ELECTRONIC VERSION Ver. 1.04 December 2010 Cite as: Morris,GregoryL.andFan,Jiahua.1998.ReservoirSedimentation Handbook, McGraw-Hill Book Co., New York. Electronicversionoftheoriginal1998publicationbyMcGraw-HillBook Co. This fully-searchable version contains minor corrections. Gregory L. Morris and Jiahua Fan Gregory L. Morris, P.E., Ph.D. Gregory L. Morris Engineering P.O. Box 9024157 San Juan, Puerto Rico 00902 Tel. (787) 723-8005 gmorris@gmaeng.com gregmorris70@gmail.com Jiahua FAN China Inst. Of Water Resources & Hydropower Research 20 West Chegongzhuang Road Beijing, China fan-jiahua@sohu.com For Daniel, Richard And others of the next generation. CONTENTS (abbreviated) Use BOOKMARKS on left-hand side of the Adobe PDF file viewer for full table of contents and jump-to capability Chapter 1. Introduction Chapter 2. Reservoirs and Sustainable Development Chapter 3.Engineering Features of Dams and Reservoirs Chapter 4. Concepts of Reservoir Limnology Chapter 5. Sediment Properties Chapter 6. Erosion Chapter 7. Sediment Yield from Watersheds Chapter 8. Fluvial Morphology and Sediment Sampling Chapter 9. Hydraulics of Sediment Transport Chapter 10. Sediment Deposits in Reservoirs Chapter 11. Modeling of Sediment Transport and Deposition in Reservoirs Chapter 12. Reduction in Sediment Yield Chapter 13. Sediment Routing Chapter 14. Turbid Density Currents Chapter 15. Flushing Chapter 16. Sediment Evacuation Chapter 17. Decommissioning of Dams Chapter 18. Environmental and Regulamentory Issues Chapter 19. Case Study: Cach Hydropower Reservoir, Costa Rica Chapter 20. Case Study: Loza Reservoir Case Study, Puerto Rico Chapter 21. Case Study: Gebidem Dam and Reservoir, Switzerland Chapter 22. Case Study: North Fork Feather River, California Chapter 23. Case Study: Sefid-Rud Reservoir, Iran Chapter 24: Case Study, Sanmenxia Reservoir, China Chapter 24. Case Study, Heisonglin Reservoir, China PREFACE TO ELECTRONIC VERSION Ver. 1.04 Sediment issues at reservoirs are growing, as is the world communitys awareness to thisissue.Tofurtherthedisseminationofinformationonthistopic,theauthorsare makingtheReservoirSedimentationHandbookfreelyavailableinsearchableelectronic form. This document was prepared from scans of the original, which were converted to text and then re-formatted to match as closely as possible the original book. All figures, tables and headings fall on the same page as in the original, but paragraph breaks are not exactly the same. This version includes a correction to Figure 9.11, several known typographical andgrammaticalerrorsintheoriginalhavebeencorrected,andhopefullytheerrors introducedbythetextrecognitionsoftwarehavethemselvesallbeenfoundandfixed. The original index has not been reproduced, since the text is now fully searchable. TheauthorsgratefullyacknowledgetheassistanceofRichardMorrisandJulianand Christopher Libby in the tedious work of preparing the electronic version. Versions:1.04Reconstruction of Bookmarks in Adobe Acrobat as Table of Contents 1.03Includestheentirechapter2,whichwasaccidentlytruncatedinthepriorfile compilation process. Several typographical errors have also been corrected. 1.02Corrects a number of typographical errors. Gregory L. Morris San Juan Jiahua Fan (FAN Jiahua) Beijing PREFACE Whenthefirstauthorbeganresearchingproblemsofreservoirsedimentationmany yearsago,thelackofabook-typetreatmentofthesubjectbecameimmediatelyapparent. While there is a significant amount of literature, it is widely scattered, written in several languages, and much of it is "gray literature" published in the form of engineering reports and conference proceedings which is not readily accessible. This handbook is an outgrowth of the need for a comprehensive treatment of this complex subject. An in-depth treatment of each of the major technical topics is, of course, impossible within the contextofasinglehandbook,andthetreatmenthereinisbynomeanscomprehensiveandall-inclusive. The topics in each chapter are themselves of sufficient complexity to warrant book-length treatment. It is hoped that this conceptual overview of the major themes in each topic area, supported with a list of references and case studies, will facilitate the readers' comprehension of existing and potential problems at their own particular reservoir and watershed, and the types of remedialactionsthathaveprovedusefulatothersites.Theauthorsstronglyfeelthatthemost fundamental problem is to properly diagnose a sedimentation issue and select the conceptually appropriate course of action. Once the nature of the problem has been conceptually identified and one or more feasible courses of action identified, the approach to be used for subsequent detailed technical studies will usually be clear. The first author's research on reservoir sedimentation topic was sparked by Ariel Lugo, and a discussion of problems at Puerto Rican reservoirs several years later with A. Santiago Vazquez. Both authorswouldliketoexpresstheirspecialthankstoMaraMargaritaIrrizarrywho,duringher termasExecutiveDirectorofthePuertoRico Aqueduct and Sewer Authority, started sediment management activities at the Loa reservoir. Work at this site initiated the collaboration between the co-authors that eventually resulted in this handbook. Manyworkersfromaroundtheworld,toonumeroustonameindividually,havebeenvery helpful in providing time, insight, information, material, photographs, and permission to reproduce material.Weareparticularlyindebtedtothosewhoprovideddetailed informationoncasestudies: Esmaiel Tolouie (Sefid-Rud); Gian Rechsteiner (Gebidem); Alexis Rodrguez, Ake Sundborg, and Margareta Jansson (Cach); Xia Maiding (Heisonglin); Richard Webb and Allan Zack (Loza); and Donna Lindquist, Clay Clifton, Larry Harrison, Howard Chang, and Scott Tu (Feather River). Alain PetitjeanandJean-PierreBouchardatElectricitideFranceprovidedinformationonFrench reservoirs. Contributions by Shou-shan Fan at the U.S. Federal Energy Regulatory Commission are greatlyappreciated.RobertMacArthur,GeorgeAnnandale,RobertStrand,VicGalay,Andrea Handler-Ruiz, Tom Skelly, Rollin Hotchkiss, andKathleen Wilson were allparticularly helpful. Professionals at the following institutions also provided invaluable information and assistance: China Institute of Water Resources and Hydropower Research (Beijing), Northwest Hydrotechnical Institute (Xian), U.S. Bureau of Reclamation, U.S. Army Corps of Engineers, U.S. Geological Survey, U.S. Natural Resources Conservation Service, U.S. Forest Service, Puerto Rico Aqueduct and Sewer Authority,andLosAngelesCountyPublicWorksDepartment.Editorialsupportfrom McGraw-Hill was outstanding. Production of this handbook also represents a 2-year team effort by personnel at Gregory L. Morris&Assoc.Wecouldnothavemadeitthroughthelongordealofmanuscript preparation without the capable editorial assistance of Marco Flores, and graphic materials prepared by Elias Castro, Ada Sotto, Sebastian Garcia, Pully Torres, and Miguel Menar helped keep the officerunningthroughoutthelongwritingprocess.FormeremployeesGuangdouHuand G.T. Anderson discussed ideas and performed literature searches at university libraries. The first author greatly appreciates the support by his wife, Miriam, and sons Daniel and Richard, who graciously accepted the long hours and forfeited hiking and kayaking trips. Also gratefully acknowledged are the many years of dedication by his parents, andguidancebymany excellentteachers.Thesecondauthorgreatlyappreciatesmorethan45yearsofsupport andhelpinhissedimentresearchforwaterresourcesdevelopmentbyhiswife,Song Xiuzhen, and the support of his daughters Ying, Bing, and Xing. Wehopethatsocietywillbenefitfromourcollectedthoughtsaboutwaterandsediment management. Gregory L. Morris San Juan Jiahua Fan (FAN Jiahua) Beijing CHAPTERONE INTRODUCTION 1.1 Need For Sediment Management Most natural river reach are approximately balanced with respect to sediment inflow and outflow. Damconstruction dramaticallyaltersthis balance,creatinganimpoundedriver reach characterized by extremely low flow velocities and efficient sediment trapping. The impoundedreachwillaccumulatesedimentandlosestoragecapacityuntilabalanceis againachieved,whichwouldnormallyoccuraftertheimpoundmenthasbecomefilled up with sediment and can no longer provide water storage and other benefits. Declining storage reduces and eventually eliminates the capacity for flow regulation and with it all watersupplyandfloodcontrolbenefits,plusthosehydropower,navigation,recreation, andenvironmentalbenefitsthatdependonreleasesfromstorage.TheCamarreservoir in Venezuela (Fig. 1.1) offers an example of the consequences of sedimentation; less than 15 years were required for loss of all storage. This site differs from most other reservoirs in only one aspect, the speed at which storage capacity was lost. Sedimentmanagement was not practiced at this site. Storagelossisbutoneofmanysedimentationproblemsthatcanaffectreservoirs. Operation of storage reservoirs is severely impacted by the time half the volume has been sedimented,butseveresediment-relatedproblemscanappearwhenonlyasmall percentageofthestoragecapacityhasbeenlost.Asreservoirsageandsediments continuetoaccumulate,sediment-relatedproblemswillincreaseinseverityandmore siteswillbeaffected.Atanydamorreservoirwheresustainablelong-termuseistobe achieved, it will be necessary to manage sediments as well as water. This is not a trivial challenge. Many type of sediment-related problems can occur both upstream and downstream of dams,andsedimententrainmentcanalsointerferewiththebeneficialuseofdiverted water.Sedimentcanenterandobstructintakesandgreatlyaccelerateabrasionof hydraulic machinery, thereby decreasing its efficiency and increasing maintenance costs. Turbiddensitycurrentscancarrysedimentstensofkilometersalongthebottomofthe impoundment,eventuallyenteringdeepintakesandaccumulatinginfrontoflowlevel outlets.Localizedsedimentdepositsinthedeltaregionandstreambedaggradations upstream of the reservoir can produce flooding, cause soil water logging and salinization, impairnavigation,alterecologicalconditions,inundatepowerhousesdischarginginto deltaareas,andburyintakes(Fig.1.2).Inaridregionsthegrowthofphreatophytic vegetationondeltadepositscansignificantlyacceleratewaterloss.Thecombinationof sedimenttrappingandflowregulationalsohasdramaticimpactsontheecology,water transparency,sedimentbalance,nutrientbudgets,andrivermorphologydownstreamof the reservoir; dam construction is the largest single factor influencing sediment delivery to the downstream reach. The cutoff of sediment transport by the dam can cause stream- INTRODUCTION1.2 (a) (b) FIGURE 1.1 Fully sedimented Camar irrigation reservoir in Venezuela: (a) overview, (b) looking upstream from the spillway showing incipient formation of floodplain deposits above the spillwayelevationoneithersideofthechannel.Thewatersurfaceinthephotoisatthespillwaycrestelevation (G. Morris). INTRODUCTION1.3 FIGURE1.2SedimentedwatersupplyintakeinthedeltaregionupstreamofKhasmElGirbaDam on the Atbara River, Sudan (R. Hotchkiss). bed degradation, accelerate rates of bank failure, and increase scour at structures such as bridges.Thestreambedwillcoarsenandbecomearmored,degradingoreliminating spawningbeds.Evencoastalprocessescanbeaffected;acceleratedcoastalerosion