.4.3

Biodegradation Redox Regimes

When contaminants are oxidised, biodegradation reactions occur in a speciﬁc order deﬁned by thermodynamics. More energy is available to organisms that degrade one mole of benzene aerobically (DGr ¼ 765 kcal) compared to the energy obtained via sulfate reduction (DGr ¼ 123 kcal). In fact, the electron acceptors are preferentially utilised in decreasing order of the free energy yielded by the reaction: oxygen, nitrate, manganese(IV), iron(III), sulfate and carbon dioxide. The result of this sequential degradation is a distinct zonation of redox conditions and therefore electron acceptor distributions in contaminant plumes, grading from carbon dioxide reducing (commonly referred to as methanogenic) in the plume core to aerobic at the outer fringe or ‘‘corona’’ of the plume (Figure 7.2.2). Because of the higher energy yield, aerobic reactions tend to be faster than methanogenesis, giving rise to a state where slow reactions are conﬁned to the core while faster reactions are limited to the outer plume fringe. In most cases the plume travels at a slower rate than the reaction products (due to sorption of the former), meaning that once oxygen and nitrate are depleted at the leading plume edge, the following contaminant mass travels in an electron acceptor-depleted halo (Figure 7.2.5). Several studies have shown that the fringe removes more contaminant mass than the core. The only way degradation reactions can be maintained at the fringe is via mixing of outward

NAPL source

methanogenic Fe(III) reducing O2 reducing

flow SO4 reducing NO3 reducing anaerobic halo

Figure 7.2.5

Distribution of redox zones in a typical naturally biodegrading plume of dissolved oxidisable constituents.

432

Chapter 7.2

dispersing plume constituents with inward dispersing electron acceptors in the uncontaminated groundwater. In many cases therefore, the overall plume attenuation is controlled by this rate of mixing between the plume and uncontaminated groundwater, which is inﬂuenced by horizontal and vertical transverse dispersion.

7.2.2.5

Hydrogeology and Heterogeneity

The migration of dissolved compounds (i.e. contaminant plumes) through water-saturated sediments or rock is described by the advection–dispersion equation: 2 @C @C @ C ¼ n þD þRþl @t @x @x2 where C is concentration, t is time, n is transport velocity, x is distance in transport direction, D is dispersion, R is retardation and l is, in this context, biodegradation rate. The velocity term represents the advective component, while the dispersion term represents mechanical and diﬀusive mixing. Dispersion is strongest in the direction of travel, and much weaker in the transverse horizontal and transverse vertical directions. The result is that often contaminant plumes are only marginally wider and thicker than their source, but often of considerable length: sometimes referred to as ‘‘cigar shaped.’’ As can be seen, three of the ﬁve principal components of natural attenuation are embedded in the equation: dispersion, sorption and degradation. It should be noted that retardation is only evident in the growth and decay phases of plume life; it is in dynamic equilibrium during the steady-state phase and thus eﬀectively inactive. As previously noted, biodegradation is the only process that results in the transformation or destruction of contaminants. The groundwater transport velocity (and thus rate at which contaminants migrate away from the source) depends on (1) the hydraulic gradient driving ﬂow and (2) the ability of the porous media to conduct ﬂow. The former is a function of local or regional scale ﬂow regime, while the latter is described in terms of the permeability (k) or hydraulic conductivity (K) of the porous media. The magnitude of hydraulic conductivity is in part related to grain size (porous media) or fracture aperture (fractured rock media). Since the geological processes that result in the accumulation of sedimentary sequences are spatially variable, it follows that hydraulic conductivity is also spatially variable. Whether deposited by water or wind, sediments invariably exhibit layering, cross bedding, grain size grading and imbrication. For example, coarse-grain sediments (sands and gravels) overlie or underlie ﬁner grained media (silts and clays) in alluvial systems, and layering is often of a spatially discontinuous nature. Attempts have been made to quantify this complex heterogeneity using statistical methods, but most often it is simply accepted that geological systems are inherently variable to some degree. This assumption implies that the ﬂow

433

Forecasting Natural Attenuation

depth above datum (cm)

ug/L 60 40 20 0

0

100

200

distance across ML transect (cm)

Figure 7.2.6

300

3000 2626 2251 1877 1502 1128 754 379 5

Spatial complexity across a benzene plume emanating from a petroleum NAPL source migrating 15 m upgradient.

ﬁeld conducting contaminants may be similarly complex. Macro-scale sedimentary (buried stream channels, soft sediment faulting and discrete lenses of high- or low-K media) or rock (intense fracturing, major faulting) features impart a contribution to plume spatial complexity in addition to the more subtle eﬀects of the meso-scale features mentioned above. Additional heterogeneity may arise due to the nature of organic contaminant source areas. Often, the source of a dissolved plume is a NAPL that has invaded the subsurface as a result of leaks or spills during storage, shipping and/or handling. NAPLs (petroleum, chlorinated solvents, coal tars, etc.) are immiscible in water and move though sediments according to a diﬀerent set of relationships. The result is most often a spatially complex distribution of blobs, stringers and pools of NAPL from which species dissolve to form a plume. Since the source is distributed in a discrete fashion, the plume emanating from it is likewise complex. Furthermore, due to weak lateral dispersion, this spatial complexity may persist for many tens of metres (Figure 7.2.6). Thus, it is not surprising to ﬁnd that most plumes do not always describe a simple ‘‘cigar’’ shape.

7.2.3

Current Natural Attenuation Assessment Practices

Site characterisation for the performance assessment of MNA is fundamentally diﬀerent and more comprehensive than that required for an active remediation scheme, because it requires greater understanding of processes aﬀecting the contaminant plume and there is greater emphasis on data collected from within the plume. An initial monitoring phase is required to identify the location and extent of contaminant source area(s), the spatial distribution and concentration of contaminants, heterogeneity in the aquifer geological and hydrogeological characteristics and variations in the groundwater hydrochemistry. Figure 7.2.7 shows the eﬀect of various forms of heterogeneity in aquifer properties, hydrogeology and contaminant source distribution which must be considered in the design of monitoring networks used for MNA assessment. These aspects can all aﬀect interpretation of ﬁeld data used in the analysis of MNA. The

434

Chapter 7.2

a

b

C d

flow direction

c

d

Figure 7.2.7

C

flow direction

e flow direction

d

flow direction

d

C

flow direction

C

d

C time 2 time 1

d

The eﬀects of various forms of heterogeneities on assessment of natural attenuation by conventional protocols.

objective of the initial site characterisation for MNA assessment is to deﬁne the baseline conditions and establish the conceptual site model (see below). This is necessary to ascertain whether MNA is likely to be a viable remediation option and, if so, to provide a reference state from which its performance can be monitored over time. A typical network of monitoring wells used to evaluate MNA for contaminant plumes will include wells that assess uncontaminated background groundwater quality upstream of the plume, the contaminant composition in the source area and groundwater quality along the plume ﬂow path. Monitoring wells positioned in the ﬂow path ahead of the plume are required to correctly to deﬁne the downgradient extent of contamination and transverse to the plume to deﬁne the lateral extent of contaminant migration. The monitoring well network is usually installed in phases since the extent of plume migration is unknown until the wells have been drilled. Current NA assessment often focuses on assembling converging and supporting lines of evidence to establish the presence and long-term viability of biodegradation by indigenous microorganisms. In current technical protocols or guidance documents (e.g. Refs. 4–6) developed for the assessment of MNA three lines of evidence are typically sought. These are: (1) historical data

Forecasting Natural Attenuation

435

indicating spatial and temporal trends in the plume development, that is, establishing whether the plume has either reached steady state, is growing or is shrinking; (2) geochemical data demonstrating the occurrence of contaminant biodegradation processes in situ, by identifying decreases in contaminant concentrations which are correlated with a decrease in electron acceptor or donor concentrations, increase in metabolic by-product concentrations and increase in daughter product concentrations; and (3) microbiological data demonstrating the presence and activity of microorganisms capable of the required degradation process.

7.2.3.1

Establishing Whether Natural Attenuation is Occurring

The data required to demonstrate the occurrence of attenuation processes using the lines of evidence concept is largely based on the collection and analysis of groundwater samples from the contaminated site. Additional data on aquifer geochemical and hydrogeological properties are also required from appropriate site investigation to support quantitative prediction of MNA performance using various data reduction techniques (see next section). Groundwater samples are collected from the contaminant plume and uncontaminated aquifer and are analysed for a range of chemical species (or indicator parameters) which (1) document the type of degradation process occurring, (2) allow estimates to be obtained of contaminant mass loss contributed by each degradation process and (3) enable predictions to be made of MNA performance, plume length and long-term behaviour. The indicator parameters measured in a typical MNA assessment for groundwater will include the dissolved contaminant species of interest, their known organic metabolites or breakdown products (daughter compounds) and a combination of dissolved O2, NO3, NO2, SO4, HS, Mn21, Fe21, CH4, H2, CO2, alkalinity and total dissolved inorganic carbon (TDIC). In addition, pH, Eh (redox potential), temperature and electrical conductivity are also frequently measured at the time of sampling, to provide a more complete understanding of the groundwater chemistry and to provide input data for reactive transport modelling, if required. Further information on the parameters which are required for MNA assessment of speciﬁc contaminant matrices is provided in relevant technical guidance documents (e.g. Refs. 4–7). Primary and secondary lines of evidence are normally obtained from the distribution of dissolved reactants in groundwater along the plume ﬂow path, as determined by sampling groundwater monitoring boreholes located in the plume source area, uncontaminated and contaminated sections of the aquifer. Demonstrating MNA using primary lines of evidence requires historical site data that show that a contaminant plume is shrinking, stable or growing at a rate slower than that predicted by conservative groundwater ﬂow calculations.6 This line of evidence may demonstrate that a plume is being attenuated (e.g. by dilution and sorption), but it may not indicate contaminant mass loss through biodegradation. The latter can only be demonstrated from hydrochemical data by the secondary lines of evidence. Tertiary lines of evidence are used to more rigorously interpret data obtained from secondary lines of evidence,

436

Chapter 7.2

particularly if the primary and secondary lines of evidence are inconclusive in demonstrating that MNA is occurring at a suﬃcient rate.5,6 Tertiary lines of evidence include data from microcosm studies which may be used to support an MNA assessment, particularly where information is required on a speciﬁc degradation mechanism, site-speciﬁc environmental factors (e.g. contaminant toxicity) which may limit biodegradation or if contaminant degradation processes are diﬃcult to deduce from the background groundwater chemistry (e.g. degrading plumes in naturally anaerobic aquifers). Evaluating the eﬀectiveness of MNA is an ongoing commitment and a longterm monitoring strategy is required to ensure that initial predictions of its performance meet the site-speciﬁc cleanup objectives.8 Because the MNA assessment of contaminant plumes largely relies on groundwater quality data and understanding the spatial and temporal distribution of indicator parameters, which deﬁne the relevant processes and their eﬃcacy, it is vital that due consideration is given to good practice in the groundwater sampling eﬀorts and the design of an appropriate monitoring well network used in the analysis. Procedures must be used which ensure that the groundwater samples collected are representative of the formation being sampled and that samples are not biased by design of monitor networks, sampling methods or post-sampling handling.9,10 Traditional MNA assessment protocols call for a series of monitoring wells (at least 4–6) to be installed along a ‘‘representative ﬂow path,’’ ideally the plume centreline. However, in many cases aquifer heterogeneity results in spatially variable source distribution and/or contaminant transport along preferred (or spatially discrete) ﬂux paths, resulting in plumes that lack a unique centreline.11 Using data from wells that do not represent a centreline can lead to over- or under-prediction of NA rates (Figure 7.2.7), and in any case a high degree of uncertainty in the delineation of the plume geometry, structure and temporal behaviour. Moreover, monitoring well design can have a signiﬁcant inﬂuence on the result of an MNA assessment. Traditionally, monitoring wells ﬁtted with single screens several metres long are used in site investigations for contaminated groundwater, but this design is inappropriate for the collection of data required in an MNA assessment for groundwater and can result in the erroneous interpretation of plume behaviour. Wilson et al.11 compared diﬀerent methods used to estimate the degradation rate for a 50year-old plume of phenols in a sandstone aquifer instrumented with single screen monitoring wells and multilevel samplers (MLS). They found that a ﬁrstorder plume-scale degradation rate estimated using conventional MNA analysis techniques and the single screen monitoring wells was two orders of magnitude higher than that obtained using data from the MLS and the electron balance methodology.12 The former estimate implied that the plume would attenuate over a much shorter time scale than 50 years, whereas the latter analysis implied plume expansion could be expected. The analysis highlights the emphasis that is placed in MNA assessments on deducing the spatial distribution of degradation processes over scales much less than a metre, in order to reliably estimate the parameters needed to make realistic predictions of plume length. This important information can only be obtained by using a

Forecasting Natural Attenuation

437

groundwater monitoring well network which includes MLS designs in the MNA assessment. This is the approach invoked in the Corona concept of plume interpretation and prediction.13 It is also the basis of the quantitative analysis of plume behaviour and MNA performance developed within the CoronaScreen screening model for groundwater MNA assessment.14

7.2.3.2

Establishing Whether Natural Attenuation is Suﬃcient

A range of data reduction and interpretation techniques are available to evaluate the performance of MNA, using both qualitative and quantitative analysis of the site data. At a simple level, solute concentrations in groundwater samples from the monitoring network can be contoured to produce a two-dimensional map of contaminants and indicator species in groundwater. These plots can be completed in plan or section view along the plume ﬂow path. This provides a visual analysis of the plume size, shape and spatial distribution of associated degradation processes in the aquifer. When combined with repeat surveys of the same monitoring wells, this analysis can be used to identify the temporal variation in contaminant distribution to establish the plume status (e.g. expanding, steady state, shrinking) and relative importance of degradation processes as the plume source term evolves. This understanding is helpful in completing mass balances for plumes to estimate degradation rates.12 However, the resolution of solute distributions and interpretation of plume behaviour achieved with this analysis are determined largely by the quality of the monitoring well network and plume migration over time. It can be diﬃcult to deduce detail (e.g. preferential ﬂow paths and heterogeneity) in the contaminant distribution or conﬁrm the plume status using a limited number of monitoring wells ﬁtted with single screens and for slow-moving plumes where biodegradation is slow.11 Additional visual tests include plotting concentrations of contaminants and indicator parameters or ratios of these species as a function of distance along the plume ﬂow path or time for diﬀerent monitoring wells.4 This will provide a qualitative assessment of the extent of contaminant attenuation in general and long-term behaviour, correlated with either the spatial distribution of relevant degradation processes or preferential attenuation of contaminants within the plume. Most of these techniques are used to conﬁrm attenuation with the primary line of evidence. The assessment of NA using the second line of evidence requires the estimation of contaminant degradation rates, by quantitative analysis of the spatial and temporal variation in either the contaminants or indicator species. When contaminant concentration data are used, degradation rates are commonly estimated with methods which involve graphical regression or the correction of measured concentrations to assumed tracer species in the plume.6 The analysis is completed as a function of distance along the ﬂow path or time for individual monitoring wells and the eﬀect of sorption and aquifer hydrogeological properties on attenuation is also considered. Statistical analysis of contaminant concentration data used in this way can be undertaken to evaluate plume behaviour, but this usually requires extensive time series data to be statistically signiﬁcant.6,8

438

Chapter 7.2

When geochemical indicator species are used in the analysis, estimates of contaminant degradation can be obtained by calculating the aquifer biodegradation capacity.12 This allows the contribution of each degradation process in contaminant attenuation to be deduced and, by comparison with the maximum contaminant concentration in the plume, can be used to establish if the aquifer is able to assimilate the contaminant loading from the source area. This analysis must be repeated for successive groundwater quality surveys to identify the plume behaviour and performance of MNA over a realistic time frame (e.g. annual cycle and consecutive years), should the contaminant ﬂux from the source area change over this period. The assessment methods described above typically rely on the analysis of data from a limited number of monitoring wells located along a presumed centreline (representing the highest contaminant concentrations) for the plume. Valid interpretation of MNA in these circumstances requires that the plume centreline is known, is temporally and spatially invariant and can be instrumented accordingly to provide data of the required quality for the analysis.11 Figure 7.2.7 shows many typical situations where the implicit assumption of a plume centreline is not valid. Therefore, the application of such data analysis techniques (and assumptions therein) can only be justiﬁed by the conceptual site model, which itself is largely derived from data collected from such idealised monitoring networks. This circular argument clearly presents a problem with respect to the type and extent of monitoring installations that should be considered for MNA assessments at most sites (particularly those with unknown or complex source histories) and the appropriate analytical methods which can accommodate uncertainty in both the conceptual model and data inputs obtained from the site investigation and monitoring well network. This emphasises the need to adequately characterise the plume source during the initial site investigation and to design a downstream monitoring well network which is appropriate for the aquifer geological conditions and hydrogeological setting, to correctly understand the plume behaviour and reliably estimate mass loss. More sophisticated methods for the quantitative assessment of MNA performance, which take account of these fundamental problems, involve the calculation of mass balances for plumes using more elaborate monitoring well networks and improved conceptual basis of the attenuation processes involved. The analysis can consider either the contaminants or indicator species, or both, according to the available data. Two approaches are available. In the ﬁrst approach, multiple monitoring wells are arranged in a series of transects across the plume at several locations along the ﬂow path. This monitoring array is then used to estimate the mass ﬂux (Section 2.3.2) of contaminants passing through the transect or ‘‘control plane’’ at each location. Comparison of the mass ﬂux crossing successive control planes along the ﬂow path then allows attenuation to be estimated. Typically, vertical proﬁles of contaminant concentration obtained from MLS will be used in this analysis (e.g. Refs. 12 and 15), but it is also possible to apply the approach using contoured isopleths plots developed with single screen monitoring wells.6 An alternative approach, again using solute mass ﬂuxes, is the electron balance methodology.12 In this case the

Forecasting Natural Attenuation

439

plume is conceptualised as a ‘‘box model,’’ where various mass inputs (e.g. contaminants, electron acceptors, reaction products) arising from diﬀerent transport and degradation processes (e.g. advection, dispersion, in situ production) are calculated using the same site investigation and groundwater quality data required in the simple methods. These mass inputs are compared in a plume-scale mass balance which then allows important MNA performance indicators, such as predicted plume length, current and future plume status (should source inputs change), source term and degradation rate, to be estimated. This methodology is built into the CoronaScreen decision-support model, developed within the CORONA project. The signiﬁcant diﬀerence and innovation in this approach is that all monitoring wells are used in the analysis (accommodating spatial variability in solute concentrations within the plume) and interpretation is based on relevant transport and degradation processes which are combined to derive mass inputs to the plume. A key advance made with the CoronaScreen model is that whilst the underpinning conceptual and theoretical basis is relatively sophisticated, the data input requirements can be obtained with a much reduced monitoring network, comprised of a few MLS and strategically placed single screen monitoring wells. This allows for obvious cost savings on the monitoring and data collection eﬀort, without sacriﬁcing accuracy in the predicted outputs. The main improvement in accuracy and understanding of MNA performance gained from the use of mass ﬂux-based (rather than concentration-based) approaches in the diﬀerent methodologies described is that less bias or potential error is introduced into the analysis. The data inputs are based on an entire monitoring well network or a signiﬁcantly increased number of monitoring wells, heterogeneity in contaminant mass distribution is taken into account and the plume geometry is not assumed. The important result is that the performance assessment of MNA is more rigorous and uncertainty in the prediction of plume behaviour is reduced.

7.2.3.3 7.2.3.3.1

Steps in Natural Attenuation Assessment Conceptual Site Models

The ﬁrst phase of NA assessment generally involves an assessment of historic site data. A step that is often overlooked, but one of the ﬁrst tasks that should be started for a new site, is the development of a conceptual site model (CSM). The CSM is a site-speciﬁc representation of the contaminant distribution, hydrogeological conditions and NA processes, which contribute to the understanding of whether MNA is appropriate for a given site or contaminant release scenario. Greater initial eﬀorts in developing a technically defensible CSM will provide increased conﬁdence in the selection of MNA, with reduced long-term monitoring costs. By comparing the subject site to analogous settings (geology, contaminant, mode of contaminant release, etc.), the type, location within the plume and relative importance of attenuation mechanisms can be anticipated. This can

440

Chapter 7.2

inﬂuence where best to install monitoring wells and which monitoring devices to use. CSMs should be considered a living device that is constantly updated as new data is gathered. With respect to MNA assessment, there are a number of aspects of a CSM: geology, hydrogeology, geochemistry, microbiology, mode of contaminant release and relevant attenuation processes. Site records, maps and local knowledge are often suﬃcient to build a basic geological model. Reﬁnements to the geological component of the model will come from the analysis of sediment or rock core, trial pits or drill cuttings: this may be important if the site is comprised of highly heterogeneous sediments or fractured rock. The hydrogeological, geochemical and microbial components (water table maps, average water velocity) may be approximated from local or regional data, but site-speciﬁc data will almost certainly be required. An idealised source distribution model can be generated if the mode of contaminant release is known (i.e. release of a known volume of NAPL with certain properties from a known location at a known rate). However, given the complex nature of two-phase ﬂow, the exact distribution of NAPL in the subsurface can never be known with certainty. In the early phase of site investigation for MNA assessments, it is be desirable to make an initial appraisal of the magnitude of the problem. Three key questions are the following. How long will the plume take to reach and maintain a steady-state condition? What will the concentration of contaminant(s) be at certain points along the plume? How will contaminant concentrations at compliance points change over time? It is likely at this stage that insuﬃcient data will be available to use a complex numerical model to simulate the plume history. Groundwater velocity, ﬂow direction, contaminant concentrations and background electron acceptor concentrations may be known from a few locations, but other key parameters such as dispersivity and sorption may have to be estimated from the literature. A preliminary assessment using simpliﬁed or screening level models may be a more appropriate approach because these models require few input parameters and provide an approximate answer. Subsequent phases of site investigation result in the acquisition of data that help resolve the question of risk. These data can also be used to reﬁne the conceptual site model and improve parameter estimates for input to simpliﬁed (or complex) simulation tools.

7.2.3.3.2

Application of Screening Models

Screening models that incorporate analytical solutions of the mass transport equation have been used to explore various basic characteristics of a plume. These models assume biodegradation proceeds at a uniform pseudo ﬁrst-order

441

Forecasting Natural Attenuation

rate for all the electron donor/acceptor couples. Solutions are available for vertical and horizontal line, point and planar sources of ﬁnite, inﬁnite or decaying duration. As such, a wide range of source/plume scenarios can be approximated for a homogenous aquifer domain with uniform properties by an analytical solution. Input requirements are generally limited to groundwater velocity, source geometry, source concentration, electron acceptor concentrations and longitudinal and/or transverse dispersivity. These inputs could also be used in a numerical model, but simpliﬁed models are easier to operate and run quicker. Given similar levels of uncertainty, they may be a preferable means to explore the eﬀect of key factors on the length of a plume (e.g. sensitivity of plume length to changes in velocity, concentration or dispersivity).

7.2.3.3.3

Available Screening Models

There are three screening models currently available: BioScreen,16 NAS17 and BioChlor.18 The ﬁrst two are intended for use on oxidising plumes (e.g. aromatic hydrocarbons) while the latter is for reductive dechlorination (e.g. chlorinated solvents). All three models use a modiﬁed form of the Domenico19 analytical solution for a degrading contaminant plume transported from a vertical plane source. From a relatively short list of input parameters (those for BioScreen are compiled in Table 7.2.1; inputs for NAS are similar), a proﬁle of contaminant concentrations along the theoretical oxidising plume centreline is calculated. From these proﬁles, the maximum plume length can be obtained. Alternatively, plume length (assuming it is known) can be input and dispersivity values and/or decay rate can be estimated by curve ﬁtting. BioChlor is used in a similar manner, except individual ﬁrst-order decay rates and concentrations for parent and daughter products must be entered q (i.e. for PCE, TCE, cis-1,2-DCE and VC in the case of chlorinated ethenes). Table 7.2.1

Input parameters for BioScreen.

Parameter

Comments

Groundwater velocity

Or hydraulic gradient, hydraulic conductivity and porosity

Longitudinal dispersivity Transverse horizontal dispersivity Transverse vertical dispersivity Retardation factor First-order biodegradation constant Source area width and thickness Source contaminant concentration

Or bulk sediment density, fraction of organic carbon and water/solid partitioning coeﬃcient Or electron acceptor concentrations in the case of instantaneous reactions Combined concentrations of all hydrocarbons in plume

442

Chapter 7.2

7.2.3.3.4

Advantages of Existing Screening Models

Natural attenuation screening models have a number of advantages. They typically run very quickly, allowing for output determination and sensitivity analysis, in a matter of minutes. These models also require relatively few input parameters: mostly those that can be measured in the ﬁeld or estimated from ﬁeld data. Due to their simplicity, screening models do not require any special user expertise. They are usually based on spreadsheet applications, which make them relatively compact and easy to disseminate. These models are also costeﬀective; a number are currently available freely from websites (e.g. BioScreen and BioChlor can be obtained from the USGS website).

7.2.3.3.5

Disadvantages of Existing Screening Models

Parameters such as source concentrations, groundwater velocity and retardation can be easily measured or estimated from semi-empirical relationships. However, values of dispersivity are diﬃcult to measure, and establishing the source geometry requires a signiﬁcant site investigative eﬀort. As such, estimates of plume length, etc., are subject to a degree of uncertainty that may be unacceptable. Indeed, the BioScreen model has largely fallen out of favour primarily because it has been used without a clear understanding of the implications of inputting inaccurate estimates of key attenuation parameters. It is not necessarily a ﬂaw with the underpinning analytical solution, but rather the NA assessment philosophy embodied in the model. In general, the less data available and the more heterogeneous the site, the less accurate will be the representation by any model.

7.2.4

CORONA: A New Natural Attenuation Assessment Philosophy

The CORONA approach to NA assessment diﬀers from the traditional approaches in that it does not seek to obtain plume centreline concentration data or gather lines of evidence. Rather, CORONA seeks to quantify and rank the key NA processes inﬂuencing plume transport. A conceptual model of the site should always be built, starting with a comparison of possible plume analogues. For example, the maximum length of a plume degrading by oxidation processes will be controlled by mixing of electron acceptors at the plume fringe, which is controlled signiﬁcantly by vertical transverse dispersion. Monitoring well installation and data collection would be designed to obtain data to allow estimation of dispersivity, inward electron acceptor ﬂux gradients and outward contaminant ﬂux gradients.

7.2.4.1

Core and Fringe Controlled Plumes

By focusing on process identiﬁcation and quantiﬁcation, the CORONA approach involves a determination of where within the plume the majority of

443

Forecasting Natural Attenuation

flow

fringe processes: higher redox - aerobic ornitrate reducing oxidation

Figure 7.2.8

Concentration Depth

core processes: low redox - methanogenesis, fermentation or reductive dechlorination

Pollutant

O2 NO3

Schematic of a plume showing location of fringe and core processes.

contaminant biodegradation occurs. Generally, fringe processes control the attenuation of plumes consisting of contaminants that are oxidised (e.g. petroleum hydrocarbons), whereas processes occurring in the core control the attenuation of plumes comprised of reducible contaminants (Figure 7.2.8). It is important to note that the reactive plume fringe will become increasingly thicker along the plume ﬂowpath. In some cases, both plume core and fringe processes may contribute signiﬁcantly to overall mass degradation, so NA assessment should account for both. Plumes derived from coal tar NAPLs are an example of dual process systems, because both oxidation and fermentation may contribute signiﬁcantly to overall carbon turnover. Because chlorinated solvents degrade by reductive dechlorination (which requires highly reducing conditions), plume length is typically controlled by core processes only.

7.2.4.2

Preferred CORONA Site Instrumentation

To obtain the data necessary to identify and quantify key processes, MLS wells equipped with sample ports that provide high resolution (25–100 cm spacing) can be installed to allow quantiﬁcation of dispersive electron acceptor and plume gradients at the plume fringe (in the case of oxidising plumes), and therefore transverse vertical dispersivity (Figure 7.2.9). Because the focus is on process quantiﬁcation rather than spatial contaminant distribution, it should be possible to make a reasonably accurate estimate of NA performance from a smaller number of high-resolution MLS wells.

7.2.4.3

CoronaScreen Models

The CoronaScreen models are predicated on the hypothesis that the identiﬁcation and quantiﬁcation of key attenuation processes will yield better NA

444

Chapter 7.2 profile 6 mm HDPE tubing

5 cm

200 um mesh nytex screen

LDPE cable tie Cross-drilled holes (0.7 mm)

Points across the upper plume fringe – 27 @ 20-25 cm spacing

Open bottom m

Points across the lower plume fringe – 27 @ 20-25 cm spacing

Figure 7.2.9

Schematic of a MLS installed in a rock aquifer to monitor a phenolics plume. Materials are HDPE, and the annulus is ﬁlled with ﬁne sand.

assessment than evaluation of spatial concentration data. It was found during the CORONA project ﬁeld investigations (six diﬀerent contaminant plumes in unconsolidated and bedrock aquifers), that diﬀerent processes contributed to varying extents to NA. It follows that the parameters that have the most impact on plume length should be estimated with a greater degree of rigour. This diﬀers from the conventional approaches embodied in MNA assessment protocols and the BioScreen model, which rely on concentration data collected along the plume centreline and estimates of dispersivity.

7.2.4.4

Output Goals

As stated earlier, knowledge that is useful at the early stages of site management include some estimate of maximum plume length, concentration at key distances from source (compliance boundaries, receptors, sensitive monitoring locations) and the time taken for a plume to reach maximum length. The primary goal of a screening model is to obtain these useful estimates with as high a degree of accuracy as possible. It should be understood that the outputs described below are calculated from the point where ﬁeld data are collected. Therefore, if data are collected some distance downgradient of a source zone, the maximum plume length is the sum

Forecasting Natural Attenuation

445

of that returned by the models and the distance from the source to the monitoring point. For example, if input data are collected from a monitoring point located 100 m downgradient of a petroleum spill source, and the model returns a plume length of 500 m, the actual total maximum plume length is 600 m. The model automatically performs this calculation.

7.2.4.4.1

Maximum Plume Length

As described above, plume evolution involves growth, steady-state and decay phases. A plume reaches steady state when contaminant mass ﬂux downgradient is balanced by mass attenuated through natural processes, predominantly biodegradation. At steady state, the plume will have reached its maximum length unless concentrations in the source area increase. Some estimate of this length is useful when considering impacts to potential receptors or other regulatory compliance point. For example, if a given plume is expected to migrate no more than 500 m from source, it is unlikely to impact an abstraction well 800 m away, and thus poses no risk. The maximum plume length is the distance from source to the location of some speciﬁed concentration. Several alternative concentrations could be used, for example the maximum allowable concentration (e.g. 5 mg l1 for benzene), the analytical detection limit (varies with analytical method) or zero concentration (not attainable in a practical sense). Diﬀerent regulatory systems may have their own criteria for deﬁning plume boundaries: the user is referred to local or national regulations. Alternatively, a working deﬁnition may be negotiated with local authorities. Examination of numerical model results suggests that there is no signiﬁcant diﬀerence in plume length for the diﬀerent plume limit deﬁnitions. Nevertheless, the CoronaScreen models all deﬁne maximum plume length as the point where concentrations are reduced to zero, which provides the most conservative estimate.

7.2.4.4.2

Concentration at Compliance

One of the analytical model outputs from the CoronaScreen model is concentration plotted along the plume centreline. From these plots, concentrations at a given distance from the source can be estimated. The analytical model can also output vertical concentration proﬁles of contaminant and electron acceptor at a user-deﬁned distance from the source. These can also be used to indicate contaminant concentration at any point along the modelled plume ﬂow path.

7.2.4.4.3

Time to Reach Steady State

If there is stoichiometrically more contaminant (electron donor) mass than electron acceptors, a plume will grow in length until attenuating mechanisms oﬀset contaminant ﬂux. An estimate of how long it will take for a plume to reach steady state (i.e. maximum plume length) may be useful in developing MNA sampling frequency and inﬂuence site management decisions. As an

446

Chapter 7.2

example, if a plume is estimated to require decades to reach steady state, it may be desirable to implement some engineered remediation, while a plume estimated to reach steady state in a few years might be left on its own.

7.2.4.5

Model Descriptions

The CoronaScreen suite consists of three simpliﬁed reactive transport models (Analytical, Electron Balance and Travelling 1-D) that operate within an Excel spreadsheet environment. Data is input into topic-related section ﬁelds: each model extracts the required inputs (see Table 7.2.2), and performs the relevant calculations using VBA macros. Output is displayed as maximum plume length or concentration at a speciﬁed distance from source, and time to reach steady state. A database of common contaminants that can be added to the relevant input section is contained within the spreadsheet. It is also possible to add contaminants to the database provided the user has a balanced half reaction equation for each species. The Travelling 1-D model uses PHREEQC (geochemical speciation and reactive transport code), which is embedded in the CoronaScreen environment and automatically called when the Travelling 1-D model is used. Each of the three models is predicated on a fundamentally diﬀerent premise, providing the opportunity to assess NA from diﬀerent theoretical and conceptual perspectives. The Analytical model is similar to that used in BioScreen and NAS, except that both electron acceptor and donor concentrations are

Table 7.2.2

Parameters required by the CoronaScreen models.

Parameter

Travelling 1-D

Analytical

Electron Balance

Groundwater velocity Vertical dispersivity Horizontal dispersivity Longitudinal dispersivity Plume width Plume thickness Plume fringe thickness Background EAa concentrations Plume EA concentrations Background EDb concentrations Plume ED concentrations Porosity Aquifer bulk density Fraction organic carbon Distance: source to ‘‘source’’ well Distance: source to MLSc well

|

| | | | | | Optional | | | |

| | |

a b c

| | | |

| |

| | | | | | | | | | | |

EA ¼ dissolved electron acceptors or their redox couples: O2, NO3, SO42, Mn21, Fe21, CH4. ED ¼ electron donors (i.e. plume constituents). MLS ¼ multilevel sampler.

Forecasting Natural Attenuation

447

converted to stoichiometric electron equivalents to implicitly represent multiple contaminants and electron acceptors, consistent with reaction stoichiometry. This is an important innovation in the ability of NA screening models to simulate complex plumes of mixed organic and inorganic contaminants. The convention is also used in the Travelling 1-D and Electron Balance models.

7.2.4.5.1

Analytical Model

The Analytical model is based on a closed-form analytical solution of the transport equation that simulates advection, dispersion and biodegradation of a ﬁnite dimension continuous source emanating from a vertical plane.20 The model makes the following assumptions: (i) biodegradation reactions are instantaneous, which for most groundwater velocity regimes is a valid assumption;21,22 (ii) the plume is at steady state; (iii) velocity is uniform and constant; and (iv) the source is continuous and spatially/temporally uniform. Nearly all the required inputs are parameters that can be measured in the ﬁeld (Table 7.2.2), with the possible exception of plume width. Vertical dispersivity can be either estimated based on literature values or calculated from reactive zone thickness derived from vertical concentration proﬁles obtained from multilevel wells or drive point proﬁling. This reactive zone thickness is a key input for the Electron Balance model, and is discussed in greater detail below. Because the shape of vertical proﬁles will change with distance travelled, accurate estimation of vertical dispersivity requires that the proﬁles be adjusted for the distance between source and observation. In addition to estimating plume length, the model calculates the time required to reach steady state. Time to steady state is calculated using the input velocity, which is otherwise not used by the model (velocity is not needed to calculate plume length). A plot of contaminant concentration along the plume centreline is generated, from which concentration at some compliance point downgradient of the source can be estimated. Vertical proﬁles of electron acceptors and contaminants at user-deﬁned distances from the source are also generated, from which an evaluation of expected vs. observed proﬁles could be performed to check or reﬁne the conceptual site model.

7.2.4.5.2

Electron Balance Model

In the Electron Balance model,12 the plume is represented by a box with inwardly dispersing electron acceptors and outwardly dispersing electron donors on the lateral, top and bottom sides (Figure 7.2.10). The model compares the equivalents of electron acceptors and electron donors entering the plume ‘‘box,’’ and based on the stoichiometric demand for electron acceptors,

448

Chapter 7.2 source Electron Donor input

vertical dispersive Electron Acceptor mixing

background advective Electron Acceptor input plume

residu

als

horizontal dispersive Electron Acceptor mixing

Figure 7.2.10

Schematic of plume representation by the Electron Balance model.

iteratively increases the plume length until both electron donors and electron acceptors are balanced. At that point, there are suﬃcient electron acceptors for biodegradation of all the electron donors: i.e. the plume is at steady state and reached its maximum length. The Electron Balance model has the most input parameter requirements of the three models. Because it considers dissolved inorganic carbon and methane in background groundwater and the plume, the model can directly account for fermentation in the plume core. This means that both plume fringe and core degradation processes for oxidisable contaminants can be represented. The Electron Balance model assumes that: (i) (ii) (iii) (iv)

the source is inﬁnite and temporally invariant; plume velocity is uniform in space and time; bioreactions are instantaneous; and reactive fringe thickness (RFT) does not change with travel distance.

It is well accepted that horizontal transverse dispersion is greater than vertical transverse dispersion (generally assumed to be a factor of 10). It is therefore reasonable to assume that horizontal RFT will similarly be greater than vertical RFT. The Electron Balance model can convert vertical RFT (estimated from MLS data) to vertical dispersivity using a simple equation derived by algebraic manipulation of the aforementioned analytical solution: pﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ CEAmax RFT ¼ 1 patv L CEDmax ½1 þ ðCEAmax =CEDmax Þ It then assumes that horizontal dispersivity is an order of magnitude greater than the vertical dispersivity, and calculates the horizontal RFT using the same equation. Of course, RFT increases with travel distance: the equation accounts for this by taking as input the distance between the source and the observation

449

Forecasting Natural Attenuation

point. Thus, uniform dispersivity, consistent with theory, is obtained from diﬀerent MLS data collected along a given plume. It is important to note that assumption 4 (uniform RFT with distance) is in violation with the above. However, the error introduced into estimates of maximum plume length decreases when RFT is estimated from data collected some distance from the source (roughly half the anticipated plume length is ideal).

7.2.4.5.3

Travelling 1-D Model

The Travelling 1-D model uses the numerical geochemical speciation and transport code PHREEQC to dispersively mix a one-dimensional column of water, oriented transverse to the plume axis, until the plume fringe bioreactions reduce plume concentrations to zero. The time taken to reach this condition is translated into distance (i.e. plume length) using site groundwater velocity (or plume velocity if the contaminant is retarded). Only the top plume fringe is modelled; it is assumed that the same fringe processes occur at the bottom fringe. Figure 7.2.11 shows conceptually how the model functions, although the model does not actually simulate advection of the water column. Near the source, the water column is assumed to be occupied by electron acceptor(s) above the background/plume interface only, and electron donor only below the interface (Figure 7.2.11 idealised graphs). As time proceeds, electron donors and electron acceptors are consumed and the reactive front becomes less sharp. At steady state, there is only a dispersive proﬁle of electron acceptors, contaminants having been completely consumed. The model assumes instantaneous reactions, and since initially electron donors are much greater than electron acceptors, the reaction front, i.e. plume fringe, moves outward from the plume

depth

Travel Time = simulation time in 1D model

EA

Reaction front

Reaction front

ED

depth

Conc.

Max transverse extent of ED plume

Plume fringe ED=EA

Position of 1D model relative to steady plume

Flow

Steady State Plume length = Travel Time * Velocity

Figure 7.2.11

Conceptual diagram of how the Travelling 1-D model works. Transport is modelled only vertically.

450

Chapter 7.2

core (Figure 7.2.11). Transverse mixing and reactions proceed until the plume reaches its maximum extent and the plume fringe retreats towards the plume centreline. The model has four assumptions: (1) the source is continuous; (2) velocity ﬁeld is steady and uniform; (3) reacting species (both contaminant and electron acceptor) are mobile and unretarded; and (4) longitudinal dispersion does not signiﬁcantly inﬂuence maximum plume length. Assumption 3 means that reactions with immobile mineral phases (e.g. iron(III)) cannot be considered. While retardation cannot be included, it does not impact on maximum plume length (it does impact the time to reach steady state). Assumption 4 is valid provided the ratio of plume length to longitudinal dispersivity is greater than 30 (i.e. most of the time). The Travelling 1-D model has the fewest input data requirements of the three CoronaScreen models (Table 7.2.2). However, the dispersion coeﬃcient used to mix electron acceptors and plume is deﬁned by the sum of mechanical mixing (velocity transverse vertical dispersivity) and diﬀusion, and therefore the model is sensitive to vertical dispersivity.

7.2.5

Recommendations

These concluding remarks summarise the role of NA in managing contaminated land and groundwater pollution, and oﬀer some recommendations on how to forecast NA with conﬁdence. Groundwater pollution from point sources is only one of many environmental threats facing society. It can be technically diﬃcult and expensive to clean up, and so society’s scarce resources must be used wisely on the highest priority sites. Risk-based decision-making is increasingly being used to identify the sites with the greatest potential to harm health or the environment. An intrinsic part of this process is understanding the extent to which pollution will naturally decay: clearly natural processes are more sustainable and cheaper than engineered interventions, if they will reduce risk suﬃciently. With appropriate monitoring, these processes of NA are often invoked as the remedial strategy for a contaminated site under the title of MNA. MNA can be a remedial strategy in two ways. It can be the only strategy adopted, because the understanding and modelling of future behaviour suggests that NA will be suﬃcient to protect health and the environment. Frequently, MNA is part of a multifaceted strategy, in which engineered clean-up is used to remove some of the pollution, most commonly the source zone. MNA is then invoked as the strategy to deal with the residual pollution, perhaps the groundwater plume, or remnants of the source.

Forecasting Natural Attenuation

451

The NA processes which occur on any site are the result of interactions between the often unknown mixture of pollutants, the hidden and heterogeneous aquifer materials and the microorganisms present. Forecasting the outcomes of this complex interaction requires a good scientiﬁc understanding and high-quality site investigations. A broad overview of the scientiﬁc processes has been given above, and further details are in the references, but additional knowledge and further research are still required to interpret many of the more complex sites. Good site investigation is the main tool in forecasting NA with conﬁdence. The data collected must be suﬃcient to build a clear CSM, identify the key processes and parameters and then quantify these. It is an old adage but worth repeating, that money spent on ground investigation is repaid many times over by reducing uncertainty. We strongly recommend the use of some multilevel samplers to obtain vertical proﬁles of pollutant concentrations, the other chemicals involved in NA processes and redox conditions. These proﬁles can also be interpreted to calculate vertical dispersivity, which is rarely measured. Vertical dispersion is a very important process which controls mixing between the plume and background water, and hence the rate at which oxidants (electron acceptors) are mixed into the plume and made available to microorganisms. With the new understanding of the importance of vertical dispersion gained in the CORONA project, the project team have developed new mathematical expressions to describe the growth and shape of plumes which are attenuated by oxidation of the pollutants in them. A new tool for making ﬁrst estimates of the length of such plumes, CoronaScreen, has been constructed from these relationships and made freely available. Three diﬀerent methods of calculation are included in the Excel spreadsheet, and manuals and a guidance document are also available as free downloads from www.shef.ac.uk/corona.

Acknowledgements The CORONA concepts and CoronaScreen tool were mainly developed in a research project principally funded by the European Union within the 5th Framework Programme under contract EVK1-2001-00087; the Environment Agency for England and Wales and CL:AIRE provided additional ﬁnancial support. The authors are grateful to the other members of the CORONA research team for their contributions to the project and for hospitality and interesting discussions in a variety of locations. Some of the technical material in this chapter was originally published in the guideline document for CoronaScreen.14

References 1. R. M. Davison, G. P. Wealthall and D. N. Lerner. Source Treatment for Dense Non-aqueous Liquids. R&D Technical report P5-051/TR/01, Groundwater Protection and Restoration Consultants Ltd for Environment Agency, Bristol, 2002 (ISBN 1 857054 830).

452

Chapter 7.2

2. Committee on Source Removal of Contaminants in the Subsurface, Contaminants in the Subsurface: Source Zone Assessment and Remediation, National Research Council, National Academies Press, 2004 (ISBN: 030909447X). 3. M. D. Einarson and D. M. Mackay, Environ. Sci. Technol., 2001, 35(3), 66A–73A. 4. Environment Agency, Guidance on the Assessment and Monitoring of Natural Attenuation of Contaminants in Groundwater, R&D 95, UK Environment Agency, Almondsbury, 2000. 5. ASTM Guide for Remediation by Natural Attenuation at Petroleumeum Release Sites, American Society for Testing and Materials, Philadelphia, PA, 1998. 6. T. H. Wiedemeier, H. S. Rifai, C. J. Newell and J. T. Wilson, Natural Attenuation of Fuels and Chlorinated Solvents in the Subsurface, Wiley, 1999. 7. T. Buscheck and K. O’Reilly, Protocol for Monitoring Intrinsic Bioremediation in Groundwater, Chevron Research and Technology Company, Health, Environment and Safety Group, 1995. 8. M. J. Barcelona, Site characterisation: what should we measure, where, when and why?, Proceedings of the Symposium on Natural Attenuation of Ground Water, 20-25, US EPA/600/R-94/162, 1994. 9. American Petroleumeum Institute, Evaluation of Sampling and Analytical Methods for Measuring Indicators of Intrinsic Bioremediation, API Soil & Groundwater Research Bulletin No. 5, 1998. 10. S. F. Thornton, S. Quigley, M. Spence, S. A. Banwart, S. Bottrell and D. N. Lerner, J. Contam. Hydrol., 2001, 53, 233–267. 11. R. D. Wilson, S. F. Thornton and D. M. Mackay, Biodegradation, 2004, 15, 359–369. 12. S. F. Thornton, D. N. Lerner and S. A. Banwart, J. Contam. Hydrol., 2001, 53, 199–232. 13. D. N. Lerner, P. Bjerg, J. Datel, A. Gargini, P. Gratwohl, C. Holliger, P. Morgan, T. Ptak, R. Schotting, H. Slenders and S. F. Thornton, CORONA: conﬁdence in forecasting of natural attenuation as a risk-based groundwater remediation strategy, ﬁnal report of the EU research project EVK1-2001-00087, University of Sheﬃeld, UK, 2005 (available from: www.shef.ac.uk/corona). 14. R. D. Wilson, S. F. Thornton, A. Hu¨ttmann, M. Gutierrez-Neri and H. Slenders, CoronaScreen: process-based models for natural attenuation assessment, Guidance for the application of NA assessment screening models, University of Sheﬃeld, UK, 2005 (available from: www.shef.ac. uk/corona). 15. M. Einarson, Multilevel groundwater monitoring, in Practical Handbook of Environmental Site Characterisation and Groundwater Monitoring, ed. D. M. Neilsen, CRC-Taylor Francis, 2nd edn., pp. 807–848. 16. C. Newell, J. Gonzales and R. McLeod, BIOSCREEN Natural Attenuation Decision Support System, version 1.3, EPA/600/R-96/087, US EPA, Washington, DC, 1996.

Forecasting Natural Attenuation

453

17. F. Chapelle, M. Widdowson, J. Braunner, E. Mendez and C. Casey, Methodology for estimating times of remediation associated with monitored natural attenuation, 03-4057, US Geological Survey Water Resources, 2003. 18. C. Aziz, C. Newell, J. Gonzales and P. Haas, Biochlor: natural attenuation decision support system, US EPA, Washington, DC, 2000. 19. P. Domenico, J. Hydrol., 1987, 91, 49–58. 20. P. A. S. Ham, R. J. Schotting, H. Prommer and G. B. Davis, Adv. Water Resour., 2004, 27, 803–813. 21. R. Borden and P. Bedient, Water Resour. Res., 1986, 22(13), 1973–1982. 22. C. Gramling, C. Harvey and L. Meigs, Environ. Sci. Technol., 2002, 36(11), 2508–2514.

CHAPTER 7.3

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use: Management and Policy Implications of Scientiﬁc Realities STEPHEN FOSTERa AND LUCILA CANDELAb a

IAH President, International Association of Hydrogeologists, PO Box 9, Kenilworth, Warwick CV8 1JG, UK; b Technical University of Catalonia, Dept of Geotechnical Engineering & Geoscience, Gran Capita`, s/n Ed. D-2, ES-08034 Barcelona, Spain

7.3.1

Why is Agricultural Land-use the Greatest Challenge Facing the New EC Water Directives?

In many regions of the European Union (EU) the main recharge areas of groundwater bodies form valuable tracts of farming land and are extensively utilised for agricultural production: either crop cultivation or animal husbandry. Very extensive land areas are often involved and this means that: large volumes of groundwater replenishment (originating as excess rainfall and excess irrigation inﬁltrating this agricultural land) are potentially aﬀected by agronomic practices; and relatively large numbers of individual agricultural land-users (potential polluters) are implicated. Agricultural land-use is an inherently ‘‘leaky activity,’’ and one which is always potentially prone to leaching of nutrients and pesticides: especially when practised on the more permeable soils that tend to coincide with the most important aquifer recharge areas. And most of the EU has witnessed radical changes in agronomic practices over the past 20–40 years, associated with 454

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

455

(largely successful) attempts at increasing agricultural productivity (to achieve self-suﬃciency in grains, oils, milk and meat) and, in the case of Mediterranean Europe, to expand export production of vegetables and fruits. Common trends include the replacement of traditional crop rotations by near monocultures of higher-value crops across extensive areas: selected according to prevailing climatic conditions, guarantee price and/or market opportunities. This intensiﬁcation of agricultural production (in some cases with increased cropping frequency) has been sustained by the application of ever-increasing quantities of inorganic fertilisers and a wide spectrum of synthetic pesticides, together (in areas of drier climate) with new irrigation schemes and increasing irrigation water-use eﬃciency.

Irrigated Cultivation in Mediterranean Spain The most intensive land-use in Europe, in terms of water consumption and agrochemical application, corresponds to areas of irrigated multi-cropping. In Spain irrigated farmland represents 13% of the total cultivated area but contributes 50% of agricultural income.1 Some 25–30% of this irrigated area uses groundwater, especially in the Mediterranean coastal zone. In most areas where irrigated land directly overlies an unconﬁned groundwater body, a substantial proportion of the total groundwater recharge is produced by ‘‘irrigation returns’’—either representing supplementary recharge or recycled water depending on whether it is derived from surface water or groundwater irrigation. During the last 15 years or so, traditional ﬂood irrigation has been widely replaced by more eﬃcient systems (pressure tube distribution to microsprinklers or drip nozzles). Very high (but locally variable) application rates of (almost exclusively) inorganic fertiliser predominate, although rates appear to have stabilised or declined somewhat in recent years. In addition the use of plant protection products (notably insecticides) has increased greatly both in absolute quantity and in range of active ingredients.

Arable Farming in Eastern England The land-use of this extensive region underwent major changes during the period 1950–1970 which had a number of components: conversion of large areas of permanent pasture to tilled land under arable cultivation; major increases in the proportion of land dedicated to cereal cultivation, from rotations of less than 20% to monocultures of more than 60% in many areas; and substantial increase in the application of inorganic fertilisers to sustain more continuous cereal cultivation, on average from less than 50 kgN ha 1 yr 1 to more than 100 kgN ha 1 yr 1.

456

Chapter 7.3

The net eﬀect of these changes during 1940–1980, coupled with some intensiﬁcation of grassland-based livestock rearing, was a 20-fold increase in fertiliser nitrogen use to achieve a threefold increase in food production. In subsequent years there has been a progressive move from spring-sown cereals (following autumn ploughing and winter fallow) to winter-sown cereals (with minimal tillage implying improved nutrient management but increased herbicide use), together with introduction of substantial areas of oil-seed rape and oscillating amounts of more traditional crops like sugar beet, potatoes and peas. Moreover, in recent years the amount of nitrogen fertiliser applied to winter-sown cereal and oil-seed crops has reduced somewhat, and an incipient trend towards the return of permanent pasture for sheep rearing is evident in some areas. In the past agricultural land-use has been essentially ‘‘uncontrolled’’—although over the past 50 years changes have been driven by government subsidies (as guarantee prices for certain produce), market opportunities and agro-technical innovation. Moreover, until the early 1980s there was general complacency about risks to groundwater quality. This was because ﬁeld investigations by agricultural scientists produced misleading results since they were based on monitoring drainage from relatively heavy soils—for which a substantial proportion of nitrogen losses were gaseous (following soil denitriﬁcation) with much reduced leaching losses compared to those subsequently proven for more aerated soils (typical of groundwater recharge areas). There is thus a legacy of decades of laissez faire agricultural land-use to overcome: with a substantial proportion of the groundwater bodies deﬁned during the initial phase of EC Water Framework Directive implementation not achieving ‘‘good chemical status’’ as a result of the eﬀects of past agricultural land-use (Figure 7.3.1). More recently, and in eﬀect since the promulgation of the EC Nitrates Directive (1991), groundwater quality considerations have begun to enter into consideration, but as yet there has been little clear incentive for (and absolutely no constraint on) farmers in this regard. This situation arises because of: lack of understanding of diﬀuse groundwater pollution and individual capacity to inﬂuence it; and the fact that fertilisers represent a very small proportion of agricultural costs. On a widespread basis across the main recharge areas of important groundwater bodies, there is a pressing need for more groundwater-friendly regimes of agricultural land-use to be promoted in the interest of future groundwater quality protection (and of signiﬁcantly diluting existing diﬀuse pollution). For this it will be necessary to integrate groundwater considerations (and costs) fully into emerging action plans for EU Common Agricultural Policy (CAP) reform.

457

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

100 80 60

Figure 7.3.1

7.3.1.1

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

Year

40 20 0 1985

mg/l

160 140 120

Preliminary assessment of excess nitrogen loading and groundwater nitrate concentrations in the Jucar Basin of eastern Spain.2

How does Agricultural Land-use Impact on Groundwater Quality?

Agricultural soils contain large, but widely varying, quantities of nitrogen in organic form: often amounting to more than 2000 kgN ha 1. This is oxidised by soil bacteria to soluble nitrate (at rates varying with soil temperature and humidity) and is then susceptible to leaching below the root zone. Inorganic nitrogen fertilisers are added to increase the immediate availability of nitrate for plant growth, while manures (which also contain large quantities of less readily available nitrogen) are applied to replenish soil organic matter. Nitrate applied to cultivated land is subject to complex soil processes: it may be taken up directly by the growing crop, incorporated into the soil nitrogen pool, reduced and lost either as NH3 or N2 gas or leached as NO3 or NH4 to

458

Chapter 7.3

groundwater. Leaching happens whenever excess rainfall occurs at times when nitrate is present in the soil, as a result of heavy fertiliser application and/or natural oxidation of residues. Thus while only a small proportion of the nitrate leached in a given year is derived directly from inorganic fertilisers, the overall rate of nitrogen mineralisation and leaching normally relates in a general way to fertiliser application rates, although some leaching will occur even when no nitrogen is applied and/or the land is fallow. All pesticide compounds are, to greater or lesser degree, chemically tailored to be toxic and persistent. However, until the mid-1980s, there was not much concern about the possibility of leaching to groundwater, since agricultural scientists argued that certain other processes would predominate: soil sorption of higher molecular weight compounds (such as chlorinated hydrocarbon insecticides), introduced from the 1950s and characterised by low solubility but great persistence; and volatilisation of (subsequently introduced) lower molecular weight more soluble compounds (such as most herbicides).

Nitrate Leaching from Typical Crop Cultivation Regimes On the Chalk recharge area of eastern England greatly increased rates of nitrate leaching result from the conversion of pasture land to continuous cereal cultivation, even in situations where both receive low applications of inorganic fertiliser. In arable soils, nitrate is readily leached by excess rainfall at times when it is present in excess of plant requirements, whereas the existence of soil compaction and a root system under pasture land lead to greater nitrate uptake and probably also to signiﬁcant soil denitriﬁcation losses. The former conclusion was corroborated by many vadose zone proﬁles beneath long-standing arable land in the drier parts of England3 with nitrate leaching usually much in excess of 50 mg N03 l 1. The leaching of nitrate from dryland agricultural soils is dependent on a complex interaction of soil type, cropping regime and rainfall inﬁltration. In intensive cereal cultivation 30–70 kgN ha 1 yr 1 can be leached to groundwater from fertiliser applications of 100–150 kgN ha 1 yr 1, and higher leaching losses occur from heavier applications to potatoes and oil-seed crops. Nitrate leaching is particularly marked in areas of intensive irrigated agriculture (horticulture and citriculture) in Mediterranean Spain where sandy soils are predominant. In areas where irrigation rates exceed 1000 mm yr 1, leaching to groundwater of 50–250 kgN ha 1 yr 1 has been estimated to occur,4 equivalent to 50% of that applied even given an irrigation eﬃciency of around 0.7. More traditional agricultural cropping has a fertiliser eﬃciency of less than 0.65 and the impact of this has already been experienced in the Valencia region, where the drinking water supply of 400 000 population exceeded the EC-MAC of 50 m NO3 l 1.6

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

459

Moreover, in a ‘‘fertile soil’’ most modern pesticides have half-lives of less than 1 year (and many of less than 1 month), and thus it might be assumed that soil residues would thus be eliminated (by aerobic biodegradation or chemical hydrolysis) preventing groundwater contamination. However, the property of greatest importance in respect of pesticide leaching is ‘‘mobility in soil solution,’’ which varies inversely with aﬃnity for organic matter and/or clay minerals.7,8 If mobile pesticides are present at times of excess rainfall or irrigation, they are likely to be leached into the vadose zone or, where preferential ﬂow is signiﬁcant, rapidly to the water table. Once below the microbiologically active soil zone they will be very much more persistent than suggested by the manufacturer’s quoted ‘‘half-life in a fertile soil.’’9 This scientiﬁc reality has important implications for the protection of groundwater quality.

Pesticide Leaching From Typical Crop Cultivation Regimes The EC-MAC for pesticides in drinking water (0.1 mg l 1) has already been considerably exceeded in many British public water supply boreholes, although concentrations in excess of 1.0 mg l 1 have seldom been recorded. The pesticides most frequently encountered to date are all herbicides: atrazine, simazine, mecoprop and isoproturon.10 Detailed monitoring of the Chalk water table at sites in Hampshire under continuous arable cultivation using regular applications of isoproturon (to winter cereals) or atrazine (to summer maize) has revealed some penetration of these pesticides into groundwater via preferential ﬂow immediately after the onset of recharge. Signiﬁcant pesticide concentrations were also detected in the upper part of the vadose zone in matrix transport but were attenuated within 5 m of the surface.11 In Mediterranean Spain insecticides are more heavily applied than other types of pesticide, as a result of the prevailing climatic conditions. The irrigated citriculture and horticulture belt has long used heavy applications of a range of insecticides and herbicides, and Beltra´n et al.12 investigated the vadose zone mobility of organochlorine pesticides (such as tetradifon), organophosphorous pesticides (such as dimethoate) and phenoxyacid herbicides (such as MCPA), by installing soil suction samplers to depths of 3.5 m below orange groves. Concentrations up to 1.0 mg l 1 of the more mobile compounds (notably dimethoate and MCPA) were recorded at depths of up to 3.5 m within 10 days of application, suggesting that some pesticide was reaching the water table at 4 m depth by preferential ﬂow, although none was detected in the water wells of the area. A similar experiment in an intensively cultivated horticultural area northeast of Barcelona showed glyphosate presence in detectable concentration at more than 2 m depth13 and Candela5 has detected residues of organochlorine pesticides (such as lindane and heptachlor) in groundwater at concentrations of more than 10.0 mg l 1 in the same area.

460

7.3.1.2

Chapter 7.3

Are all Types of Groundwater Body Equally Threatened by Agricultural Practices?

Across the EU diﬀerent topographical, pedological and groundwater conditions, associated with diﬀerent hydrogeological settings (or typologies), lead to wide variation in: the potential for generation of contaminant pressures from nutrient and pesticide leaching (as a result of diﬀering types of agricultural and nonagricultural land-use in aquifer recharge areas); and the capacity for natural contaminant attenuation (or the intrinsic aquifer pollutant vulnerability), which depends largely on the thickness and character of the vadose zone. In a broad sense this is illustrated by Figure 7.3.2, which is based on some widely occurring hydrogeological typologies in the EU countries, although in practice the range of typologies encountered is signiﬁcantly greater that those illustrated. Among groundwater bodies at most serious risk from agricultural activity are: the coastal and alluvial aquifers of Mediterranean Spain, because they have a coarse-granular vadose zone of less than 20 m thick and relatively low natural recharge rates, and their recharge area is often utilised for intensive irrigated agriculture with double or triple cropping; and the extensive outcrops of highly permeable ﬁssured porous limestones and sandstones in northern Europe, which have been subject to increasingly intensive monocultures of cereal and oil seed crops sustained by heavy agrochemical applications. The question of ‘‘how far can natural subsurface contaminant attenuation go in providing the required level of groundwater quality protection?’’ is an important component of the more general question posed above, and one which merits more detailed discussion in relation to: the extent and stability of natural subsurface denitriﬁcation in the aquifer system concerned; and the capacity of the aquifer concerned for subsurface pesticide retention and degradation. Denitriﬁcation has been the subject of considerable research14 because if active on a widespread basis (and without other complication) it can have a major beneﬁcial eﬀect on groundwater quality. Evidence comes from the conﬁned parts of some British aquifers, but in the vadose zone of the major aquifers the generally aerobic conditions and persistence of high nitrate concentrations to depth imply that denitriﬁcation cannot be widely active, despite the presence of potentially denitrifying bacteria.9 It has been suggested that denitriﬁcation is

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

Figure 7.3.2

461

Eﬀect of hydrogeological setting on the general susceptibility to agricultural impacts and the approach to groundwater protection.

more likely in the zone of water table ﬂuctuation, but measurement of the normal gaseous products (N2, N2O) suggested that it is not signiﬁcant in relation to the overall nitrogen ﬂux. However, where the vadose zone includes strata rich in organic carbon, the process is likely to become more signiﬁcant. The factors inﬂuencing pesticide attenuation in the vadose zone are transit time and attenuation capacity, which interact with the persistence and mobility

462

Chapter 7.3

of the pesticide concerned. Leaching of pesticides into groundwater is to be expected under sandy soils of low organic matter with a water table within 5 m depth, where mobile and persistent pesticides are applied at a high rate. But there is considerable evidence that ‘‘natural preferential ﬂow’’ is a major factor in the transport of mobile (and in some cases sorbed) pesticides through the vadose zone of a wide range of aquifer types (especially consolidated formations).7,15

7.3.1.3

What can be done to Make Agricultural Cropping more ‘‘Groundwater Friendly’’?

A number of measures can contribute to constraining agricultural land-use practices in the interest of groundwater quality, but their implementation requires improved institutional capacity for cross-sectoral action. Certain amongst these are actually facilitated by the new EC directives themselves, whilst others will be required to fulﬁl the broader requirement of the same legislation. These approaches are introduced below with comments on their potential applicability, possible impediments and ultimate limitations. In the longer run it will be essential to promote an appropriate combination of measures using the powers and obligations of the new EC directives (2000 and 2006) in combination with EC CAP reform to move from ‘‘universal crop production subsidies’’ to ‘‘selective support of catchment sensitive farming,’’ including the introduction of land management regimes which on balance at the local scale are ‘‘groundwater quality friendly’’—and in turn to keep under review the need for potential adaptation in response to accelerated climate change.

7.3.2

Guidelines on ‘‘Best Agricultural Practice’’

The dissemination of improved agricultural practices (via rural extension services of the appropriate government agency) which take account of aquifer vulnerability and groundwater quality concerns will have the joint beneﬁts of: tending to reduce the leaching of nutrients and pesticides from agricultural land to groundwater; and raising farmer awareness about their inﬂuence on groundwater and role in its protection. In this context it will be important to have a groundwater quality monitoring system installed that is capable of demonstrating the corresponding improvements. The sort of improved practices envisaged include: rationalising inorganic fertiliser regimes and reducing initial applications (e.g. avoiding nitrogen fertiliser application when sowing autumn cereals and oil seeds in view of the natural availability of soil nutrients at this time);

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

463

reducing autumn ploughing and winter fallow, which are associated with high rates of soil nutrient leaching, by using autumn ‘‘cover crops’’ and direct drilling techniques; curbing the practice of ‘‘disposal’’ of agricultural manures and sludges to grassland (way in excess of normal fertilisation needs) in areas with intensive livestock rearing; and where groundwater is used for irrigation, checking its nitrogen and potassium concentrations and allowing for this source of nutrients when considering the rate of fertiliser application. While an element of self-interest can usually be identiﬁed in ‘‘selling’’ to farmers the need for control over leaching losses—through reducing irrigation water and agrochemical use, avoiding pollution of their own groundwater supply, etc.— the reality is that these losses are usually only a minor item in overall crop production costs and the potential ﬁnancial saving does not generally represent a signiﬁcant incentive for control. Moreover even under ‘‘best agricultural practice,’’ where intensive monocultures are practiced and aquifers are relatively vulnerable to pollution, this will generally not be suﬃcient alone to reduce the average rate of leaching losses to within drinking water and aquatic ecosystem guidelines. The EC Nitrate Directive (1991) constituted an attempt to provide an instrument for control of nitrate leaching from agricultural soils in the interest of groundwater quality, and involved the declaration of Nitrate Vulnerable Zones (NVZs) in which constraints were placed on inorganic fertiliser and organic slurry application rates according to ‘‘best agricultural practice.’’ But in most instances such measures were not suﬃcient alone to reduce the average nitrate concentration of groundwater recharge to below 50 mg l 1, since they did not provide for the introduction of a more mixed land-use regime with interspersed ﬁelds of lower cultivation intensity.16,17

7.3.3

Reducing Overall Cultivation Intensity

An earlier English scheme for Nitrate Sensitive Areas (NSAs) included stricter measures than those required by the NVZs with provision for: conversion of some intensively cropped arable land to low-intensity pasture land; and ensuring a high proportion of autumn-sown crops to reduce fallow during periods of excess rainfall. These measures proved successful in signiﬁcantly reducing nitrate leaching losses,18 albeit that the beneﬁts came with substantial time-lag in most aquifer systems and the negotiation of compensation paid to farmers was repeatedly contested. Since pasture is less prone than cultivated land to nitrate leaching, it would appear to oﬀer a useful option for reducing overall rates of nitrate leaching to a

464

Chapter 7.3

groundwater body. Where grass is cut and removed for animal fodder, the use of nitrogen is relatively eﬃcient and leaching is modest even at relatively high fertiliser application rates—but leaching rates on well-drained soils increase abruptly to elevated levels when grassland productivity is intensiﬁed by highdensity grazing sustained by large fertiliser applications. Moreover, organic nitrogen accumulates in pasture land and can be oxidised and leached at high rates following ploughing and reseeding.19 In some areas more direct action has been taken by water supply utilities, who have been prepared to enter into private legal and ﬁnancial agreements with the farmers of a speciﬁc groundwater protection area to modify their activities in the interest of ‘‘harvesting’’ a high-quality groundwater supply through such actions as: converting existing arable land to permanent pasture or woodland; restricting animal grazing densities generally and not grazing certain ﬁelds; and reducing the application of agrochemicals or adopting a completely organic farming regime. This type of agreement will be more readily negotiable if some form of ﬁnancial support from government to encourage groundwater protection is also forthcoming. To date the best examples of constraining or changing agriculture in the interest of groundwater quality in the EU countries have involved action in speciﬁc protection zones (e.g. historically in Sussex, England, and Bavaria, Germany, and more recently in Jutland, Denmark, and Lower Saxony, Germany) (e.g. Refs. 17 and 20). These zones have (variously) been established by regulatory action alone, by purchase and conversion of agricultural land to pasture or woodland, or by private ﬁnancial and legal long-term agreements on land-use changes and cultivation practices.

7.3.4

Constraints on Pesticide Manufacture, Sale or Use

A wide range of pesticides are currently in use across the EU, with more than 600 diﬀerent active compounds being applied on cultivated agricultural soils,21 and there have been numerous reports of concentrations of ‘‘active pesticide compounds’’ in groundwater exceeding 0.1 mg l 1 in most countries where intensive agriculture is practised.22 The evaluation of associated groundwater quality risk is proving costly and problematic because of this wide range of compounds, the fact that some break down to toxic derivatives (metabolites) and the need to work at very low concentrations (with careful sampling to avoid volatile loss). In view of these diﬃculties, an essential prerequisite is to identify the most likely types and sources of pesticide contamination, and the most probable mechanisms of transport from the land surface to groundwater.7 Such

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

465

information is essential for the speciﬁcation of sampling protocols and monitoring networks, and to prioritise and rationalise investigation work. The removal (or refusal) of pesticide registration on grounds of expected or proven mobility to (and toxicity in) groundwater is the most eﬀective tool with which to protect groundwater quality in the interest of potable water supply and the environment. And although registration was originally introduced to protect crop consumers and agricultural workers, in recent years there have been examples of action on these grounds. Some alternatives to a complete ban include: statutory or negotiated control of the application of a speciﬁc pesticide (often with an element of farmer compensation), but this can only really be implemented at a local scale (e.g. in public water supply ‘‘safeguard zones’’); and systematic substitution of a given pesticide (which has been proven to be highly mobile and persistent in groundwater) for a certain application (e.g. atrazine and simazine for extensive defoliation, and some carbamates as soil insecticides). Various mitigation measures to control the risk of groundwater pollution can also be used for speciﬁc individual pesticides compounds such as: changes in the recommended timing and rates of application; and improving the targeting of pesticide application through changes in spraying technique. Given the formidable problem of adequate monitoring of large numbers of agricultural pesticides and their metabolites, numerical modelling is being increasingly applied to assess the potential mobility of pesticide compounds to groundwater, and thus priorities for constraint on use and for groundwater monitoring. Numerical codes (such as LEACHM, PRMZ, GLEAMS, SWAP, PEARL) have been used to classify the groundwater contamination potential from pesticides,23 some having been developed by regulatory agencies for pesticide management. Certain problems have occurred with their use because of the limited amount of validation work that has been done and a tendency to underestimate the presence of preferential transport of pesticides from the soil to groundwater through the vadose zone.

7.3.5

Improving Irrigation Water Use Eﬃciency

Where irrigation is practised, there exists the possibility of controlling soil moisture so as to maximise nutrient uptake and to restrict deep percolation (thereby controlling agrochemical leaching). This is most practicable where most plant moisture requirements are provided by irrigation, and is less feasible where supplementary irrigation is required: but even here the maximisation of nitrate uptake can be assured by providing optimum moisture levels at times of rapid plant growth, and thereby reduce soil nutrient residues. Moreover,

466

Chapter 7.3

denitriﬁcation losses (rather than nitrate leaching) become more signiﬁcant in irrigated cultivation, especially on ﬁner-grained soils. The improvement of irrigation water use eﬃciency (through the introduction of more advanced irrigation technology) can reduce immediate nutrient and pesticide leaching, and also improve crop yields and save on energy costs for pumping. It is often advocated as the panacea to all groundwater problems, since it also oﬀers the opportunity for real water resource savings through reducing non-beneﬁcial evaporation and other water losses. However, the relation between irrigation practices and groundwater recharge is a complex one. It is extremely important to note that irrigation returns often represent the major component of groundwater recharge in the more arid climates—and this will represent ‘‘new recharge’’ in cases where surface water is the source of irrigation and in eﬀect ‘‘recycled water’’ (reducing net groundwater abstraction) in the case where groundwater irrigation is practiced. This scientiﬁc reality has very important implications for groundwater resource management, since in some cases improvements in irrigation water use eﬃciency may be largely achieved through reduction in irrigation returns to groundwater, and as such do not result in ‘‘real water resource savings’’. Thus, in areas of intensive groundwater development, future resource management should be based on accounting for ‘‘consumptive water use’’ (rather than groundwater extraction per se) and this will pose new challenges for the administration and monitoring of groundwater use for agricultural irrigation. Moreover, in the longer run improvements in irrigation water eﬃciency can result in major increases in the salinity of groundwater recharge.

7.3.5.1

What are the Main Policy Implications for Groundwater Body Quality Protection?

During the last 20 years or so it has become increasingly evident that agricultural land-use practices exert a dominant inﬂuence on groundwater. For example a recent British assessment showed that during 1973–2003 public groundwater supplies totalling more than 1500 Ml per day required treatment or blending or were abandoned (resulting in capital costs alone in excess of h500 million) as a result of quality deterioration due to diﬀuse agricultural pollution and tighter EC water quality standards.24 Thus a major challenge to the successful implementation of EC directives will be achieving improved harmonisation between agricultural practices and groundwater resource conservation—on which only patchy progress has so far been made. Various types of agricultural cropping (even if conducted under ‘‘best agricultural practice’’) will not achieve drinking water or ecosystem guideline quality as regards average quality of groundwater ‘‘harvested’’ from the corresponding land. Thus other more radical measures will widely be required to satisfy the ‘‘good chemical status’’ of groundwater bodies including: at farm level—more balanced land-use regimes and improved nutrient management plans;

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

467

at groundwater supply ‘‘capture zone’’ level—a need to maintain or convert a proportion of land to low-intensity agricultural activity (woodland, pasture or recreational usage) with minimal agrochemical application; and at ‘‘groundwater body’’ level—similar land-use management approaches, but with less constraint on the precise location of areas under lowintensity use. It will be evident that the introduction of appropriate measures to control diﬀuse groundwater pollution will be highly dependent upon the presence of monitoring networks (and supporting investigations) adequate to demonstrate unequivocally their eﬀectiveness, not only in a direct regulatory sense but also to the stakeholders involved.25 The required constraints on agricultural land-use will normally have to be applied in ‘‘designated areas’’—the EC Water Framework Directive makes speciﬁc provision for the statutory declaration of ‘‘groundwater protected areas’’ with the implication that: these are likely to comprise the main recharge area of groundwater bodies used as an important source of public drinking water supply; and additional ‘‘safeguard zones’’ around individual groundwater sources will be required to stabilise or reduce potable water-supply treatment needs, but these will not be mandatory.

The Critical Issue of Adequate Groundwater Quality Monitoring The management objective of the EC directives, coupled with the large response times of many important European aquifers, means that substantial investment in groundwater quality monitoring networks will be required. Much of the routine monitoring that has been carried out in many countries to date has been focused on ‘‘external receptors’’ (mainly drinking water wells and springs), and in terms of groundwater body management has to be regarded as a ‘‘post-mortem’’ activity, since it is a very tardy and insensitive indicator of incipient pollution especially in the deeper aquifer systems. This leads to the inescapable conclusion that monitoring and assessment eﬀorts need to focus much more on the quality of contemporary recharge to groundwater bodies, since not to do so will: compromise ability to obtain an early warning of potential new groundwater quality problems; and make the timely demonstration of the eﬀectiveness of management measures impossible. Having said this the direct monitoring of groundwater recharge quality at ﬁeld scale is not readily possible on a routine operational basis, and researchlevel techniques for soil leachate and vadose zone sampling in permeable

468

Chapter 7.3

proﬁles (developed during 1975–1990) are required for this purpose, and even then the interpretation of results generated by these techniques may not be straightforward due to the potential complication of soil disturbance and preferential vadose zone ﬂow. The favoured solution as regards routine monitoring is to put much more emphasis on sampling groundwater close to the water table in the ‘‘recharge zone’’ (Figure 7.3.3), which is a much more sensitive indicator of incipient pollution. Both exist to a varying degree (and based on variable criteria) in many EU countries, and clear beneﬁts should accrue from a harmonisation of approaches. Such ‘‘protected areas’’ and ‘‘safeguard zones’’ could equally apply to important groundwater-dependent ecosystems (especially EU Natura 2000 sites), with the caveat that there is likely to be a much higher level of uncertainty in what will constitute a ‘‘signiﬁcant impact’’ in terms of groundwater contamination. Groundwater management measures involving the establishment of protection zones are of necessity long term, in view of the associated investment in water supply infrastructure, and thus need to be based on stable ‘‘groundwaterfriendly’’ agricultural cultivation regimes. But complicating factors in this regard include: the long ‘‘response times’’ of many aquifers, which mean that the full beneﬁt of control measures on agricultural activities will only be obtained in the long run and that the requirement of the EC Water Framework Directive (2000) of reversing negative trends in groundwater chemistry by 2015 will not be feasible for groundwater bodies of large storage (although this must not be accepted as a reason for not introducing appropriate control measures); and

Figure 7.3.3

Detection of groundwater quality trends in aquifer recharge by sampling installations in the zone of water table ﬂuctuation.

Diﬀuse Groundwater Quality Impacts from Agricultural Land-use

469

the need to stimulate ‘‘adaptive capacity’’ in the water sectors to cope with the impacts of accelerated climate change on agricultural cropping and in turn on groundwater recharge and quality. If adequate groundwater quality management measures are not (or cannot be) widely introduced, then the implication is that advanced treatment will be needed to secure the quality of public drinking-water sources and that wetland ecosystems will not be adequately protected. In reality water supply treatment should not be thought of as an alternative to source protection, but as a complementary measure to ensure drinking water quality depending upon the level of protection that can be guaranteed for raw water quality. Certain other processes may also need to be taken into consideration to get a ‘‘fully rounded’’ picture of the diﬀuse groundwater pollution pressure arising from essentially rural areas: non-agricultural use of pesticides (usually at high rates) for railway, highway and airﬁeld defoliation; eﬀects of aerial fallout to agricultural land from industrial process (past and present) and the potential for soil accumulation and eventual leaching of certain types of synthetic organic pollutants; and irrigation of agricultural land with treated urban wastewater and/or the land application of sewage sludge, both of which could under some conditions result in the leaching of synthetic organic chemicals (such as industrial chemicals, cleaning solvents, antibiotics, hormonal compounds, disinfectants, fragrances, etc.).

Acknowledgements The opinions and judgements expressed in this paper are those of the authors alone, but they wish to acknowledge the encouragement of many groundwater and agricultural specialists in England and Spain with whom they have worked over the past 30 years, and interest in the subject has been further stimulated recently by discussion with Dr Philippe Quevauviller (EC-DGE) and with various colleagues from the British Geological Survey at Wallingford (notably John Chilton) and from the Universidad Polite´cnica de Catalun˜a.

References 1. MAPA, Plan Nacional de Regadı´ os—horizonte 2008, Ministerio de Agricultura, Pesca y Alimentacio´n, Madrid, 2001. 2. MIMAN, Jucar Pilot River Basin, provisional Article 5 report pursuant to the EC Water Framework Directive, Ministerio de Medio Ambiente/ Confederacio´n Hidrogra´ﬁca de Ju´car Report, Valencia, 2004. 3. S. S. D. Foster, A. C. Cripps and A. K. Smith-Carington, Nitrate leaching to groundwater, Philos. Trans. R. Soc. London, Ser. B, 1982, 296, 477–489.

470

Chapter 7.3

4. J. Guimera, O. Marfa, L. Candela and L. Serrano, Agricult. Ecosyst. Environ., 1995, 56, 121–135. 5. L. Candela, Hydrogeology, 2000, 3, 85–91. 6. E. Sanchı´ s, Estudio de contaminacio´n por nitratos de las aguas subterra´neas de la provincia de Valencia, Editorial Graﬁcuatre, Valencia, 1991. 7. S. S. D. Foster, P. J. Chilton and M. E. Stuart, J. Inst. Water Environ. Manag., 1991, 5, 186–193. 8. J. Tindall and B. Kunkel, Unsaturated Zone Hydrology, Prentice Hall, New York, 1999. 9. S. S. D. Foster, Quart. J. Eng. Geol. Hydrogeol., 2000, 33, 263–280. 10. D. C. Goody, M. E. Stuart, D. J. Lapworth, P. J. Chilton, S. Bishop, G. Cachandt, M. Knapp and T. Pearson, Quart. J. Eng. Geol. Hydrogeol., 2005, 38, 53–63. 11. P. J. Chilton, M. E. Stuart, D. C. Goody, R. J. Williams and A. C. Johnson, Quart. J. Eng. Geol. Hydrogeol., 2005, 38, 65–81. 12. J. Beltra´n, F. Herna´ndez, I. Morell, P. Navarrete and E. Aroca, Sci. Total Environ., 1993, 132, 243–257. 13. J. Caballero, L. Candela and M. T. Condesso, Field and analytical work undertaken in the Maresme and Canary Islands areas for determination of priority pesticides in aquifers, EC-DGXII Water Pollution Research Report 31, Brussels, 1995, pp. 67–70. 14. S. F. Korom, Water Resour. Res., 1992, 28, 1657–1668. 15. K. Beran and P. Germann, Water Resour. Res., 1982, 18, 1311–1325. 16. P. J. Chilton and S. D. Foster, Control of groundwater nitrate pollution in Britain by land-use change, NATO-ASI Series G 30, 1992, pp. 333–347. 17. Water4All, Sustainable Groundwater Management: Handbook of Best Practice to Reduce Agricultural Impacts on Groundwater Quality, EU Interreg IIIB North Sea Programme Publication, OOWV, Oldenburg/AK-Print, Aalborg, 2005. 18. M. Silgram, A. Williams, R. Waring, I. Neumann, A. Hughes, M. Mansour and T. Besien, Quart. J. Eng. Geol. Hydrogeol., 2005, 38, 117–127. 19. A. P. Whitmore, N. J. Bradbury and P. A., Agricult. Ecosyst. Environ., 1992, 39, 221–233. 20. R. Thomsen and L. Thorling, Trans. Am. Geophys. Union, 2003, 84, 63. 21. EEA, Groundwater Quality and Quantity in Europe: Data and Basic Information, European Environment Agency Technical Report 22, Copenhagen, 1999. 22. I. Heinz, Economic analyses concerning the EC Drinking Water Directive (80/778/EEC): the parameters for pesticides and related products, EC-DGXI, unpublished report, Brussels, 1995. 23. P. Sorensen, B. Mogensen, A. Gyldenkaerne and A. Rasmussen, Chemosphere, 1998, 36, 2251–2276. 24. UK-WIR, The cost of groundwater quality deterioration and tighter standards, UK Water Ind. Res. News, 2004, 33, 1. 25. S. S. D. Foster, Evaluating Diﬀuse Groundwater Pollution Threats: A Key Challenge for the New EC Water Directives, IAHS Publication 297, 2005, pp. 1–8.

8. Integrated River Basin Management

CHAPTER 8.1

Integrated Management Principles for Groundwater in the WFD Contextw PHILIPPE QUEVAUVILLER European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium

8.1.1

Introduction

As highlighted in Chapter 1, the complex nature of groundwater calls for emerging groundwater problems to be addressed through integrated management approaches designed to change the way people view and use the resource.1 This involves an appreciation of three eﬀective levels of integration: integration within the hydrologic cycle (the physical process) including the temporal dimension; integration across river basins and aquifers (spatial integration); and integration of socioeconomic considerations at regional, national and international levels. A fourth level of integration concerns the way scientiﬁc knowledge is used (see Chapters 2.1 and 11.3). The great natural variation in groundwater conditions obviously aﬀects actual management needs and options as well as perceptions. In many cases, variations between areas necessitate highly localised approaches to management; in other cases, regionally based approaches are suﬃcient.1 The management scale therefore needs to be carefully evaluated against available policy choices and options in each particular setting. This is not as straightforward as, for example, ‘‘surface’’ river basin management, since boundary conditions, three-dimensional structure and temporal responses are much more complex for aquifer systems. This being said, the general principles of integrated water resource management provide a basis for addressing groundwater management in the context of a strategic framework. In this chapter, focus will be on Water Framework w

The views expressed in this chapter are purely those of the author and may not in any circumstances be regarded as stating an oﬃcial position of the European Commission.

