,
" ~
.
I
Evacuation of sediments from reservoirs
Rodney White
~
I ThomasTelford I
\ I
i
.
..,, :-
'\
Published by Thomas Telford Publishing, Thomas Telford Ltd, 1 Heron Quay, London E14 4JD. URL: http://www.thomastelford.com Distributors for Thomas Telford books are USA: ASCE Press, 1801 Alexander Bell Drive, Reston, VA 20191-4400, USA Japan: Maruzen Co. Ltd, Book Department, 3-10 Nihonbashi 2-chome, Chuo-ku, Tokyo 103 Australia: DA Books and Journals, 648 Whitehorse Road, Mitcham 3132, Victoria First published 2001
.
,
~
Front cover shows reservoir sedimentation in Zimbabwe
I•
- j i
1
--I ,i
i
:-1 ~I
-j
I
-I
I
A catalogue record for this book is available from the BIitish Library ISBN: 07277 2953 5 © Rodney White and Thomas Telford Limited, 2001
All rights, including translation, reserved. Except as pennitted by the Copyright, Designs and Patents Act 1988, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the Publishing Director, Thomas Telford Publishing, Thomas Telford Ltd, 1 Heron Quay, London E14 4ID.
II :--1
j
I
-I-~
I
This book is published on the understanding that the author is solely responsible for the statements made and opinions expressed in it and that its publication does not necessarily imply that such statements and/or opinions are or reflect the views or opinions of the publishers. While every effort has been made to ensure that the statements made and the opinions expressed in this publication provide a safe and accurate guide, no liability or responsibility can be accepted in this respect by the authors or publishers .. Typeset by Apek Digital Imaging, Bristol, UK Printed and bound in Great Britain by MPG Books, Bodmin, Cornwall
-I
Preface In many areas of the world the life span of reservoirs is determined by the rate of sedimentation which gradually reduces storage capacity. Eventually, this process destroys the ability of the scheme to deliver the benefits for which it was built. Many major reservoirs are approaching this stage in their life. There are various options available for positively managing sedimentation in reservoirs.
I. Minimising sediment loads entering reservoirs There are three common ways of achieving this objective: • catchment conservation prograrIh'TIes to minimise sediment yields. Land use practices, agriCUltural methods and engineering measures to control erosion all feature in this category. o upstream trapping of sediments. Check dams and veget~tion screens can be used to intercept sediments on their way to downstream reservoirs . • bypassing of high sediment loads. The principle here is to fill the reservoir at low to medium flows when sediment concentrations are low and to bypass high flows, with their high sediment content, around reservoirs. This can be achieved using bypass channels or tunnels or by having the reservoir 'off line'.
2. Minimising deposition
of sediments in reservoirs
There are two main ways of passing sediments through reservoirs without deposition: sluicing - the process of passing sediment laden flo09 waters through the reservoir. This method involves the reduction of water levels in the reservoir during the flood season and is applicable mainly to very fine sediments (clays and silts). " density current venting - this method has the attraction that it is not necessary to lovver water levels but is only applicable in very exceptional circumstances where sediment-induced density currents carry very fine sediments towards the dam. The number of cases where density current venting has been successful is minimal. a
3. Removing accumulated sediments from reservoirs Hydraulic and mechanical methods are available for removing sediment which has already accumulated in reservoirs:
iii
EVACUATION OF SEDIMENTS
• flushing - the process of re-entraining deposited sediments and passing the sediment laden flow through low level outlets in the dam. This method involves the reduction of water levels in the reservoir, it consumes significant quantities of water but is capable, under certain circumstances, of removing coarser sediments (mainly sand sizes). It removal of sediments using dredging or mechanical means this method is feasible but usually requires reservoir levels to be maintained at low levels for extended periods of time. It is expensive in itself, quite apart from the loss of benefits from the reservoir during dredging operations. The disposal of large quantities of sediment often presents problems. All the above methods can be used to extend the useful life of reservoirs. However, their technical, economical and environmental feasibility depend on a number of specific factors including: It
• It
• • • • • •
the availability of suitable engineering facilities at the dam to control water levels and outflows the availability of 'surplus' water and its value if used for other purposes the predictability of dver flows, including seasonal variations the characteristics of the sediments entering, and within, the reservoir the availability of disposal sites for dredged sediments the effects on the downstream reach of evacuating sediments through the dam th~ effects of sediment management on the normal operation of the scheme and the financial and social consequences of the measures taken the effects of sediment management on other reservoirs within the region institutional and political problems among the affected stake-holders.
The objective of making reservoirs more sustainable using sediment management techniques is clearly laudable. However, the techniques are not applicable to all reservoirs and some dams will inevitably need to be either raised to regain storage or decommissioned and possibly replaced elsewhere. However, there are fewer and fewer good dam sites available and new dams can have sedous environmental and social consequences. This book is concerned principally with one of the methods of removing previously deposited sediments from reservoirs, namely the flushing of sediments through purpose-built outlet works within the dam. This technique can be applied to existing dams (with adaptation of the engineering works) and to new dams. However, the technique is only effective under certain favourable conditions and is not applicable universally. Dams designed within the last ten years or so, have sometimes incorporated design features which will allow flushing to be undertaken when appropriate. However, these designs have been based on considerations which are site specific. The purpose of this book is to give guidance on the necessary hydrological, hydraulic, sedimentological and topographical features for successful flushing. It is based on a review of recent research and field experience
iv
PREFACE
worldwide and draws together this existing knowledge into a concise manual for practising engineers. The book begins by assessing the scale of the problem of reservoir sedimentation. It assesses the volume of storage that is likely to be lost to sedimentation and compares this volume with the net volulne of storage that is likely to be required to meet continuing demand. The book provides a review of the current state of knowledge of reservoir flushing, and then considers the worldwide experience of flushing to date. Areas of the world are then identified where flushing is likely to be most useful. The final section of the book describes the more detailed investigations which must be carried out when considering sediment flushing at a particular dam site.
Rodney lVhite developed his interest in hydraulics at Leeds University from where he gained his first degree in 1962 and his PhD in 1965. He joined the Hydraulics Research Station, as it was, in 1965 and specialised in flo1;V measurement and sediment transport during his early career. He led the River Engineering Department of HR Wallingford before becolning the Research Director in 1990. In more recent years he has been a consultant to the firm with a remit to develop and apply new technologies, particularly with regard to sediment related issues. His research has resulted in internationally accepted theories that explain the movement of sediment in rivers, the resistance of naturally fonned alluvial· channels, the equilibrium, size of natural channels and their plan form characteristics. He has extensive practical experience of sedim.entation in rivers ., and reservoirs worldwide. He has written several books and many scientific papers based on his research and on his specialist consultancy assignments and he is currently the editor of the International Association of Hydraulic EngineeringResearch (IAHR) lournal of Hydraulic Research.
v
Acknowledgements This book describes work which was funded principally by the Department of the Environment, Transport and the Regions (DETR) as part of the Partners in Innovation programme. The work was undertaken jointly by HR V/allingford, as the lead partner, TAMS UK, Binnie Black and Veatch, and LAWGIBB. It is a pleasant duty to acknowledge the valuable contributions made by Laurence Attewill and Atila Bilgi of TAlVIS UK, Ed Atkinson and Andrew Nex of ILl{ Wallingford, John Ackers, Chris Scott and Robert Jones of Binnie Black and Veatch, and Richard Wingfield and Mary-Ellen Cromack of LAWGIBB. HR Wallingford is an independent specialist research, consultancy, software and training organisation that has been serving the water and civil engineering industries worldwide for over 50 years in more than 60 countries. \Ve aim to provide appropriate solutions for engineers and managers working in: • water resources II»
o
• • • • o
•
h~gation
groundwater urban drainage rivers tidal waters ports and harbours coastal waters offshore.
Address: Internet:
Howbery Park, Wallingford, Oxon, OXlO 8BA, UK http://www.hrwallingford.co.uk
vii
Notation DDR
DSOT D50B F~VR'
H fiush
Hmax
LTCR lv!in LV
P sand
Qf Qm Qs S
SBR SBR d SSR TE
Tf TIVR \tV
¥Ibed
the ratio of the height of water at the dalTI during flushing to the maximum height of water at the dam (to reservoir retention level), both measured above original river-bed level (nondimensional) 50 percentile size of sediment in transport (IILrn) 50 percentile size of river-bed material (mm) the ratio of the natural width of the flushing channel and a representative bed width for the reservoir (non-dimensional) the height of water at the datu during flushing, measured above original river-bed level (m) the maximum height of water at the dam (to reservoir retention level), measured above original river-bed level (m) the sustainable storage capacity divided by the original storage capacity of the reservoir (non-dimensional) the average sediment inflow rate (t/yr) the interval between flushing operations (yrs) proportion of total sediments in motion which exceed 0-06 rom in size (sand and coarser material) the flushing discharge (m3/s) mean annual flow (m 3/s) the sediment transporting capacity of the flow in the incised flushing channel (tis) the longitudinal energy gradient trliough the reservoir (nondimensional) the ratio of sediment flushed to sediment depositing (nondimensional) the specific value of SBR related to flushing with maximum reservoir drawdown (non-dimensional) .~ sand-size ratio, DsoTIDsoB the trapping efficiency of the reservoir, i.e. ratio of sediments retained within the reservoir to sediments~.;,entering (nondimensional) . the duration of flushing (days) the ratio of the natural top width of the flushing channel and a representative top width for the reservoir the bed 'width of the incised flushing channel (m) the representative bottom width of the reservoir, taken as the bottom width of the reservoir just upstream of the dam (m)
ix
EVACUATION OF SEDIMENTS
Wtep Wmin
a p
"¥
x
the representative top width of the reservoir, taken as the top width of the reservoir just upstream of the dam (m) the lesser of Wand Wbed (m) the angle of the side slope of the incised channel formed during flushing (zero is horizontal) (degrees) the density of the deposits expressed as weight of dry material per unit volume (tlm3 ) constant related to the sediment type (non-dimensional)
I
Contents
Dlustrations
1.
Executive summary 1.1. 1.2.
2.
xv
Review of sedimentation 2.1. 2.2.
2.4.
2.5.
2.6.
1
Introduction, 3 Summary of conclusions, 4 1.2.1. Review of sedimentation in reservoirs, 4 1.2.2. Research into factors Wllich influence sediment flushing, 7 1.2.3. vYorldwide experience of sediment flushing, 9 1.2.4. Geographical areas suited to flushing, 11 1.2.5. Site-specific investigations and design considerations, 13
reservoirs
15
Summary, 17 World total reservoir storage, 17 2.2.1. ICOLD World register of dams, 17 2.2.2. Other sources, 18 2.2.3. Conclusion, 18 vVorldwide distribution of existing storage, 18 2.3.1. Global water resources, 18 2.3.2. Geographical distribution, 19 World demand for more storage, 19 2.4.1. Population, 19 2.4.2. Irrigation, 21 2.4.3. Hydropower, 21 2.4.4. Conclusion, 23 Distribution of demand for more storage, 23 2.5.1. Europe, 23 2.5.2. North America, 24 2.5.3. South and Central America, 25 2.5.4. Aflica, 26 2.5.5. Asia and Oceania, 27 2.5.6. Summary, 28 Rate and distribution construction of new
xi
EVACUATION OF SEDIMENTS
2.7. 2.8. 2.9. 2.10.
xii
reservoirs, 28 2.6.1. Worldwide, 28 2.6.2. Distribution of construction of storage, 30 2.6.3. Comparison of storage construction with demand, 30 Rate and distribution of loss of storage, 31 2.7.1. Rate of loss of storage, 31 2.7.2. Distribution of loss of storage, 32 Trends in the rate of loss of storage, 34 Reservoir size and rate of loss of storage, 35 Requirements for new storage, 36
3.
Research into factors which influence flushing 3.1. Introduction, 39 The mechanism of flushing, 40 3.2. The development of criteria for successful 3.3. flushing, 42 3.3.1. Sediment balance, 42 3.3.2. Sustainable reservoir capacity, 47 3.3.3. Evaluation of flushing criteria, 50 3.3.4. Practical criteria for successful flushing, 50 Summary of the requirements for effective 3.4. flushing, 58 3.4.1. Hydraulic conditions required for efficient flushing,· 58 3.4.2. Quantity of water available for flushing, 59 3.4.3. Mobility of reservoir sediments, 59 3.4.4. Site-specific factors, 60 3.4.5. Constraints on the ultimate capacity achievable by sediment flushing, 60 3.4.6. Economic assessment, 60 3.4.7. Summary, 60 Numerical models, 61 3.5.
37
4.
Worldwide experience of sediment flushing Introduction, 65 4.1. Flushing, 66 4.2. Worldwide experience of flushing, 67 4.3. 4.3.1. Overview, 67 4.3.2. Flushing teclmiques, 67 4.3.3. Sediments flushed, 70 Case studies of reservoir flushing, 71 4.4. 4.4.1. Summary, 71 4.4.2. Findings, 81 4.4.3. Summary of findings, 88
63
CONTENTS
5.
Geographical areas suited to flushing 5.1. Worldwide variation in erosion rates, 93 5.1.1. Factors that affect erosion, 93 5.1.2. Estimates 9.f global sediment yield, 93 5.1.3. Maps of global variation in sediment yields, 100 5.2. Climatic zones of the world, 101 5.2.1. Introduction, 101 5.2.2. Precipitation regimes and their seasonal variation, 101 5.2.3. Koppen classification, 108 5.2.4. Relationship between climate zone and erosion rates, 115 5.3. Geographical areas suitable for flushing, 120 5.3.1. Introduction, 120 5.3.2. Factors affecting erosion rates, 120 5.3.3. Sediment delivery ratio, 122 5.3.4. Hydrological characteristics, 123 5.3.5. Areas of the world which are best suited to reservoir flushing, 123
6.
Site-specific investigations and design considerations
125
7.
References
131
8.
Bibliography
141
Appendices Appendix 1. Appendix 2. .8..Dlue:nWlX 3.
91
149 Reservoir data, 151 Numerical model case study, 163 Flushing case studies, 171 Erosion, 211
251
xiii
I!' ..
I Illustrations
Tables
Table 1.L Table 1.2. Table 1.3. Table 1.4. Table 1.5. Table 2.1. Table 2.2. Table 2.3. Table 2.4. Table 2.5. Table 2.6. Table 2.7. Table 2.8. Table 2.9. Table 2.10. Table 2.11. Table 2.12. Table 2.13. Table 2.14. Table 2.15. Table 2.16. Table 2.17. Table 2.18. Table 2.19. Table 2.20. Table 2.21. Table 2.22. Table 3.1.
Demand for storage Geographical demand for new storage Demand for new storage, South and Central A.merica Demand for new storage, Africa Gross storage requirements to 2010 Distribution of reservoir storage volume Growth in 'world population Prediction of global demand (after Shiklamanov) Growth in irrigation area Growth in energy generated by hydropower Comparison of actual andeconolTlJcal potential energy Annual gro\tvth rates and increase in storage European growth in irrigation and hydropower North American growth in irrigation and hydropower South American growth in irrigation and hydropower Annual growth rates, South and Central America African growth in irrigation and hydropower Asian and Oc~anian growth in irrigation and hydropower Annual growth rates, Asia and Oceania Regional demand for new storage of construction of new storage Distribution of storage increase Regional sedimentation rates Extent of sediment data Distribution of sediment rate and storage loss Ringlet reservoir, sedimentation Gross requirement for new storage Application of sediment balance and long-term capacity ratios to existing reservoirs
5 5
6
6 8
19 20 21 21 22 22 23 24 25 26
26 27
28 29 29
29 30 31
32 33 34 35
46
xv
EVACUATION OF SEDIMENTS
Table 3.2.
Table 3.3. Table 3.4. Table 3.5. Table 3.6. Table 4.1. Table 4.2. Table 4.3. Table 4.4. Table 4.5. Table 4.6. Table 5.1. Table 5.2. Table 5.3.
Table 5.4. Table 5.5. Table 5.6. Table 5.7. Table 5.8. Table 5.9. Table 5.10. Table 5.1l. Table 5.12. Table 5.13. Table 5.14. Table 5.15.
Figures Figure 2.1. Figure 2.2. Figure 2.3.
xvi
The relative importance of the discharge used for flushing and the duration of flushing for a particular volume of flushing water The influence of sediment size on the amount of sediment removed from reservoirs Effect of the sediment size ratio on sediment balance Effect of the proportion of sand and coarser material on extending the life of reservoirs Application of constraint criteria to existing reservoirs. Summary of reservoirs flushed Summary of experience in flushing Distribution of flushing experience by purpose Summary of flushing techniques Detailed list of reservoirs subject to flushing Summary of key flushing parameters Continental variations in sediment yield (Mahmood, 1987) Continental variations in sediment yield (Jolly, 1982, taken from Gregory and Walling, 1973) World maximum recorded suspended-sediment yields greater than 2000 tlkm2/yr (Jolly, 1982, from Gregory and Walling, 1973) Rates of sediment yield for the world's maj or rivers at ocean level, excluding basins with an area less than 10000 km2 (Mahmood, 1987) Values of sediment yield in excess of 10000 t/km2/yr (Walling and Webb, 1983) Colombia, SON, elevation 65 m India, 13 oN, elevation 22 m Wadi HaIfa, Sudan, 22°N, elevation 160 m England, 51·5°N, elevation 5 m Calgary, Canada 51°N, elevation 329 m Italy, 42°N, elevation 131 m Greenland, 81·5°N, elevation 35 m Antarctica, 66·5°S, elevation 30 m Reasons for combining climates into homogeneous climatic groups (Jansson, 1988) Countries classified into climatic zones showing number of river basins in each zone (modified from Jansson, 1988) Growth in world population Comparison of growth rates Historic growth in reservoir storage
52 54 55 56 58 68 69 70 70 72 74 94 94
96 97 99 112 112 113 113 114 114 115 115 117 119 20 23 30
ILLUSTRATIONS
Figure 2.4. Figure 2.5. Figure 3.1.
Figure 3.2. Figure 3.3. Figure 3.4.
Figure 5.1. Figure 5.2. Figure 5.3. Figure 5.4. Figure 5.5. Figure 5.6. Figure 5.7. Figure 5.8. Figure 5.9.
Storage lost to sedimentation Reservoir size and rate of loss of storage Longitudinal profiles during flushing: (a) flushing with full drawdown; (b) flushing with insufficient drawdown; (c) final conditions after a long period of flushing with insufficient drawdown Channel widths formed in reservoir deposits during flushing Cross-sections of flushing channels: (a) Heisonglin reservoir, China; (b) Sanmenxia reservoir, China Simplified reservoir geometry for application of capacity criterion: (a) actual reservoir plan; (b) fitted reservoir plan; (c) simplified reservoir plan and sections; (d) simplified reservoir elevation; (e) enlarged section immediately upstream of dam Global patterns of sediment yield: (a) after Strakhov (1967); (b) Fournier (1960) Global patterns of suspended sediment yield: (a) from Lvovich (1991) in \Valling and Webb (1996); (b) from Walling and Webb (1983) Annual precipitation for 1998 in mm per month Precipitation distribution during winter 1998 (December to February) Precipitation distribution during spring 1998 (March to May) Precipitation dist.ribution during summer 1998 (June to August) Precipitation distlibution during autuIllll 1998 (September to November) . Climates of the world according to the Koppen classification Number of basins within sediment yield classes in climatic groups
33 35
41 44
48
49 102 104 106 107 109 110 111 116 118
xvii
e u
rn a y
;'.: '
-- ..
....
.
I. Executive summary 1.1.
INTRODUCTION There are around 40 000 large reservoirs worldwide used for water supply, power generation, flood control, etc. Between a half and one per cent of the total storage volume is lost annually as a result of sedimentation and 300 to 400 dams, at the cost of around £5 million per dam, would need to be constnlcted annually to maintain current total storage. The introduction of flushing systems in some old dams, where appropriate, and in the design of new dams could save 10 per cent of these costs, i.e. £200 million annually. This book provides guidelines on the design aspects of flushing systems and indicates where such systems could be used beneficially. ' The benefits attributable to dams and reservoirs, most of which have been built since 1950, are considerable and they have improved the quality of life worldwide. These benefits can be classified under three main headings.
Irrigation About 20 per cent of cultivated land worldwide is irrigated, some 300 million hectares. This irrigated land produces about 33 per cent of the worldwide food supply. Irrigation accounts for about 75 per cent of the world water consumption, far outweighing the domestic and industrial consumption of water. Hydropower About 20 per cent of the worldwide generation of electricity is attributable to hydroelectric schemes. This equates to about 7 per cent of worldwide energy usage. Flood control and storage Many dams have been built with flood control and storage as the main motivator, e.g. the Hoover dam, the Tennessee Valley dams and some of the more recent dams in China. In many areas of the world the life span of these reservoirs is determined by the rate of sedimentation which gradually reduces storage capacity and eventually destroys the ability to provide water a.Tld power when sedinlents clog low level outlets. Many major reservoirs are approaching this stage in their life. One way of preserving reservoir storage is to flush sediments through purposebuilt outlet works within the dam. This technique can be applied to existing dams (with adaptation of the works) and to ndw dams. However, the technique is only effective under certain favourable· conditions and is not applicable universally. The alternative is to build more dams to replace the
3
EVACUATION OF SEDIMENTS
depleting storage of the existing stock. However, there are fewer and fewer good dam sites available and new dams can have serious environmental and social consequences. Dams designed within the last ten years or so, have sometimes incorporated design features which will allow flushing to be undertaken when appropriate. However, these designs have been based on considerations which are sitespecific. The purpose of this project was to produce a generally applicable design manual which provides guidance on the necessary hydrological, hydraulic, sedimentological and topographical features for successful flushing. It is based on a review of recent research and field experience worldwide and draws together this existing knowledge into a concise manual for practising engineers. The book starts by assessing the scale of the problem of reservoir sedimentation. It compares the volume of storage that is likely to be lost to sedimentation and compares this volume with the net volume of storage that is likely to be required to meet continuing demand. The book provides a review of the current state ofknowledge of reservoir flushing, and from this proceeds to consider the worldwide experience of flushing. Areas of the world are then identified where flushing is likely to be most useful. The final section of the book describes the more detailed investigations which must be carried out when considering sediment flushing at a particular dam site.
1.2.
SUMMARY OF CONCLUSIONS
-.i
1.2.1. Review of sedimentation in reservoirs World storage The best estimate of world storage in reservoirs (excluding natural lakes used as storage for power and irrigation) is 6815 km3 • Distribution of storage The worldwide distribution of existing storage and storage under construction, as determined from the International Commission on Large Dams (ICOLD) Register (1998), is shown in Table 2.1. The Americas, together with northern Europe and mainland China, account for 70% of the existing world stock of reservoir storage. Demand for more storage The world population in 1990 is estimated to have been 5286 million, growing at an annual rate of 1·5%. This rate of growth is forecast to decline in the coming decades so that the predicted future world population is as shown in Table 2.2 and Figure 2.1. Water demand is expected to continue to grow at a faster rate than that predicted by population growth alone. Much of this demand will be satisfied by
i
-~
4
J
EXECUTIVE SUMMARY
Table 1.1.
Demand/or storage
Period
Increase in net storage: km3
Annual growth rate: %
increased surface and groundwater abstraction; water re-use and no direct linkage between overall demand and water storage can be assumed. An estimate of the growth in total water demands by Shlklamanov is given in Gleich (1993), which shows that t.h.e rate of growth in demand is consistently higher than population growth rate and that the contribution of storage to the total supply is greater still, as shown in Table 2.3. From the rates of growth for population, water consumption, irrigation area and hydropower, the following growth rates for demand of storage are postulated, and are shown in Table 1.1. The forecast future demand for storage is shown in Table 1.2.
Distribution of demand Europe. Although the dema.'1d for new storage is sensibly zero in much of Western Europe, it does appear that for the region as a whole there is a small demand, of the order of 1% per annum, for new storage for hydropower, mainly concentrated in Eastern Europe. North America. Although the data show that the energy generated by hydropower, as well as the area of land under irrigation, continued to ,grow through to the 1990s, the fact that no new storage was constructed in that period ..;"'.... ~;".::, ....'v .. ..., that the data are influenced by operational factors. Therefore, although Table 1.2.
Geographical demand/or nevv storage Demand for new storage: km3
Region
49
51
54
South and Central America
467
495
424
Africa
167
203
248
5
EVACUATION OF SEDIMENTS
Table 1.3.
Demand for new storage, South and Central America
there remains a large undeveloped potential resource in Canada, environmental pressures will probably preclude any further development. South and Central America. The data show that the energy generated by hydropower, as well as the area of land under irrigation, grew strongly through the 1980s and into the 1990s. Furthermore, it is estimated that only 21 % of the economically-feasible hydropower potential has so far been developed, so that short to medium term growth is unlikely to be constrained by shortage of sites. From the rates of growth for population, water consumption, irrigation area and hydropower, the growth rates for demand of storage, shown in Table 1.3, are postulated.
The data show that the energy generated by hydropower, as well as the arya of land unde~ irrigation, grew weakly through the 1980s and into the 1990s at a rate well below the rate of growth of population. This trend is likely to continue, despite strong demand and great potential, so an annual growth rate of 2% is postulated.
Africa.
Given the rates of growth for population, irrigation area and hydropower, the growth rates for demand of storage, shown in Table 1.4, are postulated.
Asia and Oceania.
New reservoirs under construction Worldwide. The historic rate of construction of storage worldwide is shown in Table 2.13 and Figure 2.3. The overall growth rate for the century as a whole has been 6·5%. It is interesting to note that neither of the World Wars nor the Depression made any serious impact on the rate of growth: in this context the Table 1.4. Period 2000-2010
')0')0..')010 ~
6
Demand for new storage, Africa Annual growth rate: % 2-0
EXECUTIVE SUMMAR.Y
apparent fall in the rate of construction during the 19908 is dramatic and may, in part, be due to inadequate data. The distribution of the growth of new storage is set out in Table 2.14.
Rate and distribution of loss of storage In order to assess the variation in the rate of loss of storage around the world, data from approximately 2300 dams in 31 countries have been gathered and analysed. The summary of the results of the analysis is given in Table 2.15. The estimates of annual loss of storage owing to sedimentation have been used in conjunction with the gross storage volume data available in the ICOLD World Register of Dams to estimate the magnitude of the sedimentation problem. The results of the analysis are displayed in Figure 2.4. In summary, the analysis shows that by the year 2000 approximately 567 km3 (10% of the current gross available storage in the world) has been lost to sedimentation. From the data available from the 1325 registered dams under construction, it can be seen that the average gross storage volume of new reservoirs is approximately 370 M.m3 •
Rate of loss of storage The rate of loss of storage for a given reservoir is dependent on the rate of erosion of the catchment. In regions where the catchments have remained stable, e.g. N orthem Europe and North America, the rate of loss of storage is constant. In regions where deforestation has occurred the· rate of catchment erosion and consequently the rate of loss of storage increases.
Reservoir size and rate of loss of storage The highest rates of loss of storage are found in the smallest reservoirs and the lowest rates in the largest. Of the 1105 reservoirs studied, 730 have a storage volume of less than 1233 M.m3 and an average rate of loss of storage in excess of 1% per annum. At the other extreme, 23 of the reservoirs studied had a storage volume in excess of 1233 M.m3 and an average rate of loss of storage of 0·16% per annum.
Requirement for new storage New storage will be required in the future both to satisfy increasing demarld generated by the growing world population and to replace the storage lost owing the next decade to sedimentation. The estimate of the gross storage required is shown in Table 1.5.
1.2.2. Research into factors which influence sediment flushing For effective flushing the following factors need to be considered and satisfied.
Hydraulic conditions required for efficient flushing Riverine conditions must be created in the resel'loir for a significant length of time. The reservoir level must be held low throughout the flushing period, possibly with minor fluctuations in level to activate sediment movement. To achieve this:
7
EVACUATION OF SED1MENTS
Table 1.5.
Gross storage requirements to 2010 Storage volume: km3
Continent Gross requirement
New demand 2000-2010
Loss to sedimentation up to 2000
49
54
103
North America
0
112
112
South America
467
17
484
Aftica
167
35
202
Asia and Oceania
315
349
664
998
567
1565
Europe I.
Total
I i
• the hydraulic capacity of the bypass must be sufficient to maintain the reservoir at a constant level during the flushing period • flushing discharges of at least twice the mean annual flow are required • flushing volumes of at least 10% of the mean annual run-off should be anticipated.
Quantity of water available for flushing There must be enough water available to transport the required volume of sediment. This has the following implications. • Reservoirs where the annual run-off is large compared with the volume of the reservoir are suitable for sediment flushing. • Reservoirs where there is a regular annual cycle of flows and a defined flood season are suitable for sediment flushing. This favours sites in monsoon areas and sites where flood flows are generated by annual snowmelt in the spring and summer months. • Reservoirs where release of significant quantities of water for flushing does not significantly affect the ability to satisfy water demands at other times of the year.
Mobility of reservoir sediments The nature and quantity of river sediments are important factors in determining whether the quantity of water available for flushing is adequate to remove the desired quantity of sediment from the reservoir. • Graded bed sediments produce conditions which are the most conducive to the efficient flushing of sediments. Such conditions are typical of gravel rivers with a varying bed material composition. In large rivers this situation is found where the longitudinal bed gradient is between, say, 0·001 and 0·002. In smaller rivers the equivalent range nlay be 0·002 to 0·005.
8
EXECUTIVE SUMMARY
c
From the point of view of sediment alone, delta deposits of fine sand and coarse silt are the most easily flushed. Coarser material is difficult to move and tends to deposit at the upstream end of the reservoir. Finer Inaterial which deposits in the body of the reservoir outside any incised channel will not be available for reworking during flushing.
Site-specific factors The most suitable conditions for flushing are to be found in reservoirs that are approximate i.T1 shape to the incised channel which develops during flushing. Long, relatively narrow, reservoirs are better suited to flushing than short, wide, shallow reservoirs. Summary Reservoirs in tb.e 'upper and middle reaches' of rivers are likely to be best suited to sedinlent flushing for the following reasons. \I
fII
II
In the lower reaches, reservoirs are likely to have inundated areas that have previously been flood plains and these areas would not be reached by the incised flushing channel which is inevitably of limited width. The longitudinal slope available for the flushing channel is relatively small, thus limiting the amount of sediment transport. Reservoir volumes in the lower reaches are likely to be larger compared with . run-off and hence water availability becomes a restraint on the mean ap. .11ual sediment flushing. .-
1.2.3. \j\/orldwide experience
of sediment flushing
The findings from the review of worldwide experience of flushing can be sum..rnarised as fonows.
The hydrology and sedimentology of the catchment The hydrology and sedimentology of the catchment need to be understood fully in the planning of flushing facilities for new or existing reservoirs and to provide the background for analyses of past sedimentation and flushing perfonnance.
The storage capacity of the reservoir Successful hydraulic flushing is more likely to be practicable in reservoirs which are small hydrologically, with a storage capacity less than 30% of the mean annual inflow. The smaller the reservoir, the greater the chance of it being successfully flushed and the the likely residual storage capacity. The sediment deposition potential Flushing is vital for the preservation of long-term storage in reservoirs the sediment deposition potential is greater than 1 to 2% of the original capacity. Even in reservoirs with a potentially long life, consideration should be given to possible eventual decommissioning problems when deciding whether or not to flush.
9
EVACUATION OF SEDIMENTS
The shape of the reservoir basin The shape of the reservoir basin can have a large impact on the practicability of effective flushing and the residual storage capacity. Narrow steep-sided reservoirs in valleys with a steep longitudinal slope are the easiest to flush. Wide valleys, where the impoundment covers fonner floodplains, can be flushed less effectively, because the deposits tend to consolidate and are remote from the flushing channel. The low-level outlet facilities provided For effective empty flushing with full draw down , the low-level outlets must be both low enough and of sufficient capacity to allow the drawdown to be controlled during the'time of year when flushing is undertaken. Proportionately - larger outlets are required for flood-season flushing than for flushing outside the . flood season. Operational limitations Operational considerations, such as water and power demands, can inhibit the ability to flush successfully, but they must not be allowed to prejudice the longterm preservation of an important resource. The deployment of full or partial drawdown Full drawdown and empty flushing have been found to be much more effective than partial drawdown~ The scope for enhancements to flushing Fluctuations in water level and discharge during flushing are beneficial to the promotion of bank slumping and increasing the rate of sediment discharge. Also, the deployment of lateral and longitudinal diversion channels has been successful in promoting flushing in reservoirs which are large hydrologically or contain significant proportions of deposition in areas remote from the main flushing channel. Downstream impacts Downstream impacts can act as a constraint in the planning and operation of flushing. In some cases flushing may be ruled out, whereas sluicing, which approximately preserves the seasonal distribution of sediment load, maybe a practicable alternative. Value of sediment flushing The degree of success in flUShing should be judged by whether it makes a worthwhile difference to the beneficial uses of the reservoir, rather than simply by whether it meets numerate objectives, such as a long-term balance between inflows and outflows, or the retention of a certain percentage of the original storage volume.
10
!"
•
y
EXECUTIVE SUM MARY
1.2.4. Geographical areas suited to flushing E.rosion rate The erosion rate depends on a complex interaction of the following factors. • Climate: precipitation and run-off, temperature, wind speed and direction . ., Geotechnics: geology, volcanic and tectonic activity, soils. e Topography: slope, catchment orientation, drainage basin area, drainage density. • Vegetation. • Land use and human impact. factors are discussed in Appendix A4.1. It is not easy to generalise between areas of high and low erosion rates depending on their geographical location. Estimates of average global rates of denudation have ranged from 0·06 to 0·16 mm/yr (Morris and Fan, 1997). This is equivalent to estimates of between 15 and 20 x 109 trKTn?/yr (vValling and vVebb, 1996). Areas with sediment yield over 1000 t/km2/yr are 8·8% of the total land area and account for 69% of the total sediment load. Regions with less than 50 t/km2/yr account for about half of the land area and 2·1 % of the sediment yield. Case studies of erosion rates are presented in Appendix A4.2.
Transport
of sediment
In order for reservoir flushing to be needed, it is necessary for sediment to be eroded in the catchment, transported down the river system and deposited in the reservoir. The efficiency of the transport process is expressed by the sediment delivery ratio, which is the proportion of sediment eroded from the land that is discharged into rivers (Morgan and Davidson, 1986). The sediment delivery ratio is generally higher for sediment derived from channel-type erosion which delivers sediment to the main channels of the transport system more quickly and directly than in the case of sheet erosion. The poor correlation between sediment yield and erosion rates makes it difficult to estimate the sediment load entering a reservoir on the basis of erosion rate within the catchment (Morris and Fan, 1997). Most studies that have attempted to relate the delivery ratio to catchment characteristics have found that the delivery ratio decreases as the catchment area increases (Walling and Vvebb, 1983).
Climatic zones An understanding of the precipitation regimes throughout the world may allow the definition of climatic zones based on temperature and precipitation regimes. This may permit definition of areas of high and low erosion rates. It is difficult to classify distinct climatic zones as they tend to merge into one another rather than have sharp boundaries, but a number of general models have been produced.
II
EVACUATION OF SEDIMENTS
There have been many climatic classifications produced but one of the most common is based on the original Koppen classification, with eight climatic regions based on four temperature zones and one moisture zone and the seasonal domination of air masses. Details of this classification are given in Chapter 5 and a discussion of alternative classifications is given in Appendix A4.3. The eight Koppen climatic regions are as follows. • Tropical wet: classification Af. • Tropical wet and dry: classification Aw, Am and BS. e Tropical desert: classification BW. e Mid-latitude wet: classification Cf and Df. • Mid-latitude winter dry: classification Cw and Dw. • Mid-latitude summer dry (Mediterranean Climate): classification Cs and Ds. • Polar wet and dry: classification ET. • Polar desert: classification EF.
Hydrological characteristics Experience has shown that low reservoir water levels provide the most effective conditions for sediment flushing. To allow water levels to be lowered requires confidence that rainfall can be relied upon to refill the reservoir. It follows that well defined wet and dry seasons will be favourable for a sediment flushing regime. Such a climate is defined by Koppen as tropical wet and dry: Aw, Am and BS. Also, there are areas in the mid-latitudes where spring snowmelt provides a regular and predictable annual pattern of high flows. River discharges must also be sufficient to transport sediment loads through the reservoir. Regions of low precipitation like the Sahara and other desert environments therefore will not be suitable for flushing even if they exhibit a defined seasonal effect. The availability of water will also affect the duration and discharge rate of the flow required for flushing. Where there is a limited amount of water it is better to use a high discharge for a short period of time than a low discharge for a long period of time. This increases the amount of sediment that is removed. Areas of the world which are best suited to reservoir flushing It is not possible to define precisely which specific areas of the world will provide conditions for successful flushing. In reality there is a spectrum of conditions ranging from those sites where conditions are ideal to those sites which are quite unsuited to sediment flushing. From the Koppen classification of climatic zones and the mid-latitude spring snowmelt effect, the requirements for successful flushing are most likely to be met in the following locations: . • parts of Central America extending into South America • areas in North and South America where the rivers are fed by the Rockies and the Andes • parts of Central Africa from the Ivory Coast in the west to Sudan in the east
12
EXECUTIVE SUMMARY
• areas in Central Asia where the rivers are fed by the Himalayas, including Pakistan, India and Nepal & parts of Asia, including Calnbodia, Vietnam and Thailand.
1.2.5. Site-specific investigations and design considerations There are many detailed factors which need to be evaluated on a site-specific basis before the technical viability and economic soundness of sediment flushing can be confirmed. Chapter 6 provides details of the nature of these site-specific investigations, including design considerations for the sediment bypass itself. There are numerous stages for such investigations, as follows.
Site investigations Site investigations are required to identify t.~e most compact and efficient geometry for the flushln.g outlets and the energy dissipation works. The reservoir itself requires a detailed survey to establish its topography. Hydrological investigations Inflows to the reservoir need to be established with confidence. This involves the acquisition of historical records of river flows going back at least 30 years and preferably longer and/or the development of a longer sequence from rainfall records using catchment modelling. Sediment investigations The amount and nature of the sediment entering, or likely to enter, the reservoir needs to be established. This requires measurements of sediment transport rates in the rivers feeding the reservoir over many years to establish the results with the confidence that is required. In the case of existing reservoirs, information about the amount of sediments entering the reservoir can be augmented by surveys of the amount and nature of the material settling within the reservoir. is required, however, to allow. for the amount of material, mainly fine, which passes through the reservoir without deposition. Bed material sampling should be undertaken in the reservoir and in the rivers which feed the reservoir. A sound knowledge of the nature of these sediments, including their size, specific gravity and degree of compaction, is an essential requirement to provide inputs for numerical models which simulate sediment moven1ent, see below.
Hydraulic modelling Numerical (computer) modelling of the way sediment is likely to behave within the reservoir and the amount and nature of the sediment which will be passed to the downstream reach is the cornerstone of any detailed evaluation of flushing facilities. Computer simulations of reservoirs ideally use representative, long-term sequences of water and sediment inflows to the reservoir. The models are capable
13
EVACUATION OF SEDIMENTS
of looking at the effectiveness of various aspects which affect reservoir sustainability,over periods of up to 50 or 60 years, including: • measures to reduce the amount of sediments entering reservoirs, such as catchment conservation or upstream storage • measures to manage the sedimentation process within reservoirs, such as variations in the operating rule curves for the reservoir • measures to evacuate sediment from the reservoir, including dredging and sediment flushing.
System simulation modelling System simulation modelling is required to evaluate the conflicting demands of hydropower production, irrigation and other requirements, and must be able to assess the impacts of the various reservoir operating strategies.' The simulation model must be able to replicate the outputs of water and power under a range of operating strategies so that an optimal economic and technical solution may be identified. In addition, it nlust be possible to take account of the effects of other reservoirs upstream and downstream of the one under consideration. Economic and financial analysis The main aim of economic and financial analyses is to assist in the identification and selection of the most favourable sediment management option. For each option the most important factor, from the economic viewpoint, is to define the 'with' and 'without' project cases. These-will illustrate the net economic impact of the availability of water resources over time, including any seasonal variations. Evaluation of the impact of alternative investment phasing is also important.
J
J
14
•
t
f
atio •
er\f Irs
2. Review of sedimentation in reservoirs 2.1.
SUMMARY In this chapter a summary of the total volume of reservoir storage, and its distribution is given. An attempt is made to quantify the future demand for new storage, especially for hydropower and irrigation, and this estimate is compared with the historic rate of reservoir constnlction throughout the twentieth century. The rate of loss of storage due to sedimentation is made, so as to arrive at a prediction of both the net and gross future storage requirements. '
2.2.
WORLD TOTAL RESERVOIR STORAGE
2.2. I. ICOLD World register of dams The most recent ICOLD World register of dams was published in 1998 and was compiled from data collected from member, and some non-member, states in 1996. ICOLD required, in their circular instruction for reporting dam data,that respondents should include all dams with a height greater than 15 m and dams between 5 ill and 15 m in height with a storage of 3 M.m3 . or ' more. The introduction to the register qualifies the data as follows. s Japan reported only dams greater than 30 m high. = Russia reported mainly hydropower dams. '
o
.c_
""
_"
'' . '''., . . Some countries failed to respond 8....1J.d for these countrie$ .data:vvasretained from the earlier edition. --
The register gives the total number of dams reported by the 80 member countries and the 60 non-member countries as 25 410. No exact "sUlTilllary of storage volume is provided but in the introduction it is stated that the total volume of storage is 6000 km3 • From the analysis of the data in the register, the total gross storage volume of the reservoirs reported by ICOLD is 6465 km3• This includes 490 km 3 of storage registere9 as under construction. In order to estimate the total world storage it is necessary therefore to assess the extent to which the register under-reports the total number of dams and the number of dams less than 5 m high (and their storage). It is evident from the res!:ister that the ratio of dams less than 30 m high to the total number of dams varies from about 90% in the case of India to 5%~in the case of China. From this it can be infened that many countries, but China in particular, under-report dams
17
EVACUATION OF SEDIMENTS
in the 15 m to 30 m range. Therefore, it would seem reasonable to add a 20% allowance for under-reporting. Postulating an average storage volume of 10 M.m3 per dam, this will increase the total storage by 50 km3 More difficult is the assessment of dams in the range 5 m to 15 m with a storage of less than 3 M.m3 and all dams less than 5 m high. The 1995 National Inventory of Dams maintained by the US Army Corps of Engineers lists 74053 dams over 2 m high with at least 60 000 m3 capacity, compared with the ICOLD record of 6375 dams. The storage contributed by the 67 678 small dams not included in the ICOLD register are estimated at some 12 km3 , some 5% of the total. If the US data can be taken as typical for other countries, an allowance of 300 km3 should be made for storage provided by small dams. Thus· the total storage could be assessed at 6815 km3 •
2.2.2. Other sources The estimated total capacity of the world's reservoirs is given in Water in crisis (Gleich, 1993) as 7000 km3 , lJNESCO estimated in 1974 that the total storage of all reservoirs with capacities of 5 km3 and above to be 4050 km3 • This estimate was used by Mahmood (1987), who assumed an allowance of 20% for the storage provided by reservoirs less than 5 km3, to estimate total reservoir storage at 4880 km3 in 1987, when the total nurnber of registered dams was approximately 20 000. Increasing the storage pro-rata with the increase in number of dams gives a present day storage of 6345 km3 •
2.2.3. Conclusion The best estimate of world storage in reservoirs (excluding natural lakes used as storage for power and irrigation) is 6815 km3 ,
2.3.
WORLDWIDE DISTRIBUTION OF EXISTING STORAGE
2.3.1. Global water resources The total world fresh-water resources are estimated at 35 million km? (Morris and Fan, 1997). Of this, approximately 70% is locked up in the polar icecaps, glaciers and permafrost, and approximately 30% is stored as groundwater. The available water in lakes, rivers and swamps only accounts for 0,30% of the global fresh-water resources, Natural lakes are estimated to contain 91 000 km3 , while manmade lakes and reservoirs contribute 7000 km3 • The water stored in natural and manmade lakes and reservoirs is equivalent to 820/0 of the global annual precipitation of 119 000 km3 and is twice the global annual run-off of 47000 km3 ,
18
SEDIMENTATION IN RESERVOIRS
2.3.2. Geographical distribution The worldwide disuibution of existing storage and storage under construction, asdetermined from the ICOLD register is shown in Table 2.1. The Alnerica's together with Northern Europe and mainland China account for 70% of the existing world stock of reservoir storage.
2.4.
WORLD DEMAND FOR MORE STORAGE
2.4.1. Population The world's population in 1990 is estimated to have been 5286 million, growing at an annual rate of 1·5%. This rate of growth is forecast to decline in the coming decades so that the predicted future world population is as shown in Table 2.2 and Figure 2.1. Water demand is expected to continue to grow at a faster rate than that predicted by population growth alone. This is because the present -per capita Table 2.1.
Distribution of reservoir storage yolume
Region
North America
Number of dams
I
1498
South America Northern Europe
7205
I
2277
I I !
1845 1039 938
Average size of reservoir:
Fraction of world total
Gross storage: km3
M.m3
I I
I
29%
256
16%
694 412
15% I
3220
145
2%
45
Sub-Saharan Africa
966
575
9%
595
North Africa
280
188
3%
652
China
1851
649
10%
351
Southern Asia
4131
319
5%
77
44
148
2%
3364
277
117
2%
424
Pacific Rim
2778
277
4%
100
Middle East
895
224-
3%
250
25422
6464
100%
254
Southern Europe
Central Asia South-East Asia-
World total
19
EVACUATION OF SEDIMENTS
Table 2.2.
Growth in world population
Year
Population: millions
Annual growth rate in following' decade: %
1990
5286
1·53
2000
6158
I
1·34
I
I
2010
I
7032
1·15
I
i
2020
7887
2030
8671
0·72
2040
9318
0·54
2050
9833
0·95
demand in much of the developing world is constrained by lack of availability and is lower than that in the developed world: the total growth in demand is a combination of population growth and per capita growth. Much of this demand will be satisfied by increased surface and groundwater abstraction and water reuse. No direct linkage between overall demand and water storage can be assumed. An estimate of the growth in total water demands by Shiklamanaov is given in Gleich (1993), which shows that the rate of growth in demand is consistently higher than population growth rate and that the contribution of storage to the total supply is greater still, as shown in Table 2.3. Nearly 70% of the world demand for water is for irrigation. The bulk of the world's storage is for irrigation and hydropower purposes, or a combination of 12000
1·8
(J)
c
~
8000
'E
c0
~ '3
0..
0
a..
6000
-... ---------- -Population
10000
--- -- --
......
------ .....
-...----- -- ---
..... ------
--
_
1·6 1·4
~
..c:
0·8
4000
0·6 0·4
2000 0·2
__~--~--------~----~O
OL-------~------~--------~ 1990 2000 2010 2020
Year
Figure 2.1.
20
Growth in world population
*'
1·2 1Y
2030
2040
2050
~
0)
Cii ::I c c
<:
,I
SEDIMENTATION IN RESERVOIRS
Table 2.3.
Prediction of global demand (after Shiklamanov) Rate of growth of demand: %
Year
Total
.,
,
I
Reservoirs
1950-1960
3·88
13-47
1960-1970
2·67
11·12
1970-1980
2·51
6·16
1980-1990
2·21
3·54
1990-2000
2·31
2·61
,1 ~
~ !
.,
the two. It is pertinent therefore to examine the historic growth in irrigation and energy generation.
2.4.2. Irrigation The gro\vth in demand of water for irrigation is illustrated by the growth in irrigated area, given in Table 2.4. These data give only an ~indication of the growth of demand for water since they are a record of the irrigated area actually planted and are subject to annual variations in water availability: the effects of the drought in the 1980s is apparent. Moreover, no data are available on the proportion of the total demand met from storage reservoirs compared with other sources.
2.4.3. Hydropower Historic groV'lth \-Vorldwide power consumption is growing at a faster rate than population growth as nations industrialise. The share of energy generated by hydropower is difficult Table 2.4.
Growth in irrigation area
Year
Irrigated area: million ha
Annual growth rate in following 5 years: %
1975
190
2·22
1980
212
1·02
1985
223
1·23
1990
237
1·87
1995
260
21
EVACUATION OF SEDIMENTS
Table 2.5.
Growth in energy generated by hydropower
1990
2 112000
1995
2474000
3·21
to predict as the viability of new schemes is influenced heavily by the price of oil and gas and the safety and environmental concerns associated with nuclear power. The cost of new hydroelectric schemes increases in real terms as the best sites are used up. The historic energy generated by hydropower is shown in Table 2.5. As with the data for irrigation, the link between these data and reservoir storage is tenuous: the energy generated is affected by water availability and does not directly reflect either the growth of installed capacity or storage. The high growth rates in the 1970s is a reflection of the quadrupling of oil prices in 1972.
Potential The potential for new hydropower development is indicated in Table 2.6, which compares average energy generated by hydropower with the estimated economically-feasible energy in each continent. These data (Table 2.6) show that future hydropower development is unlikely to be constrained by the lack of suitable sites within the foreseeable future.
Table 2.6.
Comparison of actual and economical potential energy Average annual production: GWh/yr
Region
Percentage developed
Europe
525000
800000
66
North America
627000
1000000
63
South and Central America
491000
2325000
21
67 000
1000 000
7
750 000
3700000
20
2440000
8825 000
28
Africa Asia and Oceania Total
22
Economically feasible potential: GWh/yr
..
SEDIMENTATION IN RESERVOIRS
5 -- - - -_ •••• Water demand
4
- - - - -
Hydro demand
- ' - ' - Irrigation demand
--- ----- --- ----- ---- ..... ....
- - - Population -
--_._-- .... __ .. _---------._------_ .. . ...
1 1970
1975
1980
-
-
Storage
---
.--.---.... ..
1985
1990
1995
2000
2005
2010
2015
2020
Year
Figure 2.2.
Comparison of growth rates
2.4.4. Conclusion The rates of growth of population, global water demand, irrigation area and hydropower generation are compared in Figure 2.2. From the rates of growth for population, water consumption, irrigation area and hydropower, the following growt...1-t rates for demand of storage are postulated and are shown in Table 2.7.
2.5.
DISTRiBUTION OF DEMAND FOR MORE STORAGE
2.5.1. Europe Population The population of Europe (including Russia) is predicted to grow from 722 million in 1990 to a maximum of 730 million in the year 2000, and thereafter decline to 723 million in 2020 and to 678 million in 2050. The overall effects of Table 2.7. Period
Annual growth rates and irlcrease in storage Annual grovlth rate per annum: %
Increase in net storage: km3
2000-2010
1·56
998
2010-2020
1·39 .
1032
?Cnn_"il~il
23
EVACUATION OF SEDIMENTS
Table 2.B.
European growth in irrigation and hydropower
Year
Irrigation area: million ha
1975
27·3
3·3
1980
32·1
1985
Energy: GWh/yr
Annual growth in following 5 years: %
535 000
4·03
2·1
652 000
1·40
35·7
0·0
699 000
-0·32
1990
35·7
-6·8
688 000
1·39
1995
25·1
I
Annual growth in following 5 years: %
I
737 000
these population changes on water demand will be negligible, although changes in the regional distribution of population will result in regional shortages and surpluses which may generate demand for reservoir storage.
Irrigation and hydropower The historic growth in irrigation and hydropower in Europe is shown in Table 2.8. The negative irrigation growth since 1990 reflects the abandonment of uneconomic or marginal irrigation schemes in Russia since the break up of the USSR. Although the connection between irrigation area and storage is far from direct, these data suggest that there is no demand for new storage for irrigation. The low growth in energy generation in the 1980s probably indicates water shortages. Conclusion Although the demand for new storage is sensibly zero in much of Western Europe, it does appear that for the region as a whole there is a small demand, of the order of 1% per annum, for new storage for hydropower, and this is mainly concentrated in Eastern Europe. However, against this must be considered the relative shortage of suitable new sites and the strong opposition to new reservoirs from environmental groups. It is therefore concluded that the overall demand for new storage in Europe will grow at about 0·5% per annum.
2.5.2. North America Population The population of North America is predicted to grow from 278 million in 1990 to 306 million in the year 2000, rising to 358 million in 2020 and 389 million in 2050 with the growth rate declining from nearly 1% at the present time to 0-250/0 by 2050. As with Europe, the overall effects of these population changes on water demand will be negligible, although changes in the regional disttibution
24
~.
!
.~
·i ~<;
f
SEDIMENTATION IN RESERVOIRS
i
Table 2.9.
North American growth in irrigation and hydropower
Year
lITigation area: million ha
Annual growth in following 5 years:
18·0
2·8
520 000
1·0
20·7
-1·9
547 000
2·2
1985
18·8
0·2
611 000
0·0
1990
19·0
1·3
614000
2·4
1995
20·3
1975 1980
I
Energy: GWhlyr
%
Annual growth in following 5 years: %
I
693 000
of population will result In regional shortages and surpluses which may generate demand for reservoir storage.
Irrigation and hydropower The historic growth in irrigation and hydropower in North America is shown in Table 2.9. The overall growth rate of irrigation is 0·6% per annum. Although the conIlection between irrigation area and storage is far from direct, these data suggest that there is little demand for new storage for irrigation. Growth in energy generation has been stronger, at an average rate of 1·4%.
Conclusion , Although the data show that the energy generated by hydropower, as well as the area of land under irrigation, continued to grow through to the 1990s, the fact that no new storage was constructed in that period suggests that the data are influenced by operational factors. Therefore, although there remains a large undeveloped potential resource in Canada, environmental pressures will probably preclude any further development.
2.5.3. South and Central America Population The population of South and Central Amelica is predicted to grow from 440 million in 1990 to 523 million in the year 2000, rising to 676 million in 2020 and 839 million in 2050 - with the growth rate declining from 1·74% at the present time to 0·71 % by 2050. Because of the relatively low per capita consumption at the present time, water demand could increase at a rate well in excess of these rates.
Irrigation and hydropower The historic growth in irrigation and hydropower in South America is shown in Table 2.10.
25
EVACUATION OF SEDIMENTS
Table 2.10.
South American growth in irrigation and hydropower
Year
In'igation area: million ha
Annual growth in following 5 years: %
Energy: GWh/yr
1975
11·7
0·8
117 000
Annual growth in following 5 years: % 11·2 I
1980
12·2
1985
14.. 8
1990 1995
I
i
16·7 19·8
I
3·9
199 000
7·2
2·4
282 000
5·3
3·5
365 000
5·5
I
477 000 I
The overall growth rate of irrigation is 2·67%. Although the connection between irrigation area and storage is far from direct, these data suggest that there is a steady demand for new storage for irrigation. The overall growth rate of hydropower has been 7·2% Conclusion The data show that the energy generated by hydropower, as well as the area of land under irrigation, grew strongly through the 1980s and into the 1990s. Furthermore, it is estimated that Qnly 21 % of the economically-feasible hydropower potential has so far been developed, so that short to medium term growth is unlikely to be constrained by shortage of sites. From the rates of growth for popUlation, water consumption, irrigation area and hydropower, the growth rates for demand of storage shown in Table 2.11 are postulated.
2.5.4. Africa Population The population of Mrica is predicted to grow from 663 million in 1990 to 832 million in the year 2000, rising to 1348 million in 2020 and 2141 million in 2050 - with the growth rate declining from 2·77% at the present time to 1·53% by Table 2.11.
26
Annual growth rates, South and Central America
SEDIMENTATION IN RESERVOIRS
Table 2.12. Year
African growth in irrigation and hydropower In'igation area: million ha
Annual growth in foUowing 5 years:
1975
9·4
1·2
1980
10·0
1·4
1985
10·7
1·3
1990
11·4
1·5
1995
12·3
Energy: GWh/yr
Annual growth in following 5 years: %
37000
10-9
%
I
I
62 000
45000
I I
-0·9 5-g
43 000 57 000
-6·2
I
2050. Because of the relatively low per capita consumption at the present time, water demand could increase at a rate well in excess of these rates. Irrigation and hydropower The historic growth in irrigation and hydropower in Africa is shown in Table 2.12. The annual growth rate has been remarkably constant over this 20-year period at an average rate of 1·35 % per annum - about half the population growth rate.; The hydropower data reflect more upon the extremity of the 1980s drought than upon the growth of new hydropower capacity.
Conclusion The data show that the energy generated by hydropower, as well as the area of land under irrigation, grew weakly through the 1980s and into the 1990s at a rate well below the rate of growth of population. This trend is likely to continue, despite strong demand and great potential, so an annual growth rate of 2% is postulated.
2.5.5. Asia and Oceania Population The population of Asia together with Oceania is predicted to grow from 3213 million in 1990 to 3785 million in the year 2000, rising to 4784 million in 2020 and 5787 million in 2050 - with the growth rate declining from 1·65% at the present time to 0·6%- by 2050. Because of the relatively low per capita consumption at the present time, water demand could increase at a rate well in excess of these rates. Irrigation and hydiOpOWer The historic growth in irrigation and hydropower In Asia is shown In Table 2.13.
27
EVACUATION OF SEDIMENTS
Table 2.13. Year
Asian and Oceanian growth in irrigation and hydropower Inigation area: million ha
Annual growth in following 5 years:
1975
123·2
1·9
1980
135·0
1985
I I
%
I
Energy: GWh/yr
%
I I
Annual growth in following 5 years:
235000
4·9
1·2
291000
3·7
143·2
1·5
342000
5·2
1990
155·5
3·4
437000
3·2
1995
182·0
511 000
The annual growth rate has been reasonably constant over this 20 year period at an average rate of 1·97% per annum - well above the population growth rate. The average growth rate of hydropower over this 20-year pedod is 4·25%, about three times the population growth rate, reflecting the strong growth of the economies of many of the Asian countries in this period. It is unlikely that this differential is likely to persist in the future. Conclusion
From the· rates of growth for population, irrigation area and hydropower, the growth rates for demand of storage are postulated and are shown in Table 2.14.
2.5.6. Summary Based on the above, the forecast future demand for storage is shown in Table 2.15.
2.6.
RATE AND DISTRIBUTION OF CONSTRUCTION OF NEW RESERVOIRS
2.6.1. Worldwide The historic rate of constluction of storage worldwide is shown in Table 2.16 and Figure 2.3. The overall annual growth rate for the century as a whole has been 6·5%. It is interesting to note that neither the two World Wars nor the Depression have made any serious impact on the rate of growth: in this context the apparent fall in the rate of construction during the 1990s is dramatic and may in part be due to inadequate data.
28
SEDIMENTATION IN RESERVOIRS
Table 2.14.
Annual growth rates, Asia and Oceania
Peliod
Annual growth rate:
%
2000-2010
2·0
2010-2020
1·5
2020-2030
1·0
Table 2.15.
Regional de117.andfor new storage Demand for new storage (kIn3)
Reglon . 2000-2010
2010-2020
I
I I
2020-2030
I
Europe South and Central America
I
Africa Asia and Oceania
Year
1900
51
467
495
167
203
248
281
213
1032
939
315
I
I
998
Total
Table 2.16.
49
54
I
424
Rate of construction of new storage Cumulative storage: ]crn 3
I
I I
11
Annual rate of increase in following decade:
%
6·7
I
1910
21
1920
63
1930
121
1940
252
1950
I
11·6 6-7
I I
7·6 5·1
414
I I
I
11·2
1960
1196
.\
9·8
1970
3035
I
4·5
1980
4708
1990
5581
2000
5976
I
I I
I
1·7 0-7
I I
i
29
1
EVACUATION OF SEDIMENTS
6000
12
A I 5000 CO)
E .::
4000
I
E "0 CD Ol
I
I
\
/
I '"
\
\
\
\
\
\
I
::::l
>
I
I
I
............ , "'............"
3000
I ,
,
~
.e til
ta
;e
2000
I 'J
I
I
I
I
I
I
"
\
\
\
\
I
1
\ \ \
\ \
- - - World gross storage - - - - -
1.
10
Annual growth rate
1000
,, ,, ,, .........
.......... .....
2 ,
oL-~~~~==~~~~==~--~----~--~__-L__~O 1900
191.0
1920
1940
1930
1950
1960
1970
1980
1990
2000
Year
Figure 2.3.
1
Historic growth in reservoir storage
2.6.2. Distribution
1 1
of construction of storage
The distribution of the growth of new storage is set out in Table 2.17.
2.6.3. Comparison
of storage construction with demand
A comparison of the growth of storage constructed in the 1990s and planned from 2000 onwards shows that in South America and in Africa actual construction falls considerably short of demand. In the case of planned construction in the decade 2000-2010 this may be explained by inadequacies in the data.
1
I 1
Table 2.17.
Distribution of storage increase
,1
Annual growth rate in following decade: % ear
I
Europe
North America
South America
Africa
I
Asia
I
1950
0·05
9·39
6·17
5·40
10·84
1960
0·05
7·39
11,87
33·07
13·61
1970
0-05
3·69
8·24
5·85
1980
0-05
0·99
5·62
0·72
2·13
1990
0-05
0-03
1·59
0·49
1·16
2000
30
I
I
4·76
1 1
I
1 1
SEDIMENTATION IN RESERVOIRS
2.7.
RATE AND DISTRIBUTION OFLOSS OF STORAGE
2.7.1. Rate
of /OS5 of storage
The world estimate of land denudation is approximately 65 mm per 1000 years (Walling, 1984). Of this, approximately 90% is transported as suspended sediments and bed load, the remainder is in the form of dissolved matter. The rate of loss of storage due to siltation varies around the globe. Extreme examples include: • the complete loss of storage of the Wetzman reservoir on the River Gail in Austria withinone year of commissioning in 1883 (Cyberski, 1973) • the almost negligible loss of storage (0·1 % per annum) found from the suriey of 95 reservoirs in the UK (White et at., 1996). In order to assess the variation in the rate of loss storage around the world, data from approximately 2300 dams in 31 countries have been gathered and analysed. The summary of the results of the analysis is given in Table 2.18. The countries have been grouped by geographic regions, taking account of the global map of sediment yield (Walling, 1984). Table 2.18.
Regional sedimentation rates
R~gion
Estimated annual loss of storage due to sedimentation:
Estimated reservoir balf-life: yrs
%
North America
0·20
250
South America
0·10
500
Northern Europe
0·20
250
Southern Europe
0·17
294
Sub-Saharan Africa
0·23
217
Northern Africa
0·08
625
China
2·30
22
Southern Asia
0·52
96
Central Asia
1·00
50
South-East Asia
0·30
167
Pacific rim
0·27
185
Middle East
1·50
~"I
~J
31
EVACUATION OF SEDIMENTS
J The estimates are based on varying quantities of data gathered, as shown in Table 2.19. The estimates of annual loss of storage due to sedimentation have been used in conjunction with the gross storage volume data available in the ICOLD World register of dams (1998) to estimate the magnitude of the sedimentation problem. The results of the analysis are displayed in Figure 2.4. In summary, the analysis shows that by the year 2000 approximately 567 km3 (10% of the current gross available storage in the world) has been lost to sedimentation. From the data available from the 1325 registered dams under construction, it can be seen that the average gross storage volume of new reservoirs is approximately 370 M.m3 • Therefore, in order to replace the volume lost to sedimentation, over 2000 average-sized dams would have to be constructed around the world.
1
2.7.'2. Distribution of loss of storage The distribution of the annual rate of loss of storage and the total volume lost to sedimentation are shown in Table 2.20. Table 2.19.. Extent of sediment data Region
Total storage used in estimating regional loss: M.m3
Gross storage used in esth'llating regionalloss: %
109980
6·0
1038913
3832
0·4
Northern Europe
938 168
3067
0·2
Southern Europe
145 162
24030
16·5
Sub-Saharan Africa
575352
252 168
43·8
Northern Africa
188473
181 760
96·4
China
649322
42804
6·6
Southern Asia
318602
92712
29·1
Central Asia
148032
Nil
N/A
South-East Asia
117371
Nil
N/A
Pacific rim
277 124
20192
7·3
Middle East
223683
9006
4·0
Global total
6464730
738552
11·4
-.