affectingtheMississippiandNiledeltasisattributedtosedimenttrappingbehinddams more than 1000 km upstream. BasedontheinventorypublishedbytheInternationalCommissiononLargeDams (ICOLD,1988)andthecurrentrateofdamconstruction,asof1996therewereabout 42,000large(over15mtall)damsworldwide.Thereareseveraltimesasmanylesser structures.Anoverwhelmingmajorityofthesestructuresaredesignedandoperatedto continuouslytrapsediment,withoutspecificprovisionsforsustainedlong-termuse. Neither current nor projected levels of population and economic activity can be sustained if today's inventory of storage reservoirs is lost to sedimentation, and, as population and economicactivitygrow,relianceontheservicesprovidedbydamsisincreasing. Reservoir-dependentsocietiesrangefromtechnicallyadvancedurbanandagricultural systems in the western United States to village irrigators on the Indian peninsula. Sudden loss of the world's reservoir capacity would be a catastrophe of unprecedented magnitude, yet their gradual loss due to sedimentation receives little attention or corrective action. Reservoirs have traditionally been planned, designed. and operated on the assumption thattheyhaveafinite"life,"frequentlyasshortas100years,whichwilleventuallybe terminatedbysedimentaccumulation.Littlethoughthasbeengiventoreservoir replacementwhentoday'simpoundmentsarelosttosedimentation,ortoproceduresto maintainreservoirservicesdespitecontinuedsedimentinflow.Therehasbeenthetacit assumptionthatsomebodyelse,membersofafuturegeneration,willfindasolution when today's reservoirs become seriously affected by sediment. However, sedimentation problemsaregrowingastoday'sinventoryofreservoirsages,andseveresediment INTRODUCTION1.4 problemsarestartingtobeexperiencedatsitesworldwide,includingmajorprojectsof national importance. Sediment management in reservoirs is no longer a problem to be put off until the future; it has become a contemporary problem. Traditionalapproachestosedimentmanagementhavenotconsideredtheneedfor sustaineduse.Largeinitialstoragevolumesanderosioncontrolhavetraditionallybeen recommended to reduce sediment inflow and delay the eventual "death" of reservoirs, but erosion control alone cannot achieve the sediment balance required to stabilize reservoir storagecapacityandachievesustainableuse.Furthermore,manyerosioncontrol programsarepoorlyconceivedandimplemented,andfailtoachievethedesiredreductions in sediment yield. As a result, reservoirs worldwide are losing storage capacity rapidly,possiblyasfastas 1 percent peryear (Mahmood, 1987).Reservoirconstruction requiressiteshavinguniquehydrologic,geologic,topographic,andgeographic characteristics, and existing reservoirs generally occupy the best available sites. Because of the high cost and multiple problems associated with sediment removal and disposal on amassivescale,thesedimentationoflargereservoirsistoalargeextentanirreversible process.Iffuturegenerationsaretobenefitfromessentia1servicesprovidedby reservoirs it will be largely through the preservation and continued utilization of existing reservoirsites,notthecontinuedexploitationofashrinkinginventoryofpotentialnew sites.Thewatersuppliesandotherbenefitsderivedfromreservoirsdonotconstitute renewable resources unless sedimentation is controlled. 1.2 ELEMENTS OF SEDIMENT MANAGEMENT Conversionofsedimentingreservoirsintosustainableresourceswhichgeneratelong-term benefits requires fundamental changes in the way they are designed and operated. It requiresthattheconceptofareservoirlifelimitedbysedimentationbereplacedbya conceptofmanagingbothwaterandsedimenttosustainreservoirfunction.Sustainable use is achieved by applying the following basic sediment control strategies: 1.Reducesedimentinflow.Sedimentdeliverytothereservoircanbereducedby techniques such as erosion control and upstream sediment trapping. 2. Route sediments. Some or al1 of the inflowing sediment load may be hydraulically routedbeyondthestoragepoolbytechniquessuchasdrawdownduringsediment ladenfloods,off-streamreservoirs,sedimentbypass,andventingofturbiddensity currents. 3.Sediment removal. Deposited sediments may be periodically removed by hydraulic flushing, hydraulic dredging, or dry excavation. 4.Provide large storage volume. Reservoir benefits may be considered sustainable if astoragevolumeisprovidedthatexceedsthevolumeofthesedimentsupplyinthe tributarywatershed.Therequiredsedimentstoragevolumemaybeincludedwithin the reservoir pool or in one or more upstream impoundments. 5.Sedimentplacement.Focussedimentdepositioninareaswhereitssubsequent removalisfacilitated,orwhereitminimizesinterferencewithreservoiroperation. Configureintakesandotherfacilitiestominimizeinterferencefromtransportedor deposited sediments. The cost and applicability of each strategy will vary from one site to another and also as a function of sediment accumulation. However, even the largest reservoirs will eventually INTRODUCTION1.5 bereducedtosmallreservoirsbysedimentationand,soonerorlater,willrequire sediment management. 1.3 HANDBOOK APPROACH Thishandbookseekstogenerateanawarenessofsedimentationproblems,outlining practical strategies for their identification, analysis and management. Basic concepts and toolsarepresentedwhich,whenappliedinanintegratedmanner,canachievewhatwe willtermsustainablesedimentmanagementinreservoirs.Sedimentationisthesingle processthatallreservoirsworldwideshareincommon,todifferingdegrees,andthe management strategies and techniques presented herein are applicable to reservoirs of all ages, types, and sizes. An understanding of these principles will also aid in the effective designandmanagementofsediment-trappingstructuressuchasdebrisbasinsand detentionponds.Althoughcomplexmathematicalandphysicalmodelingstudiesmay oftenberequiredtofinalizethedesignandoperationalproceduresforsediment management,this handbook makesnoattempttoprovideanin-depthtreatmentofthese analytical methods or sediment transport theory. Extensive treatments of these topics are already available by others (Yang, 1996; Simons and Senturk, 1992; Julien, 1995; Chang, 1988; Vanoni, 1975; Graf, 1971). Rather, this handbook presents a broader view of basic principlesandmethodsessentialfortheconceptualizationandassessmentof sedimentationissuesassociatedwithreservoirs,andforidentifyingmanagement strategies to be explored in subsequent detailed and site-specific studies. Sedimentationproblemsandmanagementtechniquesvarywidelyfromonesiteto another,andbystudyingspecificsitesonecanappreciatethecomplexityofsediment problemsandthemannerinwhichtheycanbeaddressed.Sevenchaptershavebeen devoted to detailed case studies. These cases have been selected to demonstrate a variety ofmanagementtechniquesandcoverawiderangeofgeographicandhydrologic conditions. Case study reservoirs range in size from 2 to 1700 Mm3 (2 to 1700 l06 m3). TwositesarelocatedintheUnitedStates,twoinChina,andoneeachinCostaRica, Switzerland,and Iran. Chinahas82,000reservoirswhicharelosingstoragecapacityatanaverageannual rateof2.3percent,thehighestrateoflossofanycountryintheworld(Zhou,1993). China'sYellowRiverhasthehighestsedimentconcentrationsofal1theworld'smajor river systems, and China also has half of the world's large dams. Not surprisingly, China hasconsiderableexperienceinthemanagementofreservoirsedimentation.This handbookhasmadeaparticularefforttodistillusefullessonsfromtheChinese experience. Finally,thishandbookintroducestheoverallconceptofsustainablesediment managementwiththegoalofconvertingtoday'ssedimentingreservoirsintoresources thatwillbenefitfuturegenerationsaswellasourown.Whereasthetwentiethcentury focused on the construction of new dams, the twenty-first century will necessarily focus on combating sedimentation to extend the life of existing infrastructure. This task will be greatly facilitated if we start today. CHAPTER2 RESERVOIRS AND SUSTAINABLE DEVELOPMENT 2.1WATER SUPPLY AND WATER SCARCITY 2.1.1 Global Water Resources Naturallakesandriversworldwidecontainabout93,000km3offreshwater,about 0.27percentoftheearth'stotalfreshwaterresource(Table2.1).Whiletheabsolute volumeofgroundwaterstorageislarge,itismuchlessaccessiblethansurfacewaters becausemostislocatedtoodeeporinformationsofinsufficientpermeabilityfor economic exploitation, the rate of replenishment is much slower than surface water, and extractionfromcoastalareasislimitedbysalineintrusion.Thus,groundwaterismuch lessimportantthansurfacesupplies.Forexample,totalgroundwateravailabilityin21 Arabcountriesisestimatedatonly14.8percentoftotalstreamflow(Shahin,1989). Groundwater recharge in India has been variously estimated at 15 and 25 percent of total runoff, and in the United States and Russia respectively groundwater use accounts for 20 and9.3percentoftotalwithdrawals(Gleick,1993).Inthelongrun,withdrawalsfrom eithersurfaceorgroundwatersourcescannotexceedrunoff,definedasrainfallminus evapotranspiration, except to the extent that water is reused. However, reuse potential is not large because the most voluminous water use, irrigation, is consumptive. Waterstorageinnaturallakes(91,000km3)plusreservoirs(7000km3)isslightly less than the annual global precipitation of 119,000 km3over land areas, and about twice theannualrunoffintotheoceanof47,000m3,afigurewhichincludessurfacewater, groundwater, and glacial discharges. The remaining 41 percent of global precipitation is returnedtotheatmosphereasevapotranspiration.Runoffisunevenlydistributedboth geographically and in time. While the Amazon basin accounts for 15 percent of global TABLE 2.1 World Water Resources Relative amount of Volume, 103 km3Total water, %Fresh water, % Saline water:1,350,939.497.47 Oceans1,338,000.096.54 Lakes85.40.01 Groundwater12,854.00.93 Fresh water:35,028.62.53100.00 Antarctic ice21,600.01.5661.66 Other ice and permafrost2,764.00.207.89 Groundwater10,546.00.7630.11 Lakes91.00.010.26 Swamps11.50.000.03 Atmosphere12.90.000.04 Rivers2.10.000.01 Source: After Shiklomanov, 1993. