Sustainable Reservoir Development and Management

S O M E O T H E R TITLES P U B L I S H E D BY The International Association of Hydrological Sciences (IAHS)

in the Series of Proceedings and Reports (Red Books)

Scientific Basis for Water Resources Management. Proceedings of a symposium held at Jerusalem, September 1985 edited by M. Diskin Publ.no. 153 (1985), price £27.50 ISBN 0-947571-50-7

Water for the Future: Hydrology in Perspective. Proceedings of a symposium held at Rome, April 1987 edited by J. C. Rodda & N. C. Matalas Publ.no.164 (1987), price £32.50 ISBN 0-947571-06-X

Hydrology 2000. Report of the Hydrology 2000 Working Group (August 1983-August 1987) edited by Z. W. Kundzewicz, L. Gottschalk & B. Webb Publ.no. 171 (1987), price £14.50 ISBN 0-947571-41-8

Systems Analysis for Water Resources Management: Closing the Gap between Theory and Practice. Proceedings of a symposium held during the Third IAHS Scientific Assembly, Baltimore, Maryland, May 1989 edited by D. P. Loucks & U. Shamir Publ.no. 180 (1989), price £29.50 ISBN 0-947571-91-4

Hydrology in Mountainous Regions. II— Artificial Reservoirs; Water and Slopes. Proceedings of two symposia held at Lausanne, August 1990 edited by R. O. Sinniger & M. Monbaron Publ.no. 194 (1990), price £32.50 ISBN 0-947571-62-0

The Hydrological Basis for Water Resources Management. Proceedings of a symposium held at Beijing, October 1990 edited by Uri Shamir & Chen Jiaqi Publ.no.197 (1990), price £39.00 ISBN 0-947571-77-9

Hydrology for the Water Management of Large River Basins. Proceedings of a symposium held during the XX IUGG General Assembly, Vienna, August 1991 edited by F. H. M. van der Ven, D. Gutknecht, D. P. Loucks & K. Salewicz Publ.no.201 (1991), price £35.50 ISBN 0-947571-97-3

Modelling and Management of Sustainable Basin-Scale Water Resource Systems. Proceedings of a symposium held during the XXI IUGG General Assembly, Boulder, Colorado, July 1995 edited by S. P. Simonovic, Z. Kundzewicz, D. Rosbjerg & K. Takeuchi Publ.no.231 (1995), price £48.50 ISBN 0-947571-59-0

Application of Geographic Information Systems in Hydrology and Water Resources Management. Proceedings of the HydroGIS'96 Conference held at Vienna, April 1996 edited by K. Kovar & H. P. Nachtnebel Publ.no.235 (1996), price £58.00 ISBN 0-947571-84-1

Sustainability of Water Resources under Increasing Uncertainty. Proceedings of a symposium held during the Fifth IAHS Scientific Assembly, Rabat, Morocco, April-May 1997 edited by D. Rosbjerg, N.-E. Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F. Rasmussen Publ.no.240 (1997), price £60.00 ISBN 1-901502-05-8

FRIEND'97—Regional Hydrology: Concepts and Models for Sustainable Water Resource Manage­ment. Proceedings of the third FRIEND conference held at Postojna, Slovenia, September-October 1997 edited by A. Gustard, S. Blazkova, M. Brilly, S. Demuth, J. Dixon, H. van Lanen, C. Llasat, S. Mkhandi & E. Servat Publ.no.246 (1997), price £50.00 ISBN 1-901502-35-X

Hydrology, Water Resources and Ecology in Headwaters. Proceedings of the HeadWater'98 Conference held at Meran/Merano, April 1998 edited by K. Kovar, U. Tappeiner, N. E. Peters & R. G. Craig Publ.no.248 (1998), price £75.00 ISBN 1-901502-45-7

Orders Please send orders and enquiries concerning Red Books to:

Jill Gash, IAHS Press, Institute of Hydrology, Wallingford, Oxfordshire OX10 8BB, UK

telephone: +44 1491 692442, fax: +44 1491 692448/692424; e-mail: jilly@iahs. demon. co.uk

August 1997 Special Issue of Hydrological Sciences Journal

Sustainable Development of Water Resources edited by S. P. Simonovic

is available as a single issue, price £26.00 or US$40.00

Orders Please send orders and enquiries concerning Hydrological Sciences Journal to:

Frances Watkins, IAHS Press, Institute of Hydrology, Wallingford, Oxfordshire OX10 8BB, UK

telephone: +44 1491 692405, fax: +44 1491 692448/692424; e-mail :frances@iahs. ciemon.co.uk

All prices include delivery by surface mail. Payment may be made by credit card (VISA, MASTERCARD, EUROCARD). Invoices to customers in the USA will be in US dollars and US dollar payments will be banked through the IAHS Treasurer in the USA.

IAHS Members are offered 50% discount on Hydrological Sciences Journal (HSJ) bought for their personal use and 25% discount on other IAHS publications. To make our publications more affordable worldwide an 80% discount on HSJ and many other publications is now offered to IAHS members in financially disadvantaged countries. Individual membership of the Association is free and open to anyone who endeavours to participate in IAHS activities. Further information on membership may be obtained from: Jill Gash, IAHS Press, Institute of Hydrology, Wallingford, Oxfordshire OX10 8BB, UK [telephone: +44 1491 692442. fax: +44 1491 692448/692424; e-mail: jilly@iahs.demon.co.uk]. Information about IAHS may be found at: http://www.wlu.ca/~wwwiahs/index.html

Sustainable Reservoir

Development and

Management

by the IAHS/ICWRS Project Team (July 1993-July 1998)

Edited by

Kuniyoshi Takeuchi

Michael Hamlin

Zbigniew W. Kundzewicz

Dan Rosbjerg

Slobodan P. Simonovic

The River Environment Fund of Japan provided financial sponsorship for this publication

IAHS Publication no. 251 in the IAHS Series of Proceedings and Reports

Published by the International Association of Hydrological Sciences 1998 IAHS Press, Institute of Hydrology, Wallingford, Oxfordshire OX10 8BB, UK IAHS Publication no. 251 ISBN 1-901502-60-0

British Library Cataloguing-in-Publication Data. A catalogue record for this book is available from the British Library.

IAHS is indebted to the employers of all the Editors for the support and services provided.

The designations employed and the presentation of material throughout the publication do not imply the expression of any opinion whatsoever on the part of IAHS concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.

The use of trade, firm, or corporate names in the publication is for the information and convenience of the reader. Such use does not constitute an official endorsement or approval by IAHS of any product or service to the exclusion of others that may be suitable.

Files of the chapters provided by the editors were formatted in house style and assembled by Penny Kisby (IAHS Press, Wallingford, UK).

Printed in The Netherlands by Krips repro Meppel.

Preface

V

Reservoirs reduce peoples' dependence on the natural availability of water in rivers and streams. In this role they have been important tools for the development of civilization. They have been built throughout human history, a process accelerated and augmented by modern technology in the 20th century. In many parts of the world more reservoirs are still needed to provide a higher quality of life for the population, which may possibly double in less than a century. In recent years, it has become a common perception that reservoirs must serve for the protection of both human beings and the environment during the life time of this generation as well as of future generations. How do we construct and manage reservoirs to meet these requirements? This is the crucial question of sustainable reservoir development and management.

Much has been said about the concept of sustainable development, a UN-supported principle since its formulation in 1987 by the World Commission on Environment and Development (WCED). Although it has been widely accepted as a basic principle, concrete workable criteria by which the principle can be applied in practice are as yet unidentified. This is the problem we face in almost all plans for the construction of new dams and the management of existing reservoirs, and our societies are still waiting for a rational and scientifically-based answer. The International Commission on Water Resources Systems (ICWRS) of the International Association of Hydrological Sciences (IAHS) considered it a duty to review the problem, scientifically and practically, and provide guidelines for sustainable development and management of reservoirs.

In July 1993 during the 4th Scientific Assembly of IAHS in Yokohama, Japan, the ICWRS Bureau decided to look jointly into this important issue as a Commission Project under the title "Sustainable Reservoir Development and Management". A project team was formed consisting of, respectively, the ICWRS Bureau officers: Lev S. Kuchment, Russian Academy of Sciences, Moscow, Russia; Zbigniew W. Kundzewicz, Polish Academy of Sciences, Poznan, Poland (then at WMO, Geneva, Switzerland); Dan Rosbjerg, Technical University of Denmark, Lyngby, Denmark; Slobodan P. Simonovic, University of Manitoba, Winnipeg, Canada and Kuniyoshi Takeuchi, Yamanashi University, Kofu, Japan; the Chairperson of the Water Resource Management Section (1993-1997) of the International Association for Hydraulic Research: Peder Hjorth, Lund University, Sweden; and three Japanese scholars in the field of water resource systems management: Shuichi Ikebuchi, Kyoto University; Toshiharu Kojiri, Kyoto University (then Gifu University); and Norio Okada, Kyoto University. In addition, the following three experts agreed to contribute a case study each: Nils Roar Saelthun, Norwegian Institute for Water Research, Oslo, Norway;

vi Preface

Hussam Fahmy, National Research Institute, El-Qanatir, Egypt and; E. D. Andrews, US Geological Survey, Boulder, Colorado, USA. Research grants were sought, and the project team is most grateful to the River Environment Fund of Japan, for awarding JP Ï 1 700 000, which made it possible for the project team to meet in Kyoto in 1996.

A Mini-Workshop on the project held in Boulder, Colorado, on 9 July 1995 during the XXI General Assembly of the International Union of Geodesy and Geophysics, signified the real start of the project: specifying the programme and the role of each member. Based upon those discussions and subsequent correspondence, a draft report was compiled in August 1996 and discussed at the International Workshop on Sustainable Reservoir Development and Management in Kyoto on 1 November 1996. This event was attended by about 20 specialists and organized during the International Conference on Water Resources and Environment Research. Further work was suggested, and another round of discussions took place. A draft final report was compiled in March 1997 and discussed in Rabat, 28-30 April 1997, during the 5th Scientific Assembly of IAHS. At this stage, Professor Michael Hamlin, a former President of ICWRS, joined the team and agreed to act as an additional editor for the complete volume. As a result of the discussions in Rabat, some major revisions were agreed in order to make the document comprehensible to a wider audience. Another three rounds of improvements and editorial review took place before it was finally submitted for publication in July 1998.

It took five years to complete the project since the first discussions, or three years from the start of the real work—much more than expected! We believe, besides the busy schedules of the team members, that this is a manifestation of both the topic complexity and the difficult process of reaching consensus between different views on the possible approaches towards sustainability of reservoir management.

The project team is indebted to many individuals and institutions. First of all, the team would like to express its deep gratitude to the individual contributors to the report. All the authors and co-authors are listed in the table of contents. Without their contributions, this project could not have been completed. Also, we thank the many individuals who provided valuable comments and suggestions. Last but not least, we appreciate the continuous encouragement of the IAHS Bureau, including the financial support of US$5000 from the IAHS 1996 Interdisciplinary Research Fund. With all this support the project became feasible.

The project team wishes this report to serve for the improvement of current practice and for stimulation of further research on sustainable reservoir development and management.

Editor-in-chief: Kuniyoshi Takeuchi Co-editors: Michael Hamlin

Zbigniew W. Kundzewicz Dan Rosbjerg

July 1998 Slobodan P. Simonovic

Contents

vii

Preface v List of contributors x

CHAPTER 1 Introduction

(edited by K. Takeuchi & D. Rosbjerg)

1.1 Objective K. Takeuchi, Z. W. Kundzewicz, D. Rosbjerg & S. P. Simonovic 1

1.2 Background and scope K. Takeuchi, Z. W. Kundzewicz, D. Rosbjerg & S. P. Simonovic 1

1.3 Current status of reservoirs 4 1.3.1 Number, size and shape of reservoirs K. Takeuchi 4 1.3.2 Land efficiency discussion of large reservoirs K. Takeuchi & S. P. Simonovic 7 1.3.3 Reservoir sedimentation K. Takeuchi 9

1.3.4 Factors controlling the future needs of reservoirs K. Takeuchi & S. P. Simonovic 10

1.4 Critiques of present reservoirs P. Hjorth, Z. W. Kundzewicz, L. S. Kuchment & D. Rosbjerg 14 1.4.1 Introduction 14 1.4.2 Environmental effects of reservoir construction 16 1.4.3 Socio-cultural and institutional problems 20 1.4.4 Problems related to the use of reservoirs 22

1.4.5 Concluding remarks 25

References 28

CHAPTER 2 Sustainability and Reservoirs

(edited by K. Takeuchi & Z. W. Kundzewicz)

2.1 Notions of sustainable development 31 2.1.1 Introduction K. Takeuchi & Z. W. Kundzewicz 31 2.1.2 Sustainable development: perspective of WCED and UNCED K. Takeuchi &

Z. W. Kundzewicz 31

2.1.3 Indicators of sustainability Z. W. Kundzewicz 36 2.1.4 Follow-up discussions of sustainable development K. Takeuchi, S. P. Simonovic

& Z. W. Kundzewicz 37

2.2 Sustainable development and management in the reservoir context 43 2.2.1 Introduction K. Takeuchi & Z. W. Kundzewicz 43 2.2.2 Integrated water resources management K. Takeuchi & P. Hjorth 44 2.2.3 Notion of multiple objectives P. Hjorth, Z. W. Kundzewicz & S. P. Simonovic 47

vin Contents

2.2.4 Risk- and uncertainty-related considerations P. Hjorth, Z. W. Kundzewicz &

K. Takeuchi 51

2.2.5 Systems view S. P. Simonovic 54 2.2.6 Sustainability criteria for possible use in reservoir analysis S. P. Simonovic 55 2.2.7 Rationale of the checklist for sustainable reservoir development and management

K. Takeuchi 58 References 60

CHAPTER 3 Comparative Assessment of Reservoirs with Non-reservoir

Alternatives (edited by Z. W. Kundzewicz)

3.1 Introduction Z. W. Kundzewicz &. P. Hjorth 63

3.2 Multipurpose reservoirs versus alternatives Z. W. Kundzewicz, P. Hjorth & K. Takeuchi 64 3.2.1 Water demand and supply 66 3.2.2 Energy 72 3.2.3 Flood control 76 3.2.4 Other purposes 77

3.3 Final remarks Z. W. Kundzewicz &. P. Hjorth 78

References 80

CHAPTER 4 Design and Management of Reservoirs

(edited by S. P. Simonovic & K. Takeuchi)

4.1 Introduction S. P. Simonovic 81 4.1.1 Structure of the chapter 81 4.1.2 Dimensions of reservoir analyses 82

4.2 Hydrological input for reservoir design and management 84 4.2.1 Time series analysis of reservoir inflows L. S. Kuchment 84 4.2.2 Use of inflow forecasts for efficient management of reservoirs K. Takeuchi 89

4.3 Methodological contributions to the design and management of sustainable reservoirs 95 4.3.1 New method for the design of a sustainable reservoir A. S. Kotula &

S. P. Simonovic 95

4.3.2 Methodology for reassessment of existing reservoirs S. P. Simonovic 106 4.3.3 Methodology for net benefit allocation for reservoir redevelopment N. Okada &

H. Sakakibara 112 4.3.4 The least marginal environmental impact (LMEI) rule for reservoir sizing

K. Takeuchi & M. M. Hufschmidt 120

4.4 Conclusion S. P. Simonovic 124

References 126

Contents ix

CHAPTER 5 Case Studies

(edited by D. Rosbjerg)

5.1 Introduction D. Rosbjerg 129

5.2 The Alta hydropower development scheme N. R. Saelthun 130 5.2.1 Introduction to the Alta case 130 5.2.2 The river 130 5.2.3 Undercurrents in society 132 5.2.4 The Alta hydropower scheme 134 5.2.5 "The Alta affair" 136 5.2.6 Impacts of reservoir development 138 5.2.7 Design and management of reservoirs 140 5.2.8 Comparative assessment of the reservoir with non-reservoir alternatives 141 5.2.9 Sustainability issues 141

5.3 Aswan High Dam H. Fahmy 142 5.3.1 Introduction to the Aswan case study 142 5.3.2 Inundation of populated areas 144 5.3.3 Sedimentation problems 147 5.3.4 Conclusion on the Aswan case study 150

5.4 Management of annual peak flows to restore aquatic resources in Green River, Utah E. D. Andrews & L. A. Pizzi 151 5.4.1 Introduction to the Green River case study 151 5.4.2 Historical development of Colorado River basin water resources 153 5.4.3 Hydrology of the Green River basin 155 5.4.4 Ecological and geomorphological alteration of the Green River 161 5.4.5 Anticipated benefits of increased flood magnitude and duration 163 5.4.6 Conflicts and opportunities for increased flooding 164 5.4.7 Conclusion on the Green River case study 166

5.5 Projects undertaken to achieve sustainable development of dam reservoirs in Japan S. Ikebuchi, N. Okada & H. Sakakibara 167 5.5.1 Introduction to the Japanese reservoir policy 167 5.5.2 Projects related to sustainable development of reservoirs 167 5.5.3 Legal arrangements for social impacts of dam construction projects 169 5.5.4 Water quality conservation measures for dam reservoirs 170 5.5.5 Rehabilitation of an existing reservoir 171 5.5.6 Improvement of the management and operation of dam reservoirs 173 5.5.7 Regional promotion and regional planning centred around dams 174 5.5.8 Other issues 177 5.5.9 Conclusion on the Japanese study 178

5.6 Concluding remarks on the case studies D. Rosbjerg 179

References 180

X Contents

CHAPTER 6 Checklists and Concluding Remarks

K. Takeuchi, Z. W. Kundzewicz, D. Rosbjerg, S. P. Simonovic

6.1 Checklists for sustainable reservoir development and management 183

6.2 Concluding remarks 187

List of Contributors

Edmund D. Andrews, US Geological Survey, 3215 Marine Street, Boulder, Colorado 80303, USA [e-mail: eandrews@usgs.gov]

Hussam Fahmy, National Research Institute, PO Box 6, El-Qanatir 13621, Egypt [e-mail: nwsru@link.com.eg]

Peder Hjorth, Department of Water Resources Engineering, Lund University, Box 118, S-221 00 Lund, Sweden [e-mail: peder.hjorth@tvrl.lth.se]

Maynard M. Hufschmidt, 19191 Harvard Avenue, Apt. 107E, Irvine, California 92612, USA

Shuichi Ikebuchi, Water Resources Research Centre, Disaster Prevention Research Institute, Kyoto University, Gokasho, Uji 611-0011, Japan [e-mail: chief@wrcn2.dpri.kyoto-u.ac.jp]

Toshiharu Kojiri, Water Resources Research Centre, Disaster Prevention Research Institute, Kyoto University, Gokasho, Uji 611-0011, Japan [e-mail: tkojiri@wrcn2.dpri.kyoto-u.ac.jp]

Agnes S. Kotula, ACRES International, 4342 Queen Street, PO Box 1001, Niagara Falls, Ontario, Canada L2E 6W1 [e-mail: akotula@nf.acres.com]

Lev S. Kuchment, Water Problems Institute, Russian Academy of Sciences, 10 Novo-Basmannaya, Moscow 107078, Russia [e-mail: kuchment@iwapr.msk.su]

Zbigniew W. Kundzewicz, Research Centre of Agricultural and Forest Environment, Polish Academy of Sciences, Bukowska 19, 60-809 Poznan, Poland [e-mail: zkundze@man.poznan.pl]

Norio Okada, Disaster Prevention Research Institute, Kyoto University, Gokasho, Uji 611-0011, Japan [e-mail: okada@@wrcn2.dpri.kyoto-u.ac.jp]

Leslie Abrams Pizzi, US Geological Survey, 3215 Marine Street, Boulder, Colorado 80303, USA [e-mail: lapizzi@brrcrmail.cr.usgs.gov]

Dan Rosbjerg, Department of Hydrodynamics and Water Resources (IS VA), Technical University of Denmark, Building 115, DK-2800 Lyngby, Denmark [e-mail: dan@isva.dtu.dk]

Nils Roar Saelthun, Norwegian Institute for Water Research, Brekkeveien 19, N-0411 Oslo, Norway [e-mail: nils.saelthun@niva.no]

Hiroyuki Sakakibara, Department of Civil Engineering, Yamaguchi University, Tokiwadai, Yamaguchi 755-8611, Japan [e-mail: sakaki@rmsl.dpri.kyoto-u.ac.jp]

Slobodan P. Simonovic, Natural Resources Institute and Civil and Geological Engineering Department, University of Manitoba, Winnipeg, Canada R3T 3N2 [simon@ce.umanitoba.ca]

Kuniyoshi Takeuchi, Department of Civil and Environmental Engineering, Yamanashi University, Kofu 400-8511, Japan [e-mail: takeuchi@mail.yamanashi.ac.jp]

Sustainable Reservoir Development and Management. Report by the IAHS/ICWRS Project Team. IAHS Publ. no. 251, 1998. ]

CHAPTER 1

INTRODUCTION

1.1 OBJECTIVE

The sustainability principle, to be discussed in detail in Chapter 2, calls for an integrated consideration of economic, social and environmental issues related to development, while addressing both intra- and inter-generational equity.

The objective of this report is to translate the general sustainability principle into the reservoir context and to provide a set of concrete planning, design and management criteria that could be used for selection of a reservoir option that compares favourably to other reservoir and non-reservoir alternatives.

In order to fulfil this objective, the present report has the following aims:

(1) To review the current status and anticipated future needs for reservoirs in order to develop a new approach that can address existing criticisms of reservoirs.

(2) To analyse the general notion of sustainable development and to interpret it in the context of reservoirs.

(3) To discuss non-reservoir alternatives to be integrated with reservoir options for sustainable water resources management.

(4) To present selected reservoir design and management techniques that can help implementation of the sustainability principle.

(5) To show selected examples of design and management of existing reservoirs highlighting different aspects of sustainability.

(6) To present a checklist which may guide sustainable reservoir development and management in practice.

1.2 BACKGROUND AND SCOPE

Water storage is the basic means of controlling water availability over time and in space. Storage of water, needed in vast quantities in virtually every human activity, is essential for human beings as well as other living species.

2 Introduction

The bodies of camels and cacti are adapted for the function of storing water, indispensable for sustaining life. Different storage media have existed naturally in the world or have been developed by humans and utilized throughout history. While groundwater aquifers, lakes, snowpacks, soil moisture etc. are natural water reserves, a manmade reservoir is a structural means for storing water. Humans have been building reservoirs over thousands of years. The oldest ruin of a dam is Sadd El-Kafara 30 Ion south of Cairo, Egypt (length 104 m, height 11 m and storage capacity 0.57 1 0 6 m 3 ) built around 2800 BC for water supply and believed to be destroyed by its first flood (Biswas, 1970). One of the oldest existing reservoirs in the world is Mannoh-Ike in Japan (height 32 m, area 1.4 km 2 and storage capacity 15.4 10 6 m 3 ) that was built prior to the 9th century AD and is still used for paddy field irrigation.

There are a great number of large dams all over the world, most of which have been built in the 20th century. The present rate of construction is of the order 250 dams a year (ICOLD, 1988). It is projected that the world population may reach 8 billion by 2020 and 10 billion by 2050 with dramatic consequences for enhanced urbanization and increased food and drinking water requirements to provide safe living conditions for all people. The demands for irrigation water, municipal and industrial water supply, flood control, hydropower generation and navigation become tremendous. Undoubtedly, reservoirs are among the most important physical means to meet such needs.

There is increasing concern about the social and environmental effects of reservoirs. Some experts claim that large reservoirs built from the late 1950s to the 1970s, mainly for hydropower, are economically inefficient and have strong negative ecological impacts. The same holds for some irrigation reservoirs in arid and semiarid regions. Many undesirable cases are reported by Pearce (1992). In developed countries, the environmental concern is increasing to such a level that practically no new major reservoirs may be constructed and in some cases existing dams are going to be removed to restore the original flow regime of the river for natural aquatic habitats. The construction of the Three Gorges Dam in China was finally approved in 1992 and is now underway. The dam will be 185 m high and the reservoir, on the average, 2 km wide and 600 km long. The primary purposes are flood control, hydroelectric power generation and navigation. While the economic expectations are large, some believe that the benefits are overestimated and are concerned with adverse socio-cultural and environmental effects, including resettlement of more than a million people and potential sedimentation problems (Dai Qing, 1996).

Criticism of reservoirs does not come only from extreme environmentalists. It is reflecting the global concern for the Earth environment. The World Commission on Environment and Development (WCED) introduced the sustainable development principle (WCED, 1987)

Introduction 3

which subsequently was turned into the UN action plan, Agenda 2 1 , by UNCED (1992). The concept of sustainable development with intra- and inter-generational equity motivated various leading policy agencies to reconsider the environmental impacts of their projects. For instance, the World Bank changed its funding policy for water projects (Cernea & Le Moigne, 1989), and the US Bureau of Reclamation shifted their major concern from development to management in the 1990s (Beard, 1994). With these new policies, financial and other institutions, national and international, are now expressing support for the implementation of the sustainability principle.

It is a plain fact that more freshwater is necessary for the future development of society and support of the increasing world population. On the world scale, new reservoirs will be needed as one of the most effective means of acquiring and controlling water. However, they have to be built and managed in a sustainable way. How should the necessary reservoirs be designed, constructed and managed to meet the requirements for sustainable development? How can existing reservoirs be managed efficiently, reliably and flexibly under increasing uncertainty? How can the reservoirs be rehabilitated and their life time prolonged for increased safety and better performance? The answers are not simple, but much needed.

There have been attempts to apply the general sustainability principle to water management, but very few have addressed reservoirs in particular. Partly this may be due to the difficulties in assessing the needs for, and the environ­mental impacts of reservoirs. Emphasis on environmental concerns often results in an oversimplified view and one-sided arguments against reservoirs. Instead, appropriate attention should be given to more environmental care in the management of reservoirs. This report intends to provide the scientific background for application of the sustainability principle to reservoirs.

In the rest of Chapter 1, the current status of the basic statistics of reservoirs is reviewed to identify where we are and what the problems are. Section 1.3 provides the information on the current status and the expected future needs of reservoirs and Section 1.4, various critiques of present reservoirs. A fairly comprehensive review of the criticisms is made since this constitutes the basis for the new proposal of this report.

In Chapter 2, the notion of "sustainable development" is reviewed and translated into the reservoir context. Section 2.1 discusses the key components of the principle and the related follow-up discussions, and Section 2.2 translates the basic principle into reservoir design and management strategies. Furthermore, based upon these interpretations, it presents some concrete proposals and discussions on criteria and an introduction to a checklist that may be useful in practice to achieve sustainable reservoir development and management.

The following two chapters, 3 and 4, discuss the two basic components of sustainable reservoirs, i.e. integrated water resources management, which

4 Introduction

includes the non-reservoir option as an important component, and the systems approach for design and management of reservoirs. Chapter 3 compares reservoirs with non-reservoir alternatives attempting to judge their advantages and disadvantages in an objective way. Typically, a reservoir serves multiple purposes in the presence of conflicting objectives, so that when comparing alternatives, a number of single-purpose projects must be compared to a multi­purpose reservoir. Demand management is considered important wherever it is possible as an alternative to or a combinatorial part of reservoir options.

Chapter 4 presents the engineering and systems analytical techniques for efficient use of existing reservoirs. Section 4.1 discusses some basic principles of reservoir analysis, as well as basic requirements imposed by the sustaina-bility paradigm. Section 4.2 illustrates the importance of the hydrological input for reservoir design and management. Section 4.3 illustrates four possible ways to address the requirements of sustainability in reservoir analysis: De Novo programming for reservoir development, reassessment of existing reservoirs, cost/benefit allocation for reservoir redevelopment and the Least Marginal Environmental Impact (LMEI) rule for reservoir sizing.

Chapter 5 presents case studies of existing reservoirs in Norway, Egypt, USA and Japan, the current status of reservoir management and the directions of the future managerial innovation. They are not necessarily examples of sustainable reservoirs that this report considers as model cases. Rather, it is admitted that there are no model cases. The case studies show how each reservoir is struggling with the changing needs of society which may serve as lessons in other situations.

Chapter 6 concludes this report with the authors' views on the current status of reservoirs and the directions in which to go in the form of a set of checklists to be used as a practical reference at respectively the planning, design, construction, operation and maintenance stages.

1.3 CURRENT STATUS OF RESERVOIRS

In this section, the current status of reservoirs is reviewed in respect of numbers and sizes, land efficiency with respect to hydropower generation, sedimentation and the future needs of reservoirs.

1.3.1 Number, size and shape of reservoirs

According to the World Dam Register (ICOLD, 1988), there were 36 235 large dams (defined as higher than 15 m) in the world in 1986 in a total of 133 countries (79 ICOLD member countries and 54 non-member countries). There were 427 dams in 1900 and about 29 900 dams were built between 1951 and

Introduction 5

H H - f r H 1 1 1 1 1 1 1 0 3 10- 2 10-' 10° 10' 10 2 10 3 lu 4 10 5

INUNDATED AREA A [km2] Fig. 1.1 Inundated area and gross storage capacity of reservoirs in the world (after Takeuchi, 1997).

1982. The rate of dam construction is decreasing in the long term, but in recent years it has not necessarily decreased. The average number of dams built during 1983-1986 was 267 annually. The total number of dams now must be approaching 39 000. The number of dams under construction in the world was about 1242 in 1994; 6.36% of which were dams higher than 100m. For comparison, dams higher than 100 m constituted 1.1% of the 36 235 dams existing in 1986 (Veltrop, 1995).

Following the Russian tradition of global hydrology, Avakian (1990) estimated the total volume of reservoir capacities and the total inundated area in the world to be of the order of 6000 km 3 and 400 000 km 2 , respectively.

6 Introduction

The storage element corresponds to around 5 % of the total precipitation on the land (119 000 km 3 ) or 13% of the total runoff from the land (46 000 km 3 ) (UNESCO, 1978) and the inundated area corresponds to the area of France. Thus quite a large part of the world's freshwater is stored in manmade reservoirs and the proportion is still increasing.

Figure 1.1 shows the relationship between the gross capacity (V; 10 6 m 3 ) and the inundated area (A; km 2) of 7936 reservoirs (7602 dams from World Dam Register 1988 (ICOLD, 1988) and 334 dams from the Yearbook of Dams 1990 (Japan Dam Association, 1990), selected from those that have both capacity and inundated area data). Although some data are not necessarily reliable as the average water depth becomes greater than 1000 m or less than 10 cm, it clearly demonstrates the general tendency of the V-A relation of reservoirs in the world. The average relation is:

V=9.IA1-1 ( i . i )

which was derived by the least squares method with respect to the distance normal to the regression line (instead of the distance in the vertical coordinate to the line). The coefficient of determination was R2 = 0.869 and the root mean square error or the standard error of estimate was SEE = 0.4137.

Although this line is a combined average of many reservoirs in many locations, a standard topography can be conceived where the land surface is covered by valleys of a common shape. Namely if the longitudinal gradient of the river is 9 and the valley is ideally prismatic with h — a xb, where x is the horizontal distance normal from a river and h is the elevation of the land surface above the riverbed. In such an open book shape, shown in Fig. 1.2, the A-Vrelation of reservoirs (1.1) implies:

Aozh*% and V°cho)S (1.2)

where h0 is the dam height (Takeuchi, 1997). A high power in (1.2) means that progressively more land is inundated when reservoirs become larger. Thus there is little scale merit in constructing large reservoirs as far as the environmental effects due to inundated area are concerned. It may be fair to say that from the ecological point of view there is a scale demerit in large

o Fig. 1.2 A hypothetical valley with cross-section h = ω and gradient q (after Takeuchi, 1998).

Introduction 7

reservoir construction. Takeuchi (1997) demonstrated regional differences of land efficiency in reservoir size. He noted that in many large reservoirs in continental countries like the former USSR, Brazil, Thailand and Mexico, more land was inundated than the global average indicated by (1.1) corresponding to the same storage.

1.3.2 Land efficiency discussion of large reservoirs

Table 1.1 shows some of the largest reservoirs of the world. It is quite striking that the Akosombo Dam on the Volta Lake alone stores 150 km 3 of freshwater inundating an area of 8500 km 2 . It is also noticeable that most large reservoirs are primarily for hydropower generation. The High Aswan and the Cabora Bassa dams are primarily for irrigation and secondly for hydropower generation but would not be so large if hydropower were not included. This is true in most of the gigantic reservoirs. They tend to become gigantic to obtain enough head, to produce economically justifiable energy. A typical example is the Srinagarind Reservoir of the Mae Klong River, Thailand, which has a total capacity of 17 745 1 0 6 m 3 but an effective storage of 7481 10 6 m 3 . The percentage of dead storage is as much as 58%.

There are many criticisms of large reservoirs not only from the social and ecological but also from the economic point of view. In some cases it is difficult to verify that the long-term benefit obtained by reservoir construction exceeds

Table 1.1 Some large dams in the world and the Japanese total.

Name of dam (river, country) Capacity Inundated area Purposes (1cm3) (km2)

1 Bratsk (Angara, Russia) 169.0 5470 HNS 2 High Aswan (Nile, Egypt) 162.0 6500 IHC 3 Kariba (Zambezi, Zambia) 160.0 5100 H 4 Akosombo (Volta, Ghana) 148.0 8480 H 5 Daniel Johnson (St Lawrence, Canada) 142.0 2000 H 6 Guri (Orinoko, Venezuela) 135.0 4250 H 7 Krasnoyarsk (Lena, Russia) 73.3 2000 H 8 WAC Bennet (Mackenzie, Canada) 70.3 1650 H 9 Zeya (Amur, Russia) 68.4 2420 HNC

10 Cahora Bassa (Zambezi, Mozambique) 63.0 2580 IHC 11 Kuibyshev (Volga, Russia) 58.0 6150 HNIS 12 Rybinsk (Volga, Russia) 25.4 4550 HNS Total of 2575 Japanese reservoirs in 1994 18.7 1250

H: Hydropower; I: Irrigation; C: Flood control; N: Navigation; S: Water supply. 1-10 are the ten largest reservoirs of the world in terms of gross capacity. 11 and 12 are two large reservoirs in terms of inundated area. After: ICOLD (1988) and Japan Dam Association (1996); effective digits were adjusted.

8 Introduction

the cost of destroying the natural regime and the human activities developed over hundreds of years. If the marginal environmental and socio-economic loss of a reservoir is calculated, few would exceed the marginal benefit of the pro­duction at the margin of the capacity because dA/dh is very large at h = h0.

Table 1.2 compares the hydroelectric energy generated in the eight countries where the reservoirs listed in Table 1.1 belong. The Japanese figure includes not only energy generation at dam sites but also run-of-the-river and pumped generation schemes. Although the climate, topography and other conditions for hydropower generation are very different in different countries, it is still a surprising fact that the hydroenergy single-purpose Akosombo Reservoir alone inundates an area of tropical rainforest nearly 7 times more than the total of the 2575 reservoirs of Japan. Ghana produces about l/20th of the hydroelectric power of Japan, a factor which is more than 100 in land surface efficiency for electricity generation. Five reservoirs in Russia listed in Table 1.1 inundate more than 15 times the total of Japanese reservoirs. Russia as a whole produces less than 3 times the hydroelectricity of Japan. Other countries show similar trends.

Japanese high efficiency in hydroelectric power generation with respect to inundated area is due to the extensive development of run-of-the-river type hydropower generation. This takes water from a river into an artificial channel or tunnel leading to a site where penstocks can be built and electricity can be generated with a high head. In this way large amounts of hydroelectricity can be generated without constructing a major dam and with only a small water intake weir. This type of hydroelectric generation is possible only in a mountainous country with steep slopes and considerable rain all year round and is impossible in relatively flat continental countries.

The first study of land efficiency on hydropower generation was presented by Goodland (1990) which showed the installed capacity of hydroelectric power generation per unit inundated area, under normal conditions, in kW ha' 1

at various dam sites. As he admitted, the actual electric energy generated is a

Table 1.2 Production of hydroelectric energy in 1997.

Country Hydroelectric energy generated (GWh)

Brazil 250 000 Canada 331 000 Ghana 6 100 Japan 91 300 USA 296 000 Former USSR 228 000 Venezuela 63 000 Zambia 8 100

After: World Atlas and Industry Guide (1997); effective digits were adjusted.

Introduction 9

better indicator but the data are not necessarily available. Comparison of Tables 1.1 with 1.2 is the best we can do so far.

It is important to note that a simple comparison, as above, may be misleading if it is narrowly interpreted. First, not all the giant reservoirs are constructed for hydropower alone. An evaluation should be based on the overall outcomes and impacts. Second, the alternatives available for each country to secure electric energy are restricted in various respects, for natural, social and economic reasons. Each country has to generate power by whatever means available. Third, the land efficiency cannot be assessed only by inundation. Inundation may save other kinds of degradation of land environment, such as by thermal pollution, increased C 0 2 , nuclear risk potential etc. Thus, the trade-off goes beyond the simple consideration of reservoirs.

1.3.3 Reservoir sedimentation

Reservoirs receive sediment continuously from upstream. However small the sedimentation rates are, sediment accumulates and will eventually fill the whole reservoir if not deliberately released. Table 1.3 lists some statistics available on reservoir sedimentation. According to Milliman & Meade (1983), the world average annual sediment discharge to the sea is 13 billion tonnes, and according to Meybeck (1988), 2-5 billion tonnes are trapped behind dams each year. If this figure is correct, the 6000 km 3 of world reservoir storage will be filled on average in 1400-3600 years assuming that 1.2 tonnes of sediments occupy 1 m 3 . This sounds very far into the future if only average conditions are considered.

Lake Mead in USA is losing storage at the rate of 0.33% each year (Gottschalk, 1964) implying 300 years before it is completely filled. In the case of the Aswan High Dam, according to Shahin (1985), the mean annual suspended solid load passing through the Aswan site was, prior to the

Table 1.3 Reservoir sedimentation rates.

Reservoir Total capacity Annual sediment Life time (10 6 m 3) (10 6m 3year"') (years)

729 Japanese reservoirs* ( > 106 m 3) 17 300 40.8 395 Lake Nasser (Aswan High Dam) 162 000 109.0f 1600 Lake Mead (Hoover Dam) 34 900 0.33%$ 300 World reservoirs 6 000 000 2-5 10' t year i 1400-3600

* Kobayashi (1996); effective digits were adjusted. t Said (1993). X Gottschalk (1964). t Meybeck (1988).

10 Introduction

construction of the Aswan High Dam, 125 million tonnes, 98% during floods. After construction, 2.5 million tonnes are now flowing downstream from the dam. This implies that 30 km 3 of dead storage set aside for sediments, out of the total 164 km 3 , will be filled in some 300 years and the whole dam will be filled in 1300 years. The estuary sand bank at the mouth of the Nile retreated 4 km in 30 years since the completion of the Aswan High Dam in 1964 (Biswas, 1992). Sedimentation, together with the sea water intrusion, is one of the major reasons that Vietnam opposes the development of dams on the main channel of the Mekong River.

Sediment yields, however, differ greatly region by region. In East and Southeast Asia where about two thirds of world sediments discharge to the sea, there are many reservoirs filled in a much shorter time. In northeastern Thailand, a number of reservoirs built for irrigation purposes were filled soon after completion with sediments derived from deforested areas. There are a number of dams in Japan, mostly smaller than 10 7 m 3 , built on very steep mountains for hydroelectric power generation that often experience a rapid filling by sedimentation, for example within 10 years of completion. Although the head for hydropower generation does not change, small effective storage results in little temporal adjustability of power generation.

Various sediment flushing techniques such as sediment outlet gates and bypass channels have been developed. Most methods are still at an experimental stage. The example of the Dashi-Daira Dam of the Kurobe River, Japan, provides an important lesson (Investigation Committee on Sediment Release Impact of Dashi-Daira Dam of the Kurobe River, 1995). It is a concrete gravity dam 76.7 m high with 901 10 6 m 3 storage for hydropower generation completed in 1985 with two built-in 5 m x 5.5 m sediment discharge gates at 30 m from the bottom. After large landslides upstream, attempts were made to release sediments from the gates in 1991. This had an unexpected impact. The sediment contained a lot of humus and logs which under anaerobic conditions resulted in the production of hydrogen sulphide gas which was carried 28 km to the sea. The sediments were responsible for destroying many aquatic weeds in the sea and putting local fishermen out of work. This is not necessarily an extraordinary case, similar experiences have been reported elsewhere in Japan.

1.3.4 Factors controlling the future needs of reservoirs

Main controlling factors for the future needs of reservoirs would be the growth of developing countries, the environmental conservation movement in developed countries, the need for hydroelectric power energy and the anticipated climate variability and change, which will be briefly reviewed below.

Introduction 11

Developing countries In developing countries where the population growth is expected to double in the next half century the need for additional water resources is tremendous. It is necessary for food production for sustaining the population, for grazing pasture for cattle for higher food standards, for supplying mega-cities where more and more people and activities are concentrated and for hydroelectric power necessary for industrialization. Most of the extra water supply will come from surface water, because groundwater depletion can cause serious land subsidence in humid regions and non­renewable groundwater extraction in arid regions is too costly. Surface water can only be stored and transferred by means of reservoirs. Regardless of the environmental concern, their development will be indispensable. The only issue will be how to construct reservoirs which are acceptable in terms of sustainable development.

In a humid region, reservoirs are also necessary for flood control. Loss of life, property and agricultural products cause serious social problems, as seen in Bangladesh, China, Philippines, Korea and also many developed countries such as Japan, USA and several European countries.

Developed countries The situation is quite different in developed countries where the population is becoming stabilized and the most economic dam sites have already been developed. The municipal water demand is still increasing due to increasing standards of living, but agricultural and industrial water demands are becoming stabilized because there is little further expansion of irrigated land and the rate of recycling of industrial water has been increasing (77% in Japan, 1994).

Water shortages are, however, fairly common in developed countries, too. Some Japanese cities are suffering from water shortages almost every year, due to frequent dry spells on a large regional scale. A large part of Japan, especially the southwestern part, had only 50-60% of the average annual precipitation in 1978, 1989 and 1994, and some experts speculate that this is connected with global climate change. Such a tendency naturally creates the demand for further water resources development for large urbanized areas.

However, it seems that the public are not sympathetic to the construction of more dams. On the contrary, the trend is that more and more people demand better management of water for safe and efficient water use. Efficient allocation of water rights among users, a priority use for urgent needs such as a temporary re-allocation of agricultural or hydropower water rights for urban water use during drought periods, the emergency import of water from other regions as well as various water saving operations are now options being increasingly supported by broad public opinion. The municipal water use in northern Kyushu, which has experienced some of the worst droughts that have occurred in Japan several times in the last 20 years, is less than 300 litres per day per capita, the lowest in Japan excluding the cool region of Hokkaido (National Land Agency, 1995). This clearly reflects the recent global

12 Introduction

environmental movement. Recycling of goods, concern for the life cycle of artificial goods, preservation of nature and the like are becoming a common virtue in the consciousness of the people. Water is no exception.

Nonetheless there are about 600 new dam construction plans endorsed by various levels of governments in Japan. Many of them are in conjunction with flood control, which is the responsibility of the central government and the local government has the incentive to get subsidies. Its 100% implementation may, however, be far in the future if indeed realized at all. A large number of rainfall-induced disasters in Japan are localized landslides and mudflows for which large reservoirs are of no help.

At the 18th ICOLD Congress, Beard (1994), the Bureau of Reclamation Commissioner at that time, stated that "the dam building era in the United States is now over". He continued: "We will emphasize water conservation, demand management and efficient use, including reuse, whenever possible. Every problem we must address has a common theme. That is: there isn't enough water in the river. This sounds elementary, but it isn't. Most western streams [in USA] are over-allocated and under stress. Excessive use has been condoned—even encouraged—and legitimate in-stream uses have been ignored or prohibited. To solve these problems, we cannot build new reservoirs. Instead, we will have to encourage the movement of water from one use to another. We believe conservation, demand management, efficiency improvements, and reuse offer the best opportunities for doing this, if structured to provide real economic benefits to all participants."

The Elwha Dam (132 m high; built in 1913) and the Glines Dam (64 m high; built in 1927) on the Elwha River, Olympic, Washington will be demolished based on the Elwha River Ecosystem and Fisheries Restoration Act of 24 October 1992. Although the Congress has not yet made an appropriation for the removal costs, the service of the dams for hydroelectric power generation, extending three quarters of a century, will come to an end, and the original flow regime of the river for natural fish habitats, such as salmon on which the S'Klallam Indians used to live, will be restored (Sumi, 1996). There are in fact rather special reasons for the decision, the main one being the Treaty of 1855 between the US Government and the S'Klallam Indian Tribe which ensured the tribe the fishing rights in exchange for the right to the land (Nagase, 1998; http : / /www. olympus. net/personal/sklallam/index. htm).

Although the decision regarding Elwha River rhymes with Beard's (1994) statement, it is by no means evident that the statement has been generally accepted in the USA. Nevertheless, in developed countries an environmental concern is now prevailing, which has a great impact also on reservoir construction for water supply, flood control and hydropower. The social preference for human adjustment instead of continued emphasis on the control of nature reflects a movement of society towards a new paradigm for human existence.

Introduction 13

Table 1.4 Technically exploitable hydropotential and hydrogeneration in 1997.

Region Technically exploitable hydropotential (GWh year"1)

Hydrogeneration in 1997 (GWh)

Hydropotential used (%)

World 14 000 000 2 460 000 18 Europe* 1 190 000 511 000 43 North/Central America 1 500 000 668 000 45 South America 2 600 000 450 000 17 Former USSR 2 510 000 228 000 9 Asiat 2 260 000 236 000 10 China 1 920 000 167 000 9 Japan 134 000 91 000 68 Africa 1 670 000 67 000 4 Australia/Oceania 270 000 41 500 15

After: World Atlas and Industry Guide (1997); classification and effective digits were adjusted. * Excluding former USSR, t Excluding China and Japan.

Hydroelectric power Hydroelectric power is the other major potential source of reservoir needs. Especially in developing countries, hydroelectric power is one of the key elements for economic development and an important means of getting foreign currency. It is also expected to be a clean energy source together with the use of biomass, geothermal, wind, wave, solar and other methods of power generation. According to Table 1.4, after the World Atlas and Industry Guide (1997), the technically exploitable hydroelectric power is estimated as 14 10 6 GWh year"1 and that currently exploited is 2.46 10 6 GWh year"1, 18% of the potential. The potential is particularly large in many developing countries.

Note that the primary electricity is the gross energy used for producing electricity. It was assessed at the equivalent of 1.0 efficiency for wind and hydroelectric generation (at the heat value of electricity 1 kWh = 3.6 10 6 J), at 0.33 for nuclear and 0.1 for geothermal.

It should be noted, however, from the figures in Table 1.5, that the energy consumption of the world is so large that even a full development of

Table 1.5 Commercial energy production in 1991 (in peta J = 10 J).

Total Liquid Gas Solid Primary electricity: Hydro Nuclear Geothermal Hydro

and wind

334 890 93 689 132 992 76 275 8 049 22 669 1261 100% 28.0% 39.7% 22.8% 2.4% 6.8% 0.38%

After: World Resources Institute (1994).

14 Introduction

hydropotential would barely solve the energy problem. The world energy consumption is estimated as 330 000 peta J in 1991 by the World Resources Institute (1994), of which hydroenergy is 8000 peta J or only 2.4% and nuclear energy is only 6.8% based on gross energy or 2.0% based on generated electricity. The proportion of hydroenergy is even less in developing countries because of the construction of fewer large dams. It would therefore mean significant new construction if their huge hydropotential were to be fully utilized. But, at the same time, it means that even when those countries become developed, hydropower can only play a minor role in global energy production.

This raises the question of how hydropower development can justify its severe ecological effects when it can only meet such a small proportion of the total energy demand. It is said that the impact of reservoirs is not a global issue by comparison with fossil fuels and nuclear energy which have such a direct impact on the global climate and environment. It would be true that the extinction of some species or the decrease of numbers of large mammals as a direct result of reservoir construction does not affect the energy and water circulation of the earth, but such arguments are not tenable with sustainable development and the principle of inter-generational equity.

Climate variability and change Climatic variability and change due to global warming could be a decisive factor in the future needs and potential of water resource development in any region. Although its exact timing, magnitude, seasonal patterns and regional characteristics are as yet unknown, some climatic variation and change seem to have already started with the more frequent occurrence of extreme events all over the world. Climatic variability and change would have a major impact on water resources in the 21st century and later. But the counter measures to minimize the effect of significant climate change would not be just more construction of reservoirs or the extension of existing technologies. It would necessitate an increase in social adaptability to a large variation of climate and eventually the gradual but drastic shift of land-use patterns and population distribution. For the as yet uncertain climatic variation and change, we should not start by constructing additional large reservoirs, large dikes etc. but should prepare for future scenarios by improving our ability to cope with the current problems.

1.4 CRITIQUES OF PRESENT RESERVOIRS

1.4.1 Introduction

Availability, variability, and reliability are key issues to users of water resources. Reservoirs are an obvious means to improve the characteristics of

Introduction 15

water resources in this regard. Consequently, reservoir development has played an important role in water resources management.

The advantages of reservoir development are well known and widely publicized, but increasing concern and resistance to reservoir development indicate that negative impacts have frequently been disregarded or inadequately assessed within current planning practice. As new reservoirs tend to be larger than the older ones, impacts, both negative and positive, become more evident, and conflicts have developed. The positive impacts are fairly obvious and are usually fully appreciated in the planning, but the negative impacts, especially intangibles, play a less prominent role.

In order to make the long-term balance of impacts more realistic, the negative impacts must be given more attention, and more effort has to go into planning remedial measures. As a consequence of neglecting these issues, some projects have developed into conflicts that are very serious, so serious that some water resource managers think that in many countries it might soon become impossible to gain public acceptance for the construction of large dams.

A truly consultative and participatory planning process is paramount to the reduction of adverse impacts. In this regard the policy for localization is crucial. The same human activities will have quite different ecological and social impacts depending on where they are located and how they are coordinated. In general, we need to place more emphasis on the institutionalization of the decision process at the expense of the formal techniques of planning.

Although extra storage facilities may increase confidence among users, they can also create additional problems depending on how the system is managed. Unless the management system is able to ensure that the flow will be controlled in accordance with the needs of the users, and according to agreed criteria, governance problems may emerge and encourage users to apply "first capture strategies", i.e. a "use it or lose it" attitude where environmental and social concerns are ignored.

In parallel to the increasing awareness of the need for including remedial actions in the planning of new reservoirs a growing interest in the rehabilitation of degraded ecosystems due to reservoir construction has become obvious. Thus, the concept of reservoir sustainability must address both construction, management and rehabilitation issues.