Current gross storage in region: M.m3
1
I
North America
1 844 530
, i
South America
32
1
SEDIMENTATION IN RESERVOIRS
7000
Projection assuming no new dams
6000 5000 C'l
E
.::£
Q.j
01
4000
~
~ m
3000
;0
~ ::;:f'-_- -'-,~.= =~ ; .; ; J.'- ':-" "L. .-: " -;'. .l:. . -_-.l-: ~ _- L_- -JL. -. _J-. _. . l. -_- -L. ._- L_.- -J Sedimentation
.-.=..-= ..- ""_1.
1900
1970 1980 1990 2000 20iO 2020 2030 2040 2050
1910 1920 1930 1940 1950 1960
Decade
Figure 2.4.
Storage lost to sedimentation
Table 2.20.
Distribution of sediment rate~ and storage loss
Region
North America South America
I )
Gross yolunie in 2000:
Annual sedimentation:
km3
1845 973
I I
Ian3
Lost to sediments: %
Total storage loss: km3
3·69
7·9
112
1·04
2·5
I
17
Northern Europe
822
1·88
6·8
Southern Europe
135
0·25
5·6
6
574
1·32
7·8
32
Northern Aflica
188
0·15
2-4
3
China
526
14·93
45·8
Southern Asia
233
1·66
13·1
Central Asia
132
1-48
26·9
0·35
8·0
Sub-Saharan Africa
South-East Asia
I
I
117
I
I
230 31
I
29
6
I
I
Pacific rim
232
0 ·75
7·6
IS
Middle East
199
3 ·36
27·7
38
Giobal total
5976
30·85
11 ·8
567
33
EVACUATION OF SEDIMENTS
It is interesting to note that the annual loss of world storage estimated by this study, 0-48% per annum, is almost half that estimated by Mahmood (1987), who quotes a figure of 1%. Although it is not clear how Mahmood estimated his figure, it can be argued that the figure determined by this study may underestimate the problem. This is further supported by estimates made by other authors, such as Goldsmith and Rildyard (1984), who estimate the annual loss of storage in central Europe to be 0·5%, compared with the estimates for Europe by this study of 0·2% and 0·17% per annum.
2.8.
TRENDS IN THE RATE OF LOSS OF STORAGE The rate of loss of storage for.a given reservoir is dependent on the rate of erosion of the catchment. In regions where the catchments have remained stable, e.g. Northern Europe and North America, the rate of loss of storage is constant. In regions where deforestation has occurred, the rate of catchment erosion and consequently the rate of loss of storage increases. This phenomenon is clearly visible in a number of locations, such as the Aberdare Forest in Kenya where the natural vegetation is gradually being lost and replaced by small holding subsistence farms. This has resulted in an increase in the rates of erosion. As part of this study, data from 16 reservoirs in the US have been analysed to detect any evidence of change of loss of storage over the life of the reservoirs. The reservoirs ranged in size from 175 000 m3 to 23 km3 , with catchments ranging from 9 km2 to over 100 000 km2 • With the exception of Fort Peck (23 000 M.m3) and Great Falls (68 M.m3), reservoirs that showed a definite trend of reducing sediment accumulation rates with time, no discernible trend could be established. The data from Ringlet reservoir in Malaysia (ICOLD, 1997) shows clearly the dramatic effects of deforestation. The 183 km2 catchment has been gradually changed from forests to plantations and holiday facilities, which has resulted in the specific sedimentation increasing by an order of magnitude from the midsixties to the present day (see Table 2.21).
Table 2.21. Year
34
Ringlet reservoir, sedimentation Specific sedimentation: m 3/km2/yr
Annual sedimentation: . M.m3/yr
1965
165
0·03
1980
275
0-05
j
SEDIMENTATION iN RESERVOIRS
4·0 3·5
E
::::l
c: c:
...
CiS
3·0
Based on data from 1105 US reservoirs
c.. ~
2·5
1ri
CI
e:! 0
t;
2·0
'0 Co" tfl
10S
.Q (l)
CI
e ID
1·0
~
0·5
t
"i"":
0.0 1 0·001
••
ttlt'
0-01
.
,
, "
It I
0·1
10
100
1000
10000
Aveiage size of reservoir: M.m3
Figure 2.5.
2.9.
Reservoir size and rate of loss of storage
RESERVOIR SIZEAND RATE OF LOSS OF STORAGE In the USA it is estimaJed that 1235 M.. m 3 of sediments are deposited in the resen-oirs annually (Glymph, 1973). The total storage of reservoirs in the USA is estimated at 627300 M.m 3 (Morris and Fan, 1997). This represents a loss of O· 20% of the storage volume annually to sediments. However, the rate of loss of storage varies substantially and shows a striking inverse relationship between the rate of loss of storage and reservoir capacity, as shown in Figure 2.5. The highest rates of loss of storage are found in the smallest reservoirs and the lowest rates in the largest. Of the 1105 reservoirs studied, 730 have a storage volume of less than 1233 m 3 and an average rate of loss of storage in excess of Table 2.22.
Gross requirement for ne"Y\/ storage
Continent
demand
2000-2010 Europe
49
South America
467
Africa
167
35
202
Asia and Oceania
315
349
664
Total
998
567
1565
35
EVACUATION OF SEDIMENTS
1% per annum. At the other extreme, 23 of the reservoirs studied had a storage volume in excess of 1233 M.m3 and an average rate of loss of storage of 0.16% per annum.
2.10.
REQUIREMENTS FOR NEW STORAGE
New storage will be required in the future both to satisfy increasing demand generated by the growing world population and to replace the storage lost due to sedimentation. The estimate of the gross storage required in the next decade is given in Table 2.22.
36
earc •
In
ue
fa tor u Ing
30 Research into factors which influence flushing 3.1.
INTRODUCTION This review of the factors which contribute to the efficiency of sediment flushing operations' for reservoirs is based both on field experience, which has been built up over the past 30 years or so at dams that have fluspjng systems LTl regular operation, and on research findings using simulation models, either numerical or physical, which in tllln rely on fundamental experimental data concelning the detailed physics of the movement of sediments in water. Field installations provide information which includes: • the hydrological conditions at the site " details of the flushing system 0._ the way in which the flushing system is operated the development and re-erosion of the sediment delta • details of the development of the incised channel during flushing • sediment inputs, throughputs and outputs. (9
The data are very valuable in looking at the efficiency of flushing operations. They represent 'real' situations and there are no scale effects or other simulation deficiencies to mask the findings. However, tt:lere are shortcomings: • there are only a few reservoirs being flushed at present and these do nbfcover the full range of conditions where flushing might be considered: i.e. the data forms a sparse matrix • the data are rarely comprehensive enough to cover the complex situation found in the field • the records are short in terms of sediment deposition and they do not, therefore, necessarily represent the hue long-term situation o the historic development of flushing systems started, for economic reasons, with very modest installations of low capacity - which turned out to be relatively inefficient. There is thus a dearth of information for installations with high capacity, more efficient, flushing systems: Fortunately, there have been major advances in our understanding of sediment transport by water and this has facilitated the development of numerical simulation models. It has also facilitated a better understanding of the strong and weak points in physical model simulation techniques. These models have also
39
EVACUATION OF SEDIMENTS
had the benefit of the field data in order to check their validity. Simulation techniques can thus be used to: • extend the range of information beyond that covered by field data • systematically look at the importance or sensitivity of individual variables. This holistic approach, using field data, models and fundamental knowledge of the sediment transport process, has enabled a clearer understanding of the requirements for efficient flushing systems to be developed.
3.2.
. THE MECHANISM OF FLUSHIN.G Shen (1999) quot~s numerous researchers, Lai and Shen (1996), Albertson et al. (1996), Shen· and Lai (1996) and Morris and Fan (1997), who have studied sediment flushing. Based on this work, Shen quotes three stages of flushing. (a) When the water level in the reservoir is high, the water velocity in the
reservoir is too low to move much sediment. Only close to the flushing outlet are the flow velocities high enough to erode sediment, and a flushing cone is formed close to the flushing outlet. (b) At intermediate water levels, water velocities at the upper end of the reservoir increase and sediment is transported towards the flushing outlet. There remains a flushing cone close to the outlet. (c) When the water level falls to the top of the flushing outlet, scouring velocities can be generated throughout the length of the reservoir. Retrogressive erosion of previously deposited sediment occurs. Shen (1999) concludes that stage three removes far more of the deposited sediments, thereby regaining storage capacity. He also comments that stage three 'uses more water'. Indeed, stage three is the only realistic scenario for removing significant quantities of previously deposited sediments from reservoirs. When flushing is attempted without drawing down water levels, the high flow velocities at the outlets are very localised and the impact is insignificant. The water level in a reservoir must be drawn down close to the bed elevation at the dam before flushing can be effective (Figure 3.la). Many authors have confirmed this with observation, theory or modelling, including Mahmood (1987), White and Bettess (1984) and Atkinson (1996), However, moderate lowering of water levels during flushing will still increase flow velocities significantly at the upstream end of the reservoir, where bed levels will be above the water level at the dam (Figure 3.lb). Large sediment volumes will be scoured from these upstream reaches and will re-deposit nearer the dam. Eventually, bed levels upstream from the dam will rise to the water level during flushing and then significant sediment quantities will be transported through the low level outlets (Figure 3.1c). Thus, flushing represents an extreme change in reservoir operation. It requires draw down of the reservoir so that the velocity and volume of the flow are
40
FACTORSWHICH !NFLUENCE FLUSHING
Maximum water level
----------
Original bed level
'.
(a)
Maximum water level
(c)
Figure 3.1. Longitudinal profiles during flushing: (a) flushing with full drawdol,w1.; (b) flushing yvith insu:fficient drm·vdown; (c) final conditions after a long period of flushing with insufficient drawdown
41
EVACUATION OF SEDIMENTS
sufficient to scour and remove sediment. This raises many technical, economic and environmental issues. • The shape of the reservoir may preclude the formation of a scouring channel that is capable of removing significant quantities of sediment. • River discharges may be insufficient to transport large sediment loads through the reservoir. • The drawdown of reservoir level reduces the capacity to generate power and the release of high volumes of water for flushing may also reduce the annual water yield from the reservoir. • . There is a need to be able to predict ahead so that the flushing operation is not undertaken if it may jeopardise future power or irrigation supplies. ..... The environmental consequences of passing sediments, that may have been in , the .reservoir for .s.ome considerable time, to the downstream reach.
-.,
- \
-Ij
I
~
!
-I !
3.3.
THE DEVELOPMENT OF CRITERIA FOR SUCCESSFUL FLUSHING
Sediment flushing is not universally applicable. If the conditions are right, flushing represents an efficient and economical way of preserving reservoir storage. If the conditions are not right, attempts to flush sediments from the reservoir will be disappointing. This section considers those factors which determine whether flushing is likely to succeed.
J J
3.3.1. Sediment balance The sediment balance considers the quantities of sediment entering the reservoir compared with the quantities which might be flushed through the dam. The analysis makes some simplifying assumptions but is useful in deriving one of the indicators which determines whether flushing operations are, or will be, successfuL This indicator is the sediment balance ratio, SBR.
Long-term equilibrium conditions If flushing water levels are close to bed elevations at the dam (either as in Figure 3.la or as in Figure 3.1c) and long-term equilibrium conditions are to be reached, then the sediment mass flushed must, in the long-term, balance the sediment mass depositing between flushing operations. This balance can be expressed as: (1)
where Qs is the sediment transporting capacity (tis) of the flow in the incised flushing channel, n is 86400 (seconds per day), Tf is the duration of flushing (days), N is the interval between flushing operations (yrs), Min is the sediment inflow rate (tiyr) and TE is the trapping efficiency of the reservoir.
42
J.
FACTORSWHICH INFLUENCE F.LUSHING
Efficiency of flushing More generally, the non-equilibrium state of sedimentation in a reservoir can be expressed as a sediment balance ratio SBR, the ratio of sediment flushed to sediment depositing, which can be defined as:
(2) The transporting capacity Qs' (tis) will be a function of discharge, channel roughness, width and slope, and the properties of the deposited material. A low sediment balance ratio indicates low flushing efficiency and the continued build up of deposited s~~iments in the reservoir. A value of SBR> 1-0 indicates high efficiency and the lpng -term stability of sediments within the reservoir.
The relevant parameters 1. The rate at which sediment is flushed, Qs' The only method for predicting Qs during reservoir flushing which has been widely tested, is an empirical equation derived by Tsinghua University and reported in IRTCES (1985) and 110rris and Fan (1997):
(3)
where Qf is flushing discharge (m3/s), S is the longitudinal energy gradient through the reservoir, W is channel vvidth (m) and "¥ is a constant related to the sediment type:
1600 (530, Atkinson (1996)) for loess sediments 650 (225, Atkinson (1996)) for other sediments 'with median size finer than O-lmm • 300 (lao, Atkinson (1996)) for sediments with median size larger than 0-1 mm,and III 180 (60, Atkinson (1996» for conditions of flushing with a low discharge. IJ
o
Discrepancies between predictions of sediment load derived from Equation (3) and observations were relatively small for the flushing data from China, on which the method is based_ Discrepancies \vere within a range of half to twice in 87% of cases, which is very good for sediment transport predictions. Atkinson (1996) compared the method with observations from reservoirs in India, USA and the former USSR. Equation (3) was found to overestimate sediment loads by a factor of t.'ree and even more \vhere flushing was not performed annually (as is common practice in China) but leSS frequently where sediments were much greater than the 0·1 mrn threshold. The threefold correction suggested by Atkinson is recoITunended when conditions differ from those typical in China (see above).
43
EVACUATION OF SEDIMENTS
The coefficient, '1', varies with particle size, the larger the sediment size the smaller the value of W. It follows, therefore, that fine sediments are more easily flushed from reservoirs than coarser sediments. 2. The longitudinal energy gradient, S. This is a parameter which depends on the degree of draw down at the dam during flushing. The maximum energy gradient is obtained when the drawdown is maximum and under these conditions it approximates to the slope of the original river bed prior to impoundment. It is a parameter which can be controlled by the choice of the amount of drawdown and which influences the efficiency of flushing operations because of its effect on sediment transport rates, see Equation (3)~ 3. The bed width of the incised channel, W. Equation (3) requires the bed width of the incised channel, W, to be input. Channels formed by flushing in reservoir sediment deposits correlate well with flushing discharge. Figure 3.2 shows the relationship and the data from which it was derived. The fitted line is described by the equation (in SI units): _.0
(4) In some cases, channel bed widths may be constrained by the reservoir width: In general, though, the width of the incised channel is determined by the flow and is independent of sediment size. Vl't = 12·8 0.0.5
1000
•
+
Sanmenxia Guanting
E
@ Guernsey
~
Baira
EB
• +
.s=
:0 .~
1i5 c: c: ctS
..c: ()
•
100
"'0
~
(/)
.0
0
10~--~~~~~~--~--~~~~~--~~~~~~----------
1
10
Flushing discharge,
Figure 3.2.
44
1000
100
0.: m3/s
Channel widths formed in reservoir deposits during flushing
J
FACTORSWHICH INFLUENCE FLUSHING
4. The discharge used for fIushing~ Qf. The discharge used for flushing reservoir sediments is, in the case of maximum drawdown, the incoming river flow because there is little storage in the .reservoir to cause attenuation. Any flushing system must be capable of passing these river flows while maintaining low water levels at the dam. 5. The duration of the flushing operation, Tf •
This is under the control of the reservoir operating authority. Factors to be considered include: • the current availability of water • the need to safeguard future supplies of water • the need to miniwise the loss of present ~'1d future po\ver supplies.
6. The number .of years between flushing operations, N. Again, this is under the control of the reservoir operating authority. Factors to be considered include: /II
Ii
e
the need for the operation in the ~ght of the success of previous years' operations hydrological forecasts for the follovving months likely demands for power and irrigation water supplies for the following months, taking into account alternative sources of supply.
7. The trapping efficiency of the reservoir, TE. The trapping efficiency of a reservoir depends on many factors. Brune's (1953) curves give good guidelines in general terms. However trapping efficiency depends on the volume of water in the reservoir which in tum depends upon reservoir water level. 'Vater levels in most reservoirs, particularly those used f6r annual water storage and supply, vary throughout the year. The flushing of sediment demands large fluctuations in reservoir water level and hence the application of Brune's (1953) curves does present some problems. 1
Evaluation of the sediment balance ratio, SBR Equations (2) and (3), together with an estimate of trapping efficiency taken from Brune's (1953) curves, can be used to derive the sediment balance ratio, SBR. For long-term equilibrium to be obtained, values of SBR comfortably in excess of unity are required owing to uncertainty in the prediction methods and input parameters. Discharge flushing, Qf' and the duration flushing, ·will depend on the reservoir operation chosen, and the can initially be set at the river slope before impoundment. Predicted values for SBR can be used to guide the choice
45
EVACUATION OF SEDIMENTS
of the inputs to repeat predictions. For example, a low SBR may imply flushing should be performed at a time of higher discharge and a high SBR may imply that flatter slopes can be expected upstream from the dam (as shown on Figure 3.lc). If a value of SBR well above 1 cannot be achieved, then flushing is not feasible. In the comparison with data presented in Table 3.1, SBR was computed for a slope defined as the drawdown water surface elevation below maximum water surface elevation divided by reservoir length. Table 3.1. Application of sediment balance and long-term capacity ratios to existing reservoirs Reservoir
. Initial . capacity: M_m3
Country
Sediment balance ratio
Long-term capacity ratio (LTCR)
(SBR)
Estimated from reservoir surveys
Calculated
Calculated
Reservoirs flushed successfully 1
Baira
India
9-6
0·85
0-85
Gebidem
Switzerland
9-0
Approx_ 1-00
0-99
7
Gmund
Austria
0-93
0-86
0·98
21
Hengshan
China
13-3
0·75
0-77
3
Palagnedra
Switzerland
5-5
1·00
1-00
33
Santo Domingo
Venezuela
3·0
0-97
1-00
11
.
7
1
I i
-
I
Reservoirs flushed unsuccessfully Guanting
China
2270
Low
0-20
0-2
Guernsey
USA
91
Low
0-26
1·0
Heisonglin
China
8·6
0·23-0-35
0·30
Approx_ 0-70
Ichari
India
11·6
Approx. 0·35
0-36
7
Ouchi-Kurgan
Former USSR
56
Low
Approx.
7
China
9640
Sanmexia
0·10 0·31
0-39
3-4
!
46
Sefid-Rud
Iran
Shu1caozi
China
1760
<0·26
0·13
4
9·6
Low
0·39
4·6
FACTORSVVHICH INFLUENCE FLUSHING
3.3.2. Sustainable reservoir capacity The flushing of sediments depends on many factors. Prior to the construction of the reservoir, the entire annual flows were used to transport sediment through the reach and water depths were the natural river depths throughout the period. Once the reservoir is impounded, depositional areas are created and only a proportion of the annual flow is available for flushing sediments. Thus, some pennanent long-term deposition is inevitable in reservoirs. The question to be asked is: what proportion of the original storage volume can be retained by flushing?
Long-term capacity ratio~ LTCR The long-term capacity ratio, LTCR, is defined as the sustainable storage capacity divided by the initial storage capacity of the reservoir. Flushing will cause a channel to be scoured into the reservoir deposits. In most cases~ this channel will be narrower than the reservoir and so substantial deposits will remain in the reservoir. In the long term these deposits will rise to an elevation close to the maximum water level, leaving the volume created by the incised flushing channel as the only storage volume remaining in the reservoir. This storage volume is defined as the sustainable reservoir capacity. Figure 3.3 illustrates the process, it shows cross sections at two reservoirs ~vhere flushing has maintained a relatively small sustainable reservoir capacity. If a trapezoidal cross-sectional shape is assumed for the incised channel, the sustainable reservoir capacity volume is't.lJ.en determined from: minimum bed elevations at each point in the longitudinal profile, as shown in Figure 3.1 - these can be determined from the water level during flushing and the sediment balance calculations described above e the ma."{imum water level, .. the bed width of the incised channel, which can be calculated using Equation (3) and t..he flushing discharge (bed width Inay be constrained by reservoir width) €I the side slope steepness of the incised channel (if the side slope is shallower L~an the reservoir side slope, then the reservoir widths may constrain the width of the sustainable section). e
In the list above only the side slope steepness is not known. In wellconsolidated sediments, near vertical channel sides can occur, while slopes as low as 2·5% have been observed for poorly consolidated material. Therefore, a technique to predict this slope is vital to a reliable prediction of sustainable capacity. Atkinson (1998) recommends, with some reservations, the use of the following expression:
,..,.,Ian a. = 06'"1 . .Jp4,7
(5)
'where CL is the angle of the side slope (zero is horizontal) and p is the density of 3 the deposits expressed as weight of dry material per unit volume (tlm ). p ca.TJ. be
47
EVACUATION OF SEDIMENTS
. ---I
-I I 1
-II
i
i
1750
I
E C
~I
0
~
> Q) Q5
~l
"C Q)
a:l
1745
I 1
-I
I
1740~----------------~-------------------L--------------__~
o
100
200
300
(a)
!
-\
I
(Before flushing)
0
Q5 Q)
I
-i
Sept. 1973 - (after flushing)
320
~ ~
"C
J
I
325
E C
I
I
i
I
-I
-I
315
m
~
310
!
r)
~I
Aug. 1960 (at end of construction period)
305 0
1000
2000
3000
4000
-
)
Lateral distance: m (b)
i 1
~.
Figure 3.3. Cross-sections of flushing channels: (a) Heisonglin reservoir, China; (b) Sanmenxia reservoir, China
predicted from the composition and age of the deposits using Lane and Koelzer's (1953) method. Atkinson (1998) found that slopes computed using Equation (5) could be in enor by as much as a factor of ten and clearly this method needs to be treated with caution. A simple criterion for assessing sustainable reservoir capacity can be developed by fitting a simplified reservoir shape as shown in Figure 3.4. A cross section just upstream from the dam can be taken as representative of the entire reservoir, and then the area of the trapezoidal flushed section can be compared to the original cross-section area. The ratio of these areas then gives a long-term capacity ratio (LTCR), which is an estimate of the reservoir capacity that can be sustained in the long term by flushing.
J
J
----~. : ----------------------------------~----------------------------
48
J
FACTORSWHICH INFLUENCE FLUSHING
(a)
(b)
-
~ River channel
.-/-
Dam
------
Section Section
"'C7
""=7'
J
(e)
Full supply level Original river-bed levels (d)
""':;:-------or---------,-----""""7'
~\ B_~ __
I I
lA' ),t
Full supply lever
_ _ _ Water level during flushing
,-- - - - - - - - -
--It- . "I"
Bed width,
Reservoir bed elevation at dam
WbO!
~ Flushing channel width, Lt'~ (e) Long-term capacity ratio, LTCR, is approximated to:
Area B Area A plus Area 8
Figure 3.4. Simplified reservoir geOl7'letry for application of capacity criterion: (a) actual reservoir plan; (b) fitted reservoir plan; (c) simplified reservoir plan and sections; (d) simplified reservoir elevation; (e) enlarged section immediately upstream, of dam
49
EVACUATION OF SEDIMENTS
If the long-term capacity ratio, LTCR, is greater than 0·5 then flushing is likely to be successful in terms of maintaining live storage in the reservoir and is likely to be economic if the shortfall in generating capacity during the flushing period is not too severe.
3.3.3. Evaluation
of flushing criteria
The sediment balance ratio, SBR, is a measure of the propol1ion of the incoming sediments which may be flushed from the reservoir. The long-term capacity ratio, LTCR, is a measure of the proportion of the initial storage capacity which may be retained by flushing from the reservoir. Some of the factors which determine the values of the sediment balance ratio, SBR, and the long-term capacity ratio, LTCR, are inherent characteristics of the site. These include: • the shape and size of the reservoir • the imposed hydrological conditions • the imposed sediment inputs. Some of the factors are controllable. These include: • the operation of the reservoir between flushing operations • the design of the flushing system, including elevation and capacity /I the operation of the flushing system, including discharge and duration. A review of the literature on reservoir flushing, produced information from 14 reservoirs where flushing had been attempted and where sufficient data were available to test the criteria (Atkinson, 1996). The 14 reservoirs can be divided into two categories: six where observations indicated that flushing would sustain a long-term capacity in excess of half the original capacity, and eight where it would fail to do so. Table 3.1 presents the results of the application of the two assessment criteria. The criteria performed very well in distinguishing between the six reservoirs where flushing was successful and the eight where it was not. The predicted LTCR also proved to be a good indicator of the long-term capacities that were estimated from the observations. The sediment balance ratio, SBR, was, in most cases, not a constraint to successful flushing.
3.3.4. Practical criteria for successful flushing There are several factors which impose constraints on sediment flushing. These include: • the operation of the reservoir between flushing operations as dictated by its usage • the design of the flushing system, including elevation and capacity, as dictated by economics and hydrological factors
50
J
r'"
FACTORSWH1CH INFLUENCE FLUSHING
It
the operation of the flushing system, including discharge and duration, as dictated by external demands for energy and water during the flushing period.
It is useful to review the factors which constrain successful flushing at a reservoir, and so assess whether they can be overcome, for example by enlarging outlets in the dam. Four main constraints are identified and these have been assessed against field data for reservoirs that have been successfully and unsuccessfully flushed. These constraints are presented below.
Incomplete drawdown of water levels during flushing It may not be possible to flush sediments from outlets close to the initial bedlevel upstream of the dam. Existing installations may have been built at a higher level. There may be serious engineering/economic problems in building such outlets at projected new installat~ons. Under these circumstances, the capability of removing sediment from the reservoir by flushing is reduced. By taking water height as elevation above the base of the dam, a drawdown ratio is expressed as: (6)
\Aihere Hfjush is the height of water at the dam during flushing and Hma,'( is the maximum height of water at the dam (to reservoir retention level), both measured above original bed-level. DDR less than about 0·7 indicates some degree of constraint owing to insufficient drawdown.
Insufficient flushing flows to develop a long-term sediment balance Flushing flows can be constrained for many reasons: ., the flushing outlets may be too small and may restrict discharge ., the river discharge may not be adequate at the appropriate time for flushlIJ.g e the water may have to be retained in the reservoir for needs in the immediate future, thereby restricting the duration of the flushing operation. Under these circumstances the amount of sediment enteIing the reservoir will exceed that which is removed by flushing until the storage is reduced to the point where a new, but unsatisfactory, sediment balance is achieved. The retained reservoir storage will be low compared with its original value and it will have a much lower natural trapping efficiency to match the constrained flushing flows, see Equation (1). . For a given flushing discharge, Qf' the maximum rate at which sediment can be flushed from the reservoir occurs under the conditions of maximum drawdown. The sediment balance ratio llsed to assess the adequacy of flushing flows should t.l}erefore be based on this maximum drawdown. This specific value
51
EVACUATION OF SEDIMENTS
of the sediment balance ratio is designated SBR d and is calculated using the original river slope, that is for conditions of full drawdown. SBR d < 1·0 indicates a constraint due to the inadequate capacity to flush sediment. The amount of sediment removed during the flushing period depends, for a particular reservoir, on the flushing flow and the duration of flushing. The relative importance of discharge and duration are illustrated in the following example. By using Equation (4) to determine flushing channel widths and then substituting in Equation (3), this yields values of sediment discharge rates, Qs' The total quantities of sediment removed during the flushing period are then obtained by considering the duration of flushing. Table 3.2 shows an example of the results from this type of analysis. It is based on a notional reservoir with a bed slope during flushing of 0·0006 and a flushing water volume of 8·64 x 109 m3• It is assumed that the reservoir contains 0·1 mm sand and hence the coefficient in Equation (3) is 100 (Atkinson, 1998). If there is a restriction on the quantity of water available for flushing, it is clearly better to use a high discharge for a short period than a low discharge for a long period. This increases the amount of sediment that will be removed. The penalties of extended flushing periods are considerable, quite apart from the Table 3.2. The relative importance of the discharge used for flushing and the duration of flushing for a particular volume offlushing water Flushing discharge: m 3/s
Flushing duration: days
I
Mean sediment concentration: ppm
Sediment removed per day: Mt
Sediment removed during flushing period: Mt
500
200
9600/5760*
0·4110·25* .
82/50*
1000
100
14550
1·26
126
1500
67
18560
2·41
160
2000
50
22060
3·81
191
2500
40
25220
5·45
218
3000
33
28130
7·29
243
3500
28·5
30860
9·33
267
4000
25
33430
11·55
289
4500
22·2
35880
13·95
5000
20
38220
16·51
330
5500
18·2
40470
19·23
350
6000
16·7
42640
22·10
369
I I I
* Using Atkinson (1996), see Section 3.3.1(1).
52
I
I
310
FACTORSWHICH INFLUENCE FLUSHING
hydraulic efficiency of removing sediments from the reservoir, in terms of loss of potential generating capacity and of volumes of water for irrigation. Experience suggests that a flushing discharge of twice the mean annual flow is desirable and that the volume of water used for flushing should be not less than 10% of the mean annual run-off.
Reservoirs that are too narrow to develop an efficient flushing channel The reservoir could be too narrow for the natural width of the flushing channel to develop during flushing. This is particularly the case where flushing discharges are high. To check this, the bottom width of the flushing channel should be compared with the bottom width of the reservoir. A flushing width ratio, FWR, is defined as:
FvVR
=VV/W
(7)
bed
where W is computed from Equation (3) and V/bed is a representative bottom width in the reservoir. If the reservoir can be approximated to the shape shown in Figure 3.4~ the representative bottom width should be taken as that which occurs just upstream of the dam. FWR> 1·0 is required unless the side slopes are shallow (see below - TVVR greater than about 2 would indicate that F'YVR is not a constraint).
Reservoirs that. are too wide for the flushing channel to reach tie perimeter The natural top width of the flushing channel may be less tha..TJ. the representative top width of the reservoir and under these circumstances accumulated sediments vviU remain along the perimeter of the incised channel forming a high level terrace. If the top width of the section scoured by L.~e flushing channel is not restricted by the reservoir sides, then the top width of the flushing channel is a constraint. top width ratio is defined as: TVVR
=[Wmin + (2 X Hrnax X tan a)J/~Vtop
..
(8)
where vVmin is the lesser of Wand }Vbed and V/top is a representative top width in' the reservoir. If the reservoir can be approximated to the shape shown in Figure 3.4, the representative top width should be taken as that which occurs just upstream of the dam. TVVR> 1·0 is required.
The effect of the size
of sediment in the reservoir deposits
The nature of the sediments entering and depositing in reservoirs influences whether flushing is practicable. The width of the incised channel formed during flushing is determined by the flushing discharge and it is independent of sediment size for the raD.ge of sediment sizes generally found in reservoirs, see Equation (4). However, the sediment tral1.Sport rates are dependent upon sediment size as indicated by the constant, 'Ii, in Equation (3). Coarser materials are more difficult to remove from reservoirs than finer materials.
53
EVACUATION OF SEDIMENTS
Table 3.3. The influence of sediment size on the amount of sediment removed from, reservoirs Sediment removed during the flushing period: Mt
Flushing conditions
Sediments with a mean size less than 0·1 mm
Sediments with a mean size greater than 0-1 mm
Conditions with low flushing flow
N/A
N/A
N/A
50
1750
740
330
N/A
m3/s
Duration: days
Loess sediments
500
200 20
Flow:
I
5000
I
I
By taking the example given in Table 3.2 and by considering the specific combinations of flushing discharges of 500 m3/s and 5000 m3/s for 200 days and 20 days respectively, the effects of sediment size are as given in Table 3.3. The sizes of the sediments deposited in reservoirs are an important factor in deciding whether flushing will be effective. The effect of widely graded sediments In many rivers there is a mixture of sand and gravel in the bed material. In these circumstances, there may be a fairly small proportion of the bed material consisting of fine sand, but owing to its high transportability, a large proportion of the material being transported by the river is fine sand. When an impoundment is introduced to such a river, the material depositing is dominated by fine sand together with finer cohesive material that is transported in the river as wash load. These deposits are relatively mobile compared with the general river-bed sediments and hence are amenable to flushing. A parameter that can identify these favourable circumstances for achieving a sediment balance by flushing is the sediment size ratio, SSR: (9) Table 3.4 presents the relationship between this ratio and the number of days of flushing required annUally. It is assumed that flushing is performed at a discharge of twice the mean annual flow, as recommended above. Full drawdown of water levels is also assumed and the calculations were performed for a series of assumed sediment size ratios, SSR. In preparing the table, the updated Ackers and White sediment transport predictor, Ackers (1993), was used and the riverbed material was divided into ten fractions. The equation derived by Tsinghua University and reported by IRTCES (1985) could not be used as it cannot be applied to sediment transport rates in rivers. There is no universal relationship between the sediment size ratio, SSR, and the number of days of flushing required annually because the reservoirs compared are of different sizes. If a general rule is to be made, then it may be suggested that rivers where the SSR is less than about 0-03 are generally suited for flushing.
54
FACTORSWHICH INFLUENCE FLUSHING
It should be noted that flushing has been proposed at the Tarbela reservoir, but not, to our knowledge, at the other two reservoirs. The Tarbela and Tungabhadra reservoirs are large (>3000 M.m3), while the planned Rooiport reservoir is about 800 M.m3 • The parameters required to determine the SSR can be derived as follows. • DSOT: the 50 percentile. size of the sediment in transport in the river can be obtained, if possible, from sediment sampling during periods of high river discharge or (if the reservoir has been constructed) from samples taken from deposits. Otherwise values can be obtained either from estimates derived from ·other rivers in the region or by prediction using the bed material grading . • Dsos: the 50 percentile size of the river-bed material can be obtained from
representative bed material samples. In each case, silt and finer material can be excluded as it is usually not a constraint to a sediment balance. Table 3.4.
Effect of the sediment size ratio on sediment balance Flushing period required: days
c. -ment Size ratIO Sill (SSR)
Tarbela, Pakistan
1
Tungabhadra, India
Rooiport. South Africa
(> 100)
(> 100)
(> 100)
0·4
62
94
(> 100)
0-2
38
I
58
84
0-1
23
I
30
58
0-06
15
Ii
21
46
0-05
13
19
43
0-04
11
17
40
15
36
0-03
9 .
I
I
55
EVACUATION OF SEDIMENTS
The sediment sizes will also affect where in the reservoir the· material will settle. Some of the silt, the sand and the coarser material tend to deposit in a delta at the upstream end of a reservoir, while the finer silts and the clay can deposit throughout a reservoir. After a period of flushing an incised channel will be . formed in the deposits in the delta, which will quickly refill with incoming sand and coarser sediments. Thus, most of this coarser material will be flushed from the reservoir when the incised channel reforn1s during the subsequent flushing operation. In contrast, the finer deposits formed nearer the dam will be distributed across the reservoir, and so will be only partially removed by the flushing of an incised channel. The impact of this process will be to extend reservoir life at sites with less fine sediment, even when the long-term capacity achieved by flushing is quite smalL Table 3.5 quantifies this. effect. It gives predictions of increases in reservoir life due to flushing for various values of the proportion of sand and coarser materials, P sand ' and for flushing discharge at the three reservoirs listed in the Table 3.4. Psand is defined as the proportion of the liver sediment load that consists of sand and coarser material. In each case, the following assumptions were made: • there is a single flushing period of 30 days annually • sediment inflow to the reservoir for all material (wash load and bed material load) can be described by the simple relationship: Concentration = Constant x Discharge1.2
Table 3.5. Effect of the proportion of sand and coarser material on extending the life of reservoirs PropOition of coarse sediment
Factor by which reservoir life is extended
Qllushin/ Qmean I
0·2 0·4
Tarbela
Tungabhadra
1
1·9
1·4
1
2·5
1·8
Rooiport 1-4 1·8 !
I I
0·6
1
3·8
2·7
2·7
0·8
1
7·6
5-4
5·5
0·9
1
15·2
10·8
11·0
0·2
2
2·7
1·6
1·6
0·4
2
3·6
2·1
2·1
0·6
2
5·4
3·2
3·1
2
10·9
6·3
6·2
2
21·7
12·7
12·5
0·8
i
I
0·9
56
FACTORS WHICH INFLUENCE FLLJSHING
$
Q
Sensitivity to the exponent in this equation was slight (about 5% when the exponent \-vas doubled to 2.4) the silt deposits downstream from the main sedimentation delta (this is a conservative assumption, if a proportion of the silt is known to deposit in the delta then that proportion can be included in P sand) a sediment balance is achievable.
The analysis technique outlined in Atkinson (1998) was used and the assumed proportion of sand and coarser material, P sand ' was varied from 0·2 to 0·9 in each case. These results indicate that where a large proportion of the material deposits in the delta, say P sand >O·8, then flushing for 30 days annually can greatly extend reservoir life. This would apply even at sites where flushing does not produce an acceptable reservoir volume in the very long term. Sensitivity to other flushing periods "vas found to be slight, for example reducing the period to 10 days only reduced the factor by which reservoir life is extended by between 2% and 120/0. (It has been assumed that a sediment balance can still be achieved with L.ie reduced period of flushing.) Summarising both the analyses presented in this section provides the following conclusions . ., The sediment sizes in transport in the river can be of paramount importance to the success of flushing in a reservoir. • From the point of view of achieving a sediment balance, a large factor is required benveen the sediment sizes being transported in the river and the sizes found in the river-bed material. Such conditions are typical for gravel rivers with a widely-varying bed material composition. • If a sediment balance can be assured, then a predominance of fine sand, and other material that deposits in the delta at the head of a reservoir, ensures that flushing greatly extends reservoir life. $ Therefore, from the point of view of sediment size alone, delta deposits of fine sand and coarse silt are the most likely to produce success in flushing a reservoir. Coarser material may inhibit a sediment balance arid finer material will deposit in the body of the reservoir outside any incised channel and so will not be available for reworking during flushing.
Evaluation
of criteria at existing reservoirs
Application of these criteria to the 14 reservoirs presented above is given in Table 3.6. Unfortunately, there is insufficient data readily available to include the sediment size ratio, SSR, and the proportion of sand and coarser sizes, Psand ' in this table. When compared with field data, the criteria are able to distinguish reasonably well between the reservoirs where flushing is successful and those where it is not: almost all the criteria were met for the six successfully flushed reservoirs (figures in bold) and at least one criterion was not met for each of the eight other reservoirs.
57
EVACUATION OF SEDIMENTS
Table 3.6.
Application of constraint criteria to existing reservoirs DDR value
SBRd value
FWR value
TWR value
India
0·68
24
3-4
1·6
Gebidem
Switzerland
0·93
20
6·7
1·5
Gmtind
Austria
0·89
58
5-2
1·3
China
0·77
Approx.4
0,1
7·1
Palagnedra
Switzerland
1·00
33
1·4
1·0
Santo Domingo
Venezuela
1-00
11
1·4
1·8
0·04
0·5
Reservoir
Country
Reservoirs flushed successfully Baira
Hengshan
I
I
Reservoirs flushed unsuccessfully
I
Guanting
China
0·81
0·3
Guernsey
USA
0·44
3-2
1·4
0·26
Helsonglin
China
0'77
Approx.1
0·06
0·8
Ichari
India
0·31
33
9·9
1·4
Ouchi-Kurgan
Former USSR
0·14
110
Approx.2
Approx.0·3
Sanmenxia
China
0·75
4·8
0·26
0·9
Sefid-Rud
Iran
0'96
4·3
0·3
0-1
Shulcaozi
China
0·37
15
1·0
2-1
I
The results in the second part of Table 3.6 indicate that at two reservoirs, Ichari and Shuicaozi, changes to the outlet structures at the dam could potentially remove all constraints to successful flushing, while at the other reservoirs, site conditions constrain the success of flushing.
3.4.
SUMMARY OFTHE REQUIREMENTS FOR EFFECTIVE FLUSHING
For effective flushing the following factors need to be considered/satisfied.
3.4. I. Hydraulic conditions required for efficient flushing Riverine conditions must be created in the reservoir for a significant length of time. Flushing is most effective when the reservoir is fully drawn down to a level approaching the conditions which applied prior to impoundment. The reservoir
58
FACTORSWHICH INFLUENCE FLUSHING
level must be held constant at as Iowa level as possible throughout the flushing period. To achieve this: ED
the hydraulic capacity of the bypass must be sufficient to mairitain the reservoir at a low level during the flushing period.
3.4.2. Quantity of water available for flushing There must be enough water available to transport the required volulue of sediment. For a given quantity of water used for flushing, it is more efficient hydraulically to use a high discharge for a short period than to use a low discharge for an extended period. This has the following implications. Reservoirs ~vvhere the annual run-off is large compared with the volume of the reservoir are suitable for sediment flushing. o Reservoirs where L.~ere is a regular annual cycle of flows and· a defined flood season are suitable for sediment flushing. This favours sites in monsoon areas and sites where flood flows are generated by annual snowmelt in the spring and summer months. • Reservoirs where the release of significant quantities of water for flushing does not significantly affect t.he ability to satisfy water demands at other times of the year. &
Flushing discharges of twice the mean annual flow are recommended and the quantity of water required for flushing is unlikely to be less than 10% of the mean annual run-off. This is based on worldwide experience from reservoirs which are being flushed on a regular basis, see Atkinson (1996, 1998), Basson and Rooseboom (1997a and 1997b) and Mahmood (1987), together with detailed numerical modelling of proposed flushing systems, see Attewill et al. (1998) for example. Note: • flushing discharges of at least twice the mean annual flow are required e flushing volumes of at least 10% the mean annual run-off should be anticipated.
3.4.3. Mobility of reservoir sedjments The nature and quantity of river sediments are important factors in determining vvhether the quantity of water available for flushing is adequate to remove the desired quantity of sediment from the reservoir. (j)
Graded bed sediments produce conditions which are the most conducive to the efficient fiusr.ting of sediments. Such conditions are typical of gravel rivers with a varying bed material composition. In large rivers this situation is found where the longitudinal bed gradient is between, say, 0·001 and 0·002. In smaller rivers the eauivalent range may between 0·002 to 0·005. 1. ...
59
EVACUATION OF SEDIMENTS
.. From the point of view of sediment size alone, delta deposits of fine sand and coarse silt are the most easily flushed. Coarser material is difficult to move and tends to deposit at the upstream end of the reservoir. Finer material which deposits in the body of the reservoir outside any incised channel will not be available for reworking during flushing.
3.4.4. Site-specific factors The most suitable conditions for flushing are to be found in reservoirs which approximate in shape to the incised channel which develops during flushing. If the reservoir is too narrow, the incised flushing channel cannot develop its full equilibrium width. If the reservoir is too wide, large areas of sediments will remain on the flanks of the incised channel.
J J
J
J
• L()ng, relatively nan"ow, reservoirs are better suited to flushing than short, wide, shallow reservoirs.
3.4.5. Constraints on the ultimate capacity achievable by sediment flushing In cases where the amount of water available for flushing, combined with considerations of the shape of the reservoir, and the nature of the sediments within the reservoir, lead to a restriction on the amount of sediment which can be removed, the long-term capacity ratio should be assessed in order to check what percentage of the original reservoir capacity can be retained long term by sediment flushing.
1
1
• The greater the sustainable live storage the more attractive flushing systems become. Subject to the economic circumstances, flushing systems will normally be worthwhile if a sustainable live storage of more than, say, 35% of the original live storage can be achieved by flushing.
3.4.6. Economic assessment
1
• A full economic analysis covering the whole life costs and benefits of the flushing system should be undertaken. .
1
3.4.7. Summary Reservoirs in the 'upper and middle reaches' of rivers are likely to be best suited to sediment flushing for the following reasons.
60
1
To be worth doing, the benefits of sediment flushing measured over the anticipated lifespan of the works, must exceed the penalties of loss of power during the draw down period and possible loss of stored water for irrigation and other uses.
1
I
1
FACTORSWHICH INFLUENCE FLUSHING
o
Q
o
3.5.
In the lower reaches, reservoirs are likely to have inundated areas that have previously been flood plains and these areas would not be reached by the incised flushing channel which is inevitably of limited width. The longitudinal slope available for the flushing channel is relatively small, thus limiting the amount of sediment transport. Reservoir volumes in the lower reaches are likely to be larger compared with the mean annual run-off and hence water availability becomes a restraint on sediment flushing.
NUMERICAL MODELS The previous section described the factors which influence the efficiency of . sediment fiuslling and gave guidance on some of the hydrological and design parameters which need to be satisfied. Detailed analysis of specific sites requires the use of numerical models that can provide much firmer estimates of flushing performance. Numerical models can take into account many details that are precluded from the simpler desk calculation techniques. These include: e 4D
c
e
details of the reservoir topography 'details of the long-tenn development .of sedimentation using representative flow sequences details of the annual/monthly/daily operational n11es for the reservoir, in terms of required releases, rule curves for water levels and maximum rates of change of water levels, etc. details of the sediments in motion, including graded sediments where these are a factor of importance.
The calculation methods described in Section 3.4 rely on several assumptions and can only provide an approximate estimate for the design and operation of reservoir flushing systems. A more accurate method is numerical modelling, albeit with a requirement for much more input data. A one-dimensional model is often suited to the simulation of reservoir sedimentation. More complex two-dimensional or th..ree-dimensional models will, in general, require too much data and computational time because simulations are usually required to cover periods of 50 t() 100 years into the future and have a time step of aday or an evep shorter ;p~rlod. For each time step in a 'cme-dimensional'1inodel, th€---\vater levels and flow :. ",.,:' conditions are predicted from discharges and/or changes in storage, 8.J.1.d hence sediment concentrations within each of, typically, 10 size fractions are routed through the reservoir. Bed level changes are determined, using the concept of sediment continuity, from the cbanges in concentrations through the reservoir. These changes in bed level are used to update the bed elevations stored in the model. Usuallybed elevations are stored as full cross sections rather than single
61
EVACUATION OF SEDIMENTS
values, so additional rules are required to determine how deposition or erosion is distributed across the sections. White and Bettess (1984) and Basson and Olesen (1997), as well as other authors, present one-dimensional numerical model applications. The White and Bettess (1984) model has now been combined with reservoir survey analysis software, which uses the accurate Stage-Width Modification Method (SWIMM), to form the PC software RESSASS (REServoir Survey Analysis and Sedimentation Simulation). Two- or three-dimensional models can be used to assess the localised impact of flushing near low-level outlets. Atkinson (1996) briefly presents threedimensional modelling in an idealised reservoir to investigate the extent of influence achieved by flushing. Such modelling would usually provide little useful information on the feasibility of flushing, but may prove invaluable as a component of the design process.
J
J 1 1
1
1 1 1
1 1
1 1 J~
1
1
I 62
1 1 I
I WI e Imen e
fit
0
len e G
..
flu hing
4. Worldwide experience 4. J.
of sediment flushing
INTRODUCTION The main purpose of this chapter is to answer the following questions: how many reservoirs are being flushed? where are they? • are they used for water supply (potable and irrigation) hydropower or flood detention? ., what methods are employed for flushing (the facilities and the operational regime)? Q what is the nature of the sediment? • how successful is the flushing? iii what constraints (operational, economic and environmental) affect the flushing sy'stems? /I what downstream effects occur? e
e
The answers to these questions will allow general findings to be made with regard to the success (or failure) of current flushing operations, the factors that influence the outcomes, and will provide an indication of the trends in the designs for flushing systems. The principal method of appraising worldwide experienGe_has been a comprehensive literature review, drawing on the reference lists of previous reviewers and including fresh searches of library references available orieD and the Internet. Appendix 3 contains descriptions of a number of case studies, where sufficient information has emerged from the literature searches. For each case study, the history and physical features of the reservoir are described, the sedimentation evidence reviewed and the flushing measures which have been implemented are described. In the overall stock of dams worldwide (over 40 000 with dams higher than 15 m according to Morris and Fan, 1997), flushing, in one form or another, must have been attempted in many hundreds, probably th"Ousands of dams. Unfortunately, the amount of accessible documentary evidence amounts to only about 50 cases, with substantial quantitative and qualitative data readily available for only about half of these. Inevitably, the degree of science applied to the design and execution of the flushing process must have varied considerably, while the degree of success would depend on factors such as:
65
EVACUATION OF SEDrMENTS -
i
.j
~
• whether the reservoir and dam were designed taking account .of local sedimentation data, with facilities to enable flushing to take place • the fundamental suitability of the reservoir and dam for undertaking successful flushing • the degree of operational flexibility to allow an effective regime of sediment fl ushing to be undertaken as needed • the application of sufficient know ledge and experience to allow the optimum flushing regime to be developed.
I
4.2.
The physical factors, such as the hydrological setting, reservoir basin geometry and outlet pipework elevation and discharge capacity, that influence the suitability of the reservoir for successful flushing are discussed in detail in Chapter 3.· A few of the main points are given below, as these help to explain some of the experiences of flushing performance.
FLUSHING
Flushing is a technique which, by using a suitable combination of the drawdown (water level lowering) and increased flow in the reservoir, allows previously deposited sediments to . be discharged from the reservoir basin into the ..downstream ri~er or irrigation system. Flushing is undertaken over a relatively short period - usually a few days or weeks and would typically be annual, although there are some cases where it is undertaken once every few years. Flushing may be undertaken with the reservoir effectively empty ('empty flushing'), so that riverine conditions are established, or with the reservoir paItially drawn down (,pressure flushing'). It may be undertaken either during the flood season, as is most common, or outside it. Flushing can be distinguished from sediment 'routing' techniques, which aim to pass the bulk of the sediment load without deposition in the reservoir. Examples of these techniques are:
J r
J
J J
J
• 'sluicing' by drawdown through the flood season • 'sluicing' by drawdown during the main annual floods • density current venting. Routing - particularly sluicing - results in the seasonal pattern of sediment outflows largely following the pattern of sediment inflows, whereas flushing typically compresses the annual sediment load, which may occur over two or three months, into a few days or weeks. Inevitably, there is a potential overlap between the techniques, such as in cases where a significant propoltion of the annual sediment load is passed without deposition, but where flushing is relied upon to erode those sediments deposited during the sluicing operations or during floods outside the sluicing period.
66
-{ J.
WORLDWIDE SEDlMENT FLUSHING
4.3.
WORLDWIDE EXPERIENCE OF FLUSHING
4.3. I. Overview Table 4.1 lists the reservoirs for which the literature search has revealed evidence of flushing, although in a few cases it appears that SOlne form of sediment routing (sluicing or density current venting) may be the major method of sediment discharge. Excluding those cases where there is no hard evidence of flushing, leaves 50 cases, for which the locations and purposes are summarised in Table 4.2. In many cases, the purpose is not provided and in some cases multiple purposes apply, so that the total numbers are not equal to the sums of the purposes. (In a few instances where the reference cites two or three reservoirs or parallel for flushing, only a single case is included in this operated in table.) By far the greatest number of examples is in China, but this is not surprising, because of the size of the country, the numbers of reservoirs (18 800 dams higher than 15 ffi, according to the 1998 ICOLD vVorld register of large dams), and the high sediment yield, particularly in the basin of the Yellow River. It is notable, however, that 42% of the reservoirs listed in Table 4.1 in China, which contains 52% of the dams higher than 15 m and 30% of those higher than 30 m (ICOLD, 1988). "". . . ,............. , it may also be noted that the majority of the examples are from ........ ,..".............. . ., with high sediment yields. Attempts to relate L;e number of examples of flushing in different countries to their stock of large dams and their typical sediment yields would not be fruitful, because of the relatively small sample sizes in most cases, together with a number of other factors which come into play, such as: available to those countries for t.1.e resources - fin~ncial and technical researching and dealing with sedimentation problems • policies for open dissemination of the lessons learned from sedimentation and flushing experience 11 the financial resources to allow attendance at international conferences • languages in which technical papers might be written or receive publication. G>
Of those flushed reservoirs for which the purpose is k..l1own, Table 4.3 lists the numbers falling into each purpose or combination of purposes.
4.3.2. Flushing techniques Of the 54 cases included in Table 4.1, 50 were to involve significant while three were predominantly routing (sluicing) and one flood storage r,:;;>cpr,J'f"'I1r was considered to be essentially uncontrolled, so that the mode would Of the 50 fi ushing a closer resemblance to sluicing, rather than t-!l1C'hl1"\ ......
67
EVACUATION OF SEDIMENTS
Table 4.1.
Summary of reservoirs flushed ii0
Reservoir/dam
Baira
India
Barenburg
Switzerland
Bajiazui
China
Cach!
' Former USSR
Chiyu Dalingkou
I China
H
0
""0
0
::lE
3 .§~ '" ,t:tl
F
E
Q
F
J.6
i
F F
F
F
F
M&F(l997)
F
F
China
F
F
Fergoug
Algelia
F
Ferrera
Switzerland
-.
F F
F
Gebidem --; ...
~wi cierI and
H
1968
2·)
F
F
F
F
H
1945
0·7
F
F :
Guanting Guernsey Heisonglin Hengshan
~ China
!
:
China
Honglingjin
China China
FWH
1953
182
HI
1927
IF IF I
1959
4·3
F
P
61
FRD
1966
84
1960
38
F R
H
1975
0·2 76
1938
30?
=±1
Sudan
HIW
1964
Kunda Pal am
India
H
Liujixia
China
Jensanpei
Taiwan !China
Jiaojiazhuang Khashm El Girba
F D FR
I
:
M
l
La
M&F (1997)
F
Yes Bhargava et at. (1987) UNESCO (1985)
F
I 141-435
P
China
IF
1974
8·4
Naodehai
China
IF
1942
63
U
Nebeur
Thnisia
Ouchi-Kurgan
1961
04
P
PaJagnedra
Former USSR Switzerland
PD FR
1952
2·8
F
F
Prieto
Puerto Rico
F
Rioni
Former USSR
F
Sakura
Japan
6·0
H
1956
F
M
1960
USA
Sanmenxia
China
F
Sanshenggong Santa Maria Santo Domingo
China Guatemala Venezuela
H
1974
0-7
Sefid-Rud
Iran
HI
1962
35
Shiaodaokuo
China
Shimalin
China
Shuicaozi
China
Warsak Yanouxia
Pakistan China
i
I
F F F F
i
i
9 M
Yes SNCOLD (1982)
Lo
M&F(1997)
I
M&F(l997)
ht
Yoon, 1992
!
i M&F(l997)
M
Yes M&F(1997) M&F(l997)
l
I F F F
L N
FM LoP I
8-1] 15-45
La
HI
F
1960
0·8
F F
68
R
Zemo-Afchar
Former USSR
H
! 1927
Zhenziliang
China
I
1958
F
Drawdown Full Partial
M&F(1997) Yes K&C (1979) Yes M&F(l997) M&F(l997) M&F(1997)
F 1958 I 1·9
Qian (1982) Yes UNESCO (1985)
F
F H
Intake forebay
M&F(l997)
Yes C&Z (1992) Yes UNESCO (1985)
L
I
F FRD II F
22
2 in series
Yes UNESCO (1985) P&D(1988)
M&F(1997)
M
F
San Gabriel
I
Yes . Hwang (1985) M&F(1997)
Yes Jowett (1984)
F
DF
I
I
Yes M&F(1997)
F F
HI H
Yes J&M (1963)
LI F P
Nanqin
Mode Flushing Density current Routing/sluicing Uncontrolled
5880
FIFI
USA
1924
(1982)
Yes B&P(1986) I
Morris
1954
i SNCOLD
Yes M&F(l997)
Puerto Rico
H
M&F(l997) M&F(1997)
La
F F FR
M&F(l997) Yes M&P(l997) Yes i R&S (1982)
F
i
W
11
i
Yes UNESCO (1985) Yes Zbang et al. (1976)
New Zealand
i
Qian (1982) 38 17-21
F
M&F(l997)
La
Mangahao
Hydropower Irrigation Flood control Water supply Multipurpose
L
F iF F F
Loiza (Carraizo)
Purpose
I
1
P
I
i
I
FD
1953
i Algeria
12-18 I
Parallel resrs
M&F(1997)
1
1
F
F
India
Iril Emda
EI
F H
M&F(1997) M&F(1997)
I
F
China , Austria
Hongqi Icbmi
F 1985
M&F(l997) Yes M&F(I997) , M&F(l997)
25
F
L
I
M&F(1997)
F
China
China
U
Comment
Principal reference
Yes J&K(1984)
F
Donfanghong
Grimsel
:l~
~.g E
I
Dashikau
Guanshan
,I
~~
17
Dashidaira
Groiind
jL
E-
'00:
c c
!:
;::
0 0·1
1966
c
China ' China
Genshanpei
68
Costa Rica
Chirurt
1981
H
""0
.~
i
~0
..,
~~
.;:.'"
c~ <.)0-
Sediment removal
~
~ ~ I:l..
Country
P P
23-83 592
i
10-67
F
I Season Flood Early flood Late flood Non-flood
Yes IWHR(1983) Yes Mahmood (1987) M&F(1997) I Yes UNESCO (1985) Yes Zhang et (d. (1976) Enhancements Mechanical Lateral channels Longitudinal channels Piping to induce lateral erosion Fluctuating pool level
3 in series
~.
WORLDWIDE SEDIMENT FLUSH!NG
Table 4.2.
Summmy of experience infiushing Numbers of reservoirs flushed
Country
Hydropower
Irrigation or water supply
Flood control
Algeria Austria
1
China
2
Costa Rica
1
Former USSR Guatemala India I
i
II
2
I
3
Iran
Total
1
1
I
I
I
4
I
5
1
21
15
1
1
I
I
4
2 1
I
1
3
1
1
l Japan
Unknown
I
1
I
1
1
New Zealand
1
1
Pakistan
1
Puerto Rico
I
Sudan Switzerland
I
I 1
I
1
1 1
1 1
2 1
3
2
5
2
3
I
Taiwan Tunisia US A
Venezuela Totals
I
1
I
1
I
I
1
19
1
I ,
11
5
I
25
50
cases, five were considered to involve a degree of sluicing and in five densitycurrent venting was considered to be an important contributor to sediment removal. Table 4.4 Slll1lll1arises worldwide evidence regarding the flushing techniques, covering whether the drawdown is partial or complete, the season when it is undertaken and whether it is enhanced by techniques such as lateral channels or a fluctuating pool level. The total sample number for the amount of drawdown and the flushing season is taken as the number flushed in Table 4.1. In all cases where a flushing season is given, it is also stated whether the drawdown is full
69
EVACUATION OF SEDIMENTS
Table 4.3.
Distribution offlushing experience by purpose
Hydropower, irrigation and water supply
Flood control
Multipurpose
25
Total
Table 4.4.
Summary offlushing techniques
Drawdown for flushing
Flushing season
Enhancements
Full
21
Early flood
j
2
Mechanical
5
Partial
7
Flood (not specific)
I
8
Lateral channels
4
Late flood
I
3 . Longitudinal channels
Non-flood
I
1
Not stated
1
36
Not stated
22
i
2
Lateral piping
1
Fluctuating pool
2
I
Total sample
50
Total sample
I 50 I Total number*
12
* Reservoirs where one or more enhancements are recorded
or partial. In six of the cases where an enhancement is recorded, little or no additional data on the flushing operation is available. Enhancements to flushing were apparently attempted at a total of 11 reservoirs out of 50 cases. Excluding the case where fluctuations in the pool level was the only such measure, this leaves ten cases where the enhancement required physical activity by labour and earthmoving plant within the reservoir basin. This represents 20% of the cases. Deliberate fluctuations in reservoir level during the flushing operation are probably undertaken more widely than indicated in the table.
4.3.3. Sediments flushed Most of the references provide little or no information on the sediments being flushed, although it is clearly the case that most of the sediments readily removed
70
WORLDWIDE SEDIMENT FLUSHING
by flushing are silts and fine sands. A proportion of finer materials nlay either be discharged with the water passing through the impoundment under normal operations~ or may be discharged as a density current. Coarser sands, gravels and cobbles are likely to be deposited in the upstream part of the reservoir basin. Given suitable flushing (or sluicing) conditions, these can be drawn down into the lower part of the impoundment and ultimately discharged downstream, but this tends to be a longer-term process, generally associated with a permanent rise in the form of a delta and braided channel at the upstream end of the basin and with associated permanent loss of reservoir storage capacity.
4.4.
CASE STUDIES OF RESERVOIR FLUSHING
4.4.1. Summary Over 20 case studies are included in Appendix 3. These appear in chronological order of constnlction completion or first impounding, in order, to some extent, to illustrate the development of knowledge in sediment problems and remedial measures, including flusriing. The salient features and key findings from the case sIDdies are given below, in the same sequence. Table 4.5 summarises the mrun descriptive and quantitative information on these and a few other reservoirs. Table 4.6 presents the main quantitative infonnation for the case studies only, including the key ratios concerning reservoir volume, annual inflow and sediment load and flushing discharges and volumes, where available. Also included in Table 4.6 is a subjective assessment of whether flusPing has been successful, 'which is discussed furt.~er below. JiJangahoo reservoir (New Zealand, 1924) This example suggests that sedimentation was a consideration in the design, but that no specific planning for sediment flushing was included, because it was expected that more dams were to be built upstream. Fortunately, a generous lowlevel outlet could be made available by recommissioning gates on the diversion tunnel, which had remained out of use after problems 25 years earlier. This was successful in removing a large proportion of the accumulated sediment, after which annual flushing has been undertaken.
Guernsey reservoir (USA, 1927) This is a small reservoir in relation to its annual inflow (4·3%), but it does not have low-level flushing facilities and apparently there was no attempt at flushing for the first 30 years of its operation. Partial draw downs for annual flushing between 1959 and 1962 scoured sediment from the upstream part of the basin and redeposited it in the lower basin, having a very small effect on the total volume of sediment deposited in the basin. Fortunately, the construction of
71
Table 4.5.
Detailed list of reservoirs subject to flushing
Reservation/dam
Country
Purpose
Year built or sll\rlcd impounding
Year modified
I Calchmenl
Max depth: m
Surface area: km'
i area: km'
Original capacity: M.m'
Basin 1enlllb:km
Annual inft ow: M.m'
I
1990
I
Designed for flushing?
AM",' sediment inHow:Mt
;
India
Baita
Cachi
Costa Rica
Gebidem
Switzcrland
Hydropower
I
1981
Hydropower
1966
Hydropower
1968
5] (dam height) 35 (diversion lunnel)
1996
785
69
200
lB
4-1
I 3·24
I
6
14
I
24
.0.3 (Atkinson, 1996)
(Atkinson, 1996)3500 (J984ref) 54
1500 (appro~.)
9'()
420
Yes, via diversion tunnel
()'81
Single bottom oUllel suitably localed
0·5
Yes, two flushing lunnels below bydropower 'nlake
I Gmund
Austria
Hydropower
1945
Gaunting
China
Flood control
1953
1967
156·3
311
0·J24
0·94
0·93
43400
43
229
30
2270
i
135
0·210·7
1250
73 (19505) 7 (19805)
Yes
Generously-sized bOllom outle~ but flushing not practicable because of downstream impacts
13 (B&P
I USA
Guernsey
Heisonglin
Chin.
lITigation & hydropower
1927
Irrigation & flood control
1959
42000
29
9·6 (1963)
24
91
!
value)
J.7
2100
No
(1927-57)
370
I
30
2·9
8·6
0-70
14·2
Probably, as lhe disch.tEe capacity of]O mS/s is much grelltctlban the mean inflow I
Hengshan
China
Irrigalion & flood control
1966
163
65
I'()
13·3
15·8
1·18
IYes, although main flushing outlet is 15 m above original river-bed
Honglingjin
China
Irrigalion (probnbly)
Icbari
India
Hydropower
Jensanpei
Taiwan
Waler supply (sugar cane)
1938
Khashm EI Gima
Sudan
lITigation, hydropower
1964
I !
