ReportNo. 44977-UG\ The Inspection Panel InvestigationReport Uganda: Private Power Generation (Bujagali) Project (Guarantee No. BOl30-UG) August 29,2008 About the Panel The Inspection Panel was createdinSeptember 1993 by the Boardo fExecutive Directors o f the World Bank to serve as an independentmechanism to ensure accountability in Bank operations with respect to its policies andprocedures. The Inspection Panel is an instrument for groups o f two or more private citizens who believe that they or their interests have been or could be harmed by Bank-financed activities to present their concerns through a Request for Inspection. Inshort, the Panelprovides a linkbetween the Bank and the people who are likely to be affected bythe projects it finances. Members of the Panel are selected"on the basis of their ability to deal thoroughly andfairly with the request brought to them, their integrity and their independencefiom the Bank's Management, and their exposure to developmental issues and to living conditions in developing countries."1 The three-member Panel is empowered, subject to Board approval, to investigate problems that are alleged to have arisen as a result o f the Bank having ignoredits own operatingpolicies andprocedures. Processing Requests M e rthe Panelreceivesa Request for Inspectionit is processedas follows: The Paneldecides whether the Requesti sprimafacie not barredfrom Panelconsideration. The Panelregistersthe Request-a purely administrative procedure. The Panelsends the Requestto BankManagement,which has 21working days to respondto the allegationsofthe Requesters. The Panel then conducts a short 21 working-day assessment to determine the eligibility of the Requestersandthe Request. Ifthe Panelrecommendsaninvestigation, andtheBoardapprovesit,the Panelundertakes afidl investigation, which i s not time-bound. Ifthe Panel does not recommendan investigation, the Board of Executive Directors may still instruct the Panelto conduct aninvestigation ifwarranted. Three days after the Board decides on whether or not an investigation should be canied out, the Panel's Report (including the Request for Inspection and Management's Response) is publicly available through the Panel's website and Secretariat, the Bank's Info Shop and the respective BankCountry Office. When the Panel completes an investigation, it sends its findings and conclusions on the matters allegedinthe Request for Inspectionto the Boardas well as to BankManagement. The BankManagementthenhas six weeks to submit its recommendationsto the Boardon what actions the Bankwould take inresponseto the Panel's findings andconclusions. The Board then takes the final decision on what should be done based on the Panel's findings andthe BankManagement'srecommendations. Three days after the Board's decision, the Panel's Report and Management's Recommendation are publicly available through the Panel's website and Secretariat, the Bank's Project website, the Bank's Info Shop andthe respectiveBank Country Office. 'IBRDResolution No. 93-10; IDA ResolutionNo. 93-6. .. 11 Table of Contents ABOUT THE PANEL .................................................................................................................................. I1 TABLE OF CONTENTS ........................................................................................................................... I11 ABBREVIATIONSAND ACRONYMS ............................................................................................... VI11 ACKNOWLEDGEMENTS ......................................................................................................................... X EXECUTIVE SUMMARY ........................................................................................................................ XI SUMMARYREPORT ............................................................................................................................ XVI CHAPTERI ................................................................................................................................................... 1 INTRODUCTION ......................................................................................................................................... 1 A. EVENTSLEADINGTHEINVESTIGATION .................................................................................................. TO 1 .................. ........................................................................... 4 3. Eligibility of the Request. .......... 4. TheInvestigation......................................................................................................................................... 7 5. Bank Operational Policies and Procedures Applicable to the Project B.REQUESTFORINSPECTIONRELATEDTOTHEPRIORBUJAGALIPROJEC CHAPTERI1 ............................................................................................................................................... 14 THE PROJECTAND ITS CONTEXT .................................................................................................... .14 A. POVERTY AND ENERGY INUGANDA: THEPOWER SUPPLY CRISIS............................................................. 14 I.ShortagesandLack of AccesstoElectricity................................................................................... I 4 2. Current and Planned Sources of Electricity ............................... I 6 B.ENVIRONMENTAL AND SOCIAL CONTEXT AND SETTING ........................................................................... 17 1.Hydrology of Lake Victoria and the Victoria Nile, and Hydropo 2. Lake Victoria and theImpact of DecliningLake Levels.............. 3. Bujagali Falls and Surrounding Habitats....................... .................................................................. 26 4. TheProject's Socio-economic and Cultural Setting C. THEPROJECTDESCRIPTION ....................................................................................................................... 29 I.TheResettlementProgram .......... ...... 31 2. Cultural Resources............ ........................................................................ .32 D.WORLDBANKINVOLVEMENTINUGANDAPOWERSECTOR, THELAKEVICTORIA ANDTHENILERNER BASIN ............................................................................................................................................................. 32 CHAPTERI11 .............................................................................................................................................. 35 ENVIRONMENTALISSUES..... ............................................................................................................... 35 A. THEENVIRONMENTAL COMPONENT OF THE SOCIAL AND ENVIRONMENTAL ASSESSMENTS ..................... 36 I.Adequacy of thestudies............................................................................................................................. 36 2. Environmental Management Plan sting and Future Hydro Projects ... 111 B.MITIGATIONMEASURES:KALAGALA THE OFFSETAGREEMENT ............................................................... c.SAFETY 50 OF DAMS ...................................................................................................................................... 54 CHAPTERIV ............................................................................................................ ................................ HYDROLOGICALAND CLIMATE CHANGERISKS ........................................................................57 57 A.INTRODUCTION .......................................................................................................................................... 57 B.THEHYDROLOGY ............................................................... 58 1.Appropriateness of HydrologicalData Series used in Project Design ....... OF LAKEVICTORIA AND THE VICTORIA NILE ................. .............58 2. Lake Victoria WaterLevels and Power Plant Operations on the Victoria Nile........ ..................... c.IMPACT 61 OF HYDROLOGIC RISKON ENERGYOUTPUT ................................................................................... 65 D.POTENTIALIMPACTOFTHEPROJECTONLAKEVICTO........................................................................ 68 E.CLIMATE CHANGERISKS........................................................................................................................... 71 1. ThePAD.............. 2. TheEconomic Study ................. 3. TheSSEA............. 4. Other Documents 5. Conclusions on Climate ChangeRisk................................................................... CHAPTERV .............................................................................................................................................. 77 ECONOMICAND ENVIRONMENTALANALYSIS OFALTERNATIVES ..................................... 77 A. INTRODUCTION .......................................................................................................................................... 77 1.The Context: Power Sector Developments and thePower Supply Crisis............... ........ 78 2. Demand Forecasts and Electricity Tariffs ......... ............................................................................. 80 B.ECONOMIC .................................................................................................. 85 1.BankPolicies.................................................. ANALYSIS OF ALTERNAT~~ES ........................ ..................................................................86 ~ 2. Terms of Reference for the Economic Study 3.1 Conventional Thermal Plants (201 1-2020) ..... ......................................................... 3.3. Small and Medium Scale Alternatives 3.4. Oil Resource c.THEPROJECTCOSTS.................................................................................................................................. 95 1.Bujagali Project Costs 2. Karuma Project Costs D.ASSESSMENTLEAST OPTIONSFOREXPANDINGPOWERGENERATION OF COST AND RELATED CONSIDERATIONS......................................................................................................................................... 102 1.Tariffs andAffordability 103 2. A Criterionfor Economic Acceptability of the Project: Internal Rate of Return Analysis ......................105 3. Macroeconomic Considerations in the Analysis of Alternatives 4. Externalities ...................................~..~..............,......................... E.ENVIRONMENTAL ANALYSISOFALTERNATIVES ..................................................................................... 109 1.Hydropower in Comparison to Alternative Power Generation Technologies within the Region............109 2. Hydropower Location Alternatives within Uganda 3. Alternative Project Configurations at Bujagali...... CHAPTERVI .................................................................................................................. ..........................118 ECONOMICEVALUATION:POVERTYREDUCTIONAND RISK ...............................................118 A. AFFORDABILITY AND POVERTYREDUCTION........................................................................................... 118 B.FINANCIALAND GOVERNANCE RISKS..................................................................................................... 121 1.RevenueProjections and theInstitutional Framework 2. Infiastructure Funds ......................~.....................................................................~..............................,,.,124 C. THEPOWERPURCHASE AGREEMENT ASSOCIATED ............................................................... AND RISKS 125 1.ThePower PurchaseAgreement ............................................................................................................. 125 2. Risks and ConsequencesAssociated with the Project PPA. 127 iv 3. Risk Mitigation Measures .................128 4. Conclusions on Distributi ...................................................................................................... 130 5. Disclosure of the PPA .......131 CHAPTERVI1 .......................................................................................................................................... 133 INVOLUNTARYRESETTLEMENT ..................................................................................................... 133 A.THEREQUEST'S CLAIMS AND MANAGEMENT ........................................................................ 133 B.CHANGINGCONTEXT: FROMTHEPRIORPROJECTTOTHEPRESENTBUJAGALIPROJECT.......................... RESPONSE 135 1.Management's decision to assesspast resettlement activities andprepare actionplans ....................... 136 sments of Past ResettlementActivities and Action Plans ...... 1.Method to assesslivelihood restoration .................................................................................................. 145 ..................................................................................................................... 144 2. Real orperceived unfulfilledpromises made in theprior project 3.1. Fishing...................................................................................................................................................... 3.3. Conclusions on fishingand agriculture............. 5.Land titles................................................................................................................................................ 151 6. VulnerablePeoples 8. Investment resourcesfor livelihood restoration 7. Housing, Public Services, an ............................................................................................... 154 9. Costs of Project Delay on Displaced Persons.................................................................................. 156 D.SHARING INPROJECTBENEFITS C I O. Overall Conclusions on Livelihood R AND ......................................................... 159 E.INDIGENOUS PEOPLES .............................................................................................................................. 161 CHAPTERVI11 ......................................................................................................................................... 163 CULTURALAND SPIRITUAL VALUES ............................................................................................. 163 A.INTRODUCTION ........................................................................................................................................ 163 B.THE2001RESETTLEMENTAND COMMUNITY DEVELOPMENT PLAN(RCDN)AND THE ACTION CULTURALPROPERTY MANAGEMENT................................................................................................ PLAN 165 c.PREPARATION OF THE PROJECT ............................................................................................................... 167 D.BUSOGA SPIRITUAL 168 E.BUSOGA CULTURAL DOMAIN.................................................................................................................. DOMAIN.................................................................................................................. 172 F.PANEL'SANALYSISPHYSICAL CULTURAL RESOURCES ........................................................................ - 172 173 2. Consultation I.Avoidance................................................................................................................................................ I74 3.Mitigation................................................................................................................................................ 176 G UNDERSTANDINGTHELOCAL . CULTURAL AND SPIRITUAL CONTEXT ...................................................... 178 H.PANEL'SANALYSIS- CRITICAL NATURAL HABITATS ............................................................................. 181 I.THECULTURALPROPERTY ....................................................................... 185 J.OPPORTUNITIES TO ADDRESS CULTURAL AND SPIRITUALISSUES ............................................................ MANAGEMENT PLAN (CPMP) 186 CHAPTERIX ............................................................................................................................................ 188 SYSTEMICISSUESAFFECTINGPOLICY COMPLIANCE ............................................................ 188 A.LEGACY 188 B.INCORPORATINGCLIMATE CHANGEINTOPROJECT ISSUES .......................................................................................... FROMPRECEDINGPROJECTS DESIGN .................................................................... 188 c.TIMELYDISCLOSUREINFORMATIONWITHINOF THE PROJECTCYCLE ..................................................... 189 D.TRANSPARENCY 190 E.CRITICAL NATURAL ISSUES PARTNERSHIPS............................................................... AND PUBLIC-PRIVATE HABITATS SACREDPLACES - - GUIDANCETO STAFF....................................... AND 190 ANNEXES .................................................................................................................................................. 192 V ANNEX A TABLE OF FINDINGS ..................................................................................................................... 192 ANNEX BINDEPENDENTREVIEW OF CONTRACTUAL ARRANGEMENTS ....................................................... 205 ANNEX c SPIRITUAL SIGNIFICANCE INBUSOGA CULTURE .......................................................................... 220 ANNEX DBIOGRAPHIES ............................................................................................................................... 227 FIGURES FIGURE 1WATER FLOWS FROM LAKE VICTORIA TO NALUBAALE. AND BUJAGALI KIIRA .............................. 20 FIGURE2 LAKEVICTORIA WATER LEVELS JANUARY 1900TO JANUARY OF2005 (SOURCE: LAKE VICTORIABASINCOMMISSION)..................................................................................................................... FROM FIGURE3 LEVELSTHELAKEAND OUTFLOW (2000-2005) (LAKEVICTORIA BASINCOMMISSION, 2006..59 AT 63 AND ............................................................................ 64 FIGURE5 WEIGHTED AVERAGE RETAILTARIFF DECEMBER -NOVEMBER 2006.................................. FIGURE4 COMPARISONOF WATER LEVEL OUTFLOW 2000 79 FIGURE6 GENERATION FORECASTS ............................................................................................................... 81 FIGURE7 SALES FORECAST UGANDAFOR ..................................................................................................... 82 FIGURE 8 PROJECTEDELECTRIC~Y TARIFFPATH (2000-16) ....................................................................... 104 FIGURE9 SITESOFCULTURAL SIGNIFICANCE (SOURCE: 2001RCDAP,P 107) . .......................................... 171 TABLES TABLE1SUMMARY OFWATERBALANCE LAKE FOR VICTORIA.................................................................... TABLE2 CHANGES INLAKELEVELS. FLOW ...........59 63 TABLE3 POWERSECTORPERFORMANCE(2001-05)...................................................................................... AND RELEASES OVER AND ABOVE THE AGREED CURVE 79 .................................................................. 84 TABLE5 ASSUMED GENERATIONCAPACITY EXISTINGIN2011..................................................................... TABLE 4 FORECASTSTECHNICALAND COMMERCIAL LOSSES OF 88 TABLE 6 ECONOMIC .............................................................. 95 TABLE 7 BUJAGALI EPCCOST EVOLUTION................................................................................................... COST ESTIMATE FORBUJAGALI AND KARUMA 97 EPC BUJAGALIPAD............................................................................................................................................ 101 TABLE 8 COMPARISONOF BUJAGALIAND KARUMA COST ESTIMATES: ACRES (2001) STUDY AND TABLE9 ECONOMIC ................................................................................. 101 TABLE10ESTIMATESTHEWEIGHTEDAVERAGE RETAILTARIFF PATH2011-2016.................................. COMPARISON OF SUPPLY PrucES OF 105 TABLE11CRITERIA AND WEIGHTINGSAPPLIED INMULTIPLE .................................. 111 TABLE12RESULTS OFMCAAND RISKANALYSISINRANKORDER........................................................... -CRITERIA ANALYSIS TABLE13ALTERNATIVE SITES CONSIDEREDFORLARGE HYDROELECTRIC PROJECTSWITHIN UGANDA ....111 112 TABLE14RANKINGOFOPTIONSWITHIN THE ENVIRONMENTAL CATEGoRY ............................................... 113 TABLE15RANKINGOFOPTIONSWITHIN THE SOCIO-ECONOMIC CATEGoRY .............................................. 113 PICTURES PICTURE 1PANEL .CRMUMEETING WITH REQUESTERS ................................................................................ 8 PICTURE 2 BUJAGALIFALLS .......................................................................................................................... 26 ............................................................................................... 28 PICTURE4 PANELTEAM AT BUJAGALIDAMCONSTRUCTIONSITE ................................................................ PICTURE 3 RAFTING AT THE BUJAGALIFALLS PICTURE5 KALAGALA FALLS........................................................................................................................ 29 51 PICTURE 6 MONITORINGTHE LEVELS OF LAKE VICTORIA ............................................................................. 60 PICTURE 7 PANELMEETING WITH PEOPLETO BE RESETTLEDUNDER THE INTERCONNECTIONPROJECT PICTURE 8 PANELTEAM MEETING WITH PEOPLERESETTLED AT NAMINYA .........................................................140 142 PICTURE 9 NABAMBA BUJAGALIPERFORMINGSPIRITUALCEREMONY......................................................... 170 PICTURE 10UNDATEDPICTURERECEIVEDBY PANEL EXPERT SHOWINGSPIRITUAL MEDIUM NFUUDU PERFORMINGSPIRITUALCEREMONY ........................................................................................................... 175 vi Maps IBRD36274: Uganda-PrivatePower Generation(Bujagali) Project IBRD36276: Uganda-PrivatePower Generation(Bujagali) Project -Project Site vii AbbreviationsandAcronyms AESNP AES NilePower Corporation AfDB African Development Bank APRAP Assessment o f Past Resettlement Activities and Action Plan BEL Bujagali Energy Limited BIP Bujagali InterconnectionProject BIU Bujagali ImplementationUnit BNWPP Bank-Netherlands Water Partnership Program BP BankProcedures BST Bulk SupplyTariff CDAP Community Development Action Plan CIDA Canadian International Development Agency CPMP Cultural Properties Management Plan CRMU Compliance Review and Mediation Unit(CRMU) o fthe African Development Bank (AfDB) DANIDA Danish International Development Agency DSP Dam Safety Panel DWD Directorateo fWater Development EAP Environmental Action Plan EIRR Economic Internal Rate o f Return EMP Environmental Management Plan EPC Engineering, Procurement and Construction EPRP Emergency Preparedness and Response Plan ERA Electricity Regulatory Authority ERT Energyfor Rural Transformation ESMAP Energy Strategy andManagement Action Program FIRRI Fisheries Resources Research Institute (now known as National Fisheries Resources Research Institute- NAFIRRI) G C M General Circulation Models GoU Government o fUganda HPP Hydroelectric Power Plant IA Implementation Agreement IBP International Best Practice IDA InternationalDevelopment Association IFC InternationalFinance Corporation IOH Institute o f Hydrology o f the United Kingdom IPCC Intergovernmental Panel on Climate Change IPP Independent Power Project IUCN World Conservation Union L C local government council representatives LVEMP Lake Victoria Environmental Management Project MIGA Multilateral Investment Guarantee Agency ... Vlll NAPE National Association o f Professional Environmentalists NBI Nile Basin Initiative NBTF Nile BasinTrust Fund NELSAP Nile Equatorial Lakes SubsidiaryAction Program NGO Nongovernmental organization NPV Net Present Value OP Operational Policy PAD Project Appraisal Document PAP Project Affected People PCN Project Concept Note PCDP Public Consultation and Disclosure Plan PEAP Poverty Eradication Action Plan PPA Power Purchase Agreement PSDO Power Sector Development Operation RAP Resettlement Action Plan RCDAP Resettlement and Community Development Action Plan ROW Righto fWay SEA Social and Environmental Assessment SEAP Social and Environmental Action Plan SSEA Strategic/Sectoral Social and Environmental Assessment T-Line Transmission Line TOR Terms o f Reference UETCL UgandaElectricity Transmission Company Limited UMEME Electricity distribution company Units andMeasures USD United States Dollars USh Uganda Shilling GWh Gigawatt hour kWh Kilowatt hour MW Megawatt i x Acknowledgements The preparation o f this Report would not have been possible without the support and valuable contributions o f many people and organizations. The Panel wishes to thank the Requesters and the communities who met with the Panel in the Project area. The Panel especially expresses its appreciation to the representatives o f the Requesters for arranging visits with affected people and for showing the Panel areas o f concern to them. The Panel also wishes to thank NGOs and other people and organizations with whom the Panel met. The Panel expresses its appreciation to the many national and local government officials inUganda with whom the Panel met. They provided valuable insights and information. The Panel is also grateful to the World Bank Executive Director's office for Uganda for the assistanceprovided. The Panel also wishes to extend its gratitude to World Bank Staff in Washington D.C. and in the Kampala office. The Panel wishes to thank Bank Management and Staff for their assistance inobtaining documents, providing the Panel with information, responding promptly to written requests, and assisting with logistical arrangements. The Panel i s grateful for the expert advice provided by Prof. Theodore Downing, Prof. Richard Fuggle, Mr. Graham Hadley, Prof. Peter Pearson and Prof. Carlos Tucci. The Panel also wishes to thank all the distinguished experts with whom the Panel met in Uganda. The Panel appreciates the professionalism exhibited bythem at all times. Finally, the Panel wishes to convey i s gratitude and appreciation to its former Chairperson, Edith Brown Weiss, and to the Secretariat for its resourceful handling o f this investigation, particularly to Peter Lallas, Executive Secretary, Dilek Barlas, Deputy Executive Secretary, Tatiana Tassoni and Serge Selwan, Operations Officers, Eduardo Abbott, expert consultant, and Shehan de Sayrah, Christine Curella and Trudy Rebert for their expertise and professional assistance. X Executive Summary Introduction. In March 2007, the Inspection Panel received a Request for Inspection related to the Private Power Generation Project, commonly known as Bujagali Hydropower Project (the "Project"). The Request was submitted by the Ugandan National Association o f Professional Environmentalists (NAPE) and other local organizations and individuals. The Project consists o f the construction o f the Bujagali hydropower plant on the Nile River near the Bujagali Falls, downstream from the existing Kiira - Nalubaale Hydropower Plants. It i s designed to provide an increase o f 250 MW o f power generation capacity to the national grid in Uganda. The Project would inundate Bujagali Falls and other natural habitats, which are sites o f cultural and religious significance to a large community o f people, and involve displacement and resettlement of people and families from their lands. The Project i s a Public-Private Partnership betweenprivate sponsors and the Government o f Uganda (GoU) that i s supported by private lenders and multilateral and bilateral development agencies. World Bank Group support includes a partial risk guarantee from the International Development Association (IDA, also referred to as the Bank), loans from the International Finance Corporation (IFC), and a guarantee from the Multilateral Investment Guarantee Agency (MIGA). A separate request relating to this Project was submitted to the independent recourse mechanism (IRM) o f the African Development Bank (ADB). The Panel and the IRM collaborated by sharing experts and conducting a joint field mission. The conclusions of the Panel and the IRM are, however, independent and based on different applicable policies. Context. Uganda i s facing a serious power supply crisis. The Panel wishes to place on record that it considers energy a crucial factor in Uganda's development. However, as this Report shows, energy production requires considerable care in order to ensure that social, economic and environmental aspects are properly considered, in line with Bank policy, to adhere to sound development practices and avoid situations where costs, including social and environmental costs, outweigh the benefits expected from what are usually sizable investments. The Request raises a number of environmental, hydrological, social, cultural, economic and financial concerns, and contends that a failure o f the Bank to follow its own operational policies and procedures in the design and appraisal o f the Project will result in serious harm to the people living in the Project area and to the environment. The Management Response o f April 2007 states that experienced Bank staff and consultants were engaged to work on the preparation o f this Project, that economic, financial, safeguard, technical and other analyses were done to a high standard, and that they took x i into account the findings o f the Panel's 2002 Investigation o f the previous Bujagali project inthe designand preparation o f the Project. Environmental Issues. In the context o f the Bank's policies on Environmental Assessment, OP 4.01 and on Natural Habitats, OP 4.04, the Panel found areas of compliance, including that the Project had been appropriately classified as category "A" and that the Kalagala Falls had been established as an offset for the natural habitats to be inundated by the Project. The Panel also found that the Bank complied with the procedures set forth inBank policy OP 4.37 on the Safety o f Dams. However, there were several important areas o f non-compliance with Bank policies. The Project did not appoint an independent panel o f environmental experts, as required under Bank policy for this type o f complex project, nor did it support needed capacity building for implementation o f social and environmental aspects o f the Project. The Project Social and Environmental Assessment (SEA) did not adequately make reference to the Strategic/Sectoral Social Environmental Assessment (SSEA) o f the separate Nile Basin Initiative, which analyzed issues such as climate change and cumulative effects. As a result, important information required under Bank policy was not disclosed in a timely manner as an integral part o fthe Project's documentation. In addition, neither the SSEA nor the SEA addressed the cumulative effects o f the existing and planned projects in a meaningful way. And while the Kalagala Falls have been established as an offset, inlight o f institutional weaknesses there i s no evidence that this offset site will be maintained in accordance with appropriate conservation and mitigationmeasures inconformity with sound social and environmental standards. Hydrological and Climate Change Risks. The Panel examined a range o f issues, including the impact o f hydrologic risk on energy output, the potential impact o f the Project on the levels o f Lake Victoria, and the risks from climate change. The Panel noted the substantial body o f analysis under the Project, and found that the hydrologic data sets used in Project design constitute a reliable data series and an appropriate baseline for analysis in compliance with OP 4.01. The Panel also found, however, important areas o fnon-compliance with OP 4.01 and OP 10.04. In particular, there is a discrepancy between the PAD and the Economic Study as to which water release regime will be in effect once Bujagali becomes operational. This brings into question the data basis for the Project's economic analyses and is likely to have resulted in a more positive conclusion to the Economic Study than would have been warranted. The SEA also considered that the Project's area o f influence ends downstream o f the Kiira-Nalubaale dams and therefore did not assess the Project's potential impacts on the changing levels o f Lake Victoria, as it should have. This i s particularly important because the lowering o f water levels in the Lake, as has occurred in recent years, brings significant social and environmental impacts. In addition, the PAD'S categorical assertion, without any reference to risk and uncertainty, that there will be no adverse effect on water release due to climate change during the Project life fails to express a potential risk factor, which was identified inthe SSEA, as required. xii Economic and Environmental Analysis of Alternatives. The terms o f reference (TOR) for the Economic Study called for a comprehensive update o f the first round o f the Bujagali Project. The Panel found, however, that the TOR encouraged a focus on relatively large grid-connected plant and did not draw attention to the evaluation o f smaller scale or off-grid alternatives. Ina country where only 5 percent o f the population is connected to the grid, it would be reasonable to expect attention to be paid to such options which might in theory more directly address local and rural poverty. The Panel found that the information in the Economic Study and the PAD on these options did not demonstrate full compliance with OP 10.04's requirement to evaluate alternatives. As part o f its assessment o f least-cost options, the PAD asserts that tariff rates may drop by up to 10 percent. The Panel found, however, that this should have been qualified to take into account the increases in Engineering, Procurement and Construction (EPC) costs and transmission costs after the Economic Study was prepared. The issue o f electricity tariffs and affordability i s o fhighimportance to the people andcommunities. It is the Panel's view that Management did not ensure that cultural and spiritual matters o f high significance at Bujagali Falls were adequately considered in Project preparation, and when comparing the Bujagali and Karuma alternatives. Alternative project configurations were unduly narrowed on the basis o f a-priori judgments rather than exploring all technically feasible options, including those that would not involve flooding the Bujagali Falls and thus have lower social and environmental costs, and laying them out ina systematic way along with their economic, social and environmental benefits and costs, so that judgements on optimal alternatives could be made with a full understanding o fthe trade-offs involved. Economic Evaluation: Poverty Reduction and Risks. The Panel examined poverty reduction issues and the risks and consequences associated with the Power Purchase Agreement (PPA) in the context o f OP 1.00 on Poverty Reduction and OP 10.04. The quantitative assessments o f costs and benefits in the Economic Study suggest that the Project would have largely positive direct impacts on Uganda's economy and thus appears to have complied with the requirement to show that the Project is likely to contribute to "broad based growth." The Panel notes, however, that the tariff figures provided in the Economic Study are likely to be based on an underestimate o f the cost o f electricity with the Project. In addition, much o f the expected direct benefit from Bujagali, especially in the early years, i s likely to be experienced by the better-off urban households, and electricity would still be too costly for many, especially poorer households. Neither the Economic Study nor the PAD, however, provides estimates o f the economic impact o f the Project on low- income households. The Panel found that, as with the previous Bujagali project, the PPA capacity charge is not related to output, so that the payment by the government-owned power purchaser, UETCL, will be the same under low hydrology as high hydrology, and also invariant to reduced plant availability. More generally, the Panel found that the introduction o f a cost- ... X l l l based formula in the 2005 PPA (as amended) represents a significant shift in risk away fiom the Project investors and lenders to UETCL. The high allocation o f risk to the power purchaser and eventually the GoU increases the possibility that the GoU will have to makepayments under its guarantee and/or increase tariff subsidies. Inthis context, the Project may not achieve the broad objective o f sustainable development and poverty reduction embodied in Bank Policies. The potential consequences are described in the Report. Involuntary Resettlement. This Project involves the rather unusual circumstances o f an ongoing, incomplete resettlement program developed under the prior plan for a dam at Bujagali Falls, based on a policy no longer applicable. The policy now applicable, OP 4.12 on Involuntary Resettlement, nevertheless has the same overall objectives, and both the old and new policy call for preparation o f a Resettlement Action Plan (RAP). Management chose to develop an "Assessment of Past Resettlement Activities and Action Plan (APRAP)", rather than a new RAP, with the justification that affected people had already been relocated and others had already received compensation under the prior project. The Panel considers that the overriding issue i s whether the approach taken meets the objectives and requirements of Bank policy. The way the APRAP was substituted for a full RAP, however, had far ranging consequences in terms o f complying with Bank policy. The Panel found that the critical policy requirement to census all displaced persons was neglected - a decision undermining much o f the policy objectives. The public consultation process was truncated, and the APRAP failed adequately to assess andupdatethe previous RAP to ensure compliance with Bank standards. The Panel found that the effects on the people o f the original displacement, ando f the ensuing delay, were not fully reflected inthe APRAP. On the critical question of livelihood restoration, the Panel concluded that the Project did not comply with the mandate o f Bank policy to improve or at least to restore, in real terms, the livelihoods and standards of living o f the people displaced by the Project. Many affected people also believe that other promises made under the prior project were not kept. The Panel did not find any evidence that Bank Management violated the provisions o f the Bank's policy on Indigenous Peoples, with regard to the Basoga people. Cultural and Spiritual Values. The Project i s moving into a neighborhood long inhabited with strong, complex cultural and spiritual traditions. To the Basoga people, the Bujagali Falls area, which i s to be inundated by the Project, i s inhabited by ancestral spirits. The Panel notes that studies prepared in 2001 for the prior project mapped individual and community level spirits. Problems emerged, however, with the so-called "appeasement of community spirits", which failed to lead to a lasting solution. Under the current Project, the consultation process has not yet led to satisfactory outcomes and mitigation efforts, required by Bank policy, and cannot be considered xiv completed. The Panel found that the Bank misjudged the Bujagali Falls as a cultural resource o f importance only to those living in the vicinity of the Falls, and that the consultation process excludedkey spiritual leaders o fthe wider Basoga community. Inaddition, Bank Management failed to prepare a Cultural Properties Management Plan, as requiredby policy. Such a plan should have identified Bujagali Falls not as a localized cultural site but as a significant cultural resource for the whole Kingdom o f Busoga, triggering rigorous safeguards for specific avoidance, consultation and mitigation as required under the Bank's Policy. OP 4.04 on Natural Habitats also contains provisions that are relevant to these issues. It provides that the Bank does not support projects that in the Bank's opinion involve the significant conversion or degradation o f critical natural habitats, a crucial question which requires considered judgment. The Policy states that "critical natural habitats" include "areas initially recognized as protected by traditional local communities (e.g., sacred groves)." Management took the view that the Project i s not significantly converting or degrading a "critical natural habitat," without providing adequatejustification for this determination. Considering the known spiritual importance o f the Bujagali Falls area, without such an explanation, one could also arrive at the opposite, i.e., that the inundation may be regarded as resulting in the significant conversion o f a critical natural habitat which would be in violation o f OP 4.04. The Panel found that there is an overriding need to address this issue to ensure compliance with Bank policies. Systemic Issues Affecting Policy Compliance. In closing, the Panel would like to emphasize that energy i s a crucial factor in Uganda's development. As this investigation shows, generating energy for development in a way that i s economically efficient, socially equitable and environmentally sustainable i s hugely complex and one ofthe major sustainable development challenges o f today. The results o f the Panel's investigation illuminate some o f these complexities in the Bujagali Project, which may also be relevant to other similar projects. These include addressing legacy issues from preceding projects, difficulties in achieving transparency on economic and other impacts in public-private partnership projects, incorporating climate change issues into project design, and issues regarding the application o f Bank policy on natural habitats to sacred groves and sacred places. xv Summary Report I. Introduction Right at the outset, the Panel wishes to go on record that it considers energy a crucial factor in Uganda's development. The findings o f this Report do not dispute this fact but show that energy production requires considerable care in order to ensure that social, economic and environmental aspects are properly considered, inline with Bank policy, to achieve sound development practices and avoid situations where costs, including social and environmental costs, outweigh the benefits expected from what are usually sizable investments. On March 5, 2007, the Inspection Panel received a Request for Inspection related to the Uganda: Private Power Generation Project, also known as Bujagali Hydropower Project (the "Project" or the "Bujagali Project"). The Project includes the proposed construction o f the Bujagali hydropower plant as well as the Interconnection Project, (which will construct transmission lines, a new substation at Kawanda, and the extension o f the substation at Mutundwe financed by the African Development Bank (AfDB)). The Request was submitted by the Ugandan National Association o f Professional Environmentalists (NAPE) and other local organizations and individuals. The Request o f 2007 followed a similar Request in 2001 when the Panel was asked to investigate the prior Bujagali Hydropower Project. The Panel issued its Investigation Report on May 23, 2002. Due to difficulties encountered by the former project sponsor, the first Project was terminated in September 2003. The subject o f the 2007 Request is the current, second effort o f the Government o f Uganda (GoU) to develop the Bujagali Hydropower Plant. The Project The Project consists o f the construction o f the Bujagali hydropower plant on Dumbbell Islandon the Nile River, at the Bujagali Falls, about 8km downstream from the existing Nalubaale and Kiira Hydropower Plants. The Nile River has its headwaters in Lake Victoria. Under the Project, a powerhouse complex providing a maximum capacity o f 250MW and a rock filled dam o f about 30 meters highwith spillway and other associated works i s to be developed. Adjacent to the powerhouse, a high voltage substation, the Bujagali Substation, through which all power generated from the Project will flow, i s to be constructed on the west bank o f the Victoria Nile. The reservoir will inundate the Bujagali Falls. The overall Bujagali Project also includes the construction o f 100 km o f transmission lines, a new substation at Kawanda and the extension o f the substation at Mutundwe financed by the African Development Bank (AfDB)under the InterconnectionProject. A private 'sector company, Bujagali Energy Ltd. (BEL), i s to develop the Project. `BEL i s xvi responsible for financing, constructing and operating the Project "on a Build-Own- Operate-Transfer basis." The Project is a Public Private Partnership between the private project sponsors, the GoU, multilateral and bilateral development agencies, and commercial lenders as beneficiaries o f the proposed IDA Guarantee. The total Project cost i s estimated to be around US$798.6 million. The International Development Association (IDA) supports the Project through a partial risk guarantee o f US$l15 million. The Project i s also financed through, inter alia, International Financial Corporation (IFC) loans and a Multilateral Investment Guarantee Agency (MIGA) guarantee. In total, the World Bank Group's financial support to the Bujagali Project i s around US$360 million. IDA'SBoard o f Executive Directors approved the IDA Guarantee on April 26,2007. The Claims of the Requesters The Request contends that the Bank has failed to follow a number o f its operational policies and procedures inthe design and appraisal o f the Project, and that this will result in serious harm to the people living in the Project area and to the environment, in particular the Nile River and Lake Victoria, and to the customers o f the generated electricity and to Uganda citizens ingeneral. Hydrological and Climate Change Risks, Cumulative Impacts. The Requesters claim that the Project Social and Environmental Analysis (SEA) does not properly address hydrological changes and their effect on power production, nor the potential impacts o f climate change which they claim will lead to drier conditions, lower lake levels, and therefore lower power production. The Requesters contend that the SEA lacks an analysis o f the cumulative effects i s based on "jlawed assumptions and computations" related to hydrological risks, and does not adequately consider Project alternatives. The Request also asserts that the guarantee that the Kalagala Falls will be put aside as an offset and not be developed for hydropower i s not bindingon the GoU. Economic Analysis, Options and Affordability. The Requesters raise the concern that the electricity from the Bujagali hydropower plant will not be affordable, will not meet the needs o f the majority o f Ugandans, and will reduce the funding available for rural electrification. They state that there i s no evidence that a comprehensive economic analysis o f the Project has been done, and that alternative energy options have not been adequately studied to provide evidence that Bujagali dam i s the least-cost option. Disclosure and Consultation, Use of Data, and Dam Safety. The Requesters hrther claim that the Power Purchase Agreement (PPA) between the Project Sponsor and the GoU, crucial to the Project's economic viability, was only released shortly before they submitted their Request. They also allege that the public version in Kampala was not the actual version used to negotiate loans. The Requesters say that no evidence exists that the PPA was debated and approved by the Ugandan parliament. They also raise concerns about the Project consultation process and the use o f old and inconsistent data in key Project documents. The Request also claims that the Sponsor has failed.to adequately xvii address dam safety issues or determine whether Bujagali would be able to withstand a failure o fthe Nalubaale dam. Social and Cultural Issues. The Requesters claim that the Project did not recognize the presence o f indigenous peoples in the Project area nor deal adequately with cultural and spiritual issues o f the affected community. The Requesters say that the compensation and resettlement frameworks need to be updated to reflect the current economic situation and that the sponsor needs to create a detailed compensation and detailed community development actionplan. Management Response Inits response ofApril 5,2007, Management maintainsthat the proposedproject isbeing developed to provide the needed capacity in a least-cost manner. With respect to the previous Bujagali project and the Inspection Panel's 2002 investigation, Management notes that an action plan was prepared and approved by the Board on July 17, 2002. A matrix describing the 2002 Panel's investigation findings and the status o f implementation o fthe action plan was includedinthe Management Response. Related to Kalagala Falls, Management claims that the GoU has reiterated a commitment to the offset and that the Bank will include the commitment as part o f the indemnity agreement. Management also reports that a Dam Safety Panel was created to provide advice and ensure consistency with Bank policy and that the Project's legal agreements will require the preparation o f an Emergency Preparedness and Response Plan (EPRP), which includes failure scenarios for Nalubaale, Kiira and Bujagali dams. Management acknowledges that past resettlement was not completed. To address these issues, the Assessment o f Past Resettlement Activities and Action Plan (APRAP) and Community Development Action Plans (CDAP) were undertaken to assess the current conditions. BEL and the Bujagali Implementation Unit (BIU) are resolving outstanding issues. In response to the claims that the Basoga people in the project area should be considered indigenous people, Management asserts that the Basoga are not considered indigenous people under the Bank's definition because the Basoga do not meet criteria such as marginalization and vulnerability in addition to criteria on ancient origin, self definition and land. Management maintains that experienced Bank staff and consultants were engaged to work on the preparation o f this Project and that economic, financial, safeguard, technical and other analyses were done to a high standard. Project analyses considered a wide range o f electricity demand scenarios and the impacts o f both low and high hydrology scenarios. Management regards the environmental and social work carried out thus far to have appropriately considered the issues that emerged in the previous Bujagali investigation and the new issues outlined in the current request relate to resettlement, cumulative impacts, and consultations. xviii Management claims that the Project will bring benefits to many. According to Management, providing least-cost power i s expected to increase the number of connections o f residential users to the national grid, including in rural areas, and will allow industrial and commercial users to increase output and efficiency and, therefore, profits. Management also states that the Project will bring local job opportunities during constructionand following tourism development inthe Kalagalaoffset. The InvestigationReport and the Applicable Policies and Procedures This Report concludes the Panel's investigation into the matters alleged in the Request for Inspection. Panel Chairperson Werner Kiene served as the Lead Inspector for the Panel's investigation. The Panel was assisted in its investigation by expert consultants Prof. Theodore Downing, anthropologist, Prof. Richard Fuggle, environmental specialist, Mr. Graham Hadley, economic and commercial consultant, Prof. Peter Pearson, economist, and Prof. Carlos Tucci, hydrologist. The Requesters submitted their Request for Inspection to the World Bank Inspection Panel as well as the Compliance Review and Mediation Unit (CRMU) o f the African Development Bank (AfDB). The Panel and the CRMU coordinated their field investigations o f the Bujagali projects and shared consultants and technical information duringthis investigationinorder to enhance the efficiency and cost effectiveness o f each o f their investigations. While this collaboration between the Panel and the CRMU worked to the mutual benefit o f both parties, each Panel focused its compliance review on its own policies and procedures and each Panel has made its own independent judgments about the compliance o f its Management and staff with its respective policies and procedures. The Panel wishes to express its appreciation to the CRMU for this fruitful andprecedent-setting cooperation. The Panel reviewed relevant Project documents and other relevant materials provided by the Requesters, Bank Staff, Government officials, local authorities, individuals and communities living in the areas affected by the Project, as well as scholarly literature. The Panel organized a site visit in collaboration with CRMU o f the Af-DBinNovember- December 2007. Duringits mission, the Panel met with Requesters and other individuals and communities, local and national government authorities, representatives o f the Busoga Kingdom, spiritual and religious leaders, representatives o f civil society, and representatives of inter-governmental organizations, relevant experts and others. The Panel also interviewed Bank Staff in Washington, D.C. and Kampala. The Panel wishes to express its sincere gratitude and appreciation to all those with whom it met for their time and cooperation. With respect to this Project, the Panel assessed whether the Bank complied with the following applicable operational policies and procedures: OP 1.00 Poverty Reduction OP/BP 4.01 Environmental Assessment OP/BP 4.02 Environmental Action Plans xix OP/BP 4.04 Natural Habitats OP 4.07 Water Resource Management OP/BP 4.10 Indigenous Peoples OP/BP 4.1 1 Physical Cultural Resources OP/BP 4.12 Involuntary Resettlement OP/BP 4.37 Safety o f Dams OP/BP 7.50 Project on InternationalWaterways OP/BP 10.04 Economic Evaluations o f Investment Operations World Bank Policy on Disclosureo f Information 11. Context Electricity and Power Needs inUganda: The Power Supply Crisis Uganda i s experiencing serious power capacity constraints. Only five percent of the total population, less than one percent in rural areas, has access to grid-supplied electricity. The Panel notes the critical importance of providing affordable electricity to the people of Uganda, as an integral element of national development and of Uganda's poverty reduction efforts. Uganda's main source o f electricity currently i s the Nalubaale/Kiira dam complex, located just below the source o f the Nile River at Lake Victoria. The complex consists o f two dams, Nalubaale and Kiira. Over recent years, however, the electricity produced by Nalubaale and Kiira has dropped substantially below capacity due to the hydrological limitations on the release of water into the Nile from Lake Victoria, and the interactions between the dams and the water levels o f the Lake. Two additional large hydropower projects are being proposed and/or developed along the Nile in Uganda: the Bujagali Hydropower Plant and the Karuma dam, downstream from Bujagali Falls. A critical issue raised by the Request is whether the Bujagali dam, if built, will meet its economic projections and provide affordable electricity to the people of the country, in comparison to other alternative means for doing so. Duringits visits to the Project area, the Panel heard strong expressions of concern from local people and their representatives that they will not benefit from the Project but will, nevertheless, have to bear its social, economic and environmental costs. In addition, they are concerned that, if Project costs are not properly estimated and accounted, the burden of below-capacity production will be passedto the people of Uganda. Environmental and Social Setting The Project area i s home to several ethnic groups living in and around the Project site, including the Busoga and Busanga people whose lives and livelihoods will be affected. The Bujagali damwould create a reservoir that floods an area o f 388 hectares, require the taking o f238 hectares o f land and would require additional takings for transmission lines, all o f which involve displacement and resettlement o fpeople and family from their lands. xx The dam's reservoir would inundate Bujagali Falls and other natural habitats, which are sites o f cultural andreligious significance to surroundingpeoples. Prior to the construction of the existing dams on the Victoria Nile, the amount o f water flowing from Lake Victoria was naturally determined by the level o f water inthe Lake- the higher the level o f the lake, the more water poured out from the Lake into the river. However, the successive development o f the Nalubaale and Kiira dams at the entry point from the Lake to the upper Nile changed all that. At the time that Nalubaale dam started operating, it was agreed that it must be operated in accordance with the Agreed Curve (a mathematical relationship between Lake levels and outflow) that stipulated how much water should be released from the Lake. The Agreed Curve aims to ensure that the outflow from the lake mimics the natural conditions o f the Lake before the Nalubaale dam (formerly Owen Falls dam) was constructed. A significant question raised by the Request is the extent to which the proposed Bujagali Damwill or might create incentives to depart from the Agreed Curve, and contribute to a lowering o f Lake water levels and corresponding serious impacts for the Lake's riparian states. An important related question i s the extent to which the future hydrology of Lake Victoria may be influenced by climate change. Since the Lake's water balance i s dominated by rainfall and evaporation over the surface o f the Lake, the Requesters are concerned that even relatively small long-term decreases in rainfall and/or increases in temperature could have significant impacts on Lake levels and on outflows via the Victoria Nile and, inturn, on the economic and politics o f operating the dams. Another important element of the Request i s the potential impacts o f the Project on the economy o f the area around the waterfalls, including through fishing and tourism. In addition, the Project i s being proposed for an area long inhabited with strong, complex cultural and spiritual tradition. Although the peoples o f other ethnic groups inhabit the project area, the Basoga claim spiritual dominion o f both sides o f the Nile, its islands, the water and its waterfalls. The Bujagali Falls dam would inundate places o f high cultural and spiritual significance to localpeople 111. EnvironmentalIssues (1) Adequacyof the SEA Inthe Requesters' opinion, the social and environmental studies supporting the Project are generally inadequate, based on old data that do not reflect the current situation o f the Project area, in violation o f the Bank Policy on Environmental Assessment OP/BP 4.01. Management responds that the Project i s a new operation and, as a result, social and environmental aspects have been reassessed. It adds that the current Project was designed to buildupon earlier data and additional studies were undertaken as needed, to confirm or update that baseline. The Panel notes that the Project Sponsor contracted international consultants to prepare therequired SEA for the Bujagali Hydropower and Bujagali InterconnectionProject with xxi appropriate input from Bank Management. The Project has appropriately been classified as category "A", the category under Bank policy used for projects with the most serious levelof impacts. This complies with OP 4.01. The Panel finds that, apart from the omission o f an Environmental Management Plan (EMP), the SEA includes the elements required by Annex B o f OP 4.01. The Project i s fully described and set in an appropriate policy, legal and administrative framework. However, the fact that the EMP is not an integral part of the SEA that has been disclosedis a deficiency. This is not incompliance with the requirements of OP 4.01. OP 4.01 also requires that when there i s inadequate legal or technical capacity to carry out key EA-related functions (e.g., review o f EA, environmental monitoring, inspections, or