Environmental losses and other severe problems related to the construction of reservoirs have recently become an issue of great concern to the World Bank (Goodland, 1990). The new World Bank policy will undoubtedly have a significant impact on future development projects. Williams (1993) identified a number of mistakes which have occurred in the past at different stages of reservoir design, construction or operation. He considered planning, economic, engineering and political mistakes. The typical mistakes were related to the failure to adequately consider options, effects, or elements of the

16 Introduction

problem. Lessons from these mistakes could lead us closer to sustainable reservoirs.

In a recent review of large dam projects the Operations Evaluation Department of the World Bank undertook an internal desk study of 50 large dams assisted by the Bank. Using available data for projects completed between 1960 and 1995, the dams were classified according to their economic justification and whether they satisfied the impact mitigation and management policies existing at the time of their approval, or could have been planned so as to satisfy policies that the World Bank had introduced over the intervening years. It was concluded that while 90% of the dams reviewed met the standards applicable at the time of approval, only 26% were implemented so as to comply with the World Bank's current, more demanding policies. The review also concluded that the mitigation of adverse social and environmental consequences of large dams would have been both feasible and economically justifiable in 74% of the cases. The analysis, conclusions and recommen­dations are summarized in World Bank (1996).

1.4.2 Environmental effects of reservoir construction

Inundation effects Reservoir schemes may require inundation of large areas of fertile agricultural land, forest, meadow and pasture, though careful siting can minimize such losses. The natural landscape is damaged, and wildlife can suffer substantially, including endemic species and rare species endangered by extinction. Fish migration is impaired due to reduced connectivity of ecosystems. Also, proliferation of weeds can adversely affect water quality, increase disease vectors and water loss through transpiration, and create clogging problems. Serious difficulties have been encountered in the Aswan, Kariba, Pa Mong and Brokopondo reservoirs (Biswas, 1984). Decomposition of inundated vegetation at the bottom of a reservoir consumes large volumes of oxygen. Anaerobic decomposition of organic material in the bottom water may become toxic to aquatic life. Logging of forests before impoundment helps avoid some of these problems. A large body of water may induce some water-related diseases, though implementation of precautionary and mitigatory measures can help.

In the past, several adverse effects (e.g. those related to water quality aspects and ecosystems) have not been given due attention. The aquatic environment shifts from being riverine to lacustrine and the species composition changes accordingly. Fish populations and species composition are sensitive to changes of water regime, water level fluctuations, depth, velocity, temperature, sediment load, and bank characteristics. Due to the amount of biomass normally left in a reservoir, after construction, fish production tends to increase in the short run but may decrease in the long run,

Introduction 17

unless adequate fish management practices are implemented. The fragmentation of biotopes by reservoirs, canals, and associated roads has been shown to have negative impacts on the ecosystem. In particular, barriers to the movement of animals and plants are created. Conversely, if interbasin transfer of water is introduced, natural species may be threatened by new predators or competitors.

Steep shorelines of reservoirs have frequently been eroded followed by a reduction of species diversity due to an associated decline of the vegetation. If shores have gentler slopes, large tracts of unpleasant, messy land may be exposed at drawdown. However, there are examples from Europe and other parts of the world that, if well planned and managed, reservoirs can become highly appreciated recreational assets (Schultz, 1991).

Groundwater levels tend to rise upstream of a dam. Frequently, this also has an impact on the groundwater quality. Evidently, these changes will put some stress on the existing vegetation, which eventually will adapt by changing the composition of species represented. A frequent measure to increase the inflow of water to the reservoir is to improve the drainage of the upstream reaches. This measure is one of the mechanisms which cause a reduction of upstream wetland acreage.

Another type of undesired effect associated with reservoirs in some parts of the world is the creation of favourable habitats for parasitic and waterborne diseases, such as schistosomiasis, malaria, filariasis and river fluke infection. An example is the infiltration of blood flukes into Sudan as a result of the development of the Gezira irrigation scheme. Another case is the construction of the Volta dam, where the increase of the shoreline created a serious potential mosquito breeding area (Lambrecht, 1981).

Atmospheric effects Existence of reservoirs causes increased evaporation, and there is also a problem of seepage where reservoirs are underlain by permeable strata. Davis (1985) estimated that even if additional evaporation from the surface of reservoirs amounts to only 500 mm on average per year, the global total will make up 200 km 3 of additional evaporation from the 400 000 km 2 area of existing reservoir surfaces worldwide. This is quite a conservative estimate, probably the actual figures are higher as annual evaporation from some manmade reservoirs may exceed 2500 mm. However, to some extent evaporation and seepage are water management losses rather than hydrological losses: evaporated water may enhance the rainfall and seepage may contribute to aquifer recharge.

Reservoirs can have a perceptible influence on the local climate of adjacent territories. The main causes of this influence are a decrease in the albedo and an increase in the heat capacity of water in comparison with that of the land. The differences between the albedo of water and the land surface amount to 3-7% in the boreal zone, 8-10% in the steppe zone and 20-30% in the arid zone. The increase of evaporation from the water surface leads to increased

18 Introduction

atmospheric humidity. The difference between mean monthly air humidity over the reservoir surface and the adjacent land surface can reach 15-20%. The temperature and humidity differences between water and the land surface create atmospheric pressure gradients which lead to the formation of a breeze circulation.

The wind speed over a reservoir surface is usually 20-60% larger than the wind speed over the adjacent land. The maximum wind speed over a reservoir surface is observed during the night, and the minimum during the day. The number of days with strong winds increases significantly after the construction of a reservoir. The cloudiness conditions also change (decrease in the number of days with low cloudiness). During autumn and winter in regions with moderate climates the frequency of fogs over reservoirs may increase considerably, and fogs and humid air may cover large areas. Such fogs or advections of humid air can essentially worsen the climate, especially if a big city with industrial pollution is situated not far from a large reservoir, e.g. Krasnojarsk in Siberia.

Downstream effects Downstream of a dam water levels will usually be lowered, and traditional flooding of the river banks will be very different. The associated impacts tend to be rather on the negative side: loss of wetlands and their associated environmental services and loss of productivity of the river banks associated with a reduction of vegetation diversity. This in turn may reduce the ability of the river valley to serve as a corridor for wildlife, resulting in a reduction of diversity among animal species. A frequent result is a drop both in species diversity and in fish productivity. The river estuary, which is the most ecologically productive part of the water course has often been harmed by reservoir development.

Horowitz & Salem-Murdoch (1993) consider the problems related to reduced natural flooding. The adverse impacts include: drastic reduction of recession cultivation implying a heavy reliance on costly irrigated production, a decline in the quality of riverine pasture and in the number of stock that can graze during the dry season, a reduction in fish harvesting, and a trans­formation of the natural flood plain reducing its capacity to support migratory birds and other wildlife. In addition a decline in flooding, caused by the dam, may have devastating effects on the aquifer from which villagers downstream obtain most of their domestic water supplies. There are, however, also reports on improved downstream conditions, e.g. Chang & Crowley (1997).

Morphological effects The soil erosion, which is a major contributor to reservoir sedimentation, leads to loss of valuable topsoil and causes silting, sedimentation, and turbidity problems in downstream areas. Although it is unavoidable that storage reservoirs will receive some sediment, the upper watersheds should be managed in such a way as to maximize the design life of reservoirs (UNEP, 1989). Due to various short-term, income-generating

Introduction 19

activities (e.g. deforestation, intensive agriculture and irrigation) 25 billion tonnes of soil are lost worldwide each year. It is calculated that over a 20-year period, a $4.5 billion year"1 investment in soil protection would reduce the annual cost of lost agricultural production by $26 billion year"1 (Lazarus, 1990). In addition, increased soil investments would also produce benefits outside the agricultural sector, for example reducing sedimentation in hydroelectric dams, improving water quality, and increasing fish catches.

Siltation frequently turns out to be a major problem. There are examples of reservoirs where the rate of accumulation of sediments was so rapid that a major reduction of reservoir capacity occurred within a few years of the construction of the associated dam (UNEP, 1989). The Sanmenxia Dam on the Huanghe River in China provides an especially dramatic illustration. The river impoundment began in 1960. But within the very short period of 7.5 years of operation, the reservoir lost 35% of its total storage capacity due to sedimentation.

Erosion upstream in the catchment area leads to sedimentation or landslips which can impair storage. Downstream of the dam there is a scouring tendency, as sediment-free water is discharged. Attention should also be paid to deltaic and coastal changes. Sediment control, first of all via source control (i.e. watershed management), is extremely important. Structural measures embrace channel improvement and stabilization, debris and sediment basins and diversions. Sediment transport monitoring is essential to recognize the process and its dynamics, and the efficiency of sediment-reduction measures which have been introduced. Sediment transport also has a quality dimension through contaminants associated with sediments.

Dam safety Concerns may be raised about the reliability of particular dam and reservoir projects. Reservoirs serve to improve the reliability of water supply, flood protection, power generation, low flow abatement, navigation, etc. However, no system can guarantee complete reliability, and the issue is therefore how to render the probability of failure sufficiently small. Dams have been designed to cope with extreme events for some recurrence intervals (modified by a safety factor). Design floods used in the design of spillways of large dams should ensure that the dam crest is not overtopped. However, in an extremely rare event, every dam can fail. Risk perception of society changes with time. Having achieved a higher level of wealth, societies are more and more risk averse and want to avoid disasters leading to high losses, even if a possibility of such a disaster is extremely low. Therefore, the risk of failure for many existing large dams has been reassessed and, in consequence, some upgrading measures are taken. Moreover, reservoirs do not serve humanity indefinitely: there is always a design life, after which a reservoir cannot fulfil its original functions without a redevelopment process. If we admit the very possibility of dam failures, the issue is how to make these failures acceptable, and how to reduce their severity. Recent discussion of risks of Himalayan

20 Introduction

reservoirs caused by slope instability, high sediment discharge, extremes in flow, and seismic activity can be found in Bandyopadhyay & Gyawali (1994).

There have been a disquieting number of safety incidents that have threatened the integrity of large dams during the last two decades. Williams (1993) points out that, for example, the Koyna Dam in India initiated an earthquake (approximately 6.0 in magnitude) that seriously damaged the dam and killed 200 people in the area, which had not previously been seismically active. Goldsmith & Hildyard (1984) report that the first seismic activity to affect a reservoir was in California in the late 1930s (Boulder Dam and Lake Mead). Later, major earthquakes occurred at four major reservoirs; at Hsinfengkian in China in 1962 (magnitude 6.1); at Kariba in Zimbabwe in 1963 (magnitude 5.8); at Kremasta in Greece in 1966 (magnitude 6.3); and at Koyna in India in 1967 (magnitude 6.5). In addition, a flood caused by a landslide, probably triggered by seismic activity, at the Vaiont Dam in Italy in 1963 killed 2000 people. These problems have been considered by a UNESCO working group (UNESCO, 1971).

1.4.3 Socio-cultural and institutional problems

Socio-cultural effects In some reservoir projects, the number of displaced people is very large, and this causes not only immense economic, but also psychological, social and cultural problems. People are often resettled in less fertile marginal land where they overuse resources and jeopardize sustainability. The construction of the Kariba Dam on the Zambezi required the resettlement of some 75 000 Shona tribes people, and the creation of the Aswan High Dam necessitated the displacement of more than 100 000 people in Egypt and Sudan, together with the loss of property important to our cultural and historical heritage. In the case of Ban Chao Nen in Thailand 8000 inhabitants had to be removed from the reservoir area, presenting several psychological, social, and cultural problems in addition to the economic problems (UNEP, 1989). The Three Gorges project in China will result in the resettlement of a million people. In other cases the number of resettled people can be relatively small, but serious problems may occur in cases where the rights of indigenous people are affected.

Institutional issues Howe & Dixon (1993) argue that water projects are frequently poorly conceived. They suggest that the problem is endemic to the whole process, from project identification to reservoir design and construction, and that the incentives in many cases appear to be to build a "monument". Thus, they infer that given the nature of the existing institutions, the net effect of the process tends to be a selection of capital-intensive projects and the neglect of good maintenance programmes.

Introduction 21

The need to go from "monument-building" to an integrated regional approach is underlined in the World Bank policy on dams and reservoirs (Goodland, 1990), where it is stated that benefits from dam and reservoir projects increase when they become regional development projects which integrate, for instance, power generation, irrigation, and municipal water supply with catchment area management and rural development. Designing water projects in the context of overall river basin and regional development plans normally reduces the potential for adverse environmental effects and inter-sectoral problems. The need for catchment area management and improved land use (e.g. discouraging settlement in flood-prone areas) should also be systematically considered.

Brookshire & Whittington (1993) show concern about the role of institutions. They argue that one of the most vexed problems has been the inability of water managers to think creatively about the nature of institutions. Dixon et al. (1992) note that the most severe critics see collusion between the vested interests of construction companies, senior officials and politicians as the driving force behind large dam projects. Discussing dam projects in India, they quote one observer who states that "between the claims and the performance, there has been a wide gulf. Indeed, in some cases, had the claims not been so exaggerated and the social, ecological, and economic costs so understated, it is doubtful if the dams would have ever been built".

Wynne (1992) observed that the institutionalized exaggeration of the scope and power of scientific knowledge creates a vacuum in which there should exist a vital social discourse about the conditions of boundaries of scientific knowledge in relation to moral and social knowledge. Social commitments are necessary to define the boundaries of, and to give coherence to, scientific knowledge. Whenever events expose the ignorance that always underlies scientific models used in public policy, the dominant response is invariably to focus on improving the scientific model. However, although this is important, it is not enough. A response of at least equal importance ought to examine critically the social commitments built on existing knowledge.

Evidently, sustainable development is not an easy task and there are scores of difficulties to be overcome. To summarize major obstacles, as identified by governments and the international community (World Bank, 1988), we may list: fragmented sector policies; weak or non-existent institutions and inadequate coordination among sector agencies; lack of adequately trained or motivated manpower; use of technologies inappropriate for the conditions in developing countries, and lack of knowledge of lower-cost technologies; lack of community involvement; inadequate operation and maintenance; problems with resource mobilization and utilization, including cost recovery.

These difficulties contribute to non-sustainable development and use of water resources and foster a short-sighted policy which may result in an even worse situation. This is especially true if the adverse environmental impacts are synergistic or irreversible, or both.

22 Introduction

1.4.4 Problems related to the use of reservoirs

Hydropower Hydroelectric energy offers numerous advantages (WMO, 1994). It is derived from a continuously renewable resource powered by the energy of the sun. It is an ideally renewable source based on runoff, which is a natural element in the indefinitely repeated hydrological cycle. Hydropower is non-polluting; neither heat nor gaseous pollutants (noxious or greenhouse gasses) are released. It allows high efficiency, of the order of 90%, to be achieved, which is far higher than that of fossil-fuelled thermal plants which have an efficiency of between 30% and 40%. Hydropower is based on a mature technology comprising a reliable and flexible operation, and its equipment is easily adaptable to local conditions. Hydropower is highly responsive to changing demands for electric power, virtually in seconds. Storage allows the response to fluctuations, such as within-day, peak hour, within-week and seasonal, to be achieved despite vagaries of natural flows. No fuel costs are involved. Pumping to an upstream reservoir is a hydropower variant, which has a potential for energy storage that can be used during peak hours. It is noticeable that water use in hydropower is counted at high, very impressive figures. This can, however, be misleading as through multiple reuse water may pass through turbines many times on its way to the sea. Use of water for hydropower production can often be combined with other uses, e.g. for irrigation purposes.

Despite the above positive merits there are, nevertheless, many examples of projects where adverse effects have become significant. In the case of the multipurpose Aswan High Dam Reservoir, Dixon et al. (1992) found a profound impact on the Nile. Most environmental impacts were anticipated in advance and in some cases corrective measures have been taken. The authors conclude that with any major dam project such as the Aswan High Dam, there are bound to be surprises; some environmental effects will be greater and others less than anticipated. According to Horowitz & Salem-Murdoch (1993), the construction of the Manantali Dam on the Senegal River has turned out to be a subject of great concern. They report that if the dam is managed as planned, it will severely affect downstream production, income, and employment due to adverse environmental impacts.

Irrigation Irrigation is an important means of improving the productivity of land, and thus of solving world food problems for the increasing population. In the world, irrigated agriculture occupies some 15% of total arable land but produces some 33% of global agricultural yields. Irrigation accounts for between 70% and 80% of world water use, in some countries this share is even higher, e.g. in Pakistan it is around 90%. However, experience shows that irrigation projects have to a large extent failed to produce what was expected. Rehabilitation projects and improved operation are necessary in many existing areas under irrigation.

Introduction 23

Good design, including adequate drainage is an absolute prerequisite for sustainable irrigation projects. Most likely, sustainability would be much better served if money and resources went into rehabilitation and improved operation in existing areas under irrigation rather than into the construction of new schemes. The above points are underlined by the Indian experience of large dams (Dogra, 1986). In the Sixth Plan Document, it is admitted that the returns from the massive investment made in irrigation have so far been very disappointing, both in terms of the expected increase in yields and in terms of the financial returns. Irrigated land in India was claimed to yield at least 4-5 tonnes of grain per ha per year, but such yields have only been achieved in experimental and demonstration plots. On a commercial scale, the average yield on irrigated land is barely 1.7 tonnes per ha per year. Under such conditions, most irrigation projects are unable to recover even their working expenses.

The environmental risks associated with waterlogging and salinization brought about by irrigation projects are also evident from the Indian experience. Dogra (1986) observes that in the middle of the last century, when the construction of the massive canal network in the then undivided Punjab was first undertaken, it was realized that the new water courses were causing waterlogging and salinization. As much as one third of the water transported in the canals was seeping out into the subsoil, thus raising the water table over vast areas at the rate of one or two feet per year. Eventually, the water reached the surface, giving rise to stagnant puddles. Evaporation tended to raise the salt content in the top three feet of the soil, in some cases to the intolerable level of one part per hundred. After just 20 years, as the saline groundwater invaded the root zone, crop yields started to fall. Tragically, Dogra (1986) concludes that India's planners do not seem to have learned anything from the 19th century experience. Vast amounts of fertile lands are still being lost to waterlogging and salinity.

Actually, soil salinization and water logging due to inadequate drainage facilities are a classic environmental problem. In the Euphrates valley in Syria and the lower Rafadam plain of Iraq, more than 50% of irrigated land suffers from these effects (El Gabaly, 1977). Of the total irrigated area in the world, about four fifths lie in the arid and semiarid regions of Asia. In such regions waterlogging and salinization must always be feared if irrigation is not accompanied by adequate drainage. On a global scale, more than 200 000 ha are lost every year to waterlogging and salinity (UNEP, 1989).

Water supply and sanitation Clearly, the problems related to lack of clean water and sanitation have reached a critical stage in many parts of the world, not only from a social point of view, but also from an economic and development perspective. There is, however, much room for water conservation in the industrial and domestic sectors. There is a vast under­utilized potential for stepping up agricultural and industrial production,

24 Introduction

primarily through more efficient utilization of water management structures already created (UNEP, 1989).

Frequently, an increase in water supply is developed mainly to improve the services to those who are already adequately served (de Rooy & Doyle, 1992). According to UNCED (1992), we need a reallocation of water supply: away from industry and the wasteful urban rich segments and towards those, who do not yet have healthy and safe water. It may be interesting to note that in some areas in the USA and Mexico, urban water supply is forcing irrigated agriculture out of production (Hirschleifer & Milliman, 1967; Lees, 1974; Arizona Water Commission, 1977; Hjorth, 1985; Grigg, 1994).

Rapid industrialization and population growth on ecologically fragile land creates serious physical problems. The Valley of Mexico, e.g. in which Mexico City is located, has undergone serious adverse transformations as a result of industrialization and overdevelopment. The valley has lost almost all of its lakes, which have turned into large salt basins (Schteingart, 1989). In the state of California, a culture based on, and absolutely dependent on a sharply alienating, intensely managerial relationship with nature, has led to a situation where many of the northern streams have been dammed. California's wetland acreage has been reduced to 5 % of its historical extent, habitats for millions of birds and other animals have been lost, and numerous opportunities that are dependent on water staying in streams have been eliminated (Reisner, 1989).

In Arizona, the Arizona Water Commission (1977), in its Sixth Annual Report for 1975-1976, described some of the assumptions underlying its planning. The Commission indicated that "non-agricultural uses of water would generally grow without being constrained by lack of water supply. This assumption recognized that these uses have an economic advantage over agriculture, when it comes to the amount that can be paid for an acre-foot of water, and if necessary, can buy out farmland solely for the purpose of obtain­ing a water supply". Welch (1985) cites a 1971 study by the Arizona Depart­ment of Planning and Economic Development which suggests that, if all water use in the state was diverted to municipal and industrial purposes, the state could support a population of 20 million. Of course, such a diversion would bring the present rural lifestyle and the communities that depend on agriculture to an end. It would also totally neglect values of water that are non-economic.

Low flow augmentation Low flow augmentation is frequently used to remedy negative environmental impacts on river stretches which have had their low flows diminished by various development projects. The problems addressed may include fish spawning, salt-water intrusion, navigation etc. In other cases, reservoirs are used to maintain a minimum flow required to keep the concentration of discharged effluents below some predetermined standards.

Low flow augmentation has the seemingly negative economic impact that the water used to augment flows is lost for "economic productive" activities. It should, however, be taken into account that few, if any, industrial projects

Introduction 25

are free from water pollution problems. Although conditions in many developing countries may permit relatively liberal effluent and emission standards, it should be remembered that the cost of cleaning up water, once it has been polluted, is high; hence the urgent need to introduce water quality standards capable of taking into account the cumulative consequences of the industrialization of developing countries (UNEP, 1989).

Wildlife preservation also falls into the category of ecological effects that could be mitigated by means of low flow augmentation. For instance, the Kafue hydroelectric project in Zambia, in its original design, would have had a serious impact on 90 000 lechwes, a species of unique small antelope. Their movement patterns have been dictated mainly by the grazing conditions provided by the flood cycles. By interfering with the flood regime the reservoir would have posed a grave threat to this species. However, the dam was redesigned to allow additional storage to permit the discharge of water needed by the lechwe antelope during the critical months of March and April in dry years (World Bank, 1975).

Flood control Obviously flood control would contribute to economic develop­ment if it was able effectively to prevent losses of agricultural and other resources. However, a study of past events reveals that flood control projects have turned out to be remarkably ineffective in damage prevention (Arnold, 1976). Flooding events may have become somewhat less frequent, but on the other hand, the damage from each flooding event has become much more significant.

Despite the billions of dollars devoted to flood control, the annual losses from floods in the United States seem not to show any decline. This is not because the dams are totally ineffective, or because man in trying to control nature has been over-presumptuous. Flood losses are still high because the very modifications we make in natural systems encourage people to use land that was formerly seen as much too hazardous for the location of homes and factories (Abler et ai, 1972).

Flood control reservoirs belong to the category of structural flood mitigation measures. However, no structural protection scheme can guarantee absolute safety. No matter how big a reservoir is, it may happen that a flood wave cannot be contained in the reservoir and that spillway conveyance is not adequate. Thus, the safety against flooding will always be less than 100%, as is also the case for safety against structural damage, as well as the reliability of water releases from a reservoir.

1.4.5 Concluding remarks

Reservoirs are built to serve important uses. However, they are not free from adverse effects which sometimes can be very critical. It is essential to foresee

26 Introduction

and assess such adverse effects in each particular project and to account for them in the decision process. Perhaps the most significant impacts are the direct effects of impoundment: loss of land due to inundation, involuntary relocation of people, loss of cultural property and historical heritage, and disturbance to the ecosystems by storing water and creating barriers.

Although reservoir developments of all types are progressing at a rapid rate and many beneficial effects are being recorded, the environmental impacts of some projects have not been what water resources planners and political leaders had expected. Serious problems emerging from time to time have drawn the attention of all concerned to the fact that water projects may yield a mixture of desirable and undesirable effects. It has become clear that several of these projects have been planned and designed without taking into account the complex relationships between people, water, environment and development (UNEP, 1989).

The World Bank policy (Goodland, 1990) gives some advice on how possible remedial measures could mitigate the impacts. Among other things the policy suggests:

• Careful siting can minimize losses due to inundation. The value of lost timber and other resources and foregone use of inundated land should be estimated in the economic analysis. Biotic surveys normally are essential; plant and animal extinction can be prevented or minimized by careful project siting. Animal rescues, replenishment, and relocation can be useful. Canal and other crossing facilities are often essential.

• To prevent unnecessary losses to fisheries, it may be necessary in the planning process to give special attention to spawning areas, aquaculture and improved fishing methods in the planning process.

<» Fish propagation schemes in reservoirs may mitigate losses and produce more fish protein than before the project.

o There should be a weed management plan. The potential to use weeds for compost, biogas, or fodder should be investigated. Multiple-level outlets in the dam can avoid the discharge of anaerobic water. Utilization of forest for timber before reservoir filling reduces project contributions to greenhouse gases.

• To reduce siltation problems, catchment area management should be encouraged where appropriate.

» Sometimes, management of water releases can minimize damage caused by the changed hydraulic regime downstream, in order to partially replicate natural flooding regimes.

There are still too many examples of water being mismanaged by institutions having dated missions and using inappropriate evaluation models. Consequently, water is overused and misused at the same time as significant

Introduction 27

sections of the population are barred from reasonable access to water to cater for their basic needs. While physical targets may have been achieved, overall environmental, social, and health objectives may not be met because the link between physical targets and, for example, health objectives are seldom explicitly inferred.

Institutional development, community awareness, increased involvement of women, and linkages to other sectors such as health, education, communications, and nutrition are all themes that experience singles out as important. Furthermore, if a meaningful community participation is not at the heart of development plans, then programmes will have little long-term positive impact on the environment and on the lives of women and men.

Although the economic growth generated by reservoir development could produce the means for impact prevention and mitigation and, thus, the possibility to improve welfare and the quality of the environment, this will only happen if an essential prerequisite is met: a sound policy framework drawing on the full knowledge base of society rather than upon the technical knowledge of a limited number of regulators. We need an approach to identify, evaluate and compare a range of possible strategic options for widely varying river catchment conditions which can resolve the complex ecological, economic and social interactions in water resources systems.

While we must acknowledge that treatment of existing problems is urgently needed, there is also a strong demand to focus on prevention at source, enabling the precautionary principle to be applied within the decision­making process, from policy to project level. We must focus on the agents and activities which deplete the water and environmental resources or otherwise damage them, rather than wait for problems to emerge.

Because of the difficulties involved in finding routes to sustainable development, it seems important to start now and pursue policies with commitment and steadfastness, so as to increase our chances of obtaining results in terms of testable cumulative knowledge. In this effort it will be essential to employ a plurality of methods. A model for such planning is offered by the apprenticeship process extended to the drainage basin and its social and environmental system as a whole. Consequently, emphasis should be placed on flexibility, adaptability, provision of options for the future, and on preparedness for surprise, which amounts to the recognition of uncertainty.

In a recent workshop jointly organized by IUCN (the World Conservation Union) and the World Bank it was decided to establish a World Commission of Dams by November 1997 to assess experience with large dams and to propose if and how they can contribute to sustainable development (Dorcey et al, 1997). The terms of reference are:

• to assess the experience with existing, new and proposed large dam projects so as to improve (existing) practices and social and environmental conditions;

28 Introduction

• to develop decision-making criteria and policy and regulatory frameworks for assessing alternatives for energy and water resources development;

• to evaluate the development effectiveness of large dams;

• to develop and promote internationally acceptable standards for the planning, assessment, design, construction, operation and monitoring of large dams and, if the dams are built, ensure that affected people are better off;

• to identify the implications for institutional, policy and financial arrange­ments so that benefits, costs and risks are equitably shared at the global, national and local levels; and

• to recommend interim modifications, where necessary, of existing policies and guidelines, and to promote "best practices".

The agenda of the commission should build on an evolving paradigm for large dam development. The development process should be based on the analysis of multiple criteria, including food, water, energy, foreign currency, health, employment, human rights, equity, sustainable use of natural resources, and conservation of natural ecosystems and their genetic stocks. The analysis should involve consideration of the long-term and of quantitative and qualitative values. Decision-making should be more transparent and accountable and made through consultation with multiple stakeholders, including local communities, numerous authorities and government departments, industry, and non-governmental organizations. Choices should be made by considering multiple and integrated development options, including demand management, a run-of-the-river hydropower scheme, conjunctive use of surface water and groundwater, and the development of traditional local water management and agricultural practices. A report of the World Commission of Dams is expected within two years.

REFERENCES

Abler, R., Adams, J. & Gould, P. (1972) Spatial Organization. Prentice/Hall International, Inc. , London. Arizona Water Commission (1977) Sixth Annual Report, 1975-1976. Phoenix, Arizona. Arnold, M. (1976) Floods as man-made disasters. The Ecologist 6(5), 169-172. Avakian, A. B. (1990) Reservoirs of the world and their environmental impact. In: The Impact of Large

Water Projects on the Environment, Proc. UNESCO/UNEP Int. Symp., Paris, 1987, 29-36 . Bandyopadhyay, J. & Gyawali, D . (1994) Ecological and political aspects of Himalayan water resource

management. Water Nepal 4, 7 -24 . Beard, D . P. (1994) Address at 18th ICOLD Congress. USCOLD Newsletter, November 1994, 12-15. Biswas, A. K. (1970) A History of Hydrology. North Holland Publ. Co. , Amsterdam. Biswas, A. K. (1984) Environmental consequences of water resources development. In: Proc. Fourth

Congress of Asian and Pacific Division oflAHR, Change Mai, Thailand, vol. 3 , 61-82 . Biswas, A. K. (1992) The Aswan High Dam revisited. Ecodecision 3, 67-69 .

Introduction 29

Brookshire, D . S. & Whittington, D . (1993) Water resources issues in developing countries. Water Résout: Res. 29(7), 1883-1888.

Cernea, M. M. & Le Moigne, G. (1989) The World Bank's approach to involuntary resettlement. In: Water Power & Dam Construction Handbook. Reed Business Publishing, Surrey, UK.

Chang, M. & Crowley, C. M. (1997) Downstream effects of a dammed reservoir on streamflow and vegetation in East Texas. In: Sustainability of Water Resources under Increasing Uncertainty (eds D. Rosbjerg, N . -E . Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F. Rasmussen), IAHS Publ. no. 240, 267-275 .

Dai Qing (1996) Yangtze! Yangtze!—Debate over the Three Gorges Project. lapanese translation by K. Sumi & Hu Weiting, Sankyo Damu, Tsukiji-Shoten, Tokyo.

Davis, G. H. (1985) Water and Energy: Demands and Effects. UNESCO Studies and Reports in Hydrology no. 42.

Dixon, J. A., Talbot, L. M. & Le Moigne, G. J . -M. (1992) Dams and the Environment Considerations in World Bank Projects. World Bank Technical Paper no. 110, The World Bank, Washington DC.

Dogra, B. (1986) The Indian experience with large dams. In: The Social and Environmental Effects of Large Dams, vol. 2 : Case Studies (eds E. Goldsmith & N. Hildyard), Wadebridge Ecological Centre, Camelford, UK, 201-208.

Dorcey, T . , Steiner, A. , Acreman, M. & Orlando, B., eds (1997) Large Dams: Learning from the Past, Looking at the Future. Proceedings of a Joint IUCN/World Bank Workshop, Gland, Switzerland, 1 1 -12 April 1997.

El Gabaly, M. M. (1977) Problems and effects of irrigation in the Near East Region. In: Arid Land Irrigation in Developing Countries (ed. E. B. Worthington), Pergamon, Oxford, UK, 239-249.

Goldsmith, E. & Hildyard, N . , eds (1984) The Social and Environmental Effects of Large Dams, vol. 1: Overview. Wadebridge Ecological Centre, Camelford, Cornwall, UK.

Goodland, R. J. A. (1990) The World Bank's new environmental policy for dams and reservoirs. International Journal of Water Resources Development 6(4), 226-239.

Gottschalk, L. C. (1964) Reservoir sedimentation. Chapter 17-1 in: Handbook of Applied Hydrology (ed. V. T. Chow). McGraw-Hill , New York.

Grigg, N. S. (1994) Planning and coordination in water management. In: Proc Int. UNESCO Symp. on Water Resources Planning in a Changing World, Karlsruhe, 28-30 June 1994, section I, 83-90.

Hirschleifer, J. & Milliman, J. W. (1967) Urban water supply: a second look. American Economic Review 57(2), 169-182.

Hjorth, P. (1985) On Water in Mexico (in Swedish). Department of Water Resources Engineering, Lund University, Lund.

Horowitz, M. & Salem-Murdoch, M. (1993) Development-induced food security in the Middle Senegal Valley. Geojournal 30(2), 179-184.

Howe, C. W. & Dixon, J. A. (1993) Inefficiencies in water project design and operation in the third world: an economic perspective. Water Resour. Res. 29(7), 1889-1894.

ICOLD (1988) World Dam Register 1988. International Commission on Large Dams, Paris, France. Investigation Committee on Sediment Release Impact of Dashi-Daira Dam of the Kurobe River (1995)

Analyses of Results and Proposals. Final Report to Toyama Prefecture. Japan Dam Association (1990) Yearbook of Dams 1990. Kajima Press, Tokyo. Japan Dam Association (1996) Yearbook of Dams 1996. Kajima Press, Tokyo. Kobayashi, E. (1996) Current measures against sediments in dam reservoirs. In: Yearbook of Dams 1996,

Japanese Dam Association, 21 -27 . Lambrecht, F. L. (1981) Dangerous Development. The IFDC Reports no. 9, Ottawa. Lazarus, D. S. (1990) Save our soils. Our Planet 2(4), 10-11 . Lees, S. H. (1974) Hydraulic development as a process of response. Human Ecology 2(3), 159-175. Meybeck, M. (1988) Personal communication. Cited in: World Resources Institute (1988) World Resources

1988-1989. Basic Books, Inc., New York. Milliman, J. D. & Meade, R. H. (1983) Worldwide delivery of river sediment to the ocean. Journal of

Geology 91(1), 1-21. Nagase, M. (1998) Background of Elwha Dam removal in the USA (in Japanese). Journal of Japan Society

of Civil Engineers 83(1), 43 -46 . National Land Agency (1995) Water Resources in Japan. Ministry of Finance Press, Tokyo, Japan. Pearce, F. (1992) The Dammed. The Bodly Head, London. Reisner, M. (1989) The Next Water War: Cities Versus Agriculture. Issues in Science and Technology 5(2),

98 -102 .

30 Introduction

de Rooy, C. & Doyle, B. A. (1992) Focus on Africa: water and sanitation in the 1990s. In: Waterfront, issue 2 , a biannual publication, UNICEF, New York, pp. 1, 12-17, 20.

Said, R. (1993) The River Nile: Geology, Hydrology and Utilization. Pergamon Press, Oxford. Schteingart, M. (1989) The environmental problems associated with urban development in Mexico City.

Environment and Urbanization 1(1), 40-50 .

Schultz, G. A. (1991) Hydrology of manmade lakes. In: Hydrology of Natural and Manmade Lakes (eds G. Schiller, R. Lemmela & M. Spreafico), IAHS Publ. no. 206, 139-150.

Shahin, M. (1985) Hydrology of the Nile Basin. Elsevier, Amsterdam. Sumi, K (1996) History of dam construction in USA (in Japanese). In: Wliy did USA Stop Building Dams?

ed. Parliament Members Assoc. for Realization of Check Mechanism of Public Works, Tsukiji-Shokan, Tokyo, 22-54 .

Takeuchi, K. (1997) On the scale diseconomy of large reservoirs in land occupation. In: Sustainability of Water Resources under Increasing Uncertainty (eds D. Rosbjerg, N. -E. Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F. Rasmussen), IAHS Publ. no. 240, 519-527.

Takeuchi, K. (1998) Reply to Discussion by D. Hansen on "Least marginal environmental impact rule for reservoir development". Hydrol. Sci. J. 43(2), 160-162.

U N C E D (1992) Agenda 21: Programme of Action for Sustainable Development. United Nations Conference on Environment and Development, Rio de Janeiro, Brazil, 3-14 June 1992.

U N E P (1989) Sustainable water development and management. A synthesis. International Journal of Water Resources Development 5(4), 225-251 .

UNESCO (1971) Working Group on Seismic Phenomena Associated with Large Reservoirs. Report of First Meeting, UNESCO, 14-16 December 1970, SC/CONF 200.4, Paris.

UNESCO (1978) World Water Balance and Water Resources of the Earth. Studies and Reports in Hydrology no. 25 , USSR Nat. Comm. for IHD.

Veltrop, J. (1995) Benefits of dams to society. USCOLD Newsletter, March 1995, p. 3. W C E D (1987) Our Common Future. World Commission on Environment and Development, Oxford

University Press, UK. Welch, F. (1985) How to Create a Water Crisis. Johnson Books, Boulder, Colorado. Williams, P. B. (1993) Dam safety analysis. In: Damming the Three Gorges. What Dam-Builders Don't

Want to Know (ed. G. Ryder), Probe International, Toronto, 107-111. W M O (1994) Guide to Hydrological Practice, 5th edition. W M O Publ. no. 168. World Atlas and Industry Guide (1997) The International Journal on Hydropower & Dams 4, Supplement. World Bank (1975) Environment and Development. The World Bank, Washington DC. World Bank (1988) Water and Sanitation: Toward Equitable and Sustainable Development. The World

Bank, Washington DC.

World Bank (1996) World Bank Lending for Large Dams: A Preliminary Review of Impacts. Operations Evaluation Department, Précis no. 125.

World Resources Institute (1994) World Resources 1994-1995. Oxford Univ. Press, New York. Wynne, B. (1992) Uncertainty and environmental learning—reconceiving science and policy in the

preventive paradigm. Global Environmental Change/Human and Policy Dimensions 2, 111-127.

Sustainable Reservoir Development and Management. Report by the IAHS/ICWRS Project Team. IAHS Publ. no. 251, 1998. ^ \

CHAPTER 2

SUSTAINABILITY AND RESERVOIRS

2.1 NOTIONS OF SUSTAINABLE DEVELOPMENT

2.1.1 Introduction

Sustainable reservoir development and management belong to a broader area of sustainable development. This chapter reviews the general principle of sustainable development as given by the World Commission of Environment and Development (WCED, 1987) and discusses alternative notions and interpretations of the principle. There have been numerous reports, articles and documents regarding the concept of sustainable development since it was accepted as the basic principle for harmonizing development and the environment by the international community. Section 2.1 forms the basis for translation of the principle to the water resources context, and the subsequent Section 2.2 refers to the reservoir context.

While the focus of this chapter is on quantifying sustainability, it is realized that mathematical or technical descriptions alone are not sufficient to fully measure the notion. Sustainability involves other aspects that deserve intensive discussion and require reaching beyond the scope of what may be quantifiable or measurable. But unless one can measure or describe in precise terms what one is trying to achieve, it becomes rather difficult (if not impossible) to determine how effective one is at achieving sustainable development, or even at comparing alternative plans and policies with respect to their sustainability.

2.1.2 Sustainable development: perspective of WCED and UNCED

Sustainable development The term "sustainable" has been used in scientific literature for many centuries. It is an old concept in fishery, forestry, groundwater and other areas indicating the rate of use of renewable natural resources to ensure the continuous supply of resources and their maximum use, where the terminology was such as maximum sustainable yield (MSY) in fishery, maximum allowable cutting (MAC) in forestry and the like. Even in "National Wealth", Adam Smith mentions the capital cost for sustainable agricultural land use (Smith, 1776).

32 Sustainability and reservoirs

Since the 1970s, the concept of sustainability has been reintroduced in societal policy discussions as the basic principle of harmonizing environment and development. It was used by Coomer (1979) as sustainable society and by Allen (1980) and the International Union for the Conservation of Nature (IUCN), the World Wildlife Fund (WWF) and United Nations Environment Programme (UNEP) in the early 1980s (cf. IUCN, 1991) as sustainable development. Reflecting such preceding uses, "sustainable development" was introduced to the UN community by WCED in 1987. The definition of sustainable development given by WCED (1987) is as follows:

"Humanity has the ability to make development sustainable—to ensure that it meets the needs of the present without compromising the ability of future generations to meet their own needs. "

Although the issue of harmonizing development and the environment is so difficult and controversial, this definition is broad enough to be accepted by everybody. Actions implied are not clear but its conceptual framework is clear, specifying the very basic criteria that development ought to satisfy. The definition consists of three conceptual components; needs, generations and equity, implying that the development is necessary because of human needs but that intra-generational and inter-generational equity should hold.

Concerning the limit of development, the WCED (1987) definition continues as:

"The concept of sustainability implies limits: not absolute limits, but limitations imposed by the present state of technology and social organization on environmental resources and by the ability of the biosphere to absorb the effect of human activities. "

It says that inter-generational equity should be considered since the environment has an absorption limit and over a considerable time span its overuse can bring about irreversible changes, which can make the Earth environmentally unsuitable for human habitation. But it also says that the limitation is not fixed but depends on the state of technology and social organization which manage environmental resources. How far the limit can be expanded and human activities can be designed and managed within that limit is the practical question posed to science and society.

One of the criticisms of the term "sustainable development" is that it is an oxymoron (Greek "oxys" meaning keen and "moron" meaning foolish). It consists, by construction, of a form of antithesis in which two contradictory terms are brought sharply together: development, that is growth, cannot be indefinitely sustainable in a finite world with limited resources. However, in the use of this term, development does not necessarily mean growth in the quantitative sense. It is rather a qualitative process which does not have any limit. Development is a continuous process of changing state from one form to another seeking to meet ever changing objectives. Sustainable human existence, if ever possible, must be an endless process of changes in social,

Sustainability and reservoirs 33

cultural and industrial states within a certain limit of the sustainable use of energy and resources. With this definition, sustainable development is not an oxymoron. However, sustainable development is inherently an imprecise concept which may mean different things to different people. It is difficult, if indeed possible, to delineate the borders between "sustainable" and "non-sustainable" in a unique way.

Needs and equity The specification of needs and equity is not straightforward since understanding of these notions may differ at different times and under different socio-cultural conditions. The needs of the present generation are, however, the most obvious ones. The elimination of poverty, uneven distribution of wealth, and environmental destruction and contamination are the most serious and urgent needs everywhere, especially in developing countries. In order to achieve such an objective, every aspect of the economy, technology, education, public investment, institutional and organizational system and the like has to function in the most efficient way. Equity among the current generation is also important because a large portion of developmental needs is, in fact, the need for equity between the rich and the poor, the majority and the minority, and the strong and the weak. It extends to the demand of societal fairness between the receivers and those affected, authorities and the public, professionals and laymen, informed and uninformed, loud and silent. It calls for social justice in development.

On the other hand, the needs of future generations are uncertain. Without knowing the future conditions of nature and society, it is impossible to speculate on the needs and desire of future generations. Population, technology, social preferences, value systems as well as natural and environmental conditions all change and future needs depend on them. Sustainable development, however, calls for attention to the needs of the environment. Regardless of the social conditions and preferences in the future, the very basic stage of human survival, i.e. the Earth environment, should be suited to human beings. The identification of suitability may be controversial but environmental quality that supports an ecological system with bio­diversity, i.e. sufficient individual populations to ensure the natural evolution of each species, is essential. The inter-generational equity therefore claims that development for the needs of the present generation should not narrow the possibility of future generations satisfying their needs. The present generation is obliged to pass on the Earth environment in good condition for future generations to develop their life to their satisfaction. We have no right to use up all the assets of the Earth environment for ourselves. Inter-generational equity is a notion related to the conservation of nature but it specifies who does what for whom.

Rio Declaration and Agenda 21 The idea of sustainability was furthered in the Rio Declaration on Environment and Development issued at UNCED

34 Sustainability and reservoirs

(1992). Among the 27 principles declared, there is the elaboration of the definition of sustainable development. The goal of the concept seems clear in those principles. They include:

Principle 1 : Human beings are at the centre of concerns for sustainable development. They are entitled to a healthy and productive life in harmony with nature. ...

Principle 3 : The right to development must be fulfilled so as to equitably meet developmental and environmental needs of present and future generations.

Principle 4: In order to achieve sustainable development, environmental protection shall constitute an integral part of the development process and cannot be considered in isolation from it.

Principle 5: All states and all people shall cooperate in the essential task of eradicating poverty as an indispensable requirement for sustainable development, in order to decrease the disparities in standards of living and better meet the needs of the majority of the people of the world. ...

Principle 8: To achieve sustainable development and a higher quality of life for all people, states should reduce and eliminate unsustainable patterns of production and consumption and promote appropriate demographic policies. ...

The principles repeat that the human being is the centre of concern, and has the right to develop a higher quality of life but that it should be done and is only possible in the way where environmental protection is treated as an equally important and integrated part of development. The question is how?

As concrete actions, Chapter 18 of Agenda 21 (UNCED, 1992) entitled: "Protection of the quality and supply of freshwater resources: Application of integrated approaches to development, management and use of water resources" provides the methods of management of freshwater resources. There is strong emphasis on the need for integrated management. It says that in order to accomplish integrated management "four principal objectives should be pursued, as follows:

(a) to promote a dynamic, interactive, iterative and multi-sectoral approach to water resources management, including the identification and protection of potential sources of freshwater supply, that integrates technological, socio­economic, environmental and human health considerations;

(b) to plan for the sustainable and rational water utilization, protection, conservation and management of water resources based on community needs and priorities within the framework of national economic development policy;

Sustainability and reservoirs 35

(c) to design, implement and evaluate projects and programmes that are both economically efficient and socially appropriate within clearly defined strategies, based on an approach of full public participation, including that of women, young people, indigenous people and local communities in water management policy-making and decision-making;

(d) to identify and strengthen or develop, as required, in particular in developing countries, the appropriate institutional, legal and financial mechanisms to ensure that water policy and its implementation are a catalyst for sustainable social progress and economic growth."

As concrete actions for integrated water management, it has a clear emphasis on demand management. Among 16 activity areas suggested such as national action plans, integrated water conservation, interactive database, integrated quality and quantity management, public participation, cooperation at all levels, it includes the following, as formulated in UNCED (1992):

• optimization of water resources allocation under physical and socio­economic constraints;

• implementation of allocation decisions through demand management, pricing mechanism and regulatory measures;

• promotion of schemes for rational water use through raising public awareness, educational programmes and levying of water tariffs and other economic instruments;

• mobilization of water resources, particularly in arid and semiarid areas;

• development of new and alternative sources of water supply such as seawater desalination, artificial groundwater recharge, use of marginal-quality water, waste-water reuse and water recycling;

• promotion of water conservation through improved water-use efficiency and wastage minimization schemes for all users, including the development of water-saving devices.

In short, the diagnosis of Agenda 21 for sustainable water resource management is the holistic approach based on:

(a) consideration of various alternative means and components of water resource management in every stage of planning, design, construction and operation,

(b) multidisciplinary approach to include engineering, biology, economics, sociology, health, laws, public administration etc. and

(c) multi-sectoral exercise to include all levels and sectors of legislative and governmental units, cultural and interested groups, indigenous people, women and young people, non-governmental organizations and similar groups.

Agenda 21 (UNEP, 1992) expressed a concern that commonly used indicators do not provide adequate measures of sustainability. Chapter 40,

36 Sustainability and reservoirs

devoted to information for decision making, contains the recommendation (paragraph 40.4) that indicators of sustainable development "need to be developed to provide solid bases for decision-making at all levels and to contribute to a self-regulating sustainability of integrated environment and development systems." The issue of development of indicators for sustainable development has been addressed at both the national and international levels; in the latter case by governmental as well as non-governmental organizations. The United Nations Commission on Sustainable Development (UNCSD) established a multi-year thematic programme on such indicators. Some outcomes reported so far are presented in Section 2.1.3 .

2.1.3 Indicators of sustainability

Work on measuring sustainability in water resources has been conducted with two main emphases. The first focuses on sustainability indicators (a good source of information is the Compendium of Sustainable Development Indicator Initiatives and Publications available on the internet at http://iisdl.iisd.ca). Sustainability indicators can be defined as conditions strictly associated with a sustainable development so that their presence is indicative of its existence. The second emphasis is on the development of criteria for measuring sustainability. In this context a criterion is defined as a standard on which a decision may be based.

Indicators monitor progress towards sustainable development. Their aim is to steer action, to assist the decision makers at all levels and to increase their focus on sustainable development. They can also be useful as public relations instruments.

The goal of a single aggregated universal indicator, or a small number of universal indicators is futile, as the choice of indicator may depend on the purpose of study. Consequently there should be a whole menu of indicators, recognizing the need for flexibility, as conditions and priorities may differ from country to country and from application to application.

In its work on indicators of sustainable development, the UNCSD followed the systemic framework consisting of driving force, state and response. The term "driving force" is being used more or less interchangeably with such terms as input, stimulus, excitation, stress, pressure, load or endogenous variable. It indicates human activities, processes and patterns that influence sustainable development. The driving forces can be directly related to water, such as the emission of pollutants, waste discharges or consumptive water use, or indirectly related as demographic growth. The term "state" is as intuitively expected; it refers to the indicator of the state of sustainable development. The response indicators include effects of policy and management options, regulations, controls, adaptations and other responses.

Sustainability and reservoirs 37

The UNCSD has compiled a working list of indicators of sustainable development. There are a number of indicators in the list which lend themselves well to measuring the sustainability of water resources. Table 2.1 gives a roster of indicators present in the working list of the UNCSD Department for Policy Coordination and Sustainable Development (UNCSD, 1995) which can be used for evaluation of the sustainability of water projects. However, in the planning process, unlike in the post evaluation, one can only speculate about future developments of the situation.

The UNCSD indicators are primarily national (although countries may also wish to use indicators at state and provincial levels). In the case of appraising reservoir sustainability, the catchment scale could be more appropriate, but often it does not coincide with administrative boundaries. A large reservoir may influence areas greater than the catchment (e.g. the relocation of people) and this should be reflected in the spatial scale of considerations. In order to use it for, say, comparison of the state with and without reservoirs, many other indices should be added.

Indicators are quantitative or qualitative variables that can be measured or described and demonstrate trends. They cannot, however, be directly used in decision-making. Their major role is that they provide analytical, communication, warning and mobilization and coordination functions. They can be used as a measure of the potential for management considering sustainability. To be useful, the condition shown by various indicators must be compared with a past or a desired future in order to quantify the extent to which projects may contribute to sustainable development.

2.1.4 Follow-up discussions of sustainable development

The UN leadership for sustainable development has been widely accepted and followed by various scientific communities, governmental bodies and many individuals. There have been numerous proposals for translating the basic principle and measuring sustainability to establish concrete action guidelines. Some try to provide alternative definitions or auxiliary definitions and some to clarify the principle. The Committee on Water Research (COWAR), a now disbanded body of the International Council of Scientific Unions* (ICSU) and the International Union of Technical Associations (UATI), published a report "Water in our common future—A research agenda for sustainable development of water resources" (Jordaan et al., 1993) which "summarized the research needs arising from the concept of sustainability." IAHS is one of the members of ICSU and has been working on sustainable water management from the beginning. What follows

* N o w called the International Council for Science.