1960
1364
<42
16-6
43·2
11-6
5300
0·77
I 1975
37
1955
II
5·7 (1976-84 Apparently not for basin, but lIushing provided for sediment mean) 2·2 (median) exclusion at inlBke Flushing gallery added 1955
7·0
10·6
950
84
& water supply
Loiza (Carraizo)
Pucno Rico
Water supply
1954
Mangahao
New Zealand
Hydropower
1924
534
23
2·7
9
Probably, as lbe discharge capacity of 28 m'/s is much greater Ihan Ihe mean inllow
27
449
Seven 7 m )( 7·3 m bottom outlets controlled by radial gales
0·38
i
I
!
No, upstream reservoir buill as sediment trap
I !
Chillll
Nan'lin
Naodehai
\ China
Irrigation & flood control
1974
Originally flood control. lanerly also
1942
453
1970
4-5
29
4501
I
121
0·69 or 0·531
168
265
16
10·2
Yes
I
I irrigation
I
Yes, lowcring by 5 m during Hood season. through 8 bonom oUllets
35
17
564
IS 000
13
138
55
2·6
5·5
3041 (199 in SINCOLD 1982)
0·08
NO! known, but bollom outlel available
1966-71 & 1990
688400
55
120
9640
43000
1600
Yes, bu! facilities bad to be upgraded in Stages
None
427
65
()'2
Yes, apparently
56200
50
Probably, us thare are five lowlevel outlets
Ouchi-Kurgan
FonncrUSSR
Hydropower & irrigation
1961
Plllagnedru
Swil.7.crland
Hydropower
1952
1974
Sanmenxia
China
Muhipurpose .
1960
I
I
(m340)
•
Venezuela
Sanlo Domingo
Hydropower
1974
Irrigalion (primary) & hydropower
1962
Hydropower
1958
! Irrig.tion&
1960
l'()
3'()
450
82
25
1760
5000
28
"
9·6
514
()'63
No (no bottom outlet)
42
170
21 100 (1961-70)
15·3
Apparently nol
6600
Approx.5
53-8
342
Approx.0·2 -[
I
Sefid-Rud
lImn
I
!
Shuicaozi
I
China
Pakistan
Warsak
67340
hydropower
lemo-Akhar
Fonner USSR
Hydropower
1927
Zilcnziliang
China
lrri~li"n
1958
(prohably)
18
I
* Key: A6.1, e\c. Alkinson (1996) Chl9, elc. Morri. & Fall (1997) DETR This "'I,on LTCR Long-tcrlll capacily ratio (Atkin.on. 1996)
<14
1740
I
I<
36-6
Protr.bly. as the discharge capacity of 57 m'/s is much greater th..... the mean inflow
S~"\Iillh"11t:nit1R
Mudd studics'!
e.\pcricm:e
Rcfcrcm:cs
C'lS&!
LTCR:
Slud i l!~"
Mt::m mit:
~ri~inall\'
t!$limalt:d a.__ 0'0'12 M.m.l/vr (.lbmn 4%
interrupt ion lO thcrt!aftc:"
mCf':lhs
E;ilim;JlI.:d lhat 18%
nUW! thtl~uyh
wiltKJut deposition.
pas~c:; by dCtUiit!l current "cnlin~ and 18% dcposill!L.!
Yes. :t((c:rrm.'hlern~ futlnd
Used Llivcniiun lUllltcl. ck::uin!! O·J8 M .m·' in 40 huurs go:nt:rJliun~ ann"'utli nashing P(\'Poscd
o(ori~inal 'l';r"~e): bUI ()·45 M.nr' ao:urnulalcd in firsl I~
54~
A6.1 DETR
JM''''''' (19921: R:unfre7. & RUdri;u.:7.\I99Z): M".,.is & Fan 11~9il
Commenced 1973 :lnd undcruIkcn I~ [imc~ in 13 \lars. reduced lr.lppinl:: i'rt1m 8~~ [('I ~7C;r. (fi~,t1tt:.'i on Id:t nc~d - ufter nushln~ SI:utr.:L.!'.')
~5
Cil 19 DETR
da.rifyin~
Virtually nu sccJimcnl .3CCUI11UI:ltiun.
h..:C:lU~C
of gOl".;e:·lyre
1;l!omt.·try J.nd annual tlu~h.in:;
0·2 Mllyr initially. ".'ducin~ to 0·07 Mllyr ufter upsl=m ","ervoir built 1967 350 M.m' dCpt"ilCd 1953-60: sUbsequenrly m~ny up~ueam l"C'ScrYoirs constructet.!. 5ubs(:lntially n:ducing sedimcm inl1ow:,. B&P value for annual scdim1!nl inllow recommended for plill1ning pu"",ses in 1986
R~:;.crJl1ir cmpltcJ flu!-J. day~ per yc:v ;,md about J ~·t.nl" of WOlter used; unt!crta~cn rrum ('Iutsct; hypa.'is tunnel considered. but ~jcch",-d un C(tst grounds
Phy!\il:al
Flusbing undeOUlken intermillCnlly 194:1-60: annuully Ih."",rLOr
Yc.,
DUWUl1s ",,,I. (1982): SNCOLD (1982), UNESCO 1I9H5): Alkin.on 1.1996); Mom, & F:m ( 1997)
1·62 M.m·' deposilion in fir~t :; yt!Ol'S of op.!r.ltion (6% slur"lI" los., p
Ao.3 DETR
Only one nu..t1ing operalion (1954) reported, removing 10% of annu:::al inUow, parlly hy venting clen:oity currt!nt!'
IRTCES (1985); UNESCO (1935); Atkinson (1996); Moni, & F:ln 11997); Binnic & Partners (1986) (some dot. incon5istent betw~n source.c;)
A7.1 DETR
20
AlIen1fued in Four y~ 19'9-62: nc;u ccmsidc::n.=d ccnntlmic:li ur en'ectiv~. 0lS rt!covcreti
J.r~"Cki
A7.2 DETR
26
&: Murphy (1963): UNESCO (1985): Mahmood (1987): AllUn",n 11996); Moms 8< Fan (1997)
Zhan;., 01. (1976): Xia i1~SO): UNESCO (198;1: Atkinson (1996); I-Io.,.is ," Fan (1997)
, From 1962. llc::nsity cum:nt venting and tklod Se3.C;lln sluicing r<:duced lrap efficiency 10 about 15%: lalenll erosion IAO<:hniquc successfully implemented from 1980. ra:ovt!ring some losl storage: long·u:nn c:::1pxity c~pccled !Q be 30-35% of ori¥inal
3·19 M.m·· dopa,ilcd 196C~73. rcuching deplh of 27 m :II dom
A7.J CIt25
DETR
AM
IRTCES (19S5): UNESCO 119S'): Atkin.'iOn (1996):
Emplied & nu..hod ror 37 days in 1974. removing O·M M.mJ of depo.sits: 52 days in 1979 removcd 1·03 M.m"
Moms & F:tn (1997} (su,,"..: d~b inc,,"si~lent bt:"tween
I I
77
DETR
SOLH't.~)
0·57 M.m"' deposited per year 1960-63 in impound!ng mode. representing 3·5% s[omgc loss per ye:u-
\Varer 1evel lowered in nood lensan, resuttin!! in suh"t~\ntial reduction in r:ue of "crag. In.'. 10 0",,' M.m; p
ZI,.ng er al. (1976): IRTCES (1985); UNESCO (198;): Alldnson (1996)
S.:c!imcnlO1tion reached spillway crest after one year; 85% tr:::apping much gre-Jler (h:L"l indicated by Brune's curvcs
Flushed annually by fully-op.ning spillway gaI.s: problems with :abr.1sion dumage to spilhv:::ay ~d roller
Bharg.Y:1., al. (lq87): Mohan ., al. (1982); Alkinson (19961
A7A DETR
Hwong (1985); Paul.l:: Dhillon (19881: Monis & Fon
DETR
( '.~,!):
99
Rieno..1 &: S
rut
~9 ·3% ofstorn2C lust 1927-'7 when ~ime"I
A6.! Ch~1
DETR
36
.lllLicipatcd long.lcrm c:1pacit:{ :wout 3S%
SLOr:J~t:
loss
~·26
M.m' 1938-55. repre!>enting 3-4% per
Flushing commenct:!d 1955 for 1·5 month." aa1l1uillly, virtuaity mesting ~ub~qucnl sedimcntarion. hut not re.... toring c3f':Jcity. minor mising of im['lOunding level :1bout 1942 and 1958
year I Cap~ity
seriously de~ie1l!d
L..:.lS{ 53t':~ ot' ..:a9:.u.:i(y aw:le.s blocKctl
(I'1~7)
Flushing opo<"Jtions in 1971 .nd 1973 c.ch ",moved 85 Mt
lCJ5J-~~~
:.hrcc {lOU mm
low·lcv~J
Mechanic:Ji mt:thods auemptcd unsuccessfully in 199J.; pass-through planned and .,petted to
HEC6 (Morris & Hu. 1992)
~...'!!im"nt
UNESCO 1198S): EI H.gToyeb (1980): El FailhSiI:ld (1980); CFGB (J 973 '" 1982)
DETR
Webb &: Soi~r·L.orc7. (1997): Morri • .l; Fun (1997)
Ch2Q
JOW"lr (1984); Alkinson (\ 996)
OETR
.~cdimc:H.:ltk'll':;
59% of Slara!e 1o", by 1958: prublem inc,••• in~I)· ",riOu.. by mid-I 96C.'
Slor.gc loss 53;~ by 1983 (~ppa ...:ntly bas.a ~n WL of t rS In. whereas nVL = ! 24 enl: life span then l!:tpl!::wd :c be 2000 if RUlOhing nca in~tigatt:d
CilPllCity reduced
LO iiOOUL
50% hy ! 950.
bUl
:ccuvcr:t.! to
about B
I
Flu.hod in 1969 through low·l".,.cl dive"ion tunnel anLi 75% of accumulated .!
Dun~ity current venting commenced
I
p~ing :lbout
Chen", Zhao (1992): Mmris 3< Fon (1997)
DETR
64% of annual scdimonlload in 1977-84: ~xp"ritnen,"1 ftushing from 198J. with gaud rcsuIL'Y: (.'Uncludcd Rushing ,hould O~ under!okcn for 4 clays overy 3-4 years
I I
IRTCES (19851; L'NE5CO (19851; A.kinson (1996)
ilo!totn oulict> ung.l<:ci prior to 1970.• 0 fl\1Sllinll appears
1
Bod level, nJ>" up to 23 m by 1969, ,edimen. volume oppe-.Jrs to have stabilised at 30 M.m.' sir:<:c t953
19i7;
to have been j);&IUr.:tI
ISluicrd for ~ month., annually sine. 1963 I
1973 Hood c;tuscd l·S M.:rr' dC(l<~.liit:on (33% of origin:J.1 bmtom \lur!el
IRTCES (1985): UNESCO (19851: Arkinson (1996)
. Lie::hi & Hacoerli (I 97O'J: SNCOLD (1996)
Flushing (aioe'!! by m""banic;ll pl.nt) II/iS-03179 rcmovc:u 2,.", ~I.mJ: 1760 m 10n1: scdiment bv!)a.5.S tunnel
.slor~f!} and .liuhnK:~c.:d
119~2):
AtKinscn
I
A7.5 DETR
10
A6.3 OETR
100
11.7.6 Ch14 DETR
39
A6.6
100
:~d;:i:lQ~~~;\~nl~~i~:~:dl~: villUmly full cupaelty can
low .. It:'IIct Out[CL;; flushed for 4 months annually; six devl!!0r>ment r.rage~ de~ri;,cd in li1ernture
I o·sa M.m·· depo;'.e!! in two years 1975-78: o·n "'I.m' in
Only on. 1Iu.
four years 197-1-78
I
K;umdick & Ch.:!nlot
Yes
~eu !~,:c!s ::.t
rmp!:::,'!·.:n:et.!
::''tpcrimt:~t::~!y from 19€5; b:.1t !imi!:!c by
(1~79);
Atkinson (1996)
DETR
A7.7
Toio"i •• r al. (1991): Mahmood (IY'7); Aikin,'on (1996);
Flus hing (:::about 4 month"/yrJ commcncdi in 198C: after 7 years 25% of lost storage h:r.tJ been reco\'cred: from t 99~ fluodplain erosion ~nn:lllced using diversion ciumnel!:: ~."(pe.;:~~d mat long·umn staragt: CJposcity could be up to 90% uf original
Sc',ere. causing less or 2· t % of !.he :.tor.lge c:lp:!clty ~r year up 10 19S0 (Imp efficiency 73%); mas I nf sedim"", relc:JSC' occurred in d..:n!ittj' currents
8,: 3 ~'l.m·l {~3% d' 5tarJg~) tos~ :~1:::1-81 ; dam oni:- "] m ~dow impoundin~ tc.:\'ct
Zhang &: long (1980): UNESCO (19851: Atkinson (19961: Mom, &: F,n (1997)
Rehanili!.lLion from 1966 includod con;[ruttion of Inrger
Severe. with 18110 ,'vlt dep<>.
DETR
I
nigh elcv:1tion uf !ipillway and short dl.!rltiun :;u1nual1y lo abOUt one third of inflow
i
I
(WeHR i
!9~~):
IRTCES
(19~51: liNESCO (IQ·i 5r
A'kin,on ( 1996)
I I
A7.Pt
DETR
DETR right b.u:k. lc:lding to puw'.:r
int~~.;s
I
76%
1
4·3 M.m.' dcp0:iitl"C.I r.1vUl!. rcp(!!..:cntin~
p~r
yt::tr 195lJ-6 1 in impoundln~ 12% ${Qi.!g..! )o!t; per YC:lr
I :
hnpiemcnlcd from 193'J. with. full dr:l'.\'do','In :md ~pp~Jr!:d to k..:ep SitU:lLion :;tubll! up 'D J953 . n:rnoving :tbout l M.rn' f':r yt:lr I \\:'alcr !e\lcllowcn:d in Ihmtl.,ea.'ion. rt:~uILin~ in :;ub!:.lambl ;eduction in iJ.tc orstoiJ~~lo~ t~)O · 77 M.m' pc:- ."~".U' Itjrjl-1J: tC'chnique:s ·.!!i5t!'nt1i!lIy rm.!li;"1gJ~bicir.g
I
13
ChD
I
'fUC~ {19S5): Atkinson i IY96): l.iN ESCO Morris &: Fan (1997)
Zh:;.ng
t:(
tTl.
!!~76j
DETR
{Ii.j~j,;
I·I
39
~~
m
~
()
C
~ Table 4.6. Reservoir
Year
Capacity: M.m3
C
;;:.,'
-~
Mean annual flow: M.m 3
I
Mean or median sediment inflow: Mt
Mangahao Guernsey
1927
Zemo-Afchar
1927
Jensanpei
1938
8·1
Naodehai
1942
168
265
16
Gmund
1945
0·93
135
0·07
Palagnedra
1952
5·5
199
Guanting
1953
2270
Shuicaozi
1958
Heisonglin
91
Flushing discharge:
m 3/s
Flushing volume: M.m3
Ratios: % CIl
SC: tlm3
VII
Q/I
2100
1·7
125
54
6600
5
450
30
4·3
1·9
188
2·6
215
0·5
Atkinson (1996)
9·5
0·7
7·5
0·08
2·8
1·5
1250
13
182
0·6
9·6
514
0·63
50
1·9
6·6
307
1959
8·6
14·2
0·71
10
61
8·3
2221
Sanmenxia
1960
9640
43000
1600
22
Warsak
1960
170
21 100
15·3
Ouchi-Kurgan
1961
56·4
15000
13
-~:
~
-1. - --'
25
0·45
4·32
2800
-
-l - j -
-- - L.~
583
Current or recent estimate
26
0·3
Comment
o
~rn
3:
a.>
()
m
()
::l
Z
tI)
Y
Insufficient data
N
Note I
70
Y?
98
85
Y
100
100
Y
Assisted by bulldozers
N
Note 2
Insufficient data
Y? Y
Note 3
17
50
Y
Note 4
0·8
9·0
10
N
0-4
23
L-J
1
'-
-' -~
Vl
Y
30
---L-J'
-t
45
30
589
""T1 Vl
m
Y? Original capacity not given
39
0·8
~
C"-.
20
"
63
J~"-=",,,
LTCR%
V
Q
S
1924
-1~",,";,,,j
o o
z
Summary of key flushing parameters
10
40
L_-J
L-......:
Y? Note 4
--L-J
,_.
I --.
J
L.- . ..J_
.I
\
Sdid-Rud
1962
1760
Khashm EI Girba
1964
950
5008
50
100
R4
7100
1067
35
2·8
63
2]
1966
1].3
15·g
Cach(
1966
54
1500
Y? Note 6
84
17 0·8
3·6
71
3393
429
20
3
2·}
5·6
Y
Note 7
Y -
0·5
75
1·5
9·0
Y Note 5
--
I-Jt>ngshan
1968
75
8·8
-
Gebidclll
13
147
--.
..
0·7
99
99
Y
100
96
Y
75
Y
36
34
N?
Note 9
85
85
Y
Note 10
-
~-
SanlO Domingo
1974
3·0
450
0·2
10
0·7
6·7
70
Nunq'in
1974
10·2
121
0·5
14
8-4
5·2
365
1975
11·6
5300
2·2
0·2
19
1981
2·4
2700
0·3
0·]
13
Note 8
---.. lchari - -.
naira
-
100
9
117
0·3
NOles: I. Flushing diseharge given as 120-1L10 ml/s at 12-13 III drawdown; Q and V from Atkinson (1996) 2. Value or sediment inflow ('or planning, as recommended in 13 innie & Partners (1986) report; flushing not acceptable due to downstream constraints 3. Sediment management includes sluicing during 1100d scason and lise of lateral channels I 4. Main sediment management technique is sluicing during flood season 5. By using longitudinal flushing channels, anticipated Ihal long-term capacity could be up to 90% 6. More inrormalioll may be available in 1980 references by EI Hag (1980) and El Faith Saad (1980) 7 . Much greater Ilushing disc\larges possible via a higher outlet (abollt 25% of water deplh above base of dalll) 8. Mllch t~realer Illlshing discharges possible if needed 9. No hollom ourlet; llushing via gated spill.way only 10. Mean anllual !low taken as mean of two values reported in li rerallll'e
~ 0
;0
r
0
~
0 n1 (/I
rn
0 3: fl1
Z
-I 11
r C
VI
I
Z (.)
"
. :..
.:0,
EVACUATION OF SEDIMENTS
I
upstream reservoirs controlling most of the catchment appears to have reduced the need for flushing.
Zemo-Afchar reservoir (Former USSR, (927) After over a decade with limited drawdown, apparently having little effect and allowing about 75% of the storage capacity to be lost, active flushing was undertaken between one and four times per annum. The -data are ambiguous, but suggest that the long-term accretion has been arrested and some reversal achieved. Jensanpei reservoir (Taiwan, 1938) The capacity of this reservoir is believed to be about 30% of the mean annual inflow, which may be considered as the boundary between hydrologically small and large reservoirs. In the first 18 years of its operation, over 60% of the storage was lost, but this was then arrested and an equiliblium maintained through annual flushing. This is undertaken during the latter part of the non-flood season, which coincides with a period of no water demand from the industrial consumer. Naodehai reservoir (China, 1942) This was originally an uncontrolled flood detention reservoir, with a capacity of 63 % of mean annual inflow, but gates were later added to the bottom outlet to allow some impounding of clear water for irrigation. The mode of operation probably bears more resemblance to sluicing than flushing, with riverine flow est~blished for much of the year when flows are low. During floods, which contain most of the annual sediment load, the water levels rise and there is some deposition over the flood plains, which will dry and consolidate between floods and which will not be amenable to subsequent erosion. During the first 30 years of operation, the available storage volume has ranged between 58% and 80% of the original capacity, with the lowest value having occurred in 1950. It appears to be dominated by massive deposition in the largest floods, followed by a period of progressive erosion, but the data are too limited to judge the degree to which any active flushing may be practised through the operation of the gates. Gmiind reservoir (Austria, 1945) This reservoir is hydrologically very small, at under 1% of the mean annual inflow. There. is a coarse bed load, which is bypassed around the reservoir in a tunneL Periodic flushing was undertaken initially and the storage loss reached 20% by the early 1960s. Annual flushing commenced in 1960, an additional bottom outlet was constructed in 1963 and -sediment inflows were substantially reduced after 1967 by the construction of an upstream dam. As a result, the storage loss stabilised and a modest recovery occurred. Flushing is greatly aided by the ability to control inflows by releases from the upstream reservoir. Palagnedra reservoir (SWitzerland, 1952) The reservoir is hydrologically small, at less than 3% of the mean annual inflow. Sediment inflows appear to have varied substantially from year to year, with
76
VVORLDVv'IDE SEDIMENT FLUSHING
major floods contributing the majority of sediment. It suffered sediment deposition amounting to 27% of the original capacity by 1968 and 69% by 1978 (the latter probably largely the result of the August 1978 flood). A sediment bypass tunnel was constructed in 1974 and a major flushing operation lasting four and a half months was undertaken in 1978-79, aided by bulldozers, clearing most of the deposited sediment. It is not known what flushing operations have been undertaken since, but it appears likely that annual flushing would enable most of the capacity to be retained.
Guanting reservoir (China, 1953) This is a hydrologically large reservoir, at about 1·8 times the mean annual runoff. The rate of sedimentation has reduced progressively over the years, from an initial value of about 3% per annum to 0·3% in the 1970s, apparently due to the construction of a vast number of reservoirs within its catchment and the diversion of :bighly turbid flows for warping the agricultural land. Only one case of partial drawdown is reported, in 1954, which removed only 10% of the annual sediment inflow. Downstream impacts, including a water supply intake, two hydropower schemes and the effects of accretion on flood risks, prevent substantive flushing being undertaken as the primary means of sediment management.
Shuicaozi reservoir (China, f 958) Shuicaozi reservoir is hydrologically small, at than 2 % of the annual inflow. With no botton1 outlet available for flushing, only partial drawdown was undertaken by way of the spillway, which was found to be of limited efficacy. In spite of six flushing operations between 1965 and 1981, sedimentation was very severe, consuming 85% of the oliginal storage capacity by 1981. An improved fiuslIing procedure was adopted 1984 and was found to be successful, but no further data are available. I-Jeisongfin reServoir (China, 1959)
This is a small reservoir, but it is hydrologically large with a capacity/inflow ratio of 61 %. Following sedimentation problems at many earlier Chinese reservoirs, this reservoir was used as a test-bed for sediment management techniques. Initially, it was operated only with impounding, resulting in average siltation of 60/0 per annum over the first three years. Subsequently, the mode of operation was changed to emptying the reservoir during the flood season, when the water is most turbid, and impounding only during the non-flood season. This technique which is essentially sediment routing or sluicing - reduced the trap efficiency to 15%, but further improvement was required in the longer term. Starting from 1980, experiments were carried out, resulting in the development and routine use of lateral erosion techniques. Between 1980 and 1985 deposits amounting to about 6% of the storage capacity were recovered by this method, at a very high sediment/water ratio of 23 %. It is expected that a long-term balance bet\veen sediment inflovvs and outno\vs can be maintained at Heisonglin reservoir, with a residual capacity of
77
J EVACUATION OF SEDIMENTS
J about 30% of the original capacity. The sediment released from the reservoir is used beneficially for agricultural warping.
]
Sanmenxia reservoir (China, 1960) This is a large dam, built on the world's most silt-laden river, so it was originally designed with sediment management measures. However, the assumptions upon which these were based proved to be highly optimistic and major sedimentation problems arose very quickly, threatening increased flood risks up to 260 km upstream of the dam. Hydrologically, Sanmenxia is just small enough to be classified as 'small', with a capacity/inflow ratio of 22%. After about five years 40% of the original capacity had been occupied by sediment deposition. The reduction and control of sediment deposition and the development of a sustainable sediment. management . regime became a high priority for the Sanmenxia reservoir. The development occurred over six stages between 1962 and 1978, and the final technique is largely one of sediment routing by sluicing through the high-flow season and impounding for irrigation and hydropower in the low-flow season. Since 1975, the net storage capacity below elevation of 330 m (10 m below top water level) has been stable in the range of 50-55% of the original capacity at that elevation. The lessons learned at Sanmenxia have guided subsequent projects, including the Three Gorges.
J
Warsak reservoir (Pakistan, 1960) This is a hydrologically small reservoir at less than 1% of the mean annual inflow, on a river with a bed load which includes gravels and cobbles. The total annual sediment load is equivalent to an annual accretion of about 8% and, after 20 years of operation, the reservoir was filled with sediment up to conservation level, except for a channel leading to the power intake. There is apparently no substantial bottom outlet to the dam, and flushing, which was attempted by way of the gated spillway on five occasions between 1976 and 1979, was unsuccessful. The Warsak reservoir has apparently reached a broad equilibrium, with virtually no residual live storage. Ouchi-Kurgan reservoir (Former USSR, 1961) The Ouchi-Kurgan reservoir is hydrologically small, with a capacity of less than 1% of the mean annual inflow. Soon after construction it has been flushed annually, but apparently with only a limited drawdown. The volume of deposited sediment has stabilised at 50-55% of the original capacity since 1968. Sefid-Rud reservoir (Iran, 1962) This reservoir is hydrologically large, with a capacity/inflow ratio of 35%. Sedimentation was a serious problem over the first 17 years of its operation, reducing the storage capacity at a mean annual rate of 2·1 % and reaching a minimum of 63% in 1982-83, before recovering as a result of flushing measures. The flushing measures comprise emptying the reservoir from October to Febluary, outside the irrigation season, then refilling it dUling the early part of the flood season in time for the strut of irrigation in May.
1 1 r
1
1 ]
1
1
I I
1
1 I
I I I
78
L
WORLDWIDE SEDIMENT FLUSHING
Recovery and maintenance of storage capacity has been aided by lateral erosion, by piping and by the use of a longitudinal diversion channel. · It is anticipated that, by creating a new diversion channel each year, it would be possible to reach a Iong-telTIl storage capacity of 900/0, compared with 75% by flushing alone.
Khashm EI Girba reservoir (Sudan, 1964) · No information is available on the ratio of the reservoir capacity to the annual inflow. The limited data available on flushing operations suggests that it has been successful. Hengshan reservoir (China, 1966) This is a small reservoir on a steep stream and is hydrologically large, having a capacity/inflow ratio of 84%. In the first -eight years of operation, 24% of tl]e original capacity was occupied by sediment deposits. Flushing, wpich was undertaken with the reservoir empty in 1974, 1979, 1982 and in 1986, was effective in recovering lost storage and indicated that flushing every few years would be sufficient in this case. Although the Hengshan reservoir is hydro- . logicaUy large, flushing was probably effective because of the steep valley gradient and side-slopes. Cach{ reservoir (Costa Rica, 1966) The Cacri reservoir is hydrologically small, at 3·6% of the mean annual inflow and the al.1-Dual sediment load would have a deposited volume rather over 1% of the original storage volume. For the first seven years it was operated without flushing and trapped 82% of the incoming sediment load. Between 1973 and 1990 flushing was undertaken 14 times. It was effective in the lower part of the basin, but less so in the upper part, which is progressively filling with sand and coarser material. Overall, however, flushing has been considered successful in maintaining the storage capacity at the Cacm reservoir.
r
r
Gebidem reservoir (Switzerland, 1968) Gebidem is a hydrologically small reservoir, with a capacity/inflow ratio of 2·1 %. Sediment inflows are high due to glacial activity, with stone sizes up to 100 rnm, and the potential to absorb over 4% of the original storage per annum. The reservoir has been flushed annually in the flood season and this has resulted in virtually the entire storage capacity being preserved. This is attributed to the gorge-like geometry of the basin and the steep valley slope. There have been problems with downstream sediment accretion, where the valley slope reduces, which were expected to be overcome by deploying greater flushing discharges. Santo Domingo reservoir (Venezuela, 1974) This is hydrologically small, at less than 1% of the merul annual inflow, and contains two brfu'1ches, being built at the confluence of two rivers. Sediment loads \vere expected to be high in relation to the original storage capacity, at about 8% per annum, and model studies had been undertaken which suggested that the installed flushing facilities would be sufficient. For the first four years the
79
EVACUATION OF SEDIMENTS
reservoir was operated without flushing and the total sediment accretion was about 25%. The first flushing operation, in 1978, with full draw down , was estimated to have removed 50-60% of this accretion in three or four days. Over a further three weeks, with assistance from bulldozers, the original storage capacity was virtually restored, with an estimated 3-5% loss remaining. It was concluded that, in the future, flushing should be undertaken annually, preferably towards the end of the high-flow season. It was also considered that empty flushing should occasionally be inten"upted and be followed by a short period of pressure flushing to concentrate sediment removal on the immediate areas of the bottom outlets. Nanqin reservoir (China, 1974)
Nanqin is a hydrologically small reservoir, with a capacity/inflow ratio of 8%. Initially it was used for flood detention only, then from 1976 flows were impounded to a middle level for irrigation. By 1983,53% of the storage capacity had been lost, following whIch, from 1984, an improved regime of sediment management was put into place. The steep longitudinal slope of the reservoir was suitable for density current venting, which was practised from 1977, achieving a trap efficiency of 36%. The first empty flushing was undertaken at the end of the 1984 flood season and was highly effective, removing all of the sediment deposited that season, together with an additional amount deposited earlier, equivalent to 7% of the original storage. J'he operating rules that were subsequently derived, maintain the high pool level during the flood season to trap bed load deposits near the upstream end and prevent them armouring the more erodible deposits further downstream. Empty flushing is to be undertaken every three or four years, at the end of the flood season, and density current venting is promoted. It was estimated that a longterm storage capacity of the order of 75% of the original capacity could be maintained by following these rules. Ichari reservoir (India, 1975) This is hydrologically very small, with a capacity of only 0·2% of the mean annual inflow. The annual sediment load is highly variable, with a median value having the potential to consume about 20% of the original storage per annum. This was borne out in the first year of operation, when the storage capacity was reduced by 23%. Over the first six years the total loss of storage was 60%. The sediment ranges up to cobbles and has severely damaged the spillway roller bucket. Although the dam includes facilities for excluding coarse sediment at the hydropower intake, there is apparently no low-level outlet for flushing sediment from the dead storage of the impoundment. Flushing by way of the gated spillway was undertaken annually from 1976 and it appears likely that an approximate equilibrium, with the dead storage entirely filled with sediment, existed from about 1980. The long-term storage is likely to be about 35% of the original capacity.
80
WORLDWIDE SEDIMENT FLUSHING
Baira reservoir (India, 1981)
The Baira reservoir is hydrologically very small with a capacity 1.1 the order of 0·1 % of the mean annual run-off from the catchment. In the first 18 months of operation, almost 20% of the original capacity had been consumed, representing at least double the average annual sediment load assunled during the design. The construction period· diversion tunnel had been fitted with gates to facilitate flushing and model studies had suggested that this wouldbe capable of removing virtually all the deposited sediment. The first flushing operation was successful, removing over 80% of the deposition in 40 hours, and it appears that annual flushing should be effective in maintaining a large proportion of the original storage capacity.
4.4.2. Findings The findings from this review of t.~e case histories, together with the limited information on a number of other reservoirs where flushing has been undertaken, can be considered under the following subhea?ings: , the hydrology of the catchment • the sedimentology of the catchment " the storage capacity of the reservoir o the sediment deposition'-potential 8 the shape of the reservoir basin e the low-level outlet facilities provided • operational limitations e whether full or partial drawdown is to be deployed e the scope for enhancements to flushing o downstream impacts o the critelia for judging the success of flushing.
Hydrology. "';' The hydrology of the catchment needs to be properly researched and understood, as it is central to the consideration of the other issues which affect the practicability and likely success of flushing. It is necessary to k.TJ.OW the typical patterns of run-off within the year, together with the ranges of variations encountered within the year and from year to year. This information is important both for 'broad-brush' assessments and for mathematical model simulations of reservoir sedimentation. If local fio\v gauging records are inadequate, additional expert hydrological appraisals will be needed. These might make use of national or regional hydrological paran1eters, supplemented by techniques such as flow gauging data transposition (from within or outside the drainage basin) and flow gauging data record extension, using correlations with longer periods of rainfall records. The data intervals used need to take account of the size of the catchment, varying from an hour or less when considering local floods in very small
81
1 EVACUATION OF SEDIMENTS
1 .catchments, up to. perhaps a week fer very large catchments, particularly these where a large part o.f the flo.o.d flew is go.verned by glaciers and sno.wmelt. Upstream reservo.irs, that are already in existence o.r planned in the future, can have the fo.llo.wing effects en water inflews to. the reserveir: • en the to.tal annual inflo.w, if flews are diverted fer irrigatio.n er transferred into. ether catchments, o.r subject to. increased evapo.rative lesses in sto.rage • en the seasenal distributien ef inflo.ws (unless the upstream reservo.irs are few in number and all hydro.lo.gically small) • en the availability and co.ntro.l o.f flushing flews when they are needed . . In the case o.f Guanting reservo.ir, fer example, the censtructio.n o.f ever 300 reserve irs upstream and increased water use fer irrigatien between the 1950s and 1970s, reduced the tetal river flews entering the reservo.ir. At the Gmund reservo.ir, the censtructio.n o.f an upstream reserveir allo.wed a high degree o.f co.ntrel to. be exercised ever inflews during flushing. Acco.unt also. needs to. be taken o.f ether land-use trends affecting the catchment hydro.lo.gy, such as urbanisatio.n and defo.restatien.
Sedimentology The cellectio.n o.f useful sediment data is a vitally impo.rtant issue, because large errers in predictien can be made if proper acceunt is net taken o.f the large variatio.ns in sediment co.ncentratio.n which naturally o.ccur. An insufficiently frequent and rigo.reus sediment data co.llectio.n pro.gramme is" liable to. underestimate severely the large co.ntributien to. the annual sediment lead - beth suspended and bed lead - which derives fro.m the highest discharges. Wherever po.ssible, the available sediment data sheuld be tested against experience o.f reserveir sedimentatio.n in the area, to.gether with natio.nal and regienal data en sediment yields fer the seils and geelo.gical co.nditio.ns feund in the catchment. The ultimate ebjective ef the sediment studies is to. o.btain reliable values fer the mean annual sediment lead and the degree o.f variability fro.m year to. year, tegether with particle size distributio.ns and to. derive hydro.graphs that give the seasenal prefiles o.f sediment lead which co.mplement the seaso.nal flew hydro.graphs. Catchment changes also. need to. be taken into. acco.unt, as in the hydro.lo.gical studies. Reductio.ns in sediment leads entering reservo.irs due to. the develepment o.f upstream reservo.irs have been repo.rted fer the Guernsey, Gmund and Guanting reservo.irs. At the Guanting reservo.ir, fer example, the reductio.n in sediment leads, due to. reserveir and irrigatio.n develepment between the 1950s and the 1970s, was much greater than the reductio.n in annual inflo.ws. Co.nversely, fer the Mangahae reserveir, the expected develo.pment o.f additienal upstream reservo.irs, to. reduce sediment leads, did · net o.ccur. Where the censtructio.n o.f upstream reserveirs prevides a respite in sediment lead, it sheuld be remembered that this may be ef limited duratio.n, as the upstream reserveirs fill with sediment, er as sediment flushing and sluicing measures are implemented.
82
! ,.--L
1 _1
1
1 1 1
1 1
'vVORLDWIDE SEDIMENT FLUSHING
Storage capacity The key storage capacity parameter is the ;hydrological size' which is the ratio of the storage capacity to the mean annual inflow. Table 4.6 lists reservoirs with ratios of between 0·1 % and almost 200%. A ratio of 30% may be considered as an approximate boundary between hydrologically large and small reservoirs. If the ratio is less than about 30%, there is a reasonable prospect of having sufficient flow available to allow the reservoir to be emptied for flushing annually, generally in the early part of the flood season, so that it can be reliably filled in the later part of the flood season. The smaller the ratio, then the more practicable flushing becomes, from a water resource standpoint. Smaller ratios, of perhaps 5% or less, allow more rapid emptying and refilling of the reservoir, and so suit relatively short periods of flushing. Su.bject to tlle other constraints (in particular u~e adequacy of the bottom outlet and the suitability of the basin shape), most hydrologically small reservoirs appear to have been flushed successfully, examples being Gmiind, Palagnedra, Cacm, Gebidem, Santo Domingo and Baira. However, this is not a guarantee of successful flushing if other requirements are not met, such as at Guernsey and Warsa..1(. If the ratio is much larger than 30%, then it becomes increasingly difficult to schedule a flood-season flushing regime that will still meet the water storage objectives, which generally require the reservoir to be full by the end of the flood season. Successful flushing has nevertheless been undertaken at reservoirs witJ.1. a higher ratio, an example being the Hengshan reservoir, aided in that case by its small size and narrow steep Valley. Once the ratio approaches or exceeds 100%, it is clear that an impounding reservoir (other than one used solely for flood control) is designed for the carryover of water from one year to the next, to cover shortages in drought years. Annual empty flushing of such reservoirs is not possible, but there may be the possibility of empty flushing once every decade or so, if this is beneficial. There may also be some benefit in flushing with partial draw down , at the lowest annual water levels. Reservoirs which are initially hydrologically large and impnlctlcable to flush, may become practicable to flush as sedimentation reduces the storage capacity, allowing an acceptable residual capacity to be sustained in the long term. The Heisonglin reservoir probably falls into that category. o
0
Sediment deposition potential As well as the relationship between reservoir storage capacity and mean annual inflow, it is relevant to consider the potential accretion which would result if there were 100% trapping of the sediment load. A potential accretion rate of 1-2% per annum probably represents a reasonable boundary between reservoirs where flushing should be started early and those where it might be delayed for perhaps 20 years. Potential rates of 5% or higher certainly spell danger, requiring flushing to be fully planned in the design and implemented from the outset. Of the reservoirs listed in Table 4.6, for which this information is available, 50% have
83
EVACUATION OF SEDIMENTS
potential rates (in tonnes of sedimentlm3 of water) above 6-7 % (corresponding to potential accretion of about 5% or more), the highest being over 20%. Very few reservoirs in Table 4.6 have potential accretion rates of less than 1-2%, but this is not really surprising, as there is clearly less need for, and interest in, flushing where the potential rate of accretion is very low. Hydrologically large reservoirs, except where sited in the areas of the highest sediment yield, tend to have low rates of potential annual accretion, so may not be expected to need active flushing during their economic life. The Guanting reservoir· would appear, on the basis of the raw figures, to now fall into this category, the sedimentation problems having arisen because of higher historic sediment loads, which were nearly all concentrated into the reservoir arm that had a lesser storage capacity. Nevertheless, in the case of large dams, proper consideration needs to be given to the long-term conditions, including the practicability of decommissioning and the problems associated with any eventual release of the reservoir deposits.
Basin shape Narrow steep-sided 'gorge-like' reservoirs are clearly more amenable to effective flushing, pruticularly where the longitudinal gradient is steep. To some degree, this is often offset by poorer mobility of the coarser bed load present at many of these reservoirs. Several of the hydrologically small reservoirs listed in Table 4.6, such as Gmund, fall into this category and have been successfully flushed. The Hengshan reservoir, Which is hydrologically large, but steep-sided and with a steep valley gradient, has also been successfully flushed. Broader reservoir basins are vulnerable to sediment deposition over the flooded floodplains, leading to two problems with flushing: • when the deposits are exposed during draw down of the reservoir, they tend to dry out and consolidate • they are isolated from the flushing flows, which tend to be in line with the original watercourse, so are not subject to significant erosion. The operational regime (for example, whether the reservoir is empty or full at the time when the flood anives) can have an impact on the vulnerability of the former floodplains to progressive deposition. As a result, the residual storage capacity resulting from flushing broader reservoirs is generally limited by the geometry of the channel that can be eroded by the flushing flows. The eroded channel gradient, width and side-slope angle are generally functions of the sediment characteristics, valley gradient and the flushing discharge, but may also be limited by other features of the geometry of the basin. Guernsey, Naodehai, Heisonglin and Sefid-Rud reservoirs are examples of broad reservoir basins where the basin geometry acts as a constraint on the efficacy of flushing. The flushing of reservoir basins, which include the valleys of tributaries, is also likely to be influenced by the relative magnitudes of the tributary flows. An
84
WORLDWIDE SEDIMENT FLUSHING
example of a reservoir with limited inflows available in a significant tributary valley is the Guanting reservoir.
Low-level outlets The primary requirement, even in reservoirs where empty flushing may not be desirable for operational reasons or considered likely to be necessary for many years, is that there should be effective low-level outlets near the bottom of the basin. \Vithout these, there is no possibility of undertaking empty flushing if and when required in the future. If the lowest outlet is at mid-height, for example, this provides a pelmanent constraint to flushing, limiting it to the less effective partial drawdown. This is the case at the Guernsey, Shuicaozi, Warsak and Ichari reservoirs. J.4. common rule of thumb for successful flushing is that the discharge capacity of L.~e low-level outlets should be sufficient to Dass at least twice the meaTl annual . inflow at a drawdown of the pond level by at least 50%. This drawdown elevation may allow sediments to be effectively scoured from Ll.e upstream half of the reservoir length, although some of the coarser material would be expected to be redeposited in the downstream half. However, this criterion should not be ta.l(en to imply that only partial drawdown is needed. Full drawdown clearly has the potential to be more effective. If the reservoir is hydrologically smail, so that a relatively short flushing .. period is possible, this discharge capacity criterion is probably a reasonable one for flushing outside the flood season. In a larger reservoir, where the time taken to draw the reservoir down would be longer, this discharge capacity may be inadequate from an operational standpoint, even outside the high-flow season. There would also be the risk of the drawdown being interrupted and extended by periods of higher flow, which would also tend to result in further deposition. For full draw down flushing in the flood season, the discharge capacity would probably have to be significantly greater than the above rule-of-thumb value. For t~e design or checking of a flushing system, consideration therefore should be given to actual hydro graphs at the proposed time of flushing, simulating the time ta..ken for the reservoir to empty for flushing, then to refill under a range of flow conditions, together with th~f'robabilities of being able to keep the reservoir pool at the required level during the flushing period ... The Sanmenxia reservoir is an example of a reservoir where obtaining an adequate discharge capacity through the low-level outlets for flushing (or sluicing) through the flood season was vital for the success of the sediment managemen t measures. ~
.
Operational considerations The relevant operational considerations regarding flushing are mainly those associated with the lack of water supplies and/or the reduction in operating head, as they affect hydropower generation or irrigation supplies. These considerations vary considerably from site to site. For example, if the irrigation supplies are abstracted downstream and the irrigation system is designed to accept turbid water for warping the agricultural land, such as at the Heisonglin reservoir, there
85
EVACUATION OF SEDIMENTS
may be no disruption and even a benefit from flushing. Seasonal demands will also influence the time when flushing is most convenient and the amount of time and water which can conveniently be used. At the Jensanpei and Sefid-Rud reservoirs, industrial water supply and irrigation demands respectively favoured flushing during the low-flow season. In the final analysis, of course, operational considerations must not be allowed to prevent effective flushing to be undertaken, if that is needed for the preservation of the resource for future generations.
Drawdown The case histories . show that full draw down to achieve empty flushing is preferred, but there are cases where this is not practicable, owing to limitations imposed by the available flushing facilities or from operational considerations. The available outlets may prevent full draw down due to either their elevation (as at the Guernsey, Shuicaozi, Warsak and Ichari reservoirs) or because of insufficient discharge capacity, or a combination of reasons. In cases where a limited discharge capacity prevents full draw down during the flood season, it may be possible and more effective to undertake flushing at lesser discharges during the non-flood season. Enhancements Of the enhancements listed in Table 4.1, fluctuating water levels during flushing have been reported to be beneficial at the Gebidem and Santo Domingo reservoirs. In practice, the technique has probably been more widely employed than reported and would be expected to be always beneficial in terms of enhancing sediment outflows. Fluctuations in flushing discharge have also been found to be beneficial to the encouragement of slumping failures of the channel banks. The other enhancements all involve human intervention, preferably widl earthmoving plant, on the reservoir ·deposits. In some of the cases where bulldozers have been used to shift sediment towards the main 'channel and increase its rate of disposal, the intervention was experimental in the first flushing and may not be found worthwhile after a regular flushing regime has been instigated. The two cases (Heisonglin and Sefid-Rud) where lateral channels, lateral piping and longitudinal channels have been successfully deployed (and where sufficient information is available to judge) are hydrologically large reservoirs with significant areas of deposits over the fonner floodplain. Without these measures, there would be a significantly worse prognosis for the preservation of storage in the long term. Downstream impacts There have been severe impacts on the downstream aquatic environment in a number of cases, principally where heavy deposition or high suspended sediment concentrations affect the habitat and the survival of fish and other wildlife. In this respect, Sh011 peliods of flushing are particularly problematic. Sediment routing,
86
.!
WORLDWIDE SEDIMENT FLUSHING
by sluicing through most of the flood season, on the other hand, is "much more benign environmentally. The potential adverse environmental impacts downstream should always considered in the light of the alternatives, such as an acceptance of long-term sediment accretion, requiring the further use of natural resources for such things as the construction of additional reservoirs or the development and use of alternative sources of power. In some cases, reservoir operators are subject to regulatory limits on downstream sediment loads or concentrations, wrJeh have to be taken into account in the detailed planning of every flushing operation. Because of the variable nature of the phenomenon, unless a very large factor of safety is employed vvith implications for the duration of flushing needed, occasional noncomplial"lce is almost inevitable. The downstream impacts can be rnitigated substantially if there is dilution available a short distance downstream, for example, from hydropower releases or at the confluence wit.~ a larger river, such as in the case of the Rhone downstream of the Gebidem reservoir. The control of flushing - for example, in response to downstream concentration monitoring can be aided if inflows to the reservoir can be controlled, for example, by releases from an upstream reservoir, as at the Gmund reservoir. Other downstream interests that may be affected by sediment releases include: recreation, such as boating or swimming • water supply intakes III hydropower intakes reservoirs. @
Ill)
Criteria for the success of flushing Past attempts to define objective criteria in order to judge the success or otherwise of flushing have been: G
Q
L~at
there should eventually be a balance, over time, between the sediment inflow to and outflow from the reservoir that the sustainable long-term capacity of the reservoir s.hould be at least a certain proportion, typically 40-500/0 of the original reservoir capacity.
Criteria such as these, however, are probably unnecessarily restrictive. Whether an absolute sediment balance is required depends on the potential rate of loss of storage. In the case of a large reservoir which might lose 2% of its capacity per annum, flushing which lowers that rate to 1% might be considered successfui, as it would lengthen substantially the useful life of the reservoir. Similarly, an ultimate equilibrium long-term capacity of only 200/0 might be considered perfectly acceptable at a reservoir vvhere this was expected at the outset and where alternative solutions for water supply, power generation or flood control are less attractive.
87
.
,
, I
EVACUATION OF SEDIMENTS
J The key test is that a practical compromise should be achieved between the processes of sedimentation and the requirements for beneficial use of the reservoir. The long-term beneficial use may be much diminished from what might have been expected at the time when the reservoir was designed and built, but in comparison with the poorer or even catastrophic outcome that may result without intervention, any tangible improvement from flushing must be judged a success to some degree. In these terms, therefore, the flushing undertaken at most of the reservoirs presented in the case studies can be considered successful. Several of these are cases that were considered unsuccessful by Atkinson (1996), based on the more restrictive criteria listed earlier. The reservoirs at which flushing is not possible or has been unsuccessful are probably limited to the following: Guernsey Guanting Warsak Ichari
Effective flushing is not possible, due to limitations of low-level outlets, but apparently it is not needed, because of vastly reduced sediment inflows. Major flushing is not feasible because of downstream constraints; the reservoir is required to contain sediment and prevent significant . downstream sedilnentation. Flushing is not possible because of a lack of low-level outlets; a lack of live storage limits hydropower generation to the 'run of the river' . No low-level outlet, so cobbles and gravels must pass down the spillway, damaging the roller bucket.
J. J J
J
4.4.3. Summary of findings The findings from the review of the worldwide experience of flushing can be summarised as follows. • The hydrology and sedimentology of the catchment need to be fully understood in the planning of flushing facilities for new or existing reservoirs and need to provide the background for analyses of past sedimentation and flushing performance. • Successful hydraulic flushing is more likely to be practicable in reservoirs that are hydrologically small, with a storage capacity less than 30% of the mean annual inflow. The smaller the reservoir, the greater the chance of it being successfully flushed and the greater the likely residual storage capacity. • Flushing is vital for the preservation of long-term storage in reservoirs where the sediment deposition potential is greater than 1-2% of the original capacity. Even in large reservoirs with a potentially long life, consideration should be given to possible eventual decommissioning problems when deciding whether or not to flush. • The shape of the reservoir basin can have a large impact on the practicability of effective flushing and on the residual storage capacity. Narrow steep-sided the easiest to flush. reservoirs in valleys with a steep longitudinal slope
are
88
J
j
j
I
...l
J 1
WORLDWIDE SEDiMENT FLUSHING
"
" • o
s
•
e
Wide valleys, where the impoundment covers former floodplains, can be flushed less effectively, because the deposits tend to consolidate and 'are remote from the flushing channel. For effective empty flushing with full drawdown, the low-level outlets must be both low enough and of sufficient capacity to allow the draw down to be controlled during the time of year when flushing is undertaken. Proportionately larger outlets are required for flood-season flushing than for flushing outside the flood season. Operational considerations, such as water and power demands can inhibit the ability to flush successfully, but they must not be allowed to prejudice the longterm preservation of an important resource. Full drawdown and empty flushing have been found to be much more effective tJ."'1an partial drawdown. Fluctuations water level and discharge during flushing are beneficial to the promotion of bank slumping, increasing the rate of sediment discharge. The deployment of lateral and longitudinal diversion channels has been successful in promoting flushing in reservoirs that are hydrologically large or u1.at contain significant proportions of deposition in areas remote from the main flushing channel . Downstream impacts can act as a constraint in the planning and operation of flushing. In some cases flushing may be ruled out, whereas sluicing, which approximately preserves the seasonal distribution of sediment load, may be a practicable alternative. The degree of success in flushL.'1g should be judged by whether it makes a worthwhile difference to the beneficial uses of the reservoir, rather than simply by whether it meets numerate objectives, such as a long-term balance between inflows and outflows, or the retention of a certain percentage of the original storage volume.
89
ro· hi it d to
I
5. Geographical areas suited to flushing 5.1.
WORLDWIDEVARIATION IN EROSION RATES
5.1. I. Factors that affect erosion The erosion rate depends on a cOlnplex interaction of the following factors:
. (a) Climate (i) precipitation and run-off (ii) temperature . (iii) wind speed and direction. (b) Geotechnics (i) geology (ii) volcanic and tectonic activity (iii) soils. (c) Topography
(i)
slope
Oi) catchment orientation (iii) drainage basin area (iv) drainage density. (d) Vegetation
(e) Land use and human impact These factors are discussed in Appendix A4.1. It is not easy to generalise between areas of high and low erosion rates as it depends on their geographical location.
5. 1.2. Estimates
of global sediment yield
The estiInates derived from more than a dozen studies of global average rates of denudation have ranged from 0·06 to 0·16 rnmlyr (Morris and Fan, 1997). Estimates for the aggregate worldwide sedilnent yield of between 15 and 20 x 109 t/yr have been given (\Valling and \Vebb, 1996). Areas with sediment yield over 1000 tlkIn2/yr are 8·8% of the total land area and account for 690/0 of
93
EVACUATION OF SEDIMENTS
the total sediment load. Regions with less than 50 tJkm2/yr account for about half of the land area and 2·1 % of the sediment yield. Case studies of erosion rates are presented in Appendix A4.2.
Continental variations in erosion rates A number of estimates have been made of sediment yield on a continental basis. Results from two such studies are presented in Tables 5.1 and 5.2. Sediment yield in Asia is four times larger than South America even though South America experiences the highest rates of run-off in the world. The sediment yield for basins in Asia is over twice the world average and contributes
Table 5.1.
I
~I
Continental variations in sediment yield (Mahmood, 1987)
Area I
mm
km3
North America
756
15·8
15·4
Asia
740
25·7
25·0
Africa
740
19·7
1600 790
South America Europe
Sediment: Mtlyr
%
Run-off: km3
%
Precipitation
Yield: tlkm2Jyr
6·6
17·1
1460
10·9
84
10·8
28·0
6350
47·4
380
19·2
4·2
10·9
530
3·9
35
27·0
26·2
11·8
30·5
1790
13·3
97.
7·5
7·3
2·7
230
1·7
50
60
0-4
28
3000
22·4
1000
13420
100·0
165
I
1
7·0 I
Australia 791
%
7·1
2·5
6·9
6·5
Oceania I i
Total
102·8
I
100·0
38·6
100·0
Table 5.2. Continental variations in sediment yield (Jolly, 1982, taken from Gregory and Walling, 1973) Suspended sediment: tllcrn 2Jyr
Continent
Suspended sediment: 10 Mtlyr
27
550
600
16160
Australia
45
230
Europe
35
330
96
1990
63
1220
Mrica Asia
North America
i
~j
J J
J
I
South America
J
94
f
AREAS SUITEDTO FLUSHING ·
approximately 80% of the world sediment total (Jolly, 1982). The largest sediment yields occur in Oceania at 1000 t/km2/yr including ntllnerous catchments in. New Zealand, New Guinea and Taiwan with sediment yields two to three times the world average. There are considerable differences between the continental figures produced by the two studies. The table produced by Mahmood (Table 5.1) distinguishes between yield rates in Australia and Oceania and this leads to the highest sediment yield rates in Oceanian rivers at 1000 t/km2/yr. This is hidden in the study by Gregory and Walling (Table 5.2), where Australia is taken to include Oceania. This produces a higher rate for Australia than found in the study by Mahmood but a much lower rate than for Oceania. The next highest sediment yield is produced by Asia with a yield of 380-600 tlkm?/yr. The lowest rates of sediment yield occur in Australia (owing to aridity) in the Mahmood study, at 28 tfJs:m?/yr, and in Africa, at 27 tlkm?/yr, in the study by Gregory and Walling.
Vaiiations in sediment yield by drainage basin The areas where extreme erosion rates occur are, of course, hidden in the statistics that define areas according to continent. Various attempts have been made to identify the drainage basins worldwide which have the highest sediment yields. Table 5.3 presents the results of one such global study. Table 5.3 shows high erosion rates mainly in Asia, Oceania, the USA and in Eastern Europe. The two highest values of sediment yield, both around 8000 tf1
e
Asia, especially in China and India Oceania, especially in New Guinea South America, especially in Peru and Colombia.
These rivers do not produce the world's highest sediment yields as the list excludes drainage basins smaller than 10000 km2 which are likely to produce the highest yields per unit area.
Areas of high erosion rates Further COITh.'TIents on areas of particularly heavy erosion: by various researchers, are given below. o
Areas of high erosion include mountainoLls areas, such as the Andes, Himalayas anYd Karakorams, parts of the Rocky mountains and the African rift
95
J EVACUATION OF SEDIMENTS
1 Table 5.3. World maximum recorded suspended-sediment yields greater than 2000 tlkm 2/yr (Jolly, 1982, from Gregory and Walling, 1973) River
Location
Average annual yield: tlkm2/yr
I
J
l
r--'-
Ching
Changchiashan, China
8040
Lo
Chuantou, China
7922
Waipaoa
Kanakanaia, New Zealand
6982
Tjatabon
Java, Indonesia
6250
Lo-Lo
Luyang, China
6068
Pietracuta, Italy
4570
Semani
Urage, Kucit, Albania
4150
Soldier
Pisgash, Iowa, USA
4072
Shkum Bini
Paper, Albania
3590
Kosi
Chatra, India
3130
Yellow
Shenhsien, China
2957
Indus
Kalabagh, Pakistan
2498
Santa Anita
Arcadia, California, USA
2374
Eel
Scotia, California, USA
2292
Marecchia ...,;. .
.
J
1 1 I
valley, and areas of volcanic soils, such as Java, South Island of New Zealand, Papua New Guinea and parts of Central America (Morris and Fan, 1997). • The Pacific Asiatic-Australian sector demonstrates the most intensive rates of erosion. Figures in the range of 10 000 to 50 000 tJkm2jyr have been reported · at stations in China, Taiwan, the Philippines, Indonesia, Java, Kenya, New Guinea and New Zealand (Walling, 1994) due to active tectonics and volcanism, steep slopes, high precipitation amounts and intensities, high and irregular lun-off, . dissected mountain relief composed mainly of sedimentary rocks, and human influence by agriculture and logging (Dedkov and Moszherin, 1992). • Taiwan discharges more sediment to the ocean per unit area than any other country in the world. Streams draining the central range produce suspended sediment yields of 13 760 tlkm2jyr. One small basin exports 31 700 tJkm2jyr (Li, 1976). The sediment discharge of Taiwan is nearly five times larger than that from the continent of Australia, even though it is 210 times smaller. Lower values of 11-12000 tJkm2 have been reported in Java (Walling and Webb, 1983). • In New Zealand values reach between 20 and 28 000 tlkm2jyr with a mean value of around 2000 tJkm2jyr. The highest mean annual sp~cific suspended-
96
.---L
1 ·1 1 f
1 j
1
AREAS SUITEDTO FLUSHING
Table 5.4. Rates of sedilnent yield for the world's major rivers at ocean level, excluding basins with an area less than 10 000 km 2 (IYlalirJ'lood, 1987) Continent
River
Country
I
I
Drainage area: million km 2
Run-off: cm/yr
Oceania
New Guinea
Purari
0·031
248
South AmeLica
Peru
Chira
0·02
25
Sediment: tlkm 2/yr
I
Yield:
ppm
2581
1039
2000
I
8000
I
Asia
China
Daling
Asia
China
Haiho
Asia
China
Yellow
India
Damodar
Asia I
North America I USA Asia :
I II
I Copper
5
1800
36 000
0·05
4
1620
40500
0·77
6
1403
22041
01.02
50
1400
2800
65
1167
1795
1128
1720
1083
1057
917
928
01.06
I
I
I
66
1·48
Gange sIB rahm
Bangladesh
0·02
I
Asia
Vietnam
Hungho
012
103
South America
Colombia
Magdelena
9.24
99
043
100
616
619
80
500
625
0061
126
492
390
097
25
454
1849
31
27
310
1143
Asia
I Burma
Irrawaddy
i
USA
Susitna
Oceania
New Guinea
Fly
Asia
Pakistan
Indus
Asia
I India I China
I
Orinoco
·99
111
212
191
Colorado
9.64
3
211
6750
Europe
Mekong
Brazil
I
0·17
I
!
I I I
I Vietnam I China
340
59
Pearl
9.44
69
157
228
Amazon
~'15
102
146
143
I
a·l1
6
145
2286
II
C·1
I
30
130
433
(·09
I
S4
111
204
I
t
Negro Rhone i
203
·79
Brazos
~
531
t
326
Mexico
South America
I
214
North America
I USA II Ar2:en tina I France
I
66
Venezuela
North America
246
9.07
South America
South America
46
I
6833
Po
I
I
I I
241
Italy
Asia
°r 1·94
I
4
Europe
Asia
I
Yangtze
I I Liaohc .
USSR
~; ~
005
I
Godavari
4l,,:)Ld.
1\
I
..
North America
Asia
I
I
i
I
I
I
.'
.
97
T
1 EVACUATION OF SEDIMENTS
Table 5.4.
continued
Continent
1
Country
River
Drainage area: million krn 2
Run-off: crnlyr
Sediment: t/km 2/yr
Yield: ppm
North AmeIica
USA
Mississippi
3·27
18
107
602
Africa
Tanzania
Rufiji
0·18
5
94
1889
North America
Canada
Fraser
0·22
51
91
179
Europe
Romania
Danube
0·81
25
83
325
Africa
Mozambique
Limpopo
0·41
1
80
6600
North America
USA
Yukon
0·84
23
71
308
North America
Canada,",.
1·81
17
55
327
North America
USA
0·02
60
50
83
' ''''M ackenzie Hudson
"
"
"
1 1 i
1 I
~
Asia
Iraq
Tigris-Eupha
1·05
4
50
1152
Europoean AItic
USSR
Indigirka
0·36
15
39
255
Africa
Egypt
Nile
2·96
1
38
3700
South Amelica
Argentina
La Plata
2·83
17
33
196
Africa
Nigeria
Niger
1·21
16
33
208
Asia
USSR
Amur
1·85
18
28
160
Oceania
Australia
Murray
1·06
2
28
1364
Africa
South Africa
Orange
1·02
1
17
1545
Africa
Mozambique
Zambesi
1·2
19
17
90
Asia
' India
Mehandi
0·13
52
15
30
European Artic
USSR
Yana
0·22
13
14
103
European Artic
USSR
Sev. Dvina
0·35
30
13
42
North America
USA
Columbia
0·67
37
12
32
Africa
Zaire
Zaire
3·82
33
11
34
South Amelica
Brazil
Sao Francisco
0·64
15
9
62
European Artic
USSR
Kolyma
0·64
11
9
85
European Artic
USSR
Ob
2·5
15
6
42
1 1 j
I
-1
" I'
98
European AItic
USSR
Yenisei
2·58
22
5
23
European Artic
USSR
Lena
2·5
21
5
23
North America
Canada
St Lawrence
1·03
43
4
9
i
1 I
..J..
1 T
AREAS SUITEDTO
sediment yield is 53 SOD t/la.n2/yr for the Huangfuachan river (3199 k1112), a tributary of the Yellow River in China (Walling and Webb, 1983). Table 5.5 lists a number of basins with very high yields in various countries, which were reported in 1983 and which exceed the values of record yields published in 1973 and shown in Table 5.3. For tributades of the Yellow River, highly erodible loess, lack of vegetation and the semi -arid climate are the major controlling factors. The semi-arid climate is a factor in the Kenyan example, but severe disturbance due to agriculture is also a factor. For Java and New Guinea steep relief, high rainfall and agriculture are important and in New Zealand the steep relief, high rainfall up to 9000 mm1yr, and tectonic activity play a role (vValling and Webb, 1983).
Areas of low erosion rates Global minima below 2 t/k.rn?/yr have been documented. Douglas (1973) cites a yield of 1·3 tlkm2/yr for the Brindabella catchment (26·1 km2) and 1·7 t/k.m 2/yr for the Queanbeyan River (172 km2) in the southern Tablelands and Highlands of Table 5.5. 1983)
Values of sediment yield in excess of 10 000 tlkm2/yr (l;Valling and vVebblec
Country
River
China
Dali
China China
I
Kenya
II
t
Mean annual sediment yield:
Source
tfkrn'1jyr
96·1
25600
MOli and Meng (1980)
Dali
187
21 700
Mou and Meng (1980)
Dali
3893
16300
j Mou and Meng (1980)
Perkerra
1310
(Unknown)
Taiwan
Drainage area: km 2
I!
I
(Unknown)
I
19520 31 700
Dunne (1979) Li (1976)
Java
Cilutung
600
12000
I I Hardjowitjitro (1981)
Java
Cikeruh
250
11 200
Hardjowitjitro (1981)
New Guinea
Aure
4360
11126
Pickup et aI. (1981)
Waiapu
1378
19970
Griffit.~s
Waingaromia
175
17340
Griffiths (1982)
Hikuwai
307
13 890
Griffiths (1982)
South Island
Hokitika
352
17070
Griffiths (1982)
New Zealand
Cleddau
155
13 300
Griffiths (1981)
f
North Island
[
New Zealand
New Zealand
I
J I
(1982)
99
EVACUATION OF SEDIMENTS
New South Wales, Australia. Values of less than 1 t/km 2/yr have been reported in Poland (Branski, 1975) (from Walling and Webb, 1983). Areas of low sediment yield are usually flat, arid with inadequate streamflow to transport large sediment volumes, or arctic regions with low relief, little precipitation and human impact (Morris and Fan, 1997) or low mountains of the temperate zones that are underlain by crystalline rocks and covered by dense deserts e.g. Scandinavia, the Urals, the mountains of South Siberia and the Trans-Baikal region (Dedkov and Moszherin, 1992). The results quoted for Poland appear to be anomalous.