management o f mitigatory measures), the Project includes components to strengthen that capacity. This requirement to support needed capacity building, which is important in the implementation of the social and environmental aspects, has not been complied with inthis Project The Panel also finds that an independent panel o f internationally recognized environmental specialists has not been appointed for the Project (or a single panel to cover both the Hydropower and the Interconnection projects). As the Project i s contentious and involves complex multidimensional environmental concerns, appointment of an environmental panel of international experts i s warranted and the lack of such a panel is not incompliance with OP 4.01. (2) Disclosure of Project Documentation The Requesters contend that the Project SEA does not address significant issues relating to hydrology, climate change and cumulative impacts. The Management Response cites the Strategic/Sectoral Social and Environmental Assessment (SSEA) o f the separate Nile Basin Initiative as the source o f data and analysis on issues of climate change and cumulative effects o fthe Project. The Panel notes that the Bujagali SEA makes only a passing reference to the SSEA, and the SSEA makes no mention o f the Bujagali SEA. It is clear from reading the two reports, and the lack of cross-references between them, that they do not form part of the same suite of documents. The Panel i s o f the view that, inthe interests o f efficiency, an EA may, in principle, refer to and/or incorporate, as appropriate, other relevant studies. However, as the purpose o f both the sectoral and project specific EA is to disclose information relevant to a decision, the fact that one study i s reliant on another must be clearly stated and disclosed inproject documentation. Without this, information important to a project i s obscured even ifit i s disclosed independently, which weakens or undercuts the achievement o f the key elements o f OP 4.01 on informed decision-making, public consultation and disclosure. The Panel findsjustifiable the Requesters complaint that some aspects o f the Project, i.e. effects o f climate change and the cumulative effects, have not been properly addressed in xxii the Project SEA. The Panel acknowledges that the necessary studies have been conducted and disclosed, albeit independently, and considered by Management and referred to specifically in the PAD. However, the failure to disclose the SSEA or its relevant parts as an integral part of the Bujagali Hydropower Project's documentationina timely manner is not consistentwith OP 4.01. (3) Cumulative Impacts Inthe Requesters' opinion, the SEA does not discuss cumulative impacts, and BEL did not attempt to identify issues arising from building a cascade o f dams on the River Nile, especially with respect to the health o f the Lake Victoria. Management argues that cumulative impacts o fthe currently proposed Bujagali project are addressed as part o fthe Project's SEA and inthe SSEA. The Bank's OP.4.01 Annex A states that a "[s]ectoral EA pays particular attention to potential cumulative impacts of multiple activities." The analyses in the SSEA allow a comparison amongst the various proposed portfolios o fpower development options inthe Nile Equatorial Lakes Region. They do not, however, provide a systematic examination of the potential consequences of the Nalubaale and Kiira facilities, the Bujagali Project, and the planned Karuma project all being situated on the Victoria Nile between Lake Victoria and Lake Kyoga. In addition, there i s no examination o f the impact o f additional transmission lines between the hydropower stations and Kampala. Although section 14 of the SSEA is headed "Assessment of Cumulative Impact" the Panel finds that the analyses are not sufficiently backed by evidence and include opinions rather than careful fact-based examinations of the additive effects of impacts from present and foreseeable projects. The SEA seems to address cumulative effects in more detail. However, it makes statements that are not substantiated by data or factual analysis. There is no determinations o fhow many people stand to be affected, how much agricultural land i s to be lost, the extent to which riverine forest habitat will be lost, or the extent to which tourism will be affected. The Panel finds that neither the SSEA nor the SEA has addressed the cumulative effects of the existing and planned projects in a meaningfulway. This is not incompliance with OP 4.01. (4) Transmission Lines The transmission lines that will transport electricity from the hydropower site pass through areas where people live, wetlands, and the ecologically important Mabira Forest. As noted above, the SEA fails to address the cumulative effects o f transmission lines; neither does it propose mitigation to reduce additive effects. The Panel was not furnished with documentation indicating that the Project considered ways to mitigate or reduce the amount o f land taken for the second (Bujagali) transmission line. Rather, the Project assumed that the size o f the existing right o f way needed to be doubled, which is technically incorrect. The Panel finds that the failure to xxiii consider mitigation measures, which would reduce the social and environmental impacts of the transmission line, does not comply with OP 4.01 and OP 4.12. (5) Environmental Impacts on Fisheries and Aquatic Systems The Requesters express concern as to the accuracy o f the surveys o f endemic fish species and claim that the data on which the EA i s based i s flawed and outdated, in non compliance with OP 4.01 on Environmental Assessment and OP 4.04 Natural Habitats. Management responds that the Project builds on relevant work conducted for the prior Bujagali Project and on updated information gathered in further field studies and analysis, including studies on fisheries conducted for the prior project and updated for the current Bujagali Project. Based on its review of relevant research studies, the Panel observes that the status of the fish species inhabiting both Lake Victoria and the Victoria Nile is disputed and that ongoing research is desirable. However, significant effort has been devoted to study these fish in the reaches of the Victoria Nile that will be affected by the Bujagali Hydropower Project. Studies undertaken by the Ugandan National Fisheries Resources Research Institute (NAFIRRI, previously known as FIRRI) show that fish ladders suggested by the Requesters would not be scientifically justifiable because a barrier in the upper reaches up to Dumbbell Island would not significantly affect the stability o f fish populations in Lake Victoria and neither would a fish ladder be relevant. The studies undertaken by, and the formal indicative position of, NAFIRRI are persuasive and the conclusions logically drawn. Bank Management exercised appropriate diligence inusingthese documents inits decision-making. The Panel consequently finds Bank Management acted consistently with the provisions of OP 4.01 and OP 4.04 in so far as these relate to assessment of the likely consequences of the Bujagali Hydropower Project on fish stocks in the Upper Victoria Nile and Lake Victoria. (6) Mitigation Measures: The Kalagala Offset Agreement Mitigation measures for the Project call for the Kalagala Falls to be established as an appropriate offset for the natural habitats that would be inundatedby the Bujagali project. Kalagala Falls originally had been identified as the site o f a potential future hydropower project. In2006, however, the Government stated that: "The Government position on the site is that it continues to be frozen for development purposes." This offset is now provided for inthe IndemnityAgreement betweenthe Bank and the GoU. The Requesters express concerns about the agreement between the World Bank and the GoU because, in their opinion, this agreement i s not a guarantee that Kalagala Falls will never be developed for hydropower. At the time o f its Response, Management claimed that the offset provision related to the Kalagala Falls to be included in the Indemnity Agreement ". .. will be binding throughout the life of the Indemnity." xxiv The Panel found that that the "Kalagala Offset" has come to be accepted as a site to be used to "offset" a variety o f the features that are to be lost by inundating the Bujagali rapids, but there is almost no mention o f the core purpose o f a conservation strategy for lost natural habitats as provided by Bank policy on Natural Habitats. During its investigation visit, the Panel observed uses at Kalagala Falls that are not necessarily consistent with this conservation purpose. The World Bank stance has been nevertheless clear in the sense that : "...the long term protection of the Kalagala Falls and the preclusion of development of hydropower potential at Kalagala is a necessary offset for World Bank Groupparticipation in theproposedproject." The Panel wishes to note and highlight the Bank's efforts in cooperation with the government to develop the commitment to set-aside and protect Kalagala Falls as an offset to impacts produced by the dam. Although certain important issues in this regard are noted in the Report, the Panel notes and appreciates that the action to develop and strengthen this commitment in light o f issues raised inthe Panel's previous investigation report andrelevant Bank policy. The Panel finds that there is evidence that an offset has been created to meet the requirement of OP 4.04. On the other hand, the Panel finds that there is evidence that the offset site i s not being subject to appropriate conservation and mitigation measures inconformity with sound social and environmental standards. The Project i s thus not in compliance with OP 4.04 on this point. Given present institutional weaknesses and lack o f proper training arrangements, the Panel finds that the capacity o f local institutions to plan and manage the Kalagala offset has not been developed and that no provision has been made to rectify this. As a consequence the Kalagala offset may not achieve the purpose for which it was set aside, and this is not consistentwith the provisions of OP 4.04. The Panel notes with concern that the proposed Environmental Mitigation and Monitoring Plan i s silent on the need for monitoring of enhancement and offset plantings. Also, monitoring of replacement plantings has not been included in the terms of reference of the witness NGO that has been appointed to monitor Project compliance with IDA conditionalities. This is not consistent with the provisions of OP4.04. (7) Safety of Dams The Request claims that safety issues possibly emanating from the existing Nalubaale dam at the Owen Falls are not taken into consideration in the Bujagali dam design. Management responds that a Dam Safety Panel has been established to provide advice through design, construction, filling, and start-up to ensure that the project i s consistent with Bankpolicies. The Panel visited the Nalubaale complex inDecember 2007 and was shown the cracks in the powerhouse as well as the routine measurements o f structural movement and o f pore- xxv water pressure that are undertaken and reported. The Panel i s satisfied that Eskom (Uganda) i s undertaking and reporting the monitoring o f the Nalubaale complex that the Bankrequires. The Panel notes that the cracks are inthe powerhouse structure and not in the wall o f the dam. The Panel finds that Management has complied with the procedures set forth inOP 4.37. IV. Hydrological and Climate Change Risks According to the Requesters, BEL'S SEA does not adequately address the issues o f possible hydrological changes affecting power production at the Nalubaale, Kiira and the proposed Bujagali facilities, especially at a time when Lake Victoria water levels are declining. Inits Response, Management states that the impact o f hydrological flow rates on the planned Bujagali dam has been addressed extensively in the SEA and in the Project Economic Study. According to these studies, "the proposed 250MWproject is not expected to significantly alter or affect the hydrologv of Lake Victoria or the Victoria Nile." Analyses used to assess the hydrology o f the Lake comprise 106 years o f data, including several hydrological cycles. Management acknowledges that in recent years the Government has over-abstracted water for power generation because o f a general drought, lack o f generation investments to use available water more efficiently, and a demand growth o f eight percent. However, Management also states that the "GoU has steadily decreased hydropower generation in an effort to return to theAgreed Cuwe operating regime." (1) Appropriateness of HydrologicalData Series used inProject Design Observers generally divide the history o f Lake Victoria's water levels into three main periods. In general, the period before 1960 i s characterized as a period o f relatively low water levels. Between 1960/61 and 1999, Lake Victoria level rose, while starting in2000 and until very recently, lake levels decreased to a level observed before the 1960s. The Project Economic Study concluded that the whole period o f record from 1900 should be used to determine the future dependable flow for power generation at hydro power stations on the Victoria Nile. The Panel's hydrology expert concluded that the hydrologic data sets used in Project design constitute a reliable data series and its variability over time is a natural condition, which can be observed in other hydrologic series of different parts of the world, when the hydrologic series is long enough. The Panel finds that this provides an appropriate baseline for analysis of environmental and economic issues, incompliance with OP 4.01. (2) Lake Victoria Water Levels and Power Plant Operations on the Victoria Nile The Agreed Curve has been used to specify the outflow that should be released from Lake Victoria down the Victoria Nile following the construction o f Nalubaale dam. After 2000, the entry into operation o f Kiira dam (in a side-channel constructed parallel to xxvi Nalubaale dam) increased generation capacity. Since these two dams operate in parallel, the system required more water to flow downstream and through the turbines to generate energy. Inthe period 2001-2005, the increase o f the water release o f the Lake above the Agreed Curve has resulted in increased energy production downstream but hadnegative upstream effects o f lake depletion and resultingimpacts. The Panel notes that the Agreed Curve constrains the ability to use the lake to store "excess" water for later use when inflow exceeds outflow. During its field visit in December 2007, the Panel was given documentation showing what appears to be a new release policy - the "Constant Release" rule. The Panel received information suggesting that this new rule, which allows for a constant release to be applied when the lake level fluctuates within a certain range, has been ineffect since June 2006, and it i s the basis for the analysis inthe Economic Study. (3) Impact of Hydrologic Risk on Energy Output The Requesters and Management Response include contrasting statements as to whether the Economic Study adequately addresses the Project's economic viability in relation to hydrological risks. The PAD states that the assessment o f the energy output was based on the flow released from Lake Victoria through the Nalubaale/Kiira dam complex in accordance with the Agreed Curve, and that the Project i s designed in accordance with the Agreed Curve release rule. The Economic Study states that it adopted the "Constant Release" rule to determine the energy generation capability o f the hydro options considered. This discrepancy between key Project documents brings into question the data basis for the Project's economic analyses, and i s likely to have resulted in a more positive conclusion to the Economic Study than would have been the case under the Agreed Curve scenario. This is inconsistent with OP 10.04. In March 2007 an internal Management Review had proposed that the PAD should confirm that the plant would be operated under Lake Victoria's Agreed Curve release strategy, rather than under a constant release regime, "and should conJrm that this regime does not affect the conclusions of the economic evaluation of theproject.. .." The PAD does not appear to have followed this latter recommendation. In the Panel's view, the provisions of OP 10.04 require Management to provide an accurate picture of the Economic Study (based on the Agreed Curve), and indicate whether this affectsthe relevant conclusions. The Panel notes that this contradiction in Project documents has a material implication not only for the economic viability of the Project and the provisions of OP 10.04, but also on the lake levels of Lake Victoria, since different operational rules result in different time-profdes and variance of water levels. While the Panel recognizes that, over a certain period o f time, the mean outflow under the "Constant Release" rule may be identical to that under the "Agreed Curve" rule, the time-profile and variance inlake levels under the two regimes will be different. xxvii (4) Potential Impact of the Project on Lake Victoria The Requesters are concerned about over-draining o f Lake Victoria, and believe that the issue o f the long-term health o f the Lake has not been addressed in Project documents. Management indicates that the operation o f the Bujagali/Kiira/Nalubaale system i s not expected to affect the hydrology of Lake Victoria because the water released from Lake Victoria and the timing o f these releases will still be controlled by the operation o f the Nalubaale and Kiira dams. Bujagali will use the same water already utilized by the parallelupstream dams ofNalubaale and Kiira. The Panel notes that the SEA study was based on the assumption that the Project's upstreamarea o f influence ends downstream o f Kiira-Nalubaale dams. The SEA did not take into account the potential impacts o f the Project on Lake Victoria. The SEA expected and the PAD stated that the Project would be operated in accordance with the Agreed Curve. The Panel notes that this approach does not take into account the contradiction between the PAD and the Economic Study regarding the Project's operation rule and the recent history o f 2003-2005 when the Nalubaale-Kiira system was operated above the Agreed Curve, which contributed to a severe depletion ofthe Lake. The Panel also notes that the operation policy o f Lake Victoria could be other than the Agreed Curve, using the lake as reservoir regulating the flow. However, the Panel observes that such change in operating regime and its impact upstream and downstream need to have been assessed inthe Project environmental assessment. The Panel notes the importance of assessing such a situation and extending the area of influence of the Project to Lake Victoria. This i s also important because the lowering of water levels in Lake Victoria brings significant social and environmental impacts upon the Lake ecology and the people and countries that rely on it for resources and livelihoods. The Panel notes that the SEA study considered that the Project's area of influence ends downstream of the Kiira-Nalubaale dams. As a result, the Panel finds that the SEA analysis did not comply with OP 4.01 in defining the area of influence of the Project because the Project impacts on the changing levels of Lake Victoria were not assessed. In light of its relevance to the analysis of the Bujagali Project, the Panel notes the importance of making the structure for governance of water releases from Lake Victoria clear and transparent to all stakeholders. (5) Climate Change Risks The Requesters aver that the project preparation and assessment reports do not address climate change and its possible impact on power production at Bujagali. Management, on the other hand, claims that climate change aspects were addressed in different studies, such as the SSEA, which includes a detailed analysis o f the impacts o f climate change in the Nile Equatorial region comprising Bujagali. xxviii The PAD states that both the Economic Study and the SSEA conclude that there will be "no adverse effect on water release due to climate change during the life of theproposed project." The Economic Study states inthe main text that the influence o f climate change was not found to be significant enough in the medium term. Further brief discussion i s included inAppendix B o f the Economic Study. In the Panel's view, the brevity of this discussion of a highly complex issue with the potential to influence significantly the Project's economic outcomes does not demonstrate compliance with OP 10.04's paragraph 5, which requires proper assessment of the robustness of the Project with respect to environmental risks. In contrast, the SSEA assesses potential impacts on hydroelectric generation and examines whether such impacts might affect new power options being evaluated. An independent review o f the hydrology o f Lake Victoria, financed under the Bank- Netherland Water Partnership (BNWP) which also peer-reviewed the Project Economic Studywith respect to hydrological risk, statesthat "there is considerable variability in the results of the individual models and caution should be used when applying these results to make operational decisions." The SSEA appraisal appears to be the result of a thorough, detailed study that draws on its own analysis and a range of other internationalstudies. The Panel finds that the possible effect of climate change on hydropower projects on the Victoria Nile has been seriously considered in the SSEA. This analysis meets the requirements of OP 4.01. As noted above, however, the SSEA was not properly disclosed as a Project document. It is important to note that the results of the SSEA analysis show that there are few identifiable hydrological risks to the hydro-power options studied, and overall for the Northern and Central West regions o f the Nile Equatorial Lakes there i s a higher probability o f increases in runoff, and thus power generation, than determined from historic flow data. The Panel notes, however, that the Economic Studydoes not cite or draw on the results o f the SSEA. Management does not appear to have ensured that the Economic Study drew on the much more thorough analysis in the SSEA. The Panel finds that this does not comply with paragraph 5 of OP 10.04. Considering that the PAD draws on the authority of both studies, particularly the SSEA, the Panel finds it surprising that the PAD concludes that, "[. ..] therewill be rzo adverse effect on water release due to climate change during the life of theproposedproject." The Panel is aware of the limitation of the known technology in evaluating climate change scenarios and that the analysis o f climate change i s an evolving science, where gaps remain. Indeed, this situation makes all the more troubling the PAD'S categorical assertion, without any reference to risk and uncertainty, that there will be no adverse effect on water release due to climate change during the Project life. This failure to express a riskfactor is not consistentwith OP 10.04. The Panel notes the importance of continued attention and analysis to the effect of climate change on flows and hydropower generation on the Victoria Nile. xxix V. Economic andEnvironmentalAnalysis of Alternatives The Requesters argue that energy alternatives to Bujagali were not adequately addressed inthe SEA and that the Economic Study does not include an adequate assessmentofthe economic alternatives to support the statement that the Bujagali dam is the least costly option. Management believes that the economic, financial, safeguard, technical, governance, and other required analyses meet high professional standards and are in compliance with applicable Bank policies. Management adds that these analyses take into account the findings o f the previous Bujagali InspectionPanel Report and result from the overall project due diligence, which adequately takes into consideration best practice. Bank Economic Evaluation policies applicable to this Project are OP/BP 10.04 on Economic Evaluation o f Investment Operations and OP 1.00 on Poverty Reduction. OP 10.04 provides that "For evey investment project, Bank staff conduct economic analysis to determine whether the project creates more net benefits to the economy than other mutually exclusive optionsfor the use of the resources in question." The Policy then sets out specific provisions in seven areas: criterion for acceptability, alternatives, non- monetarybenefits, sustainability, risks, poverty and externalities. (1) Demand Forecasts and Electricity Tariffs The forecasting o f demand and its interaction with likely tariffs i s a necessary element in the process o f analysing project alternatives. Thus, the analysis o f the future "expansion path" o f an electric power system should explore both the likely evolution o f the demand on the system and the least cost means o f satisfying that demand through existing plant andnew investments. The Requesters argue that the demand forecast analysis for the Project i s unrealistic, as only a small part o f the population o f Uganda can afford electricity that i s unsubsidized. Management notes that the risks related to future uncertainties o f variables have been evaluated. The Inspection Panel Report on the first Bujagali project criticized aspects o f the load forecasts used for that project, including the assumption o f narrow ranges. In the Panel's judgment, Management addressed demand forecasting for the current Project seriously, in that it commissioned a detailed, sophisticated review in 2004, which stressedthe importance of a thorough revision of the load forecasts. The forecasts for the current project show a much broader range between the high and the low cases. This reflects inparticular significant variations around the base assumptions about residential connections and the rates o f growth in household income and commercial and industrial GDP. All other assumptions remain the same as for the base forecast, however. The economic study assumes that connections o f new consumers will rise significantly in one year after Bujagali's commissioning. The Panel notes that although the availability of reliable electricity supply at the time the Bujagali plant is commissioned might reasonably be expected to stimulate new connections, the xxx Economic Study appears to assume a more sudden increase in connections than seems likely to occur. A more gradually phasedtrajectory of connections to the grid after 2011 would seem more plausible, both for the base forecast and the low and high variants. Given the difficulties inherent in reducing commercial and technical losses in the electricity system in Uganda, and in particular in light o f the challenges recently experienced by the electricity distribution company (UMEME), the Panel finds that the demand forecast should have varied the assumptions on losses and the collection ratio (i.e. the ratio between UMEME's billed sales collected and billed sales) as part o f the sensitivity analysis and o f a more complete appraisal o f risks, in conformity with OP 10.04. Indeed, somewhat lower values might also have been appropriate for the base forecast, as an alternative to assuming that the targets set for the electricity distribution concession would be fully achieved. (2) EconomicAnalysis: Alternatives Considered The PAD states that major generation alternatives to Bujagali considered inthe Economic Study include: small and medium-sized hydropower projects, large hydropower projects studied beyond the feasibility stage (i.e. Karuma), thermal options, bagasse based cogeneration and geothermal. (a) The Geothermal Potential The Request claims that Ugandapotential for geothermal energy i s up to 450MW but that hydropower generation studies took precedence over thermal energy. Management Response states that a "detailed review of geothermal prospects was conducted aspart of the project analysis of alternatives." The analysis concluded that only 10 percent o f the potential 450MW claimed by the Requesters i s feasible and a geothermal 40MW plant was assessed in the least-cost analysis. The Economic Study reaches its conclusion by questioning existing estimates o f temperature for Katwe and Buranga contained in a 2005 paper, whose authors' affiliations include Uganda's Department o f Geological Survey and Mines. The Panel notes the statement in the Management Response that additional studies and shallow drilling are included under the ongoing Power IV Project, to assist the Government in assessing geothermal prospects at several sites in Western Uganda. The additional information resulting from this work would help resolve conflicting views regarding geothermal potential in Uganda, and may have a significant bearing on the economic analysis of alternatives. (b) Small and Medium Scale Alternatives Inthe Requesters' view, only a limited energy potential at various hydropower sites has been developed. Management responds that the Bank i s supporting development o f mini- hydro potential and states that projects providing power to the grid or suitable for grid xxxi connection were considered in the economic study. However, the TORfor the Economic Study encouraged a focus on relatively large grid-connected plants and did not draw attention to the evaluation o f smaller scale or off-grid alternatives. The Economic Study does not discuss any other renewable sources o f electricity, such as municipal solid waste, solar or wind. In a country where only 5 percent of the population i s connected to the grid and there i s widespread poverty, it would be reasonable to expect attention to be paid to small and/or distributed generation options (not only hydro) which might intheory more directly address local and rural poverty. It i s noteworthy that the Management Response to the Request contains a much fuller discussion and appraisal o f the smaller scale and/or distributed generation options than was contained in the Economic Study and the PAD. The Panel notes that the information in the Economic Study and the PAD relating to knowledge about and the potential of smaller scale and/or distributed generation alternatives did not clearly establish that the available studies and data had been identified and evaluated in a way that would have enabled decision-makers to decide whether further consideration was required. The Panel finds that the Economic Study and the PAD did not demonstrate full compliance with OP 10.04's requirement in paragraph 3 to evaluate alternatives. (c) OilResources In January 2006, an oil company announced "the existence of a working petroleum system in the Albertine Basin," while warning that it was too early to determine its size or potential commerciality. Other reports convey more scepticism about the scale of the discoveries. While the oil resource discovery was at a very early and unproven stage at the time when the Economic Study Final report was completed (February 2007), the Panel finds that the existence and potential of this resource should have been reviewedinthe discussionof alternative supply options. (3) Project Costs (a) Bujagali Project Costs The PAD acknowledges that by the time of its publication, estimates of Bujagali's Engineering Procurement and Construction (EPC) costs were substantially higher than those for the prior Bujagali project. The Panel notes that power plant costs have increased in real terms internationally. Nevertheless, because EPC costs form a key element in determining the Project's economic and financial viability, the Economic Study and the PAD should have supplied fuller explanations of the details of this cost increase, supported by appropriate analysis and quantitative evidence. Inadditionto the cost increase noted above, there is evidence o f significant cost increase duringand after the appraisal process for the current Project. Management responded to a question from the Panel about differences between the cost estimates used for the Economic Study and the PAD, stating that: "it .was not practical to consider restarting xxxii this analysis when each newhefined estimate of project costs became available, since the new estimates were such that all parties involved in the study considered that they would not [to] alter the conclusions of the study." The Panel finds that, although certain parts of the analysis were carried out thoroughly, to meet all of the requirements of OP 10.04, the PAD should have included an explanation and supporting evidence of why the substantial project cost variations would not alter the conclusions of the Economic Study. (b) Karuma Project Costs The Requesters claimed that the Karuma construction costs were inflated to gear the analysis o f alternatives infavour o f Bujagali. Management stated that the analysis has in fact shownthat Bujagali has lower constructioncosts than Karuma. Project documents estimate costs for Karuma and Bujagali as o f 2001 and as o f 2007. A comparison o f the rate o f increase in the EPC cost estimates during this period suggest that Karuma's EPC cost estimates grew by a smaller percentage than those o f Bujagali. Therefore, the Panel observes that the updating of the EPC cost figures in the PAD does not obviouslydisadvantageKaruma relative to Bujagali. At the same time, the Panel found conflicting and incomplete reports on cost estimates for Karuma at the time o f the prior project. Thus, the Panel could not fully assess these estimates. (4) Assessment of Least Cost Options for Expanding Power Generation and Supply The Economic Study devised and compared alternative generation expansion plans with andwithout Bujagali as a candidate plant. The process of testing the sensitivity of the least cost expansion plans with and without Bujagali appears to have been carried out thoroughly. The assumed increase of 10 percent for the "high Bujagali capital cost scenario" compared with the "base scenario", with an assigned probability of only 20 percent, was inappropriately low. Nevertheless, a sensitivity test suggested that the Economic Study's conclusions that Bujagali was the least-cost option were robust for an increaseof almost 50 percent in capital costs. The PAD states that the Economic Study for the power system as a whole suggestedthat, when compared with the assumed tariff underlying the demand forecast, "[. .] the tariff . may drop by up to 10% in real terms after the commissioning of the proposedproject." The Economic Study suggests that fi-om 2011 the average long term cost o f supply, 16 cKWh, i s 1.2 c K W h lower than the assumed constant tariff level (a 7 percent difference). The PAD'S statement simply asserts that the Economic Study shows that the tariff may drop by up to 10 percent, without qualifying the statement in light of the increases in EPC and transmission costs after the Economic Study was prepared and that were recorded in the xxxiii PAD. The issue o f electricity tariffs and affordability is o f high importance to the people and communities. The Panel finds that, in order to comply with the requirementsof OP 10.04, the PADshould havequalifiedits statement about the projecteddrop intariffs to takeintoaccountthe impactofEPCandtransmissioncostincreases. The PAD presents its own estimates o f the projected levels o f the weighted average retail tariff path, based on a financial analysis that i s different, and presumably later, from that o f the Economic Study. The PAD does not compare these figures with those in the Economic Study; neither does it explain why they differ. It also does not comment on any implications. The Panel notes that the Project's impact on tariffs and their affordability was known to be a key concern. In this light, the Panel considers that the relationship between the estimates in the Economic Study and those from the PAD'Sfinancial analysisshould havebeenpresentedmoreclearly andtransparentlyinthe PAD. (5) Externalities Paragraph 8 o f OP 10.04 requires the economic evaluation to take into account domestic and cross-border externalities, which are in large part environmental. The Economic Study states that a field missionto UgandainJuly 2006 was carried out to collect data on the environmental and social costs o f the Bujagali and Kanuna projects. It adds that the Economic Study for Bujagali also used data gathered inthe preparation o f the SEA. The Panel finds that the limited presentation and discussion of these costs in the EconomicStudy did notsucceedin demonstratingfull compliancewith OP 10.04. In the Panel's view, to meet all the requirements of Paragraph 8 of OP 10.04, the Economic Study should have examined, in more detail, the potential of changes in damage from pollutants other than C02, such as sulphur and nitrogen oxides, particulatesandnoise,even ifit mighthaveproveddifficult to value them. (6) EnvironmentalAnalysisofAlternatives The discussion below reviews the analysis o f alternatives to the proposed Bujagali hydropower facility by looking into the evaluation o f these options with a focus on environmental and social considerations and the decision-making process that led to the selection o fthe Project design. (a) HydropowerinComparisonto other Technologies Within the Region An analysis ofpower options within the Nile EquatorialLakes Regions, funded under the Nile Basin Initiative and done as part of the SSEA, indicated that among four options considered for Ugandan base-load supply, the most appropriate was large-scale hydropower. This conclusion put a focus on large-scale options in the analysis of alternatives, both within Ugandaand at Bujagali Falls inparticular. (b) Hydro-powerLocationAlternativeswithinUganda xxxiv Twelve alternatives at seven different sites in Uganda were considered for large-scale hydroelectric projects inthe Nile Equatorial Lakes Region study. Only two alternatives- Karuma and Bujagali-were found to be both cost-effective and socially and environmentally acceptable. The Panel finds that Management did not ensure that cultural and spiritual matters were properly considered when comparing the Bujagali and Karuma alternatives, as required by OP 4.01. This i s especially relevant in light of the significant cultural and spiritual importance of the Bujagali Falls to the Busoga people. The lack of proper consideration of cultural and spiritual matters in this comparison had important consequences, inthat it appears to have led to the conclusion that there was little difference between the Bujagali and Karuma sites and that therefore the economic and financial aspects of the options should become the determining factor in selectingthe preferred option. (c) Alternative Project Configurations at Bujagali The SEA undertaken for the prior Bujagali project included an analysis o f alternative impoundments to utilize the drop provided by the falls at Kyabinva, Bujagali, Buyala and Busowoko. For each alternative the power that could be generated, costs, and both socio- economic and environmental impacts were evaluated. This analysis was revisited for the SEA in 2006 for the second Bujagali project. The SEA for both the prior and current Bujagali project conclude that the optimal least-cost option for generating large-scale hydro-power at the Bujagali site, without major socio-economic or environmental consequences, would be to construct a 30m highdam across Dumbbell Island. The Panel notes that a range of alternatives have been considered in these studies. The Panel is concerned, however, that the analysis unduly narrowed its consideration of alternativeson the basis of apriori judgments rather than exploring all technically feasible options, including those that would not involve flooding the Bujagali falls and thus have lower social and environmental costs, and laying them out ina systematic way along with their economic, social and environmental benefits and costs, so that judgments on optimal alternatives could be made with a full understanding of the trade-offs involved. This is not consistent with OP 4.01's provisions that feasible alternatives should be explored systematically to meet the basic Project objectives, and may have led to inadequate consideration of alternatives that met Project objectives while avoiding the social and environmental costs associatedwith flooding the Bujagali Falls. VI. Economic Evaluation: Poverty Reduction andRisks The Requesters believe that the Project i s economically risky, a risk that has been worsened by changing hydrology. They are concerned that because o f the cost increase in the Project, the majority o f Ugandans will not be able to afford unsubsidized electricity from the Bujagali dam and, as a result, the Project will undermine Uganda's efforts for poverty eradication. Inaddition, the Requesters indicatethat the population livinginrural areas i s far from the national grid and will not benefit from the Project. Management claims that the Project i s expected to "have positive impacts on poverty alleviation in Uganda" directly through the availability o f power and indirectly through employment xxxv creation. It adds that the Project will help Uganda to continue its broad based growth in support o fpoverty reduction. (1) Affordability and Poverty Reduction The Economic Studyprovides quantitative assessments o fboth costs and benefits, which suggest that the Project would have largely positive direct impacts on Uganda's economy and enhance national economic activity. In this sense, and bearing in mind the reservations about the cost estimates of the Economic Study, from a macroeconomic perspective, the analysis appears to have complied with the requirement inOP 1.00 to show that the Project is likely to contribute to "broad based growth." Interms of the affordability of electricity generated under the Project for the people of Uganda, Management Response acknowledges that "end-user tariffs in Uganda almost doubled in 2006" and that the "increased price still does not fully cover the cost of generation, transmission and distribution, estimated at US&Z.Vkwh, requiring government subsidiesfor the difference." Still, Management claims that, with the Project, the cost ofpower would fall to US$16/kWh in2006 money. The Panel notes, however, that the US$16/kWh figure provided in the Economic Study is likely to be an underestimate of the cost of electricity with the Project. As explained in this Report, the Bujagali Engineering-Procurement-Construction (EPC) costs used in the Economic Study were nearly a fifth below the EPC values cited in the PAD. Further, the transmission cost estimates used inthe Economic Study were low. The Management Response does not mention these differences in cost estimates or make clear their implications for the tariff estimates of the Economic Study, on which the estimate of US$l6/kWh and Management's above statement about improved affordability are based. Much o f the expected direct benefit from Bujagali especially in the early years, i s likely to be experienced by the better-off urban households and particularly the industrial and to a lesser extent the commercial sectors and their stakeholders. As noted previously, the Project supplies to the grid but only five percent o f the population, and less than one percent in rural areas, i s connected. Existing poorer households that could afford to connect would benefit from the delivery o f a more reliable and possibly relatively cheaper service. Nevertheless, the electricity would still be very costly for poorer households and too costly for many. The TORfor the Economic Study discuss the calculation o f the ERR for Bujagali, outline the broadrange o fbenefits and costs to be included, and say that "the direct impact of the project on poverty alleviation will be identijied by estimating the economic impact of the project on low income households." The Panel did not find evidence in the Economic Study or the PAD of any estimates of the economic impact of the Project on low- income households. The Panel considers that such analysis, in addition to the broader macroeconomic analysis undertaken in the Economic Study, should have xxxvi been made during appraisal to provide a better understanding of whether the objective of poverty reductionenvisaged by OP 1.00 would be achieved. (2) Revenue Projections and the InstitutionalFramework Sensitivity tests were performed on the base case financial projections to 2016. The tests cover five "downside risks" and three "upside potentials" scenarios. The PAD states, however, that, "[elach of the sensitivities is considered in isolation, with all other assumptions in the base case remaining unchanged." It would have been helpful to have applied these tests using a more comprehensive probability-based analysis (e.g. the "Monte Carlo procedure"), which would have enabled wider distributions o fthe values o f each variable and their simultaneous variation to be taken into account, along with other variables such as changes inthe USh/US$ exchange rate. More specifically, the PAD projects specific amounts o f GoU support to power utilities that will be needed over the period 2005-2016, and indicates that the government is projected to collect net revenues o f $US217 million over this period. The PAD states that "[t]he power sector will be a drain on the Treasury until the proposed project is commissioned but a net contributor after." The Panel notes that this statement inthe PAD appears misleading and seriously at odds with the projectedrevenue stream of the Bujagali Project, given a large shortfall until2022 between revenue to be raised by the tariff for Bujagali proposed inthe PAD, and the requirements o f the capacity charge, as also indicated in the PAD. The Panel Report provides additional detail on the revenue gap that UETCL, in particular, will face, which may lead to large, urgent demands on the GoU Treasury and potentially on the Bank via its Guarantee. The possibility o f both higher Project costs and significantly lower revenues will have a major bearing on whether the GoU guarantee o f capacity payments under the PPA agreement i s likely to be triggered. The likely tariff variations and the possible revenue shortfalls or surpluses and their implications for UETCL, UMEME and government net revenues are key sustainability concerns; they matter for the future o f the power sector, for electricity consumers, actual and potential, and for the GoU's ability to invest in other key sectors and services. On the institutional side, the PAD recognises as a critical risk the possibility that UMEMEterminates its concession. The PAD further notes that following large increases in tariff rates, UMEME billing collection rates declined. The decline in fee collection rates suggests that UMEME's actual performance i s likely to remain potentially vulnerable to tariff increases from a variety o f causes, both external and internal. There are also risks that the technical and commercial losses will not be reduced as projected in the PAD. The Project revenue forecasts assume recovery rates will rise from 54 percent o f the energy sent out in 2006 to 75 percent by 2013 and thereafter. The Panel expert considers that it would have been realistic to use a lower forecast recovery rate. xxxvii Bank Management lists various approaches taken to address the potential risks to UMEME, including arestructuringo fits concession contract to protect it fkom the impact o fpower shortages and reduced revenue streams. Itremains to be seen, however, whether the requirement o f OP 10.04, to verify that the institutional framework is or will be in place to ensure that the Project functions as designed, can be met. As noted above, UMEME faces vulnerabilities, and the restructuring might have weakened their incentives to achieve the targets for reduced losses, enhanced collection rates and new connections envisaged in the load forecast and economic evaluation in the Economic Study. (3) Infrastructure Funds The Country Economic Memorandum cited in the PAD states that, "Special or extra- budgetary infrastructure funds have increasingly been started as a means to "protect" public funds from funding specific targets." O f the five funds listed, three are in the electricity sector. The Memorandum then says, "In general the proliferation of Extra- budgetaryfunds poses a serious fiscal threat in a poor country with weak governance systems and capacity. [. ..]Uganda is no exception: the TariflStabilization Fund which was designed to smooth tar#s until the Bujagali hydropower project comes on stream is already being utilized to subsidize higher tarus from thermal power generation. This Fund is also being used to fund selective rural electrijkation projects, despite the existence of a separate Rural Electrijkation Fund. Fiscal liabilities and contingencies created through extra-budgetary funds are not accounted for in the Government's budget." In light of these comments and of the scale of the revenue requirements, the financial risks accepted by UETCL and the Government, and the scale of the subsidies and guarantees involved in Bujagali, the Panel notes that Management should have explored further ways of managing and addressing these financial and governance risks, in the interests of project sustainability in accordance with OP 10.04. (4) The Power Purchase Agreement and Associated Risks Physically and in its electrical, economic and social impacts, the Project closely resembles the prior Bujagali project. Although there are some changes in the loan and guarantee structures, the key contract documents (PPA and Implementation Agreement) are similar inmany respects, but with some important differences highlighted below. In the Panel's opinion, a meaningful analysis o f the adequacy o f the current financial arrangements requires a comparison with those established under the prior Project. This section examines the current PPA, signed inDecember 2005 and amended and restated in 2007 (the 2005 PPA), and associated documents, and compares it in certain aspects with the PPA for the prior project (the 1999 PPA). xxxviii (a) Terms of the PPA Ingeneral terms, apowerpurchase agreement is a long-termcontract between a generator o f electricity and a purchaser. In the present Project, the PPA is a 30 year contractual arrangement between the Project Sponsor, BEL, and the GoU's entity in charge o f transmission, UETCL. Under the PPA, the Sponsor i s to sell the contracted capacity o f 250 MW exclusively to UETCL. The terms o f the PPA are critical in understanding how financial and economic risks o f the Project are allocated, including who would bear the risk o f low water flow and, correspondingly, low energy output (below capacity) o fthe hydropower facility. In the Panel's opinion, the introduction of a cost-based formula in the 2005 PPA, instead of the maximum capacity charge specified in the 1999 PPA, is probably the single largest adverse contractual change for the power purchaser (UETCL) and its guarantors. The new contractual basis for the Project represents a significant shift inriskaway fromthe projectinvestors andlenders to the power purchaser. Inshort, the cost provisions andtheir effects canbedescribed as follows. The formula for determination o f the monthly capacity payment (charge) is in Annex D to the 2005 PPA. It i s very complex, since the components are defined rather than assigned a specific price, and all are subject to variation. Inbroad terms, the components are: development costs; engineering, procurement and construction (EPC) costs; tariff debt service reserve; working capital, taxes and fees payable by the Sponsor- all o f these constituting "Tariff Project Costs" (TPC)-plus equity repayment and return; debt repayment; GOU Equity (representing past development costs), and Operation and Maintenance(O&M) fees. Some o f these components are treated as pure pass-through (fees, and elements o f the O&M charge). Others are carefully defined as to the make-up o f their "base" cost, and in some cases - including EPC costs - increases on the base are subject to a quantified percentage "cap". The costs are subject to accountants' inspection. However, the fact remains that, leaving aside debt repayment, the Sponsor has considerable scope to shape the base costs and insome cases the increases also, to deliver a higher capacity charge. In addition, potential for considerable delay i s built in to the determination o f the capacity charge (before which payments are to be made on an interim basis). In fact, up to 26 months may elapse after the start o f operations before there i s a determined capacity charge and, curiously, there are no specific provisions for dispute resolution for this particular item. As was the case with the 1999 PPA, the capacity charge is not related to output, so the payment will be the same under low hydrology (when the output may be halved) as with high hydrology. Of course, hydrology i s outside the Sponsor's control. But the payments also remain relatively invariable incases o f reduced plant availability, which i s underthe Sponsor's control. A percentage reduction inavailability (say, 5 percent) would have to be sustained for an entire year before there was an equivalent reduction in the monthly capacity charge. xxxix The Panel finds that for the Sponsor and its lenders, the terms and conditions of the 2005 PPA, especially those set forth in Annex D, seem to represent a low-risk (though potentially disputatious) means of managing and recovering costs which are, by definition, subject to uncertainty. For UETCL, the power purchaser, and its guarantors, by comparison, it means that there is no ceiling on capital costs and whether or not the Project delivers the direct economic benefits offered over 30 years, interms of costs and tariffs, i s to a significant extent, outside their hands. (b) RisksandConsequencesAssociated with the PPA The increased risk borne by the power purchaser and its guarantors (the GoU and the World Bank); discussed above, has significant consequences. Although some matters are discussed elsewhere, it i s important to highlightthem succinctly inthis section. The risks to which the Project i s exposed, how the risks are shared, mitigation measures and possible consequences, are summarized below. Capital cost escalation. If the capacity payment is set higher than present estimates, or rises subsequently, either tariffs must increase or additional subsidies are to be paid by GoU to UETCL, as discussed elsewhere inthis Report. Currency depreciation. For the current Project as for its predecessor, capacity payments are denominated in US dollars. As noted in the Inspection Panel's 2002 Investigation Report on the prior Bujagali project, a 10 percent per annum depreciation o f the Uganda Shilling (USh) against the US dollar (USD) would double the price o f the Project to Uganda in seven years. This would lead to tariff increase or additional GoU subsidies to UETCL. Prolonged low hydrology. Substantial uncertainty remains about future hydrological conditions, as discussed in detail in this Report. The PAD illustrates how the cost o f a unit from Bujagali rises dramatically in a "low-water" year. A "levelized" tariff may be set ex-ante, but if the actual hydrological pattern falls below that assumed for the "levelized" tariff, then the capacity payment shortfall will widen and the consequences will be those described above. Lower demand growth. Demand growth projections rest both on continuing growth o f demand from existing customers, and a highrate o f new connections/customers, such that the number o f customers almost doubles by 2012. If this growth does not occur, UETCL's revenues would fall, with the already discussed consequences. Lower or static proportions of supply costs recovered from customers. It has been assumed that this ratio will have risen to 75 percent by 2013. If it were to remain at the 2006 rate (54 percent), sector revenues would be 28 percent lower. Affordability. While the capital costs and total costs for the power plant have increased significantly in real terms (including 8 percent since the PAD was issued and the final price was fixed), the Economic Study o f the Project assumes that Bujagali's introduction xl will allow a reduction in (real) retail tariffs o f at least 5 percent compared with current levels. As noted above, this tariff reductionmay prove to be too optimistic. Collection rates. As described previously, there are risks as to whether the distributor (UMEME) will be able to reach and maintain high collection rates and to reduce the technical and commercial losses. The Economic Study forecasts that by 2012, UMEME will have reduced its technical losses to 16 percent and its commercial losses to five percent. Failure to achieve these reductions in losses may impair the GoU's ability to fully cover the costs o f new energy investments through the tariff system. If the risks noted above arise, this may (in the absence o f subsidies) result in a tariff increase which would affect the affordability o f electricity. Construction Delay. Despite Liquidated Damages provisions penalizing the contractor, the costs o f delay would likely in practice be shared via the 2005 PPA with the power purchaser, UETCL. The main consequence o f delay would be to defer expected consumers' benefits from the Project. In the Panel's opinion, overall, this may be regarded as one o fthe lesser, or more manageable, economic risks. Withdrawal of the Developer/Operator. This risk has been mitigated when compared with the 1999 PPA. The contractor is bound for the construction phase, and subsequently would be replaceable as operator if not so easily as investor. The Panel notes that the 2005 PPA provides for the Project to be bought out ifnecessary. Poor Plant Performance. Although when compared with international best practice, the 2005 PPA seems generous to the owner-operator in the scale o f penalties for low availability, this may be regarded as a low-risk. In the extreme, existing provisions for Company Default provide a safety net. (c) RiskMitigation Measures As described in the previous section, there have been important changes between the 1999 and 2005 PPAs that have had the effect o f increasing the risk on the purchaser as compared to the project sponsor. In the Panel's opinion, however, some other changes represent potential improvements regarding reduction o f risk. Some o f the changes most relevant for project costs and risks are: Award of the Project by Competition. The Panel acknowledges Management's statements that competitive solicitation o f Independent Power Producer (IPP) projects i s an international best practice aimed at ensuring the lowest market price consistent with technical fitness to carry out a project. This procedure i s a marked improvement over the prior project. Inthis case, however, the benefits o f competition were largely lost by post- bid negotiations, which allowed the price to rise by at least 28 percent before it was established. Further, the recent amendments to the PPA provide specific contractual scope for further upward revision. Buy-back in case of L o w Hydrology. Both the 1999 and 2005 PPAs and Implementation Agreements provide for a buyback o f the plant by UETCLunder default xli conditions and certain force majeure events. In general terms, these provisions follow international norms. However, the 2005 PPA adds a new provision: UETCL may terminate the PPA and buy back the plant inthe event o f 30 consecutive months o f "low water". In this scenario, the cost o f power from Bujagali, per unit, may become prohibitively high, and it may be preferable for the public authorities to assume control, stop paying the fixed capacity charge, smooth tariff effects and ensure that funds were available for alternative generation. While this new provision i s to be welcomed, the Panel notes two areas o f concern: first, the low water trigger may have been defined too restrictively from the power purchaser's perspective. Second, the terms and conditions for the buy-out, which appear to allow the Sponsor to set the price broadly to equate to capacity payments foregone, seem relatively generous to the Sponsor, given that the plant will be inreal trouble ifthis scenario occurs. (d) Conclusions-Distribution of Risks It is clear from the review