38 Sustainability and reservoirs

Table 2.1 Indicators of sustainable water resources development and management.

Driving force indicators: Social:

- employment rate (%) - population growth rate (%) - net migration rate - % of people without access to sufficient and safe drinking water - % of people without access to adequate sanitation - rate of growth of urban population (%) - transport fuel consumption per capita (1)

Economic: - real GDP per capita growth rate (%) - ratio of net resource transfer/GDP (%)

Environmental: - annual energy consumption per capita (J) - annual withdrawals of groundwater and surface water as % of available water - annual domestic consumption of water per capita (m3) - land use - annual fuelwood consumption per capita (m3) - livestock levels per km2 dryland - population living below the poverty line in dryland areas (%) - annual use of agricultural pesticides (t km"2) - annual use of fertilizers (t km"2) - arable land per capita (ha per capita) - irrigation % of arable land - annual deforestation rate (km2 year 1) - annual roundwood production (m3)

State indicators: Social:

- population living in absolute poverty (number and %) - population density (people per km2) - infant mortality rate and maternal mortality rate (per 1000 live births) - life expectancy at birth (years) - % of population in urban areas - area and population of marginal settlements (km2, number) - human and economic loss due to natural disasters (number and US$)

Economic: - GDP per capita (US$) - EDP per capita (environmentally adjusted value added) (US$) - share of manufacturing value added in GDP (%) - total ODA given or received as percentage of GDP (%) - debt/GDP (%) - debt service/export (%)

Environmental: - groundwater reserves (m3) - average concentration and range of faecal coliforms in freshwater bodies

(no. per 100 ml)

Sustainability and reservoirs 39

- average BOD and COD and range in water bodies (mg l"1) - area affected by soil erosion (km2) / erosion index - land affected by desertification (km2) / desertification index - drought frequency (or intensity) - area affected by salinization and waterlogging (km2) - timber stocks (m3) - forest area (km2) - wood as % of energy consumption - threatened species as % of total native species

Institutional: - mandated EIA (yes/no) - programmes for national environmental statistics (yes/no) - sustainable development strategies (yes/no) - national councils for sustainable development (yes/no) - representatives of indigenous people in national councils for sustainable

development (yes/no) - existence of database for traditional knowledge information (yes/no) - representation of major groups in national councils for sustainable

development (advisory bodies in decision process)

Response indicators: Social:

- % GDP spent on health - infrastructure expenditures per capita (US$)

Economic: - investment share in GDP (%) - participation in regional trade agreements (yes/no) - ratio of consumption of renewable energy resources over non-renewable

energy resources (%) - environmental protection expenditure as % of GDP - (environmental) taxes and subsidies as % of government revenue - amount of new or additional funding for sustainable development

given/received since 1992 (US$) - programme of integrated environmental and economic accounting (yes/no) - debt relief

Environmental: - waste-water treatment (% of population served, total and by type of treatment) - land reform policy (yes/no) - cost of extension services provided and investment in agricultural research

(US$) - area of land reclaimed (km2) - reforestation rate (km2 year 1) - protected forest area as % of total land-use area - protected area as % of total land area

Extracted from UNCSD (1995). GDP: gross domestic product, EDP: environmentally adjusted value added domestic product, ODA: overseas development assistance, BOD: biochemical oxygen demand, COD: chemical oxygen demand, EIA: environmental impact assessment.

40 Sustainability and reservoirs

is a partial collection of the wisdom about sustainability interpretation from IAHS-related literature.

In the 1980s, IAHS launched a Hydrology 2000 Working Group consisting of young hydrologists with the mandate to look into the perspectives of development of hydrological sciences. Although the word "sustainable" was not used in the Hydrology 2000 report (Kundzewicz et al, 1987) which was pub­lished in the same year as the WCED report, the authors considered the idea of sustainability. They stated that "fresh water should not continue to be viewed as a cheap natural resource, or as a principal means of disposing of municipal, rural and industrial wastes. It is important to rely on "income" from water resources and not to deplete the water capital of the Earth, which has been inherited by our generation". This is essentially what sustainability means.

In the XXI IUGG General Assembly, Boulder, 1995, the ICWRS/IAHS symposium on "Modelling and Management of Sustainable Basin-Scale Water Resource Systems" (Simonovic et al., 1995) was held, where the importance of geographical information systems, various decision support systems and modelling tools for the sustainable management of water resources was illustrated by Smithers & Walker (1995), Schultz & Hornbogen (1995), Schumann (1995) and Kojiri (1995) amongst others.

In the Fifth IAHS Scientific Assembly, Rabat, 1997, a symposium on "Sustainability of Water Resources under Increasing Uncertainty" (Rosbjerg et al., 1997) was held where the focus was on the increasing uncertainty in the entire hydrological regime with particular reference to continents such as Africa. Orange et al. (1997) showed the climatic variability of drying Central Africa during the entire 20th century. Many papers in these proceedings such as Lorup & Hansen (1997), Van der Zel (1997), Sehmi & Kundzewicz (1997), Akiwumi (1997), dramatically report on the nature of African water problems and various depressing experiences. They indicated that the climatic and population pressure superimposed on the lack of knowledge, data and an integrated planning approach made the situation serious and that this is where more scientific efforts need to be directed.

Also in the symposium on "Human Impact on Erosion and Sedimentation" (Walling & Probst, 1997), there were major discussions on sedimentation in reservoirs and their long-term geomorphological effects. Based on 633 large reservoirs greater than 0.5 km 3 in the world, Vorosmarty et al. (1997) estimated the mean discharge weighted residence time to be two months and the estimated retention of sediment in the globe to be at least greater than 16% and, with 35 000 reservoirs, possibly greater than 25%.

Such efforts are further highlighted by the special issue of Hydrological Sciences Journal on "Sustainable Development of Water Resources", edited by Simonovic (1997b). The editor considers that the groundwork on defining sustainability, finding ways to measure it and identifying the steps for implementation has been completed and that the concepts presented in the papers have matured. Since they are the major efforts of IAHS on sustainable

Sustainability and reservoirs 41

water resources development, each of the 10 contributions is briefly described below under three general classifications :

Relationship between sustainable development and water resources management The first contribution in this group is by Falkenmark (1997). It takes as its starting point today's situation in which water illiteracy is widespread among those expected to cope with the global water crisis. The paper proposes some simple explanatory models, to be used for explaining and visualizing the fundamental interaction between people and water. It also discusses environmental sustainability criteria and their consequences in terms of the capacity to support water dependent populations.

The second contribution by Kundzewicz (1997) discusses the significance of water availability for sustainable development. The problems of water resources of vulnerable areas are tackled for the examples of arid and semiarid lands, mountains and small islands. The paper advocates a holistic perspective of water resources development.

Burn (1997) presents the development of a framework for data collection network design that considers sustainable development goals. Important components of the framework include a focus on hydrological information, the preservation of long-term gauging stations, and the adoption of integrated ecosystem monitoring.

Criteria for sustainable water resources management Bender & Simonovic (1997) present consensus as a flexible measure of sustainability. Consensus as a sustainability measure describes the level at which stakeholders are satisfied with the solution to a question.

McMahon & Mrozek (1997) debate the relevance of entropy to the economics of resource use. This paper places entropy as a limit to economic growth.

Loucks (1997) also focuses on the measurement of the relative sustainability of renewable water resource systems. Commonly used measures of reliability, resilience and vulnerability are combined in an index and used as a measure of changes in relative sustainability over time.

Matheson et al. (1997) present general fairness measures that may be used as criteria for sustainable project selection. Sustainable development, fair allocation objectives and empirical distance-based measures of fairness, and their evaluation are discussed.

Practical implications of sustainability Savic & Walters (1997) present the need for efficient pipe networks for urban water supply and drainage in the context of global sustainable development. The role of computer-based analysis and design tools in the planning and operation of such systems is described.

42 Sustainability and reservoirs

Das Gupta & Onta (1997) argue that an understanding of the behaviour of a groundwater system and of its interaction with the environment is required to formulate a sustainable management plan. A brief review of two case studies is provided to illustrate how a systems approach can be used in water allocation satisfying some of the technical and environmental constraints.

Takeuchi (1997) deals with the environmental impacts of large storage reservoirs. The paper presents the least marginal environmental impact rule which states that the reservoir size should not exceed the size whose marginal negative environmental impact is equal to that of any alternative means that can provide the same level of incremental gain in the global objectives.

IAHS has also co-sponsored various symposia and conferences. At the International Symposium on Water Resources Planning in a Changing World held in Karlsruhe, 1994, Loucks (1994) stated that "Most definitions of sustainable development include three broad notions: justice to nature, justice to future generations, justice within our generation." He continues "The sustainability of societies and ecosystems can only be realized by adaptation to changing circumstances and involves the processes of change, substitution and replacement. The continued existence of you or me or of some particular product or technology is not a necessary condition for sustainable development. [I 'm sorry to say!] A sustainable economy can only be realized if there is innovation, new ideas, and new technologies."

In the Third IHP/IAHS George Kovacs Colloquium on Risk, Reliability, Uncertainty and Robustness of Water Resources Systems, Paris, 1996, Nachtnebel (1998) suggested that the change of entropy caused by development (e.g. reservoir construction and accompanying and following changes) can serve as an indicator of sustainability. That is, this indicator assesses the amount of work needed to restore the original situation (decommissioning). The entropy would certainly be the basic indicator of environmental effects if it were measured. However, measurement would be extremely difficult and translation of this concept into a workable operational measure is not straightforward. Sugawara (1970) proposed to charge water supply not by the amount of water used but by the entropy increase in the quality change given to water. Naturally, the proposal was not implemented partly because measuring the quality of waste water at each household is not practicable.

At the IAHS co-sponsored International Conference on Water Resources and Environment Research held in Kyoto in 1996, Shamir (1996) reviewed numerous studies and pointed out that the various definitions of sustainability always included the two aspects, environmental quality and the interests of future generations. About the interests of future needs, he said "Future objectives and needs are subjective values, which should be left to the discretion of future generations to set. But they are not present at the time decisions must be made. Therefore we must stand for them and do our best to

Sustainability and reservoirs 43

represent them. If we cannot forecast their objectives and needs, we should assume that future generations will have values and needs similar to ours, and use conservative projections of the means which will be at their disposal."

The Scientific Committee on Water Research (SCOWAR) of the International Council of Scientific Unions" prepared a report on trends and needs in water resources research (SCOWAR, 1998). They considered the scientific issues to which existing interdisciplinary knowledge and mechanisms were insufficient. Sustainable reservoir development and operation was among the eight selected topics and it was stressed that physical sustainability, environmental concerns and efficient and sound operation and maintenance were the key areas for further research.

2.2. SUSTAINABLE DEVELOPMENT AND MANAGEMENT IN THE RESERVOIR CONTEXT

2.2.1 Introduction

Sustainable water resources management is, from the methodological viewpoint, a generalization of the notion of integrated water management which has been known as a concept for decades. However, it requires a clear commitment to broader objectives explicitly including ecological conservation and intra- and inter-generational equity issues. As a result, it calls for new instruments, such as demand management which has not been seriously exercised in the past. Management of water demand, using pricing mechanisms and other regulatory measures is a key instrument for sustainable freshwater resources management as proposed in Agenda 21 , Chapter 18 (UNCED, 1992). They are important everywhere in the world including the countries where supply increase is mandatory for water resources management. Without a coherent demand management policy, supply tends to create more demand and inefficient use. Chapter 3 will discuss the details of various means of demand management as non-reservoir options.

Sustainable water resources management may require a change of admini­strative system from an orientation towards supply to the development of management-oriented organizations. It may require a new allocation of ad­ministrative power endorsed by new legislation. The administration itself may need to be reorganized to achieve the objectives of sustainable development.

Sustainable reservoir management is a subset of sustainable water resources management and requires every notion of integrated management, multiple objectives, risk and uncertainty matters, systems approach and the like which should eventually be translated into workable criteria. Sections

N o w called the International Council for Science.

44 Sustainability and reservoirs

2.2.2-2.2.6 try to cover all the aspects that are required for sustainable reservoir management.

In Section 2.2.7, the rationale for the checklist given in Chapter 6 is presented. It is proposed that this checklist is used for planning, design, construction, and operation and maintenance of sustainable reservoirs.

Indicators (Section 2.1.3), criteria (Section 2.2.6) and the proposed checklist (Section 6.1) will be useful as a reminder of the important basis to which constant reference should be made during decision-making on reservoirs. It should not be overlooked simply because of the complex, urgent and conflicting reality of the design process.

2.2.2 Integrated water resources management

Integrated water resources management basically requires consideration of the system design and operation at basin scale. The considerations should include:

• integrated hydrological control of water and material transport;

• land use and vegetation management;

• upstream and downstream economic integrity; and

• collection and dissemination of hydrometeorological information.

Basinwide integration A river basin is the natural physical unit within which any water resources management works are interrelated and dependent on each other. Flood control, water use, sediment control and many other reservoir functions affect hydrological, socio-economic and environmental conditions of the basin from the headwaters to the estuary. Upstream versus downstream conflicts in socio-cultural and economic matters require large regional scale management and mutual understanding. The upstream basin management controls the water quantity, quality and sediment inflow into the reservoir. Reservoirs may impact on sedimentation, water quality, groundwater levels, river bed erosion, estuary retreat and marine culture, condition of navigable water, scenic, cultural and recreation values and many others. The water use downstream, its withdrawal and discharge, affects water level, infiltration, evaporation, and water quality. All these factors should be considered and coordinated from the planning stage throughout the whole lifetime of a reservoir as mutually dependent with integrated basin-scale hydrological and environmental cycles.

Reservoir sedimentation is a key factor for controlling the physical sustainability of a reservoir. In order to control sediments to and from reservoirs, bypass (compensation) channels, release gates, sediment check-dams upstream and other elements of infrastructure may be needed. Compensation channels are especially useful to curb deepening the river bed

Sustainability and reservoirs 45

downstream arising from erosion. Although the average reservoir lifetime is not very short (cf. Section 1.3.3), there are many reservoirs that may be filled by sediments in a few decades. It is obvious that the basin management upstream is the most fundamental remedy for sediment control in the long run. There are examples of reservoirs, such as the Meiktila Reservoir in Burma constructed nearly two millennia ago, which was sustainable for a long time as laws introduced a severe penalty for removing a tree in the drainage basin of the tributary to the lake. Once the regulations became less onerous, accelerated sedimentation of reservoirs began.

Evapotranspiration losses are important indices of reservoir sustainability. Human successes in suppressing evaporation are virtually non­existent. This means that evaporation from a reservoir should be given due consideration at the planning stage. Reservoirs with large inundated areas are not desirable not only due to the loss of much land, severe disturbances in ecosystems and the need for substantial relocation of people, but also because of the high water losses by evaporation, especially in arid and semiarid regions of the world. Evapotranspiration problems extend to the upstream catchment. As an example, the Bhumibol Reservoir in Thailand, built in 1964 with a capacity of 13.5 billion m 3 , has experienced a decrease of reservoir inflow since its completion. Some claim that the effect of climate change has been responsible but the major reason is believed to be the expansion of irrigation and other major uses of water in the upstream catchment, that is developments which, in fact, caused an increase of évapotranspiration upstream.

An example of integrated urban water management is a combination of artificial infiltration of urban rainfall underground and to cisterns in houses, which contributes to flood control, low flow augmentation and water supply, considerably reducing the dependency of urban areas on upstream reservoirs. More detail is provided in Section 3.2.

Information integration The integrated perspective also embraces the use of meteorological forecasts which can improve reservoir operation. If meteorological forecasts are used in an effective way, the same reservoirs may function better, satisfying more needs and thus possibly offsetting a part of new developmental needs. The techniques used for measurement of hydrometeorological variables, such as satellites, radar, automated and telemetered observation networks; and data transmission and processing techniques, have progressed rapidly. Together with the scientific advancement in hydrometeorology to obtain knowledge about physical processes, the advanced observations render complex numerical models operational. These may yield accurate forecasts. Yet, in general, reservoir operation is not taking sufficient advantage of advanced technology. This may be partly because reservoir operation is so sensitive to human society that even a small error creates an unacceptable impact on users and

46 Sustainability and reservoirs

accordingly the existing forecasting accuracy is still below the acceptable level. The other reason being that reservoir operators do not fully realize the benefit that may be obtained with the available level of forecasting accuracy. Section 4.2.2 shows an example by Takeuchi & Sivaarthitkul (1995) of how the introduction of accurate forecasts virtually expands the capacity of a reservoir and offsets some developmental needs for new reservoirs.

Integrated water resources management necessarily increases the number of system components to be jointly considered and the amount of data to be collected, stored, analysed and used. The necessary information typically includes hydrometeorological records, past operational experiences, GIS with various layers, such as topography, land use, social, physical, chemical, administrative and other components. Integrated water management is possible only with the aid of an advanced computer system with an adequate knowledge base and a capable inference engine to update, process, analyse and utilize the available information. Computer graphics and man machine dialogue devices are useful for the development/creation of public awareness in water resources planning and management,

Artificial intelligence and other computer-aided decision support systems lend themselves particularly well to applications in the field of reservoir operation (Simonovic, 1996a,b). One of the reasons is the complexity of integrated management. However, in addition, reservoirs are often located in remote mountainous areas with difficult access. They are therefore difficult to maintain, in particular in winter, during heavy rain or snowfalls and during natural hazards such as floods. At the same time, the amount of accumulated past data is very high and growing fast. The proper use of these data is essential for improved management of water resources. Thus, both in reservoir operation and design, the introduction of automated decision support systems is anticipated. However, its introduction is currently being cautiously received in field operation offices.

From the reliability and robustness viewpoints, it should be noted that dependence on computer knowledge base and decision support systems, coupled with automated remote controls, can be quite risky. If electricity goes off, if machines malfunction, if unwelcome surprises appear and so forth, human ability to control automated systems is very limited, as evidenced in examples of aircraft crashes and disasters in nuclear power stations. There are many efforts to get around such problems and to pre-program for such emergency cases within the system. Indeed, a cautious approach to artificial intelligence systems is justified until truly intelligent robots with self-learning, self-thinking and risk-conscious minds become available. Human ability to control emergency cases such as gate operation of reservoirs should always be permitted and appropriate arrangements (including training) made. Under such circumstances, integrated large systems should be subdivided into manageable sizes.

Sustainability and reservoirs 47

2.2.3 Notion of multiple objectives

Water resources planning and management for maximization of net economic benefits is no longer sufficient. The introduction of the principles of sustainability expands the range of issues that must be incorporated in the objectives of water resources planning. The main modifications are required to account for:

• non-economic objectives (example: requirement of no net loss of ecosystem productivity);

• needs of future generations;

» distribution of costs and benefits;

• balancing inequalities (within, and between, generations);

• increase in efficiency; and

• avoidance or minimization of irreversible effects.

Sustainability places a significant weight on replacing single objective optimization with a multi-objective analysis where a set of trade-offs (nondominated, noninferior, compromise, Pareto optimal solutions) can be identified. Selection of the best compromise solution is a political decision. Trade-offs are an inherent part of the consensus reaching process. Simonovic (1989) has shown that the idea of replacing the best compromise solution with the most robust solution is an appropriate one to be used in the multi-objective analysis for sustainable development of water resources.

Necessary institutional setting In order to render integrated water resource management possible, it is most important to establish an institutional setting that can allow various policies to be implemented. Integrated water resource management includes basinwide multi-objective management, optimal allocation of water with economic and environmental evaluation, the use of demand management with pricing mechanisms and regulatory measures, the development of alternative sources of water supply including waste-water reuse and water recycling, water conservation through improved water-use efficiency, water-saving and effluent control. Many of these measures could be implemented within the current administrative and institutional structure. But in practice they would all need a drastic modification and rearrangement of current administrative sectoring, regulations, water rights etc. to make the multi-sectoral approach possible. As regards reservoir operation, the allocation of water rights set separately for each reservoir is an obstacle to utilizing available water in an efficient way and to managing the water resources in an integrated manner at the basin scale. Joint operation of a set of reservoirs makes it possible to benefit from coordination. Even during severe droughts, enforcement of a temporary transfer from less urgent agricultural

48 Sustainability and reservoirs

water use to more urgent municipal water use is not easy. Still more problematic is the introduction of new demands such as environmental use. The reliability and robustness required can never be achieved within rigid administrative constraints, sectionalism and regulations. The opportunity cost of administrative rigidity is very high creating inefficient water use, difficulty for risk management and extra demand for reservoir development.

Sustainable development requires improvement in efficiency of use of environmental, economic, manpower, time and other resources. Efficient use of existing reservoirs should precede any new reservoir construction since creating a new major element of infrastructure is expensive, affects people and the environment and also uses the irreversible dam site resources. There must be a great possibility of increasing efficiency in flood control, water supply, etc. by improvement of operation of existing reservoirs, integrated with such measures as predictions of hydrometeorological phenomena as well as demand patterns and socio-economic preferences. The rational allocation of water rights according to realistic priorities, i.e. earmarking more water to the most urgent need, is of utmost importance. Optimality is a difficult concept in a multi-objective scheme but it can be colloquially stated as efficient and rational water use without critical institutional constraints.

Notion of equity and democracy One other aspect of sustainability considerations of reservoirs is the notion of equity, which refers to other passengers of the spacecraft Earth at the present time and in future generations. Intra-generational equity requires consideration of a range of issues belonging to the realm of social systems, organizational levels, institutional arrangements and public participation, which can be collectively called human dimensions of sustainable reservoir development. In particular, one should be cognisant of:

• respect of the opinions of reservoir site communities;

• care for involuntarily displaced people;

• benefits to the poor, the weak, the illiterate;

• fairness between upstream and downstream residents; and

• democratic and collective process for decision making and conflict resolutions.

A decision on the construction of a reservoir requires a broad consultation. People affected by the project need to be given the opportunity to express their opinion which, in turn, should be seriously taken into account in the decision­making process. The process of mass displacement needs to be considered with due concern in order to at least maintain the living standard of the involuntarily displaced people. Simple monetary compensation is not an adequate solution for those people who may have never before managed a

Sustainability and reservoirs 49

large amount of money in their lives. They need to be taken care of until their new life is stabilized. Socio-economic fairness in upstream and downstream development is one of the most important issues in any basin, where water quantity and quality are the limited resources to be shared for development.

Public participation, including that of women, young people, indigenous people and local communities is essential to sustainable reservoir development. As mentioned in equity considerations above, this is the basic condition for equity criteria to be satisfied. It also plays other important roles. It is a challenging issue to identify how the current political movement of regionalization and the overall weakening of government intervention affects such needs of governmental coordination and where the new leadership in achieving integrated water resource management comes from. It may be pointed out that a strong government usually has strong sectionalism, which creates the need for coordination whilst acting as a barrier to coordination. Regionalization and the weakening of central government would weaken sectionalism and, as a result, create a more flexible basis for integrated basinwide management. As for the source of new leadership for integrated management in a regionalized weak government, public awareness and citizen participation would play an important role. Information disclosure (e.g. via various computer-aided decision support systems) would be indispensable in allowing the public to have a correct understanding of alternative courses of action and their consequences and to reach wise judgements.

A democratic institutional structure is an important prerequisite for sustainable development. Planning of non-sustainable schemes is frequently developed behind locked doors, and the concentration of uncontrolled power allows the few to close the door to the many and gives birth to the monument-building syndrome. Increased public involvement will create more forethought in the decision-making process and there is a reason to expect a more fair allocation of benefits and costs. Pursuing sustainable development of water resources will require major changes in both substantive and procedural policies. The diverse policy questions raised include:

• How should the methods and processes of impact assessment and planning be used?

• What should be the role of the market as opposed to direct regulatory mechanisms?

• What should be the role of public and interest groups in the management of the resource?

• How much should be invested in managing the resource and how should this be financed?

It is difficult to specifically consider the notion of inter-generational equity in reservoir design and planning. Its most concrete aspects are:

50 Sustainability and reservoirs

• conservation of nature, ecological systems and cultural heritage;

• dam safety;

• quality of reservoir water;

• sediment control;

• rehabilitation; and

• reversibility.

Most of them are important for the contemporary generation, too. However, dams are expected to serve not only for the present but also for a long time into the future. Rehabilitation programmes for existing reservoirs may improve their sustainability (cf. Section 4.3.2). Other aspects of inter-generational equity listed above are also discussed in this section.

A release from a reservoir which is close to the natural flow pattern of the hydrological variability including floods and droughts may be considered beneficial from the viewpoint of natural habitats of aquatic fauna and flora. This would surely help the conservation of nature, possibly at the expense of the economic use of water. The balance between economic use and the environmental use of water is shaped by the people. For example, after many years of discussion, it was decided to release water from the Glen Canyon Dam in Colorado in order to create artificial floods. It was not an easy decision to take. The hydropower loss by this operation is not marginal although the water is recaptured again downstream by the Hoover Dam.

Consideration of hydropower When analysing hydropower development, impacts on sustainability of the use of energy (production, urbanization etc.) should also be included. Electricity rates have frequently been found to be inefficient and serving to subsidize industrial activities that have significant impacts on the overall sustainability (Goodland et al., 1992). These authors recommend that hydropower development should not be considered unless:

• the price of electricity has reached the long-term marginal cost;

• most energy conservation and efficiency measures are substantially in place;

• all economically perverse subsidies and other incentives have been rescinded; and

• rehabilitation, re-development and expansion of existing sites have already been accomplished.

Agreeing with the principle of this approach one could note that the process of reservoir planning takes a long time. Typically, it is based on a projection of energy demand some time into the future. Suppose that there are still reserves

Sustainability and reservoirs 51

in energy conservation, yet even full implementation of these reserves, whose efficiency is highly uncertain, may still be insufficient for meeting future energy demands. The search for reserves in energy conservation is to be encouraged. Brazil has probably saved more than US$ 1 billion in new generation capacity because of recent major improvements in the electricity tariff structure.

A proposed dam should have a high rate of power production per area inundated. The acceptable efficiency of the use of the area should depend on the value of the ecosystems and production systems. Goodland et al. (1992) suggest that if the ecosystem to be flooded is an intact primary tropical forest, then the acceptable ratio should be set higher, e.g. to 100 kW ha"1. In the case of impounded agricultural areas or degraded lands, the acceptable ratio could be respectively lower. However, it does not seem possible to devise general guidelines for acceptable limits in quantitative terms.

Thompson (1994) compared two hydroelectric projects of different sizes—a mini-hydro and a micro-hydro in the Himalayas. The smaller scheme compared favourably to the bigger one in terms of its sustainability: people with tiny incomes seem to look sufficiently well after a microscale hydropower station. Thompson (1994) advanced the hypothesis that improving sustainability and enhancing resilience are very close to each other. Synghal (1994) sketched how a small, community-owned reservoir system may contribute to sustainable development by triggering community welfare (activating initiative, assuring water supply, increasing agricultural production, livestock, fuel and timber, improving family and community income, advancing education, health and recreation infrastructures). The reservoir in question, increased the availability of gainful employment in the locality without the need for the population to migrate to a city. It is easy to see that the case presented by Synghal (1994) can be conveniently described by a number of indicators of sustainable development included in Section 2.1.3 describing social, economic, environmental and institutional indicators of sustainable development. In the same reference, Synghal (1994) compared a system of 34 small reservoirs giving comparable benefits at a total cost of one third of that for a project with a single large reservoir.

2.2.4 Risk- and uncertainty-related considerations

Dam safety is partly an efficiency issue in the economic sense and partly an inter-generational equity issue. A reservoir must be safe and serve as long as possible. Takahasi (1990) considered several cases of dam failure: the South Fork Dam, Pennsylvania (of height about 22 m (72 ft) and length about 284 m (931 ft), completed in 1852) broke in 1889. The dam burst, triggered by the blockage of the spillway by trees etc., resulted in a major disaster with a substantial loss of human life (2209 people killed in Johnstown). The

52 Sustainability and reservoirs

Malpasset Dam, France (arch dam, height 66.5 m, dam length 222 m, storage capacity 50 million m 3 , completed in 1959) broke in 1959 due to rock movement on the bank of the dam on the first occasion that full capacity was reached. The Teton Dam, Columbia River, USA (earth-fill dam, height 93 m, dam length 930 m, storage capacity 356 million m 3 , completed 1975) broke in 1976 initiated by piping flow on the bank during the first filling, before water storage reached the full capacity. In the case of the Viant Dam, Italy (arch dam, height 262 m, dam length 190 m, storage capacity 169 million m 3 , completed 1961), a volume of 300 million m 3 of earth from a landslide after heavy rain in 1963 created overtopping of 30 million m 3 of water, which killed 2125 people and destroyed 595 houses downstream. These are clear cases of unsustainable reservoir design. Increasingly risk-averse societies wish to reduce the probability of rare, yet highly serious, disastrous events. They are willing to pay increasing costs for securing safety.

Sustainable management of water resources necessitates consideration of longer time scales over which various anticipated and unexpected uncertainties may occur. Nobody knows with certainty the degree of population growth, urbanization, industrialization, climate change, societal preferences and value system changes, and technological innovation over the longer term, say, half a century. And it is precisely changes in these processes that strongly influence the hydrology, objectives, constraints, evaluation criteria, operational domain and many other conditions under which water resource systems must operate. There currently exists no firm design methods to incorporate such uncertainties or to increase the ability of the system to adjust to new conditions, to accommodate the unexpected, and perhaps there never will be.

Expanded spatial boundaries, lengthened time scales, comprehensive multi-objective analysis and other issues related to sustainable water management are placing immense demands on science. A number of questions raised by the sustainable development perspective of water resources reveal major deficiencies in the knowledge of the behaviour of a wide range of natural and human systems under consideration. Recognizing the fact that many of these deficiencies cannot be eliminated in the short term, makes it evident that risk and uncertainty are inherent concepts related to sustainable water management.

It is common to study various predictions and scenarios. They are meaningful when the probability and conditions of occurrence of each scenario are, to some extent, scientifically identifiable and the decision can be made on such probabilities. In reality, however, the probability of occurrence of each realization is too uncertain to be relied upon. Scenario analyses are therefore useful for selecting policy alternatives for the relatively short and known future. In reservoir management these scenarios are not so much concerned with the physical design, except for the design in stages with the possibility of add-ons, but rather the operation and management that reservoirs can cope with and adapt for unknown future conditions. Some increase in resilience and

Sustainability and reservoirs 53

robustness in reservoirs can be assured, in physical terms, if the following conditions for distributed systems are fulfilled:

• many small reservoirs rather than a single giant one;

• many alternative water sources and storage facilities; and

• water transfer and exchange lines and networks.

However the major sources of resiliency and robustness reside in managerial flexibility in reservoir use. Water allocation, integrated operation with other reservoirs and sources, demand management policy and other reservoir use policies all help to make the physical reservoir capable of meeting unknown or crisis situations.

Flexibility means the requirement to meet changing constraints and changing quality criteria, and accounting for side effects which were not considered at the design stage. This criterion is especially important for sustainable development in the context of a longer time scale to be considered and a wide spectrum of possible future uncertainties to be managed. Flexibility is not necessarily the adaptability of a system to meet two or more scenarios, but rather the adaptability to meet unexpected occurrences. Flexibility is a notion, germane to resiliency and robustness, which relates to the time necessary to adjust to the unexpected occurrence and the degree of surprise to which a system can be adjusted. One can distinguish physically-based flexibility and managerial flexibility. The physical flexibility of a dam is not very high, except for add-on design. However, as a reservoir can serve many purposes in many ways, the managerial flexibility can render the storage useful or useless, with a range of intermediate levels.

Sustainable water resources planning and management include, by definition, long-term consequences in the analysis. This implies examining not only the longer term consequences of proposed developments but also the possibilities of modifying or even reversing the consequences of past commitments. It is especially difficult for government plans to be modified. A dam construction process may sometimes take more than three decades including planning, authorization, residents' agreement, financing, land acquisition and construction, while societal needs and the economic situation may change in both directions. The dam may even become no longer necessary before completion of the infrastructure. In most countries, it is very difficult to modify or to cancel past commitments of the government. For sustainable reservoir development, however, a flexible decision-making process adjustable to such changes is indispensable.

Physical reversibility is difficult, but there are some examples, as in the case of Elwha (see Section 1.3.4), coming into the picture now. Since dam sites are limited, reversible dam construction, if ever possible, is a rational way to ensure sustainable river use. The technology required for dam removal may not be very simple. But it would be a matter of cost and

54 Sustainability and reservoirs

economic justification rather than a technological challenge. Any very large dam that contains huge amounts of sediment would take decades to remove and little justification is likely under normal circumstances. Small dams would have more chance. If reversibility is included as a managerial option for reservoirs, the original plan has to include the cost of removal and the financial means have to be identified from an early stage. In this sense, one has to consider the life cycle of a dam, in the context of environmental impacts and economy.

2.2.5 Systems view

Reservoir analysis, design, planning and operation are difficult and complex problems. They require expertise from numerous fields like engineering, economics, physics, chemistry, biology and zoology. No one can be expert on every subject. But when we turn to experts for help, they often seem confused and isolated, arguing with each other and looking only at those pieces of a reservoir problem that happen to fit their own particular specialities.

Systems thinking (often called systems approach, systems view etc.) and its practical application known as "systems analysis" have been recognized as essential tools for reservoir analysis. Probably the most important contributions of systems thinking as it applies to reservoirs are (a) provision of a way of tackling those big, complicated, real-world reservoir problems which do not fit neatly into various specialities and (b) provision of a way for decision makers and the general public to get a clear picture of how reservoirs and their environment work.

The systems approach is a discipline for seeing entities. It is a framework for seeing interrelationships rather than things and seeing patterns rather than static snapshots. It is a set of general principles. It is also a set of specific tools and techniques.

We live and work within social systems. Our research is exposing the structure of nature's systems. Our technology has produced complex physical systems. But even so the principles governing the behaviour of systems are not widely understood. A systems approach is needed now more than ever before because we are becoming overwhelmed by complexity. Probably, for the first time in history, human kind has the capacity to create much more information than anyone can absorb, to foster far greater interdependency than anyone can manage, and to accelerate change far faster than anyone's ability to keep pace.

Problems related to reservoirs (Section 1.4) are examples of system breakdowns. Complexity can easily undermine confidence and responsibility. Systems thinking is one cure for this type of helplessness. It allows us to see the structure that underlines complex situations. Through proper implementation of a systems approach to reservoir analyses we can expect:

Sustainability and reservoirs 55

• To understand reservoirs and their interactions with the surrounding environment much more easily. The basic rules of how reservoirs work apply to other kinds of system (social, political, economic, ecological, physical) and vice versa.

• A systems approach to help us identify "high leverage points" in the reservoir systems where the effective strategies for influencing them will have a greater chance of success.

2.2.6 Sustainability criteria for possible use in reservoir analysis

In order to describe in more detail the behaviour of water resource systems, with regard to sustainability, performance indices (PI) are used, classified as resilience, vulnerability, grade of service, availability, quality of service etc. In addition to performance indices, figures of merit (FM) are also used for the analyses of system behaviour. They are defined as functions of performance indices. Some attempts are made to measure sustainability using different FMs. One idea views sustainability as a combination of high resiliency and low vulnerability (Duckstein & Parent, 1994). Another idea involves identifying a new FM as a weighted statistical index to directly describe sustainability (Loucks, 1997). Yet another concept used to assess planning decisions in terms of sustainability includes entropy (McMahon & Mrozek, 1997).

Most criteria for sustainable decision-making are not yet in operation. Environmental integrity, for example, is difficult to define in an operational form. Inter-generational equity is difficult to comprehend except as a combination of more measurable terms such as reversibility, resiliency or robustness. Unlike the above concepts, economic efficiency is better understood and is commonly applied.

Four practical criteria developed for sustainable water resources decision­making with potential application to reservoir developments: fairness, reversibility, risk and consensus are recommended by Simonovic et al. (1997) and Bender & Simonovic (1997). The sustainability definition emphasizes the integral treatment of three subsystems: economic, social and ecological. In general, developmental decision-making related to reservoirs becomes progressively more complex with the growing recognition of the comprehensive linkages between the natural (ecological), economic and human (socio-political, inter-temporal) subsystems to be considered. The selection of inter-temporal fairness, reversibility, risk and consensus as sustainable decision-making criteria is an attempt at addressing some of these issues.

Fairness Fairness or equity, is an important consideration in the selection of large-scale reservoir project alternatives. Consideration of the fairness of

56 Sustainability and reservoirs

impacts of a reservoir project is important both to ensure the maintenance of social well-being and to secure project acceptance by affected stakeholders. The fairness of a distributive situation can be quantified using a variety of distance-based approaches which result in both intra-temporal and inter­temporal fairness measures (Lence et al., 1997).

The distance-based measures (Lence et al., 1997) are grouped according to whether they are essentially measures of proportionality, quality or need. A set of required principles and characteristics for distance-based distributive fairness measures are necessary and these can be extended for inter-temporal consideration. General formulations are developed for intra-temporal fairness and inter-temporal fairness. Overall fairness may be interpreted as a combination of equity, equality, and need-based fairness objectives. The overall fairness of a reservoir alternative can be used in the process of selecting project alternatives for implementation.

Reversibility Reversibility is viewed as a measure of the degree to which the aggregated set of anticipated and unanticipated impacts of a reservoir can be mitigated (Fanai & Burn, 1997). Reservoir projects that are highly reversible should allow the users of the affected system to continue their normal use. A high degree of reversibility requires the imposition of the least amount of disturbance to the natural environment. The reversal of adverse effects is often not technically feasible, but mitigation plans, or the provision of substitute resources, can help to reduce the negative effects.

The reversibility framework developed by Fanai & Burn (1997) considers all alternatives of the reservoir project of interest and determines the degree of reversibility associated with each project alternative. An alternative that is more reversible is superior to alternatives that have a lesser degree of reversibility. The framework for measuring reversibility involves several tasks. These tasks include: (a) identifying and categorizing impacts; (b) classifying the impacts, if necessary; (c) specifying units of measure for the purpose of quantifying each impact; (d) specifying weights for each impact; and (e) applying a formula to obtain indices of reversibility. The framework results in measures of reversibility that can be used as a part of the evaluation and selection of reservoir alternatives.

Risk Risk exists when there is the possibility of negative social, environmental or economic impacts associated with a reservoir. Risk can be defined as the product of the magnitude of a negative consequence and the probability of occurrence of that consequence (Simonovic, 1997a). Risk can be estimated using combinations of historical and empirical data, heuristic knowledge and cultural perceptions. A composite measure of risk is influenced by the weighting that is given to the various components of the risk measure.

Kroeger & Simonovic (1997) have developed an algorithm for the evaluation of risk that can be used in the process of selecting reservoir

Sustainability and reservoirs 57

project alternatives. The intent of the risk measure is to involve project stakeholders in the process of quantifying the risks associated with each project alternative. The algorithm for quantifying the risk measure includes the following steps: (a) identifying the risks that are relevant for the analysis; (b) estimating the probability of each of the risks occurring for each alternative; (c) calculating the risk magnitude for each risk for each stakeholder group; (d) estimating the risk separately for each alternative and each stakeholder group; (e) comparing the alternatives by combining the risk estimates from the stakeholders; and (f) analysing the future joint estimates of risk.

Consensus Consensus, as the concept for promoting sustainability in reservoir analyses is a criterion quite unlike the others described previously (Bender & Simonovic, 1997). Consensus has no units of measurement. It is measured in a brief moment of time, but may implicitly consider future events and uncertainties. Consensus is a high level criterion, dependent on value judgements which may in turn depend on lower level indicators derived from facts concerning problem characteristics. The definition for consensus in Webster 's Dictionary is "a general agreement in opinion". It relies on a qualitative and subjective opinion, and the qualifying condition is a general agreement. How well do they need to agree? There may not be an adequate answer to this question, but a consensus approach may indicate more than just when to stop an exploration of alternatives.

If consensus leads to sustainability, what is consensus sustainability in an operational form? The following definition for consensus as it relates to sustainability is suggested: "Consensus is an equitable compromise which is robust with regard to (a) reservoir management uncertainties, and (b) stakeholder perspectives". Reservoir management uncertainties include data uncertainty, model uncertainties, and technological uncertainties. Stakeholder perspectives are related to the value systems. This definition is not yet operational, but its constituent parts might be manageable. There are some assumptions which also need to be made. It is assumed that the appropriate stakeholders have been included in the decision-making process. By stakeholder, we refer to interested parties which may be impacted in some way by any decision that is made (a political choice). The second major assumption is that all stakeholders voluntarily cooperate in the decision­making process.

The decision-making process for a reservoir involves the creation of alternative choices (in the design and operations framework) that appear to satisfy developmental criteria, and are financially feasible and institutionally acceptable. The addition of four new criteria to be used in decision-making is placing much greater weight on replacing single-criterion optimization with multi-criteria analysis. Sustainable reservoir decision making is subject to four major sources of complexity regarding the application of multi-criteria

58 Sustainability and réservoirs

analysis: (a) addition of sustainability criteria to the payoff table; (b) quantification of different criteria; (c) obtaining preferences from decision­makers regarding the different criteria; and (d) avoiding overlap between different criteria. These problems are being addressed in current research efforts. Meanwhile, this publication provides an alternative practical approach, the checklist, which serves as a vehicle for considering some of the points relevant to reservoir sustainability.

2.2.7 Rationale of the checklist for sustainable reservoir development and management

The aspects discussed in Section 2.2 are transformed into a checklist to be examined in various steps of reservoir planning, design, construction and operation. It is a collection of concrete actions to be taken during those steps. Although the list itself is presented in Chapter 6 as part of the conclusion of this report, the present section provides the rationale behind it. It is hoped that the checklist will serve as a guideline for implementation of the concept of sustainable reservoir development and management.

A sustainable reservoir is a reservoir designed and managed in accordance with the principle of sustainability. It is designed and managed as an integral part of the holistic system of society, land, air and water. The checklist was prepared based on the following rationale:

(1) Existing reservoirs are fully utilized. Existing reservoirs, as well as those under construction or authorized to be constructed, are used or will be used efficiently, integrated with other physical and managerial components of water resource systems. Efficient use of existing reservoirs can offset or postpone the need for new reservoir construction or reduce their necessary size. Efficiency of reservoir operation may be enhanced, among others, by such means as conjunctive use with alternative sources, lengthening the life of a reservoir by sediment control, use of hydrometeorological forecasts and priority and flexible allocation of water rights among users.

(2) Alternatives are exhaustively examined. The option of constructing a reservoir should be considered together with many other alternative ways, managerial or physical, that can equally attain the objectives to be achieved. This should include, as extensively described in Chapter 3, economic incentives, legal arrangements, modification and restructuring of existing water resource systems, use of other existing as well as new sources, efficient water distribution and use and many other demand management methods.

Sustainability and reservoirs 59

(3) Selection of a reservoir option should be made under the sustainability criteria. A reservoir option has to be selected under the sustainability criteria, as discussed in the previous sections of Chapter 2, specifically considering future generations, ecological soundness, fairness to the indigenous people, robust performance, sedimentation and other long-term consequences, in addition to economic efficiency, reliability and structural safety.

(4) Reservoir size is determined using the least marginal environmental impact rule. If the reservoir option is selected, the size should not exceed the level where the marginal socio-environmental impact is the least among all other alternatives that can provide the same level of satisfaction in objectives. There is no reason to increase a reservoir size if alternative means can provide the same objective with less negative impacts. This is the procedural means to ensure sustainable reservoir design, explained in detail in Section 4.3.4.

(5) Democratic decision-making process is followed. The decision-making process on reservoir development should be open to all the indigenous people, affected people, interest groups and other public with full information disclosure related to the plan and be carried out in a democratic way at all levels of decision making. In the event that a reservoir is built, the life of the people involuntarily displaced should be taken care of until they and their community regain their existing vitality and viability.

(6) Mitigation measures are fully taken. If reservoirs are built, construction, inundation and operation should be planned and implemented, by using all feasible mitigation measures, to minimize the negative environmental impacts and ensure the quickest recovery of the damaged ecological system.

(7) Reservoir is post audited over the full life cycle of its existence. The reservoir should be developed and managed considering its full life cycle from the planning stage to the time when it is filled by sediments. The environmental situation, sedimentation and the reservoir's use and impacts in the whole basin should be continuously audited and proper measures and modification have to be taken to make it function properly over its total lifetime.

(8) Systematic approach. New technology and tools are emerging everyday. Various data are

60 Sustainability and reservoirs

acquired and are accumulating. A systems approach provides the greatest possibility to use these available resources for reservoirs so that they perform better with higher efficiency and reliability and to make the decision-making procedure more democratic.

REFERENCES

Akiwumi, F. A. (1997) Conjunctive water use in an African river basin: a case study in poor planning. In: Sustainability of Water Resources under Increasing Uncertainty (eds D. Rosbjerg, N. -E. Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F. Rasmussen), IAHS Publ. no. 240, 495-502.

Allen, R. (1980) How to Save the World. Kogan Page, London.

Bender, M. J. & Simonovic, S. P. (1997) Consensus as the measure of sustainability. Hydrol. Sci. J. 42(4), 493-500.

Burn, D . H. (1997) Historical information for sustainable development. Hydrol. Sci. J. 42(4), 481-492.

Coomer, J. (1979) The nature of the quest for a sustainable society. In: Quest for a Sustainable Society (ed. J. Coomer) . Pergamon Press, Oxford.

Das Gupta, A. & Onta, P. R. (1997) Sustainable groundwater resources development. Hydrol. Sci. J. 42(4), 565-582 .

Duckstein, L. & Parent, E. (1994) Systems engineering of natural resources under changing physical conditions: a framework for reliability and risk. In: Natural Resources Management (eds L. Duckstein & E. Parent). Kluwer, Dordrecht.

Falkenmark, M. (1997) Society's interaction with the water cycle: a conceptual framework for a more holistic approach. Hydrol. Sci. J. 42(4), 451-466.

Fanai, N . & Burn, D. H. (1997) Reversibility as a sustainability criterion for project selection. International Journal of Sustainable Development and World Ecology 4(4), 259-273 .

Goodland, R. J. A. , Juras, A. & Pachauri, R. (1992) Can hydroreservoirs in tropical moist forest be made environmentally acceptable? Energy Policy 20(6), 507-515.

IUCN (1991) Caring for the Earth, a Strategy for Sustainable Living. IUCN, U N E P & W W F , Earthscan, London.

Jordaan, J. , Plate, E. J. , Prins, E. & Veltrop, J. (1993) Water in Our Common Future. Committee on Water Research (COWAR), UNESCO International Hydrological Programme, Paris.

Kojiri, T. (1995) Planning and management of water resources systems under uncertain future information using fuzzy optimization. In: Modelling and Management of Sustainable Basin-Scale Water Resource Systems (eds S. P. Simonovic, Z. W. Kundzewicz, D. Rosbjerg & K. Takeuchi), IAHS Publ. no. 2 3 1 , 311-318.

Kroeger, I. H. & Simonovic, S. P. (1997) Development of a risk measure as a sustainable project selection criteria. International Journal of Sustainable Development and World Ecology 4(4), 274-286.

Kundzewicz, Z. W. (1997) Water resources for sustainable development. Hydrol. Sci. J. 42(4), 467-480.

Kundzewicz, Z. W. , Gottschalk, L. & Webb, B. (1987) Hydrological sciences in perspective. In: Hydrology 2000 (eds Z. W. Kundzewicz, L. Gottschalk & B. Webb), IAHS Publ. no. 171, 1-7.

Lence, B. J., Fùrst, J. & Matheson, S. (1997) Distributive fairness as a criterion for sustainability evaluative measures and application to project selection. International Journal of Sustainable Development and World Ecology 4(4), 245-258.

L0rup, J. K. & Hansen, E. (1997) Effect of land use on the streamflow in the southwestern highlands of Tanzania. In: Sustainability of Water Resources under Increasing Uncertainty (eds D. Rosbjerg, N . -E . Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F. Rasmussen), IAHS Publ. no. 240, 227-236.

Loucks, D. P. (1994) Sustainability: what does it mean for us and what can we do about it? In: Proc. Intern. Symp. Water Resources Planning in a Changing World, Karlsruhe, 28-30 June 1994, section I, 3 -12 .

Loucks, D . P. (1997) Quantifying trends in system sustainability. Hydrol. Sci. J. 42(4), 513-530.

Matheson, S., Lence, B. & Fùrst, J. (1997) Distributive fairness considerations in sustainable project selection. Hydrol. Sci. J. 42(4), 531-548.

McMahon, G. F. & Mrozek, J. R. (1997) Economics, entropy and sustainability. Hydrol. Sci. J. 42(4), 501-512.

Sustainability and reservoirs 61

Nachtnebel, H. P . (1998) Irreversibility and sustainability in water resources systems. In: Risk, Reliability, Uncertainty and Robustness of Water Resources Systems (eds J. J. Bogardi & Z . W. Kundzewicz). International Hydrology Series, Cambridge University Press, in press.

Orange, D . , Wesselink, A. J. , Mahé , G. & Feizoure, C. T. (1997) The effect of climate changes on river baseflow and aquifer storage in Central Africa. In: Sustainability of Water Resources under Increasing Uncertainly (eds D. Rosbjerg, N . -E . Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F. Rasmussen), IAHS Publ. no. 240, 113-123.

Rosbjerg, D . , Boutayeb, N . -E . , Gustard, A. , Kundzewicz, Z. W. & Rasmussen, P. F . , eds (1997) Sustainability of Water Resources under Increasing Uncertainly. IAHS Publ. no. 240.

Savic, D. A. & Walters, G. A. (1997) Evolving sustainable water networks. Hydrol. Sci. J. 42(4), 549-564. Schultz, G. A. & Hornbogen, M. (1995) Sustainable development of water resources systems with regard to

long-term changes of design variables. In: Modelling and Management of Sustainable Basin-Scale Water Resource Systems (eds S. P. Simonovic, Z . W. Kundzewicz, D. Rosbjerg & K. Takeuchi), IAHS Publ. no. 2 3 1 , 329-338.

Schumann, A. H. (1995) Flexibility and adjustability of reservoir operation as an aid for sustainable water management. In: Modelling and Management of Sustainable Basin-Scale Water Resource Systems (eds S. P. Simonovic, Z. W. Kundzewicz, D . Rosbjerg & K. Takeuchi), IAHS Publ. no. 2 3 1 , 291-297.

SCOWAR—Scientific Committee on Water Research of the International Council of Scientific Unions (1998) Water resources research: trends and needs in 1997. Hydrol. Sci. J. 43(1), 19-46.

Sehmi, N . S. & Kundzewicz, Z. W. (1997) Water, drought and desertification in Africa. In: Sustainability of Water Resources under Increasing Uncertainty (eds D. Rosbjerg, N . -E . Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F . Rasmussen), IAHS Publ. no. 240, 57 -65 .

Shamir, U. (1996) Sustainable management of water resources. In: Proc. International Conference on Water Resources and Environmental Research, 29-31 October 1996, Kyoto, vol. 2 , 15-29.