" 5.1.3. Maps
1
of global variation in sediment yields
A number of maps have been produced to illustrate global variations in sediment yields. The maps of Strakhov (1967) and Fournier (1960) are based on 96 and 60 observations respectively and are shown in Figures 5.1 a and 5.1 b. There are large discrepancies between the maps with values on the Fournier map frequently of an order of magnitude greater than on the Strak...lJ.ov map. A later study by Walling and Webb (1983) was based on 1500 stations with basin sizes from 1000 to 10 000 km? The highest values are associated with the loess areas of China and the Cenozoic mountain areas around the Pacific Margins. High values occur in mountainous areas, mediterranean, semi-arid and seasonally humid climates (Walling and Webb, 1983). Low values occur in the desert regions, areas of low relief and glaciated regions (Walling and Webb, 1996). A better likeness can be seen between the maps produced by Lvovich et aI. (1991) and by Walling and Webb (1996), both based on drainage basins of between 1000 and 10 000 km2 (Figures 5.2a and 5.2b) In this study we have used the map produced by Walling and Webb to generate rates of sediment yield in every country of the world. These are tabulated in Appendix A4.4. Despite the use of a relatively small unit, such as the country, highly variable rates were found within many countries. A number of large countries would clearly be better sub-divided into smaller homogenous regions but this is a task beyond the scope of this study. Jansson (1988) makes a number of observations on the 1983 map (Figure 5.2b). • Small rivers in Taiwan, Java and Borneo and in the mountain areas of Pakistan and Soviet Central Asia have high yields. Thailand and Cambodia have low values while the Philippines are intermediate. Low values occur in Africa except for the mountains in northern Africa and pal1s of South Africa and Lesotho. • In South America the values for large drainage basins are low. In the north, . smaller basins which drain the Andes have high values. In the Andes in Venezuela and Colombia, the small rivers have . extremely high sediment yields. Western Central South America has "fairly high levels while the mountains of northern Argentina and Bolivia have high values. The lowest rates of .s~diment yield occur in northern Chile, in the mountains east of the desert.
100
",
i
1
- ,.
1
AREAS SUITEDTO FLUSHING
o
5.2.
In northern USA and in Canada there are low sedin1ent yields, ' except in Alaska and the mountains, in southern USA and along the coast' of California (Jansson, 1988).
CLIMATIC ZONES OFTHEWORLD
5.2./. Introduction An understanding of the precipitation regimes throughout the world may provide a key to the definition of areas of high and low erosion rates. We briefly describe these in the section which follows. It is difficult to classify distinct climatic zones as they tend to merge into one another rather than have sharp boundaries but a number of general models have been produced. These are discussed in Appendix
A4.3.
5.2.2. Precipitatjon regimes and their seasonal variation High annual precipitation Figure 5.3 shows the global mean annual precipitation for 1998 in rnm per month. This shows that annual precipitation values are greater than 150 mm per month along the eastern edge of Asia in Vietnam, China, Bangladesh and NepaJ and into the Pacific Islands of Japan, Malaysia, Taiwan and Papua New Guinea; High rates in Mrica of over 100 rom per month occur in the western centraf region around Cameroon, Gabon and Zaire. In South America values over 100 mIn per month are found predominantly in Brazil, extending northwards into Guyana and Surinam. In Central America values over 100 Dl1n per month occur in Guatemala, Honduras and Nicaragua and in North America they are limited to a region north of the Gulf of Mexico and to a narrow strip along the northwestern coast stretching into Canada.
Low annual precipitation .Precipitation rates under 10 IT.L..TD. per month occur in the Saharan region of Africa, Israel and countries of the Middle East, NIongolia and nearby regions of central Asia and in the northern regions of Greenland and Canada.
Seasonal variation Precipitation distribution during winter (December to February). Figure 5.4 shows the global distribution of precipitation during the northern winter of 1998. During this period there were precipitation values of over 100 mm per month along the western fringe of North America into Canada, a region to the north of the Gulf of !vIexico, a region of Central America, mainly in Brazil, southern Africa in the region of fvlozambique, Mada£ascar, Tanzania, Zambia and Angola. In Asia, high p;ecipitation values- occur in the Pacific Islands and in the northern . part of Australia.
101
m
~
()
C
~
.......
. .. .. .. . . . .. . .. ... .. . .. .. ...... ...... .. ...... . ...-............... ... . .. .. ., .. .............. . . . . .. . . . .. . . ............................. . .. . .................... . . . . . . . .. . .. . . . . . . .. . . . . . ·1········································· --'
o z o-n
..
•
..
"!j
•
•
•
•
•
•
•
..
..
•
•
..
.. . .
It
•••
.,
..
i
~~:::~; ~:~ ~~ ~mwm lj;m\ij:i{:":"""
(/)
m
o
3: m
Z
~
(/)
t km-2 yr- 1
240
100
50 10
,0
Arid regions
(a)
Figure 5.1.
't"
I
'
'
Global patterns of sediment yield: (a) after Strakhov (1967); (b) after Fournier (1960)
.
"
1
1
,-
)
,"
1
,., '
L--- '
' . L ,,--
'" L.--,- ,
'
AREAS SUITEDTO FLUSHING
103
m
< » ()
0
~
c
~ 0
z 0
"m
Vl
-0
3:
m
Z
-I Vl
,
~
Suspended sediment yield: t km-2 yr- 1 5
20
..
200 1000 5000
~
(a)
Figure 5.2. Global patterns of suspended sediment yield: (a) from Lvovich (1991) in Walling and Webb (1996); (b) frOln Walling and Webb (1983)
1--
l- - ·
~
l--.
~ .
1--
1
"I
"
~,
~ .~
xv
b
o
Sediment yield: t km-2 y r -1
1000
»;0 m »
750 500
Vl Vl
250 100 50
O
VI
Deserts and permanent ice (b)
),
~
I@ -;
0 I~ C
Vl
o
Vl
Figure 5.2_
continued
I
-
Z G)
m
-Nean ,monthl}!. precipitation JGPCpmonitoring) _ forthe.- yeor lJan -Dec) -1~98 -.nmm/month 90N..---------r------,-.-----,------....,..------r-o------.
~
()
C
~
o z o "m (/)
o
3:
m
Z
-I
30N - t - - - - -- - t - " I
(/)
EQ~--------~----~
3~~---------+----~--~
6~~--------~----------~----------r----------+--------~~-------~~
9~~--------~----------~----------~---------+----------~--------~ 1BO 12ml 60W (I GOE 1eo 1 Figure 5.3.
10
25
50
75
tOO
150
Annual precipitation/or 1998 in mm per month
1 : ~
OD
300
0400
800
800 1000
. .
.
.I
\
\
_..__ ..__ _.. _#_ . . __ _
~_. _
...._... . .. ~_"'
:..a ._~
__
~ ; ~ !a.'
. . ..;:J~ 5~ ~~
Ea~-----------~-----~~~
» m » :;;0
VI VI
C -l
9~~----------~----------~----------~~------------~----------~~--------~ 1fiD 12ml !)OW 0 tiOE 120£ 1SO
1
1Q
25
50
15
100
1M
200
300
400
m
o
-l
o r"C VI
I:
Figure 5.4.
Precipitation distribution during winter 1998 (December to February)
Z C)
1 EVACUATION OF SEDIMENTS
Precipitation distribution during spring (March to May). Figure 5.5 shows the global precipitation distribution during the nOl1hern spring of 1998. In spring, high precipitation values extend to roughly the same area with a gradual northwards movement of peak values. In Africa the high values are now more concentrated on the central-western coast around Cameroon and Gabon and values in Asia are higher in eastern China and Japan with the islands of the Pacific still experiencing high values. Precipitation distribution during summer Oune to August). Figure 5.6 shows the global precipitation distribution during the northern summer of 1998. This shows high precipitation values in America extend from-the northern regions of South America into Central America. In Africa, high values are founq in central regions and in the westenl zone from Guinea, Ivory Coast, Cameroon, Nigeria and into Zaire. In Asia, the monsoon period brings high amounts of rainfall to India, Tibet, China, the other eastern Asian countries and to the Pacific Islands. Precipitation distribution during autumn (September to November). Figure 5.7 shows the global precipitation distribution during the nOl1hem autumn of 1998, with regions of high precipitation now more isolated in America and confined to a narrow strip along the western coast, Central America and the northern region of South America. In Africa, high values occur in the western region around Cameroon and Gabon. In Asia, the highest values are again in the islands of the Pacific stretching into Japan, India, Vietnam and Cambodia.
The information shown in the four seasonal maps is available on the Internet. It would be of interest in the present context to derive a world map showing the degree of seasonal polarity of precipitation, as this may correlate with erosion rates and sediment yields. However, this exercise was not possible within this study.
5.2.3. Koppen classifIcation There have been many climatic classifications produced but one of the most frequently used is the Koppen classification, with eight climatic regions based on four temperature zones and one moisture zone and the seasonal domination of air masses. Details of a version of the Koppen classification by Pidwimy (1999) are given below and a discussion of alternative classifications is included in Appendix A4.3.
Tropical wet e e
Koppen classification Af. Maritime tropical air masses all year.
The climate has consistent high daily temperatures ranging from 20-30°C. Monthly temp~rature averages range from 24-30°C . . The annual range of monthly temperatures is about 3°C. Precipitation is uniform with a total over
108
1 1 1 1 1
1 I
J..
1 1 1 1 1 I J-
1
1
J
1
, 1
-I
Mean ~onthlyprecipnutron (GP9C monitoring) for :Dlpnng (Mor.Apr»M(]y) 191)6 Inmm/rnonth
[O-~-----------+---~'-
3(}S - i l - - - - - - - - - I - - - --
-+.i
» rn » ;0
Ul Ul
C -l
D~~----------~-----------F~---------+~--------~----------~~----------, 1 12[}11( 150'1/ 0 1 ZOE 1BO
m
o
-I
o ."
1
10
25
50
15
, 00
t flO
200
300
.4QO
600
800
100'D
r
C
Ul
:c Figllre 5_5_
Precipitation distribution during spring 1998 (March to JV/ay)
zC)
m
o
~ ()
Nean. ,monthl.r' Pfacipitation(GPC.C m ,' on,itoring,) for summer (Jun',Jul,Aug) 1998 In mm/montft
C
~
9{JN------...----~--.....--...------.......------..,._----..-;...,.....----_.,
o
z o -n Vl
m
OON
o
3: m
Z
-I
Vl
EC~----------~--~~
3~~----------r-------~~
6(}S-l-------+------+-------+------+--------t----~-____1
90~J.BO------12.j..rn-'I-----6+0.,,-t-----..f..D------a+oE------1+20-E------IllBO
1 Figure 5.6.
1
Precipitation distribution during summer 1998 (June to August)
L- ' " I..--. ' ' \-. " L,
' t_ " L- " L-."
1-- ' ' L- '
L-
' L-
' 1_
L- '
.1_
M~.' on month,~
precipitation (GPee monitoring)
for autumn lS"P.Oct"No·~r} 199B in mm/month 30N-v------~~'-t-----~-~--F'--,------r_-----....,..------
__-----""""lI
m· D - - - - - - - - I-----.;..,:-;+-I\I
3~) -~------------1----------~~),~~---------1--~~~~~~~-~-----------~
» m » ;::0
6liS ·u---·-----l·-----------::--+-~----_il_~,..,.,...;..~~,-_I---------II_--------_I
Vl Vl
..
90S -\-------+------""i-~----_Ii__------_I--1BO
12mI'
-
0
M't.'
tiOE
---__I------~
120E
1BO
C -l m Q -l
o
."
10
50
75
100
ffl
200
300
400
600
800
1000
r C
Vl
I
Figure 5.7.
Precipitation distribution during autumn 1998 (September to November)
Z C)
EVACUATION OF SEDIMENTS
Table 5.6.
Temp: °C
Colombia, 5 oN, elevation 65 Tn Jan.
Feb.
27
27
Mar. , Apr.
28
28
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Yr
27
27
27
27
27
27
27
27
27
[ -
Pptn: mm
554 / 519
557
620
655
655
572
574
561
563
563
512
, ,
6905
J 2000 mm. The region lies within the effects of the intertropical convergence zone all year. Convergence and high maritime humidities create cumulus clouds and thunderstorms regularly. A typical monthly distribution of temperature and rainfall is given in Table 5.6.
....
,
.-J
J
Tropical wet and dry • Koppen classification Aw, Am and BS. • Maritime tropical air masses during high sun season and continental tropical air masses during low sun season.
- I
!
-!
.._ \
The climate has distinct wet and dry periods. The seasonal pattern is due to the movement of the intertropical convergence zone. The wet season coincides with the high sun and the presence of the convergence zone. The dry season is due to more stable air associated with the presence of the subtropical high zone during the low sun season. During the rainy season the climate is similar to the tropical wet climate. During the dry season, semi-desert conditions prevail. Some regions experience intensification of rainfall due to monsoons and orographic uplift. A · typical monthly distribution of temperature and rainfall is given in Table 5.7.
J
J i
J
Tropical desert • Koppen classification BW. • Continental tropical air mass all year. This region is found near the tropics usually, but not always, on the western side of continents and covers 25% of all land area. It is characterised by: • low relative humidity (10-30%) and cloud cover • low frequency and amount of precipitation • high mean annual temperature
J J It
Table 5.7.
India, 13 oN, elevation 22 m Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Yr
Temp: °C
27
27
28
29
29
27
26
26
26
27.
27
27
27
Pptn: mm
5
2
9
40
233
982
1059
577
267
206
71
18
3467
J
J I
112
J:
J
AREAS SUITEOTO FLUSHING
Table 5.8.
}Vadi Haifa, Sudan, 22°1V, elevation 160117. Jan.
I Feb.
Mar.
Apr.
May
June
July 32
33
1
0
Temp: °C
15
17
21
26
31
32
Pptn: mm
0
0
0
0
, 1
0
[
f
Aug.
I Sept. I Oct.
l
Nov.
30
28
22
0
1
0
[
I Dec. I I
Yr
17
35
0
, 3
• high monthly temperatures ... . e high diurnal temperature rahg~s e high wind velocities. . . "'-.=-• .J
~,r': ~
.'
_
_" ......-
The region is influenced {~y "'upp~r ail<:~:$ta.bility and subsidence owing to the presence of the subtropicalhfgh pressure:~'zone . Temperatures are highly variable daily and annually. vVith theayerage monthly temperatures ranging fronl 29-35°C and the average diurnal rangejs :}Jetween 14-25°C. A typical monthly distribution of temperature and rainfall is tsiven in Table 5.8.
Mid-latitude wet "Koppen classification Cf a#d~f.~: . .;, :."}~.7:'· :' NIaritime tropical in sUII1IIl~r:~'d:maritime p~Iru
o
In the Northern HeInisphere the ~regi6n is from -60oN to between 25 and 30 0 N mainly on the western side of continents. In the Soutt1.em Hemisphere the climate.;,· spans from the south-eastern tip of South America, New Zealand and the south-~ east coast of Australia. Summer is dominated by thunderstorms produced by daily: heating. Monthly average temperatures range from 21-26°C. Frontal weather associated with the mid-latitude cyclone dominates the climate of more polar areas and is more frequent in all regions in winter. .c; .. . . . . Precipitation is fairly evenly distributedthr~ughog"t~~~!JJ~ year with .variable annual totals depending on the latitude and the coiltitIentalposition of the regions. A typical montpjy distributi6n. 'o(~~mperature ·am(r9:lrifall is given in Table 5.9. IY1jd-latitude winter dry . o
o
Koppen classification C\vand Dw. Maritime tropical air iflasse's ·in summer and continental polar air masses in winter.
This region is charactelised by a strong seasonal pattern in temperature and precipitation. The region is located in the interior of the continents in the midEngland, 51 ·5°N, elevation 5 m
Table 5.9.
I Jan . Temp : cC Pptn: rnm
I [
Feb.
4
4
54
40
I Mar. I APr.' , ,
7
9
' 37
38
I I
May
I June
12
16
46
46
I July i
I
18
56
f
I I I I
Aug.
17
59
I Sept. I oct./ Nov. I Dec. I Yr I ,
15 50
I I
11
,
7
I
I
5
10
I
48
, 5"95
!
57
I
64
113
EVACUATION OF SEDIMENTS
Table 5.10. 1
Calgary, Canada, 5JON, elevation Jan.
I
Mar.
Feb.
Apr.
i
Temp: °C
i
-10
-9
-4
4
17 I 20
, 26
35
i
!
i I
Pptn: mm
I
I
i i :
i
i
I
10
52
July i Aug.
June
May
I
32J m Sept.
Oct.
Nov.
Dec.
Yr
-7
4
15
444
I
13
17
I
15
11
5
-2
88
58
! 59
35
23
16
I
latitudes. The continental location causes a large annual temperature range. Summers are hot and humid with intense summer convectional storms. Continental polar air masses are associated with cold, dry! weather in the winter. A typical monthly distribution of temperature and rainfall lis given in Table 5.10.
Mid-latitude summer dry (Mediterranean climate) . • Koppen classification Cs and Ds. i • Sunlmer dominated by continental tropical alr, winter dominated by maritime polar air masses. Found on the western sides of continents between 30 and 40 o N. Precipitation falls mainly in the winter due to the mid-latitude cyclone. During the summer these areas are influenced by stable subtropical highsi, producing dry, warm weather. A typical monthly distribution of temperature and rinfall is given in Table 5.11.
Polar wet and dry
.
• Koppen classification ET. I • Maritime polar in summer and continental polar or Arctic in winter. i
I
Cold winters, cool summers with a summer rainfall regime. The areas experiencing this climate are the North American Arctic coast, Iceland, coastal Greenland, the Arctic coast of Europe and Asia and the Sou~hern Hemisphere islands. Annual precipitation is less than 250 mm with precipit4tion during the summer. A typical monthly distribution of temperature andrainfal~ is given in Table 5.12.
J
Polar desert • Koppen classification EE I • Continental Arctic and continental polar air masses. These regions occur in continental areas of thd high-latitudes, such as Greenland and Antarctica. No solar radiation is received for about half the year while during I
Table 5.11.
Italy, 42°N, elevation 131 m
I Jan. I I
Temp: °C
8
I Julyl
Feb.
i
Mar.
Apr.
May
8
I
10
13
17
22 I 24
88 I 77
72
63
48
June
I
Aug.
I Sept. Ii Oct. i
Nov.
Dec.
Yr
24 , 21
16
12
9
15
22
128
116
106
881
I
Pptn: mm
76
!
114
I
14
i
70
J
I
AREAS SUlTEDTO FLUSHING
Table 5.12.
Greenland, 81·5°N, elevation 35 m
Temp: °C Pptn: mm
I
I Oct.
Nov., Dec.
Yr
-8
1-
19
-24
I -26
-16
21
I
16
35
I
204
I Feb. I Mar.
Apr.
May
June
July
Aug.
Sept.
-30 /-30
-33
-23
-11
0
4
2
I
8
5
3
5
12
19
Jan.
23
20
I
I
37
the summer insolation is high with long days, however the albedo of the snow surfaces reflects up to 90% back. Average monthly temperatures are generally below O°C. A typical monthly distribution of temperature and rainfall is given in Table 5.13.
. 5.2.4. Relationship between climate zone and erosion rates A map of climatic regions based on Koppen is presented in Figure 5.8. Jansson (1988) took sediment yield data from 1358 drainage basins and cOlTelated sediment yields with the Koppen climate classifications. For each Koppen classification the values of sediment yield have high standard . deviations due to the strong influence of a few extremely high yields.
• In the cold steppe climate (BSk) the range in value of annual sediment yield is betvveen 1 and 16 300 t/km2 • Only 13 out of the 75 rivers have yields of more than 600 tJkm2; these are mainly in China and Argentina with one in South Africa and one in ~he USA. • The two warm temperate humid climates with no dry period (Cfa and Cfb) have high standard deviations. For the Cfa climate, Taiwan and Italy are responsible for the high standard deviations. The ten rivers with the highest sediment yield are in Taiwan with a range of 2605-18 339 tJkm2 while 106 of the rivers have yields of less than 300 tlkm2 • In the Cfb climate there are 144 rivers with less than 100 tf1lffi12 and 38 rivers (23 in New Zealand) which yield more than 500 t/km2, • The cool Mediterranean climate (Csb) has limited data with a wide variation 2 in yields. Four rivers in Australia have yields of less than 2 t/km and four rivers in the USA yield more than 1000 t/km2• ., The boreal climate without a dry period (Dfb and Dfc) has a low sediment 2 yield. The median value for Dfb is 33 tJkm2 and the mean is 104 tlkrn . There
I.
Table 5.13.
F
Antarctica, 66,5 oS, elevation 30 m
I Jan. Temp:
°c
Pptn: mm
[ Feb., Mar.
I APr.
, May
I June i July i Aug.
I -2 I -5 1-10 1-14 I -16 1-16 1-17 I. 13 I 19 I 51 ! 44 ! 92 i 67 I 77
Sept.
I Oct. I Nov. I Dec.,
Yr
-17
-17
I
-14
1
-7
1
-3
1-12
I 95
52
I
43
I
46
I
26
I 625
115
EVACUATION OF SEDIMENTS
~----------~~~--------------------------------~~~~
Figure 5.B.
A Tropical rain climates f no dry period w dry winter s dry summer m monsoon rains, dry period
C Warm temperate rain climates a warmest month >22 b at least 4 months >10 c 1-4 months> 10 d coldest month below -38
B Dry climates s steppe w desert h hot, mean temp. >18 k cold, mean temp. <18
D Boreal climates
"
~
E Snow climates T tundra F always frost
Climates of the world according to the Koppen classification
are eleven mountainous rivers in Romania and two Soviet rivers in the Caucasus area with more than 500 t/km2• The median value of Dfc climate is 9 t/km2 with a mean of 76 t1km2 • The data are highly variable, with a standard deviation of 3·5 times the mean. • The boreal climate with dry winter (Dwb) has too few data points to give reliable values. The snow climate ET consists of two populations - rivers in the fonner Soviet Union with up to 50 t/km2 and rivers in Alaska and Iceland with yields of between 375 and 4000 t/km2 (Jansson, 1988). The data from the climatic regions were amalgamated by Jansson (1988) into simplified homogeneous climatic groups, depending on their characteristics. The groups are set out in Table 5.14. The classification of sediment yield for each climatic group shows a number of distinct features which can be seen in Figure 5.9. Over half the rivers in the Af group have sediment yields greater than 1000 t/km2, with only one river exhibiting a value less than 100 t/km2 • The groups Cwa and Cs have the next largest percentages in the highest sediment yield class, with more than 50% of the basins having sediment yields above 100 tIkln2• Dfa-d and Dwc have 67% and 100% of the rivers respectively with sediment yields less than 100 t/km2 • Cfb
116
AREAS SUITEDTO FLUSHING
Table 5.14. Reasons for com,bining climates into homogeneous climatic groups . (Jansson, 1988) Climates
I Af,Cf
Climatic group
Reasons for grouping
Af
Cf is found in tropical mountains. Similar precipitation conditions as Af
AW,Am, Cw, outside Argentina
Aw
Cw is found in tropical mountains. Cw (Argentina excluded) has a median yield of 246 rll
BSh,BSk
BS
Similar rain conditions. Not many low sediment yield values in BSh. Many low and many high values in BSk
BW
Similar rain and sediment yield conditions
I I I BWh, BWk Cfa
I
I
-.
I
Few low values
Cfa
cefb in contrast has many low values)
Ctb, Cfc
Ctb
Df
Df
Csa, Csb, Cs
Cs
As rain and erosion is in winter, the temperature of the warmest month, whether Csa or Csb, is of no significance. Extremely low values in Australia both of Csa and Csb
Dsb, Ds
Ds
Dsb and Ds are found in mountainous areas in Csa or Csb but have more snowmelt run-off than Csa or Csb
l
I
Cfc similar climate as Cfu Mountains in Cfa or CfD but Df climates have more snowmelt erosion than Cfa or etb
.-
Cwa, Cw in Argentina
Dfa, Dfo, Dfe, Dfd
I
Cwa
Cw in Argentina has similar rain conditions as Cwa in Argentina. Cwb not included because rain in warm month may be of importance and few data from only one country
Dfa-d
All have snowmelt run-off in spring when vegetation cover is sparse. Temperature similar to snowmelt erosion
Dwe
Similar snO\vmelt and rainfall conditions. Dwa and Dwb hU"'e different rainfall and snowmelt conditions from Dwc and Dwd
I
D we. D\vd
ET, ET (Mt)
__::"._0
I
I I
I
ET
Similar climate
has about 450/0 within the two lowest classes. Low values are found in Central and Western Europe, with high values tn the Southern Hemisphere and in southern-most Europe. Generally, a tropical Af climate has high sediment yields while Aw has more variable values. In arid climates BS has high values, while the desert group BW has low and intermediate values. There are few rivers with yields less than 10 tf1.i.
117
EVACUATION OF SEDIMENTS
Sediment yield classes: Vkm 2
c:::J
0-10
CJ
11-50 51-100
h',~; :J,hl
101-500 Snow
501-1000 _
>1000
Boreal
1 r--'-
Warm temperate
:
,--
At
Tropical and arid ---'-
Figure 5.9.
Number of basins within sediment yield classes in climatic groups
rates to be made for each country based on climatic classification. Table 5.15 lists the climate zones in each country and indicates the number of rivers that occur in each zone. A similar exercise has been carried out in this study for a more comprehensive list of countries in the world, the results of which are shown in Appendix A4.4. The updated map of climatic zones produced by Alexandersson (1982) based on the Koppen classification, was used to determine the climate classes that occur in each country and the rates of sediment yield were based on the map produced by Walling and Webb (1983) (see Appendix A4.4). There is a reasonable similarity between the data listed in Table 5.15 and the results we have obtained using the world climate map. Differences arise mainly because the data in Table 5.15 are based· on a number of river basins in each
118
AREAS SUITEDTO FLUSHING
Table 5.15. . Countries classified into climatic zones showing nwnber of river basins in each zone (modijiedfrom Jansson, 1988) Climate zone
Country (nLlmber of rivers/zone)
Albania (312) Cfb/Csa Algeria (27/1) CsafBSk Argentina (22/13/5/5/2) BSkiCwafCw/ET (Mt)IBSh Australia (4/3/1/1/1/1) CsbfBSh/Csa/CfbIBWhiAm Austria (4/3) Df/Cfb Bolivia (3) Cwa Brazil (12) Aw Bulgaria (24/6/6) Cfb/CfafCsa Cameroon (311) AwlBSh Canada (49/18/2) DfblDfclBSk Chad (4/1) -- ~ AwlBSh Chile (2112) ET (Mt)/Cs China (1015/4/3/1/1) BSkJCfaIDwa/Dwb/Cwa/BWk Colombia (1512) Cf/Aw Costa Rica (10/6/1) Awl Af/Cf Cuba (1) Aw Czechoslovakia (7/5) Cfb/Dfb Denmark (2) Cfb Ecuador (1012) Cw/Aw El Salvador (2) Aw Finland (16/3) DfclDfo France (4) Cfb Germany (55) Cfo Great Britain (9) Ctb Greece (414/1) Csa/Cfb/Cfa Haiti (1) Aw Honduras (1) Cw Hungary (6/2) Cfb/Cfa Iceland (8/1) ET/Cfc India (5/5/2/1/111) CwalAwfBSfET (Mt)/CwfBSh Iran (6/2/1) Csa/BSk./BWh Iraq (1) BWh Israel (2/1) CsaIBSh Italy (26114/811) CfalCs;?}CfbIDf Java (6) Af Kenya (611) CwaIBSh Lesotho (9/5) Df/Cfn Madagascar (1) BSh Malaysia (3) Af Morocco (14/2) CsaIBSk Nepal (1) Cwa New Zealand (42) Cfb BWh Niger (1) Nigeria (11n) BShlAw Pakistan (1) Ds Panama (2) Af Papua New Guinea (3/2) Af/Cf Peru (2) Cw Philippines (5) Am Poland (53/28) Dfb/Cfb Romania (50/412) Dfb/Ctb/Cfa South Africa (17/8/7/4/2/1/111) I' BSklCwb/CtbfBWklBWhlCsb/CfaJBSh Soviet .(55/351221 IS!) 41 121 10/9/5/4/3/2/111) , DfblDfclD wcfDflDs/BSklETICfaJDwblDwdIDfaJCslDfdlCtb
I
I
119
·
EVACUATION OF SEDIMENTS
Table 5.15.
continued
Country (number of riverslzone)
Climate zone
Spain (18/3) Sweden (14/3/2) Switzerland (9/5) Taiwan (16111) Tanzania (1) Thailand (25/13111) Trinidad (1) Tunisia (6) Turkey (1) USA (83116/14/12/916/4/212) USA-Alaska (7/5) Venezuela (4/312) Yugoslavia (28/311) Zimbabwe (111)
CsaJCfb
- 1
Dfc/CfblDfb Df/Cfb
CwaJCfa Bsh Aw/CwaJAm
i
I
1
BS Csa Ds CfalBSk/CsblDfblDfaJDflDsb/CfblB sh
DfclET Cf/AwIBsh Cfb/CfaJCs
BShlCwa
f
I
1 I .....I-
country rather than on a global assessment of which climatic zone the country fits
In. This study emphasises the diversity of climate classes and rates of sediment yield within many individual countries and shows that this base unit will often not be small enough to produce homogeneous conditions. On a global scale, a compromise must be reached between generalisation and accuracy and the country provides a manageable unit for some purposes.
1 1 ,
1 i
5.3.
GEOGRAPHICAL AREAS SUITABLE FOR FLUSHING
5.3.1. Introduction There are a number of factors common to areas suitable for the application of reservoir flushing techniques. A first prerequisite is that there must be ,a high to medium erosion rate within the catchment. Secondly, the sediment must then be transported down the river system to the reservoir resulting in the requirement for its removal for flushing. These two prerequisites are discussed in the first two sections of this chapter. The hydrological characteristics required for successful flushing are then considered.
5.3.2. Factors affecting erosion rates Factors causing high erosion rates have been outlined earlier in this book and the main aspects are summarised below.
120
~
AREAS SUITEDTO FLUSHINC;
Precipitation High rates of erosion occur in regions where there is high intensity of rainfall. It is not just high precipitation totals that result in high erosion rates but it is the relationship between precipitation and vegetation. Global relationships between erosion rates and precipitation show variable results.
Geology The geology is an important factor deteITPjning the susceptibility of the rock to the effects of erosive forces. Erosion rates are generally highest in areas of soft sedimentary rocks.
Soils The key characteristics of a soil that influence erosion rates are texture, structure, organic matter content, shear strength and infiltration capacity. High erosion rates occur where the texture of the soil is high in silt and fine sand and low in clay, and where the structure is compacted and the organic matter, shear strength and infiltration rates are low. All these factors cause high run-off rates, leading to erosion of the soil. Slope The gradient and length of the soil sUlface influence the velocity and direction of run-off and therefore its erosivity. High erosion rates occur where there are long, steep slopes resulting in movement of water downslope at a high velocity.
Drainage basin area There is generally an inverse relationsrLip between sediment yield per unit area and catchment area. Higher rates occur in small drainage basins due to a higher overall slope, higher percentage of erodible material a.lld less opportunity for eroded material to be deposited further down the catchment.
Vegetationlland use Vegetation depends on the interaction of a number of factors including rainfall, temperature, soils and topography. The presence of a vegetation cover reduces the erosive power of rainfall by dissipating its energy, increasing infiltration, reducing ·the velocity of nln-off and by holding soil particles together. High erosion rates therefore occur where there is sparse vegetation cover either due to natural climatic conditions or due to land-use practices.
Human impact Activities such as deforestation, urbanisation and agriculture all affect the erodibility of the soil. Current erosion rates are more than two and a half times the historic, mainly as a result of human influences.
121
EVACUATION OF SEDIMENTS
5.3.3. Sediment delivery ratio Only a proportion of the sediment that is eroded will be transported down the catchment to be deposited in a reservoir. The efficiency of the transport process is expressed by the sediment delivery ratio, SDR, which is the proportion of sediment eroded from the land that is discharged into rivers (Morgan and Davidson, 1986). This measure is required to convert the estimated soil erosion within a basin into a value of sediment yield. Values vary from about 3 to 90%, decreasing with greater basin area and lower average slope (Morgan and Davidson, 1986). There are a number of factors that affect the sediment delivery ratio from a basin. The size of the drainage basin has an influence on the sediment delivery ratio with more opportunity for deposition and lower overall slopes in larger drainage basins leading to lower ratios. The following factors also influence the sediment delivery ratio.
Erosion processes The sediment delivery ratio is generally higher for sediment derived from channel-type erosion which delivers sediment to the main channels of the transport system more quickly and more directly than from sheet erosion. Distance from basin outlet Channel networks with a high drainage density are more efficient for transporting sediment than basins that have a low channel density, meandering low gradient channels, or those clogged with debris. Soil and vegetation Finer particles are transported more easily than coarse particles, therefore higher delivery ratios occur for soils with fine grained erosion particles. However, silts tend to be more erosive and produce higher delivery ratios than clays. DepOSitional features The presence of a depositional area decreases the sediment delivery ratio. Most of the sediment eroded from the steep uplands of basins may be redeposited at the base of slopes.
Catchment size and slope Large, gently sloping catchments will have lower delivery ratios than smaller and steeper catchments. The poor correlation between sediment yield and erosion rates makes it difficult to estimate the sediment load entering a reservoir on the basis of the erosion rate within the· catchment (Mon"is and Fan, 1997). Most studies that have attempted to relate the delivery ratio to catchment characteristics have used an inverse relationship with catchment area (Walling and Webb, 1983).
122
AREAS SU ITEDTO FLUSHING
5.3.4. Hydrological characteristics In addition to the factors causing a high rate of sediment inflow, discussed above, there are some specific hydrological characteristics of the catchment above a reservoir site that are required for successful flushing. Experience has shown that low reservoir water levels provide the most effective conditions for sediment flushing. To allow water levels to be lowered requires confidence that rainfall can be relied upon to refill the reservoir. It follows that well defined wet and dry seasons will be favourable for a sediment flushing regime. Such a climate is referred to in the Koppen classification as 'tropical wet and dry'. River discharges must also be sufficient to transport sediment loads through the reservoir. Regions of low precipitation, like the Sahara and other desert environments, will therefore not be suitable for flushing even if they exhibit a defined seasonal effect. The availability of water will also affect the duration and discharge rate of the flow required for flushing. As stated in Chapter 3, where there is a limited amount of water it is better to use a high discharge for a short period of time than a low discharge for a long period of time. This increases the amount of sediment that is removed. Geographical regions suitable for successful flushing must, therefore, provide a large enough annual run-off compared with the volume of a reservoir to allow use of a sufficient proportion of the water for flushing.
5.3.5. Areas of the world which are best suited to reservoir flushing It is not possible to define precisely which specific areas of the world will provide conditions for 'successful' flushing. In reality there is a spectrum of conditions ranging from tllOse sites where conditions are ideal to those sites which are quite unsuited to sediment flushing. From the hydrological a11d hydraulic conditions neCeSSfuy' for successful reservoir flushing the luost likely locations in which to use this technique are those which are within the Koppen climate classification tropical wet and dry: classifications Aw, Am and BS. Also, there are areas in the mid-latitudes where spring snowmelt provides a regular and predictable annual pattern of high flows. From the Koppen classification of climatic zones and the mid-latitude spring snowmelt effect, the requirements for successful flushing are most likely to be met in the following locations: parts of Central America extending into South America o areas in North and South America where the rivers are fed by the Rockies and the Andes Q parts of Central Africa from the Ivory Coast in the west to Sudan in the east 19 areas in Central Asia where the rivers are fed by the Himalayas, including Pakistan, India and Nepal " parts of Asia including Cambodia, V-ietnam and Thailand. o
123
Si e
In
.m:i
pecific
sti •
d sIgn
e
I
n
an era ons •
: .' ,
Site-specific investigations and design considerations Chapters 2 to 5 inclusive are concerned with general issues relating to the flushing of sediment from reservoirs. Chapter 2 assesses the scale of the problem of reservoir sedimentation. It assesses the volume of storage that is likely to be lost to sedimentation and compares this volume with the net volume of storage that is likely to be required to meet continuing demand. Chapter 3 provides a review of the current state of knowledge of reservoir flushing, and Chapter 4 considers the worldwide experience of flushing. Chapter 5 identifies areas of the world where flushing is likely to be most useful. Thus far it is possible to identify locations where sediment flushing is likely to be useful. However, there are many detailed factors which need to be evaluated on a site-specific basis before the technical viability and economic soundness of sediment flushing C8..s.'1 be confirmed. This final chapter provides details of the nature of these site-specific" investigations, including design considerations for the sediment bypass itself. There are numerous stages for such investigations as follows.
Site investigations Flushing outlets have to be able to withstand high velocityfiows with high concentrations of sediment. Such flows are pJghly abrasive and expensive steel lining will normally be required to avoid undue damage to the structures. Hence it is important that the site allows for the construction DT relatively compact flushing facilities, either orifices within the dam itself or relatively short tunnels or channels. Energy dissipation works will normally be required at the downstream side and it is an advantage if these facilities can be shared with other outlets such as high head spillways or irrigation outlets. It is advantageous if the flushing facilities discharge to the downstream channel well away from any power station outlets as any local deposition of sediments will increase tailwater levels and reduce power output. The reservoir itself requires a detailed survey to establish its topography. This is required to check whether the reservoir basin is a suitable shape for sediment flushing and also to provide input data for detailed modelling of the sedimentation process within the reservoir.
Hydrological investigations It has been stated that there are certain requirements for successful sediment fiushin2: which are related to the amount of water available and its reliability year on yea"r and season by season. Hence inflows to the reservoir need to be
127
I
-f
!
EVACUATION OF SEDIMENTS
-!
established with confidence. This involves the acquisition of historical records of river flows going back at least 30 years and preferably longer. Records of river flows can often be extended further back in time by considering local rainfall records, which often go back 100 years or more, and by undeltaking catchment modelling to convert rainfall into run-off. The ideal situation for sediment flushing is an annual inflow of water of at least three times the volume of the reservoir (original volume in the case of existing reservoirs) and an annual hydro graph which shows distinct wet and dry seasons. Sediment investigations The amount and nature of the sediment entering or likely to enter the reservoir needs to be established. This requires measurements over many years to establish the "results with the confidence that is required. There are variolls approaches to this task. Most commonly, sediment transport is measured at a gauging stavon not too far upstream of the reservoir and a relationship between flow rate and sediment transpolt rate is established. The long hydrological record is then used to compute the total amount of sediment passing the gauging station by integrating over the period of the record. There are some dangers in doing this because there is no unique relationship between flow rate and sediment transport rate for fine sediments, the quantities of sediment being determined by the amount being washed off the catchment not the capability of the river to transport them. Bed load is difficult to measure and is often estimated as 10% of the total sediment load. An alternative approach is to calculate the bed load using established predictive techniques. . In the case of existing reservoirs, information about the amount of sediments entering the reservoir can be augmented by surveys of the amount and nature of the material settling within the reservoir. Care is required, however, to allow for the amount of material, mainly fine, which passes through the reservoir without deposition. Bed material sampling should be 'undertaken in the reservoir and in the rivers which feed the reservoir. A sound knowledge of the nature of these sediments, including their size, specific gravity and degree of compaction, is an essential requirement to provide inputs for numerical models which simulate sediment movement. Hydraulic modelling Sophisticated numerical (computer) modelling of the way sediment is likely to behave within the reservoir and the amount and nature of the sediment that will be passed to the downstream reach is the cornerstone of any detailed evaluation of flushing facilities. One-dimensional models with quasi two-dimensional simulation of the incised channel that develops during sediment flushing are the most appropriate tools. These models are computationally efficient and are capable of making long-term simulations, decades rather than hours or days. They have reached reliability levels which permit thenl to be used 'cold' when
128
~-
-/
j
:4 J
J
1
SITE-SPECIFIC INVESTIGATIONS
new reservoirs are being investigated. When used on existing reservoirs they have the added benefit of measured sedimentation data for verification purposes. Computer simulations of reservoirs ideally use representative, long-term · sequences of water and sediment inflows to the reservoir. The models are capable of looking at the effectiveness of various aspects which affect reservoir sustainability over periods of up to 50 or 60 years, including: • measures to reduce the amount of sediments entering reservoirs such as catchment conservation or upstream storage • measures to manage the sedimentation process within reservoirs such as variations in the operating rule curves for the reservoir measures to evacuate sediment from the reservoir including dredging and sediment flushing.
Q)
System simulation modelling System simulation modelling is required to evaluate the conflicting demands of hydropower production, irrigation and other requirements, and must be able to assess the impacts of the various reservoir operating strategies. The simulation model must be able to replicate the outputs of water and power under a range of operating strategies so that an optimal economic and technical solution may be identified. In addition, it must be possible to take account of the effects of other reservoirs upstream and downstream of the one under consideration. Economic and financial analysis The main aim of economic and financial analyses is to assist in the identification and selection of the most favourable sediment management option. For each option the most important factor~ from the economic viewpoint, is to define the 'with' and 'without' project cases. These will illustrate the net economic impact of t1.e availability of water resources over time, including any seasonal variations. Evaluation of the impact of alternative investment phasing is also important. The greatest challenge in the evaluation of projects \vhich promote sustainability of reservoirs is to assign realistic values to the benefits of extending reservoir life. This is beyond the scope of this study. Work, however, is progressing in this direction (Palrnieri, 1998).
129
7. References Ackers, P. (1993). Sediment transport in open channels: Ackers and \Vhite update. Proceedings of the Institution of Civil Engineers Water, Maritime and Energy, 101, 247-249. Albertson, M. L., Malinas, A. and Hotchkiss, R. (eds) (1996). Proceedings of the International Conference on Reservoir Sedimenation, Fort Collins, Colorado, USA. Anderson, H. W. (1975). Sedimentation and turbidity hazards in wildlands. In: Watershed Nfanagement. Proceedings of a Symposium of the Irrigation and Drainage Division. ASCE, New York, pp. 347-376. Atkinson. (1984). Consolidation of reservoir.peposits. In: Sedimentation in reservoirs in the Tana River basin, Kenya, R. Wooldridge (ed.). Report OD 61, HR Wallingford, UK. Atkinson, E. (1996). Feasibility of flushing sediment from reservoirs. Report OD 137, HR Wallingford, UK. Atkinson, E. (1998). Reservoir operation to control sedimentation: techniques for assessment. Proceedings of the Conference of the British Dams Society, Bangor. Attewill, L. J. S., \vrute, W. R., Tariq, S. M. and Bilgi, A. (1998). Sediment man(!gement shldies of Tarbela Dam, Pa..1dstan. Proceedings of the Conference of the British Dam Society, Bangor. Basson, G. R. and Olesen, W. (1997). Modelling flood flushing. International Water Power and Dam Construction, 49, No.6. Basson, G. R. and Rooseboom, A. (1997a). Dealing with reservoirsedinuntation: guidelines and case studies. Bulletin 115, ICOLD, Paris. Basson, G. R. and Rooseboom, A. (1997b). Dealing with rese:rvoir sedimentation. "Vater Research Commission Report No. TT91197, Pretoria. Bhargava, D. N., Narain, L., Tiagi, S. S. and Gupta, P. P. (l987).Sediirientation problems at low dams in the flimalayas. Water Power and Dam. Construction, Jan., 30-33. Binnie & Partners. (1986). Report on Guanting reservoir - sedimentation. Report for Government of PRC (Beijing Municipal Engineering Administration Division), under assignment by ODA. Bordas, M. P. and Canali, G. E. (1980). The influence of land use and topography on the hydrological and sedimentological behaviour of basins in the basaltic r~gion of South Brazil. In: The influence of man on the hydrological regime with special reference to representative and experimental basins. International Association of Hydrological Sciences, Publication No. 130, pp. 55-60. Branski, J. (1975). Oeena denudacji dorzecza Wisley na podstawie wynikow pomiarow rumowiska unoszonego. Prace Instytuto Nfeteorologii i Gospodarki ¥Iodnej, 6, 1-58. efficiency of reservoirs. Transactions of the A,m.erican Geophysical Bnme, G. 11. (1953). Union, No.3, Washington, DC, 407-418. Budyko, !vi. 1. (1956). Heat balance of the Earth's surface (in Russian). Giclrvn1eteoizdat, Leningrad, 255 pp.
133
EVACUATION OF SEDIMENTS
Chen, J. L. and Zhao, K. Y. (1992). Sediment management in Nanqin reservoir. International Journal of Sediment Research, 7, No 3. Comite Francrais des Grands BalTages (CFGB) (1973 and 1982). EDF-CNEH, Savoie Technolac, 73373 Le Bourget-Du-Lac, CEDEX Chang, M., Roth, F. A. and Hunt, E. V. (1982). Sediment production under various forest-site conditions. In: Recent developments in the explanation and prediction of erosion and sediment yield, D. E. Walling (ed.). International Association of Hydrological Sciences, Publication No. 137, pp. 13-22. Chaudhry, M. R. (1982). Flushing operations of Warsak reservoir sediment. Proceedings of the Pakistan Engineering Congress, 58, Paper No. 454,Lahore. Church, M. A. and Slaymaker, H. O. (1989). Disequiliblium of Holocene sediment yield in glaciated British Columbia. Nature, 337, 452---454. Cyberski. (1973). Accumulation of debris in water storage reservoirs of central Europe. Dawans, P. H., Charpie, 1., Giezendanner, W. and Rufwenacht, H. P. (1982). Le Degravement de la Retenue de Gebidem. 14th Congress on Large Dams, ICOLD, Rio de Janeiro, 1982. Dedkov, A. P. and Moszherin, V. 1. (1992). Erosion and sediment yield in mountain regions of the world. In: Erosion, debris flows and environment in mountain regions, Proceedings of the Chengdu Symposium, July 1992, IAHS Publication No. 209, pp. 29-36. Dendy, F. E. and Bolton, G. C. (1976). Sediment yield runoff drainage area relationships in the US. Journal of Soil and Water Conservation, 31, No.6, 264-266. Douglas,!. (1973). Rates of denudation in selected small catchments in Eastern Australia. University of Hull, Occasional Papers in Geography, No. 21, 127 pp. Dunne, T. (1979). Sediment yield and land use in tropical catchments. Journal of Hydrology, No. 42,281-300. Dunne, Dietrich and Brunegngo. (1979). Rapid evaluation of soil erosion and soil lifespan in grazing lands of Kenya, hydrology of areas of low precipitation. IAHS publication 128, proceedings. of conference, Canberra, Dec. El Faith Saad, A. (1980). Sedimentation and flushing operations of Roserires, Sennar and Khashm El Girba reservoirs. International Seminar of Experts on Reservoir Desiltation, Communication No.2. Tunis, 1980. El Hag Tayeb (1980). The limited experience of desilting in Sudan. International Seminar of Experts on Reservoir Desiltation. Com. 14. Tunis. Fournier, E (1960). Climate and erosion. P. U. E, Paris, 201 pp. Fredriksen, R. L. (1970). Erosion and sedimentation following road construction and timber harvest on unstable soils in three small Western Oregon watersheds. US Forest Service Research Paper. PNW 104, 15 pp. Gleich, P. H. (ed.). (1993). Water in crisis. Oxford University Press, New York. Glymph, M. (1973). Summary: sedimentation of reservoirs. Gogus, M. and Yener. (1997). Estimation of sediment yield rates of reservoirs in Turkey. ICOLD 19th Congress, Florence, Vol. 3, Question 74, Reply 78. Goldman, S. J., Jackson, K. and Bursztynsky, T. A. (1986). Erosion and sediment control handbook. McGraw-Hill, New York. Goldsmith, E. and Hildyard, N. (eds). (1984). Environmental and social effects of large dams. Vols. 1-3, Wadebridge Ecological Centre, UK.
134
, REFERENCES
Gregory, K. J. and vValling, D. E. (1973). Drainage basin form and plVcess. Edward Arnold, London Griffiths, G. A. (1981). Some suspended sediment yields from South Island catchments, New Zealand. Water Resources Bulletin, 17, 662-67l. Griffiths, G. A. (1982). Spacial and temporal variability in suspended sediment yields of North Island basins, New Zealand. Water Resources Bulletin, 18, 575-584. Hardjowitjitro, H. (1981). Soil erosion as a result of upland traditional cultivation in Java Island. In: Proceedings of South-East Asia Regional Symposium on Problems of Soil Erosion and Sedimentation, T. Tingsanchali and H. Eggers (eds). Asian Institute of Technology, Bangkok, pp.173-179. Hwang, J. S. (1985). The study and planning of reservoir desilting in Taiwan. International, 10, 7-13.
~Vater
International Commission on Large Dams (ICOLD). (1988). World register of dams: updating. ICOLD, Pa..ris. International Commission on Large Dams (ICOLD). (1997). General report on question 74. ICOLD 19th Congress, Florence, Vol. 3. International Commission on Large Dams (ICOLD). (1998). World register of large dams. CD ROM. In[ernational Research and Training Centre on Erosion and ~edimentation (IRTCES) (1985). Lecture notes of the training course on reservoir sedimentation. IRTCES, Tsinghua" University, Beijing, Chh"la. Institute of vVater Conservacy and Hydroelectric Power Research (IWHR). (1983). The rational operation of drawdownfiushing in Shuicaozi reservoir on the flU River. Summary in English of Paper SR8301, IWHR. Jaeggi, A. L. and Kashyap, B. R. (1984). Desi1ting of Baira reservoir of Baira Siul project. Irrigation and Power, Oct., 375-380. Jansson, M. B. (1988). A global survey of sediment yield. Geographical Annals, No. 70, Series A (1-2), 81-98. Jansson, M. B. (1992). Suspended sediment outflow from Cach! reservoir during the flushing in 1990. In: Sedimentological Studies in the Cach{ reservoir, Costa Rica, IVI. B. Jansson and A. Rodriguez (eds). UNGI Report No.8l. Department of Physical Geography, Uppsala University, Sweden. Jantawat, S. (1985). An overview of soil erosion and sedim,entation in Thailand. In: Soil erosion and conservation, S. A. El-Swaify, W. C. Moldenhauer and A. Lo(eds). Soil Conservation Society of America, Ankeny, Iowa. pp. 10-14 Jarecki, E. A. and Murphy; T. D. (1963). Sediment withdrawal investigation - Guen'lsey reservoil: Paper No. 91, Miscellaneous Publication 970, US Department of Agriculture. Jowett, I. (1984). Sedimentation in New Zealand hydroelectric schemes. Water International, 9, 172-176. Jolly, J. P. (1982). A proposed method for accurately calculating sediment yields from reservoir deposition volume. In: Recent developments in the explanation and prediction of erosion and sediment yield (Proceedings from the Exeter symposium, July 1982), IAHS Publication No. 137. Jowett, I. (1984). Sedimentation in New Zealand hydroelectric schemes. 172-176,
~Vater
International, 9,
135
EVACUATION OF SEDIMENTS
Judson, S. and Ritter, D. F. (1964). Rates of regional denudation in the United States. Journal of Geophysical Research, 69, 3395-3401. Krumdieck, A. and Chamot, P. (1979). Sediment flushing at Santa Domingo reservoir. Water Power and Dam Construction, Dec., 172-176. Lagwankar, V. G., Gorde, A. K., Borikar, D. A. and Patil, K. D. (1995). Trends in reservoir sedimentation in India. In: 6th International Symposium, on River Sedimentation, New Dehli, pp.91-111.
J
J,
Lai, 1. S. and Shen, H. W. (1996). Flushing sediment through reservoirs. Jountal of Hydraulic Research, IAHR, 34, No.2. Lane, E. W. and Koelzer, V. A. (1953). Density of sediments deposited in reservoirs. Report No. 9: A study of methods used in measurement and analysis of sediment loads in streams, St Paul l!¢te~. States Engineering District, St Paul, Minnesota, US.
I' ~.
~
Langbein, W. B. and Schumm, S. A. (1958). Yield of sediment in relation to mean annual . precipitation. Transactions of the American Geophysical Union, 39, No.6, 1076-1084.
, !
Li, Y. H. (1976). Denudation of Taiwan Island since the Pliocene Epoch. Geology, 4, 105-107.
.~
Liechi, W. and Haeberli, W. (1970). Les Sedimentations dans Ie Bassin de Compensation de Palagnedra et les Dispositions plises pour Ie Deblaiement des Alluvions. 10th Congress on Large Dams, ICOLD, Montreal. Lvovich, M. I., Karasik, G. Ya, Bratseva, N. L., Medvedeva, G. P. and Maleshko A. V. (1991). Contemporary intensity of the world land intracontinental erosion. USSR Academy of Sciences, Moscow. Mahmood, K. (1987). Reservoir sedimentation: impact, extent and mitigation. World Bank Technical Paper, No.7!. ISSN 0253-7494, ISBN 0-8213-0952-8. Mohan, J., Tiagi, S. S. and Rajvanshi, U. S. (1982). Effects of sedimentation at Ichari reservoir. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 26, pp. 409-418.
J J
I !
,-1..
Morgan, R. P. C. and Davidson, D. A. (1986). Soil erosion and conservation, Longman Group, UK. Morris, G. L. and Hu, G. (1992). HEC-6 Modeling of Sediment Management in Loiza Reservoir, Puelto Rico. In: Hydraulic Engineering Water Forum '92, M. Jennings and N. Bhowmik (eds), ASCE, New York, pp. 630-635. Monis, G. L. and Fan, J. (1997). Reservoir sedimentation handbook: design and management of dams, reservoirs and watershedfor sustainable use. McGraw Hill, New York ..
i r
Mou, J. and Meng, Q. (1980). Sedilnent delivery ratio as used in the ' computation of the watershed sediment yield. Beijing, China. Neil, C. R. and Mollard, J. D. (1982). Erosional processes and sediment yield in the upper Oldm,an river basin Alberta, Canada. Recent developments in the explanation and prediction of erosion and sediment yield. Proceedings of a symposium held during the first lARS scientific assembly at Exeter, 1982, Publication No. 137. Noble, E. L. (1965). Sediment reduction through watershed rehabilitation. In: Proceedings of the Federal Inter-Agency Sedimentation Conference, 1963. US Department of Agriculture, Miscellaneous Publication 970, pp. 114-123.
136
j
,
REFERENCES
O'Loughlin, C. L., Rowe, L. K. and Pearce, A. 1. (1980). Sediment yield and water quality responses to clearfe1ling of evergreen mixed forests in western New Zealand In: The influence of man on the hydrological regime }vith special reference to representative and experimental basins. International Association of Hydrological Sciences, Publication No. 130, pp. 285-292. Ongweny, G. S. (1979). Patterns of sediment production within the upper Tana basin in eastern Kenya. In: The hydrology of areas of low precipitation (Proceedings of the Canberra . Symposium, December 1979), IAHS Publication No. 128. Palmieri, A. (1998). The role of technical solutions in the context of international developnzent and dam investments. International Atomic Energy Agency, Consultant Meeting, Thematic Planning for Technical Cooperation Isotope Hydrology in Danl Management, Vienna. Paul, T. C. and Dhillon, G. S. (1988). Sluice dimensioning for desilting reservoirs. Water POrver & Dam Con.struction, May, 40-44. Pickup, G., Higgins, R. J. and vVarner, R. F. (1981). Erosier. and sediment yield in Fly River drainage basins. In: Erosion and Sediment Transport in Pacific Rim Steeplands, T. R. H. Davies and A. 1. Pearce (eds). International Association of Hydrological Sciences, Publication No. 132, pp. 438-456. Pidwimy, M. (1999). Climatic regions of the world: clim.a.tology - the geographic study of the atmosphere, introduction to physical geography II. Department of Geography, Okanagan University College, http://www.artslouc.bc.ca/geog/G121Iln.html. pp. 1-9. Qian. N. (1982). Reservoir sedimentation and slope stability technical and environmental effects. 14tll Congress on Large Dams, leaL!!, Rio de Janeiro. __ RaInfrez, C. and Rodriguez, A. (1992). History of the Cachf Reservoh; Costa Rica. UNOI Report No. 81, Department of Physical Geography, Uppsala University, Sweden. Rhoades, E. D., Welch, N. H. and Coleman, G. A. (1975). Sediment-yield characteristics from unit source watersheds. In: Present and prospective technology for predicting sediment yields and sources. US Department of Agriculture. Publication No. ARS-S-40, pp. 125-129. Rien6ssl, K. and Schnelle, P. (1982). Sedimentation of small reservoirs in the high Alps. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 54, pp. 65-82. Rooseboom, A. and Van Harmse. H. R. (1979). Changes in sediment load of the Orange River 1929-1969. Hydrology of areas of low precipitation. IAHS publication 128, proceedings of conference, Canberra, Dec. Salas et al. (1997a). Capacity situation in Spanish reservoirs. ICOLD 19th Congress, Florence, Vol. 3, Question 74, Reply 53. Salas et at. (1997b). Sediment yield at Spanish reservoirs and its relatior.ship with the drainage basin area. ICOLD 19th Congress, Florence, Vol. 3, Question 74, Reply 54. Shahin, M. IV!. A. (1993). An overview of reservoir sedimentation in some African river basins. In: Sediment problems: strategies for monitoring, prediction and control, lARS Publication No. 217, Wallingford UK, pp. 93-100. Sharma, K. D. and Chatterji, P. C. (1982). Sedimentation in Nadis in the Indian arid zone. Journal of Hydrological Sciences, 27, 345-352. Shen, H. "'vY and Lai, 1. S. (1996). Sustain reservoir useful life by flushing sediment. International Joumal of Sediment Research, IRTCES, 11, No.3. Shen, H . \v. (1999). Flushing sediment through resen'oirs. Journal of Hydraulic Research, 37, No.6.
137
EVACUATION OF SEDIMENTS
Strakhov, N. M. (1967). Principles of Lithogenesis, Vol. 1, 245 pp. Oliver and Boyd, Edinburgh. Subramanian, V. (1982). Erosion and sediment yield in Krishna River Basin, India Recent developments in the explanation and prediction of erosion and sediment yield. Proceedings of a Symposium held during the first lARS scient~fic assembly at Exeter, 1982; Publication No. 137.
Swiss National Committee on Large Dams (SNCOLD) (1982). General paper ICOLD 141h Congress, Rio de Janeiro, pp. 889-958. Ta Wei Soong, M. and Yean Zhao. (1994). Flood and sediment characteristics of the Lower Yellow River in China. Water International, 19, 129-137. Tabuteau, M .. M. (1960). Etude graphique pour les consequences hydro-erosives du climat mediterrane. Association Geographique Fran{:aise Bulletin, 295, 130-142. Tolouie, E. (1993). Reservoir sedilnentation and de-siltation. PhD thesis, University of Birmingham, UK. United Nations Educational, Scientific and Cultural Organisation (UNESCO). (1985). Methods of computing sedimentation in lakes and reservoirs. A contribution to the International Hydrological Programme, IHP-II projectA.2.6.1 Panel (Stevan Bruk, Rapporteur) (essentially the same as IRTCES, 1985). United Nations Education and Scientific Organisation (UNESCO) (1990). UNESCO sourcebook in climatology for hydrologists and water resource engineers. UNESCO Publication, Catalogue No. 89598, Paris. Ursic, S. 1. and Dendy, E E. (1965). Sediment yields from small watersheds under various land uses and forest covers. In: Proceedings of the Federal Inter-Agency Sedimentation Conference 1963. US Department of Agriculture. Miscellaneous Publication. No. 970, pp. 47-52. US Global Change Research Information Office. (1999). Soil erosion and sediment erosion, pp. 1-3. Web site: http://www.gcrio.org/geo/ Walling, D. E. (1994). Measuring sediment yields from river basins. In: Soil erosion research methods, R. Lab (ed.). Soil and Water Conservation Society, Ankeny, Iowa. Walling, D. E. and Kleo, A. R. (1979). Sediment yields of rivers in areas of low precipitation, a global view, pp. 479-493. The Hydrology of Areas of Low Precipitation, lARS Publication No. 128, Wallingford, UK. Walling, D. E. and Webb, B. W. (1983). Patterns of sediment yield. In: Background to palaeohydrology, K. J. Gregory (ed.). John Wiley, New York, pp. 69-100. Walling, D. E. (1984). Muddy waters move mountains. The Geographical Magazine, 262-267. Walling, D. E. and Webb, B. W. (1996). Erosion and sediment yield: global and regional perspectives. IARS Publication No. 236, Wallingford, UK. Webb, R. M. T. and Soler-Lopez, L. R. (1997). Sedimentation history of Lago Loiza 1953-1994. USGS, San Juan. White, P., Labadz, 1. C. and Butcher, D. P. (1996). The management of sediment in reservoired catchments. Chartered Institution of Water and Environmental Management Journal, 10, June, 183-189. White, W. R. (1982): Sedimentation problems in river basins. Project 5.3 of the International Hydrological Programme. UNESCO.
138
REFERENCES
vVhite, vY. R. and Bettess, R. (1984). The feasibility of flushing sediments through reservoirs. Proceedings of the Symposium 'Challenges in African Hydrology and Vlt~1ter ResoLlrces\ Harare, July, 1984. IAHS Publication No. 144. Wilson, L. (1969) . Les relations entre les processus geomorphologiques et Ie climat modern comme methode de paleoclimatologie. Revue de Geographie PhysiqLie et de Geologie Dynamique, Ser. 2, 11, 303-314. Xia, M. (1986). New desilting methods used in Heisonglin reservoir: lateral erosion on floodplain . In: Collected Research Papers, Northwest Institution of Hydraulic Research, 2, Part 2, Yangling.
Yoon, Y. N. (1992). The state and the perspective of the direct sediment removal method from reservoirs. International Journal of Sediment Research, 7, No.2, 99-116. Zhang Hao, Xia Mai-ding, Chen Shi-ji, Li Zhen-Wu Xia Heng-Bin ,Jiang Nai-Sen and Lin Benwen (1976). Regulation of sediments in some medium and small sized reservoirs on heavily silt laden streams in China. 12th Congress on Large Dams, ICOLD, ~lexico. Zhang Qishun and Long Yuqian. (1980). Sediment problems of Sanmenxia reservoir. 1st Symposium on Reservoir Sedin'Lentation, Beijing, pp. 707-716.
139
Be lio rop
r I I
I II
I I
8. Bibliography
j
1 !
~
! i
Abdelhadi, :M. L. (1995). Environment and socio-economic impacts of erosion and sedimentation in Northern African countries. 6th Symposium on River Sedimentation, New Delhi, Vols. 1-4, pp. 1141-1152. Ackers, P. and Thompson, G. (1987). Reservoir sedimentation and influence of flushing. In: Sediment transport in gravel bed rivers, C. R. Thome, 1. C. Bathurst and R. D. Hey (eds). John Wiley and Sons, pp. 845-868. Ackermann, Wbite and Worthington (eds). (1973). Geophysical 1'IIonograph No. 17, American Geophysical Union, Washington. Adams, 1. (1980). High sediment yields from major rivers of western southern Alps, New Zealand. Nature, 287, 88-89. Ahmed, T. E. S. (1992). Sedimentation in Sudan mUlti-purpose reservoirs. 5th International Symposium on River Sedimentation, Karlsruhe. Aitken, A. P. (1981). Aspects of erosion and sediment transport in Java. Problems of Soil Erosion and Sedimentation, South-East Asia Regional Symposium, Thailand. Ando, N., et al. (1994). SedLllent removal project at :NIiwa Dam. ICOLD 18th Congress, Vol. Ill. Question 69, Reply 27. Annandale, G. W. (1987). Reservoir sedimentation. Elsevier Science Publishers, Amsterdam. Anon. (1982). Sedimentation of small reservoirs in the high Alps. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 5. Anon. (1982). Estimation of silting load wid capacity loss rate in irrigation reservoirs in Korea. ICOW 14th Congress, Rio de Janeiro, Question 54, Reply 12. Anon. (1982). Some problems concerning reservoir sedimentation on Romanian rivers. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 14. Anon. (1982). Sedimentation aspects of the :tYnssouri river dams. ICOLD 14th Congress,· Rio de Janeiro, Question 54, Reply 15. Anon. (1982). Environmental impact a.l1d control of reservoir sedimentation. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 23. Anon. (1982). Studies on sediment transport in mountain streams of Eastern Alps. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 28: Anon. (1982). Reservoir sedimentation - technical and environmental effects in Indonesia. ICOLD 14th Congress, Rio de Janeiro, Question Reply 31. Anon. (1982). Investigation of average long-term suspended sediment discharge in Brazil. ICOLD 14th Congress, Ria de Janeiro, Question 54, Reply 32. Anon. (1982). Silting of reservoirs in southern Brazil - first results and methodological approach. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 33. Anon. (1982). Sediment deposition in reservoirs on the River Tana, Kenya. ICOLD 14th Congress, Rio de Janeiro, Question 54, Reply 37.