o f the Project documents that the greatest share of economic risks lies with the power purchaser. The capacity charge may be adjusted upwards if the developer/operator hits unforeseen costs, but not downwards if demand or supply conditions deteriorate for the purchaser. The Panel notes that infact the lenders especially but also the investors are held harmless against all or most eventualities. However, in a crisis o f non-affordability in Uganda such as might be producedby currency devaluation or very low hydrology, the investors and lenders may also be at risk, ifthe money to pay the capacity charge is not available. Inthese circumstances, buy-out may provide the best solution. The Panel observes that the high allocation of risk to the UETCL, the power purchaser, and eventually the GoU increases the possibility that the Project may not achieve the broad objective of sustainable development and poverty reduction embodied in Bank Operational Policies and Procedures. This also increases the possibility o f the Bank (IDA) Guarantee being called. The Panel is concerned that any additional GoU resources that are spent in the financing of the development and operation of this Project may lead to decreased resources available for social and other prioritydevelopment programs. VII, InvoluntaryResettlement The Requesters claim that resettlement under the Project i s not complete. They raise multiple, interrelated involuntary resettlement issues, including loss o f livelihood, under- compensation, inability to obtain secure land titles, lack o f consultation, and request to share inProject benefits. Management believes that this Project has been well prepared in accordance with Bank policies. At the same time, Management in its Response "agrees with the Requesters' contention thatpast resettlement is incomplete." xlii (1) The Assessment and Action Plan The Panel notes that this Project involves the rather unusual circumstance o f an ongoing, incomplete resettlement program which was developed under a previous Bank-financed operation and was based on a policy that is no longer applicable, OD 4.30 on Involuntary Resettlement. The policy now applicable to the Project, OP/BP 4.12, nevertheless has the same overall objectives, and both the old and new policy call for the preparation o f a Resettlement Action Plan (RAP) consistent with the policy objectives and in compliance with specific policyandprocedural requirements. Inthe current Project, Management chose to develop and build on an "Assessment of Past Resettlement Activities and Action Plan" (APRAP), rather than to develop a new RAP. The justification for this approach was that affected people had already been relocated and others had already received compensation under the prior project. The Panel observes that such an "Assessment" i s not a resettlement instrument referenced in Bank policy. Setting aside questions o f terminology, the Panel considers that the overriding issue i s whether the TORand subsequent Action Plan meet the objectives and requirements o f Bank policy on Involuntary Resettlement. Accordingly, to achieve compliance, the APRAP should have included the elements o f a RAP as defined in the policy (and used by Management inthe T-Line part o f the Project). The way an Assessment and Action Plan was substituted for a full RAP had far ranging consequences. Following the TORS, BEL prepared an assessment o f the progress in the execution o f the Bank-approved old RAP, and recommended recovery activities where it observed gaps. The assessment did not include an evaluation o f the impact o f the delay on the socio-economic conditions o f the Project or an assessment o f whether or not the previous Sponsor's complied with either the former or current Bank's resettlement policy objectives. Consequently, the new Sponsor's resettlement responsibility to the people who were in the process o f being resettled was circumscribed to certain outstanding commitments that the new Sponsor wished to recognize. As reviewed below, the critical policy requirement to census all displaced persons as o f the project baseline was neglected-a decision undermining much o f the policy objectives. The public consultation process, an integral part o f a RAP, was truncated, predefining the consultations to on-going issues, rather than including all aspects o f the Project. The Panel also found that the approach to consultations with people who had moved and had been compensated i s not consistent with Bank policy on Involuntary Resettlement. (2) Baseline Socio-Economic Data One central requirement o f Bank Policy on Involuntary Resettlement i s to develop socio- economic data on affected communities and households, as a basis to assess risks o f impoverishment and develop measuresto safeguard affected people, including vulnerable groups, against these risks. The assessment o f risks and related mitigation measures xliii should be based on an accurate census survey with details on current occupants, displaced households, livelihood, expected loss (total and partial) o f assets, and vulnerable groups. The Panel could not find an adequate "socio-economic suwey of theproject-affected area at the hydropower site to characterize the socio-economic conditions and livelihoods of the people living in the eight project-affected communities" as required by the SEA'S TOR. Situations not adequately considered at the time o fthe prior project, or that arose in the interim period, were not appropriately dealt with because o f the lack o f an adequate baseline assessment. This does not comply with OP 4.12. This led to action plans that didnot meet the policyobjectives and requirements. The Panel notes that the survey conducted by BEL cannot be considered a census of economic or social conditions as defined inOP 4.12. Inthis sense, the Management's claim that the Project took the first Panel's report findings into account in the preparation of the current Project i s not accurate because significant weaknesses in the process of gathering baseline data information were similarly identified in the 2002 Panel Investigation Report. The Panel also found no formal monitoring or evaluation report supporting the assertion that the involuntary resettlement was "largely completed,'' the reason stated for forgoing a full RAP preparation, as required by OP 4.12. The Panel finds that the hydropower APRAP failed adequately to assess and update the previous 2001 RAP and provide additional new information as required to complete the RAP requirements to current standards (OP/BP 4.12). This does not comply with OP 4.12. (3) Livelihood restoration The restoration o f livelihoods o f displaced people i s a core objective o f OP 4.12 on Involuntary Resettlement. The policy provides that "displacedpersons should be assisted in their efforts to improve their livelihoods and standards of living or at least to restore them, in real terms, to pre-displacement levels or to levels prevailing prior to the beginning of project implementation, whichever is higher." In its investigation, the Panel learned that livelihoods o f affected people have been disrupted for some seven years, stemming back to the beginning o f relocation and resettlement actions under the prior Bujagali dam project. Duringthis period, many o fthe people that were originally displaced were essentially left in limbo, and did not receive key elements o f the resettlement process to which they were entitled under Bank policy. Also, as a consequence o f the project's "hiatus", certain o f AESNP's commitments to regulators and the communities under its resettlement and community development plans were not fulfilled The Panel observes that the effects on the people o f the original displacement, and o f the ensuing delay, have not been fully reflected in the APRAP. Specific issues relating to livelihood restoration are reviewed inmore detail below. xliv (a) Methodto Assess LivelihoodRestorationand Address Project Delay The Panel notes that the methodology used to assess livelihood restoration did not compare the 2006 livelihood status o f the resettlers to their previous conditions. Nor did it set anew 2006 baseline for future actions. This methodology was ambiguous as to what was and was not being measured and, as a result, it produced only a list o f unfulfilled promises left over by the prior project. In the Panel's view the methodology used to assess livelihood restoration inthe context of this Project, while suggestive of issues, cannot substitute for an economic analysis of the livelihood risks and restoration. The Panel also finds that Management did not assess and include into the APRAP a methodology for restitution of the unintended socio-economic costs incurred by displaced persons resulting from project stoppage/delay. This is not consistent with OP4.12. (b) Real or Perceived Unfulfded Promises inthe Prior Bujagali Project At the hydropower site, the APRAP survey found that the people believethat a number o f promises made by the previous Sponsor were left unfulfilled. Management claims that BEL and the BIU "are now resolving all outstanding issues" and have committed to address the issues left unfulfilled by the previous sponsor. The Panel notes a lack o f method for deciding what promises were or were not made, which would or would not be honored and the timeframe for completing the resettlement activities, while the Bank's safeguard policies require that the resettlement plan define clearly these activities and provide a schedule for their implementation. The Panel notes that lack of clear communication with affected people to address the concerns of the displaced persons with regards to the commitments made by AESNP, risks leaving the Bujagali project with contentious, unresolvedissues. (c) Specific LivelihoodRisks: Fishing and Agriculture During its investigation, the Panel learned that fisherman who were relocated at the time of the prior project have faced severe obstacles to restoring and maintaining their livelihoods. Among other problems, they were settled much farther from the fishing areas, lacked transport to get there, and have had their access even to these areas restricted by fencing connected with Project activities. There are also questions as to whether they were paid for fish ponds that were taken. There i s a strong belief that promises to restore their livelihoods were not kept, and feelings o f great frustration. The Panel notes that the 2006 APRAP contains a two-page "plan" to address livelihood restoration infishing, developed by BEL, sets laudable general goals such as training that will address preparation o f fisherman for change in the river characteristics following impoundment. However, this planning i s not associated with any studies on the economics and nutritional importance o f fishing despite being called for in TORSo f the SEA. Moreover, no additional support was allocated to what was called an underestimated, critical activity: the 2006 budget for fishing activities remains at the 2001 level. xlv During its investigation, the Panel also learned of productive and locational disadvantages for relocated farmers, e.g.: land fertility was not considered in livelihood restoration planning, but surfaced as a major concern in subsequent consultations; diminished ability to cultivate cash crops (coffee, vanilla); far distances between residences and agricultural land. The Panel observes, however, that the approach taken to restore damaged agricultural livelihoods, set forth in the APRAP, follows a pattern similar to that for fishing. No baseline census o f the displaced persons and a socio-economic analysis was carried out. The Panel has found that insufficient information was available to permit the new Sponsor to assess whether or not landlessness increased or decreased under this strategy. The Panel further notes that the 2001 RAP lacked any livelihood restoration plan or budget for agricultural activities. The Panel finds that the 2006 APRAP attempts to mitigate this situation, but its provisions will most likely be insufficient to meet Bank policy requirements. Management failed to ensure that the Project would institute or assure financing to mitigate these losses, exposing the displaced to on-going impoverishment risks that are now approaching eight years. The Panel finds that the Project failed to provide adequately for loss of livelihood associated with the loss of fishing and agriculture, innoncompliance with OP4.12. (d) Compensation The Panel notes that the agro-economic analysis o f livelihood restoration i s weak, particularly with reference to compensation. Underestimation o f the establishment periods for coffee and other crops, including vanilla and cocoa, made it economically unfeasible for the displaced to reestablish their lost incomes. The Panel concurs with the APRAP's findings, which validate the claims of the project affected peoples (PAPs), that full replacement value compensation may have not taken place in the prior project. (e) Land Titles Most o f the displaced lacked security o f land titles before displacement, but they may have had established, informal security with usufruct rights recognized by others. According to the APRAP, while many people who were interviewed stated that they received land titles, it also appeared that some Project affected persons (PAPs) did not receive the titles. During its visit to the Project area, the Panel team witnessed Project- generated insecurity among displaced persons as a consequence o f resurveying and proposed readjusting o f the boundaries within the settlement. The Panel expects that this situation will be dealt with during the implementation o f the APRAP. The Panel finds that the APRAP conclusion related to the necessity of issuing land titles to people resettled under the prior project is consistent with OP 4.12. The Panel notes however that there seems to be no agreed timetable for the issuanceof these titles. xlvi (f) Vulnerable Peoples The APRAP determined that there was no proper identification of vulnerable people up until 2007. The Panel notes that a group o f vulnerable people, the landless tenants and sharecroppers, may have been left out from receiving compensation as a result. The APRAP notes that "the situation of tenants and sharecroppers (who were compensated only for crops as they did not own land) appears to be worse in this respect than that of landowners)." The Panel notes that the absence of a focus on livelihood risks to the vulnerable is evident inthat none of the proposed assistancemeasures addresses the vulnerable tenants/sharecroppers or children. Additionally, the proposed assistance measures do not address the question of sustainability beyond the limited Project support. The Panel finds the Project is out of compliance with the vulnerable peoples provisions of OP 4.12. (g) Housing and Electricity for Affected People Housing: The APRAP states that the houses built for the resettled population met with the design criteria that were set out in the 2001 RAP and were therefore generally compliant with the commitments made. It states that the resettlers felt that the houses were better than the ones they had, but still complained about deficiencies in the buildings. During its field visit, the Panel verified that the standard of living of the displaced households who resettled in Naminya and Nansana has improved with respect to housing. On the other hand, the APRAP discovered some shortcomings in housingcondition and the Panel observed physical problems and deteriorationwith some of the houses and structures. The Panel is concerned that no physicalaction is planned with regard to houses at the resettlement site (apart from repairing the taps from the rainwater harvestingsystem). Electricity: A highvoltage line crosses Naminya. Throughout the process to conduct the Assessment, numerous displaced persons, those who took cash compensation, and local leaders stated that they believed AESNP made a commitment to provide electricity to Naminya and other communities. The APRAP says that "it does not seem" that such a commitment was planned under the 2001 RAP. On this point Management Response states that "BEL together with UMEME is exploring possibilities for the provision of electricity. BEL will also finance a feasibility study for electrical distribution to the resettlement community, which may convince UMEME toprovide a supply." However, during its visit, the Panel learned that AES had made the following commitment to affected communities in 2001: "AESNile Power is committed to provide step-down transformers in eight villages in the affected area and in the new resettlement land allowingfor access topower by residents who have never had the opportunity." The Panel has found evidence that displaced persons were also told that "you have a right to electricity as do all Ugandans." Given the context and previous expectations, this broad statement may have reasonably been interpreted as a promise to deliver xlvii electricity connections to affected households. The Panel notes that this is an outstanding controversy of highimportance to the affected communities. (h) Investment Resources for Livelihood Restoration The Panel's review o f the limited scope o f the livelihood restoration programs indicates that they may be under-budgeted. As information on livelihood conditions o f the displaced persons, including those who were economically or physically displaced but took cash compensation, has yet to be determined, the costs o f livelihood recovery are unreliable. As livelihood restoration instruments develop, Bank policy provides that Management i s to monitor the resettlement budget to ensure sufficient resources. (i) Overall Conclusions on Livelihood Restoration The Panel's review o f the livelihood assessment method and other Project data shows that the Bujagali Project i s facing substantial problems in measuring, monitoring, and mitigating livelihood risks, especially among vulnerable peoples. The Panel finds that the Project i s in non-compliance with the mandate of Bank Policy on Involuntary Resettlement to improve or at least to restore, in real terms, the livelihoods and standards of livingof the people displaced by the Project. (4) Sharing inProject Benefits and Community Development Sustainable development, the sharing o fproject benefits, i s one o fthe principal objectives o f the Policy on Involuntary Resettlement. The Panel believes that it is likely that the community development programs, once executed, will provide positive benefits for Uganda. However, the Panel identified compliance issues related to the Community Development Action Plans (CDAP). Lack of Focus on Displaced Persons: The Panel notes that the CDAP, though an important demonstration o f the Sponsor's corporate social responsibility, i s not necessarily related to benefit sharing for displaced persons as required by the objectives o f OP/BP 4.12. Lack of Program Specificity: The problem identified by the first Inspection Panel Report over five years ago persists. The Panel finds that in the area o f sustainable development and benefit sharing, the CDAP focuses almost entirely on short-term exercises; its targets are poorly laid out; and it makes no significant or systematic effort to ensure that resources are directed to institution building or social fundamentals rather than only short-term constructionprojects. Imbalances in Allocations between the T-line and HPP: CDAP budgets show sharp differences. The T-line has a higher number o f physically and economically displaced peoples than the HPP, but a smaller proportion o f the resources devoted to CDAP activities. The Panel finds that budget of the two components were not properly coordinated and this may leadto social discord among the displaced. xlviii Decrease in Investment Resources to this Effort. The previous Panel also found that "the net present value of the resources to be contributed over a 35-year period seems very low.'' While the decision to reduce investment resources i s not a compliance issue, the current Panel does not understand why Management decided to further reduce its effort. Even discounting for inflation, eliminating the second phase raises questions as to Management's responsiveness to the previous Panel's findings. The fact that the same problems are surfacing with two different sponsors is of concern to the Panel. The Panel finds that with limited funding, broad criteria for eligibility and lack of specificity, the CDAPprograms do not assure compliance with OP 4.12. (5) Indigenous Peoples The Requesters claim that the provisions o f OP 4.10 on Indigenous Peoples have not been applied to the Project because the SEA does not consider the Basoga inhabitants o f the Project area as indigenous people, in spite o f the fact that the Third Schedule o f the Constitution o f the Republic of Uganda expressly considers the Basoga as such. The Response states that Management respects local legislation but draws a distinction between the definition o f indigenous people according to the Constitution o f Uganda and that provided in OP 4.10. Under the Ugandan Constitution, in order to be considered an Ugandan citizen by birth-regardless of socio-economic status-one must belong to one o f the 56 "indigenous communities" listed in the above-referred Third Schedule (or have a parent or grandparent who does); while under the Bank Operational Policy, the term indigenous is used "in a generic sense to refer to a distinct, vulnerable, social and cultural group" (emphasis added) possessing "in varying degrees" the characteristics listed inparagraph 4 o fthe OP 4.10. Although the Basoga people meet some o f the criteria necessary to be regarded as indigenous people inthe context o f Bank-financed projects pursuant to OP 4.10, they are a large and influential group with political, social and economic standing in Uganda's society, and the Panel did not find any indication that they are regarded as a "marginalized and vulnerable segment'' o f the population that i s unable to "participate in and benefitfrom development." The Panel didnot find any evidence that Management violated the provisions of the Bank's policy on Indigenous Peoples, with regard to the Basoga people. VIII. CulturalandSpiritualValues Inits earlier Investigation report, the Panel indicated the efforts o f the Bank to address the cultural and spiritual issues that the project raises, and Management's good faith attempts to mitigate these issues. At the same time, the Panel also noted the importance o f including all key stakeholders in consultation and taking steps to minimize the possibility o f disturbance to the local communities that might arise from excluding any faction from such consultations as the project went forward. For the purpose o f the present Investigation Report, the Panel conducted a careful research and analysis o f xlix relevant materials, including numerous studies by the Cultural Research Center in Jinja, which focuses on Busoga culture. The studies prepared in 2001 for the prior Bujagali Project mapped individual and community level spirits. These studies also identified a general protocol for moving spirits according to the tradition o f the Busoga. The 2001 Cultural Properties Management Plan (CPMP) sets out a six month, US$125,000 program o f consultation, compensation o f individuals for disturbed graves and shrines (amasabo), appeasement and relocation o f the Bujagali spirits. Three individuals were identified as stakeholders for consultation about the spirits at Bujagali Falls. Problems, however, emerged with the so-called "appeasement of community spirits." Later, the implementation o f the CPMP stopped for the next four years. During the preparation o f the present Project, BEL committed to detailed consultation with locally affected communities regarding cultural properties management work undertaken by the prior project, with follow-up and a revised CPMP, as necessary. BEL'S consultations led it to conclude that, rather than a localized cultural site, the Bujagali Falls are o f spiritual significance to the Kingdom of Busoga as they are considered a place inhabited by spirits. The Kingdom's leadership expressed support for the Project and BEL committed to continue consultations to determine what needs to be done prior to the flooding o f the Falls. For the Basoga, the traditional religious structure i s distinct from the cultural structure. BusogaSpiritualDomain Although the peoples o f other ethnic groups inhabit the Project area, the Basoga claim spiritual dominion o fboth sides o f the Nile, its islands, the water and its waterfalls. Their language, Lusoga, predominates in this area, on the East bank o f the River Nile. The Basoga share a common dialect and ideological, spiritual history, sharing a cluster of eight or more high status spirits, including Budhagaali, the spirit residing at the Bujagali Falls site. To the Basoga, the project area-like their entire region-is inhabited by ancestral spirits and living humans who are constantly interacting - from birth to death andbeyond. The key elements o f Busoga spiritual cosmology are: (a) the spirits are innumerable, powerful and frequently cross over into the world o f the living and may do both good and bad; (b) they inhabit the same world as the living and are associated with animate and inanimate objects throughout the landscape; (c) they can move freely without the need for humanpermission; (d) they have differential power, influence, and interests; (e) they are hierarchical, somewhat comparable to the ancient Greek Pantheon; (f) they influence the health, well-being and the livelihood o f the living; (8) more powerful spirits communicate through mediums who do not view themselves as capable o f negotiating or predicting spirit behavior-they are mediums o f the spirit who possesses them; and (h) the mediums are selected by the spirits, not by the cultural (political) leaders. 1 Busoga CulturalDomain Interms of cultural structure, the Busoga Kingdom is a cultural institution that promotes popular participation and unity among the people o f Busoga through cultural and development programs for the improved livelihood o f the people o f Busoga. Unlike the typical monarchies inAfrica, the Busoga did not have a central authority at the advent o f British rule. Nevertheless, it had developed small principalities, each with its own hereditary ruler. These principalities where later to be consolidated under a King called "lsebantu Kyabazinga" who ruled the Busoga Kingdom. This secular institution, which is a stakeholder on Busoga cultural issues, makes no claims to hold spiritual power. The Panel fmds that Management and the Sponsor have increasingly recognized and involved the Kyabazinga Institution as an important guardian of the Busoga cultural tradition. The Panel also recognizes that the Kyabazinga Institution i s not empowered to speak as surrogates inconsultations for the Basoga spiritual stakeholders. Panel's Analysis-Physical Cultural Resources DuringProject design, BEL'Sconsultations ledit to conclude that, rather than a localized cultural site, Bujagali Falls i s o f spiritual significance to the Kingdom o f Busoga as it i s a place inhabited by spirits. A CPMP, in compliance with OP/BP 4.11, should have identified Bujagali Falls as a significant cultural resource, triggering rigorous safeguards for specific avoidance, consultation and mitigation as required under the Bank's Policy. Interms o f avoidance, the Panel can findno evidence, since the initiation o fthe Bujagali Project o f Management considering avoidance o f the significant cultural resource impacts at Bujagali Falls. The Panel finds that Management failed adequately to consider or implement alternatives to avoid the project-related impacts on Busoga spirituality and culture inviolation of OP/BP4.11. As for consultation, the Panel considers that the consultation methodology used in the Resettlement and Community Development Action Plan (RCDAP) was detailed, but structurally flawed. First, the survey included mostly laymen many o f whom were not sufficiently knowledgeable o f the traditional religion. Second, it excluded key spiritual leaders (baswezi abadhagaali) o f the Busoga clan. The consultations did not recognize that mediums o f the Nabamba Budhagaali spirit derive their power through recognition by the traditional clan priests (rnuswezi) as agents o f their believers. A medium o f the high Busoga spirits is incapable of commanding hisher followers, meaning that the appropriate consultation strategy i s participatory, as this i s common among traditional religions. Third, the survey was limited to the people inthe project area, many of whom were non- Busoga migrants who had moved into the area following a disease-linked depopulation. Most of those who believe in the significance of the Bujagali Falls spiritual site do not live in the immediate vicinity of the Project. The terms o f reference for the cultural consultations were not revised after interviews discovered that the spiritual sites in the project area were o f major significance to a religious tradition that extended beyond the immediate area o fthe study. li The limitedconsultation creates on-going uncertainties as to affected people's acceptance o f the project's cultural resource impacts. The Panel finds that the Project failed adequately to consult with the Busoga spiritual clan leaders associated with one or more high status Spirits about the significant cultural patrimony of the Bujagali Falls. This is not incompliance with OP 4.11. The Sponsor's approach had been to identify three interested "stakeholders" in the "Bujagali spirit(s)" and fund either appeasement or relocation ceremonies. The Sponsor focused on obtaining written consent from three stakeholders that compensation hadbeen adequate and that construction of the dam at Dumbbell Island and the resulting inundationcould proceed. Following a ceremony financed by the first Sponsor on September 28, 2001, to relocate the Bujagali spirits, Management claims that all three interested mediums acknowledged in writing that compensation had been adequate and construction o f the dam could proceed with the partial inundation o f Bujagali Rapids as a result. The witness NGO contradicts this account; they claim that the October 2, 2001 negotiations with the NabambaBujagali withheld his endorsement. Another medium, Lubaale Nfuudu, felt the spirits had been moved to a temporary location, on his property and will be relocated again nearby the project site. The Nabamba Bujagali medium seems to have remaining claims over the site. The Panel notes that 2001 Project documents identify the Lubaale Nfuudu as a diviner (muswezi) who asserts that the spirit Lubaale i s the father o f Nabamba Budhagaali spirit. He conducts occasional ceremonies with busweszi at the Bujagali Falls to communicate with Lubaale, one o f the highest spirits within Busoga cosmology, but different from the Bujagali spirit. This opens the possibility that Bujagali Falls, as a cultural property may be the site o ftwo highspirits o fthe Busoga, not one. Panel interviews with the Nabamba Bujagali, cultural experts, the Sponsor, and Management show that the consultation process has not yet led to satisfactory outcomes for all and that mitigation efforts cannot be considered completed. Misidentifying the Bujagali Falls as a local cultural resource, misaligning its consultation strategy, and failing to prepare a new Cultural Property Management Plan compounded errors and muddled mitigation. Resultant problems included loss of objectivity of the Sponsor, impatience, assignment of pecuniary motives to stakeholders, cost cutting, culturally inappropriate mitigation efforts, and most importantly, a misunderstanding that the Bujagali Project is ensconced in a long- term relationship with its new neighbors and their spirit world. The Panel finds that Management unnecessarily and inappropriately took sides in a spiritual controversy of a religion in which millions of Ugandans believe. The Panel finds this action by Management to be non-compliant with the OP4.11. lii With reference to the islands, the Sponsor felt it was impossible to locate graves located there with certainty and, therefore also impossible to exhume and relocate remains. The new Sponsor took over the mitigation strategy developed by the previous one to hold an inter-denominational remembrance service to honor the memories o f those buried in the islands. N o consultation or ethno-archaeological work had established the provenance o f the remains to determine the culturally appropriate mitigation. The Panel obtained information that the islands may be the location where previous spiritual media are buried. Noting that appropriate consultation and mitigation has yet to be done for the Bujagali Falls spiritual site, the Panel observes that the island areas must be included in the mitigation strategy to reach compliance with OP/BP 4.11. The Panel finds that Management assumed that what they called the "Bujagali spirits" were restricted to the Project construction and flooding area, in contravention to the BP 4.11 requirement that they work with and assist the Borrower to identify the spatial and temporal boundaries of the cultural resources affected by the project. This did not comply with avoidance and mitigation requirements of OP/BP4.11. Narrowing its size, location, and scale, Management discounted the significance o f what should have been identified as the Bujagali Falls spiritual site to all o f the Busoga, not just to those living in close proximity to the Project area. It appears that Management defined the project-affected-people under OP 4.1 1 on Physical Cultural Resources as those covered under OP/BP 4.12 on Involuntary Resettlement. Inthe case o f the Bujagali project, the groups are distinct. Consequently, the Panel finds that the culturally and spiritually affected people were not adequately identified as required by Bank policy. Critical NaturalHabitats Given the importance that the Requesters attach to the spiritual aspects o f the Falls, the Panel examined in detail the Bank's consideration o f this issue in light o f different policies. Inthe Project, these issues have mainly been considered under the Bank policy on Physical Cultural Resources (OP/BP 4.11). However, OP 4.04 also contains provisions that are relevant to these issues, as discussed below. Project documents recognize that the inundation o f the Bujagali Falls will destroy a natural habitat o f significance to the people o f Uganda, and identify specific actions to offset this impact. At the same time, Management takes the view that the Project i s not significantly converting or degrading a "critical natural habitat" as defined in OP 4.04. The Panel analyzes the various dimensions o f that decision in light o f provisions contained inthe Bank policy. Since OP 4.04 states that the "Bank does not support projects that, in the Bank`s opinion, involve the signijkant conversion or degradation of critical natural habitats," the Panel reviewed what constitutes a critical natural habitats. Annex A o f OP 4.04 defines "critical natural habitats" as liii "(i)existingprotected areas and areas oficially proposed by governments asprotected areas (e.g., reserves that meet the criteria of the World Conservation Union [IUCN] classifkations [footnote omittedl), areas initially recognized as protected by traditional local communities (e.g., sacred groves) and sites that maintain conditions vitalfor the viability of these protected areas (as determined by the environ-mental assessment process; or, ... (ii) sites identi3ed on supplementary lists prepared by the Bank or an authoritative source determined by the Regional environment sector unit (RESU). Such sites may include areas recognized by traditional local communities (e.g., sacred groves)." (Emphasis Added) Thus OP 4.04 indicates that socio-cultural factors do have abearing on the assignment o f "criticality" to a natural habitat. The Panel further observes that there i s substantial literature and practice recognizing the important relationship between sacred places and the conservation o f natural habitats and protected areas, a subject o f much attention in recent years. IUCN Guidelines for Protected Area Management Categories, referred to in the definition o f Critical Natural Habitat under OP 4.04, state that a Category I11 Protected Area i s an "[alrea containing one, or more, specijk natural or naturaUcultura1 feature which is of outstanding or unique value because of its inherent rarity, representative or aesthetic qualities or cultural significance." (emphasis added). The Panel notes that "areas initially recognized as protected by traditional local communities (e.g. sacred groves)", as referred to in OP 4.04, include areas recognized as protected for their cultural significance and ecological functions by traditional peoples. In the Bujagali Falls area, Project studies and the Panel have identified islands, sacred groves, rocks, waterfalls, and numerous Busoga spiritual sites. The persistent resistance to disturbance o f the site by the Busoga spiritualists and the expressed concerns o f the Kyabazinga Institutions i s evidence that Bujagali Falls are a natural habitat o f great importance to the Basoga that i s being protected by them, as provided in OP 4.04. The Panel found evidence and documents describing the cultural and spiritual significance o f the Bujagali Falls site to the Busoga people. Inaddition, studies conducted by AESNP for the prior Bujagali project suggest a strong ethno-botanical use o f the Bujagali Falls project area, in particular the islands, for healing and mental well-being. These studies include an ethno-botanical survey with these numerous healers to identify the flora associated with their practices. As mentioned above, OP 4.04 states that the Bank does not support projects that, inthe Bank's opinion, involve the significant conversion or degradation o f critical natural habitats. The Panel notes that this aspect o f the text ("in the Bank's opinion") indicates, inter alia, the need for and importance o f the considered judgment o f the Bank on this crucial question. This phrasing does not imply or give Management a blank check to apply or not certain policy provisions to a specific project but rather requires Management to form and provide expressly an opinion on the issue in question, which must be consistent with the objectives of the applicable policy. This is particularly liv relevant in view o f the controversy surrounding these issues in the present Project. The Panel did not find sufficient documentation that would have permitted Management to make such a consideredjudgment. The Panel finds that the Bujagali Falls area i s a sacred place, like a sacred grove, recognized by the Basoga, a traditional local community, for its high cultural and spiritual significance and inter-related ecological features and values. Inthis context and for the reasons described above, the Panel finds that the Bujagali Falls area may be regarded as a critical natural habitat for purposes of OP 4.04. The Project entails flooding o f the Bujagali Falls area. Bank policy regards inundation as a form o f significant conversion or degradation. Inlight of the above, the Panel finds that the Project record does not provide sufficient discussion as to why the area was not considered a critical natural habitat. Nor do Project documents explain the Bank's "opinion" that the Project would not involve significant conversion or degradation o f a critical natural habitat. Considering the known spiritual importance of the Project area, without such an explanation, one could also arrive at an opposite conclusion, Le. that the inundation may be regarded as resulting in the significant conversion of a critical naturalhabitat which would be inviolation of OP 4.04. The Panel finds that omitting the reasons behind an opinion of not declaring the Falls a critical natural habitat is not consistent with the objectives of OP/BP 4.04. The Panel finds that there is an overriding need for the Bank to address these issues in a coherent and well-founded manner to ensure compliance with Bank policies. The Cultural Property Management Plan (CPMP) It remains uncertain whether or not key stakeholders (consulted and as yet to be consulted) in the spiritual community comprehend the fact that their sacred site will be inundated and inaccessible for their traditional ceremonies. This issue extends well beyondthe two spiritual mediums. Management was also on untested grounds by substituting an abbreviated procedure, not provided for inBank Policy whereby the new Sponsor would find out what remains to be done from the previous plan, which was assumed to be correct. The prior Sponsor's plan was designed under OPN 11.03, a policy framework that had been replaced by 2006. The Panel finds that insufficient competence was dedicated to an examination of this issue for the Appraisal. There are livelihood impacts directly associated with the disruption o f the cultural resources sites that were ignored. Contemporary ethnographic accounts and the RCDAP 2001 describe many categories o f traditional practitioners (diviners, interpreters, gourd players, immunizers, exorcists, dispensers, herbalists, caretakedmediums, bone sitters, and more) who require payment in money or in-kind for their services, as in any other religion. Within the context o f a traditional society, these transactions are substantial, and they should have been included in the CPMP as specified in OP 4.11. The Panel finds lv that Management failed to prepare a Cultural Properties Management Plan, assuming that the work of the previous Sponsor was sufficient to meet OP/BP 4.11 guidelines. Insummary, the Project misidentified the Bujagali Falls spirits as localized, with Project impacts limited to people nearby the Project site. The TOR for the Cultural Properties Management Plan omitted the need for consultation with the spiritual leaders (baswezi) o f approximately 340 Busoga clans with spiritual ties to the cultural property that was to be affected by the Project. The Panel finds that Management is in non-compliance with OP 4.11, by misjudgingthe size, location, scale as well as the nature and magnitude of the cultural and spiritual significance of Bujagali Falls. The Panel also finds that Management did not consult with key stakeholders throughout the Project cycle and is, therefore, in non-compliance with OP 4.11. The Panel also finds that mitigation measures were not adequate because the scope of the impact and the consultation processwere incomplete. Opportunities to Address Cultural and Spiritual Issues The Panel observes that there are important opportunities available to address the cultural and spiritual issueswithin the context o fthe Busoga and the specifics o f OP/BP 4.1 1. The Panel's investigation o f Busoga culture suggests that the task ahead i s one o f restoration o f cultural harmony and developing an appropriate consultation process, not simply o f appeasement. Management's cultural resource strategy o f the prior project has focused on closure, relocating, or appeasing the spirits, compensating when necessary, documenting spiritual appeasement through signed certificates, and setting a finite timeline (originally six months in 2001). The current Project continued this strategy o f appeasement by honoringthe memories o f those buried on the island. Such a service might prove valuable for some residents in the project area, but does not appear to have been developed through consultations with the Busoga spiritual stakeholders. Similarly there does not exist yet a long-term strategy for sustaining a relationship between believers and the Project, nor have arrangements been negotiated allowing worship at alternative sites in the future. Panel interviews with Basoga cultural experts revealed that an outcome o f a spiritual consultation may be for the spirits to stay inplace and permit the project to proceed. The Panel finds that Management has thus far failed to support negotiations that would allow enduring coexistence with spiritual elements of Busoga traditional religion and the Bujagali dam. IX Systemic IssuesAffecting PolicyCompliance This investigation, like some earlier ones, has revealed certain systemic issues that have affected the Bank's overall compliance with its Policies and Procedures. The first o f these, noted at the beginning o f this Report, i s the need for considerable care to apply Bank policies inthe complex area o f energy production, to promote sound development practices and ensure Project costs, including social and environmental costs, do not exceedbenefits. Others are summarized below. lvi Legacy Issues from PrecedingProjects This investigation encountered a situation o f adverse effects on people due to a failure to assess, correct and complete resettlement actions initiated in the previous effort to develop the Bujagali dam. When the implementation o f this earlier project was halted, following withdrawal o f the sponsor, manyo f these people were essentially left inlimbo, and they did not receive key elements o f the resettlement process to which they were entitled under Bank policy. The experience with the Bujagali Dam highlights the significant problems that may arise when actions of previous projects are not carried to completion or corrected in accordance with Bank policy. The Panel notes the importance to affected people of timely actions to address any such situations that might arise. Incorporating Climate Change into Project Design The Panel Report indicates that important studies were done to analyze the question o f climate change. At the same time, the Panel discovered that the conclusion was drawn fiom this analysis, as presented to the Board o f Directors in the PAD, was that "[. . .] there will be no adverse effect on water release due to climate change during the life of theproposedproject." The Panel is troubled by this conclusion- - it failed to include a risk or uncertainty factor, was inconsistent with the underlying analysis, and appears to provide an overly optimistic reading o f the potential effects o f climate change. The Panel considers that climate change requires a change in mindset towards thinking in probabilistic rather than deterministic terms, recognizing the inherent uncertainty that surrounds climate related issues, and avoiding categorical, deterministic statements. The approach noted above i s not inline with the objectives ofBankpolicies in support o f informed decision-making. The Panel notes, in this regard, the Bank's increased role in supporting action to address climate change, and its systems-level efforts to ensure that climate change risks are mainstreamed and integrated into Bank's strategic analysis and project decision making. The proper reporting of risks is of central importance in this larger context. Timely Disclosure of Information within the Project Cycle The Requesters have expressed concern that it was not possible for them to bring the Request at an earlier time because o f the lack of transparency and disclosure during the discussions o f reviving plans for a second round o f investment in the Bujagali dam project. This point finds support in the record o f disclosure o f Project documents. Project files show that the Bank was involved in the preparation o f this Project since early 2005. However, the Project Information Document, which i s supposed to be issued early inthe lvii Project cycle to provide factual information to the public about a project as it evolves, was not issued until January 30, 2007. The Project appraisal took place shortly thereafter inMarch 2007, and the Board approved the Project on April 26 ofthe same year. While the Panel notes ongoing efforts to streamline procedures, this should not be at the expense of providing adequate information to the public in a timely way. Related to this, the Requesters have also raised concerns about the implications o f the Project moving forward to such a degree during the investigation o f their claims, which they note mightresult insignificant issues o fnon-complianceand harm. The Panel observes that these concerns have given the impression to affected people that the Project i s afait accompli, notwithstandingthe possibility of findings of non- compliance and harm. The Requesters have expressed concern that this could prevent the Project fkom addressing significant findings inthis regard. The Panel notes that this i s an important process and systemic issue raised by the present Request, particularly in projects where it i s alleged that irreversible harm may occur as a result of Bank's non compliance. Transparency Issues and Public-Private Partnerships During its field investigation, the Panel noted considerable concern among Ugandan citizens and a number o f their representatives about the lack o f transparency on the economic impacts o f the project. While realizing the complexity o f this project, and the resulting agreements that were made between private and public partners, it i s o f concern to the Panel that so little i s known about the impact o f these agreements not only by the average Ugandancitizen, but also by persons inposition to comprehend the implications o f the various arrangements made. Given the increase in private-public partnerships, and issues relating to access to information in this context, IBRD and IDA might incur reputational risks that are thus far not adequately handled. Similar issues were raised with regard to the prior Bujagali project and other projects reviewed by the Panel inthe past. In this regard, the Panel notes the importance of clarifying Bank policy concerning the disclosure of all project-related documents. This i s of particular relevance in public-private partnership projects where some of the documents may be concluded among private parties relying on Bank financial support. Inthe present context, the Panel found that there was anundulyoptimistic assessment o f the costs, benefits and risks o f the Project, including under-estimation o f its capital costs, o f its likely impact on tariffs, and o f key risks. In each case, Bank Management was substantially dependent on the work o f others. The Panel also found that the assessment o f alternatives was insufficiently transparent, making it difficult for Bank Management authoritatively to address claims that it was inadequate and biased in favor o f the Project. As it stands, the net benefits of the Project could be substantially less than Bank Management has claimed. lviii Critical Natural Habitats and Sacred Places- Guidance to Staff As described above, OP.4.04 defines critical natural habitats to include existing and proposed protected areas, "areas initially recognized as protected by traditional local communities (e.g., sacred groves)" and sites that maintain conditions vital for the viability o f these protected areas. Internal guidance to staff for the application o f the Natural Habitats policy, by comparison, describes "critical natural habitats" as "those Natural Habitats which are either legally protected, ofJicially proposedfor protection, or unprotected but of known high conservation value." Inpractice, this particular guidance seems to suggest a more limited interpretation and application o f the policy than a plain reading o f its terms would warrant. As a result, areas recognized as sacred and protected by traditional local communities, but considered to be lacking a uniquebiodiversity and/or official protection, may not have been regarded as "critical natural habitats." As described inthe Panel's Report, the Project provides an illustration o f an overly restrictive application o f the Policy that puts the Bank at risk o f a serious violation o f its policy. The Panel notes that, in contrast to this apparently narrow application o f the Policy, there i s a strong and increasing recognition over the years, for example through the IUCN process, o fthe importance o f sacredplaces both for their spiritual and cultural values, and for and as part o f broad conservation objectives, both individually and collectively. The Panel also addressed these same provisions o f OP 4.04 in its recent investigation o f the Cambodia forest project. In that Investigation Report, the Panel noted the presence o f spirit forests and spirit trees important to the cultural identity o f local people, and stated that ``. . .Thus, there are many areas within the general forest estate that need to be considered as critical natural habitats..." [emphasis added] The Management Response to the Panel's Report does not dispute the Panel's finding. The Panel considers that such internal guidance given to staff working in Bank- financed projects involving natural habitats and possibly critical natural habitats, like the current Project, may have sent an inadequate and overly-narrow signal on the application of the Policy. Project stakeholders would benefit from clarification on these matters. lix ChapterI Introduction A. EventsLeadingto the Investigation 1. On March 5, 2007, the Inspection Panel (the "Panel") received a Request for Inspection (the "Request") dated March 1, 2007, related to the Uganda: Private Power Generation Project, also known as the Bujagali Hydropower Project (the "Project" or the "Bujagali Project") (Guarantee N0.B-0130-UG). The Ugandan National Association o f Professional Environmentalists (NAPE) and other local organizations and individuals (collectively, the "Requesters") submitted the Request to the Panel.2 2. The Inspection Panel registered the Request and notified it to the World Bank Board of Executive Directors and to Management on March 7, 2007.3 OnApril 5, 2007, Management submitted its Response to the Request (the "Management Re~ponse").~ 3. The Project provides for the construction o f the Bujagali hydropower plant on Dumbbell Island on the Nile River, which has its headwaters in Lake Victoria. The Bujagali Dam will be located about 8km downstream from the existing Nalubaale and Kiira Hydropower Plants. The Project i s to be implemented by Bujagali Energy Limited (BEL), a private sector ~ o m p a n yThe Project's main . ~ objective i s to provide least-cost power generation capacity that will eliminate power shortages at the time o f its commissioning. The Project would represent an increase of250 MW o f installedpower generation capacity to the national grid. Request for Inspection Re: Lodging a Claim on the Proposed Bujagali Hydropower Dam and InterconnectionProjects inUganda, March 1,2007, (hereinafter "Request"). The Request i s available at the Panel's website: http://www.inspectionpanel.org. The Notice ofRegistrationis available at http:/lwww.inspectionpanel.org. Bank Management Response to Request for Inspection Panel Review of the Uganda: Private Power Generation Project (Proposed), April 5, 2007, (hereinafter "Management Response"). The Management Responseis available at http://www.inspectionpanel.org. The Project Appraisal Document (PAD) describes BEL as "a special purpose company incorporated under the laws of Uganda by the project sponsors, who will be responsible for financing, building and operating theproposedproject on a Build-Own-Operate-Transfer basis. BEL will sell electricity to UETCL under a 30 year PPA. Theproject sponsors are: (a) Industrial Promotion Services (Kenya) Ltd. (IPS (K)) 7, the Kenya subsidiary of IPS, the industrial development arm of the Aga Khan Fund for Economic Development (APED); and (b) Sithe Global Power LLC (US) (Sithe Global), an international development company formed in 2004 to develop, construct, acquire and operate strategic assets around the world, which is controlled by Blackstone Capital Partners, an affiliate of the Blackstone Group. Reservoir Capital Group, LLC, a privately held investmentfirm, and Sithe Global's management are also Sithe Global's shareholders." Project Appraisal Document for the Private Power Generation (Bujagali) Project inthe Republic o fUganda. Report N o 38421-UG, April 2,2007 (hereinafter "PAD"), p. 29. 1 4. This i s the second effort o f the Government o f Uganda (GoU) to develop the Bujagali Hydropower Plant. As described in more details below, in 2003 the earlier Bujagali Project was abandoned and the Government terminated its agreements with the World Bank Group, the other financiers and the sponsor AES Nile Power (AESNP)6, a US company. 5. The dam would create a reservoir that floods an area o f 388 hectares, requiring the taking o f 238 hectares o f land to construct the dam, 52 hectares for transmission lines, all o f which will involve displacement and resettlement o f people and family from their lands7The dam's reservoir would also inundate the Bujagali Falls and other natural habitats, which are sites o f cultural and religious significance to the Busoga peoples. 6. The Request raises a number o f environmental, hydrological, social and economic concerns related to the Project as designed, and contends that a failure o fthe Bank to follow its own operational policies and procedures in the design and appraisal o f the Project will result in serious harm to the people living in the Project area and to the environment, in particular the Nile River and Lake Victoria. Management indicates in its Response that it takes seriously the Requesters' concerns. It also states that the Project was well prepared and the Requesters' concerns properly addressedincompliancewith the applicable Bank policies. 7. The Requesters' claims and Management Response are briefly summarized below and thoroughly examined inthe following chapters o f this Investigation Report.* 1.The Request 8. The Request raises a number o f concerns regarding the Project, in relation to: hydrological risks and climate change; environmental assessment, cumulative impact assessment and terrestrial and aquatic fauna; the proposed Kalagala Falls offset; economic analysis, options, and affordability assessment; information disclosure, transparency and openness regarding the Project; dam safety; indigenous peoples, cultural and spiritual issues; compensation, resettlement and consultations. 9. According to the Requesters, the claims they present in the Request constitute a violation o f several Bank Operational Policies and Procedures, including OP/BP 4.01 (Environmental Assessment), OP/BP 4.04 (Natural Habitats), OP/BP 4.02 (Environmental Action Plans), OP 4.07 (Water Resource Management), OP/BP 4.10 (Indigenous Peoples), OP/BP 4.11 (Physical Cultural Resources), OP/BP 'Alsoprocesso referred to as "AES" inthis Report. The f resettlingthese people commenced in2001 at the time of the previously proposed Bujagali dam project, as discussed inChapter VI1 (Social Compliance-Involuntary Resettlement, pp. 137-8) even though the dam was not constructed and no flooding had yet occurred. The Panel notes that in August 2005 NAPE published a report restating its concerns about the Project. This report canbe found at http:llwww.ifitransparency.orgldoclnapereport.pdf. 2 4.12 (Involuntary Resettlement), OP/BP 4.37 (Safety o f Dams), OP/BP 7.50 (Projects on International Waterways), OP/BP (Economic Evaluations o f Investment Operations), OP 1.OO (Poverty Reduction), and World Bank Policy on Disclosure o f Information. 