Simonovic, S. P. (1989) Application of water resources systems concept to the formulation of a water master plan. Water International 14(1), 37-50 .

Simonovic, S. P. (1996a) Decision support systems for sustainable management of water resources 1. General principles. Water International 21(4), 223-232.

Simonovic, S. P. (1996b) Decision support systems for sustainable management of water resources 2. Case studies. Water International 21(4), 233-244.

Simonovic, S. P. (1997a) Risk in sustainable water resources management. In: Sustainability of Water Resources under Increasing Uncertainly (eds D. Rosbjerg, N . -E . Boutayeb, A. Gustard, Z . W. Kundzewicz & P . F. Rasmussen), IAHS Publ. no. 240, 3-17.

Simonovic, S. P . , ed. (1997b) Hydrol. Sci. J. 42(4). Special issue on "Sustainable Development of Water Resources" .

Simonovic, S. P . , Kundzewicz, Z. W. , Rosbjerg, D . & Takeuchi, K., eds (1995) Modelling and Management of Sustainable Basin-Scale Water Resource Systems. IAHS Publ. no. 231 .

Simonovic, S. P . , Burn, D . H. & Lence, B. J. (1997) Practical sustainability criteria for decision-making. International Journal of Sustainable Development and World Ecology 4(4), 231-244.

Smith, A. (1776) National Wealth. Smithers, H. A. & Walker, S. (1995) Experiences in updating an integrated decision-support system for

sustainable water resource management in northwest England. In: Modelling and Management of Sustainable Basin-Scale Water Resource Systems (eds S. P. Simonovic, Z. W. Kundzewicz, D . Rosbjerg & K. Takeuchi), IAHS Publ. no. 231 , 329-338.

Sugawara, M. (1970) Water resources and negentropy (in Japanese). Water Science 14(2), 1-20. Synghal, S. B. (1994) Smaller is better. Water Nepal 4, 40 -45 . Takahasi , Y. (1990) River Engineering (in Japanese). Tokyo Univ. Press, Tokyo. Takeuchi, K. (1997) Least marginal environmental impact rule for reservoir development. Hydrol. Sci. J.

42(4), 583-598 . Takeuchi, K. & Sivaarthitkul, V. (1995) Assessment of effectiveness of the use of inflow forecasts to

reservoir management. In: Modelling and Management of Sustainable Basin-Scale Water Resource Systems (eds S. P. Simonovic, Z . W. Kundzewicz, D . Rosbjerg & K. Takeuchi), IAHS Publ. no. 2 3 1 , 299-309 .

Thompson, M. (1994) Huge dams and tiny incomes. Water Nepal 4, 191-195. U N C E D (1992) Agenda 21: Programme of Action for Sustainable Development. United Nations Conference

on Environment and Development, Rio de Janeiro, Brazil, 3-14 June 1992. UNCSD (1995) Work programme on indicators of sustainable development. United Nations Department for

Policy Coordination and Sustainable Development, Division for Sustainable Development, unpublished manuscript.

62 Sustainability and reservoirs

Vorosmarty, C. J. , Meybeck, M . , Fekete, B. & Sharma, K. (1997) The potential of neo-Castorization on sediment transport by global network of rivers. In: Human Impact on Erosion and Sedimentation (eds D. E. Walling & J.-L. Probst), IAHS Publ. no. 245, 261-273 .

Van der Zel, D. W. (1997) Sustainable industrial afforestation in South Africa under water and other environmental pressure. In: Sustainability of Water Resources under Increasing Uncertainty (eds D. Rosbjerg, N . -E . Boutayeb, A. Gustard, Z. W. Kundzewicz & P. F. Rasmussen), IAHS Publ. no. 240, 217-225 .

Walling, D . E. & Probst, J . -L. , eds (1997) Human Impact on Erosion and Sedimentation. IAHS Publ. no. 245.

W C E D (1987) Our Common Future. World Commission on Environment and Development, Oxford University Press, UK.

Sustainable Reservoir Development and Management. Report by the IAHS/ICWRS Project Team. IAHS Publ. no. 251, 1998.

CHAPTER 3

COMPARATIVE ASSESSMENT OF RESERVOIRS WITH NON-RESERVOIR ALTERNATIVES

3.1 INTRODUCTION

Until recently, the objectives in planning and management have typically embraced some or all of the following issues: economic development, self-sufficiency, equity and environmental quality. Now, the purpose of sustainable development has emerged which may embrace elements of the above objectives. There are a number of purposes related to these objectives for which a storage reservoir can be built, such as fulfilling the needs of society regarding water supply, energy, transport and protection from floods. However, there are also other, non-reservoir, ways to attain these aims. Usually, there exist a spectrum of means to achieve a target, with differing values of quality criteria such as cost, time, reliability, by-products, etc. A multidisciplinary consideration is therefore needed where alternative options are identified, including non-structural solutions and several variants of structural measures. Each alternative may have its advantages and disadvantages. One has to evaluate all the alternative solutions, both structural and non-structural, weighting their pros and cons in an objective way.

The basic questions which come immediately to mind when studying alternatives follow the general systems approach where three elements are identified: objective to be reached, quality criteria and constraints:

• Is the objective of concern attainable by the means considered under the constraints assumed?

• What are the possible alternatives?

• What are the values of a vector of quality indices describing various alternatives in a comprehensive way?

The constraints in question may pertain to such aspects as permissible environmental impact, funding, social acceptance and political effects.

Examples of quality indices which could be used when comparing alternatives may relate to such matters as socio-economic and financial feasibility, related investment and operational costs; intervention in the natural regime, stress to ecosystems and humans, use of energy and raw materials, characteristics of waste and pollution problems and safety, risk and reliability issues. It is also necessary to examine the opportunities for reversibility

64 Comparative assessment of reservoirs with non-reservoir alternatives

(flexibility) and rehabilitation. Can the original "unengineered" state be reconstructed at all, and if so, then at what cost?

When comparing alternatives a comprehensive holistic perspective is necessary, where not only short-term benefits but also long-terms impacts and side effects are thoroughly evaluated. The time horizon of concern may extend to the design lifetime of a reservoir and beyond. Although, typically, it is not environmental objectives that play the principal role in taking decisions on creating a reservoir, these issues can be accommodated as either quality indices or constraints and are discussed in Section 2.1.3 . Conservation objectives need to be taken into account, such as conservation of nature and resources (both renewable, which should be utilized up to their sustainable yield, and non­renewable, where recycling should be considered, and transition to renewable resources). Protection of natural and cultural heritage needs to be secured. Limits to the load which the environment can absorb without adverse consequences should be determined. The viable alternatives should be revealed, made transparent to the public, subject to public discussion and, finally, the decision as to how to solve the problem should be accepted by society.

It would be instructive to elaborate on a number of case studies, where the criterion indices of existing alternatives may be evaluated in a straightforward way. One could build a matrix of different objectives and criteria, identifying less sustainable and more sustainable examples of reservoirs and comparing them with non-reservoir alternatives.

As an alternative to development projects to meet foreseen higher demands for energy or resources, such as water, it is necessary to consider also the possibilities of demand management. In the past, the focus has been on the supply side, developing new sources of water to meet a higher demand. Now, a change of this approach must be promoted, i.e. a focus on the demand side where demand management would provoke improved efficiency of water use.

3.2 MULTIPURPOSE RESERVOIRS VERSUS ALTERNATIVES

In the present study, some alternative means of achieving different purposes such as, for instance, water supply, energy generation and flood control will be compared. There exists, however, an intrinsic difficulty in comparing multi­purpose reservoirs that serve various, often conflicting, objectives and non-reservoir alternatives. Power generation is the only objective of a thermal power station, while a reservoir may have many purposes in addition to hydropower generation, such as flood control, water supply, navigation, recreation, water quality, wetlands, water habitat and scenic beauty. Meeting all the different multipurpose uses of a reservoir would require a number of separate non-reservoir projects for energy, water supply, flood protection and transport,

Comparative assessment of reservoirs with non-reservoir alternatives 65

such as, respectively, a thermal power plant, a water transfer scheme, river regulation or flood plain, and enhancing road, rail or air transport capacity. Therefore, a reservoir serving multiple purposes should be seen in a broad perspective, and compared with a combination of alternative projects.

Having too much or too little water has been an eternal problem. A storage reservoir can be helpful in curing both these maladies—water surplus and water deficit. Killing two (or more) birds with one stone is possible!

While reservoirs can be conceived with one or more purposes in mind, it is the multipurpose reservoirs that have become more common nowadays. It is not uncommon that a reservoir serves as many as six different purposes. It may happen that a reservoir designed for a particular purpose, serves other purposes in the exploitation phase (e.g. reservoirs originally designed for flood protection later serve primarily water supply, recreation or navigation).

There are a number of large reservoir projects, both existing and planned, which have a particularly bad reputation. A prior criticism may get softened if it is agreed that some target must be achieved and all alternatives that lead to reaching this target have significant disadvantages.

The Three Gorges scheme on the River Yangtze in China is one of the largest mega-projects of all time (Pan Jiazheng & Zhang Jinsheng, 1993a,b). Its purposes are: to control floods, to generate electricity and to improve navi­gation. At present, the flood danger is high, dike bursts are inevitable during large floods and massive losses cannot be avoided. It is foreseen that the project will improve the flood control from the present 10-year return period to a 100-year return period and will guarantee safety of the main dikes. The planned electricity generation renders the project the largest in the world (17 680 MW of installed capacity with 84 TWh of annual energy generation). A significant improvement to navigation on the River Yangtze which carries 78% of the country's river freight is expected. The present dangerous shoals and rapids would disappear. The annual shipping capacity could grow fivefold, saving considerable amounts of fuel and reducing the emission of exhaust gases. Water supply for south-north transfer would also be available. It is estimated that the project would pay for itself by the revenues from power sales alone in as little as 20 years. Mitigation and compensation measures are foreseen which will reduce the adverse effects. It would aim, for instance, to achieve sustainable sediment control by releasing muddy water through sediment release facilities.

The adverse impacts are very large. There is the need to resettle more than a million people. Farmland loss may occur and inundation of cultural relics and historical landmarks will occur and seismic hazards cannot be ruled out. The natural landscape will be considerably affected. Several rare/endemic species are in jeopardy.

Looking at the determination of Chinese authorities one notes that it is not easy to achieve sustainable development in a dynamically developing country subject to strong pressure on flood protection, energy extension, enhancing transport and water supply.

66 Comparative assessment of reservoirs with non-reservoir alternatives

At the other extreme, let us consider an environment-friendly mutation of multi-objective reservoirs of small size recently endorsed in Japanese cities where more and more infiltration and storage facilities have been introduced as part of comprehensive urban water and environmental management. This is basically the combination of artificial enhancement of infiltration of urban rainfall underground and of small scale storage of rainwater in buildings (cisterns) with the aim to serve flood control, low flow augmentation and water supply.

In large cities, a significant portion of rain falling on the urban area is immediately drained to the sea or downstream basin through culverts, conduits and rivers. It does not infiltrate nor contribute to the city water supply, while much water for municipal supply is being transferred from upstream basins. The water balance of an urban area becomes therefore quasi desert-like; having no rain input to subsurface water nor to the municipal water supply. As a result, in large cities, there are higher flow peaks during floods, drying-up during low flows and increased water supply dependence on upstream basins.

By promoting infiltration in residential areas, roads and public buildings by artificial infiltration facilities such as open hole conduits, sink ponds and wells, together with temporary retardation ponds, flood peaks can be lowered and the daily discharge in small urban rivers increased. By saving water in house gardens, basements of buildings, park ponds and the like, urban rainfall can be utilized as a water resource which reduces the urban water dependency on upstream rural regions.

In this way, old springs and small streams once dried up in Japanese urban areas are flowing again and a number of scattered small water storage facilities now serve as an emergency water reserve in crisis situations such as earthquakes and, at the same time, create popular water-oriented landscapes. This management is therefore sometimes described as an urban oasis plan and considered as an example of killing three birds with one stone, reducing the need of flood control storage, introducing a natural condition to urban streams and providing water resources to urban areas (Musiake et al., 1987).

The City of Winnipeg provides another example of extensive use of temporary retardation ponds as a flood control measure which is integrated with the recreational use of that space and is highly regarded for its aesthetic value with more expensive properties being developed around the pond.

3.2.1 Water demand and supply

Reservoirs are used to store water in order to maintain continuity of water supply to municipal, industrial and agricultural users. Water cannot be replaced by any other substance. Water is indispensable to sustain life and a prerequisite for virtually every form of human activity. Planning for the future, with increasing population (possibly doubling by 2050) and rising

Comparative assessment of reservoirs with non-reservoir alternatives 67

aspirations of people for a better quality of life, one typically assumes that the total water demand will grow. Yet, a fundamental question may be raised: do we need more water everywhere and do we have to develop new projects in order to meet the anticipated increased water demand? A viable alternative can be to formulate and to evaluate demand reduction methods and strategies as complements to, or substitutes for, supply augmentation projects when striving towards optimal, or satisfactory allocation of water resources among growing and competing or conflicting uses.

One may wish to re-think the need for growth in water supply as a result of the foreseen increase in municipal, industrial, and agricultural water demands. The negative results of pre-feasibility studies may lead to abandoning plans for the development of irrigated agriculture rather than to create a reservoir. Lack of sufficient water availability identified in a pre-feasibility study for a nuclear power plant project in Klempicz, Poland, was one of the reasons for abandoning plans for this investment.

In some references, e.g. Kindler & Russell (1984) one can distinguish between the term water demand and water requirement; the former being the willingness to purchase (varying with the price) whereas the latter does not depend on price. A companion notion is that of demand elasticity, understood as the percentage change in demand corresponding to a percentage change in price.

Water demands (WMO, 1994) fall into a number of categories such as agricultural, municipal, domestic, industrial, commercial, livestock, thermo­electric power, pollution abatement (e.g. low flow augmentation), navigation, flood control, recreation, aesthetics, tradition, fish and wild life conservation. Industrial demands consist mainly of water for processing, washing and cooling, with steel, chemical, paper, petroleum and food processing industries as the principal users. In the US, some 3% of industries are responsible for 95% of water withdrawals. Commercial demands include, for instance, hotels, restaurants, offices, car wash, laundries, and public buildings, such as schools, hospitals and prisons.

While considering water supply expansion, it is worthwhile analysing the possibility of demand management with the consequent improvement in the efficiency of water use. In the past, the concept of water demand management was not very popular in the world but now, as the gloomy perspective on water scarcity jeopardizes the development of vast areas, it becomes essential.

There are a number of tools and technologies of water demand manage­ment (WMO, 1994) falling into three broad categories: economic (water pricing), structural (altering existing structures to achieve better control by metering, recycling, leakage control, etc.), and socio-political (policy decisions influencing water conservation, standards and public awareness).

Improvements in the institutional, organizational and management spheres should be sought. They can be very effective as, for example, water manage­ment in river basins, rather than in administrative units. It is very unfortunate

68 Comparative assessment of reservoirs with non-reservoir alternatives

that typically water management is fragmented as multiple agencies deal with water. This may lead to ignoring interactions and externalities.

The idea of demand management is to "reduce per capita or per unit of activity use rates". It fits well with the concept of minimizing MIPS (material input per unit service), devised by the Wuppertal Institute for Climate, Environment and Energy in Germany. The MIPS, or ecological rucksack, embraces the total volume of water (and other raw materials) used to produce a unit product or a unit service, computed "from cradle to grave". For example, the total water use in soft (bituminous) coal production is 12 tonnes of pumped water per tonne of soft coal in open cast mining of lignite. In the case of paper production the water part of the ecological rucksack may reach a value of 90 to 1. This sort of consideration may be advanced further to distinguish between the consumptive use and the return flow (including the quality dimension). There is a considerable interest in Europe in the concept of Factor Ten, that is improvement of resource productivity by a factor of 10 in the next 30 to 50 years, which seems technologically feasible, yet would require costly investments.

An important avenue to achieve water savings is by water pricing. Historically, water had zero price in many of its uses (notion of "free commodity"). Pricing has been based on the costs of producing and distributing water (capital, operation costs, water gathering, purification and delivery), and not the cost of water itself. There are several ways to establish prices (Kindler & Russell, 1984), ranging from the interaction of supply and demand in an open market to an administrative decision. In fact, usually the approaches used are mixtures of both; even in market economies prices for irrigation water provided by public agencies may be nominal and unrelated to the full costs of supply and the value of water. Demand management may use price as a management tool by, for example, the volumetric pricing of urban supply and wastewater treatment on the basis of marginal supply and disposal cost, with rising excess rates, multiple season tariffs (dry and wet seasons) and temporary drought surcharges. It is a basic principle of economics that the price of a commodity controls, except in perfectly inelastic cases, the level of consumption. An increase in water price would curb consumption, which in turn may lead to the postponement of the need for expensive investments in water supply extension. However, it may not be easy to increase significantly the price of water in the short term; a gradual shift may be needed. Moreover, provision of water supply to the poor may require subsidies.

An example of the impact of a rise in the price of water on water demand has been observed in recent years in the eastern part of Germany where water consumption dropped significantly after the re-unification of Germany. Schumann (1995) noted that in the first three years after re-unification the domestic water consumption in the area of the former German Democratic Republic decreased by 27%, while industrial water consumption fell by as much as 54%, as a result of more realistic, cost-based water pricing and effluent

Comparative assessment of reservoirs with non-reservoir alternatives 69

charges. A more widespread use of the "consumer pays-polluter pays" prin­ciple augurs well for management of water, an increasingly scarce resource.

The Japanese experience of increasing the recycling rate of industrial water has been remarkable. It is an illustrative example of the importance of a governmental lead in demand management. In 1965, the average recycling rate in all the manufacturing factories with more than 30 workers was 36 .3% and 15 years later in 1980, the rate had doubled to 73.6%. During 15 years, the value of industrial products in monetary terms nearly tripled, implying that the freshwater requirement per unit value of industrial product was reduced to one sixth. The current recycling rate has stabilized at around 77% and the total amount of industrial water use has also stabilized.

Such a remarkable improvement of the recycling rate in the 1960s and 1970s was the outcome of various factors, but the most decisive was groundwater abstraction control by the government. In the 1960s, due to over-abstraction of groundwater for industrial use, groundwater levels dropped considerably in many cities and sea-water intrusion and land subsidence occurred. Under these circumstances, a government control of abstraction was found necessary. Meanwhile, industrial production was expanding rapidly and had no time to wait for public industrial water supply to be planned and built. Under a strong lead from the government, the factories themselves had to build their own recycling systems to acquire the necessary water. The effluent regulations introduced slightly later also accelerated recycling. Recycling appeared to be the most economical and practical solution to meet the increasing demand for industrial water.

Domestic water supply has also been subject to demand management practice under the municipal water supply agencies all over Japan. In 1961 the price of water supply in Tokyo was, together with it's sewage charge, 22 J P Ï m"3 for a 22 m 3 month"1 user and 26 JP Ï m"3 for all supplies above 22 m 3 month"1. In 1990, it was 195 JP Ï m"3 for a 22 m 3 month"1 user, 404 JP¥ m 3 for 200 m 3 month"1, 545 JPÏ m"3 for 1000 m 3 month 1 and 667 JP Ï Ï m"3 for a 100 000 m 3 month"1 user, despite the growth in the strength of the Japanese yen (Shoichi Fujita, personal communication).

However, Japan is still a country with increasing seasonal water supply problems (and high flood potential). The country is already virtually "stuffed with dams" (Hadfield, 1994). Despite successes in water demand control, still more recycling, re-use and conservation measures are needed in order to cope with droughts which now seem to happen more and more frequently. Municipal water use rose, on average, from 169 1 person 1 day"1 in 1965 to 338 1 person"1 day"1 in 1991. Groundwater reserves cannot offer a solution. They are over-utilized and land subsidence endangers the development.

It is important to note that simple methods of estimating future demands (straightforward extrapolation) may result in over-prediction of demands. There have been several failures to adequately predict future water demands, as for instance in Sweden in the mid-sixties, after new environmental

70 Comparative assessment of reservoirs with non-reservoir alternatives

legislation was introduced. The projections of water use proved to be largely incorrect and the schemes which were built were based on higher expectations of water demands. New legislation encouraged Swedish industrial users to introduce water recycling in order to lower the costs of complying with new legislation. Thus water withdrawals and effluent discharges dropped drastical­ly. Even more dramatic was the difference between future demand scenarios and the actual demand in countries-in-transition of central and eastern Europe after the collapse of the old socio-economic and political systems.

Improvement in efficiency of water use can also be achieved in industry by re-cycling, re-use, and water-saving technologies. There is still much room for improvement of water-efficient technologies. Cooling towers help reduce water consumption in the thermoelectric power industry. However, even if less water is required, much higher consumptive use comes about—and evaporation losses grow considerably. It should be an imperative to minimize the consumptive use in general. Potential for re-use should be increasingly exploited with due concern on quantity and quality aspects of return flow. Irrigation provides return flow of poor quality (residuals of agricultural chemistry and salt). Industrial or thermopower cooling systems have a significant re-use potential, as the quality problems may often be associated only with temperature rises. In-stream uses without withdrawals should be promoted provided there is no reduction in water quality.

Efficient technical means are needed to reduce household use and leakage in mains systems. Use of water meters may curb water wastage. Water meter­ing introduced locally in Poland at household level led to high water savings. If clients pay only for the metered volume of water they really use, water supply agencies become more interested in minimizing their conveyance losses which may be quite substantial. Considerable water savings can be achieved at house­hold level by water-economic constructions of flush toilets, washing machines and dishwashers. A more effective design of a flush toilet in 300 000 households in Mexico made it possible to expand the number of people served by 250 000 without a new source of water (Feder & LaMoigne, 1994). Recycling needs to be promoted and different forms of water re-use, such as for instance using lower grade water. "Grey" water recycling may offer significant cost savings for homes and industry. There is a potential for "grey" water use at household level. Domestic demands for water in a developed country consist of less than 50% of drinking water quality, where bath and shower exceed 30%. Potable water quality is not needed for many domestic and industrial applications where recycling and "grey" water systems (using, of course, colourless and odourless water) would be sufficient, e.g. more than 50% of domestic uses including 29% for toilet flushing and 15% for clothes washing. Domestic "grey" water may embrace effluents from wash basins, showers and baths, and even from washing machines (second and third rinsing water).

There is still a considerable potential for socio-political measures to enhance water savings. Appropriate legislation, institutional arrangements and

Comparative assessment of reservoirs with non-reservoir alternatives 71

a scheme of effective enforcement of law are necessary. Regulations may seasonally limit water use for certain purposes (lawn or garden watering, car washing). Measures which lead to the improvement of the fee collection rate are desirable, as they aid in paying off the capital investment. Inefficient water management can be considerably improved by privatization, where the efficiency of collection of fees rises and so does the level of service. Institutional arrangements in the spirit of decentralization and subsidiarity principle may improve efficiency via local and community control.

Promoting water saving habits has proved to be successful in some cases. The efforts undertaken in Arizona to replace water-thirsty grass lawns with native vegetation (cacti), well adapted to the local climatic conditions, can serve as a success story (Nathan Buras, personal communication).

However, it would be Utopian to expect that demand management alone will solve the water supply problem. Even if efficiency of water use increases, growth of water demand will occur in several countries due to population pressure, economic growth and higher aspiration of nations as far as living conditions are concerned. Growing pressure on water supply may occur even in wet areas. Moreover, although water conservation is the measure which should be vigorously promoted, it cannot solve the problem of smoothing natural variabilities which lead to non-uniform availability of water in nature. Water may not be available in the time and in the place when and where it is needed. If water demands grow, the system performance of water supply systems to agriculture, industry, municipal uses, measured by such criteria as reliability, vulnerability and resilience becomes worse.

Water needs grow continuously worldwide. Low-cost water supply options are being exploited in many areas and no potential for a substantial and inexpensive expansion may exist. The marginal cost of water from new sources is now getting considerably higher (e.g. due to lack of good quality sources and the need for higher treatment costs). Economic costs of new projects can be expected to double or even treble and the environmental costs of tapping new water will also grow. Developing new sources may require such more costly elements as, for instance, long distance water transfer, pumping of deep groundwater or water treatment e.g. from a highly polluted river.

An important source of water in many locations is groundwater which is typically less polluted than surface waters. Using groundwater, however, is not sustainable if the rate of groundwater withdrawal exceeds the rate of recharge. This is particularly true if water is mined i.e. where the recharge rate is much lower than the abstraction rate. Water transfer is another option. However, long distance transfers of large volumes of water, planned recently in the former Soviet Union, were found to be non-sustainable and subject to heavy criticism, based on the adverse effects of former projects of this type. Neither are the proposals for towing icebergs likely to be economic. Desalination of sea-water still requires substantial investment and high energy inputs and is, therefore, a viable large-scale alternative only in rich countries.

72 Comparative assessment of reservoirs with non-reservoir alternatives

If there is no possibility of diverting a significant volume of water from existing water bodies, while keeping them in dynamic equilibrium at a satisfactory level, storing surface water in a reservoir during high flows and releasing it during low flows is, in many cases, the only viable alternative.

Water demands grow rapidly in large cities (so also does flood damage at other times). As groundwater reserves are not a solution in many areas due to over-exploitation of aquifers, upstream reservoirs with minimum agricultural, industrial and municipal pollution are an attractive source of water supply for thirsty downstream cities.

There are ample possibilities worldwide to achieve considerable water savings and this is particularly important in the domain of irrigation, where water is often misallocated and the efficiency of its use is poor, as a result of underpricing. A unit of water used by industry brings far more profit than irrigation, yet the latter low value users are allowed to consume large quantities of water at subsidized prices. The "consumer pays" principle expressing the basis of cost recovery is not always relevant in the case of water and, as a result, cost recovery is very low as compared to other infrastructure sectors. In consequence, there is much water wastage and depletion of resources.

Much effort needs to be put into the development of water-efficient crops in agriculture, irrigation with treated waste water or with brackish water. Optimizing water allocation to plants (delivery where, when and how needed) and cultivation of plants with lower water requirements are challenging issues. Rehabilitation and upgrading of existing irrigation schemes are urgently required.

Notwithstanding the criticisms of low efficiency and adverse impacts of many irrigation schemes, one should stress that according to experts, feeding the growing human population without irrigation is not possible. There is simply no alternative.

3.2.2 Energy

The continuous rise of global energy use has taken place since humans learned to use fire. Economic development is accompanied by an increase in energy demands. Even now, as energy use efficiency improves in the world due to conservation measures, and the Total Primary Energy to Gross Domestic Product ratio goes down, the global energy requirement is rising due to the rise of population and living standards, and economic growth. Economic growth of 3 % per annum now causes slightly lower incremental energy demand (2.9% increase per annum). Per capita energy consumption in developed countries does not grow now—a decreasing tendency can be noted, for instance, in the UK and the USA, as higher efficiency is achieved. The global per capita energy use is now growing very little (from 2179 kWh in 1990 to 2188 kWh in 1992) and the global per capita use of fuel is slightly

Comparative assessment of reservoirs with non-reservoir alternatives 73

decreasing (from 2004 kg of black coal equivalent in 1990 to 1993 kg in 1992). However, developed countries passed through a period of very rapid growth of energy use some time ago. Industrialized countries have undergone a "jump" in the per capita energy use and it is now possible for them to stabilize and even to decrease their needs. Such growth is still ahead for many developing countries, and it is the increase of per capita energy use of developing countries that is responsible for the global rise. Moreover, population growth in these countries is rising, thus sharply increasing the total national energy needs. Per capita energy consumption in some less developed countries may be two orders of magnitude lower than in the developed ones. If a higher energy use is a necessary condition of growth, the question emerges: how to fulfil these higher energy demands?

Further improvement in efficiency of energy use is needed in countries in transition in central and eastern Europe (CEE), where despite the considerable progress achieved, the energy consumption per unit of GNP is still far higher than both in the developed and in the developing countries. There are a number of projects which are aimed at lowering the industrial energy consumption and of course improving efficiency of combustion (fluidized beds) and heating systems, e.g. district heating and better housing insulation. Subsidized energy prices do not help to reduce energy consumption and do not conform with the "consumer pays" principle.

Thus, apart from planning the extension of the energy potential for the future, vigorous actions aimed at demand management need to be taken. The possibilities of energy savings and elimination of wastage are far from being fully utilized. A lot more can be achieved at the global scale by such measures as more energy-efficient technologies, heating, lighting, and transport. Improvement in the insulation of buildings and a change in life style (e.g. reduction of air conditioning and the acceptance of a broader range of interior temperatures) help in energy savings. It is an imperative to examine whether all the existing measures (pricing at the marginal cost covering the capital and operation, use of all possible conservation and efficiency measures, elimination of subsidies which distort economic calculations and expansion and rehabilitation of existing sources) have been exploited, before going ahead with a new power extension project which typically incurs high costs.

Hydropower converts gravitational energy of falling water into electrical energy in turbine generators. Gross hydropower potential of the world amounts to 2300 GW (with over half of it in developing countries), with only 460 GW currently installed (87 GW in developing countries). The potential for hydropower expansion is limited by the availability of suitable sites and of capital. There is a whole spectrum of hydropower, from small plants of a few MW exploiting the energy of small streams and existing dams to gigantic projects with tens of thousands of MW of installed power.

Water power is one of the sources of energy, though typically not the major one, except for some countries, such as Norway, Canada, Brazil,

74 Comparative assessment of reservoirs with non-reservoir alternatives

Austria, Switzerland, Columbia, Portugal, and New Zealand. Much more energy worldwide is produced from fossil fuels (coal, oil, gas), of which 2 1 % are used for electricity production, with oil dominating where there is a high value of transportation.

Other alternative energy sources, such as solar, tide, wave and wind energies are ubiquitous and vast, but very diffuse, and therefore expensive to develop. Sources of renewable energy such as geothermal, heat pumps, energy forests, wood and wood waste, peat, straw and landfill gas play a relatively minor role in the global energy balance though they may be of vast importance locally. The energy sources for global electricity production are: fossil fuels (63.9%), hydro (20.5%), nuclear (15.3%) and geothermal (0.3%).

Pan Jiazheng & Zhang Jinsheng (1993b) note that China has exploitable hydropower resources of 378 GW which would enable annual power genera­tion of the order of 1923 TWh. The installed capacity of all power plants is 156 GW. In 1991 only 38 GW of energy came from hydropower: correspond­ing to some 10% of exploitable hydropower potential. Severe shortages in power supply have occurred and pressure for new sources is growing. There are plans for the production of 240 GW (80 GW hydropower) by the year 2000 and concepts or plans for many projects for the future already exist.

Water use in hydropower is non-consumptive and non-polluting. This sector of energy production is considered advantageous over coal energy development where water use occurs at all stages from mining to cooling and is consumptive and polluting, and where finite and limited non-renewable resources are being exploited. However, there are many cases in which the benefits envisaged by the planners of hydropower have not materialized. The potential for trans­forming the economy of the surrounding regions has often been overestimated.

Disadvantages of fossil fuel burning are the emission of carbon dioxide which augments the already rising concentration of greenhouse gases in the atmosphere, thus enhancing the global warming and gaseous pollution of the atmosphere with such noxious gases as S 0 2 and NO x . Moreover, burning fossil fuels means depleting non-renewable resources whose reserves are finite. According to recent assessments, the oil and carbon reserves will not last for many generations to come, unless considerable new deposits are found. The full range of adverse effects of coal energy should be considered, embracing also respiratory problems, damage to forest ("Waldsterben"), corrosion to buildings (including historical monuments), automobiles, etc. resulting from gaseous pollution, and also mining and coal power disasters.

Water is used in all stages of fossil energy production. Consider an example of coal energy production where water is used in coal mining, processing, slurry pipelines, liquefaction and cooling of the power plant. Regulations typically provide the maximum allowable increase of water temperature in a river. There are significant further problems with quality of runoff from coal mining (brackish water containing products such as aluminium and heavy metals).

Comparative assessment of reservoirs with non-reservoir alternatives 75

The disadvantages of coal do not occur in nuclear energy, where the available fuel could cover the demand for millennia with no significant production of greenhouse gases. Nuclear energy production has grown faster in the last decades than energy produced from other sources. Some declare it as sustainable energy. However, there exists a concerted public aversion world­wide, which has grown, particularly after the Chernobyl accident. Accidents in nuclear power stations which, albeit rare, cannot be excluded, may cause the release of gigantic volumes of radioactive material contaminating the environment (air, soil and water) over very large areas and for a very long time period, related to the decay time of radionuclides. The problem of nuclear waste is serious. Further, the issue of decommissioning a nuclear power plant with acceptable safety measures is very costly and difficult. For these reasons nuclear energy is now, and likely to remain so, being questioned by a significant part of society. Nuclear power plants impose stringent conditions related to water. Both low and high flows are of concern. There is a need for a highly reliable water supply for emergency cooling. Properties of low flows are therefore important in the design of cooling systems, while flooding may affect safety through failure of the control system.

Energy saving programmes are highly commendable. However, in strategic planning one has to consider, at some point in the future, the need for expansion of the energy potential. It is interesting that it is gas energy which is being seriously considered as the most welcome option for energy expansion rather than nuclear and hydropower in a number of European countries. Gas is a far better fuel than coal, oil shale or oil, as far as emissions of S 0 2 , NC^ and C 0 2 are concerned. Nuclear and hydropower, whilst not producing noxious gases such as S 0 2 and NO^ nor greenhouse gases such as C 0 2 , are still difficult to promote due to the strong opposition of environmentalists who emphasize the disadvantages of reservoirs. Relocation of many people, destruction of fertile land which may give shelter to rare or endemic species, destruction of connectivity of ecosystems and damage to the natural landscape are typical arguments against reservoirs. Fish populations and species composition are sensitive to water fluctuation, changes in river regime, depth, velocity, temperature, sediment load and bank characteristics. A dam is typically a barrier to fish. Reservoirs impose water quality problems downstream. There is slower flow, less aeration; lower entrapment of oxygen as compared to turbulent flow and thus less self-purification. Sediment or corrosives may endanger moving parts of turbines.

The provision of a reservoir for power generation should be compared with the other solutions, in a process of fair and comprehensive evaluation of environmental and social costs and impacts. In the case of coal power stations, one should also consider the finiteness of the fossil fuel. Gaseous emissions and their broad consequences (acid deposition and acid rain, public health effects, impact on biodiversity, greenhouse effect) should be considered along with the cost of long-distance transport, water use in both quantity and quality

76 Comparative assessment of reservoirs with non-reservoir alternatives

aspects and the considerable rate of accidents in the mining industry. On the other hand, elimination of coal as the principal energy source would cause severe social problems in mining areas. There is an urgent need for public participation in the process of decision making in order to achieve a satisfactory solution.

3.2.3 Flood control

Extreme floods have been a major calamity since the dawn of time. They jeopardized settlements which have traditionally been located near rivers. Despite progress in technology, the danger of floods has not been eliminated and average annual flood losses worldwide have soared to several billion US dollars. They have increased for several reasons. Because of anthropopressure there is a tendency to use additional land in flood plains, and to create the infrastructure there. Flood plains indeed attract development due to their flatness, high soil fertility, proximity to water and availability of construction materials. Moreover, adverse changes due to urbanization have been observed in many watersheds, such as the increase in impervious area, deforestation and regulation (e.g. shortening and straightening) of water courses. The net results are higher runoff coefficients, faster and higher peaks of the flood hydrograph (system response to intensive precipitation), and accelerated transport of water in the river. Human memory has a tendency to forget rare events and their consequences relatively quickly; people do not care about floods except for a short time after a calamity.

Constructing reservoirs where the excess water can be stored allows a more uniform time distribution of streamflow to be reached and thus is one of the remedies to alleviate the flood problem. Flood control reservoirs, and other structures (e.g. check dams, small structures storing flood runoff in upper portions of a basin), detain and store a portion of flood water, thus flattening the destructive flood peak.

There exist also a number of non-structural flood protection measures which allow one to accommodate floods, to live with them rather than trying, in vain, to eliminate them. Working with nature is more sustainable than working against nature. Adapting the "living with floods" attitude, one can try to modify (reduce) susceptibility to flooding and its impacts. This can be achieved by such means as appropriate zoning, building codes, flood proofing, regulation for flood hazard areas development leaving flood plains with low-value infrastructure, e.g. riparian forests which are subject to frequent flooding. Off-stream reservoirs (polders) consisting of a flood plain and a dike provide storage for excess water. Resources of flood plains are important in aesthetics and recreation and contribute to the restoration and enhancement of

Comparative assessment of réservoirs with non-reservoir alternatives 77

wetlands. Permanent evacuation of existing infrastructures in flood-prone areas may be necessary. Not only protection of wild rivers but even programmes of re-naturalization of regulated rivers (removing dikes in a portion of a river) are of interest in some developed countries. Other non­structural measures include flood insurance, that is a division of risks and losses among a higher number of people. Effective flood protection systems, consisting of forecasting, warning, dissemination, evacuation, relief and post-flood recovery, can also substantially reduce losses. Improving information and education on floods is also necessary.

An important flood protection measure is the so-called source control; that is watershed management including land use and soil conservation to minimize surface runoff, erosion and sediment transport. Construction or active surface mining may cause sediment runoff 2000 times higher than that of a forest without timber logging, and cropland 200 times higher than natural forest. Source control, realized by such means as trapping water in the catchment, enhancing retention, flood-plain management, retarding ponds, and other runoff-reducing means counteract the effects of urbanization (growth of flood peak, drop in time-to-peak of a hydrograph, drop in roughness coefficient and in storage potential) and of channelization (faster flood conveyance through shortened and straightened rivers). In order to decrease surface runoff and to curb erosion, such measures as for instance afforestation, arranging shelterbelts and riverine forests can be useful.

Flood control reservoirs play an important role in flood mitigation. They can retain water during an event of abundant flows. Flash floods require having storage available at any time, while seasonal floods fill storage made available during certain periods of the year. Reservoir operating policy decides upon the volume of discharge, based on the season, volume of water stored and measur­ed or forecast inflow. However, discharging water following a forecast of a very high inflow, and thus emptying a necessary storage for accommodation of the forthcoming flood wave, may be dangerous if for some reason the forecast flood does not materialize. It may cause a manmade flood and the wasting of water which will be badly needed later in a time of low inflows.

3.2.4 Other purposes

Damming a river, which results in more uniform flow conditions, may serve to improve navigation, through a navigable stage in the reservoir and/or downstream. Water demand for navigation refers to a welcome in-stream non-consumptive use. It may be somewhat polluting, so the need for control comes about. Duration of navigable periods in dammed rivers may grow in comparison to non-regulated rivers. Low flow problems endangering minimum navigable depths are less critical, and flood problems, which may disable navigation if the prescribed clearance (bridges) cannot be achieved, are

78 Comparative assessment of reservoirs with non-reservoir alternatives

also reduced. However, as flow in a dammed river is slower the duration of the ice cover is longer and the problems of ice jams may occur in cold climate areas. The effect of strengthened wind can also be considerable. In comparison to road, rail and plane transport, water transport is the cheapest option, the least polluting and does not require very expensive and ecologically-unfriendly infrastructure (motorways, railways).

Reservoirs change the landscape. They may adversely affect the original natural scenic beauty. This is often a strong argument against large projects nowadays. However, reservoirs may also offer a new, welcome, landscape dimension. Constructing a reservoir may create new wetlands, water habitat for flora and fauna, including pisciculture, facilities for recreation, and a new landscape dimension which is aesthetically pleasing. It can also serve to improve water quality (low flow augmentation, dilution of pollution, assuring biologically inviolable flow, salinity control, flushing accidental contamination, recharge of groundwater). A specific flow regime which can be controlled due to a reservoir may be desirable because of the requirements of biota, scenic and recreational purposes, historical interests or religious activities and cultural events.

3.3 FINAL REMARKS

Sustainability considerations require a comprehensive assessment of costs, benefits and impacts, including side effects, delayed and cumulative impacts for all alternative feasible options (both "reservoir" and "non-reservoir" ones). The internalization of costs of damage to the environment, remediation and safety is necessary.

Reservoirs are built to serve important uses. However, they are not free of adverse effects which sometimes can be very critical. It is essential to foresee and assess the adverse effects in each particular project and to account for them in the decision process. Perhaps the most significant common impacts are the direct effects of an impoundment, such as; following the loss of land due to inundation, the involuntary relocation of people, the loss of cultural property and historical heritage and the disturbance to the ecosystems.

Traditionally, development planning used to follow a "top-down" scheme as objectives and activities were often defined by experts of central authorities. Construction of several large reservoirs has been important for political reasons. They serve as a demonstration of power (the monument-building syndrome) aimed at enhancing safety and robustness, capability of accommodating vagaries in the natural variability of weather and wetness conditions, and liberating humans from such calamities as floods and droughts. They follow the spirit of the philosophy—"Man conquers Nature", "Nature obeys Man". Bandyopadhyay & Gyawali (1994) quoted a prominent

Comparative assessment of reservoirs with non-reservoir alternatives 79

politician from Nepal who had declared publicly at a mass meeting that his country "will be like Singapore if we could implement Pancheswor and Karnali hydroelectric projects".

A statement which has been attributed to great leaders of nations, Napoleon and Churchill, is that every single drop of water flowing in the Nile should be retained and used rather than being wasted in the Mediterranean. This attitude proves that these leaders appreciated the high value of water but their judgements were a one-sided oversimplification of a complex situation.

Recently, the structure of water resources planning has been re-thought to enhance the participatory approach. Action programmes may be initiated at the local level. Local participation in the decision-making process in matters of siiS . .uiable development and consensus building are essential. If a need for a water supply extension is envisaged, it is necessary to agree upon the approach to be chosen; whether a reservoir, water transfer, or just increased water saving measures, such as for instance pricing or water re-use enhancement. Those affected by the project need to be involved in the process of planning and decision making for new reservoir construction and operation of both existing and new reservoirs. They should be adequately informed, their opinion canvassed and by taking part in the decision-making process they should accept the solution as their own and feel a kind of responsibility. A broad consultative process and persuading a large part of the society to support the concept may ease its implementation. Subsidiarity principles and a participatory approach help to internalize measures and to promote efficiency and savings thanks to community control. In order for the participatory approach to function properly, more information is needed for the general public about dams and reservoirs, their advantages and disadvantages and also the pros and cons of other, non-reservoir alternatives.

Decentralization and an increase in the participation of stakeholders is very welcome (local authorities, private ownership, cost sharing). However, participatory approach and consensus building are terms easier said than done. Even in negotiation between government agencies the need for a difficult compromise solution, reflecting different priorities, may not come about.

Due to the complexity of impacts, siting reservoir projects is continuing to raise much concern. In fact, two types of sites are relatively easily acceptable. The first type are canyons in which reservoirs do not rise above the top and streams with waterfalls that fish never ascend, i.e. where problems concerning migratory fish do not exist. The second type are in-stream, low head axial turbines projects, where no flooding is involved. Such a site has a relatively low volume of biomass, whose decay neither contributes significantly to greenhouse gases nor impairs fish and water quality.

Large dams are in practice irreversible. There is a need therefore for flexibility (prudent stepwise design policy). Collingridge (1980) distinguished four technical indicators of inflexibility: capital intensity, large scale, long lead time, and major infrastructure needs early on.

80 Comparative assessment of reservoirs with non-reservoir alternatives

In order to enhance flexibility one strives towards consideration of alterna­tive designs of reservoirs. Comparing alternative designs: different locations, different construction options, different sizes, one may arrive at the most acceptable and sustainable solution. Specialists studying the possible climate change impacts recommend that in order to cope with overwhelming uncertainties, flexibility of the project should be enhanced. A dam should not be constructed now in anticipation of being needed in several decades. However, if it is being built now, it may be useful to "design-in" the possibility of further augmentation, should the need arise in the future.

Synghal (1994) compared two alternative plans proposed independently by different agencies for the same region in South Asia. The first plan with one large reservoir was compared to a more decentralized plan with 34 reservoirs. The latter option, while better in terms of higher flood storage, surface water area for recreation, and lower area of inundated bottom, costs only one third of the centralized option. The only advantage of the large reservoir over the set of smaller units was in energy production.

REFERENCES

Bandyopadhyay, J. & Gyawali, D . (1994) Ecological and political aspects of Himalayan water resource management. Water Nepal 4, 7 -24 .

Collingridge, D . (1980) The Social Control of Technology. Open University Press, Milton Keynes, UK. Feder, G. & LaMoigne, G. (1994) Managing water in a sustainable manner. Finance & Development, June

1994, 24 -27 . Hadfield, P. (1994) The revenge of the Rain Gods. New Scientist, 20 August 1994, 14-15. Kindler, J. & Russell, C. S., eds (1984) Modelling Water Demands. Academic Press Inc. Loucks, D . P. (1994) Sustainability implications for water resources planning and management. Natural

Resources Forum 18, 263-274. Musiake K., Ishizaki, K., Yoshino, F. & Yamaguchi, T. (1987) Conservation and Revitalization of Water

Environment (in Japanese). Sankaido, Tokyo. Pan Jiazheng & Zhang Jinsheng (1993a) The Three Gorges Project goes ahead in China. Water Power and

Dam Construction 45(2), 14-16. Pan Jiazheng & Zhang Jinsheng (1993b) Hydropower development in China. Water Power and Dam

Construction 45(2), 12-13. Schumann, A. H. (1995) Flexibility and adjustability of reservoir operation as an aid for sustainable water

management. In: Modelling and Management of Sustainable Basin-Scale Water Resource Systems (eds S. P . Simonovic, Z . W. Kundzewicz, D. Rosbjerg & K. Takeuchi), IAHS Publ. no. 2 3 1 , 291-297 .

Synghal, S. B. (1994) Smaller is better. Water Nepal 4, 40 - 45 . Williams, P. B. (1994) Learning from the mistakes of large-scale water development. Water Nepal 4, 36 -39 . W M O (1994) Guide to Hydrological Practices, 5th edition. W M O Publ. no. 168.

Sustainable Reservoir Development and Management. Report by the IAHS/ICWRS Project Team. IAHS Publ. no.251, 1998. g j

CHAPTER 4

DESIGN AND MANAGEMENT OF RESERVOIRS

4.1 INTRODUCTION

4.1.1 Structure of the chapter

The design and construction of manmade reservoirs has some important adverse environmental and social impacts including loss of useful land, relocation of population, damage to historical values, disturbance of ecological systems, reduction in river self-purification capacity, sedimentation and increased risk of water-borne diseases. Can a manmade reservoir be sustainable? This chapter concentrates on reservoir design and management issues as they relate to the notion of sustainability. The objective is not to provide ready-made solutions for sustainable reservoir design and management, but to present some of the ideas that are developed to incorporate principles of sustainability in reservoir analysis.

After some definitions and an overview of reservoir analysis problems, reservoir purposes and tools for reservoir analyses, the chapter is organized into three sections. The importance of the hydrological input for reservoir design and management is addressed in the first section. Some insights into time series analysis and the use of hydrological forecasts in improving reservoir performance are discussed in the second part of this section. Four contributions are provided in the second section in order to illustrate possible ways of addressing the requirements of sustainability in reservoir analysis. In the first contribution a new method for designing sustainable reservoirs, based on the De Novo programming approach, is presented as an idea for replacing the optimization of a given reservoir with the design of an optimal reservoir. Reservoir storage reallocation and reassessment of reservoir operational rules are considered to be the two main problems related to existing reservoirs to be addressed within the framework of sustainability. The second contribution provides a method for reassessment of existing reservoirs, which is based on the combined use of simulation and optimization. The main objective of the approach is to determine: (a) the active reservoir storage requirements based on the current demand; and (b) the best management strategy for the reservoir under consideration. The third contribution presents the cost and benefit allocation methodology for redevelopment of reservoirs. The methodology has been illustrated with two case studies: (a) the addition of new water users to an existing water supply system, and (b) the addition of new water users to a

82 Design and management of reservoirs

system with additional reservoirs. The fourth contribution provides an innovative procedure developed to guide and lead the reservoir planning process based on the least marginal environmental impact (LMEI) rule. The LMEI rule is used to decide the size of reservoir only to the extent that the extra unit increase of dam height has the least negative environmental impact solution among all the economically feasible alternative means (reservoir or non-reservoir). The chapter ends with a conclusions section.

Methods presented here are not necessarily those recommended but rather serve as examples that attempt to relate reservoir planning and management with the principles of sustainable development. Other innovative ideas are being explored and will appear in the near future. One area of particular potential includes methods based on the use of fuzzy sets (e.g. Kojiri & Sakakima, 1993; Kojiri, 1995; Bender & Simonovic, 1996; Despic & Simonovic, 1997).

4.1.2 Dimensions of reservoir analyses

Large reservoirs are used to help development and management of natural resources so as to sustain the economic growth of a region and improve the overall welfare of its citizens. Reservoirs contribute to increased crop production, more efficient water use, and energy availability. However, the growth of population has resulted in increases in demand and use of water resources. Reservoirs are the most important elements of complex water resource systems used to respond to increases in demand (Simonovic, 1992).

Reservoirs are used for spatial and temporal redistribution of water quantity and quality, and changing the water's ability to generate hydropower. The main characteristic of reservoirs is multipurpose use. By building a dam, the storage created may be used for flood control, water supply (municipal, industrial and agricultural), low flow augmentation, hydroelectric power generation, navigation and recreation.

There is no single type of reservoir analysis problem but a multitude of decision problems ranging from determining optimum reservoir storage capacity to selecting optimal reservoir operating policies. Different problem structures and conflicting reservoir purposes require a complex mathematical description. The decision variables, objective functions and constraints vary for different types of reservoir problems. Their correct formulation is required to address the trade-off between different water use purposes (hydroelectric power generation, municipal and industrial water supply, irrigation, etc.) and recreation or flood control.

Reservoir problems contain inherent uncertainty. The random nature of reservoir inflows and other related hydrological variables is the next important characteristic of reservoir analysis. The concept of reservoir reliability is

Design and management of reservoirs 83

probably one of the most important aspects involved in making meaningful decisions regarding reservoir active storage and release policies.

Reservoir analysis problem In its simplest form the reservoir analysis problem can be stated as: How large does the reservoir storage need to be to provide for a given demand with an acceptable level of reliability? This problem is also known as the determination of reservoir storage capacity. This is one of the main activities within the reservoir design procedure.

Other variations of this problem are possible, such as determining a reservoir release for a given capacity. If the time interval is longer the problem is called long-term operations planning. This problem is solved during the project planning phase.

The same problem may be addressed within a considerably shorter time frame when the project operation is the main concern. In that case the problem is called real-time reservoir operations.

In all three cases the basic problem remains unaltered. The relationship between inflow characteristics, reservoir storage capacity, reservoir release and reliability must be found.