143
EVACUATION OF SEDIMENTS
Avendano Salas, C., Sanz 1v10ntero, E. and Gomez Montana, J. L. (1997). Sediment yield at Spanish reservoirs and its relationship with drainage basin area. ICOLD 19th Conference, Florence 1997, Vol. 3. Bister. (1991). Ageing of dams. ICOLD, Vienna, Question 65. Biswas, A. K. (1991). Water for sustainable development in the 21st century perspective. Water Intemational, 6, No.4, 219-224
a global
Bollinne, A. (1985). Adjusting the universal soil loss equation for use in Western Europe. In: Soil erosion and conservation, S. A. El-Swaify, W. C. Moldenhauer and A. Lo (eds). Soil Conservation Society of America, Ankeny, Iowa, pp. 206-213. Borland, W. M. and Miller, C. R. (1958). Distribution of sediment in large reservoirs. Journal of the Hydraulics Division, ASCE, 84, No. HY2. Borland, W. M. and Miller, C. R. (1960). Distribution of sediment in large reservoirs. Transactions ASCE, 125. Brune, G. M. (1953). Trap Efficiency of Reservoirs. Trans A117,erican Geophysical Union, 34, No. 3,407-418. Bruk, S. and Zebidi, H. (1996). Reservoir sedimentation. Proceedings of the St Petersburg Workshop (May 1994), International Hydrological Programme, UNESCO, Paris. Buff, B., et al. (1997). Survey of sedimentation in dams in the free sate of Saxony, Federal Republic of Germany. ICOLD Conference. Butcher, et al. (1992). Assessment of catchment erosion in the southern Penines using reservoir sedimentation monitOling. International Association of Hydrological Sciences, No. 210. Buttling and Shaw. (1973). Predicting the rate and pattern of storage loss in reservoirs. 11th ICOLD Congress, Madrid, Q40-R37. Central Board of Inigation and Power. (1981). Sedimentation studies in reservoirs. Technical Report No. 20, Vol. I and II. Central Board Irrigation and Power, New Delhi, India. Chain. (1986). Reservoir sedimentation and desilting in Taiwan. Proceedings Conference Water Forum, Vol. 1, World Water Issues in Evolution, USA. Chen, 1. L. (1994). Sedimentation studies at Three Gorges. International Water Power and Dam Construction, Aug., 54-58. Coleman, D. J. and Scatena, F. N. (1986). Identification and evaluation of sediment sources. Drainage Basin Delivery IAHS. Pub. 159, Wallingford, UK, pp. 3-18. Crowder, B. M. (1987). Economic costs of reservoir sedimentation: a regional approach to estimating cropland erosion damages. Journal of Soil and Water Conservation, 42, No.3, 194-197. Dai, D. (1991). River sediment problems in China. In: Water conservancy in China. Water Resources and Electric Power Press, Beijing. Delft Hydraulics, et al. (1992). The control of reservoir sedimentation Ministry of Development Cooperation, The Netherlands.
a literature review.
Dendy, F. E. (1967). Sediment deposition in reservoirs in the United States. ARS #41-137, US Department of Agriculture.
Paper
Dendy, F. E., Champion, W. A. and Wilson, R. B. (1973). Reservoir sedimentation surveys in the USA. In: Man-nwde lakes: their problems and environmental effects, W. C. Ackerman, E. F. White and E. B. W0l1hington (eds). Geophysical Monograph, No. 17, AmeIican Geophysical Union, Washington, DC.
!
I
~I
i
144
BIBLIOGRAPHY
Dendy, F. E. and Champion, W. A. (1978). Summary of reservoir deposition surveys made in the USA through 1975. US Department of Agriculture, rvIiscellaneous Publication No. 1362. De Vera and Rewtarkulpaiboon. (1981). Analysis of sediment yield and estimated soil loss of selected Philippine watersheds. Proceedings of South·East Asian Regional Symposium orz Problems of Soil Erosion and Sedimentation, Bangkok, Thailand. Dunne, T. (1979). Sediment yield and landuse in tropical catchments. Joun-zal of Hydrology, No. 42, 281-300. Engelman, R. and Leroy, P. (1994). Sustaining water: population and the future of renewable water supplies. Population Action International, Washington DC. Fan, J. and Morris, G. L. (1992). Reservoir sedimentation II: reservoir desi1tation and long-term storage capacity. Journal of Hydraulic Engineering, 118, No.3, ASCE, 370-384. Fan, 1. and Morris, G. L (eds). (1993). Notes on sedinzent .'1-zanagement in reservoirs: natiortal and international perspectives. Federal Energy Regulatory Commission, Washington DC. Fan, J. (1995). An overview of preserving reservoir storage capacity. International ~Vorkshop on Reservoir Sedil7'zentation, San Francisco. Federal Regulatory Commission, Washington, DC. Fan, J. and Pan, S. (1996). Control of sediment deposition by flood flushing. Proceedings of the International Conference on Reservoir Sedimentation, Fort Collins, Colorado. Fan, S. (1988). Twelve selected computer stream sedimentation models developed in. the United States. Interagency Sedimentation Work Group and Federal Energy Regulatory Commission. Fan, S. (1989). Reservoirs. In: Civil Engin.eering guidelines for planning and designing hydroelectric developments, American Society of Civil Engineers. Fan, S. (1990). lvIajor sedimentation issues and activities at the selected Federal Agencies. Federal Energy Regulatory Commission. Fan, S. (1999). Sediment management at hydropower reservoirs. International Journal of Sediment Research, 14, No.2, 391-397. Fleming, G. (1969), Design curves for sllspended load estimation. Proceedings of the Institution of Civil Engineering, 43, No. 1. Frenette, M. and Julien, P. Y. (1986). Advances in predicting reservoir sedimentation. Proceedings of the 3rd International Symposium River Sedimentation, U. lVIississippi, Oxford. Galay, V. (1987). Erosion and sedimentation in the Nepal Himalaya. "Vater and Energy Commission Secretariat. Kathmandu. Garde, R. J. and Kothyari, V. C. (1987). Sediment yield estimation. Journal of Irrigation and Power, 44, No.3. Gogus and Yalcinkaya. (1992). Reservoir sedimentation in Turkey. 5th International Symposium on River Sedimentation, Karlsruhe. Goldsmith, E. (1985). Sedimentation - the way of all dams - the social and environmental effects of large dams. European Ecological Action Group, Vol. 1, pp. 226-239. Gong, S. and Jiong, D. (1979). Soil erosion and its comrol in small watersheds of the loess plateau. Saentia Sinica, 22, No.1, 1302-1313.
opec Precipitation Visualiser see http://www.dwd.de/research/gpcc/visu_gpcc.htm1 Graf, "V. H. (1984). Storage losses in reservoirs. vverter Power an.d Dam Construction, 36, No.4, 37-40.
145
EVACUATION OF SEDlfVlENTS
Griffiths, G. A. (1979). High sediment yield from major rivers of western southern Alps, New Zealand. Nature, No. 282, 61-63. Hadley, R. F. and Shawn, L. M. (1976). Relation of erosion to sediment yield. Proceedings of the 3rd Federal Interagency Sediment Conference Water Resources Council, Washington DC, pp.1.l32-1.139. Harrison, L. L. and .Fan, S. (1996). Sedil1'lent management solutions for FERC licensed reservoirs. Draft Staff Paper, Federal Energy Regulatory Commission. Harrison, L. L. and Fan, S. (1997). Needs for sediment management at non-federal reservoirs. International Water Power Conference. Holeman, J. N. (1968). Sediment yield of major rivers of the world. Water Resources Research, 4, No.4. Hotchkiss, R. H. (1999). Reservoir sedimentation: research needs and refocusing perspectives. International Journal of Sediment Research, 14, No.2. International Association for Hydrological Sciences (IAHS) (1979). The hydrology of areas with low precipitation. IAHS Publication 128, Wallingford, UK. International Association for Hydrological Sciences (IAHS) (1986). Challenges in African hydrology and water resources planning. IAHS Publication 144, Wallingford, UK. International Association for Hydrological Sciences (IAHS) (1986). Drainage basin sediment delivery. IAHS Publication 159, Wallingford, UK. International Commission on Large Dams (ICOLD) (1984). Deterioration of dams and reservoirs. International Commission on Large Dams (ICOLD) (1989). Sedimentation control of reservoirs. Guidelines. International Commission on Large Dams (ICOLD) (1997) . ICOW 19th Congress, Florence, Vol. 3, Question 74, Reply 34. Jiang, J. and Fu. (1997). Problems on reservoir sedimentation and measures for reducing sediment deposition in China. ICOW 19th Congress, Florence, Vol. 3, Question 74, Reply 5. Julien, P. Y. (1995). Erosion and sedilnentation. Cambridge University Press, Cambridge, UK. Lauffer, H. and Sommer, N. (1982). Studies on sediment transport in mountain streams of the eastern Alps. ICOLD conference, Vol. ill, Question 54, pp. 431-451. Liu, X., Yueping, H. E., Zhang, X., Qin, Y. and Shuyou, C. A. O. (1999). A study of sediment problem in the Gongzui reservoir. International Journal of Sediment Research, 14, No.2. Meade, R. H. and Parker, R. S. (1985). Sediment in rivers in the United States. National Water Summary 1984, USGS Water Supply Paper 2775, Washington. Milliman, J. D. and Meade, R. H. (1983). Worldwide delivery of sediment to the oceans. Journal of Geology, 91, No.1, 1-21. Mizuyama, T. (1989). Sediment yield and river bed change in mountain rivers. Proceedings of the IAHR, International Workshop on Fluvial Hydraulics of Mountain Regions, Trento, Italy, pp.1-5. Monis, E. L. (1995). Reservoir sedimentation and sustainable development in India: problem scope and remedial measures. 6th Symposium on River Sedimentation, New Delhi. Newson, M. D. (1998). The role of geomorphology in monitoring and managing river sediment systems. Journal of the Chartered Institute for Water and Environmental Management, Feb.
146
-f
".t.,"," ;,
.1
BIBLIOGRAPHY
,,~
1
I1 f
1
Palmieri, A., Shah, F. and Dinar, A. (1998). Reservoir sedimentation and the sustainable management of dams. World Congress of Enviromnerztal and Resource Economists, Venice', June, pp. 23-27.
~
Partheniades, E. (1965). Erosion and deposition of cohesive soils. Journal of Hydraulics Division, ASCE, 91, No. HYl, 105-l38.
j
Patzkowski, L. G. (1992). Rock creek - cresta sediment management project: an overview of geotechnical and sediment characterization studies. 9th. Annual Dredging and Placer Mining Conference, Reno, Nevada.
i i
1 j
I I
A
I ~
£ !
~ ~
f
Pemberton. E. L. (1978). Reservoir sedimentation. Proceedings of US-Japan Binational Seminar on Sedimentation and Erosion, Honolulu, Hawaii. Pitt, J. and Thompson, G. (1984). The impact of sedimentation on reservoir life - challenges in African hydrology and water resources. Proceedings of the Harare Symposium, IARS Publication No. 144. Ritchie, J. C., Cooper, C. M. and McHenry, I. R. (1986). Sediment accumulation rates in lakes and reservoirs in the :Mississippi river valley. 3rd Symposium on River Sedimentation, Jackson, Mississippi.
~
Roehl, J. "V. (1962). Sediment source areas, delivery ratios and influencing morphological factors. IAHS, Publication No. 59.
I
Rogers, R. S. and Schumm, S. A. (1991). The effect of sparse vegetation cover on erosion and sediment yield. Journal of Hydrology, 123, Elsevier Science Publication, Amsterdam.
! ~
Roosebooffi! A. (1975). Sediment transport in rivers and river Pretoria, South Africa ..
basins.~
Thesis. University-"of
Roosebooffi, A. and Mulke, F. 1. (1982). Erosion initiation. Proceedings of Symposium on Recent Developments in the Explanation and Prediction of Erosion and Sediment Yield, IAHS Publication No. 137. Rooseboom, A. (1992). Sediment transport in rivers and reservoirs - a South African perspective. Water Research Comrrjssion Report No. 297/1/92 South Africa. Sanderson, M. (1990). UNESCO Sourcebook in Climatology for hydrologists and water resources engineers, ·UNESCO. Scheuerlein, H. (1993). Estimation of flushing efficiency in silted reservoirs. Proceedings of the 1st JeHE, Vol. 1, Part A, USA. Shangle, A. K. (1991). Hydrographic surveys of Indian reservoirs. Report by Central Water Commission for Workshop on Nlanagement of Reservoir Sedimentation, New DeLhi. Strand, R. 1. and Pemberton, E. L. (1982). Reservoir sedimentation. US Bureau of Reclamation, Denver, Colorado. Swiss National Comrnittee on Large Dams (1982). General paper. ICaLD 14th Congress, Rio de Janeiro , pp. 889-958. Tolouie, E., West, 1. R . and Billam, 1. (1991). Sedimen.tation and desiltation in. the Sefid-Rlld reservoir. British Geomorphological Re.search Group meeting, St Andrews. United Nations Educational, Scientific and Cultural Organisation (UNESCO) (1977). Atlas of World Water Balance 1977. Karzoun. VI Editor-in-Chief, Paris. United Nations Educational, Scientific and Cultural Organisation (UNESCO) (1978). World Water Balance and Water Resources of the Earth. Leningrad.
147
EVACUATION OF SEDIMENTS
United States Department of Agriculture (USDA) (1943). The control of reservoir silting. Miscellaneous Publication No. 521. Varma, C. V. J., Rao, A. R. G. and Natarajan, S. (1992). Sedimentation of Indian Reservoirs. An assessment. 5th Symposium on River Sedimentation, Karlsruhe, pp. 919-923. Veltrop, J. A. (1992). The role of dams in the 21st century. US COLD, Denver. Walling, D. E. (1989). The erosion problem. International Journal of Sediment Research, 14, No.
1, 1-11. Walling, D. E. and Probst, J. L. (1997). Human impact on erosion and sedimentation. IAHS Publication No. 245, Wallingford, UK. Water Power. (1979). World's largest capacity reservoirs. Intenwtional Water Power and Dam Construction, 31, Nov., 98. White, W. R. and Rofe, B. H. (1996). Responsibilities and lists associated with large reservoirs. Intenwtional Conference on Aspects of Conflicts in Reservoir Development and Management, London. Wolman, M., et al. (1989). Erosion control and reservoir deposition. Proceedings of Bilateral Seminar on Problems in the Lower Reaches of the Yellow River, China. pp. 139-163. Kluwer Academic Publishers, The Netherlands.
148
p
Appendix I. Reservoir data ! !
This appendix provides data which ICOLD has obtained from member countries. Table A1.1 provides, on a country by country basis, data on reservoir capacity development in the twentieth [century. Table A1.2 provides sedimentation data for individual dams.
lSi
Table AI.I. Region
Country
Notes
Africa Africa Africa Africa Africa Aldca Al'rica Africa Africa Africa Africa Africa Africa Africa Africa Africa Afiica Africa Africa ArriclI AfriclI Afril:a Africa Africa Africa Africa Africa Africa Africa Africa Africa Africa
Angola Benin Botswana Cameroon Congo Congo (ORC) Ethiopia Gabon Ghana Guinca Ivory Coast Kenya Lesotho Liberia Madagascar Mal awi Mali Mlluritius Mozambiljue Namibia Nigeria Scnegal Seychelles Sierra Leone South AI'rica Sudan Swaziland Tanzania Togo Uganda Zambia Zimbabwe
Africa
Total
C. Asia
Kazakhstan Kirghizstan Tadjikis(an Uzhekistan
I
r
Volume
3 213 967
575352
595
12 11 7 14
89786 23303 28182 6761
7482 2118 4026 483
3
I
2 I
c c
1910--19
1900--9
1920--29
~
1930--39
1940-49
i950-59
Av. vol No. Volume No. Volume No. Volume No. Volume No. Volume No. Volume
630 9446 1734 867 77 231 14325 1592 44 22 5319 443 2624 328 220 220 150279 37570 237 119 37917 1724 162 2428 1970 281 0 0 431 43 I 4 13 440 6720 61 7 57 102 9517 51 662 40414 898 11520 5760 I 1 22 22 57 30583 5587 1117 42 250 1135 1135 I I 200 200 47 16 187117 878
IS 2 9 2 12 8 I 4 2 22 15 7 I 10 4 2 9 6 13 45 2 2 I 539 5 6
f
<1900
Totals No.
C. Asia C. Asia C. Asia
m
[COLD data - world total storage volumes (M.m 3 )
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
' 0 0
0 1267
0 0
0
I
0
0
0
0
0
0
0
0
0
0
0
0 I 0
0
0
0
0
0
0
0
0
I
125
0
1960-69
No.
Volunic
No.
3
142
4
44
0
I
0
I 5
2 0 I I
I
0 8 2417 7
146 0 36 35 1925
0
0
0 0 I 0 0
3
Volume
7 2
293
<>
ISO 000 237 9 16 3 0 0
2
I
I
2 I I 0
8 1536 4
3 I I 6
36 33 295 15079
I
I 0
1100 16 0
I 2 2
I 2 I
1980-89
1970-79
No.
Volume
Volume
5 I
5124 1710
4
10915
0
0
0
()
0
0
0
0
0
0
0
2
6
0
I 0
2 0
2
7
14
()
0
5 0
12 0
11 0
164 0
26 I
679 931
22 I
3534 0
15 0
173 0
413(> 24 85 3410
3 4
0 692
I
1600
I
0
0
0
I
2
192
I
220 0
0
0
0
87
14 2 0
28444 170 0
3 9 I
8339 2155 4
0 I I
0 70 1950
2
13
0
0
0
0
I
I
125 0 3
16053 0 23
124 0 2
2194 0 177
I
0
()
n 0
(,
1134
()
I
I 1270
I 0 1
I 1270
44 0
5099 0
0 0
I
41 ()
0 183266
2 36
6 1431
51
0 1016
0 82
0 1201.1
0 5
66
6440
164
358850
235
124018
219
46580
132
30900
17
1326
1
630
56888 723 180 2928
29878 20060
1
4160
4 4 3 6
6 3
I
3 2 4
2380 10542 1564
I
2390 140 13 300
3
15830
I
0 0
I 2
41 39
0 2
0 39
0 5
0 56
6
16
11
164
31
1696
27
4967
26
251
96
144
l'
3 0 I
I 0 0
2
8 I
2170 4 1273 98 9331 250
I I
25
0 21
0 4
0
54260 222 14856
2
200 0 0
104 2 I I
0 0
3 6 23
4
2269
I
0
()
C. Asia
Total
44
148032
3364
0
0
0
0
0
0
0
0
0
0
0
0
2
4790
17
60719
13
52207
9
14486
0
0
China
China
1851
649322
351
4
2
1
1
2
8
1
6
I
15
3
24057
177
26113
479
256549
583
103536
220
58090
120
57278
260 123667
1851
649322
351
4
2
1
I
2
8
I
6
1
15
3
24057
177
26113
479
256549
260 123667
2 1
1579 6500 70 48 340 126
0
0
0
0
0
0
0
0
0
0
0
0
2
3158
0
37 0 0 4
2101 0 0 267
39 0 0, 2
826 0 0 157
82 0
77
41
2368
41
983
China
Total
S. Asia S.Asia S.Asia S. Asia S. Asia S. Asia
Afgh anistan Danghtdesh India Nepal Pakistan Sri Lanka
4010 3 69 46
3158 6500 279548 144 23436 5816
S. Asia
Total
4131
318602
- )
I
45 0 0 I
3287 0 0 101
40 0 0 0
6658 0 0 0
35 0
2
2446 0 27 87
85
2560
46
3388
40
6658
I
213 0 0 9
25601 0
0
550 0 7 0
1329
414 0 19 12
36
557
224
30088
446
I
I)
~
o z
3 I I 4
707 0
61
I 0 8
0 0
Volumt!
0
I
144 0
No.
0 22 1824 4656 50 1135 I
1 2
No.
Volume
C
No.
I I
()
ute
1990-1998
583
103 536
220
58090
120
57278
() 0 6500 55938 1128 0 0 7961 16 621 8
0
0
0
0
0
0
0
57076 1178 3 0 15209 II 527 8
38901 144 184 2727
116 0 21 0
1045 0 48 0
683
0
85119 0 0 0
71020 1152
72 812 1200
41956
137
1093
683
85119
()
0
o " m
V1
o
3: m Z -f V1
.Mid. E"sl Mid. Easl Mid. Easl Mid . 1;",1 Mid.E'lsl (vlid.E:ISI Mid. Easl Mill. Hasl Ivlid. ['ast Mid. Fasl Mid. E'lSI
Aruh.'nill
Iii
Azerbaidj'lIl CYI'I1IS
17
l.d>:IIIOII Saudi Ambia Syri'l Turl,ey
5 311 41 1i25
I7H9 183K5 2B6 2S!!1 32935 ':>7183 203 161 2114 342 109233
'lblal
895
22:161B
5'2 14 66
(iL."Ofgia
11';111 Inil(
n
JII.-.I:1I1
8
---'-'- - - - - - - - ---Mid. Easl
- - --
-----.-
N. "f,-il' a N. Arri.:a N. Africa N. A!'rica N. Afrim
1\lgcri;1
EUypl
Lihya t\'tllruccCl
"ruilisia
- --.- - - - N. Afri':3 'lllial _- _·--0_._.N. AIII('ri~a Call:llla N.A llwric'l USA
--
---
N. AlIler;":1 'lhlal N. ElIr
Pacilic l~illl j'al'ifiL: \(jill P:tdlkRil\l Pacific Ilim Parili.: Rim 1"lcilkRill1 P;ll'ilic Rim Pacilie Rim Pa.:ilic Rilll Pa~ilic Rim P;wilic Rim Pacilic Ililll
Ln W
- --
PHeili. Rim 'lillal
---
0
0
()
II
()
(I
0
0
0
0
0
I)
I)
II
1/
()
0
0
0
0
0
0
0
0
0
0
0
0
175
()
()
2511
n
0
3 0
5 0 0
-- --
I)
0 n II I0 0
0
()
0
0
0
n
0
II
(I
t)
II
I)
0
I
I
'l
0
0
tl
n
0
n
IJ
0
I)
(I
0
I)
()
0
0
10 132
48 1638
652
3
5 6115
6T!4
517 670
9 665
47.'i~
- -I 3
()
1929 19 Inn 160SY
()
0 141 15
0 17 14 0 0
0 0
Il
0 0 0
9 14 0
403 0
(I
0
t)
n
II
II
I
II
0
0
6 I
1414 331
0 5 6
I
'2
7
6104
\0
1911
5
381
36~63
414
I
II 0 2 158 92 0 6
I 0 2 35 27 0 3
281
22
9')
19
214·2
71
1S1.S
90
5202 153
484 H7 2 9ti 1077 7(-,5 70 86
89855 10 775 153 74348 21 OKS 161161 34312 16515 333 3301 6714 2808
186 124 17 774
33 2
240 10
14 I
627 95
21 6
144 30
21
5102 24
21 20
2050 2 I
1 7 0
17 10 0
3 43 15 50
._-
:ms
- - -- . 277 124
0
I 26 14 ()
I
I)
1221 15
\I
543 83
14 8 I 2 3
0
/I 612
0 0 1532 54
30110 5 177
6
9 15 I 9
II
168
43
68429
\06
6 I II
259 1620nO 290
2238
31
0
I
8
6423 0
41i 563 1137
0 29 61
473 0 0 5143 199
tl
0
40 0 4 37 12 93
6')696 349
84923
925 336912 1672
592755 1393
I 0 6 38 9 I '2 I
25 0 5726
-II 0 21 826
60S II
654
0
~ - --
()
906 2\J 31 j
I '2 14 75 34 Il 4
51:2() 162 201 1859
0 14 2
.5
3 18
10929 3320 74J
S
II II
954 276 10281l 27
1006
17
166
24457 117
19631
--
3 5 7639 2755 850 0 307
2 4 16 90 49
-~--.--
t-506874
r--
0 485 35 2994 1811 225700 47 5329 18216 0
3 3
5 7 2953 3998 SKI \3 0 5\0 462 79
3 101 9 19 4 53 7 58
204 11 765 883 173475 41 7927 19 7~(' 38!li!
Y 0 II 93 76 4 2 2 12 0 I 2 75 10 IJ 4 30 6 32
315 264 553
425
226517
382
195444
R6 II
45850 64
16
2
0 ')
0 :2 37 3 20
"
0\7 3 68
t----.
16 5
1187 71
51 7
S'S\() 1608
93 12
121'13 5901
,II 22 9 9
106 255 86 2332 322
4 37 62 7
3(1'l3 832 1361 2112 94
2 137 102 12 \0
128 3504 191 16350 nll5
4 3 8
322 424 198
I 208 140 18 12 I 8 I 6
2556 6444 1138 4316 2907 0 875 1!50 367
336
341)20
5(11)
37,197
4
2
2
52
II
.6
51
0
0
I
II
I
12
2
31
3
57
4
160
I
I
6
332
I 4
31 12 II
100
53
493
7J
5655 135
~449
142
6(i74
44H
3
4371
0 2 14
885 12 632
0
3 I)
8
II
0
0
60
0 39 0 \0 995 248SS t--
n
()
0
0
[I
II
0
"I -0 0
122
5815
15 195
1i32 )-125
0
0
0
757 172 306
2111
4057
0
0
0 I 3 54 6 4 I
0 1900 72 3133 181 420 5
0 0 0
0 0 0 672 100 0 0
1
0
3 2 96 48 7 0
3
277
3984
132
340 65 0 735 1I2U 9154 724 162265 20 11028 2510 541
I
116 155495 641 16 MIl
2539 2054 639 1575
(--
0
264
335
168 972
2
40
43811
455767 51 \07
0
2777
221
23
35 151 55 114 161
23
41 178
432092 142 160663 1251
I 2 I
--
122
/0
(- -
3526 459
5
64 050
9 69
1-17 211691 128 Tl8 125221 1544
0 J2
5
8 I 0 15 7 151
52
133
0 6
18 13 3
0 76 14 3490
9
8551 26000
32428 52495
20
3 I 21 4
593 13 231 6 1206 1756 28 21330 80 U 0 165 U 6 18437 295
7
0 I
--
787 2182 17
8176 55 61520 294
·0
:!I)
22 490 192 III 77
15H
I)
0
0
75.) 0 0 595 197
II
II
II III
9
165
0
II
22061
3
0
0 482
86
106 0 27 0
7 8 4 3 8 I I 0 12 6 72
0 0
1
6 0 0 I 0
(I
131 190 29 350 5568 3053 II 160 I 327 12241
3 I 13 29
'/31 5000 0 373 0
I
5
I
0 '2
1
- - - - - --- I - : - - - - -
0 2
2 2 0 6
0 0 0 2
56
5
0
2298'2
98
(I
0
0
0
21
52Y
(I
()
6800
314
7(1
5 0
0 .5 I 0
0 ' :\
0
5'/
2
188
0
3 0
0
120
18
248
J I
0
CiK
4
7 16 '2 5 3
18
412
()
14
56 117 0 tl
'2 I I
I 160011 I
2
-- ---
9JS 168
0
I)
n
5
2
(I
'1277
35 15 0 3
0 II
0 312
0 1
9 321l 69 91 28 19[1 20 517
.--- - -
0
0
2679 994 H66 6"1 79 1220 9259 53496 3598 731 637 206 3:l216 46 B8'1 6864
(J
II
()
19:1 3'16 28 14 149 62 173 31 26 6\0 1029 163 52 R040 7 201 234 '1 13
~
-
0
0
0
·12 12247 85 17679 82 246 III 747 296 19284 332
256 674 5364 142 36R6 28K 30994 ._r---'t - - 1-- - --. - . 252 Iii 0 0
0 I 0
3
II 0
--
-.
0
t)
1597 81
--._0
0
()
0 0 0
I 3~6 860
16 5 20 3
UK
'Ihiw;ln
0
1111
IS
Ukr:linc
I
U
47:1
55 569 311
Sw'-'lIcn
N. Z"'II:md I'. N.Guillca l'elll Phjlipllincs
18~
111
Rllssia
J{orl~il
289
16
Slnwnia
Auslralia Chik Fiji Indonesia Japan
' -'
4732 48 Ifl71100 27833 3]6 28 I~ 54\) 15~ l11fi5 II
a
I I
0
7205 IM'14530
lkigilll\l [klllll;1rk I'illiand Fr:mc.: Gl'I"IIHIIIY kdand Ireland Lllvi" Lillluania Luscmhllrg Mllld'lI'ia Nl'lherlalld,; Nlww:IY Polalld
---- [ - . _ - N. Europe TOlal ---.--- - -- --- . -
99 6 12 92 8tl
112 1081 6 20ti 49<) 4399 25 32 7
68 17 77
1322 463 640 ()
4 0
4K
0
0
0
1 51 8 5 6 12 2 II
2780 2U944 2RO 10493 52 631 92 4011
0
0 4044 653 327liU
6 ,9
1. I 3
<)
95
II
I 5 0 0
---II
I 0
46
n
0 175 236 111805
5
"
\I
0
()
II
0
0 JOUO 0
I
3
69
38307
93
43341
92 14 2 27 165 '187 6 19
9808 1222 153 9277 1579 63911 3139 614
36 7
271H 291
0 2
0 10011
27 198 68 3 0
14515 3004 217K 6726 0
\I 56 0
41997 16119
II
0
690 2425 505
3
157
6
52
3
299
3581U
3115
29888
75
4'1738
._257 -
10. 202 170 12 14 2 5 4
574 3586 5453 1053 8978 33) 1069 3003
I)
1)95
4 6 9
525
70958
531
II
28 116034
II
» -0 -U
m
Z
o
X
m
~ n
Table Ai.i. Region
c ~
continued
Country
Volume
No.
1920-29
1910-19
1900-9
<1900
TOlals
NOles
1930-39
1940-49
I
1950-59
Av. vol No. Volume No. Volume No. Volume No. Volume No. Volume No. Volume
No.
1960--69
Volume No.
Volume ------
S. America S. America S. America S. America S. America S.America S. America S.Americ:n S.America S. America S. America S.America S. America S.America S. America S. America S. America S.America S.America S.Amcrica S.America S.America S.America
Antigua Argentina Bolivia Brazil Colombia Costa Rica Cuba Dominican R. Ecuador EI Salvador Guatemala Guyana Haiti Honduras Jamaica Mexico Nicaragua Panama Paraguay Surinam Trinidad Uruguay Vene..:uela
5 130380 372 540386 10913 2288 3755 2373 70()3 2430 460 44 I 9035 220 121253 1252 5184 33690 20 48 12335 155467
5 1304 53 922 223 254 77 216 226 486 115 22
1498 I 038913
694
I
a, d
a, b d
S. America Total
100 7 586 49 9 49 II 31 5 4 2
I 8 2 536 4 5 4 I 4 5 74
0 0 77 0 0
0 0 1874 0 0
0 0 3 0 0
0 0 179 0 0
0 0 16 0 0
0 0 540 0 0
3 0 35 0 0
32 0 2029 0 0
Albania Austria Bosnia Bulgaria Croatia Czech Rep. Macedonia Greece Hungary Italy Ponugal Romania Slovakia Spain Switzerland Yugoslavia
S. Europe
Total
e
306 148 25 180 29 118 18 46 15 524 103 246 50 1\87 156 69
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
117
10
4144
6
I
544 0
I 15
0
92
0
3220
0
0
0
14 0 100 9
I
1057 161 1665 0 0 4
0
0
0
0
2 0
14 3 16 0 0
I I
0
0
0
0
8
0
0
0
0
0 220 39
0
0
20
2658
0
1 0
624 0
I 0 0
5 0 0
0 I 44
0
0 0 9678
447 0 11825 250 0 5
17 2 103 10 2 15
0 6 620 0
0
I
I I
0 0
0 0
0 0
0 0
0 0
0 0
0 I
0 3
92
1966
II
296
27
5228
47
2323
0 0 I 0 0
0 0
17706 1841 56477 4094 4360
0 0 53 0 0 I 10 0 0 3 113 1 0
145162
45
85
182
2
----
62
----
0 7 0 0 0 4 0 2 0 12 5 0
0 0 0 0 0 7 0 0 0 17 0 0 0 57 3 0
0 3 0 0 0 7 0 0 0 16 0 I 0 44 5 0
30
84
76
0 0 0 0
0 35 0 0 0 28 0 0 0 38 0
I 0 656 94
0 4 0 0 0 2 0 0 0 79 I 0
0 41 1I 0
852 138
0 5
0 0 0 3 0 0 0 1375 22 0 0 636 303 0
0 4 0 0 0
10 I 1 0 57 3 2 1 45 10 2
2344 136
I 5
3 4 I
2
1057 6330 380 460
3
Volume
No. ~ ~
----
21 2 3
34593 0 85063 85 13
2 3
175 465
0
0
0
0
()
Volume
----
I 29 3 0
197 54 38298 1999 0 0
I
0 44
0
0
(I
20357
I 2
0 &04
0
I
21000
(I
532
0 4
0 3981
58 141283
44
1940
4
10
1
7085
0
133 I
16573 23 4000 12690
65
15293
21
1
0
0
I 42 2 2946 16 142726
0 3
357
0
0
0
8800 345
I I 3
I 29 0
I I 16
202
31466
322
147118
I
2
0 26
0 3126
2631111 191 350268
m
o 3: m Z
--i
66377 -------
1 30 3 19 5
2
3922
424
21629
JO
577 712 2289 2145 151 1567 1211 5875 19 2159 1627 2062 611 19353 1703 703
125 26 4 54 4 23 3 3 6 43 15 74 9 195 18 II
3857 700 814 2865 24 729 76 2721 30 634 899 6097 428 7444 343 1502
629
42764
613
29163
73 25 9 78 4 23 8 4 5 77 13 29 13 211 47
98 29 7 23 14
11 2 7 3 33 21 81 7 186 9 28
596 389 343 1034 231 187 19 1246 12 1070 1103 2674 131 7723 17 1620
9 9 I 4 2 5 0 26 0 31 21 20 7 139 2 2
29 94 5 43 0 253 0 2924 0 1273 644 728 258 5037
559
18395
278
0 2 0 0 0 0 2
2
..
0
4 0 0 0 0 0
42
0 28 0 34 0 38 0 3
1100 0 1355 0 6120 0 1865 0 143
11330
109
10587
()
o ..."
(.I')
(.I')
2 3
3910 143
II
I
31 6 22 I 0
9530 18 I 6 60 161247 14 3162 2 4
145 I 2 0
3
I
27 0 91 10
0
13 2 3 0 116 22 2 7 159 32 10
0 65
No.
18282
698 123 2221 608 404 359 447 0 3174 3011 10 408 9093 893 178
0 2 0 4 0 0 0 34 7
Volume
68
0 183 0 16 0 31 0 0 0 882 291 0 0 2073 278 168
22 0 0 0 192 4 41 0 715 31 14 2 2427 459 4
No.
~
5 79705 0 135220 3821 2220 2957 1141 190 1430 0
z
I
34020 425 16 0 20 0 560 4754
~217tO o
4204 151 98590 1593 51 782 0 12 0 0
Volume
o
utC
1990-1998
I 0
I
13 0 0 1 6 0 0 6 52 6 0
~~~~
I
615 0 3856 3 0 7 0
I
17 19 143 46 35 29 93 312 4 24 74 72 37 48 26 63
5061 2841 3576 8324 1014 3454 1669 14354 62
3 0 35
I
-~-
S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe S. Europe
No.
1980--89
-~~~~~~
I
I
1129 110 226 313 1037 8423 20 12 2467 2101
1970-79
_"
,.
S.E. A~ia S.B. A$iu S.E.Asia
Brunei C:unhmlia
::i.E. Asia
Millays; ..
S.E.Asi:1
S.E.l\siil
M,YIIIIHlur Sil1gllllure Thailand
S.n.Asia
VkWiltll
S.r:. Asia
'2 '2
Lnos
45
2325
22 119 7(nn 491 465
204
75 78534
15 385
I
16.5
165
277
117 311
424
I 59 5 :I
23'1
7030 28 ')(,0
1blal Worltllllllli
25432 6464730
I 14
18
2442
1 II I
28 18536 165
33
21 18(i
I 4
2
-1-
S.I1.Asia
I I
6
I
'11
~
0
0
0
254 1208 11308 1355
0
0 II
()
0
(J
0
Il
0
.,'--
9986
0
0
0
0 0
0
2
0
2
96
0
96
()
0
I
75
0 I
0
75
2
0 '2
44
0
44
4 I I 20
27
7030 6657 B27 )0 30976
I
44
19 3
19720 1423 17 27 ROI
I 135
~~
49005
I
223
n
0
0
3ri
1150
0
37
1J73
0 0
---0
0
600 411117 808 57579 979 131 139 911 162241 2734 781847 4793 I 838603 5425 I 673793 '1426 872998 1867 J9478K I32S 488566
NOles: :\ AI'IlClllill:t ~.,c1udesYm·yr~la 21 OIlO M.llll underconslruclion - indudcd in Paraguuy b Paraguuy ex<:1udt'sllllil)U 24 (Jon M.II1) (1983) - included in Bl1Izil c Zmllbiu .:xclmJes Kuritm 1806()O M.m' (1959)- included In Zimbubwil d Ul1IglIUY .:xllld.:s SUlll~ Grande 5500 M.ll1) (1979) - included ill Argenlinn e Yugu~lavia excludes Djcrdall 1 255 M.m' (1972) and Djerdap 11868 M.m' (1987) - included ill Homania r 'lllgU exdudes Nangheln 1710 M.nl) (1988) - il1c1uded io lIel1in [tala cXlracl.::t1 from ICOLD World ,.esister of ifillll.r (1998)
» '1J -0
m
Z
o X
EVACUATION OF SEDIMENTS
Table Al.2. CountTY
Reservoir sedimentation rate data Reservoir
Catchment: km 2
Capacity: M.m'
Survey dales Stan
India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India India
Uri Maharashta - Koyna Maharashta - Khadakwasala Maharashta - Yeldari Maharashta - Ramtek Maharashta - Dama Maharashta - Gima Maharashla - Ekruk Maharashta - Mhaswad Maharashta - Visapllr Maharashl3 - Mangi Maharashta - Asolmendha Maharashta - Nalganga A.P. - Sriramsagar A.P. - Nizamsagar A.P. - Himatyasagar A.P. - Cembum Tank A.P. - Kaddam A.P. - Ramappa Ll!ke A.P. - Lakhamvaram A.P. - Dindi A.P. - Palair A.P. - Manjira Kanataka - Tungbhadra Kanataka - Bhadar Tamilnadu - Lower Bhawani Tamilnadu - Vaigai Tamilnadu - Mettur Tamilnadu - Upper Bhawani Tamilnadu - Setnur Tamilnadu - Aliyar H.P. -Pong Punjab - Bhakra Uttar Pradesh - Mamtila U.P. - Ramganga U.P. - Dhukwan Uttar Pradesh - Tehri Bilhar - Panchet Hill Bilhar - Maithon M.P. - Gandhisagar M .P.-Tawa GUjrat - Ukai Gujrat - l(adana W.B. - Mayurakshi W.B. - Kangsabati Orissa - Hirakud Kerala - Malampuzha Kerala - Peec.hi
India China China China China China China China China China China China China China China China China China China China China China China China China China China China China
156
12750 892 507 7362 212 404 4729 412 1243 412 304 246 316 91751 21694 1308 993 2656 184 268 3919 1687 16770 28 180 2435 4200 2253 42200 34 10826 195 12562 56980 20720 3134 21340 5188 1087.8 6294 23025 5980 62224 25520 1860 3790 83395 148 107
2797 110 934 117 237 609 94 87 43 34 93 76 3172 841 108 106 124 82 60 74 57 51 3751 239 933 195 2709 101 235 109 8579 9869 1133 2450 106 3540 1581 1349 7740 3645 8510 1543 608 1135 8105 228 113
Total
591737
78412
Sanmenxia - Yel10w River Sanmenxia II Hongshan Guating - Yongding River Fenhe Liujiaxia - Yellow River Yanqouxia - Yellow River Qingtongxia - Yellow River Bapanxia - Yel10w River Danjian kou Celian ZhenziJiang Naodehai Gondzui Bikou Shimen Hongsiba Wangyao Fengjiashan Miugong Dongxia Shixiakoll Chang shan tao WenYllhe Bajiazui Yang maowan Zhaikou Luhun
688000
16200 9640 2560 2270 721 5720 216 606 49 16000 200 36 168 357 521 105 34 203 389 183
43402 181 800 182700 285000 215900
77 175 348 105 496 120 185 1320
End
TOlal sedimentation: M.m'
Vol. lost: 2
1961 1870 1963 1914 1910 1965 1871 1888 1902 1957 1918 1963 1970 1930 1927 1956 1958 1919 1909 1943 1928 1966 1953 1963 1953 1958 1934 1965 1957 1962 1974 1958 1956 1974 1907
1986 1940 1983 1987 1941 1979 1991 1990 1988 1989 1987 1985 1984 1975 1976 1978 1977 1975 1975 1976 1977 1977 1985 1974 1983 1983 1984 1985 1982 1981 1986 1987 1990 1986 1980
17·5 23-9 84·1 14·5 2·1 49·5 26·4 45 ·3 29·6 3·3 :27-0 4·2 794·9 533·6 28·6 2·1 45·9 2·g 2·0 2·2 1·1 18·7 588·2 31·0 37·8 22-4 528·3 3·6 27 ·6 3·1 422·J 915·8 273·5 97·0 47·2
0·6% 21 ·7% 9·0% 12-4% 0·9% 8·1% 28·1% 52·0% 68·8% 9·6% 29·0% 5·5% 25·1% 63·5% 26·5% 2·0% 37·1% 3·4% 3·3% 3·0% 2·0% 36·7% 15·7% 13 ·0% 4·1% 11·5% 19·5% 3·6% 11·8% 2·9% 4·9% 9·3% 24·1% 4·0% 44·5%
1956 1955 1960 1974 1972 1977 1955 1965 1957 1955 1957
1985 1979 1976 1980 1984 1984 1975 1972 1984 1977 1982
185·2 155·4 359·1 9·6 541·2 70·2 62·0 5·7 1334·9 8·0 25·4
11·7% 11·5% 4·6% 0·3% 6-4% 4·6% 10·2% 0·5% 16·5% 3·5% 22·5%
7514
9-6%
1960 1960 1960 1953 1960 1968 1958 1967 1975 1968 1960 1959 1963 1967 1976 1973 1960 1972 1971 1960 1959 1959 1960 1959 1958 1970 1970 1960
1978 1989 1987 1994 1989 1989 1968 1980
5450 5690 670 630 330 1410 J61 566 15 1130 205 29 2 206 218 28 7 77 63 97 41 35 47 20 249 17
33·6% 59·0% 26·2% 27·8% 45·8% 24·7% 74·5% 93-4% 30·6% 7·1% 102·5% 80·6%. 1·2% 57·7% 41·8% 26·7% 20·6% 37·9% 16·2% 53·0% 53·2% 20·0% 13·5% 19·0% 50·2% 14·2% 4·3% 4·7%
N/A
1986 1983 1973 1986 1987 1986 1988 1986 1990 1990 1989 1983 1988 1986 1988 1990 1990 1990 1983
8 62
Annual sedimentation
Notes
% M.m'/ann
550 784 673 571 938 166 747 534 357 835 336 1588 601 619 547 446 96 910 267 112 14 102 652 1158 300 398 250 5347 102 838 2800 554 388 2580 30 1400 587 1029 975 268 725 393 1667 216 593 2465 9505
7·01 0·70 0·34 4 ·20 0·20 0·07 3·53 0·22 0·44 0·34 0·10 0·39 0·19 56·78 11 ·86 0·58 0·10 2·42 0·05 0·03 0·07 0·02 1·70 18·38 2·82 1·26 0·90 10·57 0·18 1·10 0·16 35·17 31 ·58 8·04 8·09 0·65 7·26 6·39 6·48 2245 1·60 45 ·10 10·03 3·10 0·82 49·44 0·36 \·02
0·03% 0·31% 0·45% 0·17% 0·03% 0·58% 0·23% 0·51% 0·80% 0·30% 0·42% 0·25% 1·79% 1·41% 0·54% 0·09% 1·95% 0·06% 0·05% 0·09% 0·04% 3·34% 0·49% 1·18% 0·14% 0·46% 0·39% 0·18% 0·47% 0·15% 0·41% 0·32% 0·71% 0·33% 0·61% 0·21% 0·40% 0·48% 0·29% 0·04% 0·53% 0·65% 0·51% 0·07% 0·61% 0·16% 0·90%
604
357·29
0-46%
440·1
302·78 196·21 24·81 15·37 11·38 67·14 16·10 43·54
1·87% 2·04% 0·97% 0·68% 1·58% 1·17% 7-45% 7·18%
62·78 8·91 2·07 0·09 10·30 21·80 )·87 0·27 4·28 3·32 3·34 1·71 1·21 I·S1 0·69 7·78 0·85 0·40 2·70
0·39% 4·46% 5·75% 0·05% 2·89% 4·18% 1·78% 0·79% 2·11% 0·85% 1·83% 2·22% 0·69% 0·52% 0·66% 1·57% 0·71% 0·22% 0·20'11.
17
354·0 369·3 88·1 152·8
Mt/ann
% vol.
m'/km /yr
I
2
N/A
j i J J
APPENDIX I
Table Al.2.
continued
Country
Re~ervoir
c~,_, Ic",,"" I km z t.Lm}'
Survey dates Start
I 42804
I
China
I
Total
I I
End
I
ChilUl China
Total-on3!) reservoirs 1981 80986 83 357 reservoirs with 460000 M.m' storage willI an aver.lge annual loss af 2·3%
Netherlands
HlIringvUet
Romania Romania Romania Romania Romania Romania Romania Romania Romania Romania Romania Romani:! Romania Romania Romania Romania Romania Romania
Pangamli V:lduri Balea
I Tumu
21 11 21 62
Total K...pelilny Hrieov Nosice
Slovakia
Total
Japan
29 dams 13.5 d:lms 169 dams 15S dams 152 dams
J.:lI:l!ln
207
!
S 8 36
I
Spain Spain Saain Spain Spain Spain Spain Spain SP:1!tl
202 26 149 123 570 308 138 86
23
, T:lb'lls - Buendia ; ucar - Alcon Ebro - La Tranqu:::ra Sur - R~(Ie"'o.do CaL:1.lonia -'"Riudecanll5 S
0·49
0·94%
0·75 3·60 450 11·04 1431 1·32 0·96 1·08
0·17% 0·16% 0·15% 0-22% 0·25% 0·22% 0·38% 0·51%
37·55
22-4%
6-3%
Tot;u-Gains
Spain Spain Spain Spain Sp;z.in
28·4%
\.,5
24·7%
6·9%
2737
Spain
14·969
1870 579
1520 2 84 2 3 35 4 7 4
Tagus Ta!
275 379
Jucar - .~eIlQS Iuco.r - Bus(!o Ebro G:tl!iu~uen luco.r - A:ouillo de Dan Bias NOlie -Alfilorios Gu:!d:llqtlivi~ - G'.lado.lmena
I
220
I
32 57 17 3 1311 8 4-
1955 1942 1965 1973 1969 1984 1956 1962 1954 1909
557-9 64·7 56·9
-92·742 -10·835 --4·962 -4·03 -3·724 -3·478 -2·014 -2 -I·n'::: -0·504 -t:!6
1957 1958 1960 1974 19i5 1960 1973 1959 1948 1954 1972
1960 19:.8 1979 1912
1927
330 690
H,)
1973
2:2
1350
3J..7
1960 1990 1969
9
1984 1995 1993 199[ 1977 1990 1985 1%9 1990 1983
32·4 85·5 320·1
198J 1976 1994
1992 1981 1993 1994 1979 1968 1984 1979 1991 1980 1994 19S0 1979 [981 1988 1994 1 1989
0 0·()75
0·121 0·154
o·un 0·189 0·285 0·326 0·561 0·629
0·691 0·723 0·7:58
0·14 0·08 0·28
1
I
1·66% 1 0·91% 0·77%
1
45·00
0·34%
-2·2%
-0·47 -O·5S -0·07 -0·19 -0·05 -0·01
-4·69'0
-5·26 j
-0·\9%
-3·20 -0·20 -0·18
~·7*
-Q~:!2
-8·3% -5-4% -3·0% -0·6% -0·7% -1·4% -2~1%
0·00 0·00 0·00 0·01 0·00 0-01 (}O2 0·02 0·03 0·02 Q.1O 0·02 ()·O2 0·05 0·01 0·02 0-1 I
0·0% 3·8% O·I%8-6% 5·6% 0-5% 6-7o/c 4·7%
12-8% 2·0% 1·2% 44% 25~3~c
0. 793
0·6%
0-807 0-S4t e·9! 0·965 0·994
\0·-1% 21·0% 9-1% 4·4% 10·6%
H),
0·3%
I
3
0-220/0
I
-0·29% -0·[0% -0·67% -0·15% -0·38% -0·10% -0·02% -0·21% -0·06% -0·03%
-IS·8%
I
I 3·21%
6,65
32-4% 58·1%
825
\994 1994 1994 19941994
I
!
4·825 2·089 8·055
t3 200 1112
Spain
1994 1994 1994
I
I
67·07
425 reservoirs studied in 1979 Jucar - Alarcon Guadatquillir - EI Pint:ldo S~ra - Sllntomera Sur - Guadalteba Guadalquivir - Ameen!! Guadalquivir - NegrJ.tin Douro - Barrio. de Luna Guudaiquivir - Los Hurones l'ugus - i'lorbollon Guadian:l - Gasset
Spain
1991193 i98!/90 1971180 1961170 1951160 1941/50 1931/40 tol930
1-3.5 0·52 0·96 043 0·63 [.63
Total 729 clams
Spain Souin Spain Spain Soain Spain SOllin
Spain
1992 1989 1992
3·9 6·91 13·05
25·6% 41-2% 44·0% 53·5% 35·8% 32·3% 21·0%
68·3
Jupan
Spain Spain
1957 1962 1963
1·2 0·132 H 54 :5-15 11·47
I
0·11 O·CO 0·21 0·C3 0·05 0·07 0·07 0·07 0·05 0·16
[ISH
Japan
Spain
2·14% 0·69% /-17% 0·31% 2·41% )·50% 0·64% 1·20%' 3-93% 4<12% 4<88% 1·97% 10·30% 4·40% 4·46% 3·98% 2·94% 2·62%
i7 322
23i7 2942 4911 5672 599 251 [88
27 dams
Spain
0·24%
0·3% 1·4% 2-9% 6·5% 9·8% 10·8% 22·7% 36·5%
35 dams 24 dams
Spain
!
53 443
JaPan Jacun Japan Japan lapan Japan
Sp:!in
2·41 0·14 0·03
1·0%
Ramnieu Raureni Govan ' Babeni
Romania
0·99%
49·3% 11·0% 28·0% 7·5% 50·5% 25·5% lQ.8% 24·1% 74·6% 74-l% 78·1%
n
Slovnkiil Slovakia Slovakia
800·CO
70
Dnesti
I
1·/9%
14·2%
3·3 0·55 2·8 0·03 4·37 1·3 0·8 1·3 1·32
!
% vol.
! 1500
1987 1981 19&7 1987 1986
1982 1983 1986 1986 1986 1986 j 1986 19S~ 1 1986 1976 1986 1974 1986 1977 1986 1975 1986 \978 1986
Notes
Mtlann
510·71
1999
I~I
~edimcntJtion
M.mJ/a,.'l11
I 28·1%
\970
1966 1966 \967 1968 1970 1973
Annuul rn"lkm!/yr
12013
1964 1965 1963 1963 1965
10 0 9 .5 7 5 2 2 I \3 j IJ tI
Garleni Lilteci Bucau Oiesti Cerbureni
I",
%
7
.5
Vanarori RaCOV!
1 Vol.
1000 I
I
T~,'
sedimcntation: M.m"
0·00% O':H% 0-00%
0·480/" 0·09% 0·02% 042% 0·23%
0·64% 0·07% 0·17% 0·14% 0-60% 0-04% 0·15% 0-40%
0-03
1-14% 0·:6%
0·25 0·05
0·02%
2·~%
157
J EVACUATION OF SEDIMENTS
] Table Al.2. Country
Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain Spain
,
continued Reservoir
Catalonia - Foix Guadalquivir - Gergal Guadalquivir - Bennejales Jucar - Forata Segura - Argos Ebro - Santa Maria de Belsue Segura - Anchuricas Sur - La Vinuela £bro - Santolea Ebro - Moneva Guadiana - Valuengo Jucar - La Toba Guadalquivir - CubiJIas Segura - La Cierva Catalonia - San Pons Douro - Burgomillodo Jucar - Sichar Guadalquivir - La Bolera Douro - Linares del Arroyo Jucar - Guadalesl Guadiana - Torre de Abraham Ebro - Las Torcas Guadalquivir - Cala Ebro- Pena Guadalquivir - La Minilla Guadalquivir - El Tranco de Beas Jucar - Maria Cristina Norte - Penarubia Jucar - Beniarres Sur - Guadalhorce Tagus - CazaJegas Guadalquivir - Bernbezar Guadalquivir - Puente Nuevo Tagus - Guajarez Tagus - Entrepenas Guadalquivir - Torre del Aguilla Segura - Cenajo DOUTO - Agueda Ebro - Cueva Foradada Jucar - Benageber Guadiana - Zujar Ebro - La Estanca de AJcaniz Ebro - Sotonera Jucar - Ernbarcaderos Catalonia - Sau Guadalquivir - Guadelen Tagus - EI Burguillo Sur - Conde del Guadalhorce Guadalquivir - Guadalcacin Ebro - Ribaroja Segura - Talave Segura - Valdeinlierno Tagus - Gru;el y Galan Douro - Santa Teresa Tagus - Riosequillo Guadalquivir- Bornos Ebro-Oliena Guadalquivir - La Brena Guadalquivir - Guadalmellato Guadalquivir - Pedro Marin . Segura - Alfonso XII Segura - Puentes Jucar - Contreras Ebro- Yesa Guadalquivir - Dona Aldonza Tagus - San Juan Ebro - Barasona Segura - La Fuensanta Ebro - Talarn - Tremp Ebro - Mequinenza Guadiana - Cijara
Spain
Total-Loss
Turkey Turkey Turkey Turkey Turkey Turkey
Caygoren Demirkopru Buldan Kerner Yalvac Kararnanli
158
Catchment: km 2
300 1058 190 117 1221
737 167 800 187 760 60 761
64 965 1334 469 965 1955 1665 559 375 439 1060 644 3495
323 16952 1281 268
763 31 I 1846 1858 1361 2694 465 1195 420 852 1042 2344 2181 3766 863 1250 1201
7456
Capacity: M.m;
6 36 104 39 12 13 8 170 49 10 20 II 21 8 25 15 52 56 58 16 60 9 59 22 60 500 23 12 31 134 11 347 289 25 8911 70 472 22 29 228 309 14 189 9 177 173 209 78 77 219 47 25 924 496 49 215 101 116 163 19 42 32 872 471
23162 71 235 258 1530 1670
Start
End
Total sedimentation: M.mJ
1928 1979 1958 1969 1970 1931 1957 1986 1932 1939 1959 1935 1956 1929 1957 1953 1960 1967 1951 1965 1974 1946 1927 1930 1956 1945 1920 1961 1971 1972 1949 1963 1972 1971 1956 1947 1960 1931 1926 1955 1989 1944 1963 1952 1963 1954 1931 1921 1917 1969 1918 1897 1961 1960 1956 1961 1959 1935 1965 1954 1916 1884 1975 1960 1955 1955 1932 1933 1916 1966 1956
1983 1985 1978 1983 1991 1980 1979 1994 1993 1984 1985 1980 1990 1987 1968 1989 1976 1979 1980 1989 1988 1979 1984 1989 1984 1990 1991 1994 1991 1991 1990 1994 1994 1982 1979 1992 1992 1980 1992 1992 1994 1971 1986 1983 1979 1977 1991 1991 1969 1982 1993 1984 1990 1989 1970 1990 1985 1991 1992 1977 1985 1985 1994 1986 1977 1992 1993 1991 1990 1982 1983
1·158 1·305 1·395 1·542 1·666 1·742 1·759 1·799 1·851 1·991 2·131 2·174 2·299 2·429 2·591 2·603 2·729 2'828 2·954 3·008 3·15 3·195 3·603 3·619 3·64 3·675 3·75 3·788 3-831 3·97 4·052 4·899 4 ·97 5·391 5·611 5·643 6·403 6·582 6·617 6·663 7·122 7·133 7·288 7·539 8-495 9·718 10·936 11·051 11·972 12·224 12·344 12·473 12·84 13·387 14·024 14·815 15·18 15·869 16·323 17-893 18·184 18·726 19·595· 20·78 220439 24·258 24·764 25 ·273 69·592 92-822 138· 111
19·3% 3·6% 1·3% 4·0% 14·2% 13·4% 22·0% 1·1% 3·8% 19·9% 10·7% 19·8% 10·9% 32-4% 10·4% 17·4% 5·2% 5·1% 5·1% 18·8% 5·3% 35·5% 6·1% 16·8% 6·1% 0·7% 16·1% 31·6% 12·4% 3·0% 36·8% 1·4% ].7% 21 ·6% 0·1% 8·1% 1-4% 29·9% 23·1% 2·9% 2-3% 51·0% 3·9% 83·8% 4·8% 5·6% 5·2% 14·2% 15·5% 5·6% 26·2% 49·9% 1·4% 2·7% 28·9% 6·9% 15·0% 13·7% 10·0% 94·2% 43-3% 59·3% 2·2% 4·4% 97' 6% 15·0% 34·9% 10·8% 27·0% 6· 1% 8·3%
0·02 0·22 0·07 0·11 0·08 0·04 0·08 0·22 0·03 0·04 0·08 0·05 0·07 0·04 0·24 0·07 0·17 0·24 0·10 0·]3 0·23 0·10 0·06 0·06 0·13 0·08 0·05 0·11 0·19 0·21 0·10 0· 16 0·23 0·49 0·24 0·13 0·20 0·13 0·10 0·18 1·42 0·26 0·32 0·24 0·53 0·42 0·18 0·16 0·23 0·94 0·16 0·14 0·44 0·46 1·00 0·51 0·58 0·28 0·60 0·78 0·26 0·19 1·03 0·80 1·02 0·66 0·41 0·44 0·94 5·80 5· 12
0·35% 0·60% 0·07% 0·28% 0·68% 0·27% 1·00% 0· 13% 0·06% 0·44% 0·41% 0·44% 0·32% 0·56% 0·94% 0·48% 0·33% 0·42% 0·18% 0·78% 0·38% 1·08% 0·11% 0·29% 0·22% 0·02% 0·23% 0·96% 0·62% 0·16% 0·90% 0·05% 0·08% 1·96% 0·00% 0·18% 0·04% 0·61% 0·35% 0·08% 0·46% 1·89% 0·17% 2·70% 0·30% 0·24% 0·09% 0·20% 0·30% 0·43% 0·35% 0·57% 0·05% 0·09% 2·07% 0·24% 0·58% 0·24% 0·37% 4 ·09% 0·63% 0·59% 0 ·12% 0·17% 4 ·43% 0·40% 0·57% 0· 19% 0·36% 0·38% 0·31%
859
3·7%
32·40
0· 14%
24·6 563·6 31·5 209·8 5·9 9·9
19·0% 69·2% 68·5% 56·3% 68·4% 39·7%
0·88 14·45 0·98 4·66 0·23 0·38
0·68% 1·77% 2·14% 1·25% 2·63% 1·53%
Survey dates
23323 1510 6590 180 2500 133 164
130 814 46 373 9 25
\1971
I ::~
1954 1973 )973
1999 1999 1999 1999 1999 1999
Vol. lost: %
Annual sedimentation m"/km 2/yr
583 2193 5467 1865 1697 2311
M.mJ/ann
Mtlann
Notes % vol.
J - -J
j
'-1 -1 1
1 j
1
]
J
1 J 1
J J 4
]
J 1
1
APPENDlX I
j
II
Table Al.2.
continued
C(luncry
Turkq
Turkey Turkey Turkey Turkey Turkey
Sdc:vir Cubuk- I Bayi ndir Hilfanli Kcssikkopru Alcinapa Sevhan Kmalkaya Cip Surgu
Turkey
TOlal
Pakistan
Tarbela
Germany Germeny
Saxonian Reservoirs Bal c!eney Ba\'aria - Forggensee Bavaria - Saalachsee (flushed) B :lv:ui~ - Sylven5rei nsee
Turk~y
Turkey Turkey Turk~y
G~r:nany
Germolny Germany Germany Malaysia Malaysia Malaysia
C,,",mom, 1c ".o; ,> km! M_ m~ _
R
~~
721 660 70 261 70 360 589 19 :!54 11 30 236 2i5
7'
14;;
940 I US
R i ng! ~l
Tunisia Tunisia Tunisia Tunisia Tunisia
Kasseb EI Kebir Mellegut: Nebhana Bezlk Chiba L:tkmess .sou Her1.'na Iuumine Lebna Sisi Saad Sidi Salem Silianu Marg'Jellil
Tunisia
Totai
42122
~rasri
UK
95 reservoi rs surveyed
l[~ iy
Csdore Valley S
r~l y
To~i
USA USA US A USA US A USA
USA
I
Total
Morroco Morroco Morreco Marreco MOCTecO Morroco Morroco
NakhLa Mohamed V Lalla Takerkousc
~~O!"TCCO
EI Kansera
Moue KiTalcabi Ibn Batouca My. YaU.>sef Mar.sour Ed D!lhoi Bin EI Ou idane Hassan Addakhi1 Y B Tachfine
Morroco Morroco Morr.x o SM3 Abdeilah Morroco Mor.oc o o EI Makhazine Mor.oco I Hnssanler ;':[c;m:;co 1 AI M:mira
I
508 8 150 4 32
I""
1965 :960 1966
1 I
1913 i959
I
236
33-9 12-6 3·5 3936-0 28-1 14-2 429·7 42-1
55-8% 215·0% 50-0% 29·6% 43-9% 35·8% 28·4% 37·1 % 39·4%
5377
59·7%
2900
20·3%
18·3
3·6'70
2408
1-6% 100-L% i·3%
7
1379 303 1464 3760 2364 752 519 1379 3i O 3388
65-8%
8 8
7 11 8 130 29 209 ~55
iO 110
1776
I
I I
12-38 23·74 232-;6 54· 12 6· 23 7·43 10·02 6·06 11 ·00 i ·50 6·00 9 1·80 56·1 0 38-50 44·00
607·64
I
15· 1% 91 ·3'70 70·1 % 62·6%
96·5% 94·5% 125-2% 83-9% 9·4%
5-8% 20·7% 43·9% 10·1'70 55·0% 40·')%
1203 875
109 980
107 49920 1710 45oiO
13
I
0·84%
0-06 0·22 0·09
;49
G' ll
I0
5
1 ' 15%
2-5t %
Q·4[0/0 I
0·20% I
165 275 550 1360
0·030 0·050 0· 101 0·249
4228 1200 514 1918 1952 3516 2465 3811 1282 1196 5051 603 18L 3365 4911
0-427 0·3252 5·29 1·64 0· 164 0·225 0·313 0·202 0-5 0·5 0·5 5-4 J.3 3·5 5·5
0·42% 0·1 2% ()'05% 0·19% 0·20% 0-35% 0·25% 0·38% 0-13% 0· L2% 0·5 1% 0·06% 0·02% 0·34% 0-49%
27-79
Oo();%
I
6
0· 10% 360 160
723 96 330 43 436 198 592 1484 369 320 51]9 807
I I! ·8
1
I 196 1 1967 1935 1927 198 1 1979 1970 1972
1974 197:2 ;07 1 i979 271 1987 2724 I 1976
O - I~
14·3 23·5 20·9 23-6 l804 19· 1
1066 2563
3-4%
0·02 0· 15 0·97 4·49 20·69 57-92 134-44-
18·6
424 1
3-')%
218-69
6·03 256·9 1
46·8'70 35·4%
26·5 64-66 6-96 5·6 22 62-88 99·32 20·96 16-49
li ·6O/o
0·23 - L1· 17 0·50 1·22 0·S7 0·58 1-10
22 97
4 ·59'9
4SB
i 1987 1990 1988 1980 1989 1989 1'790 1988
,990 1989 : 1985 1 1990 11 990 I :937
0 2 23 94
1
25·7% 19·3% 16·5% 1304% 9·1% 3·6%
19·6% 16·2% 1·3% 11·1 % 10·6<;;, 6-7% 5-7o/c
5·1%
I
33-99
4 ·'2L~
to
!