10.InDecember 2001, the Board o f Executive Directors approved an IDA Guarantee to support an earlier proposal for the Bujagali Hydropower Project.' InJuly 2001, before Board approval, NAPE submitted a Request for Inspection to the InspectionPanel inrelation to this previous Bujagali proposal and the Owen Falls Extension (Kiira). After approval on a non-objection basis by the Board o f Directors, the Panel conducted an investigation o f the issues raised in the 2001 Request.lo 11. The Bank cancelled the IDA Guarantee after AESNP pulled out o f the Project. The Requesters noted, inter alia, that "performance shortfalls, controversies related to social, economic and environmental aspects, evidence of corruption"' had contributed to the cancellation. The Requesters claim that due to increased electricity demands and the inability o f Nalubaale and Kiira to supply enough electricity to meet those demands, the GoU "has revived and is in the process of fast-tracking the Bujagali hydropower dam project under different proponents ...this has resulted in many shortcuts being taken to ensure that the project is approved as ast as possible, ignoring outstanding and new concerns raised on theproject."' zf 12. The Requesters raise several concerns related to hydrological risk, climate change, and cumulative impact assessments. They claim that the Project's Social and Environmental Assessment (SEA) does not address hydrological changes and their effect on power production. The Requesters claim that Kiirahas contributed to an over-drawing o f water from Lake Victoria and that the SEA does not address the long-term health o f Lake Victoria. They claim that changing hydrology may be a major limitation on Bujagali's power production and that the SEA does not examine the potential impacts o f climate change, which they claim, will lead to drier conditions, lower lake levels, and therefore lower power production. The SEA also lacks an analysis o f the cumulative effects o f having a cascade o f dams along the Nile. Finally, the Requesters claim that because the analysis i s based on "flawed assumptions and c~mputations"'~related to hydrological risks, the Project's economic viability i s at risk. In addition to these concerns, they claim that the guarantee that the Kalagala Falls will be put aside as an offset and not be developed for hydropower i s not bindingon the GoU. Also referred to as the "prior BujagaliProject." loInspection Panel Investigation Report, Uganda: Third Power Project (Credit No. 2268-UG), Fourth Power Project (Credit No. 3545-UG), and Bujagali Hydropower Project (PRG B 003-UG), 23 May 2002. The Report is available at the Panel's website: http://www.inspectionpanel.org. 11 '*Request,p. p. 1. Request, 1. l3Request,p.4. 3 13. The Requesters also claim that the SEA does not adequately consider, inter alia, Project alternatives such as small hydro, the Karuma Dam, geothermal, efficient lighting, and wind power. In their view, the analysis o f alternatives was overly pessimistic while the hydrological data for the Bujagali Project was overly optimistic. The Requesters also raise concern that the electricity from Bujagali will not be affordable, will not meet the needs o f the majority o f Ugandans, and will reduce the money available for rural electrification. 14. The Requesters raise other concerns as well. They claim that the Power Purchase Agreement (PPA), a key agreement related to the Project's economic viability, was only recently released and they allege that the public version inKampala was not the actual version used to negotiate loans for the Project. The Requesters say that no evidence exists that the PPA was debated and approved by the Ugandan Parliament. Related to safety, the Requesters claim that the Sponsor has failed to adequately address dam safety issues or determine whether Bujagali would be able to withstand a failure o f the Nalubaale dam. Furthermore, the Requesters claim that the Project did not recognize the presence o f indigenous peoples inthe Project area nor did the Project deal sufficiently with cultural and spiritual issues. The Requesters say that the compensation and resettlement frameworks need to be updated to reflect the current economic situation and that the Sponsor needs to create a detailed compensation and detailed community development action plan. Furthermore, the Requesters raise concern over the consultation process, the use o f data, which they claim i s old and inconsistent, and the quality o f the Environmental Impact Assessment (EIA) related to fauna. The Requesters also complain that the World Bank failed to respect the Constitution o f Uganda because it did not consider the Busoga people living in the Project area as Indigenous Peoples for purposes o f this Project, and thus did not apply the provisions o f OP 4.10, which is aimed at protecting vulnerable minorities in Bank-funded Projects. 15. In addition to the letter from NAPE, the Request also includes a letter from the people who were displaced by the prior Bujagali Project and resettled in the Naminya area. The Naminya residents claim that they were promised many things as part o f the terms and conditions o f their resettlement, but that many o f those promises remain unfulfilled. They outline unfulfilled promises and problems related to land titles, a primary school, a health center, water, housing, latrines, electricity, sources o f income and food, a community center, a market, environmental protection, employment, and infrastructuremaintenance. 2. ManagementResponse 16. Management submitted its Response to the Request for Inspection on April 5, 2007. Management maintains that the Project is being developed to provide needed power generation capacity in a least-cost manner. Inthe past three years, Uganda has been suffering severe power shortages due to lack o f generation capacity, prolonged drought inthe region, increases in annual electricity demand, 4 and technical losses in the distribution system. Management indicates that currently the country's growth i s strained by the electricity crisis, which has caused routine power cuts affecting small and large businesses. Management maintains that the Project is expected to eliminate power shortages by 2011 by providing an additional 250MW o f generation capacity to the national grid. 17. With respect to the Inspection Panel's 2002 investigation o f the prior Bujagali Project, Management notes that an action plan was prepared and approved by the Board on July 17, 2002. The action plan related to the sectoral EA, cumulative impacts, the Kalagala offset, load forecast scenarios, affordability risks, power generation alternatives, a socio-economic survey, the community development plans, and compensation for tourism aspects of the Panel's investigation. The response includes a matrix describing the 2002 Panel's investigation findings and the status o f implementation o f the action plan. Management notes that if the Bujagali Project had been constructed under the prior Project, the reduction in Lake Victoria water levels due to over-drawing may not have happened and power would have been produced at a lower cost than Uganda i s currently paying for supply from thermal plants. InManagement's view, the Project i s overdue and Uganda i s paying a high price for the delay brought about by the failure o f the prior attempt. However, Management maintains that the GoU has learned lessons from the prior experience and the GoU i s better able to understandthe concerns o f stakeholders. 18. In response to the issues raised by the current Request, Management states that they take the Requesters' concerns seriously and that they believe the Project "adheres closely to Bank policies and more importantly, that the project developers and jnanciers have been conscientious in pursuing the welfare of project affected persons as well as Uganda as a whole."14 Related to Kalagala Falls, Management claims that the GoU has reiterated a commitment to the offset as part o f the Indemnity Agreement and that Management will engage with the GoU prior to the termination o f the Indemnity Agreement about identifylng mechanisms or instruments to continue the GoU obligation for the Kalagala offset. Management also reports that a Dam Safety Panel was created to provide advice and ensure consistency with Bank policy and that the Project's legal agreements require the preparation o f an Emergency Preparedness and Response Plan (EPRP), which includes failure scenarios for Nalubaale, Kiira and Bujagali dams. 19. On the social concerns raised, Management acknowledges that the past resettlement program was not completed. To address these issues, the Assessment o f Past Resettlement Activities and Action Plan (APRAP) and Community Development Action Plans (CDAP) were undertaken to assess and address the current conditions. Management indicates that BEL and the Bujagali l4 Management Response, 5147. 5 ImplementationUnit (BIU)'' were resolvingoutstanding issues. Inresponse to the claims that the Busogapeople inthe Project area should be considered indigenous people, Management asserts that the Busoga are not considered indigenous people under the Bank's Indigenous Peoples Policy definition. 20. Management maintains that experienced Bank staff and consultants were engaged to work on the preparation o f this Project and that economic, financial, safeguard, technical and other analyses were done to a high standard. Project analysis considered a wide range o f electricity demand scenarios and the impacts o f both low and high hydrology scenarios. Management regards the environmental and social work carried out thus far to have appropriately considered the issues that emerged inthe previous Bujagali investigation and the new issues outlined inthe current Request related to resettlement, cumulative impacts, and consultations. They note that the environmental and social documents were disclosed along with the economic and financial analysis on December 21, 2006. Additionally, Management claims that Project preparation took into account the Inspection Panel's investigation findings o f the issues raised in the 2001 Request for Inspection. 21. Management claims that the Project will bring several benefits. Providing least- cost power i s expected to increase the number o f connections o f residential users to the national grid, including in rural areas, and will allow industrial and commercial users to increase output and efficiency and, therefore, profits. The Project will bring local job opportunities during construction and tourism development in the Kalagala offset. Additionally, the Project i s expected to have environmental benefits since the same water already released through Nalubaale and Kiira dams will be used for Bujagali, thereby reducing the pressure to over- extract water from Lake Victoria. 3. Eligibilityof the Request 22. To determine the eligibility o f the Request and the Requesters, as set forth inthe 1993 Resolution establishing the Panel16and the 1999 clarification^,'^ the Panel reviewed the Request for Inspection and Management Response. The Panel Chairperson at the time, Prof. Edith Brown Weiss, together with Executive Secretary Peter Lallas and expert consultant Eduardo Abbott visited Uganda from April 18-25, 2007. During their visit, the Panel Team met with the Requesters, other members o f civil society and locally affected communities, Bank staff, national and local authorities, Project authorities, members o f Parliament and others. l5The BIU is a unit o f the Uganda Electricity Transmission Company Limited (UETCL), the country's national transmission company. The BIU i s responsible to monitor the resettlement program under the Project. l6InternationalDevelopment Association (IDA)Resolution 93-6, dated September22, 1993. l7Conclusions o fthe Board's Second Review o fthe InspectionPanel, April 1999. 6 23. The Panel determined that the Request fulfilled the eligibility requirements for inspection. The Panel recommended an investigation to the Board o f Executive Directors because the Request and the Management Response contained conflicting assertions and interpretations o f the issues, facts, compliance with Bank policies and procedures, and actual and potential harm. 24. On May 18,2007, the Board approved the Panel's recommendationto conduct an investigation into the matters alleged inthe Request for Inspection. The Request, Management Response, and the Panel's Report and Recommendation were made public shortly after the Board authorized the inspection sought bythe Requesters. 4. The Investigation 25. The purpose o f the investigationwas to establish whether the Bank complied with its own policies and procedures inthe design, appraisal and implementation o f the Project, and whether, if instances o f non-compliance were found, they caused, or were likely to cause, harm to the Requesters and the people they represent. 26. The Panel conducted a two-part investigation. The first part involved detailed research into Bank records related to the Project, interviews with Bank Staff both inWashington DC and inKampala, Uganda, and a review ofrelevant documents and scholarly literature. The second part took the form o f an in-country fact- finding visit. To assist in the investigation, the Panel retained five consultants, who are internationally recognized experts on the environmental, social, economic and technical issues raised inthe Request. The Panel was assisted inits investigation by Prof. Theodore Downing, anthropologist, Prof. Richard Fuggle, environmental specialist, Mr. Graham Hadley, economic and commercial consultant, Prof. Peter Pearson, economist and Prof. Carlos Tucci, hydrologist. 27. Panel Chairperson Werner Kiene, Panel Members Tongroj Onchan and Roberto Lenton, Executive Secretary Peter Lallas, Operations Officer Serge Selwan, and the expert consultants Ted Downing, Richard Fuggle, Peter Pearson and Carlos Tucci visited Uganda from November 27 until December 7, 2007. During its mission, the Panel met with Requesters and other individuals and communities, local and national government authorities, representatives o f the Busoga Kingdom, spiritual and religious leaders, representatives o f civil society, and representatives o f inter-governmental organizations, relevant experts and others. The Panel also interviewed Bank Staff in Washington, D.C. and Kampala. In addition to Kampala, the Panel visited the towns o f Entebbe and Jinja, the villages and areas o f Mutundwe, Kigwanya, Nakuwade, Nansana, Wasiko, Nimanye, Naminya, Kalagala, and Bujagali falls, the dams o f Kiira and Nalubale, and Mabira forest. 28. This Report presents the results o f the Panel's investigation regarding the issues the Requesters raised intheir submission to the Panel. 7 29. Collaborationwith the Compliance the African Development Bank ( Request for Inspection to the World Bank Inspection Panel as well as the CRMU o f the AfDB, as the AfDB i s co-financing the Project. The Panel and the CRMU coordinated their field investigations o f the Bujagali Project and shared consultants and technical information duringthis investigationinorder to enhance the efficiency and cost effectiveness o f each of their investigations. This collaboration between the Panel and the CRMU worked to the mutual benefit o f both parties. Each Panel focused its compliance review on its own policies and procedures and each Panel has made its own independent judgments about the compliance o f its Management and staff with its own policies and procedures. Accordingly, while there may be common elements and language in the respective reports, the findings in this report are based on the independent judgment o f the Panel and exclusively on the World Bank's Operational Policies and Procedures. Picture 1Panel CRMUMeeting with Requesters - 30. The Panel wishes to express its thanks and appreciation to the CRMU for this fruitfulandprecedent-setting cooperation. 5. BankOperationalPoliciesandProceduresApplicableto the Project 31. With respect to this Project, the Panel assessed whether the Bank complied with the following applicable Operational Policies and Procedures: OP 1.00 Poverty Reduction OP/BP 4.01 Environmental Assessment OP/BP 4.02 Environmental Action Plans OP/BP 4.04 'Natural Habitats 8 OP 4.07 Water Resource Management OP/BP 4.10 IndigenousPeoples OP/BP 4.11 Physical Cultural Resources OP/BP 4.12 Involuntary Resettlement OP/BP 4.37 Safety o fDams OP/BP 7.50 Project on InternationalWaterways OP/BP 10.04 Economic Evaluations o f Investment Operations World Bank Policy on Disclosure o f Information B. Requestfor Inspectionrelatedto the prior Bujagaliproject 32. In 2001 the Panel received a Request for Inspection related to the prior Bujagali Hydropower project. The Request for Inspection related to the Uganda: Bujagali Hydropower Project concerned three Projects: the Third Power Project, the Fourth Power Project and the then proposed Bujagali Hydropower Project.18 The Third Power Project, referred to as the Owen Falls Extension (now known as Kiira), supported by IDA, included the construction o f a powerhouse, the installation o f two 40-megawatt generating sets, the provision o f remedial works at the Owen Falls Dam, and the provision o f technical assistance to the Uganda Electricity Board. The OwenFalls Extension i s now known as Kiira and with the Owen Falls dam, known today as Nalubaale, forms the Nalubaale-Kiira system addressed in various parts o f this Report. The Fourth Power Project, financed by IDA as well, aimed at expanding Uganda's power supply to meet the country's electricity demand and to strengthen its capabilities for managing the energy reform and privatization process. The Bujagali Hydropower Project involved the joint participation o f the IDA and the International Finance Corporation (IFC) to develop a 200-megawatt run-of-the-river power plant at Bujagali Falls, a small reservoir, and a rock fill dam spillway, as well as the construction o f approximately 100kilometers o ftransmission lines and associated substations. 33. At the time the Request was filed, the World Bank Group's package o f financial assistance had not been approved. The Bujagali Project's financing plan envisioned an equity contribution o f US$111.3 million from a private sponsor, the AES Corporation, as well as contributions from other financiers such as the AfDB (US$55 million) and export credit agencies (US$219.5 million). The proposal envisioned that a privately owned and operated Project company, AES Nile Power (AESNP), would construct the hydropower plant on a build-own-operate- transfer basis and would sell electricity to a fdly state-owned company under a 30-year power purchase agreement. 34. Project preparation raised strong concerns from parts o f the country's civil society because of the cultural and spiritual significance o f the Bujagali Falls to the Busoga peoples, the involuntary resettlement o f people living on the shore of the These projects, subject to the 2002 Investigation Report, are hereinafter referred to as the "prior Bujagali project" or the "prior project". 9 falls and along the transmission lines, and other reasons. Inthis environment, the Request for Inspectionwas submitted to the Panel on July 27,2001. 35. The Request: The Requesters claimed that the Bank's failures in the design, appraisal, and implementation of the prior projects had materially affected their rights and interests and were likely to jeopardize their future social, cultural, and environmental security. More specifically, they alleged that the Owen Falls Extension and the proposed implementation o f the Bujagali Hydropower Project had resulted-r were likely to result-in social, economic, and environmental harm to the local population, such as negative effects on tourism activities, adverse impacts on fisheries, and increased electricity tariffs. The Request cited the failure to require an environmental assessment o f the Owen Falls Extension, the lack o f a cumulative environmental assessment related to the existing and proposed dams, and an inadequate involuntary resettlement plan (including inadequate compensation arrangements). The Request questioned the Bank's supervision o f the involuntary resettlement of people inthe project area. Some o f the displaced people claimed that they had been intimidated so as to help guarantee their support for the project. Moreover, the resettlement process had already started in 2000 and it was not brought into compliance when the Bank approved financing for the project. The Request also claimed that the economic and technical analysis, especially the analysis o f economic alternatives, and particularly with respect to the Owen Falls Extension, was inadequate. They also alleged inadequate consultation and disclosure o f information. According to the Request, the World Bank was not in compliance with its own policies and procedures on Environmental Assessment (OD 4.01), Involuntary Resettlement (OD 4.30), Natural Habitats (OP/BP 4.04), Safety o f Dams (OP 4.37), Poverty Reduction (OD 4.15), Indigenous Peoples (OD 4.20), Forestry (OP 4.36), and World Bank policy on Disclosure o f Information. 36. After determining the eligibility o f the Request and Requesters, the Panel recommended that the Board o f Executive Directors authorize an investigation o f the matters raised in the Request. On October 26, 2001 the Board approved the Panel's recommendation. On December 28, 2001, while the ongoing investigation was underway, the Board approved a guarantee facility not to exceed US$115 million to support the Bujagali Hydropower Project. 37. The Panel Investigation Report: The investigation focused on environmental, economic, social, and spiritual issues regarding the prior projects to determine whether the Bank followed its own policies and procedures. 38. The Panel's investigation covered issues related to cultural property, particularly in relation to recognized spiritual forces in the project area. The Panel acknowledged the Bank's efforts at consulting local people and religious leaders, as well as the good faith attempts to mitigate the cultural consequences o f losing the Bujagali Falls, which have a highly religious significance for Uganda's Busoga people., However, the Panel expressed concern that no arrangements had 10 been made to minimize the realistic possibility o f disturbance to the local communities. 39. The Panel's investigation revealed that the socio-economic baseline studies were deficient, displaced peoples were not involved in the preparation o f the environmental assessment for the Third Power Project and an environmental advisory panel was not instituted and consulted. The Panel therefore found only partial compliance on the part o f the Bank with its policy on environmental assessment (OD 4.01). The Panel's Report also pointed out that the Bank had failed to perform a sectoral environmental assessment for the Third Power Project, which not only constituted a violation o f the terms and conditions under which the Board o f Executive Directors had approved the credit, but also a failure to comply with the sectoral environmental assessment requirement o f OD 4.01. The Panel also found that the Bank was not in compliance with O D 4.01 with respect to the Bujagali Hydropower project. In this context, the Panel expressed concerns that a cumulative impact assessment o f hydroelectric projects on the Nile was not properly completed. In terms o f dam safety issues, the Inspection Panel found the World Bank in compliance with its policy (OP 4.37). Under the Fourth Power Project, the Panel found that appropriate consultations were not carried out to meet the requirements o f the applicable Bank policies. 40. Another topic discussed in the Panel's report concerned the protection o f the Kalagala Falls as a natural habitat in view o f its religious, cultural, and tourism importance. The Panel concluded that the Bank had failed to ensure adequate mitigation measures to preserve Kalagala Falls as an offset, thereby failing to comply with its policy on Natural Habitats (OP/BP 4.04). 41. The InspectionPanel also focused on the economic and financial appraisal o f the prior project and found that the forecast o f future electricity demands and the analysis o f tariff affordability used by the project were flawed and, therefore, not consistent with the Bank's policy. The Panel raised several concerns as to whether sufficient consideration was given to project alternatives, to the project's risks and the mitigation thereof during the appraisal o f the project. The Panel identified, as a key area o f concern, the narrow range o f the load forecast. It concluded that a wider range o f the load forecast would have been needed to fully satisfy the requirements o f the Bank's policy on economic evaluation o f investment operations (OD 10.04). Furthermore, the Panel also found that the economic appraisal lacked a thorough examination of the institutional risk o f a delayed or under-performing privatization o f the distribution system and its impact on the robustness o f the prior Bujagali project's affordability. Such an examination was needed for full compliance with OD 10.04. 42. Another concern raised by the Panel was related to the power purchase agreement between the GoU and the private investor, AESNP. Included in the agreement was a clause which required the Ugandan government to buy all the power that could potentially be produced, based on the plant's capacity for 30 years, 11 regardless o f whether the power was actually produced or needed. In this regard the Panel also highlighted two strategic risks o f the agreement to the Ugandan Electricity Tariffs Committee and its guarantors: (a) the shortfall in the projected demand for electricity; and (b) the non-affordability o f the electricity rates. The report also suggested two possible additional risk mitigation measures to provide flexibility as well as a mutually acceptable way o f sharing and reducing stranded costs. 43. The Panel questioned whether a depreciation o f the Ugandan currency against the U.S. dollar, leading to an increase in the electricity tariff, would be affordable for Uganda's population and pointed out that the effects o f any currency depreciation should have formed part o f the risk analysis with regard to affordability in the prior project appraisal document. 44. The Panel noted serious problems in the initial implementation o f the Resettlement Action Plan (RAP) especially in the determination o f legitimate claimants and the valuation o f land and crops. It also found that the Bank's community development program set out neither long-term targets nor projects for institution building. In this respect, the Panel found the Bank not in compliance with its policy on involuntary resettlement (OD 4.30). 45. Regarding cultural and spiritual issues, the Panel indicated the efforts o f the Bank and Management's good faith attempts to mitigate these issues. At the same time, the Panel also noted the importance o f including all key stakeholders in consultation and taking steps to minimize the possibility o f disturbance to the local communities that might arise from excluding any faction from such consultations as the prior project went forward. With respect to the indigenous peoples policy, the Panel found the Bank's policy on indigenous peoples (OD 4.20) did not apply. 46. The Request also alleged that the Bank failed to disclose relevant documents related to the prior projects. The Panel found that by failing to disclose the November 2001 report entitled "Economic Review of Bujagali Hydropower Project," the Bank had failed to comply with its obligation under BP 17.50 on disclosure o f information. 47. The InspectionPanel found no evidence o f serious efforts on the part o fthe World Bank to actively engage with project-affected groups or NGOs and accordingly found that the Bank was not in compliance with its policy on environmental assessment with respect to public consultations regarding the Fourth Power Project. 48. Management Action Plan in response to the Panel's findings: In response to the Panel's findings, the Management o f the Bank proposed inits report a number o f actions to remedy instances o f noncompliance. The actions included a commitment to amend the agreement between Uganda and the Bank with respect 12 to the protection o f the Kalagala Falls. Furthermore, the GoU reaffirmed its commitment not to develop the Kalagala Falls for hydropower but to set it aside exclusively as a natural habitat and for tourism. Management also agreed to provide support for multi-stakeholder consultations on the three hydropower Projects and to promote informed and comprehensive discussions. 49. In its response to the Panel's investigation report, the Bank Management also affirmed its support for a strategic and sectoral environmental assessment, as well as social assessments that would be a prerequisite to any future World Bank financing o f investments in Uganda's power generation facilities. The Bank would also monitor future growth inelectricity demand and the implementation o f agreements to be signed with tourism operators. The Bank would further support measures to address reemployment o f Ugandan citizens affected by loss o f tourism-related jobs. Management agreed to assist the government in examining alternatives in power generation and proposed financing o f geothermal exploration and possible drilling inwestern Ugandan areas. 50. On the topic o f social issues addressedby the Inspection Panel, the Bank agreed to request that AESNP conduct surveys that would support implementing and monitoring the Project's Community Development Action Plan. 51. Subsequent developments: On June 17, 2002, the Board o f Executive Directors met to discuss the Panel's Investigation Report and the Management Report and Recommendations inresponse to the Panel's findings, and endorsed Management recommendations. However, in 2003 the execution o f the Project was halted due to financial difficulties o fthe Project sponsor. 52. Following Board approval o f the prior project, the project encountered several difficulties which eventually led to a pull-out o f AES and termination o f the project with the Government in September 2003.19 At the same time, the Bank discussed with the Government o f Uganda "its options, transition arrangements including the integrity of the Project site and intellectual property, and the maintenance of a unit to monitor the project, including the environmental and social aspects." 2o In January 2004, the Government "initiated a transparent and competitive process soliciting the interest of prospective private sponsors in the Bujagali Hydropower Project. This led to the selection of a new project sponsor consortium (Industrial Promotion Services (Kenya) of the Aga Khan Group and Sithe Global (US)) in April 2005."21On October 3, 2005, Management sent to the Board a Project Completion Note summarizing the Project and explaining why it was not implemented. 53. The current investigation, addressed in the present Report, focuses on the second round effort to develop and complete the Bujagali Hydropower Project. l9 Project Completion Note, p.1, 5. *'Project 2o Project Completion Note, p.1,516. Completion Note, p.l,17. 13 Chapter I1 The Project anditsContext 54. This chapter presents an overview o f the economic, social and environmental context relevant to the Project and this investigation. The discussion considers the electricity crisis in Uganda and presents an overview o f the environmental and social setting in and around the Project area. It concludes with a more detailed description o f the Project that i s the subject o f this Panel's investigation and o fthe World Bank's involvement in related projects in Uganda, Lake Victoria, and the Nile River Basin. A. Poverty and Energy inUganda: The Power Supply Crisis 55. Uganda i s among the world's poorest countries, with poverty strikingparticularly rural areas. Inrecent years the country has experienced economic growth but the fast-growing population rate (the third fastest in the world) is one of the main challenges to the fbture economic growth o f the country. Poverty has been increasing inrural areas along with a rise ininequality.22 56. A 2006 Bank Poverty and Vulnerability Assessment indicates that high poverty levels stem primarily from limited access to land and other assets, high rate o f disease (though Uganda has made good progress in fighting HIV/AIDS), lack o23f control over productive resources by women, highfertility rates, and insecurity. Though improvements have occurred, the poorest people have still very limited access to essential services, including education, health services, water and sanitation, roads and electricity, especially inrural areas.24 1.Shortages andLackofAccess to Electricity 57. The power sector i s presently experiencing serious capacity constraints inrelation to needs and demand, and Uganda i s facing a major power crisis. This has strained the recent economic growth, as both consumers and businesses, in particular manufacturing and processing industries and high-value agriculture, have suffered prolonged cuts o f service. 58. According to Project documents, the crisis and these severe power shortages are considered to be rooted in four main factors: 1) delay in developing power infrastructure, 2) low levels o f water in the Lake Victoria, caused by regional 22United Nations Consolidated Appeal for Uganda 2008, Dec 10 2007, p.2. Available at http:llochaonline.un.or~lcap200Siwebpage.asp?Page=1632. See also PAD, p. 1 23WorldBank Africa Region, Uganda Poverty and Vulnerability Assessment, Report No. 36996-UG, October 12,2006.13.47 24UNDP, Human Development Report 2007-2008, HumanDevelopment Index. Uganda Poverty Assessment, p. 23. 14 drought and over abstraction o f water for hydropower 3) high levels o f losses o f the power distribution system and 4) a substantial increase o f about 8 percent o f the annual demand for power.25 The Report considers these and other factors in detail insubsequent chapters. 59. Access to electricity in Uganda i s generally very low. Only five percent o f the total population, less than one percent in rural areas, has access to grid- supplied electricity.26 Around 72 percent o f electricity i s consumed by twelve percent o f the population living inthe Kampala metropolitan area, the capital, and in nearby towns, Jinja and Entebbe. Electricity i s very costly, particularly for poorer households. According to a World Bank study, poor urban dwellers consume little ifany electricity, while most rural households are not close to a gridconnection: "electricity use by households in Uganda is stunningly low, but even worse in rural areas."27 The Panel notes the critical importance of providing affordable electricity to the people of Uganda, as an integral element of national development and of Uganda's poverty reduction efforts. 60. Management states that the Government i s addressing the power crisis through a "power sector strategy" which aims at promoting legal, regulatory and structural sector reforms, increasing the role o f the private sector inits operations and future development; providing adequate, reliable and least cost power generation to meet increasing demand and guarantee increased access; and scaling up rural access to electricity. Thus far, the GOU has promulgated a new Electricity Act and established the Electricity Regulatory Authority (ERA) in 1999, established a Rural Electrification Agency in 2002 and granted concessions on power generation and distribution. 61. The Rural Electrification Agency established the Rural Electrification Programme to expand electricity coverage, but the ability o f these communities to afford this electricity may be an issue. As a result, biomass is projected to remain the principal source o f energy for people in rural areas.28 An additional study conducted in 2006 reported that the cost o f connections, especially in rural areas are very high, mainly because o f the low capacity o f the national distribution company, lack o f planning methodology and tools, and lack o f appropriate information and ability to compare technical options.29 62. Issues o f pricing and affordability are critical to access to electricity, inparticular the tariff rates (cost) o f electricity to users (including families and households). 25PAD, p. 4. 26Energy Sector Management Assistance Program (ESMAP), Sub-Saharan Africa: Introducing Low Cost Methods in Electricity Distribution Methods, Technical Paper 104. 106, October 2006, p. 1. [hereinafter, "ESMAP"] 27Uganda - Moving Beyond Recovery: Investment and Behavior Change for Growth, Report No. 39221- UG, World Bank, Sept 2007, V. 1, p. 25. 28UNDP, Uganda Human Development Report 2005, Linking Environment to Human Development: A Deliberate Choice, Section 4.10, p.49. 29ESMAP,p. 2. 15 Chapter VI o f this Report identifies key factors affecting the cost and affordability o f electricity. 63. In the present context, a critical issue raised by the Request i s whether the Bujagali dam, if built, will meet its economic projections and provide affordable electricity to the people o f the country, in comparison to other alternative means for doing so. During its visits to the Project area, the Panel heard strong expressions of concern from local people and their representatives that they will not benefit from the Project butwill, nevertheless, have to bear its social, economic and environmental costs. In addition, they are concerned that, if Project costs are not properly estimated and accounted, the burden of below- capacity production will be passed to the people of Uganda. In their own words, they fear "being takenfor a ride" by a project that does not meet their needs, harms things o f importance to them, and enriches somebody else. This issue i s examined inmore detail inlater sections of this Report. 2. Current and Planned Sources of Electricity 64. Uganda's main source o f electricity i s the Nalubaale-Kiira dam complex, located just below the source o fthe NileRiver inLake Victoria. The complex consists of two separate dams: the Nalubaale dam constructed inthe 1950's across the upper Nile (also referred to as the Owens Falls dam); and Kiira dam, constructed in 2000 in a side-channel artificially created next to the main flow o f the Nile, and nearlyparallel with the Nalubaale dam. 65. The combined potential operating capacity of the Nalubaale and Kiira dams i s 380 MW. Over recent years, however, the actual electricity produced by Nalubaale and Kiira has dropped substantially below capacity, reaching 120MW (equivalent to water discharges of 750m3/s) between August 2006 and 2007. This contrasts with a 380 MW peak system demand and a 290 MW base load demand, only 50 percent o f which i s met by the current power supply. Unmet energy demand in 2006 amounted to 364 GWh.30 66. A key reason that these two dams have been performing so far below their capacity relates to the release o f water into the Nile from Lake Victoria, and the interactions between the dams and the water levels o f the Lake. These issues, which are at the technical core of this Investigation as they relate to the Bujagali Dam, are described inmore detail below. 67. To increase Uganda's hydroelectric power capacity beyond that provided by Nalubaaleand E r a , the Bujagali HydropowerPlant, addressedinthis Report, has been approved and i s under construction. In addition, Karuma dam, a run of the river plant significantly downstream from Bujagali Falls and upstreamofthe limit o f Murchison National Park, has been proposed. Other existing and potential sources of energy include small and micro-hydropower sites, bagasse (cane 30PAD, Annex 1,17, 11, p. 48-50. 16 residue) from sugar factories, biomass resources, geothermal, wind power, municipal solid waste, newly discovered oil resources, as well as approaches to conserve energy and reduce losses at all levels.31 The question o f alternative sources o f energy for Uganda i s considered in some detail in later sections o f this Report.32 68. In2005 and 2006, to increase power supply, the government leased two 50 MW thermal plants, while in 2007 IDA financed an additional temporary 50 MW under the Bank-funded Power Sector Development Operation (PSDO). The PAD also sets out an Interim Generation Expansion Plan from 2006 to early 2011 (when the Bujagali project would be commissioned). About 44 MW o f mini- hydropower capacity and 15 MW o f co-generation (usingbagasse) were planned for commissioning between 2007 and 2009, while reliance on 150 MW o f diesel and fuel oil power generation was expected to continue until 2011.33 B.Environmental and Social Context and Setting 69. The Bujagali dam i s within the Nile Equatorial Lakes region, consisting o f a number o f interconnected lakes providing a natural storage for the Nile River, including Lake Victoria -the largest (69,000kmz) - and Lake Albert, Lake Kyoga and Lake Edward, which are linked to Lake Victoria by the Victoria Nile. The Bujagali hydropower facility would be located on the Nile River about 8 kilometers downstream (north) o f Nalubaale and Kiira and the source o f the Nile at Lake Victoria. Since the flow o f the Victoria Nile i s regulated by Lake Victoria and i s relatively steady from season to season, the Bujagali dam i s designed as a "run-of-the-river" dam.34 70. The sections below provide an overview o f the hydrology o f Lake Victoria and the Victoria Nile and the environmental and social setting relevant to understanding the Project. 1.Hydrology ofLakeVictoria and the Victoria Nile, and Hydropower Implications 71. The hydrology o f Lake Victoria and the Victoria Nile i s a key influence on the potential energy output o f hydropower plants on the Victoria Nile. The Lake's 31The Government has reported that an oil resource was discovered in western Uganda but that no impact on power generation is predicted until2011. 32 The Request and Project documents provide different descriptions and views on the availability and otential o f these alternative sources o f energy. p3PAD, p. 26 and PAD, Annex 1, p. 66. 34 Hydropower projects can be either storage projects or run-of-the-river projects. Storage projects are usually built on rivers with significant variability in flow, whereas run-of-the-river projects suited to rivers with a fairly steady flow. Storage projects aim at capturing river flow during high flow periods and releasing it during low flow periods; run-of-the-river plants, by contrast, rely on a river's natural flow. While run-of-the-river projects sometimes have a small amount o f storage to regulate flow during a 24-hour period to help meet peak power demands, the volume o f water that needs to be stored, and consequently the area that needs to be flooded, i s generally much smaller than that o f storage projects. 17 water balance35 i s the essential link among lake levels, water flows and hydropower production. As explained below, changing lake levels affect water flows inthe Victoria Nile River, which inturn affect hydropower generation. 72. Studies o f Lake Victoria have shown that rainfall i s the principal contributor to the Lake's inflow, and evaporation and outflow via the Victoria Nile are the principal contributors to the Lake's outflow. The Lake's water balance is thus dominated by rainfall over the lake, evaporation from the lake, and outflows via the Victoria Nile. When the net volume o f inflow into the Lake (that is, rainfall plus basin inflow minus evaporation, referred to as the "Net Basin Supply") exceeds the net volume o f outflow via the Victoria Nile, the amount o f water stored inthe Lake and thus the Lake's level will increase. However, when the Net Basin Supplyi s less than the amount outflowing via the Victoria Nile, the amount o fwater stored inthe lake will decrease and the Lake's water level will drop. 73.The Net Basin Supply in any given time period i s determined by climatic variables, and varies considerably from season to season and year to year. The outflows from the Lake to the Victoria Nile, however, are subject to human control. Prior to the construction o f dams on the Victoria Nile, the amount of water flowing from Lake Victoria was naturally determined by the level o f water inthe Lake-the higher the level ofthe lake, the more water that pouredout from the lake into the Victoria Nile. However, the successive development o f the Nalubaale (formerly Owen Falls) and Kiira dams at the entry point from the Lake to the upper Nile changed all that. Before the completion of the Nalubaale dam in 1959, the outflow from Lake Victoria into the Nile was at Rippon Falls, a rock barrier at the outflow point o f the Lake that naturally regulated the water levels. During construction o f the dam in the 1950's, this rock barrier was blasted and lowered, providing more outflow o f water to the dam. Since 1959, when the Nalubaale dam started operating, the dam has regulated the outflow o f Lake Victoria into the Victoria Nile, transforming the Lake into a qua~i-reservoir.~~ At 35 The water balance of a lake (or other body of water) establishes that the difference between the total volume o f water flowing into the lake inany given time periodand the total volume flowing out during that same time period will be equal to the change in the volume o f water stored in the lake during that time period. The volume o f water flowing into the lake generally consists o f surface water inflow (from rivers or streams or direct runoff), groundwater inflow (from aquifers), and precipitation in the form o f rainfall or snowfall on the surface o f the lake, while the volume flowing out will consist o f the evaporation from the lake, storage losses including seepage, and water flowing out via downstream rivers or streams. If inflow volume exceeds outflow volume, the difference between the two will be added to the volume o f water stored inthe lake, while ifthe lake's outflow volume exceeds its inflow volume, the difference betweenthe two will reduce the water stored in the lake by that amount. Mathematically, this relationship can be expressed as C= S+G+P- (E+L+R), where C is the change in the volume of water stored in the lake or reservoir, S is the volume o f surface water inflow (from rivers or streams or direct runoff), G is the volume o f groundwater inflow (from aquifers), P is the volume o f precipitation (rainfall or snowfall) on the surface o f the reservoir, E is the volume o f evaporation Gom the reservoir, L i s the volume o f storage losses including seepage, andR i s the volume o f water flowing out via rivers or streams. 36 Daniel Kull, 2006, "Connections Between Recent,Water Level Drops in Lake Victoria, Dam Operations andDrought." Available at htt~:iiwww.irn.org/programsinileivdDO60208vic.~df 18 that time, it was agreed that the dams must be operated inaccordance with the so- called Agreed Curve, described below. 74. The Agreed Curve i s the result o f agreements between Egypt and the British Government that stipulated how much water should be released from the Lake. The agreements provide that the Nalubaale dam should be operated pursuant to a rating curve (the Agreed Curve)37aiming to ensure that the outflow from the lake mimics the conditions o f the Lake before the Nalubaale dam was constructed. This meant restoring the natural behavior o f the lake by allowing fluctuation in rainfall and evaporation to determine the amount o f water flowing out. With the Agreed Curve, low water levels determine a lower outflow to the Victoria Nile and thus a lower input o f water to the hydropower plants; conversely, highlake water levels determine a higher outflow to the Victoria Nile and thus a higher input o fwater to the hydropower plants. 75. In the 1990s, the need to increase power production led the government to examine possible alternatives for additional power generation. According to 1990 studies, two "feasible" options were identified: the Owen Falls Dam extension project (Kiira), which was designed to operate inparallel with Nalubaale, and the construction, downstream o f Nalubaale, o f the Bujagali hydropower plant.38The former, Kiira, was chosen39;its turbines are only a few meters lower than those o f the Nalubaale and use the same water drop (referred to technically as "head"), from Lake Victoria, plus some additional "head", which results in increased relative production capacity. A canal above Nalubaale diverts water to Kiira in a way that allows the two dams to control the water level and the outflow from the Lake.40Because Kiira operates in parallel to Nalubaale, power generation from Kiira requires additional outflows from Lake Victoria over and above those required for Nalubaale. One study states that when Kiira was built, the discharge into the Victoria Nile was increased "possibly by as much as 50?4", and "it became impossiblefor Uganda to adhere to the Agreed 37 Rating curve is the relationship between river level and flow. The "Agreed Curve" relationship is: Q = 132.924(h -8.486) "1.686 where Q is discharge in cubic meters per second and h i s water level (stage) in meters at the Jinja Pier. 38 Acres International Ltd. "Proposed extension to Owen Falls Generating Station: Feasibility Study Report," Oct. 1990. 39 The Panel's expert considers that this decision was biasedby the hydrologic series that was used. During the period o f 1961-1989 the mean outflow o f water was 1,200 m3/swhile before1961 the mean outflow was 660 m3/s.The hydrological studies analyzing the two possible alternatives, the Owen Falls dam extension (Kiira) and the Bujagali Falls dam, concluded that the flow observed before 1961 would not occur in the h t u r e and, as a result, considered as valid only the hydrological series 1961-1989. This led to the decision o f constructing the Owen Falls dam extension (Kiira) rather than the Bujagali plant and to rely on data showing an averagely high discharge o f water. According to the Panel expert, this solution was less expensive but relied on a mistaken assumption: as noted in the text since 2000 the Lake Victoria's water levels have decreasedand the NalubaaleKiira system's energy output has beenlower than planned. 40 Kull2006, p. 4. It was also inthis context, amidst an electricity crisis and rising demand for electricity, the Bujagali dam was f i s t proposed inlate 1999 early 2000. 41 Kiwango and Wolanski, "Papyrus wetlands, nutrients balance, fisheries collapse, food security and Lake Victoria level decline in2000-2006." Wetlands Ecol. Management, Nov. 2007. p. 90. The study notes that other sources attributed the decrease o f the lake level between 2000 and 2006 "to both lack of rain and 19 Figure 1Water Flows fromLake Victoria to Nalubaale, Kiira and Bi jagali4* G LakeVictoria I 1 1knl I Le Ripon Fa'k I I 2 km NALUBAALE KIIRA I I + BUJAGALI I I 76. Over the past 100 years the water levels o f the Lake Victoria have shown significant changes in regimes. In general, the period before 1960 i s characterized as a period o f relatively low water levels and outflows to the Victoria Nile. Between 1960/61 and 1999, Lake Victoria rose, nearly doubling the average outflows in comparison to the previous period. In contrast, from 2000 until recently, lake levels again decreased to reach a level observed before the 1960s. Experts are divided as to the causes o f the recent drops o f levels o f the Lake Victoria. Bujagali Project documents claim that a "main cause of the dro in lake level in thepastfew years was the exceptionally dryperiod 2003-2005.' & On the other hand, it i s clear that from 2003 to 2005, the water from Lake Victoria was over-abstracted (that is, released above the Agreed Curve) to expand power generation and meet the increased demand for e l e ~ t r i c i t ySome . ~ ~ authors conclude that this over-abstraction i s the main cause o f the low levels of excessive water extraction at Kiira dam, although their relative contribution was not quantified." (90) I t should be noted, however, that had Kiira been commissioned at a time o f sufficiently high Lake levels, the discharge according to the Agreed Curve could have beenhigh enough to permit the operation ofbothKiira andNalubaale at capacity. 42This figure was adapted from the figure "Hydrology and Lake Victoria" on p. 19 of the Technical Briefing "UGANDA: Bujagali Hydropower Project." Presentedby IDA,IFC, and MIGA on April 2,2007. 43Management Response, Annex 111, 1[ 10. Management also states the "analysis o f Lake Victoria water levels during 2003-2005 period concluded that the main origin of the drop in the lake level during this timeframe i s an exceptionally dry period.. ." Management Response, Annex 1, p. 18. 44The PAD, for example, states that during the period 2003 and 2005, ". . . the power demand in Uganda required a sustained release that was above the net inflow, thus accelerating the drop in lake level, and automatically increasing the departure from the Agreed Curve." (PAD, Annex 10, 7 9.) The Request notes, among other things, that "[w]ithout.doubt, Kiira has contributed substantially to the over-draining of Lake Victoria, causing a lot of miseiy and economic loss to Uganda and neighboring countries." (Request, p. 2) 20 the Lake's waters.45 In Management's view, both drought and over-abstraction are to blame; as stated in the Management Response, changing water levels are the "consequence of this low inflow [due to lack of rainfall in the recent dry period], combinedwith the over-release of waterfor power generation."46 77. Since lake levels determine river flows, which in turn determine the amount of water that flows through the power turbines, lake levels and power generation are strongly inter-linked. The World Bank - Netherlands Water Partnership (BNWPP)47 background description for the "Victoria Nile -- Independent Hydrological Review" activity, which was carried out in 2006 as part of the "River Basin Management" window of the Partnership states that the amount of power that can be suppliedby water from Lake Victoria: ". . .depends importantly on the Lake level and its management through the operating regime of the hydroelectric facilities, which until recently dependedupon an Agreed Curvegoverning water releasesfrom the Lake for power production. If Uganda over-draws the Lake for power production, as it has been doing, this could impact on the usable volumes of water relative to what they would have been under the "Agreed Curve" policy. Also, when the Lake level deteriorates, it can affect, and apparently has affected a number of other economic activities that the riparian countries depend upon, such as agriculture, fishing and transportation."48(emphasis added) 78. The same 2006 World Bank - Netherlands Water Partnership source notes that Uganda i s gradually reducing its hydropower output to be more aligned with the Agreed Curve, but inthe meantime "is suffering major day-long power cuts that are adversely affecting economic, public service and household activitie~."~~ In its response to the Request, Bank Management states that "[slince the end of 2005, the GoU has steadily decreased hydropower generation in an effort to 45 Kull2006, estimates that that the level drops are 45 percent due to drought and 55 percent to the over- releases (p. 7). 46Management Response, Annex 1, p. 18. 47 The Bank-Netherlands Water Partnership Program (BNWPP) aims at improving water security by promoting innovative approachesto Integrated Water ResourcesManagement (IWRM), and contributing to poverty reduction. The BNWPP currently operates through a framework of 14 sub-programs or windows. Eachwindow i s a sub-component o f a broad framework that embraces comprehensive, cross-sectoral water management; water-user participation; transparent and efficient institutions; the treatment of water as a social and economic resource; the importance o f water to the natural environment; and the link between water management and poverty alleviation. Available at http://www-esd.worldbank.ore/bnwd (Last accessed on June 30,2008). 48World Bank - Netherlands Water Partnership (BNWPP), background description for the "Victoria Nile- Independent Hydrological Review" activity, htt~://www-esd.worldbank.ore/bnw~/index.cfm?displa~dis~lay~activit~&~D=439, on 23 Julyaccessed 2008 49World Bank - Netherlands Water Partnership (BNWPP), background description for the "Victoria Nile- Independent Hydrological Review" activity, httD://www-esd.worldbank.org/bnwD/index,cfm?displav=disolavactivitv&AID=439, accessed on 23 July 2008 21 return to the Agreed Curve operating regime. Waterflows for power production are being scheduled so as to return to the Agreed Curve as soon as reasonably possible."50 79.A significant question raised by the Request i s the extent to which the proposed Bujagali Dam will or might create incentives to depart fi-om the Agreed Curve, and contribute to a lowering o f Lake water levels and corresponding serious impacts for the Lake's riparian states. This issue, and related issues o f hydrology andwater flow, is addressedinChapter IV o fthis Report. 80. An important related question i s the extent to which the future hydrology of Lake Victoria may be influencedby climate change. Since the Lake's water balance i s dominated by rainfall and evaporation over the surface of the Lake, the Requesters are concerned that even relatively small long-term decreases in rainfall and/or increases in temperature could have significant impacts on Lake levels and on outflows via the Victoria Nile and, in turn, on the economic and politics o f operating the dams. An analysis o f potential climate change effects, and the extent to which they were taken into account in Project analyses in line with Bank Policies, is includedinChapter IV o fthis Report. 2. Lake Victoria and the Impact of Declining Lake Levels 81. Lake Victoria i s the largest fresh water lake in Africa and a most important natural resource. The Lake and its 3,450km o f shoreline are shared by Kenya, Tanzania and Uganda; its basin includes Rwanda and Burundi, which are part o f the upper watershed draining into the Lakethrough the Kagera River. The Lake is part o f the Nile River Basin system that i s shared by ten countries, including the aforementioned countries as well as the Democratic Republic o f Congo, Egypt, Ethiopia, Eritrea, and Sudan.51 82. Lake Victoria i s an inland transport linkage for Uganda, Kenya and Tanzania. It i s also a major natural reservoir and source o f water for domestic, industrial and commercial purposes, serving major cities, towns and urban and rural centers within the basin. The Lake is considered the "largest inland water Jishery sanctuary in Africa", its fishery resources supporting livelihoods for around three million people involved in the fisheries industry.52Lake Victoria and the rivers flowing fi-om it also are seen as a major potential source for hydropower generation, as discussed earlier. Management Response, Annex 1, p. 18. 51World Bank - Netherlands Water Partnership (BNWPP), background description for the "Victoria Nile- Independent Hydrological Review" activity, http://www-esd.worldbank.orgibnwppiindex.cfm?dis~la~dis~lay~activitv&~D=439, accessed on 23 July 2008 52East African Community (EAC), Lake Victoria Basin Commission, Special Report on the Declining o f Water Levels of Lake Victoria, Arusha, Tanzania, January 2006, p. 2, [hereinafter "EAC Report 2006"l. 22 83. Many studies have examined the extraordinary ecology, wildlife and habitats o f the Lake Victoria region, its importance to the life and livelihoods o f its people, and also the history o f change and biodiversity loss in the region.53Over time, factors and pressures on the natural systems include intensified fishing methods, the introduction o f non-indigenous species such as the Nile Perch, pollution and eutrophication o f the Lake itself from agricultural and industrial activity, and the loss ofriverine migratory routes important to potamodomous fishes due to, among other things, the construction o f dams such as those at Owens Falls.54To these shouldbe added a potentially new set and scale o f impacts associated with climate change (discussed later inthe Report). 84. Various studies have also addressed changes that have occurred over time in the lake level, its chemistry, ecology, sedimentation and water quality.55The Lakes Basin Development Authority, established by the Government o f Kenya in 1979 to spearhead development inthe Lake Victoria Basin Catchments area inWestern Kenya, recommends the following measures to restore the health o f the lake: 9 Catchmentprotection - re- afforestation, agro forestry, soil and water conservation and good agricultural practices promotion in the catchments. 9 Develop a sustainable regional conservation and managementplanfor jshery resources. Enhance environmental-friendly fish harvesting practices, protection of breeding sites, enforcement of quality control and all otherfisheries regulations. Rehabilitate and maintain waste treatmentfacilities in all municipalities and industries in the region, so as to reducepollution and eutrophication in the lake. Industries should endeavour to initiate cleaner production technologies as a way of safeguarding andprotecting the environment. 