Reservoir purposes Generally, in large multipurpose projects, the total reservoir storage is divided into three principal segments: (a) the flood control storage; (b) the active storage; and (c) the dead storage. The flood control storage capacity is used to reduce flood peaks and minimize potential down­stream damage. The active storage is used to regulate streamflow and provide water supply for various purposes. The active storage capacity is the principal source of water for municipal, industrial and irrigation water supply. It also provides water for hydroelectric power generation, low flow augmentation, navigation and recreation. The dead storage capacity is used for sediment control, and, in some cases to increase the elevation of the active storage.

Tools for sustainable reservoir analysis The specific character of sustainable reservoir analysis calls for new tools. Occasionally they are only new in that they are actually being put into practice (optimization of the operation of multipurpose reservoir systems). In other cases they are new in that they implement techniques that are well established in the literature but have not yet been fully utilized (expert systems for monitoring, design, and operation of reservoir systems). Often they are new in that they are modified to address new problems arising from the application of sustainability principles to water management (reallocation of reservoir storage and reassessment of reservoir operating strategies). In still other situations they are new in that they introduce new ideas and implement new technologies for solving water management problems.

Meeting the growing diversifying demands of sustainable reservoir design and management requires use of very sophisticated, and yet flexible,

84 Design and management of reservoirs

techniques capable of providing answers to many questions within the context identified earlier in this text. Four main categories of tools foreseen to be more actively used are:

• information systems (spatial and temporal data processing including statistics, database management tools and geographic information systems);

• systems analysis (simulation, optimization and multi-objective analysis);

• artificial intelligence (expert systems, neural networks, object oriented programming, fuzzy analysis, etc.); and

• technological tools (computer graphics, sound, animation, etc.).

All the tools to be used in sustainable reservoir analysis will be incorporated into the decision support framework (Simonovic, 1996).

4.2 HYDROLOGICAL INPUT FOR RESERVOIR DESIGN AND MANAGEMENT

Hydrological information is the basis of reservoir design and management. This information includes time series of reservoir inflows, extreme discharges and hydrological forecasts. The measurement methods for collecting hydrometeorological data are rapidly progressing. Technology offered by radar, satellites, transmission networks is improving on a daily basis. So are the data processing and forecasting techniques. The general opinion is, however, that reservoir management is not taking advantage of the advanced technology for data collection and processing.

The following two sections of this chapter will discuss some important issues in time series analyses and hydrological forecasting for sustainable reservoir management.

4.2.1 Time series analysis of reservoir inflows

The primary hypothesis commonly used in reservoir analyses is that the historical records of river runoff provide sufficient information to predict the most important characteristics of the reservoir inflow regime in the future. Interpretation of "the most important characteristics" and "the future" depends on the purposes of runoff control, the choice of time interval and storage requirements. In one case, "the most important characteristics" may be the mean and the variance of annual runoff series only, but in another case, they may include daily runoff regime characteristics over a long period.

If the reservoir storage is small compared to the annual inflow volume, the relatively short representative inflow series can be used for some reservoir

Design and management of reservoirs 85

studies. For example, a 26-year historical runoff record has been utilized for planning the Volga and Kama hydropower reservoirs in Russia; planning of the Angara hydropower reservoir was based on a 29-year runoff record. However, at present, most water management specialists consider the available historical runoff series (50-100 years) as being too short for capturing all possible combinations of runoff generation events and insufficient for the analysis of the behaviour of water resource systems. It is also important to stress that the direct utilization of the historical runoff records for reservoir planning does not account for the stochastic nature of runoff and other hydrometeorological variables affecting inflow and the state of a reservoir. As a result, the present practice of water resource systems analysis is based on using synthetic inflow series. They are usually generated with the help of mathematical models of runoff time series. Parameters of mathematical models are estimated using the historical runoff records, and the Monte-Carlo simulation procedure. The generation of random hydrological series, for as long a period of time as desired, enables the water resource system planner:

» to analyse the system response to possible combinations of runoff sequences;

• to choose the system configuration that can produce the desired output; and

• to estimate how much risk and uncertainty is associated with the selected configuration.

The Monte-Carlo method is well known and applied in many fields of science. However, the efficiency and the reliability of this approach for simulating the runoff series can be guaranteed only under the following assumptions:

• the mathematical model yields satisfactory simulations of the historic data; and

• the mathematical model can predict with the necessary accuracy the required characteristics of runoff in the future.

In practical applications, these requirements are not often taken into consideration. However, they may have a profound impact on the applicability of the Monte-Carlo approach for the generation of synthetic runoff series. Two main conditions for providing satisfactory predictions using the Monte-Carlo approach are:

« the model's adequacy for describing the natural process; and

• enough data for estimating model parameters.

The more sophisticated the model is, the more difficult its validation is, and the more data are required for the estimation of its parameters. That is why it is necessary to be very careful in choosing the structure of the model.

86 Design and management of reservoirs

Most of the mathematical models used for runoff series generation are based on the assumption that the runoff is a stationary random process. Such an assumption can be true only if temporal change of the climate and watershed characteristics do not lead to significant alterations of runoff.

The time step is of crucial significance in choosing the structure of a stochastic model for runoff sequences generation. The smaller the time step, the stronger the correlation between successive events. In this case, longer historical records must be used for estimation of the model parameters.

The simpler models are usually adequate to describe annual runoff fluctuations. The autocorrelation in annual runoff of many rivers is very small and synthetic data can be generated by using the statistical distribution of annual runoff. The three-parameter normal and gamma distributions are commonly used for the generation of annual runoff series. For most rivers, the correlation coefficient between the annual runoff in adjacent years is in the range 0.2-0.4. In this case, the annual runoff fluctuations can be described by a simple autoregressive model (a first-order Markov process) which can be introduced in the form of the following recursive equation:

x(t) = rx{t - 1) + (1 - r)x + a x (1 - r 2 )e (4.1)

where x(t) is the annual runoff for year t; x(t - 1) is the annual runoff for the preceding year (t - 1), x is the mean annual flow for the historical record interval; <JX is the standard deviation of the historical runoff records; r is the first-order correlation coefficient; s is an independent, often normally distributed, variate with zero mean and unit variance. A special procedure is required to convert a normally distributed variable into a, e.g. lognormal or gamma distribution.

If there is a tendency for annual runoff values to cluster (distinct periods of high or low runoff), the simple autoregressive model can fail to capture these variations. More complex models including several parameters are then recommended. They can be of the autoregressive type with serial correlation of several orders or, in more general case, they can be from the family of ARMA models (Box & Jenkins, 1976):

x ( 0 = i<j),x(? - 0 -tQjCJa(t - y > ( 0 (4.2) j=l j=0

where a(t) are independent, normally distributed variables with zero mean and unit variance; <)),• and 0 ; are unknown coefficients. To construct the model of runoff series it is necessary in addition to use the mean value of x(f) and the standard deviation a a , to.choose the proper value for p and q, and then to determine the coefficients (j), and 9y. Historical series of annual runoff are seldom longer than 100 years. Therefore, the number of model parameters should not exceed 4 or 5.

Design and management of reservoirs 87

The use of a shorter time step (month, week, day) increases the length of the required historical runoff series. However, the dependence between runoff increases too, and the models have to include more parameters. The models can be further complicated when it is necessary to generate simultaneously several runoff series that may exhibit cross correlation. This will be quite common when the reservoir has several hydrological inputs with different water regimes.

The use of models with large numbers of parameters can significantly complicate the test of model adequacy and decrease the reliability of the generated series. In the literature, the test of adequacy of time series models is limited to the comparison of the main statistical characteristics (mean, variance, skew, coefficients of auto- and cross-correlation). In many cases, the number of characteristics used in the comparison is smaller than the number of model parameters. A small change in the historical sample, or data errors, can result in a considerable change in model structure and choice of parameters. Significant uncertainty is introduced into generated runoff time series due to a variation in the statistical parameters of the historical sample which is caused by the short length of historical series. The statistical similarity of the simulated and historical series can also be insufficient to achieve similarity in the marginal distributions of temporal sequences in a chosen season or month, as well as in the joint distributions for each pair of time intervals. The realistic simulation of the season variability often leads to unrealistic fluctuations in the yearly series. To obtain reasonable consistency of annual and seasonal statistical runoff characteristics, as well as the characteristics of runoff at different gauging stations, disaggregation procedures have been proposed (Valencia & Schaake, 1973; Tao & Delleur, 1976). The efficiency of this approach in the selection of an optimal model depends on the length of the historic series, the time step, and the serial correlation. However, in many cases, because of short historical records, an attempt to construct monthly or daily data can lead to pseudo series, which can only increase the risk and uncertainty in reservoir analyses.

A practical solution can be the use of the "fragment method" proposed by Svanidze (1980) for simulation of runoff series. In this method, the recorded daily series are divided according to the corresponding annual values and, as a result, "fragment" series are obtained. Then, the Monte-Carlo procedure is applied separately to the annual runoff values and the "fragment" series. The different combinations of the fragments and the annual runoff values enable us to generate a wide variety of daily (or monthly) runoff series. This will not provide all possible series, however it will eliminate unrealistic fluctuations caused by poorly estimated model parameters or inappropriate model structure.

An effective way of overcoming the difficulties in constructing synthetic runoff series (associated with autocorrelation of runoff values) is the use of stochastic models of precipitation series. The precipitation records are often

88 Design and management of reservoirs

longer, and usually exhibit considerably lower autocorrelation than runoff series. This leads to more stable and reliable precipitation models which can be applied for simulating synthetic series of the desired length. These series can be used as inputs into the rainfall-runoff models (often called dynamic-stochastic models). If the rainfall-runoff models are physically based, there is also an opportunity to take into account temporal change of watershed characteristics, and to substitute the runoff stationarity hypothesis with the precipitation stationarity hypothesis. The main shortcoming of this approach is the use of computed runoff instead of observed. The accuracy of the computation process depends on the quality of the physically-based model for runoff generation, and the quality of input data. The accuracy of physically-based models is usually about 15-20%. However, in most cases a more reliable calculation of statistical characteristics of runoff series is achieved than by using poorly constructed stochastic models of runoff series.

Eagleson (1972) was the first to suggest dynamic-stochastic models of runoff generation using the physically-based approach. Assuming that the duration and the average rate of effective precipitation are distributed exponentially, and the overland flow is described by the kinematic wave equations, Eagleson obtained analytical formulae for calculation of runoff. These relationships express the dependencies between the statistical characteristics of the maximum runoff and the statistical characteristics of the rainfall. They also take into consideration the hydraulic characteristics of a watershed which may include the impact of human activities on the watershed. The development of Eagleson's approach and its practical application is described by Leclerc & Schaake (1972), Wood & Harley (1975), Chan & Bras (1979) and Diaz-Granados et al. (1984) who included in the runoff generation model the description of infiltration.

A dynamic-stochastic model of rainfall and snowmelt runoff generation (including the description of snow melting and infiltration into frozen and unfrozen soil, overland and channel flow) has been suggested by Kuchment & Gelfan (1991). The Monte-Carlo procedure used in this work for simulation of the meteorological inputs, and numerical method for solving differential equations describing dynamic processes in the watershed, enable a significant extension to the range of conditions for which dynamic-stochastic models can be applied.

The absence of reliable stochastic models of series of meteorological inputs seems to be the main factor restricting the development of dynamic-stochastic models for runoff generation. The research in stochastic characteristics of precipitation and temperature is an important part of applied climatology and there is a comprehensive bibliography in this field. Many investigators use the same mathematical models for precipitation and temperature series as for runoff series (Markov process, ARIMA, etc.). In most cases, precipitation models contain the description of two random

Design and management of reservoirs 89

processes: occurrence of precipitation, and the variation of its volume (or the average rate). In a number of papers, precipitation models are based on describing precipitation as a cluster process where the occurrence of clusters and the structure of the clusters are considered separately (Kavvas & Delleur, 1981). However, the stochastic properties of rainfall series depend significantly on the climatic and physiogeographic characteristics of a given region and, in many cases, it is difficult to construct stable stochastic models of these series.

It is also necessary to stress that the application of Monte-Carlo procedures for dynamic-stochastic modelling, including complicated systems of differential equations, requires considerable computational power. A compromise has to be found between the reliability of estimation of statistical characteristics and the complexity of the model. The application of detailed physically-based models (Abbott et al., 1986; Kuchment et al., 1986) enables simulation of a wide range of possible variations in runoff factors and human activities in a watershed, for different possible combinations of meteorological inputs. Taking into account the complicated non-linear behaviour of hydrological systems, the simulation and analysis can provide a prediction of events which are not observed in the historical series of runoff. The estimation of the return period associated with these events is not always possible. However, in many cases, discovering the possibility of unusual events can be very important.

4.2.2 Use of inflow forecasts for efficient management of reservoirs

This section presents the experimental results of the quantitative assessment of the use of forecasts in reservoir operation and an investigation of forecasting accuracy (Takeuchi & Sivaarthitkul, 1995; Sivaarthitkul, 1996). Similar attempts have been reported in the literature. They include the value of knowledge on the true distribution by Klemes (1977), the effect of sampling error for determining reservoir size by Phatarfod (1977), the worth of forecasts in reservoir operation by Yeh et al. (1982), the use of forecasts for real-time operation of the High Aswan Dam by Bras et al. (1983) and Stedinger et al. (1984), and the value of forecasts in water supply reservoirs by Mishalani & Palmer (1988). The following research represents an extension of reported work for sustainable reservoir management. Special emphasis is made on the possible reduction of the necessary reservoir storage using forecasts of various accuracy and on the relation between the necessary level of forecast accuracy and reservoir characteristics (such as size of reservoir, hydro-climatic condition and reservoir operation objectives).

90 Design and management of reservoirs

Introduction

Sustainability requires that once reservoirs are built, they should be managed in a manner as efficient and environmentally friendly as possible. The basic assumption of the work presented here is that the use of precipitation or inflow forecasts is a powerful tool in achieving sustainability. If hydrometeorological forecasts are used in an effective way, the efficiency of reservoir use will increase, reducing the need for a new development.

Measurement techniques of hydrometeorological phenomena have progressed rapidly, and include the use of satellites, radar, many kinds of automated and telemetered observation facilities and data transmission and processing techniques. Technological advancement of hydrometeorology is also making more accurate flow forecasting possible. However, the management of large reservoirs is not yet benefiting fully from these technological advances. This may be partly because reservoir operation is an activity with a high degree of public scrutiny. Even minor errors can create unacceptable impacts on the user. Forecasting accuracy is therefore still incapable of practically addressing the problem. The other possible explanation may be that reservoir operators do not fully realize the benefits that may be obtained from an increased level of forecasting accuracy.

The study here intends to show the relationship between the level of forecasting accuracy and the expected benefits of using it. The following is an example of sensitivity analyses for the quantitative assessment of the benefits derived from the use of forecasts under various reservoir conditions and various levels of forecast accuracy.

A method of assessment

Outline of the methodology The basic strategy for assessing the value of hydrological forecasts is a "with-and-without comparison". The reference point is the reservoir system performance (expressed in expected losses) without forecasts. To get this performance, the system was operated by the policy derived from the steady-state solution of the Stochastic Dynamic Program (SDP). The monthly transition probabilities of inflows necessary for SDP calculation were estimated from the historical monthly inflow records. In the simulation stage, the current inflow was assumed as the historic mean monthly inflow. The performance with forecasts, on the other hand, was calculated for the same reservoir system using the inflow forecasts. Here the forecasts were used as the inputs to the Deterministic Dynamic Program (DDP) assuming that they are perfect over the period for which they are available. At the end of the forecasting period, where forecasts are no longer available, the reservoir performance (expressed as

Design and management of reservoirs 91

the expected losses) associated with the remaining storage is assumed to be the one associated with the steady-state solution of the SDP used in the without-forecast case. The release decision based on the DDP solution with the forecasts was implemented and the storage at the end of the current time period is updated with the actual inflow. At the beginning of the next time period, the same procedure is repeated with the new forecasts. This procedure is continued until the end of the simulation. The two performances are compared under different levels of forecast accuracy (R2) and different forecast lead times ( I ) .

Forecasting model and its accuracy The relation between forecasts and true realizations for different forecasting accuracy may be expressed in a number of ways. The method used is one of the simplest, which serves only to evaluate the value of forecasts in a qualitative way. A forecast lead time T made at time t, Q,(T), is considered as a sum of the true value at time t + T, Qt+T and a random error component e,(7), i.e.

for lead time one (T = 1):

0 ( 1 ) = G, + . + s,(D = & + 1 + a a ( l ) • e,+ l (4.3)

and for lead time greater than one (2<T<L),

Q,(T) = Qt+T+zt(T)

= Q,+T + PsCO c£l(T)/ae,(T- 1) [Q(T- 1) - Qt+T_{] (4.4)

+ a £ , ( r )Vl -p s ( r ) 2 - e V / .

where e,+T is an independently normally distributed random variable N(0,1); p s ( J ) is the serial correlation of errors zt(T) and s,(T - 1); and oEt(T) is the standard deviation of an error et(T).

The overall level of accuracy is expressed as:

Rf(T)2 = 1 - (oSl(T)/ot+T)2 = 1 - c,(T)2 (4.5)

where ct(T) — a & ( 7 ) / a , + r ; and where R2 is a measure of the forecasting accuracy; and the value c,(T) is the relative magnitude of the standard deviation of the forecasting error, cr e,(7), to that of the inflow itself, a , + r , thus indicating the magnitude of forecasting error. R2 = 1 (c = 0) means the perfect forecast and R2 = 0 (c = 1) means that the error variance is the same as the variance of the historical flow.

This model does not consider the correlation between the forecast made at time t and the revised forecast made later at time t + 1, but it is quite important in practice and is discussed by Takeuchi (1990).

92 Design and management of reservoirs

N

Y i i GULF OF THAILAND

Fig. 4.1 The Mae Klong River basin, Thailand.

Sensitivity analyses

Cases of analyses A Y-shaped stream having two parallel reservoirs in the upper branches, modelled from the Mae Klong River, Thailand, with the Khao Laem and Srinagarind reservoirs (Fig. 4.1) was analysed. The time interval of analysis was monthly and the storage capacities S considered were equal to 22%, 50% and 100% of their respective mean annual inflow (Qa). The objective was to minimize the quadratic form losses in water supply (WS), flood control (F) and hydropower generation (P). In addition to the hydro-climatic condition of the Mae Klong River in the tropical savannah zone, the Tone River, Japan, in the humid temperate zone was analysed for comparison, rescaling the mean to that of the Mae Klong. The inflows to two reservoirs, Ix

and I2, were assumed to have the lag one serial correlation rx(\) = r 2 ( l ) = 0.8 in the tropical savannah and 0.2 in the humid temperate zone. The lag zero cross correlation rn(0) and the skewness were assumed as observed. High and

Design and management of reservoirs 93

low cases of coefficients of variation of monthly flows (Cv = 0 .3 , 0.5) were also considered. The forecasting accuracies considered were:

p g = 0 and R2 constant; p s = 0.8 and R2 constant; p e = 0.8 and R2 decreasing with lead time T.

The lead times considered were T = 0 to 6 months.

Results of sensitivity analyses A 100-year simulation was conducted for each case. Monthly inflow series were synthesized for each hydro-climatic condition and used as the real inflow series (Qt) in the simulation analyses. Four sets of forecasts were generated for each set of accuracy parameters. Thus all the results of the sensitivity analysis were based on the average of four cases. The following are some observations on the results from the simulations:

• A small reservoir is more sensitive to forecasting errors than a large reservoir. The absolute gain obtained from good forecasts is higher for a small reservoir, but the relative gain (percentage reduction in losses) is larger for a large reservoir.

• Perfect forecasts (R2 = 1) reduce the losses a great deal. The good but not perfect forecasts decrease the expected losses up to a certain lead time but, in general, increase the losses with forecasts beyond that lead time. The longer lead time is useful for a large reservoir where over-year storage is available.

• With the lower water demand case (SW = 0.48<2J, the expected losses are smaller than with the higher demand case (SW = 0.95<2fl), especially in the case of a large reservoir. The relative gain with the use of forecasts is not significantly different.

• Where the coefficient of variation is higher (Cv = 0.5), the expected loss in absolute value is higher than in the case with the lower value (Cv = 0.3). The relative gain with the use of forecasts is not significantly different.

• In a humid temperate hydro-climate, the expected losses are smaller and a higher forecasting accuracy is necessary than in the tropical savannah.

• Where a requirement for hydropower generation is added, a higher accuracy of forecasts is required even in the case of a large reservoir. This is because the reservoirs tend to keep the water level high to get a high head for power generation, and accordingly the wrong predictions are likely to cause floods and wasteful spills.

Implication of the results on reservoir size Figure 4.2 shows the comparison of expected annual losses for different reservoir sizes, obtained from simulations using various levels of forecasting accuracy. The upper line is the

94 Design and management of reservoirs

WITH AND WITHOUT FORECASTS COMPARISON

co <

CO so O

1000

800

600 H

400

200

0.2 0.4 0.6 0.8 1.0

STORAGE CAPACITY, S Fig. 4.2 Magnitude of "virtual capacity expansion" in the case of tropical savannah climate (C„ = 0.5, WS = 0.96g„).

case without forecasts, the middle line is with a forecast of accuracy R2 = 0.64 and the lower line is the case with perfect forecasts. This diagram shows the value of forecasts expressed in terms of storage capacity. Suppose the current reservoir capacity is 83% of the mean annual flow and, without any forecast, the expected annual loss is 200 thousands units. With an inflow forecast of accuracy R2 = 0.64 the diagram shows that the same performance level can be obtained with a storage capacity of only 60% of Qa. This simply means that the introduction of accurate forecasts "increases" the reservoir capacity. We will name this "the virtual capacity expansion". If a new reservoir is being planned we can avoid the construction of a large reservoir and reduce potential environmental destruction. In the case of an existing reservoir, we could either gain more from the same reservoir or could reduce some developmental needs for new reservoirs. The improvement of hydrometeorological forecasting techniques is thus important in sustainable reservoir development and management.

Conclusions

• Reservoir size, climatic conditions and the purposes of the reservoir system are important factors in identifying the necessary accuracy of inflow forecasts. Reservoir size seems to be the main controlling factor.

Design and management of reservoirs 95

• Small reservoirs can gain more from accurate forecasts, but they are more vulnerable to forecasting errors. They need short-term forecasts with high accuracy.

• Large reservoirs are not so sensitive to short-term forecasting errors and can take advantage of good long-term forecasts. For large reservoirs with hydropower generation, a high accuracy of short-term forecasts is also required.

These findings may be used in establishing a national strategy for hydrological forecasting research. In those countries with a large number of small reservoirs, highly accurate short-term forecasts have to be obtained. The introduction of less accurate forecasts should be treated with caution. The best use of historical observations would be better than using less accurate hydrometeorological forecasts. In those countries with large reservoirs, long-term forecasting techniques should be developed, and even if the forecasting accuracy is low their implementation is recommended.

The improvement of hydrometeorological forecasting results in the "virtual capacity expansion" and higher reservoir sustainability by:

• reducing the size of new reservoirs and at the same time minimizing the environmental destruction;

• gaining greater benefits from existing reservoirs; and

• offsetting anticipated developmental needs for new reservoirs.

4.3 METHODOLOGICAL CONTRIBUTIONS TO THE DESIGN AND MANAGEMENT OF SUSTAINABLE RESERVOIRS

4.3.1 New method for the design of a sustainable reservoir

The efficient use of available resources will serve to develop reservoir design and operation strategies, which are also representative of sustainable resource management. System flexibility relative to the performance of a project must be explored in full so as to maximize system efficiency. This outlook will mean the treatment of some resource levels as "soft" (decision variables) instead of "hard" (constants), in the pursuit of allocating system boundaries in an optimal fashion. An innovative method for reservoir design has been derived from Milan Zeleny's De Novo programming (Zeleny, 1981, 1986), which emphasizes optimal system design in terms of its design characteristics as well as its performance. De Novo programming allows the decision-maker to establish an ideal system performance by defining the metaoptimum. In objective space, this point is identified by the simultaneous achievement of

96 Design and management of reservoirs

optimal criteria levels. By varying system flexibility, the feasible space defined by reservoir constraints is then modified so as to include the metaoptimum into a feasible set of decision alternatives. Hence, this new approach addresses the conflict between objectives, which has dominated the attention of most planners. It is shown that if sufficient system flexibility does not exist, the best compromise solution may be improved with respect to some or all system objectives. It is likely that physical constraints will prohibit the inclusion of the metaoptimum into the feasible set of alternatives; however, the potential of modifying or constructing the feasible region to improve the best compromise solution renders this approach beneficial.

Introduction

The systems approach to the design of water resources systems has become a powerful and common technique among water resources decision-makers (Yeh, 1984; Simonovic, 1992). The problem formulation process involves the identification of pertinent objectives, and the definition of the system constraints in terms of mathematical relationships. System constraints in general must be addressed in order to comply with the resource capacities of the system. The optimization of objectives is then carried out with the application of a suitable mathematical technique. In the classical systems approach to decision-making, the structure of the problem is determined by the identification of all resource levels, and assumed as given by the analyst responsible for the identification of all suitable decision alternatives worthy of further consideration by the decision-maker(s).

Construction of reservoirs for the purpose of municipal, industrial, recreational, irrigation, hydropower and other purposes may involve flooding large areas of land. Other impacts incorporate changes in the regional economy and social well being. In the past, intangible benefits and/or costs accrued by people directly or indirectly have not been incorporated into the planning stage of reservoir design. Through problems encountered after the implementation of reservoir projects and criticism of inefficiency, equity and fairness, researchers have proposed that new project selection criteria be fully implemented at the planning and design stage. Issues raised by sustainable development have therefore begun to evolve with the hope of developing an awareness of the effect engineering projects have on the environment and society. An approach to the optimization aspect of a proposed project which may result in a system performance that achieves or is moved closer to the metaoptimal system performance, is De Novo programming.

Milan Zeleny first introduced De Novo programming to the single optimization problem of production scheduling, and later extended the concept to a multi-objective framework. While designing the optimal system resource portfolio, De Novo programming strives to attain a metaoptimal system

Design and management of reservoirs 97

performance, where the metaoptimum is defined as a combination of desired or stipulated criteria levels. These may be attained through optimization of individual objectives such that a series of optimal criteria will produce a metaoptimum. The multi-objective system formulated by the De Novo approach will incorporate the metaoptimal performance via additional constraints while minimizing the cost of doing so. In this manner, De Novo programming is able to deal with multi-criteria decision-making frameworks.

With the goal of analysing the De Novo programming as a potential sustainable design approach in a water resources context, the following contribution investigates the application of De Novo programming for a sustainable reservoir design problem. Furthermore, the application explores the case of failure of the De Novo program to attain the metaoptimum due to insufficient system flexibility, and focuses on the improvement, if any, in the best compromise solution.

De Novo programming approach

De Novo programming involves a process of allowing resource capacities to become decision variables of the problem itself. In the single objective environment, the approach includes one additional constraint in the De Novo formulation which limits the total "budget" to be "spent" on the purchase of the unknown, or to be designed, resource portfolio (Bare & Mendoza, 1988, 1990). The resulting solution includes the value of the optimal objective function, as well as the values of the optimal capacities of system resources. This resource portfolio will produce an optimal objective function value while producing no slack or surplus variables in the "soft" constraints. Since the problem is essentially limited by the budget level, the higher the budget the higher the achievement level of an objective function. In this manner the De Novo formulation designs the optimal system as opposed to optimizing a given system, providing the decision-maker with the confidence of knowing that the optimal solution is the best that can be achieved.

Mathematically, the transformation of a linear single objective optimization problem into De Novo formulation can be stated as:

maximize Z — ^ (4.6)

subject to:

/ = 1, 2, m

x > 0 j = 1, 2, n

98 Design and management of reservoirs

where xy are decision variables; Cj are objective coefficients; atj are constraint matrix coefficients; bj are right-hand sides of constraints, assumed as given constants; m is the number of constraints; and n is the number of decision variables.

Suppose that k of the m constraints indicated by set M are "soft" while the remaining / (/ = m - k) constraints indicated by set L are "hard", then the De Novo model for the single objective case becomes:

n

maximize Z = ^JcJxj (4.7)

subject to: n

Y,auxj-bi i e L

.7=1

TV,-*, -b,<0 i e M JL-i 'J J '

Xj>0 y = 1,2, ...,n

k

Y^Pibj^B i e M / = !

where p-t is the per unit cost of a given resource bt to be designed; B is the total "budget" allowed to be "spent" on the purchase of the designed resource portfolio.

Zeleny also provides an interpretation of De Novo programming to suit a multi-objective framework. Given P objectives, the model can be written as:

maximize {Zu Z 2 , Zp, ZP} (4.8)

subject to

n n

/=! j=\

Xj>0 j = 1, 2, n

Individual optimization of each objective will yield individual optimal values (Z*) which are needed to complete the De Novo formulation. The

interpretation of the budget concept and the identification of k "soft" resources and / "hard" resources from the previous single objective solution enables the De Novo formulation to be significant in dealing with the multi-objective problem. The final De Novo formulation becomes:

Design and management of reservoirs 99

minimize fi = ^_ipjbl (4.9)

subject to:

n

< b, i e L

Ta,ixj b,<0 i e M

x > 0 j = 1, 2, n

Z, > Z * z2 >z In this manner the focus from finding an optimal (single objective) or best

compromise (multi-objective) solution is shifted to the design of a system that will provide the best performance given a certain budget, or a system which will perform at its ideal, while minimizing the cost of doing so.

De Novo programming approach for reservoir design— problem formulation Reservoir construction is a common and viable alternative in many regions of the world when a need for a reliable source of water exists. In the planning stage of design, a required active storage, or size of reservoir, has to be attained so as to ensure the satisfaction of the reservoir purposes as formulated by the planning authority. The determination of size facilitates further decisions concerning the physical requirements of reservoir location, and physical attributes of the dam itself. Since the number of alternative size solutions that will achieve the requirements is great, it is necessary to develop evaluation criteria and constraints, which will be used to discriminate among various alternatives. This process therefore finds the optimal size of required reservoir storage. The development of the reservoir as a "system" can be achieved through the generation of pertinent system objectives, and con­straints, governing the decision process. A typical system model requires the minimization of the construction cost which is directly related to the required active storage K of a reservoir, subject to three general sets of constraints.

In order to provide an evaluation of the De Novo programming approach in a water resources context, a hypothetical reservoir design problem is constructed. It is assumed that the reservoir design process must be evaluated on the basis of two criteria: the minimization of the required active storage, as well as the minimization of the sum of absolute deviations of calculated storage values from the target storage. The latter objective reflects the importance of the reservoir in providing benefits related to the regulation of the resulting lake level. For example, the reservoir may be required to satisfy multiple user demands, such as maintaining the lake level constant during the

100 Design and management of reservoirs

summer months. Doing so would benefit those using the lake for recreational or navigational purposes. The aesthetic appeal of the lake may also be linked to a high reservoir level. Other demands include flood protection for downstream locations, the maintenance of fish spawning conditions, or the increase in hydropower generation reliability. Without particular identification of any single reservoir purpose, it was assumed that deviations from a target storage profile were to be minimized.

Additional constraints on the system bind the water supply, storage and downstream release within each time period to upper and lower limits. Mathematically, the complete reservoir model can be summarized as follows:

minimize \K, 'f^S, - S]|| (4.10)

subject to:

ST = S,_i + /, - WS, -Rt- EVAP, - INF,

S,<K

S0 = ST

ç < ç < ç um'm — *-7 — "-"max

LB,„ < (DTARmy WS, < UBM

RMINM <R,< RMAXM

{t= 1, 2, 7} {m = 1,2, . . . , 1 2 }

where EVAP, and INF, are considered negligible; 5 m i n and S m a x are the lower and upper bounds on the reservoir storage within time period t respectively (assumed constant for all time periods); LB, and UB, are the lower and upper bounds on the fraction of water supply provided, relative to monthly demand target levels for that time period D T A R , , respectively (assumed constant for all time periods); RMIN, and RMAX, are the minimum and maximum allowable reservoir release for the satisfaction of downstream needs respectively (assumed constant for all time periods); t is the monthly time index; m is the monthly time index for variables which are allowed to vary only throughout 12 months, as opposed to over the entire planning horizon; and T is the length of the planning horizon in terms of months.

Numerical experiments—Shellmouth Reservoir The data used in the optimization of the above model for illustrative purposes, are mostly based on the characteristics of the Shellmouth Reservoir, better known as Lake of the Prairies, located near Russell, Manitoba, Canada. The Shellmouth Reservoir was constructed (mainly for flood control, water supply and recreation) in

Design and management of reservoirs 101

1971, and has been in full operation since 1972. Since this date, inflow records into the reservoir have been available on a daily basis. The 180-month-long inflow sequence chosen for the model, represents the first 15 years of the 18-year operating history of the Shellmouth. Minimum and maximum reservoir releases are based on those corresponding to the minimum release required for survival of downstream aquatic life, and maximum operational capabilities with two release gates open, respectively. Data used to represent the demand for water for municipal purposes at the location of, or downstream from, the reservoir are based on the projected demand for water from the Shellmouth for the year 2040. This number is based on the extrapolation of the demand for water licenses between the Shellmouth Reservoir location and the city of Winnipeg from the present to the year 2040. Last, the operational rule curve for the Shellmouth Reservoir is used as the set of target storages in the hypothetical case.

The reservoir sizing model is a system of two conflicting objectives subject to six sets of constraints. Optimization of either of the objectives will yield respective optimal values as well as a set of optimal values for each decision variable. Due to the inherent conflict between the objectives, optimization of each will produce different solutions for the decision variables. Optimization of the reservoir model while minimizing the required active storage, yields an optimal reservoir active storage of 320 1 0 6 m 3 , and an associated value of the second objective (sum of deviations) of 32 320 1 0 6 m 3 . Optimization of the reservoir model while minimizing the sum of deviations from target storage yields an optimal sum of deviations of 14 950 10 6 m 3 , and an associated value of the first objective (reservoir active storage) of 477 1 0 6 m 3 .

It is clear that an ideal solution, which will minimize each objective is not feasible. In this case traditional multi-criteria decision-making would call for the identification of the set of nondominated points, which contains all alternative trade-offs between the two objectives worthy of consideration. However, our method proposes the De Novo formulation. To achieve this task it is necessary to identify system flexibility by the selection of soft resources (new decision variables), and the ideal system performance to be achieved by the De Novo reservoir model. Also, it is necessary to establish an idea of the budget, which will form the new objective of the De Novo model as introduced in (4.9). The latter task will undoubtedly precipitate a discussion of per unit costs of all soft resources.

The De Novo formulation considered here comprises the following assumptions:

» all resources except for the storage targets and the lower bound on the water supply are allowed to become soft resources (to be designed by the De Novo formulation);

« the ideal performance or the metaoptimum is defined as the combination of the optimal objective values found by the individual optimization of each

102 Design and management of reservoirs

objective; and

• the per unit cost coefficients to be used in the expression of the budget are assumed to be unity for the purpose of clarity and simplicity.

The final reservoir model formulated using the De Novo approach becomes:

minimize B = { 5 m i n + 5 m a x UBm + RMINm + RMAXJ (4.11)

subject to:

S, = StA + I, - WSt - Rr

St<K

S0 = ST

ç < ç < s '-'min — u t — '-'max

LBm < (DTARJ1 WS, < UBm

RMINm <R,< RMAXm

K = 320 1 0 6 m 3

'YJS, -S*\ = 14 950 10 6 m 3

t=]

{t= 1 ,2, ...,T}{m= 1,2, 12}

All resources highlighted in bold introduce flexibility into the model. Minimization of the budget objective with respect to the De Novo model provides the optimal solution for the budget which will ensure the achievement of the metaoptimum (equality sign in last two constraints) via the design of an optimal resource portfolio, if such a solution exists. However, for our case study the solution of this model can be shown to yield unfeasibility indicating that there is not enough system flexibility to include the metaoptimum into our feasible solution space. It appears therefore that formulating the reservoir design problem using the De Novo approach was not very useful. However, if the metaoptimum cannot be reached there is still a possibility for improvement by moving the optimal solution closer to the metaoptimum. This can be achieved by determining the set of nondominated points (trade-offs between the two objectives) and then selecting the best compromise solution. In this work the weighting method has been used to obtain nondominated points.

Two sets of nondominated points are generated and compared graphically in Fig. 4 .3 . One set represents the solution using a conventional reservoir model, and other using the De Novo reservoir model. It is evident that the effect of the De Novo formulation serves to push out the frontier of the set of nondominated points toward the metaoptimum. Since each point represents a

Design and management of reservoirs 103

tu O

> Q

100 150 2 0 0 2 5 0 3 0 0 3 5 0 4 0 0

R e q u i r e d Ac t ive S t o r a g e (O F 1 ) 1 0 6 m 3

. G i v e n R e s e r v o i r S y s t e m De N o v o R e s e r v o i r S y s t e m !

Given M e t a o p t i m u m a D e N o v o M e t a o p t i m u m [

4 5 0 5 0 0

Fig. 4.3 Comparison of the given and De Novo nondominated sets of point, with respective metaoptima.

decision alternative of a respective model formulation and Fig. 4.3 shows the performance of each decision alternative with respect to both objectives, it can be concluded that all alternatives which exist in the De Novo set of nondominated points exhibit an improvement in one or both objectives.

Furthermore, it is now possible to analyse the exact improvement in the best compromise solution. A Compromise Programming approach (Zeleny, 1973; Simonovic, 1989) identifies the solution that provides the minimum weighted "distance" from the ideal, referred to here as the metaoptimum. Identification of one compromise solution on each of the nondominated sets can be accomplished by the minimization of the following distance metric with respect to the formulations as given (4.10) and the De Novo model (4.11). The distance metric written for a minimization problem is of the following general form:

/=i

where Z,- is the ith objective function value; Z* is the minimum value of the ith objective function; Z*' is the maximum value of the ith objective function; s is a constant used to express the decision-makers' concern with the maximal deviation, typically the problem is solved for three levels of s (s = 1, s = 2, s = 100) which yield the three best compromise solutions; a,-is a weight used to represent the decision-makers' relative importance of the ith objective function; p is the number of objectives (p = 2). The choice of a,- is subjective and is usually given in advance by the decision-makers.

z"-z* (4.12)

104 Design and management of reservoirs

Table 4.1 Compromise solutions for s = 1, 2, 100, and a = 1.0 for the given reservoir system, distance measured with respect to the given metaoptimum.

s Optimal OF1 Optimal OF2 (10 s m 3) (106 m 3)

1 403 20 202 2 392 21 486 100 457 15 949

Table 4.2 Compromise solutions for s = 1,2, 100, and a = 1.0 of the De Novo reservoir system, distance measured with respect to the De Novo metaoptimum.

s Optimal OF1 Optimal OF2 (10" m 3) (106 m3)

1 336 15 159 2 302 19 557 100 409 9 782

It should be noted that the ideal performance is measured as that combination of objective function values corresponding to the metaoptimum in Fig. 4 .3 . Analysis has been performed for values of s = 1, s = 2 and s = 100. Tables 4.1 and 4.2 summarize the results. Figure 4.4 shows the location of compromise solutions.

Comparison of the given resources and the De Novo designed soft resources sheds light on the required system design, which results in the improvement between the best performance of the compromise solutions.

Discussion of results

The improvement in the best compromise solution is the total value obtained by adding the reduction in each of the two objective functions. As a result all impacts associated with the two objectives will decrease. An improvement in the required active storage of the reservoir can only be achieved by decreasing the necessary size, while an improvement in the second objective is strictly dependent on the timing of cumulative reservoir releases. The reordering of all the soft resources is the key to achieving this complex task.

Had an optimal solution to the De Novo reservoir formulation been obtained, the analyst/decision-maker would have been forced to explicitly consider the per unit cost coefficients of the De Novo objective. However, the units of these cost coefficients have not been discussed so far. Zeleny's application of De Novo programming focuses on the minimization of a monetary budget value, which is used to obtain the soft resources. In this sense, one can view the cost coefficients as dollars per unit of a certain soft resource. In the water resources context this may not be directly applicable

Design and management of reservoirs 105

considering the difficulty in assigning costs to various operational, and/or institutional constraints. The interpretation of some right-hand sides of constraints as resources may not even be natural. Nevertheless, there is nothing stopping us from viewing the right-hand sides as decision variables, with the budget being interpreted as some governing objective (perhaps a penalty) which must be considered. It may be more appropriate therefore to change Zeleny's terminology: soft resources to soft constraints or additional decision variables, and budget to governing penalty or benefit function. In this manner De Novo programming applied to reservoir design measures the level of improvement in reservoir performance through system flexibility.

De Novo programming provides the potential for improvement in reservoir design. This is directly related to the amount of system flexibility that is intro­duced. How should one decide which right-hand-side terms should become decision variable? It is expected that certainly limitations on physical constraints cannot become flexible. Can economic, environmental, or social restrictions be manipulated to the extent that it will benefit the reservoir design? This should be answered on a case by case basis; however, if this approach is to succeed in the field of water resource development, it is evident that a new outlook on the decision-making process must take place. This is a recognition of the value to be gained from the De Novo reservoir design approach.

Conclusions

The proposed approach focuses on incorporating objectives and constraints representative of sustainability criteria and sustainability indicators into a

4 5 0 0 0 - - — — _ , - - _ - , ,

0 100 2 0 0 3 0 0 4 0 0 5 0 0

R e q u i r e d Act ive S t o r a g e (O F 1 ) 1 0 6 m 3

Given R e s e r v o i r S y s t e m D e N o v o R e s e r v o i r S y s t e m 1

A G iven M e t a o p t i m u m a D e N o v o M e t a o p t i m u m

x D e N o v o B e s t C o m p r o m i s e So lu t ions 0 Given B e s t C o m p r o m i s e So lu t ions L I

Fig. 4.4 Location of compromise solutions for given and De Novo reservoir systems.

106 Design and management of reservoirs

reservoir sizing problem. In recent years, sustainability of project development has become an important issue. Decision-makers are faced with more complex decision-making processes in which social, environmental, and economic concerns are playing roles of equal importance. Researchers feel that it is not only necessary to evaluate the performance of a project in terms of sustainability criteria, but that these criteria should also be reflected in the design of the project itself. This requires the analysis of existing relationships between system components and their impacts. If this information can be obtained and appropriate mathematical relationships derived, sustainability of a project can be measured and incorporated in the design of a system.

Present research on sustainable development is only beginning to harness the scope of the definition of sustainability. It is likely that concepts of physical, economic, social, and environmental capacity will be derived so that an evaluation of system performance will be possible. Applying De Novo programming to a model, which includes more flexible resource capacities, such as a sustainable reservoir sizing problem, would lead to an optimal system design, as well as a corresponding optimal system performance. An overview of the expected De Novo formulation is as follows:

minimize required active storage (and other)

subject to:

physical requirements < physical capacity

economic requirements < economic capacity

environmental concern < environmental carrying capacity

social concern < social capacity

and budget constraint

û*environmental capacity + £>*social capacity < economic capacity

where a and b are per unit cost coefficients.

4.3.2 Methodology for reassessment of existing reservoirs

In the future, in many regions of the world, instead of planning and designing new reservoirs, more emphasis will be placed on the management of existing reser­voirs. Reservoir storage reallocation and reassessment of reservoir operational rules are considered to be the two main problems to be addressed by reservoir research. The reservoir management process can be very dynamic (Israel & Lund, 1992), reflecting the role of reservoirs in a region's economic and social life.

Design and management of reservoirs 107

Introduction

Reservoir storage reallocation and reassessment of reservoir operational rules are considered to be the two main problems related to existing reservoirs to be addressed in the near future. Ford (1990) first, and then Johnson et al. (1990) provided excellent publications on reservoir storage reallocation. Reassessment of reservoir operations does not attract the same attention as the reservoir storage reallocation (Israel & Lund, 1992). One possible reason could be a lack of a proper methodology for the evaluation process. Historically, systems analysis (including simulation, optimization and multi-objective analysis) has been considered sufficient to provide adequate tools for addressing different reservoir problems. Many of the available systems analysis tools can be applied meaningfully to the reassessment of reservoir operations, even though they were not developed specifically for that particular purpose. However, the reassessment process has its own characteristics and therefore requires a development of appropriate tools.

Unfortunately, reservoirs have been constructed and managed to serve a static environment. Historically, the resolution of many water controversies arising from dynamic conditions has involved the construction of new facilities or modification of existing ones. Some of the rationales for reassessment of reservoir management strategies are external, and some are internal to the project. Three general groups of needs for reassessment are:

» physical;

• economic; and

• environmental.

Changing physical conditions The main need for reassessment of manage­ment strategies is the availability of additional hydrological data. The overall understanding of processes related to reservoir operations increases with additional hydrological data. Reservoir inflow is of primary importance. However, additional information on precipitation, evaporation, seepage, and land use are valuable for the reassessment process. Any change in physical conditions related to the dam itself, as well as associated structures will require modification of management strategies.

Changing economic conditions Reservoirs play a very important role in the economy of a region. Any change in the economy is usually reflected in a change of water use and vice versa. The changes may be an increase in demand for water, or a change in the demand distribution with time, or a change in the water quality requirements. Another aspect of changing economic conditions may be reflected in the change of reservoir purposes. New purposes may be added, some eliminated, and some may never develop

108 Design and management of reservoirs

to the originally planned level. All this information will play an important role in the quantification of reservoir management objectives.

Changing environmental conditions This group includes the needs rising from changes in societal priorities toward the environment. Being major structures reservoirs impose stress on the environment, and their management is closely scrutinized by the public. Management strategies have an impact on the water quality in a river downstream, and the water quality in the reservoir itself. Beside water quality issues, impacts on the environment may include changes in the habitat of fish and other wild life. One additional need for reassessment of reservoir management strategies may come from the impact which changing environmental conditions imposes on the people living in the close vicinity of reservoirs. The need for reassessment identified in relationship to changing environmental conditions may imply change in the objectives used to plan reservoir management.

Identified needs and their implications are already indicating that there is a considerable difference between planning of new and existing reservoirs. One major difference is that from the relatively unconstrained analysis of new projects, the analysis of existing projects is substantially constrained. Physical reservoir characteristics, like active storage, spillway capacity and outlet capacity, are often difficult to alter and this leaves very little flexibility for finding new reservoir management strategies. Other differences, as presented by Israel & Lund (1992) include:

• legal differences;

« differences in data;

• historical differences; and

• different public participation.

Major implications of the differences between the planning and the management of new and existing reservoirs, are in modelling and analysis. Appropriate modelling procedures are necessary to take into account the changing objectives, the availability of additional data, the effect of existing constraints, and must address the uncertainty of the future role of the reservoir. In many ways, an appropriate modelling procedure can be derived using the same expertise, models, and procedures as for any new project. Simulation, optimization and multi-objective analysis are capable of supporting the reassessment process.

A methodology

A methodology for reassessment of a multipurpose reservoir includes: (a) the database; and (b) the modelbase. The database component stores all the

Design and management of reservoirs 109

reservoir, environmental, social, economic and hydrological data. Besides providing storage for data, within this component an autoregressive moving average (ARMA) model has been incorporated to assist in generating synthetic sequences of inflow data. The modelbase component is based on the combined use of simulation and optimization.

The methodology proposed in this example for reassessment of the management strategies for an existing reservoir is based on the combined use of simulation and optimization. The main objective of the approach is to determine:

» the active reservoir storage requirement based on the current demand; and

• the best management strategy for the reservoir.

The first part of the objective is achieved by using a simulation algorithm named RESER (Simonovic, 1992). The second part of the main objective is attained by using reservoir yield optimization (Simonovic, 1987) without predefined demand for water from the reservoir. Therefore, the approach developed in this example provides a recommendation on how to use a reservoir, based on the optimal compromise between the physical, economic and environmental conditions.

Simulation The simulation model used within the proposed approach is based on the continuity equation and set of probabilistic criteria. The model makes use of a direct search technique for finding the minimum required reservoir capacity. Using the initial storage value provided by the user, the monthly operation of the reservoir is simulated. By simulating the reservoir operation under given inflow and demand conditions, four probabilistic measures are cal­culated and compared with desired values. The four probabilistic criteria are:

• monthly reliability (frequency of water shortage) of reservoir water supply (a) ;

® yearly reliability of water supply ((3);

• yearly vulnerability (amount of water shortage) of water supply (y); and

• monthly vulnerability of water supply (8).

The first two criteria are used to control the number of reservoir failures. The latter two, however, are used to take into account the severity of reservoir failures. If these four criteria are not met, the reservoir size is increased by a step size. The simulation procedure is then repeated. This process continues until the reliability constraints are satisfied. From the use of simulation and direct search, the model arrives at a value of optimal required reservoir storage capacity.

Within the reassessment process the simulation model is used to indicate the availability of reservoir capacity to be used for the new purposes. Four

110 Design and management of reservoirs

calculated reliability values are also provided for the user in order to help in evaluating how efficient is the current use of reservoir storage. Due to the use of the direct search technique, the storage is known at all time periods during a simulation, so the storage-dependent losses can be directly calculated and storage failures monitored.

The simulation model uses the following information as input: the reservoir storage-elevation curve; monthly inflows; monthly evaporation; monthly seepage; the capacity of reservoir outlet structures; monthly water demand; starting reservoir storage; the increment for increase in storage; reservoir maxi­mum capacity; and minimum reservoir storage reserved for sediment deposition.

Reservoir yield optimization In the reassessment process for an existing reservoir, the physical characteristics of a reservoir are known. The mathematical model for the reservoir yield optimization includes the three following principles (Simonovic, 1987):

• equal treatment of the reservoir users;

• use of failure magnitude and number of failures in the analysis; and

• multipurpose use of reservoir active storage for municipal, industrial and irrigation water supply, and hydropower generation.

Originally, the reservoir yield optimization model had been developed to assist in the long-term comprehensive water management planning. Using only the character of different water demands, represented by the relative demand coefficients, the model provides planners with the optimal value of reservoir yield. For the purpose of the reassessment of reservoir management strategies, the original model has been modified to find the best combination of reservoir uses and monthly reservoir releases for different purposes. The best combina­tion is the combination which maximizes the reservoir yield. Maximization of the reservoir yield provides the planner with the physically "most appropriate solution". If the reservoir performs better for the combination of new purposes with respect to the current ones, the transition should be performed to allow for operating the reservoir in such a way as to maximize its yield. This transi­tion may require the identification of new water sources or modification of the physical system used for water supply in a region. The modification may require new structures or innovative solutions. New structures can include increasing the dam height, building new reservoirs, developing new aquifers, water transfers or connecting existing reservoirs so that they can perform as a system. Innovative solutions may include nonstructural measures such as demand management, supply management, change of the regional economy, diversification of industrial production, and the controlled distribution of population. Once more, note that the economic criteria are incorporated in the reassessment process indirectly through the relative water demand coefficients and relative weights assigned to different demands for water.