82-94
3-7% 3-0%
7
0·14 0·0 1 I
I
1
I 11 138 700 5331 29249 74499
6400
120·8)
770 154 93-6
34·2%
i
178
1 1.50 %
I 1969 1925 1954 1965 1960 1'765 1966 1968 1976 1983 1986 1981 1981 1987 1990
82 26 332 86 6
IMO !5 C:("J
1-31%
135
043
I 1996
no
1-64% 3-57')I> 147% 1·65% 0-90% 1·37% 0·83% 1·05 % 1·09 %
1·00
0·10
3·0%
Notes
Muann % vol _
0-10 98·40 0-85 0·44 9-99 1·56 0·1 1 0·93
I
I I
sedi m~n cJc ian
I M_m-'/~nl1
m-'/km%/yr
28·0
1· 1
Annual
%
1 1990 1980 19~
425 1132 94 12 39 345 2265 i6 454897 477 14B
4570 3780 9SCO 1821) 1670 28500
Val. la st:
3-644
I
I
I
TOl;ll sedilO~maci o n: M.m!
J.6
I
380 93
190 reservo irs 0 co 12 -33 m' 257 reservoirs 12·33 to L23·3 mJ 283 reservoirs 123·3 to L233 mJ 176 reservoirs L233 co 12 330 m~ 107 reser/oir5 12 330 to 123 300 mJ 69 reservoi rs 123 300 co I 233 (JOO m' 23 reservoirs> I 233 000 m'
USA
1998
I
Tunisi a Tunisia Tlmisia Tunlsia Tunisia Tu nisia Tu nisia Tanisin Tunisia
Italy
I
I
101 271 10 300 855 84 64 127 53 390 418 99 8950 18 250 1040 1110
T~~ ~si:\
1999 1999 1999 _1999 1999 1999 1999 1999 1999 1999
183 183 183 183
Ringlet Ringlet
To~1
1965 1936 1965 1959 1966 1967 1956 1972 1965 1969
1974
14300
Torn I
M~ lolysia
61 6 7 5980 95 3:! 120U 148 10 71
End
9006
I Ring let
Malaysia
Survey dates Smn
3093
3-22
I I
I
1·31
0-97 1-77 3-09 3·33 i -54
3·56% 2-00% 1·02% 0·81% 0·43% 0·23% 0·1 6%
I
0-20% 1·770/0
1-54% 0·52% 0-37% 2-02% 0· 13% 0·56% 0-66% I 0·12% j 0·3 6% 0·30% Q·35 ',t.. 0-38% 1·22% 1 0.28 %
159
EVACUATION OF SEDIMENTS
Table Al.2.
continued
Country
Reservoir
Mormeo Morroeo
ldriss ler I Abde1moumen
MOITOCO
Total
Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria Algeria
Ghrib DjorfTorba Bough7..0ul Bouhanifa 5MBA Chcurfas K'Sob Bakkhadda Foum EI Gharza Beni Bahdels Oued Fodda Ighil Emda La Chaffia Zardesas Sarno Foum E1 Gheiss Hamiz 1& 2 Mef\'rouch Reservoirs on the Atlas Mountains Reservoirs all flatter land
Algeria
Tolal
TaiwlIn Taiwan Taiwan Taiwan Taiwan Taiwan Taiwan Taiwan Taiwan Taiwan
Shihmen TenKee WuSheh Tsengwen Pai ho Gen Shai Pei A Kung Ticn Lu Liao Tapu Ku Kuan
Catchment: km 2
I
Stlllt
End
1972 1981
1986 1987
169 1120 55 73 235
1974
31
1939 1974 1952 1938 1974 1954 1965
1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986 1986
3680 1300
1217 216
135195
10351
9099 34118 28571
1414 246 1132 443 1277 873 288 1453
309 255 150 713 25 7 36
30·94 1·38
2944
1525
Indonesia Ind{lIlesia Indonesia
Karangkates Selorejo Wonogiti
2050 238 1262
343 62 730
Indonesia
Total
3550
1135
Sudan Sudan Sudan
Sennar - Blue Nile (dis of Roseirc~) EI Girbll- Atbara river (Nile lrib·) Roseires - Blue Nile
Sudan
Total
I
771·08
7·4%
112·6 33·6 22-3 21·4 9·4 12 13·16 8·1 20·5 6·5 92·3 34·9 7·3 16·5 0·7 1·4 6·65 J·4
66·7% 3·0% 40·9% 29·3% 4·0%
I
42·5% 4·0% 47·7% 10-3% 63·6% 34·9% 4·3% 14·8% 3·2% 43·6% 34·8% 2·0%
Egypt
Aswan High Dam
Ethiopia
Koka - Awash & Mojo riven;
CYPl1ls Cyprus Cyprus Cyprus Cyprus Cyprus Cyprus
Galini Petra Kalokhorio Lynlhroclhonda Lymbia Kophinou Akrounda
Cyprus Venezuela
2·21 0·23
)]1 85 21 59 213 41 198 122 433 i36 2098 2498 573 938 234 187 640 648 1100 2800
1982 1982 1985
930 1300 3354
167000 I
I 0-42%
J·OI 2·90 0·60 0·56 1·18
0·60% 0·26% 1·10% 0·77% 0·50%
0·28 0·03 0·49 0·14 0·93 3·19 0·50 0·27 0·34 0·03 0-48 0-03
0·90% 0·01% 1·14% 0·21% 0·64% 3·19% 0·29% 0·24% J·55% 0·93% 2·49% 0·04%
10·14
0·08% 0·60% 0·16% 0·31%
9
I
1·7%
4-41
I 0·39%
,
560 740 1127
60·2% 56·9% 33·6%
9·18 28·46 59·32
0·99% 2·19% 1·77%
2427
1 43·5%
96·96
I 1·74% i
2100
1·3%
80·77
0-05%
1986 1990 1985
1964
1990
I
1961
1981
I
I
]7·00
340
10
0·216
0·026
JI·9%
0·00%
Sanlo Domingo
427 307 1202 148 2350 1400 1082 4S85
"
1976
1978
0·575
15 50 8 143 109 32 229 II
1905 1913 1918 1932 1947 1937 1941 1960
1930 1959 1933 1961 J961 1965 1%5 1%7
0·188 0·92 0·138 8·5 0·3 3·8 41·83 1·03
1·3% 1·8% 1·7% 5·9% 0·3% 11·8% 18·3%
56·706
9·5%
I
I
I
I
19·2%
9·4<;i-
24
17 63 125 15 125 364 118
0·29
9·58%
0·01 0·02 O·OJ 0·29 0·02 0·1.:f. )·78 0·\5
0·05% 0·04% 0·12% 0·21% 0·02% 042% 0·78% : ]·35'k
2·42
i I
i
i 040%
1
1
I
166
1 1 1
I
19
TOlal
597
8
2·04 0·10 2·27
4·10% 3·90% 3·03% 1·14% 7,13% 14·33% 5·55%
I
I
1
3·0% 1·9% 1·2%
8·2% 7·8% 9·1% 5·7% 28·5% 43·0% 16·7%
mo
I
0·67%
I
1·67
0·00192 0·0024J 0·00628 0·00232 0·00739 0-00159 0·00370
I
1
0·67% 0·44% 0·72% 0·59% 1·51% 1·44% 1·43% 0·01% 2·54% 2-48%
0·023 0·031 0·069 0·041 0·026 0·004 0·022
I
1
0-49%
!
26 37 24 9 33 12 26
l..esna Pikhowkc Lubachow OtmuchQw Turawa Porabka Roznow I Myczkowce
1
0. 18 % 0·11%
1
10·22 I·J6 9·07
1925 1964 1966
5584
I
% vol.
2·06 1·12 j·07 4·23 0·38 0·10 0·52 0·00 0·23 0-42
I 1977 1970 1981
IMtlann
43·27
13 2700 1900 4900 8800 14300 9400 16300 30 2200 600
1938
NOles
I
15·5%
409 1963
1998
i
M.mo/ann
m>lkm /yr
2·5% 0·6%
4
Total
Annual sedimentation 2
9 17
Total
160
1953 1939 1972 1937
Vol. lost: %
M.m~
2633 763 592 219 481 27 11 32 8 104 708
Taiwan
:
1948 1978
203 43 63 145 100 171 111 22 3 19 70
Taiwan
Poland Poland Poland Poland Poland Poland Poland Poland
Tota! sedimentation:
Capacity: M.ml
1 1
II
i
1
r" .1
1
i
I
APPENDIX I
Table Al.2.
continued Catchment: kml
I
I [
ICaDadty: I Survey I
~l.m'
! I Start
Avisio P"rnegg Pt.mtebbn Steyerdurchbrudl Tar:ento
Austria Austria Austria Austria Austnn Austrit, Austria Austria
Wet;:man Margarilze - study rt!f. further
ITotal
956 6150 10 575 62 324
0·85 0·[5 0-60
8177
4·10
Switzerland Switzerland
Kallna..:h Petolle.
Switzerland
Total
2621
New Zealand New Z:!\]and
Roxburgh MutOlhina
8826
I
28+1
5740 34
i7:
::[~C~land ~lomi Total
S'lutn Africa South Africa Suuth Africa SO(1tb Africa South Africa South Africa South Africa South Africa
Afhasini
End
M.il,.'
1925
, 189() 1927
1892 ISS]
1908 1884
II: ~~~ !~~~ I 1913
1919
: 1872
1886
1961
1979
[966 193.;1.
1980 1959 1979
Vol.losl; %
100·0% 70·0%
0·35 0·25 0·71 0·15 0·6
84·0% 100·0% 100·0%
3·81
93·0% 55·6% 100,0%
Annual sedimentation
m)/km~/yr 261 3704 1545 0·03 193 1852
12$
56
M.m}l:mn
Notes
I Mtlanll
% vol.
0·25 0·23 0·02
12·:50% 46·70% 3·80%
0·0\ 0·60
8·00% 100·00%
1·14
2HI%
0·17 0·07
945%
7·13%
714%
IOH 49·7
47·5 0·1
i 1939
~~:
26 liS
A1!cmanskraal Bt:ervld Boegoeberg Bon Accord Bronkhorstpruit Bulshoel(
9t
'9
Darlil1l!:ton Drid -
years
26·28 2-982 9·5 0·0296
ill
2~~~~~:
1923
1995 1995 1986 1995 1995 1986 1995
1917
1995
1935 1922
1995
9·8
1995
139·9
1995 1995 1995 1993
5
1975 [929
4
1925
S9 6
1950
to
E!!monc Elnndsdrift. Erfenis Aornkraal Gamlmpoort
1952 1960
36
124 169 to
Somh Afri:a
7 208 50
42
i973 1937
1977 1960 1957 1969
H:u·tbefspoort
Hazdmere K:Ilkfolltein Kamn1
91
19~4
195 18 355 36
]9!5 1977
7
[(l:l%rric
Kommandodri ft KGPpi~~
Kiommenellenboog
74 41 9 73
10.5 14
Leeu Gllmka
]g4
Marico NooU!!edacht Phalaborwa ba.crll"e Pietersfont.t!in
'17 79 9
PUnl!alaDCOn
S~'u(h.Ar.';ca
Poo;;jie'
E'rinsrivier Ril!tvld Rust de Wimer S(ompcirlf.
11 27
55 33 2536
V:lU Rr:-:e'.'e!c?u$ Welbedacht Wentzel Windsor
SQuth Africa
To!:!!
1923 [954
1960 1956 1911 1955 1970 1924 1959 1939 1933 196::! 1966 1963 1917 1933
1934 1965 1990 1938 1936
Vaaiha.ctZ
S"uth Airica SO'Jtn Africa SQuth Arrie:!
1938
196& [973
"
1995
1995 1995 .1995
5673 22
South Afric:l SoulhAfric:a South Africa South Africa SDuth Africn South Africa South Africa South Africa
Toml
1
1
~:~:
20·0% 21·0% 19·2%
165 75 66 22
[4600
I
1·39%
g:~~~~ I
0·43%
0·80% 0·53%
0·0007 2·0537
j
I·O!% I
II'
1·3% 263 0·25 0·l4% 0..()9% 396 3·14 0·8% 1 7920 36j0 9 rellrs I 28·22 -----------4------+---~--~~--+_------+_--~----~----~--_4----~--0·09% 382 0·8% 3-39 30·471125 8870 ! 3329
Bmzil
SuutnAfrica S·~'I.!th Africa SomhAfricn South Africa Suuth Africa South Africa South Africa South Arne:! SoudlAfrica SilUlh Afril.:a South Afric:t South Africa South Africa South Africa South Africa S')llth At'ricn South Afric:! SuuchAfric:! Soui.hArrica SourhAfrica South Aftica Seurh Afric:! SomhAfrica South Africa South Africa South Arrka SOllth Africa Scutt: Africa
! 18~2
j
1360 J261
~~: ~~::~~~ I ~~:~~~
B.'aZi!
2·()O 0·50
dUlt:$
112 '6
19:!.5 1973 1934
5
1950
1995 1995 199.:5 1995 1995 1995 19~6
1995
J995 1995 1995 1995 1995
1986 1995
4·1 4C.~
6·643 14·J
\·3 6·8913 I I
3·5 2·5 ::!3·1 16·g 10·1 5.:l..5·9
2·9 41·2
H 0·9 0·435 1.:1.·7 12 2·3 114
11·1%
0·66 0·44 0·39 2::!·75
24·}%
9-6% 13·2% 45·3<;7.
1·13%
2-19% 0·62% 1·98% 0,32%
0·38% 0·93% 040% 0·49% 0·64%
0·15%
19·9%
0·6" 0·05 0·02 002 0·38
0·18% 0·14'1& 0·30% 0·29% 0·51%
29·5%
0·14
0·35%
24·59£·
0·06 0·50
0·61%
12·2% ij,}'e
16·9'7~
55·1% 5·9% 11·9% 0·7%
6·4
0·11 0·58
0·37% 0·66% 0·66% 0·61% 042% 0·33% 0·24% 0·27% 0·13%
0·29 0·34
7·3 22-656
Q·033 210
0·16 1·92
3;1·4%
71-2%
5·) 0·64.;1.3 1-1968
0·22 0·02 0·19 0·01 0·01 0·23 0·06 0·14
7j.S
3-2
0·60
7·9% 82·8% .;!.g.!% 36·0% 35·7%
10·3%
3 0·6 55
1996
174'* 20·7%
34.:2% 10·3% [0·0%
0·5516
1995 1995 1995 1995 1995 1995
11·7%
20
1995 1995 1995 1995 1986 1986 1995
39·9% 29·7%
6·1 36·6
1985 1995
16·0% 23·0% 7·3%
3J,,1~;
22,)% 2·2"% 90·[% 66·3';7~
5·2t;~
4·49'c IHj% 0·1%
1·07 0-22 0-48
0·05 0·02 0·10 0·02
2·50 0·06 0·07 0·01
0·02 0·21
0·01
1·90%
0·68% 1·02% 1·53% 0·13% 0·19% 0·03% [·20%
0·&3% 0·10% 2·82% 0·8:5% 0·10% 0·08% 0·39% 0·02% 0·15% 0·65%
8·3~:tt
3·6~
41-(;
3g4%
0·71
3!·4
39·7%
O·57'V~
967 1·3
0·45 440
:m·J%
{J·O::!,
O·3Y'lc
~3·1';·i
O·OS
1·35~1:
1·07
0·66';:;'
1577
)·93'ij,
0·34%
161
EVACUATION OF SEDIMENTS
Table Al.2. Country
continued
Reservoir
Catchment: km 2
ICapacity: M.m)
Survey dales
I
Kenya Kenya Kenya Kenya
Kindaruma on Tana Kamburu on Tana Gitaru on Tana Masinga on Tana
9810 9520 9540 7335
Start 1967 1974 1978 198]
7 150 20 11560
I
End
Total sedimentation: M.m3
IVol.%lost:
11·7
1980
I
Annual sedimentation
7·8%
m;/km'/yr
M.m 3/ann
246
2-34
I
Kenya Portugal Portugal Portugal Portugal Portugal Portugal Portugal Portugal Korea
Santa Luzia Montargil Campilhas Burgaes Idanha Arade ROKo
Lake Kariba
1
50 164 22 0 79 29 96
I Total
Lake Cabora Bassa
1960 1961 1966 1962 1963 1973 1980
1·56%
440 85
7
26 years
0·02 0·02 0·03 0·00 0·15 0·11 0·07
0·03% 0·02% 0·15% 0·26% 0·19% 0·39% 0·08%
0·5% 0·0% 1·6% 5·8% 3·2% 7·1% 1·0%
&239
1-4%
0·41
0·09%
0·77012
11·1%
0·03
0·43%
I
I
116000 M.m' - no re-survey data available. but estimates = 1600 to 16000 yrs to !ill dead storage
650000 1000000 (inc. Kariba)
I
310 21 289 33 404 497 215
0·25 0·05 0·347 0·019 2·54 2·05 0·983
Sedimentation may I
in
I
NOles: 1. Reservoir is dis of Lake Wular 2. Catchment area not permanantly covered by snow only 3. 1·2 M.mJ of sediments dredged 4. Values calculated from hydrographic survey results 5. Approlt 50% of sediment is polluted 6. Example of the effect of deforestation 7. Typical losses from 95 reservoirs surveyed 8. dis of cascade of 3 dams & 100's of check dams - dredged at a rate of 600 000 m'Janl1 since J985 9. Flushed since 1955 at an average of 328 000 m·l/ann 10. Reservoir will not be able to function after 2010 unless rules changed 11. Reservoir Hushed in 1978 for 3 weeks clearing 620000 m3 of sediments
162
% vol.
I
11737
1
Beaggog
Notes
Mtlann
I
I
I
I
I
I
Appendix'2o Numerical model case study A2.1. TARBELA DAM, PAKISTAN The Tarbela dam, a key component of the Indus basin scheme in Pakistan, was completed in 1974 for the purpose of irrigation and hydropower. With an annual sediment inflow into the reservoir of over 200 million tonnes, the live storage is being rapidly depleted and liPless action is taken hydropower generation could cease within a decade, with irrigation releases declining over the next 30 years. The overall purpose of the feasibility study ca..rried out in 1998 by HR vVallingford and TAMS UK, was to determine a strategy for the economical preservation of h1e assets at Tarbela on a more sustainable basis.
A2.1.I. History Tarbela dam was constructed in the 1970s to help regulate the seasonal flows of the upper Indus both for irrigation of the Indus plains downstream and for the generation of hydropower. It is still, 30 years on, the only major storage reservoir on the Indus and, as such, plays a key role in the provision of dry season releases of water for irrigation. Tarbela irrigation releases amount to 11 600 M.m3 , or 50% of the WAPDA (Water and Power Development Authority of Pakistan) total, with a corresponding agricultural revenue of Rs 2·8 billion. In addition, with an installed capacity of 3478 MW and a finn electrical energy of 14·8 GWh/yr the Tarbela dam provides 32% of both Pa.1dstan's total power and energy needs with a corresponding annual revenue of Rs 6 billion. It is, therefore, a strategic national resource whose continuing future efficient operation is of paramount national interest.
A2.1.2. Sediment The Tarbela dam impounds the waters of the Indus, which carry a heavy sediment load. This is the case particularly in the spring and summer when the rnelting snows cause heavy erosion of the uplal'1d catchment. I\t1ost of the sediments brought down by the Indus are trapped in the Tarbela reservoir. Thus, with an average annual sediment inflow into the reservoir of approximately 240 Mt per year, the live and dead storages of the reservoir have diminished by 16% and 21 % to 9000 1V1.m3 and 1360 M.m3 respectively in 1997.
163
EVACUATION OF SEDIMENTS
The accumulation of sediment within the reservoir causes two major problems. • A loss of live storage which results in a gradual reduction in the regulated yield of the reservoir. This in tum results in a reduction in the water available for agriculture and a reduction in the firm energy available from the project. • The physical effect of sediment, which includes the risk of blocking the outlets, particularly in the event of an earthquake, and erosive action of sediment laden water on the dam's outlet works and turbines, which will result in increasing maintenance costs to the point when the scheme will eventually become inoperative. Unless remedial action is taken, the reservoir will be largely filled up with .sediment oy the year 2030, giving the project a useful life equal to that estimated at the time of the original design. However, in view of the size of the investment already made in the Tarbela project, and of its critical national impol1ance outlined above, there is clearly a need for a programme of future actions to maximize the economic returns from this resource.
A2.2. ENVIRONMENT
A2.2.1. Hydrology/climate The source of the river Indus is situated in the Tibetan Plateau, at an elevation of 5500 metres above sea level. From there it flows across some of the highest mountain ranges in the world before emerging onto rain-fed lower-lying country. Downstream of Tarbela, the Indus flows along a broad valley until it reaches Attock Gorge, some 51 km downstream. On leaving the gorge the Indus flows onwards for a further 1600 km to its mouth on the Arabian Sea. The Indus basin upstream of Tarbela Dam, an area of 169 650 km2, consists of two distinct hydrological regions. Over 90% of this basin lies between the Karakoram and Himalayan mountain ranges; the meltwaters from the snow and ice that cover approximately one quarter of this mountainous portion of the basin cont1ibute a major part of the annual flow reaching the Tarbela Dam. Seven of the ten highest lnountains in the world reside within the catchment. The remainder of the basin, about 11 700 km2, lying immediately upstream of the dam, is subject to monsoon rainfall, primarily dllling the months of July, August and September, the run-off from which causes sharp floods of short duration that are superimposed on the slower responding snowmelt run-off. Climate in the Indus basin is subtropical and semi-arid in the headwaters. It is divided to form two distinct seasons: kharif (summer), extending from April to September; and rabi (winter), coveling the remaining months. The annual rainfall averages around 900 mm of which two-thirds fall between June and October.
164
:f
r ]
APPENDIX 2
-l :]
1 t l
14 000
I
12 000
j
Rainfall
10 000 IJl
~
E
---I
8000
iti
3: 0
E
6000 4000
2000 I-
J J
J
A
s
o
N
o
Month
Figure A2.1.
Case study - inflow hydrograph to Tarbela reservoir, Pakistan
A2.2.2. Indus River flows The average Tarbela inflow hydrograph, see Figure A2.1, shovvs t.'1e contribution to the run-off made by rainfall and snowmelt. It is estimated that the monsoon (rainfall) contribution to the total run-off is approximately 10% of the whole. The mean annual inflow into the Tarbela reservoir is 81 km3• Variability of mean annual flows from year to year is small, wiLl. a coefficient of variation of only 15%.
A2.2.3. Sediment inflows The mean annual sediment inflow is 240 NIt. Of this, approximately 40 Mt of very fine sediment passes through the reservoir and· 200 Mt -. of the coarser fractions deposit within the reservoir. The vast majority of the annual sediment load enters the reservoir during the high flow season, l\1ay to September.
A2.2.4. Sediment deposition Since the reservoir was first impounded in 1974, a high proportion of the annual sediment inflows into D.1.e reservoir have been deposited to fann a delta which has been advancing towards the dam. This has been monitored by detailed surveys which have been carried out annually since 1979, covering the whole of the reservoir. The delta profile is sensitive to the way the reservoir is operated, in particular to the minimum pool level and the length of time the minimum pool level is
165
EVACUATION OF SEDIMENTS
maintained. During the initial years of operation, until the year 1988, the reservoir was drawn down close to the minimum operating level of 1300 ft every year and, consequently, the delta advanced towards the dam. As the rate of advance of the delta downstream is related to the extent to which the reservoir is drawn down, the policy since this time has been to operate the reservoir with a higher minimum water level. This, however, has encouraged the deposition of sediment further upstream, in the middle reaches and within the live storage.
A2.2.S. Numerical sediment modelling A numerical model was used by HR Wallingford to simulate reservoir sedimentation in the Tarbela reservoir. The model took the original reservoir cross-sections and a 60-year sequence of water and sediment inflows into the reservoir. It computed the sediment profile at each cross-section using equations that relate sediment movement and flow for a range of sediment sizes. The output included: • • • • •
changes in bed topography as sedimentation deposits and erodes volumes of sediment being deposited and eroded changes to live and total storage curves as sedimentation progressed daily discharges passing to the downstream reach sediment loads passing to the downstream reach.
A2.2.6. Verification The model was verified by simulating the observed sediment deposition from the time the reservoir was impounded to 1996 and comparing the profiles predicted by the model in 1996 with those observed. The model gave excellent predictions in the 20 km immediately upstream of the dam, see Figure A2.2.
A2.2.7. Reservoir operation policies Several scenarios were developed for the operation of a flushing system at Tarbela and appropriate runs of the model were carried out to explore the influence of different assumptions about the way in which the reservoir and the flushing system should be operated.
A2.2.B. Reservoir flushing Five model runs were carried out to simulate reservoir flushing, in which the flushing level, the flushing period and the date flushing commences were varied. The results show that for the conditions pertaining at this particular site: • flushing provides a substantial long-term live storage with only a small annual reduction
166
APPENDIX 2
455 435
E
c:
.2
415
ca> Q)
m
395
"0
--1974observed _. 1996 observed ·
Q)
co 375
- - - - 1984 mode! ••••••• 1996 model
355 335
0
10 000
20 000
30 000
40 000
50 000
60 000
70 000
80 000
90 000
100 000
Distance from dam: m
Figure A2.2. Pakistan
It
It
Case study -
veri,;ication of numerical model, Tarbela reservoir,
low-level flushing is more effective than high-level flushing flushing over a 3D-day period is more effective than over a 20-day period.
A2.2.9, Typical numerical modelling results Figure A2.3 shows the throughput of sediments to the year 2056 without the introduction of a flushing system. It assumes an operating system which gradually raises minimum water level year on year until the live storage of the 500 c::::J Silt 1 c:::J Silt 2 ~ Slit 3 ~ Silt.:;. :::::::: Siit 5 ~ Sand 1
~
Sand 2 _
DiSChaigel120 000
450 100000
r
i'! ::~
,.., E
80 009
c::
~
'E (Ii
~
60 000
250[
~
(3 CIl
]
200r
g
150
'6
40000
c
-<
ca::l C
c:
c
-<
100 50 OL-~
__-L__~__~__L-~__~__~__~~~~__- L_ _~____~~_ _~I 0
1975
1985
1995
2005
2015
2025
2035
2045
2055
Year
Figure A2.3. Pakistan
Case study -
sediment throughput without flushing, Tarbela reservai,;
167
EVACUATION OF SEDIMENTS
500
120000 c:::::J Silt 1 c:::J Silt 2 E::El Silt 3
Sand 1 _
Silt 4 c=J Silt 5
Sand 2 -
Discharge
450
c
400
'E
350
~
-0
Qi ':;"
100000
80 000
'"E ~
'E
300
a.;
C 250 OJ E
60000 . ~
200
'5
'5 OJ
~
u
C/)
C/)
(i'l ::!
c c
«
40000
150
(i'l
::! C C
«
100
20000
50 o~~--~--~--~~--~--~~~~--~--~~--~--~--~--Uo
1975
1985
1995
2005
2015
2025
2035
2045
2055
Year
Figure A2.4. Pakistan
Case study -
sediment throughputs with flushing, Tarbela reservoir,
reservoir is reduced to approximately 20% of the value at first impoundment. The sediments passing through the reservoir are mainly the finer fractions and these either pass through the machinery during the generation of power or over the spillways towards the end of the flood season. Figure A2.4 shows the equivalent results with the introduction of a low-level flushing system and an appropriate operating rule for reservoir levels. In this case, the long-telID throughput of sediment matches the incolning sediment quantities and the reservoir live storage settles to a value of approximately 50% of the original live storage. 14000
____ D1 -1350 ft min
-L:r12000
D5 -
~ D4 -1500 ft min - 0 - F4 - High flush from 2021
.......
1350-1500 ft min
--0- F6 - Flush from 2006
-+-
10000
85 -
11 10
F7 - Short flush from 2006
9
8asha in 2016
8
.......................... . . :. . . . :. . : i
4:·ir.···~·~.·.-;-;-:-F:.~-·~·-.~.-.o-.
~
u.
~
0- - 0 -
7
~
,I
6000
.....~
~ -
4000
.() -
-0
-
-
0-
-
0 -
4
-0- -0- -
-----~-----x------x-~~~..~--~--~---..~ ~ -&'--A-=--
2000
~
::J
-1
3 2 1
O~
__
1996
~
__
2001
~
____
2006
~
2011
__
~
____
2016
~
2021
__
~
__
2026
~
____
2031
~
2036
__
~
__
2041
~
____
2046
~
2051
__
~O
2056
Year
Figure A2.S. Pakistan
Case study -
II
..-1
prediction oj Juture live storage, Tarbela reservoir,
I j
168
~ j
J
APPENDIX 2
Figure A2.5 shows the live storage volumes attainable with the chosen flushing regime, compared with some of the other operational policies that could be pursued. These results were subsequently analysed along with the costs and the benefits associated with each scenario in order to determine the most appropriate course of action at Tarbela.
169
Appendix 30
Flushing case studies A3.1. MANGAHAO RESERVOIR (New Zealand, 1924) The Mangahao reservoir and 20 MW hydropower scheme "vas constructed in 1924 on the Mangahao River, in the south of New Zealand's North Island (Jowett, 1984). It was recognised at the outset that sedimentation of the reservoir would be problem, and a second reservoir was formed upstream to act as a sediment trap. As it was originally intended to continue building more reservoirs upstream for sediment trapping, no specific sediment management facilities were installed in either of the reservoirs. No information is available regarding the catchment area, basin size, annual run-off and sediment inflow.
A3.I.I. Sedimentation By 1958, the original live storage capacity of the reservoir had-reduced by 59% and the intake structure and screens were aLTJ10st buried. In addition, the highlevel outlets of the upstream sediment trapping reservoir had failed, with the result that water and sediments were being discharged through the low-level diversion tunnel. The problem became increasingly serious by the mid 1960s, tr-l1'eatening the continued operation of the power station.
A3.I.2. Hushing In 1969, it was decided to attempt sediment- flushing at Mangahao reservoir through the low-level diversion tunnel. The tunnel had not been used for 25 years, owing to problems with the gate, but had been used routinely between 1925 and 1944. For a period of 24 hours 'nothing happened, then on the second day silt began to extrude from the tunnel and the reservoir emptied, leaving a crater-like depression in the 13 metres of sediment which had overlain the tunnel entrance' (Jowett, 1984). A month of flushing, using water released from the upstream reservoir, resulted in 0·88 M.m3 of sediment being scoured from the reservoir basin, equating to 75% of the sediment that had accumulated since 1924. It is reported that large logs and tree roots had to be cleared from the entrance, but that much debris passed through the tunne1. Since the initial flushing operation the reservoir has been emptied and flushed annually (Jowett, 1984). The power station is closed for three weeks for the flushing operation, reducing the annual energy produced by 4%. The annual
EVACUATION OF SEDIMENTS
flushing of the reservoir has resulted in the removal of most of the sediment from -Mangahao Reservoir and a considerable amount from the upper reservoir. No information is available regarding the flushing discharges used, although the fact that -the original diversion tunnel is used suggests that the available discharge capacity is generous.
A3.I.3. Downstream impacts Large banks of sand and silt formed immediately downstream of the Mangahao reservoir during flushing. Studies have shown that the effect of flushing on substrate and invertebrates was minor, but noted the temporary increase in turbidity that was disturbing to recreational users.
A3.2. GUERNSEY RESERVOIR (USA, 1927) Guernsey reservoir on North Platte River, Wyoming, is used prinlarily for . irrigation, but also provides hydropower. It is impounded by an earthfill dam, which was completed in 1927 . The dam height is 41 nl, the reservoir length is 23·5 km and the original storage capacity was 91 M.m3 (Jarecki and Murphy, 1963), representing only 4% of the estiInated mean annual inflow.
A3.2.1. Sedimentation From the time of construction, about 66% of the total catchment fell within the catchment of the Pathfinder dam, reducing sediment inflows to Guernsey reservoir substantially. After the constluction of Glendo dam in 1957, only 40/0 of the catchment was expected to be contributing significant sedilnent inflows to Guernsey reservoir. Until 1957, the reservoir was subject to a high sedimentation rate, losing 39% of its oliginal capacity over a period of 30 years, with deposits cOlnprising 170/0 sand, 61 % silt and 220/0 clay. The maximuln depth of deposit repol1ed was about 12 m.
A3.2.2. Flushing Partial draw down (by 12-13 m) was carried out at Guernsey reservoir annually between 1959 and 1962 and data were collected to determine inflow and outflow rates and sediment movement within the reservoir. No definitive details are available in the references concelning the ·bottom outlet or other flushing facilities, although Morris and Fan (1997) suggest that the 'overflow spillway' was used, which is possible if there are large spillway gates. The elevations of the power intake and the sediment deposits in the vicinity of the dam are about 10 In above the original bed and the aInount of draw down would be consistent with an outlet at about this elevation. Flushing discharges were typically in the range of 120-140 In3/s, cOlTesponding to about double the nlean annual inflow.
172
J
APPENDIX 3
Although sediment was scoured from the upper portions of the reservoir during the four years of drawdowil, n10st of this was apparently redeposited in the lower part of the basin nearer to the dam, and the suspended solids concentration in the water discharge from the reservoir never exceeded 0·8 gil. From the inflow and outflow data during the period 1957-62, it was estimated that only 144 000 m 3 of accumulated sediment was relTIoved from the reservoir basin, the equivalent of less than 0·2% of the original capacity. The long time before first flushing was probably a factor in reducing the erosion of the deposits (Atkinson, 1996). It was concluded (Jarecki and Murphy, 1963) that, with future annual drawdowns following a similar pattern, only about 2% of the original capacity could eventually be recovered. As sediment inflows had been severely reduced by t.1}e construction of the upstream dams, this appeared to represent a satisfactory state of affairs. .
AI2.3. Downstream impacts No problenls are mentioned in the references, probably because the outflow sediment concentrations and efficacy of flushing are low.
foJ.3. ZEMO-AFCHAR RESERVOiR (Former USSR, 1927) The Zemo-Afchar hydropower reservoir, completed in 1927, is located just downstream of the confluence of two rivers. No data are available with respect to the original storage capacity of the reservoir, although the basin length is given as 8 km along one tlibutary and 1·8 km along the other (UNESCO, 1985).
A3.3. I. Sedimentation During the first two years of operation, the storage capacity reduced by 22% a year and during the following eight years a further 32% of the capacity was lost. Only 4% more was losfduring the next 18 years (1937-54), suggesting that an equilibrium had been reached.
A3.3.2. Flushing No details of the flushing facilities are given, but they are sufficient to pass over double the Inean annual flow when the reservoir is emptied. Plior to 1939, the reservoir apparently operated with a lilnited annual draw down of 2·3 ill, but this was not effective. Between 1939 and 1966, it is reported that 38 flushing operations (between one and four per year) were undertaken with full drawdown. A wide range of results were obtained, due to influences such as the duration of flushing and the amount of accumulation
173
EVACUATION OF SEDIMENTS
1
1
between flushing operations (Morris and Fan, 1997). Peak concentrations were reported in only two events, but were very high at 270 and 370 gil. Each flushing event comprised two stages: partial drawdown, as the reservoir was being emptied, followed by total drawdown, with flows along the bottom of the scoured channel through the basin. The duration of flushing varied from 8·5 to 65·5 hours, with a mean of 18·5 hours, and was canied out mainly in the month of April, Mayor November. The volumes scoured each year ranged between about 0·5 and 2 M.m3, with an average of approximately I M.m3 • This is substantially less than the reported mean annual sediment inflow, suggesting that most of the annual sediment load passes through during routine operation. The data presented by UNESCO (1985) suggests a reduction in the total volume of sediment contained in the reservoir since full draw down flushing began, although there is some ambiguity in the plotted data. General conclusions drawn (UNESCO, 1985) from the flushing operation at the ZelTIo-Afchar reservoir were:
l JI j
! ~ i
J
J
• an optimum flushing discharge between 400 and 500 m3/s produces the most effective evacuation of sediment (higher discharges cause a greater water depth at the dam, reducing the effectiveness of flushing) • during the process of sediment flushing, the most active erosion occurs in a period of 8 to 10 hours after effective erosion starts • when the effectiveness of flushing starts to fall, it can be restored by temporarily raising the water level for a short period.
J A3.3.3. Downstream impacts No information available.
A3.4. JENSANPEI RESERVOIR (Taiwan, 1938) The Jensanpei reservoir was built in 1938 for the purpose of water supply to the sugar cane industry. The reservoir had an original storage capacity of 7 M.m3, which is probably of the order of a third or less of the average annual inflow to the reservoir. This was raised to 7·7 M.m3 in about 1942 and to 8·1 M.m3 in about 1958, presumably by raising the impounding level.
A3A.I. Sedimentation The erosion rate in south-west Taiwan is particularly high, due to climatic conditions and geological conditions of soft and erosive rock. In an 18-year period from 1938 to 1955, the storage depletion at the Jensanpei reservoir due to silting was 4·26 M.m\ an average annual loss of 3·4% of the storage capacity . (Hwang, 1985).
174
- I
J
J J
J J J
J J
J
APPENDIX 3
A3.4.2. Flushing L., 1955 a 1·5 m diameter flushing tunnel was built through the base of the darn, as a result of the sedimentation problems, and annual flushing commenced. Flushing is arranged by emptying the reservoir between May and July and allowing free flow through the reservoir. This suits the water'demands, as sugar mills do not use water between May and October. Between 1955 and 1980, the sediment volume contained in the Jensanpei reservoir remained almost constant, showing that the adopted flushing regime was highly effective, albeit retaining only about 45% of the enlarged capacity.
A3.4.3. Downstream impacts No infonnation available.
A3.5. NAODEHAI RESERVOIR (China, 1942)
The N aodehai reservoir is a flood detention reservoir situated on the Liuhe River in an arid region in the north-eastern part of China. As the main purpose of the reservoir was for flood control, it was built initially with ungated outlets near the river's original river-bed. Control gates were subsequently installed (apparently 1970) to preserve clearer water for irrigation in the non-flood season. The reservoir was designed to attenuate a peak: inflow flood of 3500 m3/s to 1640 m3/s (UNESCO, 1985). The original design flood storage volume was a little less than the mean annual inflow.
A3.5. f. Sedimentation The heavily silt-laden Liuhe River has an annual average sediment concentration of 77 gil, so that detention floods resulted in the deposition of sedilnent deposits on the floodplain wit.~in the basin. High volumes of deposition were reported during floods in 1949 and 1963. Table A3.1below gives some data for flood peaks which occurred over a few successive days in 1963.
A3.5.2. Fiushinoo The flushing wpich occurs in a reservoir of this sort is essentially uncontrolled. Although there is some scope for control since the installation of gates on the outlets, no information is available to judge if their use has had any on sedimentation in this case. UNESCO (1985) shows the reported variation in available storage capacity in N aodehai reservoir, reducinQ: from 168 M.m3 1942 to a minimum of 97 M.m3 3 in 1950 and varying up to about 134l\1.m in 1972. Overall, the storage loss has ranged between 20% and 42% and it appears to be dominated by massive deposition in the largest floods, followed by a period of progressive erosion.
175
J EVACUATION OF SEDIMENTS
J Table A3.1.
Quantity of deposits during flood peaks in 1963, Naodehai reservoir ·
Date
July 20-22
July 23-27
July 28-31
Max. water level (m)
84·57
88·62
83·08
Max. inflow discharge (m3/s)
]928
7980
]160
Max. outflow discharge (m3/s)
760
2470
440
Max. silt discharge of inflow (tis)
617
3280
388
Max. silt discharge of outflow (tis)
67
168
]47
23 ·0
67·3
12·9
9·9
11·3
10·4
]3·1
56·3
2·5
Silt quantity in inflow (Mt)
J
J I
Silt quantity
i~
outflow (Mt)
Quantity of sediment deposited (Mt)
J
-,
Between 1950 and 1958, for example, there was progressive erosion along the valley bottom, vi11ually reaching the original 1940 thalweg. However, this did not influence the levels of deposition over the flood plain, which continued to lise. It is doubtful whether the uncontrolled erosion of sediment deposited in the basin that occurs duling operation could be significantly enhanced without mechanical intervention.
J
j
A3.5.3. Downstream impacts No information available.
A3.6. GMOND RESERVOIR (Austria, 1945) The Gnliind reservoir, used for hydropower, was formed by the construction of the first arch dam in Austria (1943-45). The height of the dam, whose crest also forms the spillway, is 37 nl, the maximunl impounded depth is 30 m (Rienossl and Schnelle, 1982), the length of the reservoir is 940 m and the maximunl width is 200 m, giving the reservoir a surface area of 12·4 ha and an original storage capacity of 0·93 M.m3, which is only 0·5% of the average annual run-off. The reservoir includes a bedload trap at the upstream end of the basin, leading to a tunnel that bypasses the reservoir. This has required substantial periodic repairs as a result of abrasion, firstly with a steel-plate lining and subsequently with a lining of basalt slabs. Bedload reduced substantially afterthe completion of Durlassboden reservoir in the upper catchment in 1967.
176
1
1
J.
1 ~
APPENDIX 3
A3.6.I. Sedimentation vVith the bypassing of bedload, settlement of suspended-sedilnent still posed a problem for the GmUnd reservoir. After the Durlassboden reservoir was commissioned, the annual sediment load entering Gmund reservoir reduced from O· 2 Mt to 0·07 Mt, comprising mainly sands and gravels. Since the 1960s, the annual sediment load has been estimated to be equivalent to about 16% of the reservoir's volume, illustrating the need for effective flushing measures. Aggradation of sediment in the Gmund reservoir increased between 1948 and 1960, reaching a maximum of about 0·2 M.n13 (over 20%) in the early 1960s. From then until 1981, as a result of annual flushing and the construction of Durlassboden reservoir, the total sediment volume was generally less, with a typical value of about 0·15 M.m3 •
A3.6.2. Flushing The bottom outlet passes around the right abutment of the dam in a curved tunnel, with an inlet elevation 28 m below the crest. (It presumably occupied the diversion tunnel used during construction.) A second outlet was added in the middle of the dam during the reinforcement of the dam in 1963, with an inlet elevation 27 m below the crest. During the first period of flushing, from 1948 to 1960, flushing was not executed every year, but was carried out depending on the amount of sediments accuTIlulated. However, from 1960 flushing was carried out every year. Initially, flushing proved to be difficult due to the low flow from the GerIos stream which was the only flow available. There was also a problem with the positioning of the entrance to the bottom outlet, apparently some distance upstream of the dam. In most instances flushing was carried out for a week. The efficacy of flush1."1g improved from 1964, after t."1e addition of the second outlet. From 1967 and the beginning of the operation of Durlassboden reservoir, tIle period required for flushing reduced to a one day. Flushing efficiency was also increased by the increased flow, by using the turbine water released from Durlassboden reservoir as well as the natural flow from the Gerios stream. The flushing operation for the reservoir since 1967 is as follows: the reservoir is drawn dovvn by 9 m one week before flushing is to commence the newer bottom outlet is op~ned the evening before flushing day and water is discharged until the drawdown is 14 m \I the original bottom outlet is opened and the reservoir is emptied ~ flushing ·is then performed through the night with the natural flow from the GerIos stream III the following day flushing continues with the turbine water released from Durlassboden reservoir (25 m?/s) over a period of 3 to 5 hours. s
@
It has been found that between 15 000 m 3 and 30 000 m 3 of sediment is scoured in this "'lay, with a sedin1ent/water ratio as high as 5%.
177
EVACUATION OF SEDIMENTS -r
J The construction of Durlassboden reservoir upstream of Gmund reservoir has -_been a key factor, reducing the incoming sediment load and aiding the flushing process. The idea of separating the bedload from the suspended load has not proved successful in the long run (Rienossl and Schnelle, 1982), due to the maintenance costs resulting from wear on the long bypass tunnel.
A3.6.3. Downstream impacts No information available.
-- f
J -,
J.
i i
f
~,
i
A3.7. PALAGNEDRA RESERVOIR (Switzerland, 1952) Palagnedra reserv()ir was completed in 1952 and forms part of a hydropower scheme on the Melezza River in the southern slopes of the Alps. The water is impounded in the reservoir by a 72 m high concrete arch dam, which also forms the free overflow spillway. The impounded depth at the dam is 55 m and the length of the reservoir is 2·6 km, giving the reservoir an original storage capacity of 5·5 M.m3 , which is less than 3% of the average annual run-off. A 1760 m long diversion tunnel, with a discharge capacity of 225 m3/s, was constructed in 1974, from upstream of the basin to downstream of the dam, with the _main purpose ot allowing sediment-laden flows to be bypassed.
I I
~
I
J
I I
J.
A3. 7.1. Sedimentation The average annual sediment inflow to Palagnedra reservoir is 0·08 Mt (Atkinson, 1996), but this masks a wide annual variation. The sediment type is largely silts with some coarser materials. It is believed that there were no particular problems with sedimentation in the first seven years of operation, although there are no specific records to confirm this (SNCOLD, 1982). In 1961 a flood with a peak discharge of 500 m 3/s carried about 0·5 M.IU3 of alluvium into the lake (9% of the storage capacity), after which accurate routine observations were n1ade. These showed that sedimentation continued steadily, at a rate which was thought to have increased as a result of the effects of the flood. By the end of 1968, the volume of the deposits had reached 1·47 M.m3, or 270/0 of the original capacity. Records of annual sedimentation are available for the period 1969 to 1976 (SNCOLD, 1982) and show annual deposition of between 10 000 and 240000 m 3 • In August 1978 an unusually large flood originated in the catchment area of the 3 Melezza River. This flood, with a peak discharge of 1000 m3/s, caused 1·8 M.m of deposition in the reservoir, which was equivalent to 33% of the original storage voluine. The intakes of the two bottom outlets were covered by Inaterial deposits 12 m and 26 In thick respectively and were temporarily out of service.
178
I
J,
1. 1.
1
1
1
APPENDIX 3
The materials deposited during the flood ranged from silt, sand and gravel to 1·4% wood. In the period 1953-78 a total of 3·79 M.m3 had been deposited, representing 69% of the original volume and an average annual accretion rate of 0·15 M.m 3, or almost 3 % of the capacity per annum.
A3.7.2. Flushing The flushing of the alluvium following the 1978 flood was accomplished in two different phases and required a period of four and a half mont.1.s, commencing in mid-November 1978 and finishing by the end of March 1979. Phase '1 was accomplished by flushing th.rough the upper of the outlets, about 44 m below the spillway crest, with low flow of approximately 300 lis, coming from a lateral valley. phase ran until the end of December and evacuated about 0·3 M.m3 of materiaL Phase 2 started with the opening of the lower bottom outlet, at the base of the impoundment, at the beginning of January. The water used for flushing was successively increased from 1 to 1·5 m3/s by using part of the river run-off. During this three-month phase it was estimated that approximately 2·1 M.m3 of material was flushed from the reservoir. Flushing was assisted by the use of bulldozers and shovels to remove the wood buried in the alluvium and to push the fflaterial into the eroded channels. During the fiusping operation it is understood that L1}e balance· of the inflows was passed through the sediment diversion tunnel, which bypasses the reservoir basin. No information is available on subsequent flushing.
A3.7.3. Downstream impacts No information available.
A3.8. GUANTING RESERVOIR (China, 1953)
Guanting reservoir, on the Yongding River upstream of Beijing in northern China, was built to provide flood protection, river regulation and hydropower. The river regulation supports a major downstream water supply abstraction for the city of Beijing, together with downstream hydropower schemes. It is reported (UNESCO, 1985) that the reservoir provided flood detention from 1953, "vith the sluice gates normally partly open and water levels in the basin fluctuating widely. From 1955 it has operated as an impounding reservoir, with the level varied to releases to the downstream river and to provide storage for impending floods. The is impounded by a 45 m high dam. reservoir basin comprises two aIms, the Yongding arm being about 30 km long and the Guishui
179
EVACUATION OF SEDIMENTS
arm about 30 km to its confluence, with the Yongding about 5 km upstream of the dam. The original surface area at the flood storage level was 229 kln2 • The catchment area is 43 400 km2, which has a semi-arid continental climate, with dry and very cold winters and 750/0 of the annual rainfall occurring between June and September. The average annual rainfall (1951-84) is 420 mm, with a range of 278-545 mm. The mean annual run-off in the period 1925-85 was 1250 M.m3 (Binnie and Partners, 1986), although it was also noted that there had been a progressive decrease in flow between the 1950s and the 1970s, probably due to the construction of about 300 reservoirs in the catchment and the increased use of water for irrigation. Only about 2% of the total catchment drains to the Guishui arm of the reservoir. The original storage capacity was 2270 M.m3, which is about 80% greater than mean annual run-off. This original storage capacity comprised 600 M.m3 of dead storage, 660 M.m3 of 'benefit' storage (for river regulation and hydropower) and 1010 M.m3 of flood storage. Of the total original storage, about 60% lay in the Guishui arm. The dam has a gated spillway, which was under reconstruction in 1986 (Binnie and Partners, 1986), to give a discharge capacity of 2950 n13/s. There is an 8 m diameter bottom outlet tunnel, through which flows are controlled by two sets of four sluice gates, with invert levels 27 In and 39 In below the flood storage level. The maximum discharge capacity is about 560 m3/s, which is about 14 times greater than the mean run-off from the catchment.
A3.8.1. Sedimentation About 40% of the total catchment is classified as loess areas, with friable soils and sparse vegetation cover. There has been a wide range in annual sediment loads in the rivers entering the reservoir, with a nlaximum of 132 Mt and minimum of 1·3 Mt in the period 1951-84 (Binnie and Partners, 1986). As shown in Table A3.2, there was a Table A3.2.
Guanting reservoir, sedbnentation Sediment deposition: M.m3
Mean annual sediment inflow: Mt
Mean trap efficiency: %
In period
Cumulative
73
73
352
352
1961-70
19
96
151
503
1971-80
11
96
85
588
1981-84
7
100
23
611
1953-84
28
611
611
Period
1953-60
i
I !
8]
!
•
•
180
APPENDIX 3
progressive reduction in the estimated annual sediment inflow to the reservoir betvveen the 1950s and the early 1980s. During the first years of operation, up to 1957, it is reported (UNESCO, 1985) that 268 Mt of sediment was deposited in the reservoir, cOlTesponding to a trap efficiency of 63%. Sediment discharge during that period comprised erosion of the emptied bed, which occurred prior to i.mpounding in 1955, and density current venting thereafter. The reduction in the amount of sediment entering Guanting reservoir since the 1950s is attributed to the construction of some 300 reservoirs, with a total storage capacity of 1500 M.m3, within its catchment (UNESCO, 1985) and the warping of agricultural land by the diversion of highly turbid irrigation flows. By 1986 it was apprehended that the sediment deposition in the Yongding arm of the reservoir basin was about to create operational problems, principally by blocking the downstream end of the Guishui ann and isolating a major"part of the residual benefit storage.
A3.8.2. Flushing and other remedial options The original operating rules required that each of the eight bottom outlet gates should be opened for a short time at least four times per year (Binnie and Partrrers, 1986), to minirnise the build up of sediment in the immediate area. By 1986 t~e general sediment level had reached 17 m above the invert of the lower gates within 250 m distance of the bottom outlet, and occasional blockage was reported. One incident of blockage occurred in 1962, resulting in minimal outflow for the first few minutes after a sluice gate was opened, but eventually the area was flushed clear of sediment. Since then, it is reported that the sediment level is monitored and the gates are each opened for about 20 1I1jnutes in tum whenever the level is more than 0·5 ill above the lower gates, with a minimum interval of one month. These operations remove only local accumulations and no attempt is made to flush a significant proportion of the sediment inflovv. Partial blockage of the gate occurred in 1974, which was sluiced away after the gate was raised above 1 m. Atkinson (1996) reports what appears to have been a more substantial flushing operation in October 1954, when a discharge of 80 m3/s was passed for five days at a ponded depth of 8 ill and removed about 10% of the annual sediment inflow. This was prior to the period of impounding and very early in the life of the reservoir, so it cannot be taken as representative of what might now be achieved by flushing. NUIT1erOUS options have been considered for managing sedimentation of Guanting reservoir (Binnie and Partners, 1986). Flushing would not be acceptable, because of the ad verse effects on the small impoundments for the downstream hydropower schemes and the intake for the water supply to Beijing, together with the rise in flood levels which would occur due to sediment
181
EVACUATION OF SEDIMENTS
accretion raising bed levels. Proposals were made for a two-phase approach comprising: (a) divelting sediment-laden flows from the Yongding arm of the basin into the
Guishui arm, to store sediment in its dead storage zone for a period of about eight years (b) subsequently raising the flood storage elevation and developing a system of polders to contain the bulk of the sediment inflows at the upstream end of the Yongding arm. No information is available on whether these or other proposals were adopted, or on sedimentation and flushing experience since 1986.
A3.B.3. Downstream impacts As noted eariier, large-scale flushing could not be undeltaken (even if it were technically possible), because of impacts on downstream works and flood levels.
A3.9. SHUICAOZI RESERVOIR (China, 1958)
The Shuicaozi reservoir is located on the Yili River in south-west China and forms part of a hydropower scheme involving four hydropower stations in cascades on two adjacent river basins (IWHR, 1983). The reservoir is 6 Ian long with a 37 m high dam, giving the reservoir an original storage capacity of 9·58 M.m3, which is less than 2% of the average water inflow to the reservoir of 514 M.m3 •
A3.9.1. Sedimentation The mean annual sediment inflow was estimated at 0·63 Mt (UNESCO, 1985). With no bottom outlet for sediment flushing available, sediment deposition was severe, amounting to,85% of the original storage capacity of the reservoir by 1981. The remaining 1-4 M.m3 of volume was insufficient for flow regulation, which required 3·6 M.m3 ,
A3. 9.2. Flushing Between 1965 and 1981, six flushing experiments were conducted on drawing down the water level in the Shuicaozi reservoir to erode sediment deposits, The outflow Inust be flushed down the spillway, which has a crest elevation about 17 m higher than the original river bed (11 m below the maximum hnpounding level).
182
"
1,
APPENDIX 3
The general procedure is that, after drawing down the reservoir, about 50 m 3/s is released from the upstrealll reservoir fOl~ flushing. The duration of effective flushing is generally about one day, during which time typically about 0·2 M.IU3 of sediment can be removed by between 1 and 3% of the annual flovv. This is only about one third of the average annual sediment inflow. It was noted that, in the vicinity of the dam, the amount of erosion was about 2 m, but 4 krn upstream there was none. The quantity of sediment flushed is reported to be limited by a combination of factors, including consolidation of the fine silts and the deposition of bedload in the upper part of the reservoir basin, but an inlportant factor must be the relatively high elevation of the spillway through which the flushed discharge must pass. IWHR's (1983) characterisation of the flushing in Shuicaozi reservoir can be interpreted as follows. Retrogressive erosion of the bottom of the basin commences at the upstream end of the pool, whose position is controlled by the spillway level. This limits the erosion capacity of the flow. • In the early stages of flushing, the sediment derives mainly from the deepening of the main channel. o Subsequently, the sediment derives mainly from collapsing of the banks. • 1"1 the top portion of the delta, on deposition, the main channel is scoured res lIlting in a progressive increase in accretion on the flood plains. • In the lower part of the delta and in the reach in front of the dam where the river valley is narrow, unconsolidated silts migrate towards the main channel and the level of the entire cross-section can be lowered. $
In the light of these observations and theoretical studies into the behaviour of the sediments, nVHR (1983) proposed the following refinements to the flushing procedure. • Flushing should con1IDence after achieving the maximum p~acticable. drawdown. e In the early stages of flushing, ~he maximum practicable discharge should be deployed to deepen the main channel. The discharge should then be reduced, to lower water levels and to induce more bank instability. e Finally, the discharge should be raised again to flush out the sediment which has accumttlated from bank sliding, etc., and enlarge the cross section of the main channel. C)
This procedure was adopted in the 1984fiushing operation and proved successful in increasing the average sediment concentration and reducing the water consumption.
183
J EVACUATION OF SEDIf'lIENTS
A3. 9.3. Downstream impacts No information available
} A3.IO. HEISONGLIN RESERVOIR (China, 1959)
Heisonglin is a small reservoir, with an initial storage capacity of 8·6 M.m3 , which occupies a valley with an average slope of 1% located in a hilly region on the upstream_reach of the Yeyu River, a tributary of the Yellow River. The 45 m high earth dam was constructed in 1959, for the purpose of impounding water for irrigation and flood protection. The average annual inflow to the reservoir is 14.2 M.m3 , which is 65% greater than the original basin capacity. . Following severe sedimentation problelns at numerous reservoirs built in China in the 1950s, a decision was made in 1961 to use Heisonglin reservoir as an experimental site, to study sedimentation behaviour in detail and to develop appropriate sediment management techniques (MOITis and Fan, 1997).
A3.! 0.1. Sedimentation The mean annual sediment load is reported as 0·71 Mt, and this has the potential to cause accretion of the order of 8% per year if it is all trapped in the reservoir. The annual amount is both irregular and seasonal, with 87% of the mean annual sediment load entering the reservoir during July and August, which represents about the first half of the flood season. In terms of discharge, however, these two months account for only about 25% of the annual inflow. (This pattern is pretty typical of most reservoirs with the potential for significant sedimentation problems.) The sediment is lnainly silt and derives largely from high rates of gully erosion in the catchment. For the first three years of operation, up to June 1962, the reservoir was operated purely as an impounding reservoir, with no flushing, resulting in serious siltation of 1·62 M.m3 , representing an average rate of 6% per annum.
J
I
i
J J
I !I
,1
J ,
J
J
!
II
I
J
AJ! 0.2. Flushing The reservoir is equipped with a single 2 m x 1·5 m bottoln outlet with an invert level 12 m below the original base of the reservoir basin and a discharge capacity of 10 m3/s, for the purpose of making downstream releases. Starting in 1962, the mode of operation changed to involve emptying the reservoir during the flood season and impounding water only during the nonflood season. Turbid density ClUTents were also released. These measures lnanaged to reduce the trap efficiency of the reservoir to 15%, but the reservoir was still losing capacity. As a result of additional work, the lateral erosion technique was developed and first used at this site in 1980 by Xia Mai-Ding and his co-workers. This technique was found to be capable of arresting sediInent accumulation and recovering a portion of lost storage.
184
J J
I
J J
i
J
APPENDIX 3
The stable long-term capacity which can be sustained at Heisonglin by deploying all the sediment management techniques is estimated as 2·5-3·0 M.m 3 , representing about 20% of the mean annual inflow and about 30% of the original storage capacity. Further descriptions of the flushing techniques follow.
Drawdown flushing The lowering of the reservoir pool in July initiates the erosional processes associated with flushing. A large and highly concentrated discharge of sediment occurs at the transition from drawdown to flushing. Smaller amounts of sediment are removed thereafter, by base flow and by the smaller inflow events that can be released through the bottom outlet. Due to the operational objectives and constraints at Heisonglin, in particular the supply of irrigation water at mal1ageable rates and the limited discharge capacity of the bottom outlet, flushing ,"vith large flows to widen the. main channel cannot be undertaken. Detention flushing vVhen floods entering the reservoir storage during the drawdown period exceed the release rate to irrigators, the pool retains the silt-laden water for petiods of hours or days. As the silts settle slowly, if the flow can be discharged within a couple of days, approximately 70% of the sediment can normally be evacuated with the water releases. In addition, as the reservoir is emptied at the end of each flood and riverine flow is again established, deposits of silt are scoured in t..l-J.e same manner as for the initial draw down .. Lateral erosion Attempts were made to create a longitudinal channel running parallel to the main flushing channel for hydraulic scouring of the floodplain deposits. Accidental overflow from this channel initiated lateral erosion, which formed a gully across the deposits. From this, it was recognised that erosion of the deposits would proceed much faster by directing the flow along the high lateral slopes from the sides of the reservoir towards the main channel. Because of the high gradient that can be achieved by lateral drainage, even small discharges were highly effective in eroding deposits of non-cohesive silt. This technique was applied for a total of 6·8 months between 1980 and 1985. Using a flow of only 0·2 m 3/s, 816000 m 3 of deposits were eroded into the main channel. This equated to a remarkably higher sediment/water ratio of 0·23.
A3.IOJ. Downstream impacts A unique feature of Heisonglin is that sediment balance is achieved in the reservoir, while both the water and sediment are diverted to beneficial use. Between floods, the river flows freely through the reservoir outlet, and is diverted into llTigation intakes downstream. Seasonal floods exceeding the capacity of the irrigati;n intakes are temporarily impounded and released to irrigators at the maximum capacity of the downstream irrigation diversions, which, along with
185
j, EVACUATION OF SEDIMENTS
1, the irrigation canals, are designed to accept very high concentrations of sediment, in order to enhance soil fertility. , Environmental penalties are clearly either non-existent or very minor.
1 A3.11. SANMENXIA RESERVOIR (China, 1960) Sanmenxia dam was completed in September 1960 and was the first to be built on the middle reaches of the silt-laden Yellow River. The 96 m high concrete gravity danl, which controls a drainage area of almost 700 000 km2, was planned as a multiple-use project for flood control, hydropower, inigation, navigation and ice jam control. The maximum historic flood at the site was 36000 m 3/s in 1843 (Morris and Fan, 1997). It was designed originally with a full reservoir level of 360 m, giving 65 000 M.m3 of storage capacity, inundating 3500 kn12 of floodplain and requiring the relocation of 870 000 people. To reduce these impacts during the first stage of construction, the dam was built to an elevation of 350 m with a maximum operating level of 340 m and an original storage capacity of 9640 M.m3 • Because of the high sediment loads in the Yellow River, the original plans included two sediment control measures: • the reservoir was to impound water continuously, but release 350/0 of the sediment inflow as turbidity currents, through 12 outlets at an elevation of 300 m (40 m below top operating level) " • total sediment inflow was to be reduced by 3% annually by soil conservation works in the catchment, resulting in a 60% reduction over 20 years. , These figures proved to be extremely optimistic (Monis and Fan, 1997).