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.2 FIGURE 2.1Water withdrawals in the United States over time, as compared to population (data from USGS Water Summary and Census of Population). runoff,Australia,withaslightlylargerlandarea,accountsforonly0.7percent (Shiklomanov,1993).SeasonalvariationisparticularlyevidentinareassuchasIndia, projectedtobetheworld'smostpopulatedcountrybyyear2050,wherethemonsoon season lasts 3 months or less and the remainder of the year is essentially dry. 2.1.2 Water Scarcity Theamountoffreshwatercirculatingthroughthehydrologiccycleisfixed,butthe demandforwaterusehasgrownrapidlyduringthetwentiethcenturyandfreshwateris becoming an increasingly scarce resource. Modern society is, above all else, a hydraulic society.Populationandeconomicactivityareincreasinglytiedtothediversionoffresh watertohumanuse,particularlyasagriculturalexpansionbecomesincreasingly dependentonirrigation.WaterwithdrawalsintheUnitedStatesgrewmorerapidlythan populationthroughthetwentiethcentury,butwithdrawalshaveslowedsincethe1970s, reflectingtheincreasingefficiencyofwaterusebecauseofscarcity(Fig.2.1).Recog- nitionthatthereisinadequatewatertopollutefreelyhasledtoincreasinglevelsof RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.3 wastewatertreatment,greaterefficiencyinuse,andrecyclingintheindustrialsector. Scarcity has also led to increasing irrigation efficiency in the agricultural sector, and new plumbing codes are gradually taking the United States from 20-L (5-gal) to 6-L (1.6-gal) flushtoilets.Municipalwaterrationinghasbeenimposednotonlyindryareassuchas southernCaliforniabutalsoinhumidareasrangingfromFloridatoNewYork. Increasingefficiencyofuse,combinedwithlowratesofpopulationgrowth,arealso generally characteristic of highly industrialized nations in Europe. ThepatternsofincreasingwaterscarcityevidentintheUnitedStatesarealso occurring on a global scale, but in many regions the scarcity is much more acute. Most of the world's population growth is occurring in less developed countries, especially in areas of low rainfall with high and increasing dependence on irrigation, which greatly increases per capita water requirements since food and fiber must be produced from runoff diverted to agricultural use rather than natural precipitation on rain-fed cropland. Thus, in Sweden itispossibletomaintainoneoftheworld'shigheststandardsoflivingwithwater withdrawalsofonly620m3/capita/yr,whereasinirrigation-dependentArabcountries withdrawalsof1205m3/capita/yrareneededtoproduceanindividual'sannual requirementof375kgoffruitsandvegetables,35kgofmeatsandpoultry,125kgof cereals, plus 55 m3 for domestic use (Falkenmark and Lindh, 1993). Totalrunoffcanbeusedtogiveacrude,andadmittedlyoptimistic,estimateofthe total volume of water that could be made available if water supplies are fully developed forhumanuse,leavinglittleifanywatertosustainnaturalaquaticorwetland ecosystems.EnglemanandLeRoy(1995)usedapercapitarunoffof1700m3/yrasa benchmarkfor"waterstress"and1000m3/yrasabenchmarkfor"waterscarcity,"and combined national data on total runoff with the 1994 United Nations medium population projectionbycountry(summarizedinTable2.2bycontinent)tocomputerunoffper capita.Thiswascomparedtothebenchmarkvaluestoexaminetrendsinwater availability.Byyear2050some58nationscontainingnearly4.4billionpeoplewillbe experiencing either water stress or scarcity, as compared to a population of less than 0.4 billionsoclassifiedin1990(Fig.2.2).Populationandwateravailabilityforseveral countries are compared in Table 2.3, where it can be seen that the United States, despite large deserts, as a whole is a water-rich nation. Israel, with only 461 m3/capita, employs the world's most technologically advanced and efficient water management system. Thecontinuouslysustainableorfirmyieldfromunregulatedstreamdiversions generallyrangesfromzeroto20percentofthemeanannualdischarge,withthelower values applicable to more arid and smaller watersheds. Reservoirs are necessary to divertTABLE 2.2Projected Global Population Growth Population, millions by year Continent1950199020202050 Growth rate 1990-2050, annual % Europe549722723678- 0.1 North America1662783583890.6 Oceania132740460.9 Asia1,403 3,1864,7445,7411.0 Latin America1664406768391.1 Africa2246331,3482,1412.1 World Total2,5215,2867,8899,8341.0 Source: United Nations 1994 medium projection of population. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.4 FIGURE2.2Currentandprojectedworldpopulation,showingthepopulationinnationshaving different levels of runoff per capita (after Engleman and LeRoy, 1994) TABLE 2.3 Runoff per Capita for Selected Countries Population millionsRunoff per capita, m3 Country Runoff depth, mm 1990205019902050 Israel104 4.78.9461241 Algeria724.955.7690309 Egypt5956.3117.41,046502 Ethiopia9847.4194.22,320566 Morocco57324.347.91,151585 Haiti3966.518.61,696593 Nigeria33396.2338.53,203910 Pakistan582121.9381.53,8381,227 India634850.61639.92,4511,271 Dominican Republic4107.113.22,8131,519 China2931155.31606.02,4241,743 Turkey26056.1106.33,6191,910 United Kingdom49257.461.62,0901,947 Mexico18284.5 161.54,2242,211 France33856.760.53,2623,059 Germany56179.464.22,5203,113 Iran7258.9163.14,4284,972 Japan1448123.5110.04,4284,972 United States253249.9349.09,9157,101 Russia (former U.S.S.R.)248280.4318.919,49317,142 Brazil816148.5264.346,80926,291 Source:Adapted from Engleman and Leroy (1994). agreaterpercentageofthesurfacerunofftohumanuseonacontinuousbasisby equalizing irregular streamflow; water is stored during periods of high flow and delivered duringperiodsoflowflow.Aswaterdemandincreasesrelativetorunoffthereservoir storagevolumeavailableforflowequalizationmustincrease,andasdemandcenters RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.5 growitalsobecomesnecessarytodevelopincreasinglydistantwatersources.Because themosteasilydevelopedresourcesareusedfirst,eachadditionalcubicmeterof sustained water yield typically requires more effort to extract than the previous one due tothe law of diminishing returns. The cost of each new project, the marginal cost of water supply, tends to increase over time. Biswas (1991) states that the unit cost of water from the next generation of domestic water supply projects is often 2 or 3 times higher than the presentgeneration;nearlyalltheeasilyavailablesourcesofwaterhavealreadybeenor are now under development. With many of the best irrigable areas already in use, the real cost of new irrigation schemes in Asia has also doubled or tripled between the mid-1950s and mid-1980s (Table 2.4). Insummary,thereisacleartrendtowardincreasingwaterscarcityrelativetoits traditionaluses.Reservoirsedimentationwillexacerbatethisscarcitybyreducingthe yield from already-developed supplies. 2.2 IMPORTANCE OF RESERVOIRS Dam construction has been undertaken on a large scale worldwide. Data collected by the International Committee on Large Dams (ICOLD) are limited to "large" dams, defined as those over 15 m tall, as reported by member nations. As of 1986, the most recent year for whichdataareavailable,therewereabout39,000largedamsworldwide,includingthe 36,237structuresintheICOLDinventoryplusapproximately3000structuresinthe formerU.S.S.R.notincludedinthedatabase(ICOLD,1988).However,amuchlarger number of impoundments are created by structures less than 15 m tall. An example of the largenumberofsmallerimpoundmentsissuggestedbydetailedU.S.datacontainedin the National Inventory of Dams maintained by the U.S. Army Corps of Engineers. As of 1995 the inventory included 74,053 dams meeting minimum criteria of over 2 m tall with at least 60,000 m3 of storage volume. Of these, only 5460 dams (7 percent) exceeded the ICOLD 15-m height criteria and are classified as large structures. The size classification and storage capacity for dams in the United States (Table 2.5) illustrate the large number ofsmallerimpoundments.DatacompiledintheUnitedStates(Dendyetal.,1973) indicatethatsmallerstructurestendtofillwithsedimentmuchmorerapidlythanlarger ones. Thepresentdayandfutureimportanceofdamstosocietyhasbeensummarizedby Veltrop(1992).Damsareauniquetypeofinfrastructureinthatthecommoditythey produce, a regulated water supply, tends to increase in value with time as water supplies becomeincreasinglyscarcerelativetodemand.Theimportancetosocietyoftoday's reservoirscanbeexpectedtoincreaseovertimeaspopulation,economicactivity,and irrigationdemandgrows.Whilemodernhydraulicsystemsconsistofmanyelementsto appropriatebothsurfaceandgroundwatersupplies,inmanyregionsreservoirsarethe singlemostimportantcomponent.However,uncontrolledsedimentaccumulationmakes storage reservoirs the key non-sustainable component of modern water supply systems.Intermsofconsumptiveusevolume,irrigationisthemostimportantuserofwater from reservoirs. Irrigated acreage has been expanding at the rate of about 30 percent per decade.One-thirdoftheglobalharvestcomesfromthat17percentoftheworld's croplandreceivingirrigation,andirrigationdeliveriesworldwidenowequal5timesthe averageflowoftheMississippiRiver(Postel,1993).Irrigatedacreagebycountyis summarized in Table 2.6.Irrigation accounts for 69 percent of global water withdrawals RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.6 RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.7 and 89 percent of consumptive use (Shiklomanov, 1993). At least half of the increases in agricultural productivity attributed to the "green revolution" in less developed nations has beenattributedtotheavailabilityofirrigation(Svendsen,1991).Reservoirssupplya large(althoughunquantified)portionofirrigationsupplies,andleakagefromirrigation canals supplied from reservoirs constitutes an important source of groundwater recharge in areas as diverse as India (Dhawan and Satya Sai, 1991) and the Caribbean (Quiones-Aponte, 1991). Approximately24percentofallelectricalgenerationcapacityworldwideisfrom hydropower,yetonthegloballevelonly14percentoftheeconomicallyexploitable potential has been developed (Gleick, 1993). In countries with very high levels of RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.8 hydropowerexplorationtheUnitedStates,Germany,Switzerland,andtheUnited Kingdom between 70 and 75 percent of the economically exploitable potential has been developed.Thisprobablyrepresentsanupperlimittowhatisachievable,andisitself probablyexcessiveinlightofpresentknowledgeoftheenvironmentaldamagesto riverine systems and fisheries associated with hydropower development. There is no accurate information on the extent to which municipal and industrial uses depend on reservoirs but it is known to be both large an increasing rapidly as the global populationsimultaneouslygrowsandbecomesmoreurban.Forinstance,in1950there werefewerthan100citieswithapopulationexceeding1million,but660citiesare projected to reach this size by year 2025. Other important beneficial uses from reservoir storageincluderecreation,floodcontrol,navigation,coolingwatersupply,reservoir-based fisheries, and ice jam control. Whereasthetwentiethcenturyhasbeencharacterizedbyarelativeabundanceof water and many water resource development projects, the twenty-first century is expected to be characterized by increasing water scarcity in response to continued population and economicgrowth.Thereissufficientwatertoprovidefoodandfiberfortheincreased world population, but, because of the uneven distribution of water and population growth, thewater-scarcecountriesinareassuchasnorthernAfricacanexpecttoseepopulation outstriptheavailableirrigationsupplies,makingthemincreasinglydependentonfood imports. It has been projected that China will become a major importer of food to satisfy apopulationthatisnotonlygrowingnumericallybutisalsoimprovingitsdietas incomesincrease.Thiswillsignificantlyincreasedemandforfood,fodder,andfiber (Brown, 1995). While The Economist (Anon., 1995) editorializes that market rather than Malthusianforceswilldealwiththeproblemoffoodscarcityonthegloballevel,itis alsorecognizedthatincreasingdemandcanmeanhigherfoodpricesandpoorpeople mayeatevenless.Incountrieswithoutadequaterainfallorirrigationsupplies,those countriesfacingwaterscarcity,higherfoodpricescanbecomeacriticalbalanceof payments and security issue. Efficiency increases can reduce the requirement to add new capacity,butbothpopulationandeconomicgrowthcannotbesustainedfroma diminishingwaterresourcebase,andtheprogressivelossofwatersuppliesbyreservoir sedimentation can only exacerbate the problem of global food security. 