473

474

Chapter 8.1

Directive (WFD) River Basin Management Planning (RBMP) requirements which basically covers (1) objectives that management planning are designed to address (namely ‘‘good status’’ objectives to be achieved by 2015); (2) the way diverse types of measures ﬁt into the overall management planning (WFD programme of measures); and (3) the criteria against which the success or failure of speciﬁc strategies or interventions can be evaluated (e.g. compliance with environmental quality standards). A management planning framework is to be conceived as a ‘‘living’’ or iterated document that can be updated, reﬁned and if necessary changed as information and experience are gained. This is exactly the purpose of the RBMP of the WFD and its related Groundwater Daughter Directive (see Chapter 3.1) that are further discussed in the present chapter along with general principles of integrated management which are inspired by outputs from of a European Union (EU)-funded project, the EUWATERMAN project.3

8.1.2

Challenges Linked to Groundwater Management

The complexity of development of integrated approaches for groundwater management is that initiatives may lead to the production of massive data collection and planning eﬀorts that are out of date before they are completed. As a result the need for integration must be balanced against practical limitations and the importance of immediate action to address speciﬁc problems. In sum, systemic perspectives and adaptive management approaches will be more eﬀective than a ‘‘fully integrated’’ or ‘‘comprehensive’’ approach.1 A systemic perspective requires a strong conceptual understanding of the natural groundwater conditions but should also cover a broad array of physical, social, economic and institutional factors aﬀecting water management needs and options. This is very demanding on the authorities charged with making groundwater management eﬀective. Such institutions require strong and stable institutional arrangements along with the necessary legal authority to make and enforce policy. They need to be knowledge-driven with broad access to data and information, and they require personnel capable of developing a broad interdisciplinary understanding of water management issues.1 As described in Chapter 3.1 and below, these principles are underpinned by the EU groundwater policy system. Another challenge is the need for ﬂexibility. Because social, economic and hydrological systems are dynamic rather than static and the factors directly or indirectly aﬀecting groundwater conditions vary greatly among regions, management approaches must be ﬂexible. This implies the development of a policy framework that enables management approaches to be tailored to reﬂect the social, economic and physical resource conditions prevailing in diﬀerent management areas.2 National frameworks that attempt to specify management details, e.g. spacing of wells, speciﬁc prices for water, will often mandate approaches that are inappropriate or unworkable at the local level. In contrast, policy frameworks that focus on broad principles and provide administrative or

Integrated Management Principles for Groundwater in the WFD Context

475

legal mechanisms enabling local managers to work out details are likely to be more workable and eﬃcient.1 This also implies that eﬀective participatory planning involves specialists and decision-makers in groundwater management. The complexity of management tends to increase with scale, group size and condition variability. Management activities carried out at the smallest scale and lowest administrative level at which they can eﬀectively be carried out are therefore easier to tackle.1 This principle has, however, to be balanced against the institutional importance of resource boundaries reﬂecting the physical scale at which groundwater systems function and, in this respect, clear management units are as important for the development of eﬀective management institutions as they are for scientiﬁc understanding (the river basin or water body levels in the case of the WFD; see Chapter 3.1). This being said, it should be clear that local management initiatives need to reﬂect aquifer dynamics and ﬁt within a management framework that recognises the aquifer (i.e. the groundwater body as deﬁned by the WFD) as the primary unit for management of the resource. Ideally, for large aquifer systems, a single institutional entity would have responsibility for developing the overall aquifer management framework and ensuring that local approaches are consistent.

8.1.3

River Basin Management Principles

8.1.3.1

Water and Its Environment

The river basin constitutes the basis of river basin management (RBM) as conceived within the WFD (Article 3, River Basin Districts). It can be deﬁned as the geographical area determined by the watershed limits of the system of waters, including surface water and groundwater (see WFD deﬁnition). Strong interactions exist between groundwater and surface water in the basin, between water quantity and quality and between land and water, upstream and downstream, which turn river basins from a geographical area into a coherent system. River basins are open systems with sometimes ill-deﬁned boundaries regarding aquifers, i.e. watershed limits often do not correspond exactly with aquifer limits. Moreover, river basins interact continuously with the atmosphere (precipitation and evaporation, airborne pollution) and the receiving waters (seas and sometimes lakes). River basins fulﬁl many important functions, such as water supply for households, industry and agriculture, navigation, ﬁshing, recreation and ‘‘living space.’’ Economic and social development and even life itself cannot be sustained without suﬃcient water at the right time and place and of the right quality.3 In addition, with respect to surface water, water has shaped and continues to shape the environment, eroding mountain areas, transporting sediment and creating delta areas. It is an essential element of nature while being subject to variability caused by human activities or natural causes, e.g. climate change, which can lead to ﬂoods or droughts. RBM has to tackle all these issues, i.e. it is much broader than traditional water management and it includes land-use planning, agricultural policy and

476

Chapter 8.1

erosion control, environmental management and other policy areas.3 It also covers all human activities that use or aﬀect fresh- and groundwater systems.

8.1.3.2

River Basin Management Objectives

In the light of sustainable developmentz requirements, RBM can be conceived as ensuring the multifunctional use of rivers and their basins for the present and future generations.3 Since the capacity of river basins to accommodate diﬀerent uses is always limited, priorities have to be set, in particular basic human needs have to be safeguarded, i.e. water supply for drinking and basic hygiene which is a component of the WFD protection framework (Article 7) and the new Groundwater Directive (see Chapter 3.1). Moreover, environment protection should be given a full place in RBM; this is actually a key feature of the WFD RBMP which is essentially directed towards achieving environmental objectives of ‘‘good status’’ by 2015. Apart from that, priorities also depend on the natural, social and economic conditions in the pertinent basin. Four diﬀerent management levels can be distinguished according to Mostert et al.:3 operational management, planning, analytical support and the institutional framework. Only operational management aﬀects river basins directly. Planning (including policy formulation and implementation) is a means to improve and support operational management. Analytical tools support both planning and operational management. All three take place in and are inﬂuenced by the legal and institutional framework, which is therefore the fourth level in RBM.3 The following sections provide more details about these components, placing them in the WFD context.

8.1.4

Operational Management

Operational management may have a direct impact on the river basin by means of river regulation, constructing and operating water supply infrastructure, reforestation projects, aquifer artiﬁcial recharge, etc. In the context of the WFD, this operational management is closely linked to the required programmes of measures, which is itself linked to the eﬀective implementation of a series of parent directives listed in its Annex VI, e.g. emission controls of agricultural or industrial pollutants, abstraction controls, codes of good practices (e.g. Best Available Technologies, Best Environmental Practices), construction and/or rehabilitation projects and desalination plants. RBM may also address the behaviour of the diﬀerent users or managers by explicitly forbidding, regulating or allowing certain activities (legislative or administrative instruments), by oﬀering economic (dis)incentives (economic or ﬁscal instruments) and by providing information (see Annex VI of the WFD). Diﬀerent resources are necessary to apply these instruments, such as ﬁnancial, personnel, legal resources, appropriate policy directives and data.3 Table 8.1.1 gives an z

Development meeting the needs of the present without compromising the ability of future generations to meet their own needs.

Integrated Management Principles for Groundwater in the WFD Context

Table 8.1.1

477

Types of operational RBM instruments for groundwater management (adapted from Ref. 3).

Type

Characteristics

Instruments

Concrete activities

Direct interference by the managers in the river basin

Regulation

Inﬂuencing other managers or users by means of forbidding activities or explicitly allowing them Inﬂuencing other managers or users by means of ﬁnancial (dis)incentives; market mechanisms

Aquifer protection Regulation of groundwater abstraction Artiﬁcial recharge, etc. Standard setting Permitting Compliance monitoring Sanctioning Charges (taxes, levies, etc.) Subsidies (ﬁnancial contribution, etc.) Tradable water use and pollution prevention and limitation Public information Non-binding plans Voluntary agreements Technical guidance Charges (taxes, levies, etc.) General taxes Staﬀ training Legislation Planning, etc.

Economic instruments

Communication and awareness raising

Inﬂuencing other managers or users by providing information

Financing

Supporting the previous instruments by providing the necessary ﬁnances Supporting the previous instruments by providing the other necessary resources (personnel, legal competencies, policy directives)

Capacity building

overview of the diﬀerent types of operational RBM instruments with respect to groundwater. As stressed in the introduction, the natural variability of groundwater implies that appropriate RBM systems will require a mix of instruments, depending upon local or regional conditions. A number of current issues are discussed below.

8.1.4.1

Pollution Control

In a sustainable world, pollution control would be limited, i.e. emissions of pollutants would be close to zero. The main issue is how to approach the ideal and how to solve urgent pollution problems and ensure that further pollution risks are prevented or limited. As discussed in Chapter 3.1, the WFD and its daughter directive rely on parent legislation regarding programmes of measures for preventing or limiting inputs of pollutants into groundwater. In particular, the legislative framework requires that point and diﬀuse sources of pollution are controlled through a combined approach based on emission controls using best available techniques, relevant emission values or best environmental practices

478

Chapter 8.1

Table 8.1.2

Types of plans and policies relevant to river basin management (adapted from Ref. 3).

Strategic vs. operational character

| |

|

Policy scope

Geographical scope

Character Validity Time horizon Detail Planning process

| | | | | | | | | | | | | | | |

Setting aims and goals (policy plans, policies) Setting short- or medium-term targets, strategies and/or speciﬁc guidelines for operational management (operational plans,strategies) Setting prioritising and/or specifying, scheduling and ﬁnancing operational activities (programmes) Only surface or groundwater, quantity or quality Only some uses (e.g. hydropower) All water, all uses Water and land Whole basin Sub-basin(s) Part(s) of sub-basin(s) Administrative area National/international Purely informative, politically binding or legally binding Short, medium, long term Short, medium, long term Global plan, detailed plan Top-down closed planning Top-down participatory planning Bottom-up planning

(in the case of diﬀuse pollution) which are set out in relevant legislation (dealing with industrial, urban or agricultural sources of pollution). This is complemented by a water quality approach based on the establishment and compliance to groundwater quality standards, and the requirement to identify and reverse any statistically and environmentally signiﬁcant pollution trends.4 Table 8.1.2 highlights strengths and weaknesses of diﬀerent approaches. There is no universally best approach: it depends on factors such as the urgency of pollution problems, the substance concerned, the pollution source, the capacity of the managers, etc. In practice, the diﬀerent approaches are often combined, e.g. minimum uniform emission standards combined with more stringent pollution controls if the water quality so requires.3

8.1.4.2

Voluntary Agreements

A great concern in all regulatory instruments is enforcement. Personnel and equipment are often insuﬃcient for frequent monitoring, sometimes the diﬀerent bodies responsible for enforcement may not co-operate eﬀectively and

Integrated Management Principles for Groundwater in the WFD Context

479

political forces and lobbying may prevent strict sanctioning.3 Voluntary agreements and other communicative instruments may oﬀer a partial solution as foreseen in the WFD and the new Groundwater Directive through the term ‘‘environmental negotiated agreement’’ (see Annex VI, part B of the WFD, and recital 23 of the Groundwater Directive), in particular with regard to agricultural activities. They are based on the co-operation of the (ground)water users or polluters: the latter are not forced but persuaded to do or not to do something. In this context, users and polluters may be willing to agree on quite ambitious goals, which may go beyond regulatory incentives (e.g. the legal framework of the WFD and Groundwater Directive acting as mandatory requirement). Voluntary agreements require a high degree of organisation of the diﬀerent user groups. Finally, it should be clear what happens if individual users or whole user groups do not comply with a voluntary agreement, and this is where obviously regulatory instruments remain necessary.3 This pattern concerns not only the groundwater regulation but also all the parent directives which have to be eﬀectively implemented to enable the WFD environmental objectives to be achieved.

8.1.4.3

Cost Recovery

Another operational issue is related to recovery of costs of water services as required under Article 9 of the WFD. This principle takes into account the polluter pays principle and requires EU member states to establish water-pricing policies by 2010, ﬁxing adequate contributions of the diﬀerent water uses, disaggregated into at least industry, households and agriculture. This policy depends on the price elasticity (the sensitivity of water use/pollution to the costs of the user/polluter), which is generally low in the case of drinking water use and high in the case of irrigated agriculture (the major water user in many countries).3 Charges that reﬂect the full economic and environmental costs of water use and pollution are economically eﬃcient since they confront the water user/ polluter with the real costs and promote an integral assessment of the costs and beneﬁts. Moreover, they solve the ﬁnancing problems of the providers of the water service concerned. However, this principle has to consider social, environmental and economic eﬀects, as well as geographic and climatic conditions of the region or regions aﬀected, and the WFD opens the possibility for not fully following the cost recovery principle (opening the possibility of funding of particular preventive or remedial measures), providing that the achievement of the ‘‘good status’’ objectives are not compromised. This might be the case, for example, if water would become too expensive for poor populations. Indeed, very high charges and especially rapid increases may decrease the willingness to pay and may result in massive political opposition.3

8.1.4.4

Institutional Structure

Mostert et al.3 illustrate diﬀerent instruments for operational management that are applied in an institutional structure which consists of formal and informal

480

Chapter 8.1

working rules. Operational rules provide a framework for operational management, e.g. emission standards and policy directives (such as the WFD and the new Groundwater Directive). Collective choice rules deal with how operational rules should be developed, e.g. permitting and planning procedures embedded into the WFD RBMP. Constitutional rules determine who is entitled to make collective choice rules, setting up the organisational structure for RBM and allocate tasks and competencies (in the context of the WFD, these are the river basin district authorities). In this context, Mostert et al.3 distinguish three basic RBM models. The hydrological model in which the organisational structure for water management is based on hydrological boundaries. In its extreme form all water management is in the hands of a single entity: the ‘‘river basin authority’’. This is actually the case of the WFD. The administrative model is in many respects the opposite of the hydrological model. In this model water management is the responsibility of provinces, municipalities and other bodies not based on hydrological boundaries. The co-ordinated model falls somewhere between the hydrological and the administrative model. In this model water management is not performed by river basin authorities, but there are river basin commissions with a co-ordinating task. Each model has advantages and disadvantages. In the hydrological model, administrative procedures coincide with hydrological boundaries and there is the least chance of upstream–downstream conﬂicts. However, since river basin authorities usually deal with water management only, this model may isolate water management from other relevant policy sectors, and intersectoral coordination may become a problem. Note, however, that such co-ordination is actually required in the context of the WFD RBMP. In the administrative model water management, land-use planning and other relevant policy sectors can be kept together (but not necessarily). A major disadvantage is the serious risk of upstream–downstream conﬂicts and the lack of a platform to discuss these problems. An example of co-ordinated model is the river basin commissions. The diﬀerent bodies participating in these commissions could each individually ensure co-ordination between water management and other policy sectors, and together, in the commission, they could co-ordinate their water management.

8.1.4.5

Infrastructure vs. Regulation, Financing and Empowerment

It seems obvious to say that it may not be advisable to combine responsibilities for constructing and operating infrastructure with regulatory responsibilities, i.e. meaning that managers would have to control their own operations as managers themselves should be strictly controlled.3 Secondly, tasks, competencies and

Integrated Management Principles for Groundwater in the WFD Context

481

ﬁnancing should go hand in hand. Without suﬃcient competencies and ﬁnancing, managers cannot perform their tasks properly. Either their competencies and ﬁnancing should increase or, if that is not possible, their tasks have to be redeﬁned. Thirdly, the institutional structure for RBM is also a means of empowerment: it gives or denies legal rights and inﬂuences the balance of power between the diﬀerent parties involved. For RBM to be eﬀective and equitable, it is important that the diﬀerent water users, local communities, environmental NGOs (non-governmental organisations), etc., can participate actively in RBM.3

8.1.4.6

Decentralisation

RBM is closely linked to decentralisation, i.e. government authorities are brought as close as possible to individual citizens, allowing for local variation in response to local circumstances and preferences for the notion of ‘‘subsidiarity’’ (a principle that is fully embedded into the EU Treaty). This is also more eﬃcient as decentralised government tends to be less bureaucratic— simply because of its size—and better informed about local circumstances.3 Decentralisation is not possible, however, for tasks such as establishing the institutional structure and formulating policies that apply to a country as a whole. However, decentralised governments should be involved because of their superior information on local conditions and because of their (usually) closer contacts with the population. Decentralisation may also not be possible if the decentralised governments lack the necessary management capacity. Solutions could include local capacity building and advisory services by specialised central governments.3

8.1.4.7

Privatisation

Privatisation is only possible for speciﬁc services such as the construction and operation of water supply and wastewater treatment infrastructure—not for regulatory functions and policy making. Diﬀerent forms exist, e.g. contracting out speciﬁc activities such as construction, billing and laboratory services. Privatisation may also imply that private ﬁrms get a contract to operate and/or maintain the infrastructure for a speciﬁc water service. The infrastructure remains in the hands of government, and after expiration of the contract another ﬁrm may be contracted. The opposite also occurs: government is responsible for operation and maintenance, but for budgetary reasons it does not own the infrastructure but rents it from a private ﬁrm. Finally, private ﬁrms may own, operate and maintain the infrastructure. A midway option between the public and private provision of water services is the provision by publicly owned private companies. These are not dependent on the government budget, can apply commercial double-entry bookkeeping, and can take out the necessary loans. In theory they can function as eﬃciently as private ﬁrms. Since their stakeholders are public authorities, the incentive to misuse a monopoly position and save on environmental measures may be smaller.3

482

8.1.5

Chapter 8.1

Planning

Whereas operational RBM constitutes the functional core of RBM, planning linked to policies (such as the WFD) has an important supportive role to play. As important as the plans and policies themselves is the way in which they are prepared: the ‘‘planning process’’ is a means to improve and support operational management. This section is oriented towards policy formulation and implementation in line with the WFD RBMP required under Article 13 and related information required under Annex VII of the directive.

8.1.5.1

Functions of Plans and Policies

Plans and policies can support operational RBM in several ways.3 Firstly, planning helps to assess the present situation in the basin, which is the basic aim of the analysis of pressures and impacts and economic analysis required under Article 5 of the WFD, and measures required to meet predeﬁned targets (e.g. the WFD environmental objectives). It helps to orient operational management and set priorities. Secondly, it is often hardly possible to carry out policy analysis and organise public participation for each individual operational decision, and planning may provide the necessary framework. Thirdly, open and participatory planning processes may result in more public support or acceptance of the resulting plan/policy and (by extension) operational management. Fourthly, plans and planning may have a co-ordinating eﬀect, i.e. bringing diﬀerent river basin managers into discussion with each other with resulting plans and policies acting as common focal points.

8.1.5.2

The Planning Process

Planning requires extensive technical and scientiﬁc information, but it can never be purely technical or scientiﬁc. The following steps are highlighted by Mostert et al.3 1. Identiﬁcation of planning needs, possibly involving some preliminary research. 2. Analysis of the institutional RBM framework and identiﬁcation of the diﬀerent operational decisions that can be taken, the bodies responsible for these decisions and their management capacity. 3. Identiﬁcation of the other stakeholders and their main interests. 4. Preparation of a process design, describing the scope of the planning exercise; the diﬀerent phases; the diﬀerent groups to be involved in each phase and the means to do so; the necessary research in each phase; and the project organisation. 5. Implementation of the process design, resulting in the adoption of a plan. 6. Implementation of the plan. After a while the plan and its implementation can be evaluated, and the process can start again. This is the way the RBMP of the WFD is established, with a

Integrated Management Principles for Groundwater in the WFD Context

483

ﬁrst cycle planned for the period 2009–2015, followed by a review (taking into account scientiﬁc progress) and a second six-year cycle, which is a continuous and iterative process.

8.1.5.3

Planning Systems

Plans and policies relevant to RBM can diﬀer on many dimensions: policy sectors, geographical scope, available funding, etc. (Table 8.1.3).3 These diﬀer from country to country and from basin to basin, but still some general guidelines may be given. River basin planning should consider diﬀerent interrelations within water systems (surface water and groundwater quantity and quality), the basin characteristics and their socioeconomic environment. This does not mean that each individual plan should have such a broad scope. Rather, the thinking should be in terms of planning systems:3 sets of interrelated types of planning, consisting of strategic and operational plans (e.g. linked to diﬀerent regulatory frameworks concerning industrial, urban or agricultural activities). The more strategic a plan is, the more important it is that it covers complete river basins and all relevant policy sectors. Operational plans have to go more into detail and usually cover only one policy sector or part of a sector. The types of plans will depend on speciﬁc features, e.g. if in a speciﬁc basin there is one very urgent, very obvious issue, such as pollution of drinking water sources, there may be no need for integrated strategic planning that provides a complete integrated description of the basin and sets long-term goals. The resources could much better be used for making and implementing an operational plan that sets speciﬁc and concrete targets, proposes operational measures, and creates the necessary support linked to the speciﬁc feature. Generally speaking, plans should be designed, taking into consideration the management capacity of the countries and basins. Too much planning might be diﬃcult if too few resources are available for each planning exercise, coordination between the plans can become problematic and transparency for the citizen is reduced. Moreover, resources that are spent on planning cannot be spent on operational management.3

8.1.6

Analytical Support

River basin management is a complex task. Therefore, tools helping to assess the present situation and assist the development and evaluation of solutions are important. Two types of support may be distinguished:3 (1) support to operational management (e.g. the WFD programme of measures) and (2) support to strategic policy-making and planning (e.g. the RBPM cycles). A second distinction is between (support) systems for monitoring, data collection and processing, oriented towards making facts and ﬁgures about the present situation and about possible trends; and tools or systems to support decisionmaking with a view to the future, typically oriented to the ex ante identiﬁcation,

484

Table 8.1.3

Chapter 8.1

Regulatory approaches to pollution control related to groundwater (adapted from Ref. 3).

Approach

Description

Strength

Weakness

Product policy

| Explicit forbidding or allowing the use of speciﬁc products

| Administrative costs and demands and enforcement can be low (depending on product) | Appropriate for diﬀuse pollution (e.g. pesticides) | Standards concerning environmental performance are steps towards ultimate aim of closing substance cycles

| Production process often more important than product

| Quality standards for product Process standards

Uniform emission standards

Water quality approach

| Prescribing speciﬁc processes

| Standards concerning the environmental performance of production processes | Emission standards applying to all emission in a certain area (usually country) | Emission standards, usually set in a permit, that reﬂect the quality of the receiving water in relation to the prevalent quality standard

Solution for reducing emissions of dangerous substances quickly to zero | Most eﬃcient: pollution reduction eﬀorts concentrated where need is largest

| Administrative costs and demands on capacity of managers are high | Prescribing speciﬁc processes may hinder the application of cleaner technologies | Too strict in some cases, too lenient in others | May promote end-of-pipe solutions | Diﬃcult to link emissions to water quality; many data needed (water quality, emissions, transport and persistence, dilution, etc.) and good models | Administrative costs and demands on capacity of managers are high | Not appropriate for pollutants accumulating in the environment; diﬃcult to deal

Integrated Management Principles for Groundwater in the WFD Context Table 8.1.3 (continued ) Approach

Description

Strength

485

Weakness with synergetic eﬀects | Unequal treatment of polluters in diﬀerent basins and between diﬀerent parts of the basin (upstream– downstream, tributaries) | Diﬃcult to ensure minimum level of pollution control

analysis and evaluation of alternative allocations, policies or plans. These distinctions are not absolute. Operational management and strategic policymaking interact, and data collection and ex ante analysis support each other. The development of information and computer technology over the last 30 years has enabled the design and application of a wide array of systems and modelling tools for supporting water managers. Most eﬀorts in the ﬁeld have so far concentrated on the technical and physical aspects of the (physical) river systems itself, and little attention has been paid to the development of systems and tools covering relevant aspects and processes in the river basin as a whole.3 This can be explained by the fact that broader notions of integrated water are relatively recent, and by the complexity of monitoring and analysing the interaction between natural and socioeconomic systems at the scale of a river basin. The situation is changing, however, with new research trends and development of multidisciplinary synergies.

8.1.6.1

Analytical Support for Operational Management: Main Challenges

As described in several chapters of this book, many analytical tools have become available to support operational management. With respect to groundwater, eﬀorts remain necessary to harmonise monitoring and analysis methods used by diﬀerent organisations, especially in international basins. A second challenge is to make the information available to anybody involved or interested. Development in database technology, often in combination with internet applications, can provide powerful tools for data retrieval and map visualisation, which is the case of the Water Information System for Europe (WISE) for EU water policies (see Chapter 11.3).

486

Chapter 8.1

A more advanced type of operational support is to combine on-line monitoring with computer models in order to predict future conditions of the system. Examples are early warning systems, both for water quantity issues (ﬂoods, droughts) and for water quality issues (accidental spills). Flood early warning systems are already installed in many major basins in the world. An even more advanced form of support is the automation of infrastructure operation, such as weirs, pumps and sluices. In most cases such tools do not replace human operators: the tools provide the necessary information, but the operators decide. This information is generated using monitoring data, often combined with computer models that describe the behaviour of the natural system (water levels, discharges, etc.). The main challenge is to develop support systems that describe not only the natural system but also the use functions related to this system, thus enabling a weighing of all aspects involved.3

8.1.6.2

Analytical Support and the Strategic Level: New Directions

At the level of strategic planning and policy-making, eﬀorts so far are mainly related to the development of speciﬁc tools for speciﬁc problems in speciﬁc river basins, e.g. options for managing and cleaning up heavy metal pollution in a given groundwater body. Challenges for developing more generic and comprehensive tools at the river basin level are enormous as there is a lack of data and theories that may fully describe complex processes taking place in a groundwater body or groups of groundwater bodies within a river basin, taking socioeconomic issues into consideration. This does not allow one to include all relevant issues in a single model or tool. Yet, given the crucial importance and complexity of management at the basin level, it is of utmost importance that investments are made in the further development of analytical approaches and associated tools. Some possible tool development orientations are highlighted by Mostert et al.3 Tools for supporting integrated management and analyses at the river basin level describing not only the diﬀerent aspects (quantity and quality) of the physical system, but also interactions with the socioeconomic system. Tools facilitating the linkage of (aggregated) strategies at the basin level and strategies at the regional and local levels to take account of processes and implementation aspects that have a regional rather than a basin-wide character. The challenge is to develop a family of tools operating at diﬀerent geographical scales and levels of aggregation, linked to each other for overall consistency. Tools or models describing the costs and beneﬁts of speciﬁc actions to the various actors involved, also helping to explore the possibilities for exchanges between actors, to assess the need to involve other actors in the process and possibly to identify potential linkages to other issues that would turn in a win–lose situation into a win–win situation.

Integrated Management Principles for Groundwater in the WFD Context

487

Support systems and tools that are better tuned to the dynamic and increasingly participatory nature of policy processes, i.e. accessible to non-specialists. For interactive learning settings there is a need for more ﬂexible and transparent tools. Alternatives to the traditional tools based on ‘‘objective’’ system analytical approaches should be explored, e.g. striving to distinguish between ‘‘objective’’ knowledge and subjective judgements. Perceptions of problems and solutions are inevitably aﬀected by diﬀerences in interests of participants, and arguments put forward in policy debates typically contain a mixture of ‘‘objective’’ facts and subjective viewpoints or perceptions. Argumentation analysis may be supported by tools speciﬁcally designed to describe, visualise and analyse policy arguments. Another novel approach is to use gaming as a vehicle for learning. In a policy game participants interact as if they were playing the role of diﬀerent parties involved in a real-world issue. Such games can be very instructive to both participants and observers as they include parts of the social and psychological dynamics of real policy processes, which cannot be included in more traditional systems. Policy games are generally supported by computer-based tools that take account of physical and other aspects in the process. New opportunities linked to developments of information and communication technology, e.g. geographical information systems (GIS) and interactive interfaces, allowing use of support tools by a broader group of users, and the development of the internet. This is clearly the aim of WISE which is brieﬂy presented in Chapter 11.3.

8.1.7

International River Basins

8.1.7.1

The Challenge

A special type of RBM is the management of international river basins, which are usually larger than national basins and less homogeneous. Natural and socioeconomic conditions, culture and language often diﬀer signiﬁcantly between the diﬀerent parts of the basin, and consequently upstream–downstream conﬂicts may occur. More importantly, however, international basins are by deﬁnition located in diﬀerent states. Consequently, international co-operation is needed in order to best manage the basin and prevent or solve upstream– downstream conﬂicts. This co-operation can be made more eﬀective when required by law, e.g. the requirement for coordination is explicitly embedded into the WFD RBMP (paragraph 2 of Article 13 of the WFD). A major problem in the management of international basins is the so-called ‘‘lowest common denominator’’. Few obligations can be imposed on countries without their own consent in the absence of an international regulatory framework such as the WFD (which imposes coordination towards the achievement of common environmental objectives). In the absence of such

488

Chapter 8.1

international law, many international agreements simply reﬂect the commonalities in the national policies of the states concerned or are very procedural and vague.3

8.1.7.2

International Basins at the Global Level

Considering that international river basins constitute some 47% of the earth’s land area (nearly 60% for Africa and Latin America, all excluding Antarctica), one might be inclined to conclude that the objectives and actions to be taken for managing international river basins are now agreed upon and clearly deﬁned.3 Moreover, one might suppose that such agreement would be reﬂected in recent international treaties. However, while important steps towards the codiﬁcation and development of international water management law have been taken, they have mainly taken place at the regional level and, in some cases, at the river basin level.5 At the global level the normative system for the management of international river basins focuses on the discretion of states and their sovereignty, rather than on their particular responsibilities in the process towards attaining sustainable water management, even if cooperation among those states is encouraged in conformity with existing agreements. Compliance regimes have now been included or are being developed in most multilateral environmental agreements, e.g. a procedure that entails that, at the request of a state, the commission coordinates negotiations among the parties and makes recommendations for an equitable solution to the dispute. While these recommendations are not binding in law, the parties to the dispute are to consider them in good faith. Such a procedure remains short of the compliance regimes included in multilateral environmental agreements in that it does not provide an automatic peer review system.3 It may, however, provide a mechanism through which the normative content of the international regime for river basin management may be enhanced. The past and present dialogue on sustainable water management taking place under the auspices of the EU may provide a valuable model for further developments at the international level.

8.1.7.3

International River Basin Organisations

A major aspect of many river basin treaties is the establishment of a river basin organisation. Two types of national river basin organisations have been mentioned earlier in this chapter: river basin commissions with a primarily coordinating task and river basin authorities with decision-making and policy powers. The same type of organisations can be found in relation to international basins.3 River basin commissions may co-ordinate monitoring and research eﬀorts, add legitimacy to the monitoring and research results and in this way provide a common, generally agreed upon factual basis for management. They furthermore oﬀer the basin states a platform for co-ordinating their policy and management. River basin commissions can also prepare RBM plans and programmes, but after adoption by the Commission they still have to be adopted

Integrated Management Principles for Groundwater in the WFD Context

489

by river basin countries or a ‘‘ministers’ conference.’’ River basin commissions may also oversee the implementation of the plans, programmes, but implementation remains the responsibility of the basin countries. Finally, river basin commissions can play a signiﬁcant role in resolving riverrelated international conﬂicts. They constitute a relatively informal forum for discussion, may help in selecting fact ﬁnders and arbitrators or may even do fact-ﬁnding or act as arbitrators themselves. They can be organised in diﬀerent ways, always involving a plenary commission, usually consisting of national representatives and meeting once or twice a year. The work of the plenary commission can be supported by working groups, project groups and/or expert groups. Moreover, the commission may have an independent secretariat, in charge of preparing the meetings, publishing annual reports, giving information to the public, etc. Alternatively, the secretarial function can be performed by one of the states concerned or may alternate between the member states. River basin authorities have decision-making and policing powers. They can independently adopt plans, programmes and bylaws and implement them or enforce their implementation. International river basin authorities usually have a limited scope. The structure of river basin authorities can be the same as that or river basin commissions, with a plenary commission, subgroups and a secretariat.

8.1.7.4

Interbasin Co-operation (Twinning)

A last issue concerning river basin management is co-operation between river basins, both as a form of development co-operation and co-operation between comparable partners. In this respect, besides ﬁnancial support provided by international donors for development assistance, twinning initiatives may lead to long-lasting co-operative relationships between two river basin organisations, involving exchanges of information and experiences.3 Twinning may be focused on assistance or speciﬁc activities, e.g. research projects. Typical activities are short visits including site visits and presentations, and long-term staﬀ exchange aiming at mutual learning with respect to the operational, policy and institutional aspects of RBM.

8.1.8

Public Participation

8.1.8.1

At European Level

The A˚rhus Convention establishes a number of public rights (for citizens and their associations) with regard to the environment. Public authorities (at national, regional or local level) are to contribute to allowing these rights to become eﬀective. In the light of this convention, the WFD includes a speciﬁc article on public information and consultation (Article 14) on the basis of which member states have to encourage the active involvement of all interested parties (including users) in its implementation, in particular in the production, review and updating of the river basin management plans. This enables the

490

Chapter 8.1

general public to have access, on request, to background documents and information used for the development of the draft river basin management plan. This new citizen right is an important step towards integration of socioeconomic awareness into the policy-making (development, implementation and review) process. With regard to groundwater management, four groups stand out and should, as a basic principle, be involved in management initiatives:1 local stakeholders—water users and others whose interests are directly aﬀected by groundwater management and whose actions often determine the eﬀectiveness of any given initiative; policy-makers—those who have the ability to inﬂuence the institutional environment within which management approaches must evolve; public-sector organisations—these stakeholders often have their own internal agendas and control large programmes that either directly or indirectly have major impacts on water resources; and private-sector organisations—these stakeholders are often major water users whose interests may or may not coincide with those of local stakeholders. Stakeholder involvement and education are essential for any attempt to manage groundwater resources. It cannot, however, concern each individual but rather groups representing communities which may have a major impact on the resource (e.g. large water users such as municipalities, agricultural sector) and those whose interests will be signiﬁcantly aﬀected by management regimes (these groups are not mutually exclusive). The principle of stakeholder involvement is to start by being as inclusive as possible. The involvement and education will be all the more eﬃcient if it is linked to a legal base, thus mixing stakeholder organisations with policy makers guiding discussions in relation to policy development, implementation and review needs. This is exactly the case of the CIS Working Group on Groundwater (linked to the implementation of the new Groundwater Directive); see Chapter 4.1. Public participation plays an essential role in planning and policy-making. It can be seen as a legal right of individuals and social groups, often resulting in procedural requirements for decision-making. Public participation can also be seen as a means for empowering individuals and groups and developing local communities. Finally it can be seen as a means of improving the quality and eﬀectiveness of decision-making.3 All three views are equally valid and complement each other. Public participation as a legal right is based on the notion that individuals and groups aﬀected by decisions should have the opportunity to express their views and become involved in decision-making. Often three ‘‘pillars’’ of public participation are identiﬁed: access to information, involvement in the decision-making process (e.g. possibility to comment), and access to justice (right of legal review and redress). The danger of a purely legal approach to public participation is that it may become nothing more than an administrative requirement. Moreover, litigation is often time-consuming and expensive.3

Integrated Management Principles for Groundwater in the WFD Context

8.1.8.2

491

At International Level

River basin commissions generally provide for public access to information through oﬃcial (approved) information (including action plans and programmes), working documents and drafts and ﬁnancial and personnel information.3 The dissemination is more or less diﬃcult according to the cases (restrictions are frequent). The internet has proved to be a useful tool for making information widely accessible. In some cases, information leaﬂets, press releases and communications are made available to the public. When the international legal regime for a basin does not provide for access to information, the public depends exclusively on the speciﬁc national legal systems for access. The second pillar—public participation—appears in diﬀerent forms for international rivers. Examples are the consultation of the public on a (draft) river basin action plan, a role for the public in monitoring conditions of waters (e.g. through the submission of data), the possibility to report to the commission on speciﬁc matters concerning the river basin, and even public involvement in the negotiation of international agreements. A number of river basin commissions have made the meetings of their plenary body and/or subsidiary bodies more open. In most cases the general public cannot participate in the meetings, but sometimes NGOs can have observer status. Their admittance is usually made dependent on speciﬁc criteria of recognition, such as the international character of the NGO concerned (international objectives and goals and international membership). Pragmatic modalities are found in those situations where public participation is not formally, or insuﬃciently, provided for. Such modalities include (1) representatives of NGOs as members or experts in the national delegations; (2) participation of NGOs in national preparatory meetings for the plenary meetings of the joint body and/or its subsidiary bodies; and (3) special consultative meetings with NGOs organised by the river commission. Moreover, several river basin commissions invite NGO members as experts to their meetings. Participation as an observer or expert in meetings of the commission automatically involves access to information which otherwise might not have been disclosed. Provisions on access to justice in international environmental treaties—the third pillar—are not only scarce: they predominantly refer to a national legal system. Often they do not provide minimum rights, but set out the nondiscrimination principle (citizens from other states are to be given equal access to an administrative or judicial review procedure in accordance with the national legal system).3

8.1.9

Conclusions

It is one thing to know how groundwater should be managed at the river basin scale, but another to actually implement RBM principles in an eﬃcient way. RBM involves conﬂicting interests and is therefore ultimately a political process. The main question is how recommendations may be formulated to

492

Chapter 8.1

best ensure this political process. These recommendations will be more inﬂuential if they reﬂect and refer to principles generally agreed upon. However, this is in itself not enough as general principles are open to diﬀerent interpretations, and many statements of principles have only been agreed upon because they have no direct consequences in practice. This is the actual diﬀerence between non-legally binding guidance documents and enforceable provisions contained in the regulations. The value of non-legally binging recommendations should not, however, be underestimated as they often contribute to achieving common understanding on speciﬁc legal provisions which would otherwise remain prone to various interpretations. Recommendations will be more convincing if they are not merely normative (‘‘shall,’’ ‘‘should’’) but also give a rationale (‘‘in order to’’). This rationale should refer to truly generally agreed upon principles, or more profanely, it should appeal to the largest possible number of parties involved in RBM.3 Recommendations will be more inﬂuential if they are the following. They are ‘‘realistic’’. They do not have to be politically feasible in the short term—they can also try to make the unfeasible feasible—but they should start from political realities. Recommendations should show how the practice can be improved rather than just paint an ideal. They should integrate vision and politics, dream and reality. They reﬂect the diﬀering hydrological, socioeconomic and cultural contexts and are technically/scientiﬁcally sound. They should either be truly generally applicable, or clearly distinguish between diﬀerent contexts. They are formulated succinctly and should be concrete, i.e. preferably not containing phrases such as ‘‘as appropriate’’ or ‘‘in some countries/ basins’’ but instead indicate when and in which type of countries and basins something is appropriate. Finally, recommendations should not merely promote a speciﬁc national system of RBM. Not only would this conﬂict with the fourth and possibly the third criterion, but this would also create a lot of unnecessary opposition from countries whose RBM system has not been selected as the ‘‘model.’’ The principles of developing recommendations in the form of guidance documents have been followed under the Common Implementation Strategy (CIS) of the WFD (see Chapter 4.1), and considerably helped the development of the new Groundwater Directive (see Chapter 3.1). They will hence represent a clear support for the development of the WFD RBMP.

References 1. J. J. Burke and M. H. Moench, Groundwater and Society: Resources, Tensions and Opportunities, United Nations, 2000 (ISBN 92-1-104485-5). 2. M. Moench, Approach to groundwater management: to control or enable, Economic and Political Weekly, 1994, 24 Sept., A135–146.

Integrated Management Principles for Groundwater in the WFD Context

493

3. E. Mostert, E. Van Beek, N. W. M. Bouman, E. Hey, H. H. G. Savenije and W. A. H. Thissen, River basin management and planning, Proceedings of the International Workshop on River Basin Management, The Hague, 27–29 October 1999, IHP-V, Technical Documents in Hydrology No. 31, UNESCO, 1999, pp. 24–55. 4. Directive 2006/118/EC of the European Parliament and of the Council of 12 December 2006 on the protection of groundwater against pollution and deterioration, Oﬃcial Journal of the European Communities, L 372, p. 19. 5. S. Burchi and K. Mechlem, Groundwater in international law, FAO Legislative Study, FA0/UNESCO, 86, 2005.

CHAPTER 8.2

System Approach to Environmentally Acceptable Farming RAMON LAPLANAa AND NADINE TURPINb a

CEMAGREF, Unite´ Ader, 50 avenue de Verdun, FR-33612 Cestas, France; b UMR Me´tafort-Cemagref-AgroParisTech-ENITA-INRA, 24 avenue des Landais, BP 50085, FR-63172 Aubiere Ce´dex, France

8.2.1

Introduction

Implementing the European Union (EU) Water Framework Directive requires that regulators at the local scale can choose among all potential measures those that are the most eﬀective for water pollution mitigation. On watersheds, the pollution in water is the result of point and non-point sources. For the latter, the diﬃculty in measuring the individual emissions renders the design of mitigating policies particularly problematic. When the measurement of individual emissions is impossible or too expensive, the literature focuses on two basic options for regulation design. The ﬁrst option bases the regulation on a collective performance variable (for water the ambient concentration of pollutants in environmental media). Following the pioneering work of Segerson and Miceli,1 the set of polluters can be considered like a team whose joint product is the level of pollution observed in the environmental media. The second option bases the regulation on individual variables related to pollution ﬂows that are more or less easy to monitor such as inputs, productions or agricultural practices. The challenge is designing inputbased incentives that achieve environmental goals at reasonable cost.2 Non-point source pollutions from nutrients, pesticides or hazardous substances in groundwater exhibit speciﬁc characteristics such as a long lag between human polluting actions and the resulting pollution level that can be measured in groundwater. Because of this speciﬁcity, the ﬁrst option in the regulation (target-based instruments) will be ineﬃcient, and most regulators turn towards the second option and implementation-based instruments. 494

System Approach to Environmentally Acceptable Farming

495

The ﬁrst framework to design implementation-based instruments assumes that the soils, climate or farm locations are of great importance for both farmers’ proﬁt and pollutant transfers to water. The models resulting from this approach are generally soil management based. They are often combined with a hydrological model to describe water and pollutants transfers through a watershed, and are coupled with a simple economic model of some technically deﬁned types of farms. The coupling of the physical and economic models allows a cost-eﬃciency analysis of diﬀerent policies. This framework led to several important advancements: there are large diﬀerences in emission levels from one farm to another, especially when their productions, like pigs, milk or cereals, diﬀer;3 management decisions interact with soil and climate conditions with signiﬁcant consequences for proﬁt and emissions.4 Thus, a regulating option can be of high value for some types of farms in one given watershed and of no relevance elsewhere.5 All these models consider representative agents. The second framework focuses on the description of the heterogeneity of farms along a watershed or a country. When facing heterogeneous agents, the regulator can choose to apply uniform instruments (these instruments are noncontingent to the heterogeneity of the farms). To mitigate NPS nitrate pollution, these instruments can be taxes on nitrogen fertilisers, quotas on nitrogen fertilisers or uniform reductions of the production level. On the other hand, the regulator can build diﬀerentiated regulations to take into account the agents’ heterogeneity. These regulations will induce self-selecting constraints, for two reasons. First, the agent’s type can be unobservable. Second, even with observable types, institutional constraints prevent ﬁrst-degree discrimination. The following discussion can be considered as guidelines to compare good farming practices at the watershed scale. It focuses on main key elements useful for decision making (best management practices (BMPs) and critical areas deﬁnitions, cost/eﬀectiveness approach, acceptability and integrative analysis grid).

8.2.2

Integrated River Basin Management with BMPs to Mitigate NPS Pollution

In order to achieve a successful application of the concept of integrated water resource management (IWRM), one important step concerns the development of the programme of measures to be applied. To propose directions for programmes of measures, decision-makers need operational tools to compare scenarios and solutions on a cost/eﬀectiveness approach.

8.2.2.1

What are Best Management Practices?

Voluntary approaches have been developed from the early 1990s for environmental regulation in most OECD countries. At this date, the existing

496

Chapter 8.2

instruments for environmental regulation were split into two main categories: command and control instruments were based on direct prescription from the environmental agencies; and economic instruments relied on incentives such as taxes, subsidies and tradable permits. The relative eﬃciency of these instruments is still under debate. In practice, they face implementation diﬃculties: command and control instruments are criticised because of their lack of ﬂexibility and because they induce high costs for policy-makers and for polluters; implementing economic instruments generally leads to political pressure from interest groups that usually results in low levels of the incentives that hamper their eﬃciency. To get round these diﬃculties, while improving the acceptability of the environmental policies, decisionmakers designed voluntary approaches. Voluntary approaches are deﬁned by the OECD as schemes whereby ﬁrms make commitments to improve their environmental performance; they include negotiated agreements, voluntary public programmes and unilateral commitments.6 Integrated river basin management faces populations of actors with diverse objectives and opportunity cost functions, scattered on heterogeneous watersheds, and deals with diﬃculties of measuring the individual contributions to ecological processes subject to discontinuities and threshold eﬀects.7 Mitigating non-point source pollution and restoring water quality require the cooperation of most of the actors on the watershed. Thus, a high level of acceptability of the policies is needed and voluntary approaches may be of interest when they are combined with other instruments.6 The combination with other instruments is of absolute necessity to ensure that the targeted environmental level is higher that what would happen in the absence of regulation.8 In Europe, agri-environment measures (AEMs) have been designed to encourage farmers to protect and enhance the environment on their farmland.9 Agri-environment commitments have to go beyond usual good farming practices (GFPs), which are deﬁned as encompassing mandatory legal requirements and a level of environmental care that a reasonable farmer is expected to apply anyway. AEMs can be designed at national, regional or local level and have at least one of the following objectives: (1) to reduce the environmental risks associated with farming and/or (2) to preserve nature and cultivated landscapes.9 The AEMs are chosen in sets of BMPs (this set is larger than the currently adopted AEMs). The term ‘‘best management practices’’ has several meanings in the literature today, ranging from a combination of land treatment practices, production practices and technologies10 to components of the farming system to reduce possible farm-generated contamination of soil, water, air and biological resources.11 In this chapter, BMPs for water quality restoration are any kind of cropping method, agronomic technique or landscape ﬁxture that potentially reduces water pollution and is proposed to farmers on a contractual basis.12 This deﬁnition includes implicitly schedule of activities, prohibitions or maintenance procedures but does not consider enforcement or education

System Approach to Environmentally Acceptable Farming

497

measures. It associates the concept of changes in the farming system, the way these changes are designed by people outside the farm (by local technicians usually) and two main assumptions: the BMP is supposed to have an impact on the on-farm emissions, but the real impact may depend on the farm characteristics (including location on the watershed); and the adoption of the BMP is subject to some contract-based process (the negotiation is led by the regulator). The implementation of BMPs can greatly modify the production systems because most often bundles of BMPs are adopted together. Usually, BMPs are designed such as:

being scientiﬁcally sound; using all expertise available; being technically feasible at farm level; being adapted to local conditions; being eﬀective towards water quality standards; having low implementation costs for farmers; and being acceptable by the population (farmers and users of the watershed).

8.2.2.2

BMP Typology

More often, BMPs are categorised by environmental objective (water quality, erosion, biodiversity, landscape, etc.). They can be classiﬁed also by type of pollutant or by geographical area aﬀected. Table 8.2.1 lists some examples classiﬁed by type of pollutant.

8.2.2.3

How to Design Acceptable BMPs ?

Because of the diﬀuse nature of pollution by nutrients, pesticides and hazardous substances, their mitigation requests the cooperation of any potential polluter. Voluntary approaches based on BMPs seem appropriate in this particular case of pollution control, to reach a pre-deﬁned environmental target. Case studies provided by the OECD (2003) some ten years after voluntary approaches had been ﬁrst designed, combined with existing theoretical literature, opened the way to a draft analysis of the room for manoeuvre in pollution mitigation at the watershed scale.

8.2.2.3.1

Why do Firms Participate in Voluntary Approaches?

The ﬁrst drawback stressed by the OECD from case studies is that the environmental objective of many voluntary approaches schemes is not diﬀerent from what would have happened in the absence of any regulation.6 Fortunately for IWRM, the Water Framework Directive ﬁxes environmental objectives all over Europe, and thus these objectives are not subject to local negotiations

498

Chapter 8.2

Table 8.2.1

Examples of potential BMPs.a

Improving fertilisation practices:

application of all the manure produced within the watershed, then adjustment of inorganic fertilisation to meet crop needs;

decrease of mineral nitrogen amounts; use of fertilisation guidelines to adapt the amount of spread nutrient closely to plant requirements.

Modiﬁcation of rotation:

local crop rotation with additional catch crops during winter period; green fallow; change from maize to meadow, alfalfa or ryegrass–maize rotation. Modiﬁcation of soil structure and porosity to reduce erosion and P transfer:

catch crop implementation; mulching on maize ﬁelds; grass under permanent cultures and vegetative ﬁlter strips (VFS). Improving drainage water quality on acid sulfate soils:

control drainage; lime ﬁlter drainage. Improving pesticides management:

weed control by a combination of mechanical and chemical measures; implementation of warning system and purposeful selection of fungicides and dosages;

application of herbicides in rows and mechanical weeding between rows; insect pest control related to population level; use of models to select the less harmful pesticide for the environment depending on the on-ﬁeld condition.

Composite BMPs:

improvement of cattle feeding to reduce the amount of nutrients in their eﬄuents plus amount of fertilisers brought to the plants close to plants requirements;

speciﬁc technical BMPs targeted to each kind of soil. Economic policies:

a

increase of rye grass and clover instead of corn silage to optimise gross margin; tax on mineral nitrogen; modiﬁcation of milk quota; mandatory quota of bought mineral nutrients; optimally diﬀerentiated policies on production level.

Source: agriBMPwater project.

System Approach to Environmentally Acceptable Farming

499

anymore and the main drawback of voluntary approaches is not a diﬃculty for non-point source pollution mitigation at the watershed scale. Of course, it is obvious that no ﬁrm is directly interested in participating in any environmental scheme that will generate additional costs and limit proﬁts. Nevertheless, some empirical evidence suggests that some farms adopt BMPs even when the associated premium is below their implementation costs.13 Let us examine the potential reasons for this adoption. Dupraz et al.13 suggests that farmers include the environmental quality on their own farmland in utility function: adopting BMPs that improve this environment in a noticeable way (improving birds’ habitat for example) increases their welfare even if they have to bear some additional monetary costs. But improving the quality of groundwater will not be noticeable without additional measures within a few years, even if all the restoration measures are adopted, and this simple scheme cannot be applied. The OECD lists some additional reasons for adoption. Large ﬁrms may be interested in improving their image, which in some cases can improve the motivation of their employees or their quotation. Up to now, this motivation has seldom been demonstrated,14 and is only slightly relevant for family farms. A voluntary reduction of pollution can allow a ﬁrm to avoid a more restrictive regulation in the future.1 While taking the initiative, the ﬁrm can aim at anticipating the evolution of public regulation; with voluntary approaches, the ﬁrm can choose itself the de-polluting technique and thus mitigate the pollution at lower costs than those involved in a restricting policy.15 Even when individual consumers are not directly aware of the environmental quality linked to the production process of the products they buy, there is a threat of social protest for the environmental quality as a whole. A ﬁrm may wish to anticipate this protest16 and adopt a voluntary agreement scheme to be covered from future potential risks. Anton et al.17 demonstrate that the potential pressure from future investors and the possibility of future trials are the two main reasons for ﬁrms to adopt environmental management techniques. Firms may adopt voluntary agreements for strategic reasons: when they develop a new technology that leads to a better environmental performance, this technology will provide advantages to these ﬁrms over their competitors if the technology becomes the basis for a new regulation.18 Moreover, very recent work demonstrates that when the ﬁrms are imperfect competitors, the current environmental regulation strongly modiﬁes the competition rules.19 This point suggests that the voluntary approaches analysis cannot be studied without considering the local competition rules and regulation in force. At this stage, we can consider that the farms on a watershed may adopt voluntary agreements schemes for reasons that vary from one farm to another. Moreover, when the adoption is subsidised, usual schemes favour uniform compensation: the farmers will receive the same amount of money for a

500

Chapter 8.2

given BMP, whatever their adoption costs. Recent work suggests that the relationship between adoption costs, environmental performance and regulation in force is not straightforward.20 The farmers who have the lower costs of adoption will adopt the BMPs ﬁrst. As the farmers are not uniformly scattered within a watershed which, in general, is not physically uniform either, the costs and acceptability of the BMPs have consequences on their environmental eﬀectiveness also: for a given watershed, if the farmers located on the more sensitive areas have higher costs of adoption, they will not implement the BMP unless the associated premium is very high, and the BMP eﬀectiveness on this watershed may dramatically drop.12

8.2.2.3.2

Adoption of BMPs, Threshold Eﬀects and Beneﬁts

The greater diﬃculty that local regulators face with voluntary approaches is the share of costs and beneﬁts that may occur from the adoption of the BMPs. Onfarm cost assessment is diﬃcult because of farms’ heterogeneity. It is obvious that farms have a wide range of production factors (such as soils, climatic situation, management skills, genetic value of herds) and that farmers’ objectives are very diﬀerent. This results in a wide range of technical choices, such as the degree of production intensiﬁcation, the amount of inputs used and the techniques implemented. The heterogeneity of farms has consequences on their behaviour when facing a regulation, and on the amounts of non-commodity outputs (in this case, good quality of water) they produce, i.e. the same technical choice on two diﬀerent farms may result in diﬀerent non-commodity outputs. On-farm heterogeneity and hidden reasons of adoptions are usually grouped into what economists call asymmetries of information (farmers have a better knowledge of their farm characteristics and on their own behaviour than has the regulator). In practice, public regulation of agri-environmental processes has to cope with two diﬀerent problems: on the one hand, there is asymmetric information between the regulator and the farmers on the adoption cost and the eﬀective eﬀort of the farmers; on the other hand, the regulator and the farmers share uncertainty in the relationship between farming practices and environmental quality. These two diﬃculties often cumulate into the agri-environmental schemes and may lead, when threshold eﬀects occur, to no eﬀective environmental eﬀect and to farmers’ discouragement.13 Threshold eﬀects are deﬁned by Muradian7 as a sudden change in any property of an ecological system as a consequence of smooth and continuous change in an independent variable. The examples are numerous in the ecological literature; for water quality, the most frequent examples are pollution from pesticides when protected areas disappear, eutrophication in rivers related to phosphorus loads and groundwater contamination depending on the location of intensive practices on karstic areas. The existence of discontinuities in the ecological processes underpinning natural resources renewal generates nonconvexity phenomena that are thoroughly analysed in natural resources economics.22,23 The exploitation of these natural resources, when threshold eﬀects

System Approach to Environmentally Acceptable Farming

501

occur, is characterised by the existence of multiple equilibria, and thus requires the design of dynamic management policies.23–26 Through lack of enough precise information, local regulators are often brought round to design AEMs without considering threshold eﬀects, which decreases the eﬃciency of their policies and leads to waste of public funds: empirical studies depict more and more often the adoption of BMPs, after much eﬀort from involved agents and sometimes with important public subsidies, without any modiﬁcation of the environmental quality, because threshold eﬀects have been neglected.7 Lastly, cost–beneﬁt analyses are currently improved and provide very useful assessments of the beneﬁts linked with any agri-environmental schemes (see Ref. 27 for a well-known example), but they never explain how the beneﬁts can be shared among people in society. If the improvement of the water quality beneﬁts only local inhabitants, they can be charged a fee that compensates the costs of mitigation, for example on their water bill. In this case, the only remaining problem for the regulator is the share of mitigation costs among the population when the eﬀectiveness of the scheme is not noticeable during the ﬁrst years of application. Of course, the fairness of the scheme will condition its long-term acceptance by the population.28 But most often, the beneﬁts of an AEM are scattered along a large range of people: more ﬁshers enjoy the quality of the river, families enjoy good water quality for bathing, sportspeople appreciate fresh water, ecological associations value a good ecological equilibrium, tourists may appreciate a pleasant river in the landscape and the restoration of water quality may also indirectly create employment. But there is no way for regulators, who bear the costs of the restoration, to extract part of their costs from all these people and thus to recover the costs. For this reason, cost-eﬀectiveness is usually preferred to cost–beneﬁt analysis when dealing with IWRM.

8.2.2.3.3

Design of BMPs in Practice

Usually, the design of a restoration plan starts with agri-environmental audits at farm and watershed levels, including the diﬀerent uses for water, its quality and available quantity depending on the diﬀerent periods of the year. This diagnosis is often performed by consultants with few relationships with the other potential users on the watershed. To improve the appropriation of diagnosis by all the actors on a watershed it is highly recommended to involve them at the earliest stages of the restoration plan. The design of BMPs can be a good step in the procedure to begin an active cooperation. Many BMPs have already been tested and experienced in various watersheds throughout the EU. Appropriate BMPs can be locally designed through interviews with administrations, professional advisors and elected representatives in order to describe the history of environmental measures tested on the watershed, share experience from other regions and deﬁne practices that could match the local situation.

502

8.2.3

Chapter 8.2

How to Build a Cost-eﬀectiveness Grid of Bundles of BMPs

We shall describe with a simple example of a watershed how a cost-eﬀectiveness analysis can help regulators in designing mitigating policies. The Don watershed (71 706 ha) is located in the western part of France, in Pays de la Loire region. Farm production is mainly cattle breeding (dairy and meat production), where cereals are grown for both grain and forage. Grasslands, associated with dairy production, account for around 50% of the total agricultural area. The water coming from the Don watershed is connected to two pumping stations for drinking water, supplying around 150 000 people. In this particular watershed, a small range of farmers showed that they can adopt costly BMPs without compensation.12

8.2.3.1

Eﬀectiveness Assessment

Basically, implementing a BMP on a given area will have short- and long-term consequences on water quality, while modifying speciﬁc discharge, pollutant pathways, nutrient cycles, and so on. The environmental eﬀectiveness of a given BMP can be deﬁned as the evolution of water quality led from the BMP implementation on a watershed or on some speciﬁc areas within this watershed. The eﬀectiveness should be considered as the diﬀerence between the baseline scenario and the modiﬁed scenario, each system being in equilibrium. The environmental eﬀectiveness (and associated costs) of BMP mitigation programmes is greatly improved on targeting agri-environmental schemes on speciﬁc rural areas; these speciﬁc areas are more often understood as critical areas.

8.2.3.1.1

Deﬁnition and Delineation of Critical Areas

Actually, groundwater contamination depends on several geomorphological, hydrogeological and hydrological variables such as topography, sequence of soil layers, thickness of soil layers, soil textural properties, soil depth, humus content, hydraulic conductivity function, biological mixing parameter, organic matter, ﬁeld capacity and water retention characteristic. These variables need to be analysed to deﬁne diﬀerent potential risk geographic areas and their delineation where BMP implementation should be the most eﬀective. The Water Framework Directive requests the identiﬁcation of heavily modiﬁed water bodies, where more emphasis should be put on restoration measures. The concept of critical areas developed here is diﬀerent. Even for slightly modiﬁed water bodies, it is obvious that all the components of the watershed do not contribute at the same level to the NPS pollution process. Besides, the least costly way for the economy (or for speciﬁc economic sectors, here agriculture) to achieve well-deﬁned environmental objectives for water resources often requires targeting the measures to speciﬁc areas where they may be more eﬀective, or cheaper, to implement.

System Approach to Environmentally Acceptable Farming

503

The thematic Strategy on the Sustainable Use of Pesticides suggests the designation of areas where use of pesticides has to be reduced. The implementation of the Nitrate Directive (Council Directive 91/676/EEC, Article 10) requests the elaboration by each member state of a code of good farming practices, applicable to designated nitrate vulnerable zones. The deﬁnition of these priority zones, named ‘‘critical areas’’ will strongly depend on the level of integration of the management plan. If only physicists are involved, the programme will have a natural science theoretical aim and a critical area can be deﬁned as ‘‘the minimum area, where feasible measures can be applied, needed to reach the desired quality standard of the considered pollutant at the outlet or pumping station’’ (agriBMPwaterw deﬁnition). When many stakeholders participate in the diagnosis, an operational deﬁnition can be adopted and the critical areas are ‘‘the sets of areas where feasible measures can be applied needed to reach the desired quality standard of the considered pollutant at the outlet or pumping station.’’ More often, physicists, stakeholders and economists are involved in the restoration plan. In this case, critical areas can be deﬁned as ‘‘the set of areas where feasible measures can be applied to reach the desired quality standard of the considered pollutant at the receptor at the least social cost.’’

8.2.3.1.2

Characterising Methods

Although linearly presented, the analysis is iterative: initial analysis is based on existing information, and will be upgraded as new information and knowledge are gathered. The use of a spatialised hydrological model is of importance to select, among all the watershed areas, some of them where the implementation of BMPs is expected to be more eﬃcient. These models need to be calibrated ﬁrst on a baseline scenario. Of course, no hydrological model will provide immediate delineation of critical areas. There is a need to rank the speciﬁc pollutant loads from each unit area with respect to the others. A sensitivity analysis will provide great help at this stage for the interpretation of ranking the diﬀerent areas according to their potential eﬀect on the BMP eﬀectiveness. Once the diﬀerent unit areas from a watershed are ranked, their speciﬁc simulated eﬀectiveness has to be combined, so that each BMP delineates the areas deﬁned as critical according to the natural science deﬁnition. To go further in the delineation of critical areas, the stakeholders’ and ﬁrms’ interests can be taken into consideration. The areas where potential BMPs are modelled to be most eﬀective may diﬀer from the areas where the same BMPs are more liable to be implemented. Then, the diﬀerent areas have to be ranked according to both eﬀectiveness and acceptability criteria, before delineating the ‘‘critical areas’’ according to the operational deﬁnition. The same procedure can be applied to design critical areas according to the welfare economic deﬁnition, the candidate areas being ranked according to a w

The AgriBMPWater project received funding from the European Commission in the 5th Framework Programme.

504

Chapter 8.2

cost-eﬀectiveness ratio, their potential acceptability being also considered. Note that the delineation of critical areas according to the two last deﬁnitions is an iterative process which is often time consuming. Most studies use the physical deﬁnition of critical areas only.

8.2.3.1.3

Assessment Method

Eﬀectiveness can be estimated through the introduction in previously validated models of pre-designed BMPs as alternative practices. Each BMP eﬀectiveness can be determined as the ratio between the initial state and the estimated state after BMP implementation, both systems being in equilibrium. Basically due to lack of monitoring, modelling is the appropriate approach to assess the eﬀectiveness. Even though distributed, physically based models do not in principle require lengthy hydro-meteorological records for their calibration, but they do require considerably more input parameters than the simpler lumped models. Again in theory, the parameters and their spatial distributions could be measured in the ﬁeld, but the expense of such a survey is obviously not realistic and would prohibit practical implementation of the models. It is therefore necessary to reduce the number of direct measurements and to employ more indirect evaluations readily available from ﬁeld studies. As the parameter values should be characteristic for the spatial resolution used in the model, the sampling and evaluating of the parameters represent a supplementary diﬃculty. Many hydrological measurements, for example, are made at the point scale and may or may not be representative of conditions at the grid scale used in the distributed models. In this regard, parameter evaluation from data provided by remote sensing techniques or satellite information is potentially of great help. However, while these techniques can currently give surface distributions of watershed properties such as topography, land use or vegetation, they do not provide information on soil type and subsurface soil conditions. Against the above assessment of some of the major diﬃculties associated with data provision, it is clear that the choice of a model is directly conditioned by the way the problem of data provision is handled. Precise guidelines should therefore be speciﬁed right at the beginning of the coding eﬀort, rather than in the process of development. As the reality shows that most natural watersheds are often poorly deﬁned in data, three avenues are chosen for the data provision and the type of model chosen. The ﬁrst concerns the need to reduce the number of system parameters to a strict minimum. Even though this point seems obvious, still too many simulation codes suﬀer from the problem of over-parameterisation. The second data provision criterion concerns the structural ﬂexibility of the modelling code. The model should be able to match the sophistication of the solution with the speciﬁc project requirements or the availability of data. In this regard, two categories of input data should be considered, i.e. those data which are absolutely necessary to drive the modelling

System Approach to Environmentally Acceptable Farming

505

system, and those data which are useful in the sense that their knowledge improves the precision of simulation. Moreover, the ﬂexibility of the model architecture should be able to accommodate diﬀerent parameter evaluation techniques. As the parameter values are estimated from either direct or indirect measurements, the code should be capable of running the speciﬁc conﬁguration out of a wide class without any need for work at the level of the software. The last point of importance for a sound data provision strategy concerns the pre-processing of the rough ﬁeld data. The pre-processor of the modelling code should include tools which are capable of aggregation, disaggregation and/or interpolation (in space and time) of various hydrological and hydro-meteorological input data.

8.2.3.2

Implementation Costs

‘‘The main policy conclusion from the work thus far in the joint working party is that improving the environmental performance of agriculture in many countries involves costs that would be lower in the absence of commodity production-linked support measures. In other words, it is not suﬃcient to show that policies have been eﬀective in achieving a desired environmental outcome; it is also necessary to evaluate the economic costs and beneﬁts of such achievement, and demonstrate which combination of policies and market actions would achieve the same or better environmental outcome at lower cost.’’29 Implementing agri-environmental schemes leads to several kinds of costs. Adopting BMPs induces on-farm costs (increase of labour, machinery, time for management operations, investment for technological change, increased risk of yield loss, etc.). IWRM often requires pollution reduction from point source polluters also, who bear de-polluting costs. Choosing to adopt or reject the policy induces transaction costs for the farmers (mainly information costs). Designing and implementing the policy creates transaction costs for the regulator for negotiations with the potential polluters, in some cases, for scientiﬁc analysis of polluting phenomena, administrative costs for policy design, for control of implementation and enforcement. In some cases, because of competition eﬀects of diﬀerent policies, designing an AEM will lead to a lower eﬃciency of other policies. When the adoption of BMPs is subsidised, tax-payers have to bear additional taxes (these taxes have a social cost because either the tax level is raised or the amount of money devoted to pollution mitigation cannot be used for other purposes like schools or roads). Because of the jointness between commodity output and non-commodity output supply, BMPs designed for mitigating pollution may lead to a modiﬁcation of the commodity outputs supplied, which in turn may

506

Chapter 8.2

modify other non-commodity outputs like rural viability or landscapes. Thus consumer proﬁt may also be aﬀected. It is often the case that cost–beneﬁt analysis focuses on static comparisons of a baseline and a modiﬁed situation. Of course, the nature and size of the costs evolve in time, because of changes in the legislation in force (which may compete with the agri-environmental policy) or of the economic conditions. The supply of non-commodity outputs by farms is encompassed by farm heterogeneity. The connectedness of non-commodity outputs with commodity agricultural outputs has been widely analysed in the literature, erosion in economics of scope partially explaining the specialisation of farm production. When designing a policy, a regulator has to make choices on how to distribute the non-commodity production eﬀort among the heterogeneous producers, because the ratio between commodity and non-commodity outputs diﬀers among them. On the Don watershed, we carried out an analysis of the costs associated with several agri-environmental policies with a model for on-farm proﬁt which includes a description of the heterogeneity of the farms, a model for tax-payer surplus and a model for the damage cost (see Ref. 30 for details of the modelling). On this watershed, the regulator compares the costs associated with the environmental policy design with the beneﬁts that can be withdrawn from a reduction of the treatments of the water distributed to the population. For simplicity reasons, we neglected the consumer surplus and most of the transaction costs. We compared BMPs such as reduction of mineral fertilisation (BMP1), reduction of organics fertilisation (BMP2) or both (BMP3 ¼ BMP1+BMP2) towards economic instruments, either uniform ones (uniform extensiﬁcation, uniform tax on nitrogen, nitrogen quota at farm level) or diﬀerentiated policies (the farmers are taxed or subsidised with a rate depending on their production yield). The diﬀerentiated policies are designed such that 25, 50 or 75% of the farms beneﬁt from the regulation (diﬀerentiated 1, 2 or 3 in Figure 8.2.1). Figure 8.2.1 depicts the implementation cost of each measure and the proportion of farms that beneﬁt from it (i.e. which have higher proﬁt after regulation than before, or that declare themselves ready to implement them even if it is costly). The costs are expressed as the sum of producer plus tax-payer surplus. Black diamonds depict schemes where the total costs for tax-payers and farmers exceed the beneﬁt from damage reduction: this is the case for BMPs and nitrogen quota. Unexpectedly, some farmers declare themselves ready to implement the proposed BMPs, even if they are costly for them. Mandating a tax on the amount of nitrogen bought by the farms, or mandating a uniform extensiﬁcation would lead to beneﬁts from damage reduction greater than the implementation costs, but would disadvantage all the farmers. On the contrary, the diﬀerentiated policies result in implementation costs lower than the beneﬁt from the damage reduction, and beneﬁt to a share of the farmer population; only when they are designed such that more of 95% of the farmers beneﬁt from the regulation (not depicted in Figure 8.2.1) do they result in costs greater than beneﬁts.

507

System Approach to Environmentally Acceptable Farming acceptability (% of the watershed area)

80 differentiated3

70 60

differentiated2

50 BMP1

40 BMP3

BMP2

30

differentiated1

20 10 uniform ext,

N tax

N quota cost (euros/kg of avoided N)

0 -5

-4

-3

Figure 8.2.1

-2

-1

0

1

2

3

4

5

6

7

Implementation costs (costs for farmers and tax-payers) for diﬀerent mitigating policies on the Don watershed.

The ﬁrst result underlined by this study is that there are many ways to mitigate non-point source pollution from farming activities on a watershed. Most of them have been simulated as eﬀective enough to reach the EU threshold of 25 mg l1 of NO3. The second result highlights the diﬀerence between the feasibility of technical BMP implementation and the acceptability by farmers. Obviously there is a large place here to improve the environmental advice strategies. Third, the cost investigation suggests that optimally diﬀerentiated regulations are the best way to conciliate eﬀectiveness, implementation costs and acceptability of mitigating instruments. In our application case, the loss of welfare related to private information is lower than the cost of information required to implement the ﬁrst best production levels and obviously such a programme with adjusted subsidies is cost-eﬀective to implement the Water Framework Directive.

8.2.3.3

Building of the Comparison Grid

Once all the previous steps have been completed, the integration of the diﬀerent elements is a very useful decision tool. The integration is performed through a synthetic diagram that depicts on each watershed the contracted area, the eﬀectiveness of the BMP, the associated costs and either the current participating area or the potential area where the BMP is acceptable. On the Don watershed, the technical BMPs that have been compared are: BMP1 ¼ decrease of the inorganic nitrogen spread over all the crops, BMP2 ¼ manure spreads on grasslands instead of corn, BMP3 combines both

508

Chapter 8.2

Effectiveness (% of initial N loads)

Costs (% of max cost) 100

100

90

95

eff-BMP1

80 eff-BMP2

90 70

eff-BMP3b

85

60 Regulator's objective

80

50 40

75

30 70 20 65

10

60 10

0

20

30

40

50

60

70

80

90

0 100

eff-BMP3 cost-BMP1 cost-BMP2 cost-BMP3b accept-BMP1 acceptBMP2-3

% of agricultural area where the BMP is implemented

Figure 8.2.2

Analysis grid: application to Don watershed.

BMPs 1 and 2 and BMP3b is BMP3 with an adjusted inorganic fertilisation close to crop requirements. When comparing the size of their implementation area and their simulated eﬀect on water quality on the watershed, it is easy to notice that a regulator with the objective of reaching the EU threshold of 25 mg l1 of NO3 has to implement these BMPs on a large range on the watershed area (60% of the agricultural area for BMP3 and 85% for BMP1). It is now possible to compare this necessary implementation area with the area where the farmers declare themselves ready to implement each BMP: clearly, on the Don watershed, there is no way to conciliate the potential area of BMP implementation (37% of the agricultural area for BMP3 and 45% for BMP1) with the simulated necessary area. Only BMP3b, which requires a high technology level and the capacity to adapt the fertilisation each year depending on the previous climatic conditions, could conciliate the regulator’s objective and a low level of implementation, but its acceptability (not depicted in Figure 8.2.2) is too low. A regulator who relied on the voluntary adoption of the technical BMPs would never reach the objective of meeting the EU threshold of 25 mg l1. Thus there is a need to design other BMPs. This conclusion is strengthened by the diﬀerence noted within the acceptability analysis between the BMPs’ acceptability and their feasibility.

8.2.4

Conclusions

Concerning the response to water quality issues, sustainable BMPs, scientiﬁcally consistent and transparent, suitably designed, cost eﬀective and acceptable, should integrate the complexity of the farm system: diﬀerent spatial and

System Approach to Environmentally Acceptable Farming

509

temporal scales and multiple objectives to reach in response to social demand. They demonstrate a wide array of pollutant mitigation capabilities. But their cost eﬀectiveness can be aﬀected by several factors such as environmental conditions, farmer awareness, willingness to pay for environmental improvements and farm evolution trajectories. However, as shown in the simulations of the AgriBMPWater project, most BMPs proposed do not allow the environmental standards to be reached. Hence, we must imagine more restricting measures, which can induce a lower acceptability by farmers. Regarding the deﬁnition of programme of measures, there is a consensus around the participation of all concerned actors in the design process. Today, the value of participatory approaches for sustainable use of water at watershed scale is well identiﬁed and a wealth of good farming practices has been developed to help solve real-life problems. Nevertheless, little has been done to apply this method to the identiﬁcation of issues and assessment tools in the context of agricultural activity and its impact on local environmental, social and economic stakes. In this respect, cross-societal partnerships will have valuable experience to contribute. Many other improvements can be foreseen: until now, focus has been only on the potential eﬀects and costs of particular BMPs. Obviously, the application of a speciﬁc BMP generates eﬀects on other practices at the farm level. Developing a joint approach that incorporates the economical, sociological and physical aspects of the modelling through the building of a decision support system is in our opinion the key for future research in the area of mitigating non-point source pollution from human activities and a promising way to help EU member states enforce policies focusing on water issues. Although the design of policies that promote good farming practices may improve the sustainability of agriculture, this scope is not wide enough regarding the diﬀerent demands from society and the challenges that the community has to face. Farming activity, with the support of the CAP ‘‘second pillar,’’ has the potential to supply non-commodity outputs that will increase the sustainability of society as a whole, and combining environmentally friendly practices and environmental technologies may lead to an increase in competitiveness. Indeed, new paradigms in rural economics such as multifunctional agriculture or in engineering such as the design of environmental technologies are required to integrate systemic and comprehensive approaches based on biosciences and human and social sciences, so that farming activity can take a signiﬁcant part of future bioeconomy.

References 1. K. Segerson and T. J. Miceli, J. Environ. Econ. Manag., 1998, 36(2), 109–130. 2. J. S. Shortle and R. D. Horan, J. Econ. Surveys, 2001, 15(3), 255–289. 3. J. S. Schou, E. Skop and J. D. Jensen, J. Environ. Manag., 2000, 58(3), 199–212.

510

Chapter 8.2

4. A. Kampas, A. C. Edwards and R. C Ferrier, J. Environ. Manag., 2002, 66(3), 281–291. 5. N. B. P Polman and G. J Thijssen, Agricult. Econ., 2002, 27(1), 41–49. 6. Voluntary Approaches for Environmental Policy: Eﬀectiveness, Eﬃciency and Usage in Policy Mixes, OECD, Paris, 2003. 7. R. Muradian, Ecol. Econ., 2001, 38(1), 7–24. 8. M. Glachant, Voluntary agreements in a rent seeking environment, in The Handbook on Environmental Voluntary Agreements, ed. E. Croci, Kluwer, 2005. 9. Agri-environmental Measures: Overview of General Principles, Types of Measures, and Application, European Commission, Directorate General for Agriculture and Rural Development, Unit G-4, Brussels, 2005. 10. C. Santhi, R. Srinivasan, J. G. Arnold and J. R Williams, Environ. Model. Software, 2006, 21(8), 1141–1157. 11. R. F. Cullum, S. S. Knight, C. M. Cooper and S. Smith, Soil Tillage Res., 2007, 90(1/2), 212–221. 12. N. Turpin, R. Laplana, P. Strauss, M. Kaljonen and F. Zahm, Int. J. Agricult. Res., Gov. Ecol., 2006, 5(2/3), 272–288. 13. P. Dupraz, I. Vanslembrouck, F. Bonnieux and V. Huylenbroeck, Farmers’ participation in European agri-environmental policies, 10th Congress of the European Association of Agricultural Economists, Zaragoza, 2002. 14. A. Alberini and K. Segerson, Environ. Res. Econ., 2002, 22(1–2), 157–184. 15. J. J. Wu and B. A. Babcock, J. Environ. Econ. Manag., 1999, 38(2), 158–175. 16. T. Hommel and O. Godard, Economie Rurale, 2000, 270, 36–49. 17. W. R. Q. Anton, G. Deltas and M. Khanna, J. Environ. Econ. Manag., 2004, 48(1), 632–654. 18. S. C. Salop and D. T. Scheﬀman, The American Economic Review, vol. 73(2), Proc. 95th Annual Meeting American Econ. Assoc., 1983, 267–271. 19. M. David and B. Sinclair-Desgagne, J Regul. Econ., 2005, 28(2), 141–155. 20. J. Videras and A. Alberini, Contemp. Econ. Pol., 2000, 18(4), 449–461. 21. P. Dupraz, K. Latouche, N. Turpin. Eﬀets de seuil et coordination des eﬀorts agri-environnementaux. Territoires et enjeux du de´veloppement re´gional, Lyon, 9–11, March 2005. 22. P. Dasgupta and K. G. Maler, Environ. Res. Econ., 2003, 26(4), 499–525. 23. F. Wirl, Resour. Energy Econ., 1999, 21, 19–41. 24. T. Mitra and S. Roy, Econ. Theory, 2006, 28(1), 1–23. 25. D. Rondeau, J. Environ Econ. Manag., 2001, 42(2), 156–182. 26. M. Toman and C. Withagen, Accumulative pollution, ‘‘clean technology, ’’ and policy design, Resources for the future, Discussion Papers, 2006. 27. R. Costanza, R. d’Arge, R. de Groot, S. Farber and M. Grasso, Nature, 1997, 387(6630), 253–260. 28. E. Fehr and B. Rockenbach, Nature, 2003, 422(6928), 137–140. 29. Agriculture and Environment: Lessons Learned from a Decade of OECD Work, OECD, 2004. 30. P. Bontems, G. Rotillon and N. Turpin, Environ. Res. Econ., 2005, 31, 275–301.

CHAPTER 8.3

WATCH. Water Catchment Areas: Tools for Management and Control of Hazardous Compounds THOMAS TRACK,a STEVE SETFORD,b SHARON HUNTLEY,b CLAUDINE VERMOT-DESROCHES, c JOHN WIJDENES,c DAMIA BARCELO´,d MONICA ROSELL LINARES,d PETER WERNER,e JENS FAHL,e HANS-PETER ROHNS,f CLAUDIA FORNER,f JESPER HOLM,g DOUGLAS GRAHAM,g ECKARD HITSCHh AND JOSEF LINTSCHINGERh a

DECHEMA eV, Theodor-Heuss-Allee 25, DE-60486 Frankfurt am Main, Germany; b Cranﬁeld Centre for Analytical Science, Cranﬁeld University, Silsoe, Bedfordshire MK45 4DT, UK; c Diaclone, 1 Bd Fleming-BP 1985, F-25020 Besancon Cedex, France; d Consejo Superior de Investigaciones Cientı´ ﬁcas, Instituto de Investigaciones Quı´ micas y Ambieltales de Barcelona, Dept. de Quı´ mica Ambiental, Jordi Girona, 18-26, E-08034 Barcelona, Spain; e University of Technology Dresden, Institute of Waste Management and Contaminated Site Treatment, Pratzschwitzer Str. 15, D-01796 Pirna, Germany; f Stadtwerke Du¨sseldorf AG, Quality Control Water, Wiedfeld 50, D-40589 Du¨sseldorf, Germany; g DHI – Water and Environment, DK-2970 Horsholm, Denmark; h Salzburg AG Centre Wasser, Hagenau 1, A-5101 Bergheim, Austria

8.3.1

Introduction

In many European regions groundwater is an important resource for drinking water supply. It strongly interacts with the overlaying soils and related surface water systems. As groundwater quality is a parameter that reacts slowly and with some retention to environmental impacts, it needs special attention for 511

512

Chapter 8.3

sustainable management. This becomes even more important in the light of European legislation: the Water Framework Directive (2000/60/EC), the new Groundwater Daughter Directive (see Chapter 3.1) and the proposal for a ‘‘Soil Framework Directive.’’ The water legislation requires, for example, the setting up of monitoring programmes and measure plans to keep and bring water resources to a good quantitative and qualitative status. Beyond this the increasing spread of anthropogenic compounds, from trace amounts to massive pollutions, requires advanced detection strategies, an improved understanding of environmental fate and behaviour as well as intelligent strategies to ensure a sustainable management of groundwater resources in terms of monitoring, early warning and decision support. As target pollution for this project, methyl tert-butyl ether (MTBE) was selected. MTBE as a fuel oxygenate replaced lead beginning in the 1980s but mainly starting in the early 1990s. Compared to other fuel constituents and some of the most common groundwater pollutants MTBE is more persistent and can be found in groundwater systems in all concentration ranges from the nanograms per litre level as ‘‘trace concentrations’’ up to the grams per litre level at ‘‘hotspots.’’ In addition MTBE is widespread in surface waters often up to several hundred nanograms per litre. Being widespread in aquatic environments poses the risk to groundwater and surface water to be aﬀected in their quality status. WATCH, a European Union-funded project within the 5th Framework Programme, faces these challenges by providing tools for management and control of hazardous compounds in groundwater catchment areas (Figure 8.3.1). The developments of WATCH are the following:

Strategy

Tasks

Pollution detection

Surveillance and monitoring techniques

WP1

Env ronmental i fate of BTEX, MTBE/ETBE

WP2

Data assessment

Monitoring & early warning

Monitoring and decision support system

Evaluation in case studies

subsurface near-infrared reﬂectometer sensor with associated data collection and transmission capabilities for the online remote identiﬁcation of free-phase hydrocarbons;

WP3 WP4

Protection and sustainment of water resources

Figure 8.3.1

The WATCH (water catchment areas: tools for management and control of hazardous compounds) project approach.

WATCH. Water Catchment Areas

513

fully automated purge and trap coupled to gas chromatography/mass spectrometry (P&T-GC/MS) method for the trace analysis (ng l1 to mg l1 levels in water, mg kg1 in sediments) of MTBE, its main degradation products (tert-butyl alcohol (TBA), tert-butyl formate (TBF)) and other oxygenated additives (ethyl tert-butyl ether (ETBE), tert-amyl methyl ether (TAME), diisopropyl ether (DIPE)); a competitive MTBE enzyme-linked immunosorbent assay (ELISA) testkit with a sensitivity of standards down to 0.05–0.5 mg l1, and an application range down to 50–500 mg l1; an immuno-ﬂow through and a magnetic bead format assay with application ranges o50 mg l1 are under development; elaboration of MTBE degradation pathways including inhibiting and stimulating eﬀects; integrated management protocol including a conceptual model protocol and a remedial action protocol; and demonstration of project developments in two case studies help to protect and sustain groundwater quantity and quality in Europe.

8.3.2

Near-infrared Fuel Leak Sensor

At the start of WATCH, a prototype fuel leak sensor, developed by Cranﬁeld University, had been developed to the point of ‘‘proof of principle’’ in identifying insulating oil leakages from subsurface power distribution cable joint-bays. The WATCH project extended this technology via the following objectives: to assess device performance as an early warning tool for identiﬁcation of free-phase hydrocarbons from leaking underground storage tanks (USTs); to produce a rugged unit amenable to low-cost manufacture; to assess device performance towards diﬀerent fuel types, soil types and degrees of matrix saturation; to develop cheap compatible systems for onboard data acquisition and transmission; to ﬁeld-test arrays of the sensor in contained environments; to assess the sensor as an in-borehole device for monitoring free-phase hydrocarbon contamination of drinking water; and to evaluate the socioeconomic relevance of the fuel leak sensor in the early detection of hydrocarbon fuel releases.

8.3.2.1 8.3.2.1.1

Description of Results The Sensor Head

The sensor assembly is housed in a sealed metal tube (70 15 mm) incorporating a polycarbonate window (Figure 8.3.2). A reﬂective hydrocarbon-speciﬁc

514

Chapter 8.3

Figure 8.3.2

Hydrocarbon sensors and exploded view of sensor head.

Figure 8.3.3

Light-emitting diode signal emission and reﬂection.

membrane is positioned over this window, which turns translucent on contacting free-phase hydrocarbons. The device may be placed in the interstitial space between double-skinned tanks or directly within soil matrices. A near-infrared (NIR) light-emitting diode and adjacent light-receiving phototransistor are built into the sensor head. The amount of membranereﬂected light received by the phototransistor is continuously monitored as a voltage. On contacting free-phase hydrocarbon, the reﬂectance of the membrane changes, reducing the amount of incident light upon the phototransistor, culminating in an increase in output voltage (Figure 8.3.3). The electronic circuitry is mounted on a small printed circuit board which ﬁts inside the sensor assembly. A ﬂying lead connects the sensor to the data acquisition/power supply. The sensor provides two outputs: a digital signal that triggers when the reﬂected signal drops below a predetermined level, and an

WATCH. Water Catchment Areas

515

analogue voltage that is directly proportional to the reﬂected NIR signal level. Multiple sensors may be linked to create a distributed system for monitoring large-scale hydrocarbon repositories. Water and alcohol-based solvents have no eﬀect on membrane reﬂectivity.