Design and management of reservoirs 111

The main input into the optimization program consists of the physical and inflow information for a reservoir and the relative demand requirements for each user. In addition the program requires information on supply reliability for each user. For the purpose of reassessment four sets of relative coefficients are required. One set is needed for each of the three users (municipal and industrial water supply; agricultural water supply; hydropower generation) representing the monthly percentage of their individual total demands. These coefficients may be determined by analysing the past demands or by making projections for the future. The fourth set of relative coefficients divides the total amount of water available for use from the reservoir active storage amongst the three principal users.

Work is in progress on developing a decision support system for a multipurpose reservoir reassessment by adding a user-friendly -interface and an expert system module for capturing the existing experience in reservoir operations to the modelbase and database.

Contribution to sustainability The reservoir reassessment system is capable of addressing a number of important issues related to sustainability. By reassessing the reservoir use the spatial and temporal scales of the reservoir analysis are changed. Considering the best physical use for the existing reservoir requires an evaluation of alternative resources for meeting the water demand. Alternative sources require a considerably larger spatial scale to be used. The use of a longer historical sequence of reservoir inflows for the reassessment of reservoir storage allocation will affect the temporal scale.

Through the ability to generate numerous hydrological scenarios (ARMA model within the database component) the reassessment methodology is capable of supporting "what if" analysis. Different hydrological scenarios can be examin­ed and an appropriate decision reached for the reservoir storage allocation.

Maximization of reservoir yield provides the decision-maker with the physically "most appropriate solution". This solution can be used instead of using the reservoir for the purpose for which it was designed. As presented, the optimization model is inherently dealing with the changing objectives to find the reservoir yield as a product of the best compromise between different users of the reservoir storage. The methodology can react to changes identified earlier by adjusting the relative demand coefficients. The final benefit of the proposed reassessment approach is that it provides a direct response to physical, economic, social and environmental conditions surrounding the reservoir.

The methodology deals with future uncertainty at two levels. Firstly, within the simulation, four reliability criteria are used to evaluate the reservoir storage necessary to satisfy the current reservoir demand. Secondly, the developed optimization model belongs to the group of implicit stochastic techniques. The main input into the model may be a historical sequence of monthly reservoir inflows or a synthetically generated sequence of inflows.

112 Design and management of reservoirs

An example— Wonogiri Reservoir in Indonesia

The methodology presented in Section 4.3.2 for reservoir reassessment has been applied to the Wonogiri Reservoir in central Java, Indonesia (Simonovic & Qomariyah, 1993). Due to the attainment of self-sufficiency in rice production and industrial growth in the region, a reassessment of Wonogiri management strategies to take into consideration the needs for hydropower production and industrial and municipal water supply was undertaken. The analysis of the active storage of the Wonogiri Reservoir demonstrates that the present storage capacity is greater than that required for the irrigation of 25 320 ha. Reallocation of the existing active storage to include hydropower generation and municipal and industrial water supply, without ignoring the present function of the reservoir, is feasible. The water supply/irrigation/ hydropower demand reliability combination of 40%/40%/20% generates the "best" trade-off between the three water users resulting in an overall reservoir yield of 97%. Therefore, a modification of Wonogiri Reservoir management is strongly recommended to include the other two purposes.

The analysis was done efficiently, but the lack of a good usermachine interface was acknowledged. It was also obvious that the reassessment procedure, and the methodology in general, can benefit from incorporating a module containing the expertise related to a particular reservoir operation.

4.3.3 Methodology for net benefit allocation for reservoir redevelopment

Introduction

In this section, an example of the net benefit allocation scheme for reservoir redevelopment is presented. In multipurpose (or multi-participant) water resources development projects that include reservoirs, it is a common practice for a basinwide planning authority to evaluate the economic feasibility of the project as a whole by carrying out a cost-benefit analysis. If the project is justified, then a crucial question is how to allocate the total costs to the agencies (representing specific users) involved in the project. As will be explained later, for a conventional type reservoir development, there is a reasonable methodology on which costs can be allocated. But for the redevelopment of existing reservoirs, not only the costs but also the benefits from the redevelopment have to be allocated and there is no current methodology. Due to the newly emerging need for redevelopment, therefore, the conventional cost allocation scheme has to be extended to net benefit allocation. What follows is a methodology for dealing with the net benefit allocation by the use of game theory.

Design and management of reservoirs 113

Cost allocation problem

We start with cost allocation. In order to explain the cost allocation scheme, the following terms are defined.

(a) Definitions

» Player /—The agencies or users which are involved in a project.

• Coalition—Any groups formed by any players. A special coalition formed by only one player ("acting alone") is also included.

» Grand coalition A7—The largest coalition (in size) which includes every player.

» Partial coalition S—The subset of grand coalition N (S çzN).

9 Alternative cost C(S)—Since an acting-alone coalition means constructing a reservoir by each player, the cost for this hypothetical reservoir can be calculated. The cost is called an alternative cost. The alternative cost for coalition S is denoted by C(S), while the total cost for the project is C(N). C(S) is called a cost function of the alternate project joined by S (smaller in membership size than AO.

(b) SCRB method

In multipurpose reservoir projects, the SCRB (Separable Costs Remaining Benefit) method (Federal Inter-Agency River-Basin Committee, 1950) is a time-tested conventional cost allocation method. The allocation equation for the SCRB method is shown as follows. At first, the separable cost SC for player i, the non-separable cost NSC, and the remaining benefit pi are defined as follows:

SQ = C(N) - C(N- {i}) (4.13)

NSC=C(N)-YlSCi (4.14) ieN

pi = min{£({/}), C({i})} -SQ (4.15)

Here, b({i}) is the benefit which player i can obtain from the grand project, in which all players participate. SC, is the cost which is saved in the case that player i does not participate in the project. In other words, SC, is the incremental cost in the case player i joins the coalition N at the end of the queue of players. NSC is the difference between the total cost of the project (C(A0) and the sum of separable costs of every player. In the definition of pi, min{£>({i}), C({i})} can be regarded as its willingness to pay, because it means the upper limit of player i's payment for the multipurpose reservoir project, pi is the difference between the willingness to pay and player i's separable cost. The cost that player i has to pay by the SCRB method is:

114 Design and management of reservoirs

x^Sq+^-NSC (4.16)

First, player / has to pay its separable cost. That implies that player / should pay at least the incremental cost. Then the residual (NSQ is allocated in proportion to player i's remaining benefit.

(c) Shapley value

Independent of such practical developments as the SCRB method, the allocation methods based on game theory (particularly cooperative game theory) have been proposed and discussed by Young et al. (1982). Game theory gives a theoretical basis to allocation methods. There are several allocation methods. In this study, we use the concept of the Shapley value. The Shapley value (Shapley, 1971) for n players is given by the following formula where xt is the cost which player / has to pay.

x , = I (n~s)l(f~l)l{C(S)-C(S-{i})} (4.17) ScNJeS n -

The Shapley value assumes that players form a coalition by joining one by one. In the case of three players, there exist six orders to form the grand coalition which includes all three players. In the SCRB method only the incremental costs to join the coalition N, at the end of the queue of players, are used to decide the amounts of allocated values. However, in the Shapley value method, player i's incremental costs are different and depend on when the player joins the coalition. Therefore, as shown in equation (4.17), the costs for every order (which are called marginal costs) are summed. The average of the marginal costs is then calculated as the Shapley value for player i.

Note that in the Shapley value, in the case of more than four players, the method accounts for all possible intermediate coalitions, whereas the SCRB method does not.

From cost allocation to benefit - cost (net benefit) allocation

As Okada & Tanimoto (1996) have shown, the practice of cost allocation uses the convention that the calculation of the alternative cost functions C(5) should be made for the same site as that of the grand project. In addition to this "single-site" assumption, there are two more essential assumptions made to calculate alternative cost functions:

(1) Players always have to provide the same level of service. For example, the agency responsible for flood control always plans to prevent the fixed scale of floods.

Design and management of reservoirs 115

Fig. 4.5 Multipurpose and multi-site situation.

(2) Players always construct a reservoir to achieve a necessary level of service. For example, the possibility of thermal power generation is not considered in the discussion of the cost allocation for hydropower generation.

These assumptions may well be justifiable if players have basically agreed to the grand project of a multipurpose reservoir subject to cost allocation. That is, if the project has been justified these assumptions are obviously appropriate. In such a case, a cost function is used to guarantee fairness and/or equity as defined by cooperative game theory.

However, more generally, we assume the situation in which the members joining the project and/or the size of a project have not yet been decided. In Fig. 4 .5 , for example, two agencies, the waterworks bureau (player 1) and the power generation company (player 2), plan to use the storage capacity of a reservoir; and there are two alternative sites for constructing the reservoir. Several patterns for using these sites can be assumed. For example:

(a) The waterworks bureau constructs and uses a reservoir only for itself, and the power generation company constructs and uses another reservoir at another site.

(b) The two agencies construct and use a large multipurpose reservoir at a single site.

(c) The two agencies construct, use and manage two multipurpose reservoirs together.

(d) Although the waterworks bureau has to obtain a stable water supply from a new reservoir, the power generation company does not necessarily have to construct a reservoir for hydropower generation. They may be satisfied with small-scale power generation (which needs a small-scale reservoir storage), or other power generation means (thermal, nuclear, etc.).

116 Design and management of reservoirs

In this situation, the single-site assumption cannot be applied. According to (d), assumption 1 (players always have to provide the same level of service) and assumption 2 (players always construct a reservoir to achieve a necessary level of service) are also not appropriate. In summary, as shown in Table 4 .3 , the problem in this situation involves site selection and variable service level. Furthermore, several projects (from (a) to (d)) can be assumed. This means that the project to be implemented has not been specified.

To cope with this problem, it is proposed to classify players into two types; one the "service provider" and the other the "(land) owner of site". One provider is a municipal water provider and the other is an electricity provider. It is assumed that any project can be implemented by forming a coalition including both owners and providers.

The proposed method allocates both costs and benefits of the project at the same time. This means that the net benefit of the project is allocated. In this case, the allocation is used not only for financial arrangement, after project justification, but also for project justification. Game theoretic allocation methods, including the Shapley value, can be extended to apply to the net benefit (benefit - cost) allocation problem. To allocate the net benefit, it is necessary to decide a characteristic function (v(S), v(N)) instead of a cost function (C(S), C(A0). A characteristic function v(S) for a coalition S in the redevelopment project can be defined as the net benefit which the respective coalition can obtain by themselves. According to the Shapley value, for example, the net benefit allocated to the player / (y,) is calculated as follows:

In the next subsection, this extended scheme of net benefit allocation is explained in a reservoir redevelopment case.

Table 4.3 Project types.

(4.18)

Project type Single/ Fixed/ Fixed/ Financial Exis multi variable variable arrangement/ new site service provision project resei

Existing/ Cost new allocation/ reservoir net benefit

level means justification allocation

Multipurpose and Single-single-site reservoir site development Multipurpose and Multi-multi-site reservoir site development (see Fig. 4.5) Multipurpose and Multi-multi-site reservoir site redevelopment (see Fig. 4.6)

Single- Fixed Fixed Financial New site (reservoir) arrangement

Cost allocation

Variable Fixed Project (reservoir) justification

New Net benefit allocation

Variable Fixed Project (reservoir) justification

Existing Net benefit and new allocation

Design and management of reservoirs 117

New purposes: Flood Control (+ Water Supply)

Site B ( N e w

Rural Area requiring flood control

Original purpose: Water Supply

New purposes: Water Supply (+ Flood Control)

Urbanized Area requiring water supply and flood control

River basin b River basin a

P r o v i d e r s Munic ipa l w a t e r (Player 1), Flood c o n t r o l O w n e r S i t e A (Player 3), S i t e B (Player 4)

(Playe

Fig. 4.6 Redevelopment project.

Net benefit (benefit - cost) allocation in reservoir redevelopment projects: an example

As shown in Table 4 .3 , when the net benefit allocation scheme is extended to reservoir redevelopment projects, the approach similar to the multipurpose and multi-site situation in Fig. 4.5 may be applicable. However, a critical difference exists between them. In reservoir redevelopment projects, the new users, which do not have user's rights in an existing reservoir, may agree to change the service level depending on the circumstances. If they can form a coalition for the redevelopment project, they may fulfil their purposes. However, if they cannot cooperate with existing users, they may choose not to provide the same level of service. In other words, the status quo of user 's rights may affect the feasibility of alternative projects. This asymmetry in the ownership right between agencies is also seen in other infrastructure development projects.

(a) Two-purpose two-site case

For illustration, a reservoir redevelopment project shown in Fig. 4.6 is considered where two agencies intend to join a redevelopment project: One is the waterworks bureau, and the other is responsible for flood control. The reservoir at Site A has been operated by the waterworks bureau, supplying municipal water for a long period of time. Now, the flood control agency recognizes the necessity of the capacity for flood control in both river basins. It is assumed that the population in river basin b (rural area) is smaller than that in river basin a (urbanized area). That implies that the benefit for flood control in river basin b is smaller than that in river basin a.

In river basin a, the appropriate site for a reservoir (Site A) is already occupied by the existing reservoir. Therefore, it becomes necessary to reallocate part of the capacity at Site A to flood control. On the other hand, in

118 Design and management of reservoirs

river basin b, the flood control agency can obtain a new reservoir at Site B. In addition to the capacity for flood control for river basin b, the capacity for compensating for the reduced portion of municipal water supply at Site A should be obtained at Site B. This project is called a "reservoir redevelopment project". A project similar to this has been implemented in the western part of Japan.

It is assumed that the municipal water supplier and the owner of Site A belong to the same organization, the waterworks bureau, and that the owner of Site A is subordinate to the municipal water supplier, because Site A is already occupied and used exclusively for municipal water supply. In such a case, the project between a new provider (player 2) and the owner subordinate to the existing provider (player 3) is infeasible. In other words, the flood control agency cannot use the reservoir at Site A without permission from the municipal water provider. As a result, the existing provider with a vested interest has power in the negotiation for the net benefit allocation.

Under such an hierarchical structure, the following three types of feasible projects can be assumed:

• A redevelopment project for both river basins: The existing reservoir at Site A is used for flood control for river basin a and municipal water supply, and the new reservoir at Site B is used for flood control for river basin b and municipal water supply (by players 1, 2, 3, and 4).

• An existing project for river basin a: The existing reservoir at Site A is used only for municipal water supply not at all for flood control for river basin a (by players 1 and 3).

• A flood control project for river basin b: The new reservoir at Site B is used only for flood control for river basin b (by players 2 and 4).

(b) Benefits and costs

For each project, the net benefit can be expressed as follows:

Br = The net benefit of a redevelopment project between two reservoirs = 6(1,AB) + b(2,A) + b(2,B) - c(12,A) - c(12,B)

Be = The net benefit of an existing project (the benefit from the reservoir only for waterworks) = 6(1, A)

Ba = The net benefit of an alternative project at site B (when player 2 does not obtain any capacity at site A and constructs a reservoir at site B only for flood control) = Z?(2,B) - c(2,B)

(c) Net benefit allocation

By taking note of the provider-owner interdependence structure, we use the concept of cooperative games with Permission Structures by Gilles et al. (1992) to formulate the net benefit allocation of this redevelopment project. As

Design and management of reservoirs 119

previously explained, player 3 is subordinate to player 1. Following Gilles et al. (1992), we assume the following principle for characteristic functions to hold:

• The coalition for the project, which is infeasible based on the hierarchical structure, cannot obtain any positive payoff.

We also set two additional principles:

• The characteristic function for only one player is always zero (v({/}) = 0).

• The coalition of two providers (players 1 and 2) or two owners (players 3 and 4) cannot obtain any positive payoff (v(12) = 0, v(34) = 0).

Therefore, the characteristic functions become:

v(l) = 0, v(2) = 0, v(3) = 0, v(4) = 0

v(12) = 0, v(13) = Be, v(14) = 0, v(23) = 0, v(24) = Ba, v(34) = 0

v(123) = Be, v(124) = Ba, v(134) = Be, v(234) = Ba

v(1234) = Br

As an example, hypothetical benefits and costs are given as follows:

Benefits: b(l, A) = 4, b(l, AB) = 4, b(2, B) - 5, b(2, A) = 10

Costs: c(2, B) = 3, c(12, A) = 2, c(12, B) = 4

Net benefits: Be = b{\, A) = 4,

Br = b{\, AB) + b(2, A) + b(2, B) - c(12, A) - c(12, B) = 13

Ba = b(2, B) - c(2, B) = 2

where ^(players, sites) is the benefit obtained from the project which corresponding providers can implement at the corresponding site, and c(players, sites) is the cost necessary for the project which corresponding providers can implement at the corresponding site.

Then the resulting Shapley value is (3.75, 2.75, 3.75, 2.75). Figure 4.7 shows the flow of the net benefit allocation. First, player 2 pays the whole cost of the redevelopment project (c(12, A) + c(12, B)). Furthermore, it has to refund some parts of its benefit from the project. The Waterworks Bureau can obtain the benefit (3.5) additional to the reserved benefit from municipal water supply (4.0).

Conclusion

In this section, the current cost allocation methods are discussed and then an extended scheme of net benefit allocation for reservoir development/ redevelopment projects is proposed.

120 Design and management of reservoirs

Project Cost

6

W a t e r w o r k s Bureau

Benef i t from unic ipal Wate r Supply

Al loca ted Va lue 7.5

Fig. 4.7 The flow in net benefit allocation.

The construction of new reservoirs will become more difficult in the future. Therefore, dam sites themselves will increasingly be viewed as rare resources. Furthermore, water and its related environmental resources are also site-specific and therefore should be regarded as land-related resources. This means that the redevelopment of existing reservoirs involves resource reallocation with a view to improving the efficiency and effectiveness of providing water resources to meet the changing needs of society. The net benefit allocation method shown in this section is applicable to various types of reservoir redevelopment projects, which need coordination between existing users with vested interests and new users.

4.3.4 The least marginal environmental impact (LMEI) rule for reservoir sizing

Introduction

The scale-diseconomy of large reservoirs and the importance of reservoir size in relation to ecological effects was discussed in Sections 1.3.1 and 1.3.2 and the checklist rationale for reservoir sizing for sustainable reservoirs was discussed in Section 2.2.7. This section describes a reservoir sizing procedure called the least marginal environmental impact rule (LMEI) which was first introduced by Takeuchi (1997).

While the concept of sustainable reservoir development and management has been increasingly clear as extensively discussed in previous chapters, implementation of such a concept needs a concrete procedure to guide any reservoir project. As such the checklist in Chapter 6 suggests a new approach for reservoir sizing, which is considerably different from the most commonly exercised economic efficiency rule. In the latter method, the net economic benefit is maximized under a given set of constraints where the socio-

B e n e h t from

Flood Cont ro l

15

C D 3.5 2 .75

Al loca ted Va lue 2 .75

Al loca ted Va lue

2 .75

Design and management of reservoirs 121

environmental conditions to be satisfied are usually given as constraints. The allowable level set for the constraints is, however, not necessarily the minimum attainable but is often a result of compromise with the socio­economic reasons being the major driving force. In addition there are often some institutionalized incentives, such as various subsidies, that play an important role in project justification so that less attention is given to the alternatives with the least environmental effects. This approach does not necessarily ensure the development of sustainable reservoirs. The LMEI rule is a modification to the traditional economic efficiency approach.

Before stating the rule project impacts are classified into economic impacts and enviromnental impacts. Economic impacts are the impacts whose receivers and whose shares are identifiable and whose costs and benefits can be allocated to particular natural or legal individuals. Most socio-economic impacts measurable in monetary terms belong to this category. Environmental impacts are the impacts whose receivers and whose shares are not necessarily identifiable and whose costs and benefits cannot be allocated to particular natural or legal individuals. Most socio-environmental quality factors, including social ease and satisfactions and cultural assets such as historical artefacts and views, and ecological factors are termed environmental impacts. They are equivalent to external diseconomies.

Both economic and environmental impacts include tangibles and intangibles and may or may not be measured in commensurate values. The economic impacts are equivalent to the net economic benefit or loss if all the economic impacts are measured or expressed with weights in commensurate values. Similarly, the environmental impacts can be expressed as environmental gain or loss if the impacts are classified into positive and negative effects to the environment which do not cancel each other.

The LMEI rule

The LMEI rule is a rule to decide the scale of reservoir only to the extent that the incremental unit increase of dam height results in the least negative environmental impact solution among all the economically feasible reservoir or non-reservoir alternatives, that can provide the same level of incremental gain in the global objectives that the extra unit increase of dam height can provide.

Specifically, suppose the current plan for the dam height is h and the level of net benefit expected is B{h). Consider whether an increase in height of dh produces an extra net benefit dB(h) = B(h + dh) - B(h) or not. The conditions to be satisfied under the LMEI rule is that among the many other alternative ways to obtain the extra net benefit dB(h), raising the dam height by dh should be the solution that creates the least environmental impact. If not the dam height increase by dh should be abandoned and the benefit should be

122 Design and management of reservoirs

derived from the solution with the least adverse environmental impact. Further increases of dam height should be considered by a repetition of this procedure dh by dh. (Note that a different magnitude of dh may lead to a different solution since the cost and benefit response surfaces are not necessarily monotonous. Some different magnitudes of dh may have to be examined to avoid local optima.)

By replacing the approach of the maximum net-benefit with a constraint on environmental impacts, this rule ensures that the reservoir option is, if selected, always the least destructive solution in environmental impacts. It will highlight the point at which the reservoir is no longer the best available solution and has to be replaced by another environmentally more desirable choice.

The following formulation will specify the concept in more detail:

(1) Suppose an action x be any action. Since any action can be reached in many different ways, a decision vector x can be defined for a set of decisions associated with an act ions . Similarly, suppose an action ar be an increase in dam height from h to h + dh, then a decision vector ar is defined as a set of decisions associated with an action an i.e.:

x = {any action}

ar = {action | dam height increase from h to h + dh}

x = {decision vector | a set of decisions associated with an action x)

ar = {decision vector | a set of decisions associated with an action ar)

(2) Let G(x), NB(x) and EC(x) be the constraint set, net economic benefit and external socio-environmental diseconomy associated with the decision vector x of action x, i.e.:

G(x) < 0: the constraints set with respect to a decision vector x.

NB(x) — {net benefit | net economic benefit by decision vector x where all the economically commensurable values are internalized}

EC(x) = {environmental diseconomy vector | external non-commensurable socio-environmental diseconomies associated with decision vector x}

(3) There must be an optimal set of decisions in any action x as well as the dam height increase ar and the net benefit and external diseconomy associated with them, i.e.:

x* = {x I max NB(x) subject to G(x) < 0}: the optimal decisions associated with an action x.

a,* = {ar I max NB(ar) subject to G(ar) < 0}: the optimal decisions associated with an action ar

NB(a*) = {net benefit | by actions ar*}: marginal net benefit by dh. Note

Design and management of reservoirs 123

that if this value is negative, the dam height will not, in general, be increased any further

EC(a,* = {environmental diseconomy vector | by actions a*}: marginal environmental impact due to the increase in height dh

(4) Then the LMEI rule selects action ar to raise the dam height by dh provided there is no such action as x that satisfies the following conditions:

NB(x*) > NB(a*) and EC(x*) < EC(a,*) (4.19)

(5) The further increase of dam height should be considered by the repetition of this procedure dh by dh.

Here the distinction between the socio-environmental costs to be counted in NB and EC is important. Every socio-environmental cost that can be valued in monetary terms has to be internalized in the net benefit NB. Every socio-environmental diseconomy that can be counted but not valued in monetary terms is counted in the environmental cost vector E C item by item, each having a different non-commensurable unit. The vector comparison required in inequality (4.19) may be difficult in theory but would often be obvious in practice. Further discussion on this matter is omitted here.

Relation between the L M E I rule and the economic efficiency rule

The LMEI rule suggests a reservoir scaling procedure, always investigating whether or not any other alternative means is available that is equally productive but less socio-environmentally damaging when every incremental height increase dh is considered for a dam. The dam height is quite often decided for geological, hydrological and economic reasons. But as pointed out earlier (Section 1.3.1), the unit increase of dam height inundates the area proportional to the 9th power of the dam height and the area inundated is most directly related to the socio-environmental impacts. The LMEI rule is therefore very conscious of dam height and suggests a procedure which considers the dam height, as being of equal importance as the problem of "to build or not to build a dam".

A major question that may arise here would be how much economic efficiency has to be given up when the reservoir option is to be abandoned to an alternative. This is a question of the relationship between the economic efficiency rule and the new LMEI rule. The answer to this question is very simple: "it is unnecessary to give up anything". This is because the rule is constrained by the condition "among all the means that can provide the same level of incremental gain" or the first inequality condition of (4.19).

Up to the reservoir height where the reservoir option is the least marginal environmental impact solution among alternatives that provide the same net economic benefits, the LMEI rule coincides with the maximum net benefit

124 Design and management of reservoirs

solution. However, once it comes to the point where the least costly environmental impact would be obtained from an alternative and not from a reservoir project increment, the reservoir project would not be developed under the LMEI rule to a scale where net economic benefits are maximized. In these cases, the LMEI rule acts as a constraint on the economically optimal development of the reservoir project, while still allowing optimal development to proceed via the alternative which minimizes environmental costs. In other words, as far as economic efficiency is concerned, there is no difference between the two. The LMEI rule just requires a reservoir planner to give up the reservoir option and to achieve the same objective by alternative means.

The LMEI rule is considered superior to a rule which maximizes net project benefits subject to an arbitrary limit on the environmental constraint related to the minimum attainable from alternative solutions. In the case of demand management alternatives, for example, environmental costs are likely to be very low; yet, under the arbitrary constraint on reservoir development, this low environmental cost alternative would never be considerable. The LMEI rule avoids these problems by tracing an environmentally least costly path of development up to the point where the marginal benefit equals the marginal cost.

Conclusions

The least marginal environmental impact (LMEI) rule was proposed to ensure the selection of the least environmentally destructive choice among the equally productive solutions. There are many disputes as to whether or not to create a reservoir in order to achieve a goal. But the most important issue would be the size and the location of reservoirs. There is no reason to create a reservoir of the size where the marginal environmental loss becomes more than that of alternative solutions elsewhere.

4.4 CONCLUSIONS

This chapter of mixed contributions deals with the design and management of reservoirs. Issues involved in implementing sustainability principles to the complex task of reservoir design and management are discussed. Various contributions, addressing sustainability in different ways, illustrate a variety of sustainability definitions. A unique solution for sustainable reservoir design and management has not been detected. However, innovative thinking and serious effort aimed at addressing sustainability have been presented.

In looking at the time series analysis of reservoir inflows a compromise has to be found between the reliability of estimation of statistical

Design and management of reservoirs 125

characteristics and the complexity of the model. The application of detailed physically-based models enables simulation of a wide range of possible variations in runoff factors and human activities in a watershed for different possible combinations of meteorological inputs.

In looking at the importance of hydrological forecasts the following conclusions have been identified:

• small reservoirs are more vulnerable to forecasting errors than large ones; and

• in the case of large reservoirs, good forecasts can reduce considerably the storage necessary to attain a given release level.

The improvement of hydrometeorological forecasting results in "virtual capacity expansion" and improves reservoir sustainability by:

• reducing the size of new reservoirs and at the same time minimizing the environmental destruction;

» gaining more benefits from the existing reservoirs; and

• offsetting anticipated developmental needs for new reservoirs.

In the consideration of a new approach for reservoir design based on De Novo programming considerable benefits in comparison to classical optimization are acknowledged. The reservoir sizing problem is formulated realistically as a system of several non-commensurate objectives. The design of the entire resource portfolio could shape the feasible decision space to achieve the metaoptimum, i.e. the optimal solution for which all objectives are optimized simultaneously. The allowable budget, or penalty/benefit function, to be spent on the purchase, or design, of the resource portfolio will then determine the feasibility of the metaoptimal solution. It is expected that with the physical resource capacities being inflexible (fixed), or the system budget being insufficient, it may not be possible to achieve the metaoptimal solution. The design of the system may shape the feasible region in such a way that a solution (or the set of non-dominated points) which is "closer" to the metaoptimum is obtained.

Reassessment of existing reservoirs has been identified as one of the realistic demands imposed by sustainability. A methodology based on the integrated use of reservoir storage simulation and reservoir yield optimization has been demonstrated to be capable of addressing different sustainability issues successfully.

One of the critical barriers in implementing the reassessment and redevelopment of an existing reservoir system is the allocation and reallocation of costs and benefits which accrue due to the modification of the system or of a newly created system. A more realistic approach than the traditional separable cost remaining benefit approach is necessary to promote various coalitions to initiate the development and redevelopment plans. The new

126 Design and management of reservoirs

methodology based on Shapley values demonstrates a mechanism to introduce more fairness in cost/benefit allocation which will surely contribute to encouraging more investment for redevelopment rather than in new development.

The least marginal environmental impact rule is proposed to ensure the selection of the least environmentally destructive choice among the equally productive solutions. Its application in reservoir design is of great potential when questions of reservoir location and storage capacity are addressed. There is no reason to create a reservoir of the size where the marginal environmental loss becomes more than that of alternative means elsewhere.

It is our wish and hope that some of the contributions presented in this chapter will be criticized, some be accepted and all will help in moving reservoir design and management to a new level. A level at which we will be able to look at the reservoirs in a more realistic way. A level at which we will be able to think about compromises. A level at which new ideas are going to be born.

REFERENCES

Abbott, M. B . , Bathurst, J. C , Cunge, J, A. , O'Connell , P. E. & Rasmussen, J. (1986) An introduction to the European Hydrological System-Système Hydrologique Européen, SHE, 2: Structure of a physically-based distributed modelling system. J. Hydrol. 87, 61-77 .

Bare, B. B. & Mendoza, G. A. (1988) A soft optimization approach to forest land management planning. Canadian J. Forestry Resources 18, 545-552.

Bare, B. B. & Mendoza, G. A. (1990) Designing forest plans with conflicting objectives using De Novo programming. J. Environ. Management 31 , 237-246.

Bender, M. J. & Simonovic, S. P. (1996) Fuzzy compromise programming. Water Resour. Res. Report no. 33, The University of Manitoba.

Box, G. E. P. & Jenkins, G. M. (1976). Times Series Analysis: Forecasting and Control. Holden-Day, Inc. , San Francisco.

Bras, R. L. , Buchanan, R. & Curry, K. C. (1983) Real time adaptive closed loop control of reservoirs with the High Aswan Dam as a case study. Water Resour. Res. 19(1), 33-52.

Chan, S. & Bras, R. (1979) Urban storm water management distribution of flood volumes. Water Resour. Res. 15, 371-382.

Despic, O. & Simonovic, S. P. (1977) Methods in evaluating qualitative criteria in water resources multi-criteria decision making. Water Resour. Res. Report no. 37, The University of Manitoba.

Diaz-Grenados, M. A. , Valdes, J. B. & Bras, R. L. (1984) A physically-based flood frequency distribution. Water Resour. Res. 20(7), 995-1002.

Eagleson, P. S. (1972) Dynamics of flood frequency. Water Resour. Res. 8(4), 878-898.

Federal Inter-Agency River-Basin Committee (1950) Proposed Practices for Economic Analysis of River Basin Projects. Technical Report, Washington DC.

Ford, T. D . (1990) Reservoir storage reallocation analysis with PC. J. Water Resour. Plan. Management ASCE 116, 402-416.

Gilles, R. P . , Owen, G. & van den Brink, R. (1992) Games with permission structures: the conjunctive approach. Int. J. of Game Theory 20, 277-293 .

Israel, M. & Lund, J. R. (1992) Managing existing reservoirs to meet new challenges. In: Water Resources Planning and Management: Saving a Threatened Resource—In Search of Solutions, ed. M. Karamouz, Water Forum '92, Baltimore, 673-678.

Johnson, W. K., Wurbs, R. A. & Beegle, J. E. (1990) Opportunities for reservoir-storage reallocation. /. Water Resour. Plan. Management ASCE 116, 550-566.

Design and management of reservoirs 127

Kavvas, M. L. & Delleur, J. W. (1981) A stochastic cluster model for daily rainfall sequences. Water Resour. Res. 17(4), 1151-1160.

Kleraes, V. (1977) Value of information in reservoir optimization. Water Resour. Res. 13(5), 837-850. Kojiri, T. (1995) Planning and management of water resources systems under uncertain future information

using fuzzy optimization. In: Modelling and Management of Sustainable Basin-Scale Water Resources Systems, eds S. P. Simonovic, Z. W. Kundzewicz, D. Rosbjerg & K. Takeuchi, IAHS Publ. no. 2 3 1 , 311-318 .

Kojiri, T. & Sakakima, S. (1993) Decision support system of reservoir operation considering weather forecast and hydrograph similarity. In: Extreme Hydrological Events: Precipitation, Floods and Droughts, eds Z . W. Kundzewicz, D . Rosbjerg, S. P. Simonovic & K. Takeuchi, IAHS Publ. no. 213 , 429-438 .

Kuchment, L. S. & Gelfan, A. N. (1991) Dynamic-stochastic models of rainfall and snowmelt runoff formation. Hydrol. Sci. J. 36(2), 153-169.

Kuchment, L. S., Demidov, V. N. & Motovilov, Yu. G. (1986) A physically-based model of the formation of snowmelt and rainfall runoff. In: Modelling Snowme/t-Induced Processes, ed. E. M. Morr is , IAHS Publ. no. 155, 27-36 .

Leclerc, G. & Shaake, J. (1972) Derivation of hydrologie frequency curve. Technical Report no. 142, Ralf Parsons Laboratory, Water Resources and Hydrodynamics, MIT, Cambridge, Massachusetts.

Mishalani, N . R. & Palmer, R. N. (1988) Forecast uncertainty in water supply reservoir operation. Water Resour. Bull. 24, 1237-1245.

Okada, N . & Tanimoto, K. (1996) Interpreting and extending conventional cost allocation methods for multipurpose reservoir developments by use of cooperative game theory. In: Proc. IEEE International Conference on Systems, Man and Cybernetics (Beijing, October 1996), 2698-2703.

Phatarfod, R. M. (1977) The sampling error of storage size. Water Resour. Res. 13(6), 967-969. Shapley, L. S. (1971) Cores of convex games. Int. J. of Game Theoiy 1, 11-26. Simonovic, S. P. (1987) The implicit stochastic model for reservoir yield optimization. Water Resour. Res.

23(12), 2159-2165. Simonovic, S. P. (1989) Application of water resources systems concept to the formulation of a water

master plan. Water International 14, 3 7 - 5 1 . Simonovic, S. P. (1992) Reservoir systems analysis: closing gap between theory and practice. J. Water

Resour. Plan. Management ASCE 118(3), 262-280. Simonovic, S. P. (1996) Decision support systems for sustainable management of water resources: 1.

General principles. Water International. 21, 223-232. Simonovic, S. P. & Qomariyah, S. (1993) Reassessment of management strategies for Wonogiri Reservoir

in central Java. Water International 18, 207-216.

Sivaarthitkul, V. (1996) Quantitative assessment of efficiency increase of reservoirs by improvement of inflow forecasts. PhD Dissertation submitted to Yamanashi University, Japan, March.

Stedinger, J. R., Suie, B. F. & Loucks, D. P. (1984) Stochastic dynamic programming models for reservoir operation optimization. Water Resour. Res. 20(11), 1499-1505.

Svanidze, G. G. (1980) Mathematical Modelling of Hydrological Series. Water Resources Publications, Fort Collins, Colorado.

Takeuchi, K. (1990) Prediction accuracy of precipitation and practicability of anticipated release operation policy (in Japanese). Japanese Society of Civil Engineers Proc. Hydraulic Engineering 34, 73 -78 .

Takeuchi, K. (1997) Least marginal environmental impact rule for reservoir development. Special issue of Hydrol Sci. J., ed. S. Simonovic, 42(4), 583-598.

Takeuchi, K. & Sivaarthitkul, V. (1995) Assessment of effectiveness of the use of inflow forecasts to reservoir management. In: Modelling and Management of Sustainable Basin-Scale Water Resource Systems, eds S. P. Simonovic, Z. W. Kundzewicz, D. Rosbjerg & K. Takeuchi, IAFIS Publ. no. 2 3 1 , 299-309.

Tao, P. C. & Delleur, J. W. (1976) Multistation, multiyears synthesis of hydrologie time series by disaggregation. Water Resour. Res. 12(6), 1303-1312.

Valencia, R. & Schaake, J. C. Jr (1973) Disaggregation processes in stochastic hydrology. Water Resour. Res. 9(3), 580-585 .

Wood, E. F. & Harby, B. M. (1975) The application of hydrological models in analysing the impact of urbanization. In: Application of Mathematical Models in Hydrology and Water Resources Systems, IAHS Publ. no. 115, 265-278 . '

Yeh, W. W.-G. , (1984) A state-of-the-art review of systems analysis in reservoir management and operation. Water Resour. Res. 20, 959-972.

Yeh, W. W.-G. , Becker L. & Zettlemoyer, R. (1982) Worth of inflow forecast for reservoir operation. /. Water Resour. Plan. Management ASCE 108, 257'-269.

128 Design and management of reservoirs

Young, H. P . , Okada, N . & Hashimoto, T. (1982) Cost allocation in water resources development. Water Resour. Res. 18(3), 463-475.

Zeleny, M. (1973) Compromise programming. In: Multiple Criteria Decision Making, eds J. L. Cochrane & M. Zeleny, University of South Carolina Press, Columbia.

Zeleny, M. (1981) Systems Approach to Multiple Criteria Decision Making: Metaoptimum. Fordham University, Lincoln Center, New York.

Zeleny, M. (1986) Optimal system design with multiple criteria: De Novo programming approach. Engng Costs and Production Economics 10, 89-94.

Sustainable Reservoir Development and Management. Report by the IAHS/ICWRS Project Team. IAHS Publ. no. 251, 1998. 129

CHAPTER 5

CASE STUDIES

5.1 INTRODUCTION

Since reservoir development always implies some negative environmental irreversible effects, sustainable reservoir development must be seen in a broad context where an overall balancing of present and future life-supporting needs is a prerequisite for decision-making. This balancing is extremely difficult, and an uncritical use of the sustainability term is obviously a risk. However, the present report may prove to be a valuable tool in the search for sustainable solutions.

The following three case studies, as well as a country policy study, all address sustainability, but they focus on different aspects of the concept. An accounting of a broad set of sustainability issues in reservoir development is intended. The three cases represent reservoirs of small, medium and large sizes, and they are located in different climatic zones on different continents.

The first case study on the Alta Reservoir in Norway in particular addresses the rights of indigenous people and the vulnerability of nature. Also, public participation in the decision-making process is described. The second case study on the Aswan High Dam in Egypt focuses on the inundation of large areas with the associated needs for moving a large population. In addition the reservoir sedimentation problem is given a thorough treatment. The third case study on the Green River in Utah, USA, particularly takes the changed downstream conditions into consideration and discusses the possible use of artificial flooding as a means for achieving sustainability. Institutional aspects are also covered in detail.

Finally, a Japanese policy study on reservoir development discusses the balancing of social, environmental and economic issues in connection with reservoir development. In this context improved performance of existing reservoirs and other alternatives to reservoir building are considered.

The case studies and the policy study are individually presented and concluded. The studies are selected to illustrate how the sustainability concept can be applied in different ways to real-life water resources development. In this respect they may serve as a source of inspiration steadily to put more emphasis on sustainability related issues in future reservoir development and management.

130 Case studies

5.2 THE ALTA HYDROPOWER DEVELOPMENT SCHEME

5.2.1 Introduction to the Alta case study

The Alta hydropower development and the bitter struggle around that scheme not only made a deep impact on Norwegian public debate, but also created interest and engagement far outside the borders of the country. One reason for this was that besides all the classic conflicts usually found in connection with hydropower development, it also encompassed the rights of indigenous people, in this case the Sami people, and highlighted the issue of civil disobedience.

The case and the public debate lasted around 20 years, from the first plans around 1968 until the power plant was in operation in 1987, and the debate to some extent still lingers on. This is certainly much ado for a power plant with an annual production of 625 GWh, around 0.5% of the hydropower production in Norway. The production of the Alta plant, by the way, has become a new unit for energy production and consumption in Norway, one "Alta".

Some important milestones along the way: 1968: The Norwegian Water Resource and Energy Board publishes plans for development

of the Alta River with total production 1150 GWh. The State Power Plants applies for development of 860 GWh. Revised application for a reduced development of 710 GWh.

1974 1976 1978 Stortinget, the Norwegian Parliament, grants a license for the development of

625 GWh. 1979: Powerful demonstrations against the plans. Sami activists hunger strike in Oslo. Start

of the access road construction hampered by sit-down and chain-gang action in Alta. Preliminary halt of construction.

1981: New start of road construction. Large demonstrations and police action in Alta, 900 taken into custody and fined.

1982: "Folkeaksjonen"—the most important action group for Alta is dissolved. 1984: Work on the dam site and the power plant commences. 1987: The plant begins operation.

The following information is compiled from several sources, among others Ryvarden & Toemmeraas (1979), Dalland (1994) and Olje- og Energidepartementet (1978).

5.2.2 The river

The Alta River, or the Alta-Kautokeino River (the upper reaches are called the Kautokeino River) is one of the largest rivers in northern Norway, in the county of Finnmark. The catchment area is approximately 6500 km 2 , and the length of the river 150 km in a direct line from the Finnish border to where it meets the Alta Fjord on latitude 70°N. Its total length is 235 km. It drains

Case studies 131

large parts of the Finnmarksvidda plateau, an undulating plain with an altitude mainly in the 400-600 m a.m.s.l. range. The vegetation is mainly birch, shrub and heather, and the area is used as reindeer grazing ground by the Sami herd owners. From the plateau it plunges down from 250 m in the Vird'nejav'ri Lake to 80 m a.m.s.l. through the Sav'co gorges. From here it continues 46 km to the sea, the first stretch through the famous Sav'co canyon, the largest and most spectacular canyon in Norway, and the part of the river commonly shown in pictures. On this lower stretch the river carries salmon; it is one of the best salmon rivers in Norway. The river is navigable with the long and sleek wooden river boats of the region for most of its length, except the Sav'co gorge and the gorge between the Vird'nejav'ri and Ladnatjav'ri lakes.

The climate of the area is continental and dry, an annual precipitation of approximately 500 mm and a runoff of 360 mm. The winters are cold, but the summers are temperate, with mean July temperature of 13 °C. The hydrological regime is dominated by the spring flood in May, see Fig. 5 .1 .

The basin is sparsely populated. The main part of the catchment falls within the municipality of Kautokeino with a population of 3000, dominantly Sami. The two largest villages in the municipality are Kautokeino, the administration centre, population 1200, and Masi, population 400, almost exclusively Sami. Both are on the river, and the agricultural areas (4 km 2 ) in the municipality are mainly around Masi. Masi is situated at the upper end of

M e a n

Jan. Feb. March April May June July Aug. Sept. Oct. Nov. Dec.

Fig. 5.1 Seasonal variation of runoff on the Alta River from mean daily values at Masi, 1967-1986.

132 Case studies

the two narrow lakes formed by the river upstream of Sav'co: Vird'nejav'ri and Ladnatjav'ri.

The lower reaches of the river fall within the municipality of Alta. The town of Alta, with a population of 8000, the majority of which is Norse, is situated at the river mouth and is the largest town in the county of Finnmark. The border between the two municipalities has been disputed, but is now just upstream of the Sav'co gorge. As a result, most of the economic compensation and license fees for the hydropower development benefit Alta.

5.2.3 Undercurrents in society

The Alta dispute became far more bitter than could be expected from the physical facts about the actual scheme and its impacts. The reason for this was that three strong undercurrents in Norwegian society surged during the Alta case:

• the nature conservation movement,

• the Sami identity and nationality movement, and

• a growing district scepticism to central government.

Nature conservation As Norwegian electricity production is 100% hydropower, and hydropower development is the main and most visible impact of the industrialized society in the pristine mountain and wilderness areas of Norway, there were increasing confrontations between conservationist groups and hydropower developers in the 1960s and 1970s. The two main battlefields before Alta were Aurland and Mardoela. Aurland is a scenic valley and a very popular hiking area in western Norway that was developed by the Oslo municipal energy works in the 1970s, against strong protests. The Mardoela River has the highest waterfall in Norway (and possibly Europe), Mardalsfossen, 297 m free fall. It was developed by the State Power Company around 1970 as part of the Grytten energy works. The water was transferred to the neighbouring valley, and the waterfall only has compensation flows during the tourist season. The conservationists, dominated by urban and academic people, joined the locals who opposed the transfer scheme. For the first time civil disobedience in the form of nonviolent sit-downs and chain-gang action was used in the environmental struggle in Norway, and the actions triggered a lasting discussion on the justification of civil disobedience.

The Sami issue The Sami is an indigenous people native to parts of Norway, Sweden, Finland and the Kola peninsula of northwest Russia. The total population is of the order of 70 000, with more than half living in Norway and the largest concentration in the county of Finnmark. They speak three

Case studies 133

languages belonging to the Finnish-Ugric group of languages. Although the languages are related to Finnish, the people and their culture are not. Their traditional lifestyle was hunting and trapping, with an increasing element of reindeer herding. The Sami society has been organized in "siidas", i.e. small communities with common use of land. The Sami people were recognized as a separate "nation" and taxed by the kings of Denmark, Norway and Sweden, and the Tsar of Russia, with occasional disputes and even wars over the delineation of the national borders in the area. "Southerners" started immigrating into the area, and encroaching on Sami territory, especially the arable areas, from 1600. The pressure on the Sami lifestyle, culture and language grew. A Sami revolt in Kautokeino in 1852 cost eight lives, including two Sami leaders who were executed. During the second half of 1800 into this century the strong national movement in Norway led to the suppression of the Sami language and Sami identity. After World War II this tendency changed, and there has been a growing nationalism in the Sami communities. Most of the Sami people are today living in mixed communities with a modern lifestyle, but a number of families still live by reindeer herding, and they constitute the core of Sami culture. The total number of reindeer herders is around 2000. The Alta River basin is to a large extent a heartland for the Sami. For more information on the Sami see, for instance, Helander (1992).

District vs. central government The "periphery" has a relatively strong political standing in Norway, with local self-government on the municipal level, and over-representation for sparsely populated areas in Parliament. Nevertheless, a growing polarity between the periphery and centre over the last decades can be discerned, and the tendency is strongest in northern Norway. To some extent this started just after World War II. When the Germans were forced out of Finnmark by Russian and Norwegian troops in 1944, they applied "burnt earth" tactics. The population, except those who hid and barely survived a harsh winter, was evacuated to southern Norway, and all buildings and infrastructures were burnt and destroyed. The postwar government, being worried that the scattered settlement structure, with small towns and fishing harbours, would become inefficient in supporting modern fisheries and industry, planned to resettle Finnmark with a more centralized structure. The inhabitants, however, liked neither the plans nor intended to wait for the Government to finalize them. They moved back en mass in 1945, and started reconstruction of their old homes with whatever was available. The Government was taken aback, and did not manage to get the supply lines running as fast as needed. Needless to say, the Finnmark people had another hard winter—and to some extent Oslo was blamed for this.

In the 1960s, the development of a modern fishery fleet made the structure of small harbour and fish processing plants in northern Norway obsolete. Small-scale farming, often combined with fishing, provided incomes that

134 Case studies

could in no way compete with the booming postwar industrialized society. As a result an increasing number of people moved from the districts to larger towns in northern Norway and to south Norway. Many felt that this was part of a central scheme to "centralize" Norway, and it kindled scepticism towards central government and strengthened the radical movements among the youth—in synergy with the 1968 political radicalization. It should be mentioned that there were strong traditions of communism in the area, which made this more than a youth revolt.

These tendencies were manifested and strengthened by the 1972 referendum on Norwegian membership in the European Economic Community. The margin against was only a tenth of one percent, but the general tendency was: centrum for, the periphery against, with northern Norway essentially en bloc against membership. The campaigns before the vote strongly polarized the existing differences.

5.2.4 The Alta hydropower scheme

The industrialization of Finnmark lagged behind the rest of the country, and after the war it was envisioned that a necessary condition to improve on this was to provide more energy. Hydropower development had strongly contributed to industrialization in southern Norway. Up to the 1970s, Finnmark was not connected to the national power grid, so all energy had to be produced locally. Even when it became connected, the transfer capacity was limited.

The Alta River had been considered an interesting river for hydropower development from earlier times, due to the concentrated fall at Sav'co. The first plans were made in 1922, but they had a much higher production than was necessary at that time. A couple of sketches were made in the 1960s, while a more detailed plan was published by the State Power Company in 1968. This was based purely on economic considerations, and had a total annual production of 1150 GWh. The main elements in this plan were as follows:

• utilization of the river reach from Sav'co to Kautokeino (90 km) as a reservoir;

• regulation of several lakes in the catchment, including the largest, Stuorajav'ri;

• regulation of the largest lake on Finnmarksvidda, Iesjav'ri, with transfer of the water to Jota'kajav'ri on the Alta River;

• two hydropower plants in Sav'co, two between Sav'co and Kautokeino.

The plan is shown in Fig. 5.2, from Koeber (1981). Extensive reservoirs were needed to give an acceptable firm power production, as the natural

Case studies 135

20 km

Fig. 5.2 The plan from 1968 (from Koeber, 1981).

runoff pattern is very seasonal in the Alta River. Of course, the result would also be large changes of the runoff regime in the lower stretches of the river.

What immediately caused an outcry was that the Sami village of Masi with 400 inhabitants and most of the agricultural land in the municipality of Kautokeino would be submerged. This of course upset the Sami, especially so as it was without precedent; very few, if any, Norwegians have been permanently relocated as a result of hydropower development. The question naturally was raised whether the Sami people were still considered second class by Norwegian officials. In addition to this, several important reindeer grazing areas would be affected by the regulation of the lake and the construction works. The plan made the Sami generally hostile to the idea of hydropower development on the river.

It soon became clear that the idea of damming Masi was politically unacceptable. It was dropped in 1971, and Masi was formally preserved against hydropower development in 1973. In 1974 the State Power Plants formally applied for regulation of Alta with a reduced plan: three reservoirs

136 Case studies

(Vird'nejav'ri, Iesjav'ri and Jota'kajav'ri), and two powerplants in Sav'co. Annual production was 860 GWh. During the executive work on the application it was decided to drop the regulation of Iesjav'ri from the plans, and a revised application was submitted in 1976 with annual production 710 GWh. The reasons for dropping Iesjav'ri were partly Sami protests and partly the rich fisheries in this lake. Another problem with Iesjav'ri was that the scheme included water transfer from the Tana River (the other large river on Finnmarksvidda) to Alta. As Tana is a border river with Finland, this would affect Finnish interests, and make the compensation discussions complicated.

During the discussions in the Norwegian Parliament, Jota'kajav'ri with the associated power plant was dropped, and a license was given for the development of the Sav'co power plant with the Vird'nejav'ri Reservoir in November 1978. A claim for reopening of the case was rejected in June 1979, and funds for the construction were granted later that year. The final plan is shown in Fig. 5.3, from Olje- og Energidepartementet (1978).