2.3 CONSEQUENCES OF SEDIMENTATIONAll rivers transport sediment as well as water, and dam construction impacts the transport ofbothsubstancesbutwithimportantdifferences.Becauseriverstransportmuchmore water than sediment it takes much longer to fill a reservoir with sediment than with water, so much more that the gradual accumulation of sediment tends to be ignored. That most sedimentaccumulatesunderwaterwhereitisnotvisiblefurtherremovestheproblem from popular, political, and engineering consciousness. But the most important difference is this: water can be easily removed from a reservoir but sediment cannot. As reservoirs age,theimpactsofsedimentationarebecomingmoresevereandbetterrecognized,and the frequency and severity of sediment-related problems is increasing. 2.3.1 Upstream ConsequencesAwiderangeofsediment-relatedproblemscanoccurupstreamofdamsasaresultof sediment trapping. Storageloss.Sedimentdepositioninthereservoirpoolwillreduceandeventually eliminate usable storage capacity, making the reservoir useless for either water supply or flood control. If the spillway capacity is based on flood storage within the reservoir, sedimentation can render the dam unsafe when this flood storage is lost.Deltadeposition.ThecoarserportionoftheinflowingsedimentloadisdepositedRESERVOIRS AND SUSTAINABLE DEVELOPMENT2.9 whereriversenterreservoirs,formingdeltadepositswhichnotonlydepletereservoirstoragebutcanalsocausechannelaggradationextendingmanykilometersupstreamfromthereservoirpool.Channelaggradationcanincreasefloodingofinfrastructure,communities,andagriculturallandsonfloodplains;increasegroundwaterlevels, creatingwaterloggingandsoilsalinization;reducenavigationalclearancebeneath bridges; and submerge upstream intakes (recall Fig. 1.2). If delta areas become heavily vegetated,theupstreamfloodlevelscanbefurtherelevatedbecauseofincreased hydraulicroughness,andthevegetationcantrapsedimentpromotingadditionalaggradation.Inaridzonesthetranspirationfromlargeareasofphreatophytic vegetation in delta areas can significantly increase water losses from the reservoir. For example,evaporativelossesfromthedeltaofElephantButteReservoirontheRo GrandeinNewMexicowereestimatedat176Mm3/yrpriortoconstructionofalow-flow conveyance channel through the delta in 1951 (Gorbach and Baird, 1991). Navigation. Both commercial and recreational navigation can be severely impaired by sedimentaccumulation,especiallyindeltaareasandinthevicinityoflocks.In navigationprojectscreatedbyachainoflow-headdamsandlocksalongariver, essentiallyopenriverconditionswillprevailduringhighdischargestherebypassingmostsedimentthroughthestructures.However,sedimentaccumulationcanoccurinlocks,approachchannels,anddeltaregions.Recreationalaccesscanbeimpairedas sediment accumulates at marinas and boat ramps. Airpollution.Inseasonallyemptyirrigationreservoirs,desiccateddepositsoffine sediment can be eroded and transported by wind, creating a nuisance and health hazardto nearby communities (Danielevsky, 1993; Tolouie, 1993). Earthquakehazard.Sedimentdepositshaveagreatermassthanwater,andsome researchindicatesthatthepresenceofsedimentagainstthedamcansignificantlyincrease the force of earthquake shaking against the structure (Chen and Hung, 1993).Sediments accumulating near the dam may be liquefied by earthquake shaking so that they flow toward and bury bottom outlets, entering and clogging any conduits that are open. At the large Tarbela dam on the Indus River in Pakistan, it was estimated that 6 to12monthswouldberequiredtorestoreirrigationandhydropowerserviceafteran event of this nature (Lowe and Fox, 1995). Abrasion.Inhydropowerfacilities,sedimentcoarserthan0.1mmwillgreatly acceleratetheerosionofturbinerunnersandPeltonwheelnozzles,andinthecaseof angularquartzsediments(e.g.,fromglaciatedwatersheds)andhigh-headoperation, damagecanbecausedbyevensmallergrainsizes.Thisreducespowergeneration efficiencyandrequiresremovalofgeneratingunitsfromserviceforrepair.Sediment sizesuptoboulderscanbepassedthroughbottomoutletsonsteepstreams.Abrasion can damage gate seals, outlet works, aprons, and spillways. Energyloss.Lossofstorageeliminatesthepotentialtocapturehighflowsfor subsequentenergygeneration.Whenaseriesofhydropowerstationsareconstructed along a river, the tailrace from one power station may discharge into the delta reach of another downstream reservoir. Delta deposition can elevate the streambed and tailracewater level, reducing the available power head and possibly flooding the power station if there is no remedial action. Intakes and outlets. Sediments can block or clog intakes and low-level outlets at dams andcanclogorotherwisedamagegatesnotdesignedforsedimentpassage.During extremefloods,depositionofmanymetersofmaterialcanoccurinafewhours. Sedimentanddebris17mdeepweredepositedinfrontofValdesiaclaminthe DominicanRepublicduringthepassageofhurricaneDavidin1979,cloggingthepower intakes for approximately 6 months. Shoreline erosion and Landslides. Reservoirs having a narrow or dendritic planform RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.10 canhavelongshorelinesinrelationtosurfacearea,andlargeerosivewavescan develop in broad reservoirs in windy area. Shoreline erosion can be substantial in areas ofloessalsoils.Forexample,thesixCorpsofEngineersdamsontheupperMissouri Riverimpound1200kmofriver,butcreate9560kmofshoreline.Averageshoreline recession rates in the lakes range from 1 to 2.5 m/yr (Dorough and Ried, 1991). While shorelineerosion,evenattheseratherhighrates,isnotathreattomaintenanceof overallreservoircapacity,itcanbecomeaseriousproblemwhenshorelineproperties are costly or when structures are affected.Landslidesanddebrisflowscanpartiallyorcompletelyfillreservoirs.Landslides totaling 177 Mm3 collapsed into Chinas Anmenzia reservoir during the first 16 months ofimpounding(Qian,1982).AsmallhydropowerdaminPuertoRicofilledbya landslideisshowninFigure2.3.Theseeventscanalsocausecatastrophicdambreak-typefloods.The265-mVaiontthinarchdamnewBellunoinVenetoProvince,Italy wasprecipitouslyfilledbya240-Mm3rockslideonthenightofOctober9,1963, displacingwaterinthereservoirwhichovertoppedthestructuretoadepthof100m and created a flood that killed 2600 people downstream (Jansen, 1983). Ecology. Changes in sediment loading and sediment accumulation within the pool can dramaticallyalterreservoirecology,affectingspeciescompositionandboth recreational and subsistence fishing. In areas where sedimentation continues unabated, open-waterhabitatwilltransitiontowetlandsandeventuallytouplandbecauseof continuedsedimentdepositionabovethenormalpoolelevationduringfloodflows.A large fraction of the organics, nutrients, and contaminants occurs in particulate form or is sorbed onto clays, and the deposition of organic sediments can playa large role in the reservoir oxygen budget. 2.3.2 Downstream ConsequencesRiverreachesdownstreamofdamssufferlargeenvironmentalimpactsduetoflow reductionandalterationofhydroperiod,reductionofsedimentload,alterednutrient dynamics, temperature changes, and the presence of the migration barrier imposed by the dam plus the upstream impoundment. Because organic material, nutrients, and pollutants suchaspesticidestendtobeassociatedwithfinesediments,damscancontrolthe mobilityoftheseecologicallyimportantmaterialsthroughthefluvialsystem,The trappingofsedimentbythedamalsohasimportantengineeringconsequences downstream. Downstream impacts have been summarized by Petts (1984). Streammorphologydownstreamofdamscanbedramaticallyimpactedbyreduction inthesupplyofbedmaterialsediment.Clearwaterintheriverchanneldownstreamof thedamwilltendtoscourthestreambedcausingittocoarsen,degrade,andbecome armored.Coarseningofthebedcanmakeitunsuitableasecologicalhabitatand spawning sites for both native and introduced species. Channel degradation can increase bothbankheightandbankerosionrates,increasescouratdownstreambridges,lower waterlevelsatintakes,reducenavigationaldepthincriticallocationsandlower groundwatertablesinriparianareasadverselyaffectingbothwetlandsandagricultural areas. Recreational use can be affected, as in the Grand Canyon where the sandbars used as campsites by river-rafters have been significantly eroded due to trapping of sand by the upsteam Glen Canyon Dam (Bureau of Reclamation 1994). Sedimenttrappingbyreservoirsreducesthesuspendedsolidsconcentration downstream,whichmayhavemanybeneficialeffects.TheSuspendedsolidslevelsof manyrivershavebeendramaticallyincreasedduetoupstreamdeforestationand development.Sedimenttrappinginreservoirsisbeneficialtoaquaticecosystems sensitive to elevated suspended solids levels, including costal marine ecosystems such as grass beds and coral reefs harmed by sediment discharged from rivers draining disturbed landscapes (Rogers, 1990).RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.11 RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.12 Many types of recreational uses also benefit greatly from reduced suspended sediment andenhancedwaterclarity.Clark(1985)examinedtheoff-sitecostofsoilerosionand concludedthatrecreationalimpairment(boating,fishing,swimming)constitutesthe singlelargestcategoryofcostsattributedtoincreasedsuspendedsolidsonriversinthe United States because of the large size of the recreational industry, which accounts for 7 percentofalljobs.Sedimenttrappinginreservoirsalsoreducestheneedfornavigation dredginginriversandriverineharbors.Sedimenttrappingbydamscanevenaffect coastalmorphology.Forexample,suspendedsedimentloadsinthelowerMississippi decreasedimmediatelyafterclosureoftheGavinsPointDamontheupperMissouri River in 1953, and in 1984 the suspended sediment discharge to the Gulf of Mexico was onlyone-thirdthepre-1953level.Reducedsedimentdischargeisamajorfactor contributingtotherapidshorelinerecessionandsubsidenceoftheMississippidelta (MeadeandParker,1984).TheerosionalhistoryoftheNiledelta,about1000km downstreamofAswanDam,issummarizedinFig.2.4.Alargefractionofthenutrients andorganicmaterialtransportedbyriversoccursasalgalcells,finelydividedorganic detritus, or in association with clays. Modification of the production and transport of this organicmaterialbythedam-reservoirsystemcanhaveimportantecological consequences downstream.Reservoirs can greatly reduce the downstream transport of Figure 2.4Erosional history of Nile delta (Khatagy and Fanos, 1993). The first Aswan Dam was completed in 1902 and the Aswan High Dam in 1967. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.13 detrital organic material used as a food source in the downstream ecosystem. Conversely, reservoirswithaprolongeddetentionperiodcandischargewaterenrichedwith limnoplankton. Insummary,damscanhavecompleximpactsonallfacetsoftheriverinewater-sedimentsystem,andtheseimpactscanbefar-ranginginbothtimeanddistance.Dam constructionrepresentsasingularlylargeandpotentiallypermanentalterationtoriver systems,andtheunderstandingandmanagementofsedimentisnolessimportantthan hydraulic control if reservoirs are to provide sustainable, long-term service at acceptable levels of environmental impact. 