8.3.2.1.2

Data Acquisition and Communication

Up to 8 sensors can be connected to the BioDAQ data acquisition system. The system supplies power to the sensor heads and records sensor outputs. BioDAQ performs on-board data processing, such as detecting alarm conditions and forwards data to other systems via a wired or wireless data link. The BioDAQ and sensor units consume microamp currents, and thus the system can operate for extended periods on AA batteries. Sophisticated power management features are incorporated, including a ‘‘sleep mode’’ function to prolong battery life. Signal outputs below the alarm threshold are date stamped and stored in the on-board memory and/or sent to the operator as required. Breach of the alarm threshold triggers communication. A RS232 serial connection with Labview software allows simple visualisation of transmitted data. Two wireless technologies have so far been implemented: a GSM cellphone modem and a point-to-point FM wireless link, with a wireless Ethernet link currently under consideration.

8.3.2.1.3

Sensor Trials

8.3.2.1.3.1 Laboratory Testing. Initial laboratory tests were performed using a 9-sensor distributed network within a tank with simulated water ﬂow and UST leakage. It was found that: the sensors perform optimally under non-saturated conditions; the sensors were able to identify the direction and rate of free-phase fuel migration; the sensors were regenerable under certain conditions; and the sensors’ suitability to ﬁeld trials was indicated. 8.3.2.1.3.2 Large-scale Lysimeter (ﬁeld) Trials. Field-based trials were performed in a 1.0 0.8 m lysimeter. The open top allowed natural ﬂow-through of rain water. Trials were performed with sand and soil. A 19 16 cm UST complete with leakage point was placed onto a gravel bed within the matrix. In a typical trial, 3 sensors were placed close to the base of the leaking vessel, with a further 4 sensors placed at uniform depths below the vessel. The sensors were linked up to the on-site data acquisition unit and sampled hourly over several weeks. The data were stored and relayed to the operator by daily e-mail. The results from trial 2 are shown in Figure 8.3.4. All sensors triggered except sensors 3 (positioned to one side under the leaking UST) and 7 (base of lysimeter, 1 m depth). Excavation of the sensors indicated that they had not

516

Chapter 8.3 04 10 02 wk1 4800 4300

s1

s3

s4

s5

3800

Signal [mV]

3300

0.5L fuel

2800 s6

2300 1800 1300 800 300

-200 17:00 05:00 17:00 05:00 17:00 05:00 17:00 05:00 17:00 05:00 17:01 03/10/02 05/10/02 06/10/02 08/10/02 09/10/02 11/10/02 12/10/02 14/10/02 15/10/02 17/10/02 19/10/02

Date/ Time

Figure 8.3.4

Fuel leak sensor data output from lysimeter trial 2.

contacted the fuel. Sensors 1 and 2, closest to the leak point, triggered within 6 h of fuel addition. The signals of all triggered sensors remained high, regardless of weather conditions.

8.3.2.2

Conclusion and Perspectives

The integrated fuel leak/immunoassay approach has implications within the European Union Water Framework Directive (WFD), by directly contributing to groundwater protection, essential given that groundwater supplies 65% of all Europe’s drinking water. The directive states that monitoring programmes should be operational as a basis for water management by December 2006, an aspiration that is supported by the early warning and monitoring tools developed in WATCH.

8.3.3

Analysis of MTBE and Related Compounds in Soil and Groundwater

8.3.3.1

Description of Results

An analytical protocol for sample handling and analysis of MTBE and its main degradation products, TBA and TBF; other gasoline additives, oxygenated dialkyl ethers ETBE, TAME and DIPE; aromatics benzene, toluene, ethylbenzene and xylenes (BTEX); and other compounds causing odour events in groundwater such as dicyclopentadiene (DCPD) and trichloroethylene (TCE) was developed (Figure 8.3.5). Three diﬀerent methods were developed to analyse target compounds in water, soil and gasoline. Innovative fully automated P&T-GC/MS was

517

WATCH. Water Catchment Areas RT:0.00 - 49.99 100

11

IS1

95 90 85

2’

5

80 75

3

Relative Abundance

70

1

65 60

10

55

11’

9

50

12

4

45 40 35

6

30

IS3

25 20

2 7 8 IS2

15 10 5 0 0

5

10

15

20

25

30

35

40

45

Time (min)

Figure 8.3.5

Total ion chromatogram (TIC) in selected ion monitoring (SIM) mode for a 1 mg l1 spiked groundwater sample. Compound identiﬁcation: 1, TBA (m/z ¼ 59); 2, MTBE-d3 (IS1, m/z ¼ 76)+MTBE (2 0 , m/z ¼ 73); 3, DIPE; 4, ETBE; 5, TBF (m/z ¼ 59); 6, benzene; 7, TAME; IS2, ﬂuorobenzene; 8, TCE; 9, toluene; 10, ethylbenzene; 11, m-+p-xylene; 11 0 , o-xylene; 12, DCPD; IS3, 1,2-dichlorobenzene-d4.

optimised for the simultaneous determination of the compounds mentioned above, which permitted the detection of concentrations at ng l1 (ppt) to mg l1 (ppb) in water and mg kg1 in sediments. The use of mass spectrometry allows unequivocal identiﬁcation of the analytes present in environmental samples. Special attention was given to the determination of polar MTBE degradation products, TBA and TBF, since not much data on method performance and environmental levels are given on these compounds in groundwater. The protocol also includes diﬀerent parameters to guarantee the quality assurance of the analytical methodology. From the data reported, it is clear that MTBE poses a problem of groundwater contamination and severe monitoring programmes at risk sites (potential gasoline point sources) are needed to know the extent of this contamination and start remediation actions. Appropriate tools are thus necessary. The data obtained from these monitoring sites will be useful to know the presence, behaviour of the MTBE contamination plume and its degradation rate under the subsoil conditions. This protocol includes guidelines for transportation of

518

Chapter 8.3

contaminated groundwater and soil samples from diﬀerent European sites to the central laboratory and the parameters of extraction, identiﬁcation and quantiﬁcation of the pollutants in these samples. In general, the application of this protocol can allow the monitoring and control of risk sites or water catchment areas. Contaminated gasoline ‘‘hot spots’’ can be identiﬁed with MTBE levels up to US Environmental Protection Agency drinking water advisory limit values (20–40 mg l1) or taking into account stricter measures, as the Swiss guideline value for groundwater of 2 mg l1 based on precautionary principle, protecting in this way end-users. The presence of TBA (MTBE and ETBE key intermediate) in drinking water merits similar (or even more) consideration than that of its parent compound due to its complete water solubility and demonstrated toxicity and carcinogenicity in rats and mice. Therefore, the present analytical methodology permits the detection of this pollutant from low levels (0.1 mg l1). In practice, diverse water monitoring campaigns were carried out in diﬀerent sites located near gasoline point sources all over Europe: Catalonia (Spain), Salzburg (Austria), Du¨sseldorf and Leuna (Germany) and Faro (Portugal). Most analysed samples contained MTBE at levels varying between 0.01 and 670 mg l1. But the highest concentration of MTBE detected in groundwater (45 mg l1) was found under a reﬁnery. Therefore, several ‘‘hot spots’’ were identiﬁed in these study sites with levels up to 20–40 mg l1 and some of them even exceeded the Danish suggested toxicity level of 350 mg l1 (Figure 8.3.6). In addition, some inter-laboratory comparisons by analysing groundwater samples from these sites were done with similar results being obtained among laboratories. In addition, the analysis of samples from FP5 SENSPOL ﬁeld

Compound

Düsseldorf (Germany) Leuna MTBE site BTEX site (Germany)

Catalonia (Spain)

Salzburg (Austria)

666 0.68 nd 1.53

3.32 0.04 nd 0.01

645 nd 0.08 0.17

0.14 nd nd <0.01

45100 nd nd nd

62 <0.05

0.41 nd

440 3.34

<0.1 nd

37000 nd

4120

0.45

0.20

4820

920

na na

0.04 nd

na na

nd nd

Oxygenate additives

MTBE ETBE TAME DIPE Degradation products

TBA TBF Aromatic hydrocarbons

BTEX Other VOCs

TCE na DCPD na nd: not detected, na: not analysed

Figure 8.3.6

Maximum detected levels (in mg l1) of target compounds found in European studied groundwaters.

WATCH. Water Catchment Areas

519

tested in Koblenz (Germany) permitted the elaboration of a protocol for immunoassay validation to compare the response between developed immunoassay methods and the GC/MS method. Of special interest was a two-year monitoring programme of a gasolinecontaminated site at Du¨sseldorf (Germany) that was carried out to determine the presence, behaviour (horizontal movement and vertical proﬁle) and evolution of 12 target gasoline additives in a total of 96 samples from 14 groundwater wells. The origin of the contamination was suspected to be a gasoline spill at a gas station. The study of the plume showed that the distribution of MTBE and TBA in the aquifer forms a similar vertical concentration proﬁle which is inﬂuenced by the groundwater ﬂow direction. And as a result of WATCH partners’ collaboration, a conceptual model for the distribution of MTBE in the aquifer was developed based on the chemical properties of the pollutants, the hydrogeological conditions of the aquifer and the microbial degradation capacity. This is one of the ﬁrst studies of groundwater contamination of MTBE and TBA together in Europe due to the diﬃculty of TBA analysis at trace concentrations. Furthermore, the wide range of application of the P&T system for analysing diﬀerent aqueous matrices (rain, domestic waste, sewer and sea waters) in the vicinity of Faro (Portugal) has permitted the observation of fuel additives’ ubiquity in the aquatic environment and their daily and seasonal variability close to urban centres.

8.3.3.2

Conclusions and Perspectives

An analytical protocol for sample handling and analysis of MTBE, its main degradation products, other oxygenated dialkyl ethers and aromatics present in gasoline, and other compounds causing odour events was achieved. This fully automated P&T-GC/MS was optimised and applied to environmental samples for the simultaneous determination of these compounds, which permitted the detection of concentrations at trace levels (ppb) in water and sediments. This protocol is ready for use by other laboratories, water suppliers and public authorities to monitor and control risk sites or water catchment areas.

8.3.4

Immunoassay for Field-based Determination of MTBE

Concerns over the health and environmental hazards associated with the addition of lead-based compounds into fuels have led to the use of alternative fuel oxygenate compounds. MTBE is the most widely used fuel oxygenate, being found in B80% of fuels at up to 15% by volume. Current studies suggest that MTBE is less toxic than lead-based additives, although longer term studies are still ongoing. However, MTBE has a particularly unpleasant taste and odour, even at trace levels. This factor, coupled to the high solubility and longterm persistence in the environment has resulted in the contamination of

520

Chapter 8.3

drinking water. Whilst many standard methods exist to determine MTBE, these are laboratory based and may suﬀer long sample turnaround times, during which a pollution incident may escalate. Correspondingly there is a demand for rapid, ﬁeld-based tests that may be used to screen suspected contaminated sites and provide timely end-user information. As part of the WATCH project, an MTBE immunoassay was developed to fulﬁl this need. Immunoassays make use of the speciﬁc binding of target analytes by biological binding proteins called antibodies. The following tasks were undertaken by the WATCH programme to develop an MTBE-speciﬁc immunoassay: production of a new antibody cell line producing monoclonal antibodies with binding speciﬁcity towards MTBE; production of immunoreagents for construction of MTBE immunoassays; determination of optimum immunoassay format; testing of immunoassay with synthetic samples and environmental samples collected from contaminated sites; evaluation of immunoassay performance; and examination of immunoassay cross-reactivity towards structurally similar compounds.

8.3.4.1 8.3.4.1.1

Description of Results Development of Monoclonal Antibodies

The usefulness of monoclonal antibodies (mAbs) stems from their speciﬁcity of binding, their homogeneity and their ability to be produced in unlimited quantities. To increase the ability to induce an immune response, MTBE was coupled to bovine serum albumin (BSA), a larger protein molecule called a carrier. Hybridomas were produced by fusing the mouse myeloma cell line X63Ag8.653 and antibody secreting cells isolated from the spleen of MTBEimmunised BALB/c mice. The myeloma cells provide the correct genes for continued cell division and the antibody-secreting cells provide the functional immunoglobulin genes. Hybridoma screening was done by ELISA using coated BSA conjugated MTBE, BSA alone or irrelevant BSA conjugated antigens. Only the hybridoma-secreting antibodies which speciﬁcally react on BSA conjugated MTBE were selected. After a positive culture supernatant was identiﬁed, the cloning of the antibody-producing cell was performed.

8.3.4.1.2

MTBE Competitive ELISA: Assay Development

A competitive immunoassay for MTBE was developed. The competitive immunoassay format exploits the speciﬁc binding of the anti-MTBE antibody

521

WATCH. Water Catchment Areas

for the target MTBE analyte. In this particular assay format, a limiting quantity of MTBE-BSA immunogen (essentially an MTBE ‘‘analogue’’) is immobilised onto a solid-phase support, such as a microtitre well wall or magnetic bead (Figure 8.3.7). During the assay, the end-user supplies a sample, containing free MTBE analyte and anti-MTBE antibody (non-limiting), either free or conjugated to a binding protein (streptavidin). The free MTBE antibody will either bind to the free MTBE or immobilised MTBE analogue. The binding process is competitive: thus the more free MTBE present, the greater the amount of antibody that will bind to this analyte and hence will be removed from the system on washing the microtitre wells. Conversely, the lower the amount of free analyte, the greater the amount of antibody binding to the analogue immobilised on the solid-phase support. The amount of immobilised anti-MTBE antibody is determined by addition of tracer material, in this case the enzyme horseradish peroxidase (HRP), either linked to biotin (which has binding aﬃnity for streptavidin) or a ‘‘secondary’’ antibody having aﬃnity for the ‘‘primary’’ anti-MTBE antibody (grown in mouse and hence having binding aﬃnity to the mouse-raised anti-MTBE antibody). Enzyme activity is determined by addition of enzyme substrate that produces a colour in the presence of enzyme that may be recorded using a simple portable photometer.

MTBE Analogue

Figure 8.3.7

primary antibody

free MTBE

secondary antibody + HRP label

substrate

product

Competitive ELISA immunoassay, with immobilised MTBE analogue. Free MTBE competes with immobilised MTBE for anti-MTBE primary antibody binding sites. Following washing, residual immobilised antibody is determined using secondary antibody with HRP label. Following a further washing step, HRP label activity is determined. The amount of product formed is inversely proportional to the amount of free MTBE present in the original sample.

522

Chapter 8.3

The overall process is simple to perform, requiring only the sequential addition of assay reagents followed by incubation and washing steps. The procedure can be completed in B2 h and allows multiple sample throughputs. The simplicity of the process allows kit usage in ﬁeld-based locations without recourse to centralised laboratories. The assay is often referred to as an ELISA.

8.3.4.1.3

ELISA Secondary Antibody Linked to HRP

Optimal concentrations of both Ag and Ab (biotinylated and native) were determined by checkerboard assay. The optimal concentrations of coated antigen conjugate and biotinylated and native Ab were all found to be 500 ng ml1.

8.3.4.1.4

Microtitre Plate MTBE ELISA Performance: Synthetic Samples

The ELISA protocol was ﬁrst developed using buﬀer solutions spiked with various quantities of MTBE. The classic sigmoidal response curves (absorbance vs. analyte concentration), characteristic of competitive ELISAs were observed (Figure 8.3.8).

8.3.4.1.5

Alternative Assay Formats

8.3.4.1.5.1 Magnetic Bead Format. In order to simplify the MTBE immunoassay for use in the ﬁeld, the immobilisation of the MTBE analogue to micro-scale 4

Absorbance [450nm]

3.5 3 2.5 2 1.5 1 0.5 0 0

Figure 8.3.8

0.005

0.05

5 50 500 0.5 MTBE concentration [ppb]

5 5000

5 500000 50000

MTBE competitive ELISA calibration curve constructed with MTBE standards diluted in buﬀer. An assay calibration range of 50–5000 ppb (ng l1) is evident.

523

WATCH. Water Catchment Areas

Ag coated tosylactivated beads 160304 Absorbance [450nm]

3 2.5 2 1.5 1 0.5 0 0

0.05

0.5

5

50

5000

MTBE concentration [ppm]

Figure 8.3.9

MTBE immunoassay with antigen coated onto tosylactivated magnetic beads. The immunoassay, when in the magnetic bead format, can identify lower levels of MTBE contamination (o50 ppb) compared to 450 ppb in the microtitre plate format.

magnetic beads has been attempted. Preliminary results (Figure 8.3.9) indicate that a superior limit of detection is possible using this assay format.

8.3.4.1.5.2 Enzyme-linked Immunosorbent Flow Assay (ELIFA). Automation of the assay using a ﬂow-through approach, in which key assay components are pre-immobilised to a membrane, was also examined in order to further simplify the assay format. The principle of the approach has been proven, but further assay reﬁnement is required to reduce the incidence of non-speciﬁc binding of enzyme label to the solid-phase support.

8.3.4.2

Conclusion and Perspectives

For the ﬁrst time, a simple immunoassay has been developed for the ﬁeld-based determination of the fuel oxygenate MTBE. A large-scale development programme was required to obtain antibodies of the desired binding speciﬁcity due to the small size of MTBE. Promising monoclonal antibody cell lines were produced and tested before selection of the ﬁnal desired clone and assay development. When working with contaminated groundwater samples, the assay range is 50–5000 ppb MTBE and thus can be used to assess site pollution without the need for costly and time-consuming central laboratory analysis. Certain other structurally related compounds can also be determined using this assay, including TAME and ETBE and the MTBE breakdown products TBA and TBF. The assay is not designed as a replacement for GC analysis. It is essentially semi-quantitative, but has uses for the rapid decentralised assessment of a site where an MTBE contamination is believed to have occurred and where rapid information to allow near-instantaneous on-the-spot decisionmaking to be made with respect to pollution incident control and management.

524

8.3.5

Chapter 8.3

Elucidation of Stimulating and Inhibiting Eﬀects on Biodegradation

Work in the WATCH project focused on early detection and risk assessment decisions of environmental pollutions in drinking water catchment areas, especially those which use groundwater as a water source. For project-related research the choice was on a very mobile hazardous compound, the gasoline additive MTBE. Gasoline contains MTBE as an antiknock compound/octane booster which replaced the very toxic lead compounds like tetraethyl lead. MTBE usage in European Union countries extended to 3 Mio tons a year. A lot of environmental data show the gasoline additive MTBE as a substance that has the potential to seriously endanger the quality of groundwater reservoirs in subsurface sediments. Its toxicity is not comparable with the replaced lead compounds, but its sensorial properties make it so problematical. Humans can taste and smell MTBE in water at very small concentrations in a parts per billion range (sensorial threshold value approx. 10 mg l1). Unfortunately its physical data, such as high water solubility (up to 60 g l1) and the Henry constant (50 (dimensionless form), CMTBEwater/CMTBEair, at 20 1C), predetermine an enrichment of MTBE in groundwater. The eﬀect is intensiﬁed by the low adsorption aﬃnity of MTBE in the soil matrix. A further negative aspect is that the ether linkage and tertiary carbon structure make MTBE relatively recalcitrant to microbial degradation and lead to fears of an accumulation of this compound in the aquatic environment. In the beginning of WATCH available knowledge regarding the behaviour of MTBE in groundwater was still very weak. Detailed investigations were necessary to estimate the real risk of moving MTBE plumes in groundwater. Validation of the recalcitrant and persistent character of MTBE should help in clarifying the expected extent of natural attenuation of these hazardous compounds in groundwater systems. Speciﬁc tasks were: development of a speciﬁc, sensitive and reproducible laboratory procedure for MTBE/BTEX quantiﬁcation; characterisation of MTBE degradation capacities under groundwater conditions; and identiﬁcation of favourable electron acceptors for biodegradation.

8.3.5.1 8.3.5.1.1

Description of Results Laboratory Procedure

A ﬁrst step was the development of a speciﬁc, sensitive and reproducible laboratory procedure for MTBE/BTEX quantiﬁcation via gas chromatographic analysis. With the GC/FID system used (6890, Hewlett-Packard) coupled with an automatic headspace sampler (HS40XL, Perkin-Elmer) it is

WATCH. Water Catchment Areas

525

possible to determine these compounds in the environment. The detection limits are 4.0 mg l1 (MTBE), 5.0 mg l1 (ETBE) and 1.0 mg l1 (each BTEX compound). For routine analyses this equipment is suﬃcient. For recurrent periodical identiﬁcation it is also possible to use a GC/MS system.

8.3.5.1.2

Soil and Groundwater Sampling

In order to estimate how natural processes aﬀect MTBE biodegradation and retardation in laboratory experiments, sediment samples had to be obtained from the groundwater zone. Those natural groundwater sediments are expected to contain a suﬃcient amount of autochthonous bacteria. Retardation eﬀects can be investigated best in original soil samples. Due to the local variation of indigenous microorganisms, signiﬁcant investigations of naturally occurring biodegradation entail using sediment samples from sites with diﬀerent hydrogeological and geochemical conditions. Therefore soil samples were taken from diﬀerent sites in eastern Germany, some of them contaminated by MTBE and BTEX. Additional sediment samples were collected from a project model site in Du¨sseldorf, Germany.

8.3.5.1.3

Biological Degradation Experiments

Comprehensive laboratory experiments on biological degradation were performed under groundwater similarity conditions. They enabled an estimation of the capacity of naturally occurring processes of MTBE degradation in groundwater with respect to favourable electron acceptors (biochemical reaction partners of the organic pollutants, e.g. oxygen, nitrate, iron(III), sulfate). An important aim was to identify electron acceptors supporting MTBE mineralisation. Degradation experiments were performed with aquifer sediment samples collected from 13 diﬀerent sites. Investigations concentrated on the search for eﬀective electron acceptors and on the possible formation of MTBE metabolites. The experiments were established under six diﬀerent conditions promoting aerobic degradation, denitriﬁcation, sulfate reduction, Fe(III) reduction, Mn(IV) reduction and methanogenesis. In additional experiments, the MTBE degradation behaviour of groundwater samples collected in a contaminated aquifer at a reﬁnery site was studied under aerobic and sulfate reducing conditions.

8.3.5.1.4

Aerobic Degradation

It was clearly obvious apparent that oxygen is the preferred electron acceptor for MTBE biodegradation. Several aerobic (oxygen-containing) experimental microcosms showed a signiﬁcant decrease in MTBE concentrations, even after multiple re-addition of MTBE. The sediments of all these microcosms came from sites which had been contaminated by MTBE. Sediments from completely unpolluted sites were unable to biodegrade MTBE in a time scale of 10 months.

526

Chapter 8.3

Aerobic MTBE-Degradation GW 140000

MTBE concentration [µg/l]

120000 100000 80000

GW 1 GW 2

60000

GW St

40000 20000 0 28.11 13.12 28.12 12.1

27.1

11.2 26.2 13.3 28.3

12.4

27.4 12.5

Time

Figure 8.3.10

Reﬁnery groundwater experiments GW 1, GW 2 (parallel control) and GW St (sterile control).

Groundwater samples collected in a contaminated aquifer near a second reﬁnery site showed eﬃcient MTBE degradation capacities. The groundwater of these wells had an MTBE contamination of 45 mg l1 and no BTEX content. Initial MTBE concentrations of 13 mg l1 in the batch experiments (after dilution with aerobic media) could be degraded completely within 2 months. Re-added MTBE (up to 80 mg l1) was completely consumed in less than 2 weeks (Figure 8.3.10).

8.3.5.1.5

Anaerobic Degradation?

Under anaerobic (oxygen-free) conditions no signiﬁcant MTBE degradation was observed, in spite of the successfully achieved anaerobic milieu and ongoing electron acceptor decrease in most microcosms. Only some experiments under iron and sulfate reducing conditions seemed to show a MTBE degrading capacity with very low reaction rates.

8.3.5.1.6

Which Other Factors Aﬀect MTBE Degradation?

Additional experiments showed the inhibiting eﬀects of BTEX (benzene, toluene, ethyl benzene, xylenes) contaminations on MTBE biodegradation. BTEX are typical gasoline constituents and range up to 50% by volume in gasoline. They are expected as co-contaminants of MTBE in many gasoline spill and leakage incidents.

WATCH. Water Catchment Areas

527

BTEX, in comparison to MTBE, are more favoured and more rapidly used substrates for the investigated groundwater sediment bacteria. The ‘‘inhibition’’ appears as an eﬀect of substrate preference causing consecutive degradation of BTEX and MTBE. Also, the postulated MTBE metabolite TBA was degraded faster by the bacteria culture than MTBE; this observation could explain why TBA does not accumulate during MTBE biodegradation in the experiments. Clarity was sought on a co-metabolic MTBE degradation during consumption of other organic carbon sources. We tested the eﬀect of the known, well degradable co-substrates sodium acetate, sodium succinate and methanol on biodegradation velocity of MTBE. Often several environmental pollutants which are not eﬀectively useful as the sole carbon source can vanish during a co-metabolic degradation of microorganisms, which degrade other available primary substrates. The eﬀect can be explained by an unspeciﬁc activity of enzymes, which are produced in the metabolism of the primary substrates. However, all these amendments, except methanol, decelerated the degradation of MTBE signiﬁcantly compared to the batch test without any addition of other carbon sources. A real enzymatic co-metabolism was not observed.

8.3.5.2

Conclusions and Perspectives

Experimental results conﬁrmed the recalcitrant and persistent character of MTBE under groundwater conditions. An eﬃcient natural attenuation of this hazardous compound cannot be expected at each site. As demonstrated in the literature and observed in our own experiments, biodegradation can reduce MTBE pollutions signiﬁcantly only in the presence of oxygen (aerobic conditions). Our experiments provided further evidence of a non-ubiquitous distribution of an aerobic degradation potential. An occurrence of a potential MTBE degradation was only found in aerobic microcosms with sediments from MTBE-contaminated sites. MTBE represents a serious threat for groundwater bodies and regarding the European Framework Directive on Water, MTBE should be a cause of alarm as a typical contaminant that accompanies production, distribution and sale of petrol products. A main demand is to include this compound in standard measuring and monitoring programmes as well as risk assessments related to drinking and groundwater questions. It is necessary to include MTBE threshold values in European drinking water regulations. Chemical research should concentrate on the search for an alternative antiknock fuel additive. Data about the consecutive degradation of BTEX and MTBE due to the preference of the BTEX compounds by microorganisms cause us to assume that a biodegradation of MTBE is not very probable if BTEX are present in the MTBE-contaminated groundwater plume. This is one reason for the signiﬁcantly longer plumes of MTBE than those of BTEX in the ﬁeld, which have been reported by several authors. This fact and the properties of MTBE, which are discussed above, have a great weight for real cases of remediation planning at contaminated sites. Natural attenuation strategies for treatment of MTBE should only be implemented under

528

Chapter 8.3

a veriﬁed availability of oxygen and the absence of other well degradable organic hydrocarbons including co-contaminants such as BTEX.

8.3.6

Development of an Early Warning and Management System for Groundwater Resources

The aim was to develop a protocol for the management of groundwater resources at water catchment scale. The protocol provides a decision support tool for water suppliers, regulatory agencies/authorities and/or their advisors for the quantiﬁcation and reduction of risk related to the water resource for the catchment area.

8.3.6.1

Description of Results

By implementing an intelligent early warning and management system, countermeasures against pollution of drinking water resources can be initiated earlier, more cheaply and more eﬀectively, thereby minimising the need for high-cost water treatment at drinking water plants. The development of the protocol for water management will increase the awareness of water suppliers with respect to possible measures to be taken to protect the safe abstraction of clean water. The dynamic reﬁnement and adaptation of monitoring will support economic eﬃciency. The deﬁnition of strategies for implementation of such a decision support tool requires an advanced understanding of the water cycle within the area of interest. The user is guided through the seven steps that are required for the basic set-up of an early warning and management support tool: initial delineation of water catchment area; conceptual model for catchment area; interpretation/understanding of conceptual model/water cycle within catchment area; identiﬁcation of risk within catchment area; analysis of risks within catchment area; optimisation of monitoring for identiﬁcation of potential pollution of drinking water; and contingency plans for protecting water supply. The protocol allows the end user to ﬁll this basic framework with casespeciﬁc information and to identify in detail additional work that is needed. Figure 8.3.11 gives an overview of this stepwise and iterative approach.

8.3.6.1.1

From the Delineation of the Catchment Area to a Conceptual Understanding of the Water Cycle

The delineation of the groundwater catchment is required for all further activities and it is the basic step of setting up a conceptual understanding of

529

WATCH. Water Catchment Areas Delineation of catchment area Preliminary delimitation of area of interest for focus of effort Sophistication of prelim. assessment

Conceptual model Analysis and interpretation of geological and hydrogeological data for area of interest

Data review and analysis

Iteration Possibly more data needed

Understanding of water cycle Interpretation of conceptual model and analysis of water cycle in catchment area

Analysis of risks Identification of risks that may cause pollution of abstracted water

Inventory of risks Inventory of risks for pollution within catchment area

Optimisation of monitoring Analysis of placement and frequency of monitoring

Contigency plans Plans for minimising effects of pollution

Figure 8.3.11

Basic set-up of the early warning and management support tool.

the groundwater system, the conceptual hydrogeological model. The conceptual model provides a reliable basis for the integration of more complex soil– water–sediments systems, and for the deﬁnition of critical control points, it is a prerequisite for the identiﬁcation, evaluation, control and management of multi-component systems. It allows a more reliable application of numerical models to evaluate risk scenarios and management options. Thus it is the basis for reliable decisions in groundwater and integrated catchment area management. By integrating the conceptual system understanding in a numerical model system, an improved, reliable understanding of the water cycle was derived.

8.3.6.1.2

From Analysis of Risks to Contingency Planning

The general procedure to come from the analysis of risks to contingency planning is a four-step approach (Figure 8.3.12). The estimation of risk-sensitive

530

Chapter 8.3

Contaminationsensitive sensitiveareas areas Contamination surfaceuse use ••surface • subsurface structure • subsurface structure ••t travel times

Recommendationsfor formonitoring monitoring Recommendations

• suitable monitoring points • suitable monitoring points • schedule • schedule • frequency • frequency Impacton onwater watersupply supply Impact • parameters st • occurrence ••11st occurrence expectedmaximum maximum ••expected

Contingencyplans plans Contingency groundwater ••iningroundwater extractionwell well ••atatextraction rawwater water ••ininraw

Figure 8.3.12

Preparation of contingency plans.

areas requires a combination of surface use scenarios with a description and prognosis of groundwater quality- and quantity-relevant processes. This enables an assessment of existing and planning of future monitoring strategies in terms of the following questions. What are representative monitoring points for risk sensitive areas? What parameters have to be measured and what is their usual operation window? When does monitoring have to start at what distance from risk-sensitive areas and potential receptors? What is the optimised frequency to survey the spreading of pollution? Based on such a monitoring programme the potential impact on water supply facilities can be estimated. When is a pollution to be expected to occur in an extraction well? What is the expected maximum concentration at the target of protection? What is the timeframe available for contingency planning and countermeasures? The availability of answers to all these questions, provided by potential pollution scenarios, allows the set-up and design of groundwater catchment area-speciﬁc measure plans to protect groundwater resources (Figure 8.3.13).

8.3.6.2

Conclusions and Perspectives

The protocol developed is of value to water suppliers, regulatory agencies/ authorities and/or their advisors for the quantiﬁcation and reduction of risk

531

WATCH. Water Catchment Areas alarm

continuous monitoring Contamination occurs in a water catchment area.

• nature of the contamination source: diffuse/point source?, singular event or continuous release? • substance(s) and their amount • physical and chemical data • ecotoxicological data • applicable law (EU and national directives) • hydraulic data • concentrations in groundwater and soil

Drilling of new monitoring wells at appropriate surveying points.

Stop of the release of the contaminant as soon as possible! NO

Are there enough monitoring points? NO

YES

Is the raw water endangered?

increased control of the raw water and the monitoring wells in the catchment area: • concentration of pollutant • control of the general chemical conditions • concentration of degradation products

NO

YES

raw water endangered: • excavation, recovery of pollutant • protection of raw water wells by physical barriers, changes in hydraulic conditions (protection wells) • changes of the raw water well itself to change hydraulic conditions

Will the contaminant reach the raw water wells?

NO

building of a treatment system with respect to following aspects: • nature of pollutant • amount of water treated per day • peak capacity needed • investment and maintenance costs • future developments (water consumption per person, population development in supplied area, possible changes in raw water quality)

YES

Is there a water treatment system present?

YES

changes in the water treatment system: • infiltration into the ground and recovery • aeration • ozonation • advanced oxidation processes (AOP) • adsorbing resins • granular activated carbon • UV-light • dilution with uncontaminated water

Drinking water according to the restrictions of law.

Figure 8.3.13

Measure plans: decision tree for contingency planning if pollution occurs in water catchment areas.

related to the water resource for the catchment area. Anthropogenic trace compounds like MTBE that are meanwhile widespread in the environment are a good example for the importance of a reliable groundwater management system that also keeps an eye on the integration of related environmental compartments. It allows those who are responsible for groundwater bodies

532

Chapter 8.3

and catchment areas to set up new and eﬀective management strategies for their water resources as well as to improve and optimise existing approaches for a sustainable use of groundwater resources. Setting up an integrated understanding of the complex interactions of processes within a groundwater body and providing an interface for interactions with soil sediments and surface water are the bases for socioecological and economical sound management planning.

Acknowledgement and Further Information The project was funded by the European Commission under the Fifth Framework Programme within the Energy, Environment and Sustainable Development Programme. The authors of the article are solely responsible for the contents of this publication. It does not necessarily represent the opinion of the European Commission. The European Commission and the partners are not responsible for any use that might be made of WATCH data. Further information on the WATCH project is available at: http://www. dechema.de/watch.

9. Groundwater Status Assessment

CHAPTER 9.1

Methodology for the Establishment of Groundwater Environmental Quality Standardsw DIETMAR MU¨LLERa AND ANNE-MARIE FOUILLACb a

Umweltbundesamt GmbH, Spittelauer Laende 5, AT-1090 Wien, Austria; Bureau de Recherches Ge´ologiques et Minie`res (BRGM), 3 avenue Claude Guillemin, FR-45060 Orle´ans ce´dex 2, France b

9.1.1

Introductory Remarks on Science–Policy Integration Needs

Developed under Article 17 of the European Union Water Framework Directive (2000/60/EC),1 the Groundwater Directive (GWD) (2006/118/EC)2 sets out criteria to assess the chemical status of groundwater bodies (see Chapter 3.1). In accordance with an analysis of pressures and impacts, European Union member states will be required to identify pollutants that are representative for groundwater bodies found as being at risk and for which threshold values should be established as (environmental) quality standards at the most appropriate scale (national, river basin district or groundwater body). BRIDGE (background criteria for the identiﬁcation of groundwater thresholds) as a policy support-oriented research project under the 6th Framework Programme of Research and Technological Developments aimed to contribute to the Water Framework Directive (WFD) Common Implementation Strategy (CIS) by developing a common methodology for deriving groundwater thresholds. To support sustainable groundwater management the methodology shall ensure that the derivation process for groundwater thresholds will be based on w

The views expressed in this chapter are purely those of the authors and may not in any circumstances be regarded as stating an oﬃcial position of the European Commission.

535

536

Chapter 9.1

common criteria, and comparability of status classiﬁcation for groundwater bodies across Europe can be assumed.

9.1.2

Towards an Integrated Management of Groundwater Resources

Groundwater is a key environmental resource and factor supporting sustainable regional development. This is recognised by the WFD and the new GWD which represent a major move in water policy towards an integrated management of groundwater resources and provide a new regulatory setting for the protection of groundwater quality. The ‘‘prevent and limit’’ concept of the former Groundwater Directive 80/68/EEC3 (to be repealed under the WFD by 2013), aiming to control point sources and to protect groundwater quality at a local scale, is nowadays complemented by the concept of managing the environmental status of groundwater resources, aiming to control widespread pollution by diﬀuse sources and to preserve groundwater quality at regional scales. According to the environmental objectives deﬁned by the WFD the ﬁnal goal of groundwater management at larger scales is to preserve good status (quantitative and chemical) to maintain groundwater in its functions to support aquatic and terrestrial ecosystems. Furthermore both the WFD and the GWD recognise groundwater as a resource for human uses.

9.1.3

The Framework: An Integrated Characterisation Process

A methodology to derive environmental thresholds for groundwater bodies has to be based on a sound conceptual model and an integrated characterisation process (see Figure 9.1.1) following three pillars: characterisation of potential pollutants and any parameters indicative of pollution, including description of the properties which inﬂuence their fate and transport, e.g. transport through and out of the aquifer, including transport in unsaturated zone, the behaviour within diﬀerent hydrogeochemical environments, ecotoxicology and toxicology and possible impacts on ecosystems; characterisation of groundwater bodies, including a description of the hydrogeochemical setting of aquifers, the background quality (natural and anthropogenically inﬂuenced) and any dependencies of water quality on quantitative aspects (like the variability of water levels due to the hydrological cycles during the year, groundwater to surface water interactions or the water balance in the long term); and characterisation of receptors, including aquatic ecosystems, dependent terrestrial ecosystems and groundwater. As a further fundamental basis the methodology has to refer to the deﬁnitions provided by the WFD, which are generally focused on possible impacts on

537

INDICATION FOR POOR STATUS Polluted groundwater (significant impacts possible)

Receptor-oriented standard alteration of groundwater quality

Variability in natural quality due to hydrochemistry

GOOD STATUS NATURAL QUALITY undisturbed groundwater

Figure 9.1.1

Quality poor due to naturally elevated concentrations, but no human impact (chemical status: good)

Level of detection/ quantification

Increasing concentration

Increasing pollutant concentration

ANTHROPOGENICALLY INFLUENCED

Methodology for the Establishment of Groundwater Environmental Quality

Groundwater quality and status: general relationship under the WFD.

receptors (associated surface waters, dependent terrestrial ecosystems, groundwater as a resource for human uses) and a qualiﬁcation of the signiﬁcance of these impacts (e.g. signiﬁcant diminution of the ecological or chemical quality of surface water bodies or any signiﬁcant damage to terrestrial ecosystems which depend directly on the groundwater body). Consequently any environmental threshold to indicate the change between good and poor chemical status has to be derived with a strong orientation to a risk-based approach and the likelihood that receptors are or might be harmed. Of course the diﬀerences between synthetic substances and anthropogenically introduced but naturally occurring substances also need to be taken into account, together with natural variations in quality both within and between groundwater bodies. Moreover, some substances, though considered as pollutants, may be present in naturally elevated concentrations. Such water may then be considered to have a poor quality (with regard to possible uses, e.g. drinking water abstraction) but still represents a good chemical status (see also Figure 9.1.1). It has to be pointed out that the environmental thresholds requested under the regime of the new GWD aim to assess the ‘‘overall health’’ of groundwater bodies and therefore are not to be compared to any quality standard under use to control local pollution of groundwater caused by point sources (prevent and limit).

9.1.4

Criteria to Assess Quality and Status of Groundwater

Groundwater quality assessment is a routine, which may aim at various possible objectives like monitoring human impacts at local scales, controlling

538

Chapter 9.1

drinking water supply or characterising pristine conditions. In contrast, groundwater status assessment as requested by the WFD is a new task, lacking experiences and deﬁned procedures. As for developing groundwater protection towards an integrated management of groundwater resources, criteria applied for quality and status assessment should be in correspondence and interlinked. To assess groundwater quality at local scales it is quite common to make use of natural background levels (NBLs) and generic reference values according to possible receptors, which are in general ecosystems and human uses. As status assessment can be understood as being an integrated assessment at larger scales of groundwater bodies, these two categories of criteria consequently have to be included. Moreover status assessment has to take into account that the general chemical quality of groundwater as well as the fate and transport of contaminants are determined by a variety of factors, where petrographic properties of rocks in the vadose and groundwater saturated zone, regional hydrological and hydrodynamic conditions and hydrogeochemical processes controlling the behaviour of natural and anthropogenic substances are of major importance. Thus status assessment would need to go beyond quality assessment and consider attenuation criteria like dilution, diﬀusion, retardation and degradation, which might be described referring to properties of contaminants, hydrogeological units, where speciﬁc hydrogeochemical processes govern, and the interaction with surface waters.

9.1.4.1

Natural Background Levels

Depending on data availability, NBLs can be deﬁned following a hierarchy of possible options. To unify the starting point for groundwater status assessment, BRIDGE has proposed a European aquifer typology, classifying 16 types (see Figure 9.1.2), and also has referenced NBLs from national studies accordingly. These NBLs might be used within the procedure to derive threshold values if no appropriate groundwater quality data are available but only the hydrogeological units of a speciﬁc groundwater body can be described. Given a groundwater body where a limited set of quality data is available, a second option of a simpliﬁed and practical approach to determine NBLs based on a pre-selection method can be employed. As a prerequisite for applying a simpliﬁed pre-selection method, common minimum requirements for groundwater quality data (e.g. deviation of the ion balance o10%) and appropriate pre-selection criteria to identify groundwater samples showing no signiﬁcant anthropogenic impact (e.g. nitrate o10 mg l 1) are to be deﬁned. Finally, given a groundwater body where a broad set of quality data is available, the third option to estimate NBLs is to apply scientiﬁcally sound methods (e.g. hydrochemical simulations, component separation by concentration separation analysis), which already have been established at national or international level.

9.1.4.2

Generic Reference Values

Besides NBLs, generic reference values according to receptors which might be harmed by groundwater contaminants are used for groundwater quality

Methodology for the Establishment of Groundwater Environmental Quality

Figure 9.1.2

539

European aquifer typology for NBLs (Research Centre Ju¨lich, draft of November 2006).

assessment. Giving a focus to ecosystems and human uses means consequently that environmental quality standards (EQSs) for surface waters or drinking water standards (DWSs) are to be transferred and linked into groundwater status assessment too. These reference values might be deﬁned at European

540

Chapter 9.1

level or at national level. With respect to the variety of possible substances contaminating groundwater it is also likely that for some substances no reference values are available at all. Again as a hierarchy of options it can be envisaged that uniﬁed European standards like, for example, the EQS for priority substances set out by the proposal of the European Commission (COM(2006)397 ﬁnal)4 are preferable to and overrule national standards. If no agreed European standards exist, national reference values can be used. Finally, for substances without established receptor-oriented reference values at European or national scale, a survey and evaluation of human toxicity or ecotoxicity data will be necessary. Depending on data availability the evaluation of these toxicity data should again be based either on agreed European procedures or on national agreements. As for ecotoxicity there are few data of tests with groundwater organisms available, but a transition of test results with aquatic organisms is generally recognised as appropriate. Considering long residence times and the relatively slow dynamics of groundwater, emphasis is to be given to ecotoxicity data of chronic eﬀects. Regarding surface water as a receptor, it has to be noted that the existing European groundwater quality standards set out by the WFD for nitrate and pesticides may not always be protective for ecological receptors. Particularly for nitrates, it is likely that protecting surface waters from eutrophication might demand lower quality standards for some groundwater bodies.

9.1.4.3

Attenuation Criteria

Pollutant attenuation may occur along the ﬂow path of groundwater, at the interface between groundwater and surface water (the hyporeic zone) and in the surface water itself. General consideration might be given to dilution, dispersion, diﬀusion, volatilisation, sorption and chemical and biological degradation. The physical, chemical and biological processes which occur in aquifers and which may act to naturally attenuate pollutants are well known and widely reviewed for the purposes of assessing point source pollution. The same processes can also act over much larger scales like groundwater bodies. It needs to be recognised that a real in-depth conceptual understanding of the groundwater and receptor system is still needed and consequently the demand on speciﬁc data increases. The current situation for the surface water interface is that knowledge is not suﬃcient at the moment and science is on its way. In comparison, the attenuation at the receptor surface water might be described easily, as in general dilution will be the major attenuating process. The estimation of quantity relationships of groundwater ﬂow against surface water ﬂow can be done rather easily by a variety of methods to estimate the base ﬂow (e.g. by hydrograph separation, temperature or water quality surveys, tracer analysis). Therefore, a pragmatic approach is to consider dilution as separate, generic criteria, whereas the description of all other attenuating processes would need in-depth investigations, which might be necessary for a

Methodology for the Establishment of Groundwater Environmental Quality

541

ﬁnal status assessment but are hardly to be considered for the derivation of threshold values.

9.1.5

Receptor-oriented Quality Standards Derived by a Tiered Approach

To combine the described criteria within a receptor-oriented status assessment, a tiered approach for deriving appropriate threshold values for groundwater bodies is recommended. Tiers may provide intermediate levels based on increasingly detailed understanding of a groundwater body. Therefore, each tier will involve increasingly sophisticated levels of data collection and analysis. An initial analysis (tiers 1 and 2) can use conservative and rather simple criteria (e.g. tier 1: check of monitoring data against natural background levels; tier 2: check against environmental quality standards deﬁned for associated surface waters), whereas further steps of a detailed analysis (tiers 3 and 4) would mean a thorough evaluation of speciﬁc groundwater body characteristics (e.g. tier 3: back calculation for an associated surface water body taking into account the contribution of the groundwater to the total pollutant burden; tier 4: taking into account further attenuating capabilities of the subsurface environment, speciﬁcally for the aquifer, the hyporeic and the riparian zone). As a result of the detailed analysis the ﬁnal separation between good and poor status by a reﬁned threshold is identiﬁed (see Figure 9.1.3). With respect to the heterogeneity of groundwater bodies on the one hand and the limited data availability in practice on the other, it seems likely that the detailed analysis will often be limited to a rather simple third tier by describing groundwater and surface water interaction in terms of quantity relationships.

9.1.6

How to Determine a Threshold Value (Example: Surface Water)

Groundwater thresholds are generally only required to be established if a groundwater body has been characterised as being at risk for a speciﬁc pollutant. Relevant risks have already been identiﬁed by member states during the process of characterisation as set out by the WFD (Article 5). Given the situation of a groundwater body which may have an impact on an associated surface water, the threshold value for groundwater should consider natural background levels, EQSs and attenuation (dilution) as described. This means that status assessment may follow a maximum of four possible tiers and the diﬀerent tiers may refer to the following sequence of considerations. Tier 1: natural background level. Pollutant concentrations within the range of NBLs would cause the groundwater body to be determined as ‘‘good status’’. For naturally occurring substances background levels

542

Chapter 9.1 MONITORED DATA

Derive NBL (according to Annex I) Tier 1

Is [pollutant] > NBL? Set threshold = NBL

OR

Tier 2

Yes

Tier 3

No

Is [pollutant] > (QS/DF)?

(according to Annex II)

Check for trends

Status = GOOD

Is [pollutant] > QS?

Tier 2a Derive TV

No

Set threshold = QS (or NBL if exceeding QS)

Check for trends

Status = GOOD No

Yes

Set threshold = (QS/DF)*AF

Set threshold = QS/DF

Check for trends

Status = GOOD

Is [pollutant] > TV? Tier 4

Rules

Is [pollutant] > (QS/DF)*AF?

1.

Use the appropriate quality standard, QS. If ecological risk use EQS.

Yes

Does appropriate Investigation show that conditions for good chemical status are not met? Yes

Status = POOR

Figure 9.1.3

No

If human health risk use DWS.

Status = GOOD

2.

If dilution factor, DF, not known assume = 1.0

3.

If attenuation factor, AF, not known assume = 1.0

4.

In check for Trends use ALL triggers-consider need for trend reversal if crossing each trigger

No Check for trends

Define Objectives and Measures

Flow chart for derivation of threshold values.

have to be considered as concentration ranges and a consequently derived threshold has to be set above this range to allow variation. For substances without a natural origin, the connected thresholds would be zero but for practicality has to be referenced by a distinct factor to the limit of detection (LOD). Tier 2: quality standards. The groundwater monitoring data are simply compared to the quality standard established for the associated surface water body. If the quality of groundwater ﬂowing into the receptor is below the receptor quality standard then the risk to that receptor must be derived from some other sources and groundwater is to be determined as ‘‘good status’’. Tier 3: considering dilution. The groundwater provides only a proportion of the pollutant ﬂow to the receptor. The concentration of a substance in the groundwater is compared to the quality standard multiplied by a dilution factor (DF). Where the dilution factor is not known it is assumed to be 1. Tier 4: considering attenuation. The pollutant might be attenuated on its path to the receptor. If further information is known on the attenuation a relating factor could be introduced. This is ﬁnally the ultimate threshold in this scheme, where exceeding monitoring data triggers speciﬁc investigations and indicates the likelihood that a groundwater body would need to be classiﬁed as being at ‘‘poor status’’ and measures

Methodology for the Establishment of Groundwater Environmental Quality

543

under the subsequent River Basin Management Plan will have to be deﬁned.

9.1.7

Compliance Regime for Groundwater Quality Standards

A standard may have widely diﬀerent impacts and implications, depending on the compliance regime selected. Therefore, a clear deﬁnition is at least necessary for a summary statistic (such as a mean, 95 percentile or a maximum value); a time period over which compliance is assessed; and an area over which the criteria are applied and interpreted via the monitoring network. Following the concept of a tiered and receptor-oriented approach the compliance regime may vary widely; for example: data aggregation over time only for monitoring stations at tier 1 and data aggregation over all monitoring stations and time within a deﬁned area (groundwater body) at tier 2; or data aggregation over monitoring stations with an associated surface water system. It has to be recognised that any quality standard (limit values, thresholds, indicators) must be associated with explicit compliance regimes that enables a straightforward assessment of compliance via the monitoring data obtained from the operational monitoring network. Therefore, the continuation of discussions between scientists, regulators and stakeholders after agreeing upon the threshold methodology is a fundamental prerequisite to develop a consistent approach for groundwater status assessment. The co-ordination to practice and the ongoing implementation of the WFD by the member states is another challenge and chance in developing a common and comparable groundwater management which pays respect to a Europe of regions and may contribute to enhanced sustainability.

9.1.8

Conclusions

This chapter highlights some of the main ﬁndings of the BRIDGE project,5 which provide a scientiﬁc background for agreeing about a common methodology for the establishment of threshold values in the light of the requirements of the new GWD (see Chapter 3.1). It does not prejudge about the ﬁnal methodology which will have to be agreed and adopted by the member states in the framework of the Working Group on Groundwater of the CIS of the WFD (see Chapter 4.1).

544

Chapter 9.1

Acknowledgement BRIDGE was funded by the European Commission, DG Research, within the 6th Framework Programme under Priority 8 (Contract No. 006538 (SSPI)— Scientiﬁc Support to Policies).

References 1. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities, L 327, 22.12.2000, p. 1. 2. Directive 2000/118/EC of the European Parliament and of the Council of 12 December 2006 on the protection of groundwater against pollution and deterioration; Oﬃcial Journal of the European Communities, L 327, 27.12.2006, p. 19. 3. Council Directive 80/68/EEC of 17 December 1979 on the protection of groundwater against pollution, Oﬃcial Journal of the European Communities, L 20, 26.1.1980, p. 43. 4. COM(2006)397 ﬁnal: proposal for a directive of the European Parliament and of the Council on environmental quality standards in the ﬁeld of water policy and amending Directive 2000/60/EC. 5. BRIDGE reports available through www.wfd-bridge.net.

CHAPTER 9.2

Pesticides in European Groundwaters: Biogeochemical Processes, Contamination Status and Results from a Case Study CHRISTOPHE MOUVET Bureau de Recherches Ge´ologiques et Minie`res (BRGM), 3 avenue Claude Guillemin, FR-45060 Orle´ans ce´dex 2, France Present address: Lyonnaise des Eaux-Suez, 11 Place Edouard VII, FR-75316 Paris ce´dex 09, France

9.2.1

Introduction

Numerous articles in newspapers and consumer magazines, as well as reports on television, are now heightening public awareness of the presence of pesticides in groundwater (GW). Statements such as ‘‘pesticides contaminate the groundwater’’ are, indeed, frequently read or heard. One might suppose that in a modern world where science plays an increasing role in shaping public opinion about major environmental issues, the scientiﬁc community has all the hard data and detailed understanding it needs to support such statements. Actually, this is not exactly the case. Descriptive data for many aspects of GW contamination by pesticides are still limited, and the basic processes governing the fate and transport of pesticides in the unsaturated and saturated zones have not yet been suﬃciently studied. The information presented below is broken down into ﬁve sections. The ﬁrst section is a review of studies conducted on the major processes involved in the transport of pesticides from the soil to and in GW. The second and third sections discuss the status of GW contamination by pesticides at the European scale, and in selected European countries. The fourth section describes a case study conducted at aquifer scale in France. The ﬁnal section consists of prospective comments and suggestions on pending research questions that

545

546

Chapter 9.2

should be addressed in the future, notably trying to link science and policy in the ﬁeld of GW contamination by pesticides.

9.2.2

Major Processes Involved in the Transport of Pesticides from the Soil to and in Groundwater

Before they are detected in GW, pesticides applied to the soil surface must pass through the soil (Figure 9.2.1), then the unsaturated zone below the root zone (UZ) and ﬁnally travel in the saturated zone (SZ) from their entry point in the SZ to a point where GW quality is being monitored (well, piezometer, waterworks). The processes controlling the fate of pesticides in the soil compartment are extensively documented in publications focusing on speciﬁc issues such as preferential transport, ﬁeld observations1 and modelling,2 the relationship between sorption and microbial degradation,3 ﬁeld studies on leaching through the root zone,4 formation of bound residues5 and volatilisation from the soil surface and plant leaves.6 By contrast, the information available on the fate of pesticides in the UZ and SZ is far more limited. The deepest layers usually studied are described as Precipitation Pesticides inputs

Volatilization Evapotranspiration

Run-off

Soil: sorption, degradation Leaching

Unsaturated zone: mainly vertical circulation, sorption, degradation

Recharge

Satured zone: mainly horizontal circulation, dilution, sorption, degradation

Figure 9.2.1

Main processes involved in the transport of pesticides from the soil surface through the unsaturated zone and into the groundwater. The greater the font size of each process in a given compartment, the greater its importance. The true compartment thickness is always the smallest for the soil, and usually the greatest for the unsaturated zone below the root zone.

Pesticides in European Groundwaters

547

subsurface soil, which usually means depths of about 40–100 cm7,8 that do not, by far, represent the entire UZ. The short review below will address only data from studies in the UZ beyond the root zone and in the SZ, with a focus on two main processes, degradation and sorption.

9.2.2.1

Sorption and Degradation

The partition (or sorption) coeﬃcient of a pesticide between the aqueous and solid phases, Kd (ml g1), is most often determined from batch equilibration studies: Cs ¼ Kd C1

ð1Þ

where Cs is the equilibrium concentration in the solid phase (mg kg1), Kd the sorption coeﬃcient and Cl the equilibrium concentration in the liquid phase (mg l1). When the sorption isotherm with several equilibrium concentrations is not linear, ﬁtting with the Freundlich isotherm is done: Cs ¼ Kfn C1

ð2Þ

where Kf is the Freundlich sorption coeﬃcient and n is the Freundlich exponent. Kd and Kf are often normalised to the mass fraction of organic carbon of the solids investigated, foc, to yield Koc: Koc ¼ Kd =foc

ð3Þ

Alternatively, Koc can be estimated from the hydrophilicity of the solute based on its octanol–water partition coeﬃcient, Kow.9 However, such empirical relations are based on assumptions that may fail under ﬁeld conditions.10 When studying transport through porous media, the retardation factor, Rf, is a measure of the eﬀective velocity of a solute compared to that of water. Rf is commonly calculated using the equation Rf ¼ 1 þ rs Kd =n

ð4Þ

where rs is the bulk density (g cm3), Kd the sorption coeﬃcient and n the porosity of the solid matrix.10 The degradation of pesticides is most often studied with 14C-radiolabelled molecules that enable easy, precise and inexpensive measurement of the mineralisation (formation of 14CO2) of the parent compound and the formation of bound residues. The pesticide metabolites (transformation of the parent compound) are commonly analysed by GC or HPLC techniques. Degradation studies conducted under laboratory conditions greatly outnumber those run under ﬁeld conditions. The duration of most experiments is of the order of weeks, during which the decrease with time in concentration of the initially added pesticide is monitored.

548

Chapter 9.2

Results from degradation studies are usually described using ﬁrst-order kinetics from the equation ln S ¼ ln S0 kt

ð5Þ

where S is the amount of pesticide at time t (mg kg1 dry soil), S0 is the amount initially added, k is the ﬁrst-order rate constant (day1) and t is the time (day). Half-life values (represented by the symbols DT50 or T1/2) are calculated on the basis of ﬁrst-order kinetics: T1=2 ¼ ln 2=k

ð6Þ

where k is the ﬁrst-order rate constant. Depending on the experimental protocol, the half-life refers to mineralisation (the pesticide is completely degraded to 14 CO2) or to dissipation (degradation+transformation+volatilisation+sorption, if the latter two are not accounted for by other means).11 For any given pesticide, the half-life values are strongly dependent on experimental conditions such as temperature, water content and initial concentration of pesticide. The half-life values are also strongly dependent on the solids in the presence of which biodegradation is studied. For instance, the physicochemical characteristics of the solids inﬂuence sorption (and in turn biodegradation), and the microbial activity controls biodegradation and biotransformation.

9.2.2.2

Biodegradation in Microcosms with Solids from the Unsaturated Zone Below the Root Zone

Most degradation studies beyond the root zone resort to solids collected at diﬀerent depths, spiked with pesticides, then incubated in the laboratory under controlled temperature and water-content conditions. Studying pesticide degradation in unsaturated conditions is hampered by the diﬃculty in setting up and running microcosms under realistic moisture conditions.12 The general trend is for the potential for pesticide mineralisation to decrease with depth. For atrazine (At), T1/2min for subsoils (0.7–10.0 m below the surface, mbs) ranges from 120 days11 to inﬁnite: no mineralisation.13 At mineralisation of the initial 0.1 mg kg1 spike after 28 days of incubation reaches 8% at a depth of 2.5 m, 1% at a depth of 5.5 m and becomes negligible (o0.2%) at a depth of 10 m.11 An increase in At half-life from 77–101 days in the surface soil (0–0.15 mbs) to 900–1700 days in the vadose zone (1.75–2.20 mbs) is reported in another study.14 For isoproturon (IPU), bentazone and mecoprop-p (MCPP) incubated under laboratory conditions with 20 topsoil (TS; 0 to 15 cm depth) and 20 subsoil (SS; 50 to 60 cm depth) samples, the mean DT50 in the SS (238 days for bentazone, 44 days for IPU and 42.3 days for MCPP) is 3.6 to 13.5 times longer than in the TS.7 Compared to the TS, degradation in the SS appears to be more

Pesticides in European Groundwaters

549

variable.7 Another study on MCPP and IPU with sediments from an aerobic sandy aquifer 0.06 to 7.68 m below the surface (the water table is at 2.56 mbs) shows an increase in DT50min from approximately 20 days for MCPP and 700 days for IPU in the plough layer, to more than 5 years below 4.73 m for MCPP and 6 years in the subsurface for IPU. No lag phase is reported for MCPP or IPU.13 A mini-review of microbial degradation of IPU and related phenylurea herbicides in and below agricultural ﬁelds conﬁrmed the general trend in decreasing potential for biodegradation below the plough layer.15 No reports of signiﬁcant mineralisation of the phenyl structure of phenylurea herbicides in samples from subsurface environments have yet been published. However, the formation of the monodesmethylated metabolite of IPU (initial concentration: 100 mg l1) has been reported in laboratory studies with samples from the unsaturated zone (10.3–10.75 mbs) and the saturated zone (18.8–19.3 mbs) of an aerobic chalk aquifer.12 Studies on biodegradation must, therefore, include transformation products in addition to the formation of 14CO2. The generally observed inhibition of mineralisation below the plough layer can be explained by various factors such as unfavourable growth conditions for indigenous microbial degraders (absence of available carbon and/or nutrients) or the lack of competent pesticide-degrading microbial populations. The latter seems to be the key factor13–15 since populations capable of metabolising a simple aromatic substance such as benzoate have been observed in the vadose zone,14 and microbiological characterisation reveals that subsoils have a viable and active population, although direct counts of bacteria are consistently lower in subsoils than in surface soils.16 Pesticide half-lives in the subsoil being much longer than in the plough layer, the overall tendency is to consider that biodegradation in the vadose zone has a negligible role in pesticide fate. However, the much longer residence time of pesticides in the UZ than in the topsoil can lead to signiﬁcant long-term degradation even if degradation rates are very slow. The mean time required for water to move down 1 m below the root zone ranges, for loess soils, from 553 days17 to 3318 days,18 and is usually estimated to be 365–730 days for chalk.19 Such travel times obviously leave time for some degradation to take place. Modelling work shows, for instance, that a half-life as long as 1825 days for MCPP combined with a residence time and a recharge rate typical of aquifers overlain by fractured clayey tills leads to ﬂuxes into the aquifer 1000 times lower than when no biodegradation is taken into consideration.20 The use of the PELMO model shows that the estimated GW concentrations of acetochlor are signiﬁcantly reduced, from 0.09 mg l1 with no subsoil degradation (default values from standard risk assessment scenarios) to o0.01 mg l1 when degradation data measured in the undisturbed ﬁeld’s unsaturated subsoil (2.6–3.1 mbs) are included.16 By contrast, the inﬂuence of long travel time through the unsaturated zone on the attenuation of pesticide ﬂuxes is sometimes used as a key parameter in considering that the impact of pesticides leaching below the root zone might be negligible on potable supplies.21

550

9.2.2.3

Chapter 9.2

Sorption on Solids from the Unsaturated Zone

Most of the data on pesticides in the vadose zone concern subsoils22,23 or sandy and/or Quaternary aquifer materials.24,25 Sorption data are also available for the European Senonian chalk26,27 or other rocks of more local distribution.28 It must be noted that all the available data are based on batch equilibrations conducted under saturated conditions, even though the solids are recovered from the UZ. Since sorption of non-ionic pesticides depends strongly on the organic carbon content of solids, which decreases with depth, pesticide sorption usually decreases with depth.14 For instance, IPU sorption coeﬃcients measured in vertical proﬁles of sandy clay loam are 3.9 times lower in the 50–60 cm layer (SS) than in the 0–10 cm layer (TS).7 The same study reports, by contrast, 1.8 times greater sorption coeﬃcients of bentazone in the SS than in the TS. The formation of non-extractable residues in batch experiments with solids from the vadose zone is reported for bentazone and IPU (for bentazone, 13.8 times greater in the SS than in the TS),7 and for At with up to 16.5% of the applied mass with solids from 172 to 200 cm depth.14 The variability of sorption coeﬃcients and formation of bound residues is higher in the SS than in the TS.7 When the organic carbon content is low, the role played by clay and iron or manganese oxides may become important.29 Some studies report signiﬁcant sorption of pesticides in the UZ.30–32 The sorption of dichlobenil and its polar metabolite, 2,6-dichlorobenzamide (BAM), by reduced layers of clayey tills sampled 17 m below the soil surface (mbs) is greater than (or at least equal to) the sorption on top soils, whereas the sorption by the oxidised clayey till is much lower than by the reduced layers.30 In a screening of vadose zone Eocene geologic materials (limestones, marlstones, clays and sands), the At, IPU and metamitron sorption coeﬃcients of two materials, a lignitic clay (sampled 3.8 mbs) and a clay with low organic carbon content (sampled 2.8 mbs), are reported to be higher than those of the TS (Figure 9.2.2).32 The complex interactions between sorption, degradation and water movement in the UZ are illustrated by the study on the fate of At in the vadose zone at a till plain site in central Indiana.33 At and its major metabolites are retained and almost completely degraded in the vadose zone. The retention of At and its metabolites is attributed to evapotranspiration (the year of the study was substantially drier than normal, with measured precipitation only 65% of the long-term average), which creates surface directed hydraulic gradients in the vadose zone. Prolonged residence time in the vadose zone provides an opportunity for biological and chemical processes to reduce At and its metabolites to further by-products. However, as a result of macropore ﬂow, small quantities of At and its degradation compounds may reach the SZ.33

9.2.2.4

Biodegradation in Microcosms with Solids and Water from the Saturated Zone

As for the UZ, most published SZ studies are based on laboratory studies with a variable (usually low) degree of similarity to ﬁeld conditions. An important

Pesticides in European Groundwaters

Figure 9.2.2

551

Adsorption coeﬃcients (Kd) of atrazine, isoproturon and metamitron and stratigraphy of two vertical vadose zone proﬁles from the Bruye`reset-Montbe´rault catchment. Vertical bars around symbols indicate the layers where the samples were taken. Each sample is identiﬁed by a code between each graph and its corresponding stratigraphic log. Note that the scale of Kd values is diﬀerent in the two proﬁles (reproduced from Ref. 32, with permission).

source of dissimilarity is the range of concentrations studied. Whereas pesticide concentrations in GW are usually in the low microgram per litre range (see Section 9.2.3), most biodegradation studies making use of GW and solids from the SZ are conducted with concentrations in the 8–100 mg l1 range.12,13,34 Threshold concentrations above which the biodegradation rate accelerates gradually due to selective growth of speciﬁc biomass are reported to be in the range of 1 mg l1 for p-nitrophenol and 2,4-D (2,4-dichlorophenoxyacetic acid)35 and 10 mg l1 for MCPP [()-2-(4-chloro-2-methylphenoxy)propanoic acid].36 By contrast, a signiﬁcantly higher rate constant for bentazone transformation at lower initial bentazone contents has been reported.37 Concentrations of around 1 mg l1 are, however, clearly required if biodegradation rate measurements are to simulate the appropriate no-growth regime expected for pesticide-contaminated aquifers. Degradation rates also vary in relation to the history of exposure of the degraders to the pesticide. In situ long-term pre-exposure of aquifer sediments results in signiﬁcantly higher degradation rates of the phenoxy acids MCPP and 2,4-D.36 Half-lives in unpolluted aquifer sediment decrease from 500 days for 2,4-D and 1100 days for MCPP at 10 1C to approximately 5 days after

552

Chapter 9.2

adaptation. The enhanced rate of degradation by adapted systems is maintained during degradation of the last residuals measured to less than 0.1 mg l1. The inﬂuence of incubation temperature can be very marked over a 5 1C interval. For instance, no bentazone degradation is measured in subsoils from 2.0 to 2.5 m (1 m below the water table) at 10 1C, but at 15 1C a half-life of 38 days is reported.37 A review of biotransformation of selected pesticides in SZ materials was published in 2000.38 The authors reviewed laboratory studies on dibromoethane (EDB; 3 studies), At (13 studies), acetanilide herbicides (propachlor, alachlor, metolachlor, propanyl; 6 studies) and a carbamate herbicide (aldicarbe; 5 studies). There is also one ﬁeld study in an aerobic aquifer for At. The initial pesticide concentrations vary from 1mg l1 to 27 mg l1. The review showed that a biotransformation potential exists for all 4 pesticides in SZ materials, although substantial diﬀerences in biotransformation occur for any given pesticide. These diﬀerences are related to diﬀerences in microbial communities and sediment properties. Of the 4 pesticides investigated, At is the least prone to systematic biotransformation. Several studies, including the one under ﬁeld conditions, do not show any transformation after periods of up to 161 days. In the case of acetanilides, 5 of the 6 studies show biotransformation: parts per billion concentrations may biotransform over a period of 5–18 months. The biotransformation of EDB and aldicarbe is observed in all the studies reviewed. The mechanisms of biotransformation are not characterised in any of the studies. It is therefore impossible to extrapolate the observed biotransformation potential to other sites and experimental conditions. The very limited biodegradation of At (and deethylatrazine and IPU) is also observed in a later study with solids and GW from the SZ of an aerobic sandy aquifer.39 By contrast, mineralisation of MCPP is reported in this study, with a strong lateral and vertical spatial variability at centimetre and metre scales.39 The vertical variability of pesticide biodegradation associated with solids from the SZ and overlaying UZ is illustrated by a 5.3-year incubation study on the rate of bentazone and At transformation conducted with humic sandy soil materials from layers below the GW table, at depths of up to 2.5 m.40 In subsoils with intermediate to high redox potential, At measurable half-lives using ﬁrst-order kinetics range from 0.16 to 1.6 years. Material from the top of the phreatic aquifer has a higher bentazone biotransformation rate constant than material from the layers just above. The presence of fossil organic material in the ﬂuviatile water-saturated sediment probably stimulates microbial activity and bentazone transformation. In addition to studies of pesticide biodegradation under laboratory conditions mimicking ﬁeld conditions, some studies focused on biodegradation and estimations of half-lives in the SZ from ﬁeld data at aquifer scale. Combining GW ages (measured by 3H/3He) and changes in pesticide concentrations with depth in an unconﬁned shallow (5 m) GW, the estimated half-lives in GW are 10 to 20 years for At, and more than 10 to 20 years for alachlor ethane-sulfonic acid.41 The chemical transport modelling of ﬁeld data from a shallow (water

Pesticides in European Groundwaters

553

table at a maximum depth of 10 mbs) sandy-till aquifer42 yields calculated halflives of 3470 days for At and 2770 days for deethylatrazine. The long observed or calculated residence times of pesticides in GW suggest that even very slow degradation rates could be of importance in determining future GW quality.14,42,43 Slow degradation rates mean that even in situations where all input to the aquifer ceases, as for example after a ban of all applications, it may take more than a decade for concentrations deep within the aquifer to drop to half of what they were during the period of At use.42

9.2.2.5

Sorption on Solids from the Saturated Zone

A study of a chalk aquifer in southern England26 shows a sorption potential for IPU within the chalk down to 7.2 mbs (Kd ¼ 0.46 l kg1, compared to Kd ¼ 4.4 in the 0.5 m subsoil). It is not clear whether this sorption is governed by the organic or the inorganic component. Since the thickness of the chalk formation (27 m) exceeds that of the soil (maximum 0.6 m) by a factor greater than the diﬀerence in Kd values of each component, the possible retardation of IPU compared to a non-sorbing solute in the movement through the chalk should not be underestimated. The fate of bentazone, IPU, 2-methyl-4,6-dinitrophenol (DNOC), MCPP, dichlorprop, 2,4-D and BAM was investigated in column experiments using a medium- to coarse-grained sand aquifer material with 0.02% organic carbon, pesticide concentrations of approximately 25 mg l1 and a temperature of 10 1C. The sorption data indicate that IPU and DNOC are signiﬁcantly retarded (Kd ¼ 0.14 and 0.32 l kg1, respectively), whereas no signiﬁcant retardation of the phenoxy acids (MCPP, 2,4-D and dichlorprop), BAM or bentazone is observed.44 In the case of IPU, the sorption values are lower than those reported in another study26 using a batch protocol and a chalk material with a higher organic carbon content (0.1–0.4%). The fate of metabolites may also be diﬀerent from that of the parent product. In the case of At, for instance, the hydroxylated metabolite sorbs very strongly to aquifer sediments whereas the parent compound exhibits no measurable sorption.24

9.2.2.6

Field Studies Involving both Sorption and Biodegradation

Studies where pesticide fate in aquifers is characterised in natural and rather well-characterised conditions have the disadvantage of not dissociating each of the concurring mechanisms (sorption, degradation, diﬀusion, dilution) but oﬀer the major advantage of real environmental conditions, including the issues of scale that cannot be addressed through laboratory studies. A natural-gradient study on transport in a saturated gravelly sand, organicrich (between 0.09 and 0.16%) aquifer showed a retardation factor (Rf) of 6.1 for At (injected at 3 mg l1) and 7.1 for alachlor (injected at 2 mg l1) over a 2 m travel length in the saturated zone.45 Due to the very low hydraulic gradient of

554

Chapter 9.2

the aquifer studied, 0.00072 m m1, and to the sorption of At, the average distance travelled by At over a year would be only about 10 m. During the 50 days of the experiment, no degradation product of At was detected (the alachlor metabolites were not analysed).45 Another experiment10 conducted in a similar aquifer except for a much lower organic carbon content (o0.015%) yields Rf of 1.2 for atrazine and cyanazine, 1.1 for deethylatrazine, and 1.3 for deisopropylatrazine, almost an order of magnitude greater than those reported in Ref. 45. Alachlor and metolachlor show Rf o1.3, while butachlor is somewhat more retained (Rf ¼ 1.65). No detectable loss solely attributable to degradation was observed for any of the substances.10 A third study46 diﬀered by the much higher concentrations injected in the aquifer, 400 mg l1, and the presence of MCPP as test compound in addition to At. The duration of the experiment (3 months) and the aquifer characteristics (Borden test site: shallow, aerobic, ﬁne- to medium-grained glacio-ﬂuvial sand, average GW velocity of 8 cm per day) were very similar to those of the two previous studies. However, the foc ¼ 0.02% was similar only to that of Ref. 10. Atrazine is slightly more retarded than chloride, Rf ¼ 1.2, and is not degraded after 96 days, whereas MCPP is not retarded but is signiﬁcantly degraded down to a concentration of 30 mg l1 after a 42–56 day lag phase. A fourth study,47 also conducted at the Borden test site, used an in situ microcosm approach where a core of aquifer material is isolated from GW ﬂow at a depth of 1.5–3 m in the ﬁeld. The biotransformation and biodegradation of glufosinate-ammonium (GLUF-NH4) was studied for 77 days at concentrations of 50–400 mg l1. For the lowest concentration, a loss of about 50% of GLUF-NH4 is observed within 20 days, presumably a consequence of biotransformation. Overall, GLUF-NH4 appears to be persistent under the conditions reﬂecting the in situ situation. Laboratory experiments show that the addition of a readily degradable carbon source promotes biotransformation and suggest, therefore, that the absence of biodegradation under in situ conditions is due to the lack of any appropriate carbon source. A continuous, natural-gradient, ﬁeld-injection experiment that lasted for 216 days in a shallow aerobic aquifer (average ﬂow velocity of 0.5 m per day) involved bentazone, MCPP, dichlorprop, IPU and BAM.48 Bentazone, BAM, MCPP and dichlorprop retardation is negligible, and only slight retardation of IPU was observed. No degradation of bentazone, BAM or IPU was observed in the aerobic aquifer during the monitoring period. The two phenoxy acids, MCPP and dichlorprop, were both degraded in the aerobic aquifer but with a lag phase that leads to their spreading, albeit at decreasing concentrations with time, beyond the 25 m long monitoring network. One study was quite speciﬁc because it involved an anaerobic aquifer.49 The aquifer was sandy and shallow (water table located 2.0–2.5 mbs), with an average pore water velocity of 17 cm per day. The substances studied were At, DNOC and 2,4,5-trichlorophenoxyacetic acid (2,4,5-T). Atrazine and 2,4,5-T (concentrations in the source well of 10–500 mg l1) were persistent during the approximately 18-day residence time in the aquifer. This residence time may, however, be too short to enable detection of degradation. In contrast, DNOC

Pesticides in European Groundwaters

555

(concentrations in the source well of 500–1500 mg l1) was rapidly removed from the water phase, likely through an abiotic reduction. Overall, the Rf values measured during the ﬁeld studies were greater than those calculated using sorption parameters from batch studies (Freundlich coeﬃcient, Kf) or from correlations between Kf and octanol–water partition coeﬃcient (Kow) and organic carbon content (foc), and from estimates of porosity and bulk density. Under-prediction of Rf when using the Kf/Kow or foc correlation may be due to the contribution of mineral surfaces to sorption when organic carbon content is very low.50

9.2.2.7

Synthesis of Published Results on Sorption and Biodegradation below the Root Zone

The number of studies using protocols very similar to actual natural conditions is extremely limited and in view of the highly heterogeneous nature of subsurface environments, such a limited body of literature makes it diﬃcult to draw general conclusions. Even though sorption and biodegradation are, in general, less eﬃcient in the vadose and saturated zones than in the soil layers, marked exceptions are reported, related to speciﬁcities of the pesticides, the solids and the biodegraders from the subsurface environments. Since the travel time through the UZ and the residence time in the SZ are usually long, even very low rates of mineralisation may be of importance for the natural attenuation of pesticides in aquifers. The complexity and heterogeneity of the processes involved hinder the prediction of the contamination of GW that might result from the use of pesticides. It is therefore necessary to examine the existing monitoring data to assess the status of GW with respect to pesticides, as presented in the following section.

9.2.3

Status of Groundwater Contamination by Pesticides at European Scale

The European Environment Agency (EEA, Copenhagen, Denmark) is the institution in charge of collecting data on GW quality at European scale and producing the related reports and databases. A short review of the information available through the EEA is presented below.

9.2.3.1

The Waterbase Data Base of the European Environment Agency, June 2006

Information from the EEA is downloadable from the EEA web site (http:// dataservice.eea.europa.eu/dataservice/). As of June 2006 (last upload spring 2004), the ‘‘Waterbase-Groundwater: quality water’’ database contains 20 221 records for data on nutrients, organic matter and selected pesticides. Data are provided on GW bodies (thickness of the GW body, depth to the GW, GW horizon, aquifer type (porous, karstic or fractured), GW body area; only 15% of the records are documented for all of these descriptors), and

556

Chapter 9.2

pressures (annual precipitation, percentage of arable land use within the area of the GW body; the database is almost complete for these two parameters). The pressure data do not include information on the pesticides applied or the soil types in the recharge area of the GW bodies. Additional data on GW body characteristics and pressures requested through the European Environment Information and Observation NETwork (EIONET) water data collection are contained in a working database at the Umweltbundesamt, Austria. These are used to prepare the indicator fact sheets (see below), but most are not yet consistent enough to be published. Possibly in line with the EEA report51 that concluded that, as of 2000, atrazine, simazine and lindane were the compounds analysed by the greatest (albeit low) number of countries and were the compounds most frequently detected, the graphs, tables and data sets presently accessible on-line from the EEA database include only these three substances. No data on pesticides more recently registered or on metabolites of old and recent pesticides are available on-line,52 even though it is recognised that metabolites are commonly detected more frequently and at higher concentrations than the parent compounds.53–58 Amongst the 25 European Union (EU) countries, the information contained in the pesticide database is extremely variable. No data are available for 12 countries, including some of the oldest EU member states, e.g. the Netherlands and Luxemburg. Data from 1993 (or even 1990) to 2004 are available for Austria, Belgium, Denmark, Germany, and Slovenia, but data from most other countries are only available since 2001. The number of representative GW bodies for which pesticide data are reported is highly variable: for instance, 184 in France, 3 in Denmark, 4 in Slovenia, 33 in the UK, 9 in Germany. The comparison of national data is hindered by the diﬀerences in, amongst other things, (i) the substances monitored, (ii) the monitoring strategy (sampling frequency and density) and (iii) the type of sampling wells. It remains therefore extremely diﬃcult to determine an overall status of pesticide contamination of GW at the EU scale. However, the database makes it possible to calculate the frequency of GW bodies whose mean concentration (annual mean values of the sampling sites) of atrazine, simazine and lindane is below or above the threshold value of 0.1 mg l1 for the EU10 or EU12 (the number of countries involved depends on the pesticide) and for each country providing data (Table 9.2.1). Values greater than 0.1 mg l1 are reported only for 6 of the 317 GW bodies in the case of At (1.9%), 1 out of 305 GW bodies for simazine (0.3%) and none of the 261 GW bodies monitored for lindane. Based on Table 9.2.1, the contamination of GW by pesticides at the European scale appears to be very limited. It must be emphasised, however, that the database includes only three substances, two of which (atrazine and lindane) have been banned in several EU countries for a number of years (a decrease in the concentration of atrazine and its metabolites has been reported for some GW bodies in Austria, France and Switzerland51,59). Furthermore, except for France, where the highest frequency of values 40.1 mg l1 is reported (2.7% of the GW bodies monitored for atrazine), the number of GW bodies for which

2 4 33

0 0

3 184 9 12 5 1

0 2.7 0 0 0 0

50

317 14 12 38

1.9 0 0 0

0 0

4 33

2

5 1

0 0 0

3 184 9

305 14 12 38

Total number of GW bodies

0 0.5 0

0.3 0 0 0

GW bodies with mean value 40.1 mg l1 (%)

Total number of GW bodies

Simazine GW bodies with mean value 40.1 mg l1 (%)

Atrazine

5 1 4 5

0 0 0

183 7

261 14 4 38

Total number of GW bodies

0

0 0

0 0 0 0

GW bodies with mean value 40.1 mg l1 (%)

Lindane

Frequency distribution of atrazine, simazine and lindane in GW bodies for the latest year (in parentheses) with uploaded data available on the EEA Waterbase database (adapted from Ref. 52).

EU 10 or 12 Austria (2004) Belgium (2004) Czech Republic (2004) Denmark (2004) France (2004) Germany (2004) Italy (2001) Lithuania (1998) Portugal (2001) Portugal (2002) Slovakia (2004) Slovenia (2002) Slovenia (2004) UK (2004)

Table 9.2.1

Pesticides in European Groundwaters 557

558

Chapter 9.2

data are reported is very low and there is no guarantee that such a low number of observations is truly representative. The picture may, therefore, be diﬀerent once the data at national scale are considered more closely (see Section 9.2.4), and the following disclaimer on the EEA web site seems to be entirely appropriate: ‘‘The data in Waterbase are sub-samples of national data assembled for the purpose of providing comparable indicators of state of waters on a Europewide scale and the data sets are not intended for assessing compliance with any European directive or any other legal instrument. Information on the national and sub-national scales should be sought from other sources.’’ One of such sources is the pesticide indicator fact sheets prepared by the EEA and presented in the next section.

9.2.3.2

The Indicator Fact Sheet of the European Environment Agency

For the preparation of the ‘‘Pesticides in GW 2004.05’’ indicator fact sheet (http:// themes.eea.europa.eu/Speciﬁc_media/water/indicators/WHS01a%2C2004.05/ WHS1a_PesticidesGroundwater_110504.pdf) based on 2000 data, only 11 of the 28 EEA member countries sent pesticide data from 1 to 39 GW bodies for 1 to 30 substances. The ﬁve substances monitored in the greatest number of countries are atrazine (10 countries), lindane (6), and simazine, hexachlorobenzene and diuron (5). The complete list of 13 substances included in the fact sheet was based on a combination of those listed for the Water Framework Directive (WFD) as well as substances thought to be the most important in terms of endangering GW. For these 13 substances, the proportion of GW wells with annual mean concentrations greater than 0.1 mg l1 is reported in Table 9.2.2. The number of countries providing data and the number of monitored sites varies greatly between substances. Information on pesticide use in the recharge area of each GW body and on monitoring and sampling strategy is a prerequisite, as yet unfulﬁlled, for more detailed interpretation of these data. The reported frequency of values 40.1 mg l1 cannot, therefore, be taken as an established ranking of substances according to their tendency to contaminate groundwaters. Nevertheless, the preponderance of triazines and one degradation product of At, shown in Table 9.2.2, are conﬁrmed by the more detailed data at national scale (see Section 9.2.4). The representativeness at EU scale of the high rate of values exceeding 0.1 mg l1 for bentazone may be questioned by the fact that only two countries provided data for this substance. The EEA indicator fact sheet suggests that the substances listed in the Priority Substances List is not entirely adequate for GW and do not cover the most important (polluting) substances. Even though signiﬁcant eﬀorts are being made by countries in investigating the situation of pesticide pollution at national level, a lot of additional eﬀort was still necessary in 2004 to provide a comparable overview at the European level. Such eﬀorts are being made, as illustrated by the improvement of data

559

Pesticides in European Groundwaters

Table 9.2.2

Descriptive statistics on data for the 13 substances selected by the EEA for its indicator fact sheet on pesticides in groundwater (data from the year 2000).