The final plan was for an annual production of 625 GWh, a little more than half that of the original plans. The total reservoir capacity was to be 135 million m 3 , approximately 6% of the annual runoff, and the submerged area was to be 2.8 km 2 . The gorges between Vird'nejav'ri lake and Sav'co were to be submerged, and the river was to be redirected for 1.7 km below the 110 m concrete gravity arch dam.

To reduce adverse effects on the salmon fisheries and in consideration of the potential for ice problems in the lower Alta, the plant is subject to strong regulation restrictions. The summer discharge is only allowed to deviate 10% from the natural runoff in the period from when the reservoir is filled until the 1 September. The winter operation is also strongly restricted.

5.2.5 "The Alta affair"

Although the hydropower developers had been losing ground throughout the 10 years of planning, and the final plans were more in line with what was considered acceptable for both the Sami, the locals of Alta, and most of the environmentalists at the start of the planning project, the Alta hydropower development had now become a matter of prestige and a case of large symbolic value for the parties involved. The decision to start construction triggered hunger strikes by Sami activists in Oslo, large demonstrations in Oslo and Alta, and mobilization for "River Saver" camps in the area where the access road was to be built. The establishment answered with a massive mobilization of police forces, including mooring a hotel ship in Alta to provide police accommodation and a request to the Defence Forces for helicopter support. Alongside the civil disobedience, a parallel struggle was

Case studies 137

20 km

Fig. 5.3 The finale scheme (from Koeber, 1981).

fought with legal actions to stop the construction, and claims for more impact assessments. Occasionally the latter line could sometimes be justified, but often it had the nature of filibustering tactics, and this was emphasized by the fact that most of the academic community was against the Alta scheme.

After the initial sit-down and chain-gang protests to stop the construction of the access road in the autumn of 1979, the construction was halted until January 1981. New sit-down protests led to 900 activists (about 30 locals) being taken into custody and fined. Some received short sentences. There were new delays on the road construction until the autumn of 1981, when it was restarted. Some of the activists were quite militant, and in March a Sami activist tried to blow up a bridge on the access road. He was badly hurt in the attempt. The leading organization, "The Civil Action against the Alta Hydro-power Development", was strictly nonviolent, and the leaders were exhausted after the prolonged struggle. The organization decided to dissolve itself during the winter of 1982. The civil disobedience died down, and the construction of the dam and the power plant proceeded without much interference.

138 Case studies

The Alta power plant started production in 1987 just in time to hit the market in a period of low consumption due to the general economic downturn, and record high inflows in the Norwegian hydropower system. As a result, the electricity prices were rock bottom for the rest of the decade, and the local partner in the Alta Hydropower Plant, Finnmark Energy Works, had to sell its shares to the State Power Company to survive. This gave the opposition, who had claimed that the Alta development was unnecessary, fuel for its arguments. Even the Prime Minister stated in public that the Alta plant should not have been built. Until now the figures have been in the red, but at present the energy balance of Norway is changing and electricity prices are escalating.

5.2.6 Impacts of reservoir development

Ecological impacts The environmental impact assessments of the Alta scheme were fairly extensive for the time, but somewhat unsystematic, and to a large extent aimed at making inventories of the state of the environment. The focus was on the effects on the salmon, secondly on the vegetation along the shores of Vird'nejav'ri and the Sav'co gorges that were to be submerged. The critical issue for the salmon was how the reservoir was going to be operated. Increased traffic and pressure on nature in the Sav'co area caused by the access road were considered potential problems.

The experience so far indicates that there are some problems in the upper salmon reaches—see more on this below. Increased algae growth in the river just below the plant has been observed. The access road has brought more traffic into the area, although the access is limited. The increased traffic is hardly a big problem; the main ecological problem in the area today is wear caused by the too large reindeer herds and the motorized terrain transport of the herders.

Morphological and geophysical effects The main geophysical problems were considered to be increased danger of ice jams in the lower reaches of the river, and erosion in the river bed. Local climatic changes were also central in the discussion, mainly changed drainage of cold air in the Masi area, and frozen fog forming in winter over open water below the power plant. Of these issues, only the ice problems have drawn any attention after the construction. The winter operation of the power plant has been decided in cooperation with the Hydrological Department at the Norwegian Water Resources and Energy Administration. In contrast to many earlier hydropower schemes, landscaping was used extensively to reduce the visual impact of roads, landfills etc. (Hillestad, 1993).

Economic effects The economics of the power plant has, as stated above, been quite shaky. Although some of the forecasts for power consumption

Case studies 139

given in the 1970s were too high, both for the region and the nation, the poor economy has to a large extent been due to an adverse combination of hydrological and economical fluctuations.

Locally the economics of the scheme have been positive. The salmon fishing has not been influenced to the extent that income has been noticeably affected, and tourism has probably increased as a result of the hydropower development. Alta became highly popular during the Alta struggle as a travel destination, and, in addition, the dam and the Sav'co area have become a popular tourist attraction, easily accessible by the new road. Reindeer herding has not been significantly affected, and the municipalities of Alta receive significant revenues from the hydropower plant, especially Alta. A number of jobs have been created in Alta by the power plant.

Socio-cultural effects The socio-cultural effects overshadow the environ­mental and economic effects of the hydropower development. One side is the general effects the Alta struggle had on the Norwegian public debate, focusing issues like the rights of the Sami people, nature conservation, civil disobedience, confidence in experts, welfare development vs. environmental protection etc. These themes were already under discussion, but the Alta case highlighted and intensified them. Even more important are the impacts on the Sami situation.

In a sense, the Sami came out as the winners of the Alta case. As mentioned above, there had been a change in the attitude to Sami rights after World War II, and there was a growing feeling of Sami identity and self confidence in Sami groups even before the Alta case, but the Sami politicians very efficiently used the Alta case for what it was worth to put Sami issues on the political agenda. One important principle had been established before the Alta case: a court verdict on another hydropower scheme, Altevann in Troms County, had given Sami use of common (state owned) land comparable standing to proprietary rights. When the Alta development scheme was strongly modified and reduced in extent during the planning stage, it was mainly due to considerations of the Sami use of the areas (and Sami opposition to the plans). In 1987 the Sami Act was ratified, giving a legal foundation for specific Sami issues, and establishing the Sami Parliament, an advisory assembly with a strong political influence, and power of decision in some fields. The Norwegian constitution was modified in 1988 to explicitly recognize the Sami people. In 1996 a new Reindeer Herding act was ratified, which grants Sami herders wide privileges in the mountain areas, even on private land.

On the cultural side, the development has been equally favourable, and Sami handicraft, music and other cultural manifestations are well established both regionally and nationally. A Sami theatre group, Beaivvas, was established in Kautokeino in 1979 in the midst of the Alta struggle, and it now has the status of a permanent theatre. The Sami language has been recognized as an official Norwegian language. A Sami regional college was established in Kautokeino in 1989.

140 Case studies

5.2.7 Design and management of reservoirs

Hydrological input The hydrological data available for the design of the Alta reservoir were in principle quite adequate, with a runoff series from the lower stretch covering the period from 1915 to 1969. Several new stations were established in connection with the hydropower planning. However, it became evident at a late planning stage that the original runoff series was unreliable on winter low flow data (many Norwegian series are, due to ice problems), and the firm power estimates had to be adjusted during the final design.

Although the operation of the Alta reservoir in principle is straightforward, a combined system of hydrological forecasting and power generation optimization has been set up for decision support. The main purposes are to improve flood handling and to optimally utilize the little leeway left within the strict operation rules. The system is based on telemetering meteorological stations, quantitative forecasts from the Norwegian Meteorological Institute, historical meteorological series for long-term planning, and hydrological models. Release scheme suggestions are established by a linearized optimization model. The actual operation is, however, not based on economic criteria alone within the existing restrictions (see below), and this limits the practical use of the optimization part.

Reservoir operation The actual operation of the reservoir is influenced by the following factors: reservoir inflow, reservoir level, energy demand (through energy prices), the legal operation rules, downstream ice conditions, and downstream ecological considerations. The two last items are not necessarily covered by the operational rules. Winter releases are decided in cooperation with external advisors, and the operation strategy has been adjusted to abate negative ecological impacts that have been identified after the start of the plant. To cater for such problems the operator has two mechanisms: to adjust the release strategy within the operation rules, and to manipulate temperature by switching or mixing between the two intakes in the reservoir, a top level intake and a deep intake.

Better use of the reservoir—ecological aspects The salmon and the salmon fishing in the river have been closely monitored. Generally the catches have been good, but a problem has been identified in the top reaches, the Sav'co area. In this stretch, which was earlier considered one of the best fishing places in the river, the catch is definitely reduced. The investigations indicate that this is a result of reduced parr production in the reach (Atlantic salmon head for the pool in which they were hatched). A possible explanation for this is that there have been rapid water level fluctuations in this stretch, due to a moderate day/ night peaking practice, and because the power plant has been used for automatic electric frequency stabilization on the grid. Such variations

Case studies 141

can leave parrs stranded. Further down the river such variations are evened out. After the discovery of reduced parr production both these practices have been abandoned.

Another possible explanation for the problems in the upper reaches is an increased growth of waterweeds (periphytic algae) that reduce the sight and thus the feeding possibilities for the parr. One solution is to reduce the weed growth through manipulating the water temperature (by using the double intake).

5.2.8 Comparative assessment of the reservoir with non-reservoir alternatives

With a reservoir capacity of only 6% of the annual runoff and the strict operation rules, the Alta hydropower scheme comes close to a run-of-river power plant, and there were some discussions about dropping the reservoir completely. The reasons for keeping it were that the increased winter production is valuable, and that the environmental gain of leaving it out was quite marginal. It is, however, not clear how thoroughly this alternative was investigated. Probably the planning was still influenced by the original concept that Finnmark primarily needed firm power and winter production.

To calculate the actual economic value of a hydropower project like this, with or without the reservoir, is straightforward—see for instance chapter 9 in Killingtveit & Saelthun (1995). If the system is small and does not to any noticeable degree influence the operation of the existing systems, the production value can be calculated by running operation simulations against simulated price series. For a larger system, it may be necessary to include the power plant in a nationwide energy production model, like the Norwegian "Power Pool Model" (Johannesen, 1993). A problem in a nearly pure hydropower system like the Norwegian system is that the economics of a single system with low or no storage capacity might be good, but the combined effect of many such systems may reduce the overall performance of the total system.

5.2.9 Sustainability issues

The main criterion for sustainability of a project like this is the presence or absence of irreversible adverse effects. In the 1970s, the term sustainability was not invented, but looking back at the planning process, from the first plan to the final version, we can see that the discussion in reality centred on sustainability issues. The first main issue was the submergence or survival of the Masi village, and with that the agriculture in Kautokeino municipality. When this was settled, the questions that guided the further planning process were:

142 Case studies

• sustainability of the traditional Sami lifestyle—reindeer herding and inland fisheries in the area;

• sustainability of the salmon fisheries on the Alta River.

In these respects, the final solution satisfies the sustainability criterion: no significant, irreversible adverse effects have emerged so far. The other questions that were raised were, to a large extent, conservation issues: whether there were rare species in the submerged areas, and whether unique prehistoric sites would be destroyed. As far as the current investigations can tell, no such damage has been identified, but of course the question always remains whether such investigations are thorough enough. The main irreversible element is of course the dam itself, and the change from an unregulated river to a regulated one.

An interesting observation is that the modifications of the plans came about as a result of the ordinary planning process, and political decisions under the influence of the public debate. The Alta actions and the civil disobedience did not have any influence on the realization of the power plant, except for delaying it by a couple of years. The actions may, however, have had other, long-term effects on public opinion and attitudes.

5.3 ASWAN HIGH DAM

5.3.1 Introduction to the Aswan case study

Egypt's hydrogeopolitical situation is unique. The Nile River constitutes the only significant water resource for the entire country. Few other insignificant water resources are available such as groundwater in the desert, and scarce rainfall at the northern boundaries of the country. As the most downstream country in the Nile basin, Egypt has no direct control on its main water resource. Egypt's land resources are no better than its water resources. About 4% of its territory is arable while 96% is barren desert. Population growth and the quest for economic development put these two crucial resources, water and land, under great stress.

Since the early years of the 20th century, several development plans for the Nile River have been suggested. These plans included different seasonal storage, over-year storage, and channelization schemes that involved more than one country in the basin. In 1952, Egypt adopted the idea of building one large over-year storage reservoir within its borders to face its critical present and future hydrogeopolitical situation. The suggested location for the Aswan High Dam (AHD) was upstream from Aswan cities and the Old Aswan Dam (OAD). The annual storage of OAD after its second heightening in 1933 was five billion cubic metres (bcm), but this storage capacity could

Case studies 143

not secure the increasing needs of socio-economic development of the country.

The AHD was built not only to meet Egypt's agriculture expansion plans, but also to achieve other objectives such as satisfying industrial power demand, flood control, and navigation improvement. Shalaby (1993) summarizes the main objectives of the AHD as follows:

• optimizing and rationalizing the Nile flow at Aswan;

« regulating and controlling the daily, monthly and yearly discharge downstream of the dam to match the actual water needs;

• protecting the Nile Valley and Delta from hazards of high floods and perils of droughts;

» generating cheap and clean hydroelectric power, sorely needed for development;

• increasing the limited cultivated area by horizontal land expansion and reclaiming new lands;

• changing the system of basin irrigation (one crop per year) to perennial irrigation (two or more crops yearly);

• expanding rice and sugarcane cultivable areas;

» improving navigation through the Nile and navigable canals;

• creating a greater flexibility in agricultural planning and crop patterns.

Construction work on the AHD started in 1960 and finished in 1970. At that time the dam was considered, and maybe still is, one of the most sophisticated monumental civil engineering works. The impounded reservoir upstream of the AHD is one of the largest manmade lakes. At the level of 182 m a.m.s.l . , which corresponds to the maximum storage capacity (162.3 bcm), the surface area of the lake is 6540 km 2 . The ratio between storage capacity and the corresponding surface area or the storage mean depth for the AHD is 24.82 m. The AHD Lake has a length of 500 km and an average width of 12 km at the maximum storage level. About 300 Ion of its length lie within the Egyptian borders, while 200 km are within Sudanese territory. The total storage capacity of the reservoir is allocated as follows: 90 bcm as live storage capacity between levels 147 m and 175 m a.m.s.l.; 31 bcm as dead storage for sediment deposition; and 41 bcm as a flood control buffer between the levels of 175 m and 182 m a.m.s.l.

The hydroelectric power station consists of 12 units with a capacity of 175 MW each, i.e. a total capacity of 2100 MW, producing on average 7 billion kWh of energy annually. After the heightening of the OAD a hydropower plant (Aswan-I) was constructed to utilize the head potential at the dam site. Since Aswan-I could not accommodate all the AHD releases for power production, a new hydropower house (Aswan-II) was added. It has four

144 Case studies

turbines, 67.5 MW each, which brings the total AHD capacity of power generation up to 2370 MW.

Egypt has devoted an entire research institute to monitor and study the impacts of the AHD. This institute is the Nile Research Institute (NRI). During the last three decades a lot of information and knowledge about the AHD reservoir and its effect on the downstream system have been accumulated. It should be noted that the scope of this case study is not intended to address all the impacts due to the construction of the AHD. Only some of the issues discussed in the previous chapters will be addressed with the focus on two topics:

• inundation of populated areas; and

• sedimentation problems.

5.3.2 Inundation of populated areas

Nubia is the name of the area that was subject to inundation due to the construction of the AHD. As shown in Fig. 5.4 it extended from the north near Aswan cities in Egypt down to the third cataract of the Nile in the south of Sudan. The maximum inundated area by the AHD reservoir inside the Egyptian borders is about 0.4% of the total area of Egypt.

More than 100 000 Nubians occupied this inundated area before the construction of the AHD; some of them lived inside the Egyptian borders while others were settled in Sudan (Said, 1993a). Although the Nubians lived in two different states, demographically, they are considered as one community. Their unique cultural heritage distinguished them from both the Egyptians and the Sudanese (Berg, 1976). Their economic activities were limited to agriculture, herding, and fishing. They used to live in small scattered communities that yearned for services regarding health, roads, social care, and schooling. The Nile played a major communication role among these communities. At the time of the AHD construction, northern parts of Nubia had already suffered from the inundation of their land and properties due to the increase in height of the OAD.

In 1963 the entire Nubian population (Egyptian and Sudanese) that was subjected to resettlement represented about 0.4% of the Egyptian population. The Nubians resettlement plan was very successful according to Said (1993b), Shalaby (1993) and Abu El-Wafa (1962). About 55 000 Egyptian Nubians were resettled in the vast fertile plain of Kom Ombo 60 km downstream of the AHD, called New Nubia. In the Kom Ombo area, 28 000 acres were reclaimed and provided with the required irrigation infrastructure. The resettlement plan included more than 25 000 houses in 33 villages that combined both modern public utilities and Nubian architecture in their design.

Case studies 145

Nearly 53 000 Sudanese Nubians were resettled in Khashm el Girba, in the east of Sudan. A new irrigation scheme called new Haifa was developed for them. Although migration to the north was very well planned, the immigrants to the south were luckier in terms of the compensation they received and services provided during the transportation and early phases of the immigration process.

The Egyptian Government in 1962 estimated the Egyptian Nubians loss of property to be six million Egyptian Pounds (L.E.). Nearly 50% of indemnities were paid in advance to help the Nubians move to their new settlements. The other 50% were used to pay part of the cost of land, infrastructure and houses. The rest of the cost is being paid in 40 annual instalments without any interest (Abu El-Wafa, 1962). According to the 1959 agreement between Egypt and Sudan, resettlement of the Sudanese Nubians was the responsibility of the Sudan Government. Nubians, as a distinctive community, being deeply attached to their lands, lost something they cannot be compensated for and suffered psychologically. "After 25 years, the social impacts of migrants have been remarkable. They were mixed with people of Aswan and Qena provinces and got used to their culture and traditions and were thus leading normal happy life" (Shalaby, 1993).

Nubia had monuments, temples, tombs, fortresses and other remains from Pharaohnic, Ptolemaic and Roman ages. Most of these treasures were already seasonally submerged due to the heightening of the OAD. Only during low storage seasons of the OAD did these historical assets re-emerge from the water. The AHD threatened these ancient sites, including the rock-cut temples

Fig. 5.4 Map of the AHD Lake and Nubian land.

146 Case studies

of Abu Simbel, with permanent loss. In response to this threat an unprecedented international campaign to salvage the Nubian legacy was set up by UNESCO. More than 25 countries including the USA, Germany and The Netherlands contributed to the salvage effort either through funds or experts. The Egyptian Government, in its appeal to UNESCO to preserve one of the most important human legacies, promised to give 50% of the duplicated discovered antiquities to the participating foreign expeditions. Also five Nubian temples were given to countries who contributed to the large-scale salvage programme (Said, 1993b).

Most of the monuments and temples were preserved by dismantling and reconstructing them on higher grounds or islands. Special attention was given to the Abu Simbel temples because they were cut in solid rock. Dismantling and reconstructing these two temples are still considered one of the contemporary engineering miracles. This miracle cost about $40 million. Moving Philae temples was another monumental accomplishment.

The salvage campaign of the Nubian legacy can be considered without doubt a milestone in history of international cooperation. Many archaeological surveys and discoveries would not have been made, had the site not been subject to development, and they were in fact made and shared with other countries. Present and future generations are and will be enjoying the Nubian legacy for a long time, and inter-generational equity will in this context probably be achieved.

Inundation of large areas after construction of a dam is considered one of the indications of the environmental destruction. Section 1.2 compares the largest 10 dams in the world with two Russian dams on the Volga River, and 2575 dams in Japan. If the mean storage depth is applied as a criterion to measure the storage efficiency of a specific dam location, then the storage efficiency of the AHD corresponds with most of the other large dams in the world. The AHD follows the average relation V = 9.2A11, where A is inundated area and V is storage volume, obtained by analysing 7936 reservoirs distributed worldwide.

Goodland et al. (1992) justify the loss of tropical forest by inundation due to dam construction if the rate of power production per area flooded is 100 kW ha"1. They suggest lower rates if the flooded area is used for agricul­ture or degraded land. In the case of the AHD, the ratio between maximum power production in kW and the maximum surface area in ha is 3.63 kW ha"1. It is a very low ratio, but it should be noted that power generation is not the single purpose of the AHD. Moreover, the percent of the hydropower generated from the AHD and the OAD (Aswan complex) to the total national power production is declining. It has declined from 70% in 1975/1976 to 24% in 1990/1991. Currently, hydropower generation from Aswan complex is a by-product, which does not influence the reservoir operation.

Goodland (1996) suggested two ratios to judge the sustainability of several reservoir sites. One of them is the inverse of the previously suggested ratio

Case studies 147

and the second ratio is the number of persons involuntarily resettled per output of electricity (oustees MW" 1). He used the normal reservoir operating area in his analysis rather than the maximum inundated area. Accordingly, the ratio of maximum power production to the normal operating area of the AHD will improve to 5.93 kW h a 4 . Although he noted, in his analysis, the difference between multipurpose reservoirs and reservoirs built mainly for power generation, he did not include that in his evaluation of the sustainability. He also ignored the nature of the inundated land as he pointed out in an earlier publication (Goodland et al., 1992). However, AHD sustainability cannot be simplified and evaluated by such ratios. Since Nubia was basically a desert with few small scattered patches of seasonal agriculture, the low ratio of 5.93 kW ha"1 may be justified. On the other hand, inundation of Nubia as an extremely valuable archaeological site may not be justified even at a higher ratio than 100 kW ha"1.

5.3.3 Sedimentation problems

The sedimentation problem of the AHD has many dimensions. The annual flood of the Nile used to convey on average about 134 million tonnes, equivalent to 92.4 million cubic metres (mem) of silt to Egypt (Abu El-Atta, 1978). This silt was assumed to be the main contributor to Egyptian land fertility, a source of building material, and a carrier of nourishment for sardine fish in the Mediterranean near the Nile estuaries. It was considered also a source of instability of the Egyptian irrigation canals. Silt sedimentation in the irrigation canals was a costly problem which the system operators used to face every year. A sudden prevention of the annual silt flow to the Nile Delta and Valley brought forward all these issues, besides reservoir siltation, downstream morphological instability, and coastal erosion. Many pre-design and post-operation studies have focused on the sedimentation problems of the AHD. Reservoir sedimentation is a significant factor in determining the reservoir lifetime which is related to sustainability. As the reservoir lifetime increases, the reservoir is considered to be in a more sustainable way of development. Evidently human needs of future generations can be met if the reservoir lifetime is very long or can be prolonged. Most of the studies for the AHD reservoir in this regard, and also the field measurements concluded that the dead storage (31 bcm) of the AHD reservoir will take 400 years or more to fill. All of the sediment received up till now has accumulated at the upper part of AHD Lake (200 km upstream of the dam) as shown in Fig. 5.5 (NRI, 1993). Specifically, Makary (1982) expected it to take 408 years to fill the AHD dead storage. Field measurements by the High Dam Public Authority from 1978 to 1990 showed that the sediment is accumulating at an average rate of 109 mem per

148 Case studies

Fig. 5.5 Longitudinal section in AHD Lake showing deposited sediment progress (NRI, 1993).

year (Said, 1993b). If the lifetime is the only criterion for dam sustainability, AHD would rank very high among the world dams.

Total agriculture land at the construction time was estimated to be six million feddans (one feddan = 4200 m 2 ) . Arable land being a limited resource (4% of Egypt's area), agriculture productivity and sustainability have been major concerns before and after the AHD construction. Therefore, the effect of silt deprivation on the fertility of the agricultural land was thoroughly investigated and monitored. Most of the silt was received by only 700 thousand feddans that were under basin irrigation out of the six million feddans. The rest of the cultivated area was under perennial irrigation that allows for minimum sedimentation of the silt. This fact and the fact that the plant usable nutrients load of Nile silt is very low (0.04% of the total silt load) made the effect of silt deprivation insignificant. Replacing this small amount of nutrients by artificial fertilizers, which for a long time had been used in Egyptian agriculture practice, solved the problem.

The real threat to agriculture sustainability emerges indirectly from withholding silt upstream of the AHD reservoir. Although the silt deposited in irrigation canals before building the AHD constituted a costly problem, it was the source of raw material for brick manufacturing. All other alternative raw materials such as clay, sand, limestone, and cement, were not commonly used in the brick industry. The sudden loss of silt forced the brick manufacturers to find other sources of material. The obvious and easiest solution was the removal of agricultural topsoil and destruction of the Nile banks. Successful

Case studies 149

legislative measures were taken against this destructive process and incentives were provided to the manufacturers for a transfer to other raw materials. The prediction of Abu El-Atta (1978): "we will be able to overcome this problem very soon" became true.

The most evident adverse impact of the AHD on ecological sustainability is the disappearance of sardine in the Mediterranean near Rosetta and Damietta Nile estuaries. Most of the silt carried by the Nile flood was flushed to the Mediterranean through the Rosetta and Damietta branches. The total amiual catch of sardine diminished from about 18 000 tonnes before the AHD to 500 tonnes in 1966, and never recovered (ASRT, 1990). The flushed sediment obviously was loaded with attractive nourishment for the sardine species. Although the increase of fish production due to the creation of the AHD lake may counter this loss economically, it may not justify the resulting ecological destruction.

The condition of ecological sustainability that requires preservation of the minimum number of individuals to ensure the natural evaluation of each species is totally violated in the case of the sardine species. On the other hand, 20 years after the construction of the AHD, the total number of fish species has increased in the Nile and its estuaries (Biswas, 1992).

One of the major environmental concerns of retaining the sediment upstream of the AHD is the degradation and erosion that might be caused downstream of the dam. Estimates were made at several locations; at Gaafra and at three regulating barrages located downstream of the OAD. The assumptions and approaches used to obtain these estimates differed significantly as well as the estimates themselves. It is very hard to compare them to each other or to the actually observed drop in water elevation at the different monitoring locations. In Table 5.1 El-Moattassem & Abdelbary (1993) showed the drop in water levels compared with 1963-observed levels, as an indication of the downstream degradation caused by building the AHD.

Their general conclusion was that the Nile bed degradation after the construction of the AHD was limited due to the controlled discharge and the existence of intermediate barrages. The replacement and rehabilitation of the

Table 5.1 Water level drops (cm) at different stations for different discharges downstream of the Nile barrages from 1963 to 1990.

Station Distance Discharge (mem day"'): downstream AHD (km) 80 90 100 120 150 200 250

Gaafra 34 80 76 73 72 60 50 37 Esna 167 73 75 77 77 73 47 -Nag Hammadi 359 84 88 90 85 75 - -Assiut 539 60 63 62 57 56 - -

150 Case studies

old barrages due to ageing may eliminate most of the degradation impacts and the effects of the drop in water levels on navigation.

5.3.4 Conclusion on the Aswan case study

This case study is a discussion of some ideas introduced in the previous chapters, to derive a set of criteria for reservoirs to be considered as a means of sustainable development. It should be emphasized that the AHD case study, presented here, is not by any means a comprehensive evaluation of its sustainability. The AHD as one of the 10 largest dams in the world has been subjected to a wild campaign of criticism which led to a very wide international debate. Participation of politicians, environmentalists, media, scientists, and technicians added to the controversy, rather than coming up with objective unbiased evaluations.

Nevertheless, there are some simple facts that created a general public sanction of the project at national level and tamed the international criticism. It is undeniable that the AHD saved Egypt from the socio-economic destruction that could have happened due to the persistent nine-year drought from 1979 to 1987. Although destructive impacts were clear in Ethiopia and other African countries, Egypt was not affected except for some reduction in power generation. It also guarded the country from the 1964, 1975, 1988 and 1996 high floods. The role of the AHD in protecting Egypt from these catastrophic events has been highly appreciated by the entire population especially the poor sectors who are the most vulnerable to the consequences of such events.

The AHD has proven to be a very economically efficient project. According to Shalaby (1993), a study for estimating the change in national income due to the presence of the AHD showed that during the first 10 years after construction the national income increased by 10 billion L.E. , that is 20 times the AHD cost.

It is obvious, without going into too many details for estimating the economic and social benefits of the AHD, that the intra-generational and inter-generational equity has been very well achieved. It was achieved because most of the benefits of the AHD, as a national project, targeted the majority of the Egyptians. These benefits were oriented towards rural communities where poor people live. The fraction of the resettled population was very marginal with respect to the total population.

In developing a list of criteria for reservoir sustainability the discussion should not be directed to comparing the sustainability of existing reservoirs to each other. The list of sustainability criteria should be chosen and derived carefully to characterize each case individually. The needs for development differ significantly from developed countries to developing countries and from arid zones to wet zones. Nevertheless, evaluation of the sustainability of

Case studies 151

existing reservoirs is quite important because it helps in consolidating and formulating the basic rules, principles and concepts of sustainable reservoir operation and management policies that can respond to changes in future human needs and objectives, and advances in technology.

The ratio between power production and inundated area may be used to measure sustainability of a single-purpose reservoir. If the reservoir is a multipurpose reservoir, then a similar criterion for each purpose or another criterion that lumps all the uses together could be derived. The phenomenon of agriculture soil removal resulting from the AHD construction, emphasized that some indirect consequences of reservoir creation can affect sustainability. Therefore, in seeking the objective of developing a comprehensive list of sustainability criteria, the effort should go deeper than the typical direct impacts of reservoir development. The impact of the AHD on fish species in the Nile system shows that positive effects may outweigh negative effects with respect to sustainability.

5.4 MANAGEMENT OF ANNUAL PEAK FLOWS TO RESTORE AQUATIC RESOURCES IN GREEN RIVER, UTAH

5.4.1 Introduction to the Green River case study

The Colorado River is one of the most highly regulated rivers in the world. The total usable reservoir storage capacity exceeds 70 billion m 3 , or approximately four times the mean annual flow. The four largest reservoirs in the Colorado River basin, each with a usable storage capacity of 2 billion m 3

or more, are shown in Fig. 5.6. The relatively large reservoir storage capacity was developed due to significant variations in seasonal and annual runoff, a semiarid climate which necessitates irrigation to raise crops, and a need to meet the legal division of water among the basin states. With the exception of relatively small tributaries, river channels of the Colorado River basin have been extensively altered by the construction and operation of dams. All of the major dams have been completed since 1936, and those upstream from Lees Ferry were only completed in the mid 1960s. The impacts of these impoundments on aquatic and riparian resources of the Colorado River have developed over a period of years and are only now becoming substantial environmental problems (Ward & Stanford, 1989).

A principal hydrological effect of the reservoirs, though by no means the only one, is a decrease in the magnitude and duration of spring flows and an increase of flow during the remainder of the year. A wide variety of physical and ecological effects have been attributed, wholly or in part, to reduced peak

152 Case studies

\ WYOMING

I Î j FONTENELLE

i v - l i r ~ - \ — ' FLAMING GORGE Y \ !

K * \

Salt Lake City

/ UTAH ' '

( \ ( \—••• (

NEVADA \ I <</̂ l BLUE MESA

\ / Vegas .• -3s f ' - v ^ . — « > . — ' • ' ^ " \ •' \ \ • .' . . - 7 : \ I NAVAJO

HOOVER I v - ' f

V ^ D A v I S -̂..ARIZONA - \ J ^ \ \ ^—.<£ ^ - - ^ 1 j NEW MEXICO

CALIFORNIA j ^ r a r k e r \ % ^ | k

T C l

A . .

flows (Stanford & Ward, 1986a). One effect, the severe decline in populations of four species of native fish, and subsequent protection by the federal government under the Endangered Species Act, could seriously limit or stop future development of water resources in the Colorado River basin. Accordingly, the desire to restore and protect riverine and riparian resources in the Colorado River basin has prompted investigations of the costs and benefits of changing dam operations and, in some cases, the reconstruction of

Case studies 153

dam outlet structures to increase the magnitude and duration of amiual peak flows.

The purpose of this section is to examine the physical and ecological basis for increased annual peak flows together with the legal and economic obstacles and opportunities. Currently, increasing annual peak flows is being studied for all major reservoirs in the Upper Colorado River basin. This article will specifically focus on the Green River downstream from Flaming Gorge Reservoir, because it has been studied in considerable detail. In addition, the physical, ecological, legal, and economic issues concerning the Green River are typical not only of the Colorado River, but of regulated rivers throughout the United States. Restoration and protection of riverine and riparian resources in the Green River downstream from Flaming Gorge Reservoir will certainly require more than an increase in annual peak flows. As demonstrated by numerous investigations concerning a variety of aquatic resources, a regime of annual peak flows, however, is essential, if the structure and function of the Green River ecosystem are to be restored. Increasing the magnitude and duration of annual peak flows faces many physical, political, and economic obstacles as well as ecological uncertainties. To date, these obstacles and uncertainties have impeded significant progress.

Authors' note During the spring of 1996, an experimental flood was deliberately released from Glen Canyon Dam for the sole purpose of restoring downstream aquatic and riparian resources in Grand Canyon National Park. This flood was the first of its kind in the United States. More than 125 scientists studied the flood's affect along a reach of almost 500 km. The relative flood magnitude, approximately half of the unregulated mean annual flood, and duration of seven days, were small compared to the proposed Green River flood discussed in this article. Furthermore, all of the affected lands are within the Grand Canyon and administered by the federal government. Thus, the economic and political obstacles to the flood, while substantial, were much less complicated than they would be for a Green River flood, as described below.

5.4.2 Historical development of Colorado River basin water resources

Construction of Flaming Gorge Dam and the resulting negative effects on the downstream riverine ecosystem are a direct result of western water law and a political climate that favoured the rapid economic development of a sparsely populated, semiarid region. In 1922, a compact by seven states divided the basin's runoff between the Upper Basin States (Colorado, Wyoming, Utah and New Mexico) and the Lower Basin States (Nevada, Arizona and California). The compact requires the Upper Basin to release 92.5 1 0 9 m 3 over any 10-year

154 Case studies

period to the Lower Basin. The division of water focuses on rights to consumptive use as the Upper Basin States may not withhold water, and the Lower Basin States may not demand water, which cannot be put to beneficial consumptive uses, which are defined as domestic, industrial and agricultural uses. This compact between the two basins was brokered by the United States, but it is consistent with state laws regarding water use.

Water law in the Western States, with a few exceptions, is governed by the doctrine of prior appropriation, granting water rights to the first person who uses water for a defined beneficial use. Water rights are retained only if the beneficial use continues and is not abandoned. Western water law reflected the western cultural value that water left in its natural water course was wasted (Getches, 1990). Thus state water law determines the ownership of water within the state boundaries, while the agreement among states, facilitated by the federal government, reflects the division of water among the states. The body of laws, contracts, court decisions, and administrative practice which govern the distribution, uses, and management of Colorado River water are known collectively as the Law of the River.

In order to provide for development of water resources in the Upper Basin, while meeting the compact obligations, the US Congress passed the Colorado River Storage Project Act (CRSP) in 1956. This act authorized the construction of Glen Canyon Dam, approximately 20 km upstream of the compact point between the Upper and Lower Basin, Lees Ferry, Arizona, as well as additional dams on the three major tributaries, the Colorado, Green, and San Juan. Flaming Gorge Dam is the principal facility on the Green River authorized by the CRSP Act. The primary purpose of these reservoirs was to provide a dependable water supply for irrigation, industrial and municipal users. Other purposes, hydroelectric power generation, recreation and flood control, were recognized, but were of secondary importance. The CRSP Act, however, required that the cost of the dam and associated facilities be paid for by revenues from the sale of hydroelectric power over a period of 50 years.

Subsequently, the Colorado River Basin Act, was passed in 1968. This act requires that participating CRSP facilities be operated to generate the maximum hydroelectric power consistent with other Congressional Directives. The several CRSP generating facilities are operated to meet peak daily electrical power demands. Flaming Gorge Reservoir has a total active storage of 4.67 1 0 9 m 3 . Since 1962, the mean annual release has been 1.79 1 0 9 m 3 . The power plant capacity is 150 MW. The mean annual electrical energy produced over the past decade is about 450 1 0 6 kWh, which has a current wholesale market value of approximately $9 million per year.

Most recently the Law of the River has grown to include laws protecting ecological and recreational resources. The Endangered Species Act, the Clean Water Act, and others require federal actions to consider and incorporate ecological values. The Endangered Species Act authorizes the Department of

Case studies 155

the Interior to identify species as threatened or endangered and thereby invoke the protections of the Act. One of the protections is that the activities of all federal agencies shall consider and provide for the conservation and recovery of endangered and threatened species (ESA Sec. 7(a)(1), 16 U.S.C. Sec. 1536(a)(1), 1988).

In 1993 the US Fish and Wildlife Service (USFWS) determined that the operations of Flaming Gorge Dam, and every major existing or proposed water project in the Upper Basin jeopardized or would jeopardize four species of native fish: the razorback sucker, humpback chub, bony tail chub and the Colorado squawfish. Thus, before allowing any additional water projects, USFWS required protection of the four endangered fish consistent with the Endangered Species Act. USFWS issued a Biological Opinion that made two substantive requirements:

• reoperation of the Flaming Gorge Dam so that the flow and temperature would more closely resemble the river's natural hydrograph, and

• legally protected instream flows from Flaming Gorge Dam downstream to Lake Powell so that adequate water is available for the four endangered fish.

In spite of the listing as endangered, and the conclusion that operation of Flaming Gorge Dam threatens the fish, progress in recovering the four native fish has been slow and neither of the requirements of the Biological Opinion has been satisfied (Bolin, 1993).

5.4.3 Hydrology of the Green River basin

The Green River drains 116 000 km 2 along the west flank of the Rocky Mountains in Wyoming, Colorado, and Utah (Fig. 5.7). Flow in the Green River has been regulated by Flaming Gorge Reservoir since October 1962. Fontenelle Reservoir, also located on the main-stem Green River, is the second largest impoundment in the Green River basin and was completed in April 1964. Tributaries to the Green River have numerous small impoundments, especially in their headwaters. Except for the Duchesne River, however, the effects of these impoundments are small and the tributaries are generally free flowing at present (1996). Several reservoirs have been proposed and are being considered on the principal tributaries.

The locations of the long-term gauging stations in the Green River basin are shown in Fig. 5.7. Water discharge has been recorded daily at most gauges for several decades. Extensive records of suspended-sediment concentration also have been collected at the gauges shown in Fig. 5.7.

The mean annual water discharge and sediment discharge prior to 1962 at gauging stations in the Green River basin are shown in Fig. 5.8. Water and

156 Case studies

112° 111° 110° 109° 108° 107°

Fig. 5.7 Upper Colorado River basin showing principal gauging stations and major reservoirs.

sediment are not contributed to the channel network uniformly across the Green River basin (Irons et al., 1965; Andrews, 1986). Furthermore, the principal source areas of water and sediment differ greatly. A majority of the annual runoff from the basin is supplied by the headwater areas. Conversely, the semiarid parts of the basin at lower elevations contribute most of the sediment. Immediately upstream from Flaming Gorge Reservoir, the unregulated mean annual flow was 44.6 m 3 s"1 or 27% of the total basin outflow. The mean annual sediment discharge at this gauge, however, was only 0.34 10 6 tons or 2.2% of the basin outflow. Similar contrasts exist throughout the Green River basin.

The large sediment-contributing areas in the Green River basin are indicat­ed, similarly, in Fig. 5.8. All tributaries joining the Green River downstream from the gauge near Green River, Wyoming, supply relatively large percentages of the basin's sediment outflow. In all instances, the sediment is contributed

Case studies 157

112° 111" 110° 109° 108° 107°

Fig. 5.8 Mean annual runoff and sediment load at selected gauging stations in the Green River basin prior to 1962.

primarily by the downstream part of the tributary drainage basins. Thus, the large water-contributing parts of the Green River basin lie around the rim, especially along the northeast divide. Conversely, the large sediment-contributing areas are located in the central and southern parts of the basin.

Flaming Gorge Reservoir is located downstream of the largest water contributing areas, and upstream of the largest sediment contributing areas of the basin. Consequently, the reservoir controls a substantial portion of the basin runoff, but only a relatively small portion of natural sediment yield from the basin is deposited in the reservoir. Approximately, 13.4 million tonnes of sediment per year, or nearly 80% of the pre-development basin sediment yield, are supplied by unregulated tributaries to the Green River downstream from Flaming Gorge Reservoir. Adjustment of the Green River channel to regulated flows has been substantially influenced by the relatively large downstream contribution of sediment.

158 Case studies

Except for relatively small evaporation losses from Flaming Gorge Reservoir, net annual water depletion (i.e. consumptive uses) in the Green River basin has not changed substantially since the construction of the reservoir. (Note: This situation is changing somewhat now due to the completion of the Central Utah Project, which is designed to divert as much as 260 10 6 m 3 of water annually or about 5% of the mean annual runoff.) Accordingly, comparisons of mean annual runoff at main-stem gauging stations for the pre- and post-reservoir periods of record show very little change. The mean annual discharge of the Green River at the Jensen gauge was 119 m 3 s~] from 1963 to 1995 compared to 122 m 3 s 1 from 1947 to 1962. Similarly, the mean annual discharge of the Green River at the Green River gauge was 154 m 3 s"1 from 1963 to 1995 compared to 165 m 3 s"1 from 1944 to 1962.

Flow regulation by Flaming Gorge Dam, however, has substantially reduced seasonal variations in flow. The magnitude and duration of annual peak discharge have decreased, while discharge during the remainder of the year has increased. Typical pre- (1949) and post- (1967) reservoir, annual hydrographs of the Green River at Jensen, Utah are compared in Fig. 5.9. River flows at the Jensen gauging station are a result of regulated flow in the main-stem Green River and the essentially unregulated flow of the Yampa River. Approximately, 60% of the contributing drainage area at the Jensen gauge is controlled by Flaming Gorge Dam. Since the construction of Flaming Gorge Reservoir the mean annual peak discharge has decreased by about 30% from 650 to 460 m 3 s"1. The effect of flow regulation extends relatively far

800

700

600

500

400

300

200

100

I I I I I I I

l G R E E N R I V E R A T J E N S E N , U T A H

1 I I I I.

\ A N N U A L H Y D R O G R A P H

~~ M/ 1 / 1967'

I I I I r 30 60 90 120 150 180 210 240 270 300 330 360

W A T E R Y E A R , D A Y S F i g . 5.9 Comparison of stream flows in the Green River at Jensen, Utah, during representative pre- (1949) and post- (1967) Flaming Gorge Reservoir conditions.

Case studies 159

downstream because tributary contributions of water are relatively small. Accordingly, the mean annual peak discharge has decreased by 29% from 880 to 625 m 3 s"4 at the Green River, Utah gauge, 470 km downstream of the reservoir during the post-reservoir period.

The durations of daily mean discharge for the pre- and post-reservoir periods are compared, Fig. 5.10(a) for the Jensen gauge, and Fig. 5.10(b) for the Green River, Utah gauge. The magnitude of flows that occur less than 10% of the time has decreased significantly since regulation of the Green River by Flaming Gorge Reservoir began. Furthermore, for a given duration within this range, the decrease in discharge is about the same at both gauges. For example, discharges equalled or exceeded < 5 % of the time at the Jensen and Green River, Utah, gauges have decreased by 25%. Thus, although the drainage area

5,000

1,000

I I ! I l I I i i i i i i i l ' i i

PRE-RESERVOIR

S

POST-RESERVOIR

- (a) i i i i i i i l l i i l l l l , l l

£ z

l lT O DC <

o w Q 5,000

10

< 3 Z z < z < LU

1,000

100

10

PRE-RESERVOIR POST-RESERVOIR

(b) J L _L _L _L J L

0.01 0.5 1 95 98 99 99.8 99.9 5 10 20 30 40 50 60 70 80 90

DURATION OF DISCHARGE (% TIME) Fig. 5.10 Duration of streamflows in the Green River before and after regulation by Flaming Gorge Reservoir at (a) Jensen, Utah, and (b) Green River, Utah.

160 Case studies

is 2.7 times greater at the Green River, Utah gauge, than it is at Flaming Gorge Reservoir, the effect of the reservoir is still substantial 470 km downstream.

The discharge at which flow fills the channel and begins to generally inundate the flood plain is the bankfull discharge. Low-lying parts of the flood plain, such as sections of abandoned river channel, however, may be inundated at discharges significantly less than bankfull. As described below, deposition of sediment and encroachment of vegetation has reduced the bankfull discharge of the Green River channel over the last 40 years. The bankfull discharge of the Green River has decreased from 720 to 610 m 3 s"1, downstream from the Jensen, Utah gauge and from 1080 to 930 m 3 s"1

downstream from the Green River, Utah gauge (Andrews, 1986). Despite the decrease of channel size, the frequency (annual series) and duration of flooding has decreased significantly along the Green River.

Regulation of flows by Flaming Gorge Dam has significantly reduced the frequency and duration of bankfull flow, and, thus, the frequency and duration of flood-plain inundation. At the Jensen gauging station, the recurrence interval of flood-plain inundation has increased from approximately every 2.2 years to 4.5 years on average. As shown in Fig. 5.10(a), the duration of bankfull discharge has decreased from 7.3 days per year to 1.5 days per year. Similarly, the recurrence interval of the bankfull discharge at the Green River gauging station has increased from once every 2.8 years to 8.3 years, on average. The duration of bankfull discharge has decreased from 7.3 days per year to 1.5 days per year (Fig. 5.10(b)). The decreased frequency and duration of flooding are significant, because the inundation of the flood plain provides aquatic habitat essential to native fish in the Green River.

Andrews (1986) computed pre- and post-reservoir sediment budgets for three reaches of the Green River, using measured daily water and sediment discharges. Prior to the construction of a dam in Flaming Gorge, a quasi-equilibrium condition appears to have existed downstream in the Green River channel; that is, over a period of years, the transport of sediment out of a given river reach equalled the supply of sediment into the reach. Since reservoir regulation began in 1962, the mean annual sediment discharge at downstream gauges has decreased substantially. The quantity of sediment in transport downstream of the Jensen, Utah, gauge at a given discharge, however, has not been affected by Flaming Gorge Reservoir. The decrease in mean annual sediment transport at the Jensen and Green River, Utah, gauges since 1962 is due entirely to a decrease in the magnitude of river flows. The decrease in mean annual sediment discharge at the Ouray and Green River, Utah, gauges far exceeds the quantity of sediment trapped in the reservoir.

Post-reservoir annual sediment budgets show that a majority of the Green River channel is no longer in quasi-equilibrium. Three distinct longitudinal zones involving channel degradation, quasi-equilibrium, and aggradation were identified. Beginning immediately downstream from the dam and extending downstream 110 km, sediment transport out of the reach exceeds the tributary

Case studies 161

contribution, and the channel is degrading. The length of the degrading reach, however, is relatively short, due to the large quantity of sediment supplied by tributaries.

In the reach between 110 and 270 km downstream from Flaming Gorge Reservoir, the quantity of sediment supplied to the reach from upstream plus tributaries approximately equals the transport of sediment out of the reach over a period of years. This reach appears to be in quasi-equilibrium, as there is no net accumulation or depletion of bed material. From 270 km downstream of Flaming Gorge to the mouth (river km 667), the supply of sediment from upstream and tributary inflow exceeds the transport of sediment out of the reach by 5.4 10 6 tonnes yea r 1 on an average.

5.4.4 Ecological and geomorphological alteration of the Green River

Since the completion of Flaming Gorge Dam in 1962, the character and function, both physically and biologically, of the Green River have changed dramatically. Nearly every aspect of the aquatic and riparian ecosystem has been affected. Furthermore, adjustment to the regulated streamflows appears to be continuing. These changes have been studied in considerable detail, and it is possible only to describe some of the most significant changes in this chapter. Andrews (1986), Andrews & Nelson (1989) and Lyons & Pucherelli (1992) describe modifications and adjustments of channel morphology and size. Graf (1978) and Fisher et al. (1983) describe changes in riparian vegetation. Stanford (1993) provides a comprehensive and detailed summary of physical and ecological changes that have affected native fish species. Nesler et al. (1988) and Tyus (1991) summarize our knowledge and understanding of the Colorado squawfish. Minckley et al. (1991) describe the ecology of the razorback sucker and propose steps towards its recovery.

Invasion and establishment of exotic species, terrestrial and aquatic, have contributed substantially to the alteration of the Green River. The relative significance of exotic species vs. regulated river flows is largely unknown. In most instances, the exotic species are better adapted to the reduced magnitude and frequency of flood flows than the native species, and, thus, their invasion is aided by a regulated flow regime. Throughout the Colorado River basin, healthy, viable populations of the native fish occur only where the natural regime of river flows has not been altered by regulation and depletion. Minckley & Mefee (1987) found that the number of species and individuals of exotic fish were substantially reduced during a flood with an exceedance probability of 0 .05%, while the native fish were essentially unaffected. The introduction and establishment of exotic species into the Green River system would have increased competition for the native species even without the

162 Case studies

storage and regulation of river flows. The effects of regulated river flows on native species are, in most instances, exacerbated by the invasion of exotic species.

The native fish of the Green River basin have evolved in a turbid river with large seasonal and year-to-year variations in flow. For a variety of reasons (Stanford & Ward, 1986b, Tyus & Karp, 1991; Stanford, 1993), aquatic ecologists and fishery biologists have concluded that backwaters along the channel margin and seasonally inundated parts of the flood plain are essential habitat for the endangered native fish. This conclusion, however, is based almost entirely upon observations since Flaming Gorge Dam was completed. No ecological studies of the native fishes in the Green River were conducted prior to 1962.

Channel margin backwaters and inundated flood plains are generally warmer, nutrient rich and more productive than the main channel. As described above, river flow regulation has substantially reduced the frequency and duration of flood-plain inundation. Accordingly, the available aquatic habitat for the native fish, also, has been substantially reduced (Minckley et al, 1991; Tyus, 1991). In addition, the remaining available habitat is shared with exotic species. Increasing the frequency and magnitude of seasonal floods in the Green River will increase available habitats for the endangered native fish, however, in all likelihood exotic fish will also occupy any additional habitat.

Tamarisk (Tamarix pentandra) was introduced into the lower Colorado River basin by the late 1800s from the Nile Valley to provide wind breaks along irrigation canals. Tamarisk trees grow to about 10 m in well-watered locations and produce an abundance of small, easily dispersed seeds. By the 1920s, tamarisk trees were established throughout most of the Colorado River basin, including the Green River, below an elevation of about 2000 m. Although tamarisk is resistant to burial by sediment and is a prolific sprouter, it does not thrive where exposed to the high flow velocities and bank scour of large floods. Thus, the tamarisk invasion has been most extensive along river reaches with a broad flood plain. Under optimum conditions, tamarisk will form dense, riparian forests along the river banks. These forests greatly enhance bank stability and encourage the deposition of suspended sediment.