9Develop a long term comprehensive and well coordinated river and lake water quality monitoring programme as a tool for water quality 53Balinva et a1(citing many studies) note the more than 100 endemic species o f large, "magnificent" (705) piscivorous haplochhromine cichlids, most of which vanished at about the time of the increase in the Nile perch; the over 200 species of mormyrids, a family including the elephant-nose fish which i s well known for their "remarkable electrongenic and electroreceptive capabilities" (705); the African lung fish, and others. See also publications listed on the home page o fNaFFIRI at http://www.firi.go.ug/. 54Balinva et al. 2003. "SeeLehmanJT(Ed)."Environmental ChangeandResponseinEastAfricanLakes." KluwerAcademic Publishers, 1998. Specific data on present day Lake Victoria may be obtained from the World Lakes Database maintained by the International Lake Environment Committee. (Available at: htttx//www.ilec.or.iu/database/database.html) This committee has also published a report on Lake Victoria' Issues specific to the health o f Lake Victoria and its Management. Report also appears in a paper presented to the Living Lakes African Regional Conference held inKisumu, Kenya inOctober 2005. See Nzomo, R. "Sustainable Development o f African Lakes, The Case of Lake Victoria." Living Lakes African Regional Conference, Kisumu,Kenya, 2005. 23 management. Regional water quality standards should be adopted and enforced. Enhance water hyacinth control, and eradication of other obnoxious weeds e.g. Striga weed in the region. A long term regional monitoring programmefor this invasive aquatic weed is important. 9 Communities should be made aware on the signijicance of environmental management and conservation. As stake holders they should participate in decision making and implementation of environmental conservation and management projects in the basin. 56 85. Similarly, the InternationalLake Environment Committee records the major threats to Lake Victoria as: "Population pressure, contributing to the existence of "hot spots '' caused by human waste, urban runofi and effluent discharges from such industries as breweries, tanning, paper and fish processing, sugar, coffee washing stations and abattoirs; 0 Nutrient (phosphorous, nitrogen) inflows, including atmospheric deposition, causing a five fold increase in algal growth since the 1960s, resulting in deoxygenation of water that threatens the survival of deep waterfish species; Residual inflows from the use of chemical herbicides and pesticides and, to a limited extent, heavy metals resulting from gold mining operations that cause localised pollution; Proliferation of water hyacinth, resulting in biodiversity and economic losses in the lake's near shore areas; Unsustainable use of the major wetlandsfor agricultural activities and raising of livestock, which has greatly compromised the buffering capacity of the wetlands; and Introduction of two exotic species (Nileperch, Nile tilapia), and use of unsustainableJishing practices and ears, altering the composition of the lake's fauna andflora species.'s B 86.Neither report mentions hydropower generation nor changing lake levels as a problem affecting the lake or deems them responsible for the problems that the 56 Lehman J T (Ed). "Environmental Change and Response in East AGican Lakes." Kluwer Academic Publishers, 1998. ''Kayombo, Sixtus; Jorgensen, Sven Erik. "Lake Victoria: Experience and Lessons Learned Brief." Published by the International Lake Environment Committee as part o f the Lake Basin Management Initiative. Available at http://www.ilec.or.jp/eg/lbmi/pdf/27_Lake_Victoria~27Februan/2006.pdf 24 Lake i s experiencing. Nevertheless, there i s no doubt that lake levels do vary over a range o f almost two meters and that this has a real impact on socio-economic conditions on the lake's shoreline58. 87. Clearly low water levels59have had serious environmental, social and economic impacts on Uganda and the other riparian states and the lives o f about 30 million people using the lake water. Water intakes for Kampala and Jinja are affected, as are docking facilities at Jinja. During the recent period o f low lake levels, fish landing and water supply structures at riparian communities were left literally high and dry and this increased the cost of living and reduced access to clean water. Manypeople began usingnon-purified water from shallow shoreline areas, which poses a health hazard from water based and water borne diseases. In addition, hydraulic and civil structures along the shoreline required modification or began being abandoned due to the low water levels. The PAD describes the impact o fthese recent low water levels inLake Victoria as follows: "Because of low water levels, these benefits have been threatened by environmental degradation manifested in reduced fish stocks, the drying out of fish breeding areas and the loss of livelihood to many fishing communities; a decline of biodiversity; increased sedimentation and nutrient loads resulting in eutrophication; the drying out of wetlands and loss of littoral habitat; increased lake transportation costs, since ports and piers are le t hanging on dry land, and water shortagesfor shoreline towns andfarmers."6 l 88. There are, moreover, serious problems in developing responsive actions to address these problems. The PAD notes that: "EHorts to regulate and manage the activities threatening the lake are clearly insuflcient at present, and widespread poverty in the basin exacerbates environmental stress."61 A study by Kiwango and Wolanski,62 referred to earlier, focused on the potential impacts o f lower water levels inLake Victoria on papyrus wetlands and the nutrient balance around the Lake and concludes by stating: "lfUganda resumes overdrawing water from the lake and permanently dries out the papyrus of Lake Victoria, the resultant eutrophication of Lake Victoria may be large-scale and could also result in the collapse of artisanal fisheries and threaten food security for the impoverished fraction of the population living on the lake's shores, while also possibly 58 Changing lake levels and their effects are well outlined in EAC Report 2006 and Lake Victoria Basin Commission, April 2006 See also Chapter I1Context. 59 Lake Victoria has undergone significant changes in both paleo- and historic-times. An authoritative source documenting changes that have occurred i s Environmental Change and Response in East African Lakes. This volume producedby the International Decade for the East African Lakes discusses changesthat have occurred in inter alia Lake Level, Chemistry, Ecology, Sedimentation and Water Quality. Specific data on present day Lake Victoria may be obtained from the World Lakes Database maintained by the International Lake Environment Committee. This committee has also published a report on Lake Victoria. Issuesspecific to the health o f Lake Victoria and its Management appear ina paper presented to the Living Lakes African Regional Conference heldinKisumu,KenyainOctober 2005. . 6oPAD, Annex I, p. 49. 61PAD Annex 1, p. 49. 62 Kiwango and Wolanski 2007, p. 95 25 exacerbating the infestation of the water hyacinth exotic weed. Global warming may also be accelerated." 89. Low water levels also lead to a decline in electricity generated from hydropower. Load shedding63disrupts industrial activity and reduces revenue from taxation; emergency thermal generation raises the cost o f electricity to consumers.64 3. Bujagali Falls and SurroundingHabitats 90. The network o f lakes within which the Bujagali project i s located is rich with floodplains and wetlands and supports a diversity o f animals and plants and many water-dependant ecosystems. It i s also one o f the most important areas in Africa for biological diversity. 91. The Project requires the flooding o f important natural habitats including the Bujagali Falls, the riverbank portions o f the Jinja Wildlife Sanctuary and the Nile Bank Central Forest Reserve, a protected area, and the island betweenthe sections o f the Bujagali rapids. In addition, the associated transmission lines would run through the important and valuable Mabira Forest, and an area o f important and productivewetlands. Picture 2 Bujagali Falls 92. The Jinja Wildlife Sanctuary, established in 1953, i s a protected area which i s home o f several bird species, reptiles and a diversity o f insects. When the Sanctuary was established, there were hippopotami in this section o f the river. 63Loadshedding is a controlled way o f rotating the available electricity between all customers. 64EACReport2006, Section 3.4.3. 26 The Nile Bank Central Forest Reserve i s a protected area held in trust by the national government for the people. 93. Many studies also document the diverse fish species, populations and fisheries in the Lake Victoria region, and the history o f chan e and loss over the last decades due to human activities and interventions!' According to the Project Environmental Assessment, the Victoria Nile "originally had a very rich assemblage o f fish dominated by riverine species." The EA states that while the dams have created a physical barrier for fish movements, "viable population of many j s h continue to exist in the Victoria Nile."66Hundreds o f species have evolved to fill almost all of the major niches available to freshwater fishes.67 Other studies, by comparison, raise significant concerns about the situation o f fish species in the upper Nile, and the potential effects o f the existing dams and the Project on these species. These studies and related issues o f environmental impacts are considered inChapter I11(Environmental Issues). 4. The Project's Socio-economic and Cultural Setting 94. Most o f the people living in and around the Project area are farmers, though fishing is also a very important economic activity for the area. In addition, the particular site o f Project site i s very attractive to tourists because o f the scenic topography o f the area and the rapids of the Bujagali Falls, which offer white water rafting opportunities. The potential impacts o f the Project on the economy o f the area, including through fishing and tourism, i s a highly important element o f the Request and i s examined inChapter VI1 (Involuntary Resettlement), which also examines the dis lacement o f people within the area that would be flooded bythe Bujagali dam.6H) 65Balinva et al. 2003. See also fh 65. 66R J Burnside International Limited, Bujagali Hydropower Project Social and Environmental Assessment Main Report, December 2006. (D102) [hereinafter "HPP-SEA"] Executive Summary, p. 18. See also Salzburger W et a1 "Out o f Tanganyika: Genesis, explosive speciation, key-innovations and phylogeography o f the haplochromine cichlid fishes." BMC Evolutionary Biology 2005, 5:17. Available at ~http://www.biomedcentral.comil471-214815117> The cichlid fishes of East Africa are "well known for their spectacular diversity and their astonishingly fast rates o f speciation ...virtually all cichlid species from Lake Victoria (-500 species) ... are haplochromines"( 1). 67Nile perch (Lates niloticus) is a large predatory fish that was introducedinto Lake Victoria by man inthe 1950s. An estimated 150-200 cichlid species from Lake Victoria are thought to be extinct as a consequence. Not all haplochromines are lacustrine (lake dwellers) and close to 200 species inhabit rivers. They are known to inhabit almost every available lake and river habitat. Rocky shores and islands are important refuges for a number o f cichlid species that were formerly not restricted to rocky substrates, but now survive there to escapeNile perchpredation. The process of resettling people in the area o f inundation and other land areas taken as a result o f the Project commenced in2000 at the time o f the prior Bujagali dam project, as discussed inChapter VI1even though the dam was not constructed and no flooding yet occurred. 27 Picture 3 Rafting at the Bujagali Falls 95. Another very important social aspect is that the Bujagali Project has a strong, complex cultural and spiritual tradition. The Busoga make up 46 percent o f those living immediately adjacent to the dam site compared to 17 percent o f the Buganda people. Although the peoples o f other ethnic groups inhabit the Project area69,the Busoga claim spiritual dominion o f both sides of the Nile, its islands, the water and its waterfall^.^' According to a 2002 census, there are about 2.7 million Busoga in Uganda, whose territory lies to the east o f the Project site. 71 Their language, Lusoga, predominates in this area, on the East bank o f the River Nile. 96. The Busoga share a common dialect and ideological, spiritual history, sharing a cluster o f eight or more high status spirits - including Budhagaali, the spirit residing at the Bujagali Falls site - who are invoked intheir specific ceremonies. The Busoga are distinct from the Buganda, Uganda's largest ethnic group - whose traditional realm reaches to the west bank of the Nile. The potential implications o f the Project on places o f cultural and spiritual significance to local people, and whether the Bank has complied with its operational policies and procedures on these matters, i s addressedinChapter VI11(Cultural Property) 69 Several ethnic groups live in and around the Project site, including the Busoga and Busanga people whose lives and livelihoods will be affected by the Project. ' OThe 2001 RAP states its baseline survey identified 22 ethnic groups living inthe project area (HPP-SEA, p. 161). The region was repopulated by migrants from throughout Uganda and other central African countries in the 1940's after being nearly abandoned by the Busoga at the turn of the century due to sleeping sickness. (Bujagali Power Project - Hydropower Facility - Resettlement and Community Development Action Plan, March 2001 (D001) [hereinafter "RCDAP 2001"], p. 98). 71 Obwa KyabazingaBwa Busoga Online: 28 C. The Project Description 97. The Project consists o f the construction o f the Bujagali hydropower plant on Dumbbell Island on the Nile River, just below the Bujagali Falls, about 8km downstream from the existing Nalubaale and Kiira Hydropower Plants. Under the Project, an intake powerhouse complex providing a maximum capacity of 250MW and a rock filled dam about 30 meters high with spillway and other associated works will be developed. On the west bank o f the Victoria Nile, adjacent to the powerhouse, a high voltage substation, the Bujagali Substation, through which all power generated from the Project will flow, i s to be ~onstructed.~~ 98. The reservoir, which will inundate the Bujagali Falls and the islands, i s to have an estimated surface area o f 388 hectares (ha) at full supply level, which will provide a total volume o f water at full supply level o f 54 million m3.The Project requires 238 ha o f land take to construct project facilities and thus will cause the involuntary resettlement of affected people living in and around the site. The flooding for the reservoir will also cause the loss o f white water rafting opportunities over 2.5 kmfrom the Bujagali Falls to Dumbbell Island. 99. The Bujagali hydropower system also includes the construction o f 100 km o f transmission lines, a new substation at Kawanda and the extension o f the 72PAD on a ProposedInternational Development Association PartialRisk Guarantee inthe amount o f up to US$115 million for a Syndicated Commercial Bank Loan and on a Proposed International Finance Corporation Financing consisting of: an "A" Loan inthe amount o f up to US$lOO million and a "C" Loan inthe amount ofup to US$30 million, and on a ProposedMIGAGuarantee inthe amount ofupto US$l15 million for Sponsor's Equity to Bujagali Energy Limited for the Private Power Generation (Bujagali) Project inthe Republic o f Uganda, April 2,2007, p. 9. 29 substation at Mutundwe, all financed by the Afiican Development Bank (AfDB) under the Bujagali InterconnectionProject (BIP).73 100. The PAD states that Dumbbell Islandwas chosen as the Project location because at this point the river i s divided into two channels, a division that provides support for the dam and facilitates the construction o f cofferdams. The embankment i s to be located across the eastern channel at the downstream end o f Dumbbell Island, and the powerhouse and spillway will be in the western channel.74 Located downstream from the Nalubaale and Kiira plants, the Bujagali dam i s to use water released from Lake Victoria that passes through the two existing hydropower plants. 101. As noted earlier, Bujagali Energy Limited (BEL) i s to develop the Project. BEL i s responsible for financing, constructing and operating the Project "on a Build- Own-Operate-Transfer basis." On December 13, 2005, BEL and the Government signed the Implementation Agreement (IA), which defines the rights and obligations o fBEL and the Government. 102. BEL i s to sell the contracted capacity o f 250MW exclusively to the Uganda Electricity Transmission Company (UETCL), which agreed to purchase the Project's contracted capacity under a 30-year Power Purchase Agreement (PPA), also signed on December 13, 2005. This agreement was amended and restated on May 25, 2007.75The Government will guarantee UETCL's payment obligations to BEL.76 103. The Project i s a Public Private Partnership between the private project sponsors, the GoU, multilateral and bilateral development agencies, and commercial lenders as beneficiaries o f the proposed IDA Guarantee. The total Project cost i s estimated to be around US$798.6 million.77 The International Development Association (IDA)78supports the Project through a partial risk guarantee o f US$115 million, guaranteeing the commercial lenders involved in financing the Project against debt service and payment defaults o f the Government inrelation to the Government's payment obligations set forth in the Implementation 73Bujagali Interconnection Project i s closely related to the Bujagali Hydropower Project and will provide the transmission infrastructure to interconnect the new Bujagali hydropower station to the national electricity grid. A loan inthe amount o f approximately $28.6 million USD from the African Development Fundof the African Development Bank (AfDB)was approved by itsBoardofDirectors on June 28,2007. PAD, Annex 4, p. 63. 75 The Requesters claim that there is no evidence that the Power Purchase Agreement (PPA) was debated and approved by the Ugandan Parliament. Inhis legal opinion dated May 31,2007, the Attorney General o f the Republic o f Uganda issued an opinion stating that the Power Purchase Agreement ``...was duly authorized, signed, executed, and delivered" and was legally binding on the Parties "in accordance with the terms and conditions contained therein", adding that "there are no more legalformalities required to be furflled to make ...thePower Purchase Agreement...more binding on the Parties." It should be noted that the PPA was amendedand restated once more onDecember 6,2007. 76 PAD, p.19. The terms o f this guarantee are included inthe Implementation Agreement. "PAD,Annex5,p.67. 78 Inthis Report, the terms "IDA" and"the Bank" are usedinterchangeably. 30 Agreement. Under an Indemnity Agreement signed between IDA and the Government, the latter would reimburse IDA o f any claims and expenses suffered ifIDAwere calleduponto makepayments underthe Guarantee Agreement. The Project i s also financed through, inter alia, an International Financial Corporation (IFC) loan and a Multilateral Investment Guarantee Agency (MIGA) Guarantee. Intotal, the World Bank Group's financial support to the Bujagali Projectis upto US$360 million. IDA'S Board o f Executive Directors approved the IDA Guarantee on April 26,2007. 104. As noted earlier, the Project i s the second effort to develop the Bujagali Hydropower Project. On December 18, 2001 the World Bank Group approved its support to the prior Bujagali Project, which was to be undertaken by the AESNP. However, AESNP later withdrew because of, among other things, the company's "weakeningjnancial position."79The Government terminated the Project-related agreements in2003. Though the Project under investigation inthe present Report i s considered a new financial operation, its design i s practically the same as that o fthe project stopped a few years back.80Inthe first effort to develop the Bujagali dam, certain activities were initiated but not completed, such as the resettlement program leadingto "legacy issues" for the current Project, discussed below. 105. The current Project presents a number o f significant social and environmental issues and challenges. Two o f these, relatingto resettlement and cultural property, are noted briefly below. These and others, including those relating to environmental impacts and the Kalagala Falls offset, are dealt with in detail in subsequent chapters o fthis Report. 1.TheResettlementProgram 106. Under the first Bujagali project, AESNP - the previous sponsor - began the physical resettlement o f people whose land was to be taken by the Project, and paid compensation as 8 art o f the Resettlement and Community Development Action Plan (RCDAP). 7The PAD states that the previous sponsor "completed the planned compensation" and that "the resettlement housing was also completed and the 34 families have moved into it." However, "several activities under the RCDAP were not completed at the time AESNP departed the project; these were primarily income generation activities."82Under the new Bujagali Project, the new sponsor BEL prepared an Assessment o f Past Resettlement Activities and Action Plan (APRAP), which identifies the new Sponsor involuntary resettlement responsibilities. l9Management Response,720. 8o As described in Chapter V, the power generation capacity o f the Project is 250 MW, while that o f the rior project was 200 MW. PAD, Annex 15, p. 142. 82PAD, Annex 15, p. 142. 31 107. To help complete the pendingincome restoration activities at the hydropower site, the Project documents state that BEL has committed to three programs: "agricultural improvement, fisheries and small business support and mi~rocredit."~~ BEL also developed a Community Development Action Plan (CDAP) which includes actions aimed at improving livelihoods o f Project affected people; improving the overall quality o f life by expanding on basic services such as water and sanitation, health and education; and providing mechanisms for dealing with vulnerable people.84These issues are addressed in Chapter VI1(Involuntary Resettlement). 2. CulturalResources 108. Inthe context o f the prior Bujagali project, AESNP prepared a Cultural Property Management Plan (CPMP), which identified the project affected sites that are culturally significant for the local population: rocks, trees and land sites associated with spiritual forces, which, the local population believes, speak through medium or traditional spiritual leaders. Under the prior project, traditional leaders stated that the spirits would have accepted changes to the landscape o f the area if appropriate ceremonial procedures were undertaken and financed by AESNP. AESNP carried out the ceremonies. According to the PAD, however, the consultations that BEL carried out with the Kingdom o f the Busoga and the Busanga people revealed that additional ceremonies were necessary. These ceremonies are to be carried out under the current Project. This issue i s addressed inChapter VI11o fthis Report. D.World Bank Involvement inUganda Power Sector, the Lake Victoria and the Nile River basin 109. World Bank involvement in the Uganda power sector, through IDA, dates back twenty years, with projects supporting, among other things, the rehabilitation o f the Owen Falls Dams (Nalubaale), the construction o f the Owen Falls Extension (Kiira) and expansion o f rural electrification. The World Bank Group's role in general aims at supporting infrastructure development and mobilization o f private investments, the Government's power sector reforms, the structuring the project financing and the implementation o f environmental and social policies acceptable to the World Bank. 110. According to Management, "Uganda 's Renewable Energy Policy and Plan provides for "off-grid" electricity options such as solar PV and micro-hydro, as well as biofuels for cooking and industrial applications. The Bank and other donors are actively supporting these programs as well. " 85 Currently the World Bank Group i s involved with three ongoing power projects, the Fourth Power Project, the Energy for Rural Transformation Project, the Private Power 83PAD, Annex 15, p. 143. 84PAD, Annex 15, p. 143. 85Management Response, p 27. 32 GeneratiodBujagali Project. The Energy for Rural Transformation (ERT) Project (FY02) has supported reparation o f a renewable energy resource database and capacity building plan 8. It i s also supporting investments in renewable energy power generation, including bagasse based cogeneration, mini-hydro, and micro- hydro. The Fourth Power Project (FY08) i s supporting geothermal exploration in western Uganda (Kibiro and Katwe), including shallow-well drilling which i s required to assess the resource. Inaddition several projects, includingthe Thermal Generation Project and the Karuma Hydropower Plant, are beingproposed.87 111. The World Bank has also supported the Lake Victoria Environmental Management Project (LVEMP), a regional project, carried out under a Joint Project Agreement involving Uganda, Tanzania and Kenya. The LVEMP was the first phase o f a longer-term program, intended to maximize benefits for riparian countries from usingresources within the lake basin for food, employment etc, to conserve biodiversity, and to build scientific and institutional capacities to stop the environmental deterioration o f the lake and its surrounding ecosystems. This effort comprised separate projects, implemented by national secretariats in the three countries and coordinated by a small regional secretariat, established in Arusha, Tanzania. 88 The LVEMP was launched in 1997 and funding for it totaled around US$ 75 million over a seven year period until 2005. 112. Nile Basin Initiative. The World Bank is also a partner o f the Nile Basin Initiative (NBI), a regional partnership led by all ten Nile Basin countries: Burundi,Democratic Republic o f Congo (DRC), Egypt, Eritrea, Ethiopia, Kenya, Rwanda, Sudan, Tanzania and Uganda. NBI was launched in 1999 as a basin- wide framework to "develop the river in a cooperative manner, share substantial socioeconomic benefits, and promote regional peace and s e c ~ r i t y . "The Nile ~ ~ riparian countries agreed on a "shared vision" to "achieve sustainable socioeconomic development through the equitable utilization oJ: and benefit from, the common Nile Basin water resources." NBI's structure consists o f the Council o f Ministers o f Water Affairs o f the Nile Basin Countries, the Technical Advisory Committee and the Nile Basin Secretariat. The World Bank has been involved in the Nile Basin Initiative since 1997 in partnership with UNDP and the Canadian International Development Agency (CIDA), "to facilitate dialogue among the NBI countries and to chair the International Consortiumfor Cooperation on the Nile (ICCON) Consultative Group Meeting in Geneva, Switzerland in June 2001. " About US$130 million were initially committed by the partners to the 86Most recent report: Fourth InterimReport for Renewable Energy Resource Information Development and Capacity Building Assessment, Kamfor Company Ltd.April 2006. 87The issueo f alternative energy generation options is further analyzed inChapter V of this Report. "PAD for the Supplemental Credit Document International Development Association Proposed Supplemental Credit To The United Republic Of Tanzania For The Lake Victoria Environmental Management Project, September 17,2004. 89Nile Basin Initiative (MI)Backgroundat http:iiwww.nilebasin.or~iindex.phD?option=com-content&task=view&id=l3&Itemid=42, (accessed on July 10,2008). 33 Initiative and a multi-donor Nile Basin Trust Fund (NBTF), currently administered by the World Bank, was established to channel these funds to NBI. 113. One of the programs carried out under the NBI i s the Nile Equatorial Lakes SubsidiaryAction Program (NELSAP), the mission of which is "to contribute to the eradication of poverty, to promote economic growth, and to reverse environmental degradation in the NEL [Nile Equatorial Lakes] region."" The NELregion includes the six countries inthe southern portion of the Nile Basin- Burundi, Democratic Republic of Congo, Kenya, Rwanda, Tanzania and Uganda-as well as the downstream riparian states Egypt and Sudan. Under the NELSAP a StrategdSectoral, Social and Environmental Assessment was prepared "to provide guidance on thepower generation options available in the region, based on an assessment of electricity demand, project costs, and environmental and social issues surrounding such projects."92 The SSEA is analyzed inChapter I11of this Report - Environmental Issues. ~ 90Nile Basin Initiative (NBI): http://www.nilebasin.org/ (accessedon July 10,2008). 91StrategiciSectoral, Social and Environmental Assessment of Power Development Options in The Nile EquatorialLakes Region, February 2007. 92PAD, p. 43. 34 Chapter I11 EnvironmentalIssues 1 4. The Request submitted to the Inspection Panel presents a number o f claims centered on the social and environmental studies supporting the Project. In the Requesters' opinion, these studies are generally inadequate and violate the Bank's Policy on Environmental Assessment (OP/BP 4.0 1). The Request develops from the overarching claim that the Project's SEA i s based on old data-some o f these data are allegedly 10 years old-which have "little or no bearing to current situation" and "do not reflect the current environmental realities" o f the Project area. In the Requesters' view, the SEA also does not take into consideration specific important aspects o f the Project, such as the hydrology o f Lake Victoria and the Lake's longterm health; the need for a cumulative impact assessment, and the consideration o f climate change effects, all o f which may have a significant impact on the production o fhydropower. 115. In general, Management responds that the Project i s a new operation and, as a result, social and environmental aspects have been reassessed. It adds however, that drawing upon former studies, the Project benefited from the baseline social and environmental data gathered for the prior Bujagali Project by AESNP. Management states that the current Project "has also retained its original environmental footprint" and the work conducted was designed to build upon earlier data and additional studies were undertaken as needed, to confirm or update that baseline.93Management considers that the baseline data gatheringwas satisfactory. Management also argues that the SEA addressed social and environmental issues related to the Project while "the broader climate change (and hydrology) aspects were addressed in different studies,"94 in particular the SSEA prepared under the NileBasinInitiative (NBIJ9' 116. Paragraph 2 o f OP 4.01 states that the: "EA is a process whose breadth, depth, and type of analysis depend on the nature, scale, and potential environmental impact of the proposed project. EA evaluates a project's potential environmental risks and impacts in its area of influence;-examines project alternatives; identij?es ways of improvingproject selection, siting, planning, design, and implementation by preventing, minimizing, mitigating, or compensating for adverse environmental impacts and enhancing positive impacts; and includes the process of mitigating and managing adverse environmental impacts throughout project implementation. The Bank favors preventive measures over mitigatory or compensatory measures, wheneverfeasible." 93Management Response, p. 22. 94ManagementResponse, p. 19. 95The Nile Basin Initiative was described inChapter I1o f this Report. 35 117. This chapter o f the Report begins by examining the general claim related to the adequacy of the social and environmental assessment studies related to the Project. The analysis o f the specific issues, especially in relation to cumulative assessment and fisheries follow. Further sections o f the chapter address issues related to the Kalagala offset and the safety o f dams. Specific issues relating to the hydrology o f the Lake Victoria and climate change are analyzed inChapter V. Issues o f involuntary resettlement, and cultural issues, are examined in Chapters VI1andVIII. A. The Environmental Component of the Social and Environmental Assessments 118. The objective and the main provisions o f the policy on Environmental Assessment have been laid out in the preceding paragraph and will serve as the guiding norm for the Panel's analysis. As needed each section will also point to other specific provisions o f OP 4.01 relevant for the discussion. 1.Adequacy ofthe studies 119. The Panel notes that the Sponsor, with input from Bank Management, has contracted international consultants to prepare the required SEA for the Bujagali Hydropower Project96and the Bujagali Interconnection Project.97The Panel has found evidence that the Regional environment sector unit as well as the Environment Department were actively involved in guidingthe preparation o f the SEAS.^^ 120. Because o f the Project's history and the prior attempt to develop the Bujagali Hydropower Project, initial social and environmental studies, prepared under the previous sponsor AESNP, preceded the SEA studies required for the present Project. The Project has appropriately been classified as category "A", the category under Bank policy used for projects with the most serious level of impacts. This complies with OP 4.01. This classification together with key environmental issues and an Environmental Data Sheet are recorded in the Project's Concept Note, Appraisal Document and implementation documents. 121. OP 4.01 requires that an EA report on a project include the following: an executive summary; a survey o f the policy, legal, and administrative framework within which the project will be undertaken; a concise description o f the project; appropriate baseline data; and an assessmento f environmentalimpacts takinginto account human health and safety and social aspects, including involuntary resettlement, indigenous peoples and cultural property. It also requires 96HPP-SEA 97Bujagali Interconnection Project, Social and Environmental Assessment Report, December 2006 (hereinafter "IP -SEA") 98World Bank Management interviews, September2007. 36 identification o f mitigation measures and residual negative impacts that cannot be mitigated; a systematic comparison o f feasible alternatives to the Project site - technological, design, and operation; and an environmental management plan to cover mitigation measures, monitoring, and institutional strengthening ifrequired. These requirements are outlined inAnnex B o f OP 4.01 122. The Panel finds that, apart from the omission o f an EMP, discussed below, the SEA includes the elements required by Annex B o f OP 4.01. The Project is filly described and set in an appropriate policy, legal and administrative framework. Baseline data are provided, as i s an assessment o f the environmental impacts o f the proposed alternative. Feasible technological, design and operational alternatives are examined. The study takes a holistic approach to environmental issues and considers natural aspects in an integrated way. The country's overall environmental policy framework, national legislation, and obligations under relevant internationaltreaties and agreements are considered. 123. The preceding finding relates to the SEA in its entirety. Separate parts o f the study are considered inthe following sub-sections o fthis chapter. 2. Environmental Management Plan 124. The first paragraph o fAnnex C o f OP 4.01 reads: A project's environmental management plan (EMP) consists of the set of mitigation, monitoring, and institutional measures to be taken during implementation and operation to eliminate adverse environmental and social impacts, offset them, or reduce them to acceptable levels. The plan also includes the actions needed to implement these measures. [footnote omitted] Management plans are essential elements of EA reports for Category A projects ... (emphasis added) 125. The Panel notes that the SEA lacks o f a detailed EMP for the Bujagali Hydropower Project. Although fourteen action plans are outlined in the SEA (seven sponsor plans and seven contractor plans) those that relate to Environmental Mitigation and Monitoring have yet to be drawn up for implernentati~n.~~ The fact that the EMP is not an integral part of the SEA that has been disclosed is a deficiency. This is not in compliance with the requirements of OP 4.01. 99See Sections 8.3.7 and 8.4.7 ofHPP-SEA and Sections 8.3.6 and 8.4.1 o f the IP-SEA. 37 3. Institutional Capacity 126. The Panel notes that the need for strengthening country institutional capacity in the social and environmental sectors"' was identified inthe Project Concept Note (PCN) under the heading "Technical Assistance to the Government," among other things to assist the Government with monitoring the environmental and social compliance aspects o f the Project. This requirement however was not carried through into the Project Identification or Project Appraisal documents. In the Panel's view adequate capacity to implement the social and environmental aspects o f a project i s critical for its success. OP 4.01 requires that when there i s inadequate legal or technical capacity to carry out EA functions, the Project includes components to strengthen that capacity.lo'This requirement to support needed capacity building, which is important in the implementation of the social and environmental aspects, has not been compliedwith inthis Project. 4. IndependentPanel of Experts 127. The Panel also finds that although there i s evidence that both the World Bank and the Sponsor have engaged the services o f independent experts to review and advise on many aspects o f the Project,lo2 an independent panel o f internationally recognized environmental specialists has not been appointed for the Project. This is not in accord with Paragraph 4 o f OP 4.01, according to which in "Category A projects that are highly risky or contentious or that involve serious and multidimensional environmental concerns, the borrower should normally also engage an advisory panel of independent, internationally recognized EA.'3 environmental specialists to advise on all aspects of the project relevant to the 103As the Project is contentious and involves complex multidimensional environmental concerns, appointment of an environmental panel of international experts is warranted and the lack of such a panel i s not in compliance with OP 4.01. loo Particular concern i s expressed and advice sought on: (a) Bujagali and transmission line project management; (b) contingent liabilities management; (c) financing second stage o f geothermal drilling; (d) funds to foster potential tourisdother investments at the Caligula offset site; (e) to assist NEMAto monitor the environmental and social compliance aspects of the project; and/or (Q funding o f community development activities that the sponsors may not be willing to finance through either debt or equity. lo'OP 4.01 113. lo*Bujagali Hydropower Development Uganda Project Review and Assessment Report for IFC, prepared by Colenco Power Engineering Ltd., February 2007 (hereinafter "Colenco 2007"); Power Planning Associates, Economic and Financial Evaluation Study, December 2006; Bujagali Hydroelectric Power Project Transmission Interconnection Study System Analysis Report for TRC Global Management Solutions LP, prepared by Siemens Power Transmission & Distribution, Inc. Power Technologies International, August 2006 (hereinafter `Siemens 2006"). `03OP 4.01 14. The policy also reads: "The role o f the advisory panel depends on the degree to which project preparation has progressed, and on the extent and quality o f any EA work completed, at the time the Bankbegins to consider the project." 38 5. Disclosure of Project Documentation 128. The Requesters believe that the Project SEA does not address significant potential impacts o f the Project in relation to hydrology, the long term health o f Lake Victoria, climate change and cumulative impacts. They also complain that the only document they had the chance to review was the Project's SEA while the "World Bank has also recently refused topublicly release information on the Nile hydrology and the impacts of Kiira Dam's operations on the levels of Lake Victoria." 129. Management states that the SEA addressed social and environmental issues related to the project while climate change and hydrology were addressed in different studies-in particular the SSEA prepared under the Nile Basin Initiative-which were all publicly disclosed. Management states that "learning from the past" the Government "implemented a stronger program of public disclosure." The Bank has disclosed the Project's Economic Study, the SEA and the SSEA, along with other environmental and social documents in the InfoShop inWashington and invarious locations inUganda.lo4It is Management's position that "many of the information-related questions of the current Request are addressed within the body of information and analysis made available to the 130. Paragraph 7 o f OP 4.01 states that a range o f instruments-environmental impact assessment (EIA), regional or sectoral EA, environmental audit, hazard or risk assessment, and E M P - c a n meet the policy's EA requirement, and these are used as appropriate. In addition, a sectoral or regional EA i s required if the Project i s likely to have sectoral or regional impacts. 131. This i s qualified by a sentence from paragraph 8(a) o f OP 4.01 : "For a Category A project, the borrower is responsible for preparing a report, normally an EIA (or a suitably comprehensive regional or sectoral EA) that includes, as necessary, elements of the other instruments referred to in para. 7." In the section headed "Disclosure" OP 4.01 also requires that "For a Category A project, the borrower provides for the initial consultation a summary of the proposed project's objectives, description, and potential impacts; for consultation after the draft EA report isprepared, the borrower provides a summary of the EA s conclusions." 132. The Panel notes that the Bujagali SEA makes only a passing referencelo6to the SSEA.lo7The latter study was managed and supervised by the World Bank and financed by the Canadian International Development Agency (CIDA) for the Nile Equatorial Lakes Subsidiary Action Program o f the NBI. It was completed in lo4Management Response, 7 39. Management Response, y24. lo6HPP-SEA, Sections 4.3.4 (p. 183-4) and 7.6.6 (p. 4367). lo'SSEA. 39 February 2007, two months after the Bujagali SEA. The SSEA makes no mention o fthe Bujagali SEA. 133. The SSEA was undertaken, according to the PAD, "to provide guidance on the power generation options available in the region, based on an assessment of electricity demand, project costs, and environmental and social issues surrounding such projects.'71o8The PCN (the first document the Bank makes public containing essential information about a proposed project and financing), preceding the PAD, had also pointed to the 2002 Inspection Panel Investigation Report related to the prior Bujagali Project stating that the Project would take note o f the issues raised inthe Panel's report, including `>ayingparticular attention to undertaking a Strategic Sectoral Environmental Assessment and Cumulative Impacts Study.""' 134. It is clear from reading the two reports, the SEA and the SSEA, and the complete lack of cross-references between them, that they do not form part of the same suite of documents-the link between them being that they both deal with the topic o f electrical power inEast Africa and were both supervised by the World Bank. Nevertheless the Management Response to the Request for Inspection gives the Nile Basin SSEA as the source o f data and analysis o f the potential effect o f climate chan e on the Bujagali Project"' as well as for Cumulative Effects o fthe project.81 135. Although the policy does not contemplate reports from one project/program being used to fulfill the requirements o f another project-in this instance studies conducted under the NBI-the Panel i s o f the view that, in the interests o f efficiency, an EA may, in principle, refer to and/or incorporate, as appropriate, other relevant studies. However, as the purpose o f both the sectoral and project specific EA i s to disclose information relevant to a decision, the fact that one study is reliant on another must be clearly stated and disclosed in project documentation.' l2Without this, information important to a project i s obscured even ifit i s disclosed independently, which weakens or undercuts the achievement o f the key elements o f OP 4.01 relating to informed decision-making, public consultation and disclosure. The Panel finds justifiable the Requesters complaint that some aspects o f the Project, that i s effects o f climate change and the cumulative effect^,"^ have not been properly addressed in the project SEA. The Panel acknowledges that the necessary studies have been conducted and Io'PAD, p. 18 states that the SSEA was undertaken to `>provide an overview analysis of the social and environmental issues surrounding possible regional power development options in the Nile Equatorial Lakes Region of Africa based on demand scenarios up to 2020, taking into account potential climate change and cumulative impactsfrom multiple investments." logPCN, Section B(e), p. 5. `loManagement Response, Annex 1, Section 4, p. 19. `I'Management Response, Annex 1, Section 6, pp. 20-1. `I2A clear statement and graphic showing the inter-relationships and entire suite o f documents that constitute the studies making up the SEA should be included as a preface to all suchrelated documents. `I3For a discussion o f climate change and cumulative effects, see Chapter V. 40 disclosed, albeit independently, and considered by Management and referred to specifically in the PAD. However, the failure to disclose the SSEA or its relevant parts as an integral part of the Bujagali Hydropower Project's documentationin a timely manner i s not consistentwith OP 4.01. 6. Cumulative Impacts of Bujagali and Existing and Future Hydro Projects 136. Cumulative effects are changes to the environment that are caused by an action (a project) in combination with other past, present and reasonably certain fitture human actions irrespective o f who undertakes such actions. Consideration i s given to effects on: (i) bio-physical components o f ecosystems and (ii) socio- economic and cultural characteristics o f the affected space. The stress is on analyzing known or probable additive or synergistic interactions, and not simply the direct effects o fthe particular action under assessment.'l4 137. The Requesters claim that the issue o f cumulative effects "remains unresolved" in spite o f the Inspection Panel's 2002 finding that "the issue of cumulative effects, addressed by Management and raised by the Requesters, is of real significance and is deserving of greater attention." Inthe Requesters' opinion, the SEA does not discuss cumulative impacts, and BEL did not attempt to identify issues, especially with respect to the health o f the Lake Victoria, arising from building a cascade o fdams on the River Nile, including Bujagali. 138 Management argues that cumulative impacts o f the current Bujagali Project are addressed as part o fthe Project's SEA and inthe SSEA. BEL'SSEA examines the cumulative impacts o f Bujagali, the hydropower plants at Nalubaale, Kiira and Karuma along with the transmission facilities on the Victoria Nile in Uganda. It focuses specifically on the reach o f the river between Lake Victoria and Lake Albert and takes into account other initiatives such as environmental offsets, natural areas, parks, reserves and so on. The SEA concludes that the socioeconomic impacts o f Bujagali, generally, would be local because the existing Nalubaale-Kiira power plants and Bujagali are separated by Lake Kyoga from Karuma Falls and other potential hydropower sites downstream on the Nile River. In addition, the SSEA analyzes the cumulative impacts o f several hydropower development alternatives under differing scenarios o f regional grid integration. It concludes that developing Bujagali and other sites in the Victoria Nile Basin (excluding Kalagala) will not have significant cumulative environmental impacts. The SSEA analyzes and ranks potential fitture power options, based uponmultiple criteria. These are: assessment o f direct, indirectlinduced and cumulative impacts o f multiple activities; additional costs and benefits through multi-purpose use o f `14See for example: Larry Canter & Barry Sadler, A Toolkit for Effective EIA Practice-Review o f Methods and Perspectives on their Application: A Supplementary Report o f the International Study o f the Effectiveness o f Environmental Assessment, (International Association for Impact Assessment 1997) at, Chapter 5. See also Considering Cumulative Effects under the National Environmental Policy Act: Handbook on Cumulative Effects Analysis,. (Council on Environmental Quality 1997), and Cumulative Effects Assessment Practitioners Guide (Canadian Environmental Assessment Agency 1999). 41 storage reservoirs; risk o f rainfall variability; and sharing o f benefits at the local and regional level. Management also claims that the studies conducted to inform the decision making process o f the first Bujagali Hydropower Project served as part o fthe informationbase for the SSEA."' 139. Annex A of OP 4.01 states that a "[s]ectoral EA pays particular attention to potential cumulative impacts of multiple activities." The Management Response draws attention to section 14 o f the SSEA. This section is headed "Assessment o f Cumulative Impacts," covers 33 pages and provides an overview o f the consequences of various portfolios o f regional power options being .adopted. However, the Panel notes that there i s neither detailed analysis o f the existing and proposed hydropower projects on the Victoria Nile nor o f the Transmission Lines linkingtheseprojects to load centers.'16 140. The analyses in the SSEA allow a comparison amongst the various proposed portfolios o f power development options in the Nile Equatorial Lakes Region. They do not, however, provide a systematic examination of the potential consequences of the Nalubaale and Kiira facilities, the Bujagali Project, and the planned Karuma project all being situated on the Victoria Nile between Lake Victoria and Lake Kyoga. In addition, there is no examination of the impact o f additional transmission lines between the hydropower stations and Kampala. Although section 14 of the SSEA i s headed "Assessment of Cumulative Impact" the Panel finds that the analyses are not sufficiently backed by evidence and include opinions rather than careful fact-based examinations of the additive effects of impacts from present and foreseeable projects. 141. The Bujagali Hydropower SEA seems to address cumulative effects in more detail. For example a paragraph o f section 7.7.3 o f the Bujagali Hydropower SEA reads: The following impacts are considered to be negative cumulative impacts of the Bujagali HPP ... all are judged to be of minor signijkance: 9 Relocation of people with compensation to accommodate theproject's construction, facilities and operations; 9 Aesthetic impactsfrom thepresence of another dam with the potential for knockon tourism impacts (potentially positive, as well, however); `I5 Management Response, pp. 6 and20. SSEA sections 14.7.1.5, 14.7.2.3 and 14.7.4 come closest to an analysis of the cumulative effects o f adding the Bujagali and Karuma Hydro-power facilities to those already existing at Nalubaale and Kiira. Section 14.7.1.4 provides a briefstatement on the potential cumulative effects o f transmission lines. 42 Some disruption of the naturalflow regime over an -8-km stretch of the river Nile downstream of and as a result of Nalubaale and Kiira: with associated impacts on aquatic organisms and communities (also potentially positive if productivity of reservoir increased); and, river users @hers) -alsopotentially positive if increasedproductivity in reservoir is reflected infishers' catches. Losses of wildlife populations and habitats, as well as agricultural lands, due to inundation of terrestrial habitats. 142. However, no data or arguments are providedto substantiate the above statements, includingthe judgment that the negative cumulative impacts o f the Project are o f minor significance. There i s no determination o f how many people stand to involuntarily lose access to their assets, how much agricultural land i s to be lost, the extent to which riverine forest habitat will be lost, or the extent to which tourism will be affected. 143. In light of the foregoing, the Panelfinds that neither the SSEA nor the SEA has addressedthe cumulative effectsof the existingandplannedprojectsina meaningfulway. This is notincompliancewith OP4.01. 6.1 CumulativeImpactsofthe TransmissionLines 144. Section 7.3.9 o fthe InterconnectionProject SEA reads as follows with respect to the cumulative effects o f transmission lines from Jinja to Kampala: Cumulative effects resulting from the proposed interconnection project include thefollowing: Ecological Features Wayleave width through Mabira and Kifu CFRs will increase from current 30 m to 65 m, butfuture potential incremental increase to 90 m (3-132 kVx 30 m) wayleave is avoided ...; Access to, and within, Mabira CFR may be improved and control measures implemented in collaboration with the NFA and UETCL, facilitating improved management of the forest; and, Recreational facilities within Mabira CFR will be relocated within the reserve and improved, resulting in a net positive benefit to the reserve and its users. Social Features and Conditions By locating the transmission line between Bujagali and Kawanda substationparallel to the `horthern route ''versus the Danish International Development Agency (DANIDA) or 66 kV transmission corridor to the south, involuntary resettlement is minimised and sensitive compensation issues are not 43 aggravated further. (Siting a new line adjacent to the DANIDA line could potentially displace somefamilies for a second time as a result of transmission line construction.); and, Landowners will receive compensation to meet World Bank Group requirements including in certain cases, a "top-up " over Government of Uganda requirements. In general, landowners may receive a small net positive benefit due to theproject. Aesthetics Visual impact of the interconnection project will be greatest in the vicinity of Lubigi Swamp where no major transmission infrastructure presently exists and along the Bujagali substation to the Tororo line connection (as seen from the eastern bank of the Nile River). Here, several transmission lines already come idout of the Nalubaale switchyard. 145. The Panel notes that these statements fail to address the cumulative effects o f transmission lines or to propose mitigation to reduce additive effects. The cumulative loss o f forest habitat from the transmission lines has not been determined and the statements relating to access and recreational facilities are not expressedinterms that allow determinationo f the overall cumulative effect o fthe multiple transmission lines. Cumulative impacts on social and aesthetic parameters are also not determined. The statement "several transmission lines already come idout of the Nalubaale switchyard" i s used to dismiss the possible aesthetic effect o f yet another transmission line rather than to examine the cumulative effect o fnumerous lines emanating from the same switchyard. 6.2 Alternatives andMitigation Measures-the TransmissionLines 146. The transmission lines that will transport electricity from the hydropower site pass through areas where people live, wetlands, and the ecologically important Mabira Forest. The Panel notes that the SEA fails to address the cumulative effects o f transmission lines; neither does it propose mitigation to reduce additive effects. 147. The Panel was not hrnished with documentation indicating that the Project considered ways to mitigate or reduce the amount o f land taken for the second (Bujagali) transmission line. Rather, the Project assumed that the size o f the existing right o f way needed to be doubled, which i s technically in~orrect.''~ Considerate planning o f the new transmission line to take into account the requiredminimumdistance from the outside phases o fthe lines to the ROW edge, the minimumhorizontal clearance required between phase conductors o f the two `17See for example Design Manual For High Voltage Transmission Lines, Rural Utilities Service Bulletin. 1724e-200, Electric StaffDivisionU.S. Department o f Agriculture, M a y 2005. 44 lines, the spans and sags o f the lines, as well as how structures o f the two lines match up with one another, could significantly reduce the width o f the required wayleave thus reducing the cumulative impact on Mabira and Kifu forest habitat as well as the number o f families to be resettled. The Panel finds that the failure to consider mitigation measures, which would reduce the social and environmental impacts of the transmission line, does not comply with OP 4.01 and OP 4.12. 7. Environmental Impacts on Fisheries and Aquatic Systems 148. The Requesters express concern that the data on which the EA i s based are dated and that such studies as were done were conducted over unrepresentatively short time periods. They express concern as to the accuracy o f the surveys o f endemic fishspecies. 149. Management responds that the Project builds on relevant work conducted for the prior Bujagali Project and on updated information gathered in further field studies and analysis, including studies on fisheries conducted for the prior project and updated for the current Bujagali Project. The Response adds that the Fisheries Resource Institute (FIRRI) conducted four surveys in2000 and additional studies were carried out by the same institute (now known as National Fisheries Resource Research Institute-NaFIRRI) for BEL in 2006. The two sets o f surveys (2000 and 2006) differ in number o f species they found but according to Management "this is to be expected" and do not necessarily indicate species loss or extinction; it maybe due to variations indata collection, migrationor location o fspecies. The overall conclusion is that the "reach of the Victoria Nile that will be affected by Bujagali is not considered to be critical habitat for any fish species of conservation importance.,,' l8 150. As noted, OP 4.01 requires a project EA to evaluate potential environmental risks and impacts o f the Project in its area o f influence and to include a process to mitigate and manage adverse environmental impacts throughout project implementation, favoring preventive measures over mitigatory or compensatory measures if possible. OP 4.04 on Natural Habitats states that the Bank "supports theprotection, maintenance, and rehabilitation of natural habitats ... and expects borrowers to apply a precautionary approach to natural resource management to ensure opportunities for environmentally sustainable development." When a project would significantly convert or degrade a natural habitat, mitigation measures have to be provided for in the Project, measures such as minimizing habitat loss as appropriate. 151. The endemic cichlid fishes o f East Africa are "well known for their spectacular diversity and their astonishingly fast rates of speciation ...virtually all cichlid speciesfrom Lake Victoria (-500 species) ... are haplochromines". I9 Hundreds 11* Management Response, p. 23. ` 1 9Salzburger W et a12005. 45 o f species have evolved to fill almost all o f the major niches available to freshwater fishes. Not all haplochromines are lacustrine (lake dwellers) and close to 200 species inhabit rivers. They are known to inhabit almost every available lake and river habitat. Rocky shores and islands are important refuges for a number o f cichlid species that were formerly not restricted to rocky substrates, but now survive there to escape Nile perch predation.12' The cichlids that inhabit rocky shores were less subject to Nile perch predation. Consequently, rocky habitats in Lake Victoria are important to the survival o f some endangered cichlids. It maybe construed that rocky habitats (rapids) are similarly important in the Victoria Nile, where the Nileperch i s also a predator. 