Substance

Number of countries providing data

Number of monitored sites

Annual means 40.1 mg l1 (%)

Bentazone Deethylatrazinea Atrazine Diuron Simazine Alachlore Clorfenvinphos Chlorpyriphos Endosulfan Hexachlorobenzene Isoproturon Lindane Triﬂuraline

2 3 10 5 5 2 1 1 3 5 4 6 2

790 209 1354 406 949 592 159 4 49 382 210 443 255

10 9 5 1 0.3 0 0 0 0 0 0 0 0

a

Metabolite of the active ingredient atrazine.

transfer from each member country to the EEA: in 2005, 13 out of 32 EEA member countries provided pesticide data, mainly for 2004. The ongoing update of the indicator fact sheet, currently in a drafting stage, indicates (A. Scheidleder, personal communication, June 2006) that the annual mean concentration of 0.1 mg l1 in at least one monitoring site is exceeded for deethylatrazine (exceedances in about 8% of sampling sites), hexachlorobenzene (B3%), atrazine (B3%), and four other substances with an exceedance rate o1% of sampling sites: bentazone, diuron, isoproturon and simazine. No exceedance has been reported in the 2004 data for lindane, alachlor, endosulfan, triﬂuralin, chlorfenvinphos and chlorpyriphos.

9.2.4

Status of Groundwater Contamination by Pesticides in Selected European Countries or Regions

The following section presents information on the monitoring of pesticides in GW (protocols and results) for ﬁve countries or regions selected on the sole basis of existing informal contacts between the author of this chapter and people involved in this topic in each of these countries or regions.

9.2.4.1

Status in Italy (Adapted from Refs. 60 and 61)

The Italian agency for environmental protection, APAT, coordinates the overall monitoring plans (technical protocols, data processing, statistical assessment

560

Chapter 9.2

and yearly report), while each region (through its environmental agency) applies the monitoring plan that lasts 3 years. The general framework for ranking the pesticides monitored is based on national uses, the list of priority chemicals, bans at the EU level, reported contamination in previous monitoring campaigns and the EPA-California approach which selects the pesticides to be monitored on the basis of water solubility (43 mg l1), soil sorption coeﬃcient normalised to organic carbon content (Koc o1900 cm3 g1), hydrolysis half-life (414 days), aerobic soil metabolism half-life (4610 days) and anaerobic soil metabolism half-life (49 days). For 2004,61 1992 sampling points in GW generated 3529 samples and 65 383 pesticide measurements. Pesticide residues were detected in 22.5% of the sampling points and 19.6% of the samples. Out of the 188 compounds analysed, 34 were detected in GW with a median of 1.7 substances per sample. Most detections (97%) were of herbicides and their metabolites. The 10 substances most frequently detected were deethylatrazine (10.5% of the 1882 samples), atrazine (10.2% of the 3228 samples), deethylterbuthylazine (9.6% of the 1892 samples), 2,6-dichlorobenzamide (8.2% of the 404 samples), terbuthylazine (8.0% of the 3128 samples), bentazone (7.1% of the 798 samples), hexazinone (6.2% of the 1063 samples), simazine (4.6% of the 3238 samples), oxadiazon (3.0% of the 1185 samples) and metolachlor (2.1% of the 2463 samples). This list diﬀers signiﬁcantly from that of the EEA fact sheet (see section 9.2.3.2). The exposure level for GW indicates that 7.4% of the samples exceeded the EU threshold value of either 0.1 mg l1 per substance or 0.5 mg l1 for the sum of all substances detected. The 2005 survey of 20 drinking water supplies conducted by the Ministry of Health indicates that, even though at least one pesticide or metabolite was found in every sample, the World Health Organisation Maximum Residue Level (WHO-MRL) threshold values were never reached. The last available update at national scale (2004 data) concludes that an overall national picture of the monitoring data is still lacking and identiﬁed gaps resulting from the fact that there is still no harmonised plan for sampling and measuring, and that laboratories’ detection limits and analytical methods, and the pesticides they analyse are very heterogeneous. At regional scale, the results of a two-year monitoring campaign of surﬁcial (mean depth of 10 m), vulnerable (coarse, sandy soil) GW wells (n ¼ 10) in two intensively cultivated areas in northern Italy58 represented a ‘‘worst-case’’ study that complemented the data at national scale. Five active ingredients of herbicides (2 triazines, 1 phenylurea, 2 chloroacetanilides) and 17 of their metabolites were analysed in 1999–2000. The 0.1 mg l1 concentration was exceeded in 59% of samples for at least one compound. Atrazine, terbuthylazine and metolachlor were the active ingredients most frequently detected. Even though it was banned in Italy in 1986, atrazine was present in 100% of the samples and 30% of the samples exceed the WHO-MRL. Deethylatrazine and deethylterbuthylazine were often characterised by greater concentrations than their parent active ingredient (the metabolites of metolachlor and

561

Pesticides in European Groundwaters

alachlor most frequently reported in GW monitoring studies53,54,56 were not analysed).

9.2.4.2

Status in Wallony (Belgium)

In 1989, a ﬁrst monitoring campaign was conducted on 25 water intakes.62 Up until 2000, the data on pesticides in GW came only from the water intakes used for drinking water supply, with the sampling frequency and the suite of substances analysed (atrazine, deethylatrazine, simazine, diuron and isoproturon being systematically analysed) varying according to the water intake. Atrazine and its metabolites represented most of the problems encountered (8% of the sites regularly exceed the 0.1 mg l1 value), other contaminations being reported for simazine, diuron, isoproturon, bromacil and chlortoluron.63 From 1994 to 2000, the number of substances analysed progressively increased to 81. For 23 compounds, maximum concentrations exceeding 0.1 mg l1 have been reported at least once, but only 10 substances (Figure 9.2.3) exhibit 1% or more of maximum values 40.1 mg l1: deethylatrazine (17.1%), atrazine (17.5%), diuron (2.8%), bromacil (2.7%), simazine (2.1%), chlortoluron Chlorinated and carbamates

Organophosphates Malathion

Parathion Ethyl Fenthion Fenitrothion Dichlorvos Diazinon

Lindane

Aldicarbe Carbofuran Prosulfocarbe Aldrine Dieldrine Heptachlore Heptachlore epoxyde

Azinphos Méthyl

Atrazine

Azinphos Ethyl

Deethylatrazine

Chlorfenvinphos

Deisopropylatrazine

Imidaclopride

Metribuzin

AMPA

Propazine

Glyphosate

Cyanazine

Sulcotrione

Terbutylazine

Clopyralide

Metamitron

Pentachlorophenol

Herbicides and insecticides of various classes

Trifluraline

0%----->-----% of sites- -->---40%

Diuron Isoproturon

Dinoseb

Chlortoluron

Dinoterb

Linuron

Dimethenamid

Monuron

Bentazone

Metobromuron

Triadiméphon

Bromacile

Pyridate Ethofumesate dichlorobenzamide Dichlobenyl Chloridazon Metolachlore Metazachlore Alachlore Chloroacetanilides

Figure 9.2.3

0.05 <= MAX <= 0.1 µg/l

Substituted ureas

Metoxuron

Dimethoate

MAX > 0.1 µg/l

Triazines

Simazine

Pendimethaline

Phytohormones

2,4,5-TP 2,4,5-T

2,4-DB 2,4-D

Lenacile

Triclopyr Fluroxypyr MCPA MCPB Mecoprop Dichlorprop

0.025 <= MAX <= 0.050 µg/l

MAX < 0.025 µg/l (non-detected)

Frequency of maximum concentrations (ranked in 4 classes) of pesticides in GW in Wallony between 2000 and 2004 (the number of sampling sites monitored for a given substance varies from 21 (prosulfocarbe) to 736 (atrazine), with a median of 445).

562

Chapter 9.2 35

18 30

16 Production interrupted or treated (106 m3, cumulated)

34 30

26

14

25 22

12

20

10 16 8

13

6 9

10

15

11 10

7

4 5 2

5 2

Number of waterworks affected (cumulative)

32

0

0 Before 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 1993 Year

Figure 9.2.4

Evolution with time in Wallony (1993 to 2004) of the volume of GW whose production has either been stopped or that has required treatment due to pesticide contamination (left ordinate), and of the number of water intakes involved (right ordinate).

(1.6%), isoproturon (2.0%), chloridazon (1.2%), aldicarbe (1.3%) and deisopropylatrazine (1.4%). Since 2000, the whole monitoring process has been harmonised. About 100 pesticides, some now banned, are determined. About 20% of the samples originated from the surveillance monitoring network, the rest being samples from waterworks before treatment. Some of the results for the period 2000–2004 are presented in Figure 9.2.4. For 24 compounds, maximum concentrations exceeding 0.1 mg l1 have been reported at least once, but only 9 substances exhibit 1% or more of maximum values 40.1 mg l1: deethylatrazine (10.7%), atrazine (7.6%; banned in 2004), 2,6-dichlorobenzamide (9.1%), diuron (2.7%; used as total herbicide), bromacil (2.6%; used as total herbicide), bentazone (2.4%; various agricultural uses), chlortoluron (1%; used on cereals), isoproturon (used on cereals) and deisopropylatrazine (1.1%). Herbicides, from agricultural or non-agricultural uses, are responsible for most of the problems encountered by the drinking water suppliers. BAM, determined only since 2003, appears to have an alarming impact on GW quality but is not considered relevant by the Belgian Ministry of Public Health for several preparations with an application rate of 5.4 kg active ingredient (dichlobenil) per hectare and per year (http://phytoweb.fgov.be/FR/Pers/ 20060303%20dichlobenil.htm). Trends are diﬃcult to identify notably because of the improvement in analytical methods. For some substances, the frequency of mean values 40.1 mg l1

Pesticides in European Groundwaters

563

decreased after 2001, down to 7% for deethylatrazine (10% before 2001), 5% for atrazine (7% before 2001), and 0.6% for diuron (B1% before 2001). The opposite trend applies to bentazone (1% of values o0.1 mg l1 after 2001 vs. o0.2% before 2001) and bromacil (2% vs. 1%). For simazine, isoproturon and chlortoluron, the percentage of mean values exceeding 0.1 mg l1 appears to be stable and low (0.3–0.6%). In the case of atrazine and its metabolites, the various actions taken to limit its use (ﬁrst, a ban as total weed-killer, then registration only in mixtures with other active ingredients) result in an improvement of the GW quality with less frequent peaks and the disappearance of atrazine in GW bodies with short residence times. The impact of agricultural or non-agricultural uses of herbicides as total weed-killer seems more important than the use of selective herbicides on agricultural soils. This is suggested by the results for BAM and bromacil, and was already suggested for atrazine in the 1990s. The monitoring network in Wallony was redesigned in 2004 in compliance with the EU WFD. The sampling points were chosen on the basis of hydrogeological parameters (direction and gradient of water ﬂow), provided enough information is available for the GW body. Otherwise, they were chosen on a purely spatial basis aiming at a representivity 480% (calculated with the GWstat codes, quo data GmbH, Siedlerweg 20, 01465 Dresden-Langebru¨ck) based on (i) the number of sampling points in the GW body, (ii) the average minimum distance between any point in the GW body and the nearest sampling point and (iii) the area of the GW body. The network density varies from 1 to 4 points per 100 km2, depending on the pressures, but there are at least 3 sampling points per GW body. Priority was given to agricultural and household water intakes over piezometers, to reduce the cost of sampling. The total number of sampling points will be around 360 for an area of 16 900 km2 and 33 GW bodies. Half of these points will come from the main WFD network; the others will be chosen, based on their representivity, from the waterworks network. Full surveillance monitoring campaigns will be conducted once every 3 years starting in 2006. During the year of monitoring (which will vary according to the GW body), the sampling frequency will be 1 for deep GW bodies, 2 for alluvial and ﬁssured GW-bodies, and 4 for karsts. This surveillance monitoring network will complement the data from the water companies (about 500 sites) whose contribution to the production of data on pesticides in GW will be lowered to about 50%, compared to 100% before 2000. The 8 pesticides systematically analysed from 2004 onwards are atrazine, deethylatrazine, simazine, diuron, isoproturon, bentazone, chlortoluron and bromacil. They were chosen because they are the ones most frequently detected by the waterworks who analyse the GW samples for a list of 100 substances established on the basis of runs of the SEPTWA model63 which is based on the GUS index,64 and of the estimation of quantities sold in Belgium. Other substances considered relevant for GW can be analysed locally, e.g. BAM, 2-methyl-4-chlorophenoxyacetate (MCPA), 2,4-dichlorophenoxyacetate (2,4-D), lindane, metribuzin and chloridazon. Glyphosate was determined in several campaigns of the surveillance monitoring network but withdrawn as being

564

Chapter 9.2

considered not relevant for GW. The quantiﬁcation limit required of the laboratories is 0.025 mg l1 for all substances, except lindane for which it is 0.010 mg l1. All analyses of the surveillance monitoring network will be performed by a single accredited (EN ISO/CEI 17025) laboratory, ISSep. The waterworks laboratories, as long as they are accredited, will continue to provide data. The impact of the contamination of GW on the production of drinking water from GW resources in Wallony is illustrated in Figure 9.2.4. Clearly, the total volume of water whose production has been stopped or that has required treatment, and the number of pumping wells involved, strongly increased at the end of the 1990s and had reached about 4.8% and 4%, respectively, in 2004.

9.2.4.3

Status in Sweden

Investigations of pesticides in GW are performed at national scale according to a well-deﬁned centralised procedure, and by various authorities at local and regional scales using less harmonised procedures. The results from the local and regional studies were included in a database (http://pesticid.slu.se; in Swedish) that contains, as of December 2005, approximately 4900 GW samples (71% from waterworks, 21% from private wells and 8% of unspeciﬁed origin) for the period 1990–2003, with a total of 305 diﬀerent pesticides. Pesticides were found in 39% of the samples and a total of 51 diﬀerent substances were identiﬁed: 95% of the samples with a detection contain less than four substances. The highest frequency of detection, 35%, was for BAM (detected in 1359 of 3884 samples), whose parent compound, dichlobenil, is no longer used in Sweden since December 1990. The other 9 substances most often detected were atrazine and deethylatrazine (DEA) (both in 12% of 3764 samples), hydroxyatrazine (11% of 179 samples), bentazone (11% of 3370 samples), dichlorprop (2% of 3287 samples), mecoprop (2% of 3349 samples), glyphosate (2% of 709 samples), AMPA (aminomethyl phosphonic acid, metabolite of glyphosate; 2% of 679 samples) and MCPA (1% of 3360 samples). Findings of glyphosate (and AMPA) were almost exclusively restricted to shallow, private wells. The ranking of substances as a function of the frequency (40.5%) of values 40.1 mg l1 yields the suite BAM (23%), atrazine (6%), DEA (6%), bentazone (6%), dichlorprop (2%), clopyralid (1.2%), mecoprop (1.1%), MCPA (1%), terbuthylazine (0.7%), hydroxyatrazine (0.6%), glyphosate (0.7%) and AMPA (0.6%). Several of the maximum concentrations were very high and likely due to point source pollution (S. Adielsson, personal communication). The current national pesticide monitoring programme, launched in 2002, is performed by the Division of Water Quality Management in collaboration with the Section of Organic Environmental Chemistry, both at the Swedish University of Agricultural Sciences (SLU). The monitoring programme is funded by the Swedish Environmental Protection Agency (Swedish EPA). The national programme includes four intensively studied, heavily farmed catchments (800–1600 ha), representing four diﬀerent regions with diﬀerent

Pesticides in European Groundwaters

565

climates, soils and cropping systems. The main focus of the monitoring programme is surface water but 16 shallow GW samples (1 site in a recharge area and 1 in a discharge area, 2 depths sampled at each site, 4 sampling campaigns per year) are collected yearly in each catchment. In 2005, 76 substances were included in the analytical procedure. The selection has been made on the basis of, for example, national sales statistics, use within the monitoring catchments (information from farmer interviews), inclusion in Annex 10 of the WFD and data from the registration procedure. A set of quality assurance procedures was used for water sampling (one person from the Geological Survey of Sweden collects all GW samples), farmer interviews, analyses (by a SWEDAC accredited laboratory: Section of Organic Environmental Chemistry, Department of Environmental Assessment, SLU) and data reporting. Approximately 20% of the total cost corresponds to sampling, collecting ﬁeld data and information on pesticide usage from the farmers, ca. 50% to pesticide analysis and 30% to quality assurance, data storage, project management and reporting. Trends and changes in patterns are identiﬁed and published in yearly reports to the Swedish EPA. The reports are available to the public on the web site of the Division of Water Quality Management, SLU (http:// vv.mv.slu.se/ShowPage.cfm?OrgenhetSida_ID¼6455; mainly in Swedish). The results from 2004 (http://www.ust.is/ness/pest/workshop2006.html) indicate that (i) no pesticides were detected in GW from two of the four catchments, (ii) 0.1 and 0.3 mg l1 of quinmerac were detected in GW from diﬀerent depths at one site in one of the catchments (together with traces of metazachlor)—quinmerac was also detected throughout 2005 at one depth at concentrations ranging from 0.05 to 0.15 mg l1 (quinmerac, only very recently used in some quantities in Sweden, was applied very near the site where it was detected in the GW; J. Kreuger, personal communication) and (iii) atrazine (traces, 0.04 mg l1), DEA (traces, 0.03 mg l1), lindane (traces, 0.02 mg l1), bentazone and metazachlor at trace levels were found in all samples from one site of one catchment. At the other site in the same catchment, glyphosate was detected once (0.18 mg l1) and metamitron also once (0.1 mg l1). In the case of glyphosate, meta data indicate that it was applied on a ﬁeld close to the GW sampling point after a long very dry period and just before a major rainfall event, a combination of circumstances creating a worst-case scenario. The results from the selected catchments conﬁrm some of the observations from the wider database (presence of triazines and metabolites, as well as bentazone), but also point to the need for assessment at local scale that may show contamination by substances of signiﬁcant local use (e.g. quinmerac).

9.2.4.4

Status in France

The French Institute for Environment (Ifen) has been publishing annual reports on the presence of pesticides in inland waters since 1998. The 2002 data (latest year with a published report,65 data for 2003–2004 are being processed) include up to 3693 sampling sites (case of atrazine) from networks set up by France’s six regional water authorities, complemented by data from

566

Chapter 9.2 Deethylatrazine * (3586, 7670) Atrazine (3693, 7846)

Deethylterbuthylazine * (607, 1331) Simazine (3651, 7731) Deisopropylatrazine * (3064, 6591) 2-hydroxy atrazine * (661, 1057) Diuron (2729, 5463) Terbuthylazine (3444, 7099) Oxadixyl (807, 1956) Amitrole (745, 1279) Chlortoluron (2605, 5147) Glyphosate (882, 1389) Isoproturon (2686, 5491) Imidaclopride (264, 813) Bentazone (840, 1969) 0

5

10

15

20

25

30

35

40

45

50

% of analyses with the substance quantified

Figure 9.2.5

Rate of quantiﬁcation (x-axis on the bar chart; % of all analyses performed), number of sites monitored and number of analyses performed (ﬁrst and second number in parentheses after the substance name on the y-axis, respectively) for the 15 substances (asterisk indicates metabolite) most frequently quantiﬁed in French groundwaters in 2002 (adapted from Ref. 65).

regional pesticide monitoring networks. In 2002, GWs were analysed for 373 pesticides (parent compounds plus metabolites, not all substances were determined at all sites), 123 of which (33%) were quantiﬁed. For the 15 most often quantiﬁed substances, the rate of quantiﬁcation (percentageof all analyses performed), the number of sites monitored and the number of analyses performed are given in Figure 9.2.5. It is noteworthy that the top six substances (deethylatrazine, atrazine, deethylterbuthylazine, simazine, deisopropylatrazine, 2-hydroxyatrazine) are chlorinated triazines or their metabolites, and that the sum of all triazines and metabolites (plus terbuthylazine) represents 85% of the quantiﬁcations. For the corresponding active ingredients, the median registration year in France, considering all preparations, goes back to the mid- to late 1970s (Table 9.2.3). The preponderance of triazines and their metabolites might result from at least two causes whose environmental implications are extremely diﬀerent. A ﬁrst explanation might be that modern pesticides are sorbed and degraded to a much greater extent than the substances registered in the late 1950s (the triazines, among others; see Table 9.2.3), and that their application rates are lower than the older pesticides. These intrinsic diﬀerences between modern and old pesticides should indeed lead to reduced leaching to the GW for the former

b

Field DT50 (days) 6067 n.a. n.a. 6067 n.a. n.a. 9067 11468 30069 1467 13568 4767 2067 1311 79–18671

Koc (l kg1) 10067 n.a. n.a. 13067 n.a. n.a. 48067 30668 5069 10067 17568 24 00067 3467 10170 132–25671

Preparations including the substance only, and the substance combined with one or more other active ingredients. Metabolite of an active ingredient. n.a., not available.

Atrazine Deethylatrazineb Deethylterbuthylazineb Simazine Deisopropylatrazineb Hydroxyatrazineb Diuron Terbuthylazine Oxadixyl Amitrole Chlortoluron Glyphosate Bentazone Isoproturon Imidacloprid

a

1978 (17) Metabolite of atrazine Metabolite of terbuthylazine 1975 (3) Metabolite of atrazine Metabolite of atrazine 1991 (30) 1981 (2) 1987, 1990 1993 (14) 1988 (13) 1995 (29) 1995 (8) 1988 (20) 1992 (3)

Substance

47 39 21 12 9 8 6 5 4 4 3 3 2 2 2

Quantiﬁcation rate (% of monitored sites)

Median year of registration, sorption coeﬃcients, half-life and quantiﬁcation rates for the 15 most frequently quantiﬁed pesticides in French GW bodies in 2002. Median registration year66 (number of registered preparationsa)

Table 9.2.3

Pesticides in European Groundwaters 567

568

Chapter 9.2

compared to the latter. A second explanation could be that the transfer time from the soil surface to the GW might be so long in many aquifers that pesticides registered only a decade or so ago have not yet reached the GW. The present situation is probably a result of the combination of these two hypotheses, and of other explanations not considered here. For 41% of the measurement points where pesticides are quantiﬁed, the levels of contamination make the resources either entirely unﬁt for drinking water supply or suitable only after speciﬁc treatment to remove pesticides. The fact that 41% of monitoring points are contaminated does not mean that 41% of all aquifer systems are contaminated (the networks cannot be considered to be 100% representative of all French GW). It is clear, however, that the quality of French GW is aﬀected by pesticide contamination. Indeed, the work conducted in preparation of the application of the WFD indicates that pesticide contamination is one of the major problems that will arise with respect to the objective of ‘‘good chemical status’’ for GW in France by 2015.72 Compared to the data from 2000,73 the preponderance of triazines over all other pesticides appears quite consistent. The same 6 substances are the most frequently quantiﬁed in 2000 and 2002. From 1997 to 2002, the quantiﬁcation rate varies between 40 and 54% for atrazine and between 48 and 54% for deethylatrazine. For dinoterbe (banned in 1997), dinoseb (banned in 1991), metolachlore and, to a lesser extent, isoproturon, the quantiﬁcation rates decreased slightly from 1998–2000 to 2002. For chloroacetanilides herbicides (metolachlor, alachlor and acetochlor), the metabolites are not determined even though their frequent presence in GW has been reported time and again,53,54 and the data on their presence can give insight to spatial and temporal trends of herbicides in GW.57

9.2.4.5

Status in the UK

The monitoring of pesticides in GW is supervised by the Environment Agency (http://www.environnement-agency.gov.uk). The Environment Agency has outlined a number of key objectives for GW monitoring, e.g. compliance to UK and European legislation, protection of the quality of GW, determination of trends in GW quality. The resources will be allocated in priority to GW bodies that are most vulnerable to pollution, support important water supplies or have signiﬁcant resource development potential, support river base ﬂows or important water-dependent habitats or are subject to measures under the WFD to return the body to good status.74 The agency has developed technical guidelines and standards for operational procedures covering all aspects of the monitoring process. The framework for GW quality monitoring will be implemented by the agency’s regions through a series of local implementation plans coordinated nationally. The need for a nationally consistent method of obtaining GW quality information arises from the need to make meaningful comparisons between all the GW bodies. The reliable, quality-assured data results of the GW quality monitoring system will

Pesticides in European Groundwaters

569

be included in reports, updated every six years, for each GW body and monitored hydrogeological unit. The guidance document ‘‘Groundwater Quality Monitoring—Network Design and Sampling Site Selection’’ gives general instructions for subdividing aquifers into appropriate GW quality monitoring units (GQMU), criteria for prioritising the assessment of the monitoring units, the information required to understand the hydrogeology and hydrochemistry of the system, the identiﬁcation of potential polluting pressures, network design, sampling point selection and record keeping.75 The known or anticipated spatial variability in GW quality, especially that derived from anthropogenic sources, is recognised as the most important factor in the selection of an appropriate number and distribution of monitoring points within each GQMU. In the ‘‘Guidance: Groundwater Quality Monitoring—Determinand Suite Selection, Sampling Frequency and Sample Collection/Handling’’ document, ﬁve suites of pesticide determinands to be monitored (organonitrogen pesticides, organophosphorus pesticides, acid herbicides, uron/urocarb pesticides, special organics, e.g. ﬂumethrin, cypermethrin) have been selected, taking into consideration potential pollution pressures (e.g. landuse criteria: arable, managed grassland, managed woodland, urban/industrial, sheep, amenity) on GW, and laboratory/analytical techniques and capability and practicalities of sample collection.76 The results of any other monitoring, where available, must be taken into account as well to determine the suite of molecules to be determined. The frequency of measurement is recommended on the basis of the hydrogeological properties (conﬁned, outcrop) and aquifer response (slow, fast). For pesticides, it is recommended that all GW quality network sites be monitored at least once during the ﬁrst 12 months of monitoring, and then at least annually for those suites identiﬁed as necessary (more frequent monitoring may be considered where the conceptual model and analysis of pressures indicate that this is necessary). Analysis is planned by determinand-speciﬁc analysis, supplemented by GC/MS scans. The detailed monitoring requirements (e.g. the location and density of monitoring points, the chemical parameters to be measured and the sampling frequency) will be identiﬁed on the basis of the hydrogeology of each GW body. Over the period 2000–2004, the number of sampling sites varied from year to year and for each substance. In 2000, for example, prochloraz was detected in only 1 sample, vs. 641 in 2004, and atrazine was detected in 441 samples, vs. 1040 in 2004. Even though the identiﬁcation of trends is of course hindered by such changes, a series of facts are supported by enough information to be brieﬂy presented. The 10 substances detected every year over the period 2000–2004 and with the highest percentage of values greater than the detection limit are atrazine (24% of the sites in 2000, to 46% in 2001), simazine (12% in 2000, to 39% in 2001), deethylatrazine (11 to 20%; determined only in 2003 and 2004), propazine (1% in 2000, to 19% in 2001), chlortoluron (2% in 2000, to 15% in 2001), trietazine (0.8% in 2000, to 10% in 2002), mecoprop (1% in 2003, to 9% in 2000), diuron (1.5% in 2004, to 10% in 2001), bentazone

570

Chapter 9.2

(1% in 2002, to 4% in 2000 and 2001) and linuron (0.2% in 2003 and 2004, to 2% in 2000). The threshold value of 0.1 mg l1 is exceeded for atrazine in 2% (2003 and 2004) to 5% (2000 to 2002) of the sites, for mecoprop in 0.5 (2002 and 2003) to 2.4% (2000), diuron in 0.2 (2000) to 1.9% (2001), simazine 0.5 (2000 and 2001) to 2.3% (2002), isoproturon in 0.1 (2004) to 1.5% (2001), bentazone in 0.3 (2002) to 1.1% (2000), chlortoluron in 0.2 (2004) to 1.3% (2001), deethylatrazine in 1.7 (2003) to 3.1% (2004), deisopropylatrazine in 1.6% (2004) and methabenzthiazuron in 1% (2002). Some years, values 40.1 mg l1, always less than 0.8%, are reported for carbendazime, chloridazon, clopyralid, cyanazine, pp 0 -DDT, dichlobenil, dicamba, ﬂuroxypyr, ﬂutriafol, MCPA, metazachlore, monolinuron, monuron, primicarb and triclopyr.

9.2.4.6

Synthesis of the Data at National and Regional Scales

Albeit limited in its spatial and temporal extension, and partly biased by the fact that some pesticides are only registered or used in certain EU countries (e.g. oxadiazon, used in rice ﬁelds, orchards and vineyards), the aforementioned review contains enough valid information for a synthetic table to be drawn (Table 9.2.4). Irrespective of the total masses of active ingredients applied and of the intrinsic vulnerability of the GW bodies monitored, the prevalence of triazines and their metabolites is evident (8 of the top 10 substances in the panel of countries and regions considered here are triazines or metabolites). The relevance of studying pesticide metabolites is often questioned on statutory or toxicological bases. The monitoring of these substances in GW deserves, however, to be pursued (and even extended to include substances with a known potential for GW contamination—the acid metabolites of chloroacetanilides, among others) because it can provide very useful insights into the overall response of aquifers to contamination pressures applied at the soil surface. Apart from triazines, the most problematic molecules diﬀer depending on the country, most likely as a result of diﬀerent major crops or agricultural practices. The herbicides are clearly the substances that pose the most acute problem for GW contamination; there is only one fungicide (oxadixyl) and no insecticides in the list of 26 substances in Table 9.2.4.

9.2.5

A Case Study: The Bre´villes Spring

The previous sections give an overall picture of GW contamination by pesticides at very small scale (laboratory studies on processes, Section 9.2.3) and very large scale (national and regional monitoring programmes, Section 9.2.4). The intermediate scale, that of functional units, is also of interest because it provides complementary information unattainable by the other scales. The case study of the Bre´villes spring illustrates the type of information that can be gained from studies at aquifer scale.

571

Pesticides in European Groundwaters

Table 9.2.4

Rank of pesticides (ﬁrst number) and metabolites (in italics) in groundwater according to the frequency (%, in parentheses) of their detection (Italy, Sweden, France, UK) or exceedance of 0.1 mg l1 (maximum values, Wallony) established from the top ten substances of the national monitoring networks in each country. Data for 2004 for Italy, 1990–2003 for Sweden, 2004 for the UK, 2000–2004 for Wallony, 2002 for France.

Substancea

Italy

Wallony

Sweden

France

UK

Mean rank

DEA Atrazine DET BAM Propazine DIA HyA Bromacil Simazine Trietazine Dichlorprop Bentazone Diuron Terbuthylazine MCPP Hexazinone Chlortoluron Isoproturon Glyphosate AMPA Oxadiazon Oxadixyl MCPA Linuron Metolachlor Amitrole

1 2 3 4

1 (10.7) 2 (7.6) n.a. 3 (9.1)

2 (12) 2 (12)

1 (47) 2 (39) 3 (21)

3 (11.5) 1 (29.7)

1.6 1.8 3 3.5 4 4.3 5 5 6 6 6 6.3 6.3 6.5 7 7 7 7 8 9 9 9 10 10 10 10

(10.7) (7.6) (9.6) (8.2)

8 (4.6) 6 (7.1)

7 (1.1) n.a. 5 (2.6) 10 (0.7) n.a. 6 (2.4) 4 (2.7)

n.r. 1 (35) 4 (11) n.a. n.r. 6 (2) 4 (11)

9 (3.0)

10 (2.1)

n.a. 9 (1) 7 (1.1) n.a. n.a. n.a. n.a.

4 (12) n.a. 7 (6) 8 (4.5)

5 (8.0) 7 (6.0)

5 (9) 6 (8)

7 (2) n.r. n.r. 8 (2) 9 (2) n.r. n.r. 10 (1) n.r. n.r.

4 (8.3) 4 (8.3) n.a. n.a. 2 (25.8) 6 (3.1) 9 (2.6) 8 (1.5) n.a. 7 (1.6)

n.a. 5 (1.3)

9 (4.2)

n.a. n.a. n.a. n.a.

10 (3.9)

10 (0.2-2) n.r. n.a.

a

DEA, deethylatrazine; DET, deethylterbuthylazine; BAM, 2,6-dichlorobenzamide; DIA, deisopropylatrazine; HyA, hydroxyatrazine; AMPA, aminomethyl phosphonic acid. n.a., not analysed in the routine monitoring programmes; n.r., never registered.

9.2.5.1

Brief Description of the System and the Methods Used (See Ref. 77 for More Details)

The Bre´villes spring is the main outlet of a 300 ha agricultural catchment located in Montreuil-sur-Epte (Paris Basin, 70 km northwest of Paris). A gauging station provides continuous discharge measurements. Weather data are provided by two rain gauges located in the catchment and by nearby national weather stations. The vadose zone is made up mostly of Lutetian and Bartonian limestone formations more than 30 m thick over 61% of the surface

572

Chapter 9.2

of the basin, and less than 10 m thick over 7% of the basin. The average thickness of the saturated Cuise sand is 13 m. The two main soil types are a luvisol, 60–90 cm deep, silty-loamy to silty-clayey soil, and a 30–50 cm deep, stony and silty clay loam soil. Seven piezometers (Pz), distributed around the recharge area, were drilled in the beginning of 2001. They are screened over the entire saturated zone, except for Pz4, which is screened only in the bottom 7 m of the 10 m thick saturated sand layer. The land use is for the most part agricultural, with 84% farm land, 10% forest, 5% prairie and 1% paved roads. There are no farmyards in the watershed and, according to the farmers, no tanks are emptied in the recharge area. The case study therefore involves only non-point source pollution. The two pesticides with the greatest cumulative mass applied during the period 1994–2004 were isoproturon (694 kg active ingredient, a.i.) and chlortoluron (352 kg a.i.). Atrazine, the only active ingredient almost systematically detected at the spring, ranks 15th, with 125 kg a.i.; it must be noted that no atrazine was applied over the period 2000–2005. Sampling campaigns have been conducted monthly in the seven piezometers since March 2001 and at the Bre´villes spring bi-monthly since October 1999 and monthly since August 2004. Sampling in the piezometers is done after pumping around three purge volumes to allow stabilisation of pH and conductivity. The samples are analysed for atrazine (At), deethylatrazine (DEA), deisopropylatrazine (DIA), chlortoluron (CTU) and isoproturon (IPU) and two of its metabolites (monodemethylisoproturon (MDIPU) and didemethylisoproturon (DDIPU)). Analyses are performed by liquid chromatography/mass spectrometry (LCQ DECA XP Plus, Thermo Finnigan). The quantiﬁcation limit for the seven products studied is 0.025 mg l1. The daily outﬂuxes of solutes are calculated by multiplying the daily concentration (Cd) measured at the spring on day d, or linearly interpolated between two measurement days (mg l1), with the mean daily stream ﬂow measured at the weir (l per day).

9.2.5.2

Main Results from the Piezometer Network (See Ref. 78 for More Details)

IPU is very rarely detected in any Pz except Pz5, which shows an almost systematic contamination always lower than 0.14 mg l1. The IPU metabolites MDIPU and DDIPU are never detected in any Pz. The rare and sporadic detections of CTU always occur, as for IPU, in the ﬁrst 4–10 weeks after application. A fraction of IPU and CTU can, therefore, reach the saturated zone soon after application. Short-duration peaks of CTU and IPU concentrations have been also observed in piezometers in a chalk aquifer.21 At and DEA concentrations vary greatly according to the Pz, from undetected in Pz4 (o0.025 mg l1 for At) to much higher values in Pz5 (0.97 mg l1 for At and 2.7 mg l1 for DEA). Figures 9.2.6 and 9.2.7 show box and whisker plots of At and DEA concentrations in each Pz and at the Bre´villes spring. In

573

Pesticides in European Groundwaters 1

Atrazine (µg/L)

0.8

0.6

0.4

0.2

0 Pz2

upgradient

Figure 9.2.6

Pz3

Pz6

Pz5

Pz8

Pz4

Pz7 spring

downgradient

Box and whisker plot of atrazine concentrations in the groundwater from the piezometers and the spring of Bre´villes catchment (45 values for each piezometer, 135 values for the spring; 54 months of monitoring).

addition to the marked spatial variability (even though the average distance between each Pz is only 300 m), a strong temporal variability is also evident for each Pz. No clear correlation exists between the location, up- or downgradient, and the level of contamination observed in every Pz. The latter results from a complex interaction between the pressures linked to the land use at plot scale, the very heterogeneous water ﬂow through the unsaturated zone, the existence of geological faults that may isolate parts of the saturated zone (as for Pz4) and the stratiﬁcation of the GW.

9.2.5.3

Main Results from the Monitoring of the Spring (See Refs. 77 and 78 for more Details)

CTU has been detected only on 1 Feb. 2001 (0.7 mg l1) and 11 April 2001 (2.0 mg l1). IPU has been also detected on 11 April 2001 (0.3 mg l1). The presence of these two substances only in 2001 and a few weeks after the treatment periods could be explained by the combination of (i) the very high rainfall of 52 mm over the 5 days following application, (ii) the location of the treated ﬁelds within the catchment, i.e. closer to the spring than those treated in other years, and (iii) the rapid transfer of part of the water in the unsaturated and saturated zone (indeed, tracer tests show a breakthrough of the tracers at

574

Chapter 9.2

Deethylatrazine (µg/L)

3

2

1

0 Pz Pz2

up upgradient

Figure 9.2.7

P Pz3

Pz6

P Pz5

Pz Pz8

P Pz4

Pz7

spring

do downgradient

Box and whisker plot of deethylatrazine concentrations in the groundwater from the piezometers and spring of the Bre´villes catchment (45 values for each piezometer, 135 values for the spring; 54 months of monitoring).

the spring 5 and 18 days after the injection 180 m upstream in the upper and lower parts of the saturated Cuise sand, respectively). Neither of the two IPU degradates studied (MDIPU and DDIPU) are detected, in agreement with observations reported for the Trois Fontaines watershed where MDIPU has very rarely been detected (concentrations lower than 0.05 mg l1) and DDIPU has never been detected.79 Contrary to the ureas, At is detected in all of the spring water samples at concentrations that vary between 0.07 and 0.43 mg l1 (Figure 9.2.8), in spite of the fact that the product is no longer applied in the catchment since April 1999. DEA is also systematically detected in the spring, with concentrations that vary between 0.14 and 1.16 mg l1. DIA has been detected only twice, on 19 Jan. 2001 (0.10 mg l1) and 1 Feb. 2001 (0.02 mg l1). A low detection frequency of DIA has been reported by other researchers.53,80 After the extraction method was changed (February 2005) in favour of liquid–solid extraction to increase DIA recovery, DIA was detected more frequently (4 times for the 9 samples analysed) but with concentrations close to the 0.025 mg l1 quantiﬁcation limit. The absence of any downward trend in At or DEA concentrations during the 5.5-year monitoring period is a result of a combination of factors. Firstly, the At and DEA adsorbed to the soil are still quantiﬁed even 5 years after the last At application at Bre´villes.81 This stock of At and DEA, possibly also present

575

Pesticides in European Groundwaters atrazine deisopropylatrazine

deethylatrazine isoproturon

chlortoluron

1.2

Concentration (µg L-1)

1.0 0.8 0.6 0.4 0.2

27/1

0/9 27/1 9 /0 27/4 0 /00 27/7 / 27/1 00 0/00 27/1 /0 27/4 1 /0 27/7 1 / 27/1 01 0/0 27/1 1 /0 27/4 2 /02 27/7 / 27/1 02 0/02 27/1 /0 27/4 3 /0 27/7 3 / 27/1 03 0/03 27/1 /0 27/4 4 /04 27/7 / 27/1 04 0/04 27/1 /0 27/4 5 /0 27/7 5 / 27/1 05 0/0 27/1 5 /0 27/4 6 /06

0.0

Sampling date

Figure 9.2.8

Time series of pesticide (and metabolite) concentrations in the groundwater of the Bre´villes spring.

deep in the vadose zone if the appropriate type of sorbents are present,30 can constitute a source for At and DEA leaching down to the saturated zone. Secondly, it might be explained by the weak inﬂuence of the deep vadose zone and the saturated zone on At and DEA degradation. Measurements done in the laboratory with samples from Bre´villes collected in the limestone under the ﬁrst metre of soil and in the sand reveal no degradation of either At or DEA during a 2-year incubation period.82 Thirdly, the transit time of the compounds through the thick vadose zone may be greater than the time that has elapsed since the application of At is halted. Tritium proﬁles within the vadose zone80 prove that the travel time of at least a fraction of the water in the vadose zone is greater than 40 years. Each of these hypotheses is equally likely and all of the mechanisms considered play their role in generating the pesticide signal measured at the spring, with the outﬂow of At and DEA not yet decreasing 5 years after applications were halted. The annual IPU and CTU ﬂuxes out of the basin between Dec. 2000 and Nov. 2001, the only year during which these two products were quantiﬁed, were 15 and 100 g, respectively, which is 0.02 and 0.23% of the quantities applied that year. These low annual output ﬂuxes observed for IPU are similar to those observed (0.05 and 0.019%) for two consecutive years in another hydrogeological system where IPU is the pesticide applied in greatest quantity.79 The end of At application in April 1999 has not resulted in a signiﬁcant drop in monthly At and DEA ﬂuxes with time. The annual ﬂuxes (December through to the following November) out of the basin are between 623 and 710 g per year, i.e. between 0.90% (calculation based on a mean annual input

576

Chapter 9.2

reﬂecting the period 1970 to 1990, with the maximum cropped acreage and approved application rates) and 2.82% (1994 to 1999: lower cropped acreage and application rate) of the annually applied quantities. When only At is taken into account, the average outﬂow is between 0.20 and 0.63% of the annually applied quantities for the same periods. The cumulative percentage of At and DEA ﬂowing annually out of the basin is within the range of literature values for the percentages of these products found at the bottom of the root zone. This conservation of mass between the root zone and the outlet reﬂects, for the Bre´villes catchment, an overall negligible eﬀect of both degradation and irreversible sorption of At and DEA in the deep vadose zone and saturated zone.

9.2.5.4

Conclusions from the Bre´villes Case Study

Despite the limited size of the Bre´villes catchment (3.0 km2), a marked spatial and temporal variability of pesticide concentrations in the GW has been evidenced but not fully explained. The fate of pesticides from the soil to the GW is still not fully understood and some mechanisms must be studied in greater detail. Regarding hydrodynamics, for instance, the spatial variability of transfer times in the vadose zone and the saturated zone requires more work. Regarding pesticides, more knowledge is needed on desorption kinetics and on the evaluation of stocks in the soil and in the vadose zone. Such developments are required not only to reach scientiﬁc goals such as mechanistic modelling of pesticides transport, but also to succeed in implementing the WFD83 and reaching its key objective: for all GW bodies to achieve good quantitative and chemical status at the latest in 2015.

9.2.6

Conclusions and Perspectives

In the ﬁeld of GW contamination by pesticide, the word ‘‘science’’ covers a wide range of disciplines (analytical chemistry, hydrogeology, microbiology, geochemistry, etc.), whereas the term ‘‘policy’’ can be focused, for the sake of simplicity, solely on the EU WFD83 and its parent legislation (included in the programme of measures in its Annex VI). The literature reviewed above enables us to draw some conclusions, identify gaps in knowledge and suggest perspectives, notably an attempt to move towards a convergence of scientiﬁc research and policy needs. The two main processes reviewed above in the unsaturated and saturated zones are sorption and biodegradation. Very few studies use protocols really similar to actual natural conditions. For biodegradation studies, the pesticide concentrations and the incubation duration are, compared to aquifers, overwhelmingly over- and underestimated, respectively. Even though it can be stated that sorption and biodegradation are generally less eﬃcient in the vadose and saturated zones than in the soil layers, marked exceptions are reported and

Pesticides in European Groundwaters

577

are related to speciﬁcities of the solids and biodegraders from the subsurface, an environment that obviously is not suﬃciently studied. Some gaps in scientiﬁc knowledge should be the focus of future work on sorption and biodegradation. It is clear that more information on the adsorption properties of a range of well-characterised solids from the vadose zone in contact with a range of pesticides (and their metabolites) is required. Another wide ﬁeld still unexplored is that of the stock of pesticides and metabolites present deeper than the soil layers. However small the ﬂuxes below the root zone, they might induce stocks deeper in the vadose zone. What are the actual masses involved? How do these stocks vary spatially? Are they really unaltered by biodegradation that proceeds very slowly but over very long periods? What is their availability for leaching by inﬁltrating water recharging the saturated zone? How long will it take for them to travel from the uppermost layers of the vadose zone to the saturated layers? Indeed, the presence of pesticides deep in the unsaturated zone has implications for the long-term protection of GW bodies. Periods of high GW levels could result in the remobilisation of these substances, including some persistent ones, long after their use has been stopped. Indeed, in the semi-conﬁned chalk aquifer of southeast England, the GW pollution by atrazine, simazine and diuron is positively correlated with periods of high GW levels.84 Furthermore, the inﬂuence that ﬂuctuations in the water table over time may have on variations of microbial transformation capacity is not clear.36 More biodegradation studies should address the factors controlling the presence of active pesticide-metabolising populations in subsurface environments14 and the potential for degradation in anoxic subsurface environments. There is also a strong need for a better understanding of microbial genetic regulation of biotransformation processes. Once the genes involved in pesticide biotransformation have been identiﬁed, the use of speciﬁc gene probes will enable ﬁeld-scale investigations whose results will be more generic.38 Preferential transport has been one of the hot topics in pesticide fate through the soil over the last decade. Proven important to the point of now being considered more of a rule than an exception, the inﬂuence of this process on the concentrations observed in GW has not yet been addressed. A very homogenous and thick vadose zone below the soil may completely smear this preferential transport, whilst a fractured, thin vadose zone may have no inﬂuence at all and enable direct transfer to the saturated zone. Once in the saturated zone, the concentration peaks resulting from preferential transport through the soil and vadose zone may again be smeared if good mixing occurs in the saturated zone, or remain unaltered, leading to strong spatial and temporal variability, in cases of stratiﬁcation of the hydrodynamic properties of the saturated zones. Dedicated ﬁeld studies, coupled with appropriate modelling, would enable these questions to be answered. It must be stressed, however, that large-scale ﬁeld experiments using pesticides are very diﬃcult to perform because of time and resources constraints, government regulations, and diﬃculties in gaining landowner consents for drilling and repeated access to agricultural plots.

578

Chapter 9.2

As shown in the Bre´villes case study, the spatial and temporal variability of GW contamination by pesticides within a 3 km2 catchment can be quite high and can be explained only in part by the spatial and temporal variability of the applications at the soil surface. Hydrodynamic factors such as fractures in the vadose zone, horizontal and vertical variability of the hydraulic conductivity in the saturated zone leading to stratiﬁcation of GW quality and geological faults isolating part of the saturated zone all play an important additional role. Such spatial variability must obviously be taken into account when designing monitoring networks aimed at gathering representative GW quality data on which the status of the GW bodies will be established. An additional issue not suﬃciently studied is that of stratiﬁcation, which might lead to diﬀerent GW quality at diﬀerent depths. In one piezometer of the Bre´villles sandy aquifer (see Section 9.2.5), the atrazine concentration meets the EU drinking water standard of 0.1 mg l1 at the depths of 13.5 and 18.5 mbs, but not at 9.3 mbs, whilst the DEA concentration below 0.1 mg l1 is observed only at a depth of 18.5 mbs. Similar marked vertical heterogeneity of GW contamination by pesticides, with 5-fold drop in concentrations over a 5 m depth increment, has also been reported elsewhere.41 This type of situation will obviously lead to serious diﬃculties in terms of interpretation. The water quality standard might, for instance, be met by the results of one sampling method but not by another in the same well due simply to the averaging characteristics of one of the two sampling methods. This stratiﬁcation problem will be all the more likely for GW bodies with heterogeneous hydrostratigraphy, and will need to be addressed when more detailed consideration is given to the harmonisation of monitoring protocols and approaches to GW pollution risk assessment at EU scale. All of these scientiﬁc issues are clearly linked to policy in the sense of the WFD. An improved scientiﬁc knowledge of the hydrodynamics of aquifers in all of their complexity (soil+unsaturated zone below the root zone+saturated zone) and on the fate of pesticides at aquifer scale is required to satisfy the request of policy-makers to see EU member states characterise appropriately the levels of GW contamination, explain cases of non-achievement of GW good chemical status and study evolutionary trends of pollutant concentrations. With respect to the monitoring data available from national and regional networks, it appears that enough information has not yet been collected at European scale to enable a clear picture to be drawn. Existing tools such as the EEA Waterbase data and fact sheets are useful in their concept but remain limited in their applicability for policy issues based on hard data. A thorough, highly informative interpretation of the available data provided by these tools remains out of reach for various reasons. The data are not yet suﬃciently linked to the characteristics of each GW body monitored (e.g. depth to the GW and GW residence time) nor to the pressures on the GW bodies (quantitative data on pesticides applied in the recharge area of each GW body). Furthermore, the issues of uncertainty in sampling and analysis, as well as possible bias resulting from the involvement of a large number of laboratories within a single member state, are not addressed.

Pesticides in European Groundwaters

579

While science can and must meet the needs of policy, it is also evident that policy can inﬂuence science, as illustrated by the scientiﬁc upgrading of monitoring strategies applied to GW contamination by pesticides that resulted from the implementation of the WFD. It is now up to scientists to make the best possible use of the new information coming out of these policy requirements, and to policy makers to take into account scientiﬁc knowledge, including the uncertainties associated with it, in their policy decisions.

Acknowledgements The following people are deeply acknowledged for providing information on the pesticide monitoring programmes in their countries: Stina Adielsson and Jenny Kreuger (Swedish University of Agricultural Sciences, Uppsala, Sweden); Ettore Capri (Istituto di Chimica Agraria ed Ambientale, Universita Cattolica del Sacro Cuore, Piacenza, Italy), Francis Delloye (Direction of Ground Waters, Ministry of the Environment, Namur, Belgium), Bob Harris and Rob Ward (Environment Agency, UK). The results from the Bre´villles test site presented here come from work funded by the European FP5 project PEGASE (contract EVK1-CT1999-00028 ﬁnanced by the EU through its 5th PCRDT), the European Union FP6 Integrated Project AquaTerra (project no. GOCE 505428) under the thematic priority ‘‘Sustainable development, global change and ecosystems’’, the BRGM research project TRANSPHYTO, the agreement 012095 with the Seine-Normandy water authority (l’Agence de l’Eau Seine Normandie) and the Centre Regional Council (Conseil Re´gional Centre) within the framework of the doctoral thesis of X. Morvan. The time required for writing this chapter was ﬁnanced by dedicated funds from the BRGM research directorate.

References 1. M. Flury, J. Environ. Qual., 1996, 25(1), 25. 2. J. Simunek, N. J. Jarvis, M. T. van Genuchten and A. Gardenas, J. Hydrol., 2003, 272, 14. 3. Y. Schnurer, P. Persson, M. Nilsson, A. Nordgren and R. Giesler, Environ. Sci. Technol., 2006, 40, 4145. 4. N. Baran, C. Mouvet, T. Dagnac and R. Jeannot, J. Environ. Qual., 2004, 33, 241. 5. G. L. Northcott and K. C. Jones, Environ. Pollut., 2000, 108, 19. 6. F. van den Berg, R. Kubiak, W. G. Benjey, M. S. Majewski, S. R. Yates, G. L. Reeves, J. H. Smelt and A. M. A. van der Linden, Water Air Soil Pollut., 1999, 115, 195. 7. M. S. Rodriguez-Cruz, J. E. Jones and G. D. Bending, Soil Biol. Biochem., 2006, 38(9), 2910. 8. J. P. Taylor, M. S. Mills and R. G. Burns, Eur. J. Soil Sci., 2004, 55, 671.

580

Chapter 9.2

9. C. T. Chiou, P. E. Porter and D. W. Schmedding, Environ. Sci. Technol., 1983, 17, 227. 10. S. K. Widmer and R. F. Spalding, J. Environ. Qual., 1995, 24, 445. 11. M. P. Charnay, T. Se´bastien, Y. Coquet and E. Barriuso, Pest. Manag. Sci., 2005, 61, 845. 12. A. C. Johnson, C. White and C. L. Bhardwaj, J. Contam. Hydrol., 2000, 44, 1. 13. L. Larsen, S. R. Sørensen and J. Aamand, Environ. Sci. Technol., 2000, 34(6), 18. 14. E. Blume, M. Bischoﬀ, T. B. Moorman and R. F. Turco, J. Agric. Food Chem., 2004, 52, 7382. 15. S. R. Sørensen, G. D. Bending, C. S. Jacobsen, A. Walker and J. Aamand, FEMS Microbiol. Ecol., 2003, 45, 1. 16. M. S. Mills, I. R. Hill, A. C. Newcombe, N. D. Simmons, P. C. Vaughan and A. A. Verity, Pest. Manag. Sci., 2001, 57, 351. 17. H. Gvirtzman and M. Magaritz, Water Resour. Res., 1986, 22, 635. 18. R. Nativ, E. Adar, O. Dahan and M. Geyh, Water Resour. Res., 1995, 31, 253. 19. S. Brouye`re, A. Dassargues and V. Hallet, J. Contam. Hydrol., 2004, 72, 135. 20. P. R. Jørgensen, L. D. McKay and J. P. Kistrup, Ground Water, 2004, 42(6–7), 841. 21. D. C. Gooddy, J. P. Bloomﬁeld, P. J. Chilton, A. C. Johnson and R. J. Williams, Ground Water, 2001, 39, 262. 22. G. Felding, Pestic. Sci., 1997, 50, 64. 23. A. Walker, I. J. Turner, J. E. Cullington and S. J. Welch, Soil Use Manage., 1999, 15, 9. 24. C. Moreau and C. Mouvet, J. Environ. Qual., 1997, 26, 416. 25. T. B. Moorman, K. Jayachandran and A. Reungsang, Soil Sci., 2001, 166, 921. 26. A. C. Johnson, C. D. Hughes and R. J. Williams, P. J. Chilton. J. Contam. Hydrol., 1998, 30, 281. 27. T. J. Besien, R. J. Williams and A. C. Johnson, J. Contam. Hydrol., 2000, 43, 91. 28. S. J. L. Bersillon, M. Detay, J. P. Feuardent and V. Vignier, C. R. Acad. Sci. Ser. IIA, 1994, 318, 1357. 29. S. W. Karickhoﬀ, J. Hydraul. Eng., 1984, 110, 707. 30. L. Clausen, F. Larsen and H.-J. Albrechtsen, Environ. Sci. Technol., 2004, 38, 4510. 31. R. M. Johnson and J. T. Sims, Soil Sci., 1993, 5, 339. 32. Y. Coquet, Vadose Zone J., 2003, 2, 40. 33. R. Bayless, Ground Water, 2001, 39, 169. 34. N. Tuxen, J. R. Lipthay, H. J. Albrechtsen, J. Aamand and P. L. Bjerg, Environ. Sci. Technol., 2002, 36, 2205. 35. C. M. Aelion, D. C. Dobbins and F. K. Pfaender, Environ. Toxicol. Chem., 1989, 8, 75.

Pesticides in European Groundwaters

581

36. L. Tora¨ng, N. Nyholm and H.-J. Albrechtsen, Environ. Sci. Technol., 2003, 37, 3095. 37. M. Leistra, J. H. Smelt, A. M. Matser, J. J. Bogte and L. J. T. van der Pas, Pest Manage. Sci., 2001, 57, 1023. 38. B. L. Hoyle and A. L. Arthur, Hydrogeol. J., 2000, 8, 89. 39. P. G. Pedersen, H.-J. Albrechtsen, PEGASE project, EU FP5 contract EVK1-CT1999-00028, 1st annual report, 2001, pp. 55–58. 40. L. J. T. van der Pas, M. Leistra and J. J. T. I. Boesten, Pestic. Sci., 1998, 53, 223. 41. R. F. Spalding, M. E. Exner, D. D. Snow, D. A. Cassada, M. E. Burbach and S. J. Monson, J. Environ. Qual., 2003, 32, 92. 42. J. Levy and G. Chesters, J. Contam. Hydrol., 1995, 20, 67. 43. S. K. Widmer, R. F. Spalding and J. Skopp, J. Environ. Qual., 1995, 24, 439. 44. N. Tuxen, P. L. Tu¨chsen, K. Ru¨gge, H.-J. Albrechtsen and P. L. Bjerg, Chemosphere, 2000, 41, 1485. 45. A. E. Springer and E. S. Bair, J. Environ. Qual., 1998, 27, 1200. 46. J. Agertved, K. Ru¨gge and J. F. Barker, Ground Water, 1992, 30, 500. 47. R. Allen-King, B. J. Butler and B. Reichert, J. Contam. Hydrol., 1995, 18, 161. 48. M. M. Broholm, K. Ru¨gge, N. Tuxen, A. L. Højberg, H. Mosbæk and P. L. Bjerg, Water Resour. Res., 2001, 37(12), 3163. 49. N. P. Arildskov, P. G. Pedersen and H. J. Albrechtsen, Ground Water, 2001, 39, 819. 50. R. P. Schwarzenbach and J. Westall, Environ. Sci. Technol., 1981, 15, 1360. 51. European Environment Agency, Groundwater quality and quantity in Europe, Environmental assessment report no. 3, 2000. 52. European Environment Agency, Waterbase—groundwater: water quality, 2006 (http://dataservice.eea.europe.eu/dataservice/viewdata/). 53. D. W. Kolpin, E. M. Thurman and S. M. Linhart, Arch. Environ. Con. Tox., 1998, 35, 385. 54. S. J. Kalkhoﬀ, D. W. Kolpin, E. M. Thurman, I. Ferrer and D. Barcelo, Environ. Sci. Technol., 1998, 32, 1738. 55. G. M. Clark, D. A. Goolsby and W. A. Battaglin, Environ. Sci. Technol., 1999, 33(7), 981. 56. J. E. Barbash, G. P. Thelin, D. W. Kolpin and R. J. Gilliom, J. Environ. Qual., 2001, 30, 831. 57. D. W. Kolpin, D. J. Schnoebelen and E. M. Thurman, Ground Water, 2004, 42(4), 601. 58. L. Guzzella, F. Pozzoni and G. Giuliano, Environ. Pollut., 2006, 142, 344. 59. M. Fielding, N. Mole, H. Horth, A. Gendebien and P. Van Dijk, Pesticides in Groundwater: Phase II, WRc Report CO 4395, 1998. 60. E. Funari and P. Bottoni, Relazione ﬁnale piani triennali di monitoraggio, Istituto Superiore della Sanita`, Rome, in press. 61. APAT (2005), Piano nazionale di controllo degli eﬀetti ambientali dei pesticide, Residui di pesticidi nelle acque, Rapporto annuale dati 2004, APAT, Rome, 2004.

582

Chapter 9.2

62. A. Demeyere, Rapport sur les recherches de pesticides dans les eaux de captage souterraines destine´es a` la production d’eau alimentaire, Belgian Ministry of Public Health and Environment (Ministe`re de la Sante´ Publique et de l’Environnement), 1990. 63. J. Rung, J. Swarcenstajn, B. Tricot, F. Delloye and D. Willock, Monitoring des pesticides en Re´gion Wallonne, Workshop Eau, Agronomie et consommateurs, Faculte´ Universitaire des Sciences Agronomiques, Gembloux (FUSAGx), Belgium, 4 May 2001. 64. D. Gustafson, Environ. Toxicol. Chem., 1989, 8, 339. 65. Ifen, L’e´tat des eaux souterraines en France, Aspects quantitatifs et qualitatifs, Etudes et travaux no. 43, 2004 (ISBN 2-911089-74-X). 66. Index phytosanitaire, Association de Coordination Technique Agricole (ACTA), 2000 (ISBN 2-85794-184-6). 67. R. D. Wauchope, T. M. Butler, A. T. G. Hornsby, P. M. W. AugustijnBeckers and J. P. Burt, Rev. Environ. Contam. Toxicol., 1992, 123, 1. 68. Commission of the European Communities, Water Pollution Research Report 27, 1991. 69. E. Dabe`ne, F. Marie´, C. Smith, Caracte´ristiques utiles pour l’e´valuation du comportement de quelques substances actives dans l’environnement, Ministe`re de l’Agriculture, de la Peˆche et de l’Alimentation, Paris, 1995. 70. A. Boivin, R. Cherrier and M. Schiavon, Chemosphere, 2005, 61(5), 668. 71. AGRITOX data base, 2006 (http://www.inra.fr/agritox). 72. M. Normand and A. Gravier, Report BRGM/RP-53924-FR, BRGM, Orle´ans, 2005. 73. Ifen, Pesticides in water, Annual report 2002, Etudes et travaux no. 36 [in French], 2002 (ISBN 2-911089-55-3). 74. Groundwater Quality: A Framework for Improved Monitoring, Environment Agency, Bristol, UK, 2002. 75. Guidance: Groundwater Quality Monitoring—Network Design and Sampling Site Selection, Agency Management System Document, Environment Agency, Bristol, UK, 2006. 76. Guidance: Groundwater Quality Monitoring—Determinand Suite Selection, Sampling Frequency and Sample Collection/Handling, Agency Management System Document, Environment Agency, Bristol, UK, 2006. 77. X. Morvan, C. Mouvet, N. Baran and A. Gutierrez, J. Contamin. Hydrol., 2006, 87(3–4), 176. 78. N. Baran, C. Mouvet and Ph. Ne´grel, Hydrodynamic and geochemical constraints on pesticide concentrations in the groundwater of an agricultural catchment (Bre´villes, France), Environ. Pollut., under revision. 79. N. Baran, C. Mouvet and M. Lepiller, Hydroge´ologie, 2000, 1, 73. 80. N. Baran, C. Mouvet and P. Negrel, Proceedings of the 3rd European Conference on Pesticides and Related Organic Micropollutants in the Environment, Halkidiki, Greece, 7–10 October 2004, pp. 109–114 (ISBN 960-91399-0-6). 81. T. Dagnac, S. Bristeau, R. Jeannot, C. Mouvet and N. Baran, J. Chromatogr. A, 2005, 1067, 225.

Pesticides in European Groundwaters

583

82. H.-J. Albrechtsen, L. Clausen and P. G. Pedersen, Abstracts of International Symposium on Non-agricultural Use of Pesticides. Environmental Issues and Alternatives, Copenhagen, 7–9 May 2003, pp. 43–44. 83. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities, 2000, p. 72. 84. D. J. Lapworth and D. C. Gooddy, Environ. Pollut., 2006, 144(3), 1031. 85. S. Roy and A. M. Fouillac, Trends Anal. Chem., 2004, 23(3), 185.

CHAPTER 9.3

Evaluation of the Quantitative Status of Groundwater–Surface Water Interaction at a National Scale HANS JØRGEN HENRIKSEN, LARS TROLDBORG, PER NYEGAARD, ANKER L. HØJBERG, TORBEN O. SONNENBORG AND JENS CHRISTIAN REFSGAARD Geological Survey of Denmark and Greenland, GEUS, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark

9.3.1

Introduction

With the European Union (EU) Water Framework Directive (WFD) the achievement of a good ecological status of surface waters and a good quantitative and qualitative status of groundwater has become obligatory. The ecological status of surface water is here deﬁned by biological, chemical, morphological and hydrological criteria.1 The WFD calls for combined management of surface water and groundwater, with proper assessment of the inﬂuence of groundwater quantity and quality on surface water ecology.2,3 Most rivers and other natural surface water systems (lakes, wetlands, etc.) derive their ﬂow from surface runoﬀ and groundwater discharges. The adverse impacts of groundwater abstraction on stream ﬂow depletion deﬁne a limit to the exploitable groundwater resources. In addition, it is important to understand better the relationships between groundwater quality in shallow aquifers and deep aquifers, and possible negative eﬀects of excessive groundwater abstraction on future groundwater quality in aquifers that are the backbone for drinking water and the aquatic environment.4–11 The trend in recent years has been to base water management decisions to a larger extent on modelling studies, and to use more sophisticated models. Models have become an essential tool for analyzing complexly managed 584

Evaluation of the Quantitative Status of Groundwater–Surface

585

basins.12 In Europe this trend is likely to be reinforced by the WFD due to its demand for integrating groundwater, surface water, ecological and economic aspects of water management at the river basin scale and due to the explicit requirement to study impacts of alternative measures (human interventions) intended to improve the ecological status in river basins.13 Large-scale surface water hydrological modelling with dynamic simulation of streamﬂow hydrographs for catchments of more than 50 000 km2 has been carried out in several studies (see Ref. 14 for a review). However, these models did not include simulation of the groundwater system except for the routing of baseﬂow in large linear reservoirs. Groundwater modelling of hydraulic heads and ﬂow patterns for areas larger than 50 000 km2 has been carried out by several researchers.15–17 However, these models were all based on steady-state approaches and did not explicitly include surface water processes. Several examples of dynamic and integrated groundwater/surface water models exist for smaller areas. To our knowledge only a limited number of scientiﬁcally reported examples cover large-scale modelling of dynamic groundwater– surface modelling with areas of several thousand square kilometres. Examples of such models are the Danubian Lowland in Slovakia18 and catchments in Kansas.7,19,20 The adverse impacts of groundwater abstraction on streamﬂow depletion and wetlands deﬁne a limit to the exploitable groundwater resource. The sustainable yield of an aquifer must be considerably less than the recharge, if adequate amounts of water are to be available to sustain both the quantity and quality of streams, springs, wetlands and groundwater-dependent ecosystems.21 Groundwater discharges to streams constitute the major source of streamﬂow during dry periods, thus the minimum ﬂow can be violated if baseﬂows are reduced due to groundwater abstraction. The abstraction will inﬂuence the ﬂow regime in a way that depends on the characteristics of the aquifer system, the depth and distance of the abstraction from the river and the seasonal/temporal variation in pumping and river runoﬀ. The ﬂow regime is vital for the temporal variability of water depth and velocity, river morphology, sediment transport and bed sediments, and consequently for the water quality and ecosystem that develops.22 Freshwater ecosystems are an integral part of the environment and human culture. Even though the WFD clearly states that the utilisation of water, including groundwater, must not negatively impact surface water ecology, an evaluation of how much the regime can be changed for a given catchment or river reach is rather diﬃcult. Traditionally, some of the methods have been used to deﬁne a minimum ﬂow, below which no direct inﬂuences should take place. However, the current trend is towards methods that consider the ﬂow regime, with some degree of ﬂow variability, to maintain the natural morphology and ecosystem, instead of methods that set one ‘‘minimum ﬂow’’, e.g. historic ﬂow, or ‘‘rule of thumb’’ methods.23 The most advanced methods to assess the link between physical/chemical variables and ecosystem state are habitat models.22,24–28 The literature and the scientiﬁc research that focus on the impacts of groundwater abstraction on groundwater quality in shallow and deeper aquifers are rather limited. Todd29 deﬁned the safe yield of a groundwater basin in a

586

Chapter 9.3

broad formulation as ‘‘the amount of water which can be withdrawn from it annually without producing an undesired result.’’ A review of diﬀerent methods for determining sustainable yield in groundwater systems can be found in Ref. 30. Sustainable yield is best determined in the context of the basin or groundwater entity water balance. Estimation of basin or groundwater entity pre-development recharge is a relevant activity in the determination of sustainable yield. However, variation in climate and land-use changes will produce uncertainty. Recently there has been a debate in the literature about the concepts of sustainable development of groundwater resources (sustainability) vs. sustainable pumping.31–33 The debate was caused by a perceived communication gap between two groups in the hydrogeological community: those concerned with sustainable pumping and those concerned with sustainability. The former group has held that sustainable pumping rates can be determined without measuring recharge (to the aquifer). The latter group holds that recharge measurements are necessary because sustainability is broader than just sustainable pumping. Due to the eﬀects recharge is likely to have on water quality, ecology and socioeconomic factors, it remains important in the assessment of sustainability. For example, recharge could aﬀect the quality of the water in the aquifer and its nutrient content, thus also impacting associated ecological communities.31 Sustainability is a goal for the long-term welfare of both humans and the environment. As a result, eﬀort should be made to estimate recharge rates as accurately as possible, when an assessment of sustainability is the objective. Especially in conﬁned aquifers groundwater recharge may be increased signiﬁcantly as an eﬀect of abstraction and decrease of groundwater heads. As it is diﬃcult to provide general numbers on the acceptable change in ﬂow regime due to groundwater abstraction, it is also very diﬃcult to give general numbers on how much recharge can be increased before the groundwater quality is signiﬁcantly aﬀected. Degradation of groundwater quality can become a severe problem due to many diﬀerent and complex processes. Returned irrigation water or downward leakage from saline aquifers can lead to poor water quality over a period of time. Saltwater intrusion may also limit abstraction. In some areas small changes in water level and release of substances when redox conditions are altered may make a proper assessment delicate and extremely diﬃcult. Prediction of these eﬀects could be achieved using numerical models where necessary, but for basin sustainable yield assessment experience values about the groundwater system resilience would be appropriate. Comprehensive water quality groundwater monitoring datasets could be used for assessment of, for example, a sustainable fraction of the predevelopment (virgin) groundwater recharge rate. The objectives of this chapter are to: describe an approach for surface water/groundwater modelling that enables assessment of the eﬀects of groundwater abstraction on surface water quantity at river basin/national scale;

Evaluation of the Quantitative Status of Groundwater–Surface

587

describe possible criteria for assessing the eﬀect of the quantitative groundwater status on surface water ecology and for groundwater quality also; and present results of assessment of sustainable groundwater abstraction at regional and national scale.

9.3.2

The National Water Resource Model (DK Model)

9.3.2.1

Conceptual Model

The DK model consists of 11 regional sub-models with a delineation based on natural hydrological boundaries.34–36 The model is composed of a relatively simple root zone component for estimating the net precipitation, a comprehensive three-dimensional groundwater component for estimating recharge to and hydraulic heads in diﬀerent geological layers (see Figure 9.3.1) and a river component for streamﬂow routing and calculating stream–aquifer interaction. The model was constructed on the basis of the MIKE SHE code and by utilising comprehensive national databases on geology, soil, topography, river systems, climate and hydrology. Four regional sub-models covering the islands of Fyn and Sjælland were applied to a heterogeneous glaciomorphological topography, with a near surface geology consisting of Quaternary deposits overlying Tertiary limestone and marls. The Quaternary deposits consist of terrestrial glacial sediments with a thickness ranging from a few metres to 150 m whereas the pre-Quaternary deposits underneath consist in general of Danien limestone in the eastern and northern parts of Sjælland and Paleocene marl and clay in the western part of Sjælland and Fyn. Much emphasis was put on a proper description of the geological model in three dimensions.34 Jylland is split up into six regional sub-models. The eastern part of Jylland is relatively hilly, with maximum elevations of approximately 100 m above sea level. The western part is gently sloping to the west. A topographical water divide is located at the boundary between the two areas, referred to as the ‘‘Jutland Ridge.’’ In a large part of Jylland, Miocene sediments are found directly below the Quaternary deposits. The last sub-model covers the Baltic Sea island of Bornholm. The topography is hilly and most aquifers are found in granite or sandstone and PreQuaternary sand. The Quaternary sequence is relatively limited on Bornholm. In terms of both geology and discretisation the Bornholm sub-model diﬀers from the rest of the sub-models.

9.3.2.2

Processes and Data

In order to achieve a proper simulation of groundwater ﬂow processes at large scale, it was decided to include the following hydrological processes in the model.34

588

Chapter 9.3 m a.s.lA

S–N cross section

B

61.00 32.80 4.60 -23.60 -51.80 -80.00

0

10 20 km

6235 N

6225

B

6215 6205 6195

Model layer

6185

Fractured clayey till Clayey till Sand Limestone and chalk Sea

6175 6165 6155 6145 6135 6125 6115 6105 6095 6085 6075

Legend

6065

> 60 45 – 60 30 – 45 15 – 30 0 – 15 Sea level

6055 6045 625 635 645 655 665 675 685 695 705 715 725 735 745

Figure 9.3.1

The groundwater model for Sjælland comprises 10 Quaternary layers of alternating sand and clayey till above the chalk and limestone aquifer: south–north cross-section (above) and topographical variation with location of cross-section (below).

Snow accumulation and melt in order to be able to take into account the delay in net precipitation due to snow. Overland ﬂow. Unsaturated zone processes including evapotranspiration. The main requirement to this description is that the net precipitation (precipitation

Evaluation of the Quantitative Status of Groundwater–Surface

589

minus evapotranspiration) is assessed correctly on a seasonal and annual basis. Groundwater ﬂow processes including hydraulic heads, ﬂow between layers and exchange ﬂow between aquifers and rivers. Because a signiﬁcant part of the country is drained with artiﬁcial tile drains, a drainage component is included for the upper phreatic aquifer. River ﬂows and water levels. The extension of rivers was determined from digitised river points. Typical cross-sections were applied based on measured ﬂow magnitudes and catchment areas. Some smaller headwater tributaries could not be incorporated in the river network. Instead these areas are drained by the drainage component of the model. The national water resource model (DK model) uses daily precipitation, temperature and reference evapotranspiration as input. The geology has been interpreted for 10 to 50 geological layers based upon several thousands of borehole logs. Groundwater ﬂow in the upper soil layers, drainage systems and rivers is described in a fairly detailed manner.

9.3.2.3

Model Code

To simulate the groundwater ﬂow system with emphasis on groundwater– surface water interaction, the MIKE SHE code37–40 was chosen. MIKE SHE is a deterministic, fully distributed and integrated hydrological modelling system, which can describe the most important ﬂow processes in the land phase of the hydrological cycle. In order to save computational time and reduce the data requirements, it was decided to disregard the complex unsaturated zone component in MIKE SHE that is based on Richards’ equation. Instead a simple root zone module was developed for calculation of daily snowmelt and net precipitation.

9.3.2.4

Model Parameterisation and Calibration

A grid size of 1 km was chosen as a reasonable compromise. The use of 1 km grids is a rough approximation with simpliﬁcation of a number of conditions important to the groundwater recharge and streamﬂow generation, but can be considered reasonable in relation to the modelling purposes.34 The guiding principle in the parameterisation was to construct a model with as few free parameters as possible.41 Thus, uniform parameter values throughout the model area were used for geological layers composed of clayey till and sand as well as for most overland parameters. Initial best estimates of hydraulic parameter values and expected ranges have been assessed based on data from ﬁeld work, previous modelling results and the literature.34,35 For the chalk aquifer underlying the Quaternary deposits measured values of transmissivity (extracted from GEUS’s national database) were a priori used to interpolate the spatial distribution of the hydraulic conductivity.

590

Chapter 9.3

Recorded groundwater heads from GEUS’s well log archive from 1970 to 1996 were used for calibration. Thus, head data from more than 20 000 wells with screens distributed over diﬀerent geological layers were used as the measure of observed steady-state hydraulic heads. Daily streamﬂow data from more than 50 river gauging stations for the period 1990–2000 were used for calibration and validation purposes. A critique often expressed against distributed models concerns the many parameter values which can be modiﬁed during the calibration process. Hence, according to Beven,42 the problem of over-parameterisation is a key characteristic of the distributed model type. In our case we have designed the basic conceptual model with the aim of making maximum use of any structural information, especially geological data, and other existing data sources. For the parameters that had to be assessed through calibration, the general policy was to maintain global parameter values, wherever possible. This implies for instance that all geological layers consisting of sand across one of the 11 model areas are given the same parameter value. Thus in spite of a potential number of diﬀerent parameter values of the order of 106 for the combined model, all parameter values, except about 10 ‘‘free’’ parameters, were assessed directly from ﬁeld data. Given the large amount of calibration and validation data, this number of adjustable parameters is small and comparable to a simple lumped conceptual-type rainfall runoﬀ model. We estimated 10 parameter values on the basis of more than 2000 hydraulic heads and daily discharge data from 50 river gauging stations. The parameter values showed robust results when the model was subject to powerful validation tests such as using model parameters assessed in one sub-model to data from other sub-models. This indicates that the present model is not over-parameterised. The performance criteria were selected in order to reﬂect the objectives of the modelling, namely to be able to simulate aquifer hydraulic heads and river ﬂows at multiple sites. The four selected criteria, RMS for head simulation, R2 for runoﬀ simulation, Fbal for water balance and F-low for low ﬂow simulation, are described in detail elsewhere.34,43 RMS values are basically calculated for each of the nine geological layers, while R2 and Fbal are calculated for each of the gauging stations used for calibration. This results in a confusingly large number of performance criteria. Therefore, global values based on averages over the nine layers (weighted by the number of observations per layer) and the 4–20 stations in each of the 11 model areas (simple arithmetic mean) were to a large extent used in the calibration and validation process. While a global value has the advantage of providing a very easy overview, considerable information is lost as compared to the distributed information contained in the individual values. Therefore, two diﬀerent ways of aggregating the performance while maintaining the distributed information from the diﬀerent layers/stations have been attempted. Inverse modelling for steady-state conditions was carried out against groundwater heads and mean river discharges as calibration targets.34,35 This was done by linking MIKE SHE and the universal inversion code UCODE.44 Selected hydraulic conductivity parameters based on sensitivity analysis from each of the 11 model areas were optimised through inverse modelling. Subsequently, the

Evaluation of the Quantitative Status of Groundwater–Surface

591

parameter values for speciﬁc yield, surface detention storage and drainage time constant were assessed through trial-and-error using dynamic simulation and discharge as calibration targets. In selected areas (e.g. Sjælland and Jylland) the optimised parameter values were then transferred and applied also to other areas.

9.3.2.5

Model Validation

The following model validation scheme was adopted:45 an ordinary split-sample test using one subset of the period 1990–2000 for calibration and the other part for validation (diﬀerent selections in diﬀerent areas dependent on the model construction process which lasted from 1996 to 2003 for the 10 models for the islands and Jylland; for Bornholm the model calibration and validation has recently been ﬁnalised); and a proxy-basin test with the same parameters as obtained from calibration of a neighbouring model. Subsequent validation tests of, for example, simulated ﬂuctuations in groundwater head compared to piezometric head observations have further documented the reliability of the DK model for Sjælland.46

9.3.2.6

A Few Examples of Model Results

The ﬁnal results of the DK model for Sjælland, Fyn and Jylland43 showed that it was possible to construct a combined groundwater/surface water model with a horizontal grid size of 1 1 km2 that yielded reliable results with respect to simulation of hydraulic heads and discharges. The ﬁnal DK model honoured the pre-established performance criteria for river ﬂows and groundwater levels in the validation tests and is therefore ready for operational use, e.g. for assessing groundwater recharge to diﬀerent geological layers and assessing impacts of alternative groundwater development scenarios on river ﬂow on a regional scale. Figure 9.3.2 shows the simulated 1 km 1 km net precipitation. Figure 9.3.3 shows the water balance as simulated for Sjælland. With the DK model well validated, it could be used for scenario calculation for assessing sustainable groundwater abstraction. This was done on the basis of the following. The DK model which was used to simulate the groundwater–surface water situation both for pre-development conditions and various scenarios of groundwater abstraction. The model was also able to simulate the eﬀects of diﬀerent climatic conditions using data from the period 1990–2000. A set of criteria or indicators established to quantify the maximum allowable impacts of groundwater abstraction in terms of change in groundwater table, recharge to deep aquifers and baseﬂow. The actual groundwater abstraction in 2000.

592

Chapter 9.3

Net precipitation mm/year 1–50 50–100 100–150 150–200 200–250 250–300 300–350 350–400 400–500 500–600 600–900

0

Figure 9.3.2

50 km

Simulated net precipitation (mm yr1) for the national water resource model for 1 km 1 km grid. The ﬁgure shows the delineation of the 11 modelling areas: six in Jylland and ﬁve for the islands Fyn, Sjælland and Bornholm.

The most diﬃcult and controversial of these was the establishment of the set of criteria characterising sustainability. This is described in the following section.

9.3.3

Criteria for Sustainable Groundwater Abstraction

Denmark has a rather unique situation. The water supply for drinking water, industrial usages and ﬁeld irrigation is almost exclusively based on groundwater. The question we have asked ourselves is if we have suﬃcient amounts of this resource and if the present utilisation of the groundwater resource is sustainable. For the country as a whole, water abstraction may be less than the exploitable water resource, which is the amount of water we can pump up, while at the same time maintaining a ‘‘good’’ status of ecosystems and ensuring that

593

Evaluation of the Quantitative Status of Groundwater–Surface Net precipitation 207 Abstraction 25 0 1 146

Shallow aquifers 62

2

Ground water discharge 13

0 10

7 Upper regional aquifer 1 36

6

13

18 Lower regional aquifer

Figure 9.3.3

Model simulated water balance for Sjælland.

groundwater quality does not deteriorate due to the pumping. A concern here is the uneven distribution of the resource across the country, with the highest net precipitation rates in the western part of the country where the population density is relatively low, and the much smaller net precipitation in the eastern parts of the country where the major cities (Copenhagen, A˚rhus, A˚lborg and Odense) are located. This regional pattern is to some extent ‘‘compensated’’ by large irrigation requirement in the western part of the country where soils are more sandy and agriculture is more intense. The signiﬁcant utilisation of groundwater, especially in the capital area, has resulted in decreased baseﬂows and in many situations dried up reaches and wetlands. Furthermore, a decline in groundwater levels of 5–10 m compared to the pre-development (virgin) situation has been seen in some areas. At the same time pesticides, nitrate and other contaminants are transported downwards though the soil layers degrading groundwater quality of shallow groundwater aquifers. Nitrates and pesticides primarily inﬁltrate the soil from agricultural land, whereas organic or metallic pollutants come from contaminated land and are released from the soil when groundwater levels decline. Within the last 5 years, pesticides have been found in 26% of waterworks wells, and in 6% of wells the limit values for drinking water have been exceeded. In around 25% of drinking

594

Chapter 9.3

water wells nitrate is found and the limit value is exceeded in 1%. A new study of the population’s attitude towards treatment of polluted groundwater vs. groundwater protection has shown that Danes are willing to pay extra to protect groundwater and that they prefer to protect groundwater rather than clean it.47 As groundwater movement is very slow this situation is critical when viewed within the WFD framework prescribing future goals for the environment and the groundwater bodies to be achieved within ﬁxed time schedules. Therefore, an assessment of the long-term sustainable exploitable groundwater that can be abstracted is important for water management and policy-making in Denmark. The factors that have to be taken into account when assessing how much groundwater can be abstracted in a sustainable manner are illustrated in Figure 9.3.4. The limits to groundwater abstraction for most Danish hydrogeological settings are deﬁned by excessive streamﬂow depletion (reduced baseﬂow) caused by pumping from groundwater abstraction wells. In some of these areas the balance between abstraction and recharge may be more or less critical, but never providing the limit for availability. The qualitative imbalance regarding abstraction and recharge represents a concern for an increased release of toxic solutes such as nickel from aquifer sediments caused by lowering the groundwater table and associated transformation from anaerobic to aerobic conditions (decreased groundwater table). For conﬁned aquifers increased groundwater abstraction will lead to an increase in groundwater recharge, implying that pollutants, such as nitrate and pesticides, located in the upper soil layers48 move faster towards the deeper aquifers where most of the groundwater is abstracted (increased deep recharge). Finally, groundwater resources are known to be vulnerable to variability or change in climate input49,50 (climate variability). These concerns on sustainability Circulation time

10 dage

Evaporation

Climate variability

Decreased groundwater table Unsaturated zone

20 days

Sea 3100 years

Groundwater table Sand

Saturated zone Clayey till

Increased deep recharge

5 years 500 years 10 000years

Figure 9.3.4

10 years

Fresh groundwater

Salt groundwater

Reduced baseﬂow, increased deep recharge, decreased groundwater table and climate variability are the factors limiting the sustainable yield when abstracting water from an aquifer system.43

Evaluation of the Quantitative Status of Groundwater–Surface

Table 9.3.1 Indicator no. 1 2 3 4

595

The four indicators used to characterise sustainable groundwater abstraction. Indicator

Factor considered

Max. abstraction ¼ 35% of natural recharge Max. increase of recharge ¼ 30% of natural recharge Max. reduction of annual streamﬂow ¼ 10% Max. reduction of low ﬂows ¼ 5, 10, 15, 25 or 50% depending on ecological objective of river reach

Decreased groundwater table (groundwater quality) Increased deep recharge (groundwater quality) Streamﬂow depletion Reduced baseﬂow

were translated by Henriksen and Sonnenborg43 into the four indicators shown in Table 9.3.1. The indicators chosen to reﬂect concerns over groundwater quality were ﬂow indicators, i.e. indirect measures as compared to more sophisticated indicators based on groundwater level and/or solute transport. The 35% and 30% limits were derived as an empirical rule of thumb based on an analysis of the actual groundwater quality and abstraction rates for Sjælland, where it had been observed that areas with intense groundwater abstraction and signiﬁcant lowering of the groundwater table often have extended problems with inorganic trace elements. The present modelling approach based on 1 1 km2 grids and model calibration and validation on sub-catchment scales of 300–2000 km2 does not allow a direct simulation of groundwater level drawdown near abstraction wells or detailed solute transport modelling. It should be remembered that the selected 35% and 30% limits were derived for the speciﬁc sub-catchment scale and settings for Danish aquifer systems viewed ‘‘as an entity’’ and not as a measure for evaluating the sustainable pumping rate from a single speciﬁc well ﬁeld. This also means that if we zoom in on a single well ﬁeld and a scale of say 30–200 km2 then the limit values would probably increase to say 70% and 60% or even higher in some areas, while it would decrease in other areas. This means that the sustainability is bounded to the scale used for the modelling purposes and the assessments. Overall the exploitable groundwater resources were assessed for aquifers at 30 to 50 m depths from where the majority of groundwater abstractions today takes place. In the translation of the abstraction–runoﬀ balancing principle it has been assessed that a 10% reduction of the average ﬂow in river systems is acceptable (indicator 3). In the literature there are ‘‘rule of thumb’’ values for the balance between surface water abstraction/eﬄuent return by reservoir compensation and surface runoﬀ. The indicator on depletion of low ﬂows (indicator 4) is based on guidelines from the Danish EPA from 1979 prescribing a maximum reduction of low ﬂows depending on the ecological objectives of the river reach, which is categorised as

596

Chapter 9.3

A (waters for scientiﬁc reference areas), B1 (salmonidae spawning and nursery waters), B2 (salmonidae waters; nursery and living areas for trout), B3 (cyprinid waters) and C-F (watercourses solely used for drainage purposes, waters where authorised waste water discharges cause the quality to be worse, watercourses where the eﬀects of water abstraction render it impossible to maintain ﬁsh water objective or watercourses markedly aﬀected by ochre discharge). According to these old guidelines, baseﬂow depletion is acceptable if it is below a 5 (A), 10 (B1), 15 (B2), 25 (B3) and 50% (C-F) reduction. These guidelines are based on knowledge more than 25 years old. However, the most important limit value of 10% for B1 is supported from similar requirements for trout waters, e.g. in the UK51 where a maximum 10% reduction in habitat area is used as a requirement for salmonidae spawning and nursery areas (B1) and when assuming a ‘‘linear relationship’’ between habitat area reduction and ﬂow reduction which is a fair assumption for the minimum ﬂow regime.