2.4 CONCEPTS OF RESERVOIR LIFE Reservoirsedimentationhasbeenmethodicallystudiedsincethe1930s(Eakinand Brown,1939),butdamengineeringhashistoricallyfocusedonstructuralissues,giving relativelylittleattentiontotheproblemofsedimentaccumulation.Thethree-volume treatiseEngineeringforDams,authoredin1945byCreager,Justin,andHinds,failsto evenmentionsedimentation,andthe1960versionoftheBureauofReclamation's publication Design of Small Dams covers the topic in a single page, but the 1987 version expanded this topic to an entire appendix (Strand, 1987).Withreasonablelevelsofmaintenance,thestructurallifeofdamsisvirtually unlimited. Most reservoirs are designed and operated on the concept of a finite life which willultimatelybeterminatedbysedimentaccumulationratherthanstructural obsolescence. For instance, Strand (1987) states: Allreservoirsformedbydamsonnaturalwatercoursesaresubjecttosome degree of sediment inflow and deposition. The problem confronting the project planner istoestimatetherateofdepositionandtheperiodoftimebeforethesedimentwill interferewiththeusefulfunctioningofareservoir.Atthetimeofdesign,provisions shouldbemadeforsufficientsedimentstorageinthereservoirsoasnottoimpair reservoirfunctionsduringtheusefullifeoftheprojectorduringtheperiodof economic analysis. Several concepts of reservoir life are summarized below, as adapted from Murthy (1977). Designlifeistheplanningperiodusedfordesigningthereservoirproject.Planning andeconomicstudiesaretypicallybasedonaperiodnotexceeding50years,whereas engineering studies often incorporate a 100-year sediment storage pool in the design.Project life is the period during which the reservoir can reliably serve the purposes for which it was originally constructed. When the reservoir can no longer serve its intended usebecauseofsedimentaccumulation,ithasreachedtheendofitsdesignprojectlife. However,operationmaycontinuewitharevisedorscaled-downprojectdesign.Given the probabilistic nature of hydrologic events, the definition of project "failure" should be basedontherelationshipbetweendecliningstoragevolumeandreliability.Serious interference with the original project design can be expected before half the capacity has beensedimented,andatmanysitessedimentinterferencebecomesseriouswhenonlya few percent of the original capacity has become sedimented.Economiclifeistheperiodoverwhichtheeconomicbenefitsfromprojectoperation exceedcosts.Continuedoperationofaprojectiseconomicallyjustifiedaslongasthe present value of the project benefits exceeds that of project costs, in accordance with the principles of engineering economy (James and Lee, 1971). Usablelifeistheperiodduringwhichthereservoirmaybeoperatedforeitherits original or a modified purpose, whether or not such use generates net economic benefits. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.14 Reservoiroperationmaybesubsidizedbeyonditseconomiclifebecauseofsocial, political,environmental,orotherconsiderations,orbecauseofmarketimperfections. Reservoiroperationmayalsobesustainedtopostponethehighcostofdamdecommis- sioning.Attheendofitsusablelife,thedammaybedecommissionedandbreached, removed, or simply abandoned. 2.5 GEOMORPHIC STAGES OF RESERVOIR LIFEReservoirlifecanbedescribedingeomorphictermsasathree-stageprocess.This approach adopts a much longer range perspective than traditional economic planning and engineeringdesignprocedureswhichuseafixedtimehorizonlimitedtotheearlystage ofreservoirlife,whensedimentsarecontinuouslybeingtrappedbuthavenotyet interfered with the operations considered in the original design. The long-term evolution ofanimpoundmentisillustratedschematicallyinFig.2.5,andeachstageisdescribed below.Pre-impoundmentsedimentbalance.Priortodamconstruction,mostriverreaches are approximately balanced with respect to sediment inflows and outflows. Sediments maytemporarilyaccumulateinsomechannelreachesbutaremobilizedand transported downstream by larger floods. While no river reach is ever totally balanced withrespecttosediment,andsomechannelsmayexperiencelong-termcyclesof aggradationanddegradationovertimescalesofcenturies,thetotalamountof sedimenttransportedthroughareachismuchlargerthantherateofaggradationor degradation within the reach.Stage1,continuoussedimenttrapping.Damconstructiondrasticallyalters hydraulicconditionsintheriver,convertingtheflowingstreamintoapool characterizedbylowvelocityandefficientsedimenttrapping.Coarsebedmaterial load is deposited as soon as stream velocity diminishes as a result of backwater from the dam, creating delta deposits at points of tributary inflow. Most finer sediments are carriedfurtherintothereservoirbyeitherstratifiedornon-stratifiedflowand accumulatedownstreamofthedeltadeposits.Thesefinersedimentsfirstfillinthe submergedriverchannel,afterwhichcontinueddepositionproduceshorizontal sedimentbedsextendingacrossthewidthofthepool.Sedimentsaretrappedduring all flood events.Stage2,mainchannelandgrowingfloodplain.Atsomepoint,thereservoir transitions from continuous deposition to a mixed regime of deposition and scour, and the rate of sediment deposition is reduced compared to continuous sediment trapping. Inwidereservoirs,thisstageisalsocharacterizedbythetransitionofsediment depositsfromhorizontalbedstoachannel-floodplainconfiguration.Thistransition willoccurnaturallywhensedimentationreachesthespillwaycrest;amainchannel willbemaintainedbyscour,anditsbaselevelwillbeestablishedbythespillway. Sedimentdepositioncontinuesonfloodplainareasoneithersideofthechannel, causing the floodplain elevation to rise above the spillway elevation, as photographed inFig.1.1b.Innarrowreaches,theScourchannelmayoccupytheentirereservoir width and the floodplain may be absent.The channel-floodplain configuration may also be created by reservoir draw down for sedimentroutingorflushing,inwhichcaseboththemainchannelandadjacent floodplainswillbesubmergedduringnormalimpoundingandthebaselevelofthe mainchannelatthedamwillbeestablishedbytheelevationofthelow-leveloutlet. Sediments will be deposited in both channel and floodplain areas during impounding. Scouringduringdrawdownwillremovesedimentfromthechannelbutnotthe floodplains,whichwillgraduallyriseinelevationassedimentcontinuesto accumulate.RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.15 FIGURE 2.5Long-term evolution of an impoundment. Stage3,fullsedimentbalance.Sedimentinflowandoutflowareessentiallyinfull long-termbalancewhentheamountandgrainsizedistributionofsedimententering thereservoirisbalancedbythematerialpassingthedam.Thekeytoachievingfull sedimentbalanceisthedischargeofcoarsematerial.Considerableupstream aggradationmayoccurabovethespillwaycrest,anddeltadepositsmustreachthe dambeforethisbalanceisreached.Sedimentmovementthroughthereachisnot necessarilythesameaspre-impoundmentconditions:sedimentdischargemay becomemoreepisodic.Sedimentsofallsizesmayaccumulateupstreamofthedam duringsmallerevents,butmajorfloodscanwashoutlargevolumesofaccumulated sediment.Inreservoirssubjecttohydraulicflushing,thesedimentreleasemaybe asynchronous with respect to the seasonality of sediment inflow.RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.16 Mostoftheworld'sreservoirs,beingrelativelyyoungand,withoutsedimentmanage-ment, are in stage 1. The structural and operational requirements for operation in stage 2 or 3 have been investigated or implemented at relatively few sites worldwide. 2.6 RATE OF STORAGE LOSS Theaveragerateofsedimentaccumulationinreservoirscanbeexpressedinvolumetric unitssuchasmillionsofcubicmeters(Mm3)oracre-feetperyear.However,for comparingreservoirsofdifferentsizesitisconvenienttoexpresstherateof sedimentationasthepercentoftheoriginalstoragevolumelostperyear.Sedimentation ratecanalsobeexpressedintermsofthereservoirhalf-life,theyearsrequiredtoinfill halftheoriginalcapacity.Becausetheefficiencyofsedimenttrappingdeclinesas reservoir capacity is reduced, the half-life does not represent half the time required to lose all storage capacity (Lajczak, 1995). The ability of a storage reservoir to fulfill its design function will already be severely compromised by the time half the storage capacity has beenlost(Dendyetal.,1973;Murthy,1977).Thus,theconceptofhalf-lifeisamuch morerealisticindicatorofthe"life"ofaconventionalstoragereservoirthanthetime required for complete sedimentation. The half-life concept is itself only a crude indicator ofthetimeframeinwhichsedimentationmayseriouslyinterferewithoperations,since therateatwhichvariousbenefitsarelostcanvarywidelyasafunctionofsediment accumulation.Storagelosswilldirectlyaffectwatersupplyandfloodcontrolfunctions, butattherun-of-riverpowerplantstoragelossmayhavelittleconsequenceaslongas sedimentsdonotentertheturbinesorotherwiseinterferewithoperations.Nevertheless, sedimentexclusionbecomesproblematicinreservoirsinadvancedstagesofsedimen-tation.Therearenoaccuratedataontheratesofreservoirsedimentationworldwide,but, fromtheavailabledata,Mahmood(1987)estimatedthatabout1percentofthe worldwidecapacityislostannually,equivalenttoabout50km3ofannualcapacityloss. As of 1986, the capacity-weighted average age of world storage was estimated at only 22 years, with about 1100 km3 of gross storage capacity having been sedimented. Dendy et al.(1973)examinedtherateofsedimentationin1105U.S.reservoirsandfoundan inverse relationship between pool volume and sedimentation rate (Table 2.7). Six percent ofthereservoirswithcapacitiesexceeding12Mm3(10,000acreft)hadannualstorage loss rates exceeding 1 percent, whereas no reservoir larger than 1200 Mm (1 x 106 acre ft) had an annual rate of storage loss exceeding 0.5 percent. High siltation rates could be TABLE 2.7Summary of Reservoir Siltation Rates in the United States Annual storage depletion, % Original reservoir capacity Mm3Acrefeet Number of censused reservoirsMeanMedian Half life at median depletion rate, years 0 0.0120 101903.562.0025 0.012 0.12310 1022572.001.2042 0.123 1.23102 1032831.020.6281 1.23 12.3103 1041760.810.5591 12.3 123104 1051070.430.27185 123 1230105 106690.230.14357 >1230106 107230.160.11455 Source:Dendy ety al. (1973) RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.17 found in all parts of the country. Crowder (1987) estimated the rate of storage loss in the coterminous48statesintheUnitedStatesat0.22percentperyear,equivalentto2020 Mm3ofannualstorageloss.Ofthistotal,486Mm3wasestimatedtobecontributedby cropland erosion. This suggests that, in the United States, significant capacity loss during thetwenty-firstcenturymayaffectmanysmallerimpoundments,butfewofthelargest reservoirs.However,sitesofallsizescanexperiencelocalizedsedimentationproblems and create problems downstream of the dam because of the cutoff in sediment supply.Sedimentationratesinotherareasoftheworldareoftensignificantlyhigherthanin theUnitedStates.AnalysisofdatafromZimbabwepresentedbyVanDenWallBake (1986) suggests that about half the reservoirs in that country are losing capacity at a rate exceeding0.5percentperyear.GogusandYalcinkaya(1992)examineddatafrom16 reservoirsinTurkeywhichhadhydrographicdataavailableanddeterminedthatthe annual rate of storage loss ranged from 0.20 to 2.40 percent, with a mean of 1.2 percent. DatabyAbdelhadi(1995)indicateannualstoragelossratesof0.7percentinMorocco and 2.3 percent in Tunisia. Storage loss at 2.3 percent annually has been previously cited forChina.Varmaetal.