152. A baseline aquatic ecolo y and fisheries survey o f the Victoria Nile was carried out by FIRRI in 2000.'2' The survey was based on quarterly surveys o f water quality, aquatic plants, invertebrate animals and fish. Twenty sites were sampled using routine field and laboratory techniques. These sites covered all habitats present inthe upper Victoria Nile from slower-flowing vegetated margins, to fast- flowing rapids. The field studies were carried out in February, May, August and OctoberINovember 2000 to assess seasonal conditions during Uganda's short and long rainy seasons, and the short and long dry seasons. Each survey used four transects, one above122and three below,'23 the Dumbbell Island site o f the Bujagali Dam. Gill nets o f graded mesh sizes as well as beach seines were used to sample habitats at each transect. This study found that the Upper Victoria Nile has two zones, each with a characteristic fish population. An upstream zone (transects 1, 2 and 3) i s characterized by swift mid-channel current with rock outcrops and rapids such as Bujagali and Kalagala. Further downstream (transect 4), a zone with more gentle flow that is uniform across the channel occurs. Inthe upstream zone fish populations are better adapted to the rocky fast-flowing habitat than the fish populations occurring downstream. The study recommended further investigation o f possible potamodrom~us'~~migratory behavior between the two zones and the role of the natural barriers (such as the falls at Bujagali and Kalagala) in separating fish species and causing different populations in the upstream and downstream reaches o f the Victoria Nile. It also recommended a feasibility study o f a fish pass at Bujagali to allow in-stream fish migrations from below to above the dam. 153. In2001 an additional more detailed investigation o f the haplochromine fishes was commissioned in response to the concern that the previous study had not adequately addressedthe potential loss o f rocky, fast-flowing habitats. Fieldwork for this supplementary study was carried out duringJuly and August 2001. Eleven 120Nile perch (Lates niloticus) is a large predatory fish that was introduced into Lake Victoria by man in the 1950s. An estimated 150-200 cichlid species from Lake Victoria are thought to be extinct as a consequence. 121The National Fisheries Resources Research Institute (and the same institute under its former name- FIN) is a reputable established researchinstitute that has specializedinthe study ofUgandanfisheries. 12'Kalange-Makwanzi. 123Buyala-Kikuba Mutwe; Matumu-Kirindi; Namasagali-Bunyamira. 124Migration within streams or rivers. 46 sites were sampled to cover the range o f fast-flowing and rocky habitats in the upper Victoria Nile, from Ripon Falls to Kakindu, 63 km downstream o f Ripon Falls. Experimental fishing was carried out at each site, with equipment designed to target the haplochromine fishes, which are generally small fish less than 100 mm in length. Angling was also carried out by local fishermen with hooks and rods. Inaddition, sets o f gill-nets (ranging inmesh size from 25 mm to 203 mm) were set overnight. 154. A final report o f this haplochromine study was included as an appendix to the AES Nile Power Environmental Assessment. Part o fthe study's conclusionreads: A total of 35 haplochromine cichlid species were recovered from the upper Victoria Nile by experimentalfishing in rocky, rapidly- flowing habitats. All of the specimens recovered from sites upstream of Busowoko werepreviously known to science, and none are listed as threatened on theIUCNRed Listfor Uganda. Thefact that they are relatively well known is probably due to the species flock in the upper reaches of the Victoria Nile being closely related, if not a continuation oJ the well-studied Lake Victoria flock. All of the species recoveredfrom the area between Owen Falls and Dumbbell Island are well known from sites in Lake Victoria, including the Mwanza Gulf (Tanzania), and the Napoleon Gulf (Uganda). Thesefindings are similar to the findings of the FIRRI (2001) study. It is concluded thatfast-flowing, rocky areas are not theprincipal habitat for haplochromine fishes in the upper Victoria Nile. Although the Bujagali Hydropower Project will result in a reduction inflow velocity in a 4 km stretch of the river, this area is not considered an important sitefor haplochromines. Reduction in flow velocities may in fact result in increased haplochromine abundance. Therefore it is concluded that the Bujagali Hydropower Project will not have a signijkant negative impact on haplochromine cichlids in the VictoriaNile.125 155. The taxonomy o f the East African haplochromines i s the subject o f ongoing debate in the fisheries literature. It has been found that species thought to be extinct are re-emerging in Lake Victoria.'26 The IUCN 2005 study The Status and Distribution of Freshwater Biodiversity in Eastern Africa'27reports: 125Haplochromine Habitat Study, Report No. AF6097170ldg/1215 Rev. 2.0 (WS Atkins International Ltd andFIRRI2001). '26Balinva,J.et a12003 p.703. 127 Will Danvall, K. Smith, T. Lowe, & Jean-Christophe ViC, The Status and Distribution of Freshwater * Biodiversity in Eastern Africa Occasional Paper of the IUCN Species Survival Commission No. 31(IUCN 2005). 47 Many of the Lake Victoria cichlids were previously thought to be extinct but, following additional and more extensive surveys, it appears that a number of these species still exist in small pockets in the lesser-known parts of the main lake and in the smaller satellite lakes (e.g., Bisini, Kanyaboli and Nabugabo). 156. The IUCN 2005 study also concedes that assessments o f the status o f fish based on the 2003 Red Lists were not representative. Two-hundred-and-fifty-two of the 901 fish taxa assessed at the global level (mostly endemic to the region) are threatened (28% of the total number of fish taxa assessed), with two species (Aplocheilichthys sp. "Naivasha '' and Barbus microbarbis) thought to be extinct. This assessment provides a signijkantly improved picture for the regional level of threat than that previously obtained from the 100 species assessed for the 2003 IUCNRedList of which 87% were assessedas either threatened or extinct. Theseearlier assessmentsfocused on the Lake Victoriafish community in an effort to highlight the apparent large-scale decline and loss of cichlid species due to the combined impacts of invasive species, eutrophication and possibly overfishing. Clearly this picture was not representative of the threatened status for fish throughout the region. 157. Based on its review of relevant research studies, the Panel observes that the status of the fish species inhabiting both Lake Victoria and the Victoria Nile i s disputed and that ongoing research i s desirable. However, significant effort has been devoted to study these fish in the reaches of the Victoria Nile that will be affected by the Bujagali Hydropower Project. 158. As the FIRRI report on its Bujagali surveys had recommended a feasibility study o f a fish pass, the Ugandan National Environmental Management Authority formally requested FIRRIto provide an indicative position on the necessity for a fishladder at the Bujagali dam. Inthe FIRRIresponse, dated September 14,2001, the Director writes as follows: The Lake Victoria and Lake Kyoga basins are connected by the Upper Victoria Nile flowing out of Lake Victoria northwards to Lake Kyoga downstream, and, eventually through Lake Albert, the Albert Nile and beyond. Thefish fauna of both lakes Victoria and Kyogafor the most part share a similar evolutionary origin. This means that many species offish in Lake Victoria have also been recorded in Lake Kyoga. 48 It is also well known that many species offish in the lakes undertake longitudinal upstream migrations on a seasonal basis for spawning.,. These migrations have been well studied in fish fiom Lake Victoria migrating to inflowing rivers and streams... The VictoriaNile with respect to Lake Victoria is an OUT-FLOWING river. It becomes IN-FLOWING with respect to Lake Kyoga. [Emphases in original] This means that it is the in-flowing influence at the Victoria Nile-Lake Kyoga mouth where we would expect upstream migration. The investigated transects of Dumbbell Island had afishfauna which was in many respects similar to the Lake Victoriafish fauna. There was a transition zone fiom the third transect downstream of Dumbbell island merging into more typically Lake Kyogafishfauna. The most downstream transects also contained the highest density of anadromous (i.e. migrant species)... It was thus noted that fiom a fish migratory point of view, the Upper VictoriaNile behaved more as an IN- FLOWING riverfor fishes in Lake Kyoga. ... their occurrence throughout the systemproved that there were riverine fish populations that breed within the river irrespective of the natural physical barriers. Such populations especially upstream were unlikely to be affected by other barriers in terms of breeding. It was observed that inspite of the present Owen Falls Dam barrier, the fishes know to be migrants occur in Lake Victoria (where they migrate UPSTREAM) and also occur in sections of the river where breeding specimens have beenfound. This indicates that thesefishes breed within the river. Thepresent Owen Falls Dam is already a barrier to assumed migration towards Lake Victoria. Migrant fishes are found upstream and downstream of this barrier but the same species occur throughout the Upper VictoriaNile towards Lake Kyoga. It is not justifiable that a fish ladder or pass would improve the stocks of migratingfish in the Upper VictoriaNile. Were this to be so (which it is not), the present Owen Falls Dam would need afish pass, as would Owen Falls Extension. This is not necessary and a Bujagali Fish ladder is not scientificallyjustifiable. (emphasis added). A barrier in the Upper reaches up to Dumbbell Island would not significantly affect the stability of fish populations in Lake Victoria and neither wouldafish ladder be relevant." 159. The studies undertaken by, and the formal indicative position of, the Ugandan NaFIRRI are substantial and professional. Bank Management exercised diligence 49 inusingthese documents inits decision-making. The Panel consequently finds Bank Management acted consistently with the provisions of OP 4.01 and OP 4.04 in so far as these relate to assessment of the likely consequences of the Bujagali hydropower Project on fish stocks in the Upper Victoria Nile and Lake Victoria. B. Mitigation Measures: The Kalagala Offset Agreement 160. The Requesters express concerns about the agreement between the World Bank andthe GoU stating that the "Government of Uganda undertakes that anyfuture proposal which contemplates a hydro power development at Kalagala will be conditional upon satisfactory EIA being carried out which will meet the World Bank Safeguard Policies as complied with in the Bujagali Project. Government and the World Bank will jointly review and jointly clear such an EIA." In the Requesters' opinion this agreement i s not a guarantee that the Kalagala Falls will never be developed for hydropower. 161. In its Response, Management claims that the offset provision related to the Kalagala Falls "will be included as a GoU obligation in the IDA Indemnity Agreement for the Bujagali project, and will be binding throughout the life of the Indemnity."'28 This, in Management's view, i s in compliance with OP 4.04 on Natural Habitats. Management also notes that, because the Bank's legal resort to enforce the Government's commitment i s not available after the termination o f the Indemnity Agreement, this agreement "includes a provision that, prior to the termination of the Indemnity Agreement, the WorldBank and the GoUwillpursue discussions to identifL mechanisms or instruments to enable the continuation of the GoUobligation to set aside the Kalagala Falls site."'29 162. According to OP 4.04 onNatural Habitats, the policy o fthe Bank i s to support the protection, maintenance, and rehabilitation o f natural habitats and their functions in the Bank's work. The policy further states that in project design and implementation "the Bank does not support projects involving the significant conversion of natural habitats unless there are no feasible alternatives for the project and its siting, and comprehensive analysis demonstrates that overall Further, it states that "if the environmental assessment indicates that a project benefits @om the project substantially outweigh the environmental costs."13o would significantly convert or degrade natural habitats, the project includes mitigation measures acceptable to the Bank. Such mitigation measures include, as appropriate, minimizing habitat loss (e.g., strategic habitat retention and post- development restoration) and establishing and maintaining an ecologically 12'Management Response, p. 22 and p. 10 7 28 129Management Response, p. 22 and p. 10 7 28 130OP 4.04 75. 50 similar protected The Bank, however, "may acce t other forms of mitigation measures only when they are technically j~stiJied."`~f 163. According to Project documents, the inundation o f the riverbank portions o f the Jinja Wildlife Sanctuary and Nile Bank Central Forest Reserve, as well as the islands between the sections o f Bujagali rapids, i s technically necessary for the hydropower project. As a result, there will be an irreversible impact to natural riverine forest as well as aquatic habitats. For natural habitats, OP 4.04 allows such impact to be mitigated by establishing and maintaining an ecologically similar protected area. The phrase "establishing and maintaining an ecologically similar protected area" has come to be known as "an offset."'33 Kalagala Falls, a site with hydropower development potential, was agreed between the GoU and the Bank to be an appropriate offset for the natural habitats that would be inundated bythe Project. Picture5 Kalagala Falls 164. Considerable correspondence pertaining to the so-called "Kalagala offset" took place between the GoU, the project sponsor and the World Bank at the time o f the 2001 proposal to develop a hydropower facility at Bujagali. This correspondence i s reproduced as Appendix D 1 o f the SEA for the Bujagali Hydropower Project. Appendix D 2 o f this SEA provides a copy o f a letter dated September 15, 2006 from the Ugandan Electricity Regulation Authority refusing a potential project sponsor permission to conduct investigations at Kalagala with a view to establishing the site's power generation potential. In this letter it is stated that: 13'OP 4.04 75. 132OP 4.04 7.5. 133Management Response, pp. 21-22,24. 51 "The Government position on the site is that it continues to be frozen for developmentpurposes."' 34 165. However, the Requesters believe that the assurances given by the GoU in the correspondence with the World Bank "...are not a guarantee that Kalagala Falls would never be developedfor hydropower. The commitment on Kalagala Falls as an `Off-set' by government of Uganda is not binding. It does not completely removeKalagala as afuture dam site."'35 166. Inorder to meet the requirements o f OP 4.04 the World Bank has conditioned its participation inthe Project as follows: "...the long termprotection of the Kalagala Falls and thepreclusion of development of hydropower potential at Kalagala is a necessary offsetfor WorldBank Groupparticipation in theproposed 167. Inthis context, an Indemnity Agreement was entered to between the Republic o f Uganda and IDA on July 18, 2007, in consideration o f IDA providing a guarantee inconnection withthe P r 0 j e ~ t .The~IndemnityAgreement providesthat l ~ Ugandashall: (a) set aside the Kalagala Falls Site exclusively to protect its natural habitat and environmental and spiritual values in conformity with sound social and environmental standards acceptable to the Association. Any tourism development at the Kalagala Falls Site will be carried out only in a manner acceptable to the Association and in accordance with the aforementioned standards" The same paragraph of the Indemnity Agreement provides, however, that ".. Uganda also agrees that it will not developpower generation that could adversely affect the ability to maintain the above-stated protection at the Kalagala Falls Site without the prior agreement of the Association. (emphasis added) In other words, the possibility o f a power generation development at the Kalagala site is not precludedbut rather subject to the Bank's agreement. 168. During Panel interviews with Bank Management, Government of Uganda officials and the Bujagali Project Sponsor,'38 it was evident that the "Kalagala 134 In a November 2007 interview with the Inspection Panel the official responsible for drafting the Kalagala Offset agreement stated that this provided a Ugandan Government commitment that no hydropower facility would be developed at Kalagala and that it effectively takes Kalagala "off the desks of planning officials". 135Request, p. 5. 136PAD, Annex 15, p. 155. 137Indemnity Agreement (Partial Risk Guarantee for the Private Power Generation (Bujagali) Project) between the International Development Association and the Republic o f Uganda, dated July 18, 2007 (hereinafter "Indemnity Agreement"). 52 Offset" has come to be accepted as a site to be used to "offset" a variety o f the features that are to be lost by inundatingthe Bujagali rapids, but there i s almost no mention o f the core purpose o f a conservation for lost natural habitats as provided by Bank policy on Natural Habitats.'39 During its investigation visit, the Panel observed uses at Kalagala Falls that are not necessarily consistent with this conservation purpose. The visit to the site served to confirm that tourism i s actively being promoted but that the natural habitats at Kalagala are not being maintained as requiredby paragraph 5 o f OP 4.04. Removal o f natural vegetation and subsequent burningand cultivation o f the western bank o f the Nile was seen, as was an apparently new structure on one o f the islands. 169. The Project SEA also reports that a rafting company has "been awarded a concession from the National Forestry Authority to operate a high quality eco- tourism Lodge on Kalagala Island, within the Kalagala-Itanda Offset area, which will involve an investment of USD 1 million in association with international partners ... "140 The Panel fmds that there is evidence that an offset has been created, to meet the requirement of OP 4.04, and notes the efforts of Bank Management to this end. On the other hand, the Panel finds that there is evidence that the offset site i s not being subject to appropriate conservation and mitigation measure^'^' in conformity with sound social and environmental standards. The Project is thus not in compliance with OP 4.04 on this point. 170. Paragraph 6 o f OP 4.04 provides that: "ifthere arepotential institutional capacity problems, the project includes components that develop the capacity of national and local institutions for effective environmental planning and management." The Panel finds that the capacity o f local institutions to plan and manage the Kalagala offset has not been developed and that no provision has been made to rectify this. As a consequence the Kalagala offset may not achieve the purpose for which itwas set aside, andthis is not consistentwith the provisions of OP4.04. 171. In addition to the Kalagala offset, mitigation measures that will be undertaken within the Jinja Wildlife Sanctuary and Nile Bank Central Forest reserve include enhancement planting on the residual islands and inthe 100 mriparian strip along the reservoir margins. This i s for erosion control and general catchments protection, but also to offset the loss o f ecological habitat on the Bujagali islands and riverbanks as a result o f the Project. This planting will be undertaken in consultation with landowners and with National Environmental Management Authority (NEMA), the government authority charged with management o f this 13*Interviews inWashington DC andinUganda, December 2007. 13' In an interview with the Inspection Panel (Entebbe, November 2007) an official of the National Environmental Management Authority stated that he understood the Kalagala Offset was primarily to accommodate tourism activities displaced from Bujagali and that the offset agreement allowed for eco- tourism development on the Kalagala islands. Replacement plantings for lost riverine forest were seen to be the responsibility o f BEL, overseenby the National Forestry Authority. 140HPP-SEA, p. 148 14'OP 4.04 76. 53 area.'42 Seedlings are to be sourced from NGOs as well as from the National Forest Authority and local peopleespecially women-will be employed to plant and tend the plantings. To encourage positive engagement o f local people fruit trees for their use will be included inthe mix o ftrees to be planted.143 172. The success o f such enhancement planting will be heavily dependent on adequate husbandry being provided until the seedlings are established and thereafter to ensure that saplings are not harvested for poles or firewood. Appropriate management and oversight o f the enhancement plantings will be required. The Panel notes with concern that the proposed Environmental Mitigation and Monitoring Plan'44is silent on the need for monitoring of enhancement and offset plantings. Also, monitoring of replacement plantings has not been included in the terms of reference of the witness NGO that has been appointed to monitor Project compliance with IDA conditionalities. This is not consistentwith the provisions of OP 4.04. C.Safety ofDams 173. The Request claims that the safety issues regarding the Nalubaale dam at the Owen Falls are not taken into consideration in the Bujagali dam design. The Requesters raise the issue o f whether the Bujagali dam would be able to survive a failure o f the Owen Falls dam. The Requesters do not consider sufficient the proposal to form a dam safety panel, because they believe a comprehensive plan and strategies to address these issues should be integrated into the Project design. They argue that these strategies are very important, since there was no Environmental Impact Assessment for the Kiira Dam or a post-construction audit for the NalubaaleDam. 174. Management responds that dam safety i s an integral part o f the review o f any hydropower development, that a Dam Safety Panel has been established to provide advice through design, construction, filling, and start-up to ensure that the project i s consistent with Bank policies. 175. For largedams such as Bujagali OP 4.37 requires: a) reviews by an independent panel of experts (the Panel) of the investigation, design, and construction of the dam and the start of operations; b) preparation and implementation of detailed plans: a plan for construction supervision and quality assurance, an instrumentation plan, an operation and maintenanceplan, and an emergencypreparednessplan; 14'HPP-SEA, Section 7.5.2.3. 143HPP-SEA, Section 7.5.2.3. 144HPP-SEA, Section 8.3.7. 54 e) prequalijcation of bidders during procurement and bid tendering.145 Paragraphs 7 and 8 o fthe Policy provide that The Bank may jnance ... diversion dams or hydraulic structures downstream from an existing dam or a DUC [dam under construction], where failure of the upstream dam could cause extensive damage to or failure of the new Bank-funded structure; ... Ifsuch aproject ... involves an existing dam ... in the borrower's territory, the Bank requires that the borrower arrangefor one or more independent dam specialists to (a) inspect and evaluate the safety status of the existing dam ... its appurtenances, and its performance history; (b) review and evaluate the owner's operation and maintenance procedures; and (e) provide a written report offindings and recommendationsfor any remedial work or safety-related measures necessary to upgrade the existing dam ...to an acceptable standard of safety.146 The OP fbrther states that The Bank may accept previous assessments of dam safety or recommendations of improvements needed in the existing dam or DUC if the borrower provides evidence that (a) an effective dam safety program is already in operation, and (b) full-level inspections and dam safety assessments of the existing dam or DUC, which are satisfacto to the Bank, have already been conducted and documented.1 2 ) 176. As part o f the dam safety review required for the Bank-supported Uganda Power I11Project148a review of the safety o f the original Owen Falls dam indicated that the 1940's design o f the dam was inadequate to meet current safety standards. Remedial work to bringthe dam up to modem safety standards was thus required: this was financed by the Bank under a supplemental credit to complete the Power I11project. In all Bank projects related to the Owen Falls Dam and to the Owen Falls Extension Project (Power 11, Power 111, Power I11supplemental credit, and Power IV), the InspectionPanel found in2002 that the provisions o f the Policy on 145OP4.37 514. 146OP4.37 M[7& 8. '47OP4.37 19. 14'The Uganda Power I11Project is also referred to as the Owen Falls Extension (now known as Kiira). Supported by IDA, the project included the construction of a powerhouse, the installation of two 40- megawatt generating sets, the provision of remedial works at the Owen Falls Dam, and the provision of technical assistanceto the Uganda Electricity Board. 55 Safety o f Dams had been fblly addressed. Appropriate professionals had been appointed for design and construction, an independent panel o f experts had been appointed and used to advise on the best way to bringold structures up to modem standards, operational and maintenance plans had been prepared and implemented, an emergency preparedness plan was in place and routine independent safety checks were being carried out. The Bank had also appointed its own experts to review the reports o f the project's independent panel o f experts. 177. Similarly, for this Project, the lenders appointed their own expert advisors to review the report o f the project's expert panel on dam safety. The Inspection Panel's expert has reviewed this report and accepts that "the situation at Owen Falls does notpose an unusual risk to theB~jagaliproject."'~~ 178. The Panel expert also studied the report commissioned by the lenders to review preliminary dam design, including an evaluation o f flood risks in the event of catastrophic failures. The report finds that the design o f Bujagali i s consistent with industry design practice. Nevertheless it recommends that further studies be conducted to determine whether any human settlements would be affected by flood waters consequent upon a catastrophic dam failure or fiom sudden increases inriver flow that mayoccur when the siphon spillway operates. 179. The Panel visited the Nalubaale complex in December 2007 and was shown the cracks in the powerhouse as well as the routine measurements o f structural movement and o fpore-water pressure that are undertaken and reported. The Panel expert i s satisfied that Eskom (Uganda) i s undertaking and reporting the monitoring o f the Nalubaale complex that the Bank requires. The Panel notes that the cracks are in the powerhouse structure and not in the wall o f the dam. The Panelfinds that Managementhas complied with the proceduresset forth in OP4.37. 149Colenco Power Engineering Ltd., Bujagali Hydropower Development, Uganda: Project Review and AssessmentReport(seconddraft), Feb. 2007. 56 Chapter IV HydrologicalandClimateChangeRisks A. Introduction 180. This chapter analyzes the issues o f hydrological and climate risk raised by the Requesters, specifically the impact o f hydrological regimes on energy output, the impact o f the Project on lake levels, and the impact o f climate change on the hydrology o f Lake Victoria and the Victoria Nile Riverandthus on energy output. 181. According to the Requesters, BEL'SSEA does not adequately address the issues o f possible hydrological changes affecting power production at the Nalubaale, Kiira and the Bujagali facilities, especially when Lake Victoria water levels decline. The Requesters state that BEL has little or no control on the manner in which Nalubaale and Kiira will be operated and cannot control the outflow o f water from the power stations upstream. Further, BEL had not taken into account Lake Victoria's diminished hydrological state and its flow regime changes. As a result, the Requesters believe power-generating capacities o f the Bujagali Dam are overestimated and the dam will not be able to operate to achieve its designed capacity under the current hydrological regime because there will not be enough water for this purpose. They also contend that the environmental studies do not assess the possible scenario o f the Bujagali Dam providing further incentives to release higher flows. 182. Management states that the impact of hydrological flow rates on the planned Bujagali Dam has been addressed extensively: an analysis o f the lake's hydrology and its impact on power generation at Nalubaale, Kiira and Bujagali, which complements the SEA, i s included in the study "Bujagali II-Economic and Financial Evaluation Study" (Section 2: Hydrology and Energy Generation o f Hydropower Plants), known as the Economic Study. 183. According to the abovementioned studies, "the proposed 250MW project is not expected to signijkantly alter or affect the hydrology of Lake Victoria or the Victoria Nile." The Bujagali Dam and its energy output are based on water releases from Lake Victoria consistent with the Agreed Curve and on the assumption o f a low flow regime occurring during the first 20 years o f operation at about 79 percent probability. The amount o f water released from the Lake and the timing o f this release will be controlled through operating the Nalubaale and Kiira facilities. Dataused to assess the hydrology o fthe Lake comprises 106 years o f data, including several hydrological cycles, which were considered adequate and sufficient to determine the hydrological risk for energy generation. 184. Management acknowledges that in recent years the "GoU over-abstracted water for power generation" because o f a general drought, lack o f generation investments and a demand growth of 8 percent. However, it also states that the 57 "GoU has steadily decreased hydropower generation in an effort to return to the Agreed Curve operating regime. Water flows for power production are bein scheduled so as to return to the Agreed Curve as soon as reasonably possible."' 5% Management also recognizes that BEL will not control water released from the Lake but argues that the Government has an interest in ensuring that the three facilities are operated inan efficient way. 185. The remainder o f this chapter i s divided in four sections. Section B analyzes the hydrology o f Lake Victoria and the Victoria Nile, with particular attention to the appropriateness o f the hydrological data used in project design and the recent changes in Lake Victoria levels and their relationship to power plant operations. Sections C, D and E analyze, respectively, the impact o f hydrology on energy output; the assessment o f the impact o f the project on lake levels, including the impact o fthe Project on lake levels, and climate change risks. B.The Hydrology of LakeVictoria and the Victoria Nile 186. This part o f the Report analyzes the adequacy o f the Project's assessment o f the hydrological risk and impact o f the Project on Lake Victoria and the Victoria Nile. To do so, the Report first examines whether the 1900-2005 hydrological data series used inProject design i s appropriately representative o f long-term lake level and flow conditions-Section 1 below. A discussion o f the changing conditions o f the Lake's water levels in the last few years (2000-2005) and the extent to which this may be related to the operation o f the Nalubaale-Kiira system follows inSection 2. 1.AppropriatenessofHydrological Data SeriesusedinProjectDesign 187. To determine whether the 1900-2005 hydrological data series used in Project design i s truly representative o f long-term lake level and flow conditions, it i s important to take into account that, as noted in Chapter 11, observers generally divide the history o f Lake Victoria's water levels into three main periods (though the hydrology o f the Lake and the outflow from the Lake Victoria have long been a topic on which hydrologists and engineers di~agree'~'). 188. In general, the period before 1960 i s characterized as a period o f relatively low water levels. Between January 1960 and June 1964, the lake level increased about 2.5m for a total volume increase o f 170 x 109m3.Between 1960/61 and 1999, Lake Victoria had much higher average inflows (around 1200 m3/s,or nearly double the average inflows in the previous period), and the Lake level rose. In contrast, startingin2000 and until very recently, Lake levels and net inflow again decreasedto a level observed before the 1960s (see Figure 2). ' 150Management Response, p. 18. 15'InspectionPanelInvestigation Report 2002,fSO. See also Kull2006, p. 10. 58 189. Between 2001 and 2004, outflows from Lake Victoria, which were 15 percent above the average inflow for 1950-2000, exceeded net inflows and lake levels de~1ined.l~~The lake cannot maintain its water level, if human controlled outflows are higher than naturally occurring inflows. I 1900 1920 1 9 4 1960 1980 LOO* Figure 2 Lake Victoria Water Levels from January 1900 to January of 2005 (Source: Lake Victoria Basin Commission) 190. As also noted in Chapter 11, the main input o f water to Lake Victoria i s rainfall directly onto the Lake's surface (significantly greater than basin inflow), and the main loss o f water from the lake i s evaporation (significantly greater than outflow down the Victoria Nile). The amounts o f direct rainfall, basin inflow, evaporation from the lake, and outflow via the Victoria Nile from 1950 to 2004 are provided inTable 1 below, which summarizes the water balance for Lake Victoria during this period. I Table 1Summary of Water Balance for Lake Victoria153 1950-2000 I 12001-2004 I I 1950-2004 I lS2 Lake levels at the Jinja gauge reached a low point o f 10.4m inOctober 2006. Since that date lake levels have beenrising. lS3Table 2, p. 8 Special Report on the Declining of Water Levels of hake Victoria, East African Community, Lake Victoria Basin Commission (2006). 59 191. After reviewing the striking variations in the net inflow o f water into the lake'j4 between the three key periods dating back to 1900, the Project Economic Study concluded that "[. ..] the whole period of record from 1900 should be used to determine the future dependable flow for power generation at hydro power stations on the Victoria Nile."'55 192. Hydrology studies conducted by Acres in 1990,'j6 which analyzed the feasibility of two hydropower plant alternatives, the Owen Falls dam extension (Kiira) and the Bujagali Falls Hydropower Plant, concluded that the low flows observed before 1961 (when the mean outflow was 660 m3/s)were not truly representative o f long-term flow conditions; as a result, these studies considered only the hydrological series 1961-1989, during which the mean outflow o f water was 1,200 m3/s,as a valid basis for Project design. Incontrast, the 1993 studies o f the Nile Basin by the Institute o f Hydrology o f the United Kingdom (IOH, Wallingford, England) concluded that the increases recorded in the period 1961- 1964 were due to an increase o f rainfall in the basin rather than an error in the hydrological series. This period was therefore considered not representative o f a 154As explained earlier, the net inflow of water into the lake is often termed Net BasinSupply (NBS). N B S = rainfall- evaporation + basin inflow, and indicates the net amount o f water which enters inthe lake after accounting for evaporation. Ina dry year evaporation increases and rainfall decreases, which can result ina near zero or negative value o fNBS. 155Power Planning Associates Ltd, Bujagali I1- Economic and Financial Evaluation Study, Final Report, February 2007 (hereinafter "Economic Study") Executive Summary, p. 4. The review was carried out by Power Planning Associates (UK),inconsultation with Coyne et Bellier (France) and ECON (Norway). The study also concluded that the Institute of Hydrologyof the UnitedKingdom (IOH) series was a reliable one since other rivers showed conditions similar to those observed inthe periods 1900-1961 and 1960-1964. 156Acres International Ltd., Proposed extension to Owen Falls Generating Station: Feasibility Study Report, Oct. 1990. 60 long term hydrology for the Lake: both low and high flows could occur in the future. Subsequent studies by EDF (Electricite de France) in 1998 and Knight Piesoldin 1999 confirmed this analysis. 193. The Project's Economic Study compared the outflow hydrologic series obtained by IOH and Acres and analyzed it based on the flow of other rivers inthe region. It concluded that the IOH series was more reliable, since the data recorded during the period 1900-1 961, and the chan es in 1960-1 964, reflected similar conditions in these other rivers. Other studiesR7 also showed that the net inflows o f water into the Lakes Victoria, Kyoga and Albert exhibited similar behavior for the period of high flow as compared to the previous period. Other authors showed that a period o f highlevels had also occurred in the 19thcentury, though most o f the time the levels were similar to those inthe period 1900-1960 or lower'58. 194. The Management Response indicates that apeer review analysis o f the Economic Studywas prepared by Prof. Juan Valdes from the University of A r i ~ 0 n a .This l ~ ~ independent review was financed under the BNWPP,16' inter alia, to "expand the knowledge on the projections pertaining to Lake Victoria hydrology, [and] provide an important second opinion on some of the key assumptions with regard to hydrology for both the proposed Bujagali and Thermal Generation operations." This independent analysis studied the hydrological series and compared it to other rivers in the regions and concluded that the variability exhibited inthe data series was natural and recommendedthe use ofthe full series inthe future analysis. 195. The Panel's hydrology expert has concluded that the hydrologic data sets used in Project design constitutes a reliable data series and its variability over time is a natural condition, which can be observed in other hydrologic series of different parts of the world, when the hydrologic series i s long enough. The Panel finds that this provides an appropriate baseline for analysis of environmental and economic issues, incompliance with OP4.01. 2. Lake Victoria Water Levels and Power Plant Operations on the Victoria Nile 196. The Requesters contend that the Project will have severe negative impact on the long-term health o f Lake Victoria because the addition o f the Bujagali 15'Johan Grijsen, Potential Impacts of Hydrologic Uncertainty and Climate Change on Regional Power Options in the Lake Victoria Basin Presentation made at World Bank Water Week, February27-March 2, 2007 (hereinafter "Grijsen 2007"), available at: http://siteresources.worldbank.org/INTWRD/Resources/Johan Griisen Hvdro1ogicuncertaintv.pdf (date accessedJuly 30,2008). lS8Nicholson, S.E., Yin, X.; BA, M.B.2000. On the feasibility o f Using a Lake Water Balance Model to Infer Rainfall: An Example from Lake Victoria. Hydrological Sciences Journal, N.1Vol 45, February, p 75-95. lS9Juan B. Valdds, Evaluation OfHydrology of Bujagali (Uganda) Hydropower Project, Sept. 17,2006. 160Further information about the B W P in "Hydrology of Lake Victoria and the Victoria Nile, and Hydropower Implications" inChapter I1o fthis Report. 61 hydropower plant to the plants already operating-Nalubaale and Kiira-will serve to increase social and political pressure for water to be released above the Agreed Curve so as to meet electrical demand. 197. This i s denied by Management, which argues'61that the Bujagali project uses the same water as the plants already operating and thus would have no additional impact on the levels of water inLake Victoria. Management indicates that: "Since the project is located downstreamfrom the NalubaaleKiira dam complex, it will use the same water that has already been released through NalubaaleKiira and, given the project's higher head, will allow Uganda's eneration output to more than double without any additional release of water."16 f 198. This section analyzes the changing conditions o f the Lake's water levels in the last few years, and the extent to which this may be related to the operation o f the Nalubaale/Kiira system. 199. As indicatedearlier, the Agreed Curve (a mathematical relationship between Lake level measured by a gauge at Jinja and outflow), has been used to specify the outflow that should be released from Lake Victoria down the Victoria Nile.'63 Between 2000 and 2006, outflow exceeded the Agreed Curve due to Ugandan demand for electricity.164 200. In the period immediately before 2000, flow releases from Lake Victoria were less than those required by the Agreed Curve, in order to minimize the effects o f floods d~wnstream;'~~water not released was thus held in storage in the Lake. In addition, since the only dam operating during this period was Nalubaale, not all the flow released went through the turbines, meaning that part of the flow was released downstream through the spillways without generating energy. During this "high Net Basin Supply"'66 period, therefore, the Lake was used as a reservoir for dampeningfloods. 201. AAer 2000, the entry into operation o f Kiira increased the generation capacity of the Nalubaale-Kiira system. Since these two dams operate in parallel to one another, the system required more water to flow downstream and through the turbines to generate energy. Unfortunately, this development coincided with a period o f low Net Basin Supply, and the lack o f inflow water combined with the need for greater releases started to decrease the lake levels. In July 2001, the 16'Management Response,pp. 18-19 16*Management Response, 140. 163The relationship is: Q = 132.924(h -8.486) "1.686 where Q is discharge incubic meters per second and and h is water level (stage) inmeters at the Jinja Pier. 164InJune 2006 outflows were cut back to alignthese with a fixed 750 m3ls discharge. 165Dropping Water Levels o f Lake Victoria, Technical Note, Ministry o f Water Lands and Environment, Directorate o f Water Development (DWD), Water Resources Management Department, 2005, p. 27 (hereinafter, "DWD 2005"). 166As noted in Chapter 11, rainfall plus basin inflow minus evaporation is referred to as the "Net Basin Supply." 62 additional volume retained in 1998-2001 period began to be released and the lake started to decrease more than it would have if its flow had been regulated by the Agreed Curve. All releases went through the turbines and the total flow was greater than the flow which would havebeenreleasedunder the Agreed Curve. 202. Table 2 below shows the yearly reduction o f water levels in Lake Victoria during the period 2002-2005 and the flows released during these years in comparison with the Agreed Curve. The large differences were in 2004 and 2005-and mainly in 2005, because in that year, the Net Basin Supply was near zero due to drought conditions, which meant that a large volume o f water from the lake was released downstream over and above the Agreed Curve amount. Table 2 Changes inLake Levels, and Flow Releases Over and Above the Agreed Curve'67 Year Lake level Increase ofmean Proportionof Net Basin Supply Net Basin Supply decrease flow releaseover total release (m3/s) as proportion o f (cm) and above the ("/.I long term mean Agreed Curve ("/.I (m3/s) 2002 6 170 14.5 2003 10 238 19.8 693 80 2004 26 538 41.5 461 53 2005 27 561 47.7 31 4 203. Figure 3 shows the lake levels and outflows for the period. It can be seen from any o f the points o f the curve that the outflows are above the Agreed Curve. For instance, in July 2004, when the Lake level was 1134.25 m, the flow released according to the Agreed Curve should have been 802 m3/s,while the actual outflow exceeded 1300m3/s. Figure 3 Levels at the Lake and Outflow (2000-2005) (Lake Victoria Basin Commission, 2006 brnpmrison of M h r Level and OuUIow 15000I __._ - T 11R600 Economic Study 63 204. The PAD states that: "Since the end of 2005, the Government has steadily decreased hydropower generation in an effort to return to the Agreed Cuwe operating regime. Waterflows for power productions are being scheduled in such a way that the return to the Agreed Curve is achieved as soon as reasonably possible.168 (emphasis added) Figure4 ComparisonofWater LevelandOutflow Cornprisonef -tar Level and OuUlow 205. An analysis 15000 173600 14oO0 o f the 13000 113550 long 1 m o term E 11wo hydrolo loo00 1135w E 9000 1130503g gic data 8000 -1 series'69 7000 1134w shows that approxi mately 5 percent o f the time the Net Basin Supply to Lake Victoria i s negative-meaning that any outflows to the Victoria Nile come from water stored inthe lake. Inthese years, if the waters released are according to the Agreed Curve, the lake level would decrease, since outflow would be greater than net inflow. The dam operation effect-that is, beyond the natural conditions-would occur when the release i s greater than that specified by the Agreed Curve. 206. The effect o f the release policy during the period 1998-2001, which resulted in holding the flow in the lake, had downstream benefits in mitigating floods. Likewise, in the period 2001-2005, the increase o f the water release o f the lake above the Agreed Curve had downstream benefits (increased energy production), butnegative upstream effects (lake depletion). 207. The Panel notes that the Agreed Curve constrains the ability to use the lake to store "excess" water for later use when inflow exceeds outflow. During Panel interviews in December 2007, responsible authorities in the Government o f Uganda noted the need for the Agreed Curve to be understood as a tool for water resource management rather than simply a mechanism to determine volumes o f water to be released, and indicated that a new mechanism for determining water release from Lake Victoria to the Nile, based on maximum benefit to all riparian countries, i s in the process o f being developed. The World Bank-Netherlands PAD, p. 37. 169DWD2005. 64 Water Partnership (BNWPP) background description for the 2006 "Victoria Nile- Independent Hydrological Review" activity, referred to earlier, states that "partly because ofpressure from the riparian states, Uganda has been sensitized to the importance of making cogent choices between reverting to the Agreed Cuwe policy, or adopting some other water management policy that would be no more harmful to its 208. In Panel discussions with the NBI, it was indicated that the notion o f managing the waters o f Lake Victoria as a resource for all riparian countries was integral to a new treaty that i s being drafted to replace the numerous existing treaties and accords that relate to use of Nile waters. This would mean a move away from water releases dependent on lake level to variable releases based on water demand management and an increase in the "balancing times" from 10 days-as at present-to seasonalor even annual accounting. 209. During its field visit in December 2007, the Panel was given documentation showing what appears to be a new release policy, whereby discharge was fixed at either 850 or 750 cubic meters per second depending upon the level o f the lake. The Panel received information suggesting that this new rule, which allows for a constant release to be applied when the lake level fluctuates within a certain range, with mean outflow the same as under the Agreed Curve rule, has been in effect since June 2006, and it i s the basis for the analysis inthe Economic Study. C.Impact of hydrologic risk on energy output 210. The Requesters state that: "Even the recently (26thFebruary, 2007) released economic analysis does not adequately address the economic viability in relation key elements assessed in the economic analysis include: [...] (iv) the hydrology of addresses the economic viability and risk analysis of the Bujagali project. [...]The to hydrological risks."'71 Management considers that "The Economic Study [...] Lake Victoria and its impact on hydropower generation...."172 Management also with relevant World Bank Group policies [...I." says "that the economic [...] and other required analyses to date are compliant 173 IFC appointed consultants to carry out the Economic and Financial Evaluation Study, inthis Report referred to as the "Economic Study," in January 2006 and the final report i s dated February 2007.'74 211. Bank Economic Evaluation policies applicable to this Project are OP 10.04 on Economic Evaluation o f Investment Operations. OP 10.04, provides inparagraph 170World Banl-Netherlands Water Partnership (BNWPP), background description for the "Victoria Nile- Independent Hydrological Review" activity, available at: httv://www-esd.worldbank.org/bnwtlv/index.cfm?disvlav=disvlayactivitv&AID=439 (accessed on 23 July 2008). 17'Request, pp. 3 4 . See also p. 7. 17'Management Response, Annex 1, p. 25. 173Management Response,733. 174Economic Study. 65 1 that "For every investment project, Bank staff conduct economic analysis to determine whether theproject creates more net benefits to the economy than other mutually exclusive options for the use of the resources in question." The Policy then sets out specific provisions in seven areas: Criterion for acceptability, alternatives, non-monetary benefits, sustainability, risks, poverty and externalities. 212, Paragraph 2 o f OP 10.04 defines the Criterion for Acceptability of a Project on economic grounds inthe following way: "a project must meet two conditions: (a) the expectedpresent value of theproject's net benefits must not be negative; and (b) the expectedpresent value of theproject's net benefits must be higher than or equal to the expected netpresent value of mutually exclusive project alternatives." Note 3 states that "standard practice has been to calculate the expected internal rate of economic return [. ..I." Paragraph 3 lays out the importance o f the analysis o f alternatives "to ensure that theproject maximizes expected net present value," while Paragraph 5 calls for an analysis o f the sustainability o f a project to make sure that its "benefits will materialize as expected and will be sustained throughout the life of theproject." 213. The hydrology o f Lake Victoria, along with the water release regime, i s a key influence on the potential energy output o f hydropower plants on the Victoria Nile. Annex 10 o f the PAD discusses the detailed review o f Lake Victoria hydrology inthe Economic Study (100 pages, including tables and charts).'75 The main objective was "to assess the performance of Lake Victoria, by deriving the longest reliable series of Net Basin Supply (or net injlow into the lake) that should be usedfor the evaluation of energy generation of the existing andforeseen hydro power projects on the VictoriaNile [. ..I." secondobjectivewas to investigate The the causes o f the recent drop in lake level because this analysis would be "[. ..] helpful in understanding the key drivers in the hydrological performance of Lake ~ i c t o r i a . ~ ~ ' ~ ~ 214. Hydrology scenarios and their probabilities: as noted earlier, the Economic Study concluded that "[. .] the wholeperiod of recordfrom 1900should be used . to determine the future dependable jlow for power generation at hydro power stations on the Victoria Nile."177To reflect the variations in Net Basin Supply among the three key periods since 1900, the study defined two hydrology scenarios for the 20 year period that would follow the commissioning o f Bujagali: the Low Hydrology Scenario (average net inflow 660 m3/s) and the High Hydrology Scenario (average net inflow 1200 m3/s).The analysis assessed the probability o f their occurrence at 79 percent for the Low Hydrology Scenario and 21 percent for the High Hydrology Scenario.17*The evaluation o f the generation alternatives used these scenarios and probabilities. Economic Study, Appendix B `76 Economic StudyMainText, 73.1, p.41. 177 Economic Study, Executive Summary, p. 4. 17' Economic StudyMainText, p. 45. The process, which averages the results of two different approaches, i s set out inEconomicStudy, Appendix B.5.2, pp. 3841. 66 215. With respect to Agreed Curve the PAD states that, "The planning of the proposed Private Power Generation (Bujagali) Project and the assessment of the energy output have been based on the flow released from Lake Victoria through the Nalubaale/Kiira dam complex in accordance with the Agreed Curve (Annex 10). [...] Theproposed project is designed to be viable with water flows in accordance with the Agreed Curve release rule, since the Nalubaale/Kiira dam complex regulates theflow of waterfrom Lake V i ~ t o r i a . " ` ~ ~ (emphasis added) 216. The Economic Study explains that operating according to the Agreed Curve means that "the lower the lake level, the lower the release, and the higher the lake level, the higher the release." This in turn has a "diminishing" effect in long dry periods because if the net inflow i s lower than the long term average, and the release i s higher than the net inflow, the lake level drops faster and the departure from the Agreed Curve i s augmented, all of which occurred in the period 2003- 2005.'80 The Economic Study goes on to state that reservoir operation modelling was carried out to calculate the "firm" release and the "firm" energy generation in each o fthe hydrology scenarios. The main options were: "(i)return to the strict to commitment to the Agreed Curve, and (ii) to follow the Agreed Curve but in a broader sense, allowingfor a constant release to be applied when the lake level fluctuates within a certain range." It also suggests that the advantage of the "Constant Release" rule i s that it "allows for a better planning of additional means ofpower generation in the country [...]3, . 181 217. Afier describing several features o f this Rule, including "the fact that the firm energy of the "Constant release" rule is much higher than for the "Agreed Curve" rule, although the mean outflow and mean energy of both operations are identical," the Economic Study states that, "Owing to all these advantages,for thepurpose of the economic evaluation of Bujagali, the `%onstantrelease" (or "Agreed Curve by steps,? rule was adopted to determine the energy generation capabili? of each of all hydro options on the Nile downstream of Owen Falls."'8 (emphasis added) The Economic Study then summarises the results of the reservoir modelling and estimates o f firm energy generation for Bujagali and Karuma for the Low and HighHydrology scenarios in Table 3-1, reproduced as Table 10.1 in the PAD. The Economic Study and the PAD confirm that these figures, modified for maintenance (Economic StudyTable 7-6), were "used in the economic evaluation and expansionplan m~delling."`~~ 218. On the one hand, the PAD states that the planning o f the Bujagali Project and the assessment o f the energy output was based on flows released in accordance with 179PAD, p. 37. 180Economic Study, MainText, p. 47. 18'Economic Study, MainText, p. 47. 18*Economic Study, MainText, p. 50. 183PAD, Annex 10, p. 99. See also Economic Study, MainText, p. 51. 67 the Agreed Curve. On the other hand, the Economic Study states that the economic evaluation o f the Bujagali Project to determine energy generation capability was based on the Constant release rule. The Panel notes that this discrepancy between key Project documents brings into question the data basis for the Project's economic analyses, and is likely to have resulted in a more positive conclusion to the Economic Study than would have been the case under the Agreed Curve scenario. This is inconsistent with OP 10.04. 219. In March 2007 an internal Management Review had proposed that the PAD should confirm that the plant would be operated under Lake Victoria's Agreed Curve release strategy, rather than under a constant release regime, "and should confirm that this regime does not affect the conclusions of the economic evaluation of the project;"'84 The PAD does not appear to have followed this latter recommendation. In the Panel's view, the provisions of OP 10.04'85 require Management to provide an accurate picture of the economic analysis (based on the Agreed Curve), and indicate whether this affects the relevant conclusions. 220. The Panel notes that this contradiction in Project documents has a material impact not only on the economic viability of the Project and the provisions of OP 10.04, but also on the lake levels of Lake Victoria, since different operational rules result in different time-profiles and variance of water levels. While the Panel recognizes that, over a certain period o f time, the mean outflow under the "Constant release" rule will be identical to that under the "Agreed Curve" rule, the variation in lake levels under the two regimes will be different.These issues are discussed hrther inSection Dbelow. D.PotentialImpactofthe Project onLakeVictoria 221. The Requesters are concerned about over-draining o f Lake Victoria, which they state causesmisery and economic loss to Uganda and neighboring countries. They believe that the issue o f the long-term health o f the Lake has not been addressed inProject documents "other than to assert that Bujagali Dam could lead to more sustainable flows out of the lake as it will `make use of the same water' released by the existing dams." However, they argue that neither the SEA nor other documents take into consideration the possibility that the opposite will happen because a new dam may create incentives to release higher flows. 222. The Response acknowledges that "since 2003 the GoUover-abstracted waterfor power generation" but notes that the government has in the past few years "steadily decreased hydropower generation in an effort to return to the Agreed Curve operating regime. Waterflows for power production are being scheduled so as to return to the Agreed Curve as soon as reasonablypossible." Management also indicates that with the operation o f the Bujagali/KiirdNalubaale system 184QER Review, March 2 2007 (hereinafter "QER 2007"). 185Which addressacceptability, the analysis o f alternatives and sustainability. 68 "generation of the same energy output as currently generated by Nalubaale and Kiira would only require 45% of the current water releasefrom Lake Victoria." Though BEL does not control the release of water, in Management's view, "it is are operated eficiently. in the interest of the GoU186to ensure that Bujagali and the Nalubaale/Kiira dams 3' 223. The Bank policy on Environmental Assessment requires a Project EA to evaluate "aproject'spotential environmental risks and impacts in its area of influence." A project EA "[plredicts and assesses likely positive and negative impacts in quantitative terms" and "identiJes and estimates the extent and quality of available data, key data gaps, and uncertainties associated with predictions ... ." The area of influence i s defined inOP 4.01, Annex A (Definitions) as Thearea likely to be affected by theproject, including all its ancillary aspects, such as power transmission corridors, pipelines, canals, tunnels, relocation and access roads, borrow and disposal areas, and construction camps, as well as unplanned developmentsinduced by the project ... The area of influence may include, for example, (a) the watershed within which theproject is located; (3) any affected estuary and coastal zone; ... (e) migratory routes of humans, wildlife, or fish, particularly where they relate to public health, economic activities, or environmental conservation ... 