9.3.4

Model Results

In the assessment of the national resource of Denmark, 50 sub-catchments (2–7 in each of the 11 areas) were delineated in order to provide a detailed enough picture for the whole country without being compromised by too large a model uncertainty. For each of the 50 sub-catchments the four indicators were calculated. To include the climate change aspects in a simple way diﬀerent net precipitation (1991–2000) inputs for average climate, dry and wet year were analysed for indicators 1–3. In this way the temporal variability in sustainable yield indicators was assessed for diﬀerent regions and settings. Finally, the indicator with the lowest value of sustainable yield was chosen for mapping the national sustainable abstraction. The summary results for the four indicators are shown in Table 9.3.2. If we take Fyn as an example, the actual abstraction for 2000 was 12.8 mm yr1 (Table 9.3.2). For this area, indicator 4 (baseﬂow reduction) is the most critical of the four indicators (available sustainable resource B10 mm yr1), with B1 (salmonidae spawning and nursery waters) as the river reaches deﬁning the sustainable abstraction. The current abstraction of 12.8 mm yr1 gave a reduction for Fyn of 11% in the minimum ﬂow situation, which is slightly above the limit value of 10%. Indicators 1 and 2 (groundwater quality) result in a less critical available resource estimate (15–17 mm yr1), based on a calculation of deep recharge of 46 mm yr1 without pumping, and 51 mm yr1 for current year 2000 abstraction. Based on these simulations and simulations of deep recharge for Fyn for 50%, 80%, 120% and 150% of the 2000 abstraction for dry and wet conditions, indicators 1 and 2 were estimated to the ranges shown in Table 9.3.2. The reduction in average streamﬂow (indicator 3) is less critical, compared to other indicators, showing an available resource range of 17–29 mm yr1. Based on the indicator results 1–4, the ‘‘worst case scenario’’ indicator is picked, which for Fyn is indicator 4, and the last two

Fyn W-Sjælland S-Sjælland N-Sjælland S-Jylland SW-Jylland SE-Jylland W-Jylland E-Jylland N-Jylland Total

12.8 7.9 6.0 39.1 26.0 66.4 30.8 36.0 13.1 14.6

Abstraction 2000 mm/ yeara

2945 3281 3207 2831 4500 5263 4705 5291 4418 5478

Area km2 10 10 8 14 47 60 26 39 23 22

Ind.4 mm/year 15–17 9–10 8–8 25–30 47–52 57–71 28–31 67–86 34–41 33–42

Ind.1 mm/ yearb 15–16 9–10 8–8 23–27 40–45 49–61 25–27 58–75 30–37 29–37

Ind.2 mm/yearb 17–29 17–28 21–27 12–23 452 40–68 41–64 450 26–38 31–41

Ind.3 mm/yearb 10 9 8 12 40 40 25 39 23 22

Available resource mm/year

30 28 26 33 180 211 118 207 102 121 1054

Available resource mill m3/y

38 26 19 111 117 349 145 190 58 80 1133

Total Abstract mill m3/ ya

Assessment of available resources for Fyn, Sjælland and Jylland (the results for Bornholm are not yet available). Results for indicator 1–3 (groundwater quality, streamﬂow decrease and climate variability), compared to indicator 4 (baseﬂow reduction).

a Abstraction for year 2000 for water supply wells, industry etc. Assumed full irrigation according to irrigation permissions (the actual abstracted volume for irrigation only amounted to 1/3 of the permission for 2000). b The range signiﬁes an estimate of the indicator value given a net precipitation input of 80% and 120% of the average value for 1991–2000. This corresponds approximately to ‘critical’ dry and wet 5-year periods estimated to occur approximately once every century, based on precipitation variations for the period 1974– 2000 (Henriksen and Sonnenborg 2003:43–45).

1 2 3 4 5 6 7 8 9 10

Region

Table 9.3.2

Evaluation of the Quantitative Status of Groundwater–Surface 597

598

Chapter 9.3

columns in the table show the result of the assessment in million m3 yr1, allowing an easy comparison of the balance between available groundwater resource and total abstraction for each region and in total for the country. In addition to the estimates for regions shown in Table 9.3.2 estimates for sub-areas e.g. vital drinking water areas (in Danish, Omra˚der med særlige drikkevandsinteresser) and for sub-areas in each region were determined. For Fyn there are six sub-areas, and the available resource for the country based on sub-areas is slightly reduced compared to the total estimate based on regions shown in Table 9.3.2. Based on the ‘‘sub-area scale’’ of the 50 sub-areas the total available resource comprises 1024 106 m3 yr1, which is 30 106 m3 yr1 or 3% less than the ‘‘region scale’’ result. This ﬁnding also indicates that the total resource estimate is ‘‘scale dependent’’, e.g. that indicator criteria have to be reconsidered if using the methodology and the indicators for diﬀerent scales. In Figure 9.3.5 the resource estimates for sub-areas and regions are shown. For the whole country exploitable groundwater resources is estimated to be 1.0 109 m3 yr1. The assessment in Figure 9.3.5 depicts areas around Copenhagen, Odense and A˚rhus as overexploited areas due to abstraction for water supply. In addition, areas with coarse sandy soils in western Jylland are also threatened by overexploitation due to irrigation demands. In most of these areas the problem of overexploitation is related to excessive streamﬂow depletion caused by abstraction above the limit value according to reduced low ﬂow in rivers (indicator 4). In other areas the limits for how much water that can be abstracted are deﬁned by the risk of increased percolation of nitrates and pesticides to depth from the contaminated shallow groundwater and/or release of toxic solutes from soil matrix (e.g. nickel) caused by lowering the groundwater table. The increased detection of contaminants in shallow aquifers over the past several decades has forced a change in groundwater abstraction patterns from shallow to deep aquifer systems. The new assessment provides a more reliable quantiﬁcation of the exploitable resources due to a direct and thorough incorporation of restrictions on streamﬂow depletion, corresponding to the deﬁned objectives for the aquatic environment for the single stream and river reaches. The uncertainty related to the new assessment has been estimated to 10% for the total exploitable resources. For the 11 regional model areas into which the model has been subdivided, the uncertainty has been estimated to 20% (in size corresponding to Danish WFD areas). For 50 sub-areas the uncertainty has been estimated to 40%. To reduce these uncertainties a more detailed model with a ﬁner grid and a more explicit analysis of the spreading of the shallow contamination towards deeper aquifers are required. Intensive abstraction impacts groundwater vulnerability, both in terms of increased risks of pollution from land surface, and also in terms of increased risks for release of solutes from the subsurface when lowering the groundwater table. It has been estimated based on groundwater monitoring data that abstraction of a critical proportion of maximum 35% of the groundwater recharge to the deep groundwater aquifers is sustainable (at depths of 30 to

599

mm/year

Evaluation of the Quantitative Status of Groundwater–Surface 80 70 60 50 40 30 20 10 0

Sustainable yield Current abstraction

Subareas Exploitation rate (%) < 75 75-90 90-110 110-125 125-150 150-250 > 250

0

Figure 9.3.5

50 km

Resource availability status. Light grey areas: water available (sustainable yield above current exploitation); gray areas: no water available (current abstraction and sustainable yield is in balance); dark grey: overexploited areas. White bars show sustainable yield and dark bar current abstraction for regions.

50 m below surface). These assumptions, based on rough best estimates for Sjælland, are important for the calculation of the exploitable resources, and should be tested for other areas (Fyn and Jylland). Furthermore, there is a need for additional detailed studies. The analysis of critical streamﬂow depletion limit values was based on ﬁgures from Danish guidelines for water supply planning from 1979. There is a strong need for new and better estimates of limiting values for critical streamﬂow depletion for both average ﬂow and low ﬂow conditions, linked to ecological parameters, e.g. using habitat models. Another issue which needs further consideration is the choice of reference scenarios in urban areas (like the capital area of Copenhagen). For example, does it make sense to base the reference situation on conditions where the current groundwater abstractions are ‘‘turned oﬀ’’ (natural conditions), in an area where creeks and headwaters were long ago drained?

600

Chapter 9.3

9.3.5

Discussion and Conclusions

9.3.5.1

Appropriateness of Approach for the WFD

The approach is considered appropriate for the WFD. It oﬀers an integration of groundwater and surface water by the use of groundwater and surface water models that can be used to analyse the interaction between these two domains. The sustainability criteria focusing on avoiding signiﬁcantly negative impacts of groundwater abstraction on both surface water ecology (criteria 3 and 4) and groundwater quality (criteria 1 and 2) are well in line with the underlying WFD principles. Furthermore, the approach provides a transparent, practical and scientiﬁcally based methodology for assessment of sustainable groundwater abstraction. As phrased by Sophocleous,7,52 sustainability depends on the entire system, ‘‘not just the trees, but the whole forest; not just the ﬁsh, but the marine food chain; not just the groundwater, but the running streams and wetlands, and all the plants and animals that depend on them.’’ Sustainability is a goal for the long-term welfare of both humans and the environment. Additionally, any scientiﬁcally based evaluation of sustainability requires model support to assess the behaviour of all the important ﬂow processes within the hydrological cycle and to assess the behaviour of the aquifer, including the interaction with surface water systems, and its sustainable exploitation. Devlin and Sophocleous31,33 support this choice of recharge to the aquifer as an important part of such an assessment of sustainability arguing that it is evident that sustainability is a function of recharge, and that recharge rates cannot be ignored. But sustainable use of groundwater in the WFD must ensure not only that the future resource is not threatened by overuse, but also that natural environments that depend on the resource, such as baseﬂows, riparian vegetation, aquatic ecosystems and wetlands, are protected. By applying two indicators, one for evaluation of inﬂuence on mean river ﬂow and the other for evaluation of reduction of river low ﬂow when pumping under present groundwater development conditions compared to the pre-developmental (virgin) situation, a sound and intuitive methodology ensures that both groundwater recharge and runoﬀ are included. Thereby, the groundwater body as an entity is encapsulated by the suite of four selected indicators. By using four indicators it is possible to view the resource either as a question of ecological sustainability or as a groundwater quality sustainable abstraction problem or even, as presented for the Danish case, by assuming that both conditions should be supported simultaneously, by selecting the most critical of the four indicators when assessing a sustainable resource. However, although the general methodology is sound and balanced with the four indicators, the argumentation for their speciﬁc limit values, e.g. a maximum abstraction of 35% of the groundwater recharge to an aquifer entity in a depth of 30–50 m below the surface in the virgin, pre-development situation or a maximum reduction of baseﬂow by, say, 10%, have to be based on proper monitoring data. Without any relation to monitoring these assessments are not valuable for

Evaluation of the Quantitative Status of Groundwater–Surface

601

marking the boundary between sustainable yield and overexploitation of an aquifer.11 Of course they may be used for overall mapping purposes, but as something that guides the WFD and the management of groundwater in reality. The challenge is to link the modelling approach with monitoring.

9.3.5.2

Linking the Modelling Approach with Monitoring

Even though the main cause of groundwater pollution is related to poor land use, intensive agriculture and insuﬃcient industrial and domestic waste treatment and disposal, there is a growing recognition that changes in groundwater ﬂow and groundwater level (and abstraction) can signiﬁcantly change the chemical composition of groundwater with detrimental eﬀects on the sustainable yield for an aquifer.10,53 Examples include sea water intrusion as a result of intensive pumping of near-coastal aquifers, the release of toxic constituents (manganese, iron, selenium, sulﬁde, nickel) as a consequence of lowering groundwater levels and subsequent entry of oxygen into the previously anaerobic environments. Furthermore, certain chemicals intentionally or unintentionally released to groundwater environments, like pesticides and petrol products, may undergo transformation and degradation processes in the subsurface rendering them more harmful to the environment and health than their original counterparts.10,53 These interrelated factors and processes are complex and diﬃcult to predict which means that site-speciﬁc and chemical-speciﬁc knowledge and data are necessary with requirements for proper and focused groundwater monitoring. Limitations in data and analyses can result in misinterpretation of groundwater conditions, primarily due to the use of an inadequate conceptual model. There is a great need for improved data collection to better estimate groundwater conditions, including long-term changes in storage by aquifers and for an appropriate hydrological study period, and for understanding future water availabilities. Long-term, systematic monitoring and assessment programmes are integral to sustainable, adaptive groundwater management.12 The EU WFD provides new requirements for the monitoring of the freshwater cycle. Errors in estimation of the water balance will aﬀect the accuracy of mass loading calculations. Estimates of the various water balance elements can be strengthened by a combined use of monitoring and modelling. This is particularly the case for groundwater recharge, which cannot be measured directly. Use of a catchment-scale integrated surface water/groundwater model like the DK model is an obvious opportunity in this respect. In the Danish monitoring programme single elements of the water balance (quality and quantity) are monitored in supply wells, monitoring wells from 70 small monitoring areas (GRUMO) and in ﬁve small catchments (LOOP). However, these sub-programmes are only integrated to a limited degree and each of them only provides windows instead of the complete picture of the state of the water environment. Therefore, a combination of the monitoring and the national model provides an opportunity to get a more complete picture of the water

602

Chapter 9.3

status. On the one hand, the model can be used to make knowledge-based (e.g. using geological data) interpolation between the various monitoring sites and provide information on variables such as groundwater recharge that are impossible to measure in practise. Furthermore, the model can help in separating variability (noise) generated by climate from anthropogenic eﬀects. On the other hand, the model is based on data from the monitoring programme both for conceptualisation of aquifer systems and for selecting suitable indicators such as 35% of the pre-development recharge being considered as a sustainable fraction of groundwater abstraction.

9.3.5.3

Strengths and Weaknesses of Approach

The sustainability criteria represent both the major strength and the major weakness of the approach. The fact that the approach is based on transparent criteria is a major advantage. Many sustainability studies in practice rely on qualitative criteria such as the deﬁnition of safe yield by Todd29 as an amount of water abstraction that does not produce undesirable results. Such qualitative (‘‘undesirable’’) criteria enlarge the room for non-speciﬁc politically oriented statements and ambiguity among stakeholders and water resources managers. We believe that it is much sounder to deﬁne quantitative criteria and then use models to assess the consequences. This will not remove the diﬀerences of interests among stakeholders, but it will make the dialogue more knowledge based and more transparent. The major weakness then lies in the establishment of the speciﬁc criteria. All of the criteria we have selected can be subject to a dispute and the knowledge bases behind some of them are arguable. The criteria aimed at ensuring groundwater quality (indicators 1 and 2) are based on large-scale monitoring data from only a part of the country (Sjælland) and it is not documented that the speciﬁc ﬁgures also apply to hydrogeological conditions in other parts of the country. It may also be argued that the indicators instead ideally should be based on a more qualiﬁed and precise analysis of the dynamics of ﬂow system development and the possible spreading of shallow contamination towards deeper aquifers, including the inﬂuence of abstraction. This was, however, not possible within the scope of the present study and would, among other things, require a lot of detailed data that were not available. The analysis of critical streamﬂow depletion limit values was based on ﬁgures from Danish guidelines for water supply planning from 1979. There is a strong need for new and better estimates of limiting values for critical streamﬂow depletion for both average ﬂow and low ﬂow conditions, linked to ecological parameters, e.g. using habitat models. Another issue which needs further consideration is the choice of reference scenarios in urban areas (e.g. in Copenhagen). For example, does it make sense to base the reference situation on conditions where the current groundwater abstractions are ‘‘turned oﬀ’’ (virgin, pre-development conditions), in an area where creeks and headwaters were long ago drained?

Evaluation of the Quantitative Status of Groundwater–Surface

603

A better evaluation of climate impacts, especially in areas where the sustainability indicators show signiﬁcant dependency of net precipitation, is important. This includes a more detailed analysis of hydrological impacts from irrigation. Coupling of advanced regional climate models with advanced hydrological models is a promising opportunity. A weakness in the approach is that water level is not part of any of the proposed four indicators; however, this may also be seen as a strength because the approach is entirely ﬂow based, and therefore much more robust and useful for assessment of groundwater aquifers as an entity. However, for optimisation of situations at well ﬁelds more detailed studies and analysis are required. In general the strength of the approach is the robust ﬂow-based methodology where the sustainability factors (in percentage of ﬂow, e.g. 35% of deep recharge for indicator 1 and 10% reduction of baseﬂow discharges from groundwater to surface water for salmon spawning for indicator 4) are assessed based on groundwater quality monitoring data, and/or ‘‘rule of thumb’’ assessments of how much baseﬂow can be reduced without impacting ecological goals. An important weakness is that the approach is designed for a speciﬁc scale (300–2000 km2) of sub-areas and that use on other scales requires reassessment of ‘‘sustainable fractions’’ for the four indicators. Furthermore, for management of licences and optimisation of abstractions in an area, more speciﬁc and physical-based indicators are necessary, e.g. those that focus on groundwater levels and/or water levels and water moisture contents in wetlands and surface water systems. The DK model is primarily set up for simulating ﬂow and not water levels, which require a much more detailed approach for deﬁning the variations in water level (grid reﬁnement, more detailed representation of rivers, drains and abstractions and a more detailed geological model, especially for the shallow ﬂow system). The scenario approach makes it ﬂexible for exploring the eﬀects of the alternative approaches and linking groundwater and surface water, groundwater quantity modelling and quality monitoring, rural and urban areas and groundwater resources and socioeconomic factors. These links govern the human decisions on how to develop and beneﬁt from natural resources and ultimately adapt to potential negative consequences of overexploited and degraded resources. This interrelatedness of the DK model and the four indicators is perhaps the greatest strength of the approach. Thus it will be very easy to carry out new scenario calculations if stakeholders and water managers want to study the eﬀects of using other sustainability criteria. A very rewarding additional output from the modelling process was that it provided a framework for quality assurance of data and hydrogeological process understanding. Quality assurance can only be fully ensured if data are used, and modelling in this respect may be considered as the ultimate data usage, because it enables consistent checks of one data type against another. During the modelling process we experienced a large number of errors in data and conceptualisation that were corrected after feedback from the numerical model. One example is that we discovered signiﬁcant water balance errors

604

Chapter 9.3

(groundwater recharge and streamﬂow discharge overestimated by around 20%) that might have aﬀected the estimates of nitrate leaching.

9.3.5.4

Novelty of This Work

The construction of a national hydrological model of the present complexity is a major task and a novelty. In particular, the task of processing all the data on geology, soil type, land use, topography, river network geometry, water abstraction and climate to ﬁt into the numerical model is comprehensive and challenging.34 Comprehensive because it involves a vast amount of data originating from diﬀerent databases, and data processing entails a considerable amount of work. Challenging because all these data have never been used together before and they inevitably will contain some mutual inconsistencies. Development of criteria for sustainable developments enabling scenario simulations based on four diﬀerent sustainability indicators applied for examining the inﬂuence on river runoﬀ (mean and minimum ﬂow) and groundwater recharge (drawdown and water quality issues) using a practical model-based approach to predict the present quantitative exploitation is another novelty. In the approach emphasis is directed toward the documentation of the predictive capability of models in order to avoid the often and sometimes with good reason questioned credibility of model-based sustainable resource assessments for groundwater aquifers and/or river catchments. The major novelty and perspective of this work are probably the combination of a comprehensive integrated groundwater/surface water model and the sustainability criteria for a tool used for scenario simulations on issues that are directly relevant for the WFD implementation.

Acknowledgements The DK model project was ﬁnanced by the Danish Ministry of Environment and Energy during 1996–2001. The assessment of the sustainable yield of the Danish water resource using four indicators was ﬁnanced by GEUS with several institutions providing data and cooperation/co-authoring of a theme report:43 DMI, DMU and DJF. The DK model was constructed with a code development and code assistance (MIKE SHE) from DHI.

References 1. M. Eisele, A. Steinbrich, A. Hildebrand and C. Leibundgut, Phys. Chem. Earth, 2003, 28, 529–536. 2. M. Sophocleous, Hydrogeol. J., 2002, 10(1), 52–67. 3. T. C. Winther, J. W. Harvey, O. L. Franke and W. M. Alley, Ground Water and Surface Water, Circular 1139, US Geological Survey, Washington, DC, 1999.

Evaluation of the Quantitative Status of Groundwater–Surface

605

4. W. M. Alley, R. W. Healy, J. W. LaBaugh and T. E. Reilly, Science, 2002, 296, 1985–1990. 5. M. Alley and S. A. Leake, Ground Water, 2004, 42(1), 12–16. 6. J. D. Bredehoeft, Ground Water, 2002, 40(4), 340–345. 7. M. Sophocleous, Hydrogeol. J., 2005, 13, 351–365. 8. L. F. Konikow and E. Kendy, Hydrogeol. J., 2005, 13, 317–320. 9. E. Custodio, Hydrogeol. J., 2002, 10, 254–277. 10. K. G. Vilhollth, Hydrogeol. J., 2005, 14, 330–339. 11. R. Llamas, Water and Ethics. Use of Groundwater, UNESCO Series on Water and Ethics, Essay 7, 2004 (ISBN 92-9220-022-4). 12. V. K. Grabert and T. N. Narasimhan, Hydrogeol. J., 2006, 14, 407–423. 13. J. C. Refsgaard, H. J. Henriksen, W. G. Harrar, H. Scholten and A. Kassahun, Quality assurance in model based water management: review of existing practice and outline of new approaches, Environ. Model. Software, 2005, 20(10), 1201–1215. 14. J. C. Refsgaard, Discussion of model validation in relation to the regional and global scale, in Model Validation: Perspectives in Hydrological Science, ed. M. G. Anderson and P. D. Bates, John Wiley, 2001, pp. 461–483. 15. W. J. De Lange, Eur. Water Pollut. Control, 1996, 6(5), 63–67. 16. South Florida Management District, Draft Documentation for the South Florida Water Management Model, Hydrological Systems Modelling Department, Water Supply Division, SFWMD, West Palm Beach, FL, 1997 (http://glacier.sfwmd .gov/org/pld/hsm/models/sfwmm/fact_sht.htm). 17. F. A. d’Agnese, M. C. Hill and A. K. Turner, Adv. Water Resour., 1999, 22(8), 777–790. 18. J. C. Refsgaard, H. R. Sørensen, I. Mucha, R. Rodak, Z. Hlavaty, L. Bansky, J. Klucovska, J. Topolska, J. Takac, V. Kosc, H. G. Enggrob, P. Engesgaard, J. K. Jensen, J. Fiselier, J. Griﬀeoen and S. Hansen, Water Resour. Manag., 1998, 12, 433–465. 19. M. A. Sophocleous, J. K. Koelliker, R. S. Govindarajy, T. Birdie, S. R. Ramireddygari and S. P. Perkins, J. Hydrol., 1999, 214, 179–196. 20. M. Sophocleous, Geol. Surv. Bull., 2004, 249, 1–102. 21. M. Sophocleous, J. Hydrol., 2000, 235, 27–43. 22. B. Clausen, I. G. Jowett, B. J. F. Biggs and B. Moeslund, Stream ecology and ﬂow management, in Hydrological Drought: Processes and Estimation Methods for Streamﬂow and Groundwater, ed. L. Tallaksen H. and van Lanen, Developments in water sciences, no. 48, 2004, ch. 10. 23. M. J. Dunbar, A. Gustard, M. C. Acreman and C. R. N. Elliot, Overseas approaches to setting river ﬂow objectives, R&D Technical Report W6-161, Environment Agency and Institute of Hydrology, Wallingford, 1998. 24. I. G. Jowett, Instream ﬂow methods: a comparison of approaches, Regul. Rivers Res. Manag., 1997, 14, 115–127. 25. N. Lamouroux and I. G. Jowett, Canad. J. Fish. Aqu. Sci., 2005, 62(1), 7–14. 26. R. E. Tharme, River Res. Appl., 2003, 19(5–6), 397–441. 27. S. E. Gergel, M. G. Turner, J. R. Miller, J. M. Melack and E. H. Stanley, Aqu. Sci., 2002, 64(2), 118–128.

606

Chapter 9.3

28. M. D. Newson and C. L. Newson, Progr. Phys. Geogr., 2000, 24(2), 195–217. 29. D. K. Todd, Ground Water Hydrology, John Wiley, New York, 1959. 30. F. R. P. Kalf and D. R. Woolley, Hydrogeol. J., 2005, 13, 295–312. 31. J. Devlin and M. Sophocleous, Hydrogeol. J., 2005, 13, 549–554. 32. H. A. Loa´icisa, Comment on ‘‘The persistence of the water budget myth and its relationship to sustainability’’ by J. F. Devlin and M. Sophocleous, Hydrogeology Journal, 2005, 13, 549–554, Hydrogeol. J., 2006, 14(7), 1383–1385. 33. J. Devlin and M. Sophocleous, Hydrogeol. J., 2006, 14(1–2), 267–267. 34. H. J. Henriksen, L. Troldborg, P. Nyegaard, T. O. Sonnenborg, J. C. Refsgaard and B. Madsen, J. Hydrol., 2003, 280, 52–71. 35. T. O. Sonnenborg, B. S. B. Christensen, P. Nyegaard, H. J. Henriksen and J. C. Refsgaard, J. Hydrol., 2003, 236, 185–201. 36. L. Troldborg, H. J. Henriksen and P. Nyegaard, DK-model Bornholm: model construction and calibration [in Danish], GEUS report 2006/31, Geological Survey of Denmark and Greenland, Copenhagen, 2006. 37. DHI, MIKE SHE Water Movement User Manual, DHI Water & Environment, Hørsholm, Denmark, 2003. 38. M. B. Abbott, J. C. Bathurst, J. A. Cunge, P. E. O’Connel and J. Rasmussen, J. Hydrol., 1986, 87, 45–59. 39. M. B. Abbott, J. C. Bathurst, J. A. Cunge, P. E. O’Connel and J. Rasmussen, J. Hydrol., 1986, 87, 61–77. 40. J. C. Refsgaard and B. Storm, MIKE SHE, in Computer Models of Watershed Hydrology, ed. V. P. Singh, Water Resources Publication, 1995, pp. 809–846. 41. J. C. Refsgaard, J. Hydrol., 1997, 198, 69–97. 42. K. J. Beven, A discussion of distributed hydrological modelling, in Distributed Hydrological Modelling, ed. M. B. Abbott and J. C. Refsgaard, Kluwer Academic, 1996, pp. 255–278. 43. H. J. Henriksen and A. Sonnenborg, Ferskvandets kredsløb [in Danish], NOVA 2003 Temarapport, Geological Survey of Denmark and Greenland Report, 2003 (download from www.vandmodel.dk). 44. E. P. Poeter and M. C. Hill, Documentation of UCODE, a computer code for universal inverse modelling, USGS report 98-4080, US Geological Survey, Denver, CO, 1998. 45. J. C. Refsgaard and H. J. Henriksen, Adv. Water Resour., 2004, 27, 71–82. 46. B. S. Christensen and T. Sonnenborg, Grundvandsstandens udvikling pa˚ Sjælland 1989–2001 [in Danish], in press. 47. B. Hasler, T. Lundhede, L. Martinsen, S. Neye and J. S. Schou, Valuation of groundwater protection versus water treatment in Denmark by choice experiments and contingent valuation, NERI Technical Report no. 543, Ministry of the Environment, Denmark, 2005. 48. J. Stockmarr, Groundwater quality monitoring in Denmark. Geological Survey of Denmark and Greenland Bulletin 7, Review of survey activities, 2004.

Evaluation of the Quantitative Status of Groundwater–Surface

607

49. T. O. Sonnenborg, B. S. B. Christensen, L. v. Roosmalen and H. J. Henriksen, Klimaændringers betydning for vandkredsløbet I Danmark, GEUS 2006/22, 2006. 50. L. v. Roosmalen, J. H. Christensen, M. Butts, B. S. B. Christensen, K. H. Jensen, J. C. Refsgaard and T. O. Sonnenborg, Climate change for a study on the eﬀects of future climate change on water resources in Denmark, in XXIV Nordic Hydrological Conference 2006, NORDIC WATER 2006, Experiences and Challenges in Implementation of the EU Water Framework Directive, ed. C. Refsgaard and Højberg, NHP Report no. 49, 2006, pp. 583–592. 51. M. Acreman, Personal communication with Mike Acreman, CEH Wallingford with reference to the Water Resources Act 1991, Environment Agency appeal by Thames Water Utilities Limited, Axford Abstraction, File No. WAT/95/22, 2003. 52. M. Sophocleous, Ground Water, 1997, 35(4), 561. 53. B. L. Morris, A. R. L. Lawrence, P. J. C. Chilton, B. Adams, R. C. Calow and B. A. Klinck, Groundwater and its susceptibility to degradation: a global assessment of the problem and options for management, Early warning and assessment report series, RS 03-3, UNEP, Kenya, 2003.

10. Modelling

CHAPTER 10.1

Conceptual Models in River Basin Management ANTONY CHAPMAN,a JOS BRILS,b ERIK ANSINK,c CE´CILE HERIVAUXd AND PIERRE STROSSERe a

r3 Environmental Technology Ltd, c/o School of Horticulture and Landscape, University of Reading, TOB2, Earley Gate, Whiteknights, Reading RG6 6AU, UK; b Netherlands Organisation for Applied Scientiﬁc Research (TNO), Built Environment and Geosciences, Business Unit Groundwater and Soil, PO Box 80015, NL-3508 TA Utrecht, The Netherlands; c Wageningen University, Environmental Economics and Natural Resources Group, PO Box 8130, NL-6700 EW Wageningen, The Netherlands; d BRGM, Water Department, 3 avenue Claude Guillemin, BP 36009, FR-45060 Orle´ans cedex, France; e ACTeon, Le Chalimont, BP Ferme du Pre´ du Bois, FR-68370 Orbey, France

10.1.1

Introduction

Approaches to the management of water, water bodies and the wider environment across Europe have been radically altered with the introduction of the European Union (EU) Water Framework Directive (WFD)1 and the newly adopted Groundwater Daughter Directive (GWD).2 The WFD promotes the integrated management of water resources based on the natural geographical and hydrological unit of the river basin rather than administrative or political boundaries. Whereas previous approaches would generally assess the chemical quality of a stretch of river or a water body such as a lake or an aquifer, the WFD and GWD have stipulated that environmental quality should be assessed holistically on a large scale.3 It is now necessary to assess the whole river– groundwater–soil–sediment system with a view to achieving ‘‘good ecological status’’ and ‘‘good chemical status’’ for all waters and to develop a strategy for achieving these goals through a River Basin Management Plan (RBMP)w. w

See http://ec.europa.eu/environment/water/water-framework/overview.html for an introduction to the WFD including river basin management planning.

611

612

Chapter 10.1

This requires a clearer understanding of the environmental, social and economic inter-relationships within large and complicated river basin systems that calls for greater development and use of conceptual models to support the implementation of this major European legislation. This chapter initially describes and deﬁnes what is meant by a model in general and a conceptual model in particular, including a discussion of the types of models that may be employed and their advantages and disadvantages. It subsequently describes the application of conceptual models in the ﬁeld of river basin management in the light of the implementation of the WFD. A particular emphasis is placed on the needs of the wide range of stakeholders that must be considered in the development of such models and the work that must be undertaken to achieve models suitable for their ultimate purpose. Practical experiences of applying conceptual models to river basin management are drawn from the work of the ongoing European Commission (EC)-funded AquaTerra projectz to inform the wider discussion of conceptual basin models. The chapter concludes with a summary of possible future directions for the use of conceptual models in river basin management.

10.1.2

Integrated Water Resource Management

Worldwide, river basins are under pressure from economic activities that aﬀect their chemical and ecological status and deplete available soil, sediment and water resources. The wide range of economic activities and the hydrological complexity of many river basins, in terms of both the functioning of the soil– sediment–water system and of the links between water quality, quantity and economic activities, make the integrated analysis and modelling of river basins diﬃcult and challenging – in particular when policy support is one of the aims of such analysis and modelling. In Europe, pollution from agriculture, together with morphological pressures (the physical alteration of the channel for water supplies, hydroelectricity and ﬂood control) are seen as the two main issues endangering the achievement of good ecological status of European river basins.4 In addition water is both an input to many industrial production processes and a sink for their pollutants and wastewater, while households also use water for consumption and cause pollution with hazardous substances. Furthermore, other economic sectors such as navigation and hydropower rely on minimum water levels for their functioning and river basin ecology is damaged both by shortages of water and water pollution. Water prices are generally low and water-eﬃcient technologies and practices are not yet fully implemented in many sectors. Additional factors such as population growth, economic growth and possible eﬀects of climate change on river ﬂow are expected to increase existing pressures on river basins. Thus there is a two-way interaction between economic activity and river basin resources that needs to be understood in z

AquaTerra (reference GOCE 505428) is an Integrated Project of the EC 6th Framework Programme (http://www.eu-aquaterra.de).

Conceptual Models in River Basin Management

613

order to support policy decisions dealing with both economic and environmental issues. Since the 1992 United Nations Conference on Environment and Development in Rio de Janeiro, it has become widely accepted that such problems require integrated water resources management at the basin scale. Integrated water resources management considers water ‘‘as an integral part of the ecosystem, a natural resource and a social and economic good, whose quantity and quality determine the nature of its utilisation’’.5 It is against this background that the WFD operates and a ﬁrst step in such integrated analyses and modelling is the development of integrated conceptual models expressing the relationships between economic actors and their use of river basin resources. For successful implementation of the WFD, a holistic understanding of both the natural system (biophysical, river–sediment–soil–groundwater) and the societal system (management, policy-making, society and the economy) as well as an understanding of their interrelationships is required.6 In addition, the future functioning of the system as a whole, including the impacts of changing climate conditions, land use practices and pollution, needs to be understood and integrated in order to adequately establish river basin management as well as groundwater quality control plans that will continue to be relevant throughout their lifespan (six years in the case of RBMPs).

10.1.3

Conceptual Models in the Context of River Basin Management

10.1.3.1

What are Conceptual Models?

The most straightforward deﬁnition of a model is that it is a simpliﬁcation of reality, created in order to assist in the clariﬁcation and understanding of some aspect of the real world.7 The key to the success of such a model is achieving an appropriate balance between simplifying a complex reality, making it both easier to understand and applicable to a wider range of circumstances, whilst preserving the most important relationships in order to obtain results that are a reliable, representative indication of the functioning of the original system.8 Such a model can be used to assess changes in physical characteristics, for example a hydrological model to assess responses to rainfall events, or to investigate changes in the socioeconomic characteristics of a system, such as a model that investigates population movements and their implications for water resource management. A conceptual model is a theoretical construct of the interrelationships between a range of known and quantiﬁable variables acting within a speciﬁed area of inﬂuence.8 In practical terms, the system investigated is structured into a number of key (decisional and biophysical) processes or subsystems. Each subsystem is characterised by a set of variables (input, characteristics and output) and by the relationships between them, the various subsystems and external (exogenous) drivers or variables, changes within which are not

614

Chapter 10.1

inﬂuenced by the functioning of the system. Depending on the nature of the model and what it aims to achieve, its area of inﬂuence could be a ﬁeld, a river basin or a country for example. Almost all models of any description begin life as a conceptual model; some are developed and expanded into a quantitative model, while others remain as a concept to aid understanding and to develop ideas. As a general principle, the more complicated the system being described in a conceptual model, the less likely it is to be developed into a quantitative model because the greater range of variables involved and the larger number of assumptions made in order to produce a workable model is likely to considerably reduce the accuracy of a quantitative assessment. The notion of a conceptual model is implicit in almost all scientiﬁc research, although it is very rarely made explicit. Almost every scientiﬁc experiment is founded on the basis of some conceptual understanding of the functioning of a system or the inter-relationship between two or more variables. The viability of this understanding is then tested either by the development and testing of a quantitative model or by an experiment, or series of experiments. Both of these options would then be compared with an observed reality as a form of quality control. A conceptual model (and potentially a subsequent quantitative or semiquantitative model) is an ideal format to assist an improved understanding of the inter-relationships between the biophysical and societal system.9 Such integrated models can assist the development and implementation of RBMPs if their development and application ﬁt with the implementation timetable of the WFD.

10.1.3.2

What Role can Conceptual Models Play?

Conceptual models can play diﬀerent roles depending on the process and context in which they are developed. For example, they can play the role of a platform for discussion and collaboration between experts from diﬀerent disciplines. As a conceptual model presents a simpliﬁed structure of the system investigated, in the form of subsystems and their relationships, it becomes easy to establish which subsystem may be of direct or partial relevance to a given discipline (or disciplines) and group(s) of experts. Consequently a conceptual model can help to identify connections between diﬀerent disciplines and thus the type of working relationships that need to be established between them in order to work more eﬃciently. Once ﬁnalised, the conceptual model helps to summarise the interaction between activities and experts. In addition, conceptual models can provide a basis for quantitative modelling: in fact the development of conceptual models is an obligatory step to the development of quantitative models. In view of this potential development role for conceptual models, diﬀerent types of quantitative models that could be developed for river basin management are presented below.

Conceptual Models in River Basin Management

10.1.3.3

615

From Conceptual Models to Quantitative/ Computer-based Models

Integrated models derived from conceptual models for water management at the river basin scale can be broadly classiﬁed into simulation and optimisation models. Simulation models recreate the behaviour of water resources systems based on a predeﬁned set of rules which can be either actual or hypothetical. Such models are used to assess the performance of water resources systems over a long timescale. The technique is ideally suited to studying the system’s response to extreme conditions and thereby to identify the components that are prone to failure. River basin simulation models play an important role in identifying the impacts of given scenarios of global climate change as well as population growth scenarios, changing demand patterns and so on. Two basic forms of simulation model exist, one type assessing river ﬂow and the other river basin quality. Optimisation models allocate water resources based on objective functions (e.g. economic, environmental or multi-objective functions). The aim of these models is to optimise the allocation of available water in a river basin for one or more end uses (or users) subject to a given set of rules. They must include a simulation component that is capable of calculating hydrological ﬂows and mass balances in order to quantify the location and quantity of available water.10 While some models can include both simulation and optimisation capabilities, typically economic models are optimisation models whereas hydrological models are simulation models, causing diﬃculties in information exchange between the two. In addition, the integration of two models may be hampered by diﬀerent spatial and temporal scales. Economic impacts may occur over an area diﬀerent to those of hydrological eﬀects, which occur on the scale from a ﬁrst-order catchment through to that of a river basin district (RBD) as deﬁned by the WFD. Temporal scales for economic models are usually longer, forecasting change over a number of years, while hydrological models tend to observe change over the course of a year or a season. Combined economic– hydrological models attempt to overcome these barriers and have, for instance, been frequently applied to the analysis of the economics of irrigated agriculture.10 The combined modelling approach has proved useful for a wide array of management purposes, including irrigation management, reservoir operation, groundwater management and basin management. The state of the art in each of these ﬁelds is described below. Models for the management of irrigation water take into account processes that inﬂuence the quantity and salinity of irrigation water in various timeframes.11,12 The aim of these models, given scarce water supplies, is to optimise the performance of the irrigation system.13 A related branch of the literature focuses on the optimisation of water productivity, paying attention to the hydrological and agronomic aspects of water use14 and examining the issue at diﬀerent scales.15

616

Chapter 10.1

Reservoirs have a role both in the management of water quality and water quantity. Improving the eﬀectiveness of the management and operation of reservoir systems is the goal of a large number of modelling tools. Historically, these models focused on single issues such as ﬂood control or downstream water quality control; subsequently, multi-objective models were introduced.16,17 An integrated approach to reservoir operation and management complicates the analysis of optimal reservoir decisions, requiring the development of advanced tools to model the eﬀects of reservoir decisions on multiple aspects of the river basin. Examples of such tools include MODSIM18 and CalSim.19 Two types of groundwater management models exist. The ﬁrst concerns supply management with an emphasis on water allocation to various users. The second concerns water quality management.20,21 Due to the value of groundwater as a source for the public water supply in many countries, a large body of literature is devoted to its competitive extraction.22 An example of a supply management model is given by McPhee and Yeh,23 who developed a combined groundwater simulation and optimisation model and decision support system. In addition to models that focus on components of the river basin, there is an extensive literature on water management at the basin level. The general framework of such basin-scale models is given by the physical constraints of the basin (e.g. ﬂow, salt balance and transport), plus demand functions for instream purposes (e.g. ecological needs and hydropower), and oﬀ-stream purposes (e.g. agriculture, industry and households). Given information on water rights and water prices, these models aim to maximise the overall beneﬁt of water use in a river basin.24 A series of studies that modelled the economic aspects of river basin water use more explicitly emerged in the 1990s. For example, Rosegrant et al.25 analysed optimal allocation of water over demand sites in the Maipo river basin (Chile), taking into account hydrological, environmental and institutional constraints. Extensions to this model, with improvements in its hydrological and agricultural production components, have been made and implemented in the Syr Darya basin,26 the Mekong basin27 and the Yellow River basin.28 For reviews of water resources management models from an allocative point of view, see Refs. 24 and 29. A second group of basin-scale models places emphasis on water quality rather than water quantity. Its aims are to ﬁnd ‘‘optimal’’ pollution levels in a river basin and to deﬁne where pollution abatement should take place, given heterogeneous impacts of abatement eﬀorts. Other studies analysed the eﬃciency of diﬀerent policy alternatives to achieve temperature reductions in order to increase ﬁsh populations,30 and the least-cost allocation of nitrate emission reductions for regions in the Rhine basin.31 Comparable analyses have been carried out for the Fox-Wolf river basin in the USA32 and for the Baltic Sea.33 Underlying these analyses are models that quantify the sources and retention of nutrients to surface waters.34,35 A review of this type of model has been made in the EUROHARP project.36

Conceptual Models in River Basin Management

617

Two important conclusions can be drawn from this brief overview. First, the diversity of quantitative models implies a diversity of conceptual models developed for addressing diﬀerent environmental issues and policy questions. Second, given that there are a wide range of sophisticated quantitative and semi-quantitative models designed to address a range of river basin management issues, it is reasonable to ask what conceptual models of river basin management can oﬀer that is not oﬀered elsewhere. The answer is that, in the same way that the WFD has required a complete rethink and redesign of river basin management, with holistic planning and a quality judgement based on ecological status rather than chemical concentrations, so the models which can provide information of relevance to the WFD and to a RBMP need to be restructured and rethought. All of the models described previously add value and help to explain the functioning of river basins or parts thereof and the development of new conceptual models does not render such models obsolete. Rather, a conceptual model is a means by which the many diﬀerent needs, demands and pressures to which water is subjected across the scale of a large river basin can be integrated and compared in order to understand and clarify the environmental, social and economic impacts of a range of management strategies on water status across a RBD in the context of a changing world.

10.1.4

Building Conceptual Models in the Context of River Basin Management: Some Principles

10.1.4.1

The DPSIR Framework as a Guide

A useful framework for guiding the development of conceptual river basin models that can help to achieve the goals identiﬁed in Section 3 is provided by the Drivers–Pressures–State–Impact–Response (DPSIR) model (see Figure 10.1.1). This framework was originally developed in the 1970s by Anthony Friend, and popularised by the Organisation for Economic Cooperation and Development (OECD) in the 1990s under the variant of a PSI model. Other organisations such the UN Commission prefer variants of a PSR model. The European Environment Agency follows the DPSIR model to provide an insight into environmental processes and the links between human activities and their impact on the environment. It treats the environmental management process as a feedback loop controlling a cycle consisting of ﬁve stages:37 driving forces (D), pressures (P), state (S), impacts (I) and responses (R).38 Economic activities (driving forces) such as industry, agriculture and tourism lead to increasing pressures on the natural environment as these activities result in use of natural resources and/or emissions (accidental or controlled) of waste to (ground) water,39 soil40–42 and sediment.43,44 The use of resources and/or emissions will change the state of these environments in terms of their quantity and/or quality: sediment, water and soil resources are depleted (erosion) and/or they are loaded (contaminated) with hazardous substances originating from economic activities. Above a certain level of depletion and/or contamination there will be an impact on the environment, such as loss of biodiversity, vulnerability to ﬂoods

618

Chapter 10.1

bio-physical system D

societal system

D P

P

P

R

S S

S

I S

Figure 10.1.1

The Drivers–Pressures–State–Impact–Response (DPSIR) approach (adapted from Ref. 45.)

and landslides, decreased chemical and/or ecological water, soil or sediment quality and/or a shortage of these resources. Several response measures, implemented at any of the DPS or I phases, could prevent this from happening or mitigate the impacts to a level deemed acceptable or tolerable by society. The relevance of the DPSIR framework as an overarching guide to conceptual model development is important from at least two perspectives. First, it ensures that both decisional and biophysical processes are considered when building conceptual models for integrated river basin management. Second, it emphasises the need to identify potential responses at the conceptual phase, as the consideration of diﬀerent policy responses implies investigating diﬀerent biophysical and decisional processes. As an illustration, if a change in agricultural subsidies is proposed as a policy response to reduce diﬀuse pollution, then it is essential to understand farm-level decision-making and the impact of subsidies on cropping pattern and fertiliser use. Understanding farm-level decision-making processes might be less relevant, however, if the policy response imposes a given cropping pattern and fertiliser use for a given sensitive catchment, in which case the decision at the farm scale is taken out of the hands of the landowner.

10.1.4.2

Investigating the Dynamics of River Basin Systems

A conceptual model needs to be built both on a good understanding of the current state of the system and also on its likely future (or futures). As time

Conceptual Models in River Basin Management

619

passes, new subsystems and decision-making processes might emerge, for example when a new economic activity putting pressures on the water environment is developed, or when emerging pollutants become more widely used and discharged into the aquatic environment. Some subsystems might disappear, possibly as a result of human activities; wetlands that can be considered as a subsystem today may no longer exist in 20 years. Others might cease to be relevant. For example many contemporary environmental issues are the subject of environmental policies designed to resolve them, or at least remediate them to the point at which they are no longer an immediate concern. Once this occurs the issue will no longer be considered as a subsystem of the conceptual model. Conversely, new policy responses, such as the introduction of environmental taxes or charges for reducing pollution to groundwater and surface water, might become relevant in the future. As a result, additional decision-making and biophysical processes might need to be understood and included in conceptual models if they are to continue to be policy relevant. Clearly, diﬀerent time horizons should be considered depending on the environmental issue considered or the type of responses that are proposed. When long-term changes such as climate change are part of the policy focus, the need to understand the long-term dynamics of the environmental, social and economic system is essential, even if uncertainties with regard to the likelihood of future long-term changes might be high.

10.1.4.3

Stakeholder Integration and Response

A stakeholder may be deﬁned as: ‘‘Someone who may be aﬀected by, or may aﬀect, a decision that has to be made or its implementation.’’46 Any conceptual model for river basin management must take account of stakeholders in the speciﬁc basin area to which it is applied, particularly those who will make direct use of it and those whose interests could be aﬀected by decisions made using it. Participation is also recommended as it enables the exchange of information, which can lead to a better understanding of the speciﬁc situation and improved co-ordination amongst a potentially disparate group of participants: in this way it also contributes to public support.47 In addition, stakeholders are likely to improve the quality of any conceptual model through their knowledge of speciﬁc local conditions. Stakeholder participation is now required in several ﬁelds, particularly in water management (e.g. WFD Article 14). It is one means of responding to the growing expectation among the general public that they will have an opportunity to express their views on issues they consider relevant to their life and work. It also reinforces the legitimacy of decision-making, provides better understanding of key issues at a local scale and increases the eﬀectiveness of measures chosen and implemented by making people feel actively involved in the process. Examples of the range of potential stakeholders who would be relevant to river basin management within the EU include regional, national and European policy-makers; institutions such as water authorities and (national or

620

Chapter 10.1

international) river basin authorities; companies responsible for the distribution and management of water and wastewater; representatives of relevant industries, agriculture and tourism (trade unions for example); academics and scientists with relevant expert knowledge; government environmental organisations; interest groups, such as national environmental non-governmental organisations or local interest groups; as well as the general public. All of these groups could be consulted, if possible, over the design of conceptual models for river basin management that could aﬀect their livelihood or standpoint on a particular issue. Clearly their views will not always coincide nor will all stakeholders necessarily have a strong viewpoint on, or background in, all of the issues that may be considered important over the scale of a large river basin.

10.1.5

Experience from the Aqua Terra Research Project

10.1.5.1

Context and Objectives

A key objective of the EU-funded AquaTerra project is to develop an integrated modelling approach for the river–sediment–soil–groundwater system which will become the base for new tools for European river basin managers. One subproject of AquaTerra, INTEGRATOR, contributed to this overall objective by combining natural and applied science work undertaken in large European river basins (namely the Danube, Ebro, Elbe and Meuse) at diﬀerent temporal and spatial scales into conceptual models for river basin management. These models were also informed by investigation and consultation through stakeholder consultations, interviews and workshops. The DPSIR framework has been used in the context of the INTEGRATOR subproject as the guiding framework for the development of integrated conceptual models for river basin management. Conceptual models have been developed for diﬀerent case study areas, namely the Geer Catchment (Meuse, Walloon region), the Kempen area (Meuse, Flanders region and the Netherlands), the central Ebro river basin (Spain), the Krsko kotlina aquifer (Danube river basin, Slovenia) and the Meuse river basin (Flanders and the Netherlands). The main objectives of these conceptual models were to develop simpliﬁed representations of the case studies that can be used to discuss integrated water management and possible policy responses; to identify the relevance of AquaTerra’s research for supporting policy discussion and decision-making in each case study area— stressing in particular the knowledge gaps that would need to be ﬁlled if support to policy decisions were at stake; and ﬁnally (not for all case studies) to develop quantitative models focusing in particular on socioeconomic subsystems and decision processes, and to apply these models for performing economic assessments (cost-eﬀectiveness or cost-beneﬁt) of various policy responses.

621

Conceptual Models in River Basin Management

10.1.5.2

Identifying the Main Environmental Issues Relevant to the Case Studies

The ﬁrst step in developing the conceptual models was to identify the main environmental issues of importance for each case study area; the assumption being that diﬀerent environmental issues would imply investigating diﬀerent subsystems and their interactions as well as diﬀerent policy responses. Detailed background reports on each river basin were prepared, giving information about the physical characteristics, population and industrial and agricultural activities in the area.48–51 Within each report, a provisional list of key issues in each basin was identiﬁed for further discussion with stakeholders in a workshop. The issues identiﬁed for each river basin are shown in Table 10.1.1. As might be expected, a wide variety of issues were identiﬁed in the river basins studied, reﬂecting their size and heterogeneity of climate, hydrology and land use. For example, soil erosion was identiﬁed as key issue in the Ebro river basin, but not in other river basins. Thus the speciﬁc processes that inﬂuence

Table 10.1.1

Main issues within the basins identiﬁed prior to the workshops (from Ref. 38). Meuse

Water scarcity

Salinisation

Flooding

Pollution by organic matter

Sediments polluted by heavy metals

Diﬀuse pollution from agriculture

Groundwater pollution due to past mining activity

Soil erosion

Hydro-morphological alterations

Danube

Elbe

Ebro

622

Chapter 10.1

erosion had to be captured in the conceptual model for this case study. At the opposite extreme, decision-making and biophysical processes relevant to diﬀuse pollution from agriculture were relevant to all river basins and needed to be captured in all conceptual models.

10.1.5.3

Integrating Stakeholders’ Views

Stakeholders’ views were sought in the INTEGRATOR subproject as a means of capturing the dynamics of the diﬀerent river basin systems and possible responses one might consider for solving the main environmental issues identiﬁed in each river basin. In particular, the subproject made use of participatory prospective assessment,38 also known in the literature as participatory foresight or participatory future studies. This approach is quite new in the ﬁeld of environmental studies and can be deﬁned as a systematic, participatory, future vision-building process aimed at informing present-day decisions and mobilising joint actions. The prospective approach explores the future in order to obtain a better understanding of how the system functions and to better understand (or plan) the complex decisions that must be made. Its application ranges from strategy development to long-term vision building in ﬁelds such as urban planning, technology and the environment.52 Traditionally, prospective methods are often based on expert consultation and/or modelling. To add a participatory dimension in these methods implies consulting a wide range of stakeholders in the scenario-building process. The method applied included three main phases.38,52 The ﬁrst (pre-scenario phase) dealt with the preparation of the scenario-building process and the development of a ﬁrst impact assessment model at the basin level. The second phase ensured stakeholders’ input and participation in a workshop to develop Storylinesy and to enrich the impact assessment model developed previously. In the third phase, the Storylines were enriched by the results obtained and submitted for evaluation to all participants in the development process. The structure of each workshop varied, but typically they took place over one day and contained three main elements: (1) an outline of AquaTerra activities within the basin in question, (2) a presentation and discussion of key issues in the basin and (3) a discussion of how key issues might change in the future under diﬀerent conditions using the DPSIR framework as basis for interaction and discussion.

10.1.5.4

Developing Simpliﬁed Representations of the Systems Investigated

A literature review, bilateral discussions with AquaTerra researchers, experts and stakeholders from the diﬀerent case study areas, as well as the outcome of y

A storyline is the product of a discussion among groups of stakeholders relevant to a particular physical and/or social phenomenon. The aim of storyline development is to produce a multifaceted understanding of the issue for further investigation.

623

Conceptual Models in River Basin Management

the interactive workshops were used to develop the building blocks of the conceptual models for the diﬀerent case studies. Simpliﬁed representations presenting relevant subsystems and their relationships were then designed for each case study, such as that for the Meuse basin shown in Figure 10.1.2. In addition, the diﬀerent subsystems were described in greater detail, as illustrated in Figure 10.1.3, which shows the detailed inter-relationships of the agriculture

Policy (e.g. WFD)

Climate change

Population growth

Economic growth Upstream water use

Energy

Navigation

Drinking water

Agriculture

Industry

River flow: quantity Surface water quality

Groundwater quality

Figure 10.1.2

The simpliﬁed representation of the Meuse river basin (in Ref. 53).

External drivers • Prices of different inputs • Agricultural policy: quotas and output prices • Environmental legislation: obligations in environmentally-friendly practices • Climate change: rainfall, droughts

Sub-system: farm Objective: profit maximization

Input • Fertiliser (price) • Hired labour (price) • Irrigation water (price)

Figure 10.1.3

Characteristics • Total farm area • On-farm labour • Machinery Relationships and functions Y = f (input, parameters)

Output • Crop production ( ) • Nitrate surplus (kg) • Pesticide surplus (kg)

A simpliﬁed representation of the agricultural system.53

624

Chapter 10.1

subsystem that could be proposed as a basis for capturing agriculture in the diﬀerent case studies.

10.1.5.5

Preliminary Lessons from the Experience of INTEGRATOR

The activities of the INTEGRATOR subproject are still ongoing. The conceptual models developed have not yet been validated by stakeholders and experts and the follow-up development of quantitative models proposed for selected case studies has not yet been initiated. However, preliminary lessons can already been drawn with regard to the development of conceptual models. The experience has re-emphasised the diversity of conceptual models and simpliﬁed representations of river basin systems. The main environmental issues at stake, the dynamics of the system and the diﬀerent policy responses that can be considered are key elements in deﬁning the boundaries of a conceptual model and its diﬀerent elements. However, the notion of a universal conceptual model for river basin management would necessitate systematic consideration of all possible biophysical and decision-making processes, rendering such a model contradictory to its stated purpose of being a simpliﬁed representation of the system considered. The INTEGRATOR project will aim to produce a pro-forma for a European conceptual model for basin management that can be adapted to speciﬁc circumstances on the basis of information reviews, discussion and consultation with relevant stakeholders. Activities undertaken as part of INTEGRATOR stress the importance of stakeholder consultation and involvement. In all workshops, stakeholders provided a considerable amount of valuable insight into key environmental issues, helping to reﬁne and improve initial understanding based on studies and technical expertise and making it more policy relevant. On some occasions, however, stakeholders showed some reluctance to provide speciﬁc input and remained very general, as providing details and speciﬁc (individual) views were seen by them as ‘‘taking responsibility’’ for their organisations without any mandate to do so. The experiences showed that the idea of a conceptual model remains very abstract for stakeholders, in particular when discussed outside of a clear policy development process, as in the case of the research activities of INTEGRATOR. Although not the only justiﬁcation for stakeholder consultation, a policy process does provide the justiﬁcation for building a simpliﬁed representation of the system under consideration that will be instrumental in analysing the implications of policy options and decisions that are directly relevant to stakeholders. Despite these challenges in communicating the notion of a conceptual model to stakeholders, the process of discussion and collaboration is an important and relevant process that needs to be pursued. In addition, as the use of conceptual models and stakeholder discussion becomes more widespread in river basin management through the requirements of the WFD, stakeholders

Conceptual Models in River Basin Management

625

will become more familiar with them and the process will become smoother and more productive. Ultimately the stakeholders in the basin are those who could have need of a model to support their decision-making. They are more likely to make use of a model of which they are aware and already have some understanding, knowing that it has been adapted, at least to some extent, to their individual needs. Communication at an early stage helps to identify what is useful in a proposed model, what is less useful, which assumptions are valid in the speciﬁc circumstances and whether the model is appropriate for the stakeholders themselves.

10.1.6

Conclusions

Models of all descriptions have a valuable role to play in river basin management by reﬁning concepts and processes at a variety of scales to their essential elements, allowing greater understanding of the functionality of a system or process and the interaction between the key components. Conceptual models, being a theoretical understanding of the functionality of a system, are the implicit starting point of most scientiﬁc investigation. The advent of the WFD, which requires holistic understanding of the physical and socioeconomic functions within a RBD, provides an ideal vehicle for the use of conceptual models to build shared understanding and co-operation among a wide range of stakeholders. However, as has been shown from recent experiences in the EC Integrated Project AquaTerra, conceptual models can have limitations. In the ﬁrst instance, no one model is suitable for all circumstances across the EU. A generic framework for developing conceptual models can be proposed, but it leads to diverse conceptual models adapted to the speciﬁc needs of a RBD, its stakeholders and the policy responses and decisions that need to be considered. Clearly, stakeholder consultation must form a key part of this construction and adaptation process. This is a clear challenge and a continuous process of learning and adaptation. The experience of the AquaTerra project supports the contention that while conceptual models have a place in river basin management there is scope for their development and improvement, from the point of view of both model design and stakeholder interaction. The best conceptual models are seen as those which provide the simplest, clearest explanations of reality, while maintaining accuracy, allowing the capacity to make predictions of future change. However, despite the desire for simplicity, a conceptual model must also be relevant. There is a danger in using conceptual models to understand the functioning of a large and complicated system such as a river basin in its entirety that the end result will oversimplify the reality and the results may be of limited value. In contrast the experience of relating the model to stakeholders demonstrated that as straightforward a model as possible may be necessary in many cases for them to make a signiﬁcant input into the consultation process.

626

Chapter 10.1

A key issue to be decided in the development of conceptual models for river basin management is the deﬁnition of the boundaries of use for such models. Will they always simply be a conceptual, illustrative tool, aimed at improving understanding of the system as a whole among disparate groups? Or will they subsequently evolve, with stakeholder consultation and input, into a more sophisticated, quantitative or semi-quantitative tool aimed at providing an assessment of river basin management strategies as well as an understanding of them? The answer may well be that it will be a basin-speciﬁc response to a basin-speciﬁc approach, each incarnation of the conceptual model being adapted to suit the particular basin to which it is applied. Like much of the work associated directly or indirectly with the WFD, conceptual models for river basin management are a work in progress.

Acknowledgements This work was partly supported by the European Commission 6th RTD Framework Programme Integrated Project AquaTerra (project no. GOCE 505428) under the thematic priority Sustainable Development: Global Change and Ecosystems.

References 1. Directive 2000/60/EC of the European Parliament and of the Council establishing a framework for the Community action in the ﬁeld of water policy, European Commission, Brussels, Belgium, 2000. 2. Draft Proposal for a Directive of the European parliament and of the Council on the protection of groundwater against pollution, COM(2003)550, European Commission, Brussels, Belgium, 2003. 3. M. Griﬃths, The European Water Framework Directive: An Approach to Integrated River Basin Management, European Water Management Online, 2002 (available at: http://www.ewaonline.de/journal/2002_05.pdf). 4. M. Menedez, M. De Rooy, G. Broseliske and S. Mol, Key issues and research needs under the Water Framework Directive: ﬁnal document, comprising Phase 1 and Phase 2, December 2005, issue date 26 January 2006. 5. UNCED, Agenda 21, 1992 United Nations Conference on Environment and Development, Rio de Janeiro, 1992. 6. I. Heinz, The economic value of water, Presentation from the International Workshop on Hydro-economic Modelling and Tools for the Implementation of the European Water Framework Directive, Valencia, 30–31 January 2006 (available at: http://www.infu.uni-dortmund.de/pdfs/paper/ Valencia06_Paper_HEINZ_FF.doc). 7. L. M. Rocher and S. Schnell, Modelling the world (available online at: http://informatics.indiana.edu/rocha/i101/pdfs/i101_lecnotes_v3.pdf).

Conceptual Models in River Basin Management

627

8. W. S. Merritt, R. A. Letcher and A. J. Jakeman, Environ. Model. Software, 2003, 18, 761–799. 9. P. N. Manley, W. J. Zielinski, C. M. Stuart, J. K. Keane, A. J. Lind, C. Brown, B. L. Plymale and C. O. Napper, Monitoring ecosystems in the Sierra Nevada the conceptual model foundation, Environ. Monitoring Assess, 2004, 64(1), 139–152. 10. D. Wisser, Water Management Methodologies for Water Deﬁcient Regions in Southern Europe, Institute of Hydrology and Water Management, University of Bochum, Germany, 2004. 11. D. Yaron, E. Bresler, H. Bielorai and B. Harpinist, Water Resour. Res., 1980, 16, 257–262. 12. G. R. Musharraﬁeh, R. C. Peralta, L. M. Dudley and R. J. Hanks, Water Resour. Res., 1995, 31, 1077–1086. 13. P. R. Onta, R. Loof and M. Banskota, Irrig. Drain. Syst., 1995, 9, 143–162. 14. D. Molden and R. Sakthivadivel, Water Res. Develop., 1999, 15, 55–71. 15. G. Kite and P. Droogers, Integrated basin modelling, IWMI research report 43, 2000. 16. R. Harboe, Water Res. Bull., 1992, 28, 103–110. 17. S. Ko, D. G. Fontane and J. W. Labadie, Water Res. Bull., 1992, 28, 111–127. 18. J. Labadie, M. Baldo and R. Larson, MODSIM: decision support system for river basin management, Documentation and user manual, Department of Civil Engineering, Colorado State University, Fort Collins, CO, 2000. 19. A. J. Draper, A. Mune´var, S. K. Arora, E. Reyes, N. L. Parker, F. I. Chung and L. E. Peterson, J. Water Res. Plan. Manag., 2004, 130, 480–489. 20. W. W. G. Yeh, J. Water Res. Plan. Manag., 1992, 118, 224–237. 21. S. M. Gorelick, Water Resour. Res., 1983, 19, 305–319. 22. S. J. Rubio and B. Casino, Environ. Res. Econ., 2003, 26, 7387. 23. J. McPhee and W. W. G. Yeh, J. Water Res. Plan. Manag., 2004, 130, 490–497. 24. D. McKinney, X. Cai, M. Rosegrant, C. Ringler and C. Scott, Modeling water resources management at the basin level: review and future directions, IWMI, SWIM paper 6, 1999. 25. M. Rosegrant, C. Ringler, D. McKinney, X. Cai, A. Keller and G. Donoso, Agricult. Econ., 2000, 24(1), 33–46. 26. X. Cai, D. McKinney and L. Lasdon, J. Water Res. Plan. Manag., 2003, 129(1), 4–17. 27. C. Ringler, J. von Braun and M. Rosegrant, Water Int., 2004, 29(1), 30–42. 28. X. M. Cai and M. W. Rosegrant, Water Resour. Res., 2004, 40(8), W08S04. 29. M. A. Massoud, M. D. Scrimshaw and J. N. Lester, Water Policy, 2004, 6, 519–548. 30. M. Watanabe, R. M. Adams and J. Wu, Am. J. Agric. Econ., 2006, 88, 617– 631. 31. R. J. H. M. Van der Veeren and R. S. J. Tol, Environ. Res. Econ., 2001, 18, 19–41. 32. J. Schleich, D. White and K. Stephenson, Water Resour. Res., 1996, 32, 2879–2884.

628

Chapter 10.1

33. I. M. Gren, P. Jannke and K. Elofsson, Environ. Res. Econ., 1997, 10, 341–362. 34. B. Kronvang, L. M. Svendsen, J. P. Jensen and J. Dorge, Hydrobiology, 1999, 410, 207–212. 35. A. N. Rousseau, A. Mailhot, R. Turcotte, M. Duchemin, C. Blanchette, M. Roux, N. Etong, J. Dupont and J. P. Villeneuve, Hydrobiology, 2000, 422/423, 465–475. 36. O. F. Schoumans and M. Silgram, Review and literature evaluation of quantiﬁcation tools for the assessment of nutrient losses at catchment scale, EUROHARP report 1-2003, NIVA report SNO 4739-2003, Oslo, 2003. 37. Questions to be answered by a state of the environment report, Technical report no. 47, European Environment Agency, Copenhagen, 2000. 38. C. Herivaux, S. Loubier, M. Bouzit, P. Strosser, A. S. Chapman, P. Bardos, E. Ansink, A. Ruijs, L. Maring, L. Gerrits and J. Joziasse, Generic conceptual representation of river basin, and methodological guidelines to construct such a representation with a participatory approach, AquaTerra deliverable Integrator 1.3, BRGM, Montpellier, 2005 (http://www.attempto-projects.de/aquaterra/). 39. EEA (European Environment Agency), Europe’s Water: An IndicatorBased Assessment. Summary, Oﬃce for Oﬃcial Publications of the European Communities, Luxembourg, 2003 (ISBN 92-9167-576-8). 40. W. E. H. Blum, J. Buesing and L. Montanarella, Trends Anal. Chem., 2004, 23, 680–685. 41. W. E. H. Blum, D. Barcelo, J. Bu¨sing, T. Ertel, A. Imeson and J. Vegter, Scientiﬁc Basis for the Management of European Soil Resources, Research Agenda, 2004 (ISBN 3-900782-47–4). 42. L. Van-Camp, B. Bujarrabal, A.-R. Gentile, R. J. A. Jones, L. Montanarella, C. Olazabal and S.-K. Selvaradjou, Reports of the Technical Working Groups Established under the Thematic Strategy for Soil Protection, EUR 21319 EN/1, Oﬃce for Oﬃcial Publications of the European Communities, Luxembourg, 2004. 43. W. Salomons and J. M. Brils (eds), Contaminated Sediments in European River Basins, SedNet booklet as ﬁnal report for the EC FP5 Thematic Network Project SedNet (EVK1-CT-2001-20002), 2004 (downloadable through: www.SedNet.org). 44. J. M. Brils, J. Soils Sed., 2005, 1, 48–49. 45. R. H. Meade, Contaminants in the Mississippi River, 1987–1992. Heavy Metals in the Mississippi River, US Geological survey circular 1133, US Government Printing Oﬃce, Washington, DC, 1995. 46. United Nations, World population prospects: the 2002 revision, Population Division of the Department of Economic and Social Aﬀairs of the United Nations Secretariat, ESA/P/WP 180, 26 February 2003 (http:// www.un.org/esa/population/publications/wpp2002/WPP2002-HIGHLIGHTSrev1.PDF). 47. J. A. Van Ast and S. P. Boot, Phys. Chem. Earth, 2003, 28(1213), 555–562.

Conceptual Models in River Basin Management

629

48. E. Ansink, A. Ruijs and P. Strosser, Danube river basin characterisation, AquaTerra deliverable I1.1e, Wageningen University, The Netherlands/ ACteon, France, 2005. 49. M. Bouzit, C. He´rivaux and S. Loubier, A case study report on the Meuse by BRGM in relation with EUPOL and BASIN, AquaTerra deliverable Integrator, 1.1c, BRGM, Montpellier, 2005 (http://www. attempto-projects.de/aquaterra/). 50. A. S. Chapman and P. Bardos, Ebro river basin characterisation, AquaTerra deliverable Integrator, 1.1f, r3 Environmental Technology, Reading, UK, 2005 (http://www.attempto-projects.de/aquaterra/). 51. L. Maring, L. Gerrits and J. Joziasse, Elbe river basin characterisation, Aquaterra deliverable Integrator 1.1d, TNO, Apeldoorn, The Netherlands, 2005 (http://www.attempto-projects.de/aquaterra/). 52. A. M. Bouzit and S. Loubier, Combining prospective and participatory approaches for scenarios development at river basin level, AquaTerra: Integrated project funded by the European Commission (6th FP), Deliverable INTEGRATOR 1.1a, Report BRGM/RP-53345-FR, BRGM, Montpellier, 2004. 53. C. Herivaux, M. Bouzit, N. Graveline, E. Ansink and P. Strosser, Synthetic report on the economic behaviours and models including a sensitivity analysis of the proposed economic models for each selected study areas and the link with the conceptual model, AquaTerra deliverable Integrator 2.4, BRGM, Montpellier, 2006 (http://www.attempto-projects.de/aquaterra/).

CHAPTER 10.2

Modelling Reactive Transport of Diﬀuse Contaminants: Identifying the Groundwater Contribution to Surface Water Quality HANS PETER BROERS, BAS VAN DER GRIFT, JASPER GRIFFIOEN AND RUTH HEERDINK Netherlands Organisation for Applied Scientiﬁc Research (TNO), Geological Survey of the Netherlands, Princetonlaan 6, PO Box 80015, NL-3508 TA Utrecht, The Netherlands

10.2.1

Introduction

The European Union (EU) Water Framework Directive1 stimulates an integrated approach of the whole soil–groundwater–surface water system. Transport models are potentially very useful to improve insight into the dominant processes which control present and future water quality and to quantitatively evaluate groundwater and surface water protection measures for river basin management plans. Basic requirements for eﬀective transport modelling are the establishment of a clear conceptual idea about system behaviour and the presence of groundwater and surface water quality data to verify model performance. This chapter aims to describe the development of such conceptual models and to highlight opportunities for verifying the validity of the model results using a sophisticated monitoring approach. The approach is targeted towards diﬀuse contaminants. Often, the contribution of groundwater to surface water contamination is largely unknown when we are planning and implementing measures which are meant to enhance surface water quality in river basins and catchments. In some hydrogeological systems, the groundwater contribution is well known, because groundwater discharge appears in well-deﬁned areas with distinct springs or 630

Modelling Reactive Transport of Diﬀuse Contaminants

631

seepage areas. In large regions in Europe, however, groundwater discharge is a spatially diﬀuse phenomenon. This is especially true in ﬂat regions of river deltas, with an intensive, artiﬁcial drainage system of brooks, canals, ditches and drains. In these kinds of areas, it is diﬃcult to obtain reliable information on the quality and quantity of discharging groundwater. One promising approach is to model the groundwater discharge to the surface water system using three-dimensional reactive transport models and to verify these models with groundwater and surface water quality data from monitoring networks.

10.2.2

Framing a Conceptual Model

Three main factors govern the groundwater contribution to surface water: (1) the distribution of groundwater travel time towards the surface water system, (2) the inputs of diﬀuse contaminants to groundwater through leaching from the unsaturated zone and (3) transformation and retardation processes in the unsaturated and saturated zones and the groundwater–surface water transition zone. Consequently, a conceptual model should include a description of how these factors inﬂuence the surface water quality of the catchment studied. Moreover, the conceptual model should address how the eﬀects of spatial and temporal variability of all of these factors would inﬂuence the outcomes. Fortunately, groundwater discharge to a stream is largely determined by mixing of water with short ﬂow paths and short residence times with water with long ﬂow paths and long residence times. This mixing concept has been described by many authors.2–5 Duﬀy and Lee5 demonstrated that small-scale spatial heterogeneity in hydraulic properties and solute input only marginally aﬀects the overall quality of outﬂowing water under most circumstances. The eﬀect of mixing of water with diﬀerent travel times strongly dominates the outﬂow concentrations. The system can therefore be described by average inputs of solutes to groundwater, and by average hydraulic properties. In fact, the system behaves similarly to an ideally mixed reservoir.3 The concentration response for a step input of a conservatively transported solute is described as Ct ¼ Cinput þ ðCinitial Cinput Þeðt=TÞ

ð1Þ

where C ¼ Cinitial for t ¼ 0 and C ¼ Cinput for t Z 0, and T¼

eD N

ð2Þ

For Dutch circumstances, the value of the characteristic or average travel time T [yr] is about equal to the thickness of the aquifer D [m], assuming an average groundwater recharge rate N of 0.3 m yr1 and a porosity e of 0.3.3 Therefore, the thickness of the aquifer and the corresponding groundwater system is one of the dominating factors for the concentration response in corresponding surface waters.

632

Chapter 10.2

This system is completely controlled by piston displacement along the ﬂow system which is controlled by the groundwater recharge ﬂux, the aquifer structure and the discharge points. For a conservative solute the contamination pattern and breakthrough are completely described by the isochrone pattern, which indicates an increase of groundwater age with depth (Figure 10.2.1). Figure 10.2.1 shows that the concentration at some time t is calculated by integrating concentrations over the depth of the aquifer that contributes to the stream. In this example, the horizontal ﬂuxes are constant over the aquifer depth according to the Dupuit–Forchheimer assumption. In more realistic situations, the integration of concentrations should be done proportionally to the ﬂuxes calculated by the ﬂow model. Interestingly, this concept equally applies for groundwater discharge in drinking water well ﬁelds and for groundwater discharge to surface water. Therefore, modelling concepts developed for predicting concentration response

Figure 10.2.1

Concentration distribution in an aquifer with constant transmissivity for a step input of C ¼ 1 in recharging groundwater at t ¼ 0.5T, t ¼ T and t ¼ 2T, and resulting travel time distribution and concentration breakthrough in the stream. Concentration at some time t is fully described by integrating concentrations over the depth of the aquifer, which is equivalent to a mixing process. Note that the shapes of the concentration breakthrough curve and the cumulative travel time curve are identical.

Modelling Reactive Transport of Diﬀuse Contaminants

633

in drinking water abstractions are also useful for groundwater–surface water interaction. Using this mixing approach of integrating groundwater concentrations over depth, we can now frame a conceptual model of dealing with the travel time distribution, with reactive processes and with the diﬀuse inputs of solutes. This new approach deals with the use of average concentration–depth proﬁles to identify important processes and to calibrate and verify groundwater transport models. Using numerical or analytical models we can estimate the travel time distribution of groundwater discharge into the surface water system and combine them with spatially averaged inputs to the groundwater for diﬀerent land use categories and for diﬀerent geohydrological situations within the catchment. Information on transformation or retardation processes should be included into these models to eventually calculate a predicted average concentration–depth proﬁle for combinations of land use and geohydrological situation. These predicted concentration–depth proﬁles should then be compared with measured concentration–depth proﬁles in a dedicated monitoring network, and adaptation to the model can be done subsequently to improve predictive behaviour. And conversely, the model can be used to design an appropriate monitoring network which enables good estimation of averaged concentration–depth proﬁles, in order to subsequently improve model predictions and reduce model uncertainty. The hypothesis is that we have good predictions of groundwater outﬂow to the surface water system when we are able to make good predictions of the average concentration–depth proﬁles. In real-world catchments, ﬂuxes from diﬀerent depths and related travel times in the groundwater system might diﬀer considerably compared with the idealistic aquifer described above and may vary with hydrological conditions7 (Figure 10.2.2). Consequently, a detailed approach for determining travel times and ﬂuxes of diﬀerent parts of the groundwater ﬂow system is a prerequisite for a sensible modelling approach, and the non-stationary character of the drainage system must be incorporated. This means that individual water courses, ditches and drains should preferably be included in the models to obtain a realistic travel time distribution for the complete catchment.

Figure 10.2.2

Fluxes through diﬀerent parts of the groundwater system under diﬀerent hydrological conditions from base ﬂow to quick ﬂow. Rising water tables under wet conditions favour shallow ﬂow towards ditches and drains, resulting in drainage from more contaminated parts of the aquifer.

634

Chapter 10.2

Additionally, a sensible approach needs input data on aquifer properties and structure, on solute inputs, on relevant biogeochemical processes and on relevant biogeochemical reactivity of the subsurface. However, for these model parameters one might use spatial aggregations, such as averaged retardation factors and averaged solute inputs, because mixing at the discharge points smoothes out heterogeneous inputs and aquifer properties.5 Many diﬀerent model codes and modelling approaches are available to be used within such a conceptual model framework. These include distributed physically based reactive transport models (e.g. Mike-SHE8,9), fully integrated groundwater–surface water models (Hydrogeosphere10) and lumped concentration models such as the multi-compartment models SWATMOD11 and SWAP12,13 and stream tube-based models.14,15 Each of these models has speciﬁc advantages, but in principle all of them are capable of dealing with the kind of conceptual model framed in this chapter. Here, we present an application of the approach using the physically based reactive transport model for the Kempen area in the southern part of the Netherlands, which is based on a combination of HYDRUS-1D,16 MODFLOW17 and MT3DMS.18 Veriﬁcation of the model was done comparing actual and predicted average concentration–depth proﬁles for areas with a speciﬁc land use–geohydrology combination.

10.2.3

Building a Regional Model for the Kempen Area

The aim of the Kempen study was to study geochemical and hydrological controls on subsurface transport of the trace metals cadmium and zinc using a reactive transport model at a regional scale. The soils of the Kempen region in the Netherlands are severely contaminated by atmospheric emissions from three zinc smelters. The Kempen model was meant to assess: (1) temporally and spatially variable leaching of metals from the unsaturated zone to groundwater; (2) temporal trends in cadmium and zinc concentrations in shallow and deeper groundwater; and (3) the metal loads of the surface water drainage network in relation to the hydrology and geology. A coupled approach for unsaturated and saturated zone ﬂow and transport was used for three diﬀerent catchments in the immediate vicinity of the smelters (Figure 10.2.3). The model contains multiple HYDRUS-1D models which calculate leaching of heavy metals to groundwater spatially and temporally, and a three-dimensional reactive transport model to simulate transport through the saturated zone (Figure 10.2.4). Details of the model set-up are given by Van der Grift and Griﬃoen.19 Inputs to the model were groundwater recharge, hydraulic properties of the unsaturated and saturated zone, water levels of the drainage system, sorption characteristics of typical soil proﬁles, average geochemical properties of geological formations and average hydrogeochemical characteristics of groundwater in these formations. A complete description of the transport model is given by Van der Grift and Griﬃoen.19

635

Modelling Reactive Transport of Diﬀuse Contaminants

Figure 10.2.3

Map showing the location of the Kempen area in the Netherlands, the six studied catchments, the locations of the zinc smelters and the interpolated zinc contents in the topsoil for 1995.

Input Budgets • Historic atmospheric deposition • Diffuse agricultural sources Ground level parameterization • water levels • soil map (flow and • transport properties)

multi HYDRUS-1D -Unsaturated flow -transport

Unsaturated zone

parametrization • geohydrology • geochemistry • hydrochemistry

MODFLOW/MT3D -3D-saturated flow -3D transport

Saturated zone

Flow results • water levels • water volumes • pathlines

Figure 10.2.4

Transport results • concentration profiles • mass fluxes • breakthrough curves

Structure of the coupled model used for the Kempen model.

636

Chapter 10.2

To test whether the mixing concept is sensible for the Kempen catchments, we used a step input for a conservative solute to derive the travel time frequency distribution and compared it with an analytical approximation based on eqn. (1). The step input includes an input concentration to groundwater of 100 from t ¼ 1950 until t ¼ inﬁnite. Figure 10.2.5 shows the concentration response for two catchments in the Kempen area and an analytical approximation with average travel time T ¼ 40 years. The overall shapes of the analytical and numerical concentration response are similar, indicating that mixing of groundwater with a large range of travel times indeed determines the groundwater contribution to surface water quality. However, a somewhat larger percentage of young water (o10 years old) is present in the catchments, and some extra tailing of old water is observed, especially in the Run catchment. The higher percentage of young water is probably related to the presence of intensive drainage systems which favour relatively fast discharge of the upper part of the groundwater in the catchment. Overall, the graphs indicate an overall average travel time of about 40 years for a conservative solute. Cadmium and zinc are typically sorbing solutes, and the transport model predicts a corresponding slower response of the groundwater contribution to surface water. Figure 10.2.6 depicts the development of the mass of zinc in diﬀerent compartments of the subsurface system. Note that the zinc deposition strongly declined after 1975 because of new production processes. The model predicts decreasing amounts of zinc in the soil compartment since 1975 and increasing amounts absorbed and dissolved in the groundwater compartment as well as discharge to the surface water. Apparently, zinc has been leaching from the soil compartment to the groundwater and surface water compartments. More detailed model results are described by Van der Grift and Griﬃoen.19

Figure 10.2.5

Concentration response for the Run and Tongelreep catchments in the Kempen area, compared with the analytical approximation for characteristic time T ¼ 40 years according to eqn. (1).

637

Modelling Reactive Transport of Diﬀuse Contaminants

10000 9000 8000

Zn (1000 kg)

7000 6000 5000 4000 3000 1960 2000 2000 1000 2050

0 soil

Figure 10.2.6

10.2.4

saturated zone: saturated adsorbed zone: dissolved

surface water seepage

Total mass of zinc in the entire Kempen area in the soil compartment, adsorbed in the saturated zone, dissolved in groundwater and discharged in the surface water system for the period 1960–2100.