1
1 1
1 I
,..l t
1 J
, -'
A3.II.!. Sedimentation The estimated sediment load prior to construction of the dam was 1600 Mt per annum, with a median diameter of 0·03 mm, representing a mean concentration of 38 gIl. Extreme concentrations of up to 940 gil had been measured in the Yellow River in the vicinity of the dam site. About 60% of the annual sediment load occuned with about 300/0 of the annual run-off in July and August. Iffilnediately after impounding began, severe sediment problems became evident. During the first 18 months of operation, 1800 Mt of sediment had accumulated in the reservoir, representing 93 % trapping and the loss of about 20% of the storage capacity. In the next four years, 3400 Mt was deposited and the total loss of storage reached 3700 M.m3 , or 400/0 of the original capacity (Monis and Fan, 1997). This rate of accretion threatened to elilninate all the project benefits, in addition to sedinlent deposits which were raising the bed elevation and flood levels in the Yellow River as fat as 260 kin upstream of the daln.
186
1
1
1
1
1
1
APPENDIX 3
A3.//.2. Development of flushing and sediment control The severe sediment problems at Sanmenxia threatened agricultural lands and riverside industrial areas and could have required the relocation of up to 1 million more people, while abandonment of the project would require removal of the dam and have severe impacts due to additional downstream sediment loads over the subsequent years. Accordingly, the establishment of a sustainable and acceptable sediment balance became a high priority objective, on which the realisation of the other project benefits depended. The water and sediment management objectives for Sanlnenxia can be summarised as: • • • • o
controlling extreme floods providing irrigation supplies, ice jam control and hydropower limiting upstream backwater deposition and rises in flood levels limiting the amount of deposition downstream of the dam preserving the long-tenn effective storage capacity.
To achieve these objectives, Sanmenxia dam required extensive reconstnlction to provide hjgh capacity bottom outlets, and the reservoir operation was changed substantially; impounding during the non-flood seasons and emptying and flushing during the flood season from July to October each year. Almost two decades were required from first impounding to bring the sediment inflow and discharge into balance, through the implementation of a series of staged sediment control measures. The six stages described below outline the operational history and development of Sanmenxia reservoir.
Stage 1 (1960-62) Impounding of the reservoir began in September 1960 and seli.ous deposition was occurring by the time the water level reached 335·5 m (4·5 m below the planned impounding level). Sediment accumulation raised bed levels by 4·5 m . near the upstream end of the reservoir and caused backwater effects 250 kID upstream of dam. This was endangering agricultural land in the floodplain of the Wei River, which joins the Yellow River near the upstream end of the impoundment, and industrial developments in Xian City. Total deposition up to the start of the 1962 flood season was estimated as 1800 Mt. Stage 2 (1962-66) The reservoir operation was changed from April 1962 to maintain a lower water level throughout the year, by using the 12 outlets at an elevation of 300 m. However, the outlet capacity proved insufficient and water levels during periods of high discharge were too high for efficient sediment release during the flood season, resulting in the trapping of a further 3400 l'vlt over four flood seasons. It was also noted that sediment trapping in the reservoir was having an effect downstream of the dam, where degradation was occurring. It became clear that additional sluicing capacity would be required.
187
EVACUATION OF SEDIMENTS
[ Stage 3 (1966-70) Prior to the start of the 1966 flood season, additional sluicing capacity was provided by the excavation of two 11 m diameter bypass tunnels, with an invert elevation of 290 m and controlled by 8 x 8 m radial gates, around the · left abutment of the dam. Four of the eight power intakes were convelted to sediment sluices and the pool level was lowered duling the flood season. This had the effect of reducing the trap efficiency to 17·5%, but did not lower the bed elevation at the upstream end of the reservoir basin, and deposition in the Wei River was still a problem. Consequently, it was decided that still more low-level sluicing capacity would be needed. Stage 4 (1970-73) In time for the start of the 1970 flood season, 8 of the original 12 river diversion outlets, filled with concrete, were reopened at an elevation of 280 m. Reservoir operation was changed to flood detention and sediment sluicing, with all outlets constantly open. Sediment release efficiency reached 1050/0, representing a yearon-year reduction in the amount of accumulated sedin1ent in the reservoir basin. The bed elevation at the upstream end of the basin fell by nearly 2 m. In 1973 five generating sets of 50 MW each were installed, replacing the original eight 125 MW sets. Stage 5 (1973-78) Once an overall sediment balance had been achieved and bed levels at the upstream end of the reservoir were under control, it was decided that reservoir operation could be modified after the 1973 flood season, to increase the project benefits. This regin1e, which is understood to have continued to the present day, provides water for irrigation, hydropower, and ice jam control during the non- · flood months. At the start of the flood season, in July, all the outlets are opened and the high-capacity bottom outlets allow a low pool level to be maintained. The high discharges can-ying the .sediment load also prevent excessive deposition in the Yellow River downstream of the dam. Stage 6 (1978 onwards) Repair of serious abrasion in the bottoln outlet was carried out, decreasing the cross-sectional area of the outlets and reducing the low-level flushing capacity. To cOinpensate for this, two additional bottom outlets were opened in 1990. Since 1980, to avoid turbine abrasion by high sediment concentrations, hydropower generation has been halted during the flood season. Long-term storage capacity Since the sediment balance was brought under control from about 1975, the net storage capacity below an elevation of 330 In (10 In below top water level) has fluctuated between about 3000 and 3300 M.m3 , representing 50-55% of the original capacity at that elevation.
188
[ [
[ [ r
L
[
r
L
r
i
i
I I I
L [
1 r
I
1
r
L
I
APPENDIX 3
A3.ll.3. Downstream impacts Between 1960 and 1964, when the reservoir was operating with a high trap efficiency and releasing essentially clear water, there was degradation of the river-bed in the lower Yellow River, posing a danger to the fioodbanks through undelmining. Between 1964 and 1973, when extensive modifications vvere being made and sedimentation in the reservoir basin was being brought under control, the limited discharge capacity of the bOttOlTI outlets meant that high concentrations of sediment were being released, but that the discharges were generally insufficient for its conveyance, leading to severe sediment accretion over a length of 300-400 kIn downstream of the dam, reducing the flood discharge capacity of the main channel and increasing downstream flood risks. Only after the excavation of the two bypass tunnels, the conversion of the four power intakes and t..~e reopening of the eight diversion outlets, could both sediment and water be discharged at the rates required to stabilise conditions along the lower Yellow River. However, channel deposition continues to occur at low discharges, and the regulation of sediment in the lower Yellow River is a problelTI that is not fully solved (Mon-is and Fan, 1997).
A3.12. VVARSAK RESERVOIR (Pakistan, 1960) Built in 1960 on the Kabul River, a tributary of the Indus, Warsak reservoir was forn1ed for irrigation and hydropower and impounded by a 76 m high concrete dam. The crest of the gated spillway is 12 m below the highest impounding level. reservoir basin is 42 km long and has an original storage capacity of 170 M.m3 • The average fu"1nual inflow to the reservoir of 21 100 M.m3 is 124 tilnes the original storage capacity.
A3./2.1. Sedimentation During the period 1961-70, the average measured suspended-sediment concentration was 727 mg/l, giving an average sediment inflow of 15·3 Mt, suggesting a maximum accretion potential of the order of 80/0 per annum. The particle size distribution of measured suspended load comprised 12% sand, 60% silt and 28% clay (Mahmood, 1987). In addition, the Kabul River carries a bedload of gravel and cobbles, which were not included in the rneasuredconcentrations. After the first year's operation, 30 M.m3 of sediment had deposited in the reservoir, increasing to 70 M.m3 five years. By 1980, after 20 years of operation, reservoir had completely silted to the conservation pool elevation, except for a 60 m wide by 6 rn deep channel on right bank, where the po\ver and ilTigation intakes are located. The reservoir deposits had an accumulation of cobbles and boulders on the surface and, in 1983, NIahmood observed
189
EVACUATION OFSEDIMENTS
gravels and cobbles up to 75 nun being passed from the reservoir with the irrigation supplies.
A3.12.2. Flushing Five flushing operations were performed during the period 1976-79. The flushing was carried out by lowering the water level to the spillway crest level. The total duration of flushing was about 20 days and these operations removed an estimated 4·2 M.m3 of sedimel1t deposits from the reservoir, amounting to about 6% of the probable sediment inflow over the same period. No information is available on any later attempts at flushing, but it appears that the reservoir has essentially reached an equilibrium condition with virtually no residual live storage capacity, and that it will not be practicable to increase the live storage unless deeper high-capacity outlets are provided at the dam.
J
A3.12.3. Downstream impacts After a relatively short period while the reservoir filled with sediment and concentrations in the downstream river were reduced, it appears that the downstream effect of the presence of Warsak reservoir is currently minor. If flushing were to be instigated, the local effects could be severe, but downstream of the confluence with the much larger Indus River (which has been starved of high sediment loads since the impounding of the Tarbelareservoir in the 1970s), the effects could be beneficial in diminishing degradation of the river bed.
A3.13. OUCHI-KURGAN RESERVOIR (Former USSR, 1961) Ouchi-Kurgan is a 17 Ian long reservoir, used for the purpose of inigation and power production, which began impounding in October 1961 and had an original storage capacity of 56·4 M.m3 (UNESCO, 1985). The catchment area is not given, but the annual run-off is given as about 15000 M.m3 (Atkinson, 1996), which is over 250 times the original storage capacity.
A3.I3. J. Sedimentation
j
J
The annual sedinlent discharge into the reservoir is reported as between 12 and 14 Mt. The volume of deposited sediment reached about 30 M.m3 by 1968 and was reasonably stable at 50-55% of the original storage capacity up to 1970, after which no further data are available. r
190
J
J
APPENDIX 3
A3.13.2. Flushing The dam has eight bottom outlets, 35 m below impounding level and 21 m below the elevation of the power intake. These are reported to have a discharge capacity of about 350 m 3/s at maximum impounding lev~l. From other data available, it appears that this is the discharge capacity for each, giving a maximum discharge capacity of 2800 m 3/s when the reservoir is full. This is much larger than the mean inflow of about 500 m3/s. Since 1963, drawdown flushing of the reservoir has been operated, which was achieved by lowering the water level by 4-5 m during the May to August flood season. The fact that the lowering is so modest suggests that it may have depended on the use of a gated spillway, as well as the bottom outlets. The available data are not entirely consistent, as a plot of sediment concentrations entering and leaving the reservoir during the 1964 flood season suggests that there would be net accretion, rather than an approximate equilibrium.
A3.13.3. Downstream impacts No information available.
A3.14. SEFIO-RUD RESERVOIR (Iran, 1962) Sefid-Rud reservoir in Iran was constructed in 1962 for irrigation and power generation. The dam is a buttress-type concrete gravity structure, with a maximum dam height of 106 m. There is a 'morning glory' service spillway and a gated auxiliary spillway (}\IloITis and Fan, 1997). The catchment area, which is largely semi~arid, with annual rainfall of 250-400 mmlyr, totals 56200 krn2 , yielding an average annual inflow to the reservoir of 5008 M.m3 . The reservoir is located at the junction of a major and minor catchment and the basin also has major and minor branches. The maximum initial reservoir depth is 82 ill, the length of the major branch is 25 km and the original storage capacity of the reservoir was 1760 M.m3, representing 35% of the mean annual inflow. The normal impounding level is 271·65 ill with a maximum of 276·25 m (presumably the estimated peak flood level). During normal operation the minimum drawdown level is 240 ill, equivalent to a draw down of about 40% of the original basin depth.
A3.14.1. Sedimentation The vast majority of the annual sediment inflows occur with high discharges in the months of March to June. Sedimentation was a serious problem in the first 17 years of operation and caused an average storage loss of 36·5 M.m3 per annum, equivalent to an annual rate of 2·1 %. The trap efficiency during this period was estimated as 730/0, with
19/
EVACUATION OF SEDIMENTS
most of the discharged sediment complising density CUlTents. The sediInent composition was 33% sand, 47% silt and 20% clay. The reservoir capacity reached a nlinimum of about 630/0 of the original in 1982-83, before recovering as a result of the flushing measures.
A3./4.2. Flushing The reservoir was built with three bottom outlets on the right-hand side (total discharge capacity 430 m3/s, elevation 191·3 m) and two bOttOlTI outlets on the left-hand side (550 m 3/s, 193·8 m). These are near the bottom of the reservoir, close to the original river-bed level and their total discharge capacity of 980 m3/s compares well with the mean annual flow of 160 m3/s. (With the reservoir level drawn down to 25%, the bottom outlet discharge capacity would be reduced by about a factor of two, so would still be three times the nlean annual discharge.) Because of the rapid and continuing reduction in storage capacity, a decision was made in 1980 that the operating regime should be changed to incorporate more pro-active sediment removal. A number of alternatives were considered, before selecting the option of annually elnptying and flushing the reservoir. The peak annual inflows to Sefid-Rud reservoir occur in the months of March to June and the inigation period is from May to September. Accordingly, the flushing programrne was designed to occur from October to February, virtually emptying the reservoir down to an elevation of 197 m, then allowing the reservoir to fill in time for the start of the inigation season. (During the first two years of flushing, the reservoir was not completely drawn down because of fears that unstable sediment would block the bottom outlets.) It would have been desirable to flush the reservoir for a further period, to take advantage of the higher flows that would further SCOUT the deposits and widen the main channel, but the uncertainty of filling the reservoir for the irrigation period dictates the closure of the outlets in February. Furthennore, the bottom outlets may not have sufficient discharge capacity to allow this to be done. The consequence of this operating regime is that the majority of the annual sediment load enters the reservoir when it is either filling or full, so is likely to be deposited until it can be subjected to erosion during the subsequent flushing period. The initial years of flushing provided very high sediment outflows and a rapid recovery of storage capacity up to about 75% of the original in 1992. This was expected to be approximately sustained with a continuing flushing regime, with the possibility of an increase to about 900/0 with supplementary measures (as described later). The gross benefits of the flushing operations, which comprise the volume of sediment removed plus the vollune of deposition averted, amounted to 320 M.ln3 over the first 10 years of flushing, equivalent to 1·9% of the storage capacity per annum. The average suspended-sedinlent content in the flushing flow was 48 gil, with a peak of up to 670 gIl. Selected statistics on the annual flushing period up to 1990 are summarised in Table A3.3.
92
:
•...
; '.r·. . r
.:~
APPENDIX 3
Table A3.3.
Summary of sediment flushing at Sefid-Rud reservoir
F1ushing year
Dra\vdown flushing duration: days
Empty flushing duration: days
'Water volume used: M.m'
Water volume used as % of annual inflow
Sediment removed: Mt
1980-81
61
0
536
10
24
1981-82
6S
0
390
11
12
1982-83
117
10
1513
26
52
1983-84
16
80
795
23
68
1984-85
19
138
1810
29
142
1985-86
18
129
1131
29
46
17
85
26
27
1987-88
24
86
1988-89
9
113
1989-90
5
103
681
351
744
10667
1986-87
Total
I
I
!
I
942
I
I
]812
22
1057
31
54
22
32
21
514
I
I
57
It was observed that sediInents were eroded during draw down and flushing by three processes: • sheet erosion • channel erosion • bank failures. Sheet erosion was the most important type of erosion during the first draw down operation, comprising sheet flow and scour of recentiy deposited fine sediments in the lower reaches of the reservoir. After the first couple of years of drawdown flushing, the amount of sediment removed by this process was less important, removing only a small part of the deposition which had occurred onto the submerged floodplain in the previous period when the reservoir was filling or full. Channel erosion was the most important process from the third drawdown period onwards, when the first full draw down was undertaken. The channel banks would be near-vertical initially, but would then fail. The rate of sediment removal was found to be sensitive to changes in discharge, which would trigger off accelerated erosion. It \vas concluded that the long-term storage recovery would be limited by the narrow width of the Inain channel in relation to the overall width of the reservoir basin. Furthermore, sediment would continue to be deposited on the submerged floodplains during impounding periods, suggesting that, after the initial period of recovery, the storage capacity would begin to decrease again in the longer term.
193
EVACUATION OF SEDIMENTS
Two novel methods were investigated to promote the removal of deposits from the floodplains and to prevent a progressive long-term loss of storage, as described below. Lateral erosion by piping
SediInents in the upper reaches of the reservoir were deposited in thick alternating layers of cohesive and non-cohesive soils and the resistance to erosion of the cohesive materials had been enhanced by the desiccation and compaction caused by annual drawdown. However, collapses in the cohesive materials were noticed near the areas of the main channel, which appeared to have been triggered by piping and washout of the sand sediments beneath the cohesive deposits . . In the 1985-86 flushing season, field experiments were set up to induce and enhance piping artificially. A pit 2 In in diameter and 4 m deep was dug into the sandy deposits 40 m away from the main channel, which was kept full of water by pumps during the flushing operation. After 15 days, the top cohesive layer had collapsed into the main flushing channel, creating a gully 3-10 m deep and 6-15 m wide. It was concluded that a hydraulic gradient of 0·25 to 0·33 between the pit and the main channel was sufficient to initiate piping in these deposits. Diversion channel
Longitudinal erosion is achieved by constructing a pilot channel parallel to the main channel. The channel is fed by water, either from a tributary or by diverting flow from the main river using a temporary diversion dam. This concept was initially tested in the 1987-88 flushing season, in which a 5 km diversion channel was formed and flows of between 1·2 and 2·2 m 3Is were diverted from a small tributary. A substantial increase in the outflow sediment concentration was achieved, so, following this success, a longer channel was built in the following flushing season, along the smaller of the two main river valleys forming the reservoir basin. The sequence of operation was: . • construct a pilot channel defining the route and connected with the stream at its upstream end • form an earth dam to divert the streaInflow to the. diversion channel • divert water to the pilot channel at a rate which is high enough to avoid overtopping of the diversion dam, but low enough to avoid the pilot channel overtopping and sholt-circuiting back to the main channel previously formed by drawdown flushing alone. The pilot channel was 7·6 km long, with an average slope of about 11200, ~d that generally followed the edge of the floodplain deposits, to allow access by earthmoving plant. Initially, starting in mid-January 1989, a flow of 1 m 3/s was passed down the pilot channel, which was increased progressively in line with the erosion of the channel to the full 12 m 3/s flow in the tributary. Retrogressive erosion was the principal means of channel development. Erosion continued until the channel was submerged by rising water levels during ·i mpounding in February, then resumed when the diversion channel was
1
-. ~
i
I I
.--l
._- /
!
:--I
£. -
APPENDIX 3
re-exposed for the next flushing season. By December 1989, after a total" of 95 days of operation, the diversion channel had reached an essentially stable condition. The eventual channel top width ranged between 50 m and 200 m, but no information is given on the depth or total volUlne eroded.
_I
Long-term predictions Studies by Tolouie (1993) estimated that, by creating a new diversion channel each year and by deploying 75% of the total flushing season inflow, it would be possible to recover lost storage and maintain a long-telID storage capacity of about 90%, compared with about 75% by flushing alone.
A3.14.3. Downstream impacts In the case of the Sefid-Rud reservoir, the intakes and canals used for diverting reservoir releases to irrigators cannot tolerate high sediment loads. To prevent downstream sedimentation problems, the sediment concentration should not exceed 5 gil dudng the irrigation season. The sluices on irrigation ba.7ages remain open during the flushing season, thus passing the sediment-laden flow with minimum interruption, while the irrigation intakes remain shut to exclude the flushed sediment from the delivery canals. No information is given in the literature regarding potential environmentat impacts downstream of the reservoir.
A3.IS. KHASHM EL GIRBA RESERVOIR (Sudan, 1964) The Khashm EI Girba dam, sihlated on the Atbara River in Sudan, was completed in 1964 and is used for power generation, irrigation and water supply. The reservoir had an original storage capacity of 950 M.m3 , but no information is available regarding the catchment area or mean annual inflow.
A3.15.1. Sedimentation The capacity of the reservoir was seriously depleted by an average annual sediment inflow of about 84 Mt (UNESCO, 1985). Morris and Fan (1997) show a photograph of a water supply intake in the delta upstream of the dam, apparently completely surrounded by sediment, but no further details are available.
A3.15.2. Flushing Little infom1ation is available on the flushing operations. Table A3.4 lists some data for flushing operations carried out in July 1971 and July 1973.
195
EVACUATION OF SEDIMENTS
Table A3.4.
Khashm El Girba reservoir, sediment flushing
Flushing period
Water used: M.rn 3
Silt inflow: Mt
Net sediment release: M.rn 3
11-14 July 1971
612
3·5
17·5
29 July-2 Aug. 1973
545
3·3
12·5
According to EI Hag (1980) and EI Faith Saad (1980) (quoted in UNESCO, 1985) the sediment outflow each July, including the flushing operation periods, was 85 Mt, which is about the same as the average estimated annual sediment inflow. Unfortunately, the 1980 references have not been obtained, so no fUlther details are available of the apparently successful flushing operations carried out at the K.l1ashm EI Girba reservoir.
A3.15.3. Downstream impacts No information available.
A3.16. HENGSHAN RESERVOIR (China, 1966) The Hengshan reservoir is used for flood control and in-igation. It is a small gorge-type reservoir, 1 to 2 kIn in length (the references disagree on this detail), located in an arid zone of scarce water supplies (Mon-is and Fan, 1997). The concrete arch dam is 69 m high and the maximum water depth is 65 m (Atkinson, 1996), giving the reservoir an original storage capacity of 13·3 M.m3 (UNESCO, 1985), which is rather less than the reported mean annual nln-off of 15·8 M.m3 •
A3.16.1. Sedimentation From 1966 to 1973, the first eight years of the reservoir's operation, 3·19 M.m3 of sediment had deposited in the reservoir, representing 24% of the original storage, with the height of the deposits behind the dam reaching 27 m. Deposits near the dam were described as fine, with a Dso of 0·02 mm, becoming coarser at a distance of 350 m to 800 m from the dam.
A. 3. 16.2. Flushing The dam has a small outlet, 2·6 m above the base of the danl, with a discharge capacity (at full impounding level) of 17 m 3/s, but there is also an outlet for flood
196
APPENDIX 3
I
Ij
i i
I I
i
Ij I
discharge 14·5 m above the river bed, capable of passing a maximum discharge of 1260 m 3/s. Flushing was first can-ied out in July 1974, when the reservoir was emptied and flushed for 37 days. During this flushing period, 0·8 M.m 3 of sediInent was removed from the reservoir. The reservoir was then impounded for five years to June 1979, before flushing for the second time for a period of 52 days during the flood season. The second flushing period removed 1·03 M.m3 of sedilnent, reducing the volume of sediment in the reservoir to 2·62 M.m3 (20% of the original storage capacity). Emptying and flushing were subsequently undertaken in 1982 and 1986. During emptying and flushing, it was reported that strong retrogressive erosion occurred as a result of lowering the water level. A channel was rapidly fonned in the floodplain deposits regressing upstream and deepening continuously. In the first 350 m from the datu, deposits on the floodplain collapsed and slid into the main channel. In the upstream reaches, where the sediments were coarser, the cross-section eroded was initially rectangular in form and was followed by the collapse of the floodplain deposits into the main channel. Outflow concentrations were reported to reach about 1000 gIl, irrespective of the flushing discharge (UNESCO, 1985). Experience at Hengshan reservoir suggests that flushing every few years is sufficient in this case, which is probably aided significantly by the high gradient of u~e original stream bed and the steepness of the valley sides. The efficiency of the flushing was high when the main channel, which had been eroded in the previous flushing, had been silted up by deposited sediments during a period of several years. It was thought that greater recovery of storage capacity could be achieved if the reservoir was to be emptied prior to the start of the flood.
A3. I 6.3. Downstream impacts No information available.
A3.17. CACH! RESERVOIR. (Costa Rica, 1966) Cacm hydropower reservoir, located on the Reventazon River, was the first major hydropower scheme on the river and several others are planned to be built downstream. The reservoir was completed in 1966 with the construction of a 76 m high concrete arch dam, it has a surface area of 324 ha, is 6 kIn in length, with a maximum depth of 69 ill, giving an original storage capacity of 54 M.m 3 • The mountainous catchlnent of 785 km2 produces a mean annual run-off of about 1500 M.m 3 , or 25-30 tirrles the original storage capacity. The catchment area is heavily vegetated, about 55% is forest and most of the remaining area is agricultural.
197
J EVACUATION OF SEDiMENTS
,J A3./7./. Sedimentation The average annual sediment inflow is 0·81 Mt, which would have a deposited volume of the order of rather over 1% of the original storage volume. The annual load is estimated to be distributed as follows: 18 % throughflow from normal hydropower and gate operations; 21 % deposited on telTaces; 7% bedload trapped in reservoir; and 54% thalweg deposits, removed by flushing. A nalTOW section of the reservoir 4 km upstream of the dam divides the basin into upper and. lower parts. The upper basin is being progressively filled with sand and coarse matelial, which is generally not removed by flushing, whereas 'the lower basin consists of a deep river channellnaintained by flushing, between a series of relatively flat river terraces, onto which fine sediment is deposited. For the first seven years it was apparently operated without flushing, with the reservoir trapping 82% of the incoming sediment. Part of the suspended load was transported by turbidity CUlTents to the area of the dam and, after several years, were starting to interfere with hydropower production.
A3./7.2. Flushing The dam has a single bottom outlet located near the thalweg of the original river channel and immediately adjacent to the intake screen, a location that facilitates flushing of sediment from in front of the intake. Flushing operations at eachi reservoir have been well documented and have been considered successful in preserving the storage capacity of the reseIVoir (MolTis and Fan, 1997). The first flushing operation was canied out in October 1973, to flush sediments that had accumulated near the power intake. Owing to the success of this operation, it was decided to carry out flushing every year during the wet . season. During the 18 years from 1973 to 1990 the reseIVoir was flushed 14 times. Flushing was cmTied out in three stages: • Slow drawdown: the reseIVoir level was lowered from 990 m (full impounding level) to 965 m at a rate of 1 mlday, with the turbines operating at full capacity and supplemented by opening the spillway gates and the bottom outlet as necessary • Rapid drawdown: the turbines were stopped and the bottom outlet opened to evacuate the remaining water from the reservoir, which typically took between 5 and 10 hours • Free flow: this typically lasted 2 to 3 days and occulTed once the reseIVoir was empty and the river was flowing freely along the original river channel.
At the end of the flushing operation, the outlet was closed up and the reservoir allowed to refill, typically taking between 16 and 21 days. The amount of
198
J J
APPENDIX 3
sediment released during each stage varied considerably from one event to another, reflecting variations in the rates of sediment inflow and different intervals between flushing operations, as illustrated in Table A3.5. Little erosion of the sandy or gravely material in the upper part of the reservoir basin was observed during flushing operations. In the lower basin, minor gullies developed across the terraces during the slow draw down period and on the terrace slopes there was a tendency for deposited sediments to be eroded by wave action. However, there was no general erosion of sediment from the surface of the terraces. The zone of ma"'{imum erosion was along the main channel, which is also where most of the incoming sediment was deposited. The slow draw down exposed channel sediments to scouring action, and finer sediments were transported nearer the dam. The erosion and release of sediment during the rapid draw down phase was reported to be spectacular (Monis and Fan, 1997). During the last few metres of rapid draw down hyper-concentrated flows were observed. A major part of the Table A3.5.
Sediment released by 14 flushing events at Cachf reservoir Quantity flushed: t
Date
Slow drawdown
Rapid drawdown
Freefiow
Total
Oct. 1973
-
-
-
-
Aug. 1974
186200
225200
-
411400
Oct. 1975
-
-
Oct. 1977
-
40700
44000
-
19500
5000
Oct. 1981
14600
348900
113400
476900
Oct. 1982
5800
111 600
250900
386300
402400
114300
May 1980
Sept. 1983
I
I
I!
28700
84700
I
24500
545400 I
I
665500
Oct. 1984
23300
604600
June 1985
-
-
July 1987
-
-
Sept. 1988
61600
627 000
577100
Sept. 1989
42400
144300
482200
r---moo
Gsoo
278 700
347 100
653000
Oct. 1990
32600
I
-
-
1265700
Note: Dashes indicate no data available; italics indicate an estimate
199
EVACUATION OF SEDIMENTS
total sediment flushed on each occasion occun-ed during the final few hours of the rapid drawdown phase and the first few hours of free flow conditions.
A3.173. Downstream impacts During each flushing operation, in which peak concentrations exceeding 400 gil have been measured, it was observed that sub.stantial amounts of sediment were deposited on lower floodplain areas and bars between the dam and the Caribbean and a stratified plume of turbid water was observed in the sea. It was expected that the riverine deposits would be eroded by subsequent floods. No studies have apparently been can-ied out on the effect of the sediment on downstream biology, although anecdotal reports from local observers ·suggest that the concentrated sediment releases ,cal:lse extreme mortality to all types of river biota.
A3.18. GEBIDEM RESERVOIR (SWitzerland, 1968) Gebideln hydropower reservoir is situated in the Swiss Alps on the Massa River, a tributary of the Rhone. The dam comprises a thin 122 m high arch. The reservoir is 1· 5 km long, with a Inaximum depth of 113 m and storage capacity of 9 M.m3• The catchment is 200 km2, of which 65% is occupied by a glacier. The average annual inflow to the reservoir of 429 M.n13 is almost 50 times the impounding volume. The annual discharge is seasonal, being dominated by snowmelt, glacial melt and sumnler storms between May and October, with negligible flows in the winter months (Morris and Fan, 1997).
A3.IB.I. Sedimentation As a result of glacial activity the sediment inflow to Gebidem reservoir is very high, with an annual average of about 0-4 M.m3 , equivalent to over 4% of the storage capacity. This is mainly granular material, ranging from very fine sand to gravel, of which about 20% is between 1 mnl and 100 nlm in diameter. The sediment load is strongly correlated with the flow hydro graph during the summer months.
A.3.IB.2. Flushing Because of the high sediment load in relation to the reservoir capacity, sediment managelnent was planned for in the initial design. Consideration was given to the alternatives of sedilnent bypassing and dredging before selecting flushing as the . most practicable and economic option. Venting of turbidity cun-ents was also considered, but was rejected because the sediments would be too coarse for it to be effective: The danl was designed with two flushing tunnels located directly beneath the . power intakes and close to the original streanl-bed level. Originally, the low-level
200
APPENDIX 3
outlets each contained two gates; a radial service gate at the downstream end, and flap gate at the upstream end that could be closed in emergencies or for maintenance of the service gate and outlet tunneL To resist erosion, the entire sutface of the outlet tunnel was lined with steel plate. After 25 years of operation, erosion of the service gate seal on the bottom outlet had become a problem, preventing an effective watertight seal to be maintained. In .1995-96 a third gate was added to each outlet, for use as the discharge control during flushing operations, allowing the original service gates to be used only fully open or fully closed, without significant wear on the replaced seals. The reservoir is flushed between May and July every year, for 2 or 3 days. Owing to the gorge-type geolnetry of the impoundment, flushing has resulted in the entire reservoir basin being kept virtually sediment free. Flushing is carried out prior to late SUnL.T..er floods, \vhen conditions favourable to flushing occur: the flow of the Massa River is low enough (less than 20 rn3/s) to allow full drawdovvn • the flow of the Rhone is large enough (greater than 40 rn3/s) to dilute and transport the sediment-laden flows, but not sufficient to pose excessive flood risks downstream • the 0° isotherm is located around 3000 m, which corresponds to stable meteorological conditions (presumably indicating that a summer storm would not interfere with the flushing operation). I)
In preparation for flushing, the reservoir level is lowered to the minimum operating level by releasing water through the turbines, which are then closed .. Drawdown flushing is then initiated over a period of two hours by first opening one gate then the second gate, progressively raising the discharge from 10 m3/s up to about 60 m3/s. It takes between 3 and 6 hours for free flow conditions to be achieved at the outlet, typically at discharges of between 10 m 3/s and 20 m 3/s. In some years, the outlet gates are interrrJttently closed and the reservoir is allowed to fill for 20 minutes, then the' gates are fully reopened; the resulting raising and lowering of the water level facilitates flushing of deposits in the vicinity of the dam. . .-. Tables A3.6 and A3.7 provide a summary of the flushing carried out at Gebidem dam from 1982 to 1993 and a sediment balaIlce for 1990-91.
A.3. f 8.3. Downstream impacts The combination of sediment release and reduced discharge due to diversion for power production has heavily impacted the Massa gorge downstream of the dam. By end of the 1992 flushing, it was found that the lined conveyance channel on the gently sloping reach just upstream of the confluence with the Rhone had completely filled with coarse This was attributed to the fact that the flushing flows Vlere typically half those originally anticipated during the design studies. It was therefore recornmended that the flushing flows be increased.
201
EVACUATION OF SEDIMENTS
Table A3.6. Year
Summary of the flushing at Gebidem, dam, 1982-93
Duration: h
Water volume used: M.m 3
Mean flushing flow : m 3/s
Gate operations *
1982
56
2·38
11 ·8
1983
48
3·38
19·6
Sediment removed: M.m 3
Solids concentration: t
0
0·143
6·0
2
0·175
5·2
6
0·178
6·0
%
I
I
1984
68
2·97
12·1
1985
49
2·50
14·2
0
0·150
6·0
1986
45
3·53
21·8
0
0·212
6·0
1987
45
3·20
19·8
13
0·192
6·0
1988
79
2·93
10·3
13
0·176
6·0
1989
49
2·49
14·1
1
0·150
6·0
1990
40
3·18
. 22·1
12
0·191
6·0
1991
96
2·35
6·8
a
0·270
11·5
1992
151
3·28
6·0
61
0·197
6·0
1993
101
2·48
6·8
?
0·260
10·5
* Number of intennittent closures and reopenings used t Where value is 6%, this is an assumed value used to estimate the sediment outflow
Table A3. 7.
Sediment balance at Gebidem dam, 1990-91 (12 months)
Parameter
Volume: m3
Annual sediment release Sediment released by flushing (95·5 h) Sediment passing through turbines
270000 70000
. ~
I·
Fate of sediment flushed
I
Deposited in gorge
81000
Deposited in aggregate works*
32000
-it
I
-
Delivered to Rhone Delivered to Rhone via turbines Total
* Aggregate supplier's extraction operation in river-bed downstream of dam
202
157 000 70000 340 000
f[
tf
APPENDIX 3
The release of sediment into the fast flowing Rhone has appeared not to have had any significant adverse effect on river morphology, but has helped maintain sediment loads in the face of a long history of gravel extraction. However, temporary high suspended-sediment loads and deposition on the river bed have caused some problems at water supply intakes, and have also been linked to fish kills in the Rhone.
A3.19. SANTO DOMINGO RESERVOIR (Venezuela, 1974) The Santo Domingo hydropower reservoir was formed by the construction of an 80 m high arch dam at the confluence of the Santo Domingo and Aracay rivers. The reservoir basin has two branches of similar length, both tributary valleys that have steep slopes, with river-bed gradients of the order of 4% to 6%. The valley bottoms are typically 30 m to 50 m wide (Krumdieck and Chamot, 1979). The maximum water depth is about 65 m, the longer branch is about 1 Ian long and the surface width is typically 100 m, giving a gross original storage capacity of about 3·0 M.m3 • The combined 427 km2 catchment of the two rivers, which is mostly covered by tropical vegetation and agriculture, produces a flow regime that allows the power station to operate continuously between April and October with an average inflow of 20-25 m3/s. In the dry season this is reduced to 5 m 3/s, so the. reservoir storage is used to support generation to suit the daily pe3.J."<: demand.
Al/9.1. Sedimentation There was scant information available to predict likely sediment loads at the time of design, but the adopted design suspended load of 160 000 m 3 per year (based . on lirnited actual data) was considered to be conservative. The bedload was estimated at half the suspended load for design purposes. Thus, there was expected to be the potential (with 100% trap efficiency) for the loss of up to 8% of the storage capacity per annum. ., During the first four years of operation, from 1974 to 1978, the scheme operator was required to continue generation without any interruptions for sedii'1lent flushing. Generally, the reservoir was held at the highest level possible at the time, as this minimised the passage of the highly abrasive sediment through the turbines, maximised the generating head and allowed for the easy release of floodwater over the spillway. Surveys of the reservoir bed carried out in February 1976 and April 1978 indicated that 0·58 M.m3 of deposition had occurred over two flood seasons, suggesting a rate of accretion of 0·29 M.m3 per annum, which was rather greater tt~an the design estimate of 0·24 M.m3 • This was attributed, at least partly, to road building and deforestation in the catchment. On the other hand, the estimated accretion over the first four years amounted to about 25 % of original storage capacity, so was a little less than might have been expected.
203
..
EVACUATION OF SEDIMENTS
A3.19.2. Flushing The dam is provided with three bOttOITI outlets for flushing sediment froin the reservoir, two in the deeper San Domingo river valley and one in the Aracay river valley. The outlets are each equipped with a 3 x 5 rn radial gate and a 3·2 x 2·5 m sliding gate. At normal top reservoir level (1585 m), the total discharge capacity of the three bottom outlets is 170 m 3/s. They were also considered large enough to deal satisfactorily with major obstructions, such as tree trunks. Hydraulic model studies during the design stage had indicated that sediment could be effectively flushed from the reservoir both under pressurised and freeflowing conditions, which_ are described below. • Pressurised flushing. In this condition,· the bottom outlets are submerged by sediment and flushing is started by inducing a 'piping' failure of the overlying sediments, which are discharged through the outlet at a high concentration; This causes sufficient sediInent to be eroded in order to provide a clear· pathway to the bottom outlet. Special attention should be paid during the operation to check on the size of sediment load and to ensure that an adequate flow is discharged, capable of transporting the material to the river-bed downstream. • Free-flow flushing. This is used when the outlets are clear of deposited sediment and usually begins when the level of the reservoir is already low and the sediment load is moving towards the central channel of the reservoir. Freeflow conditions are capable of flushing greater sediment loads, but at the cost of consuming greater volumes of water. Experiments showed that 12 000 n13 to 15 000 m 3 of sediment could be passed through the bottom outlets per day. ., First flushing operation. The first flushing of Santo Domingo reservoir took place in May 1978, after four years of operation, when the powerhouse was closed to· enable a complete inspection of the reservoir and the plant to be undertaken. It was estimated that the bottoin outlets flushed between 50% and 60% of the deposited sediment· in a period of only three or four days of freeflow flushing, at a time when the inflow was 8-10 m 3/s. The bottom outlets were covered by deposits by the fourth day, so the reservoir was allowed to fill overnight, following which flushing under pressure cleared the botton1 outlets. Some three weeks later, with inflows remaining low in both rivers, it was decided to attempt to accelerate the flushing operation by using two bulldozers to move deposits out of reach of active erosion towards the main strealns of each river. The entire flushing operation was sufficient to relTIOVe the majority of the sediment which had been deposited in the reservoir basin over the four years of operation. A subsequent topographic survey established that 0·62 M.m 3 of deposits had been removed, restoring the storage volume to 2· 85 M.m3 •
From the first flushing operation, the following conclusions were drawn (Krumdieck and Chan10t, 1979) specifically for the Santo Domingo reservoir, but also having application for other small reservoirs with heavy sedin1ent loads:
204
APPENDIX 3
flushing should take place annually, during, and preferably towards the end of the high-flow period 'I even under low-flow conditions, hydraulic flushing can be effective e flushing operations should begin when the sediment deposits are not less than 100-200 m from the face of the dam • free-flow flushing is generally more effective than pressure flushing, but freeflow flushing should be intenupted occasionally, to catTY out pressure flushing of deposits around the entrances and exits of bottom outlets (for up to 10 minutes at a time, eroding up to 5000 m 3). e
A3.19.3. Downstream impacts It was found that7 for relatively low flushing discharges, the concentration of sediments released could exceed the capacity of the downstream channel to convey theIn, resulting in accretion starting to obstruct the outlets. No information is given regarding environmental and other downstream impacts.
A3.20. NANQiN RESERVOIR (China, 1974) Nanqin reservoir, used for flood detentiQfl and irrigatioDLis situated i..11 the hillymountainous Shaanxi Province in southelTI China. The lnaxilnum depth (up to an impoundment elevation of 124 m) is 29 ill and the length of the reservoir is 4·5 km. The original storage capacity of the reservoir was 10·2 M.m3 (Chen and Zhao, 1992), which is 8% of the mean annual inflow. The reservoir's history can be divided into three phases: (a) between 1974 and 1976 it served solely as a flood-detention reservoir (b) betw'een 1976 and 1983 flows were impounded to a middle level of 110 In a..l1d released for irrigation (c) since 1984, with an improved regime of sediment management.
A3.20.1. Sedimentation The mean annual suspended-sediment load inflow is given as 0·53 Mt, the vast majority of which enters during the flood season, July to September. The annual average covers a wide annual variation, of between 0·12 Mt and 1·34 Mt in the period 1974-83 (Chen and Zhao, 1992). In 1975, the second year of operation, a major flood occurred which deposited gravel 3-4 ill thick, 1 krn upstream of the oliginal impoundment. By the end of 1983, 53% storage capacity in the Nanqin reservoir was reported to occupied by deposited sediment and it was estimated that the span of the reservoir would end by the 2000 (Chen and Zhao, 1992). (From the quoted data, this percentage loss apparently to an intermediate
205
EVACUATION OF SEDIMENTS
impoundment level of 118 m.) The Inaximum depth of deposition near the dam was of the order of 12 m.
A3.20.Z. Flushing A 3 m diameter tunnel, 3 m above the original river-bed level, was built into the dam for the purpose of sediment flushing. This has a discharge capacity of 14 m 3/s when the pool level reaches the soffit, rising to 110 m 3/s at maximum impounding level.
Density current venting Due to the steep bed slope of the Nanqin reservoir, density currents can easily "- -re'ach the dam and form a secondary reservoir of turbid water beneath the clear water layer. Sediment sluicing by density current venting began at Nanqin reservoir . in 1977. Between 1977 and 1984, out of the 2·43 Mt of suspended sediment entering the reservoir, it is reported that 1·56 Mt (64%) was discharged successfully by this method. Drawdown flushing Although the removal achieved by density CUITent venting was considered good, it was realised that more effective methods would be needed to deal with bedload deposition and to recover and preserve storage in the longer term. At the end of the 1984 flood season, an experimental flushing operation by emptying the reservoir was carried out. Flushing was carried out for a period of four days, in which all the sediment deposited in the current year was flushed out, along with 0·72 M.m3 of sediment that had been deposited in earlier years. The effective storage capacity was restored to the value that applied in 1980 (Chen and Zhao, 1992) and the maximum thickness of deposits reduced to about 6 m. Conclusions From the experience gained in the 1984 flushing test, the following operational rules were drawn up for N anqin reservoir: • the pool level should be kept high during the flood season to prevent bedload from advancing too far downstream and armouring the more erodible deposits • density current venting should be practised and the level of the turbid water reservoir kept below the floodplains at about 114 m • drawdown flushing should be undertaken at the end of the flood season once every 3-4 years, triggered by a storage depletion criterion. It was estimated that, if these principles are observed, a long-term storage capacity of the order of 7·5 M.1n3 (74% of the original) can be sustained. No details were given of the success of subsequent flushing operations.
206
APPENDIX 3
A3.20.3. Downstream impacts No information available.
A3.21. ICHARI RESERVOIR (India, 1975) Ichari dam is a 60 m high concrete gravity dam constructed across the River Tons, a tributary of the Yamuna, in 1972 (Mohan et a!., 1982). The reservoir, which is used for hydropower generation and was first impounded in 1975, is 11·3 kIn long, has a maximum depth of 37 m and an original storage capacity of 11· 55 M.m3• The mean annual inflow of about 5300 m}/s is about 450 times the original storage capacity.
A3.21.1. Sedimentation Suspended sediment inflows to the reservoir have been estimated by one or two daily samples, supported by more detailed sediment concentration profiles. There are no direct measurements of bedload entering the reservoir, but values have apparently been inferred from outflow sediment measurements and surveys of the reservoir basin. Between 1976 and 1984 the estimated total annual amounts of sediment inflow have ranged between 0·49 an.d 29 M.m3, with a median value of 2·2l\1.m3, which is about 20% of the original storage (Bhargava et al., 1987). The reservoir began impounding in March 1975 and the sediment deposited was surveyed after one year of operation, by which time it had reached the crest of the spillway, which is 16 m below the full reservoir level, reducing the storage capacity by 23%. Between then and 1981, the sediment level throughout the reservoir basin rose progressively, reaching a total storage loss of 60%. Table A3.8 summarises some key data regarding the sedimentation of Ichari reservoir (Bhargava et al., 1987). It is notable that over 90% of the very high sediment load in 1978-79 was passed downstream. In an inspection of the roller bucket of the gated spillway in 1984, severe damage of the teeth was found, including exposed concrete surfaces (which'Ii'ad been eroded sufficiently to expose the reinforcement) and steel plate armouring (some of which had been totally removed and washed away). The damage was att.ibuted to the impact of cobbles and pebbles passed through the spillway after the reservoir had silted up (Bhargava et aI., 1987).
PJ.21.2. Flushing The po'wer intake incorporates facilities for sediment exclusion (with the excluded sediment discharged downstream of dam), but no details are given of any facilities for flushing sediment from the dead storage of the reservoir basin. It is understood that the only facility for sediment flushing from the reservoir basin is by opening the spillway, which is done during the rainy
207
EVACUATION OF SEDIMENTS
Table A3.8.
Annual inflow and sedimentation datafor the Ichari reservoir
Year (June to May)
Total water inflow: M.m3
I
1975-76
Total sediment inflow: M.m3
Sediment trapped: M.m3
-
2·62
1976-77
5049
1·63
0·61
1977-78
6455
3·71
0·12
1978-79
7825
29·02
1·86
2·07
1·14
I
capacity: M.m3
I
Dead
Live
Total
3·93
5·00
8·93
3·40
4·92
8·32
4·80
8·20
4·25
6·34
1·60
3·60
5·20
2·09
.l
I
1979-80
3420
f ::
4585
4·80
0·60
1·14
3·46
4·60
5445
1·09
-{)·41
1·08
3·93
5·01
1982-83
4716
0·49
0·35
1·08
3·58
4·66
1983-84
5148
2·43
0·23
1·20
3·23
4·43
1980-81 1981-82
I
season, whenever the powerhouse is closed. The spillway gates are fully raised, to allow free flow through the reservoir along the top of the deposits. Measurements are made during these periods, froin which the quantities of sediment flushed can be calculated. It appears that flushing by this method has been undertaken annually since 1976-77 and accounts for between about 30% of the annual sediment discharge in years with low sediment loads, increasing to 70% or more in years with high sediment loads. It appears from the information given in Table A3.8 that the regin1e of annual flushing is likely to result in a fairly stable residual storage capacity of the order of 4 M.m3 (Atkinson, 1996), but no more recent data are available in the Iiterature.
A3.21.3. Downstream impacts It was reported that there is neither silting nor appreciable scouring in the
downstream reach of the liver. The sediment flushed from the reservoir is carried by the water discharged. No details are given of possible environmental impacts.
A3.22. BAtRA RESERVOIR (India, 1981)
Baira reservoir forms part of a hydropower project which utilises the combined flow of three tributaries of the River Ravi in the nOl1h west of India. The 51 m
208
APPENDIX 3
high elnbankment dam (earth core, rockfill shoulders) diverts ·the flow of the Baira River to a network of tunnels leading to the powerhouse. The oliginal storage capacity of the reservoir was 2-4 M.m3 (Paul and Dhillon, 1988) representing only about 0·1 % of the annual inflow (which is variously reported as 1900 M.m3 or 3500 M.m 3).
A3.22.1. Sedimentation The reservoir is subject to both monsoon and winter floods carrying high silt loads of up to 100 gil (Jaeggi and Kashyap, 1984). A mean annual rate of siltation had been estimated at 0·092 M.m3 , but in the first 18 months of operation, a silt volume of 0·45 M.m3 had accumulated, representing allnost 20% of the original capacity and suggesting an annual sediment load of at least 0·3 Mt (Atkinson, 1996).
A122.2. Flushing The 5 x 7 m diversion tunnel for the construction of the dam, with an upstream invert level of 1088 Ill, which is believed to be at least 35 m below the maximum impoundment level, was equipped with a service gate and an emergency gate to facilitate flushing. Model studies carried out during the design stage (albeit with a different design of flushing tunnel) had indicated that almost the entire silt content upstream of the tunnel could be flushed out. The first flushing operation. was undertaken in August 1983, with the objective of achieving the maximum possible volume of silt removal, adopting the following sequence: e e
1/
CD
e
e
the reservoir was drawn down to the minimum normal operational water level (1113 m), following which flows to the powerhouse were stopped the diversion tunnel was opened to allow a discharge of 150 m 3/s until reservoir was almost empty the diversion service gate was opened fuHy water was fed to t.~e reservoir from,the Suil and Bhaledh to clear silt from around the associated structures flushing ceased when the concentration in the discharge had decreased to about 10 gil the diversion tunnel gate was closed and the reservoir refilled.
The duration of the flushing operation (from ceasing to resl1nung power generation) was about 40 hours, with a maximum sediment concentration of 380 gil. The volume of removed was estin1ated to be 0·38 lYLm3 , representing over 80% of the which had occurred since impounding. The recommendations made future flushing were that it should be carried hours and that it would be more in out once a year for a period of APlil or iVlay, when the discharge from the Baira is about 100 m3/s (Jaeggi and
209
EVACUATION OF SEDIMENTS
Kashyap, 1984). Bearing in mind the steepness of the valley sides and the small size of the reservoir basin in relation to the annual inflow, it appears that the recommended flushing regime should be capable of maintaining a high proportion of the original storage capacity in the Baira reservoir.
A3.22.3. Downstream impacts No info1111ation available.
210
Appendix 4.
Erosion A4.1. FACTORS THAT AFFECT EROSION
A4.1.1. Definition A fundamental definition of erosion is the detachment and removal of rock particles by water and other geological agents such as wind, waves and ice (Mahmood, 1987). A broader definition would include the subsequent removal of material deposited temporarily at another point in the catchment. The rate of erosion is generally expressed as the mass of sediment removed from a given area per year (tlkm2/yr). It vruies with climatic, geological, topographic and human factors. Sediment yield expresses the quantity of material that reaches a defined point on a river draining the catchment and therefore the quantity entering a reservoir created by the construction of a dam at this point. The quantity will depend on the effectiveness of sediment transport in the basin. The majority of sediment yield studies consider only the suspended part of the total load. Bedload is generally assumed to be a minor part, representing about 10% of the total, even though in extreme cases it can vary between 4% and 60% (Jansson, 1988).
A4.1.2. Climate Precipitation The rate of erosion depends on the erosive power of the rainfall which is related to the intensity, droplet size and total quantity. High intensity, short duration events produce more erosion than long duration staTInS of low intensity. Storms with large rain drops are more erosive than drizzle with small droplets (Goldman et at., 1986). Tne effect of rainfall intensity is illustrated in Table A4.1 by data for 183 events which caused erosion at Zanesville, Ohio, between 1934 and 1942. They show that the average soil loss per rain event increases with the intensity of the storm (Fournier, 1972, reported by Morgan and Davidson, 1986). Seasonal variations in erosion rates are influenced by previous meteorological conditions. The moisture content of the soil and hence the infiltration capacity will depend on previous rainfall and this will affect the amount of run-off which in turn has a direct on erosion rates (rv10rgan and Davidson, 1986). Highest erosion rates are likely after a long dry period when there will be a supply of readily erodible material (Morris and Fan, 1997).
211
J EVACUATION OF SEDIMENTS
] Table A4.1.
]
Relationship between rainfall intensity and soil loss
Maximum 5 min intensity per rainfall: mm/hr
Number of falls of rain
Average erosion: kg/m2
40
0·37
25·5-50·8
6]
0·6
50·9-76·2
40
1·18
76·3-101·6
19
1·14
101·7-127·0
l3
3·42
127·1-152-4
4
3·63
152·5-177·8
5
3·87
177 ·9-254·0
1
4·79
0-25·4
I
]
]
], j i
Many studies have been carried out to determine a relationship between erosion and rainfall and SOine of these are illustrated in Figure A4.1. The relationship proposed by Langbein and Schumln (1958) between annual sediment yield and effective · precipitation has been widely documented and utilised (Walling and Webb, 1983). Maximum sediment yields occur at an annual effective precipitation of approximately 300 nun (i.e., semi-arid regions). In areas with higher rates of effective precipitation, vegetation growth is increased and the sUlface is protected. In more arid areas there is insufficient rainfall to move materiaL The relationship uses the term effective precipitation, defined as the annual precipitation required to generate the given annual run-off at a standardised mean temperature of 50°F and not the standard definition of precipitation minus evapotranspiration. There are many possible sources of error in this study, the main one being that the results are based on only 94 sampling points in the USA. A number of alternative relationships have been deIived. Results produced by Judson and Ritter (1964) were based on the average suspended-sediment yield and annual run-off for the seven major drainage regions in the USA. Work by Dendy and Bolton (1976) used the group-averaging technique to generalise the relationship based on .data from 500 locations within the USA. The peak sediment yield occurred at a mean annual run-off of 25 lnrnto 75 nun which is equivalent to an effective rainfall at 50°F of 450 nun to 500 mm, higher than the 300 mm at which peak yields were predicted by the Langbein and Schunun curve. Relationships based on global data show less similarity to the Langbein and Schumm curve. There is a general increase with precipitation at the lower end of the scale but values increase again when annual precipitation and lun-off exceed 1000 mnl and 500 nml respectively. The relationship produced by Wilson (1969) has two peaks at 750 mm and at 1750 mm of annual precipitation. These coincide
212
]
J
~.
APPENDIX 4
with sub-hulnid and tropical conditions and contradict the_peak demonstrated by Langbein and Schumm for semi-arid regions. The relationship proposed by , Tabuteau (1960) demonstrates a more complex pattern with a wide range of yields in all regions. A study by Walling and Kleo (1979) based on data fron1 1246 global measuring stations showed no clear pattern between mean annual precipitation and mean annual suspended-sediment yield (see Figure A4.2). A relationship between precipitation and sediment yield group-average data for drainage basins with an area less than 10000 km2 can be seen in Figure A4.3
800
r
L
>-
L
C)'
><'t'
600
E
.:s::
E
-
~ Qi
Qi
';:;"
"E CD E
E '5 200 CD
~
~
's:,
400
"E
tlJ
'6
U)
w
en
250
500
750
0
1000 1250
0
Effective precipitation: mm (i)
1250
L
>-
c::::
-s:,
a:;
1:: tlJ E
1:: 500 tlJ E
~
~
~
a:; 100
10
'5
J
1'
tlJ (j)
150
E
100
10
100
1000 10000
I
0
400
800
1200
1600
(ii)
L 1200 >-
t
~
~ 800
-
-c Q.i
':;'
'5-
C
50
w
0
en
25:
I
(i)
><'t'
'6
750
';:;"
-L
-c a:;
I
l-
tlJ
1200
1000
;;'1000
E
oX
en
800
-L
1000
I
'6
600
Mean annual run-off: mm (ii)
(a)
10 000
C\I
400
200
'E 400
'0 Q) (j)
0
200
400
600
800
1000
400
(iii)
Figure A4.1.
800
1200
1600
2000
2400
Mean annual precipitation: mm
Mean annual run-off: mm (b)
(iv)
Sedirnent yield and annual precipitation: (a) USA; (b) world
213
EVACUATION OF SEDIMENTS
and shows the first peak at a preCIpItation of 450 mm, which reflects the relationship proposed by Langbein and Schumm (1958). The two initial peaks are similar to the curve produced by Wilson (1969) but the magnitudes of the peaks are 450 mm compared with 750 mnl, and 1350 mm compared with 1750 mm, which conesponds to the troughs in these curves. The relationship between sediment yield and Inean annual run-off is similar to that produced by Douglas (1967) although the peak in semi-arid areas is less pronounced, with maximum yields occurring in areas of high annual run-off. Other factors such as relief, geology and human impact may be more important controls at a global scale than precipitation. The seasonality, intensity and type of rainfall and its effect on vegetation cover are also important measures of the effect of precipitation on erosion rates . • Run-off. Run-off results from an excess of precipitation over the sum of infiltration and evapotranspiration and is the quantity of water available to convey the products of erosion. Factors that encourage infiltration and thereby reduce run-off will decrease the quantity of erosion. Run-off may provide a better correlation with rates of erosion than precipitation does. Low run-off rates indicate aridity and hence poor vegetation cover compared with high rates that indicate an excess of water and therefore dense vegetation cover. A
10000 '0
.... :.
I
>-
~
•
. . ~: : ..
..
00
E
-
~
o
1000
•
•
i:i
Q) ':;"
c Q)
E :0 (!) C/)
"C
Q)
100
"C
c: Q) c.. C/)
:
:J C/)
(ij :J
c: c:
ns ns Q) :2 c:
10
1L-__
o
L-~
200
__
400
~
__
600
~
__- L_ _- L_ _- L_ _
800
1000
1200
~
__
1400 1600
~
__
1800
~~~
____
~
2000 3000 4000
____
~
6000
__
~
__
8000
Mean annual precipitation: mm
Figure A4.2. tion
214
Mean annual suspended-sediment yield versus mean annual precipita-
APPENDIX4
similar general relationship will exist between run-off and erosion as for precipitation and erosion; with maximum erosion "levels at intelmediate values. o Temperature. The temperature will affect vegetation growth and evapotranspiration rates. Where there are high temperatures, higher rates of evapotranspiration occur and therefore larger amounts of rainfall are required to cause erosion. High temperatures will also cause more rapid rates of vegetation growth which will reduce run-off rates and erosion. The ilnportance of temperature will depend on the quantity of precipitation. In regions with high precipitation quantities the relative importance of temperature is likely to be reduced. • Wind speed and direction. The wind speed and direction will affect the movement of soil particles. In areas where the wind speed is high and there is a lack of vegetation to hold the soil pa.'1:icles togell-J.er high rates of wind erosion are likely to occur. Wind erosion is inlportant in arid or semi -arid regions as an agent that can transport sediment from ridges into depressions which can then be transported by run-off.
1200
r-
~
•
j\ .. · .. \. //
4DO
~
~
\.of
;/0 0 \./
1
Qi
's;.
c
I
I
I
o
400
800
I
Q)
ca::l
,
1600
1200
2000
2400
2800
3200
3600
4000
Mean annual precipitation:mm .
E
'6
o~
0/\ ./
800
1200
c::
§
c::
ctI
Q)
~
800
0
400 • • • • •
------------
~.~./.
0\.••/ . 0 1L-____- L______L -____
a
400
800
~
______
1200
~
1600
____
~
______
2000
~
2400
____
~
____
2800
~~
3200
Mean annua! run-off: mm
Figure A4.3.
Mean annual suspended sediment yield versus mean annual run-off
215
EVACUATION OF SEDIMENTS
J Table A4.2.
Effect of geology type on erosion rates (Jolly, 1982)
Lithology
Sediment loss, Utah: m3/km 2/yr
Resistant: conglomerate, limestone and resistant sandstone Medium: fliable sandstone Soft: shale and gypsum
I
Sediment loss, New Mexico/ Arizona: m3/km 2/yr
143
95-143
571
523
1237
761
A4./J. Geology Rock type The geology of the area has a major impact on the rate of erosion as it determines the susceptibility of the rock to the effect of other factors. Table A4.2 shows that under the same climatic conditions geology can produce a variation of ten times in the sediment loss. Generally, suspended-sediment loads are greater in areas of sedin1entary rocks compared with crystalline rocks by a factor of 2-4 and .compared with areas of mixed rocks by a factor of 1·4 (Dedkov and Mozherin, 1992).
J
J .J J
Volcanic and tectonic activity Recent tectonic activity influences the intensity of erosion especially in mountainous areas. An increase of approximately two fold was found for an increase of earthquake activity by 1 point in USSR (Dedkov and Moszherin, 1992). High rates of erosion are seen in zones of young mountains, e.g. Himalayas and New Zealand.
A4.1.4. Soils The soils in arid and semi-arid environments with sparse vegetation cover are very different from soils in more humid regions. The key soil ·characteristics influencing erosion rates are the texture, structure, organic matter content, shear strength and infiltration capacity.
Texture The texture describes the sizes and proportions of the particles making up the soil. Soils with high sand contents are coarse textured with high infiltration rates, low run-off and relatively low erosion potential. Soils with a high content of silts and clays are fine textured, the clay binds the soil and makes it resistant to erosion. Soils high in silt and fine sand and low in clay and organic matter are the nlost erodible. Well drained sandy and rocky soils are the least erodible as they have large particles which require large forces to transport theln (Goldman et ai., 1986).
216
:
i
.,.....L
APPENDIX 4
Soil structure Soil stnlcture is the anangement of particles into The soil structure affects the soil's ability to absorb water. When the soil is compacted or crusted, water tends to run off rather than infiltrate. Granular structure is the most desirable to minimise erosion as it absorbs and retains water, reduces run-off and encourages plant growth (Goldman et al., 1986). Organic content Organic matter improves the soil structure and increases permeability, water holding capacity and soil fertility (Goldman et al., 1986). Clay content can be used as an indicator of erodibility as it combines with organic matter to form soil aggregates and it is the stability of these particles which determines the resistance of the soil. Soils with an organic content of less than 3 ·5% are highly erodible. Shear stren~J1 This is a measure of the cohesiveness of a soil and its resistance to shearing forces exerted by gravity, moving fluids and mechanical loads. Its strength is derived from frictional resistance met by its constituent particles when they are forced to slide past one another or to move out of interlocking positions. The higher the shear strength of a soil the more resistant it is to erosion (Morgan and, Davidson, 1986). Infiltration ratelpermeabiliv/ infiltration capacity is the maximum sustained rate at which soil can absorb water and is influenced by pore size, pore stability and the form of the soil profile. Soils with stable aggregates maintain their pore spaces better while soils with swelling clays or minerals that are unstable in water tend to have low infiltration capacities. Where infiltration varies with depth, the horizon with the lowest infiltration capacity is critical. Texture, structure and organic matter all contribute to the pelweability of a soil. High erosion rates occur where rates are low and large volumes of run-off are ....,"". . . ""............ ' -' .
...l ... .l,u, .............. v
,...
A4.1.S. Catchment characteristics Slope gradient and length of the slope directly influence the velocity of run-off and its erosivity. The energy and, therefore, the erosive potential of flowing water increases with the square of the velocity (Goldman et al., 1986). Long, continuous slopes allow run-off to build up momentum and the base of the slope becomes more susceptible. On a flat surface the raindrop splash is random, nTt-,,,,,,.,::l-:lC on sloping ground more raindrops are splashed downSlope than upslope moving sediment with them. Orientation of catchment Southern facing slopes in the northern hemisphere are eroded more rapidly than ·~~"',o ... facing slopes as they are hotter and drier with less dense vegetation and
1"\ .....
't"I
217
EVACUATION OF SEDIMENTS
they experience greater fluctuations in air and soil temperature. North facing slopes are cooler and more moist with less sun.
Drainage basin area An inverse relationship has been demonstrated between sediment yield per unit area and catchment area (see Figure A4.4). In larger catchments there is a lower overall slope, smaller percentage of erodible rock and more opportunity for sediments eroded from steeper slopes to be deposited in the floodplain (White, 1982). In China approximately 50% of the sediment load from the Yangtze River is produced by 13% of the catchment and 43% of the sediment in the Yellow River comes from 7 % of the area. A number of recent articles have questioned the standard relationship between suspended sediment and drainage area. Church and Slaymaker (1989) suggested that in British Colon1bia specific sediment yields increased downstream in catchment areas of up to 30 000 km2 due to remobilisation of quaternary sediments stored in the valley and channel systems (Walling and Webb, 1996). Dedkov and Moszherin (1992) proposed that river systems are characterised by positive or negative relationships depending on the relative importance of channel and slope erosion. Where channel erosion dominates, in areas with dense vegetation cover, erosion rates increase downstream. There is a positive relationship with drainage basin areas due to greater entrainment and transportation of sediment. Where slope erosion is dominant, erosion is concentrated in the headwaters and a proportion of the mobilised sediment will be deposited during transport through the system. There is therefore an inverse 105
x x
104
E
-
~
i::i 05 ";;:.. C ~ '5 Q)
o
+
103
o o o
102
1::..
o +OX
o
0
10 1
0
o +
q. 0
o
+
C/)
o
+
1::..0
~
-k
M
+
x
1::..