224. The Panel notes that the SEA study was based on the assumption that the Project's upstream area of influence ends downstream of Kiira-Nalubaale dams. The SEA does not take into account the Project's potential impacts on Lake Victoria.187 225. The SEA states that "... it is expected that theflow of the Nile downstream of Bujagali will be very similar to theflow downstream of Nalubaale/Kiira, which itselfis still regulated, as it has been since the construction of OwenFalls dam in 1954, by the agreed curve."188As noted earlier, the PAD adds that the "project is designed to be viable with water flows in accordance with the Agreed Curve release rule, since the Nabulaabe-Kiira complex regulates theflow of waterfrom Lake Victoria."189(emphasis added) Management Response,p. 18. lS7HPP-SEA, p. 55, Section 3.2 Project Area o f Influence, defines the Project's area of influence as "including areas affected by: (i) the primary project site, (ii) associatedfacilities; (iii) cumulative effects, and (iv) unplanned but predictable developments." A subsequent table (Table 3.1 Bujagali Hydropower Project Area of Influence) lists ten "primary project sites" as follows: "I.Ladwater areas for dam, its facilities & reservoir; 2. Land for resettlers' houses & livelihoods, as specged, for # I (above); 3. Resettlers' houses, if any; 4. Off-site facilities (quarries, storage, waste disposal, access roads), if any; 5. Air quality & noise effects radii (off-site); 6. Upstream water areas (below NalubaaleKiira; mainly in Bujagali reservoir) & users; 7. Downstream water regime (water quality & flows); 8. Communities (including host communities) as speciJied in PCDP; 9. Stakeholder groups (including vulnerable groups) as identified in PCDP; 10.Projectpersonnel when off-site in project vicinityhegion..." HPP-SEA, p. 361. IS9PAD, p. 37. 69 226. The Panel notes that these statements assume that the natural conditions o f lake level will be maintained inthe future. This in some way may argue for restricting the Project area o f influence upstream at Kiira-Nalubaale and not studying the impact on Lake Victoria changing levels. However, the Panel also notes that this approach-reduced Project area o f influence-does not take into account two important factors: (1) the contradictionbetween the PAD and the Economic Study regarding the Project's operation rule discussed in the previous section, and (2) the recent history o f 2003-2005 when the NalubaaleKiira system was operated above the Agreed Curve, which contributedto a severe depletion o fthe Lake. 227. The Panel notes that the operation policy o f Lake Victoria could be other than the Agreed Curve, usingthe lake as reservoir regulating the flow. However, the Panel observes that any such change in operating regime and its impact upstream and downstream need to have been assessed inthe Project's EA. The Panel also notes that not following the Agreed Curve, with releases greater than the Agreed Curve, could lead to a decrease in the lake's level during a drought period, as happened during 2003-2005. The Panel notes the importance of assessing such a situation and extending the area of influence of the Project to the Lake Victoria. As indicatedinChapter 11, the lowering o f water levels inLake Victoria brings significant social and environmental impacts upon the Lake ecology and the people and countries that rely on it for resources and livelihoods. 228. In this context, the Panel notes a recent Project Performance Assessment Report for the Uganda Power I11Project (Owens Falls Extension - Kiira), prepared by the Bank's internal Independent Evaluation Group (IEG). The Report determined that the project appraisal for Kiira underestimated the criticality o f the hydrological risk related to water level in Lake Victoria. According to the Report, the appraisal concluded that ". . . the likelihood of this risk was less than 1 percent." The Report adds that ``[uhis risk has now been reali~ed.'~~" 229. The Panel notes that the SEA study considered that the Project's area of influence ends downstream of the Kiira-Nalubaale dams."' As a result, the Panel finds that the SEA analysis didnot comply with OP 4.01 indefining the area of influence of the Project because the Project impacts on the changing levels of Lake Victoria were not assessed. 230. Inlight of its relevance to the analysis of the Bujagali Project, the Panelnotes the importance of making the structure for governance of water releases from Lake Victoria clear and transparent to all stakeh01ders.l~~ I9OIndependent Evaluation Group (World Bank), Project Performance Assessment Report Uganda Third Power Project (Credit No. 22680-UG); and Supplemental to Third Power Project Credit (No . 22681-UG) June 26,2008 Independent Evaluation Group (World Bank) 19'HPP-SEA, p. 55. 19'The Panel was variously informed that at present the ultimate authority for determining water releases was: (a) The Commissioner for Water ResourcesManagement; (b) An Inter-Ministerial Committee; (c) The 70 E.Climate ChangeRisks 231. The Requesters aver that the project preparation and assessment reports do not address climate change and its possible impact on power production at Bujagali. They also hold the view that climate models indicate hotter, drier conditions, lower lake levels and lower river flows. Management counters this, stating, "the broader climate change (and hydrology) aspects were addressed in different studies which have also been publicly di~closed."'~~ Management states that the SSEA includes a detailed analysis of the impacts of climate change in the Nile Equatorial region comprising Bujagali.194 232. The Requesters also suggest, that "No study released to date analyses the risks to Bujagali performance climate change-induced drought and other hydrological changes."' lromManagement states, however, that the analyses: "Assessed the impacts of both low and high hydrology scenarios, and separately determined that climate change is not predicted to have a negative impact on water a~ailability."'~~ 233. Climate change risk analysis i s important under various Bank policies. OP 4.01 requires that the Project EA evaluate potential environmental risks and impacts in its area of i n f l ~ e n c e , 'paragraph 5 of OP 10.04 provides that "[tlo obtain a ~ ~ reasonable assurance that the project's benefits will materialize as expected and will be sustained throughout the life of the project, the Bank assesses the robustness of the project with respect to economic,financial, institutional, and environmental risks." (emphasis added). 1.The PAD 234. On the question of climate change, the PAD states that, "The risk of climate change on the hydrology of Lake Victoria was taken into consideration: the conclusion of both the economic study and the Strategic/Sectoral, Social and Environmental Assessment (February 2007) under the Nile Basin Initiative, is that there will be no adverse effect on water release due to climate change during the lifeof theproposedpr~ject."'~~ 2. The Economic Study Department for Water Development; (d) The Ministry o f Water and Environment; (e) The Ugandan Minister o f Water Affairs. 193 Management Response, Annex 1, Section 4, p. 19. Management Response, p. 6. 195 Request, p. 4. 19`Management Response, p. 12 19' OP 4.01 72. 19*PAD, p. 28. 71 235. Inrelationto climate change and whether and how it was taken into account inthe economic study and modelling, the Economic Study Main Text states boldly that, "The possible influence of climatic changes was found not to be signijicant enough in the medium term (to 2030) to influence [in] one way or the other the hydrological sce~larios."~~~ The further discussion o f climate change in the Economic Study i s in Appendix B, where it occupies only one page and two tables. Drawing on papers by Tate, Sutcliffe et a1.,200it concludes, "For both baselines, the 2021-2050 future climatic conditions result in average future outflow smaller than observed outflow [...]. Conversely, the 2070-2099 future climatic conditions result in average outflow greater than observed outflow [...I. For the overall period 2000-2099, the Lake Victoria outflow would be of the same order than the present outflow; [...] Thus, taking the whole of the 1900- 2005 period of record to define hydrolo ical scenarios is acceptable for modelling of future hydrological conditions." O1 In the Panel's view, the brevity f of this discussion of a highly complex issue with the potential to influence significantly the Project's economic outcomes does not demonstrate compliance with OP 10.04's paragraph 5, which requires proper assessment of the robustness of the Project with respect to environmental risks. 3. The SSEA 236. The scope o f work o f the SSEA, as defined inthe terms of reference, includes the followin key task: change." O2 Further, one o f the key-elements inthe analytical approach adopted in 9 ".,.(7) Assessment of the potential impact of climate the SSEA was "Assessment of (calculated/forecasted) climatic changes and runoff due to climate According to the SSEA, this was done because one o f the major risks that were identified was climate change and its possible impact on runoff, which in turn affects the output o f the hydropower development options.204The PAD says that, "The SSEA undertook a thorough analysis of the possible climate change impacts on power development options in the Nile Equatorial Lakes Region, including Bujagali. [...] It used the best available general circulation models to assess the potential changes in temperature and precipitation in 2050 and 2100 relative to 2000. [...] Overall, for the northern and central-west regions of the study area, including Bujagali, there is a high probability of increases in runofi and thuspower generationpotential, compared to historic data. Staff believes that the SSEA incorporated the best currently available climate change science and data in its analysis."205 199 Economic Study, MainText, p. 45. See also PAD, Annex 10, p. 97. 2oo E. Tate, J. Sutcliffe, D. Conway & F. Farquharson, Water balance of Lake Victoria: update to 2000 and climate change modeling to 2100, 49(4) Hydrological Sciences-Journal-des Sciences Hydrologiques, 563, 572 (2004). 201 Economic Study, Appendix B.4, p. 33. 202 SSEA, p. 1-2. 203 SSEA,Executive Summary, p. S-5. 204 SSEA,p.2-3. 205 PAD, p.46. See also PAD, Annex 15, p. 156. 72 237. In contrast with the Economic Study, section 12 o f the SSEA presents a 12-page assessment of, "the potential impacts on hydroelectric generation that could result from climate change, and consequently whether any such impacts could affect the selection and scheduling of new power options in the portfolios being evaluated in the SSEA study." 206 This section draws on Appendix K o f the SSEA, which contains a detailed 50-page assessment and modeling, carried out by the Stratus Consulting team, drawing on results fiom the Third Assessment Report o f the Intergovernmental Panel on Climate Change (IPCC). The final section o f Appendix K includes this statement, "There are few clearly identified hydrological risks to the hydro options included in the indicative plan, and overallfor the Northern and Central West regions there is a high probability of increases in run08 and thus generation, than presently identzj2edfrom historic flow data."2o7 The Executive Summary o f the SSEA repeats the second phrase and goes on to state: "As most of thepower development options that have been retained are located in the northern part of the region, the impact of climatic change will be positive for the development of the portfolios of generation options. No sensitivity analysesfor climate change will be carried out, since they would only present higher energy availability than current conditions indicate."208 238. The Panel has examined the SSEA. The sixteen Global Climatic Models in the NCAR209 suite o f models were examined and seven were selected210as they provided the most representative results to estimate potential changes in temperature and precipitation for the Nile Equatorial Lakes region.211The main conclusions fiom the outputs predictions were that temperature, rainfall, evaporation and runoff are all predicted to increase. The study mentions similar results obtained usingother models.212 206SSEA, p. 12. 207SSEA, Appendix K, p. K-49. 208 SSEA, Executive Summary, p. S-20. SSEA also notes that: "Results show that for all regions flood flows may increase significantly, thus designs for flood discharge during constructionand over a permanent spillway should take this potential into account. Project costs would also be affected." (p. 12-12). *09National Centre for Atmospheric Research, Boulder, Colorado, USA. 210The sevenmodels selectedwere: CERG - The European Centre for Research and Advanced Training in Scientific Computation (CERFACS), France; CCSR -National Institute for Environmental Studies, Japan; CSIRO - Commonwealth Scientific Industrial and ResearchOrganization, Australia; ECHAM3 - Max Planck Institute for Meteorology, Germany; ECHAM4 -Max Planck Institute for Meteorology, Germany; HadCM2 -Hadley Model, United KingdomMeteorological Office; HadCM3 -Hadley Model, UnitedKingdom Meteorological Office. 211This determination was based on a statistical comparison of model ability to simulate current climatic conditions. For the seven selected models a spatial correlation o f 0.94 with a root mean square error o f 0.416 and a difference from the meano f -0.228 was achieved. This is a better correlation and smaller error and difference from the mean than for all 16 GCMs taken together. 212Grijsen (2007) presentsthe results o f climate analysis for the Lake area taking into account the baseline series o f 1956-1978 and for a drought scenario o f 40 percent o f the long term NBS. Rainfall increased by 73 239. Using the mean output o f the seven selected models, general directions and magnitudes o f expected variations were estimated. The expected impacts on power outputs were then considered by assessing the risk that hydrological conditions would be less favorable than conditions estimated using conventional hydrological analyses based on past records of climatic conditions. 240. Chapter 12 o f the SSEA provides an assessment o f the potential impacts on hydroelectric generation that might result from climate change, and examines whether such impacts might affect new power options being evaluated. The objective was not to define whether global warming will take place, but rather to use the results o f existing analyses and predictions in a risk analysis to allow plausible changes from climate change to influenceplanning.213 241. The results of this climate change risk assessment show that there are few identifiable hydrological risks to the hydro-power options studied, and overall for the Northern and Central West regions of the Nile Equatorial Lakes there is a higher probability o f increases in runoff, and thus power generation, than determined from historic flow data.214The peer review analysis o f the Economic Study, prepared by Juan Valdes and referred to earlier, in its analysis of the climate change simulation, notes that most o f the models showed an increase between 7 and 12 percent for precipitation, and temperature increases from 2-Cto 3.5"C in the region for 2100, but concluded "There is considerable variability in the results of the individual models and caution should be used when applying these results to make operational decisions."215 4. Other Documents 242. The Panel examined the Regional Analysis o f the IPCC. Climate Change 2007 presents three working reports: I.The Physical Science; 11. Impacts, Adaptation and Vulnerability; 111. Mitigation o f Climate Change.*16 In report 11, Impacts, Adaptation and Vulnerability, there are two important chapters: Freshwater Resources and their Management217and Africa.218The Freshwater report includes 10 percent inboth scenarios and N B S increased 32 percent for the historic scenario and 83 percent for the drought scenario. 213Details o f the approach and methods usedare provided inAppendix K ofthe SSEA. 214SSEA, p. 12-12 215Valdes, Juan B. "Evaluation of Hydrology o f Bujagali (Uganda) Hydropower Project, September 17, 2006, p. iv. 216Available at: httD://www.iDcc.chiiDccreDorts/assessments-reDorts.htm (accessedJuly 31,2008). 217Z. W. Kundzewicz, L. J. Mata, N. W. Arnell, P. Doll, et al., Freshwater Resources and their Management in Climate Change 2007: Impacts, Adaptation and Vulnerability, Contribution o f Working Group I1to the Fourth Assessment Report o f the IntergovernmentalPanel on Climate Change, 173-210 (M. L.Parry, 0.F.Canziani, J. P.Palutikof, P.J. vander LindenandC. E.Hanson, eds., CambridgeUniversity Press 2007). 218 M.Boko, I.Niang, A. Nyong, C. Vogel; et al., Africa, in Climate Change2007: Impacts, Adaptation and Vulnerability, Contribution o f Working Group I1 to the Fourth Assessment Report of the 74 an index o f vulnerability or stress to water scarcity: according to the report, the study area o f Lake Victoria is not under stress. The Panel notes that the report does not mention any major stress for the Lake Victoria basin as a result o f climate change. In addition, in its specific section related to Africa the report mentions the inability o f the climate models to represent the observed rainfall in the continent.2 243. The Panel also examined the Climate Change Impact Assessment carried out as part o f the Studyof Water Management o f Lake Victoria by Water Resources and Energy Management International Inc. (WREM), which was cited by the Requesters. This study, which used an integrated assessment methodology, concluded that "thefuture climate implies drier hydrologic conditions, lower lake levels, lower outflows, less energy generation, smaller wetland areas, and lower downstream river ~ Z O W S . ~ 'However, the Panel's expert on hydrology has ~ ~ ~ pointed out that the study was based on observed data fkom 1960-1 980, a period during which most years were highrainfallhigh flow years as compared to the 1900-1960 period. This biases the conclusions o f the study, because climate change assessments put forward a relative rather than an absolute scenario, and in this case the relative analysis put forward is with respect to a period o fhighflow. The other analyses used in the SSEA were carried out for the full 1900-2005 series. 5. Conclusions on Climate Change Risks 244. The Panel finds that the possible effect o f climate change on hydropower projects on the Victoria Nile have been considered and well evaluated in the project documents. However, the Panel reiterates that in the context o f climate change, the Bujagali SEA does not refer to the Nile Basin SSEA and does not direct the reader's attention to this important parallel study. In addition, as noted earlier, the brevity of the discussion of climate change in the Economic Study does not demonstrate compliance with paragraph 5 of OP 10.04. 245. The SSEA appraisal appears to be the result of a thorough, detailed studJ that draws on its own analysis and a range of other international studies.2 The Panel fmds that the possible effect of climate change on hydropower projects on the Victoria Nile has been seriously considered inthe SSEA. This analysis meets the requirements OP 4.01. As noted above, however, the SSEA was not properly disclosed a Project document. While both the Economic Study and the SSEA reviewed the potential influence o f climate risks, and Intergovernmental Panel on Climate Change, 433467 (M. L. Parry, 0. F. Canziani, J. P. Palutikof, P. J. van der Lindenand C. E.Hanson, eds., Cambridge University Press2007). 219 I t should be noted, however, that regional simulation analyses, as inthe report SSEA (2007), have more specific simulation and output data than the IPCC reports. 220 WREM International Inc. "Climate Change Impact Assessment - Technical Report 10," Study on Water Management o f Lake Victoria, prepared by Water Resources and Energy Management International Inc. for the Uganda Ministry of Energy and Mineral Development, September 2005, p. v. 221 Including the ThirdAssessment Report of the IntergovernmentalPanel on Climate Change (IPCC). 75 concluded that they would not exert a significant negative influence on the hydrological scenarios, the analysis of the Economic Study does not demonstrate the detailed, sophisticated analysis and modeling that underlay the SSEA appraisal. The Economic Study does not cite or draw on the results o f the SSEA risk appraisal or the detailed reviews in Section 12 and the study inAppendix K that underlay them. Management does not appear to have ensured that the Economic Study drew on the much more thorough analysis inthe SSEA. The Panel finds that this does not comply with paragraph 5 of OP 10.04. Considering that the PAD draws on the authority of both studies, particularly the SSEA, the Panel finds it surprising that the PAD concludes that, "[ ...Ithere will be no adverse effect on water release due to climate change during the life of theproposedproject." 246. The Panel is aware of the limitation of the known technology in evaluating climate change scenarios and that the analysis o f climate change i s an evolving science, where gaps remain. Indeed, this situation makes all the more troubling the PAD'S categorical assertion, without any reference to risk and uncertainty, that there will be no adverse effect on water release due to climate change during the Project life. This failure to express a risk factor is not consistent with OP 10.04. The Panel notes the importance of continued attention and analysis to the effect of climate change on flows and hydropower generation on the Victoria Nile. 76 Chapter V EconomicandEnvironmentalAnalysisofAlternatives A. Introduction 236. This chapter first addresses the economic analysis o f alternatives, including the demand forecast, the consideration o f supply alternatives, the project costs, the assessment o f least cost options for expanding power generation and the economic rate o f return on the Bujagali Project. It then examines macroeconomic impacts and environmental and social costs, and the environmental analysis o f alternatives. Chapter VI addresses poverty reduction, power sector finances and sustainability, the PPA and associatedrisks. 237. In general the Requesters argue that energy alternatives to Bujagali were not adequately addressed in the SEA. For example, in the Requesters' opinion, a hydropower project at Karuma, downstream from Bujagali, would cause less social and environmental harm than Bujagali but was not appropriately taken into consideration as an alternative option. Furthermore, the Economic Study does not include an adequate assessment o f the economic alternatives to support the statement that the Bujagali dam i s the least costly option. 238. The Request describes eleven alternatives to the Project that were allegedly dismissed because o ftheir costs and difficulties o f connection to the national grid. These options are: bagasse (sugar cane); small hydro (less than 10MW); micro hydro (less than 100 kilowatts); geothermal; municipal solid waste; solar; efficient lighting and transmission losses as demand reducing options; wind power and efficient stove and biogas digesters. The Requesters believe that rather than dismissing options for the difficulty o f connecting to the national grid, the analysis should have focused on reducingthe burden on the national grid andon developing independent grids. 239. Management believes that the economic, financial, safeguard, technical, governance, and other required analyses meet high professional standards and are in compliance with applicable Bank policies. It adds that these analyses take into account the findings o f the previous Bujagali Inspection Panel report and result from the overall project due diligence, which adequately takes into consideration best practice. Management is convinced that the analysis undertaken was appropriate and wide- ranging enough to identify and assess all potential alternatives for expansion o f Uganda's power sector. Management indicates that the analyses "Assessed a wide range of supply options, including alternative hydropower sources, such as geothermal power and thermalpower (e.g., oil based); small-scale renewable options (e.g., mini-hydro and biomass); oil imports; and other supply options."222 222Management Response, 733. 77 240. The following paragraphs present the analysis and the Panel findings with respect to the Requesters' claims on the Project evaluation o f alternatives. The analysis will examine first the economic evaluation o f options and will follow with the environmental and social evaluation. 1.The Context:Power Sector Developmentsandthe PowerSupplyCrisis 241. As noted in Chapter 11, there have been significant developments in the Ugandan power sector since the prior attempt to develop and implement the Bujagali Hydroelectric Project: continuing demand growth; the acquisition o f new high-cost stop-gap thermal generation; big tariff increases; part-privatization o f distribution; and increased dependency o f UETCL on Government funds. The Panel observes, however, that some conditions remain broadly unchanged, notably that only about 5 percent o f the population i s connected to an electricity supply and only about halfthe cost o f electricity units sent out from power stations i s recovered from customers. 242. The Project PAD states that while Uganda's main power source, the NalubaaleKiira dam complex, has a potential capacity o f 380 MW, over recent years production has dropped to 120MWbetweenAugust 2006 and 2007.223Inresponse, in2005 and 2006 the Government leased two 50 MW thermal plants and in2007 IDA financed an extra temporary 50 MW.224The PAD sets out an Interim Generation Expansion Plan for 2006 to early 2011 (the commissioning year o f the Bujagali Project). In this plan, about 44 MW o f mini-hydropower capacity and 15 MW o f co-generation (using bagasse) are scheduled for 2007-2009, while 150 MW o f diesel and fuel oil power generation are requireduntil2011.225 243. The consultants reviewed thermal generation requirements for 2006-201 0. The PAD states that "The total cost of thefossil-fuel components of the 2006-10 interim power plan is about US$700 million. By comparison, the expected economic cost of the proposed project is about US$520 million. [...I if commissioned in 2011, the proposed project would immediately displace about 738 GWh of fossil-thermal production (about 35% of total 2010 generation)-a substantial portion of the proposedproject's expected output, estimated at226 1,165 GWh and 1,991 GWhfor the low and high hydrology scenarios, respectively.7, 244. The table below, extracted from the PAD,227shows some key aspects o fpower sector performance for 2001-2005. It illustrates challenges relating to technical and non- technical ("commercial") losses and to the collection o f billed sales (indicating, as noted, that only about halfo f the electricity sent out from the grid was paid for during this period). 223This contrastswith a 380 M Wpeaksystem demandand a290 M W base loaddemand. 364 GWh of load were shedin2006 (PAD, Annex 1,777 & 11). 224Underthe Power SectorDevelopment Operation(PSDO). 225PAD, Annex 1,78 and783. 226PAD, Annex 9, p. 78. 227PAD, Annex 9, p. 79. 78 Table 3 Power Sector Performance (2001-05) Ti&& 9.2: Power Sector ?er I 20031 20041 2005 t generationfor domesticmarket (GWh) 1,4251 1,4261 1,5421 1,6871 1,827 hemtechnical losses (GW~) I 2871 2811 3011 3311 3541 Fdsales ( G W h ) 867 933 1035 1031 1075 dlection ratio 83% 83% 77% 82% 86% .led sales collected (GWh) 720 774 797 845 924 ,es collectedas % ofnet generation 50% 54% 52% 50% 51% 245. The PAD'S Figure 12.1 reproduced illustrates how end-use customer electricity tariffs (exclusive o f 18 percent Value Added Tax)229have risen since 2005, including a near doubling o f 2006 average tariffs to accommodate the high cost of thermal generation. The PAD states that tariffs are expected to fall once Bujagali i s commissioned "[ ...Iand the benefits of the loss reduction and efJiciency improvements [to be achieved by UMEME, the private distribution company that began work inMarch20051are realized. I n real terms, under the base case scenario, the projected weighted average electricity tariff declines from the present US$l7.2c/kwh to US$13.8c/kwh by 2011." 230 Figure 5 Weighted Average RetailTariff December2000 -November 2006 Figure 12.1: Weighted AverageRetailTariff December2000-November 2006 350 300 s 250 f 200 150 . I O SO [I - 12/00 12/01 12/02 ID03 12/04 04/05 06/06 11/06 4 .....................____- . j " U G / k W h ........ -. --__.... IJSSkW . - _.--_-...-.- ...... -.. ~ (excludingValue Added Tax) 228PAD, Annex 12, p. 105. 229PAD, Annex 12, p. 105, h1. 230PAD, p. 7. See also PAD, Annex 12, Table 12.2, p. 106. 79 2. DemandForecastsandElectricityTariffs 246. The forecasting o f the demand and its interaction with likely tariffs i s a necessary element in the process o f analyzing alternatives. Thus, the analysis o f the future "expansion path" o f an electric power system should explore both the likely evolution o f the demand on the system and the least cost means o f satisfying that demand through existing plant and new investments. This inturn carries implications for the tariffs needed to recover the costs and whether they are consistent with the forecast demand. 247. Inthe Requesters' opinion, the demand forecast analysis for the project i s unrealistic. Only a small part o f the population o f Uganda can afford electricity that is unsubsidized. Therefore, the Requesters are convinced that even ifthe whole country is covered by the national grid, the electricity generated by Bujagali will not be affordable by the population. The Requesters maintain that the high cost o f the Project will further limit the amount o f subsidies for electricity tariffs for users connected to the grid, leading to even higher tariffs and pushingmore people out o f an already limited power market.231 248. Management notes that the risks related to future uncertainties o f variables such as the level o f electricity tariffs, the "end user tariffpath and its affordability" have been evaluated. The Economic Study also projected three demand scenarios: base, low and high.232 Responseindicates that these were developed taking into account data of The the past several years and also the comments made by the InspectionPanel inits 2002 Investigation Report.233 249. OP 10.04 acknowledges that the Economic Study o f projects i s based "on uncertain future events and inexact data" and as such "inevitably involves probability judgments.'' The analysis must take into consideration the "sources, magnitude, and effects of the risks associated with the project by taking into account the possible range in the values of the basic variables and assessing the robustness of the project's outcome with respect to changes in these values." This analysis aims at identifying whether it i s possible to improve the project design, increasing the expected value o fthe project and reducingthe risk o f failure.234 231Request,pp. 8-9. 232Management Response, p. 7, reads: "[bb 2011, the base case generation requirement for the domestic market would be 2,208 GWh, with a spread around the base case of about 14percent above (high case) and 18percent below (low case). By 2015, the base case demand would be 2,959 GWh, with a spread around the base case of about 24percent above (high case) and 30percent below (low case)." 2332002 InvestigationReport, 7213, p. 62, reads, "In the Panel's view an analysis of the sensitivity o f the key findings of the due diligenceto a widening of the loadforecast ranges would have beenand could still be appropriateandvaluable, andwas neededinorder hllyto satisfythe requirementso fparagraph6 (Risk) . OfOP 10.04." 234OP 10.0476 80 250. As noted, the Inspection Panel Report on the first Bujagali project criticised some aspects o f the load forecasts used for that project. In the Panel's judgment, there i s evidence that Management addressed demand forecasting for the current Project seriously, in that it commissioned a detailed, sophisticated review in 2004, which stressed the importance of a thorough revision of the load forecasts.235One o f the criticisms o f the first Inspection Panel report related to the narrowness o f the range on the prior project's forecasts, given the uncertainties relating to several o f the key underlying variables. The forecasts for the current project show a much broader range between the highand the low cases, reflecting in particular significant variations around the base assumptions about residential connections and the rates of growth in household income and commercial and industrial GDP. It i s noted, however, that all other assumptions remain the same as for the base forecast. 251. The two figures below, both extracted from the Economic illustrate the base, high and low generation forecasts and the electricity sales forecasts and the ranges across them. Figure 6 Generation Forecasts Figure2-4: GenerationFmeaslts(including c d t t e dexports-GWhnet e#o c 4040 3mo 2040 0 2015 2WC ZW7 ZOW 2005 2010 2011 2012 2013 a t 4 2015 a 1 5 2417 2011 2013 2020 ~ 235Bujagali Economic Review (BER), March 16, 2004 (hereinafter "BER 2004"). 236Economic Study, MainText 72.11, p. 38. A further figure illustrates forecast peak demand. 81 Figure 7 Sales Forecast for Uganda Figure2-5: SalesFmastsfor Uganda(GWh) --0m 2ObS 2M7 iwI M43 lo10 2011 2012 2015 a14 2015 M1C 2017 H i 8 2813 2B20 252. Inrelation to new connections to the electricity grid, the Economic Study states that the growth rate inall connections over the period 2001 to 2005 was "high, averaging 9.9%per year over the period. The average number of residential consumers added per year over the period is 21,000."237For the base forecast, the Economic Study indicates that for 2006-201 0 inclusive they assume that new residential connections will be one fifth less than this-that i s 17,000 er year. Of these, they assume that UMEME, the privatized distribution will connect 12,000, their revised concession target, inurban and peri-urban areas, and that there will be 5,000 per year o f grid-connected rural consumers from rural electrification programmes. The Economic Study states that "Umeme is not expecting to connect more consumers than they are committed to in their concession in view of the shortages of generation and high tarij5 that are likely to be experienced until Bujagali comes into sewice." 239 In June 2008 UMEME was reported to have made the, as yet unsubstantiated, claim to have already exceeded their concession target of 60,000 new connections intheir first 237Economic Study, Main Text, 12.2, p. 26. 238UMEME is a company, originally owned by Eskom, South Africa (44 percent), who subsequently withdrew, and Globeleq, UK (66 percent), set up to manage the electricity distribution operation and maintenance concession. The 20-year concession began on 1March 2005, with an option to exit pending an assessment o f operations during its first 18 months (see, for example, http://www.eskom.co.za/annre~ortO6/directorre~l.htm; accessed 15 July 2008) 239Economic Study, MainText, 12.5, p. 28. 82 five years, by making 63,000 connections since taking over the distribution network in2005.240 253. For the base forecast, from 2011 to 2020, the Economic Study assumes that connections will rise after the ending o f "generation capacity constraints, which will trigger an increase in the rate of connections, both urban and rural. Over thisperiod it is assumed that 25,000 new residential consumers will be connected each year, including both urban and rural connections." 241 For the variants on the base forecast, the Analysis assumes that: for the `high' variant, there will be 20,000 new residential connections per year to 2010 and 30,000 per year thereafter, that is 18 percent and 20 percent respectively above the base values; while for the "low" variant there will be 12,000 new connections per year to 2010 and 17,000 per year thereafter, that i s 17 percent and 28 percent respectively below the base values. 254. For the base forecast, apart from the reference to the "ending o f generation capacity constraints," the Economic Study does not explain how the study arrived at the sudden jump from 17,000 to 25,000 new connections per year between 2010 and 2011, a 47 percent increase in a year, which i s then assumed to remain constant throughout the next decade. Such a sudden increase would surely prove extremely demanding for both management and workforce of UMEME, the distribution company. The Panel notes that although the availability of reliable electricity supply at the time the Bujagali plant i s commissioned might reasonably be expected to stimulate new connections, the Economic Study appears to assume a more sudden increase in connections than seems likely to occur. A more gradually phased trajectory of connections to the grid after 2011 would seem more plausible, both for the base forecast and the low and high variants. 255. In relation to losses, the Economic Study states that in 2006 estimated total losses were 39 percent, consisting o f 20 percent technical losses (transmission 4 percent, distribution 16 percent) and 19 percent commercial losses. The Economic Study assumes that technical losses will reduce to 16 percent for the base demand forecast, a target which it says UMEME have a "strong incentive" to exceed. The Study indicates, however, that "Forecasting attainable levels of commercial losses is more dfficult," and that if UMEME's programmes are carried through and supported by the courts with strong penalties, this could lead to a big reduction in commercial losses: "Residual levels of commercial losses of between 2% and 6% should be achievable, with a base demand forecast assumption of 5%, by 2012."242It is not explained why 5 percent was selected from within this range, rather than the central value o f 4 percent. 240Demand Overwhelms UMEME,NewVision (Kampala), June 12,2008, Posted to the web June 13,2008 (htt~://allafrica.comistories/printable/200806130045.html; accessed 15 July 2008). In this report UMEME also said that: "Demandfor new connectionsfrom consumers has more than doubledfiom 1,400 a month to 3,000 currently;" and (b) that UMEME had hired 1000 technicians in order to clear the backlog. This latter would represent about a doubling o f UMEME's 2007 workforce, as recorded by the Electricity Regulatory Authority. 241Economic Study, MainText, 72.5, p. 28. 242Economic Study, Main Text, 72.9, p. 35. 83 256. Table 2-9 o f the Economic Study,243reproduced below, sets out their assumed future levels o f technical and commercial losses for the base case forecast. They assume that, once reduced, losses will stay at their 2012 levels up to 2020.244For the load forecast sensitivity scenarios, however, it i s stated without explanation that the assumed values for technical and commercial losses are "as for base forecast." The Analysis also states that the current collection ratio (Le. the ratio of sales collected to sales billed) was 80 percent and that UMEME was committed under the concession agreement to improving the ratio to 92.5 percent by 2008. The Economic Study asserts that they based the demand forecast on achieving 90 percent by 2008 and 97.6 percent by 2011, remaining constant thereafter.245N o reason is given why it was thought appropriate not to test the sensitivity o f the "high" and "low" forecasts to potential variations in technical and commercial loss reduction (or in improvements to the collection ratio).246 Table 4 Forecasts of Technical and Commercial Losses Table 2-9: Forecastsof Techcal andCotmercial Losses 200s Lood 2007 2008 2009 2010 2011 to12 hckmicallosses 19.4% 20.0"` 19.5?.$ 18.8*,,* 18.17.; 17.4% 16.7% 16.0% Commerciallosses 21.7% 19.09G 17.0% 15.0% 12.0% 9.0% 7.0% 5.0.: . Totalloses 41.1% 39.0% 36.5% 33.8% 30.1% 26.4% 23.7% 21.0% 257. The PAD confirms that the Government and UMEME renegotiated UMEME's distribution and supply license in December 2006, "Since the lack of ower severely hindered UMEME's ability to meet its performance targets [. ..].7E7 The overall collection rate, which had risen from 80 percent on takeover to 92 percent by May 2006, fell to 82 percent inNovembedDecember 2006, after a substantial tariff rise in November 2006.248In its discussion o f Critical Risks, the PAD also confirms that to address the risk that UMEME terminates its concession, IDA and MIGA are providingcoverage for regulatory, nonpayment and breach o f contract risks, and that the concession was modified to protect UMEME's ability to meet its concession obligations.249 243Economic Study, MainText, 12.9, p. 35. 244 Economic Study, Main Text, 72.9, p. 35, also estimates that 70 percent of commercial losses will be converted into billed sales and the remaining 30% will drop out of the system. They say that, "this assumption was adopted in deriving the baseforecast and the high/low sensitivityforecasts." 245Economic Study, MainText, p.31. 246 QER 2007, p. 12, states: "Comparing the start and endpoints of the Base Frame development [...] the ratios of paid end-use to generation are only about one-haK the other half consumed in losses and uncollected bills. Thefuture evolution of this tremendous leakage and the impact on electricity use of its reduction are the most important factors determining generation requirements at least over the remainder of this decade." 247PAD, Annex 12, p. 104. 248PAD, Annex 12, p. 104. 249PAD, p. 23. 84 258. Given the difficulties inherent in reducing commercial and technical losses in the electricity system in Uganda, and in particular in light o f the challenges recently experienced by the electricity distribution company UMEME, the Panel finds that the demand forecast should have varied the assumptions on losses and the collection ratio( i.e. the ratio between UMEME'sbilled sales collected and billed sales) as part o f the sensitivity analysis and of a more complete appraisal o f risks, in conformity with OP 10.04. Indeed, somewhat lower values might also have been appropriate for the base forecast, as an alternative to assuming that the targets set for the electricity distribution concession would be fully achieved B. Economic Analysis ofAlternatives 259. In general, the Requesters claim "... that the absence of an adequate and comprehensive economic and alternative (options) assessment of the Bujagali dam Project violates the World Bank's Policies on Economic Evaluation of Investment Operations (OP 10.04), Poverty reduction (OP/BP 1.00), among others ....7,250 260. The Requesters state that there is no evidence that a comprehensive economic analysis o f the Project was carried out, because the study published on the World Bank website i s not "comprehensive" and a basis for determining the economic viability o fthe Project. They believe that the Economic Study i s inadequate becauseit i s based on the flawed assumption that the Project will be able to generate 250 MW, which, inthe Requester's opinion, will not happen. 261. The Requesters' contend that the SEA did not adequately analyze feasible alternatives to Bujagali, and that the Economic Study did not include an adequate assessment o f the economic alternatives to support the statement that the Bujagali dam i s the least costly option. They maintain that the analysis o f alternatives was geared toward proving that Bujagali i s the least-cost option and thus there was no balance in this evaluation. The Requesters also claim that the Economic Study does not provide costs, cost-benefit or opportunity cost-scenarios, or calculations for developing these alternatives and for deciding to reject them infavour o fthe Bujagali option. 262. Inits Response to the Panel, Management states that it "considers that the economic, financial [. ..] and other required analyses to date are compliant with relevant World Bank Group policies [. ..I.Moreover, the overall project due diligence adequately accounts for best practice as well as thefindings of the previous Bujagali Inspection Panel Management argues that the economic study includes an assessment o f the economic viability o f the Project and risk analysis,252and maintains that the 250Requestfor Inspection, p. 9. 251Management Response, 733. 252 crisis conditions on the sector and the need for emergency thermal power; (ii) demand forecast [...]; Management Response, Annex 1, p. 25 includes key points such as "(i) impact of the current power the the (iii) level ofelectricitytariffs; (iv) the hydrologyofLakeVictoria andits impactonhydropower the generation; (v) the supply alternatives and their costs; (vi) the environmental and social costs o f Bujagali and its main alternative; and (vii) the economic value o f electricity to consumers, the end-user tariff path andits affordability." 85 analysis also addresses the financial sustainability o f the power sector after Bujagali's commissioning. 1.BankPolicies 263. Bank Economic Evaluation policies applicable to this Project are OP/BP 10.04 on Economic Evaluation of Investment Operations and OP 1.OO on Poverty Reduction. OP 10.04, provides in paragraph 1 that "For every investment project, Bank staff conduct economic analysis to determine whether theproject creates more net benefits to the economy than other mutually exclusive options for the use of the resources in question." The Policy then sets out specific provisions in seven areas: Criterion for acceptability, alternatives, non-monetary benefits, sustainability, risks, poverty and externalities. 264. Paragraph 3 o f OP 10.04 refers to the Analysis o f Alternatives as "one of the most importantfeatures of proper project analysis throughout theproject cycle. To ensure that the project maximizes expected net present value, subject to financial, institutional, and other constraints, the Bank and the borrower explore alternative, mutually exclusive, designs." Paragraph 6 on Risk provides that sources, magnitude, and effects ofthe risks o f the Project have to be evaluated "by taking into account the possible range in the values of the basic variables and assessing the robustness of the project's outcome with respect to changes in these values" to improve project design where possible, increase the expected value and diminishthe risk o f failure. Paragraph 8 addresses Externalities and states that "A project may have domestic, cross-border or global externalities. A large proportion of such externalities are environmental. The economic evaluation of Bank-financed projects takes into account any domestic and cross-border externalities." 265. In Bank projects, various key parts of the PAD normally signal compliance with OP 10.04. In the Bujagali PAD these sections include Strategic Context and Rationale (Section I),Project Description (Section II), Implementation (Section 111) and especially the Appraisal Summary (Section IV), which includes sections relating to the economic and financial analyses carried out as part of the due diligence. Annexes contain hrther details o f the underlying analysis, particularly Annex 9: Economic Analysis, but also inothers that concern the hydrology o f Lake Victoria, the financial performance o f BEL and the financial performance o fthe Ugandapower sector. 2. Terms of Reference for the Economic Study 266. IFC appointed consultants to carry out the Economic and FinancialEvaluation Study, in this Report referred to as the "Economic Study" in January 2006 and the final report i s dated February2007. 267. The analysis i s summarised in Section IV, Part A of the PAD; it reviews "[ ...I Uganda's power sector, including the impact of the current power shortages, electricity demand growth, the hydrology of Lake Victoria, generation alternatives 86 and an assessment of the least cost power investment program for Uganda [and] the project's economic rate of return, the end-user tariff path and the macro-economic impact of the project." The PAD summary o f the findings o f the Economic Study statesthat:253 . The Project is needed now, delay in the proposed commissioning date (201I) would be expensive, and its implementation presents minimal economic risk to its status as the least-cost option for the . next major Ugandan grid system generation increment; The 250 MW (megawatt) Bujagali configuration is preferred over . 200MK . I t would be uneconomic to commission the Karuma hydropower project before Bujagali; Commissioning Bujagali in 2011 has a risk-adjusted net present value (benefits minus costs) advantage of US$I84 million, at a 10% discount rate, relative to the alternative of not implementing the . project; and The economic internal rate of return (EIM) of theproject is 22% in the Base Case and lies within a range of 11.3% to 26.4%, taking account of a broad range of assumptions about demand, costs and hydrology. 268. The Executive Summary o f the Economic Study confirms that their TOR,"[. .] call. for a comprehensive update of the previous due diligence work that was carried out in thefirst round of the Bujagali project [, .,],"254The TORremind the consultants o f changes in significant key factors since the 2000-2002 due diligence, giving particular attention to potentially sensitive issues including the demand forecast and affordability and hydrological risks. The "Partial List o f Studies" appended to the TORincludes the 2002 Management Report and Recommendation inresponse to the InspectionPanel Investigationreport o fthe prior Bujagali Project. 269. The TORalso have three other significant aspects: (1) while the discussion o f supply options includes potential hydro sites, from "mini to major," a range o f thermal alternatives, geothermal potential and bagasse, it does not draw attention to any other potential alternatives; (2) project costs are defined in a specific way "Forpurposes of economic analysis, the project cost is confined to incremental economic costs. For purposes offinancial analysis, theproject cost will be the tariff the sponsor proposes to the power purchaser(^).^'^^^ 253PAD, p. 26. 254Economic Study, Appendix A-Terms o fReference. 255EconomicStudy, Appendix A-Terms o fReference,716. 87 270. The TORstate that on acceptance o f the inception report, the final report i s expected within three months of contract award. The Executive Summary o f the Economic Study notes, however, that after the February 2006 interim report and a presentation inKampala inMarch 2006, "Work was then held upfor a number of months whilst the World Bank carried out an independent review of the analysis of the hydrology presented in the Interim Report. The demand forecast was also reviewed and amended to include updated GDP estimates and a detailed assessment of the assumptions of future levels of technical and commercial losses."256In December 2006, the consultants submitted the Draft Final report, and presented it in January 2007 to the Government and other stakeholders (unidentified) in Kampala and to the lenders inLondon. 271. The following sections of this Report draw closely on materials presented inthe PAD and in the Economic Study, and a range o f other documents, and examine the Requesters' claims and Management Response in light o f applicable Bank Policies outlined above. 3. Alternatives considered 272. The PAD states that major generation alternatives to Bujagali considered in the Economic Study include: small and medium-sized hydropower projects, large hydropower projects studied beyond the feasibility stage (i.e. Karuma), thermal options, bagasse based cogeneration and geothermal.257The economic consultants were required to consider the generating capacity profile during the "interim period" until2011,the proposed commissioningyear for Bujagali. Chapter 4 o fthe Economic Study discusses these arrangements, covering a range o f actual and potential thermal (oil-fired), biomass and small hydro projects. The biomass section (4.3) discusses the generation o f electricity from bagasse, including the forthcoming supplies from the co-generation plant at Kakira Sugar Works and from the SCOUL sugar estate. The consultants produced an interim generation expansion plan for 2006-2010 and a list o f those plants expected to remain operational from 2011, when Bujagali would be due to come into service. Chapter 7 o f the Economic Study contains a list with more detail, shown inthe table below.258 Table 5 Assumed Generation Capacity Existing in2011 256Economic Study, MainText, Executive Summary, p. 8. "I PAD, p. 27. . 258Economic Study, MainText, Table 7-7,77.3, p. 107. 88 Table 7-7 AssumedGeneration Capacity Exlstlnpm2011 Buseruka 1 Drpmds onihebydmlogycaseandtheopaarn~ruktorh'alnballe-Kun l oontheplaun Comrmssoncddou~mam 3.1 Conventional Thermal Plants (2011-2020) 273. The Economic Study explored a range o f thermal options for electricity generation, all o f which were expected to depend on imported fossil fuel. The conventional thermal options considered included diesel plants, gas turbines (open and combined cycle) and steam plants, ranging in size from 10-100 MW, all burningvarious forms o f oil. Coal-fired plants were excluded because of, "[. ..] the non-availability of coal in Uganda, the high ship ing and rail haulage costs and the higher capital cost of this type of plant [. .]."25fThe study assumed that the plants would be sited close to . Kampala and would not require grid connection costs beyond those o f the switchyard. 274. The thermal candidate plants also included a geothermal plant of 40 MW, since "[. .] . we do not believe that the geothermal potential for power generation in Uganda is suflciently wellproven at this time to rely on more than about 40 MF" (the analysis o f geothermal resource potential is discussed further below). The study undertook a preliminary screening analysis o f the thermal plant options, at oil prices rangingfrom US$68/bbl (the estimated 2006 price) to US$35/bbl (the then forecast for 7-8 years ahead), estimates which now appear very conservative. The results suggested that, "Provided the geothermal resource can be proven, this appears to be the most attractive of the thermal optionsfor base load operation."260 3.2. The Geothermal Potential 275. The Request claims that Uganda's potential for geothermal energy i s up to 450MW but that hydropower generation studies took precedence over thermal energy because BEL claimed that only 45MW o f the 450 are actually feasible. The Requesters think 259Economic Study, MainText, 75.1, p. 61. 260Economic Study, MainText, 75.1, p. 64. 89 that BEL'Sassessment is "premature and pessimistic," as experts they have consulted claim that the potential for sites i s greater than indicated in the SEA. The Management Response states that a "detailed review of conducted as part of the project analysis of alternatives." 61 The analysis concluded feothermal prospects was that only 10 percent o f the potential 450MW claimed by the Requesters i s feasible and a geothermal 40MW plant was assessed in the least-cost analysis. According to the Economic study, only one o f three potential geothermal resources in Uganda is potential for power development.2613 "promising"-at Kibiro-as it "a pears to be a medium grade resource" with 276. The Economic Study discusses the geothermal potential o f Uganda, drawing on a detailed 37-page review o f geothermal options in Appendix D. The study suggests that despite a long history o f interest and increased recent activity, "the exploration of these resources remains even today at apre-feasibility level of in~estigation."~~~ 277. The Panel notes that the Icelandic International Development Agency (ICEIDA) has stated that the Ministry o f Energy and Minerals Development, "with support from ICEIDA and the World Bank, carried out a drilling programme for temperature gradient measurement in Kibiro and Katwe geothermal prospects," which followed up a surface exploration in 2005. The programme's objective "was to confirm the existence of the geothermal resource and assist in positioning deep exploration wells." ICEIDA reports, however, that "now the research is drawing to a close with only a few outstanding gradient drilling boreholes in Katwe-Kikorongo. Unfortunately, no viable geothermal prospects haveyet been i d e n t ~ j k d . " ~ ~ ~ 278. In its review, the Economic Study concludes that, "historical estimates of the geothermal potential in Uganda being as much as 450 M W are substantially overstated." The study assesses the three main geothermal resource areas inUganda, that is, Katwe, Buranga and Kibiro, and interprets the first two "to be low grade resources with reservoir temperatures of only some 100°Cand consequently with nil potentialfor commercial scale power generation." Kibiro, however, "appears to be a medium grade geothermal resource with reservoir temperatures of about 22OoC," and hence i s "considered to be the only geothermal resource in Uganda with clear potentialfor power development."265 279. The study estimates the cost o f a full "greenfield"266 development o f a 40 MW binary cycle geothermal power plant at Kibiro, "at US$134 million which equates to a cost of US3350per K W installed," 267 which would take around 52 months from when the exploration drilling commenced. Both the estimated costs and the time duration lie within, but at the upper end of, recently cited ranges. For example, expert Management Response,p. 30. 262Economic Study, MainText, 15.2, p. 65. 263Economic Study, Appendix D,p . 2 , p. 216. 264See: http://www.iceida.is/englisWmain-activities/ugandd (accessed 15 July 2008). 26sEconomic Study, MainText, 75.2, p. 65. 266A.place where no suchplant or previous development exists. 267Economic Study, Main Text, 75.2, p. 65. 90 participants at a U S workshop in 2005 are reported to have suggested that "most projects currently under development have estimated capital costs between $3000 . years to put a geothermal power plant on line [.. and $35OO/kW," and the same review states that, "[. . it takes a minimum of 3 to 5 s Consequently, the Economic Study's estimates o f cost and project duration do not seem inappropriate (and, as noted earlier, the Economic Study screening analysis suggested that geothermal plant would be attractive for base loadrelative to other thermal candidate plants). 280. The Economic Study reaches its conclusion about the limited prospects at Katwe and Buranga partly by questioning existing estimates o f temperature for Katwe and Buranga contained in a 2005 paper, whose authors' affiliations include Uganda's Department o f Geological Survey and Mines.269 The Analysis also makes a comparison with developed geothermal systems elsewhere in East Africa, suggesting that high temperature volcanic systems are more likely to be found in the Eastern rather than the Western branch o f the East African rift and that, "overall the lack of fumaroles in any Uganda pros ects is a negative indication that any high temperature Jields are present." 70 The Analysis does not report whether their ? reviewers discussed these conclusions with the aforementioned paper's authors. Given the significance o f the difference in interpretation, the Panel considers that such discussions should have beenheld and reported. 281. The Panel notes the statement in the Management Response that additional studies and shallow drilling are included under the ongoing Uganda Fourth Power (Power IV) Project, to assist the Government in assessing geothermal prospects at several sites in Western Uganda. The additional information resulting from this work would help resolve conflicting views regarding geothermal potential in Uganda, and may have a significant bearing on the economic analysis of alternatives. 3.3. Small and Medium Scale Alternatives 282. Inthe Requesters' view, only a limited energy potential at various hydropower sites has been developed. Management responds that the Bank i s supporting development o fmini-hydro potential and states that projects providing power to the grid or suitable for grid connection were considered inthe Economic Study. 283. Section 5.6.2 in the Economic Study discusses candidate small and medium scale hydro alternatives in a short paragraph. After noting that the least-cost planning analysis includes six small plants, with capacities o f 3-1 3 MW, which were expected to be on stream before 2011, it states that the least cost planning analysis did not include any other small hydro projects because firm information and studies were not 268Factors Affecting Costs of Geothermal Power Development, Geothermal Energy Association for the U.S. Dept. of Energy (2005 ), p. 47 269G. Bahati, Pang ZhongHe, H. hannsson, E. M. Isabirye, & V. Kato, Hydrology and Reservoir ` Characteristics of Three GeothermalSystems in Western Uganda.34(5) Geothermics;568 (2005). 