Verifying the Model: Setting up a Customised Monitoring System

Large spatial variability is present in leaching of cadmium and zinc to the groundwater, even within areas with similar combinations of land use, soil types and geohydrological situation. Overall the model predicted strong leaching in forested areas with soils low in pH and organic carbon (Figure 10.2.7 (left panel)). But a large variability was observed within such well-deﬁned land use–geohydrology combinations, both in the model predictions as well in the observation wells. The large temporal and spatial variability of contamination pattern in the subsurface probably results from small-scale heterogeneities in groundwater ﬂow paths, from the spatially variable reactivity of the subsurface and from spatial and temporal variability of the contaminant leaching from agricultural lands and forests. Calibrating these kinds of transport models is a diﬃcult task. Point by point calibration is practicably impossible, because of the large inﬂuence of coincidence on the individual point level. However, calibration at point level is not a goal when the main aim is to describe and predict the groundwater contribution

638

Chapter 10.2 average Zn(µg/l) 0.0

0

400

average Zn (µg/l)

800 1200 1600 2000 2400

0.0

0

400

800

1200

1600

N=77 5.0

5.0

10.0

10.0

15.0 20.0

nature-recharge nature-intermediate

depth (m)

depth (m)

N=23

20.0

agriculture-recharge

25.0

agriculture-intermediate

N=9

15.0

25.0

N=25

model: weighted average of the 5 land use hydrogeology combinations

discharge

30.0

Figure 10.2.7

30.0

Modelled average concentration–depth proﬁles for zinc for ﬁve combinations of land use and geohydrological situation (left), and comparison of measured concentrations with the modelled weighted average concentration–depth proﬁle (right).

to surface water contamination, because mixing of water with diﬀerent travel times from diﬀerent depths of the groundwater systems dominates the concentration response. Instead, we analysed the measured data and the model results in an aggregated way, using average concentrations and average concentration–depth proﬁles for ﬁve spatial combinations of land use and geohydrological situation. The hypothesis is that we have good predictions of outﬂow to the groundwater system when we are able to make good predictions of the average concentration–depth proﬁles, assuming that the groundwater contribution is well quantiﬁed by ﬂux-proportional integration of concentrations over the depth of the contributing groundwater system. Concentration–depth proﬁles typically include all the important features of the conceptual model—travel time, solute inputs and transformation and retardation processes—and can be measured in a dedicated groundwater monitoring network.6,20,21 Concretely, we assessed the average concentration–depth proﬁles for the ﬁve combinations nature–recharge, agriculture–recharge, nature–intermediate, agriculture–intermediate and discharge areas (Figure 10.2.7 (left panel)). Distinct diﬀerences were present between these ﬁve combinations, with the deepest inﬁltration and highest concentrations in nature–recharge areas. This reﬂects the vulnerable soils and high atmospheric inputs in those areas. The sample size within the existing monitoring network did not allow for a comparison of measured and predicted concentration–depth proﬁles for the individual combinations of land use and geohydrological situation. Therefore, the concentrations were averaged over the whole model area, weighting for the spatial extent of the ﬁve land use–geohydrology combinations. Subsequently, a comparison was made between these spatially weighted average concentration– depth proﬁles and weighted average measured concentrations in the model area

639

Modelling Reactive Transport of Diﬀuse Contaminants

(Figure 10.2.7 (right panel)). The model results showed a good correspondence with the measured concentration–depth proﬁles for all but the depth interval of 10–20 m. Here, far too few observations (N ¼ 9) were present to acquire a reliable estimate of the actual concentrations. Overall, conﬁdence was gained in the model performance based on the limited amount of available monitoring data. As described before, the sample size within the existing monitoring network does certainly not allow for a diﬀerentiation between the ﬁve land use– geohydrology combinations. Such a diﬀerentiation was considered necessary to further improve conﬁdence in the model performance, before starting to evaluate future scenarios of implementing policy measures aimed at reducing the environmental impact of the historical pollution. The model was now used to design a monitoring network which should enable the diﬀerentiation. We used 1000 realisations of possible network conﬁgurations with 3, 5, 10, 15, 20, 30 and 50 multi-level observation wells, by randomly selecting 1000 model cells within the ﬁve land use–geohydrology combinations at speciﬁed depths of 0.5, 1.5, 3.5, 5.5, 7.5, 9.5 and 11.5 m. From these data the average concentration–depth proﬁle and the 95% conﬁdence interval around the average proﬁle were derived. Figure 10.2.8 shows the resulting concentration–depth proﬁles for monitoring networks with a sample size of 5, 10 and 20 multi-level observation wells for nature–recharge areas. The result strongly suggests that a sample size larger than 10 wells is necessary to have a reasonable conﬁdence

5 multi-level wells 2000 4000

0

20 multi-level wells 0 2000 4000

10 multi-level wells 2000 4000

0

0

-2

-2

-2

-4

-4

-4

-6

-6

depth (m)

0

depth (m)

depth (m)

0

-6

-8

-8

-8

-10

-10

-10

-12

-12

-12

Figure 10.2.8

Eﬀect of sample size on the uncertainty of the measured average concentration–depth proﬁle for zinc under nature in recharge areas, based on 1000 realisations of monitoring networks with 5, 10 and 20 observation wells. The probability of an average concentration outside the range of the error bars is less than 5%.

640

Chapter 10.2

that the measured average concentration–depth proﬁle resembles the actual average proﬁle. Even with 20 monitoring wells the resulting average concentration at 1.5 m depth might fall between 1000 and 3000 mg l1, given a model average of 2000 mg l1. Based on these results a minimum sample size of 15–20 wells per land use–geohydrology combination was advised as a ﬁrst step in the development of the monitoring network. Once these monitoring data become available, a re-evaluation of the model performance is foreseen.

10.2.5

Predictions of Groundwater Contributions to Surface Water Quality

The Kempen model was used to predict the groundwater contribution to the surface water system. The model predicts that the historical pollution loads of zinc continue to contribute to the contamination of surface waters far into the future. Given the travel time distribution shown in Figure 10.2.5, the ﬂux towards surface water of a conservatively transported solute would have shown a strong decrease when concentration inputs stopped. This is shown for a 20-year block input of a ﬁctitious conservative solute, which indeed shows a concentration decrease directly after stopping the inputs after 1970 (Figure 10.2.9 (left panels)). However, the groundwater ﬂux of zinc into the surface

Figure 10.2.9

Modelled predicted breakthrough of a 20-year conservative block front of Cinput ¼ 1000 mg l1 in the Run and Tongelreep catchments (left) and retarded breakthrough of zinc in the two catchments (right) for the period 1950–2050.

Modelling Reactive Transport of Diﬀuse Contaminants

641

water system was predicted to further increase for the coming decades (Figure 10.2.9 (right panels)). This is due to combined eﬀect of (1) the slow and spatially variable movement of the zinc front through the unsaturated zone and (2) the sorptive behaviour of zinc during transport in the saturated zone. Here, the slow response of the groundwater system is further ampliﬁed by the retarded transport of zinc.

10.2.5.1

Discussion

The extent of groundwater–surface water interaction is still one of the most important issues in implementing the new EU water regulations. We presented a conceptual model for the transport of solutes in groundwater in the Dutch Kempen area, and showed how it was implemented in a numerical reactive transport model. The conceptual model emphasises the need to obtain information about the actual concentration–depth proﬁles of the relevant solutes. Conﬁdence in the model was gained through the comparison of calculated and measured spatially averaged concentration–depth proﬁles. Conversely, heterogeneous transport calculated with the model was used to deﬁne an eﬀective monitoring scheme which eventually might help to reduce the uncertainty of model performance. We believe that similar feedback mechanisms with modelling and monitoring phases eventually will increase system understanding and will help to gain cost-eﬃciency in water management. The presented concentration–depth approach seems to be eﬃcient for transport modelling in other areas with similar hydrogeological conditions, including granular aquifers of moderate to high permeability. Typically, these areas are important agricultural environments where groundwater contamination from diﬀuse sources often is a serious problem. The approach presented is well suited for metals, nutrients and pesticides, although diﬀerent types of transport models and diﬀerent depth ranges might be relevant compared with the presented example. The presented approach emphasises the long-term inﬂuence of groundwater quality on surface water. Given the long average residence times associated with groundwater ﬂow, this is an acceptable approach for many cases. However, short-term ﬂuctuations in surface water quality might also be signiﬁcant, especially when ﬂuxes from shallow groundwater increase in wet periods7 (see also Figure 10.2.2). These eﬀects have not been considered in this chapter and require further attention when implementing the Water Framework Directive in Europe. Knowledge about the shallow part of the concentration–depth proﬁle in groundwater should be included when studying water quality ﬂuctuations at short time scales.7 Uncertainty in the Kempen example was mainly caused by the very heterogeneous leaching of zinc from the unsaturated zone. In other areas, uncertainty in groundwater and surface water quality might also be attributed to heterogeneous inputs of contaminants, to hydraulic heterogeneities that determine groundwater travel times or by heterogeneous reactivity of the saturated zone. These heterogeneities dampen out in surface water quality because of

642

Chapter 10.2

mixing of groundwater with a large range of travel times. However, these heterogeneities should be included when designing monitoring networks, because they will result in large spatial variability of concentrations measured in observation wells which typically consist of short screen lengths and acquire a local sample of groundwater quality.6 Uncertainty in model performance can further be reduced by age dating (see Chapter 5.3). Age dating helps to diﬀerentiate between the eﬀects of groundwater age and the eﬀects of reactive processes.21–23 The model results that we presented refer to the groundwater contribution to the surface water system as a whole. We did not consider transformation or precipitation processes on the groundwater–surface water interface, nor processes and transport related to stream sediments, nor processes in the surface water itself. Overall, the predicted zinc ﬂuxes in surface water were of the same order of magnitude as measured in the associated surface waters (not presented), but this is no guarantee for the validity of the results and further research on the possible attenuating eﬀect of the groundwater–surface water interface is necessary.

10.2.5.2

Policy Aspects

Until 2000, the focus in groundwater protection policy predominantly was to protect deep groundwater resources as a potential stock to secure drinking water supply. In many EU member states, groundwater and surface water professionals lived in separate scientiﬁc and administrative worlds and did not meet frequently to discuss common water quality problems. For example, in the Netherlands surface water management was in the hands of water boards and groundwater protection was the responsibility of provincial authorities. The Water Framework Directive1 involved a new paradigm to most professionals working in the ﬁeld of groundwater protection. The directive deﬁnes a new additional scope of groundwater protection, looking at groundwater from a surface water and ecosystems perspective. This new emphasis is also included in the new Groundwater Directive which will be enacted in 2007 (see Chapter 3.1). The new scope poses the question as to how groundwater quality and surface water quality are interconnected. Consequently, there is an increasing need to develop new conceptual models emphasising the interaction between groundwater and surface water, to develop customised numerical predictive tools and to design and adapt monitoring networks. The new challenge is to ﬁnd a systems approach including soil, groundwater, surface water and associated marine water at diﬀerent spatial and temporal scales. The conceptual model presented in this chapter is just one example of how to connect the groundwater and surface water worlds and further research focusing on the connection is certainly required. One of the concrete questions is how to deﬁne groundwater threshold values to verify compliance with the good chemical status requirements of the Water Framework Directive and Groundwater Directive. The example presented in this chapter clearly shows that zinc in groundwater is an actual threat to surface

Modelling Reactive Transport of Diﬀuse Contaminants

643

water in the Kempen area, although elevated zinc concentrations only appear in the upper 10 m of the subsurface. Clearly, the deﬁnition of threshold values for groundwater—which are meant to protect surface water—should be tuned on measurements and background concentrations in the upper part of the groundwater body. The example shows that diﬀerent depths should be considered when establishing threshold values for diﬀerent solutes based on their potential penetration in the aquifer, which is controlled by groundwater residence times and by reactive processes which cause transformation and retardation.

10.2.6

Conclusions

Reactive transport models are potentially very useful within the context of the European Water Framework Directive.1 Basic requirements for eﬀective modelling are the establishment of a clear conceptual idea about system behaviour and the presence of groundwater and surface water quality data to verify model performance. We presented a conceptual model for the transport of metals in groundwater in the Dutch Kempen area, and showed how it was implemented in a numerical reactive transport model. The conceptual model emphasises the need to obtain information about the actual concentration– depth proﬁles of the relevant solutes. Model predictions indicate that the historical pollution loads continue to contribute to the contamination of surface waters far into the future, especially for solutes such as trace metals which show signiﬁcant retardation in the unsaturated and the saturated zones. Conﬁdence in the model was gained through the comparison of calculated and measured spatially averaged concentration–depth proﬁles. Conversely, heterogeneous transport calculated with the model was used to deﬁne an eﬀective monitoring scheme which reduces the uncertainty of model performance. We believe that similar kinds of combinations of modelling and monitoring ultimately will form the basis for eﬀective evaluation of the eﬀect of measures implemented in river basin management plans.

Acknowledgement This work was supported by the European Union FP6 Integrated Project AquaTerra (project no. GOCE 505428) under the thematic priority ‘‘sustainable development, global change and ecosystems’’.

References 1. EU Directive 2000/60/EU of the European Parliament and of the Council of 23 October 2000, Oﬃcial Journal of the European Communities, L327. 2. M. Eldor and G. Dagan, Water Resour. Res, 1972, 8(5), 1316–1331. 3. H. C. van Ommen, J. Hydrol., 1986, 88, 79–95. 4. P. A. C. Raats, Agric. Water Manag., 1981, 4, 63–82.

644

Chapter 10.2

5. C. J. Duﬀy and D. H. Lee, Water Resour. Res., 1992, 23, 2076–2090. 6. H. P. Broers and F. C. van Geer, Groundwater, 2005, 43(6), 850–862. 7. J. Rozemeijer and H. P. Broers, Environ. Pollut., 2007, doi: 10:1016/j. envpol.2007.01.028. 8. Danish Hydraulic Institute, MIKE SHE Pre- and Postprocessing User Manual, Denmark, 1999. 9. Danish Hydraulic Institute, MIKE SHE Water Movement User Manual, Denmark, 1999. 10. R. Therrien, R. G. McLaren, E. A. Sudicky and S. M. Panday, HydroGeoSphere: A Three-dimensional Numerical Model Describing Fully Integrated Subsurface and Surface Flow and Solute Transport, Manual, HydroGeoLogic Inc., Herndon, VA, 2004. 11. M. A. Sophocleous, J. K. Koelliker, R. S. Govindaraju, T. Birdie, S. R. Ramireddygari and S. P. Perkins, J. Hydrol., 1999, 214, 179–196. 12. J. C. van Dam, Field-scale water ﬂow and solute transport. SWAP model concepts, parameter estimation, and case studies, PhD thesis, Wageningen University, 2000. 13. J. C. van Dam, J. Huygen, J. G. Wesseling, R. A. Feddes, P. Kabat, P. E. V. van Walsum, P. Groenendijk and C. A. van Diepen, SWAP version 2.0, theory. Simulation of water ﬂow, solute transport and plant growth in the soil–water–air–plant environment, Technical Document 45, DLO Winand Staring Centre, Wageningen, Report 71, Department Water Resources, Wageningen Agricultural University, 1997. 14. Netherlands Technical Committee on Soil Protection (TCB), Advise on additional groundwater protection against diﬀuse sources of soil contamination, Leidschendam, The Netherlands, 1989. 15. M. R. Thiele, S. E. Rao and M. J. Blunt, Math. Geol., 1996, 28(7), 843–856. 16. J. Sˇimu˚nek, M. Sˇejna and M. Th. van Genuchten, The Hydrus-1D software package for simulating the one-dimensional movement of water, heat and multiple solutes in variable-saturated media, Version 2, US Salinity Laboratory, 1998. 17. M. G. McDonald and A. W. Harbaugh, A modular three-dimensional ﬁnite-diﬀerence groundwater ﬂow model, USGS Tech. Water-Res. Inv., 1998, Book 6, ch. A1. 18. C. Zheng and P. P. Wang, MT3DMS: a modular three-dimensional multispecies transport model for simulation of advection, dispersion, and chemical reactions of contaminants in groundwater systems, Documentation and user’s guide, Department of Geological Sciences, University of Alabama, 1999. 19. B. van der Grift and J. Griﬃoen, J. Contamin. Hydrol., submitted. 20. H. P. Broers, Strategies for regional groundwater quality monitoring, PhD thesis, Utrecht University, 2002. 21. H. P. Broers and B. van der Grift, J. Hydrol., 2004, 296, 192–220. 22. H. P. Broers, J. Hydrol, 2004, 299, 84–106. 23. A. Visser, H. P. Broers and M. F. P. Bierkens, Environ. Pollut., 2007, doi: 10:1016/j.envpol.2007.01.027.

11. Conclusions: Further Policy and Research Needs

CHAPTER 11.1

SNOWMAN: An Alternative for Transnational Research Funding JO¨RG FRAUENSTEINa AND H. JOHAN VAN VEENb a

Federal Environmental Agency Dessau, Unit II 4.3 Terrestrial Ecology, Land Management, Regional Protection Concepts, Wo¨rlitzer Platz 1, 06844 Dessau, P.O. Box 1406, DE-06813 Dessau, Germany; b SKB, Bu¨chnerweg 1, Postbus 420, NL-2800 AK Gouda, The Netherlands

11.1.1

Introduction

11.1.1.1

Transnational Research

Over the last few decades an enormous part of the European Union (EU) budget has been spent on research programmes such as environmental research. To date a huge number of research projects on national and European level have been successfully implemented, but did they achieve the best beneﬁt as might be? Appreciable bottlenecks still exist in general within precise focusing on customer needs, dissemination of approaches, results, project outputs and last but not least in a consequent initiation and implementation of synergy eﬀects spending human and ﬁnancial resources. To put more eﬀort into a real European research area, the Sixth Framework Programme (FP6) initiated a signiﬁcant step forward to improve the coordination and coherence of European research.1 The ERA-NET scheme as a major element of the FP6 speciﬁc programme ‘‘Integrating and strengthening the European research area’’ has been established to step up the cooperation and coordination of research activities carried out at national or regional levels in EU member states and associated states through: the networking of research activities conducted at national or regional level; and the mutual opening of national and regional research programmes. The scheme will enable national systems to take on tasks collectively that they would not have been able to tackle independently. 647

648

Chapter 11.1

Previous Framework Programmes have fostered cooperation of research actors at ‘‘project’’ level by bringing together universities, research agencies and companies. The European research landscape as it is today would be very diﬀerent without this intensive cooperation between researchers. Under FP6, the ﬁrst steps have been taken to progress on coordination and cooperation at programme level by networking national research programmes and bringing together managers of ministries and funding agencies. It also aims at establishing long-term cooperation between national programmes, ultimately leading to joint transnational research programmes. The EU is providing targeted support to this process now through the ERANET scheme for the coordination of national and regional research programmes. The popularity, success and wide range of ERA-NET schemes already in operation testify to the great interest that exists for cooperation and provide excellent examples of how we can do better and more in research by pooling and rationalising resources. Over and above this, a closer coordination of national research programmes will enable a better use of the resources available and should pave the way for the emergence of a truly European research policy, integrating the positions of all EU member states and EU institutions in key areas of common interest. The ERA-NET scheme is based on a ‘‘bottom-up’’ approach: it is open to all areas of research, even those not speciﬁcally covered by the EU Framework Programmes. The initiative therefore lies with the member states. It espouses a step-by-step approach which allows the gradual development from exploratory actions to the establishment and implementation of joint research programmes. It is expected that 75 ERA-NET projects will be up and running by the end of FP6. The response has been very positive. The ﬁrst ERA-NET projects are already changing the landscape of European research in a wide range of ﬁelds. Some of them are in the process of launching the ﬁrst joint transnational programmes and joint calls. Given its considerable success with FP6, the European Commission is proposing to continue and to expand the ERA-NET scheme. Existing projects will be encouraged to go further and ERA-NET consortia could be enlarged to include new partners. The participation of national and regional ministries, who are in the process of preparing research programmes in a particular ﬁeld, will also be encouraged to participate in the ERA-NET scheme, even if their programme is not fully developed at the time of submitting a proposal. The aim is to encourage countries, in particular the new and future member states, to participate actively in the development of the ERA by joining ERA-NET projects. This should help stimulate the exchange of information and foster potential future transnational programme cooperation with these countries. In addition, a new element of the ERA-NET scheme is being proposed under FP7. Called the ERA-NET ‘‘Plus’’ module, its objective is to encourage the pooling of funds in joint calls from national programmes on a call-by-call basis. The Commission will support this pooling of funds by providing a ‘‘top-up’’ contribution to transnational calls for proposals organised jointly by the

SNOWMAN: An Alternative for Transnational Research Funding

649

programmes participating in ERA-NET. Through this new mechanism, the Commission could provide up to an additional 25–30% of the total of national contributions.

11.1.1.2

Soil and Groundwater Management

European interest in research, policy and practice related to sustainable management of the soil and water system is increasing, especially due to the implementation of the Water Framework Directive and the development of an EU thematic strategy on soil protection.2 European soils are under pressure by past and present non-sustainable landuse practices. In the near future land-use will change in many areas of Europe due to social and economic driving forces and climate change. These changes will provide opportunities to alter the way we use our land into a more sustainable direction. Therefore a thorough understanding of the varying properties of the soil (and water) systems that determine the opportunities for more ecologically eﬃcient land use and sustainable utilisation of soil and water resources is essential for the future. Hence an analysis of processes related to eight threats to soil and their interdependency—erosion, loss of organic matter contamination, sealing, compaction, decline in biodiversity, salinisation, ﬂoods and landslides—was carried out by technical working groups to explore the future research need in the ﬁeld. The research agenda for soil protection in Europe describes general research needs for all soil threats (Table 11.1.1). It focuses on the research needs for the design of adequate resource management responses in the light of changes in land use and climate changes in the EU. Due to the regional diﬀerences among the EU member states, especially as regards climate problems, the priority setting is not the same. Consequently regional related approaches are required to combine special experiences with transferable problem solutions. The environment may also be considered as an important economic opportunity for Europe. An investment in new environmental technologies will strengthen Europe’s competitiveness, while an investment in the scientiﬁc basis of land and water resource management will maintain Europe’s environmentfriendly image. Focused long-term research and development eﬀorts are crucial to further develop the necessary innovations. The Vital Soil conference4 conﬁrmed again soil contamination as one of the priority threats to soil and its uses and functions for economy and society. It has been analysed and discussed extensively in the development of the EU thematic strategy for soil protection, using a framework to evaluate drivers, pressures, state, impacts and response. The experience gained from national soil contamination policies and from discussions in EU-funded networks like CARACAS, NICOLE and CLARINET led to the conclusion that the soil environment is too complex for classic command and control type of responses for individual soil problems or contaminants. The classic instruments like ‘‘Environmental Quality Standards’’ to set limits to further degradation of water and air quality and reverse negative trends towards quality improvement

650

Table 11.1.1

Chapter 11.1

Integrated research requirement: soil.a

Main research goals

Research clusters

Sciences involved

To understand the main processes in the ecosubsystem soil; induced by threats

Analysis of processes related to the 8 threats to soil and their interdependency: erosion, loss of organic matter, contamination, sealing, compaction, decline in biodiversity, salinisation, ﬂoods and landslides Development, harmonisation and standardisation of methods for the analysis of the State (S) of the 8 threats to soil and their changes with time ¼ soil monitoring in Europe Relating the 8 threats to Driving forces (D) and Pressures (P) ¼ cross linking with cultural, social and economic drivers, such as EU and other policies (agriculture, transport, energy, environment etc.) as well as with technical and ecological drivers, such as global and climate change Analysis of the Impacts (I) of the 8 threats, relating them to soil ecoservices for other environmental compartments: air, water (open and ground water), biomass production, human health, biodiversity, culture Development of strategies and operational procedures for the mitigation of the threats ¼ Responses (R)

Interdisciplinary research through cooperation of soil physics, soil chemistry, soil mineralogy and soil biology

To know where these processes occur and how they develop with time

To know the driving forces and pressures behind these processes, as related to policy and decision-making on a local and EU basis

To know the impacts on the eco-services provided by the subsystem soil to other environmental compartments (ecosubsystems)

To have strategies and operational tools (technologies) at one’s disposal for the mitigation of threats and impacts

a

W. E. H. Blum and J. Bu¨sing, 2004.3

Multidisciplinary research through cooperation of soil sciences with: geographical sciences, geo-statistics; geoinformation sciences (e.g. GIS) Multidisciplinary research through cooperation of soil sciences with political sciences, social sciences, economic sciences, historical sciences, philosophical sciences and others

Multidisciplinary research through cooperation of soil sciences with geological sciences, sedimentological sciences, hydrological sciences, physiogeographical sciences, biological sciences, toxicological sciences and others Multi-disciplinary research through co-operation of natural sciences with engineering sciences, technical sciences, physical sciences, mathematical sciences and others

SNOWMAN: An Alternative for Transnational Research Funding

651

do not work that well in soil. The numerous experiences with contaminated land have shown that ‘‘trend reversal’’ is very costly, because it requires the application of expensive technologies and civil engineering in the many cases that autonomous restoration of soil quality is too slow or not occurring at all. This has led to the general conclusion that an integrated management approach is needed, which considers the soil and the waters it contains as one system which interacts with other parts of the environment and with the socioeconomic world through uses and functions. Integration with spatial planning, which is starting to learn to address the soil as a threedimensional system, is of utmost importance for this system-oriented approach (Table 11.1.2).

11.1.1.3

Forerunning Projects and Signiﬁcant Scientiﬁc Input

The CLARINET working group ‘‘RTD Programmes’’ recommends taking steps towards establishing a coordinated European research policy on contaminated land and water management. Some major recommendations are the following.5 Providing a platform for research programme managers to exchange information on national research priorities, funding mechanisms and knowledge dissemination. There should be a more coherent integration of national and European research activities. This could be achieved through a closer collaboration between various scientiﬁc and technological research organisations in Europe. A joint approach to the needs and means of ﬁnancing large research projects in Europe. For example, European researchers and technology developers could test and compare their products at speciﬁc demonstration sites in Europe. Networking of existing centres of excellence and competence in Europe and the creation of virtual centres through the use of new interactive communication tools. Coordination of an agenda of joint research priorities and stimulation of transnational research and technology development (RTD) projects and European peer review of programmes. Stimulation of transdisciplinary research involving more stakeholders in the projects. More attention should be paid to the dissemination of knowledge in national programmes. The focus should be shifted from pure knowledge supply to ‘‘information on demand’’. During the communication process to establish the working basis and to identify the recent research needs regarding the European soil thematic strategy, a horizontal technical working group research carried out this for the topic on contamination based on forerunning projects and initiatives mentioned above.

Ecological, economic and social drivers of soil threats (Driving forces and pressures, ‘‘D’’, ‘‘P’’) Relating qualitatively and quantitatively the 8 threats to Driving Forces and Pressures Harmonisation of methodologies for the identiﬁcation and quantiﬁcation of potentially dangerous chemicals

Identiﬁcation and quantiﬁcation of social and economic driving forces on local and diﬀuse soil pollution and their impacts

Spatial and temporal changes of soil processes and parameters (State ‘‘S’’)

Development, harmonisation and standardisation of methods for the analysis of the state

Development of fast and cost eﬀective screening methods

Identiﬁcation and quantiﬁcation of new hazardous substances in soils

Analysis of processes related to the 8 threats to soil and their interdependency

Sources, fate and behaviour of pollutants

SNOWMAN research agenda extracted from TWG Research.a

Processes underlying soil functions and quality

Table 11.1.2

Development of operational procedures for the mitigation of the threats

Analysis of the Impacts of the 8 threats, relating them to soil ecoservices for other environmental compartments Improvement and harmonisation of concepts and models for the transport of contaminants in soil and their transfer to other environmental compartments (water, air, biomass) Development of concepts and models for the direct and indirect transfer of contaminants from soil to humans

Quantiﬁcation and improvement of natural rehabilitation processes

Improvement of soil functions, contributing to natural attenuation

Strategies and operational procedures for soil protection (Responses, ‘‘R’’)

Factors inﬂuencing soil eco-services (Impacts, ‘‘I’’)

652 Chapter 11.1

Improvement of risk assessment methodologies for remediation activities, with the ﬁnal aim of developing a ‘‘ﬁt-foruse’’ toolbox for riskmodelling, including the re-use of decontaminated soil Development of harmonised methods for deﬁning ‘‘tolerable’’ loading on soil and groundwater systems

Development of techniques, e.g. containment devices for safe storage, handling and transport of harmful substances Sustainability/persistence of remediation technologies and their environmental impacts Economic models for assessing the cost– beneﬁt relationship for cleaning-up methods of contaminated soils

Improvement of methods for alternative management options, taking into account environmental, social and economic conditions

Working Group on Research, Sealing and Cross-cutting Issues, Summary and Policy Recommendations on Research, draft ﬁnal report, European Commission, BU9 3/173, B-1049 Brussels, Belgium, March 2004 (Soil CIRCA e-library: http://forum.europa.eu.int/Public/irc/env/Home/main).

a

Mobility and availability of contaminants to other environmental compartments

Deﬁnition of indicators for the assessment of soil quality

Early warning systems for soil pollution, including bio indicators

SNOWMAN: An Alternative for Transnational Research Funding 653

654

Chapter 11.1

11.1.2

An ERA-NET Bridging the Gap Between National and European Research

In almost all EU countries soil and groundwater are accepted as an essential part of the earth that has to be protected. Groundwater has many functions which are evident for human society, so a healthy groundwater is essential for an eco-eﬃcient society. Due to numerous developments groundwater is under pressure from many threats such as pollution irrigation, loss of biodiversity and others. To manage these threats many countries and the EU have developed policies to achieve a sustainable use of soil, such as:

the biodiversity treaty of Rio de Janeiro (1992); the Kyoto climate treaty (1997); the revision of EU agriculture policy (1992, 2000, 2003); the EU Water Framework Directive (2000); the EU Groundwater Directive (2007); and the EU soil strategy, including the future Soil Framework Directive (2007?).

The development and implementation of these policies need a sound knowledge base. The knowledge on soil and groundwater in Europe is traditionally well developed, in relation to agriculture. However new developments create a number of new scientiﬁc challenges related to, for example, the system approach in river basins. Knowledge development by research will be therefore inextricable from the development of sustainable use of soils and groundwater. In Europe research and knowledge development on groundwater is executed through national programmes. Table 11.1.3 gives an overview of national research programmes in Europe related to soil and groundwater in 2005.

11.1.3

The Way Forward

11.1.3.1

The Meaning of Cooperation

There seems to be many constraints in realising good cooperation between national RTD programmes. Nationalism, independency, diﬀerent procedure, diﬀerent responsibilities and diﬀerent mandates do not make it easy. So there must be a strong motivation to overcome them. Each individual country’s programme has to have a perceived beneﬁt to it from cooperation. The beneﬁts the members of the SNOWMAN consortium see in this European cooperation can be summarised in four diﬀerent groups. Inﬂuence: increasing inﬂuence on European soil policy and European R&D budgets for soil research. The priority of soil research in the European research agenda may also inﬂuence individual member states’ priorities. There is an expectation that, within several years, a European research council will be established. Good cooperation between national R&D programmes could increase the inﬂuence in the future research council. By acting together public conﬁdence in research can be strengthened.

Environmental funding law AMINAL water-sediment research programme IWT research programme OVAM Contaminated land programme VMM water-sediment research programme RPF Framework Programme for research and technological development ADEME programme on contaminated sites RITEAU GISFI KORA REFINA Research for the environment RUBIN (permeable reactive barriers) SIWAP (seepage water prediction) UFOPLAN: environmental research plan, subheading F2 soil protection and contamination Leven met water/living with water SKB Concept and principles of the system providing for functional utilisation of contaminated sites (subpart of the strategic governmental programme ‘Environment and Health’) Governmental programme for post-industrial areas MENER environment no. 8211; Energy: Resources

Austria Belgium Belgium Belgium Belgium Cyprus

Romania

Poland

The Netherlands The Netherlands Poland

France France France Germany Germany Germany Germany Germany Germany

Program name going going going going going

5

6

6 6 3

On going 4 6 5 5 6 5 4 1

On On On On On 3

Duration (years)

20–50

No information

40 25–70 79

10–50 53 50–100 100 100 100 100 100 100

50–100 100 25–100 100 100 75

Funding rate (%)

0.500

0.003

No information 2.917 0.052

1.300 3.750 0.305 3.600 4.000 No information 1.000 2.750 1.000

0.500 No information 0.200 No information No information No information

Annual budget (h million/year)

Overview of national research programmes in Europe related to soil and groundwater in in 2005.a

Country

Table 11.1.3

SNOWMAN: An Alternative for Transnational Research Funding 655

Formas programme Ha˚llbar Sanering (sustainable remediation) TUBITAK, MAG, research committee Bioremediation link programme ‘‘Infrastructure and Environment’’ programme and ‘Engineering’ programme Soil protection Soil sustainability URGENT (urban regeneration and the environment) Water quality and catchments management Land contamination

Sweden Sweden Turkey UK UK

a

10 4

On going 14 7

On going 5 1 7 On going

Duration (years)

‘‘A SNOWMAN’s navigator through research funding programmes across Europe’’, 2006.

UK UK Total

UK UK UK

Program name

(continued )

Country

Table 11.1.3

100 100

100 100 No information

50-100 No information 75 50 100

Funding rate (%)

0.183 2.200 38.043

No information 0.183 0.500

0.250 1.000 0.750 1.100 10.000

Annual budget (h million/year)

656 Chapter 11.1

SNOWMAN: An Alternative for Transnational Research Funding

657

Eﬀectiveness is seen as the degree of application of research results into practice. It is a measure of the take-up of research results into practice and the contribution to soil quality management in practice. All national programmes see an increase of eﬀectiveness as an important criterion for success. Dissemination of know-how and the economic and public added value of R&D results are important factors. The implementation of know-how in the new member states is another important criterion. Exchange of experience about methods for dissemination of know-how and improving added value is a strong reason for cooperation. Eﬃciency is the execution of R&D programmes for a minimum of costs. Through good cooperation overlaps between programmes of individual countries can be avoided and thus cost-eﬃciency increased. The national programmes could focus on diﬀerent (but complementary) research themes. Each programme could beneﬁt from the results of other programmes and avoid ‘‘reinventing wheels’’. Finance and funding: cooperation creates the possibility of co-ﬁnancing projects and increases the possibilities for research funding, whilst increasing ﬂexibility. The members of the SNOWMAN project have set goals for cooperation between the national research programmes. The goals can be divided into short-, mid- or long-term goals. These goals can also be related to the groups of beneﬁts as mentioned before.

11.1.3.2

A Stepwise Approach Towards Cooperation

The complete cycle of research and development of know-how, the innovation process, starts with the formulation of research needs and ends in the practical evaluation of applied new technologies and feedback of the evaluation results to the end-user. The major phases in this process are:

funding; programming; implementation and evaluation; and networking.

To reach the ﬁnal goal of cooperation and to manage and control the development of cooperation, a stepwise approach seems to be appropriate. The ﬁrst step of development starts with the existing networks and their involvement in the relevant steps of the R&D-process. The intensity of cooperation can be increased in short-, mid- and long-term steps (Table 11.1.4). As the cooperation increases, the development of the networks will increase too. This development of networks is not an activity in itself, but a result of the improvement of the R&D process. Therefore, this development is described in the three steps for each phase in the R&D process.

a

Increasing use of R&D results

Decrease bureaucracy in national and international programs Avoid overlap of national programs Improve exchange of know how between national programs Create a common forum for reviewing proposals Any national program is open for international submission of proposals One European address to submit proposals Keep researchers in Europe

‘‘Working together in research and development for sustainable land management in Europe: the vision of SNOWMAN’’, 2006.

Becoming the main source of soil and groundwater quality knowledge

Reaching a common vision on the research agenda on soil matters for FP7 and so on

Long term

Integrate social and economical sciences

Coherence in European, national and regional R&D programs

Mid term

Sharing procedures of funders Creating a data base of evaluators and reviewers

Improve involvement of end users in R&D. Stimulate use of best practices. Start up of common dissemination plans Improving involvement of industry in R&D

Eﬃciency

Eﬀectiveness

Opening the information on national research programs Sharing a common R&D agenda between the SNOWMAN partners

Inﬂuence

Goals for cooperation.a

Short term

Table 11.1.4

At least one speciﬁc contamination research fund per country

More research funders

Open more funding options for national researchers

Bundle, focusing national budgets

Finance and funding

658 Chapter 11.1

SNOWMAN: An Alternative for Transnational Research Funding

11.1.3.2.1

659

Funding

Three models have been identiﬁed that can be applied in the cooperation of funding programmes for fundamental and strategic research. The increase in cooperation in funding programmes can follow these three models. Step 1: national contributions. Each country funds its own national contributions to the projects of a coordinated call or programme. Special arrangements have to be made between the country which is responsible for the management of the project (leading partner) and the countries that are responsible for the ﬁnancial control. A diﬀerence can be made between the funding of the general activities of the coordinated call of programmes and the research projects themselves. General activities such as preparation of the call, selection and dissemination can be funded out of a common pot or by an additional funding organisation such as the EC. Funding national contributions is probably the most realistic model for the short term because national procedures are applicable and no money is transferred to researchers from other countries. Step 2: adoption of projects. Each participating country adopts (a) selected project(s) that ﬁt best to their national programmes. The country that adopts the project is responsible for the ﬁnancial control of that project(s). One of the participating countries has to be responsible for the management and ﬁnancial control of general activities of the coordinated call which can be funded separately as mentioned under (a). Step 3: common pot. Each participating country contributes to a common pot. Projects are ﬁnanced out of the common pot. Financial control of the projects and the programme is carried out by one of the participating countries or an independent third party. All participating countries have to agree on the procedures of the ﬁnancial control and have trust in the controlling organisation. The balance between the contribution to and expenditures per participating country out of the common pot can be taken into account. The responsibility of the ﬁnancial control of the coordinated call is a shared common responsibility.

11.1.3.2.2

Programming

In the following paragraphs a description is given of the stepwise development in the cooperation in the programming of R&D, divided into the development of the research agenda, the organisation of the programming process and the selection of R&D projects. The three steps of the research agenda are the following. Step 1: common vision. A common vision about the research agenda has been established in discussions with national programme funders and scientists in project meetings and a think tank meeting. It is mainly an expert view on the R&D agenda. This vision is the basis for the funding of a coordinated call.

660

Chapter 11.1

Step 2: coherence in European, national and regional R&D programmes. The funding of a coordinated R&D programme funded by the participating countries out of their national budgets will enforce the participating countries to create more coherence between European and national programmes. The funding of the coordinated programme must also contribute to the national programme. In this way funders will have a beneﬁt in coherence with national programmes. To create such coherence funders will have to discuss with national policy-makers, stakeholders and scientists. Coherence with the FP7 is important to increase the opportunity of additional funding by the EU for projects in the coordinated programme as well in the national programmes. Step 3: a European R&D programme on sustainable land management. The ﬁnal step will be the direct involvement of funders, policy-makers, stakeholders and scientists in the formulation of a European programme. This programme has to comply with the European R&D agenda. European networks of these groups are necessary to allow representatives from these groups to participate in the programming process. Existing networks such as the Common Forum and the SNOWMAN group can be involved already in an earlier stage. Networks of stakeholders and scientists still have to be built up, making use of the CLARINET group or participants in other former or current European R&D projects. The three steps of the organisation of the programming process are the following. Step 1: call steering committee of funders. The organisation of the cooperation between research funders can be based upon a steering committee responsible for formulating the scope and funding of the coordinated programme and projects. The steering committee consists of one representative of each of the participating funding organisations. They must have the authority to make decisions about funding projects and signing contracts. Step 2: a programme advisory committee. Academics in universities and institutions, policy-makers and stakeholders can be involved in an advisory committee for programming the RTD. To improve the involvement of networks the members of the board are selected as linking pins between programmes, RTD projects and networks of stakeholders. They can also be involved in the execution of the programme and have a speciﬁc task to communicate with their networks or with people in their networks who are especially interested in speciﬁc subjects in the projects. Step 3: a research council on soil. In the ﬁnal stage European networks are established for end-users, policy-makers, stakeholders and scientists. Representatives of these networks are forming a research council of a European R&D programme on sustainable land management. It is possible that on a national level also representatives of the diﬀerent groups of stakeholders are discussing the national research agenda.

SNOWMAN: An Alternative for Transnational Research Funding

661

Representatives of these ‘‘national research councils’’ could also act as the research council for the European programme. This situation could form an intermediate step between steps 2 and 3. The increase in cooperation in the evaluation process can follow the next three steps. Step 1: representatives of national peer review panels. The review and selection of the proposals are done by a peer review team. The peer review team consists of peer reviewers from the participating programmes. Step 2: European peer review panel. From all participating countries peer reviewers are selected to be part of a ‘‘pool’’ of European peer reviewers. From this pool the programme advisory committee will select a peer review team that reﬂects the expertise necessary to review the submitted proposals. The interests of the end users shall be considered during the evaluation procedure by organising a communication phase with the relevant stakeholders during the pre-selection phase. Step 3: European review panel. In the review panel not only experts, but also end users, representatives of solution supply organisations and researchers from institutes and universities are present. Members of the panel will also have a task in the review of the execution of the projects. Stakeholders can guide the execution of research projects and identify interesting subjects and results to communicate with their networks.

11.1.3.2.3

Implementation and Evaluation

Implementation is the process of dissemination of know-how, demonstration and application in practice followed by the evaluation of practical experience. Implementation is deﬁned as ‘‘ﬂow of knowledge’’ from research to the use in practice. Dissemination is a process that starts at the beginning of the project or even during the programming phase, the so-called early stage dissemination. This can be ensured by involving the end users of the results in the process. It is important to ‘‘turn on the receiver’’ (i.e. end user) right from the start; otherwise dissemination at the ﬁnal stage may fail the receiver. The ﬁnal stage dissemination focuses on the output of results. Besides the output of results, also knowledge management (transfer of expertise, skills, knowledge) is important. Dissemination is a part of research projects, just like collecting R&D information. The increase in European cooperation in the dissemination process can follow the next three steps. Step 1: dissemination to the scientiﬁc community. The programme, the projects and the results are communicated via scientiﬁc journals, scientiﬁc posters and websites. EUGRIS is used as the central website and database for information about the programme and projects. CONSOIL is used as the central conference for communicating the programmes and the projects.

662

Chapter 11.1

Step 2: dissemination to the professional community. During the execution of the projects members of the review panel can act as interlinks between the projects and the professional community, being policymakers, technical experts and professional end users. They can initiate special workshops on speciﬁc themes. As projects are ﬁnalised the results can be presented as reports, fact sheets, guidelines, etc. EUGRIS can be used as an internet platform for information and communication about research projects and their results. Step 3: acceptance of results in practice. Results of R&D projects must be validated and evaluated in practice. Finally these results should be accepted as state-of-the-art in the participating countries. An extension of the EURODEMO project could be a start for this step.

11.1.3.2.4

Networking

As earlier stated the development of a network is not an activity in itself, but a result of involving networks in the activities of the coordinated programme. The existing networks can act as starting points for this process. In the ﬁeld of contaminated soil, a few networks already exist: the network of funders formed in the SNOWMAN project itself; some speciﬁc networks between research institutions; the governmental network or network of policy-makers of the Common Forum; the network of industrial stakeholders, institutions and service providers in NICOLE; and the network of stakeholders and solution suppliers involved in the EURODEMO project that started recently. Important European research projects that are running or recently ﬁnished like WELCOME, CORONA and AQUATERRA are also strengthening the European networks. Local and regional authorities and service providers have international networks, but these are not speciﬁc in the ﬁeld of contaminated soil. To build these networks their activities must be of interest to (potential) participants. This means that the networks must have an inﬂuence on European research programmes and research funds and improve the eﬀectiveness and evaluation of RTD results in practice. In the long term, the ERA networks and other networks could evolve into a European RTD network connected to the European research council, related to the theme of sustainable soil management.

11.1.4

The Upcoming Coordinated Call6

The ERA-NET SNOWMAN (sustainable management of soil and groundwater under the pressure of soil contamination) has as its objective to initiate and increase cooperation for collaborative research in Europe. The overall

SNOWMAN: An Alternative for Transnational Research Funding

663

objective of the coordinated action SNOWMAN is to enhance quality, relevance and utilisation of resources in Europe regarding research in the ﬁeld of the soil–water–sediment system. This ERA-NET project will organise a transnational coordinated call for research proposals in the ﬁeld of soil and groundwater (land management) with respect to contamination. Therefore SNOWMAN took up the TWG research agenda (Table 11.1.2) to shape a scientiﬁc scope for a ﬁrst transnational research programme. SNOWMAN is to learn how to perform such a pilot call: transnational cooperation, basic principles, guidelines for proposers, selection of projects, monitoring of proposals contracted and a better dissemination of results. Delivering an established research programme, settled and agreed research funding rules and a well developed dissemination strategy. Beyond the ﬁrst coordinated call and the evaluation of lessons learned, SNOWMAN wants to transnationally fund a larger research programme, ﬁlling the gap between the European framework programme and its respective national counterparts. SNOWMAN’s phase 1 resulted in the vision and the wish to implement a coordinated funded call in the ﬁeld of sustainable land management. The consortium has carried out a survey of respective research programmes across Europe and has consulted diﬀerent research funding organisations about their interests, to derive a truly transnational research agenda in this ﬁeld. Preparing phase 2, the consortium has produced signed letters of intent and one letter of support from nine diﬀerent research funding bodies. These indicate that funders in the partner countries are prepared to contribute up to h700 000 of research funding to a transnational programme working on the principles of sustainable land management, soil system processes, tools for sustainable land management and application of scientiﬁc knowledge and contaminated land technologies.

11.1.4.1

Objectives and Projects

The project intends to organise a coordinated call for research projects, co-funded by organisations in the SNOWMAN partner countries (Austria, Belgium, France, Germany, the Netherlands, Sweden and the UK). The aims of this call are: to experience international cooperation among national research funders by the SNOWMAN partners; to underpin the research agenda of SNOWMAN; to support the implementation of sustainable land management; and to harmonise funding and research mechanisms within Europe. This will be achieved by focusing the projects on diﬀerent topics in two areas.7 Area 1: Sustainable land management – Topic 1: principles of sustainable land management. This topic aims to achieve a better understanding of the meaning of sustainable land

664

Chapter 11.1

management in practice based on a review of (mainly) SNOWMAN countries’ experiences. The topic will deliver guidance for regulators across Europe for the development and implementation of sustainable land management. Based on the review, the research needs from a stakeholder perspective will be identiﬁed. – Topic 2: soil system processes. This topic is focused upon soil processes over larger areas and lower concentrations. It should reﬂect the state of knowledge about the functioning of the soil as a system and the relationship between soil quality and land use. Of particular interest is the consideration of the resilience of soils, the attenuation of contaminants in space and time and the impacts of the attenuation processes on soil functions and microbial diversity. The topic delivers a judgement of the existing knowledge and necessary developments from a scientiﬁc perspective. – Topic 3: tools for sustainable land management. In many countries tools have been developed for sustainable land management. In this ﬁeld of work, there is an opportunity to build on the existing experience by developing and promoting the use of harmonised decision support tools which support sustainable approaches. The topic focus is the exchange of know-how and the development and harmonisation of tools for the management of contaminated land. The topic will deliver suggestions for the harmonisation of the scientiﬁc basis of these tools and recommendations for the promotion of their use. Area 2: Application of science and technologies – This area is dedicated to the application of scientiﬁc knowledge and contaminated land technologies, related to sustainability. What new technologies are available for contaminated soil and groundwater treatment? How sustainable are each of these new technologies? Are there new technologies or approaches in the scientiﬁc literature which have the potential for greater sustainability than those presently available? The topic will deliver guidance for identiﬁcation of innovative approaches to brownﬁeld development.

11.1.4.2

The Principles of the Coordinated Call

The SNOWMAN coordinated call will oﬀer grant funding (or equivalent). Grant funding provides ﬂexibility in the funding process and allows the consortium to set selection criteria focused towards maximising the scientiﬁc quality of proposals. It allows the consortium to specify that applicants should originate from the funding countries. Ownership of intellectual property rights lies with the proposer and so there is good opportunity to bring new thinking into the research process. SNOWMAN will adopt the National Contributions model for the coordinated call. Under the National Contributions model, each country will fund those components of research proposals which take place domestically.

SNOWMAN: An Alternative for Transnational Research Funding

665

The coordinated call will be advertised on the SNOWMAN website and links to this page from the websites of SNOWMAN partners and funding organisations, from EUGRIS, CORDIS, the Water & Soil Times EU newsletter, and in relevant scientiﬁc and technical journals in each of the partner countries. This task will be coordinated by a secretariat, which will also be the central point of contact for applicants. The secretariat and oﬃce base will in principle be established for the duration of the call, and so far as is practically possible will remain at a single address throughout its life. The steering committee will manage the coordinated call, and will have a role at a number of stages in the evaluation process. The steering committee will be appointed before the evaluation process begins. Proposals shall be submitted to the secretariat, at the administrative base given, and within the terms of the ‘‘Applicants Guide’’. All proposals submitted in response to the coordinated call will be treated conﬁdentially by the SNOWMAN partners, the steering committee, secretariat and the appointed reviewers. The submission procedure will involve a single-stage application, followed by an eligibility check and a two-stage evaluation process. A single common application form will be used which will be subdivided into ‘‘A’’ (summary) and ‘‘B’’ (detailed) components to feed each of the two evaluation phases. It is especially important that the funders agree participation in this process and the assessment criteria before the call is launched, so that the steering committee, which will take the ﬁnal decision, is suitably empowered or mandated by the funders. Applications will be checked for eligibility by the secretariat. This eligibility test will be an administrative procedure, and will use the following criteria: number of participants and their place of establishment; compliance with procedures; and only proposals which are fully compliant with the procedures outlined in the ‘‘Guide for Proposers’’ shall be eligible for consideration for funding. The objective of this test is to conﬁrm that proposals lie within the remit of both SNOWMAN, and the individual research funders, before any further qualitative assessment is made. The next stage will consist of two ‘‘tests’’ (Figure 11.1.1). Funding evaluation. This evaluation will happen at the individual funder level. It will conﬁrm that both the participants and the proposal fall within the funding organisations remit and/or priorities and is therefore capable of being funded in the coordinated call. Fundable participants will be those legal entities eligible to participate in the funders’ RTD programmes in the SNOWMAN partner countries (Germany, Austria, the Netherlands, France, the UK, Sweden and Flanders (Belgium). ‘‘Fit to call’’ evaluation. This evaluation will happen at the SNOWMAN consortium level and will thus be carried out by the steering committee. It will conﬁrm that the proposal makes a signiﬁcant contribution to the work area described in the call, and that a range of proposals covering the full scope of the call are taken forward to the second stage.

666

Figure 11.1.1

Chapter 11.1

Evaluations stages 1 and 2 (eligibility and fundability checks).

These two tests will be carried out in parallel, and the results combined by the secretariat. Those proposals which pass both tests will be taken forward to the peer review stage (Figure 11.1.2). Participants in the research consortium shall enter into a consortium agreement setting out, as a minimum, their roles and responsibilities, internal organisation of the consortium, intellectual property rights arrangements and a means of settling internal disputes. A project coordinator will manage each joint project. Technical implementation of the project shall be the collective responsibility of the participants, and ﬁnal payments will be made only when complete transnational outputs have been delivered. Breaches of the funding contract will be dealt with according to the rules established by the relevant research funders. Each participant shall keep accounts as required by the project research funders, making it possible to determine the use to which funding has been put,

SNOWMAN: An Alternative for Transnational Research Funding

Figure 11.1.2

667

Evaluation 3 (peer review and ﬁnal recommendation).

and the eligibility of such expenditure. Eligibility of costs, claim, audit and other ﬁnancial procedures will be determined by the research funder in each country using their usual funding rules. Project technical progress shall be periodically evaluated by the secretariat and steering committee on the basis of progress reports provided by the participants. These shall also cover the implementation of the plan for use or dissemination of knowledge. In general, projects should produce one mid-term progress report and a ﬁnal report (subject to variation within individual project plans). Final payment will be retained by funders until ﬁnal technical reports have been submitted to and accepted by the steering committee and the national funding bodies. As a general rule it is expected that new intellectual property resulting from SNOWMAN-funded projects will be placed in the public domain.

668

Chapter 11.1

Where several participants have jointly carried out work generating new knowledge, they shall agree amongst themselves the allocation of knowledge ownership, taking into account the funding contracts into which they have entered. Participants in SNOWMAN coordinated call projects will be expected to proactively promote the knowledge resulting from the work undertaken. The participants should ensure that the knowledge resulting from the work is disseminated within the period of the project. Should the participants fail to do so, the SNOWMAN steering committee may take steps to disseminate the knowledge. A major objective of SNOWMAN is the transnational delivery of a work programme and knowledge transfer of the common results amongst the participating countries more widely. The dissemination of project results should generate multiplier eﬀects within Europe. Project proposals should thus contain well thought out and detailed dissemination plans. Public access to deliverables should be given a high priority by proposers and dissemination activity should be described in project ﬁnal reports. Executive summaries suitable for web publication will also be required and as a minimum will be disseminated via the EUGRIS and SNOWMAN websites. Therefore, it will be expected that a detailed dissemination plan is carried out as an integral part of submitted proposals. Basically, clariﬁcation should be given as to how the research results will be shared. Among the project members themselves, in a way that results are accessible for each project partner and with regard to the roles of the involved funding bodies. Among the scientiﬁc community. The project results have to be communicated via scientiﬁc journals, scientiﬁc posters and websites. EUGRIS is used as obligatory as the central website and database for information about the project, results and the deliverables. CONSOIL is used as the central conference for communicating the projects. During the execution of the projects the SNOWMAN steering group can act as an interlink between the projects and the professional community, being policymakers, technical experts and professional end users. They can initiate special workshops on speciﬁc themes. As projects are ﬁnalised the results can be presented as reports, fact sheets, guidelines, etc. EUGRIS should be used as an internet platform for information and communication.

11.1.5

Conclusions

Is transnational research funding a better approach? Discussions about research and soil-related aspects at the European level have shown that an overall solution to overcome the identiﬁed soil threats will not exist. In the EU, besides uniform framework conditions, diﬀerent and even case-related or country-speciﬁc approaches in all areas of public life remain

SNOWMAN: An Alternative for Transnational Research Funding

Table 11.1.5

669

SNOWMAN motivation. Short term

Inﬂuence

Opening up the information on national research programmes Sharing a common R&D agenda between the SNOWMAN partners

Eﬀectivness

Improve involvement of end users in R&D Stimulate use of best practices Start-up of common dissemination plans

Eﬃciency

Sharing procedures of funders Creating a database of evaluators and reviewers

Finance and funding

Bundle, focusing national budgets

vital elements. Regarding soil-related research this perception is caused by many objective and subjective concomitant factors, e.g. regional, climatic, historic, geological. To consider suitable solutions, e.g. in European soil policy, cross-national cooperation will necessitate compromises. How smaller the group and more comparable the starting position so much better will be the achievement. A common state of knowledge will stimulate compromises. Transnational research helps to achieve a common state of knowledge. Transnational research cooperation seems to be a successful attempt bridging the gap between national and European funding. ERA-NET will be an exercise to ﬁnd out the right doorway and the possibility to arrange Europe and its members for global competition. Additionally, the ERA-NET scheme will generate a beneﬁt for participating countries by the sharing and disseminating of results, and transnational cooperation will strengthen synergy eﬀects within the EU. As described before, the short-term goals related to SNOWMAN are summarised in Table 11.1.5. The following list summarises the goals and achievements. The goals on inﬂuence are on track by the joint information on national RTD programmes via the website, and there is a ﬁrst draft of a joint research agenda. This agenda will further be developed by increased involvement of the national funders. The short-term goals on eﬀectiveness are less developed; however SNOWMAN is working on a joint strategy to improve the use of research results. The goals on eﬃciency, ﬁnance and funding have all been achieved by the coordinated call. It has to be emphasised that these goals could not have been achieved without the support from the ERA-NET.

670

Chapter 11.1

Acknowledgements No work can be done single-handedly. The authors would ﬁrst like to thank H. J. Vermeulen, SKB, the Netherlands, who is the author of the SNOWMAN Vision Paper as the main input for Section 3 and J. Greaves, Environment Agency, UK, who was the editor of the Principles of the SNOWMAN Coordinated Research Call as source for Section 4. We would also like to thank A. Wieland, Federal Environment Agency, Germany, for giving us support while editing the report. Last but not least thanks are extended to the whole SNOWMAN consortium.

References 1. Networking the European Research Area Coordination of National Programmes, European Communities, 2005 (ISBN 92-894-9375-5). 2. EU Communication, Towards a Thematic Strategy of Soil Protection, COM 179 ﬁnal, 2002 (http://eurlex.europa.eu/LexUriServ/site/en/com/2002/ com2002_0179en01.pdf). 3. L. Van-Camp, B. Bujarrabal, A.-R. Gentile, R. J. A. Jones, L. Montanarella, C. Olazabal and S.-K. Selvaradjou, Reports of the Technical Working Groups Established under the Thematic Strategy for Soil Protection, EUR 21319 EN/ 1, Oﬃce for Oﬃcial Publications of the European Communities, Luxembourg, 2004. 4. Vital Soil, conference organised by the Dutch Presidency and the Commission, 18–19 Nov. 2004 (ﬁnal statement: http://www.eugris.info/news. asp?Listing¼Archive, #108). 5. CLARINET: An analysis of national and EU RTD programmes related to sustainable land and ground-water management, December 2002, Umweltbundesamt GmbH (Federal Environment Agency Ltd), Spittelauer La¨nde 5, A-1090 Wien, Austria. 6. Principles of the SNOWMAN Coordinated Research Call, 2006 (http:// www.snowman-era.net). 7. Outline of research projects in the SNOWMAN Coordinated Call, 2006 (http://www.snowman-era.net).

CHAPTER 11.2

Incorporation of Groundwater Ecology in Environmental Policy DAN L. DANIELOPOL,a CHRISTIAN GRIEBLER,b AMARA GUNATILAKA,c HANS JU¨RGEN HAHN,d JANINE GIBERT,e F. MERMILLOD-BLONDIN,e GIUSEPPE MESSANA,f JOS NOTENBOOMg AND BORIS SKETh a

Austrian Academy of Sciences, Institute of Limnology, Mondseestr. 9, AT5310 Mondsee, Austria; b GSF-National Research Center for Environment and Health, Institute of Groundwater Ecology, Ingolsta¨dter Landstrasse 1, DE85764 Neuherberg/Mu¨nchen, Germany; c Department of Ecotoxicology, Center for Public Health, Medical University of Vienna, Wa¨hringer Strasse 10, A-1090 Vienna, Austria; d Arbeitsgruppe Grundwassero¨kologie Universita¨t KoblenzLandau, Campus Landau Abt. Biologie, Im Fort 7, D-76829 Landau, Germany; e Universite´ Claude Bernard Lyon 1, UMR CNRS 5023 EHF, Equipe d’Hydrobiologie et Ecologie Souterraines, Baˆt FOREL, 43 Bd 11/11/ 1918, FR-69622 Villeurbanne cedex, France; f Istituto per lo Studio degli Ecosistemi CNR – ISE, Sede di Firenze, Via Madonna del Piano, IT-50019 Sesto Fiorentino/Firenze, Italy; g Milieu- en Natuurplanbureau, Netherlands Environmental Assessment Agency, Postbus 303, NL-3720 AH Bilthoven, The Netherlands; h University of Ljubljana, Biotechnical Faculty, Department of Biology, Vecna pot 111, PP2995, SI-1001 Ljubljana, Slovenia

11.2.1

Introduction: Groundwater Science and the New Order

In the European Community (EC) ca. 75% of the inhabitants use groundwater (GW) as drinking water, for food production and for domestic and/or industrial needs.1 Therefore information about the way high-quality GW originates and/or can be protected is of interest for a broad spectrum of Europeans, from laypersons to policy- and decision-makers. The recent release of the European Union (EU) Groundwater Directive (GWD)2 oﬀers a new order which should 671

672

Chapter 11.2

ensure the sustainability of the exploitation of GW reserves and at the same time should protect this resource from overall pollution. Moreover, it is expected to oﬀer better protection of valuable to water storage sites, especially wetlands, which strongly depend on GW. The new GWD develops the legislative framework already mentioned in Article 17 of the Water Framework Directive (WFD)3 (see Chapter 3.1). It has three major objectives: (1) to maintain a good chemical status of GW exploitable resources, (2) to prevent/limit GW pollution and (3) to develop studies on pollution trends in order to improve the water quality of GW bodies. The implementation of this ambitious long-term programme, as presented in the various contributions of this book, points out the need for new research and development strategies combined with adapted regulatory policies. From this new order a strong GW science accompanied by pragmatic policy decisions is emerging where hydrology, hydrochemistry and water planning are the major actors. What is missing in the recent GWD is the integration of GW ecology information as a useful complement to GW science even if this plea was repeatedly expressed.4,5 One could argue6 that this is due to a lack of data on the functioning of groundwater ecosystems and on the practical use of ecological information for the monitoring and/or protection measures of GW bodies. However, this is not really correct because a whole corpus of knowledge which forms the core of what is now called the ‘‘new groundwater ecology’’ exists (e.g. Refs. 7–9) and is being rapidly developed by various European scientiﬁc groups. Research projects of these latter exist in various European countries and were inter alia also ﬁnancially supported by the EC, e.g. PASCALIS (Protocols for the Assessment and Conservation of Aquatic Life in the Subsurface).10 Apparently, at the present time we suﬀer from lack of eﬀective communication between scientists dealing with ecological aspects of GW and water planners and/or water policy-makers. The recent eﬀorts of the Directorates for Environment and Research of the European Commission to improve communication between these various partners is therefore a welcome initiative,11 strongly supported by many specialists12–14 and reinforced through the various contributions of this book (e.g. Chapters 1 and 2.1). In the following, we oﬀer some practical ideas of how ecological information can be usefully integrated in future European GW management policies.

11.2.2

The New Groundwater Ecology: Its Interest for Water Management Projects and/or Water Policy Planners

Groundwater ecology deals with structural and functional aspects of the organisms which inhabit the subsurface and with the relationships between these organisms and their surrounding aquatic environment.7 Groundwater ecology emerged at the beginning of the 20th century from natural history research dealing with descriptive aspects of subterranean organisms, their adaptive morphology, their physiology, systematics and biogeography as well

Incorporation of Groundwater Ecology in Environmental Policy

673

as from observation on the hydrology, geomorphology and geochemical characteristics of various parts of the subsoil. The ‘‘new groundwater ecology’’ has existed for about 25 years.15–20 It deals with the study of ecosystems (structure and processes), with organismic assemblages and their dynamics, with relationships between subterranean and surface ecosystems, with the biological diversity of subterranean animals observed at various temporal and spatial scales and with the impact of anthropogenic pollution. A very important aspect of the new groundwater ecology is the emphasis put on ecosystem management with its instrumental aspects on monitoring and remediation schemes as well as with protection measures. Scientists involved in this new approach, as compared to the previous generation of naturalists, try to enforce the link between their research and the socioeconomic aspects of water management. Groundwater ecosystems are now more and more valued through their capacity to provide services and goods indispensable for the well being of human society and for the functioning of natural ecosystems (Table 11.2.1). Within the new groundwater ecology, students are in a very favourable position to use not only scientiﬁc arguments for environmental regulation and/ or political decisions but also ethical criteria (for instance when planning strategies for conservation of GW habitats and their unique organisms).21,22 Groundwater is an invaluable resource, which should be enjoyed by present

Table 11.2.1

Services and goods provided by groundwater ecosystems (modiﬁed from Ref. 4).

Ecosystem services

Ecosystem goods

1. Self-puriﬁcation: puriﬁcation of wa-

1. High-quality, safe drinking water

ter (through microbiological and physicochemical processes) Attenuation/or elimination of chemical contaminants (natural organic compounds, organic pollutants) where soil functions as a ﬁlter and biodegradation medium; has a strong link to long-term resource availability Provision of water for the environment as a conditions for the function of surface ecosystems (springs, brooks, lakes, wetlands, wet grasslands, estuarine and near-shore marine ecosystems). Sets hydrogeochemical conditions for subterranean and/or surface aquatic communities Maintain structural complexity through the landscape

for human consumption and the availability of a reliable water source Stable supply of water for other human needs (agriculture, industry, domestic needs) Water supply for the myriad of subsurface GW organisms Water for sustainability of GDEs Cultural value through the support for the maintenance of highly adapted GW organisms, a unique part of Europe’s biodiversity GW organisms indicate ecological conditions in GW (e.g. the hydrological and biogeochemical status of the GW system)

2.

3.

4. 5.

2. 3. 4. 5.

6.

674

Chapter 11.2

and future generations of humans. The unique organisms living in subsurface habitats are the product of a long evolution and their potential loss is seen as a prejudice to our cultural heritage. This opinion may be in contrast to those of water policy planners who favour socioeconomic criteria (see the recently released GWD).2 Ecological information on GW environments, when compared to the chemical data, is apparently less precise and its inferential power is generally low. This is partly due to the nonlinear processes which dominate organismic activities and occur in many cases also in non-living systems, e.g. karst systems.23 However, considerable research has recently focused on general aspects of ecological theory and praxis, which point out the linkages between biodiversity, ecosystem functioning and services24,25 or which use experimental approaches to quantitatively assess these relationships.26 These concepts can also be applied to the subterranean organisms, which may increase in this way the predictive power of various environmental models. In order to make the argumentation for the necessity to make better use of ecological information for GW management and/or environmental regulations more persuasive we will focus our attention to the following three topics: (1) the GW ecosystem approach, useful for planning environmental policies; (2) the diversity of GW habitats and organisms with their potential interest for environmental monitoring programmes; and (3) GW-dependent ecosystems as constituents of global surface/subsurface environmental units.

11.2.3

The Groundwater Ecosystem Approach as a Framework for Planning Pollution Prevention and/or Environmental Protection Strategies

The European WFD3 only indirectly deals with the ecology of GW ecosystems as it states: ‘‘the status of a body of groundwater may have an impact on the ecological quality of surface waters and terrestrial ecosystems associated with that groundwater body’’. There is an improvement in the new GWD2 as it is mentioned in its introductory presentation the importance of protection measures for GW ecosystems. This requirement was frequently mentioned by ecologists (e.g. Refs. 4–6, 27). It represents an important success of the new regulation and merits to be quoted in extenso: ‘‘Research should be conducted in order to provide better criteria for ensuring groundwater ecosystem quality and protection. Where necessary, the ﬁndings obtained should be taken into account when implementing or revising this directive. Such research, as well as dissemination of knowledge, experience and research ﬁndings, needs to be encouraged and funded’’.2 However, within the various articles of the GWD (e.g. Articles 1 and 3) only the term ‘‘body of groundwater’’ is mentioned which makes logical reference to the volume of subsurface water, including its quantitative and qualitative aspects. Moreover, Article 1 starts with the emphasis on ‘‘the assessment of good groundwater chemical status’’. Any specialist dealing with various aspects of the subterranean aquatic environment

Incorporation of Groundwater Ecology in Environmental Policy

675

knows that the quality and quantity of ‘‘good groundwater’’ depends on the biogeochemical processes in the subsoil, with its connectivity and strength of exchanges between the subsoil system and the surrounding earth layers and with its dependence on the multiple human activities which negatively impact the water. Hence, it becomes inescapable to improve the meaning of the term ‘‘body of groundwater’’ with the scientiﬁc content of what we ecologists understand as ‘‘GW ecosystem’’. The interest of this switching of meaning is not a matter of semantics but one of paramount importance for the strategic planning of water policies (a major concern also in the present book!). An ecosystem is generally deﬁned as the integration of various aspects of the physicochemical and biological units which dynamically interplay. Groundwater ecosystems are open subsurface systems through which energy and matter is transferred and further processed. It is delineated by external boundaries that can be recognised as a material reality or can be conceptually deﬁned.28 For open systems (GW systems belong to this type) one has to identify the input and the output areas which allow the contact with the surrounding system. Figure 11.2.1 shows a conceptual model of a GW ecosystem. One can recognise three structural components: (1) the sediment matrix with various types of voids depending on the sediment or rock type (porous, karstic, fractured types); (2) the circulating GW; and (3) the living organisms. The ﬁrst two units represent the habitat used by the living component of the system. It is to the merit of Castany29 to have pointed out that the major services of GW systems rely on three functions of aquifers: (1) the capacity to store water; (2) to transport with the water through the subsoil energy and matter; (3) to allow chemical and biological changes in water and in the solid substrate. For the third function the biological role of subsurface organisms is of paramount importance, especially that of microbial communities.30 The domain of GW within the earths’ crust is gigantic; metaphorically it can be represented as a ‘‘groundwater arena’’ (Figure 11.2.2). Therefore there is a high diversity of micro- and macro-environments or niches within the subsurface. We know that subsurface water exists in both saturated and unsaturated sediment layers and the water can penetrate deep into the earth down to several thousand metres. Even at more than 1000 m deep GW microorganisms can be detected.31 Exchange activities between surface water and the GW decrease with depth. Figure 11.2.2 depicts schematically the decrease in biological and

Figure 11.2.1

Schematic representation of an ecosystem (from Ref. 28).

676

Figure 11.2.2

Chapter 11.2

Conceptual view of the GW domain (see text for details).

chemical activities related to the water circulation and the depth of the subsoil layers. Close to the surface water (shown in Figure 11.2.2 as the white area with undulating lines) matter and organisms penetrate in high concentration and/or number (the surface water is the source of energy and matter penetrating below the soil’s surface); they can be stored there for diﬀerent periods of time (the subsoil layer acts as ‘‘sink’’ compartment); it can be further released either to the deeper layers or returned to the surface environment. Therefore this superﬁcial GW layer forms a dynamic source–sink ecotone.32 Beneath these zones, GW contains less and less high-value energetic matter, the environment becomes oligotrophic and organisms are strictly specialised to hypogean life (they are called stygobites). This is the domain of GW ecosystems where leaky sink processes dominate. Finally in the very deep layers of GW systems (in porous granular aquifers this should correspond to layers located deeper than several hundred metres below the surface of the soil or in conﬁned aquifers), the GW circulation is very slow, organic matter is scarce and refractory and the aquatic domain is generally hypoxic or anoxic. There is only a minimal return of the water and matter to the surface from this area. One could name this extreme environment metaphorically the zone of the black-hole sinks (an idea of L. Kornicker, personal communication to D.L.D.). Groundwater systems within the sediment layers relevant for human water consumption (in unconﬁned aquifers this lies, generally, 25–200 m deep below the soil surface) display permanent physical, chemical and biological

Incorporation of Groundwater Ecology in Environmental Policy

677

ﬂuctuations, they are far from equilibrium and they can switch when disturbed from one dissipative state to another. The facility with which the systems can return to a previous state marks their resilience capacity. It is well accepted now that GW systems as compared to those of surface waters are more prone to change their ecological state when perturbed by pollutants; also, the resilience time is longer. Hence, it is considered that the vulnerability of many GW ecosystems is greater as compared to surface water systems and therefore GW bodies need better protection measures33 (see Chapter 2.1). Aquifers which are located within landscapes, or hydroscapes, with a huge diversity of both physicochemical and biological attributes have a high capacity to naturally restore their ecological state. Such GW systems display a high functional adaptedness and their quality in terms of providers of ecological services and goods are very much appreciated by water managers.29 Groundwater ecosystems are both objective and subjective entities. It is this dual aspect which causes apparently an insuperable diﬃculty to water managers. In some cases we can materially deﬁne the boundaries of aquifers and we can recognise the input and the output areas.28 In other cases we have to set artiﬁcial boundaries especially when we model our system. Groundwater ecosystems as deﬁned above can be investigated at various spatial and temporal scales depending on the interests of scientists or of water managers. A grain of sand with a well-developed assemblage of microorganisms (Figure 11.2.3) covered by pelicular water and a thin layer of organic matter, surrounded by a laminar layer of GW ﬂowing along its surface, already represents a ‘‘minimal’’ GW ecosystem. The granular sandy sediments in slow ﬁltration columns with their bioﬁlms and interstitial meio- and macrofauna represent an artiﬁcial small ecosystem which is useful in the laboratory for simulation of various chemical and biological processes which exist in the ﬁeld but are diﬃcult to observe.28 For GW management purposes one investigates large areas we call aquifers, ‘‘GW bodies’’ or ‘‘subsurface hydroscapes’’. Considering ﬁeld situations we have to diﬀerentiate between deep GW ecosystems where the exchange with the surface water is reduced and superﬁcial

Figure 11.2.3

A ‘‘minimal’’ GW ecosystem: microbial bioﬁlm on a grain of sand (from Ref. 30).

678

Figure 11.2.4

Chapter 11.2

The main types of aquifers with their ecotones (from Ref. 32).

systems where the environmental conditions represent a mixture between those dominating the surface waters and those of remote or deep water layers. We call these superﬁcial systems ecotones (Figure 11.2.4) and a whole package of information on their ecology is available.15,34 Porous (granular) aquifers in alluvial sediments along running water systems build within their superﬁcial layers very dynamic ecosystems. Hydrologists and water managers appreciate those systems as they act as ‘‘bank ﬁltration’’ units and produce large volumes of high-quality water for human consumption.35 The habitat is called in the ecological literature ‘‘hyporheal’’ and the organismic assemblages are of ecotonal type. A large number of insect larvae and crustaceans live here beside typical blind hypogean crustaceans. They can locally stimulate microbial activity and change the structure of the sediment through bioturbation processes.36 The dynamics of hyporheic systems are very intensive, closely related to the evolution of surface water ﬂuctuations. Strongly polluted surface water has a negative impact on hyporheic systems, and therefore their degree of vulnerability can increase at fast rates.37 The terrestrial-GW system in porous (granular) sediments is a poorly investigated ecotone. It is common in riparian (wetland) landscapes. Within this transition zone we distinguish both unsaturated and saturated sediments on which rich microbial bioﬁlms, stimulated by the arrival of a large amount of organic matter, may develop. Aquatic interstitial invertebrates like insect larvae or worms which live normally in sediment layers fully saturated with water are also able to colonise

Incorporation of Groundwater Ecology in Environmental Policy

679

semi-terrestrial environments (the vadose zone of aquifers).28 Many plants, i.e. the so-called phreatophytes, and meadow trees send their roots into this transitory ecosystem and play an important role in the elimination of nutrients, like nitrates and/or other chemical compounds.38 Groundwater ecosystems in deep aquifers, as mentioned above, have more or less diverse organismic assemblages, their biological activity is reduced and the environmental ﬂuctuations are attenuated. Once such a system is perturbed it needs a longer recovery time. Hence the vulnerability of such systems to anthropogenic pollution is higher than in the shallow ecotonal systems. One should consider the pollution of deeper layers of porous granular aquifers with organic chemicals and thereafter diﬃculties encountered to restore them back to a pristine quality.39 Finally, one is entitled to ask as to what the best strategy is to protect a porous alluvial aquifer against pollution. The answer depends on the scale at which we need to manage the water resources and on the location within a river basin. Our experience with the ecology of alluvial ecosystems along large rivers like the Danube40 suggests that the most important areas to be protected are the river banks and the river bottom in the areas recognised as major input water zones to the aquifer. The cover land area above the aquifer, and here especially recharge zones and areas where the GW table is close to the surface, have to be better protected against diﬀused pollution. Karst (ground)water is partly stored in carbonate massifs within large subterranean voids and further circulates through conduits of various diameters, from microcrevices to large tunnels. Precipitation and surface water which inﬁltrates into the karst systems traverses in many cases a non-saturated superﬁcial layer called epikarst. Further it arrives within the saturated zone and will exﬁltrate again through karst springs or karst streams. The aquatic fauna of the epikarst is represented mainly by surface-dwelling organisms; there are few stygobiotic elements. Within the saturated zone of a karst system the water can ﬂow at high velocity through large conduits or can be stored in accessory large voids. The water ﬂows in a coherent way through a whole complex of voids which deﬁnes the karst system as a drainage unit.23 The organismic assemblages of karstic systems reﬂect the hydrological system; the main conduits are traversed by surface-dwelling animals displaying also low abundances in pristine waters. In the annex voids where water is stored and only slowly further released, animal assemblages are well diversiﬁed and dominated by stygobiotic species. One can use karstic fauna, especially small crustaceans, like the copepods, as ecological tracers for the identiﬁcation of the origin of the water41 (see Section 4). The vulnerability of the karst water to pollution events is well known. Good examples are the karst areas in northern Italy and in Austria which have focused attention in the past because of epidemic diseases due to drinking water contaminated with pathogenic microbes. The principle of prevention and protection of karst systems diﬀers partly from those of the porous granular aquifers. One has to identify in the karst, in addition to the zone of inﬁltration of the water, also the major pathways of

680

Chapter 11.2

water travelling through the subterranean systems and the exﬁltration points. Polluted water in karst spreads rapidly and can be intercepted by household wells located at the surface of the karst massif, or can be collected at the karstic spring outlets. Therefore the protection strategy is more related to the local mountainous massif and to the pathways of the water circulation as compared to those of porous granular aquifers located within river basins. Finally, it is important to mention here the proposal of de Marsily42 to create ‘‘hydrogeological nature reserves’’, in a similar way that ecologists develop ‘‘natural parks’’ for the protection of valuable landscapes and their ecosystems. Here we advocate the idea of erecting ‘‘nature reserves’’ including both surface and subterranean ecosystems which allow not only the maintenance of pristine GW reserves but also intact assemblages of organisms related to a wide spectrum of habitats.

11.2.4

Diversity of Groundwater Habitats and Organisms: Their Usefulness for Environmental Monitoring Programmes

We mentioned in the previous section that a ‘‘groundwater body’’ has to be viewed as an ecosystem incorporating not only the water but also the sedimentary substrate and the organisms that live within. Therefore for the specialists dealing with GW monitoring it appears necessary to map not only the GW quality but also the characteristics of the subsurface habitats and their living organisms. For instance alluvial sediments along streams not impacted by organic pollution display diverse animal assemblages which develop within diﬀerent types of habitats.43 The situation can be completely diﬀerent in chronically polluted sediments, e.g. those which become anoxic over a wide area, and where one ﬁnds a monotone type of sediments inhabited by few animal species.44 At the European scale, three major biogeographical regions have to be distinguished for GW fauna, which are mainly the result of the Pleistocene glaciation:45 (1) in those parts of northern and central Europe, which were covered by ice shields, GW diversity within the meio- and macrofauna is low and endemic species are nearly absent; (ii) in western, central and eastern Europe, several endemic and relictual species survived, and regional biodiversity is generally higher than in the north; and (iii) in contrast to these, in southern Europe, where the climate stayed moderate during the various glaciation periods, most of the old Tertiary fauna still persists. These regions are characterised by a diverse endemic, exclusively subterranean dwelling fauna. Here, endemism and diversity are highest in the karstic areas, where evolutionary drift is generally enhanced by the high fragmentation of biotopes. In consequence, more than 50%, sometimes up to 90%, of the hypogean species are strongly restricted in their distribution, with many species recorded from one single site only.46,47

Incorporation of Groundwater Ecology in Environmental Policy

681

Considering the microbial biodiversity both cultivation dependent and independent investigations have revealed that microbial communities are dominated by diverse heterotrophic bacteria. Furthermore, modern molecular studies have frequently identiﬁed members of several mostly uncultivated lineages as well as phyla totally devoid of cultured representatives.48 An ongoing research project on the ecological assessment of GW ecosystems, funded by the German Federal Environmental Agency, at the Institute for Groundwater Ecology, in Munich, shows that in special cases the subsurface microbial communities are distinct from those found in surface soil and aquatic environments. This distinction becomes apparent at the level of the speciﬁc assembly of GW microbial communities and by their special physiological capabilities and/or ecological requirements. Specialised microbial groups, well adapted to the prevailing environmental conditions (e.g. to the availability of electron donors and acceptors) can be used for monitoring GW habitats or for bioremediation programmes.48 During the last few years we have noticed that more and more ecological information on GW habitats and their organisms is used for environmental regulations. For instance the Swiss Water Protection Ordinance not only deﬁnes water quality standards, but also ecological goals: ‘‘the biocenosis in groundwater should be in a natural state adapted to the habitat and characteristic of water that is not or only slightly polluted’’.49,50 A GW environmental assessment system requires per deﬁnition the identiﬁcation of the ‘‘pristine reference status’’ of GW bodies (as habitats). Additionally the organismic communities can be used as biological indicators. For instance based on the hydrological exchange and the availability of organic matter and oxygen at a subsurface site, Hahn51 described three types of GW habitats with speciﬁc animal communities. These types are called (1) oligo-alimonic, i.e. ones with poor supply of organic matter (OM) resulting in an extremely poor fauna (mainly stygobiotic); (ii) meso-alimonic, i.e. ones with medium OM supply supporting an abundant and diverse stygobiotic fauna; and (3) eu-alimonic assemblages, i.e. ones with a high OM supply resulting in a very abundant and diverse fauna, but in this case dominated by surface-dwelling fauna (stygoxenes) and ubiquitous species. There are also examples for a complete community shift from a stygobiotic (exclusively subterranean dwelling organisms) to stygoxenic fauna (surface-dwelling organisms colonising temporarily the subsurface habitats) resulting from organic pollution.52–54 Metazoan fauna nicely reﬂect structural conditions such as hydraulic conductivity, heterogeneity of habitats in an aquifer and/or provide information on surface/subsurface hydrological exchanges. The composition of the meio- and macrofauna can further be used in many cases as an indicator for organic pollution. Hence this type of information can be used as an early warning system for the environmental quality of areas were drinking water production plants are located or may be considered in licensing procedures for water extraction or the evaluation of wetlands.27,54–56,71,72 Due to their omnipresent distribution, microorganisms and microbial communities may represent an attractive target for biological assessments,57,58 and may serve as reliable bioindicators for GW ecosystems.50,59,60 Microbes are characterised by short generation times and comparably high metabolic rates

682

Chapter 11.2

providing a fast reaction to changes in environmental conditions, mainly reﬂected by changes in their activity and shifts in the community composition. Today, these eﬀects may be resolved by means of cultivation-independent molecular methods. Based on the central hypothesis that in low energetic habitats with stable environmental conditions (such as most GW ecosystems) the microbial communities nicely reﬂect the in situ environmental conditions, the composition of microbial communities and their physiological status can serve as (bio)indicators for the assessment of the ecological situation. Molecular ﬁngerprinting techniques used in microbial community analysis of GW oﬀer information on changes in the community composition, which are related to environmental conditions including anthropogenic pollution. As an example RNA/DNA sequence analysis may deliver information on the individual members of the microbial community from which environmental conditions can be deduced. The analysis of functional genes on DNA bases reveals the potential for individual processes within communities and groups. Comparative analysis of microbial communities (bacteria, archaea, protozoa and fungi) in pristine and anthropogenically impacted GW ecosystems will, in the near future, provide the foundation for the selection of microbial species or groups indicative for a ‘‘healthy’’ and/or ‘‘impacted’’ status.50,59,60 A selection of valuable bioindicators will subsequently be collected and with the help of modern molecular tools, such as DNA microarray techniques, easy-to-handle assessment tools may be developed.58 Another important aspect of GW monitoring schemes is related to the artiﬁcial recharge of aquifer in urban zones. It is known that the increased demand for water in cities has motivated the search for replenishment of GW reserves through artiﬁcial recharge of the aquifers. Urban GW is commonly recharged by rivers, lakes and storm water runoﬀ.61 Monitoring the quality and quantity of the water in urban sectors is a complex activity because it needs to observe many diﬀerent parameters. Therefore new observation facilities were installed. The French Field Observatory in Urban Hydrology (OTHU), for instance, has oﬀered an integrated research programme since 1999. Its longterm objective is to acquire reliable data on urban eﬄuents during rainfalls and their impact on surface water and GW. An important activity of the OTHU is to monitor the impact of artiﬁcial storm water inﬁltration on GW ecosystems.62 Present data show that storm water inﬁltration leads to an increase of the local subsurface water temperature, of the OM content and of various other chemical nutrients.63,64 The OM enrichment of GW was positively linked to the density and diversity of GW assemblages of microorganisms and of invertebrates. Hence, the biotic subterranean communities represent useful biological indicators for urban GW quality. With this knowledge, and methodologies, it is possible to assess the sustainability of GW demand in the area of Lyon. Moreover it helps specialists dealing with urban water policies.62 In conclusion, we should point out here that the development of a biological assessment system within environmental programmes is not intended to replace traditional hydrological and physicochemical standard protocols, but rather to complement them.