8
10oL-~--~---L--~--~~~~--~--~--~--~~
10-3
10-2
10-1 Drainage area: 106 km2
Figure A4.4.
218
Sedilnent yield versus drainage area
100
101
APPENDfX4
relationspip between suspended sediment yield and drainage basin area (Walling and Webb, 1996),
Drainage density Drainage density is all expression of the distribution of streams in the drainage system. It is a crude indicator of run-off and is often used as an index of the severity of erosion - areas of high drainage density being associated with elevated erosion rates. Broadvariations in drainage density on a macro-scale are associated with differences in climate. At the meso-scale, regional variations can be related to differences in rainfall volume but are complicated by lithology and relief. Micro-scale differences in soil type and frequency and intensity of individual climatic events are important (Morgan and Davidson, 1986). Vegetation Vegetation is the most important erosion control factor. It dissipates the energy of rainfall, prevents rain impact on the ground, reduces splash erosion, increases infiltration, decreases surface nln-off volumes and velocity, holds soil particles in place and maintains the soil's capacity to absorb water. The type of vegetation cover is dependent on the raiilfall, temperature, soils and topography of a region. These factors interact to produce distinctive zones called biomes. Climates with relatively mild year-round temperatures and frequent~ regular rainfall·are favourable to plant growth. Cold and dry climates are less favourable to growth and therefore more susceptible to erosion (Goldman et ai., 1986). Land use The land use of an area is influenced by the topography, geology, soils and climate of a region. These factors determine the use to 'which land is put by humans. Cultivation may decrease the erodibility of clay soils but increase that of sandy soils (Morgan and Davidson, 1986). Soil loss from hillslopes in West Africa between a gradient of O·J:) and 4° experienced mean annual erosion rates of 0·015, 0·02 and 0·003 kglm2 under natural conditions of open savanna grassland, dense savanna grassland and tropical rain forest respectively. Clearance of the land for agriculture increased rates to 0·8, 2·6 and 9·0 kg/m2 while leaving the land as bare soil produced rates of 2, 3 and 17 . The removal of rain forest produces greater rises in erosion rates than the removal of savanna grassland (Morgan and Davidson, 1986). Areas of low precipitation are more vulnerable to land use changes. Changes in one part of the ecosystem may produce changes in the basin condition and In semi -arid regions the response and recovery may take a long period of recovery time is four times that of humid areas (Walling and Kleo, 1979). on erosion rates can The effect of cultivation or lack of soil cover by be seen in Table A4.3. to bare ground can also The increase erosion rates from natural be seen graphically in Figure A4.S which shows the results of soil erosion tests under different vegetation cover at Mpwapwa, Tanzania.
219
EVACUATION OF SEDIMENTS
Table A4.3. Rates of erosion in selected countries in kg/m2/yr (after Morgan and Davidson, 1986)
China'
Natural
Cultivated
<0·2
15-20
Bare soil
28-36 I
USA
0·5-17
0·003-0·3
0·4-9·0
0·003-0·02
0·01-9·0
1·0-75·0
Nigeria
0·05-0·1
0·01-3 ·5
0·3-15
India
0·05-0·1
0·03-2·0
1·0-2·0
Belgium
0·01-0·05
0·3-3·0
0·7-8·2
UK
0·01-0·05
0·03--0·3
1·0-4·5
Ivory Coast
Human impact
It is estimated that human activities have degraded 15% (200 million ha) of the land between 72°N and 57°S. Around half of this is due to hUlnan-induced water erosion, a third due to wind erosion with most of the balance due to chemical and physical deterioration (US Global Change Research Information Office, 1999). Activities such as deforestation, urbanisation and agriculture all increase the erodibility of soil. Present rates of erosion are approximately two and a half times historic rates mainly due to human influences. With the conversion of forest to agricultural land there Inay be an increase in sediment yield at the basin mouth by three and a half times (Mahmood, 1987) - see Table A4.4. Explanation:
Ungrazed thicket
Soil lost by erosion, t per acre
Ot 1·9%
rI
Water lost by run-off per cent of rainfall
26·0%
50-4%
Figure A4.5.
220
Soil erosion at Mpl,vanga, Tanzania
;
J. . i
APPENDIX 4
Table A4.4.
Erosion rates for different land use categories (Morris and Fan, 1997)
Land Llse
Forest
8
Under natural conditions erosion rates in mountain zones are 27 times greater than in lowland areas. The influence of man has increased sediment yield from mountainous regions by 1·4 times; however, larger increases in lowland areas have reduced the difference between mountainous and low land regions to 3·2 . times. For eX3J.I1ple, sediment yields in the sout.h and middle Urals are up to ~ 30 tlk..rn2/yr, less than the neighbouring eastern part of the Russian plain where rates of up to 200 tlkn12/yr occur (Dedkov and Moszherin, 1992). Increases in sediment yield caused by human activity are demonstrated in Table A4.S.
A4.2. C.A.SE STUDIES OF EROSiON RATES
A4.2.1. Erosjon rates in Africa The annual sediment yield of rivers in Africa for drainage basins around 10 000 kIi? is between 1 tJkrn2 and 4000 t1km2 • Four-fifths of the surface area of Africa produces less than 100 t/lan2/yr, highlighting L,e regional variation in erosion rates (Shahin, 1993).
The Upper Tana basinJ eastern Kenya The area of the Upper Tana basin is 9250 km2 and can be subdivided into three areas depending on altitude: e $
10
Mount Kenya volcanic summits and the Aberdare range slopes of J\.1ount Kenya 11wea-M~singa plains.
221
I'
EVACUATION OF SEDIMENTS
Table A4.S. 1983)
Increases in sediment yield due to land use changes (Walling and Webb,
Region
I
Land use change
Factors for increase in sediment yields
Source
Rajasthan, India
Overgrazing
4-18
Utah , USA
Overgrazing
10-100
Noble (1965)
Oklahoma, USA
Overgrazing and cultivation
50-100
Rhoades et al. (1975)
Oklahoma, USA
Cultivation
5-32
Rhoades et al. (1975)
Texas, USA
Forest clearance and cultivation
340
Chang et al. (1982)
Northem California, USA
Conversion of steep forest to grassland
5-25
Anderson (1975)
Mississippi, USA
Forest clearance and cultivation
10-100
Southern Brazil
Forest clearance and cultivation
4500
Bordas and Canali (1980)
Westland, New. Zealand
Clearfelling
8
O'Loughlin et al. (1980)
Oregon, USA
Clearcutting forest
39
Fredriksen (1970)
Sharma and Chattelji (1982)
Ursic and Dendy (1965)
The altitude has a distinct effect on the climate causing a knock-on effect on the vegetation of each zone, as shown in Table A4.6. There are two rainy seasons: from March to May and from October to December. Sediment yields from the forested areas of Mount Kenya are around 20 t/km2/yr rising to 1000 t/km2/yr on grazed areas and lnore than 3000 t/km2/yr on steep cultivated areas of the basin. This shows the strong influence of land use on sediment yield. Under natural conditions cultivated areas have ground cover of crops for around eight months of the year. Soil losses from grazing lands are therefore generally higher and increase markedly as basal cover declines. In cultivated areas, rural roads yield 5-20% of the sediment yield (Ongweny,1979). Table A4.6. Altitude: m
Vegetation
Rainfall: mm
Soil
>1800
1800
Dense evergreen forest
Clay loam
>1400
1400-1800
Steep cultivated slopes
Clay loam
1100-1400 <1000
222
Variations in rainfall, vegetation and soils with altitude (Ongweny, 1979)
900-1400
Farming Marginal farming
APPENDIX4
Comparisons between the data for the Upper Tana basin and that produced by Dunne, relating mean annual suspended-sediment yield to annual run-off depending on land-use, showed good agreement. This can be seen in Figure A4.6, which shows the relationship between mean· annual suspended-sediment yield and mean annual run-off.
E.rosion rates in the grazing lands of Kenya The Athi-Kapiti plains within a 50 km radius to the south and east of Nairobi consist of a dissected plateau developed on cernozoic tuffs and lavas. The mean annual rainfall is 5 to 700 rom with grasses that cover 40-90% of the ground below a sparse (5%) canopy cover. The soils vary from planosols on the ridges through phaeozems to vertisols on the footslopes. 180 km south east of Nairobi, quaternary lavas extend northwards as a stepped plateau from the slopes of Mount Kilimanjaro where the meatl arulual precipitation is 450 mID. Soils are 10 cm to 200 cm thick sandy-clays. The ground cover is less than 10% with a sparse (10-30%) canopy of dry woodland and bush. Between the two volcanic plateaus lies a belt of precambrian basement schists north of the Amboseli basement. The region receives 300 mm of rainfall and the vegetation cover is grassland (10-40%) and bush with a canopy cover of up to 40%. The soils are sandy clays between 50 cm and 150 cm deep.
-- --
__~-L__~~~~~~--~~--~--~~--~ 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 Mean annual run-off: mm
1~~~
a
1
2
III 0 Forest
r:.
•
o
Forest> Agriculture • Agriculture> Forest
0
0
A
Grazing Scrub lorest
1-UpperTang values plotted on Dunne's curve 2-Dunne's values
Figure A4.6.
Mean annual suspended-sediment yield plotted against annual run-off
223
EVACUATION OF SEDIMENTS
In all three regions the slopes are longer than 500 m with only a small portion of gradients over 0·1. The n1aximum rates of erosion occur at the central, steepest part of the hillslopes. The sediment yield on the slopes is 2500 t/km 2/yr to 3150 tlkm2/yr, on the basement rocks 11 300 tlkm2/yr and on Kilimanjaro lavas 17 600 t/lGl1 2/yr (Dunne et a!., 1979).
Sediment loads in the Orange River, South Africa The sediment load is predominantly small particles less than 0·2 mm. The average sediment yield from sub-basins of the Orange River vaIies,from less than 10 t/km2/yr to lTIOre than 1000 t/km2/yr. The highest yields are from the solonetzic soils and valleys filled with sediments on the westelTI side of Lesotho with lowest yields from the sandy permeable regions in the north-west drainage region. Soils ~roIP- areas with a high silt and:-Iow clay content are generally the most erodible (Rooseboom and Van Harmse, 1979)
A4.2.2. Erosion rates in Thailand The rates of sediment production in Thailand range from a value of 8 t/km2/yr to 3874 t/km 2/yr with the highest rate experienced in the Lam Dome Noi River catGhment, a tributary of the Mun-Chi River. The lowest rate was in Huai Pa Tao River Basin, a tributary of the Chi River (see Table A4.7). The highest rates seen in the north-eastelTI region are due to the highest rates of deforestation (Jantawat, 1985).
A4.2.3. Erosion rates in China The erosion rates in China are highly variable. High rates occur in northern China where silt contents are high, the climate is dry and there are extensive areas of loess. The climate in southelTI China is generally \vet and warm producing a better vegetation cover and therefore lower erosion rates (Lagwanker et a/., 1995).
Table A4.7. Region
224
Erosion rates, Thailand Rate of erosion: tlkm 2/yr
NOlthern
12-2045
Central plains
20-570
East
27-356
South
30-1787
. - :!
. i
-~.
APPENDIX 4
Table A4;8.
Sediment contribution for sections of the Yellow River (Tal Wei Soong and Yean Zhao, 1994)
I
Upper
Area: km'
385000
I
Length of channel: km
1
Middle
I
Contribution: % water
Contribution: % sediment
111000
I
48-7
9-0
36-6
89·3
1
345000
1206
I
111400
!
22000
786
I
1/8000
I
1
Lower
I
3472
I
Average slope: m/m
I
11
I
1·7
Lower Yellow River The Yellow River is· the second largest river in China at 5464 km long with a drainage area of 752400 k..rn? The sediment contlibution is mainly from t.l}e middle reaches that pass through the loess plateau (see Table A4.8). The rainfall is seasonal with the majority occurring between July and October. The run-off during this period can account for 60% of the annual discharge and sediment inputs 85% of the annual sediment totals. The rainstorms are generally of high intensity and short duration and sediment delivered from the catchment reaches 91·1 kg/m3 (Tai VVei Soong and Yean Zhao, 1994).
Sanmenxia reservoir The drainage basin to the reservoir is 688 400 km2, constituting 92% of the Yellow River basin. The Yellow River drains China's semi-arid loess plateau composed of thick aeolian deposits of silty soils. Due to the high erodibility of this soil, intensive land use, inadequate soil conservation practices and virtually limitless supply of sediment, the load through the valley is high. Rates of sediment trarlsport are especially high in July and August which account for 60% of the total annual sediment yield and 30% of the annual run-off. Sediment discharge averages 1·6 billion tlyr, equivalent to an annual sediment yield of 2300 t!k..rn2 and a mean suspended-sediment concentration of 38 gil ~Aorris and Fan, 1997).
A4.2.4. Erosion rates in India A relationship between drainage basin area and average annual silt deposits was calculated by Lagwanker as follows: Catchment area above 2500 km2 Catchment area between 2500 and 100 km2 Catchment area below 100 lan2
S = 0-065 A S = 0·278 AO- 8i5 S = 0·200 A 0-887
where S = average annual silt deposits in ha per m A = catchment area in km2 The relationship was based on selected reservoir sites in various areas of India (Lagwanker et aI., 1985). '
225
EVACUATION OF SEDIMENTS
Erosio-n and sediment yield in Krishna River Basin
The Krishna River rises in the Western Ghat mountains at an elevation of 1400 m and flows 1400 k:m to the Bay of Bengal. 40% of the area is mountainous. 80% of the basin is formed on archaen and younger crystalline rocks and 20% on Deccan Traps and recent sedin1ents. The discharge is 3 m 3/s to 3400 m 3/s with the majority of the flow occurring during the monsoon in July and August. The maximum sediment yield occurs between July and October with 95% of the annual sediment load derived froln the monsoon period. The erosion rate is highest in small sub-basins of the Krishna dver where rates are up to 4000 t1km2/yr (Subramanian, 1982).
A4.2.S. Erosion rates in the Himalayas The Sapt Kosi is the third largest river with a source in the Himalayas. The three main tributaries are the Sun Kosi, Arun and Tamur. Precipitation falls 89% as rain and 11 % as snow with 80-85 % of the total rainfall in the monsoon months from June to August. 85% of the run-off occurs between June and September with sediment load 98% of the annual total. The area to the gauging station is 59 000 km2 with an average annual run-off of 53 000 M.m3 • The average annual sediment yield is 2800 t/km2/yr composed of 16% coarse sand, 29% medium sand and 55% silt and clay (Mahlnood, 1987).
A4.2.6. Erosion rates in Pakistan The River Indus carries large alnounts of sediment particularly in the spring and summer when melting snow leads to high rates of erosion. The source of the river is in the Tibetan Plateau at a height of 5500 m above sea level. There are two distinct hydrological regions in the 169 650 km2 drainage basin upstream of the Tarbela dam. About 90% lies between the Karakoram and Himalayan mountain ranges, from which the meltwaters contdbute a significant proportion of the flow. About 10% iImnediately upstream of the dam is subject to monsoon rainfall between July and September. The average annual rainfall is 890 mm with twothirds occurring between June and October. The climate is subtropical and semi-arid in the headwaters. The summer season is from April to September and the winter. from October to March. Average daily temperatures range from 7°C in January to 41°C in June. The mean annual sediment inflow to Tarbela reservoir is 200 Mt equating to a sediment yield of 1179 t/km2/yr (Attewill et al., 1998).
A4.2.7. Erosion rates in Puerto Rico The climate is wet and tropical and the topography is rugged and hilly with elevations up to 1337 m. The rainfall is 1500 mm/yr to 2800 mmlyr and is variable, being orographic in nature. The total sediment load is 37 600 t1yr of sediInent based on 650 tlkIn2/yr for a 58 km2 drainage basin. The geology is
226
APPENDIX 4
tough and easily erodible siltstone or sandstone with nlore granitic rock above Caonillas reservoir. Loiza reservoir
The region has an average rainfall of 1900 mmJyr (841 mrn run-off) and teluperature of 25°C. The dam impounds 534 kn12 of the Loiza catchment of which over half the area has slopes greater than 35°. The sediment yield of the region is high, between 1000 tfKm2/yr and 2000 ttlCIl12/yr (Monis and Fan, 1997).
A4.2.B. Erosion rates in Switzerland 65 % of the 200 krn 2 catchment above Gebidem reservoir is occupied by d'Aletsch glacier, the largest in Europe. The glacial activity leads to denudation rates of 2·5 mmlyr, more than an order of magnitude greater than unglaciated catchments in the same area. The sediment load is 400 000 m3/yr or 2000 m 3/km2jyr of cohesionless material from fine to gravel. The discharge is seasonal and dominated by snowmelt, glacial melt and summer stonns (Morris and Fan, 1997).
A4.2.9. Erosion rates in Turkey Seventeen dams in Turkey were studied to produce estimates of sediment yield depending on catchment characteristics like drainage area, soil type, erosion intensity, slope, annual average precipitation, water discharge, kinetic energy of rainfall and stream power. To find a relationship the actual rates of sediment yield 'were· determined either by using hydrographic resurvey results or using data from the sediment gauging stations of the Electrical Power Resources Survey and Development Administration. Sediment yield rates in t/km2/yr were converted to m 3/km?/yr by dividing by the density of sediment, taken to be 1·8 tim3 • This figure was factored by 1·2 to allow for 20% bedload. The relationships were (Gogus and Yener, 1997);
Qs
= 13·959*A
1,213
(r
=0·946)
Qs == 0·024*A 1.002*Ei}509*p~.994*Q~185 (r == 0-96)
Qs
=0·036*A I.039*S~O.I08*Ei).73*SO·061*p~.855*Q~122*p~.063 (r =0-961)
where A == catchment area (km2) St == soil type
E: == erosion index S == slope of telTain P r = annual average precipitation (mm)
227
EVACUATION OF SEDIMENTS
Qw = water discharge x 106 (m3/yr) P s =stream power (kgf/yr)
A4.2. f O. Erosion rates in Spain Sixty drainage basins were studied, ranging in area from 31 km2 to 16 952 km2 with specific sediInent yields from 8-4 t/km2/yr to 2703 t/km 2/yr, with an average of 429 tJkm 2/yr. The basins are divided into groups. Group I Specific sediment yield less than 150 t/km2fyr. Twenty catchments in zones where erosion processes are slow, or where carbonate lithologies predominate.
Sf = 617 AO- 67
(r
=0·77, n =20)
Group 2 150-1000 t/km 2/yr. This group contains over 500/0 of the reservoirs.
SY =202 A I -07
(r
=0·92, n =33)
Group 3 Over 1000 t/km2/yr. This group includes 7 basins which are less than 420 km2 so that eroded sediment is likely to be transported into the reservoir.
Sf =3137 Ao- s7
(r=O·91,n=7)
These relationships describe the sediment yield at the reservoirs, which is not all the material eroded. The sediment delivery ratio varied from 0·8% to 67-470/0 with most ratios less than 25%. Those catchments with the lowest specific sediment yields are not always those with the lowest erosion rates. The surface area and location of the sediment source, relief, slope, transpo11 system and vegetation cover all affect the delivery ratio. The predominant factor is the drainage basin area (Salas et aI., 1997a and 1997b)
A4.2.11. Erosion rates in Canada The 4000 km2 Oldman river basin extends eastwards from the Rocky Mountains in south-west Alberta. The area can be divided into mountains, foothills and high plains. The suspended-sediment yield averages 70 tlkm2/yr cOlnpared with a range of 20 to 100 elsewhere in Alberta and up to 350 t/km2fyr in Canada generally (Neil and Mollard, 1982)
228
APPENDIX 4
A4.3. CLIMATE CLASSIFICATION
A4.3. I. Koppen classification Of the many climatic classifications produced, one of the most durable is the Koppen classification which was originally based on eight climatic regions. Others have since refined it and a recent version published by UNESCO in 1990 is illustrated in Figure A4.7.
A4.3.2. Thornthwaite .il..n alternative climatic classification was put forward by Thornthwaite (see Table A4.9), based on the relationship between precipitation and potential evapotra...l1spiration. The classification calculates an index I which expresses the relationship between surplus moisture, 5, moisture deficiency, d,and potential evapotranspiration, pe.
1= (5 - d)lpe x 100 The zero value separates moist (positive) from dry (negative) climates. To incorporate a thermal parameter in the classification Thomthwaite used potential evapotranspiration as this parameter expresses water need as a function of temperature and length of day (UNESCO, 1990).
A4.3.3. Alisov A classification by Soviet climatologist Alisov is based on the conditions of circulation of the atmosphere. He identified seven main climatic zones which can be seen in Figure A4.8. Each zone is characterised by the predominance of the air mass corresponding to the name of the zone (UNESCO, 1990).
A4.3.4. Map
of aridity
A map of aridity based on the Thornthwaite classification was compiled by UNESCO. The aridity indices were determined in this case by examining the ratio of annual precipitation, P, to annual potential evapotranspiration, ETP. Four aridity classes were identified: Hyper-arid Arid Semi-arid Sub-humid
(PIETP (PIETP (PIETP (PIETP
< 0·03) < 0·2) < 0·5) < 0·75)
In addition to the four aridity classes, temperature was used to further define the arid regions. The subdivisions were warm winter, mild winter, cool winter and
229
m
IV
w
< » ()
(:)
c
» --I 0 Z
0
-n l/)
m
0 3:
m
Z
--I
l/)
A. T,op"=-, ,ainy clim.',"
8. Dry cli~l"
C. Humid nwtOlMtm.1 cllmal"
O. Humid mlcrotMfma' ctifNl"
~ T,opiul fllinlorn,IAf. Ami ~ T,opic.1 .....nN IAwl ~ Steppe 18&1 mo.-I18WI
D '0
o
f11M
Wafm wilh dry win'., lewl
Imon_ • upllllld .....Mll IC.I IMtdil..,_enl Humid I....".,.'. ICII
W..m whh dry
ILl""""
Cold ",ilh main w'nter lOt)
lmmCoId with lily win.er 10wi
I·.,·-t-J~~
~ ImonJOOf\ IYpe)
• . f'ollf
clime..,
Figure A4.7.
_1
Koppen climate classification
J
.J
J
J
' ]
APPENDIX4
Table A4.9. Symbol
Climatic classification by Thornthwaite (UNESCO, 1990) Moisture province
Thermal province
Annual pe: em
Index: I
Perhumid
Megathermal
>114
>100
A
B4 Humid
I I
80-100 99·8-114
60-80
Bz
85·6-99·8
40-60
B,
71·3-85·5
20-40
B3
C2
Meso thermal
Moist sub-humid
Microthennal
0-20
42-8-57·0 I
Dry sub-humid
28·6-42·7
(-33)-0
D
Semi-arid
Tundra
14·3-28·5
(-66)-(-33)
E
Add
Forest
0-14·2
(-100)-(-66)
a
60
60
120
120-"
180
5 60~----~----~~~-------4------~~~------r---'------+--------~
7
o
60
1. Equatorial 2. Subequatorial 3. Tropical 4. Subtropical
Figure A4.B.
60
120
180
120
5. Temporate
6. Subpolai 7. Polar
Climatic zones a/the earth (after B. P. Alisov)
231
EVACUATION OF SEDIMENTS
cold winter. Consideration was also given to the time of the rainy periods and to the length of the dry peliod (UNESCO, 1990).
A4.3.5. Budyko Budyko (1956) devised a climatic classification based on a 'radiational index of dryness' as a means of rating the efficiency of the moisture supply. This is illustrated in Figure A4.9. The index is given by: 1= RnlLn
where Rn is the annual average net radiation in cal/cm2 Ln is the energy (in calories) required to vaporise 1,ocal precipitation This index expresses the relative values of the heat and water balances (see Table A4.10). For each index RnlLn there is a corresponding value of the run-off coefficient.
A4.3.6. Vegetation classification Vegetation depends on a variety of factors including te~perature, rainfall, soils and topography. Different regions of the world have di~tinct vegetation types called biomes. These may be used as a surrogate classification of cliInate and lnay offer an alternative basis for classification of i global erosion rates. Combinations of temperature and rainfall have been used ~o identify nine biomes, as shown in Figure A4.10. ' Tundra This area occurs around the North Pole, mainly north of the Arctic Circle. The ground is pennafrost as there are long cold winters and ~hort 'warm' summers. There are also 'tundra' like regions, known as Alpine regions, found on the peaks of the tallest mountains at all latitudes. Trees and tall perennial plants are usually absent and the ground is covered by mosses, lichens, grasses and perennial herbs. Northern coniferous forest (taiga or boreal forest) This zone is found in North America and Eurasia. It is ch~ractelised by very cold winters, more precipitation than the tundra and longer, w~nner winters. The soil thaws and the vegetation grows abundantly. The principal plant life is drought resistant needle-leaf conifers and some deciduous trees li~e paper birch. Deciduous forests Temperate areas with abundant rainfall. The summers ~re relatively long and warm and the winters are cold. Broad-leafed deciduqus trees dominate the canopy.
232
APPENDIX4
0
:!
0
~
~ 0 ~ 0
:! 0
~
~
i
=
0
Q
..
<0
0
0 'Of
~ 0 0
~
..-."
Cl
0
~
CI
~.
e
~
~
0
¢:..
~
0 ~
0
:
:.... ~
--
Co
c..:: c..::
~
~
-e
~
C)"
)...,>
~
~
.::;
~
'';::
.~ '\::
C
,..~
0\ ~ ~
~
::::
CJ:
t;:
233
EVACUATION OF SEDIMENTS
Table A4.10.
Climatic classification by Budyko (UNESCO, 1990)
Climatic index
Vegetation type
Run-off coefficient
I
< 1/3
Tundra
0·7
1/3-1
Forest
0·3-0·7
1-2
Steppe
0·1-0·3
2-3
Semi-deselt
<0·1
>3
Deselt
<0·1
30~----~--------~~~~-----------
25
20
15 Warm temperature
o
Cold temperature
-5
-10
-15L-----~~-----L------~------~----~------~------~------~----~
o
50
100
150
200
250
300
350
Total annual rainfall: em
Figure A4.10.
234
Biomes based on temperature and rainfall (after Budyko)
400
450
APPENDIX 4
Rain forests Abundant ai110unts of rainfalL Olympic Rainforest: located on the west coast of USA, on the OlYlnpic peninsula in \Vashington State. Warm climate. Tropical rainforest Abundant amounts of rainfall. Located near the Equator. Wann climate. Grasslands The typical rainfall is 25 cmJyr to 30 crrJyr. The lowest rainfall occurs in a desert biome. The grass has roots that can penetrate into the soil to obtain some water, but the amount of water available is not enough to sustain much tree life. Biomes are explained more fully at: o
http://sheepshead.usl. edufT......aCEPTIrainforest.ht!pl
A4.4. TABLES OF COUNTRY DATA
The following tables (Tables A4.11-A4: 16) provide, country by country, data on climatic conditions and rates of sediment yield" Examples from case studies are included where available.'-The tables.also.c·ontain data on the number and storage capacity of dams in each country together. with .estimates of the loss of storage due to sedimentation.
235
m
Table A4.11. Country
Afghanistan
~
Country data for Asia Climatic classification (Koppen)
BwklBsk
n Rates of sediment yield: tlkm 2/yr (Walling and Webb, 1983)
Example rates: tlkm 2/yr (from literature)
No. of dams data for
Total capacity of dams:
% Li:lP,lL Il Y lost to sediments
c
» --I o
Annual loss of storage
M.m}
100-250/250-500
28
0
Z
800·0%
o
."
0·0%
Vl
m
Armenia
DfiDs
CJ
100-250
3:
----
Azerbaijan
DflDs
100-250
Bahrain
Bwh
<50
m
Z --I Vl
-----
Bangladesh
Am/Cwa
1128
500-750/> 1000
-
-
Bhutan
Bs
>1000
Brunei
Af
<50
Burma
Am
500-750/750-10001 >1000
Cambodia
Aw
250-500
China
ET/BwkIBs/Cwbl Cf/CfalCw/Dwl
Dwa/Dwb
50-100/250-5001 500-750/7 50-1 0001 >1000
Df
100-250
Hong Kong
Cwa
250-500
India
Aw I AflAmIAs/Cwa
] 00-2501250-5001 500-750/>] 000
3130/310115
47
784]2
9·6%
0·5%
250-500/500-750
6250112 000111 200
3
1135
1·7%
0·4%
616
804012957/1 8001 1620/1400124611571 25 600/2] 7001 16300/2300
28
42804
28·}%
]·2%
--
-
TncioI1P"i::J
---,----BwkIBsk
50-100
Iraq
Bwh
50-100125()-500
l~rael
Bwh
50-100
DfC/D IlJ/Cfa
<50/50-tOOI 250-500
Tnm
-50
. - -------- -----
--
--
Japan
729
17322
6·9%
0·2%
7
11·1%
0-4%
.Jordan
Bwh
50-100
Kazakstnn
DI11/Bsk/Bwk
<50/50-100
Korea (North and South)
ClhlOwa
500-750
Kuwait
Rwh
250-500
Kyrgy:-:tan
Df
100-250
Laos
Aw/Cf
250-500
Lebanon
Bwh
50-100
Malaysia
Af
250-500
Maldives
Csb
Island
Mongolia
Bwk/Bsk
50-100
Cwa
500-750
Cwa
>1000
Om:m
[3wh
50-100
Pakistan
Bw
50-1 OOt:!50-500
.-
--
~~---~
1
--------
-
.
-
Myanmar -------~---
Nepal ----~---
-
-
-
~-~
..
2800 ~
.
~---
-~
»-u
iJ
2498/454/1179
1
14300
Pnpu:1 New Guinea .
AI'
250-500/500-750
2581/492/11 126 ..
-------~
.
20·3%
0·8%
m Z
o X ..t:>.
Table A4.11.
m
continued
~ ()
Climatic classification (Koppen)
Rates of sediment yield: tlkm2/yr (Walling and Webb, 1983)
Philippines
Am
500-750
Qatar
Bwh
50-100
DfclDtbfDfalDfdF-· DwclDwdlET
<50/50=100
Saudi Arabia
Bwh
50-100
Singapore
Af
250-500/500-750
Sri Lanka
Af
100-250
Syria
Bwh
50-100
Taiwan
Cwa
>1000
Country
Example rates: tlkm2/yr (from literature)
No. of dams data for
% capacity lost to sediments
Total capacity of dams: M.m3
Annual loss of storage
C
~ o z o -n Vl
Russia-··
..
m
-~
~----
----
·~241128/39/141B/9/
CJ 3:
. ...
m
6/5/5
Z -I
Vl
.-~
31 700
1525
10 ......
Tajikistan
Ds
250-500
Thailand
Aw
<501250-500
Turkey
DslDf
100-250/500-750/ >1000
Turkmenistan
Bwk
50-lO0/lO0-250
United Arab Emirates
Bwh
50-100
Uzbekistan
BwkfBsk
<50/50-100
Vietnam
Aw
250-500
Yemen
Bwh
50-100
........
.
_
0·7%
.......
7·6-3874 16
9006
.. -------
--------
--------
1083/203
59·7%
1·5%
Table A4.12.
Country data for Africa
Country
Climatic classification
(Koppen)
Rates of sediment yield: tllcm 2/yr (Walling and Webb, 1983)
Example rates: tlkm1/yr (from literature)
No. of darns data for
Total capacity of dams: M.m)
% capacity lost to sediments
Annual loss of storage
L7
2632·8
15·5%
0·5%
1·24%
O·O~%
-Algeria
BWh/Csa
<50
Angola
Aw/Cwa
]00-250
Benin
Aw
50-100
Botswana
BSh
100-250
Burkina Faso
Bsh/Aw
50-100
Burundi
AflAw
500-750
Aw/Am
100-250
lslmul
Is1ancl
Aw/Am
<50
BwhlBsh/Aw
<50
Comoros
ls1and
Is1and
Congo
Aw/Af
<50
Congo, Democratic RepUblic
Aw
<50
Djibouti
13sh
50-100
Egypt
BWh
<50
--
--
--------Carneroon Verde
Central Africa Republic
Chad
--
--
38
'1
168900·0
Table A4.12. Country
m
continued
~
Climatic classification (Koppen)
Rates of sediment yield: tlkro2/yr (Walling and Webb, 1983)
Aw/Am
50-100
ExampJe rates: tlkm2/yr (from literature)
,
No. of dams data for
()
TQtal capacity of dams: M.m3
% capacity lost to sediments
Annual loss of storage
,!
Equatorial Guinea
---.
Eritrea
CwlBsh
50-100
Ethiopia
Bshl Bwh
5
Am/Aw
100-250
Gambia
Bsh
50-100
Ghana
Aw
50-100
Guinea
Aw
50-100
Guinea-Bissau
BWh
50-100
Ivory Coast
Aw
50-100
Kenya
Bsh/C
50-100/250-500
Lesotho
Ctb
250-500
Liberia
Am
50-100
Libya
BWh
<50
Madagascar
Af/Aw
250-500/500-750
Malawi
Aw/Cwa
50-100
Mali
Bwh/Bsh/Aw
<50
"Tl Vl
m
3:
m
Z -I
-----
Vl
---
'-------------
19 520/20117 600/ 20-3000/2500-17 600
--------
-----
L
o z o o
-----
Gabon
C
?:j
1
150-0
I
7·80%
1-56%
Mauritania
DwhlBsh
Mauritius
Isl.and
<50 Island ~
Morocco
Bsh/Bw
Mozambique
Aw
-
<501750-1000
100-2501250-5001
10 351-0
17
7-45%
0-42%
43·46%
1·74%
80/17
500-750
Namibia
BwlBsh
50-100
Niger
BWh
<50
Aw
50-100
At'
<50
SnoTome and Principe
Island
Island
Senegal
Aw/l3sh
50-100
Seychelles
Islands
Nigeria
33
~-~.
Rwanda
-----
Island
--
--.--Sierrn Leone
Am
50-100
Somalia
Bsk
100~-250
South Africa
Csb/Cll)/I3W
<50/50-1001
17/10-1000
100-250/250-5001 5,00-750
----
-----
--
Slldan
BWh/Bsh/Aw
<50
Swaziland
eft)
250-500
3
5584·0
;
TLltlZania
Bs/Aw
250-500/500-750
Aw
50-100
-Togo
94
» iJ -u
rn Z
o x
,.tI..
m
~
()
C
~
o z o
."
c.n m
o
3: m Z
Table A4.12. Country
Tunisia
-I
continued
c.n
Climatic classification (Koppen)
Rates of sediment yield: tlkm2/yr (Walling and Webb, 1983)
BWh/Csa
<501750-} 000
Example rates: tlkm2/yr (from Ii terature)
No. of dams data for
Total capacity of dams: M.m3
% capacity
lost to sediments
Annual loss of storage
15
1776·0
34·21%
0-07%
---
Uganda
Aw
50-100/500-750
Zambia
Cwa
50-100
Zimbabwe
Aw
100-250
-~--
---
.1
'j
Table A4.13. Country
Country data/or-Australasia Climatic
classification (Koppen)
Rates of sediment yield: t/km2/yr (Walling and Webb, 1983)
Example rates: tlkm2/yr (from literature) 28
Australia
Aw/Bsb/Cfh/CsICsa/ Bs/Bwh
<50/50-100
Fiji
Af
Island
Kiribati
Af
Island
Marshall Islands
Af
Island
Micronesia
Af
Ishmd'
Nauru
AI'
Island
New Zealand
Ctb
250-500/500-7501 >1000
Palau
Af
ISland
Solomon Islands
Af
Island
Tonga
AI'
Island
Tuvnla
Af
Island
Vanuatu
At"
Island
6982119 9701 1'7 340/13 8901 17 070113 3001 12736
No. of dams data for
Total capacity of dams: M.m3
% capacity
Annual
lost to sediments
loss of storage
4
202
19·2%
1·0%
I
---Western Samoa
~ w
Af
I----- .
Island
» iJ iJ
ITI
Z
o x
~
Table A 4.14.
m
Country data for Europe
~
Climatic classification (Koppen)
Example rates: t/km2/yr (from literature)
Albania
Ctb
500-750
4150/3590
AndolTa
Ctb
250-500
o z o "m
250-500/500-750
3:
Country
No. of darns data for
~
Total capacity of dams: M.m3
()
Rates of sediment yield: tlkI1l2/yr (Walling and Webb, 1983)
% capacity lost to sediments
Annual loss of storage
C
~
Vl
Ctb
Austria
o
6
4
93·0% -
Dtb
<50
Belgium
Cfb
<50
Bosnia and Herzegovina
Cfb
500-750
Bulgaria
Dfa/Ctb
100-250/250-500
Croatia
Ctb
100-250
Cyprus
Csa
50-100
Czech Republic
Ctb
<50/100-250
Denmark
Cfb
<50
Estonia
Dtb
<50
Finland
DfclDtb
<50
France
Ctb
<50/50-100/ 100-250/250-500
GelTOany
Ctb
<50/100-250
Greece
CsalCfb
500-750
Belarus
I
27·9%
-l Vl
7
0
11·9%
3·9%
3
236
3·0%
0·2%
-,-~-~
"J
"\ !
'I
m
Z
~--
--
111
J
Hungary
Ctb
100-250
Icchmd
efelET
<50
Ireland
cn)
50-100
Italy
Cfa/Csa
50-1001100-250/ 250-500
Latvia
Dtb
<50
Liechenstcin
Cfb
100-250
Lithuania
Dfb
Luxembourg
Cfb
<50
Macedonia
cnl
500-750
Malta
Csa
250-500
Moldova
Of
<50
Monaco
Cfb
250-500
Netherlands
Ctb
<50
Norway
orc
50-100
Poland
Otb
<50
8
597
9·5%
0·4%
Portugal
Ctb
100-250
7
440
1·4%
0·1%
Romania
CfblDtb
< 5011 00-250
18
207
32·4%
3·2%
San Marino
Ctb
250-500
--_._--
45701214
-
83
"'0
.--
Serhia and Montenegro
------
Ctb
500-750
» "'0
I
rn
Z
o
>< ..r:...
m
~ () C
~
o z o
II (f)
Table A4.14.
m
continued
o
Climatic classification (Koppen)
Rates of sediment yield: tlkm2/yr (Walling and Webb, 1983)
Slovakia
Ctb
100-250
Slovenia
Cfb
500-750
Spain
Cfb
100-250
Sweden
DfclDfb
<50
Switzerland
Cfb
50-100
Ukraine
Dfb
<50
United Kingdom
Cfb
<50/50-100
Vatican City
Cfb
250-500
Country
Example rates: tlkm2/yr (from literature)
8·4-2703
No. of dams data for
Total capacity of dams: M.m3
% capacity lost to sediments
Annual loss of storage
3
53
28-4%
0-9%
91
23323
3-7%
0-1% --
2
3
71-4%
8-6%
-----
)
,
95
-)
0-1%
3:
m Z -I
(f)
Table A4. J5.
Country data for North America
Country
Climatic c1nssi1ication (Koppen)
Rates of sediment
Antiglla amI Barhuda
At"
1.00-250
Bahamas
Af
100-250
Barbados
At'
100-250
Belize
Af
lOO-250
Cl.1nadn
ETICfc/C1b/Dfcl Dlb
< 50/50-] 001 100-250/250-5001 .500-7501750-10001
yield: tlkm2/yr
(Walling amI Webb, 1983)
Example rates: t/km2/yr (from literature)
No. of dams data for
Total capacity of dams: M.m3
_..__.
:
-
91/55/4/0-350
>1000 Costa Rica
Aw
100-250
Cuba
Af
100-250
Af
100-250
Af
100-250
EI Salvador
Aw
100-250
Greenland
Hf
50-100
Af
100-250
Guatemala
Aw
100-2501250-500
Hniti
Af
100-250
Dominica
-
Dominican Republic
-------.---
Grenada
--
-
-
----
% capacity
lost to sediments
Annual loss of storage
m
~
()
C
Table A4.15.
~
continued Climatic classification (Koppen)
Rates of sediment yield: tlkm 2/yr (Walling and Webb, 1983)
Honduras
Aw
]00--250
Jamaica
Af
100-250
Mexico
Bw/Cw/Af/Aw
50-100/100-2501 250-5001 500-750
Nicaragua
Aw
100-250
Panama
Af
100-250
Saint Kitts and Nevis
Af
100-250
Saint Lucia
Af
100-250
Saint Vincent and Grenadines
Af
]00-250
Trinidad and Tobago
Af
100-250
USA
DfblDfaiB w/Cfal
< 50/50-1 001 100-2501250--5001 500-750/7 50--10001 >1000
Country
Example rates: tlkm2/yr (from literature)
No. of dams data for
Total capacity of dams: M.m 3
% capacity lost to sediments
Annual loss of storage
3:
m
Z
-I
(/)
211
.-.. ---..
1
(/)
o
--------_.
CsblBs/Df
o z o " m
407212374122921 1167/500114511 071 71150112
1105
--~-----
109980
3·9%
0·2%
Table A4.16.
Count}")1
data for South America
Climatic classification
Country
(Koppen)
Rates of sediment yield: tlkm 2/yr (Walljng and Webb, 1983)
Example rates: tlkm 2/yr (from literature)
130/33
Argentina
Bw/Cfa/Bs
< 501250-500/ 750-1000/> 1000
l10livia
AwlBsc
100-250/250-500/ > 1000 <50/100-250
Cfa/Cwa/ Awl AI'· Brazil - - - - - - - - -1 Ct11lCsIBsIB w Chile
No. of dams data for
Total capacity of dams:
% capacity lost to sediments
Annual lo~s of storage
M.m3
]46/9
2
3829
0-8%
--0-1%
19·2%
--9·6%
100-2501>1000
----~--~----.
Colombia ------~---
50-100/> 1000
Af/Aw ..--.-
917
-
-,
Ecuador
Aw/Af
100-250/> 1000
French Guiana
A1'
<50
Guyana
AI'
<50
Paraguay ---
Cw..uAw
<50
Peru
Af/AwlBsc
100-250/250-5001 >1000
Af
<50
2000
----~---
Suriname
---UllIgtluy Venezuela
~ ~
Cra
Aw/Af
<50 <50/100-250
212
1
3
» -0 -0
m
Z
o x
~
Index
Page numbers in italics refer to illustrations.
Africa country data 239-242 erosion rates 221-224 hydropower 27 irrigation 27 population 26-27 Alisov classification 229, 231 Americas country data 247-249 hydropower 24-26 irrigation 24-26 population 24-26 annual precipitations 106, 213-214,213 Aracay River, Venezuela 202 areas erosion rate 95-100 suited to flushing 93-124 aridity maps 229,232 Asia country data 236-238 hydropower 27-28 irrigation 27-28 population 27 Atbara River, Sudan 195 Athi-Kapiti plains 223 Australasia 243 Austria 68,72-74, 76 case study 176-178 downstream impacts 178 flushing 177-178 sedimentation 177 autumn 1998 precipitation 108,111
Baira, India 68, 72-75, 81 case study 208-210 downstream impacts 209-210 flushing 209-210 sedimentation 209
bank failure 193 basin areas, erosion rate 121 basin outlet distances 122 basin shapes 10, 84,88-89 bed width, incised channel 44, 44 bibliography 143-148 biomes 232,234-235,234 boreal climates 115-116 forests 232 broad reservoirs 84 Budyko classification 232-235, 233-234
Cacm, Costa Rica 68, 72-75, 79 case study 197-200 downstream impacts 200 flushing 198-200 sedimentation 198 Canada, erosion rates 228 capacity criterion, geometry 49,49 case studies B aira, India 208-210 Cacm, Costa Rica 197-200 erosion rates 221-228 flushing 71-81, 171-210 Gebidem, Switzerland 200-202 Grofind, Austria 176-178 Guanting, China 179-182 Guernsey, USA 172-173 Heisonglin, China 184-186 Hengshan, China 196-197 Ichari, India 207-208 Jensanpei, Taiwan 174-175 Khashm El Girba, Sudan 195-196 Mangahao, New Zealand 171-172 N anqin, China 205-207 Naodehai, China 175-176 numerical models 163-170 Ouchi-Kurgan, former USSR 190-191 Palagnedra, Switzerland 178-179 Sanmenxia, China 186
251
EVACUATION OF SEDIMENTS
Santo Domingo, Venezuela 202-205 Sefid-Rud, Iran 191-195 Shuicaozi, China 182-184 Tarbela Dam, Pakistan 163-170 Warsak. Pakistan 189-190 Zemo-Afchar, former USSR 173-174 catchments characteristics 123,217-221 delivery ratio 122 hydrology 81-82, 123 orientation 217-218 sedimentology 9, 82 size 122 slope 217 Central AmeIica hydropower 25-26 imgation 25-26 population 25 channel erosion 193 China 68, 72-80,88 case study 175-176,179-186,196-197,
205-207 density current venting 206 downstream impacts 176,182-186, 189,
coniferous forests 232 constraints, flushing 60 constrictions, flushing flow 51-52 constIuction reservoirs 6-7, 28-30 continental variations erosion rate 94-96 sediment yield 94-96 cool Mediterranean climates 115 Costa Rica 68,72-75, 79 case study 197-200 downstream impacts 200 flushing 198-200 Reventazon River 197 sedimentation. 198 country classification, climate 118-120 country data tables Africa 239-242 Asia 236-238 Australasia 243 Europe 244-246 North America 247-248 South America 249 cross-sections, flushing channel 48, 48
197,207 erosion rates 224-225 flushing 175-176, 181-189, 196-197,
206 lateral erosion 185 Liuhe River 175 sedimentation 175, 180-186, 196, 205-206 storage capacity 188 Yellow River 186, 225 Yeyu River 184 Yili River 182 Yongding River 179-180 climate 211-215 classification 229-235 combining homogenous groups 117 erosion rate 93, 115-120 Tarbela Dam, Pakistan 164-165 climatic zones 12 country classification 118-120 earth 231 erosion rate 115-120 river basins 118-120, 118 sediment yield 118, 118 world 101-120 cold steppe climates 115 concentration inflow 56
252
DDR see-drawdown ratio deciduous forests 232 delivery ratios catchment size 122 depositional features 122 demand comparisons 30 demand distribution 4-7 storage 23-28 density cun-ent venting 66-69 Nanqin, China 206 deposition features 122 potential 9, 83-84 sediment size 53-57 Tarbela Dam, Pakistan 165-166 design considerations 127-130 detention flushing 185 discharge flushing 45 distIibution construction reservoirs 28-30 sediment rate 33 storage demand 4-7 storage loss 7,31-34 distIibution of demand 23-28 diversion channels 194 downstream impacts 10, 86-88
INDEX
Baira, India 209-210 Cachi, Costa Rica 200 Gebidem, Switzerland 201-202 Gmund, Austria 178 Guanting, China 182 Guernsey, USA 173 Heisonglin, China 185-186 Hengshan, China 197 Ichari, India 208 Jensanpei, Taiwan 175 Khashm El Girba, Sudan 196 Mangahao, New Zealand 172 Nanqin, China 207 Naodehai, China 176 Ouchi-Kurgan. fonner USSR 191 Palagnedra, Switzerland 179 Sanmenxia, China 189 Santo Domingo, Venezuela 205 Sefid-Rud, Iran 195 Shuicaozi, China 184 Warsak, Pa..\istan 190 Zemo-Afchar, former USSR 174 drainage basin areas 218-219,218 erosion rate 121 sediment yield variations 95 drainage density 219 drawdown 10, 85-86, 88 eachf, Costa Rica 198 flushing 40-41,41 Heisonglin, China 185 incomplete 51 N anqin, China 206 drawdown ratio (DDR) 51 dryness, radiational index 232-235 Durlassboden reservoir 176-178
earth. climatic zones 231 econoIilic assessment, flushing 60 economic factors analysis 14, 129 efficient flushing 43, 58-61 hydraulic conditions 7-8, 58-59 empty flushing 66 enhancements, flushing 10, 70, 86 environments, Tarbela, Pakistan 164-169 equiiibrium conditions 42 erosion affecting factors 211-221 controls 220-221 definition 211 human impact 220-221
land use 219-220 processes 122 vegetation 219 erosion rates 11 Africa 221-224 Canada 228 case study 221-228 China 224-225 continental variation 94-96 drainage basin area effects 121 geology effect 121 geotechnics 93 Himalayas 226 human impact 93, 121
India 225-226 Kenyan grazing lands 223-224 Krishna River, India 226 land use effect 93, 121 Pakistan 226 precipitation effect 121 Puerto Rico 226-227 slope effect 121-122 soil effect 121-122 Spain 228 Switzerland 227 Taiwan 96 Thailand 224 topography 93 Turkey 227-228 vegetation effect 93, 121-122 worldwide 93 Europe _ country data 244-246 hydropower 24 irrigation 24 population 23-24 evaluation flushing criteria 50
sediment br..lance ratio
45-46
financial analysis 129 flood control 3-4 flushing Baira, India 209-210 Cacm, Costa Rica 198-200 case study 71-81,171-210 constraints 60 criteria 42-58, 87-88 discharge 45 economic assessment 60
253
EVACUATION OF SEDIMENTS
efficiency 43, 51-52 enhancements 10, 70, 86 equilibrium conditions 42 flow constrictions 51-52 Gebidenl, Switzerland 200-201 Gmiind, Austria 177-178 Guanting, China 181-182 Guernsey, USA 172-173 Heisonglin, China· 184-185 Hengshan, China 196-197 !chari, India 207-208 influences 39-62 Jensanpei, Taiwan 175 Khashm E1 Girba, Sudan 195-196 long-term equilibrium conditions 42 Mangahao, New Zealand 171-172 mechanisms 40-42 Nanqin, China 206 N aodehai, China 175-176 numerical models 61-62 operation duration 45 optimum locations 12-13 Ouchi-Kurgan, former USSR 191 outlets 127 Palagnedra, Switzerland 179 periods 57-58 Sanmenxia, China 187-189 Santo Domingo, Venezuela 202-205 Sefid-Rud, Iran 192-195 Shuicaozi, China 182-183 site-specific factors 60 suitable geographical areas 93-124 Tarbela Dam, Pakistan 166-167 techniques 66-70 value 10 \yarsak, Pakistan 190 water available 8, 59 worldwide experience 65-89 Zemo-Afchar, former USSR 173-174 flushing channels cross-sections 48, 48 narrow reservoirs 53 wide reservoirs 53 flushing width ratio (FWR) 53 forests 232, 235 fOlmer USSR 68, 72-74, 76, 78 case study 173-174, 190-191 downstream impacts 174, 191 flushing 173-174,191 sedimentation 173,190 free flow
254
Cachf, Costa Rica 198 Santo Domingo, Venezuela full drawdown, flushing 41 FWR see flushing width ratio
204
Gebidem, Switzerland 68, 72~75, 79 case study 200-202 downstream iInpacts 201-202 flushing 200-201 sedimentation 200 geographical areas, flushing 93-124 geographical distribution 19 geology 216 erosion rate effect 121 geometry, capacity criterion 49,49 geotechnics, erosion effect 93 global sediment yields 93-100 variation maps 100, 102-105 global water resources 18-19 Gmiind, Austria 68, 72-74, 76 case study 176-178 downstream impacts 178 flushing 177-178 sedimentation 177 gorge-like reservoirs 84 graded sediments, effect 54-55 grasslands 235 growth, world population 20, 23 Guanting,China 68,72-74,77,88 case study 179-182 downstream impacts 182 flushing 181-182 sedimentation 180-181 Guernsey, USA 68, 71-74, 88 case study 172-173 downstream impacts 173 flushing 172-173 sedimentation 172
Heisonglin, China 68, 72-74, 77-78 case study 184-186 downstream impacts 185-186 flushing 184-185 lateral erosion 185 sedimentation 184 Hengshan, China 68, 72-73, 75, 79 case study 196-197 downstreanl impacts 197 flushing 196-197
INDEX
sedimentation 196 high annual precipitation 10 1 high erosion rate areas 95-99 Himalayas, erosion rates 226 historic growth . hydropower 21-22,23 reservoir 30 homogenous climatic groups I 17 human impacts erosion control 220-221 erosion rate 93, 121 humid climates 115 hydraulic conditions, flushing 7-8, 58-59 hydraulic modelling 13-14, 128-129 hydrology catchment 81-82, 88, 123 characteristics 12, 123 investigations 13, 127-128 sedimentology 9 Tarbela Dam, Pakistan 164-165 hydropower 3 Africa 27 Americas 24-26 Asia and Oceania 27:-:-28 Europe 24 historic growth 21-22, 23 potential 22-23
Iehari, India
68, 72-75, 80, 88
case study 207-208 downstream impacts 208 flushing 207-208 sedimentation 207 ICOLD World Register of Dams 17-19 world storage volume data 152-155 incised channels 44, 44 incomplete drawdown 51 India 68,72-75,80-81,88 case study 207-210 downstream impacts 208-210 erosion rates 225-226 flushing 207-210 Krishna River 226 Ravi River 208 sedimentation 207, 209 Indus basin Pakistan 164 Indus River flows 165 infiltration rate, soils 217 inflows 127-128
concentration 56 hydro graph 165 Tarbela Dam, Pakistan 165 insufficient drawdown 41 insufficient flushing flows 51-52 investigations sediment 128 site-specific 13 Iran 68, 72-73, 75, 78-79 bank failure 193 case study 191-195 channel erosion 193 diversion channels 194 downstream impacts 195 flushing 192-195 lateral erosion 194 long-term predictions 195 sedimentation 191-192 sheet erosion 193 irrigation 3, 21 Africa 27 Americas 24-26 Asia and Oceania 27 Europe 24
Jensanpei, Taiwan. 68, 72-76 case study 174-175 downstream impacts 175 fl ushlng 175 sedimentation 174
Kabul River, Pakistan
189
Kenyan grazing lands 223-224 . Khashm E1 Girba, Sudan 68, 72-75, 79 case study 195-196 downstream impacts 196 flushing 195-196 sedimentation 195 Koppen classification 12, 229-230, 230 climate 108, 112-115,116 world climate 116-118,116 Krishna River, India 226
land use erosion control 219-220 erosion rate effect 93, 121 lateral erosion 185, 194 Liuhe River, China 175
255
EVACUATION OF SEDIMENTS
Loiza, Puerto Rico 227 long-term capacity ratio (LTCR) 47-50 long-term equilibrium conditions 42 long-telm storage capacity 188 longitudinal energy gradient 44 loss rate trends 34 soil 212 storage 7, 31-32, 35-36, 35 low annual precipitation 101 low erosion rate areas 99-100 low-level outlets 10, 85 lower Yellow River, China 225 LTCR see long-term capacity ratio
major rivers, sediment yield 97-98 Mangahao, New Zealand 68,71-74 case study 171-172 downstream impacts 172 flushing 171-172 sedimentation 171 maps alidi ty 229, 232 global sediment yield 100,102-105 Massa River, Switzerland 200-201 mean annual precipitations 213 mean annual run-off 213,215 mean suspended-sediment yields 214-215, 223 mechanisms, flushing 40-42 Mediten-anean climates 114-115 Melezza River, Switzerland 178 mid-latitude summer dry 114 wet 113 winter dry 113-114 mobility, sediments 8-9,59-60 modelling hydraulic 13-14 system simulation 14 Tarbela Dam 166-169,167 Mount Kenya 221-222 Mount Kilimanjaro 223 Mpwanga, Tanzania 220 Mwea-Masinga plains 221
Nanqin, China 68,72-75, 80 case study 205-207 density CUlTent venting 206
256
downstream impacts 207 flushing 206 sedimentation 205-206 Naodehai, China 68, 72-74, 76 case study 175-176 downstream impacts 176 flushing 175-176 sedimentation 175 nalTOW reservoirs 84 flushing channels 53 New Zealand 68, 71-74 casestudy 171-172 downstream impacts 172 flushing 171-172 Mangahao River 171 sedimentation 171 North America country data 247-248 hydropower 25 irrigation 25 population 24-25 Northern coniferous forests 232 numerical models case study 163-170 flushing 61-62 sediment 166 Tarbela Dam, Pakistan 166-169,167
Oceania hydropower 27-28 irrigation 27-28 population 27 operation duration, flushing 45 operation policies, Pakistan 166 operational considerations 85-86, 88 operational limitations 10 optimum locations, flushing 12-13 Orange River, South Africa 224 organic content, soil 217 orientation, catchments 217-218 Ouchi-Kurgan, fOlmer USSR 68,72-74, 78 case study 190-191 downstream impacts 191 flushing 191 sedimentation 190
50 percentile size river-bed matelial 55 transported sediment 55
INDEX
Pacific Asiatic-Australian sector, erosion 96 Pakistan 72-74, 78, 88 see also Tarbela Dam case study 163-170,189-190 climate 164-165 downstream impacts 190 environment 164-169 erosion rates 226 flushing 166-167, 190 history 163 hydrology 164-165 Indus basin 164 Indus River 163-165 Indus River flows 165 Kabul River 189 numerical sediment modelling 166, 167 operation policies 166 sediment 163-164 sediment deposition 165-166 sediment inflows 165 sediment throughput 167-168 sedimentation 189-190 storage prediction 168-169,168 Palagnedra, Switzerland 68, 72-74, 76-77 case study 178-179 downstream impacts 179 flushing 179 sedimentation 178-179 parameters, sediment ratio 55 permeability, soils 217 polar desert 114-115 wet and dry 114 population Africa 26-27 Americas 24-26 Asia and Oceania 27 Europe 23-24 world 19-21 precambrian basement schists 223 precipitation 211-215 erosion rate effect 121 seasonal variation 101-108 precipitation distribution autumn 1998 108, 111 spring 1998 108, 109 summer 1998 108, 110 winter 1998 107 pressurised flushing 66, 204 Puerto Rico 227 erosion rates 226-227
radiational index, dryness 232-235,233 rain forests '235 rainfall intensity 212 rapid drawdown 198 rates construction reservoirs 28-30 loss of storage 7. 31-32, 35-36, 35 Ravi River, India 208 references 133-140 regional sedimentation rates 31 register of dams 17-18 Reventazon River, Costa Rica 197 Rhone flow 201 ringlet sedimentation 34 river-bed material 55 rivers Aracay, Venezuela 202 Atbara, Sudan 195 bank 193 basin 118-120,118 climatic zones 118-120, 118 Indus, Pakistan 163-165 Kabul, Pakistan 189 Krishna, India 226 Liuhe, China 175 Mangahoa, New Zealand 171 Massa, Switzerland 200-201 Melezza, Switzerland 178 Orange, South Africa 224 Ravi, India 208 Reventazon, Costa Rica 197 Santo Domingo, Venezuela 202 sediment yield 97-98 Tons, India 207 Yellow, China 186, 225 Yeyu, China 184 Yili, China 182 Yongding, China' 179-180 rock types 216 routing techillques 66-70 run-off 214-215,215,223
Sanmenxia, China 68, 72-74, 78, 225 case study 186 downstre~m impacts 189 flushing 187-189 sedimentation 186 storage capacity 188 Santo Domingo, Venezuela 68, 72-75, 79-80
257
EVACUATION OF SEDII'1ENTS
case study 202-205 downstream impacts 205 flushing 202-205 sedimentation 203 SBR see sediment balance ratio SDR see sediment delivery ratio seasonal variations, precipitation 101-108 sediment balance ratio (SBR) 42-43 evaluation 45-46 parameters 55 sediment delivery ratio (SDR) 122 sediment loads 224 sediment size 53-57 sediment size ratio (SSR) 54,55 sedimentation, rate data 156-162 Sefid-Rud,Iran 68,72-73,75,78-79 bank failure 193 case study 191-195 channel erosion 193 diversion channels 194 downstream impacts 195 flushing 192-195 lateral erosion 194 long-term predictions 195 sedimentation 191-192 sheet erosion 193 shapes, basin 10 shear strength soils 217 sheet erosion 193 Shuicaozi, China 68,72-74, 77 case study 182-184 downstream impacts 184 flushing 182-183 sedimentation 182 site-specific factors 9, 60 site-specific investigations 13, 127-130 sizes, reservoir 35-36, 35 slopes catchments 217 erosion rate effect 121-122 slow draw down 198 sluicing 66-69 soils 216-217 erosion rate effect 121-122,220 infiltration rate/permeability 217 loss 212 organic content 217 shear strength 217 structure 217 texture 216 South Africa, Orange River 224
258
South America country data 249 hydropower 25-26 irrigation 25-26 population 25 Spain, erosion rates 228 spring 1998, precipitation 108, 109 SSR see sediment size ratio steppe climates 115 storage 3-4 capacity 9, 83, 188 construction vs demand 30 demand diSllibution 4-7, 23-28 distribution 31-34 gross requirements 35-36 increase distribution 30 loss rate 7,31-36 lost to sedimentation 33 requirements 7, 35-36 trends 34 volume disllibution 19 volume predictions 168-169,168 world demand 19-23 world total 17-18 worldwide distribution 18-19 structures, soil 217 Sudan 68, 72-75, 79 Atbara River 195 case study 195-196 downstream impacts 196 flushing 195-196 sedimentation 195 summer 1998, precipitation 108,110 suspended-sediment yields 214-215,223 world maximum 96 sustainable reservoir capacity 47-50 Switzerland 68, 72-77, 79 case study 178-179,200-202 downstream impacts 179, 201-202 erosion rates 227 flushing 179, 200-201 Masse River 200-201 Melezza River 178 sedimentation 178-179,200 system simulation modelling 14, 129
tables, country data 235-249 taiga forests 232 Taiwan 68, 72-76 case study 174-175
INDEX
downstream impacts 175 erosion rate 96 flushing 175 sedimentation 174 Tarbela Dam, Pakistan case study 163-170 climate 164-165 environment 164-169 flushing 166-167 history 163 hydrology 164-165, 165 Indus basin 164 Indus River flows 165 modelling 166-169, 167 numerical sediment modellimz 166 operation policies 166 ..., sediment 163-164 sediment deposition 165-166 sediment inflows 165 sediment modelling 166 sediment throughput 167 storage prediction 168-169 tectonic activity 216 temperatures 215 textures, soil 216 Thailand, erosion rates 224 Thomwaite classification 229,231 throughputs sediment, Pakistan 167-168 Tons River, India 207 top width ratio (TWR) 53 topography, erosion rate 93 transportation 11 transported sediment 55 transporting capacity 42-44 empirical equation 43 trapping efficiency 45 trends, storage loss rate ·34 tropical desert 112-113 forests 235 wet 108,112 wet and dry 112 tundra 232 Turkey, erosion rates 227-228 TWR see top width ratio Upper Tana basin, eastern Kenya upstream depositation 71 USA 68, 71-74, 88 case study 172-173 downstream impacts 173
flushing 172-173 sedimentation 172 USSR see former USSR
value, flushing 10 vegetation climate classification 232, 235 erosion control 219 erosion rate effect 93, 121-122 Venezuela 68,72-75, 79-80 Aracay River 202 case study 202-205 downstream impacts 205 flushing 202-205 sedimentation 203 volcanic activity 216
warm humid climates 115 vVarsak, Pakistan 72-74, 78, 88 case study 189-190 downstream impacts 190 flushing 190 sedimentation 189-190 water levels flushing 8, 59 incomplete drawdown 51 wide reservoirs 53 wind direction 215 wind speeds 215 winter 1998, precipitation 107 world climatic zones 101-120 ICOLD data 152-155 Koppen classification 116-118,116 maximum suspended ...gediment yield 96 population 19-2 t storage 4 storage volume 152-155 worldwide erosion rates 93 reservoir construction 28-30 sediment flushing experience 65-89 storage distribution 18-19
221-223 Yellow River, China 186, 225 Yeyu River, China 184 yields, sediment 93-105, 213 YiIi River, China 182
259
EVACUATION OF SEDIMENTS
Yongding River, China
179-180
Zemo-Afchar, former USSR case study 173-174
260
68, 72-74, 76
downstream impacts 174 flushing 173-174 sedimentation 173 Zemo-Afchar reservoir, former USSR
72-74, 76
68,