(1992)statethatthegeneralpracticeinIndiaistodesigna reservoir for a life of 100 years, and the results of sedimentation studies indicate that the actuallifeofreservoirswillnearlycoincidewithdesignlife.Anannualsedimentation rateof0.5percentappearsrepresentativeofoverallconditionsinIndia,asdiscussed below.Nationallyorregionallyaveragedsedimentationratesobscurethegreatrangeinthe rate of storage loss among reservoirs. Within any geographic area the rate of storage loss willvarywidelybecauseofdifferencesintheratioofreservoircapacitytoinflow. Variationinsedimentyieldfromdifferentcatchments,expressedintonspersquare kilometerperyear,furthercontributestothewidevariation.Thisvariationmeansthat somereservoirswillhaverelativelyshortlives,andotherswillbeextremelylong-lived. DatafromIndiaareusedtoillustratethiscondition.Availabledatasuggestthatlarge dams in India are silting at an average annual rate of about 0.5 percent (Morris, 1995b). While this is a rather modest rate of storage loss, if it is unabated the twenty-first century willwitnessthelossofoverhalfIndia'spresentreservoircapacity.Because sedimentationratesarenotuniform,somesiteswillbecomeseverelysedimentedmuch earlier than others. The available data suggest that about 23 percent of India's reservoirs, representing24percentofthenationalreservoircapacity,maybeseriouslyaffectedby sedimentationbyyear2020,andnearlyhalfofthenation'sreservoirsmayexperience severesedimentationbyyear2100.However,theleast-affected20percentofthe reservoirs,thosewithlargecapacitytoinflow(C:I)ratios,willnotbecome50percent sedimented until after year 2500.2.7 CONCEPTS OF SUSTAINABLE DEVELOPMENT Population growth, economic expansion, mechanization, and improved engineering skills have profoundly altered the surface of the planet, and human society now appropriates or destroysanestimated40percentoftheglobalprimaryproductivity(Vitouseketal., 1986).Thelargestandmostrapidchangeshaveoccurredduringthetwentiethcentury, and trends point toward continually increasing disturbance. Looking even relatively short distancesintothefuture(e.g.,onehumanlifespan)extremeconflictsareforeseeablein someregionsbetweenthehumanpopulationandnaturalecologicalsystems,andthe abilitytofeedandprovideevenminimalacceptablelivingstandardsforasignificant fraction of the projected twenty-first century population is seriously questioned. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.18 Developmentpatternswhichproduceexploitativeandunsustainablepracticessucha globaloverfishing,widespreaddegradationoflandandwaterresources,lossofnatural habitatandbiologicaldiversity,globalclimatechangeandcontinuingrapidpopulation growththatoutstripstheavailablenaturalresourcesandinfrastructureandcreates conditionsofmassiveimpoverishment,allreduceoptionsandopportunitiesfor subsequent generation.In1987theU.N.WorldCommissiononEnvironmentandDevelopmentIssuedthe reporttitledOurCommonFuturewhichproposedtheconceptofsustainable developmentasMeetingtheneedsofthepresentwithoutcompromisingtheabilityof futuregenerationstomeettheirownneeds.Manyinterpretationsoftheconceptof sustainabilityhavebeenadvanced.Bruce(1992)proposedthreebasicelementsof sustainability:protectionoftheplanetsbasiclifesupportsystem,provisionof continuousflowofgoodsandservicesderivedfromtheearth'snaturalresources,and equitableresourcedistribution.Weiss(1993),interpretingsustainabilityfromthe standpointoftherightsandobligationsbetweengenerations,statesthat"International rights, are to receive the planet in no worse condition than did the previous generation, to inheritcomparablebiodiversityinthenaturalandculturalresourcebases,andtohave equitable access to use and benefit from the environmental system." The three principles ofintergenerationalfairnessproposedbyWeiss,access,quality,andconservationof optionsanddiversity-mayberephrasedwithinthecontextofwaterresource sustainabilityasquantity,quality,anddiversity.Allthreeelementshaveimplications from the standpoint of reservoir design and management.Quantity of water. Although the hydrologic cycle continuously distills and circulates theearth'ssupplyofwater,itistheprovisionofreservoirstoragethatallowsthe irregular flow of surface waters to be regulated to meet societal needs. However, most storagereservoirsaredesignedandoperatedonthebasisoffinitelifetimes. Sustainabilityrequiresthatthetraditionalconceptoffinitereservoirlifebereplaced with a longer-term view focusing on sustained use.Water quality. Sediment is the most abundant pollutant in runoff from disturbed and human-occupied land. Furthermore, many of the toxic materials in streams, including metals,radionucleides,nutrientspesticides,andotherorganicsubstances,aretightly adsorbedontosedimentparticles(MeadeandParker,1984).Themannerinwhich reservoir sediments are managed will directly affect water quality in the reservoir and thedownstreamriverreach,withimpactspotentiallyextendingmanyhundredsof kilometersbelowmajordams.Reservoirmanagementalsoaffectsimportant parameterssuchaswatertemperature,particulateanddissolvednutrients,and dissolved oxygen.Diversity.Damconstructionconvertsfree-flowingriversintoaseriesoflakes, drasticallyalteringtheaquatichabitatandfoodchainandeliminatingimportant migratoryspeciessuchassalmonfromtheriversystem.Whiledamconstructionis nottheonlythreattonaturalriversystems,itisthemostimportantbecauseofthe magnitude and permanence of its impacts. Virtually no river system in the world has notbeensurveyedforpotentialdamsitesandthereiscontinuedpressuretodevelop newsites.Mostriversarealreadyimpoundedandnaturalriversystemsarebeing systematicallyeliminated.Themaintenanceofdiversitydoesnotrequirethatdam construction be avoided. Rather, a balance should be achieved between the dedication ofriverstodevelopmentactivities,suchasdamconstruction,andthepermanent preservation of underdeveloped rivers and associated ecosystems for the maintenance of diversity and other benefits.The Wild and Scenic Rivers System in the United States represents a small step in the direction of maintenance of natural river systems.Nevertheless, by 1990 there were RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.19 965,000riverkilometersintheUnitedStatesinundatedbyreservoirswhereasonly 15,000 river kilometers had been protected under the Act (Graf, 1993). This represents a lopsided ratio of 64: 1 in favor of dams. Because the environmental impacts of dams can extend many kilometers upstream and downstream from the pool, the ratio of developed to protected river lengths is even more skewed toward development than suggested by the inundationmileagealone.Furthermore,manyofthedesignated"wild"reachesare themselvesfarfromnatural,beingsandwichedbetweenupstreamdamswhichdivert waterandalterhydroperiodandsedimentloads,anddownstreamdamswhichblock migration.MostcountrieslagbehindtheUnitedStatesintermsofenvironmental protection of rivers.Theconceptofdiversityalsoincludesculturaldiversity,anddamconstruction sometimes involves the inundation of areas populated by ethnic minorities, with resultant relocations and cultural disintegration.Waterresourcedevelopmentissustainablewhenitdoesnotsignificantlyor progressively diminish the resource quantity or quality over the long term, and maintains the diversity of natural habitats and ecosystems.The need for sustainable patterns of development is increasingly being recognized in the engineering community (Prendergast, 1993), and 16 different aspects of sustainability werediscussedbyPlate(1993).Essentialelementsofsustainabilitypertinenttowater resource planning and engineering are summarized by the following six concepts:1.Protecttheenvironmentandconserveresources.Manageandprotectboth naturalandman-madeenvironmentstoavoiddegradationovertime.Preserveadequate habitattomaintainthediversityofnaturalecosystems,includingnecessarymigration corridors.Adoptdesignswhichconserveresources,whichminimizeenvironmental impacts, and which do not impact natural preserves.2.Sustaininfrastructurefunction.Designandoperateessentialinfrastructurefor thedeliveryofservicesforanindefiniteperiod,orprovideforitseventualreplacement with new infrastructure at the end of its economic service life.3.Anticipatechange.Continuouslycollectdataandanalyzetrendstoanticipate change. Regularly revise strategic plans and update design and operations to reflect better data, new technology, and evolving needs. Social needs and technology change and may requirenewinfrastructure.Similarly,olderinfrastructurewhichnolongermeetsthe highest public or environmental interest should be modified or removed from service.4. Prepare for failure of engineered systems. Stochastic concepts are widely used in water resource engineering, and over a long period of time an extreme event more severe thanthedesigneventcanbeexpectedtochallengetheproject.Theconsequencesof extremenaturaleventsshouldbeconsideredinprojectdesignsothatextremestorm, flood, or drought do not have disastrous consequences.5.Considerlong-termconsequences.Avoidshiftingconsequencessuchas environmental remediation and resource exhaustion forward onto future generations. Do not assume that difficult environmental problems posed by today's actions and trends will beanyeasiertosolveinthefuture.Plananddesignaccordingtotheconceptthat preventionisalwayslesscostlythanremediation.Recognizetheinherentlimitationsof engineeringsolutions,andthelimitedabilityoftheenvironmenttosustainessential functionsandmaintaindiversityinthefaceofcontinuouslyincreasinghumandemands. Whenthelong-termconsequencesofpresentactionsortrendscannotbeaccurately predicted, and involve the possibility of large-scale or long-term social or environmental disruption,orasignificantlossofdiversity,theavoidanceofpotentiallypermanent damage or catastrophe should be the primary objective. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.20 6.Humandevelopmentanddiversity.Cultivatehumanresources,establishing motivatedtechnicalpersonnelandlocalcitizenparticipation.Aknowledgeable, informed,andconcernedpopulationprovidesaneffectivebasisforachieving sustainability.Recognize, respect and protect human cultural diversity.Watermanagementliesatthefoundationofsustainabledevelopment.Whetherour children and theirs will inherit an earth with accessible resources, diversity, and beauty is beingdeterminedbytodaysactions,whichfrequentlygivelittleconsciousthoughtto implicationsbeyondaprojectseconomiclife.Applicationofsustainabilityconcepts seeks to bring these questions into the sphere of conscious decision making. Even though our vision of the distant future is always clouded, it should not be ignored. Designers of waterresourceinfrastructurehaveaspecialresponsibilitytoconsidersustainability criteriaintheirwork,andtocommunicatesustainabilitycriteriatodecisionmakers, becauseofthelargeandlong-termimpactstoinfrastructure,developmentpatterns,and the environment caused by projects with their purview.2.8 RESERVOIRS AS NONRENEWABLE RESOURCESIfundevelopedreservoirsitesabound,thenimpoundmentscanbecontinuously constructedtooffsetstoragelossesfromsedimentation.However,iftheavailabilityof replacementsitesislimited,andthereisnoidentifiablesubstitutefortheservices providedbyreservoirs,thentoday'ssitesrepresentinfrastructurethatshouldbe developedandoperatedontheconceptoflong-termsustaineduse,asopposedtothe concept of replaceable engineering infrastructure with a limited economic life time.Theconceptofsustainableuseisalreadyappliedtogroundwatersystemswhich consist of unique and irreplaceable natural components (the aquifer and its recharge area) andengineeredcomponents(wells),Groundwateroverpumpingisawidelyrecognized