Incorporation of Groundwater Ecology in Environmental Policy

11.2.5

683

Groundwater-dependent Ecosystems: A Holistic Representation

Groundwater-dependent ecosystems (GDEs) are hydroscapes or landscapes that must have access to GW in order to maintain their ecological structure and function.65 The implementation of the WFD3 requires the provision of an adequate amount of water for the individual ecosystem types. In trying to adapt an integrated river basin management (IRBM) concept as envisaged by the WFD, in most cases much attention is given to the maintenance of riparian and instream habitats. In many instances it has been the principal concern of water allocations under environmental ﬂow considerations. Here, the sum of estimated environmental ﬂows over a year is the total annual water volume, which can be allocated for environmental purposes.66 However, providing water for the environment is more than mere allocation of GW to river ﬂow and riparian health. The assessment of environment ﬂow requirements at river basin scale becomes more complex especially if all downstream ﬂuvial and coastal requirements have to be considered.67,68,69 The following are some examples: arid and semi-arid regions, such as in Mediterranean countries (Greece, Italy, Spain), which are characterised by long dry periods interrupted by short periods of high rainfall intensities and where running waters are most of the year driven by baseﬂow, i.e. the portion of streamﬂow that is contributed by GW; wetlands depending on GW inﬂux at all times of the year; terrestrial ecosystems that show seasonal or episodic reliance on GW; and estuarine and near-shore marine ecosystems that use GW discharge. The impact of the exploitation of GW resources on GDEs is a major concern for water supply companies70 as the proportion of GW in drinking water is generally high in Europe1 and one of the best places to extract huge volumes of subsurface water for human needs are the riverine aquifers. In view of these important economic aspects of GDEs, new research should be developed in the future combining ecology with hydrological studies,16 and/or using Castany’s concept of the ‘‘global aquifer/river system’’.29 So ﬁnally, if GW policy and management systems intend to appropriately consider and protect GDEs a better understanding will be needed of (1) identiﬁcation of likely GDEs and assessment of the nature of the dependency, (2) analysis of the respective ecosystem dependency on GW and timing of the dependency, (3) GW regime required to meet the water requirements of the ecosystem and (4) the impacts of change in key GW attributes on that ecosystem. By providing water managers and policy-makers with recommendations to the above, scientists can help managers to understand what will be required for sustainable management of GDEs and to integrate them in their plans.

684

11.2.6

Chapter 11.2

Overview: The Expanded Order (Achievements and Future Needs)

The information presented here should convince the reader that GW is not only an important resource needed for the well-being of humans but also a living medium for the diversiﬁed forms of life below the surface soil; it fuels water and energy to various kinds of subsurface, and even surface ecosystems. The multifarious activity of subterranean organisms oﬀers valuable (but mostly unrecognised) services to nature and humans. Hence, we are sure that GW ecology is a new important aspect in modern GW research. However, we still need to develop new pathways for the communication between the various partners involved in GW science. Once this important initial step is achieved we have to further transmit our combined GW knowledge to a broad spectrum of interested parties. Beside laypersons, stakeholders, water managers, providers for research and technology development as well as the policy decision-makers need to be informed about the advancement in GW ecological research. Figure 11.2.5 portrays this idea, inspired from Ref. 11 and from Chapter 2.1. Within this ‘‘expanded order’’, it is important to oﬀer some hints about what ecology in the future may oﬀer for improved management and protection of the GW domain. 1. Groundwater ecology is a useful tool for water management. Stakeholders, operational managers and policy-makers will proﬁt from the use of ecological knowledge and from the experience of GW ecologists. 2. GW ecology should contribute to programmes which unify the problems of resource sustainability with those of the maintenance of GW ecosystem integrity.

Figure 11.2.5

Flowchart of the way scientiﬁc knowledge is communicated to laypersons, water managers and policy-makers (from Ref. 12).

Incorporation of Groundwater Ecology in Environmental Policy

685

3. GW organisms are especially important for the self-puriﬁcation processes of GW systems; hence the maintenance of a healthy diversity of organisms is extremely important. We need early warning and ‘‘alarm bell’’ technologies which are able to rapidly detect pollution threats or trends in GW ecosystem stress. Micro- and/or macrobiota may selectively be used as indicator organisms for giving more complete information about the qualitative state of GW ecosystems. However, the diversity of these organisms should be ﬁrst better mapped, bioindicators identiﬁed and subsequently tested for monitoring purposes. The same applies for the habitats in which the GW organisms live. The EC 7th Framework Research Programme can provide a good basis for the development of such new technologies in ecological monitoring. 4. The project PASCALIS (funded within the EC 6th Framework Research Programme)10 oﬀers important recommendations for the way we should protect the aquatic subterranean biodiversity and how we should use the diverse organisms for water management and for protection policies. It is proposed inter alia: (a) the establishment of priority lists of GW animal species and GW habitats (aquifers) to be protected; (b) the application of biodiversity data to the evaluation of the ecological status of GW bodies; and (c) the development of a European network of nature reserves for the GW domain, which will protect not only the quality and quantity of subsurface water but also diverse subterranean organismic assemblages. 5. An interesting aspect of the new GWD is the ﬂexibility oﬀered to the EU member states for deciding on the practical measures for monitoring and/ or protection of the GW environment. Hence, member states could independently, where possible or useful, integrate GW ecology in their water policies. Especially the Mediterranean countries (e.g. Italy, Spain, Greece, Portugal) with their chronic problems of water shortage and economic problems to fund environmental monitoring schemes could proﬁt from existing GW ecological knowledge. 6. The WFD and the GWD will have to account for the large diﬀerences between karstic and porous granular aquifers (e.g. self-puriﬁcation potential, water residence time, vulnerability). Hence, one needs diﬀerent strategies for their management and protection. 7. GW ecology should be horizontally linked with other strategic EU frameworks and directives (like Natura 2000, the Birds and Habitats Directive or the Thematic Soil Strategy). 8. We have to raise more public awareness on the ecology of subterranean waters and its importance for both humans and nature. Hence, it will be necessary to undertake systematic campaigns highlighting the ‘‘yet unrecognised’’ and positive contribution of the GW ecosystem services within the broad umbrella of the advantages oﬀered by GW science. 9. Finally, by incorporating ecology in the policies dealing with water management we insert positive values of nature besides the existing socioeconomic values, those dealing with the quantity and quality aspects of the water. Our eﬀorts to communicate these ideas to a broad spectrum

686

Chapter 11.2

of people, if accepted and further developed, should enrich GW science and in the long term improve the economics of GW resources.

Acknowledgements Professor Ph. Quevauviller is acknowledged for the productive exchange of ideas concerning our topic and for his support and patience during the preparation of this contribution. One of us (D.L.D.) acknowledges the Austrian Foundation for Research (FWF) which gave ﬁnancial support during the years of his groundwater ecology research. Many colleagues helped with information and logistical support, a few are here mentioned, M. Bakalowicz, L. Kornicker, T. Lu¨ders, A. Mangin, R. Rouch, F. Schiemer, C. Schweer, K. Minati, J. Knoblechner, M. Pichler and H. Ployer, who helped during the production of the manuscript. J.G. and F.M-B. acknowledge the Urban Community of Lyon and the Rhoˆne-Alpes Region for their ﬁnancial support of the OTHU projects for many years. C.G. is indebted for ﬁnancial support to the Helmholz Society, to the German Ministry of Education and Research (BMF, contract KORA 02 0462) and to the German Research Foundation (DFG, contract ME-2049/2-1) and the Federal Environmental Agency (UBA project ‘‘Biological assessment of groundwater ecosystems’’).

References 1. A. Aureli and J. Ganoulis, UNESCO project on international shared aquifer resources management,2005 (http://www.inweb.gr/).. 2. Directive 2006/118/EC of the European Parliament and of the Council on the protection of groundwater against pollution and deterioration, OJ of the European Communities, L 372, p. 19. 3. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the ﬁeld of water policy, Oﬃcial Journal of the European Communities, L 327, 22.12.2000, p. 1. 4. D. L. Danielopol, J. Gibert, C. Griebler, A. Gunatilaka, H. J. Hahn, G. Messana, J. Notenboom and B. Sket, Environ. Conserv., 2004, 31, 185. 5. Deutsche Naturchutzring, Bund, Gru¨ne Liga, Position of German Environmental NGOs on Amendments No. 1-227 of the draft report by Christa Klass, 29.10.2004 and 20.12.2004, Berlin, 2005 (http://www.wrrl-info.de/docs/ DNR-Position Klass-Reportﬁnal.pdf). 6. S. Scheuer, EU Environmental Policy Handbook: A Critical Analysis of EU Environmental Legislation, European Environmental Bureau (EEB), Brussels, 2005, p. 125. 7. J. Gibert, D. L. Danielopol and J. A. Stanford, Groundwater Ecology, Academic Press, San Diego, CA, 1994. 8. H. Wilkens, D. C. Culver and W. F. Humphreys, Subterranean Ecosystems, Elsevier, Amsterdam, 2000.

Incorporation of Groundwater Ecology in Environmental Policy

687

9. C. Griebler, D. L. Danielopol, J. Gibert, H. P. Nachtnebel and J. Notenboom, Groundwater Ecology: A Tool for Management of Water Resources, Oﬃce for Publications of the EC, Luxembourg, 2001. 10. J. Gibert, World Subterranean Biodiversity, University Claude Bernard Lyon 1, Laboratory of ﬂuvial hydrosystems ecology, Villeurbanne, France, 2005. p. 13. 11. Ph. Quevauviller, P. Balabanis, C. Fragakis, M. Weydert, M. Olivier, A. Kaschl, G. Arnold, A. Kroll, L. Galbiati, J. M. Zaldivar and G. Bidoglio, Environ. Sci. Pol., 2005, 8, 203. 12. D. L. Danielopol, J. Gibert and C. Griebler, ESPR, 2006, 13, 138. 13. Umweltbundesamt, European Groundwater Conference 2006, Proceedings, Umweltbundesamt, Vienna, 2006. 14. R. Cunningham, S. Scheuer, D. Eberhart and C. Schweer, A Critical Assessment of Europe’s Groundwater Quality Protection Under the New Groundwater Directive, Berlin, 2006 (http://www.bund.net/lob/redot2/pdf/ groundwater directive assessment 20006/212.pdf). 15. J. Gibert, J. Mathieu and F. Fournier, Groundwater/Surface Water Ecotones:Biological and Hydrological Interactions and Management Options, Cambridge University Press, Cambridge, UK, 1997. 16. J. A. Stanford and T. Gonser, Freshwater Biol., 1998, 40. 17. J. Gibert and L. Deharveng, BioScience, 2002, 52, 473. 18. C. Griebler and F. Mo¨sslacher, Grundwasser-O¨kologie, Facultas, Vienna, 2003. 19. A. Boulton, Aquat. Conserv.: Mar. Freshw. Ecosyst., 2005, 15, 319. 20. A. Baba, K. W. F. Howard and O. Gunduz, Groundwater and Ecosystems, Springer, Dordrecht, 2006. 21. D. L. Danielopol, C. Griebler, A. Gunatilaka and J. Notenboom, Environ. Conserv., 2003, 30, 104. 22. D. L. Danielopol,and P. Pospisil, World Subterranean Biodiversity, University Claude Bernard Lyon 1, Laboratory of ﬂuvial hydrosystems ecology, Villeurbanne, 2005, p.29. 23. M. Bakalowicz, Hydrogeol. J., 2005, 13, 148. 24. M. Loreau, S. Naeem, P. Inchausti, J. Bengtssoon, J. P. Grime, A. Hector, D. U. Hooper, M. A. Huston, R. Raﬀaelli, B. Schmid, D. Tilman and D. A. Wardle, Science, 2001, 294, 804. 25. S. Naeem and J. P. Wright, Ecol. Lett., 2003, 9, 567. 26. A. M. Lohrer, S. F. Thrush and M. M. Gibbs, Nature, 2004, 431, 1092. 27. H. J. Hahn and E. Friederich, Grundwasser, 1999, 4, 147. 28. D. L. Danielopol, P. Pospisil, J. Dreher, F. Mo¨sslacher, P. Torreiter, M. Geiger-Kaiser and A. Gunatilaka, Subterranean Ecosystems, Elsevier, Amsterdam, 2000, p. 481. 29. G. Castany, Principes et me´thodes de l’hydroge´ologie, Dunod, Paris, 1982. 30. C. Griebler, Groundwater Ecology: A Tool for Management of Water Resources, Oﬃce for Publications of the EC, Luxembourg, 2001, p. 81. 31. F. H. Chapelle, Ground-water Microbiology and Geochemistry, J. Wiley, New York, 1993.

688

Chapter 11.2

32. J. Gibert, F. Fournier and J. Mathieu, Groundwater/Surface Water Ecotones: Biological and Hydrological Interactions and Management Options, Cambridge University Press, Cambridge, UK, 1997, p. 3. 33. J. Notenboom, Groundwater Ecology: A Tool for Management of Water Resources, Oﬃce for Publications of the EC, Luxembourg, 2001, p. 247. 34. W. K. Jones, D. C. Culver and J. S. Herman, Epikarst, Special Publication 9, Karst Waters Institute Inc., Charlestown, WV, 2004. 35. H. Frischherz, Wiener Mitteilungen, Wasser-Abwasser-Gewa¨sser, 1979, 29, 1. 36. D. L. Danielopol, J. N. Am. Benthol. Soc., 1989, 8, 18. 37. J. B. Jones and P. J. Mulholl, Streams and Ground Waters, Academic Press, San Diego, CA, 2000. 38. M. Tre´molie`res, D. Correll and J. Olah, Groundwater/Surface Water Ecotones: Biological and Hydrological Interactions and Management Options, Cambridge University Press, Cambridge, UK, 1997, p. 227. 39. M. Alexander, Biodegradation and Bioremediation, Academic Press, San Diego, CA, 1994. 40. D. L. Danielopol, P. Pospisil and J. Dreher, Hydrological Basis of Ecologicaly Sound Management of Soil and Groundwater, IAHS Publication 202, 1991, p. 215. 41. R. Rouch, Me´m. Soc. Geol. France, 1980, 11, 109. 42. G. de Marsily, Ground Water, 1992, 30, 658. 43. J. W. Ward, G. Bretschko, M. Brunke, D. L. Danielopol, J. Gibert, T. Gonser and A. G. Hildrew, Freshwater Biol., 1998, 40, 531. 44. D. L. Danielopol, Int. J. Speleol., 1976, 8, 322. 45. A. Thienemann, Verbreitungsgeschichte der Su¨ßwassertiere Europas. Versuch einer historischen Tiergeographie der europa¨ischen Binnengewa¨sser, Springer, Stuttgart, 1950. 46. B. Sket, Biodiv. Conserv., 1999, 8, 131. 47. D. C. Culver and B. Sket, J. Cave Karst Stud., 2000, 62, 11. 48. C. Griebler and T. Lueders, Freshwater Biol., submitted. 49. GSchV, Water Protection Ordinance, SR 814.201, Swiss Federal Law, Bern, 1998. 50. N. Goldscheider, D. Hunkeler and P. Rossi, Hydrogeol. J., 2006, 14, 926. 51. H. J. Hahn, Limnologica, 2006, 36. 52. B. Sket, Proc. 6th Int. Congr. Speleol. Olomouc, 1977, 5, 253. 53. B. Sket, Biodiv. Conserv., 1999, 8, 1319. 54. F. Malard, J. Mathieu, J.-L. Reygrobellet and M. Lafont, Hydrobiologia, 1999, 58, 158. 55. P. Dumas, C. Bou and J. Gibert, Int. Rev. Hydrobiol., 2001, 86, 619. 56. T. Pipan, A Brancelj, Zool. Stud., 2004, 43, 206. 57. H. W. Pearl, J. Dyble, P. H Moisander, R. T. Noble, M. F. Piehler, J. L. Pinckney, T. F. Steppe, L. Twomey and L. M. Valdes, FEMS Microbiol. Ecol., 2003, 46, 233. 58. K. Lemarchand, L. Masson and R. Brousseau, Crit. Rev. Microbiol., 2004, 30, 145.

Incorporation of Groundwater Ecology in Environmental Policy

689

59. C. Griebler, T. Lu¨ders and J. Liebich, European Groundwater Conference 2006, Proceedings, Umweltbundesamt, Vienna, 2006, p. 130. 60. D. Hunkeler, N. Goldscheider, P. Rossi and C. Burn, Biozo¨nosen im Grundwasser: Grundlagen und Methoden der Charakterisierung von mikrobiellen Gemeinschaften, Bundesamt fu¨r Umwelt, Bern, 2006. 61. R. D. G. Pyne, Groundwater Recharge and Wells, Lewis Publishers, Boca Raton, FL, 1995. 62. S. Barraud, J. Gibert, T. Winiarski and J.-L. Bertrand-Krajewski, Water Sci. Technol., 2002, 46, 203. 63. T. Datry, F. Malard and J. Gibert, Sci. Total Environ., 2004, 329, 215. 64. T. Datry, F. Malard and J. Gibert, J. N. Am. Benthol. Soc., 2005, 24, 461. 65. B. Murray, G. C. Hose, D. Eamus and D. Licari, Austral. J. Bot., 2006, 54, 221. 66. V. Smakhtin, C. Revenga and P. Do¨ll, Water Int., 2004, 29, 307. 67. M. Sophocleous, Hydrogeol. J., 2002, 10, 52. 68. D. Eamus, R. Froend, R. Loomes, G. Hose and B. Murray, Austral. J. Bot., 2006, 54, 97. 69. W. F. Humphreys, Austral. J. Bot., 2006, 54, 115. 70. J. Petersen and U. Su¨tering, Wasser & Boden, 2003, 55, 58. 71. F. M. Butterworth, A. Gunatilaka and M. E. Bonaparte, Biomonitors and Biomarkers as Indicators of Environmental Change, Plenum Press, New York, 2001, vol. 2. 72. A. Gunatilaka and J. Dreher, Water Sci. Technol., 2003, 47, 53.

CHAPTER 11.3

Towards a Science–Policy Interface (WISE-RTD) in Support of Groundwater Management PHILIPPE QUEVAUVILLERw European Commission, DG Environment (BU9 3/142), Rue de la Loi 200, BE-1049 Brussels, Belgium

11.3.1

Introduction

As highlighted in Chapter 2.1, there is a clear need to improve the link among scientiﬁc developments and environmental policy implementation in view of a better integration of scientiﬁc outputs into policies. In this respect, the discussion of a possible ‘‘science–policy interface’’ based on a coordination of relevant programmes/projects has been ongoing since 2004 with regard to the European Union (EU) Water Framework Directive (WFD) implementation.1 This chapter gives an outline of general needs and of the diﬀerent elements of such an interface, and presents an operational web portal linked to WISE (Water Information System for Europe),2 namely the WISE-RTD web portal,3 which is also described in Chapter 4.3. The chapter also provides information on research and technology development (RTD) funding mechanisms and on WISE, and discusses coordination needs among the diﬀerent science and policy features. This is obviously not only directed at the groundwater sector, i.e. the scope of the initiative is much wider as it covers, in principle, all research- and policy-relevant features concerning water, but also the examples below are directed towards groundwater research and policy.

w

The views expressed in this chapter are purely those of the author and may not in any circumstances be regarded as stating an oﬃcial position of the European Commission.

690

Towards a Science–Policy Interface (WISE-RTD)

11.3.1.1

691

General Needs

One of the main principles of an operational science–policy interface, as discussed in the water sector, is that R&D results should be synthesised in a way that can eﬃciently feed the implementation and further review of water policies. The transfer mechanism may be diﬀerent according to the policy step, e.g. the development of daughter directives, technical support for policy implementation, reviews based on scientiﬁc progress, which type of research is required (background information or tailor-made research and demonstration).

11.3.1.2

Diﬀerent Levels of Interactions

The integration of scientiﬁc outputs into water policies may be conceived at various levels, e.g. diﬀerent user communities, policy steps. One of the main identiﬁed diﬃculties for ensuring such integration stems from the fact that there is no suﬃcient streamlining of information from, for example, the scientiﬁc community to policy decision-makers, neither is there from the latter to the former as to formulating their problems to identify scientiﬁc ‘‘inputs’’ necessary to solve them.4 Eﬀorts are ongoing in the framework of various initiatives (the HarmoniCA project5 funded by the 5th Framework Programme, in particular; see Chapter 4.3) to examine how an eﬃcient and operational ‘‘science–policy interface’’ could be developed primarily in support of the WFD implementation. This interface actually aims to meet the demand of diﬀerent levels of users and stakeholders; in particular, it should help to enhance the conﬁdence and trust of the latter towards scientiﬁc results from projects funded by the EU. Community added value could be ensured in terms of combination of the best expertise, critical mass and resources at EU level and testing against broader geographical, ecological, socioeconomic and cultural background conditions which cannot properly be covered by nationally funded research. In the ﬁrst place, a ‘‘simple’’ level of information is necessary. It is addressed to the general public, and should hence possibly take into account the language barrier, i.e. translating relevant WFD information into the 23 EU languages. In this context, explanations on the key issues of the WFD groundwater features and of the new Groundwater Directive should be explained in a publicly accessible way (Figure 11.3.1). A second level of information is required for operational managers (i.e. technical people responsible for practical aspects of policy implementation) and research and technology providers. In this context, information on relevant tools as developed through research and demonstration projects should be made accessible to users in a readily applicable way, e.g. in the form of description sheets (using a common template) and guidance. Similarly to the ﬁrst level (Figure 11.3.1), information on new methods, technologies, management solutions, etc., should be presented according to diﬀerent categories of information in order to properly guide the users (Figure 11.3.2). A third level of information needs to focus on relationships among river basin authorities and the scientiﬁc community regarding monitoring schemes

692

Chapter 11.3 Accessible (in 23 EU languages?), covering main features of the Groundwater Directive

1.

Good status objectives

5.

Figure 11.3.1

2. Characterisation, Resp. authorities

Monitoring, modelling

6.

3.

Economic issues

Measures Against pollution and deterioration

7.

4.

River Basin Management plans

Public participation

First level of information: providing the public with accessible knowledge (adapted from Ref. 4).

Access to guidance, good practices, methods and S&T information

1.

Characterisation, modelling

Figure 11.3.2

2.

Management, economics

3.

Monitoring, Compliance testing

4.

Measures (protection, Restoration)

Second level of information: technical implementers and RTD providers (adapted from Ref. 4).

and the associated data that are available on their respective river basins, as well as on decision-making based on these data (e.g. design of programmes of measures, follow-up of their eﬃciency). In addition, enabling the scientiﬁc community to use these data is absolutely crucial for further development, validation and calibration of tools (e.g. modelling tools) against real-world conditions and enhancement of their robustness and operability. Finally, a follow-up of research and technological progress is of key importance to ensure that reviews of EU legislation are carried out with a proper integration of the latest R&D ﬁndings. This information level corresponds to a kind of benchmarking of relevant activities (including discussions on future water R&D policy) that might have an impact on future reviews. In addition, this information is important to possibly avoid duplication of R&D developments either at national or EU level. This information level needs to be addressed to research programme managers and policy implementers. Examples of information needs concern projects from the EU RTD framework programmes, LIFE programme (demonstration projects)6 (see Section 3.6), national research programmes, etc. (Figure 11.3.3).

693

Towards a Science–Policy Interface (WISE-RTD) Access to scientific information in a “policy-formatted”way

1.

RTD-funding programmes

Figure 11.3.3

2.

EU-funded projects

3.

National projects

4. LIFE and other projects

Third level of information about R&D projects (adapted from Ref. 4).

To summarise, a mechanism is required to ensure that the most promising research and demonstration projects in support of the groundwater policy framework, and more generally to water policies in a wider context, are disseminated eﬃciently and applied at the appropriate level (regional, national or EU). The issue of development of a possible ‘‘interface’’ is discussed below.

11.3.2

Introduction to WISE

11.3.2.1

What is WISE?

WISE, the Water Information System for Europe, is an umbrella term for a wider initiative to modernise and streamline the collection and dissemination of information related to European water policy.7 The starting point for WISE is the WFD, a new and comprehensive piece of legislation consolidating EU water policy and introducing an integrated and holistic approach to water management. The overall concept for WISE was laid down in a document which was agreed in 2003.8 Since then, the process of implementing WISE has started but is still in the early stages. There are several aspects which illustrate what WISE is (or will be), in particular the following. WISE is the water-related component of INSPIRE. However, regarding water-related information, WISE is going beyond INSPIRE since it also covers non-georeferenced data and information (e.g. numeric data, textual or administrative information). WISE is a formal compliance reporting tool. In this regard, it facilitates the information exchange between EU member states and the Commission. This implies that formal rules are established for this part in order to avoid parallel or double reporting. However, only part of the data in WISE is relevant for compliance. WISE is water-related data available on a European level. Extensive amounts of data are being collected by European and international bodies. Exchange of data and interoperability of systems.

694

Chapter 11.3

In summary, depending on the context, WISE can refer to an initiative, a concept, a process, an information system, a set of rules or tools for reporting, a dataset or a component. If WISE is referred to as an information system, it includes all possible WISE nodes, data and viewer providers as well as the common WISE public website and their interactions. It is not a central ‘‘mega-database’’ but rather a decentralised system at EU level which will have capabilities to interoperate with existing national systems.

11.3.2.2

Why do we Need WISE?

The WISE process has been in place since 2003. It is based on the critical review of past reporting exercises in relation to water-related directives before the WFD, in particular the one based on the Standardised Reporting Directive (91/692/EEC) and the water questionnaire (95/337/EEC). Amongst the lessons learnt from the past are: the need to streamline and facilitate water-related reporting in order to avoid burdensome double reporting to European bodies; the need to move to an electronic-based reporting in order the make the process more eﬀective and eﬃcient; the need to ensure quality checking and control of the submitted data in order to enhance readability, validation and processing of the data; the need to clarify requested reporting information and reduce diﬀerent understanding of the reporting questions in order to improve comparability of the data; and the need to be able to correct and update submitted data, as appropriate, in order to ensure that changes are being communicated to the users and considered in assessments. Furthermore, there is an increasing amount of water-related data publicly available and there is a need to share such data in a more eﬀective way. Moreover, for integrated assessments, e.g. across several environmental media (air, water, soil), it is increasingly necessary to ensure interoperability of information systems. Finally, the various EU bodies (DG ENV, JRC, ESTAT and EEA) have diﬀerent interests and needs for water-related data but are committed to work together using their competences to share the data in the most eﬀective and eﬃcient way.

11.3.2.3

What is the Objective of the WISE Process?

Further to this analysis, it was common sense to develop a long-term perspective for water-related reporting which is based on the requirements and approaches set out by the WFD. The overall objective for this process is the ‘‘development of a new, comprehensive, shared European data and information management system for water, including river basins (WISE). The system should be based on the concept paper and should be fully implemented by 2010.’’

Towards a Science–Policy Interface (WISE-RTD)

695

This objective and the more detailed concept for a WISE were agreed by all relevant EU bodies (DG ENV, JRC, ESTAT and EEA) and the then 25 EU member states, Norway, Bulgaria and Romania during the water directors’ meeting in November 2003. In the beginning, WISE will concentrate on providing an infrastructure for reporting under the WFD for the 27 EU member states. However, the scope will be extended step-by-step to cover all EU water-related information and inviting other countries with which the EU shares river basins and other international organisations to cooperate on a voluntary basis.

11.3.2.4

Does WISE Already Exist?

Following the agreement of the WISE concept paper, the Commission decided to develop within a short timeframe a WFD prototype which enabled the submission of reports in accordance with Article 3 (designation of competent authorities and river basin districts, RBDs) of the directive. This ﬁrst prototype was available in June 2004 and EU member states committed themselves to submit information to the WFD prototypez until June 2005 on a voluntary basis. In the meantime, the WFD prototype has been extended for submissions of reports in relation to Article 5 of the WFD (characterisation of RBDs, analysis of pressures, impacts and economic aspects). Furthermore, a ﬁrst (map and data) viewer has been programmed. The updated WFD prototype was presented at the WISE workshop on 15–16 December 2005 in Brussels. EU member states are committed to transfer the relevant Article 5 data to WISE. However, the current WFD prototype does not reﬂect the envisaged system design and operation. The coming years will be used to set up a wider, more ambitious and more integrated information system. As a ﬁrst step, it will be necessary to deﬁne what is ‘‘WISE compatible.’’ The next steps are to develop the components, for the input and output data ﬂow. Thereby, best use will be made from existing systems such as the EIONET Water or the WFD prototype. However, if necessary, also new components and tools will be developed. In particular a WISE GIS infrastructure will have to be put in place so that it can be used for diﬀerent purposes. The WISE Implementation Plan 2006–2010 sets out the steps for this process.

11.3.2.5

Who will Use WISE?

Currently, the WFD prototype is limited to compliance data exchange between the EU member states and the Commission. In its ﬁnal form, WISE will have a variety of users. The EU bodies will use the information for three purposes: compliance checking, assessment of state of the environment and trends (SoE) and z

WISE WFD prototype is an IT system based upon Linux, PHP, Oracle (Spatial) and Minnesota Mapserver. It is developed by WRc plc (UK) on behalf of DG Environment. WISE WFD prototype is hosted by a server located at the JRC site in Ispra, Italy.

696

Chapter 11.3

policy eﬀectiveness analysis. The European Commission will use WISE to inform the public and generate summary reports on the implementation of the respective EU legislations mainly targeted to the European Parliament and the Council (see Article 18 WFD). The EU member states and international organisations will not only be data providers, but will also use the information for their own purposes, e.g. obtaining statistics on implementation across the EU. The general or interested European public (e.g. environment NGOs, industry stakeholders) will seek to obtain an overview of the status and quality of the aquatic environment and the success with which policy instruments are able to protect it. The information will therefore also be useful in the context of the public participation requirements of the WFD when interested parties in river basin districts wish to compare the situation to other river basins.

11.3.2.6

What will WISE be Used for?

The main services that WISE will provide are the following. Viewing and visualisation: the electronic submission of data will facilitate and improve the possibilities for producing paper maps or ﬁgures for reports, web mapping services and other visualisation tools. Expert assessment: the submitted data can be used for various assessments, e.g. compliance, pressure, trend, state of the environment or impact assessments. Not all data may be useable for all purposes but the system will allow the selection of appropriate data sets for diﬀerent purposes. Analysis scenarios and research: in addition to expert assessments, the data can be analysed, for example, to identify the contribution of certain pressures from diﬀerent sectors (agriculture, industry, etc.). It may also be possible to analyse water-related data with data from other policy areas (comparison of indicators) or develop prediction scenarios on possible future trends (e.g. baseline scenario, diﬀerences in water supply and demand, economic development). Also the scenarios could test various policy options on EU level to identify the most cost-eﬀective responses. Finally, the data may be useful in diﬀerent research areas, e.g. the validation of models. The data gathering in WISE is mainly driven by the core purpose of expert assessment. In other words, the requirement to submit data for compliance checking under EU legislation is the driving force for deﬁning the data or reporting needs. The other purposes are being fulﬁlled on the basis of the reported data and, in general, rarely additional data are being asked for solely for viewing or making scenarios. Furthermore, the system will enable the introduction of standardised quality control and quality checking procedures which ensure that the data improve

Towards a Science–Policy Interface (WISE-RTD)

697

with time from ‘‘poor’’ to high’’ quality and thereby improve comparability. Most of these quality controls steps may be done by the data holders, i.e. decentralised (e.g. EU member states), but will be coordinated through WISE.

11.3.2.7

What are the Next Steps in WISE Development?

The next important steps are set out by the WISE Implementation Plan for 2006 to 2010. It provides the general system design and the technical features that need to be developed. It also sets out a process for developing a WISE GIS infrastructure and deﬁnes the data sets that will be incorporated up to 2010. As a ﬁrst milestone, WISE will go public in early 2007. The oﬃcial public launching event has been held in Brussels on the 22 March 2007.

11.3.3

EU RTD Funding Mechanisms

The treaty establishing the EU indicates that Research Framework Programmes have to serve two main strategic objectives. First, they provide a scientiﬁc and technological basis for industry and encourage international competitiveness. Second, they promote research activities in support of other EU policies. To this end, Framework Programmes (FPs) are designed to help in solving problems and responding to major socioeconomic challenges faced by society. The Research Framework Programme is the EU’s main instrument for funding research and development. In this context, the European Commission has been supporting research on water for several years through its successive FPs for RTD.9 The FP aims to foster scientiﬁc excellence, competitiveness and innovation through the promotion of better cooperation and coordination. It also aims to produce advances in knowledge and understanding, and to support the implementation of related European policies. The FP is implemented through open ‘‘calls for proposals’’ and successful projects are selected after an evaluation procedure carried out with the help of external independent experts.

11.3.3.1

FP5 Research Projects

Water has been identiﬁed as a key action in the 5th Framework Programme (1998–2002), as part of the Environment and Sustainable Development Programme. The Key Action ‘‘Sustainable Management and Quality of Water’’ has invested more than h150 million in research projects directly relevant to the WFD. To further enhance the impact of EU-funded research, projects within the same thematic area were clustered together in order to improve coordination and synergies, promote integration and synthesis of results of policy needs, create platforms/forums for active communication and targeted dissemination of RTD results to key stakeholders and end-users. Areas addressed by clusters which are of direct interest to groundwater are integrated catchment modelling (CATCHMOD), management of scarce water resources (ARID), integrated

698

Chapter 11.3

urban water management (CityNet), drinking water (CLUED’EAU), etc. Speciﬁc research projects also contributed to ongoing discussions about groundwater policy development, e.g. the BASELINE project (see Chapter 5.3), and many other examples described in this book.

11.3.3.2

FP6 Targeted Research and Integrated Projects

The 6th Framework programme (FP6) for RTD (2002–2006) included two main priorities that integrated research in support of water policies, namely Priority 6.3 ‘‘global change and ecosystems’’ and the so-called Priority 8 ‘‘policy-oriented research.’’ While Priority 6.3 opened the possibility to fund research projects dealing with policy in general, Priority 8 was designed to respond to direct policy needs expressed by various EC general directorates. Water research in the context of the Global Change and Ecosystems subpriority was mainly supported by the ‘‘water cycle, including soil-related aspects’’ area which put emphasis on hydrology and climate processes, ecological impact of global change, soil functioning and water quality, integrated management strategies and mitigation technologies, and scenarios of water demand and availability. These topics were mainly implemented with the help of integrated projects (IPs) and speciﬁc targeted research projects (STREPS). Integrated projects are projects of substantial size, designed to help building up the ‘‘critical mass’’ in objective-driven research with clearly deﬁned scientiﬁc and technological ambitions and aims, while STREPS are smaller projects in terms of scale of activities, duration and partnership. Examples of IPs of direct relevance to groundwater focus on interactions between land-use change, nutrient loading, acid deposition, toxic pollution on the structure and functioning of European freshwater ecosystems (EUROLIMPACS), on integrated modelling of the river–sediment–soil–groundwater system in the context of global change (AQUATERRA), on new approaches to adaptive water management under uncertainty (NEWATER) and on new tools integrating management, technical, economic and institutional instruments for water stress areas (AQUASTRESS). In addition to those IPs, several STREPS were also funded with indirect links to groundwater policies, in particular twinning basin initiatives (e.g. RIVERTWIN, TWINBAS) aiming to improve the eﬀectiveness of the cooperation between European and other countries’ river basins for the implementation of integrated water resources management (IWRM) principles (as stated in the WFD and the EU Water Initiative), to enhance the human resources capacities of other countries and to collect and disseminate the shared knowledge for the common beneﬁt of all.

11.3.3.3

FP6 ERA-NET Projects

The ERA-NET scheme has been set up within the FP6 to coordinate national and regional publicly funded programmes. Funding bodies like ministries and research councils may submit proposals for the networking of national or

Towards a Science–Policy Interface (WISE-RTD)

699

regional research programmes or innovation programmes in sectors of their choice. The Commission funds the coordination and the EU member states ﬁnance the research activities. Typically, ERA-NET projects include exchanges of information on programmes and projects, exchanges of best practice, strategic analyses for future joint activities and programmes, joint calls for proposals, etc. This mechanism hence allows for the coordination of research programmes with relevance to environmental policies, including the WFD, but also on bilateral or international (research) programmes. The ERA-NET scheme represents a very valuable mechanism to regroup national funds at the level of programmes so that larger or more coordinated projects can be funded. Further, it allows increasing the access to scientiﬁc expertise available at regional or national level as well as cross-border cooperation at the levels of programmes and of projects. In the sector of groundwater, an example of a relevant ERA-NET project is the SNOWMAN project which aims to coordinate research on sustainable management of soil and groundwater under the pressure of soil pollution and soil contamination (thus directly in line with the WFD provisions, in particular concerning groundwater).

11.3.3.4

Projects Issued from the Scientiﬁc Support to Policies (SSP) Priority

The so-called Priority 8 ‘‘policy-oriented research’’ has been designed within the FP6 to respond to direct policy needs expressed by various EC general directorates (DGs). It hence enabled publication of calls for proposals which accommodated speciﬁc research needs which were described by the policy DGs in the forms of terms of reference. In this respect, a range of topics has been deﬁned in support of water policies in general, and groundwater policy in particular,4 the most relevant example of which is the BRIDGE project which developed the scientiﬁc basis for the development of a methodology to establish groundwater threshold values (see Chapter 9.1).

11.3.3.5

Orientations of the 7th Framework Programme

FP6 is now about to be terminated, and will be continued by the 7th Framework Programme (FP7) which began on 1 January 2007 and will run until the end of 2013. While FP6 was the Commission’s response to the requirements of the Lisbon summit in March 2000 calling for a better use of European research by creating an internal market for science and technology (the European Research area), FP7 is designed to build on the achievements of its predecessor and to move forward in the creation of a European knowledge economy and society. FP7 is to respond to Europe’s employment needs, competitiveness and quality of life. FP7 (formally adopted by the European Parliament and the Council on the 18 December 2006) covers priority areas reﬂecting EU research needs in sectors

700

Chapter 11.3

such as health, food and agriculture, information and communication technologies, nanosciences, energy, transport, socioeconomic sciences, space and security. Environment and climate change is one of these ten priorities. It focuses on knowledge on the interactions between the biosphere, ecosystems and human activities, and the development of new technologies, tools and services, with emphasis on: improved understanding and prediction of climate, earth and ocean systems changes; tools for monitoring, prevention and mitigation of environmental pressures and risks; and management and conservation of natural resources. More speciﬁcally, the research areas will address pressures on environment and climate, impacts and feedback, environment and health, conservation and sustainable management of natural resources (including groundwater), evolution of marine environments, environmental technologies, understanding and prevention of natural hazards, forecasting methods and assessment tools, and earth observation. The overall environment (including climate change) theme has a budget of h1890 million for the period 2007–2013 (on a total budget of h50 521 million).

11.3.3.6

LIFE: Demonstration Projects

LIFE stands for L’Instrument Financier pour l’Environnement or the Financial Instrument for the Environment.6 It is a European Commission ﬁnancial mechanism speciﬁcally aimed at assisting the development of environmental policy through its co-ﬁnance of demonstration projects proposed from within EU member states or certain other countries. The rationale behind this instrument is simply that innovation, be it highly technical or more like a new approach to an old problem, needs to be demonstrated to persuade other potential users of its value, and to establish that any innovations proposed actually do work in the real world. About 30% of the approximately 2500 LIFE projects co-ﬁnanced to date have water resource management within them and some are focused on ﬁnding eﬀective ways to implement water-related directives. This funding instrument thus represents in principle a natural continuation of research projects aiming at demonstrating the applicability of innovative methods, solutions and techniques on real environmental cases.

11.3.4

An Operational Web Interface: WISE-RTD

11.3.4.1

The Harmoni-CA Initiative

Harmoni-CA is a large-scale concerted action supported by the DG-RTD under FP5.5 One of the objectives of Harmoni-CA is to create a forum for

Towards a Science–Policy Interface (WISE-RTD)

701

communication, information exchange and harmonisation of information communication and technology (ICT) tools for integrated river basin management and the implementation of the WFD (see Chapter 4.3 for further details). Harmoni-CA’s role is twofold: (1) to facilitate speciﬁc activities as identifying and enhancing complementarities between diﬀerent research projects and disseminating research results, focussing on the project in the EC-supported CatchMod modelling cluster and (2) to bring together the demand and support for ICT tools and methodologies for the implementation of the WFD. Recently, a closer cooperation has been proposed between RTD and those responsible for implementing the WFD,4 in particular the following activities have been proposed and developed from 2004 to 2006: linking WFD requirements and RTD products; building of a web portal; and establishing a close cooperation within pilot river basins. The web portal primarily aimed to focus on operational managers (people responsible for practical aspects of policy implementation) and research and technology providers (which is the so-called second level as mentioned in Section 1.2). It developed into a wider information platform as described in Section 4.2.

11.3.4.2

The WISE-RTD Web Portal

The WISE-RTD has been conceived as a platform for accessing scientiﬁc information of potential use to water policy implementation.3 It will progressively be enlarged to cover speciﬁc scientiﬁc information for policy oﬃcers, RTD managers and scientiﬁc stakeholders, providing access to relevant scientiﬁc information (the third level described in Section 1.2). The web portal will be supported by personnel through the Communication Services Centre (CSC) and it can be seen as a ﬁrst step towards a sustainable communication process (see Chapter 4.3). The portal has been made publicly available along with the launching of WISE in March 2007, and will continuously develop in forthcoming years.

11.3.5

Conclusions: Needs for an Overall Science–Policy Integration Framework

This chapter highlights the needs for integration at various levels for a proper understanding and implementation of water policies, with focus on science– policy integration and groundwater policy. Diﬃculties experienced to date stem from the fact that there is no suﬃcient streamlining of information from, for example, the scientiﬁc community to policy decision-makers. In this respect, eﬀorts are ongoing in the framework of various initiatives to examine how an eﬃcient and operational ‘‘science–policy interface’’ could be developed in support of the implementation of the WFD (hence of direct interest to

702

Chapter 11.3

groundwater policy). This development is being undertaken in the framework of the Common Implementation Strategy of the WFD10 (see Chapter 4.1) and in close cooperation with the Harmoni-CA initiative5 which has elaborated the WISE-RTD prototype (considered to be one of the elements of the ‘‘science– policy’’ puzzle). The ultimate aim is to develop such an interface in a way that it could meet the demand of diﬀerent levels of users (policy-makers, industry, etc.) and stakeholders (the scientiﬁc community, academia, etc.), ensuring an eﬃcient dissemination and use of research results. A workshop was held in Ghent on 4–5 October 2004 which gathered representatives of the EU member state environment ministries and agencies, coordinators of research, development and demonstration projects and European Commission oﬃcials, the main conclusions of which were published in the open literature,1 Workshop discussions had clearly identiﬁed the ‘‘missing link’’ corresponding to an operational ‘‘science–policy interface,’’ which would allow result outputs to eﬃciently ﬂow into the policy-making process as discussed throughout the present chapter. This interfacing goal is ambitious and involves many diﬀerent actors: hence its complexity. It should also be seen in a wider framework, in liaison with other parallel activities. One of them concerns coordination of national research programmes (the so-called ERA-NET initiative; see Section 3.3). In this respect, Figure 11.3.4 illustrates a possible framework linking water policies in a broad sense to the R&D life cycle. This includes research development (links to FP7 and national research), demonstration (testing of R&D outputs in the framework of LIFE projects or projects funded with regional funds, or INTERREG), communication (through the WISE-RTD web portal) and policy review (taking policy-related research needs into account when establishing research priorities). At present, the diﬀerent parts of this diagram are not fully coordinated; the challenge will be to establish operational links so that RESEARCH EU-wide: FP7 National: ERA-NET

DEVELOPMENT

POLICY Review, Integration, Research needs

IMPLEMENTATION

DESIGN

Water policies REVIEW

INTERFACE WISE-RTD

Figure 11.3.4

Interfacing mechanism.

DEMONSTRATION LIFE, INTERREG

Towards a Science–Policy Interface (WISE-RTD)

703

the diﬀerent pieces of this puzzle constitute a general interfacing mechanism at the horizon of the FP7 (2006–2013).

References 1. Proceedings of the workshop on research and technology integration in support of the Water Framework Directive, Environ. Sci. Pol., 2005, 8(3). 2. Water Information System for Europe (WISE), http://ec.europa.eu/ environment/water/pdf/concept_report.pdf. 3. W. de Lange, G. Arnold, P. Willems, F. Provost, F. Hatterman, J. Plyson, M. Mestdagh, P. Swartenbroekx, P. Balabanis and Ph. Quevauviller, WISE-RTD webportal: a gate to scientiﬁc information for WFD implementers and water managers, poster presented at the International Conference on Monitoring Under the WFD, Lille, March 2007. 4. Ph. Quevauviller, P. Balabanis, C. Fragakis, M. Weydert, M. Oliver, A. Kaschl, G. Arnold, A. Kroll, L. Galbiati, J. M. Zaldivar and G. Bidoglio, Environ. Sci. Pol., 2005, 8, 203. 5. Harmonised Modelling Tools for Integrated River Basin Management, Harmoni-CA, EU-funded concerted action, contract EVKI-2001-00192 (www.Harmoni-CA.info). 6. LIFE programme, European Commission, DG Environment (http://europa. eu.int/comm/environment/life/life/index.htm). 7. J. D’Eugenio, P. Haastrup, S. Jensen, A. Wirthmann and Ph. Quevauviller, General introduction to WISE, 7th International Conference on Hydroinformatics, Nice, September 2006. 8. WISE concept paper, European Commission (EUROPA web page: http:// europa.eu.int/comm/environment/water/pdf/concept_report.pdf). 9. B. Schmitz, P. Reiniger, H. Pero, Ph. Quevauviller and M. Warras, Europe and Scientiﬁc and Technological Cooperation on Water, European Commission, Report EUR 15645 EN, 1994 (ISBN 92-826-6464-3). 10. Common Implementation Strategy for the Water Framework Directive, European Communities, 2003 (ISBN 92-894-2040-5). Final CIS document available at: http://europa.eu.int/comm/environment/water/water-framework/ implementation.html.

Appendices Appendix I Main provisions of the Water Framework Directive 2000/60/EC ‘‘establishing a framework for Community action in the ﬁeld of water policy’’ with focus on groundwater.

Recitals The directive contains a total of 53 recitals, which establish the main principles for water protection and recalls the water policy ‘‘history’’ which led to the adoption of a wide water policy framework. A series of recitals make explicit reference to groundwater (others also concern groundwater in a more generic way). Recital 3 makes reference to the 1991 declaration of the ministerial seminar on groundwater and to the Groundwater Directive 80/68/EEC. Recitals 4 and 5 highlight the basic arguments which led to the decision for developing the Water Framework Directive (WFD), including groundwater. The ‘‘polluter pays’’ principle is recalled in Recital 11. Recital 23 underlines that common principles are needed to coordinate European Union member states’ eﬀorts to improve the protection of Community waters in terms of quantity and quality. Environmental objectives for both surface and ground waters are referred to in Recital 25, with requirements for programmes of measures to achieve good status objectives highlighted in Recital 26. The need to take early preventive actions to protect groundwater and to reverse any pollution trends appears in Recital 28. Hydrological considerations related to surface and groundwater interactions for the purposes of environmental protection are recalled in Recital 34. Other recitals also directly concern groundwater, e.g. river basin management plans (Recital 35), characterisation of water bodies (Recital 36) and drinking water abstraction (Recitals 37 and 41). Article 1 ‘‘Purpose’’ establishes the framework for the protection of inland surface waters, transitional waters, coastal waters and groundwater, with requirements for prevention of deterioration, promotion of sustainable water use, reduction of pollution, and contribution to mitigating eﬀects of ﬂoods and droughts. Article 2 ‘‘Deﬁnitions’’ includes several deﬁnitions directly relevant to groundwater, i.e. ‘‘groundwater’’ (Deﬁnition 2), ‘‘aquifer’’ (Deﬁnition 11), ‘‘body of 704

Appendices

705

groundwater’’ (Deﬁnition 12), ‘‘groundwater status’’ (Deﬁnition 19), ‘‘good groundwater status’’ (Deﬁnition 20), ‘‘good groundwater chemical status’’ (Deﬁnition 25), ‘‘quantitative status’’ (Deﬁnition 26), ‘‘available groundwater resource’’ (Deﬁnition 27), ‘‘good quantitative status’’ (Deﬁnition 28), ‘‘pollutant’’ (Deﬁnition 31), ‘‘direct discharge to groundwater’’ (Deﬁnition 32) and ‘‘pollution’’ (Deﬁnition 33). All these deﬁnitions are obviously not repeated in the ‘‘daughter groundwater directive’’ (see Appendix II). Article 3 ‘‘Coordination of administrative arrangements within river basin districts’’ deals with the identiﬁcation of river basins (in which groundwaters that do not fully follow a particular river basin have to be assigned to the nearest or most appropriate river basin district) of competent authorities. This also includes the obligation for member states to establish international river basins and to ensure appropriate coordination with relevant non-member states for river basin districts extending beyond the territory of the Community. The list of competent authorities had to be reported to the Commission in 2004. Article 4 ‘‘Environmental objectives’’ sets principles to achieve ‘‘good status’’ for surface and ground waters. Provisions relevant to groundwater are found in paragraph 1(b) of this article. They concern requirements for member states to implement measures to prevent or limit the input of pollutants into groundwater and to prevent deterioration of the status of groundwater bodies, and obligations to protect, enhance and restore all bodies of groundwater, ensure a balance between abstraction and recharge, with the aim of achieving good groundwater status by 2015. This article also requests measures necessary to reverse any signiﬁcant and sustained upward pollution trends. Important exemption clauses have to be considered to get a complete picture of the regulation, in particular paragraph 4 regarding possible extension of the 2015 deadline if justiﬁed by arguments related to technical feasibility, cost disproportionality or natural conditions. In addition, less stringent objectives may be deﬁned by member states if this is justiﬁed by environmental or socioeconomic considerations. Temporary deterioration due to natural cause or force majeure (e.g. accidental pollution) may also result in an exemption. All the exemptions, however, are prone to veriﬁcations and are not synonym for ‘‘no actions’’ regarding environmental objectives. This is clearly expressed in the various provisions of Article 4. Article 5 ‘‘Characteristics of the river basin district, review of the environmental impact of human activity and economic analysis of water use’’ requests member states to carry out an analysis of river basin characteristics, a review of the impact of human activity (analysis of ‘‘pressures and impacts’’) and an economic analysis of water use, following speciﬁcations set out in Annexes II and III of the directive. These analyses and reviews have to be reviewed, and if necessary updated, in 2013 and every six years thereafter. Article 6 ‘‘Register of protected areas’’ requires member states to establish a register or registers of areas requiring special protection under speciﬁc Community legislation for the protection of their surface and ground waters or for the conservation of habitats and species directly depending on water. The protected areas are listed in Annex IV of the directive. The register(s) had to be

706

Appendices

produced by the end of 2004; they have to be kept by member states and reviewed/updated. Article 7 ‘‘Waters used for the abstraction of drinking water’’ sets out provision for the establishment of ‘‘protected areas for drinking water’’ which have to be protected and monitored following speciﬁc requirements indicated in Annex V of the directive. The requirement for achievement of Article 4 objectives (for protected areas) is recalled, and cross-references to pollution prevention and drinking water quality requirements (link to the Drinking Water Directive 80/778/EEC amended by Directive 98/83/EC) are inserted. Member states are required to ensure that protection will be such as to reduce the level of puriﬁcation treatment required in the production of drinking water. Linked to this, safeguard zones may be established. Article 8 ‘‘Monitoring of surface water status, groundwater status and protected areas’’ establishes requirements for the monitoring of water status (chemical and quantitative status for groundwater), requesting the programmes to be operational by the end of 2006 and developed in accordance with technical speciﬁcations set out in Annex V.2 of the directive. This article opens the possibility to lay down technical speciﬁcations and standardised methods for analysis and monitoring that may be adopted by comitology. Article 9 ‘‘Recovery of costs for water services’’ requests member states to take account of the principle of recovery of the costs of the water services, including environmental and resource costs according to Annex III of the directive, and in accordance in particular with the polluter pays principle. By 2010, water pricing policies shall provide adequate incentives for users to use water resources eﬃciently, thereby contributing to the environmental objectives of the directive. The diﬀerent uses will have to be considered, disaggregated into at least industry, households and agriculture. Water pricing shall take into consideration social, environmental and economic eﬀects of the recovery as well as the geographic and climatic conditions of the region or regions aﬀected. Funding of particular preventive or remedial measures is allowed under this article. Finally, the cost recovery and contributions by various water uses will have to be reported in the river basin management plans (unless a clear justiﬁcation for not doing so is provided). Article 10 ‘‘The combined approach for point and diﬀuse sources’’ requires member states to regulate discharges from point and diﬀuse sources into surface water, based on emission controls, emission limit values, taking account of best available techniques and, as appropriate, best environmental practices. It is, therefore, not an article tuned to groundwater for which point and diﬀuse sources are regulated under other mechanisms (in particular Directive 2006/ 118/EC; see Appendix II). Article 11 ‘‘Programme of measures’’ concerns the establishment of a programme of measures in each river basin district or part of an international river basin district falling in a member state territory, taking into account the results of the analysis carried out under Article 5 of the directive, in order to achiever Article 4 objectives. Each programme shall include ‘‘basic’’ measures, i.e. protection measures against pollution required under the legislation speciﬁed

Appendices

707

in Article 10 and Annex VI of the directive, measures to promote an eﬃcient and sustainable water use, measures related to drinking water abstraction (reference to Article 7), controls including requirements for prior authorisation of artiﬁcial recharge or augmentation of groundwater bodies, controls of point and diﬀuse sources of pollution based on prior regulation and prior authorisation or registration based on general binding rules. Speciﬁc rules are established in relation to prohibition of direct discharges of pollutants into groundwater, with possible exemptions linked to injection/reinjection of groundwater resulting from hydrocarbon or mining activities, civil engineering works, etc., providing that the discharges do not compromise the environmental objectives of the directive. The article also includes requirements for measures to prevent signiﬁcant losses from technical installations and impacts of accidental pollution incidents, for example as a result of ﬂoods. ‘‘Supplementary measures’’ complement basic measures and are listed in Part B of Annex VI of the directive. The article requests that causes of possible failure to achieve environmental objectives are investigated, relevant permits and authorisations are examined and reviewed as appropriate, monitoring programmes are reviewed as appropriate and additional measures are implemented, subject to Article 4(6) provisions (natural cause or force majeure). The programmes of measures have to be established by the end of 2009 and made operational by the end of 2012. They shall be reviewed, and if necessary updated, by the end of 2015 and every six years thereafter. Article 12 ‘‘Issues which cannot be dealt with at member state level’’ allows a member state to report to the Commission an issue which has an impact on water management but cannot be resolved by that member state. Recommendations to resolve it may be formulated by the Commission and any other member state concerned. Article 13 ‘‘River basin management plans’’ sets out binding rules concerning river basin management planning (RBMP) in each river basin district, with requirements for coordination within international river basin districts (including those extending beyond the boundaries of the Community), in accordance with Annex VII of the directive. The plans may be supplemented by more detailed programmes for sub-basin, sector, issue or water type to deal with particular aspects of water management. The ﬁrst RBPM has to be published at the end of 2009, and reviewed/updated by the end of 2015 and every six years thereafter. Article 14 ‘‘Public information and consultation’’ requests member states to encourage the active involvement of all interested parties in the implementation of the directive, in particular the production, review and updating of RBMPs. In this context, a timetable and work programme has to be communicated to the public, including users, three years before the RBPM (2006 for the ﬁrst plan), an overview of the signiﬁcant water issues two years before (2007 for the ﬁrst plan) and draft copies of the plan one year before (2008 for the ﬁrst plan). Article 15 ‘‘Reporting’’ requests member states to send copies of the RBMPs and all subsequent updates to the Commission and any other member states concerned within three months of their publication. Member states have also to

708

Appendices

report analyses carried out under Article 5 and monitoring programmes designed under Article 8 in relation to RBMP requirements. An interim report is requested of member states within three years of publication of each RBPM (the ﬁrst review will hence be in 2012) to describe the implementation of the planned programme of measures. Article 16 ‘‘Strategies against pollution of water’’ is focused on surface water, in particular requesting the adoption of speciﬁc measures regarding individual pollutants or groups of pollutants presenting a signiﬁcant risk to or via the aquatic environment, more speciﬁcally the progressive reduction of those pollutants, and cessation or phasing out of discharges, emissions and losses of priority hazardous substances. The article forms the basis for the proposal for a daughter directive (in negotiation in 2007, expected to be adopted in 2008) setting environmental quality standards for priority substances and control measures related to priority substances (inter alia emission controls), as well as reviewing Annex IX of the directive. This adoption follows the adoption of the list of priority substances published in 2001, which has to be reviewed under the new directive and at least four years thereafter, and proposals have to be formulated as appropriate. Under the new directive, the list of priority substances will become Annex X of the WFD. Article 17 ‘‘Strategies to prevent and control pollution of groundwater’’ forms the basis for the adoption of the daughter directive 2006/118/EC, which is summarised in Appendix II. Article 18 ‘‘Commission report’’ requires the Commission to publish a report on the implementation of the directive by the end of 2012 and every six years thereafter, submitting it to the European Parliament and Council. An interim report has also to be published (also to be communicated to the European Parliament and Council) following three years of the publication of each report. It also requests the Commission to publish a report on progress in implementation of Articles 5 and 8 at the latest two years after the dates referred to in those articlesw. The article asks the Commission to convene, when appropriate, in line with the reporting cycle, a conference of interested parties on Community water policy from each of the member states to comment on the Commission’s implementation reports and to share experiences, involving a range of participants. Article 19 ‘‘Plans for future Community measures’’ requests the Commission to prepare information addressed to the committee referred to in Article 21 on an indicative plan of measures having an impact on water legislation which it intends to propose in the near future. Such information had to be presented to the committee in 2003. The article also requires the Commission to review the directive by the end of 2019. Article 20 ‘‘Technical adaptations to the directive’’ opens the possibility to adapt Annexes I, III and section 1.3.6 of Annex V of the directive by w

The Commission report on Article 5 has been published and presented at the European Water Conference held in Brussels on 22–23 March 2007. The report on Article 8 will be prepared for publication in 2009.

Appendices

709

comitology, taking into account the periods for review and updating of the RBMPs. Where necessary, the Commission may adopt guidelines on the implementation of Annexes II and V in accordance with comitology rules. Finally, technical formats may be adopted for the purpose of transmission and processing of data, including statistical and cartographic dataz. Article 21 ‘‘Regulatory committee’’ sets out the comitology principle for the Commission had to adopt rules of procedure (which were adopted in 2003). Article 22 ‘‘Repeals and transitional provisions’’ provides a list of directives which have to be repealed by the end of 2007 (Directive 75/440/EEC on surface water for drinking water abstraction, Directive 77/95/EEC on exchange of information on surface freshwater quality, Directive 79/869/EEC on sampling and analysis of surface water intended for drinking water abstraction), while other directives will have to be repealed in 2013 (Directive 78/659/EEC on the quality of freshwater for ﬁsh life, Directive 79/923/EEC on shellﬁsh water, Directive 80/68/EEC on groundwater protection against pollution, Directive 76/ 464/EEC with the exception of Article 6 which has been repealed with the adoption of the WFD). Transitional provisions apply to Directive 76/464/EEC (e.g. list of priority substances replacing the list published in 1982), and provisions concern environmental quality standards and possible addition of priority substances in Annex VIII of the directive or Article 10 of Directive 96/61/EC. Article 23 ‘‘Penalties’’ concerns penalties applicable to breach of the national provisions adopted pursuant to this directive. Article 24 ‘‘Implementation’’ requests member states to transpose the directive into national laws, regulations and administrative provisions by the end of 2003. Annex I ‘‘Information required for the list of competent authorities’’ sets out elements to be considered for the reporting of list of competent authorities to the Commission (e.g. name and address, geographical coverage of the river basin district, legal status, responsibilities, membership, international relationships). Annex II provides criteria for characterising surface and groundwater bodies. Part 1 concerns surface waters and will not be detailed here. Groundwater is covered by Part 2, in particular the initial characterisation (location and boundaries of groundwater bodies, pressures aﬀecting them, general character of overlying strata, bodies having direct links with dependent aquatic and terrestrial ecosystems). The further characterisation has to detail geological and hydrogeological characteristics, characteristics of the superﬁcial deposits and soils in the catchment from which the groundwater body receives its recharge, stratiﬁcation characteristics, inventory of associated surface systems, ﬂow and exchange rates between ground and surface water, recharge rate and characterisation of the chemical composition, including inputs from human activities. It also covers a review of the impact of human activity on groundwaters (e.g. location of points in the groundwater body used for the abstraction of water and annual abstraction rates, chemical composition of abstracted water, information on discharges, land use in the catchment from which the groundwater z

An example is the development of the Water Information System for Europe (WISE).

710

Appendices

body receives its recharge), a review of the impact of changes in groundwater levels and a review of the impact of pollution on groundwater quality. Annex III ‘‘Economic analysis’’ contains related to information for the economic analysis, including relevant calculations for the principle of cost recovery of water services, and about the most cost-eﬀective combination of measures in respect to water uses to be included in the programme of measures. Annex IV ‘‘Protected areas’’ lists types of protected areas, e.g. areas for drinking water abstraction, areas designated for the protection of economically signiﬁcant aquatic species, recreational waters (including bathing waters), nutrient-sensitive areas (vulnerable zones under Directive 91/676/EEC and sensitive areas under Directive 91/271/EEC) and areas for the protection of habitats and species. Annex V contains provisions on monitoring (Part 1 for surface water, Part 2 for groundwater). Only Part 2 is described here. It covers deﬁnitions of quantitative and chemical groundwater status, and provisions to be considered for the design of monitoring networks, e.g. representativeness of monitoring points, density, frequency, groundwater ﬂows, impacts of abstractions and discharges on groundwater level, concentrations of pollutants, conductivity. Monitoring has to be designed so as to detect the presence of long-term anthropogenically upward trends in pollution, and should take account of the analysis of pressures and impacts carried out under Article 5 of the directive. Groundwater has to be monitored on the basis of a surveillance monitoring programme (supplementing the impact assessment procedure and providing information for trend studies) addressed to groundwater bodies identiﬁed as being at risk and bodies which cross a member state territory (core parameters are oxygen content, pH value, conductivity, nitrate and ammonium, but parameters indicative of identiﬁed pressures should also be be monitored). Transboundary water bodies have also to be monitored for parameters that are relevant for the protection of all the uses supported by the groundwater ﬂow. Operational monitoring has to be undertaken in the periods between surveillance monitoring programmes in order to establish the chemical status of groundwater bodies determined as being at risk and to establish the presence of any long-term anthropogenically induced upward trends of pollution. Annex VI provides lists of measures to be included in the programme of measures. Part A covers eleven directives, namely the Bathing Water Directive (76/160/EEC), the Drinking Water Directive (80/778/EEC amended by Directive 98/83/EC), the Major Accidents (Seveso) Directive (96/82/EC), the Environmental Impact Assessment Directive (85/337/EEC), the Sewage Sludge Directive (86/278/EEC), the Urban Waste-water Treatment Directive (91/271/ EEC), the Plant Protection Products Directive 91/414/EEC), the Nitrates Directive (91/676/EEC), the Habitats Directive (92/43/EEC) and the Integrated Pollution Prevention Control Directive (96/61/EC). Part B provides a nonexclusive list of supplementary measures which member states may choose to adopt in RBMPs, namely legislative instruments, administrative instruments, economic or ﬁscal instruments, negotiated environmental instruments, emission controls, codes of good practice, restoration of wetland areas, abstraction

Appendices

711

controls, demand management measures, eﬃciency and reuse measures, construction projects, desalination plants, rehabilitation projects, artiﬁcial recharge of aquifers, educational projects, research, development and demonstration projects and other relevant measures. Annex VII ‘‘River basin management plans’’ details elements to be covered by river basin management plans, in particular a general description of the characteristics of the river basin district including mapping and location of water bodies, a summary of signiﬁcant (point and diﬀuse) pressures and impacts of human activity on water status, mapping of protected areas, a map of monitoring network, a list of environmental objectives, a summary of the economic analysis of water use, a summary of the programmes of measures, more detailed programmes and management plans for sub-basins, sectors, issues and water types, a summary of public information and consultation measures, a list of competent authorities and contact points for obtaining background information. Annex VIII ‘‘Indicative list of the main pollutants’’ gives a list of 12 main pollutants, namely (1) organohalogen compounds and substances which may form such compounds in the aquatic environment, (2) organophosphorus compounds, (3) organotin compounds, (4) substances and preparations, or the breakdown products of such, which have been proved to possess carcinogenic or mutagenic properties or properties which may aﬀect steroidogenic, thyroid, reproduction or other endocrine-related functions in or via the aquatic environment, (5) persistent hydrocarbons and persistent and bioaccumulable organic toxic substances, (6) cyanides, (7) metals and their compounds, (8) arsenic and its compounds, (9) biocides and plant protection products, (10) materials in suspension, (11) substances which contribute to eutrophication (in particular nitrates and phosphates) and (12) substances which have an unfavourable inﬂuence on the oxygen balance (and can be measured using parameters such as BOD, COD, etc.). Annex IX ‘‘Emission limit values and environmental quality standards’’ covers limit values and quality objectives established under diﬀerent directives, namely the Mercury Discharges Directive (82/176/EEC), the Cadmium Discharges Directive (83/513/EEC), the Mercury Directive (84/156/EEC), the Hexachlorocyclohexane Discharges Directive (84/491/EEC) and the Dangerous Substances Discharges Directive (86/280/EEC). Annex X lists the priority substances adopted under the daughter directive developed under Article 16 of the directive. Annex XI gives maps of ecoregions found in the European Union for rivers and lakes (Map A) and transitional waters and coastal waters (Map B).

Appendix II Main provisions of the Groundwater Directive 2000/118/EC ‘‘on the protection of groundwater against pollution and deterioration’’.

712

Appendices

Recitals The directive contains a total of 25 recitals, which recall that groundwater is a valuable natural resource and as such should be protected from deterioration and chemical pollution, and introduce the diﬀerent regulatory provisions of the directive. Provisions which are not covered by operational articles include considerations about links with the Common Agricultural Policy, naturally high concentrations of substances not considered as pollution, needs to use reliable and comparable methods for groundwater monitoring and needs to undertake research on groundwater ecosystem quality and protection. Article 1 ‘‘Purpose’’ establishes speciﬁc measures as provided by Article 17(1) and (2) of Directive 2000/60/EC, in particular criteria for the assessment of good groundwater chemical status, and for the identiﬁcation and reversal of signiﬁcant and sustained upward trends, and complements the provisions on prevention and limitation of inputs of pollutants into groundwater. Article 2 ‘‘Deﬁnitions’’ does not repeat WFD deﬁnitions, and includes new deﬁnitions about groundwater quality standards, threshold values, signiﬁcant and sustained upward trend, input of pollutants into groundwater, background level (corresponding to no or only very minor anthropogenic alterations to undisturbed conditions) and baseline level (average value measured at least during the references years 2007 and 2008). Article 3 ‘‘Criteria for assessing groundwater chemical status’’ clariﬁes groundwater chemical status criteria which are based on groundwater quality standards (in reference to Annex I), threshold values to be established by member states at the most appropriate level (following recommendations of Annex II), coordinated among member states (or with a non-member states) for groundwater bodies crossing the boundary of a member state, by the end of 2008. The threshold values shall be published in the ﬁrst RBPM in 2009 (including information arising from Part C of Annex II). Amendments to the list of threshold values, whenever new information on pollutants are available, are possible in the context of the periodic review of the river basin management plans. The article also requests the Commission to publish a report on threshold values by the end of 2009. Article 4 ‘‘Procedure for assessing groundwater chemical status’’ speciﬁes provisions to be followed up for assessing groundwater chemical status, including considerations of monitoring data, and compliance to groundwater quality standards (and threshold values), opening ﬂexibility with regard to exceedance values (calling for investigation and risk assessment studies to check the impact of the exceeding point(s) on the overall status of an impacted groundwater body). The article requests member states to publish a summary of the groundwater chemical assessment in the RBMP. Article 5 ‘‘Identiﬁcation of signiﬁcant and sustained upward trends and the deﬁnition of starting points for trend reversals’’ requests member states to identify any signiﬁcant and sustained upward trends in concentrations of pollutants or indicators of pollution in groundwater bodies identiﬁed as being at risk pursuant to the analysis of pressures and impacts carried out under Article 5 of the

Appendices

713

WFD. Trend reversal shall be carried out in accordance with Annex IV requirements for trends presenting a risk of harm to the quality of aquatic or terrestrial ecosystems, human health or actual or potential legitimate uses of the water environment, through the WFD programme of measures. The starting point for trend reversal is deﬁned as percentage of the level of groundwater quality standard or threshold values in accordance with Part B of Annex IV of the directive. Trend assessments have to be reported in RBMPs, explaining the reasons for the starting points. The article also covers impact assessment of existing plumes of pollution resulting from point sources and contaminated land to verify that plumes do not expand and do not represent a risk for the chemical status of bodies or groups of bodies of groundwater (these assessments shall be reported in the RBPM). Article 6 ‘‘Measures to prevent of limit inputs of pollutants into groundwater’’ establishes provisions for preventing or limiting inputs of pollutants into groundwater, based on the programme of measures of the WFD. In this context, inputs of any hazardous substances have to prevented (taking account of exemption clauses), in particular pollutants referred to in points 1 to 6 of Annex VIII of the WFD and hazardous substances of points 7 to 9; for these latter, circumstances for which substances are considered hazardous or non-hazardous have to be identiﬁed, in particular for metals and their compounds. For pollutants which are not considered hazardous, inputs shall be limited in order to ensure that they do not cause deterioration or signiﬁcant and sustained upward pollution trends, using measures taking into account, at least, established best practices. A series of exemptions is included, referring to WFD provisions on direct discharges, technical feasibility, results of accidents or exceptional circumstances of natural cause, results of artiﬁcial recharge, as well as exemptions related to small pollutant concentrations and results of interventions in surface water management, providing that the exemptions are used only when eﬃcient monitoring of groundwater bodies is carried out. An inventory of such exemptions shall be kept by member states. Article 7 ‘‘Transitional arrangements’’ ensures that groundwater chemical status provisions have to be taken into account in the implementation of Directive 80/68/EEC from 2009 until its repeal in 2013. Article 8 ‘‘Technical adaptations’’ opens the possibility to amend Parts A and C and Annexes III and IV in the light of scientiﬁc and technical progress by comitology, taking into account the period of reviewing and updating RBMPs. Part B of Annex II may also be amended by comitology in order to add new pollutants. Article 9 ‘‘Committee procedure’’ speciﬁes that the Commission shall be assisted by a committeey. Article 10 ‘‘Review’’ requests the Commission to review Annexes I and II by 2013 and thereafter every six years. Based on the review, it shall come forward with legislative proposals to amend Annex I and/or II. y

It will be the same as the one referred to in Article 21 of the WFD.

714

Appendices

Article 11 ‘‘Evaluation’’ requires that the Commission report to be produced under Article 18 of the WFD includes an evaluation of the functioning of this directive in relation to other relevant environmental legislation, including consistency therewith. Article 12 ‘‘Implementation’’ requests member states to transpose the directive into national laws, regulations and administrative provisions by the beginning of 2009. Annex I ‘‘Groundwater quality standards’’ sets out a nitrates standard of 50 mg l1 and a standard of 0.1 mg l1 for active substances in pesticides, including their relevant metabolites, degradation and reaction products, and 0.5 mg l1 for sum of individual pesticides, without prejudice to Directives 91/414/EEC and 98/8/EC. Where groundwater quality standards could result in failure to achieve the environmental objectives for associated bodies of surface water, more stringent threshold values shall be established; this also applies to activities falling within the scope of Directive 91/676/EEC (nitrates). Annex II ‘‘Threshold values for groundwater pollutants and indicators of pollution’’ provides guidelines for the establishment of threshold values by member states in accordance with Article 3 (Part A) as well as a minimum list of pollutants (Part B) for which member states have to consider establishing threshold values (they concern substances or ions or indicators which my occur both naturally and/or as a result of human activities, namely arsenic, cadmium, lead, mercury, ammonium, chloride and sulfate, synthetic substances, namely trichloroethylene and tetrachloroethylene, and conductivity). Part C of this annex summarises information to be provided by member states with regard to the pollutants and their indicators for which threshold values have been established (including extent of interactions between groundwater and associated ecosystems, the size of groundwater bodies at risk, background levels, the way threshold values have been established, etc.). Annex III ‘‘Assessment of groundwater chemical status’’ sets out technical details concerning the assessment procedure to be used for determining the chemical status of a body or group of bodies of groundwater, with reference to outputs from the analysis of pressures and impacts, monitoring data and any other relevant information. Recommendations concerning investigations (e.g. conceptual modelling) are also given. The reporting is based on the WFD provisions of Annex V, sections 2.4.5 and 2.5, complemented by the requirement for member states to report monitoring points where values exceed quality standards, where this is relevant and feasible. Annex IV ‘‘Identiﬁcation and reversal of signiﬁcant and sustained upward trends’’ is made up of two parts. Part A concerns criteria for the identiﬁcation of signiﬁcant and sustained upward trends in groundwater bodies identiﬁed as being at risk, with technical details about the identiﬁcation procedure (e.g. hydrological, chemical and physical characteristics of the groundwater, statistical evaluation, recommendations on monitoring methods and on measurements below the quantiﬁcation limits) and considerations about ‘‘baseline levels’’. Part B sets out criteria for starting points for trend reversals which shall in principle correspond to 75% of the parametric value of groundwater

Appendices

715

quality standards or threshold values, with ﬂexibility opened for member states to set out starting points that are either more stringent (earlier stating point if necessary) or adapted to the rate of increase and trend reversibility (justifying a later starting point). A diﬀerent starting point may also be deﬁned if the detection limit does not allow for establishing the presence of a trend at 75% of the parametric value. Measures are those of the WFD programme of measures and of the Directive 91/676/EEC (for nitrates).