Decreased annual flood flows following the construction of Flaming Gorge Dam, together with an abundant contribution of sediment from unregulated tributaries, would have led to significant channel narrowing in and of themselves. Regulated river flows, however, promoted the growth of vast riparian forests of tamarisk and exacerbated channel narrowing. The invasion of tamarisk has reduced essential habitat for some of the native fish and, thus, contributed to their decline. First, low velocity areas along the channel margin which were essential to the native fish were removed as sand bars were first stabilized by tamarisk and, then, incorporated into a developing lower flood plain. Secondly, the dense riparian forests of tamarisk obstruct the connection

Case studies 163

between the river channel and inundated flood plain during high flow. Consequently, even when the river stage is high enough to inundate the flood plain, significant areas are not accessible to the young native fish (Stanford & Ward, 1993). Thus, the combination of a physically altered river together with the invasion of exotic species has had severe consequences for the native fish and riparian vegetation.

5.4.5 Anticipated benefits of increased flood magnitude and duration

Anticipated benefits of increasing the magnitude and duration of seasonal flooding along the Green River include:

• maintain and possibly increase bankfull channel size;

• expand aquatic habitat available for endangered native fish species by increasing the frequency and duration of flood-plain inundation;

• improve the competitive status of native fish relative to exotic fish by increasing the magnitude and duration of high velocity flows;

• limit the spread of tamarisk by scouring the margins of the low flow channel;

• expand within channel aquatic habitat available for native fish by building mid-channel bars that create a more topographically complex channel at low flow.

The magnitude, duration, and frequency of floods necessary to maintain and restore a given resource is subject to significant uncertainty and varies considerably from one resource to another. For example, a relatively short, large-magnitude flood every several years may be sufficient to limit the spread of tamarisk whereas longer, smaller magnitude floods every couple of years may be sufficient to expand the habitat available to the native fish species.

Although no decision has been made at this time, serious planning for an experimental flood to maintain and restore aquatic and riparian resources along the Green River downstream from Flaming Gorge Reservoir during the spring of 1997 is ongoing. Numerous issues and aspects of the experimental flood must be considered and resolved. Progress has been slowed by scientific uncertainty, conflicts within the existing Law of the River and the necessity of reaching consensus among several federal and state agencies. Even an experimental flood has the potential to set important legal and administrative precedents. Therefore, most participants have only considered those floods which they believe are possible or practical over the long term. The most significant physical, legal and economic issues are discussed in the remaining section.

164 Case studies

5.4.6 Conflicts and opportunities for increased flooding

Increasing the magnitude and duration of flooding in the Green River downstream from Flaming Gorge Dam will involve substantial economic costs and conflicts with current uses of land and water. Increased flooding will:

• limit future storage and consumptive uses of water,

• reduce the quantity of energy generated by the Flaming Gorge power plant, and

• remove agricultural land from production.

Finding acceptable resolutions to these conflicts will be difficult as responsibility for the development, management, and protection of the aquatic and riparian resources of the Upper Colorado River basin is dispersed among several government entities.

As discussed above, a discharge of approximately 500 m 3 s"1 is required downstream from the Jensen gauging station to initiate flooding and inundate a portion of the flood plain. Maximum release from the Flaming Gorge Power plant is about 125 m 3 s"1. An additional 115 m 3 s"1 can be released through tubes which bypass the power plant. Releases in excess of 240 m 3 s"1

are possible only when the reservoir is full and the spillways are available. Thus, the magnitude of discharges needed to create an aquatic habitat on the Green River flood plain downstream from Jensen, Utah, can only be achieved by combining a maximum release from Flaming Gorge Dam with a significant Yampa River flood. If water is to be secured from the Yampa River for annual peak flows on the Green River, it is apparent that the State of Colorado, where the Yampa is located in its entirety, will be precluded from fully developing the waters of the river. The Law of the River provides for full development of the states' waters on the one hand yet federal law requires the protection of endangered fish on the other hand. The Law of the River does not provide a means to balance these competing objectives (Getches, 1985).

Flaming Gorge Dam, along with the other CRSP facilities generate hydroelectric power which is distributed across several western states. The dams are federally owned and managed by the Bureau of Reclamation. The Western Area Power Administration (WAPA), formed under the Department of Energy Reorganization Act of 1977 is responsible for power marketing, and the construction, operation and maintenance of transmission facilities. The revenues from the sale of hydroelectric power are used for financing operation, maintenance, replacement and emergency work and recently, environmental and research work. Additional funds are used to repay costs and are apportioned to the Upper Basin States and used generally to repay irrigation investment costs. Hydroelectric power facilities are especially valuable as they can respond quickly to short-term peaks in demand. In

Case studies 165

contrast, the predominant alternative energy source, fossil fuel-based generation, is most efficient at relatively steady output. Accordingly, the CRSP facilities are operated to meet daily peak electrical power demand. Any restriction on operation, such as limitations on the daily range or hourly change in the river discharge, would affect the economic efficiency of the dams. Similarly, releases from the dam in excess of power plant capacity have significant costs. As described above, even moderate magnitude floods in the Green River will require bypassing 115 m 3 s"1 around the power plant. Sustaining such a release for 14 days would require about 7 .5% of the mean annual flow and cost $0.7 million in foregone power revenue.

Since Flaming Gorge Dam was completed in 1962, land uses adjacent to the Green River below the dam have changed, and expanded considerably. Previously, the Green River flood plain and riparian areas were generally free from permanent improvements due to unpredictable flooding and the existence of wetlands. The decrease in the extent and duration of annual flooding as described above, made it possible for various economic activities, primarily agricultural, to expand into lands adjacent to the channel. The annual costs of damage and loss of opportunity on these lands which would result from returning to the pre-dam flood magnitude and duration are estimated to be as much as $1 million.

Finally, the costs and benefits of reoperation of Flaming Gorge Dam to provide for peak annual flows are unevenly distributed between the current and future users of the Green River. Hydropower generation benefits consumers in 15 states. Reduction of hydropower production will have a negative effect on those consumers as well. Likewise, the benefit of attenuated, controlled flows affects landowners adjacent to the Green River who have been able to fully utilize lands previously subject to flooding. Also, to the extent that increased annual peak flows do not allow for full development of Colorado's share of water, the state of Colorado may be negatively impacted. In contrast, the benefits of increased annual flows are evident to those whose values are coincidental with the environmental values espoused by the ES A.

No governmental organization has broad, overall responsibility for the development, management and protection of the aquatic and riparian resources of the Upper Colorado River basin. Rather, responsibility for these resources is divided among a plethora of governmental entities. Furthermore, each entity tends to be focused on only one or, at most, a few resources; i.e. water supplies for irrigation, hydroelectric power distribution and marketing, recreation, biodiversity, or environmental conservation. Since 1988, a Recovery Implementation Program (RIP) for the Endangered Fish species in the Upper Colorado River basin has been in place attempting to find a cooperative, non-regulatory solution (Wydoski & Hamill, 1991). Five agencies of the federal government, three states, and several water

166 Case studies

conservation districts share authority and can substantially influence the steps taken to recover the endangered fish species of the Upper Colorado River basin. In the past, the government encouraged and facilitated compromise agreement among the several agencies, states, and interest groups by contributing enough funds to satisfy most parties. The current federal budget deficit makes such contributions unlikely in the foreseeable future.

In spite of the extensive body of agreements and laws allocating and controlling water in the Green River basin, many issues and conflicts remain unresolved (Wilkinson, 1985). In most instances, these issues will not become critical until the Upper Basin States begin to consume all of the water allocated to them by the Colorado River Compact (Garner & Michelle, 1995). Most observers anticipate a major crisis will ensue when the Upper Basin States begin to utilize all of their apportioned share of water sometime early in the next century. It remains to be seen whether the present Law of the River is sufficiently flexible and robust to deal with the competing claims for the water of the Colorado River (Bolin, 1993).

5.4.7 Conclusions on the Green River case study

Floods are essential in determining the physical and ecological character of the Green River. The decrease in flood magnitude and duration following the construction of Flaming Gorge Dam has substantially altered and damaged the aquatic and riparian resources of the Green River. Restoring and protecting these resources, which include four species of endangered native fish, will require more frequent, large and sustained floods than have occurred over the 36 years since the dam was completed.

Due to considerable scientific uncertainty concerning the ecology of the Green River prior to regulation as well as the introduction of exotic species, the necessary frequency, magnitude and duration of floods are poorly known. Determining the flood regime needed to restore and protect the Green River aquatic and riparian resources will require an extensive, long-term programme of ecological experimentation and research.

Even a partial restoration of the pre-dam flood magnitude and duration will involve significant economic costs and the loss of opportunity. Future water consumption in the Upper Colorado River basin and the quantity of hydroelectric energy generated would be significantly less than anticipated only a few years ago. Resolving these conflicts in a politically acceptable way will be difficult. The current Law of the River appears to be insufficient and the responsibility for the development, management, and protection of the Green River is fragmented and dispersed among several state and federal government entities.

Case studies 167

5.5 PROJECTS UNDERTAKEN TO ACHIEVE SUSTAINABLE DEVELOPMENT OF DAM RESERVOIRS IN JAPAN

5.5.1 Introduction to the Japanese reservoir policy

By adjusting natural flow fluctuations, reservoirs have made a large contribution to the prevention or reduction of damage from flooding and to the provision of a stable supply of water resources. This is particularly true in Japan, a country with steep topography and widely fluctuating meteorological and hydrological conditions, where dams have provided superior flood control and water utilization, based on improvements in dam construction technology.

A study of dams in Japan reveals that about 40% of the area of each water­shed is regulated. Although this means that the density of dams in Japan is extremely high compared with the rest of the world, Japan's dams can control only flood water produced by an equivalent rainfall of about 120 mm, and considering the growth of the water demand that must be satisfied, reservoir capacity per capita is still quite low. About 50% of the country's population and 75% of its assets are concentrated on downstream alluvial lowlands. Considering these circumstances, more reservoirs must be constructed and the flood control and water utilization functions of dams have to be improved. For these reasons, dams are indispensable in Japan, and their use must continue.

The construction of a dam submerges land, has significant social and economic impacts on its surrounding communities, and, temporarily harms the natural environment. This means that the consensus building process, which starts with the preparation of an initial plan and concludes with the construction of the dam, is an extremely time-consuming process. On top of this the amount of available land suitable for efficient dam construction has declined. And recently, the public's need to use dam reservoir surroundings for recreational activities has changed, as shown in Fig. 5.11, requiring dam development plans that incorporate measures to improve the surroundings of dam lakes to meet these new needs.

In response to these changes, projects related to "sustainable development of reservoirs" are undertaken in Japan.

5.5.2 Projects related to sustainable development of reservoirs

The projects are as follows:

(1) Legal arrangement for social impacts of dam construction projects. This incorporates measures for obtaining a smooth achievement of public consensus for dam construction and to reduce the socio-economic impacts.

168 Case studies

1955 1965 1975 1985 1990

k i Study: Study of nature,

»*** Î ecology, etc. around dams f »•** i and dam reservoirs y** i (Biotope, Dam)

• 4 Y

Experience: Activities around dams and dam reservoirs (camping, water sports, fishing, events, etc.)

Y ? / ,••* ; Space Utilization: Recreation at dams and

/,••** ; on suitable land around dams Y I (tennis courts, athletic parks, walking trails)

Sightseeing: Take advantage of the completion of the dam

Fig.

and dam lake to develop new sightseeing sites (field trips, strolling, industrial tours)

5.11 Schematic change of recreational needs for dam reservoirs.

(2) Projects defined as reservoir redevelopment. The following three categories are included in these projects:

(i) Recovering. This category embraces the recovery of reservoir functions, i.e. storage volume and reservoir level from functional degradations caused, for example, by sedimentation in the reservoir.

(ii) Improving the current functions of already developed (existing) reservoirs in an area wide river basin system. This category includes the increase of storage volume and level of an existing reservoir, with a view to meet the increasing demand.

(iii)Addition/reduction of new objectives (functions) to the current reservoirs and/or readjustment to new situations. "Adding/reducing" and "readjusting" reservoir functions means the addition of new objectives (functions), such as recreation and amenities, to the current functions of the reservoir, or the replacement of some objectives by others (e.g. replacement of some part of the municipal water supply by flood control).

Redevelopment activities may not necessarily be confined to single reservoirs but can be extended to an entire river basin system. To distinguish the former from the latter, they are defined as a "single" reservoir redevelopment and "networking" reservoir redevelopment, respectively.

There are a number of reservoir redevelopment activities which have been implemented, or are still under way or in the planning process, throughout the Japanese Archipelago (Sakamoto et al., 1995). The following box gives an overview of some examples:

Case studies 169

Tsugaru Dam (Aomori Prefecture) Category (i) (ii) Single Miwa Dam (Nagano Prefecture) Category (i) (ii) Single Shin-Maruyama Dam (Figu Prefecture) Category (ii) (iii) Single Amagase Dam: Category (Kyoto Prefecture) (ii) Single Matsubara/Shimouke Dams (Ohita and Kumamoto Prefecture) Category (ii) (iii) Networking Fukuchi Dam (Okinawa Prefecture) Category (ii) Networking Sasabugawa Dam (Fukui Prefecture) Category (iii) Single Asahikawa Dam (Okayama Prefecture) Category (ii) Single Kiyagawa Dam (Yamaguchi Prefecture) Category (ii) Single Kawakami Dam (Yamaguchi Prefecture) Category (ii) (iii) Single Minamihata Dam (Fukuoka Prefecture) Category (ii) Single Kayase Dam (Nagasaki Prefecture) Category (ii) Single Tsuruta Dam (Kagoshima Prefecture) Category (ii) (iii) Single Shin-Nakano Dam (Hokkaido Prefecture) Category (iii) Single Nagasaki Honkouchikoubu/Nishiyama/Urakami/Nakao/Yukiuradaini Dams (Nagasaki Prefecture) Category (iii) Networking

These projects can also be classified corresponding to redevelopment as shown in Fig. 5.12. In what follows the classification of projects is introduced according to this figure.

5.5.3 Legal arrangement for social impacts of dam construction projects

Dam construction projects are different from other ordinary public projects because they greatly affect the residents whose houses and surrounding areas may be submerged after a dam is constructed. Therefore, to promote dam construction projects, it becomes indispensable to take measures for relieving the impacts (denoted measures for reservoir areas) in conjunction with dam construction projects.

For this purpose, it is necessary to maintain the livelihood of the residents concerned and enhance their welfare by reconstructing the living environment and basic production of the reservoir areas in accordance with the Act on Special Measures for the Reservoir Areas Development in addition to the compensation offered by the dam owner. The regional reconstruction plans for the designated dams are planned by the prefectural governors concerned and approved by the Prime Minister after coordination between the involved ministries and the National Land Agency. Projects designated in the plans will be adopted as governmental subsidized projects, and a higher percentage of subsidies is permitted for a large dam by which extensive areas will be submerged. The same measures are also applied to large-scale lake development projects.

170 Case studies

Single reservoir "redevelopment

Structural -(Rehabilitation of an existing reservoir)

Nonstructural (Change of rules) "

Dredging of buttom sediment (Miwa Dam)

Increasing dam height (Shin-Maruyama Dam)

Improving discharge facilities (Asahikawa Dam)

Reallocation of reservoir . capacity among purposes (Tsuruta Dam)

Changing reservoir operation rules (Asahikawa Dam)

Transfer of ownership (e.g. water right)

Networking of reservoirs

Combination of existing reservoirs redevelopment " (Matsubara-Simouke Dams)

Redevelopment of existing reservoirs and . construction of new reservoirs (Nagasaki Honkochikoubu / Nishiyama / Urakami / Nakao / Yukiuradaini Dams)

Fig. 5.12 Classification by redevelopment means.

Also, the costs required for executing reconstruction projects are some­times partially borne by the beneficiaries in the areas downstream. In addition, as a supplement to the compensation by dam owners and reservoir area re­arrangement projects stated above, a "Fund for Reservoir Areas Development" was established for certain areas with contributions from the Government, prefectural governments and cities concerned. In certain areas the livelihoods of residents whose houses were submerged, are re-organized under this fund.

5.5.4 Water quality conservation measures for dam reservoirs

Some environmental problems have been created at certain dams, such as long-term turbid water in reservoirs due to floods, eutrophication as a result of development around reservoirs, discharge of cold water, landslides around reservoirs, and environmental degradation. Due to people's awareness of the environmental impact of development activities, these problems are considered extremely important in dam construction.

The three major water quality problems of cold water, long-term turbid water, and eutrophication in reservoirs have been studied by many organizations, and countermeasures have been examined. As a result it has

Case studies 171

been demonstrated that cold water and turbid water problems can be solved to a certain degree by installing selective intake facilities (including surface intake facilities) as required. With respect to eutrophication, various measures are being taken as shown in Table 5.2 and Fig. 5.13. Conservation measures suitable for dam reservoirs are being taken as well.

Table 5.2 Methods for improving water quality in reservoirs (Ministry of Construction, 1984).

Measures for dam reservoir catchments (control of nutrient loads into reservoir)

Measures for point sources

Measures for non-point sources

Highly treated sewerage (chemical coagulation treatment) Reduction or elimination of phosphate in detergents Excavation or bypass channel Restricting fertilizers Seepage gutters Filtering by a column of aluminium oxide Removal of phosphorus by pre-reservoir Removal of phosphorus from main inflow rivers Restricting land utilization

Measures for reservoirs

Removal of nutrient salts in reservoirs

Emergency measures for eutrophication

Chemical coagulation in reservoir Dilution and flushing Discharge from deep layer Removal of nutrient salts by removing organisms Drying and removing sedimentation at the bottom of reservoir due to lowered water level Treatment of reservoir bottom (covering bottom)

Aeration (circulation) Control of aquatic plants due to change in water level Physical measures (cutting aquatic plants) Chemical control (use of weed and algae killers) Biological control (treatment by predators)

5.5.5 Rehabilitation of an existing reservoir

Dam refreshing projects (sediment removal projects) Dams constructed as sedimentation measures are designed to guarantee a stipulated sediment capacity. But at some dams, factors such as abnormal flooding or weak geological upstream structures result in large amounts of sediments, far beyond the planned quantity, being deposited within the effective reservoir capacity. At dams in this condition, sedimentation countermeasures are implemented in order to restore the effective reservoir capacity needed to adjust flood waters or supply water to local users and to guarantee new capacity. Specific methods employed are: (a) the use of a pump dredger to

Detour channel)

9

Fig. 5.13 Schematic diagram of water quality improvement techniques in reservoirs.

Case studies 173

perform on-land excavation and dredging, and (b) the installation of permanent dredging facilities (dredging pipelines, dredging gates, etc.).

Increasing dam height By increasing the dam height of an existing dam extra effective storage capacity can be obtained for flood control or water utilization. It has the advantage that a large expansion of effective capacity can be achieved with only a slight enlargement of the dam. However, it involves crucial processes, such as the use of fairly advanced techniques to connect the new and old dam bodies and to improve the foundation, and the difficulty of maintaining normal functioning of the existing dam during the construction period must also be taken into account.

5.5.6 Improvement of the management and operation of dam reservoirs

Changing reservoir operation rules In Japan, flood water period or summer period maximum reservoir levels are established, and dam managers take steps to keep the reservoir level at each dam at or below the stipulated level from June to September. A high volume of water flows rapidly into the reservoirs during these months as a result of the extensive rain brought by the late spring and early summer seasonal rain fronts and by typhoons. When it is possible to keep the restricted flood period water level at the dam flexible, and adjust it step by step in conjunction with using the preliminary release method, it helps ease the conflict between the needs for flood control and water utilization during the same period. Thus, by allowing more water to be used, an extra water storage capacity can be guaranteed. Methods of forecasting extensive rainfall and flood water levels must become more accurate, and fully equipped evacuation and warning systems have to be provided.

Another use of dams is electric power production, but hydroelectric plants form dry river beds between the dam and the power plant outlet, and these dead water sections adversely affect both the rivers' recreational functions and the living organisms which inhabit them. Dam managers are encouraged to discharge a stipulated volume of water in order to eliminate these dead water sections. When the interrelationships between various aspects of the physical environment, living organisms, the ecosystem, flow volume, water levels, scouring of the river bed, sedimentation processes etc. are eventually elucidated, fluctuations in the river maintenance flow volume suitable for co­existence with living organisms and the ecosystem should be approved and dams used to rationally control these fluctuations.

Dam integration projects and comprehensive water utilization systems A dam integration project involves the linking of a number of existing adjacent

174 Case studies

dams with channels so that the existing storage capacity can be used more effectively by allowing dams to send their uncontrolled outflow to other dams for storage. A drought protection dam, once designed and built with sufficient capacity to deal with drought conditions as shown in Fig. 5.14, guarantees the minimum amount of water needed during drought conditions. Projects of this kind are undertaken to permit water circulation and provide the minimum required amount of water during drought conditions, but, in order to guarantee a highly reliable stable water supply during both normal and drought conditions to meet the growing water demand of the large cities, it is important to revive the idea of improving dam facilities by implementing the integrated water utilization system shown in Fig. 5.15.

In other words, future plans should be based on so-called integrated water utilization planning and management theory that considers the possibilities for obtaining water from a number of different sources such as river water from reservoirs, groundwater, recycled treated waste water, rain water, and desalinated sea water in urban regions, as well as the construction of urban storage facilities corresponding to the quality of the source water and its intended uses, and provision of double pipes and other improved lifeline facilities to supply water and link the system to water conveyance works outside the watershed through neighbouring dams or regulation ponds.

5.5.7 Regional promotion and regional planning centred around dams

Because the construction of a dam has a large impact on the society of the region and the foundations of the life of the local people by submerging their

Fig. 5.14 Dam against extraordinary drought. Time

Case studies 175

Sea WateADesalinization Plant

Use of

River

Recycled [Treated

-e Sewage Treatment

Y Low-quality Water Network

£ 7 Water storage under ground

Waste Water £ ;

High-quality Water Network

Plant

Use of Recycled Treated Waste Water

Urban Area

Water Conveyance works

Dam as measeure for extraodinary drought Water Conveyance works

River

Fig. 5.15 Comprehensive water utilization scheme.

land, dam planning should include provisions for the improvement of the surrounding environment in order to reduce the impact and to permit the dam to serve as a basic facility contributing to a revitalization of the surrounding region. Furthermore, each dam should be planned so that both the dam and its improved environment will be attractive not only to people living close to the dam, but also to those in the downstream region, allowing a growing number of people to take part in communal events near the dam, or simply to visit the facility.

Public access dam projects In response to the recently growing public need for greater access to the natural environment and recreational facilities the promotion of the use of dams and dam reservoirs and their surroundings as open space filled with water and natural greenery, as well as the efforts to preserve their natural environments, will definitely make an important

176 Case studies

contribution to regional stimulation. For this reason, "Public Access Dam" projects intend to take advantage of the creativity and innovative spirit of the local populace and provide dams accessible to the public. Encouragement is given to the use of these dams in order to increase the local people's familiarity with them and, in this way, to stimulate development of the region.

Throughout Japan, improvement projects are in progress at a total of 25 dams designated as "Public Access Dams". Figure 5.16 shows an example: the Hiyoshi Dam on the Yodogawa River system, which is now under construction, and which is combined with a memorial park downstream of the dam. Facilities in concert with the surroundings are being provided around the reservoir, and the land around the reservoir is being landscaped and reforested.

Recreational water improvement projects As a result of the recent increase in the amount of leisure time enjoyed by the people, a growing number are seeking means for spiritual repose. To increase the hydrophilic properties of reservoirs to meet the new needs, dam administrators are drawing on the resources of the private sector to create environmental water surfaces

Fig. 5.16 Hiyoshi Dam memorial park.

Case studies 177

(recreational lakes) and to promote the development of resorts around dam lakes. For example, recreational lakes with constant water levels are developed by installing a secondary dam which simultaneously improves the water quality and stores sand at the upstream end of the dam lake.

Dam site environment improvement projects These projects improve dam reservoir surroundings and contribute to improved dam management at the same time as they provide places for relaxation close to dams. Specific measures include the improvement of river banks, reforestation and the provision of footpaths on the river banks, the installation of protective fences and warning equipment, the provision of environmental conservation facilities etc.

5.5.8 Other issues

Constructing large dams on main river courses or on large tributaries is an effective way to prevent flooding and supply water to water users. However, because of the serious socio-economic and environmental impacts of the construction of such dams, it is more difficult than ever to carefully implement such projects. This is likely to become even more difficult as a consequence of the growing need to preserve the environment and promote co-existence with natural organisms and the ecosystem. This means that it is vital to establish planning and management policies for constructing small capacity dams at a number of sites on small tributaries which would have the same capacity as a single large capacity dam; but which would produce small amounts of sediment. It is also vital to carry out an integrated management of the combined capacity of these small dams. This approach would permit a regulation that provides the water volume needed for both flood control and water utilization at the same time as it would reduce the impact of dams both on the environment and on human society. Its implementation would, of course, increase the importance of linking dams and developing integrated dam management technology including forecasting technology.

Reservoir redevelopments can be interpreted as a developmental activity aimed at sustainability. For example, given that in developed countries like Japan suitable dam sites in a river basin are becoming increasingly difficult to find, it is imperative that the following conditions are met:

• growth of limits (rather than limit to growth);

• co-existence with nature;

« co-existence with future generations;

« ultra-long-range management;

• self-supporting, recycle-oriented maintenance.

178 Case studies

Reservoir redevelopments can be regarded as a typical example of a development activity where "nature is created" by paying special attention to the above conditions. Stated in another way, a new planning process and methodology should be conceived to systematically design an inter-generational, long-range, self-supporting implementation/maintenance system for river basins with manmade reservoirs.

It may be useful to list essential problems which may come about in planning and applying a methodology to the above-stated sustainable reservoir redevelopments. They are:

• water resources reallocation for a single or networking reservoir redevelopment;

• benefit allocation for a single or networking reservoir redevelopment;

• cost allocation for a single or networking reservoir redevelopment;

• inter-generational, long-range, self-supporting implementation/maintenance, multi-phase planning strategy.

In 1982, Nagasaki City experienced a large-scale flood disaster. The rapid urbanization was one of the reasons for this disaster. The plan of the Nagasaki emergency flood control project, set up to deal with just such a problem, was to create a flood control capacity in the existing reservoirs located upstream of Nagasaki. These reservoirs had been used as sources for municipal water supply. A reallocation of the reservoir capacity leads to a reduction of the amount of municipal water supply. This project planned to compensate the reduction by constructing new reservoirs in different river basins. Cost allocation became a difficulty in realizing the project. One reason was that the project involved many reservoirs. Another reason was that the reorganization of reservoirs did not modify the amount of water provided for municipal water supply. Because this project was only for flood control, the waterworks bureau had no incentive to pay. No methodology has been established for this type of cost allocation.

5.5.9 Conclusion on the Japanese study

This study has outlined the laws and regulations, technological measures, and project policies required in Japan to construct future dams, maintain the water storage functions of existing dams, and contribute to regional and watershed development. The environment surrounding dam provision may become increasingly severe, but we are confident that the new developments described above will overcome these difficulties and contribute to sustainable development.

Case studies 179

5.6 CONCLUDING REMARKS ON THE CASE STUDIES

The studies presented all addressed sustainability issues but focused on different aspects of the concept. They are definitely not ideal cases and not selected as such. Together, however, these case studies, differing substantially in size and geographical location, have shed light on the difficulties we are faced with when striving for sustainability in reservoir development.

The Alta case study showed the importance of respecting the rights of indigenous people and the necessity of public participation in the planning process. Further, the vulnerability of Arctic nature and ecology was highlighted. Local vs. national decision making became an issue for strong debate and civil disobedience which resulted in the project being delayed but not significantly modified. The impact on future planning, however, was evident leading to a much greater awareness of sustainability issues between politicians and other decision-makers.

The Aswan High Dam project has been subject to severe international criticism. The sustainability could be, and has been questioned for several reasons. The inundation behind the dam was the reason for resettlement of more than 100 000 people and a unique cultural heritage was situated on the inundated land. To balance the discussion we need to take into account the fact that a population much greater than 100 000 has since been spared from disastrous floods, and saving of the cultural treasures created a truly international effort. The lack of silty sediment deposition downstream of the dam has necessitated the use of artificial fertilizers and sardine have almost disappeared from the estuary. On the other hand, irrigation and electricity production are of outstanding importance for the population and the fishery in the Aswan High Dam Lake has become an important source of income. The case illustrates the great complexity of sustainability assessments.

The Green River case study gives a detailed assessment of how the construction of a reservoir can adversely change the downstream ecology. The decreased sediment transport and the reduced flood peaks are considered the main reasons for a severe decline of four native fish species. The ecological problem has grown substantially due to the invasion of both terrestrial and aquatic species that have adapted better to the changed conditions than many of the natural species. Proposals have been made for increasing flooding artificially both in magnitude and duration which to some extent could alleviate the negative impacts of the dam construction. The many difficulties to overcome before the management policy of the river basin could be changed are highlighted. There are strong economic interests involved and many different federal and state agencies are players. Even agreement to perform limited experiments in order to achieve a stronger scientific background for a new policy has not been reached. The Green River case study illustrates how difficult it can be to solve conflicting

180 Case studies

interests and highlights the risk that sustainability issues will carry the least weight in the decision-making process.

In Japan there is a strong need for utilization of reservoirs but also an increasing awareness of the need for sustainable development. Thus a new policy has been introduced with more emphasis than before on, among other things, rehabilitation of existing dams and establishment of comprehensive water utilization projects. The planning process has been changed in order to allow for public participation, and all new projects are now planned with due consideration of landscaping, reforestation, and recreation facilities for the public.

REFERENCES

Abu El-Atta, A. A. (1978) Egypt and the Nile After the Construction of the High Aswan Dam. Ministry of Irrigation, Cairo, Egypt.

Abu El-Wafa, T. (1962) The Social and Economic Construction of the High Aswan Dam. Ministry of the High Dam, Cairo, Egypt.

Andrews, E. D . (1986) Downstream effects of Flaming Gorge Reservoir on the Green River, Colorado and Utah. Geological Society of America Bulletin 97, 1012-1023.

Andrews, E. D. & Nelson, J. M. (1989) Topographic response of a bar in the Green River, Utah to variation in discharge. In: River Meandering (eds S. Ikeda & G. Parker), American Geophysical Union, Washington D C , 463-485.

ASRT (1990) Development of Natural Fisheiy Resources. Second Subject, Fish and Animal Resources Research Council, Cairo, Egypt.

Berg, L. (1976) The Aswan High Dam and its various consequences. In: Symposium on Nile Water and Lake Dam Projects, National Research Centre, Cairo, Egypt, session 4, paper no. 1, 1-98.

Biswas, A. K. (1992) The Aswan High Dam revisited. Ecodecision, September 1992, 67-69 . Bolin Jr, J. H. (1993) Of razorbacks and reservoirs: The Endangered Species Acts protection of endangered

Colorado River basin fish. Pace Environmental Law Review 11, 35-87 . Dalland, Oe. (1994) Alta-kroenike (Alta chronicle): in Norwegian. Davvi Girji o.s . , Karasjok, ISBN 82-

7374-073-0. El-Moattassem, M. & Abdelbary, M. R. (1993) Bed degradation and channel shifting in the Nile after the

Aswan High Dam. ICOLD 61st Executive Meeting and Symposium, High Aswan Dam: Vital Achievement Fully Controlled, Cairo, Egypt, 266-279.

Fisher, N. T . , Toll, M. S., Cully, A. C. & Potter, L. D. (1983) Vegetation along Green and Yampa Rivers, and response to fluctuating water levels, Dinosaur National Monument. Final Report, Contract CX-1200-2-B024. Biology Department, University of New Mexico, Albuquerque, New Mexico.

Garner, E. L. & Michelle, 0 . (1995) Future shock? The law of the Colorado River in the twenty-first century. Arizona State Law Journal 27, 469-506.

Getches, D. H. (1985) Competing demands for the Colorado River. University of Colorado Law Review 56, 413-479.

Getches, D . H. (1990) Water Law in a Nutshell. West Publishing Co. , St Paul, Minnesota. Goodland R. (1996) Distinguishing better dams from worse. International Water Power and Dam

Construction 48(9), 34-36 . Goodland, R., Juras, A. & Pachauri, R. (1992) Can hydro reservoirs in tropical moist forest be made

environmentally acceptable? Energy Policy 20(6), 507-515. Graf, W. L. (1978) Fluvial adjustments to the spread of tamarisk in the Colorado Plateau region. Geological

Society of America Bulletin 89, 1491-1501. Helander, E. (1992) The Sami of Norway. The Ministry of Foreign Affairs, Internet edition:

http ://odin. dep. no/ud/norny tt/uda-309 .html Hillestad, K. O. (1993) The Alta Power Plant in the Landscape. Norwegian Water Resources and Energy

Administration, Oslo.

Case studies 181

Irons, W. V. , Hembree, C. H. & Oakland, G. L. (1965) Water resources of the Upper Colorado River basin—Technical Report. US Geological Survey Professional Paper 441.

Johannesen, A. (1993) NORDEL-modellen (The N O R D E L model), in Norwegian. In: Energy Planning Models—Climate Change (eds M. L. Fossdal & N. R. Saelthun), Publ. 08-1993, Norwegian Water Resources and Energy Administration, Oslo, ISBN 82-410-0193-2.

Killingtveit, Aa. & Saelthun, N . R. (1995) Hydrology. Vol. 7 in: Hydrology Development. Norwegian Institute of Technology, Trondheim, ISBN 82-7598-026-7.

Koeber, K. (1981) Alta-utbyggingen har mange sider (The many aspects of the Alta Hydropower Scheme), in Norwegian. Oslo, ISBN 82-524-0048-5.

Lyons, J. K. & Pucherelli, M. J. (1992) Sediment transport and channel characteristics of a sandbed portion of the Green River below Flaming Gorge Dam, Utah, USA. Regulated Rivers: Research & Management 7, 219-232.

Makary, A. Z. (1982) Sedimentation in the High Aswan Dam Reservoir. PhD Thesis, Irrigation and Hydraulics Dept, Faculty of Engineering, Ain Shams Univ. , Cairo, Egypt.

Minckley, W. L. & Mefee, G. K. (1987) Differential selection by flooding in stream-fish communities of the arid American southwest. In: Community and Evolutionaiy Ecology of North American Stream Fishes (eds W. J. Matthews & D. C. Heins), 93-104. University of Oklahoma Press, Norman, Oklahoma.

Minckley, W. L. , Marsh, P. C , Brooks, J. E. , Johnson, J. E. & Jensen, B. L. (1991) Management toward recovery of the razorback sucker. In: Battle Against Extinction: Native Fish Management in the American West (eds W. L. Minckley & J. E. Deacon), University of Arizona Press, Tucson, Arizona, 303-357.

Ministry of Construction (1984) Multipurpose Dams in Japan. River Development Division, River Bureau, Ministry of Construction.

Nesler, T. P . , Muth, R. T. & Wasowicz, A. F. (1988) Evidence for baseline flow spikes as spawning cues for Colorado squawfish in the Yampa River, Colorado. In: American Fisheries Society Symposium 5 , 68 -79 .

NRI (1993) Study of Sediment Movement in Aswan High Dam Lake. Nile Research Institute, National Water Research Centre, Cairo, Egypt.

Olje- og Energidepartementet (1978) Om statsregulering av Altavassdraget i Finnmark fylke (On the state regulation of The Alta River in Finnmark County), in Norwegian. Stortingsproposisjon nr. 107, 1977-78.

Ryvarden, L. & Toemmeraas, P. J. (1979) Alta-Kaulokeino vassdragel (The Alta-Kautokeino water course), in Norwegian). Universitetsforlaget, Oslo, ISBN 82-00-05311-3.

Said, N . F. (1993a) The exodus of the Egyptian Nubians subsequent to the building of the High Aswan Dam. ICOLD 61st Executive Meeting and Symposium, High Aswan Darn: Vital Achievement Fully Controlled, Cairo, Egypt, 232-239.

Said, R. (1993b) The River Nile: Geology, Hydrology and Utilization. Pergamon Press, Oxford, UK. Sakamoto, T . , Yamazaki, Y. & Yonezaki, F. (1995) Flood control-included multi-purpose reservoir

redevelopment projects in Japan. Darn Engineering and Technology 108, 3-17. Shalaby, A. M. (1993) Socio-economic impacts of the High Aswan Dam. ICOLD 61st Executive Meeting

and Symposium, High Aswan Dam: Vital Achievement Fully Controlled, Cairo, Egypt, 158-165. Stanford, J. A. (1993) Instream Flows to Assist to Recovery of Endangered Fishes of the Upper Colorado

River Basin: Review and Synthesis of Ecological Information, Issues, Methods and Rationale. Flathead Lake Biological Station, Poison, Montana.

Stanford, J. A. & Ward, J. V. (1986a) The Colorado River system. In: The Ecology of River Systems (eds B. R. Davies & K. F. Walker), Dr W. Junk Publishers, Dordrecht, The Netherlands, 353-374.

Stanford, J. A. & Ward, J. V. (1986b) Fish of the Colorado system. In: The Ecology of River Systems (eds B. R. Davies & K. F . Walker), Dr W. Junk Publishers, Dordrecht, The Netherlands, 385-402.

Stanford, J. A. & Ward, J. V. (1993) An ecosystem perspective of alluvial rivers: connectivity and the hyporheic corridor. Journal of the North American Benthological Society 12(1), 48-60 .

Tyus, H. M. (1991) Ecology and management of Colorado squawfish. In: Battle Against Extinction: Native Fish Management in the American West (eds W. L. Minckley & J. E. Deacon), University of Arizona Press, Tucson, Arizona, 379-402.

Tyus, H. M. & Karp, C. A. (1991) Habitat Use and Streamflow Needs of Rare and Endangered Fishes, Green River, Utah. US Fish & Wildlife Service, Flaming Gorge Studies Program, Vernal, Utah.

Ward, J. V. & Stanford, J. A. (1989) Riverine ecosystems: the influence of man on catchment dynamics and fish ecology. In: Proceedings of the International Large River Symposium (ed. D . P. Dodge), Canadian Special Publication of Fisheries and Aquatic Sciences 106, 56-64 .

Wilkinson, C. F. (1985) Western water law in transition. University of Colorado Law Review 56, 317-345.

182 Case studies

Wydoski, R. S. & Hamill, J. (1991) Evolution of a cooperative recovery program for endangered fishes in the Upper Colorado River basin. In: Battle Against Extinction: Native Fish Management in the West (eds W. L. Minckley & J. E. Deacon), University of Arizona Press, Tucson, Arizona, 123-135.

Sustainable Reservoir Development and Management. Report by the IAHS/ICWRS Project Team. IAHS Publ. no. 251, 1998. j g j

CHAPTER 6

CHECKLISTS AND CONCLUDING REMARKS

6.1 CHECKLISTS FOR SUSTAINABLE RESERVOIR DEVELOPMENT AND MANAGEMENT

The following are checklists for sustainable reservoir development and management for which a brief rationale was introduced in 2.2.7. They refer to three stages of reservoir development:

• a planning and design stage;

• a construction stage; and

• an operation and maintenance stage;

and five broad areas, where disadvantages can be critical:

• the conservation of nature;

• inter-generational equity;

• intra-generational equity;

» efficiency; and

• integrated water management.

PLANNING AND DESIGN STAGE

Conservation of nature

• EIA Adequate environmental impact assessment is carried out (the necessary time and resources are invested, embracing breadth of perspective, low-probability extreme events, long time horizons and spatially remote impacts).

• Co-existence with nature Such alternatives that have fatal effects on ecological systems should be excluded, especially on biodiversity, rare species and the number of population required to ensure the natural evolution process.

• Minimum environmental impact alternative The alternative selected is the one that accompanies the minimum environmental impact as compared with other alternatives that can yield the same level of satisfaction on the objectives.

184 Checklists and concluding remarks

• Environmentally conscious design The reservoir is designed to minimize the adverse effects of storing water, segmentation of the habitat, inundation etc. by facilitating sediment discharge gates, sediment bypass channels, fish ladders, bio-recovery sanctuary (biotopes) etc.

Inter-generational equity

• Explicit consideration of future generations The impact on future generations is explicitly considered in plan formulation and evaluation.

• Benefit exceeds cost for future generations The long-term benefit will exceed the long-term loss due to long-term environmental impacts.

» Land safety and productivity The construction and planned water use will not cause long-term environmental and geomorphological deterioration of the land, neither waterlogging upstream, coast line retreat and sea-water intrusion downstream or salinity problems on irrigated land.

» Socio-cultural heritage The protection of historical and archaeological sites, cultural heritage and unique landscapes is taken care of in project planning.

Intra-generational equity

• Democratic decision-making process The construction plans and alternatives are known to the broad public and made available for public discussion with full information disclosure among all the stakeholders, in particular, the indigenous people in any inundated area. Public consent is sought at all stages in the decision-making process.

• Proper rehabilitation of involuntarily relocated people The people to be relocated are taken care of until their lives and their community in the new location have been stabilized.

• Upstream and downstream equity The benefit obtained from the reservoir is shared equally with the people in the reservoir area.

• Income redistribution The economic benefit of the reservoir output is considered for people in need thus contributing to an equalization of income distribution.

Efficiency

• Economic efficiency The project is financially feasible with satisfactory cost recovery for the capital investment.

» Use of technology favouring resilience Every possible technology is utilized to make the reservoir system resilient so as to adapt efficiently to

Checklists and concluding remarks 185

unexpected surprises in nature, as well as society, including extreme hydrometeorological phenomena, changes in the public value system and other conditions.

• Multipurpose use The reservoir is planned and operated in a multipurpose manner.

• Evaporation loss The estimated water losses (évapotranspiration and seepage) are acceptable.

• Hydropower generation The power generation component is included in the set of purposes. It is always recommended, if a reservoir is to be built, to utilize environmentally clean hydropower regardless of the prime purposes and the size of reservoir, so long as it is economically justifiable.

Integrated water management

• Thorough consideration of non-reservoir options The non-reservoir alternatives to reach the planning objectives are thoroughly considered.

• Demand management Special consideration of demand management, re­allocation of water rights, urban water storage and infiltration are fully taken into account before choosing the reservoir option.

• Integrated multi-sectoral water management The reservoir is planned with multi-sectoral cooperation in an integrated way with institutional, managerial, economic, physical and all other relevant means.

• Basin-scale water management The reservoir is planned and operated as a part of the overall basin-scale water management.

CONSTRUCTION STAGE

Conservation of nature

• Environmentally careful construction work The reservoir is constructed in the least environmentally damaging manner related to water quality, erosion, air pollution, liquid and solid waste, noise during various construction works such as access roads, bypass tunnels, rock and soil material quarrying as well as the dam construction itself.

• Biomass inundation Organic decomposition does not lead to critical dissolved oxygen deficits in the waters of the reservoir.

• Rescue of animals and plants Transferable animals and plants are relocated (in a way so as not to create serious transmigratory impacts on the receiving ecological community).

186 Checklists and concluding remarks

OPERATION AND MAINTENANCE STAGE

Conservation of nature

• Rehabilitation of surrounding areas Adequate measures are taken to rehabilitate damaged natural environment in the affected areas.

• Normal maintained flow Streamflow below the dam is sufficient to maintain the normal biotic life.

» Natural flow pattern The reservoir is operated with a provision for natural release patterns (seasonal and diurnal variations, minimum hygienic flow, low flow characteristics, flow requirements for aquatic species) with similar hydrological extremes, wherever desirable, to ensure quasi-natural biotic conditions downstream.

• Post audit The ecological impacts on aquatic fauna and flora as well as on water quality and sedimentation are continuously monitored over the long term.

Inter-generational equity

• Sediment control and safety The reservoirs are designed not to lose their capacity, safety, and other functions for centuries. This requires undertaking proper measures for mitigating reservoir sedimentation (built-in sediment discharge gates or bypass channels), safety design against overtopping and earthquakes and against structural decline with proper provision for routine structural monitoring.

• Catchment management Continuous efforts are made on land (hillslope) management, forests and native vegetation maintenance, effluent discharge control so as to avoid the adverse effects from upper basin development such as farmland reclamation, urbanization, industrialization, residential and recreation activities, on the quality and quantity of reservoir water.

Efficiency

• Combined use The reservoir is operated in the manner integrated with other reservoirs, aquifers, and storage facilities.

• Use of available forecasts Hydrometeorological forecasts and water demand predictions are, if available with reliable accuracy, properly utilized.

• Use of computer-aided information management and decision support technology The knowledge base for flood control, drought management, water quality control, navigation, hydropower etc. is properly managed and transmitted to the available decision support technology, such as expert systems, artificial intelligence and others, wherever appropriate for enhancing efficiency and reliability.

Checklists and concluding remarks 187

• Health risk reduction The reservoir operation avoids health risks due to mosquitoes, filaria, schistosomes etc.

• Maximum environmental value The new lake is managed to produce the maximum environmental value.

Integrated water management

• Integrated operation of multi-component water resource system The reservoir is operated and maintained in conjunction with all the related components of water resource systems in a comprehensive manner, for multipurpose flood control, water supply, environmental quality control etc.

• Flexible allocation of water resources The water allocation rule is flexible enough to react to the real needs of the region whenever necessary, such as during prolonged droughts, or when social needs and preferences may change in the long term.

6.2 CONCLUDING REMARKS

Reservoirs are undoubtedly one of the most important means of water resources development and management, and they have been built since earliest historical times. Human beings have been successful in achieving the necessary control of water by reservoirs for agriculture, flood control, hydropower generation, water supply, navigation and recreation. Engineering technology for reservoir construction has saved many lives and made development possible. However, it is also true that many large reservoirs were built, at least in part, without proper account of environmental, social and economic interactions. Critics who claim that many large dams are environmentally destructive, socially tragic and economically unsuccessful should be listened to carefully and their concerns scientifically analysed. New approaches must be taken.

In the 21st century plainly more water will be needed for an increasing world population which aspires to a higher standard of living, and that surface water development through reservoir construction is likely to be the major source. In these circumstances, scientific diagnoses must be subject to a broad public discussion among interested and affected parties. It is no longer acceptable simply to point out the impacts of reservoir construction in terms of some future benefits and disadvantages.

Sustainable reservoir design and management is a new concept in the sense that not only the efficiency of current performance and physical safety should be considered, but also that intra- and inter-generational needs and equity must be taken into account. The general sustainability principle has been widely acknowledged at all levels, yet its concrete implementation in individual

188 Checklists and concluding remarks

subject matters is not adequately formulated in any field. In this study, the translation of the basic sustainability concept into all aspects of reservoir planning and management has been attempted. During the study, many previous reports were reviewed and considered, and the final outcome is not necessarily new. What is needed is, in many respects, a restatement of integrated water resources management with social consideration, environmental and ecological care, and sediment control combined with democratic decision-making procedures. This has been pointed out for many years in the literature.

This report especially emphasizes the importance of socio-environmental care for the affected people and nature itself. Planners can no longer pay biased attention to the beneficial elements of development and accept adverse effects with a minimum of materialistic compensation. It is no longer acceptable to overlook the intangible negative consequences of a project. The unspoken qualitative, indirect effects should be carefully assessed and included in the overall evaluation process, since they too have an impact on the basic conditions of society, nature and future generations. It is an obligation for engineers and scientists to consider the total system response and provide an objective analysis to enable a rational, democratic and feasible solution.

Human beings change and modify the natural condition of nature. However, human beings also have the ability to protect nature for themselves and for future generations. How much of nature we have to expend in order to sustain our own existence depends on the socio-technical development level of civilization. Even the question of sustainable reservoir development cannot be answered by hydrological and water resource systems sciences alone. Sciences as a whole must work together and solve the problem as an integrated part of the dilemma of contemporary human existence.

In this report a set of new ideas has been explored in order to illustrate possible ways to address the requirements of sustainability in reservoir analysis. One contribution involves the introduction of the least marginal environmental impact rule. It proposes that the size of a reservoir should be limited to the extent that an extra unit increase of dam height has the least negative environmental impact compared to all economically feasible alternative means, reservoir or non-reservoir, which provide the same level of incremental gain in the global objectives as the extra unit increase of dam height.

Among other contributions a new method for designing sustainable reservoirs, based on the De Novo programming approach, is presented as an idea for replacing the optimization of a given reservoir with the design of an optimal reservoir. Reservoir storage reallocation and reassessment of reservoir operational rules are considered to be the two main problems related to existing reservoirs that should be addressed within the framework of sustainability. Another contribution to the report is a method for reassessment of existing reservoirs, which is based on the combined use of simulation and

Checklists and concluding remarks 189

optimization. The main objective of the approach is to determine: (a) the active reservoir storage requirements based on the current demand; and (b) the best management strategy for the reservoir under consideration.

The report also presents an original cost and benefit allocation methodology for redevelopment of reservoirs. The methodology has been illustrated with two case studies: (a) addition of new water users to an existing water supply system, and (b) addition of new water users to a system with additional reservoirs.

Other ideas emphasize the importance of non-reservoir options, especially demand management, including pricing and other economic incentives to reduce the net consumption and/or pollution of water and institutional mechanisms that promote efficient water allocation and conservation through mobilized water rights and a multi-sectoral approach. This may limit conflicts between different water users and reduce water shortages and the need for new reservoir development, thus becoming the most important contribution from the water resources sector in promoting a sustainable environment.

An important aspect stressed in the report is the necessity of continued scientific research on complex water problems. It has become more and more crucial to conduct research on sediment control, water quality control, ecological impacts and assessments and ecologically conscious release policies. Also, it is necessary to monitor reservoirs in operational use and connected river reaches for a long time after completion. It is essential to utilize advanced hydrometeorological forecasting techniques to improve the performance of reservoirs and to develop procedural mechanisms for conflict resolution which can lead society to a more environmentally efficient allocation of resources. To use a large set of available data requires knowledge of advanced analytical tools, for example, computer-aided decision support systems, artificial intelligent systems, and various operational research techniques such as optimization and simulation, including the application of statistical, stochastic and fuzzy approaches. Many other analytical and synthetic techniques need to be further developed and utilized. They all require interdisciplinary research. The scientific institutional framework, making such an interdisciplinary approach possible, is increasingly important and needs to be encouraged and further developed.

There is an urgent need for implementation of reservoir development and management based on the sustainability principle. The concrete criteria of planning, design, construction, operation and redevelopment of reservoirs depend on local, site-specific conditions of a socio-economic, cultural and environmental nature. However, the basic concept of reservoirs as components of water resource systems that must be sustainable in societal as well as environmental contexts has now become clear. This report, reflecting the views of many critics, has highlighted a number of aspects which all contribute to sustainability, and has provided new methodologies for the improvement of current practice.

190 Checklists and concluding remarks

Evidently, the 21st century will b e c o m e t h e w a t e ï S t t C S S C C f t t \ M ^ . ¥L \ ff iM survival and the Earth environment are largely dependent on the human capability for water management and this, in turn, poses a significant demand on development and use of reservoirs in a sustainable manner. The authors sincerely hope that this report will be a step towards the implementation of sustainable reservoir development and management practices.