exampleofnon-sustainablewateruse,sincethecontinuedextractionoffreshwaterat ratesexceedingrechargewilleventuallybecurtailedbythedecliningwatertableorby salineintrusionincoastalaquifers.Groundwatermanagementtodayemphasizes sustainability throughtheprotectionofaquifersfromcontamination,maintenanceofthe balance between recharge and extraction rates, and protection of recharge areas.Primereservoirsitesarenolessimportantthanaquifers.Reservoirsalsorequire uniquenaturalcomponents(damsiteshavingappropriatetopography,hydrology, geology)andengineeredcomponents(dam,deliverycanals,etc.).Replacementofthe engineeredcomponentshasnopurposeifthestoragevolumeislosttosediment accumulation. Whereas aquifers have been recognized as sustainable resources which are beingthreatenedbyoverexploitationandpoormanagement,neithertheneednorthe potentialforsustainablemanagementofreservoirsiswidelyrecognized.Anumberof factorsindicatethatreservoirsshouldbeconsideredasirreplaceableresources,noless importantthanaquifers,andshouldbedesignedandoperatedinaccordancewiththe objective of sustained long-term utilization (Morris, 1995a). Theinventoryofunexploitedreservoirsitesisneitherlargenorgrowing,and sedimentedreservoirsitescannotbereadilyrecycledforreuse.Theinabilitytoreuse reservoirsitesmakesdamsdistinctfromotherclassesfromengineeringinfrastructure, suchasbuildingsandroads,thatcanbedemolishedandreplacedwithnewand serviceablestructuresonthesamesite.Althoughdamscanbereconstructedand incrementally heightened, new construction implies new non-sedimented reservoir sites. Existing reservoirs generally occupy the best available locations with respect to water availability, site geology, competing uses and proximity to demand areas. These sites RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.21 offergeologicandhydrologicconditionssuitablefordamconstructionandarelocated appropriately for the intended uses. A flood control structure must be located upstream of theareaitwillprotect,navigationprojectsarelocatedasafunctionofriverslope,and water supply projects should be reasonably close to irrigable soils or cities because of the costoftransportandpoliticalwatertransferissues.Thedecliningquality,progressive remoteness, and increasing competition for undeveloped sites incrementally drives up the cost of development. Only hydropower sites have great latitude in their location because electricitycanbetransmittedcheaplyovervariedterrainandisareadilymarketable international commodity.Sedimentedreservoirscannotindefinitelybereplacedbynewdamsatnewsitesand manyoftoday'sreservoirsareirreplaceablebecauseofuniquesitecharacteristics.For example,the3172-Mm3SriramaSaganreservoirinAndhraPradesh,India,cannotbe replaced by a new structure further downstream because such a location would inundate partoftheirrigationareaandcouldnotdelivertotheexistinggravitycanalsystem.An upstream location is infeasible for political reasons; it would require inundation of fertile lands and villages in the neighboring state of Maharastra. Significant heightening will be extremelycostly,willinundatemorelandanddisplacemorepeople,andcannotbe repeated indefinitely. Similar constraints can restrict the options for future replacement of manyreservoirsofallsizes,rangingfromsmallerbutlocallyimportantsitestoprojects ofnationalimportancesuchasEgypt'sAswanHighDamandthefederaldamsonthe Colorado River in the United States.Competitionfornaturalresourcesisincreasingbecauseofpopulationandeconomic growth and the need to maintain environmental diversity. Dam construction is but one of manystrong,conflicting,andlegitimateinterestscompetingforriversandtheriparian landsfloodedbyreservoirs,andnewdamconstructionworldwideiscomingunder increasingopposition.Astheinventoryofun-exploitedriverscontinuestodiminish, opportunities for continued dam construction should also be expected to diminish. While the world has not run out of potential reservoir sites, areas that are rich in potential sites (South America) are not the areas of greatest population pressure and water scarcity (Asia and North Africa).Itisnotfeasibletorecoverthecapacityoflargerreservoirsbydredgingbecauseof high costs, which in the United States presently start at about $2.50/m3 and increase as a function of longer pumping distances, deeper digging depths, and increasing complexity of disposal. For example, at $2.50/m3 it would cost $83 billion to restore Lake Powell on theColoradoRivertoitsoriginalcapacityoncefullysedimented,assumingasuitable disposalsitecouldbefoundfor33km3ofsediment.AperusalofICOLDstatistics indicatesthatreservoirvolumeisoftenabout3ordersofmagnitudegreaterthanthe volume of the dam itself, and this great disparity in volume makes sediment removal and disposalfarmorecostlythanconstructionoftheoriginalstructure.Toretaincapacity, dredgingmustberepeatedforaslongasthereservoiristoremaininservice,creating recurrent costs and severe long-term disposal problems. Although reservoir dredging will become increasingly common, high cost and disposal limitations indicate it will not be a generally applicable solution to the sedimentation problem.Hydraulic flushing can be an effective mechanism for removing sediments, emptying the reservoir through low-level outlets, and allowing natural, flows to scour out deposits. However,manyreservoirscannotberemovedfromserviceforflushing,andinmany cases flushing cannot maintain the original reservoir volume. Flushing also releases large volumesofsedimentdownstreamcreatingpotentiallyseriousproblemsincluding interferencewithwaterintakes,increasedsedimentloadingondownstreamreservoirs, andadverseimpactsonfisheries,theenvironment,andrecreationaluses.Inmanycases these impacts will make flushing infeasible. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.22 Resource substitution can replace some of the services provided by todays reservoirs. Energy,ifnotsuppliedfromanotherdam,canbesuppliedfromanalternativesource (e.g.,fossil,nuclear,wind,orsolar).However,thereisnosubstituteforwater,only increasingly efficient use. While there is great room for efficiency gains, especially in the irrigationsector,itseemsoptimistictobelievethatthefloodandfiberrequiredan additionalpopulationof4billionexpectedbyyear2050canbeproducedfromawater resource base that is continually shrinking from declining reservoir capacity.Water production technologies such as desalination continue to be hopelessly energy-intensive and costly compared to the diversion and storage of streamflow. The production costofseawaterdesalinationexceeds$1percubicmeter,about10timesthecostthat municipalwaterusersintheUnitedStatespayforthedeliveryofpressurizedportable water. Even if desalination cost were greatly reduced, it would offer little benefit to non-costal areas, to less industrialized areas, or to irrigators who will require low-cost water. Desalting technology is not foreseen as a viable substitute for reservoirs. Ofallthefactorspointingtotheneedtoconsiderlong-termsustainabilityin reservoirs,perhapsthemostimportantistheexperienceatsiteswhichalreadysuffer serious sedimentation problems. Dam owners are seeking and implementing solutions to sedimentationproblemswhichwillsustainthefunctioningofthejeopardized infrastructure. This trend is expected to continue.2.9 ECONOMICS AND SUSTAINABILITY Considerationsoflong-termsustainabilityarenotapartoftraditionalprojectanalysis, which has been based largely on economic criteria and projections using a horizon on the orderof30to50years.Thistimeframeresultsnaturallyfromabasicpremiseof economic analysis, the time preference for income as expressed in the discount rate, The present value (PV) of any future value (FV), n years in the future, at a discount rate of i, may be expressed by: PV = FV / (1 + i )n(2.1 ) Largewaterresourceprojectsarecapital-intensive,andfollowingconstructionare anticipated to provide a stream of benefits that extend long into the future. Benefits such asirrigationmaygrowonlyslowlyanduncertainlyasfarmersgraduallyadopttonew practices and crops, and seek to penetrate new markets. Discounting is an essential tool to helpdeterminetheamountofpresentinvestmentthatcanbejustifiedtogeneratefuture benefits.Ithelpsfocuslimitedinvestmentresourcesintoprojectswithhighernear-term benefits,asopposedtoprojectswhoseviabilitydependsonlowreturnsthatextendfar into an uncertain future.Economicanalysisessentiallyignoresbothcostsandbenefitsmorethan50yearsin the future. Fifty-year projections are tenuous at best, and the present worth of income or costsbeyond50yearsarerenderednegligiblysmallbydiscounting.Consider,for example, the present value of an income stream $100/yr discounted at an annual rate of 7 percent, as presented in Fig.2.6. The present worth of this income stream starting year 0 (today) through year 49 is $1477. The present worth of the additional income from years 50 through 99 is $50, about 3 percent of the present value of the first 50 years of income. Economicanalysisisnotanappropriatetoolforanalyzingissuesoflong-term sustainability. Major aspects of project design already fall beyond the sphere of economic analysisandaretreatedasmarketexternalities,suchaswaterpollutionwhichis recognizedtohavesocial,economic,andecologicalcostswhicharenotaccurately reflected in the marketplace. Weiss (1993) points out that market externalities also tend to be evaluated within the context of the present generation alone, and future generations are not explicitly represented.RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.23 FIGURE 2.6Present worth value of an income stream of $100 per year, discounted at 7% annually. (This series is computed receiving $100 in year 0, $93.46 in year 1, etc.) 2.10 CRITERIA-BASED APPROACH TO SUSTAINABILITYSomeaspectsofdesign,suchasdamsafety,areestablishedonthebasisofminimum engineeringdesigncriteriawhichmustbemetinallprojects.Projectanalysisevaluates benefits and costs for various alternatives, all of which meet established safety standards. Safe projects are not compared against less costly but unsafe ones. A similar approach is applied,atleastintheory,tosocio-environmentalissues,wherelevelsofenvironmental complianceareestablishedthroughlegislative,regulatory,andpoliticalprocesses.The economicanalysiscomparesenvironmentallyacceptable(thoughnotequivalent) alternatives against one another. Concepts of long-term sustainability and sediment management have heretofore rarely been incorporated into reservoir projects, and the design strategy for China's huge Three Gorgesprojectrepresentsasignificantdepartureinthisrespect.Sustainedusewas establishedasanengineeringcriterion,andfromtheplanningstagestheproject incorporatedthosestructuralandoperationalfeaturesrequiredtoachievealong-term sedimentbalancewithnolossinprojectbenefits.Alternatively,itcouldbestatedthat project benefits were limited to those that could be sustained, as opposed to focusing on higherinitialbutunsustainableinitialbenefits.Abalancebetweensedimentinflowand outflow, in terms of both total sediment load and grain size distribution, is expected to be achievedinthismorethan600-km-longreservoirafterabout100yearsofoperation.A similarcriterion-basedapproachisrecommendedfordefiningandimplementing sustainable reservoir management at both existing and new dams. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.24 New projects. The sustainability criteria suggested for new reservoirs is to design for aminimumof1000yearsofoperation.Thismaybeaccomplishedbybalancingsedimentinflowandoutflow,byproviding1000yearsofsedimentstoragecapacity,or some combination thereof. The analysis period should equal the shorter of: (1) the period requiredtoreachsedimentequilibriumor(2)1000years.Therecommended1000-year designhorizonhasbeenselectedasacompromisebetweentheinadequacyofthe100 yeartimeframegenerallyemployedtoday,andtheimpossibilityofmakinglong-rage projections. However, a 1000-year service life is not at all unrealistic for dams. Schnitter (1994) list 12 ancient dams that have been at least 2000 years of service. While a 1000-year analysis of a physical processes in the reservoir can give only approximate results, it can help define the time frame and types of sedimentation problems to be anticipated, the recommended management strategies, and data collection needs. Existing reservoirs. Many benefits from existing reservoirs may not be sustainable as sedimentationprogresses,andsedimentmanagementtopreservecurrentcapacitymay not be feasible. Therefore, the sustainability criteria for exiting reservoirs will necessarily bedifferentfromthatapplicabletonewsites.Atexistingreservoirs,sustainable sedimentmanagementshouldseektobalancesedimentinflowandoutflowacrossthe impoundedreachwhilemaximizinglong-termbenefits.Thismayinvolvestrategiesto minimize sediment inflow, enhance sediment release, or a combination of both.Asedimentbalancewilleventuallybeachievedatallsites,theresultofeither managementornaturalphenomena.Sedimentmanagementseekstomanipulatethe system in such a way that usable storage capacity or other benefits are maximized when this balance is reached. 2.11 ACTIONS TO ACHIEVE SUSTAINABLE USE Sustainablesedimentmanagementencompassestheentirefluvialsedimentsystem consisting of the watershed, river, reservoir, and dam. It is not achieved without cost. As aminimum,itinvolvesbetterinformationandimprovedmanagement,butitmayalso include large operational and capital costs for watershed management, the construction of low-leveloutletsorbypassworks,temporaryremovalofthedamfromservicefor sedimentmanagementactivities,thereleaseofincreasedvolumesofwaterdownstream for sediment discharge, and dredging. It will frequently not be economically attractive to manageexistingreservoirstomaintaintheoriginal activestoragevolume,implyingthat firm yield, power production, or other benefits may be reduced in the long term. At some sitesthelossofbenefitswillbezero,butatothersitwillbelarge.Nevertheless, continued operation with reduced net benefits is preferred to project abandonment.2.11.1 Sedimentation AssessmentThestartingpointforallanalysisisanassessmentofthenatureandseverityofthe sedimentationproblem.Thisevaluationseekstodescribethephysicalprocessof sedimentationwithinthereservoirandthehistoricalandprobablefuturerate,location, and grain size of sediment deposition. The assessment may be performed on the basis of geomorphic analysis and modeling in combination with analysis of recent bathymetry to determinethehistoricalpatternandrateofdeposition.Factorspotentiallyinfluencing futureratesofsedimentdeliveryshouldbeevaluatedincludingupstreamreservoirand land use trends. The role of upstream impoundments is of particular importance. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.25 Sediment yield can be greatly reduced by upstream impoundments of all sizes, including farmpondsifbuiltinlargeenoughnumbers.However,ownersofupstreamreservoirs mayalsoinitiatesedimentmanagementmeasuresdesignedtopasssedimentsthrough their impoundments, thereby increasing the sediment loading on downstream sites.2.11.2 Evaluation of Sedimentation ImpactsIdentifytheanticipatedconsequencesofsedimentation.Whatbeneficialuseswillbe threatened by sedimentation and over what time frames? Consequences of sedimentation should be evaluated, not in terms of the present market price for the water supply, but on thebasisofprobableresponsesrequiredtocompensateforlossofreservoirdeliveries. Thus,thesedimentationeffectsonanirrigationreservoirthatsustainsthelivelihoodof 100,000 families might be analyzed in terms of the costs and consequences of relocating thefamilies,ratherthanthecurrentpricepaidforthewater(whichmaybeheavily subsidized and thus very low).Modestchangesinwatersupplycanbeaccommodatedwithinthecontextof increasing efficiency of use. However, when one or more reservoirs provide the primary or sole reliable source of water for a region, loss of these supplies can virtually eliminate thehydraulicbaseforthesociety.Theconsequenceswillbeespeciallyacutein nonindustrial,irrigation-dependentcountrieswithrapidlygrowingpopulations.Such areasmaylackboththehumanandcapitalresourcesneededtoachievedramatic increases in efficiency, such as a transition from traditional furrow to drip irrigation, and the industrial base required to sustain their population from imported foods.2.11.3 Identification of Priority SitesRemedial work at existing sites should be prioritized by sedimentation rate and expected consequences.Existingreservoirswithserioussedimentationproblemsshouldnotbe dismissedas"mistakes."Rather,theyshouldbeinstrumented,studied,andmanagedas fieldresearchlaboratoriesinsedimentmanagement.Itisatsuchsitesthatmanagement techniquescanbetestedandengineerstrained.Workshouldbeinitiatedfirstatthe highestprioritysites,thenatprogressivelylowerprioritysitesuntilalltherecognized problems have been addressed.2.11.4 Screening of Sediment Management AlternativesAllsedimentmanagementalternativespotentiallyfeasibleateachsiteshouldbe considered, including alternatives such as eventual dam heightening or construction of a newdamelsewhere.Ifconstructionofanewdamisanalternativethatwillbe implemented,itisessentialthatthealternativesitebeidentifiedandpreservedforthis eventualuseandadecommissioningstrategybeidentifiedfortheto-be-abandonedsite. Littlehasbeenaccomplishedifanalternativesiteisidentifiedbutsubsequently appropriated for another use because it has not been acquired and protected for eventual reservoir construction. Even site acquisition does not guarantee that a project can be built, suchasDenver'sTwinForksDamwhichwassuccessfullyopposedonenvironmental grounds.AnimportantconsiderationinthecaseofDenverwastheavailabilityofother supply options, including water conservation.The evaluation of management alternatives should not focus exclusively on the supply side,butmustalsoconsiderincreasingwateruseefficiencyanddemandside management. For example, irrigation reservoirs and their distribution systems are not RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.26 constructed to deliver water but to grow crops. Exploit every opportunity to reduce water loss,improveirrigationscheduling,enhanceagronomictechniques,reducepostharvest losses,establishanefficientwatermarket,andotherwiseimprovethenetbenefitper cubic meter of available water.2.11.5 Implementation Measures and Scheduling Afterascreeninganalysisthatidentifiesthemostpromisingmanagementstrategieson both the supply and demand sides, the most promising alternatives should be analyzed in more detail to determine technical and economic feasibility. This stage may involve field andengineeringstudies,includingpre-designenvironmentalandmodelingstudies. Implementationmaybescheduledusingassigneddatesorperformancecriteriasuchas thelossofacertainpercentageofstoragecapacityortheadvancementofthedelta depositstoacertainpointwithinthereservoir.Someprocedures,suchassediment routing, may not become effective until a significant portion of the reservoir volume has become filled with sediment, thus reducing the C:I ratio. 2.11.6 Design, Implementation, and Monitoring of Control Measures Thefinalstepinvolvesthedesign,implementation,andmonitoringofsedimentcontrol measures.Thesemayincludemodificationofthedam,watershedprotectionactivities, and demand-side management to increase water use efficiency by users.Yieldreductionduetosedimentationcanchange(orcreate)watermarketsandalter usepatterns.Assuppliesbecomeincreasinglyscarcebecauseofgrowthindemandor reductioninsupply,waterwillbedrawnawayfromlower-valueusestosustainhigher valueuses.Thus,inanirrigationdistricttheacreagedevotedtoanimalforagemay declinetosustainwaterdeliveriestohigher-valuecropsdevotedtodirecthuman consumption.Scarcitywillalsostimulateincreasedwateruseefficiencyandrecycling, suchastheuseoftreatedwastewaterforirrigation.Thistypeoftransitioninwateruse was outlined in a study of economic responses to water scarcity in Arizona (Kelso et al., 1973),whichconcludedthatasignificantreductioninwatersupplywouldreduce irrigated acreage of low-value crops but not curtail economic development.Thetruevalueofwaterisitsmarginalvalue,thepricethatoneiswillingtopayfor eachadditionalincrementinsupply,ratherthantheaveragepriceforamixtureoflow-cost and higher-cost supplies. The marginal cost of developing new supplies is typically higherthantheaveragecostofolder(cheaper)watersourcesmixedwithnewer(more costly)sources.Importantfinancialandwaterconservationbenefitsareattributedtothe use of marginal cost pricing in the water sector (Mann and Clark, 1993; Hirshleifer et al., 1970). Higher cost to users will encourage increasingly efficient use and simultaneously raise a reserve fund for new projects. This contrasts to current practice, which tends to be basedonaveragecosts,exceptfortheirrigationsectorwherewaterdeliveriesareoften heavily subsidized, an even worse strategy from the standpoint of achieving efficient use.Insummary,stainableusecriteriarequirethatprojectanalysisanddesignlookwell beyondtraditionalplanninghorizons.Itrequiresthatengineersanalyzetheabilityof dam-reservoirsystemstomanagesedimentaswellaswater,inthesamemannerthese facilitiesaredesignedandmanagedtocomplywithsafety,environmental,and operationalcriteria.Sustainabilityshouldbeconsideredfromtheinitialphaseofproject design, incorporating to the greatest extent possible those elements which will eventually lead to a long-term sediment balance. RESERVOIRS AND SUSTAINABLE DEVELOPMENT2.27 2.12 POPULATION AND FOOD SUSTAINABILITY In closing this chapter, it seems appropriate to point out that the efficient management of reservoirsandirrigationis,ofitself,insufficienttoensuresustainabledevelopmentor securefoodsupplies.Inanumberofcountriespopulationgrowthseemstoberapidly outstrippingtheavailablewaterresourcebase,thecaseofEgyptbeingparticularly instructive.EssentiallyallcroplandinEgyptisirrigated.TheNileiscompletely controlled by the Aswan High Dam where its total flow is now regulated to meet human needs.However,thetotalflowoftheNileprovidesinsufficientwatertomeetEgypt's growing needs.Flow from the White Nile will be increased significantly by projects such as the 350 m3/s Jonglei canal and others to divert water around, and thus desiccate, vast wetlands in Sudantoreduceevaporativelosses.ThiswillincreaseEgypt's55.5km3/yrallocationof waterfromtheNileby4.7km3/yr,a9percentincrease.Large-scalewastewaterreuse, recoveryofirrigationdrainage,groundwaterdevelopment,andincreasesinirrigation efficiencyarealsoinprogress.Nevertheless,evenunderanoptimisticscenario,Egypt will experience an annual water deficit of 10.5 km3 by year 2025