Subject Index abstraction economic weighting, 60–1 natural trends, 212 over-exploitation and, 8, 134, 136, 228 quality impacts, 585 See also sustainable management abstraction–runoff balancing principle, 595 acceptable daily intake (ADI), 100 acetanilide herbicides, 552, 560, 568 acetochlor, 549, 568 acid herbicides, 569 action plans and IMS, 419–20 action programme (EU, 1996), 89–90 “active involvement” requirement, 157–8 actor analyses, 158 adaptive management, 150–70, 474 adaptive regulatory impact assessment, 158 NEWATER project, 698 public participation and, 151–4, 161–9 ADI (acceptable daily intake), 100 adsorption coefficients. See distribution coefficients advection, 426, 432, 439, 447–9 AEMs (agri-environment measures), 496, 500–1, 505–6 aerobic and anaerobic conditions, 431, 527, 554, 594 See also anoxic conditions

age determination case studies, 229–35 definition of groundwater age, 219 geoindicators, 223–5 modelling techniques, 225–6, 231, 642 natural baseline studies and, 208–10 pesticide half-lives from, 552–3 using tracers, 217–23, 229–35 “age gradients”, 208 agri-environment measures (AEMs), 496, 500–1, 505–6 agriBMPwater project, 498, 503, 509 agriculture CIS group on, 126 dissemination of best practices, 462–3 Ebro basin, 41–2 environmentally-acceptable farming, 494–509 farm sector economic model, 71 groundwater pollution from, 12–13, 454–68 incentives for farmers, 497–500 non-commodity outputs, 500, 506–7 public participation, 159–61 Sewage Sludge Directive, 103 Shannon PRB studies, 130–2 simplified representation of, 623 See also fertilisers; irrigation agrochemicals, 198 airborne contaminants, 195–6, 199, 218, 469, 634

Subject Index

alachlor, 372–5, 552–4, 559–61, 568 alarm thresholds, WATCH, 515 aldicarbe, 552, 561–2 alluvial aquifers, 6, 460–1 Alsace, France, 412–17 amitrole, 566–7, 571 AMPA (aminomethyl phosphonic acid), 561, 564, 571 analogous groundwater bodies, 188–9, 439 analytes. See determinands analytical detection limit, 445 analytical laboratories certified reference materials, 383– 92 inter laboratory studies, 380–3, 518–19 methods comparison, 380 quality control inadequacies, 378 sources of error, 380 statistical control, 379–80 Analytical Model, CoronaScreen suite, 446–7 ANCORE network, 27 ANOVA (analysis of variance) method, 397, 399 anoxic conditions, 676, 680 anthropogenic effects groundwater impacts, 3–4, 51–3, 95–6, 136, 198 identifying anthropogenic impacts, 198–9 indicators of, 356, 682 time scale of, 217 Antwerp, Belgium, 412–17 AQUASTRESS project, 698 AQUATERRA project, 31–57 basin studies and sub-projects, 34 conceptual models and, 612, 620–5 Ebro basin case study, 38–43 European regulatory context, 26, 662, 698 Meuse basin case study, 34, 43–51, 620–1, 623 aquatic ecosystems, 9, 16

717 aqueous phase importance of measurements in, 293 sampling uncertainties, 392 transfer rates, 295 See also distribution coefficients; NAPL aquicludes, 6 aquifer heterogeneity, 211, 244 aquifer types and monitoring, 351–2, 364 aquifers alluvial, 679 BaSeLiNe project coverage, 195 biodegradation capacity, 438 BRIDGE classification of, 538 carbonate, 208 chalk, 224, 458–9 contamination below megasites, 409–11 cross-sections, 205, 210 definition of groundwater bodies and, 180 dispersivity, 426, 441–2, 446–51 fractured, 17, 26, 269–88, 361 freshening, 223 geological variability, 5 island aquifers, 360–1 karst, 233, 461, 500 multi-layer, 198, 263, 432 river basin boundaries and, 475 sensitivity, 114 sustainable yield, 76, 586, 601 thickness, 631 three functions of, 675 transmissivity, 589, 632 unsuitable for other purposes, 95 See also artificial recharge; groundwater bodies; transboundary aquifers aquitards, 6, 219, 226, 230 Ardèche river basin, 78 39 Ar studies, 221–2 Århus Convention, 489 ARID project, 697 artesian wells, 5

718

artificial groundwater, 382, 385 artificial recharge initial characterisation and, 126, 183 monitoring requrement, 360, 682 over-exploitation and, 62, regulatory framework for, 87, 96 artificial reference materials, 382 asymmetries of information, 500 At. See atrazine “at risk” groundwater bodies, 183, 187–90 operational monitoring, 348, 353, 361 surveillance monitoring, 355 atmospheric tracers. See airborne contaminants atomic absorption spectrometries, 385–6, 389–90 atrazine (At) biodegradation, 548, 550, 552–4 Brévilles Spring case study, 573–6 deethyl-, 552–4, 559–72, 574–6, 578 deisopropyl-, 554, 561–2, 566–7, 570–2, 575 effect of discontinuing, 465 groundwater contamination, 372, 459, 556–61, 566–7, 571 hydroxy-, 564, 566–7, 571 immunoassay screening), 369, 372–5 attenuation. See natural attenuation Australia, 75, 77 Austrian MTBE levels, 518 authorisation procedures Directive 80/68/EEC, 87–8 Urban Wastewater Treatment Directive, 100 WFD, 95–6 avoidance cost method, 62 background quality values guidelines for determining, 201–2 inorganic components, 200–2 methods of establishing, 198–214, 353 organic components, 202–6 WFD preferred term, 197 backtracking techniques, 254, 264

Subject Index

bacteria. See biodegradation; microbial populations Baltic Sea, 616 BAM (2,6-dichlorobenzamide) modelling transport of, 231 occurrence in EU countries, 560, 562–4, 571 sorption and biodegradation, 550, 553–4 base flows, river, flow modelling and, 633 groundwater ecology and, 683 maintenance role of groundwater, 3 as a sustainability indicator, 600, 603 vulnerability to over-extraction, 134–5, 585, 594–7, 603 baseline chemical composition, 193–215 natural baseline trends, 211–14 tracers and time scales, 208–11 BaSeLiNe project, 194-205 approach, 198–200 conclusions, 214–5 findings on TOC and depth, 205 tracer studies, 208–11 batch shaking tests, 294–9, 302 Bathing Water Directive (Directive 76/160/EEC), 93 Bavaria, 464 Beauce aquifer, 77 behavioural models, 63–4 Belgium contaminated megasites, 412–17 Meuse basin case study, 43–51 pesticide contamination in, 561–4, 571 bentazone, 548, 550–4, 558–67, 569–71 benzamide, 2,6-dichloro-. See BAM benzene, 258–9 biodegradability of, 424, 429–30 defining plume boundaries, 445 plume complexity, 433 See also BTEX compounds benzene, 1,2,4-trimethyl- (1,2,4-TMB), 310 benzenes, chloro-, 263 bequest value, 61

Subject Index

best management practices (BMPs), 495–501 cost-effectiveness grid, 502–9 bio-fuel industry, 72–3 bio-remediation, 423, 681 bioassays, 366, 369, 370, 375 See also immunoassays BioChlor model, 441–2 Biocides Directive (Directive 98/8/EC), 101 BioDAQ data acquisition system, 515 biodegradation of organics acceleration threshold, 551 CO2 evolution from, 310 competitive inhibition, 526–7 contaminant mass loss, 435, 438 contaminant plume effects, 254–5, 259 CoronaScreen assumptions, 447 establishing viability of, 434–9 isotopic fractionation, 329, 333 Kaplan’s scheme, 322–3 MTBE, 524–7 pesticides, 548–9, 550–5, 576–7 redox regimes, 431–2 risk assessment modelling and, 306, 308 See also natural attenuation; weathering biodiversity groundwater ecosystems, 674, 681 Rio treaty (1992), 654 biogeochemistry coupled models, 306 pesticides, 545–79 subsoil, 675 bioindicators, 653, 680–2, 685 biological parameters, geoindicators, 223 biological species. See microbial populations; viruses bioluminescence inhibition, 370 biomarkers, 322, 324 bioremediation, 423, 681 BioScreen model, 441–2, 444, 446 biosensors, 370 “Birds and Habitats Directives,” 685 Bitterfeld, Germany, 412–17

719 black-hole sinks, 676 BMPs (best management practices), 495–501 cost-effectiveness grid, 502–9 Borden test site, 554 Bornholm, Denmark, 587, 591–2, 597 boron, 417 box and whisker plots, 200, 203–4, 572–4 “box models,” 439 Brévilles catchment, 34, 570–6 BRGM (Bureau de Recherches Géologiques et Minières), 393–4, 579 BRIDGE (background criteria for the identification of groundwater thresholds) project case studies, 69, 133–4 EQS establishment, 535, 538, 543 Groundwater Directive support, 27–8 PRBs involved, 130 SSP example, 699 setting threshold values, 97, 126 work packages, 133–4 Britain. See UK British Geological Survey, 195, 469 bromacil, 561–3, 571 bromide, CRMs for, 390–1 BSEM (backscattered electron microscopy), 274 BTEX compounds analytical techniques for, 517–19 co-metabolic MTBE degradation, 527 contaminant characterisation, 322–5, 328, 330 monitoring for, 369 Ringe site, 281, 283 sites contaminated with, 248, 260–3, 330 risk assessment for, 305 building material limit values, 311 bunker C oil, 325 Bureau de Recherches Géologiques et Minières (BRGM), 393–4, 579 buried valleys, 230

720

cadmium, 293–4, 634, 636 calcium, 299–300, 385 calibration errors, 379, 383, 386, 390 calibration graphs, 385–6, 389 calibration of models, 692 DK model, 589–91 Kempen area transport model, 637 CalSim model, 616 canonical correlation coefficients (CCC), 324–5 CAP (Common Agricultural Policy), 456, 462, 509 capillary fringe, 306, 309 capillary number, 271, 277–8 capillary pressure, 275–6, 278 capture zone management, 467 CARACAS network, 649 carbamate herbicides, 552 carbamate insecticides, 465 14 C studies, 220–2, 230, 234, 547–9 13 C studies, 206, 225, 331–3 carbonate media (artificial), 382, 385 case studies age determination, 229–35 AQUATERRA project, 620 contaminated megasites, 412–17 groundwater sampling uncertainty, 392–9 case studies, BRIDGE project Castelporziano protected area, 136–40 Colli Albani, 134–6, 140 Salone-Acqua Vergine area, 136 case studies, economic Krsko, Slovenia, 66–9 Riga, Latvia, 69–71 upper Rhine valley, 64–6, 71–3 case study, pesticides, 570–6 Castelporziano protected area, 136–40 Catalonia, Spain, MTBE levels, 518 catchment areas. See river basins CatchMod projects, 142–6, 697, 701 CEN standards, 302–4, 311 certified reference materials (CRMs), 378, 381, 383–92 CFCs (chlorofluorocarbons), 220–2, 227, 230

Subject Index

chain management, 35 changes. See trends CHCs (chlorinated hydrocarbons), 329–33, 417, 441 chemical logs, 394–5, 397, 399 chemical status Castelporziano protected area, 137– 9 Colli Albani, 135 Groundwater Directive, 97 trend monitoring and, 252–9 WFD, 91, 96, 126 Chile, 76, 616 chloridazon, 561–3, 570 chlorides airborne and groundwater levels, 195–6 groundwater age determination, 234–5 inorganic ion correlation, 213 retardation, 554 chlorinated hydrocarbons (CHCs), 329–33, 417, 441 chlorinated solvents, 305, 335 See also PCE; TCE chloroacetanilide herbicides, 560 chlorobenzenes, 263, 265 chlorofluorocarbons (CFCs), 220–2, 227, 230 chlorpyriphos, 559 chlortoluron (CTU), 561–3, 566–7, 569–73, 575 chromatography liquid chromatography/mass spectrometry, 572 pesticide metabolites, 547, 572 total ion chromatograms (TIC), 517 See also gas chromatography CIS (Common Implementation Strategy, WFD), 122–4 cooperative approaches and, 104 CatchMod/Harmoni-CA links, 145 groundwater characterisation and risk assessment, 177, 186, 191–2 Groundwater Working Group (WG C), 28, 123–7, 346, 490, 543

Subject Index

IMPRESS working group, 186 Pilot River Basins (PRBs) network, 123 Working Groups, 123, 128 See also guidance documents; Monitoring Guidance for Groundwater cities, INCORE project, 317 CityNet projects, 698 CL:AIRE (Contaminated Land: Applications In Real Environments), 451 CLARINET network, 27, 649, 651, 660 clay minerals column leaching tests, 303–4 intraparticle diffusion, 301 clay tills, 270–3, 275, 277–82, 286–8 Clean Water Act (US), 195 cleanup. See remediation climate change adaptive management and, 150, 169 DK Water Resource Model, 596 geoindicators and, 223 global change scenarios, v–vi, 8, 72 groundwater management and, 469, 619 palaeoclimatic information, 225 simulation models and, 615 7th Framework Programme, 700 clopyralid, 564, 570 CLUED’EAU project, 698 clusters contaminated megasites, 410, 414–15, 418–20 EU research projects, 697 coastal locations. See saltwater Colli Albani case study, 134–6, 141 colorimetry, 372, 386 column leaching tests, 293–305 low-volatility components, 295, 300 standardised protocols, 302–4 “comitology” rules, 97 command and control instruments, 496 Common Agricultural Policy (CAP), 456, 462, 509

721 Common Forum, 660 Common Implementation Strategy, WFD. See CIS communication science–policy integration and, 22, 24 See also IC tools; ICT tools; public participation; research; stakeholders community water systems (CWS), 109, 112, 114 complexation of metals, 296, 299 Compliance and Trends activity, 126 compliance monitoring, GWR, 116 compliance reporting, WISE, 693, 696–7 COMPUTE sub-project, AQUATERRA, 34, 56 computer-based modelling. See modelling; simulation concentration–depth profiles, 633, 638–41 conceptual models AQUATERRA case studies, 621, 623 contaminated megasites, 419 CoronaScreen Travelling 1-D model, 449 defined and described, 613–14 deriving environmental thresholds, 536 DK Water Resource Model, 587, 590 groundwater contribution to surface water, 631–4, 642 limitations, 601, 625–6 monitoring design, input to, 347–50, 359, 362, 364 MNA monitoring and, 438, 442 MTBE distribution, 519 quantitative modelling and, 614–17 Ringe site, 273–5 risk assessment and, 184–6, 188 river basin management, 611–26 source-pathway-target model, 422 WATCH project tools, 528–9 conceptual site models (CSMs), 439–40, 451 conductivity levels, 137–40

722

conferences, CONSOIL, 661, 668 conferences, Harmoni-CA, 144–6 confined groundwater, 190, 594 conflicts of interest, 151, 156 conservative solutes, 640 CONSOIL conferences, 661, 668 consolidated materials, 5 Construction Products Directive (Directive 89/106/EEC), 103 consultation legally-binding, 157 Lower Saxony Ministry of Information, 160 See also public participation; stakeholders contaminant attenuation. See natural attenuation contaminant distribution Ebro basin, 42 fractured aquifers, 269–88 IPT inversion uncertainties, 252–3, 255–6 pesticide status, 555–70 pesticide transport, 546–55 contaminant plumes core and fringe controlled, 442–3 development of, 421–2, 427–9, 435 fringe behaviour, 431, 442–3, 446, 448–50 hydrogeology and heterogeneity, 432–3 integral investigation, 241–3, 249 mass loss from biodegradation, 435, 438 maximum lengths, 445, 449–50 monitoring flow path, 433–4, 436–8 plume length statistics, 255, 258 screening to locate sources, 318–20, 330 contaminants classification according to volatility, 293 cross contamination with surface water, 231 distinguished from pollutants, 194

Subject Index

integrated investigation method, 241 mass flux, 427–8, 438–9, 445 release kinetics, 299–302 statistical identification, 323–9 contaminated megasites archetypes, 412–13, 420 case studies, 412–17 categorisation, 408–10 defined, 406 integrated management, 405–20 risk-based approaches, 406–13 contamination aqueous phase emphasis, 293 characterisation at multiple sites, 240 cross contamination with surface water, 231 location and megasite classification, 409 contamination model uncertainty, 251 contingency planning, 529–31 contingent valuation method, 62, 70, 78 control charts, 380, 392 control planes (CP) industrial site application, 257–65 integral groundwater investigation and, 241–9, 253–4 length and cost, 338 monitored natural attenuation, 438 plume screening along, 319 selection, 242, 248 cooperation committees, Lower Saxony, 168 copper, 296, 299–300 CORDIS (Community Research & Development Information Service), 665 corn root worm, 72–3 CORONA approach to natural attenuation, 437, 442–50, 662 CoronaScreen model, 437, 439, 443–50 corrective action requirements, 115–16 correlation assessments, 187 cost-benefit analysis evaluating management options, 59, 62–3, 78, 486

Subject Index

adoption of BMPs, 500–1, 505–6 Latvian case study, 69–71 cost-effectiveness evaluating management options, 59, 62–3, 78 BMP schemes, 504–9 INCORE project, 318–19 IWRM schemes, 501 nitrate pollution prevention, 66–8 risk-based remediation, 422 surveillance monitoring, 356, 362, 365 cost of illness method, 61 cost of reversing a decision, 153 cost recovery for water services, 479 CP. See control planes creosote, 325 critical areas (Nitrates Directive), 502–4 critical path analysis (CPA), 275 CRMs (certified reference materials), 378, 381, 383–92 crop rotation BMP examples, 498 CSMs (conceptual site models), 439–40, 451 CSTREAM program, 244, 246, 263 CTU (chlortoluron), 561–3, 566–7, 569–73, 575 cumulative frequency plots, 200–2 CWS (community water systems), 109, 112, 114 cyanazine, 554, 561, 570 cyclic approach to remediation, 318–37 plume screening, 319–20 remediation strategy, 333–7 source identification, 320–33 cyclopentane, 307 cypermethrin, 569 Czech Republic, 372 D3/C3 weathering ratio, 322–3 D3/P3 source ratio, 322–3 damage costs, 61, 65–6, 506 Danish EPA. See Denmark Danish Geological Survey, 383 Danube river basin, 34, 620–1, 679 Darcy velocity, 245–6

723 data acquisition, WATCH project, 515–16 collection and analysis needs, 25 reduction techniques, 435, 437–9 visualisation, 696 databases EEA Waterbase, 555–8 INCORE project, 324–5, 327 See also WISE dating. See age determination Daugava River, Latvia, 372 DBE (dibromoethene), 430 DCPD (dicyclopentadiene), 516 DDIPU (didemethylisoproturon), 572, 574 decentralisation, 481 decision-based models, 63 decision making, 490, 619–20, 622, 625 decision support systems, 509, 528, 616 decision trees, 255–8 deep recharge, 594–6, 601 deethylatrazine (DEA), 552–4, 559– 72, 574–6, 578 deethylterbuthylazine, 560 degradation. See biodegradation of organics deisopropylatrazine (DIA), 554, 561–2, 566–7, 570–2, 574–5 denitrification, 458, 460–1, 466, 525 Denmark age determination, 229–31 Danish Geological Survey, 383 fractured site risk assessment project, 272 groundwater resource estimates, 598 MTBE toxicity levels, 518 National Water Resource Model, 587–92 nitrate concentrations, 227 nitrate protection zones, 464 dense non-aqueous phase liquids (DNAPL), 270–1, 425 dependent ecosystems good chemical status and, 352

724

groundwater age and, 218, 228–9 groundwater body characterisation and, 180, 183, 187 holistic representation, 683 quantitative monitoring, 359, 585 depth variations concentration–depth profiles, 633, 638 contamination, 598, 354, 368 dissolved oxygen, 394 IPT capture zones, 263 lead concentration, 396–8 microbiology, 675, 677, 679 pesticide mineralisation, 548 pesticide sorption, 550 physicochemical properties, 354 derogation Directive 80/68/EEC, 87 megasite IMS, 416–18 WFD, 92 desalination, 42 desorption kinetics, 576 determinands, CRM certification, 386 determinands for monitoring, 365 DIA (deisopropylatrazine), 554, 561–2, 566–7, 570–2, 574–5 dichlobenil, 550, 562, 564, 570 dichlorprop, 553–4, 561, 564, 571 dicyclopentadiene (DCPD), 516 diffuse pollution sources case studies, 131, 134, 159, 454–6, 572 Groundwater Action Programme and, 90 inclusion in RBMPs, 126, 477 passive sampling applicability, 371 projects examining, 26 reactive transport and, 630–4 risk assessment and identification of, 14, 186 See also agriculture; contaminated megasites diffusion in aquifers, 425–7, 432, 450 diffusive gradients in thin films (DGTs), 371

Subject Index

diisopropyl ether (DIPE), 513, 516, 518 dilution factor (DF) establishing threshold values, 542 dilution in natural attenuation, 425 dilution ratio, 221 dinoterb(e) and dinoseb, 562, 568 DIPE (diisopropyl ether), 513, 516, 518 direct discharge of pollutants, 86–8, 96 Directive 76/160/EEC (Bathing Water Directive), 93 Directive 79/409/EEC (Wild Birds Directive), 93, 685 Directive 80/68/EEC, 25, 85–8, 89, 363, 378, 536 Groundwater Directive and, 97 WFD and, 98 Directive 80/778/EEC (Drinking Water Directive), 10, 99–101, 378 Directive 86/278/EEC (Sewage Sludge Directive), 103 Directive 89/106/EEC (Construction Products Directive), 103 Directive 91/271/EEC (Urban Wastewater Treatment Directive), 11, 49, 93, 99–100 Directive 91/414/EEC (Plant Protection Products Directive), 100–1 Directive 91/676/EEC (Nitrates Directive), 93, 98–9, 378, 456, 463, 503 Directive 91/692/EEC (Standardised Reporting Directive), 123, 694 Directive 92/43/EEC (Habitats Directive), 93, 685, 710 Directive 96/61/EC (Integrated Pollution Prevention and Control (IPPC) Directive), 90, 101–2 Directive 98/8/EC (Biocides Directive), 101 Directive 1998/83/EC (Drinking Water Directive), 136, 373 Directive 99/31/EC (Landfill Directive), 102–3 Directive 2000/60/EC. See Water Framework Directive

725

Subject Index

Directive 2006/118/EC. See Groundwater Directive Directives, EU Priority Substances Directive, 123 “Soil Framework Directive” proposal, 512 status and implementation, 197 disciplines conceptual models and, 614 groundwater management and, 49, 474, vi–vii multidisciplinary teams, 51, 53, 56 pesticide contamination, 576 See also economics discriminant analysis, 324–5 dispersion, attenuation by, 426, 432 dispersivity, 426, 440–2, 446–51 dissolved organic carbon (DOC), 198, 203, 296, 298–9 dissolved oxygen, 394 distributed models, 590 distribution (sorption) coefficients, Kd leaching tests and, 295–6, 299 pesticide sorption and, 547, 551, 553, 555, 567 vapour phase diffusion, 305 See also Freundlich coefficient diuron, 558–9, 561–2, 566–7, 569– 71, 577 DK model, 587–92 DNAPL (dense non-aqueous phase liquids), 270–1, 425 DNOC (2-methyl-4,6-dinitrophenol), 553–4 DOC (dissolved organic carbon), 198, 203, 296, 298–9 domestic effluents, 87 Don catchment, France, 502–8 Doñana aquifer, 204–5, 210 “downstream” economic assessments, 77 DPASV (differential pulse anodic stripping voltammetry), 389–90 DPDP. See dual porosity/dual permeability

DPSIR (drivers-pressure-stateimpact-response) framework, 32 AQUATERRA project, 41, 620, 622 conceptual river basin models, 617–18 drivers and pressures, 39, 41, 45–6, 186–7 soil protection strategy, 649–50 Drinking Water Directives Directive 1998/83/EC, 136, 373 Directive 80/778/EEC, 10, 99–101, 378 Drinking Water Protected Area (DWPA) objectives, 348 drinking water standards (DWSs) assessments, 539–40 drinking water supply contamination costs, 61 gasoline contamination, 518 groundwater quantitative contribution, vi, 109, 345, 671 as priority use, 42, 169 River Meuse use, 48 Salone-Acqua Vergine area, 136 standards as environmental guidelines, 197 vital drinking water areas, 598 drivers and pressures. See DPSIR framework droughts, v–vi, 486 dry-cleaning, 329–30 dual porosity/dual permeability (DPDP) models, 270, 279, 288 dual porosity/single permeability (DPSP) models, 270 Dupuit–Forchheimer assumption, 632 Düsseldorf, Germany, 518–19, 525 Dutch Building Materials Decree, 311 Dutch Environmental Policy Plan 4 (2001), 35 See also Netherlands E. coli, 107, 111, 113, 115 EAF (Expert Advisory Forum) groups, 123

726

early warning systems, biological indicators, 681–5 monitoring schemes for, 17, 486 MTBE pollution, 512, 528–30 soil pollution, 653 East Midlands aquifer, 224–5, 234–5 Ebro basin, Spain, 34, 38–43, 620–1 ecology. See ecosystem effects; groundwater ecology Economic Analysis for the Final Ground Water Rule, 111, 113 economics behavioural models, 63–4 case studies introduced, 64 groundwater management and, 58–81, 319 groundwater models and, 71–3 influencing water use, 59, 73–7, 496 methodologies and tools, 60–4 optimisation models, 615 of remediation, 288, 335–6, 337–40 of risk assessment, 293 social learning and, 161 See also cost ECOSAT model, 311 ecosystem effects, 9–10 benefits and services, 673–5, 677, 684–5 Ebro delta, 41 ecological sustainability, 600 hydromorphological modification, 48 PRB network studies, 129 research needs, 29 surface waters, 584 threshold effects, 500–1 See also dependent ecosystems; groundwater ecology ecotones, 676, 678 ecotoxicity, 540 EDB (dibromoethane), 552 effective medium approximation (EMA), 275 effective transport coefficients, 275–8 EIONET (European Environment Information and Observation NETwork), 556, 695

Subject Index

Elbe river basin, 34, 620–1 electrochemical sensors, 369 electron balance methodology, 436, 438, 446–9 electron transfer in biodegradation contaminant plumes, 427–32 electron acceptors for MTBE, 524–5 establishing and modelling NA, 435–6, 438–43, 445–51 microbial specialisation, 681 electrothermal atomic absorption spectrometry (ETAAS), 385, 389–90 ELIFA (enzyme-linked immunosorbent flow assay), 523 ELISA (enzyme-linked immunosorbent assay), 370, 513, 520–2 endosulfan, 559 England. See UK enteroviruses, 111 Environment Agency, UK, 451, 568 Environment and Development UN Conference, 1992, 613 Environment and Sustainable Development Programme, 697 environmental cost assessments, 61 environmental cost-effectiveness of BMP mitigation, 502 “environmental negotiated agreements”, 479 environmental policies four generations identified, 35–7 groundwater ecology and, 671–86 learning in policy, 53, 55 stakeholder involvement, 121–2, 684 environmental pressures, 32 Environmental Protection Agency, US. See EPA Environmental Quality Standards (EQS) compliance regimes, 543 groundwater status assessments, 539–40 limited applicability to soil, 649–51 methodology for establishing, 535–44 threshold values and, 97, 135, 137, 140

Subject Index

environmental regulations. See regulation environmental taxes and charges, 74–6 environmental thresholds. See threshold values environmental tracers. See tracers enzyme-linked immunosorbent assays. See ELISA EPA-California approach, 560 EPA-PAHs, 260, 299, 301, 322 EPA (United States Environmental Protection Agency) legislation and, 107–8, 110, 195, 197 MTBE advisory limit, 518 See also Ground Water Rule; RAGS epidemiological studies, 113 epikarst, 679 EQS. See Environmental Quality Standards ERA-NET “Plus” module, 648 ERA-NET scheme FP6 projects and, 647–9, 698–9 role in international groundwater research, 654, 669 WISE-RTD and, 702 See also SNOWMAN project errors, 380–1 Estonia, abstraction taxes, 75 ETBE (ethyl tert-butyl ether), 512– 13, 516–18, 523, 525 ethane, dibromo- (EDB), 552 ethene, dibromo- (DBE), 430 ethene, perchloro- (PCE), 329–30, 332, 430 ethyl tert-butyl ether (ETBE), 512–13, 516–18, 523, 525 EU (European Union) agricultural practices and pollution, 454–62 DG ENV, JRC, ESTAT and EEA, 694–5 framework programmes summarised, 697–9 ministerial seminar on groundwater, The Hague, 1991, 85

727 national soil research programmes, 655–6 policy integration, 98 regulatory framework, 86–106 research effectiveness, 647 6th Environmental Action Programme, 23, 121 5th Framework Programme of RTD, 142–3, 240, 313, 406, 451, 512, 697–8 6th Framework Programme of RTD, 33, 43, 146, 148, 368, 535, 647, 685, 698 7th Framework Programme of RTD, 29, 146, 648, 660, 685, 699, 703 Thematic Strategy on Soil Protection, 649, 651 See also Directives; ERA-NET; WISE; individual projects EU Water Initiative, 124 EUGRIS project, 26, 661–2, 665, 668 EUPOL sub-project, AQUATERRA, 34 EURO-LIMPACS project, 698 EURODEMO project, 662 EUROHARP project, 616 Europe national environmental agencies, 559, 562, 564–5, 568 pesticide contamination, 555–70 European Environment Agency (EEA), 555–60, 578, 617 European Environment Information and Observation NETwork (EIONET), 556, 695 European Inventory of Existing Chemical Substances, 287 European Union. See EU eutrophication, 33, 48, 99, 500, 540 EUWATERMAN project, 474 evapo-transpiration, 360, 550, 588–9 Expert Advisory Forum (EAF) groups, 123 exponential models, 232–3

728

fact sheets, 558–60 faecal contamination, 17, 107, 110–11, 113–15 FAO Legislative Study No 86, 2005, ix, 104 farming environmentally-acceptable, 494–509 farm sector economic model, 71 See also agriculture Faro, Portugal, 519 FC (faecal coliforms), 17 Federal Water Pollution Control Act Amendments, 195 feedback mechanisms, policy, 153 fermentation in contaminant plumes, 443, 448 fertilisers groundwater pollution from, 12–13, 455 limiting use of, 67, 462–3, 506–8 potential BMP examples, 498, 506–8 use in England and Spain, 455–6 See also nitrates; sewage Feuerbach district, Stuttgart, 340 Fick’s second law, 305 Field Observatory in Urban Hydrology (OTHU), 682 Filter Backwash Recycling Rule (US EPA), 108 filtration, leaching tests, 303 filtration of pollutants, 6–7 financial resource allocation, 59 See also economics first-order attenuation rates, 251, 255, 262, 427, 440–1 first-order catchments, 615 first-order kinetics, 548, 552 First River Basin Management Plan, 126–7 “fit to call” evaluation., 665–6 flame atomic absorption spectrometry (FAAS), 385–6 flood warning systems, 486 flooding, 621 flow charts, 256–7, 339, 418, 531, 542, 684

Subject Index

flow direction, groundwater, 201, 242 flow rates, groundwater, 5–6 modelling, 72, 633 Shannon PRB studies, 130–2 significance in WFD, 180 See also mass flow rates flumethrin, 569 forums, Harmoni-CA, 144–5 fossil groundwater, 8 Fox-Wolf river basin, 616 fractured aquifers, 269–88 desiccation and tectonic fractures, 276 remediation alternatives, 282–7 Framework Programmes. See EU (European Union) framing conditions, 162–3 France abstraction taxes, 75 Ardèche river basin, 78 Beauce aquifer, 77 Brévilles catchment, 34, 570–6 Don catchment BMPs, 502–8 Hart catchment, 372–5 Lyon urban hydrology study, 682 megasite case studies, 412–17 Meuse basin case study, 43–51 ore processing site sampling, 394–6 pesticide contamination in, 565–8, 571 remediation costs, 338 freshwater diagenesis, 223–4 Freundlich coefficient/isotherm, 547, 555 fuel. See petroleum products fungicide oxadixyl, 566–7, 570–1 Fyn island, Denmark, 587, 591–2, 596–9 gaming, 487 gas chromatography (GC), 369, 513, 524, 547 See also GC-MS gasoline, 306, 325, 328–9, 516–18 GC-MS (gas chromatography-mass spectrometry) fingerprinting, 320–9

Subject Index

MTBE trace analysis, 513, 519 UK pesticide monitoring, 569 GDEs (groundwater dependent ecosystems). See dependent ecosystems Geer catchment, 620 generic reference values, 538–40 geographic(al) information systems (GIS) as communication tools, 164, 166, 487 DPSIR framework use, 187 WISE and, 695, 697 geographical scaling, 486 geohydrological models. See hydrogeological models geoindicators, 223, 260 Geological Survey of Ireland (GSI), 131 geology contaminant plume management, 423, 439–40 DK model regions, 587 Kempen area, 634 Rabis Creek, 229 Shannon PRB studies, 129–30 vulnerability assessments and, 14–15 German Federal Environmental Agency, 681 Germany age determination case study, 230 Bitterfeld contaminated megasite, 412–17 integral investigation method, 247–9, 258–9, 260–2, 263=265 Meuse River projects, 43–51, 367 MTBE and BTEX in soil and groundwater, 518–19, 525 nitrate protection zones, 464 public participation experience, 158–61, 167–9 remediation costs, 338 SAFIRA project, 240 Stuttgart INCORE study, 317, 324, 327–9, 340 GEUS (Geological Survey of Denmark and Greenland), 587, 589–90, 604

729 GFPs (good farming practices), 496 Giardia, 108 GIS. See geographic(al) information systems glacial tills. See clay tills glaciation in Europe, 225, 680 GLEAMS model, 465 Global Change and Ecosystems subpriority (FP6), 698 glufosinate-ammonium (GLUFNH4), 554 glyphosate, 459, 561, 563–7, 571 good farming practices (GFPs), 496 good quantitative and chemical status (WFD), 91, 130, 413, 466 See also “at risk” groundwater bodies GQMU (groundwater quality monitoring units), 569 GRACOS (groundwater risk assessment at contaminated sites) project, 26, 291–313 grain size denitrification and, 466 hydraulic conductivity and, 432 release rates and, 310 grains of sand, 677 Ground Water Rule (US, 2006), 107–16 challenges to, 108–9, 111 economic analysis, 111 risk-targeted strategy, 109–10, 113–16 summary of provisions, 114–16 Groundwater Action Programme, 1996, 89–90, 345 groundwater age. See age determination groundwater bodies comparison with analogous bodies, 188–9, 439 definition of aquifers and, 180, 347 fresh water storage in, vi, 3–4 identification and delineation, 180–2 local management of, 474–5 with lower objectives, 184 natural chemical variations, 199 pesticide contamination data, 556 regarded as ecosystems, 675

730

stratification, 354, 365, 394, 396, 573, 577–8 vulnerability to agricultural pollution, 460 See also aquifers; conceptual models groundwater characterisation work, WFD, 93, 94, 129, 177–92 Groundwater Conference, Vienna, 2006, 125 Groundwater Daughter Directive. See Groundwater Directive groundwater dependent ecosystems (GDEs). See dependent ecosystems Groundwater Directive (Directive 2006/118/EC), vi, ix, 97–8 accessibility to the public, 691 Articles 1 and 3, 674, 712 Article 6, 126, 713 Article 7, 127, 713 BRIDGE project and, 27–8 challenges, 474–5 groundwater age and, 228 groundwater baseline levels, 197 groundwater ecology and, 674 Harmoni-CA project and, 149 main provisions, 711–15 monitoring requirements, 347, 364 natural baseline trends, 211–14 Nitrates Directive and, 99 recitals, 479, 712 research needs, 29 risk-based megasite management, 412–13 vulnerability assessment, 14, 190 See also Directive 80/68/EEC (predecessor directive); Water Framework Directive (parent directive) groundwater ecology environmental monitoring and, 680–2 environmental policy and, 671–86 new groundwater ecology, 672–4 groundwater flow. See flow rates, groundwater Groundwater Forum, UK, 24

Subject Index

Groundwater Foundation, USA, 24 Groundwater Interlaboratory Programme, 381–2 groundwater management AQUATERRA contribution, 51–3 contaminated sites, 337 drivers for change, 151, 153 ecological information and, 672 economics and, 58–81, 479 effectiveness of recommendations, 492 institutional structure, 479–80 local approaches, 473–5, 481 models for, 616 public participation in, 155–61, 163 research integration and, 24–5 research needs, 28–9 research projects on, 26 See also adaptive management; integrated management strategy; river basin management; sustainable management groundwater monitoring. See monitoring groundwater protection needs, 13, 467 groundwater quality, 6–7 abstraction and, 585 AQUATERRA basin studies, 40, 47 assessment criteria, 537–41 basin-scale models, 616 causes of pollution, 10–13 as a function of age, 226–7 health risks of untreated, 112–13 land use practices and, 457–9, 462–3 operational and surveillance monitoring, 348 public participation, 159–61, 167–9 resource estimates, Denmark, 598 screening methods for, 364–5 spatial and temporal variability, 366–8, 576, 578 surface water interaction, 584–604 See also environmental quality standards; natural background level; pollution

Subject Index

Groundwater Quality Monitoring, UK Environment Agency, 569 groundwater quantity deterioration risks, 7–9 monitoring requirement, 348, 355, 359–62 preliminary assessment, EU, 88–9 See also abstraction; over-exploitation groundwater regulatory framework. See regulatory framework Groundwater Resources of the European Community: Synthetical Report, 85, 88 groundwater risk assessment at contaminated sites (GRACOS) project, 26, 291–313 groundwater sampling, 103 groundwater use, economic characterisation, 60–1 See also drinking water Groundwater Working Group (WG C). See CIS group processes, 156, 167 GRUMO sub-project, 601 guidance documents UK Environment Agency, 569 WFD, 128–9, 146–7, 155, 157, 162, 492 GUS index, 563 GWDD. See Groundwater Directive GWR. See Ground Water Rule GWstat code, 563 habitat diversity, 680–2 habitat models, 585, 599, 602 Habitats Directive (Directive 92/43/EEC), 93, 685, 710 half-life values (DT50 or T1/2) for pesticides, 459, 548–9, 552, 560, 567 halogenated hydrocarbons, 430 See also chlorinated hydrocarbons Harmoni-CA (Harmonised Modelling Tools for Basin Management) forums and conferences, 144–5 products and tools, 146–8

731 science-policy interface efforts, 691 successors, 148–9 WISE-RTD and, 700–2 HarmoniCOP (Harmonizing COllaborative Planning) project, 161, 167 Hart catchment, France, 372–5 hazardous substances. See contaminants; pollution; WATCH project heavy metals background levels, 356 biosensors for, 370 complexation, 299 leaching tests, 293–305 megasite contamination, Kempen, 417 sampling study, 394–6, 397–9 screening methods, 369, 375 sediment pollution, 621 Henry’s law constant, 305, 307, 424–5, 524 herbicides acid herbicides, 569 groundwater pollution from, 13 as total weed-killers, 563 See also pesticides hexachlorobenzene, 558–9 hexazinone, 560, 571 hierarchical cluster analysis, 327–8 high-priority pollutants. See List I pollutants hot spot zones, 241, 366 household WTP (willingness to pay) values, 62, 70 human impacts. See anthropogenic effects human waste. See sewage humic acid match, 383 humic and non-humic substances, 203 Hungary, abstraction taxes, 75 hybrid generation atomic absorption spectrometry (HGAAS), 389 hydraulic barriers, 284 hydraulic conductivity, 250, 432, 589–90

732

hydraulic fracturing, 286 hydraulic gradients, 228, 233, 244, 432, 550, 553 hydro-economic system, 59 hydrocarbons aliphatic, source identification, 321–2 early warning leak sensors, 513 GC-MS fingerprinting, 320–9 mass flow rates, 260 remote identification, 513–14 weathering, 321, 324 See also BTEX compounds; chlorinated hydrocarbons hydrochemical evolution, 230 hydrochemical facies, 134 hydroelectric power generation, 39 3 H. See tritium hydrogeochemical studies aquifer characterisation, 135–6, 536 groundwater age from geoindicators, 223 modelling, 199, 202, 206–8, 634 role of TOC, 203 See also chemical status; geoindicators hydrogeological cycle, 4–5 hydrogeological models, 185, 529, 633 “hydrogeological nature reserves”, 680 hydrogeological sensitivity assessment (HSA), 113–14 hydrogeology agricultural pollution vulnerability and, 461 contaminant plumes, 432–3 economics and, 77 monitoring parameters and, 360, 639 research needs, 28 transboundary aquifers, 104 vulnerability assessments and, 14–15, 126 Hydrogeosphere model, 634 hydrological balance, 62 hydrological cycle integrated management, 90, 473 hydrological models

Subject Index

agricultural pollution control, 495, 503–5 DK water resource model, 587–9, 604 RBMP, 480 river basin management, 585 as simulation models, 615 hydrology ecology and, 683 research needs, 28 social learning and, 161 hydromorphological modification,48, 621 hydroxyatrazine, 564, 566–7, 571 HYDRUS-1D model, 634 hypogean species, 678, 680 hyporeic zones, 540–1, 678 IC (information and communication) tools, 164–6 ice, fresh water stored in, vi ICP emission spectrometry, 394 ICPAES (inductively coupled plasma emission spectrometry), 389 ICMS (inductively coupled plasma mass spectrometry), 383, 386, 389–91, 394, 397–8 ICT (information, communication and technology) tools, 142–3 ID-ICPMS (isotope dilution ICPMS), 389 IFEN (Institut Français de l’Environnement), 565 illegal wells, 134, 154 IMAGE-TRAIN (innovative management of groundwater resources in Europe) project, 27 imidacloprid, 561, 566–7 immunoassays test kits, 369–70, 373, 513 WATCH project, 516, 519–23 impact assessment, 188–90 adaptive regulatory impact assessment, 158 megasite IMS and, 406–9 tools to assist, 188–9 See also pressures and impacts

Subject Index

IMPRESS working group, 186 in situ chemical logging, 396–7, 399 INAA (instrumental neutron activation analysis), 389, 391 incongruent dissolution, 224–5 INCORE project, 26, 240, 316–40 ISIRE program, 335–7 participating cities, 317 remediation implementation, 337– 40 remediation strategy, 333–7, 339 statistical methods, 323–9 indicator fact sheets, EEA, 558–60 indicators bioindicators, 653, 680–2 for soil quality assessment, 653 species for natural attenuation, 435, 437–8 for sustainable groundwater abstraction, 595–8, 600, 604 indirect discharge of pollutants, 86–8 inductively coupled plasma atomic emission spectrometry (ICPAES), 387 inductively coupled plasma mass spectrometry (ICPMS), 383, 386, 389–91, 394, 397–8 industrial effluents, 12, 412, 421 industrial megasites. See contaminated megasites infiltration treatment, 11 information, communication (and technology) (IC(T)) tools, 142–3, 164–6, 701 information asymmetries, 500 Information Collection Rule (US EPA), 108, 111 “information on demand”, 651 information system interoperability, 693–4 information tools, 26, 142–3, 164–6 infrared spectrometry, 512–16 infrastructure management, 480–1 inorganic constituents, risk assessment, 309–12

733 INSPIRE (Proposed EC Directive on an Infrastructure for Spatial Information in the Community), 693 Institut Français de l'Environnement (IFEN), 565 Institute for Reference Materials and Measurements, 392 integral groundwater investigation method, 241–9 applications, 247–50, 258–9, 260– 2, 263–5, 319–20 costs, 337–8 inversion problem, 244–7 multilevel integral investigation, 262–5 parameter uncertainty and, 249–59 integral pumping tests (IPTs) decision tree approach, 255–8 INCORE studies, 325–7, 330–2, 337 integral groundwater investigation method, 241–2 numerical inversion solutions, 245 parameter uncertainties and, 249–58 sources of uncertainty, 250–3 integrated characterisation process, 536–7 integrated management strategy (IMS), 26 contaminated megasites, 406–20 including soil protection, 651 integrated monitoring, 350–1 Integrated Pollution Prevention and Control (IPPC) Directive (Directive 96/61/EC, 90, 101–2 Integrated Protection and Management of Groundwater programme, 89 Integrated River Basin Management (IRBM), 473–92, 683 See also CatchMod; river basin management integrated water resource management (IWRM), 495–501, 505 developing for groundwater, 536 implementation costs, 505–6

734

river basin management and, 612–13 twinning and, 698 INTEGRATOR sub-project, AQUATERRA, 34, 56, 620, 622, 624–5 intellectual property, 666–7 interference, analytical methods, 379– 83, 386–7, 389–90 Intergovernmental Panel on Climate Change (IPCC), v, 72 Interim Enhanced Surface Water Treatment Rule (US EPA), 108 International Commission for the Protection of the Meuse (ICPM), 45 international cooperation, 104 International Hydrological Programme (IHP), 104 International Law Commission (ILC), 104 International Meuse Commission (IMC), 44, 47, 49 international organisations, river basins, 488–9 international standards, 101, 379, 392 CEN standards, 302–4, 311 ISO standards, 103, 302–3, 370, 379, 392 national standards and, 376, 540 Internet, 146–9, 491, 700–1 interoperability, 693–4 INTERREG funding, 702 intraparticle diffusion, 301–2 inverse modelling methods, 277, 590 inversion problem, 244–7, 250–1 ion chromatography, 387, 389, 391, 394 IPCC (Intergovernmental Panel on Climate Change), v, 72 IPPC (Integrated Pollution Prevention and Control) Directive (Directive 96/61/EC, 90, 101–2 IPTs. See integral pumping tests IPU. See isoproturon IRBM (Integrated River Basin Management), 473–92, 683

Subject Index

See also CatchMod; river basin management Ireland, 130–2 irrigation consequences of excessive, 8 Ebro basin, 39, 43 groundwater use for, vi, 3, 28, 603 increased efficiency of, 455, 465–9 modelling and management, 615 returned water, 586 ISIRE (in situ remediation technologies), 335–7 ISO standards, 103, 302–3, 370, 379, 392 isochrones, 243–7, 250, 252–3, 632 isopleths, 438 isoproturon (IPU), 459, 548–54, 559– 563, 566–8, 570–5 didemethyl- (DDIPU), 572, 574 monodemethyl- (MDIPU), 572, 574 isotope dilution mass spectrometry, 380 isotope evolution studies, 206, 209, 220 isotope ratio, 221 isotopic fingerprinting, 320, 329–33 Italy Milan, 335, 340 pesticide contamination in, 559–61, 571 IWRM. See integrated water resource management Jucar Basin, 457 Jutland, Denmark, 464, 587 Jylland, Denmark, 587, 591–2, 597–9 Kansas, USA, 585 karst systems, 14 case study areas, 130–1, 233 groundwater ecology, 674, 679–80 sampling location and frequency, 354, 359, 563 Kd. See distribution coefficients Kempen, Belgium/Netherlands, 412–17, 620, 634–43 kerosene, 308–9

Subject Index

knowledge production, 36, 51 Koblenz, Germany, 519 Krsko kotlina, Slovenia, 66–9, 620 85 Kr, 220–2, 231–4 Kyoto climate treaty (1997), 654 laboratories. See analytical laboratories laboratory reference materials (LRMs), 383–4, 391 lag times, 126, 494 land use concentration-depth profiles and, 638 constraints on, 467 Meuse basin case study, 46 pesticide monitoring and, 569 sustainable land management, 649, 664 See also agriculture Landfill Directive (Directive 99/31/EC), 102–3 landfill sites, 11 language barriers, 691 laser-induced fluorescence methods, 369 Latvia, 69–71, 372 lauryl sulfate, 383 leachate, Landfill Directive and, 102–3 leaching of heavy metals, 634, 636–7, 641 leaching of nitrate, 457–8, 463 leaching of pesticides, 462, 466, 566, 571 processes involved, 546, 549 UK levels, 459, 577 leaching tests, 293–305 LEACHM model, 465 LeachXS-Orchestra tool, 298 lead, 397–9 leadership, 124, 165–7 learning approaches fourth generation policy-making, 53, 55 policy games, 487 social learning, 154–5 learning cycles, 152, 154 LEDs (light-emitting diodes), 514 legal redress, 491 legislation. See regulatory measures

735 Leuna, Germany, 518 “levels of confidence”, 337 LGM (Last Glacial Maximum), 225 LIFE (L’Instrument Financier pour l’Environnement/Financial Instrument for the Environment) project, 148, 692, 700, 702 light-emitting diodes, 514 limestone, vulnerability assessment and, 14, 460 limit of detection (LOD), 542 lindane, 556–9, 561, 563–5 linear programming (LP), 71–2 liquid:solid (L/S) ratio, 294–9, 302–3, 311–12 List I and List II pollutants, 86–7 lithium timescales, 225 lithology and baseline geochemistry, 195 logging probes, 396–7 Long Term 1 (and 2) Enhanced Surface Water Treatment Rules (US EPA), 108, 110 longitudinal dispersion coefficients, 280 LOOP sub-project, 601 low flows. See base flows low-intensity agriculture, 467 low-volatility organics, 293–305 Lower Saxony, Germany, 159–61, 167–9, 464 “lowest common denominator”, 487 LRMs (laboratory reference materials), 383–4, 391 L/S (liquid:solid ratio), 294–9, 302–3, 311–12 lubricating oil, 325 Luxembourg, 43–51 Lyon, France, 682 lysimeter measurements, 297–8, 306, 308, 515–16 macroeconomic models, 78 Madrid basin aquifer, 202–3 Maipo River basin, 616 management. See adaptive management; groundwater management

736

manure, 168 mapping, 14–15, 126, 696 marine water. See saltwater Maslow, A, 167 mass balances, MNA, 435, 438–40, 445 mass flow rates integrated investigation method, 241, 243–7, 251–2, 259–62, 264 total mass flow rate, 246, 260 mass spectrometry, 220, 572 See also GC-MS; inductively coupled plasma; isotope dilution mass spectrometry mass transfer rates, 295 maximum allowable concentration (MAC), 100, 445, 459 Maximum Residue Level (WHOMRL), 560 MCPA (methyl chloro-phenoxyacetic acid), 459, 563 MCPP (mecoprop-p), 548–9, 551–4, 571 MDIPU (monodemethylisoproturon), 572, 574 mechanistic geochemical modelling, 296–8, 309–10, 312 mecoprop, 459, 561, 564, 569–70 See also MCPP Mediterranean basin, 104, 124, 149 groundwater ecology, 683, 685 multi-cropping in Spain, 455 megasites. See contaminated megasites metals complexation, 296 leaching tests, 293–305 as pollution indicators, 17, 49 metamitron, 550–1, 561, 565 metazachlor(e), 561, 565, 568, 570 methyl tertiary-butyl ether. See MTBE metolachlor, 552–4, 560–1, 568, 571 METREAU project, 392–9 metribuzin, 563, 565 Meuse river basin AQUATERRA project, 34, 43–51, 620–1, 623 SWIFT-WFD project, 367

Subject Index

microbial degradation. See biodegradation microbial populations groundwater depth and, 675 on a sand grain, 677 vadose zone, 549 MIKE SHE code, 587, 589–90, 634 Milan, Italy, 335, 340 Millennium Ecosystem Assessment Report, vi mining, 12, 47, 228, 417, 621 ministerial seminar on groundwater, The Hague, 1991, 85, 89 MIN3P model, 308, 311 mixing and dispersion biodegradation redox behaviour and, 432 transport models and, 219, 632–4, 636, 638 MLS (multilevel samplers), 366, 436–7, 443–4, 451 MNA (monitored natural attenuation), 424, 433–40, 444–5, 450 mode 1 and mode science, 36, 38 modelling AQUATERRA project, 56 combining study data, 111 contaminated megasites, 410 distributed models, 590 DK Water Resource Model, 587–92 DPDP/DPSP models, 270, 279–80, 288 exponential models, 232–3 fully integrated models, 634 groundwater discharge, 631 habitat models, 585 hydrogeochemical, 199, 202, 206–8 inverse modelling methods, 277 LeachXS-Orchestra tool, 298 mathematical models, 189 mechanistic geochemical models, 296–8, 309–10, 312 models defined, 613 natural attenuation, 251 numerical modelling and groundwater age, 225–6

Subject Index

piston flow models, 219, 232–3 quantitative vs conceptual models, 614 Rayleigh models, 329, 333 risk assessment for inorganics, 309–12 simulation and optimisation models, 615 water management and, 584 See also conceptual models; CoronaScreen; hydrogeochemical studies; numerical modelling; reactive transport models modern water. See young components MODFLOW model, 246, 250, 634 MODPATH code, 246, 250 MODSIM model, 616 molecular fingerprinting, 682 molybdenum, 299–300 MONIT InterReg project, 71 MONITOR sub-project, AQUATERRA, 41–2 monitored natural attenuation (MNA), 424, 433–40, 444–5, 450 monitoring, 16–17, 345–401 abstraction points, 136 adaptive management and, 153 chemical status and trends, 352–9 contaminated megasites, 410 customised monitoring systems, 637–42 data collection and analysis needs, 25 DK Model link with, 601–2 early warning monitoring strategy, 17, 528–32 ecological, 680–2 frequency, 356–9, 361–2, 569 gasoline contamination, 518 groundwater age and, 227 groundwater synthetic samples, 382 interpreting results, 367 operational and surveillance, distinguished, 348 quality assurance for, 378–400, 565 quantitative monitoring, 359–62 regulatory context, 345–62, 363–4

737 research needs, 28–9 triggered source water monitoring, 115 See also monitored natural attenuation; operational monitoring; screening procedures; surveillance monitoring Monitoring Guidance for Groundwater document, 124–5, 346 monitoring networks, 93 compliance regimes and, 543 cost, 355 diffuse pollution and, 467–8 early warning systems, 486 MNA, 433–4 model validation, 637–8, 642 site selection, 350–1, 353–5, 563, 578 See also points of compliance (POC) monoclonal antibodies, 520, 523 Monte Carlo simulations, 250 MSWI (municipal solid waste incineration) ash, 303 MTBE (methyl tertiary-butyl ether) analysis in soil and groundwater, 517–27 BTEX co-metabolic degradation, 527 field-based immunoassay, 513, 519–23 natural attenuation, 424–5, 524–8 MT3D-IPD model, 262 MT3DMS model, 634 multi-parameter probes (including YSI), 365, 368, 372, 394 multi-tracer approach, 222, 232–3, 235 multicomponent system modelling, 296 multidisciplinary teams, 51, 53, 56 multilevel integral investigations, 262–5 multilevel samplers (MLS), 366, 436–7, 443–4, 451 multiphase local equilibrium, 312 multiphase transport coefficients, 275–8 multiple barrier approach, 113 naphthalene, 280–2 NAPL (non-aqueous phase liquid) phase, 260–1, 269–71, 275–81, 283, 286

738

contaminant plumes from, 421–8, 433, 440, 443 See also benzene; DNAPL NAS screening model, 441 national standards, 376 National Water Resource Model (Denmark), 587–92 Natura 2000, 92, 136, 468, 685 natural attenuation, 421–51 acceptability, 423 assessment practices, 433–42 contaminated megasites, 411 CORONA approach, 437, 442–50 establishing threshold values, 542 establishing viability of, 434–9 factors affecting, 6–7 groundwater age and, 218 groundwater status and, 538, 540–1 INCORE project, 26, 240 mechanisms of degradation, 429–32 modelling, 251 MTBE, 524–7 of pesticides, 461–2 processes contributing to, 424–5 quantifying attenuation rates, 259–62, 265 remediation alternative, 319, 333–4, 423 sustainable land management, 664 volatile organics, 308 See also biodegradation; monitored natural attenuation natural background level (NBL) Colli Albani aquifer, 134–6, 140 groundwater age and, 228, 230 groundwater status and, 538 threshold values and, 541–2 See also background quality values; baseline chemical composition natural elements, 6–7, 10, 135 “nature reserves”, hydrogeological, 680 NBL. See natural background level near-infrared sensors, 513–16 Netherlands groundwater tax, 75–6

Subject Index

Kempen and Rotterdam megasites, 412–17 Kempen area model, 634–43 Meuse basin case study, 43–51 nitrate concentrations, 227 screening test development, 365 networking activities, 27 EU-funded networks, 27, 649, 662 research activities, 647 stakeholder communications, 164 See also monitoring networks NEWATER project, 698 NGOs (non-governmental organisations), 27, 152, 154, 491 involvement in policy formulation, 54, 122–3 participation in river basin management, 620, 696 nickel, 594, 598, 601 NICOLE network, 27, 649, 662 NIMBY (not in my back yard) effect, 163 NIST (National Institute of Standards and Technology, US), 392 nitrate balance model, 72 nitrate guideline values, 4, 197 Nitrate Sensitive Areas (NSAs), 463 nitrates Castelporziano, 137 DK model, 593–4 domestic effluents and, 11 Don catchment, France, 507–8 Doñana aquifer, 205 East Midlands aquifer, 234 Eastern England chalk lands, 458 global anthropogenic effect, 217 infiltration treatment and, 11 irrigation efficiency and, 465 Krsko aquifer, 66–9 Lower Saxony, 159–61, 168 Plana de Valencia area, 457–8 public participation, 159–61 Rabis Creek aquifer, 229–30 removal by phreatophytes, 679 Salone-Acqua Vergine area, 136

Subject Index

sampling frequency and application, 359 screening technologies, 370, 374 temporal evolution, 226–7 upper Rhine valley, 64–6, 71–4 Urban Wastewater Treatment Directive and, 99 See also denitrification; vulnerable zones Nitrates Directive (91/676/EEC), 93, 98–9, 378, 456, 463, 503 critical areas, 503 “no further action” approach, 319 noble gases as tracers, 220 See also 85Kr non-aqueous phase liquids. See NAPL non-agricultural use of pesticides, 469, 563 non-commodity farm outputs, 500, 506–7 non-community water systems (NCWS), 114 non-equilibrium conditions, 299, 301 non-governmental organisations. See NGOs non-point sources (NPS). See diffuse pollution sources non-transient non-community water systems (NTNCWS), 109, 112 non-volatile compounds. See lowvolatility organics North America, 286 See also USA NPS (non-point sources). See diffuse pollution sources NSAs (Nitrate Sensitive Areas), 463 “nugget” effect, 392 numerical modelling age determination using, 225–6 groundwater discharge, 633, 641 monitoring and, 360 NAPL spreading, 278–81 pesticide mobility, 465 See also simulation Observatoire de Terrain en Hydrologie Urbaine (THU), 682, 686

739 occurrence data, 110–11 OECD (Organisation for Economic Cooperation and Development), 495–7, 499, 617 on-site logging, 396–7, 399 operational monitoring, 348, 353 frequencies, 356, 358, 361–2 optimisation models, 63, 615–16 ORCHESTRA model, 311–12 organic carbon and pesticide sorption, 550, 553–4 See also dissolved organic carbon; total organic carbon organic farming, 464 organic matter, groundwater habitats, 681 organochlorine pesticides. See chlorinated hydrocarbons organonitrogen pesticides, 569 organophosphorous pesticides, 459, 569 Orlice River, Czech Republic, 372 OROGONATE (on-site remediation of groundwater contaminated by polar organic compounds using a new adsorption technology) project, 26 OTHU (Observatoire de Terrain en Hydrologie Urbaine), 682, 686 over-exploitation, 7–8, 141, 228, 601 alternative solutions, 62, 77 examples, 136, 598 surface water effects, 584 oxadiazon, 560, 570–1 oxadixyl, 566–7, 570–1 oxidative biodegradation, 429–30 oxidising-reducing potential (ORP), 429 See also redox processes 18 O studies, 234 PAHs (polycyclic aromatic hydrocarbons) as biosensors, 370 distribution and release, 299–302 source identfication through, 322, 324–7 at specific locations, 42, 248, 260–1, 281

740

“PALAEAUX” research project, 230 parameter uncertainty, 249–59 parameters DK model, 589–91 integral pumping tests, 249–59 interlaboratory comparison, 382 IWRM cost-effectiveness, 504–5 monitored in SWIFT-WFD, 368 See also determinands; modelling participatory approaches. See public participation; stakeholders participatory prospective assessment, 622 particle backtracking techniques, 254, 264 particulate organic matter (POM), 296 partition coefficients. See distribution coefficients partitioning between phases, 300 PASCALIS (Protocols for the Assessment and Conservation of Aquatic Life in the Subsurface) project, 672, 685 passive remediation, 333–4 passive samplers, 371, 375–6 Pastel UV, 372–4 pasture, 463–4 pathogens Ground Water Rule and, 107, 111–13, 116 karst system vulnerability, 679 removal by filtration, 6 sewage contamination and, 11 PCBs (polychlorinated biphenyls), 370 PCE (per/tetra-chloroeth(yl)ene), 329–30, 332, 430 PEARL model, 465 Peclet number, 278–80 peer review panels, 661, 667 PEGASE project, 26, 579 PELMO model, 549 percolation tests, 294, 297, 304, 311–12 See also column leaching tests percolation theory, 275 permeability of fractured networks, 275–7 permeable reactive barriers (PRBs), 423

Subject Index

permit procedures (IPPC and Landfill Directives), 102 persistent organics agricultural chemicals as, 13, 458, 464–5 MTBE as, 512, 519 as pollution indicators, 17 pesticides biogeochemistry, 545–79 Brévilles spring case study, 570–6 constraints on manufacture or use, 464–5 DK model, 593–4 fungicides, 566–7, 570–1 groundwater pollution from, 12–13, 455, 459, 555–70 half-lives, 459, 548–9, 552, 560, 567 indicator fact sheets, 558 indicators of anthropogenic impact, 198, 217 metabolites, 547, 553, 556, 568, 570–1, 575 non-agricultural use, 469, 563 organonitrogen and organophosphorus, 569 potential BMP examples, 498 preferential transport, 577 prescribed standards, 197 registration date, 567 sampling frequency and application, 359, 566 saturated zone biodegradation, 550–3 saturated zone sorption, 553 screening methods, 369–70, 372–3 sorption and biodegradation, 553– 5, 576–7 Strategy on the Sustainable Use of Pesticides, 503 transport mechanisms, 546–55 unsaturated zone sorption, 550 upper Rhine valley, 64–6 uron/urocarb pesticides, 569 See also acetanilide herbicides; carbamate herbicides; phenoxyacid herbicides;

Subject Index

phenylurea herbicides; triazine herbicides Pesticides in Groundwater factsheet (EEA), 558 petroleum industry, 286, 394 petroleum products biodegradation, 429–30 diesel, 329 discriminant analysis, INCORE, 325 early warning leak sensors, 513 gasoline constituents, 306, 328–9, 516–18 GC-MS fingerprinting, 320–9 stimulated remediation, 287 kerosene constituents, diffusion, 308–9 megasite contamination, 417 Riga, Lava site pollution, 69–71 threshold values, 527 See also BTEX; MTBE; NAPL pH cadmium leaching and, 294 percolation tests and, 297, 299, 303–4 pharmaceuticals, 48 PHAST model, 311 phenol, 280–2 2-methyl-4,6-dinitro (DNOC), 553–4 p-nitro-, 551 phenol(ic)s, 370, 436, 444 phenoxyacetic acid 2,4-dichloro- (2,4-D), 551 methyl chloro- (MCPA), 459, 563 2,4,5-trichloro- (2,4,5-T), 554, 561 phenoxyacid herbicides, 459, 551, 554 phenylurea herbicides, 549, 560 See also isoproturon phosphate screening technologies, 370 phosphonic acid, aminomethyl(AMPA), 561, 564, 571 phreatic zone, 409, 552 See also saturated zones phreatophytes, 10, 679 PHREEQC model, 208, 311, 446, 449 PHT3D model, 262, 311 physicochemical parameters, 137–40, 354, 368–9, 394

741 piezometry, 134, 572, 591 Pilot River Basins (PRBs) network, 128–41 CIS and, 123 river basins involved, 130 science-policy integration, 129–30, 145 Shannon PRB studies, 130–2 Tevere PRB studies, 133–4 WISE-RTD and, 14, 149 See also Mediterranean basin piston displacement, 632 piston flow models, 219, 232–3 Plana de Valencia, Spain, 457–8 Plant Protection Products Directive (Directive 91/414/EEC), 100–1 plumes. See contaminant plumes POCs (points of compliance), 407–11, 413–14, 416–17, 419, 440 point measurements and IPT, 241, 247, 253 point pollution sources case studies, 49, 134 risk assessment and identification of, 14, 186, 421-423 Groundwater Action Programme and, 90 inclusion in RBMPs, 126, 477 passive sampling applicability, 371 point source monitoring, 366 points of compliance (POC), 407–11, 413–14, 416–17, 419, 440 Poland, 412–17 policy-maker perceptions, 13 policy-making economics and, 60, 79 iterative development, 152, 162 learning approaches in, 55 stakeholder involvement, 54–5 See also environmental policies; science-policy integration “policy-oriented research”, 698–9 pollutant attenuation. See natural attenuation pollutants BAM (2,6-dichlorobenzamide), 231

742

behaving like tracers, 226 BMP classification by, 498 direct and indirect discharges, 86–8 distinguished from contaminants, 194 European Inventory of Existing Chemical Substances, 287 List I and List II, 86–7 natural background level (NBL), 134 polluter pays principle as an environmental principle, 18 embodiment in the WFD, 58, 73, 94, 479, 704 identification of pollution sources and, 320 pollution bio-fuel production and, 73 control, regulatory approaches, 484–5 control techniques, RBMP, 477–8, 495 cost of illness method, 61 economic weighting, 60–1 hot spot zones, 241 petroleum products, 69–71, 287 principal causes of, 10–13 problems in heavily polluted areas, 316 risk assessment research, 25–6 socioeconomic impact, 64–6 storm water infiltration, 682 threshold values, 27 See also contamination; nitrate; pesticides pollution indicators, 17, 356, 681 See also bioindicators; faecal contamination pollution plumes. See contaminant plumes pollution pressures, 188 pollution prevention groundwater ecology, 674–80, 685 importance, 15–16 remediation difficulty and, 4, 16 pollution sources depth variations, 263 integral groundwater investigation method, 241–9, 253–5

Subject Index

research needs, 28 source identification, 320–33 source zone delimitation, 241, 249–59 See also diffuse pollution sources; point pollution sources polycyclic aromatic hydrocarbons. See PAHs POM (particulate organic matter), 296 population groundwater use in USA, 109 growth and simulation models, 615 precipitation and, Denmark, 593 water resource requirements, v–vi porous materials characteristic ecosystems, 678 conceptual discharge model, 631 dispersivity and, 426 hydraulic conductivity and, 432 release rates, 310 Portugal, 519 potash mining, Alsace, 417 “potential impacts”, 188 potentiometry, 385, 394 PRBs. See permeable reactive barriers; Pilot River Basins (PRBs) network pre-industrial age water, 198 pre-selection method for NBLs, 538 precautionary principle, 18 precipitation (rain and snow), 593, 603 “predictor variables”, 324 preferential transport of pesticides, 577 pressure magnitude thresholds, 131–3 pressure measurements. See piezometric levels pressures and impacts, WFD, 187–91 in PRB exercises, 130 requirement and process, 98, 178–9, 183 pressures and receptors, 354 prevent/limit measures, 62, 124–6, 182, 536–7 prevention. See pollution prevention priority substances, 374–5, 540 Priority Substances List, WFD, 558–9

Subject Index

“pristine reference status”, 681 privatisation, 481 PRMZ model, 465 process management, social learning, 163–5 prochloraz, 569 product oriented policies, 35 project appraisal, 62–3 propachlor, 552 propanyl, 552 propazine, 561, 569, 571 protected areas, WFD, 93, 467–8 PSI model. See DPSIR framework public consultation, 192 public health risks, 112–13 public participation, 150–70 actor analyses, 158 adaptive management and, 151–4 German experience, 158–61 guidance document on, 155–7 IC tools and, 164–5 river basin management, 487, 489–91 social learning and, 161–9 WFD and, 157–9 public perceptions, 13 public water systems (US definition), 108–9 pump-and-treat techniques, 284, 423 PURE project, 26 purge-and-trap (P&T) techniques, 513, 516–17, 519 push-based profiling, 241 quality assurance groundwater monitoring, 378–400, 565 interlaboratory studies, 380–3 within laboratories, 379 quantitative status (WFD), 91, 141 quinmerac, 565 Rabis Creek site, 229–31 radioactive substances, 87, 356 radioactive tracers, 209 radiolabelled pesticides, 547

743 RAGS (Risk Assessment Guidance for Superfund) approach, 281, 284 Raoult’s law, 305, 307, 312 Rayleigh models, 329, 333 RBCA software, 281 RBMPs. See River Basin Management Plans reactive fringe thickness (RFT), 447–9 reactive transport models, 255, 260, 262, 410, 435 contaminated megasites, 410 CoronaScreen suite, 446–50 coupled models, 634–5 diffuse contamination, 630–43 indicators for, 435 Kempen area, 634–43 plume length and mass flow estimation, 255, 260, 262 See also MIKE SHE REBECCA project, 146 receptor-oriented status assessment, 541, 543 receptors attenuation by dilution at, 540 definition and characterisation, 536–7 megasite IMS and protection of, 406–9, 419 vulnerability and remediation options, 422 See also source-pathway-receptor model recharge irrigation and, 466 over-exploitation and, 8, 586 pesticide transport by, 546 point and diffuse, 14 sustainable abstraction and, 598, 600, 602, 604 wastewater purification by, 11 recharge mechanisms, 4–5 recharge zone sampling, 468 “recharge–pathway–discharge” model, 349, 364 redox fronts, 199, 203, 206, 213, 229 redox media, 210

744

redox potential (Eh), 429, 435 redox processes in biodegradation, 429–32, 441 reference materials, 380 artificial reference materials, 382 See also CRMs; LRMs reference methods, 380 registration of pesticides, 465 regulation adaptive regulatory impact assessment, 158 asymmetries of information, 500 baseline chemical levels and, 195 basis of site evaluations questioned, 294 definition of plume boundaries, 445 economic impacts of, 59 infrastructure management and, 480–1 monitoring requirements, 345–62 sampling methodologies, 399 regulatory framework European Union, 86–106 international cooperation, 104 USA, 107–16 See also Directives regulatory limit values, 311–12 regulatory measures first generation environmental policy, 35, 37 new technology and, 376, 692 problems in heavily polluted areas, 316 reinjection, 95 relative permeability curves, 275–6, 278 release kinetics, contaminants, 299–302 remediation assessing the requirement for, 422 biodegradation of organics, 206 bioremediation, 681 cost-benefit analysis, 63 costs, 288, 335–6, 337–40 cyclic approach to, 318–37 difficulty of, 7, 12 fractured aquifers, 269–71, 282–7 INCORE project, 26, 240, 316–40 monitored natural attenuation alternative, 319, 333–4

Subject Index

multi-scale investigation prior to, 240–1 natural attenuation forecasting, 421–51 need to prioritise sites, 291 options for point sources, 422–3 passive remediation, 333–4 prevention alternative, 4 quantifying efficiency, 265 research projects, 26 in situ, 286–7, 333, 335–7 technologies, 284–7, 664 remote sensing techniques modelling BMP effectiveness, 504 remote hydrocarbon identification, 512 “Reporting Directive,” 123, 694 research AQUATERRA contribution, 51–3 conceptual models in, 614 cooperation in a stepwise process, 657–62 “policy-oriented research”, 698–9 projects supporting groundwater policy, 25–7 results dissemination and implementation, 661–2, 684, 691–3 soil protection, 650, 654–5 “tailor-made” research, 21, 25–7, 691 transnational funding, 647–70, 697 treaties and agreements, 654 use of WISE, 696 See also RTD Framework; sciencepolicy integration: STREPS research integration, 21, 24–5 reservoir operation models, 615–16 reservoirs, underground. See aquifers residence times, 209 groundwater age and, 218, 223, 225, 232–3 pesticide biodegradation and, 550, 553 retardation factors, 251, 547, 553–4 revitalisation. See remediation RFT (reactive fringe thickness), 447–9 Rhine basin, 616 See also upper Rhine valley

Subject Index

Richards’ equation, 589 Riga, Latvia, 69–71 Ringe site, 272–5, 282–7 Rio biodiversity treaty (1992), 654 risk assessment, 177–92 economics of, 293 fractured aquifers, 269–88 GRACOS project, 26, 291–313 grouping groundwater bodies, 181, 189 inorganic constituents, 309–12, 356 Integrated Risk Assessment activity, 126 low-volatility components, 293–305 occurrence data, 110–11 passive sampling and, 375 policy implications, 13–17 remediation requirements, 422 reporting, 191–2 research projects, 25–6 screening and, 367 Shannon PRB, 131–3 surveillance monitoring and, 348, 352–3 volatile components, 305–9 See also source-pathway-receptor model; vulnerability Risk Assessment Guidance for Superfund (RAGS) approach, 281, 284 Risk-Base coordination action, 43 risk-based approaches contaminated megasites, 406–13 to monitoring, 349 remediation through natural attenuation, 421–51 threshold values, 537 WATCH project, 529–30 risk management scenarios, 410–12 risk-targeted strategy, GWR, 109–10, 113–16 RITEAU projects, 393 river basin management AQUATERRA case studies, key issues, 621

745 conceptual model/understanding, 613–17 Ebro basin, 40 environmental flow requirements, 683 international organisations, 488–9 Meuse basin, 45 principles of, 475–6, 478 surface water hydrological modelling, 585 See also integrated river basin management River Basin Management Plans (RBMP) analytical support, 483–7 available instruments, 477–8, 495–6 BMP integration, 495–501 conceptual models, 185–6, 617–20, 624 consultation requirements, 157–8, 160 derogations from WFD and, 92–3 EUWATERMAN project, 474 groundwater body identification and, 182, 187 integrated management requirements, 474, 476 management of megasites, 412, 418 operational management, 476–81, 485–6 planning, 482–4, 486–7 risk assessment and, 126–7 Shannon PRB, 131 threshold values, 543 river basins AQUATERRA studies, 34 assignment of groundwater bodies to, 181 boundaries of, 129, 528–9 dynamics of, 618–19 integrated water resources management, 90 international, 487–9 pressures on, 32–3 WISE-RTD testing, 149 See also Pilot River Basins (PRBs) network

746

river flow maintenance, vi, 3 RIVERTWIN project, 698 root zone, 546–9, 576–8 Rotterdam, Netherlands, 412–17 RTD Programmes working group, 651 RTD (research and technology development) Framework Programmes. See EU (European Union) rule-based models, 63 runoff abstraction–runoff balancing principle, 595 groundwater recharge and, 604 simulation, 590 runoff coefficients, 44 Safe Drinking Water Act (SDWA, US), 108 safeguard zones, WFD, 467–8 SAFIRA project, 240 saline intrusion. See saltwater intrusion saline water aquifers, 204, 223–4, 230 salinisation, 8, 40, 137, 230, 621 See also chlorides salmonidae, 596 Salone-Acqua Vergine area, 136 saltwater intrusion climate change and, 6 Colli Albani case study, 135–7, 140 good chemical status and, 352 NBL modelling and, 208 over-extraction and, 9 as a pollution pressure, 187 quantity monitoring and, 360–1, 365 sustainable yield and, 586, 601 See also salinisation Salzburg, Austria, 518 sampling assessment of uncertainty, 392–9 leaching tests, 297–8 MTBE biodegradation, 525 new technologies, 393–4 representativeness, 392, 396–7, 436, 563

Subject Index

See also monitoring; passive samplers; spot sampling sand grain “ecosystem,” 677 sand lenses, 270, 275, 280 sandstone media artificial reference materials, 382, 385 monitoring well design and, 436 pollution risk, 460 sanitary surveys, 108, 114 sanitation. See sewage satellite information, 504 saturated zones pesticide biodegradation, 550–3 pesticide sorption, 553–4 pesticide transport, 546–55 “scale dependent” estimates, 598 scaling differing spatial and temporal, 189, 615, 677 economic studies, 78 mutli-scale contamination characterisation, 240–66 scenario analyses in adaptive management, 152 AQUATERRA, 622 concentration-time series, 242 risk management, 410–12 sustainable extraction and, 591–2, 603 WISE, 696 science-policy integration, 21–9 environmental policy generations and, 36–8, 121–2, 535–6 Harmoni-CA, 142–9 lack of effective ecological input, 672 pesticide case study conclusions, 578–9 pilot river basin network, 129–30 research needs, 28–9, 578 See also research “science–policy interface” common approaches, 24 Harmoni-CA and, 145 SPI-Water, 148 WISE-RTD and, 690–703 See also SPI-Water project

Subject Index

scientific disciplines. See disciplines scientific research. See research Scientific Support to Policies (SSP) Priority, 699 screen-printed electrodes (SPEs), 369, 375 screened boreholes, 221, 436 Screening Methods for Water Data Information in Support of the Implementation of the Water Framework Directive, 368 screening models natural remediation and, 440–2, 444–5 See also CoronaScreen screening procedures cyclic remediation approach, 318–20 emerging technologies, 369–76 integral investigation approach, 241 for priority substances, 374–5 RAGSA methodology, 284 rapid quality estimation, 364–6 RBMPs and, 187, 189 SDWA (US Safe Drinking Water Act), 108 SEA Directive, 157 seawater. See desalination; saltwater intrusion sediment contamination, 48 semi-permeable membrane devices (SPMDs), 371 sensitivity analysis, 503 sensor design, WATCH, 513–14 See also remote sensing sensory properties. See taste and odour SENSPOL, 518 septic tanks, 11, 66 SEPTWA model, 563 sewage, 10–11, 49, 66, 469 Sewage Sludge Directive (Directive 86/278/EEC), 103 SF6 environmental tracer, 220–2, 231 shaking tests, 294–9, 302 Shannon PRB studies, 130–2 shared aquifers, 129 simazine, 369, 459, 465, 556–63, 566–71, 577

747 simulation contaminant transport, 269–72, 285, 288 coupled models, 306–7, 308, 313 nitrate pollution, 71–4 simulation models, 615–16 SIMUSCOPP program, 271, 279–81, 288 “sink” compartments, 676 Sjælland, Denmark, 587–8, 591–5, 597, 599, 602 Slovakia, 585 Slovenia, 66–9, 75 snow, 588 SNOWMAN project (sustainable management of soil and groundwater pollution), 647–70 coordinated call, 662–8 goals for cooperation, 657–62, 669–70 networks and, 27 research agenda, 652–3, 659–61 social learning, 154–5, 161–9 methods and tools, 164–5 socioeconomic impacts, 287–8, 486–7 See also economics soil erosion, 40–1, 621 “Soil Framework Directive” proposal, 512, 654 soil management based approaches, 495 soils eight threats to soil protection, 649 pesticide transport, 546–55 research programmes, 654–6, 662–3 research requirements, 650 Thematic Strategy on Soil Protection (EU), 649, 651, 685 See also sustainable land management solute transport experiments, 279 solvents, 12 See also chlorinated solvents sorption natural attenuation by, 426, 432, 440 of pesticides, 458, 553–5, 576–7 of zinc, 641

748

sorption coefficients. See distribution coefficients “source-path(way)-receptor” model, 131, 406, 419, 422 sources of analytical error, 380–1 source–sink ecotones, 676 SOWA project, 26 Spain chlorides in air and groundwater, 195–6 Doñana aquifer, 204–5, 210 Ebro basin case study, 38–43 fractured site risk assessment project, 272 irrigated multi-cropping in Mediterranean, 455, 458 Madrid basin aquifer, 202–3 MTBE in soil and groundwater, 518 tradable groundwater licences and rights, 77 SPEs (screen-printed electrodes), 369, 375 Special Conservation Zones, 136 SPI-Water project, 148 SMDs (semi-permeable membrane devices), 371 spot sampling, 367, 376 alternatives to, 371, 374 springs, as sampling sites, 354 SSP (Scientific Support to Policies) Priority, 699 stability of reference materials, 383–5, 389 stakeholders, 121–70 adaptive management and, 161 AQUATERRA project, 41, 50, 51, 53–5, 622 communication networks, 22, 27–8, 64 critical area designation, 503 economics and, 79 environmental policy and, 121 EU research programmes, 27–8, 660–2, 664, 691 groundwater policy and, 32, 467, 482, 490

Subject Index

participatory approaches and, 150–2, 154, 159–61, 169 privatisation and, 481 representativeness, 167 risk assessment data, 192 river basin conceptual models, 619–20, 624–5 social learning and, 161 sustainability of groundwater abstraction, 602–3 WISE project, 696 standardised protocols, leaching tests, 302–4 Standardised Reporting Directive (Directive 91/692/EEC), 123, 694 standards. See Environmental Quality Standards; international standards statistical models, 63–4 statistics analytical laboratories, 379–80 background quality values, 194, 200–2, 211 box and whisker plots, 200, 203–4 combining study data, 111, 141 contaminant identification, 323–9 IPT contaminant plumes, 255 sampling uncertainties, 398 steady-state plume behaviour, 427–9, 440, 445–50 steam injection, 284 step drawdown tests, 249 Stern Report, vi STOFFBILANZ software, 72 storm water infiltration, 682 storylines, 622 Strategy on the Sustainable Use of Pesticides, 503 streamflow depletion, 584–5, 594–9, 602 streamtubes, 244–5, 247, 255, 634 STREPS (specific targeted research projects), 698 strontium, 225 Stuttgart Feuerbach district, 340 INCORE study in, 317, 324, 327–8

Subject Index

Nesenbach site, 329 railway station site, 340 Ulmer Strasse site, 327–8 stygobiotic species, 679, 681 stygobites, 676 stygoxenic species, 681 subsidence and over-exploitation, 8, 14 subsidiarity principle, 197, 481 substances regulated under Directive 80/68/EEC, 86 sulfur hexafluoride. See SF6 surface water groundwater interactions, 584–604, 630–43 groundwater sampling via, 354 threshold value determination, 541–3 Surface Water Treatment Rule (US, SWTR), 107 surveillance monitoring programmes, 96, 348–9, 352–4, 562–3 frequency, 356–7 suspended matter pollution, 48, 49 Sussex, England, 464 sustainability objective, 89–90 sustainable development principles, 18, 476, 586 Sustainable Development Strategy, EU, 98 sustainable land management, 649, 654, 658, 660, 662–4 “Sustainable Management and Quality of Water”, 697 sustainable management of groundwater, 512, 591–6, 598, 600, 602–4, 683 See also SNOWMAN project sustainable pumping, 586 sustainable yield, 76, 586, 601 SWAP model, 465, 634 SWATMOD model, 634 Sweden, 564–5, 571 SWIFT-WFD project, 367–8, 371–2, 374–5 Swiss Water Protection Ordinance, 681 Switzerland, 231–3, 518 SWTR (Surface Water Treatment Rule, US), 107

749 Synthesis reports, Harmoni-CA, 148 Syr Darya basin, 616 systematic errors, laboratories, 379– 80, 389 “tailor-made” conferences and workshops, 145 “tailor-made” environmental policies, 55 “tailor-made” research, 21, 25–7, 691 TAME (tert-amyl methyl ether), 513, 516, 523 Tarnowsky Góry, Poland, 412–17 taste and odour problems, 516, 519, 524 TBA (tert-butyl alcohol), 513, 517–19, 523, 527 TBF (tert-butyl formate), 513, 517, 523 TCE (trichloroeth(yl)ene), 329, 332, 417, 430, 516, 518 TCR (Total Coliform Rule, US), 107, 114–15 TDIC (total dissolved inorganic carbon), 435 temperature biodegradation and, 308–9 vapour-phase diffusion and, 307 terbuthylazine, 560 terrestrial ecosystems, 9–10, 16 tert-amyl methyl ether (TAME), 513, 516, 523 tert-butyl alcohol (TBA), 512, 517– 19, 524, 527 tert-butyl formate (TBF), 513, 517, 523 tetradifon, 459 Tevere PRB studies, 133–40 Thematic Strategy on Soil Protection (EU), 649, 651, 685 thermal soil treatment, 284 three-dimensional characteristics, 349–51, 354, 366 three-dimensional views aquifer structure, 364, 366, 473 modelling, 263, 308, 311, 349–51, 587, 631, 634 sampling flows, 199, 201 threshold ecosystem effects, 500–1

750

threshold values alarm thresholds, WATCH, 515 BRIDGE project and, 27–8, 535 Colli Albani case study, 135–6 good chemical status and, 190, 197–8 Groundwater Directive, 97, 130, 190, 197 nitrates, 66–9, 73, 137 pesticides, 65–6, 556, 560, 570 petroleum products, 69–71, 527 pressure magnitude thresholds, 131–2 risk-based approach, 537 saline intrusion, 140 sampling depths and, 643 surface water example, 541–3 tiered approach, 541–3 See also values of compliance (VOC) TIC (total ion chromatograms), 517 time-integrated sampling. See passive samplers time-lag, 181, 189, 468 time resolution, dating methods, 220 time span, megasite IMS, 418 timing anthropogenic effect onset, 217 concentration time series, 242–3, 245, 247 fertiliser application, 462–3 pesticide application, 465 science–policy integration and, 22 stakeholder perceptions, 50–1 temporal baseline variability, 194, 366–8 temporal scales in baseline studies, 208–11 temporal scales in economic models, 615, 619 temporal scales in impact analysis, 189–90 See also residence times; travel times 1,2,4-TMB (1,2,4-trimethylbenzene), 310 TNCWS (transient non-community water systems), 109, 112 TOC (total organic carbon), 203, 205–6

Subject Index

toluene, 309 See also BTEX compounds Total Coliform Rule (US, TCR), 107, 114–15 total dissolved inorganic carbon (TDIC), 435 total ion chromatograms (TIC), 517 total organic carbon (TOC), 203, 205–6 total reflection XRF (TXRF) spectrometry, 389 toxic metals, 12 trace elements, 383, 387–90 TRACE-Fracture project, 26, 271–2, 281, 287–8 traceabliity requirement, 379, 389, 392 tracers, 208–11, 573 characteristics required, 194 chemical, 223 ecological, 679 groundwater dating by, 217, 219–23, 229–35 immunoassay, 521 multiple, 208, 232–4, 236 for natural attenuation, 437 pollutants resembling, 226 radioactive, 209, 220–1, 232–3 reactive, 225 See also indicators tradable groundwater licences and rights, 76–7 transboundary aquifers assessing pressures, 188 monitoring requirements, 348–9, 353, 361 proposed international legislation, 104 Shannon PRB, 130 WFD requirements, 184, 348–9 See also river basins:international transient non-community water systems (TNCWS), 109, 112 “transition management”, 35 transnational funding. See research transport models. See contaminant distribution; reactive transport models

Subject Index

travel times, 549, 575–6, 631–3, 636, 640 Travelling 1-D Model, CoronaScreen suite, 446–7, 449–50 treatment technique requirements, 115–16 trees, 679 trends modelling spatial and temporal, 207 monitoring, 352–9 natural and contamination, 212–14 natural and developmental, 211 triazine herbicides, 558, 560, 566, 568, 570 See also atrazine; simazine trietazine, 569, 571 trifluralin, 559, 561 trigger values, 99 triggered source water monitoring, 115 tritium (3H), 222, 227, 229 Brévilles spring, 575 groundwater ages using, 552 nuclear bomb test marker, 220 Odense Water Company, 230 Rabis Creek aquifer, 230 Swiss case study, 232–4 TWG (Technical Working Group) Research. See SNOWMAN TWINBAS project, 698 twinning, 489, 698 2,4-D (2,4-dichlorophenoxyacetic acid), 551, 563 2,4,5-T (2,4,5-trichlorophenoxyacetic acid), 554, 561 two-tracer approach, 221, 234 TXRF (total reflection XRF) spectrometry, 389 UCODE, 590 UK arable farming in Eastern England, 455–6 diffuse groundwater pollution, 466 East Midlands aquifer, 224–5, 234–5 nitrate protection zones, 463–4

751 pesticide contamination in, 565–70, 571 screening test development, 365 Wessex aquifer, 224–5 Ulmer Strasse, Stuttgart, 327–8 ultraviolet spectrophotometry. See Pastel UV uncertainties analytical and sampling, 398 groundwater sampling, 392–9 parameter uncertainties, 249–59 resource estimation, 598 uncertainty analyses, 250, 255 unconsolidated strata, 5, 8 underground storage tanks (USTs), 513 UNEP (United Nations Environment Programme) 1996 report, 13, 16–17 UNESCO, 104 United Nations Conference on Environment and Development, 1992, 613 United States Geological Survey (USGS), 197, 442 unsaturated zones in aquifer structures, 5 cotamination categories, 409 pesticide sorption, 550, 555 pesticide transport, 546–55 See also vadose zone upper Guadiana basin, 154 upper Rhine valley groundwater pollution case study, 64–6 nitrate pollution case study, 71–3 “upstream” economic assessments, 77 Urban Wastewater Treatment Directive (91/271/EEC), 11, 49, 93, 99–100 urbanisation CityNet project, 698 groundwater contamination and, 10–11 polluted recharge and, 682 uron/urocarb pesticides, 569 USA Fox-Wolf river basin, 616 Ground Water Rule (2006), 107–16

752

groundwater system demographics, 109–10 modelling surface water interactions, 585 nitrate concentrations, 227 regulatory framework, 107–16, 195 tradable groundwater licences and rights, 76 See also EPA USGS (United States Geological Survey), 197, 442 USTs (underground storage tanks), 513 vacuum extraction, 284 vadose zone agricultural leachate profiles, 458–9, 461, 467 contaminant plumes in, 312 microbial populations, 549, 679 See also unsaturated zones Valencia, Spain, 457–8 values of compliance (VOCs), 407–10 vapour-phase diffusion, 305–7 vegetative filter strips (VFS), 498 vertical dispersivity. See dispersivity vertical distribution of pollutants, 263, 265, 578 vertical variability of biodegradation, 552 vertical wells, remediation shortcomings, 286 VFS (vegetative filter strips), 498 viruses, 108 See also pathogens vital drinking water areas, 598 Vital Soil conference, 2004, 649 VOCs (values of compliance), 407–10 VOCs (volatile organic compounds), 368–9 volatile components, 305–9 See also BTEX compounds volatilisation herbicides, 458 natural attenuation through, 424–5

Subject Index

voluntary agreements, RBMP, 478–9 voluntary measures in environmental policy, 35, 37, 495–500 vulnerability to agricultural pollution, 460–1 assessments, 14–17, 126, 131, 188–9 groundwater ecosystems, 677–80 vulnerable zones, 93, 98, 230, 463–4 Wallony, Belgium, 561–4, 571, 620 wastewater treatment, 49–50, 68, 372 See also sewage; Urban Wastewater Treatment Directive WATCH project (Water Catchment Areas: Tools for Management and Control of Hazardous Compounds), 26, 511–32 Water & Soil Times EU newsletter, 665 water authorities, Spain, 40 water balance, 590–1, 601, 603 Water Framework Directive (Directive 2000/60/EC), vi, ix, 86 accessibility to the public, 691 Annex 10, 565 Annex II, 177–8, 182–3, 709–10 Annex V, 347, 476, 710 Annex VI, 479, 576, 710–11 Annex VII, 482, 711 Annex XI, 711 antecedents, 85–6 AQUATERRA project and, 33 Article 3, 475, 695, 705 Article 4, 103, 131, 177, 182, 184, 347, 361, 705 Article 5, 61, 94, 96, 129, 144, 177– 8, 182, 348, 482, 541, 695, 705 Article 7, 476, 706 Article 8, 347, 706 Article 11, 178, 183, 706–7 Article 13, 158, 482, 487, 707 Article 14, 157–8, 162, 489, 619, 707 Article 15, 191, 707–8 Article 17, 96–7, 130, 190, 197, 345, 363, 535, 672, 708 Article 18, 696, 708

Subject Index

Directive 80/68/EEC and, 88, 91, 98 DK Model applicability, 600–1 economics integration, 58, 78 groundwater action programme (1996) and, 90–1 groundwater age and, 228 groundwater baseline levels, 195 groundwater characterisation requirements, 177–92 groundwater protected areas and safeguard zones, 467 guidance documents, 128–9, 146–7, 155, 157, 162, 492 INCORE project and, 337 main provisions affecting groundwater, 704–11 monitoring requirements, 16, 346–8, 352, 361, 363–4 participatory approaches, 150 Priority Substances List, 558–9 recitals, 97, 704 science–policy integration, 23 vulnerability assessment, 14 See also Common Implementation Strategy Water Information System for Europe. See WISE water pricing, 479 water resources fresh water reservoirs, vi future requirements, v–vi holistic management, 31–2 integrated management, 89–90, 495–501 management in Spain, 40 monitoring use, 28 optimisation models, 615 over-exploitation, 7–8 sustainable abstraction, Denmark, 592–8 See also drinking water supply; groundwater quantity water scarcity, 40, 47, 126, 621, 697 water services cost recovery, 479 water tables depth variations, 5

753 over-exploitation and, 228 quality effects, 594 recharge zone sampling, 468 rising and falling, 8, 577, 598 Waterbase database (EEA), 555–8, 578 waterborne disease outbreaks, 112–13 WATERMAN project, 474 watersheds and aquifer boundaries, 475 See also river basins weathering, hydrocarbons, 321, 324 Web portal, WISE-RTD, 146–7, 149, 700–1 WELCOME (Water, Environment, Landscape management of Contaminated Megasites) project, 26, 406–7, 420, 662 well capture zones, 243–4, 253 wetlands, 9, 39, 130, 585, 678 WFD. See Water Framework Directive WHO (World Health Organisation), 10, 560 whole-organism bioassays, 370 Wild Birds Directive (Directive 79/409/EEC), 93, 685 willingness to pay (WTP), 62, 70 wireless technologies, WATCH, 515 WISE Implementation Plan, 695, 697 WISE (Water Information System for Europe), 24, 192, 485, 693–7 “science–policy interface” and, 690–703 WISE-RTD Web Portal, 146–7, 149, 700–1 WISE WFD prototype, 695 WISE Workshop, 2005, 695 Working Group on Groundwater of the Common Implementation Strategy of the WFD, 24 workshops, CatchMod/Harmoni-CA, 145–6 World Health Organisation (WHO), 10, 560 World Science Forum, 2003, 31, 56 worldwide cooperation, 104 WTP (willingness to pay), 62, 70

754

X-ray fluorescence (XRF) spectrometry, 383 Yellow River basin, 616 young components, 209–10 mixing with older water, 220, 232

Subject Index

YSI. See multi-parameter probes Zeeman electrothermal atomic absorption spectrometry (ZETAAS), 386 zinc, 296–7, 634–42