270Economic Study, Appendix D, 7D.6, p. 233. 91 available. The Analysis adds that "The impact on the least cost plan, and on the Bujagali project, of any other projects that may be developed over the coming years is likely to be relatively ~ma.11.'~~~` The PAD simply says that the costs and production characteristics o f other potential small hydro sites in Uganda are not sufficiently known at present for purposes o flongterm planning.272 284. The Economic Study review o f the potential o f other biomass alternatives other than Kakira and SCOUL, i s limited to the Analysis discussion o f Interim Supply Arrangements, where it i s asserted that, "There is some potential in Ugandafor the generation of electricity rom wood waste, coffee husks and rice husks, as identi led in the ESMAP study.[fn: Lf ganda: RuralElectrificationStrategy Study, UNDPiWorld Bank, ESMAP; Report2 1/99] r However, these biomass resources are considered to be too small and spread out to be economically justifiable for large-scale power generation within the timescale of this study.''273The Economic Studycites no authority or source for this last statement. Nor is there any indication of whether sources other than the 1999 ESMAP study were identified and consulted. The Economic Study or the PAD might have referred, for example, to the assessment o f the state o f information emerging from the study listed on the website of the Ministry o f Energy and Minerals, which says that in January 2005 MEMD appointed consultants to undertake "Technical Assistance for Renewable Energy Resource Information and Capacity Building 285, Thus, the Economic Study examines smaller scale hydro and biomass alternatives, apart from those likely to come on stream before 2011, in little more than four sentences (on pages 53 and 79-80 o f the Economic Study Main Text) and one reference, to the 1999 ESMAP study. The Economic Study does not raise or discuss any other renewable sources o f electricity, such as municipal solid waste, solar or wind, even simply to confirm that the resources might be unidentified, insufficient or otherwise unsuitable for appraisal inthe least cost expansion planning process. 286. A related issue concerns connectability to the grid. The Economic Study does not make clear whether connectability to the grid was viewed as a necessary qualification for including an option in the analysis. If it did, the Economic Study should have explained, first, why distributed generation had been ruled out on principle; and second, how the test o f connectability hadbeen constructed and applied. 287. The TORfor the Economic Study states that, in preparing the electricity demand forecast, the consultants should ". ..delineate the potential demographic and economic size of the market catchment area for an integrated grid of the type the Bujagali project would serve [...] sight should not be lost of those areas outside of the UMEME concession which are potential grid connectable demand centers which may be supplied within the framework of the Energy for Rural Transformation 271Economic Study, MainText, 7 5.6.2, p. 79-80. 272PAD, Annex 9, p. 82. 273Economic Study, MainText, 74.3, p. 53. 274See httv:i/www.energvandminerals.~o.ug/renenergy/index.html [accessed 15 July 20081. 92 Project The TOR also indicates, in relation to mini to major hydro alternatives, that the consultants should assess "whether there is newer and better information about costs and capacity/energy supply potential for serving the grid thanpreviously available."276 288. Thus, the TORtended to encourage a focus on relatively large grid-connected plant and did not draw attention to the evaluation of smaller scale or off-grid alternatives. Inaddition, as noted above, Management has stated that any project providing power to the grid or suitable for grid connection was considered in the economic study. However, in a country where only 5 percent o f the population i s connected and there i s widespread poverty which access to electricity could help to alleviate, it would be reasonable to expect attention also to be paid to small and/or distributed generation options (not only hydro) which might intheory more directly address local and rural poverty. 289. It is striking that the Management response contains a much fuller discussion and appraisal o f the smaller scale and/or distributed generation options than was contained inthe Economic Study and the PAD. 290. The Panel notes that the information in the Economic Study and the PAD relating to knowledge about and the potential o f smaller scale and/or distributed generation alternatives did not clearly establish that the available studies and data had been identified and evaluated in a way that would have enabled decision-makers to decide whether further consideration was required. Consequently, the Panel finds that the Economic Study and the PAD did not demonstrate full compliance with OP 10.04's requirement inparagraph 3 to evaluate alternatives. 3.4. Oil Resources 291. InJanuary 2006, an oil company announced that the had proven "the existence of a workingpetroleum system in the Albertine Basin,"2 while warning that it was too Y early to determine its size or potential commerciality. Six months after the date o f the PAD, this company claimed that "Preparations for commerciality are well advanced [...] Preparing for Power generation in 2010."278 Other reports convey more scepticism about the scale o fthe discoveries.279 292. The PAD states that, "The Government has also reported a domestic oil resource discovery in the Lake Albert region of western Uganda, which would need to be proven as economically viable; this is not expected to have any impact on power 275Economic Study, Appendix A, 18, p.6. 276Economic Study, Appendix A, 7 20, p. 9. 277See http:llwww.tullowoil.codtlwimedialnewsi200612006-0 1-171[accessed 15 July 20081. 278See httv:iiwww.tullowoil.comitlwlirireportspres~~~evortsvres/2007ivresentationslanalvstvisitidxlanalvstvisitid xlconclusion.vdf [accessed 15 July 20081. 279See, for example, "Oil Hangs on Crude Balance", posted 2007-11-06: htt~:i/www.mvu~anda.co.uelnewsl?more=l96 [accessed 15 July 20081. 93 generation before 2011."280The PAD later goes on to say that because o f the time required to go ffom proving the reserve to refinery construction, this discovery, "is not ex ected to have an impact on power generation options over the medium term," 81 although it does not define the medium term. f 293. The Economic Study does not appear to discuss the oil discovery at all. While the oil resource discovery was at a very early and unproven stage at the time when the Economic Study Final report was completed (February 2007), the Panel finds that the existence and potential of this resource should have been reviewed inthe discussionof alternative supply options. 3.5. Large Hydro-Electric Power Plants 294. The Nalubaale and Kiira power plants were, o f course, included in the modelling o f the power system. The coverage o f major new candidate hydro projects apart from Bujagali and Karuma is briefly addressedin four paragraphs in S. 5.6.1, which draw on the "review of the potential large hydroelectric projects in Uganda [. ..] made in the 2001 Acres study, based on existing studies such as the 1997Master Plan." The Economic Study states that apart fi-om Bujagali and Karuma, "[. ..] the only sites that were considered particularly attractive for the development of the Uganda power system were: Kalagala, Ayago, Murchison and Masindi." Kalagala i s not considered a candidate because o f its "offset" status, while the other three projects were "eliminated by Acres in 2001 for reasons that are still valid today:" Ayago and Murchison because o f environmental impacts in the Murchison Falls National Park and Masindi because it had "been studied only at a conceptual level, and its large size (up to 3000 MW) makes it a too large project for being considered in the Uganda power system in the next tenyears."282 295. The Requesters claim that Karuma i s less socially and environmental destructive than Bujagali and that inthe comparison with Bujagali, Karuma lost on economic grounds because the Economic Study for Bujagali "was based on greatly inflated costs for building Karuma." Management states that Karuma was the most likely alternative to Bujagali but also that the analysis showed that the latter i s the least cost option between the two as Bujagali has a lower construction cost. 296. The PAD states that apart from Nalubaale and Kiira, Bujagali and Karuma are the only large hydropower projects "that have been developed beyond the feasibility stage in Uganda." Management appears to have focussed on two possibilities for the dimensions o f the Bujagali option: the TORfor the Economic Study state that, "One of the Bujagali project design decisions that needs to be assessed here is project dimensioning-specijkally whether to provide for four or five 50MW units, the "' 280 PAD, p. 5. PAD, p. 27. 282 Economic Study, MainText, 75.6.1, p. 79. 94 economic viability of thefifth unit depending very much on demand and hydrology 297. Inrelation to Bujagali, "Theproposedproject costs are based on the terms of the bid for its EPC [Engineering-Procurement-Construction] contract and current estimates of the project development, environmental and social and financing costs. Its economic cost is estimated at US$520.6million (2006 money). On the same basis, the estimated global cost of the Karuma Hydropower project is US$587.8 million."284 The PAD summarises the capital costs o f the 250 MW Bujagali option and the 200 MW Karuma option inTable 6 below, (the figures are aggregates o f those inTables 5-4285and 5-6286of the Economic Study). Table 6 Economic Cost Estimate for Bujagali and Karuma T a b 9.5: EconomicCoatEstlmrtc for BujapUand Karnma C. The Project costs 1.BujagaliProject Costs 298. The Requesters argue that from an original estimate o f US$430 million, the Project cost has now jumped to US$735 million. NAPE states that, on February 28, 2007, it met with World Bank officials who acknowledged that the cost increase has been 30 percent. 299. Management claims that the increase since 2000 i s due to an increase in Engineering Procurement and Construction (EPC) costs o f around 65 percent caused by an increase in the cost of metals and increases in the cost o f oil and o f transporting equipment to Uganda. However, Management adds that the Bank group as well as the other lenders "have taken several steps to ensure that costs of Bujagali reflect current market conditions." 283Economic Study, Appendix A, 716, p. 8. 284PAD, Annex 9, p. 82. 285Economic Study, Main Text, 75.4.3, p. 72. 286Economic Study, Main Text, 75.5.3, p. 78. 95 300. The Economic Study states that they evaluated the economic cost o f Bujagali, taking into account the results o f the EPC contract negotiations reachedby January 31,2007, and the unit rates o f civil works and power plant equipment.287They do not comment here on the very large difference between the EPC costs o f the earlier Bujagali proposal and the current proposal. The PAD acknowledges that by the time o f its publication, estimates o f Bujagali's EPC costs o f US$51lmillion, excluding spares, were substantially higher than those for the prior Bujagali Project (US$3 15 million in 2000). The PAD offers three sentences o f explanation, with no quantitative information. 301. As noted above, the Management Response offers a somewhat fuller explanation, plus an assurance that the EPC contract price would be reviewed by the lenders. Management ascribes the increase inProject cost by approximately 65 percent to "(i) increase in the cost of metals by an estimated 90% over the last 5 years (metals accountfor about 40-60% ofpower generation equipment); (ii) increase in the cost of oil (140% between 2000 and 2006), which raises the cost of transporting equipment to Uganda over more than 1,000 km from the nearest port in Kenya; (iii) a tighter market for power generation equipment: higher global demand combined with consolidation among manufacturers has resulted in higher prices." The Response also notes that, the procurement o f the EPC contractor was conducted under the supervision o f the EIB and that, before finalization, the lenders, assisted by their Independent Engineer, will review the bid prices conducted by BEL'S Owner's Engineer, the EPC contract price and conditions.288 302. The Panel notes that power plant costs have increased in real terms internationally, although the index o f this for hydro plant may be less than for thermal plant because o f the higher proportion o f civil engineering costs in the former, the other more local factors referredto inthe PAD may more than offset this. Nevertheless, because EPC costs form a key element inthe determining the Project's economic and fmancial viability, the Economic Study and the PAD should have supplied fuller explanations of the details of this cost increase, supported by appropriate analysis and quantitative evidence. 303. In addition to the cost increase noted above, there is evidence o f significant cost increase during and after the appraisal process for the current Project. The Economic Study states that "Just after this report was completed, BEL informed PPA and the Bank Group of the most recent results of on-going negotiations with the EPC contractor [...] bringing the total EPC cost increase into a range of $30 to $35 millions, nominal and undiscounted." The Analysis argued, however, that an incentive scheme to accelerate commissioning was being negotiated, which would yield, "a real economic cost saving on thermal plant operation estimated at $30 to 287 Economic Study, Main Text, 15.4.2, p. 70. These figures are obtained from international bidding on similar works, taking into account the recent trend of tightening of the market in the hydro power sector and a substantial increase inthe price o f someconstruction material. 288 Management Response, p.35. 96 $40 million (in dollars of2006)." Consequently, the Economic Studyjudged the net impact onthe project's economic viability to be 304. InApril 2007, the PAD290(and later the Management Response291)cites an EPC price o f USS520 million (including spares; USS511 million without spares).292This figure suggests an increase over the value shown in the PAD'S Table 9.5 (sourced from the Economic Study Table 5-4), since as noted above, the PAD states that the US$521 million total inthat table i s based on the EPC contract bid terms and estimates of the project development, environmental and social and financing costs. The PAD does not specify which elements in Table 9.5 constitute the EPC cost, nor does it state clearly the price of the original October 2006 EPC tender.293If the EPC element of PAD Table 9.5 i s the sum of Direct Construction Costs (Civil Works, US$227 million; andEquipment, USS187 million) andEngineering and Coordination (USS28 million), then the EPC costs would be USS442 million (or US$441 million from the source, the Economic Study Table 5-4). The PAD'S US$521 million EPC figure is US$79 millionhigherthan this. 305. InJanuary 2008, a communication from Management to the Panel stated that the final EPC price was US$564.4 million. Thus, the expected cost of the `tfixedprice EPC contract" had risen significantly during the appraisal process between tender evaluation and the April 2007 PAD estimate of US$511 million. By December 2007 the final price was US%123 million (28 percent) above the Economic Study value and US$97 million (21 percent) above the tender value. 306. Overall, leaving aside financing costs, the EPC costs have changed significantly since October 2006, as indicated inthe table below: Table 7 Bujagali EPC Cost Evolution Source EPC Costs (US Percentage of Economic $million) Study Value Tender Price, October 2006 467 a 106% Economic Study,February 2007 441b 100% (estimated) PAD, April 2007 (estimated) 520` 118% Final EPC price, December 2007 564d 128% Notes: (a) Colenco Project Review and Assessment Report, February 2007, p. 12.2 (evaluated pr :e 289 Economic Study, MainText, fh 5, p. 70. 290 PAD, p. 26. See also PAD, Annex 9, p. 78. 291 Management Response, p. 30. 292 PAD, p. 16. 293 Colenco Project Review and Assessment Report, (February 2007) (hereinafter "Colenco 2007") records that the tender EPC price of the successful bidder was US$467.2 million, excluding spare parts (Colenco 2007, p. 12.2). Colenco notes that ,while the October 2006 tender price was US$467.2 million, the evaluated price was raised by US$10million, "to account for technical deviations." 97 307. The February 2007 Independent Engineer's Report estimated the "total project implementation budget," excluding financial costs, based on the EPC tender price, plus estimates of spares, environment costs, engineering and management costs and contingencies at company level but excluding transmission line engineering and associated environmental and social The total was US$624 million, (including US$60 million o f contingencies at company level), significantly higher than the estimate o f US$476 million (US$521 million minus US$28million for grid connection and US$17 million for transmission line environmental and social costs) inthe Economic StudyTable 5-4 (PAD Table 9.5). 308. The Economic Study cost estimates for Bujagali and Karuma both included an estimate for "Connection to the grid (line and substations)." For Bujagali, the Economic Study Table 5-4 (also PAD Table 9.5) indicates a cost o f US$28 million. The Economic Study states that these costs were based on the designs and the cost estimates proposed by the Project sponsor's consultants. The US$28 million was an underestimate, however.295 The May 2007 African Development Fund Appraisal Report for the Bujagali InterconnectionProject (BIP) produced an estimated total cost o f the project o f US$75 million (of which US$17 million were for resettlement/c~mpensation).~~~ PAD says that, ahead o f the competitive tender The and EPC contract, the "construction cost of the Interconnection Project is estimated at approximately US$55 million."297 This estimate i s almost double the US$28 million used inthe Economic Study and reproduced inAnnex 9 o fthe PAD. 309. Two considerations may be drawn. First, EPC costs have increased by US$123 million (28 percent) from the Economic Study estimate to the point where the contract price was fixed. Second, the Economic Study, which appears to be the only economic appraisal addressing the total project costs, uses the lowest numbers, for boththe hydropower and the interconnectionprojects. The PAD relies heavily on this study in confirming the judgement that this is the lowest cost option for generation and should enable retail tariffs to be reduced. The PAD adds a financial appraisal o f Bujagali Hydropower Project (BHP) (using higher costs), but omits Bujagali Interconnection Project (BIP) from this analysis altogether, on grounds that-as the Panel was recently informed-the Bujagali InterconnectionProject's wider role inthe system would make it "inappropriate to attribute the transmission line costs solely to the Bujagali project." 310. Inaddition, inconsidering tariff effects, the full recoverable costs o f the Project must be included. Inthis case, it i s not clear to what extent it i s intendedto recover the cost o f the BIP through the Bulk Supply Tariff. The loan repayment terms would 294Colenco 2007, p. 13.7. 295Siemens 2006 evaluated several options, recommending option "3aR." This option had estimated initial capital costs o f US$41 million, a present value o f capital cost for 2010-2030 o f U S 5 2 million, and NPV of total costs (including losses and 0 & M) ofUS$64M: Siemens 2006, Executive Summary, Table 3, p. 3- 2. 296AfDB, Appraisal Report for the Bujagali InterconnectionProject, 30 May2007,B 4.8.1 andTable 4.1, p. 21. Later communications suggestedthat the final figure might turn out to be less than U S 7 5 million. 297PAD, p. 17. 98 theoretically allow a relaxed attitude to this; but even if the total cost o f BIP i s omitted for the purpose o f tariff calculation, the Economic Study still appears significantly to underestimate costs as shown in the PAD'S Table 3. It thus seems likely that the Economic Study underestimated both the costs (for comparative purposes) and the tariff effects o fthe BHP/BIP project. 311. Management responded to a question from the Panel about differences between the cost estimates used for economic analysis in the Economic Study and the PAD, stating that: "Although it may have beenpossible to revise the analysis mid-stream to incorporate emerging new data, it was not practical to consider re-starting this analysis when each newhefined estimate of project costs became available, since the new estimates were such that all parties involved in the study considered that they would not [to] alter the conclusions of the study. In contrast, thefinancial analysis, which drew from the same data set as well as results of the economic analysis, was more nimble. The financial analysis therefore represents the most up-dated information at the time of appraisal and issuing the PAD."298 312. The Panel finds that, although certain parts of the analysis were carried out thoroughly, to meet all of the requirements of OP 10.04, the PAD should have included an explanation and supporting evidence of why the substantial project cost variations would not alter the conclusionsof the Economic Study. 313. Paragraph 41 o f the PAD states ". .. there is limited likelihood of EPC cost increases once the EPC contract is finalize."299 Section 5.4.4 o f the Economic Study explains that for the risk analysis o fthe Net Present Value calculations they defined two cases, "Low Bujagali capital cost" and "High Bujagali capital cost." The values for the lower and upper cost scenarios for the items in Table 9.5, above were aggregated to minus 5 percent and plus 10 percent o f the base capital cost, with each scenario assigned a 20 percent probability, with the base case at 60 percent. The Economic Study states, "It should be noted the relatively high probability assigned to the base cost estimate takes cognisance of the advanced stage of development of the Bujagali project and thefact that the EPC contract has already been tendered and is under the final stages ofnegotiati~n.,,~~~ 314. The judgments o f the PAD and the Economic Study may be optimistic, however, for the following reasons: (a) After the price i s set, contractors are adept at pleading unforeseen geology/geotechnical grounds to justify an increase. Inthis case, the winning bid price was significantly lower than the next best, but between the time inwhich the 298Communication to Panel, January 09,2008. The note also said that, "The remaining differences between the PPA and PAD totals given above resulted largely from exchange rate fluctuations, and hence EPC cost fluctuation, which occurred after the economic analysis was largely complete, but which were accounted for inthe financial analysis." 299PAD,p. 11. 300Economic Study, MainText, 75.4.4, p. 74-75. 99 contract was awarded and the formal price was fixed, there was an increase o f 28 percent. (b) Although the Project may be technically straight-fonvard by international standards, the challenge in Uganda o f pulling together international and local contractors inan integrated project program will be significant (c) Ina 2006 review o f the draft contract, attention was brought to some provisions, which appeared to relax the discipline on the contractor (on defect restitution, warranties, and scope to resist Liquidated Damages inthe event o f delay). It i s not clear to the Panel that these comments have been taken into account. (d) The PAD puts weight on the incentive on BEL to contain EPC costs. This may also be optimistic, inthat there i s scope for cost increases to be recovered via the PPA (see later comments on the PPA). 315. The Panel notes that these factors, as well as the increases noted in the Bujagali EPC costs, suggest that the confidence in the base scenario was misplaced and that the 10 percent increase in the "High Bujagali capital cost" scenario was insufficiently cautious. 2. Karuma Project Costs andComparative Costing 316. As noted above, according to the Requesters, Karuma construction costs were inflated to gear the analysis o f alternatives in favor o f Bujagali. Management states, on the other hand, that the analysis has showed that Bujagali has lower construction costs thanKaruma. The Panelhas reviewedthis question. 317. The Economic Study states that the economic construction costs o f Karuma with 200 MW capacity (shown inTable 9.5 ofthe PAD-Table 6 above) were evaluated on the basis o f the design and drawings in the March 1999 Project DefinitionReport issued by a company which has been promoting a project at Karuma since the 199Os, plus a February 2006 memo showing the main volumes o f works. The Economic Study states that the estimates were based on, "Unit rates of civil works and power plant equipment obtained from recent international bidding on similar works, consistent with the rates usedfor Bujagali cost estimate^."^" 318. Table 8 below compares the estimates o f EPC cost from the Acres 2001 study for the prior Bujagali Project302with those from the PAD'S Table 9.5 (Table 6 above) (and Economic Study Table 5-4). The columns showing the percentage changes between the Acres 2001 and the PAD/Economic Study 2007 figures suggest that Karuma's EPC cost estimates grew by a smaller percentage than those o f Bujagali. 301Economic6tudy, MainText, 15.5.2,p. 76. 302Acres International, "Economic Review ofthe BujagaliHydropower Project," 2001. 100 Table 8 Comparisonof Bujagaliand KarumaEPC Cost Estimates:Acres (2001) Study andBujagaliPAD Bujagali Karuma 1 Source/ Cost Acres PAD %change Acres PAD Table %change Category (2001) Table 9.51 (2001) 9.5/ Table Economic Table Economic 4.4 Study 4.6 Study EPC Cost (excl. 306.4 442 144% 416.5 465 112% Transmissionline) EPC Cost (incl. 335.3 460 137% 473.5 544 115% Transmissionline) TotalImplementation 364.3 521 143% 427.5 588 138% cost 319. Comparative costing: The PAD states that "For illustrative purposes, a comparative costingframework of the major projects described above is shown in Table 9.7 [...] It compares the economic cost of generation of the main long-term optionsfor grid system expansion (in 2006 real terms), indicating that the proposed project is the least cost option under both hydrological scenarios." 303 Table 9 Economic Comparison of Supply Prices Table9.7, Eeommk CoipPrisonofSopply Prices 320. The indicated supplyprices o f Bujagali under the low and highhydrologies are lower than those of Karuma (and the geothermal and diesel plants). The PAD says that these supply prices are, "relevant, but the ultimate cost of a system expansion program depends not only on individual project costs, but also on the required sequencing and energy/capacity contributionfrom each unit dispatching into the system, which varies 303PAD, Annex 9, p. 84. The value o f "Investment" inTable 9.7, US$683.4 million, i s significantly above the US$521 million "Total Implementation Cost" for Bujagali in Table 9.5 o f the PAD. The notes explain that the figure includes IDC (interest during construction), which was not in the Table 9.5 estimate. The notes offer no other explanation for the difference. 101 from year to year. This is why detailed least cost generation expansion plans for Uganda are derived to analyze ifand how Bujagali wouldfit under suchplans."304 321. The Panel observes that the updating of the EPC cost figures does not obviously disadvantage Karuma relative to Bujagali. 322. At the same time, the Panel found conflicting and incomplete reports305on cost estimates for Karuma at the time o f the prior project. Thus, the Panel could not fully assess these estimates, but notes that a recent report funded by the NBI and carried out as part o f the SSEA ranks Karuma ahead o f Bujagali in comparing costs, socio- economic and environmental considerations. D.AssessmentofLeastCost Options for ExpandingPower Generation and Related Considerations 323. OP 10.04 states that the "basic criterion for a project`s acceptability involves the discounted expectedpresent value of its benefits, net of costs. Both benefits and costs are defined as incremental compared to the situation without theproject." The policy also requires the economic alternative analysis to compare the project design with other designs but also to compare it with the alternative o fnot doing the project at all. The analysis also studies "the switching values of key variables [...I and the sensitivity of theproject's netpresent value to changes in those variables (e.g., delays in implementation, cost overruns, and other variables that can be controlled to some extent)" to improve the design, increase the expected value and reduce the risk o f failure. 324. The Economic Study devised and compared alternative generation expansion plans with and without Bujagali as a candidate plant. The Economic Study explains that, "The difference in present-worth value between the costs of these two development strategies is defined as the Net Present value (NPV)of Bujagali HPP."306``The least cost generation expansion analysis was undertaken for base, low and high demand forecasts: low and high hydrology scenarios; base, low and highfuel price projections; and base, low and high Bujagali cost estimates."307Seventy two cases were evaluated, 54 with Bujagali and 18 without, to explore the range o f risks, with 13 more for hrther sensitivity analysis. 325. The PAD states that, "The economic analysis conJirms that theproposedproject is the next major least-cost generation expansion option for Uganda. [...I In addition, the 304 PAD, Annex 9, p. 84. 305 Development Today: Nordic Outlook on Development Assistance, Business and Environment, Confidential Report Over-Prices Karuma Falls Projects, December 3,2003, No.1912003 306 Economic Study, Main Text, 17.1, p. 97. See also PAD, Annex 9 725, p.84, which states "A set o f least cost generation expansion plans was developed for the Ugandan power system beyond 2010 based on candidate plants described above. These plants are then entered as candidatesinthe WASP software, together . sequenceofplantsthat meet demandat the lowest combination o fcapitalandenergy cost [...I." with existing generation capacity, the load forecast andthe cost o f unserved energy. WASP then generates the 307PAD, Annex 9, p.85. 102 least-cost status of Bujagali was tested for 200 M W versus 250 MWproject size, delayed commissioning and the Karuma hydropower project preceding it."308The Economic Studyreports a sensitivity test inwhich Bujagali's capital cost was increased to determine how large an increase in capital cost would need to be so that the least cost expansion plan would no longer choose Bujagali as the next best option for the expansion o f the system. Based on these calculations, Karuma would become the best option ifthe capital cost o f Bujagali increased by 49 percent o fthe base cost estimate, the probability o fwhich was considered very 326. The PAD also states that the only cases where Bujagali i s not in the least-cost expansion plan "are those where low demand is combined with high hydrology; such scenarios have a combined probability of occurrence of only 6%."310The PAD suggests that "Because the low hydrology has a 79%probability of occurrence versus 21%for the high hydrology scenario, it would not be economic to delay theproposed project.9731 327. The process of testing the sensitivity of the least cost expansion plans with and without Bujagali appears to have been carried out thoroughly. The assumed increase of 10 percent for the "high Bujagali capital cost scenario" compared with the "base scenario", with an assigned probability of only 20 percent, was inappropriately low. Nevertheless, a sensitivity test suggested that the Economic Study's conclusions that Bujagali was the least-cost option were robust for an increase of almost 50 percent incapital costs. 1.Tariffs andAffordability 328. The PAD'S Annex 9 states that the Economic Study also showed that under the identified least cost system generation expansion plan, with Bujagali commissioned in 2011, "the resulting costs of meeting the demand forecast, as well as the incremental costs of transmission, distribution and losses, can be recovered at tar$s no higher than those on which the demandforecast itselfwas based." It also says that the financial analysis for the power system as a whole suggested that, when compared with the assumedtariff underlyingthe demand forecast, "[. ..] the tarifmay drop by up to 10% in real terms aJ2er the commissioning of thepropo~edproject."~'~ 329. The Economic Study compari~on,~'~suggests that from 2011 the average long term cost o f supply, 16 c k w h i s 1.2 c/KWh lower than the assumed constant tariff level o f 17.2 c k w h (a 7 percent difference). This estimate i s based on system costs that incorporate the EPC contract costs and transmission costs for each power station, includingBujagali. However, the costs usedhere bythe Economic Studyfor Bujagali 308 PAD, p. 27. 309 Economic Study, MainText, 77.4.4, p. 120. 310PAD p. 27. 311PAD, Annex 9, p. 86. 312PAD, Annex 9, p. 87. 313Economic Study, 79.4.3, p. 151. 103 were again based on their EPC estimate o f US$441 million, rather than the 18 percent greater US$520 million cited inthe PAD, as well as on their probably understatedfigure for transmission connection costs.314This suggests that the Economic Study's 16 c k w h estimate o f post-2011 average long term cost o f supply was an underestimate-and that the post-2011gapbetweenthe supplycost andthe assumedtariffwould havebeen smaller thanthe 1.2 c/Kwh cited earlier. 330. The PAD'S statement simply asserts that the Economic Study shows that the tariff may dropbyupto 10percent, without qualifjmgthe statementinlighto fthe increasesinEPC andtransmission costs after the Economic Studywas preparedandthat were recorded in the PAD (butbefore it was finalised). The issueo f electricity tariffs and affordability is o f such high importance to the people and communities. The Panelfinds that, inorder to comply with the requirements of OP 10.04, the PAD should have qualified its statementabout the projected dro intariffs to take into account the impact of EPC andtransmissioncostincreases.31Q 331. The PAD presents its own estimates o f the projected levels o f the weighted average retail tariff path, based on a different and presumablylater Economic Study, shown in Figure 12.2 (below) : Figure 8 ProjectedElectricity Tariff Path (2000-16) Figure 12.2: ProjectedElectricity Tariff Path(2006-16) _-.___ . . . . . . . . ...... --. .................-........... -.......................... h j c c t c d Tariff Path 332. The series in real terms (Le. at 2006 prices), and exclusive o f 18 percent value added tax, i s shown in the table The PAD does not compare these figures with those inthe Economic Study and comment on or explain why they differ. The Panel notes that the Project's impact on tariffs and their affordability was known to be 314Economic Study, p. 152, Table 9-6; 79.4.3, p. 153; and Table 9-8,79.4.4, p. 154. 315The issueso faffordability of electricity tariffs andpoverty reduction are also addressedinChapter VI. 316PAD, Annex 12, Table 12.2, p. 106. 104 a key concern. In this light, the Panel considers that the relationship between the estimates in the Economic Study and those from the PAD'Sfinancial analysis should have been presented more clearly and transparently inthe PAD. 2011 2012 2013 2014 2015 2016 kWhat 2006 prices 13.8 15.1 14.5 13.9 13.9 13.9 2. A Criterion for EconomicAcceptability of the Project: Internal Rate of Return Analysis 333. The Requesters argue that the "SEA does not give cost, cost benefit, opportunity cost scenarios and calculations for installation and development of these alternative energy options as basis for determining Bujagali as the least-cost ~ption.,'~''They add that risks to the economy related to hydrology issues, the drought in the region and so on, have not been adequately assessed in the decision making process to choose the best option. 334. Intheir opinion, there i s a need to assess what power options may help reducing the burden on the national grid at competitive costs and prices, and study the feasibility o f developing independent grids, which in the Requesters' view could be more beneficial for the people thanbeing connected to the existing national grid. 335. Management responds that the information regarding costs and benefits is included in the Economic Studyratherthanthe SEA. Management also believes that independent grids are part o f the electrification program o f the country and both grid and off-grid systems are supported under the Energy for Rural Transformation Program (ERT). The Response says, however, that the expansion o f the national grid network i s still the least-cost means o f connecting the customers. 336. The economic internal rate o f return (EIRR) on a project i s the rate at which the present value o f the project's series o f incremental economic benefits i s equal to the present value o f its series o f incremental economic costs. The PAD notes that "The benefits are a combination of displacement of more expensive thermal power in the early years of the project's life and `consumer willingness-to-pay ' for incremental electricity supply. The costs include constructing and operating the project and the incremental transmission and distribution works neededfor delivering the pro `ect's energy to end-users, as well as managing environmental and social impact^."^" The risks to the EIRR include hydrology, fuel prices (influencing willingness to pay for alternatives to grid power), the demand forecast and the capital cost o f the project, for which different cases hadbeen specified as part o f the least cost expansion planning. 317Request, p. 7. 318PAD, p. 29. Additionally, the PAD, Annex 9, p. 90 explains that, "The EIRR is calculatedover 2007 to 2061 inclusive, with project benefits and costs stabilized at the level reached by the year the proposed project's output is fully absorbed, which varies depending on the selected hydrology and demand forecast assumptions." 105 337. Potential scenarios were explored, both with and without an estimated "greenhouse gas" benefit o f US$25 per ton o f C02 emissions reduced through the displacement o f thermal capacity using fossil fuel. These calculations suggestedthat the EIRR "would be no less than 12.4% and no more than 25.8% in the series without greenhouse gas benefits (or no less than 12.9%or nor more than 26.4% with greenhousegas benefits). The EIRR or the Base Case is 22.0% without the C02 benefits and 22.9% with these benefit^."^ In an alternative approach to risk analysis the Economic Study specified probability distributions for the values o f project capital cost (Bujagali and incremental transmission and distribution), demand forecast, willingness to pay o f newly connected residential consumers, oil prices andhydrology andusedthe Crystal Ball programto run a Monte Carlo procedure, "to randomly select any combination of values for each variable within the specijied ranges over a series of 10,000 iterations [.. The PAD states that, "The results of the EIRR analysis are that the EIRR without any greenhouse gas credit has zero probability of being less than 11.3% or more than 26.4%."321TheERRappearedrelativelyinsensitiveto anincreaseincapitalcosts although, as noted, the changewas over arelativelynarrow range. 338. OP 10.04 does not requirea specific value for the ERR, althougha fi-equentlycitedrangefor the opportunity cost of capital is h m 10to 12 percent. The Economic Study confirms that they applieda test discountrate o f 10percentintheir studies, ``as directed by the WorldBank Group.'"22ThedistributionshowninthePAD(Figure 9.3,which differsvery slightlyfiom that showninthe Analysis) suggests a zero probability o fanEIRRless than 10percent anda very low probability o f an ERR less than 12 percent. The analysis o f the possible effects o f different capital costs with and without the Bujagali Project appears to have been carried out thoroughly, as citedearlierinthis chapter. 3. Macroeconomic Considerations inthe Analysis of Alternatives 339. The Economic Study states that through meetings held with various institutions in Uganda, "It was found that the tools available for analyzing the impact of power sector investments and production on other sectors are not well developed. Forecasts presented by international institutions and the GoUfor the Ugandan economy are based on extrapolations and simple accountingformulae. There are no models with relationships representing the responses of various sectors to changes in income and prices." The Economic Study states that because o f this limitation on modelling tools, the total impact o f the two cases they compared, "with Bujagali" (and Karuma in 2017) and "without Bujagali," could not be "quantijed with precision."323 Consequently, they mainly discuss the direct impacts rather than the full direct and indirect effects. 319PAD, Annex 9, p. 91. 320PAD, Annex 9, p. 92. 321PAD, p. 29. 322Economic Study, 17.2.5, p. 101. 323Economic Study, 710.1, p. 163. 106 340. Among other effects on components o f GDP, the Economic Study claims that the "without Bujagali" case will have 5 percent higher tariffs than the "with Bujagali" case, which has a "small" direct impact on households; and that the effects o f changes in power supply will be felt most in the manufacturing and mining sectors. This is because the agricultural sector uses little electricity, while energy intensity i s ten times greater in manufacturing than in the commercial sector. The Economic Study ends its review o f the government's financial position thus, "The Government will through its ownership of UECTL carry substantial risks related to the power sector through UECTL's payment obligations under the power purchase agreements. However, we assume that the Government will not have to subsidise electricity after 2010 in the Bujagali case and after 201I in the `withoutBujagali 'case." 341. Management examined macroeconomic effects M e r in March 2007. An independent consultantassembleda spreadsheetmodel o f the impacts, using data developed by the consultants for the Economic Study and two o f their power sector expansion scenarios, one with Bujagali from 2011 and Karuma from 2017, and the other largely thermal but including existinghydro capacity (and different from the Economic Study second case). The PAD states that the Project "is expected to have a positive macroeconomic impact. Compared to a thermal oil-based expansion plan, the hydro- based expansion plan is expected to save the country's balance of payments over US$700 million from 201I to 2020."324The independent consultant states that his sensitivity tests show that, "Even iffuel costs fell by 40%, while at the same time construction costs rose by 25%, the more capital intensive butfuel saving alternative I373 374. In its assessment o f the economic internal rate o f return to the Bujagali Project, the Economic Study provides quantitative assessments o f both costs and benefits, includingthose benefits associated with new connections, reductions inthe amount o f unserved energy demand and the displacement o f relatively expensive thermal generation. The findings o f Chapter 10 o f the Economic Study, which was peer reviewed by an independent hydrologist, suggested that the Project would deliver largely positive direct impacts on Uganda's economy, including enhanced electricity supplies, probably at lower cost than they otherwise would be, that would benefit industry, commerce and connected households, thus enhancing national economic activity. In this sense, and bearing in mind the reservations about the cost estimates of the Economic Study expressed in Chapter V and this Chapter, from a macroeconomic perspective, the analysis appears to have complied with the requirement in OP 1.00 to show that the Project is likely to contribute to "broad based growth." 375. Interms o f the affordability o f electricity generated under the Project for the people of Uganda, Management Response acknowledges that "end-user tariffs in Uganda almost doubled in 2006' and that the "increased price still does not fully cover the cost of generation, transmission and distribution, estimated at US#25/kWh,requiring government subsidiesfor the difference." Still, Management claims that "according to the Economic Study, Bujagali's commissioning in 2011 would enable the cost of power to end-users to fall to US#16/kwh in 2006 money. This would improve the affordability ofpower to end users.7J74 376. The Panel notes, however, that the $16/kWh figure provided in the Economic Study is likely to be an underestimate of the cost of electricity with the Project. As explained in Chapter V o f this Report, the Bujagali Engineering-Procurement- Construction (EPC) costs used in the Economic Study were nearly a fifth below the EPC values cited in the PAD.375Further, the transmission cost estimates used in the Economic Study were low. The Management Response does not mention these differences in cost estimates or make clear their implications for the tariff estimates of the Economic Study, on which the estimate of US$16/kWh and Management's above statement about improved affordability are based. 373OP 1.oo 7 1. 374Management Response,p.35. 375EPC costs used in Economic Study turned out to be more than one quarter lower than the December 2007 final EPC costs. 119 377. Much o f the expected direct benefit from Bujagali, however, especially in the early years, i s likely to be experienced by the better o f f urban households and particularly the industrial and to a lesser extent the commercial sectors and their stakeholders. The Economic Study estimates, for example, that in 2005 residential users consumed around one third o f total electricity sales, with the other two thirds consumed by commercial (12 percent) and industrial (55 percent) users.376 378. Existing poorer households that could afford to connect would benefit from the delivery o f a more reliable and possibly relatively cheaper service. New connections, in urban and gradually in rural areas, facilitated by UMEME's distribution investments and bybetter electricity availability, would meanthat increasingnumbers o f households would gain access. Nevertheless, the electricity would still be very costly for poorer households and too costly for many. Poor urban dwellers consume little if any electricity, while most rural households are not close to a grid connection: "electricity use by households in Uganda is stunningly low outside of Kampala. ,,377 379. The 2004 Bujagali Economic Review378noted that a 2002 Uganda Bureau o f Statistics (UBOS) survey, with population quintiles defined over household per capita consumption expenditure, showed no recorded spending on electricity by the bottom quintile o furban households. The mean spending on electricity inthe fifth quintile by households that consumed it was five times that o f those in the second quintile. The Country Economic Memorandum says that, "The distributional and policy implications of this coverage profile are huge. For instance, the electricity profile according to the income cut I... suggests] that any subsidy to consumption is rather the way to go as suficientpower becomes available. regressive, but also indicating that targeted subsidies379to new connections might be I 9 380. The Terms o f Reference (TOR)for the Economic Study discuss the calculation o f the ERR for Bujagali, outline the broad range o f benefits and costs to be included, and state that "This section will also identijj the direct impact of the project on poverty alleviation by estimating the economic impact of the project on low income households. 1'380 The Panel did not find evidence in the Economic Study or the PAD of any estimates of the economic impact of the Project on low-income households. The Panel considers that such analysis, in addition to the broader macroeconomic analysis undertaken in the Economic Study, should have been made during appraisal to provide a better understanding of whether the objectives of poverty reduction envisaged by OP 1.00 would be achieved. 376Economic Study, MainText, Table 2-5, p. 26. 377Uganda - Moving Beyond Recovery: Investment and Behavior Change for Growth, Report No. 39221- UG, World Bank, Sept 2007, V. 1, p. 25. 378Bujagali Economic Review, p. 42. Memorandum, ,Report No. 39221-UG, World Bank, October 2007, Vol. 11, Overview, p. 169. 379Uganda - Moving Beyond Recovery: Investment and Behavior Change, for Growth, Country Economic 380Economic Study, Appendix A, 126, p. 10. 120 B.Financial and GovernanceRisks 381. According to OP 10.04, Bank staff must verify whether "the legal and institutional framework either is in place or will be developed during implementation to ensure that the projectfunctions as designed" and whether "critical private and institutional stakeholders have or will have the incentives to implement the project successfully. " Assessing sustainability includes evaluating the project's financial impact on the implementinglsponsoringinstitution and estimating the direct effect on public finances ofthe project's capital outlays and recurrent costs. 1.RevenueProjectionsandthe InstitutionalFramework 382. Section B o f the PAD'S Appraisal summary addresses the financial analysis o f BEL and suggests that the Project's (i.e. BEL'S)ability to withstand downside scenarios is robust against a 30 percent increase inEPC costs that i s not fully recoverable by BEL, unrecoverable increases o f 25 percent in0 & M costs, a 50 MW shortfall incapacity at commissioning, and availability below 90 percent, as well as a project delay of up to 6 months.381 383. Section C o f the Appraisal Summary and Section 12 o f the PAD review the financial situation and prospects o f the power sector. They outline the challenges and risks it faces, relating to: tariffs (including the recent increases relatingto the costs o fthermal plant and UMEME's revenue requirements, and future increases needed to cover system investments); the past and future performance of both UETCL and UMEME, distribution losses and uncollected energy bills; revenue shortfalls and government support through subsidies and debt service deferment; and IDA support under the Power Sector Development Operation. 384. Figure 3 and Figure 12.2 in the PAD depict projected revenue requirements and the projected tariff path to 2016. These graphs indicate the scale o f the challenge, and the scale o f expected revenue shortfalls. The PAD suggests that "The projected revenue requirements and tarij5 converge by the time the proposed project comes on line in 2011. Electricity t a r ~ would befully cost reflective by then and subsidies would be s removed, exce t for duty exemptions on generation fuel and transmission investments. 'JBPIt estimates Government support to powerutilities at US$734 million for 2005-2011 and US$85 million for 2012-16. Over the period 2005-16, the government i s projected to collect net revues o f $US217 million: "The power sector will be a drain on the Treasury until the proposed project is commissioned but a net contributor after. 385. The Panel notes that this statement inthe PAD appears misleadingand seriously at odds with the projected revenue stream of the Bujagali Project, given the large shortfall until 2022 between the revenue to be raised by the tariff for Bujagali 381PAD, p. 31. 382PAD, p. 34. 383PAD, p. 36. 121 proposed inthe PAD, and the requirements o fthe capacity charge. This gap i s explicit inthe PAD figures, as is made clear below, and it is not clear from where else but the Treasury this gap will be bridged. Inthe Project's later years, the tariff revenues will exceed the capacity charge for Bujagali, which will relieve the Treasury o f this burden and enable the resources to be re-couped. The statement quoted above, however, appears to be about cash flow, which i s negative from the commencement o f Bujagali operations until at least 2022. The revenue gap that UETCL, in particular, will face, may lead to large, urgent demands on the GoU Treasury and potentially on the Bank via its Guarantee. The possibility o f both higher Project costs and significantly lower revenues will have a major bearing on whether the GoU guarantee o f capacity payments under the PPA agreement i s likely to be triggered. 386. Sensitivity tests were performed on the base case financial projections to 2016. The PAD'S Table 12.8 shows the resulting percentage tariff impacts. The tests cover five "downside risks" and three "upside potentials" scenarios. The PAD states, however, that, "Each of the sensitivities is considered in isolation, with all other assumptions in the base case remaining unchanged. "384 It would have been helpful to have applied these tests using a more comprehensive probability-based sensitivity analysis385, which would have enabled wider distributions o f the values o f each variable and their simultaneous variation to be taken into account, along with other variables such as changes inthe USh/US$ exchange rate. The likely tariff variations and the possible revenue shortfalls or surpluses and their implications for UETCL, UMEME and government net revenues are key sustainability concerns; they matter for the future of the power sector, for electricity consumers, actual and potential, and for the GoU's ability to investinkey sectors and services. 387. Paragraph 95 of the PAD gives estimates o f BEL'S annual project revenues duringthe life o f the senior loans (US$137-187 million) and o f "the estimated hydropower electricity tariff in nominal and levelized terms [.. The levelized tariffs for the low and high hydrologies respectively are 9.7 USckWh and 5.7 USckWh over the years 2011-2027 (Table 5). With expected outputs o f 1165 GWh and 1991 GWh in the two hydrologies, these tariffs imply a stream o f annual payments o f U S $ l l 3 million, which UETCL would need to recover through the Bulk Supply Tariff. UETCL will also presumably need to recover at least the construction investment costs o f the transmission line for Bujagali, which the PAD estimates at US$55 million.387 388. As noted, the PAD indicates levels o f 1165 GWh and 1991 GWh inthe low and high hydrology scenarios. Using those figures, the PAD also shows that in a high hydrology scenario, Bujagali's lifetime (30 years) capacity charges could be recovered through a levelized bulk supply tariff (2.5 percent per annum inflation 384PAD, Annex 12, p. 114 385A standardpracticeinOperations Researchknown as "Monte Carlo Analysis." 386PAD, p. 30. 387PAD, p. 17. 122 assumed, 2006 prices) o f 5.7chnit (Table 5, para. 95). The equivalent figure under low hydrology, calculated to have the same value, $113m, is 9.7chnit. Presumably this charge would beincludedinUETCL's Bulk SupplyTariff (BST), to bepassedon to customers via UMEMEand retail tariffs. The actual revenue generated for UETCL would however be less than $113m, (25 percent less, at a conservative estimate) because o f technical and commercial losses. By contrast, during the first 12 years o f operation (the period of repayment o f senior debt), the Bujagali annual capacity charge is actually estimated at an average o f $155m, with a peak o f $187m in2022.388 So the levelized tariff would leave UETCL with a substantial revenue shortfall in paying the Bujagali capacity charge. 389. It i s then arguable how a levelizedtariff will be set, given hydrological uncertainty.389 Whichever levelized tariff is set, there will be a significant revenue shortfall, to be paid by UETCL, against the required capacity charge up to 2022, o f $32m, plus compensation for losses, per annum on average, peakingat $74m plus in2022. Ifthe tariff were set at 8 . 4 ~but 2022 was actually a year o f low hydrology, the revenue gap that year would rise to $89m plus. UETCL's revenue shortfall should have been included in the PAD financial, cash flow and retail tariff forecasts. Moreover, the revenue forecasts assume collection rates rise from 54 percent in 2006 to 75 percent in 2013390.The Panel expert considers that it would have been realistic to use a lower forecast recovery rate. The possibility o f both higher costs and significantly lower revenues will have a major bearing on whether the GOU guarantee o f capacity payments underthe PPA is likely to be triggered. 390. The PAD says that, "The evacuation of maximum electricity output from the plant would require 100 km of transmission lines, the construction of a new substation at Kawanda, and the extension of the Mutundwe substation (the Interconnection Project)." It points out that it would be built as a separate project and "will be jnanced by ADB."391 In the PAD'S financial discussion and projections, it i s not obvious where, if at all, the expected costs o f the transmission project enter the projections and on what estimates they are based. Detailed consideration o f supply options inthe PAD'S Annex 9 appears to exclude or under-estimate connection costs -seeTable9.5whichrepeatsthefiguresintheEconomicStudy.AsnotedinChapter V, the actual bulk supply tariff which UETCL will pass onto the distribution sector, for inclusion in retail tariffs, should include an element for recovery of at least some o f the BIP costs, which the PAD estimates at US$55 million.392 391. This issue did not arise with the prior Bujagali project's evaluation because AESNP were investing in both the dam and the required transmission connection. In a communication to the Panel, Management has suggested that some elements o f the 388PAD, Annex 11, paragraph10. 389One answer might be to use the low/high hydrology probability estimate of 79/21: on the PAD 7 95 basis, this would give an ex-ante levelized tariffof 8.4ciunit. 390PAD, Annex 12, p. 117. 391PAD, Annex 4, p. 61. 392PAD, p. 17. 123 cost o f the new transmission arrangements might contribute to purposes beyond connecting Bujagali to the grid. Even so, in the Panel's view, to demonstrate compliance with OP 10.04, and in light of the varying estimates o f the costs o f the Interconnectionproject, the evaluation should have presented identifiable estimates o f the impacts on electricity tariffs and of the challenge facing UETCL in recovering these costs, on top o f the requirement to meet the capacity payments for the dam project. 392. The PAD states that "One of the biggest challenges currently facing Uganda's power sector are the high level o distribution losses (34.1 %) and non-collection rate (18%) as of December 2006. !l 11 3 3 Along with transmission losses o f up to 5 percent, "This means that at the end of 2006 approximately 49% of the energy sent out is not paid for. It will be crucial that loss numbers and collection rates improve again.9394 Collection rates rose from 80 percent at the start o f the concession to 92 percent by May 2006 but after two tariff increases dropped to 82 percent by December 2006. As noted inChapter V, the PAD recognises as a critical risk, the possibility that UMEME terminates its concession (inMay 2006 UMEMEwas consideringusingan exit clause that allowed it to exit after 18 months395)and lists various approaches taken to address this, including IDA and MIGA risk coverage, and says that the concession structure was modified to protect UMEME from the impact o f power shortages and reduced revenues,396"Under the restructured concession, there will be a downside protection for UMEME, and benefits accruing from lower losses will be shared between UMEMEand UETCLas long as thepower crisispersists. 393. The decline in fee collection rates suggests that UMEME's actual performance i s likely to remain potentially vulnerable to tariff increases from a variety of causes, both external and internal. There are also risks that the technical and commercial losses will not be reduced as projected in the PAD. It remains to be seen, however, whether the requirement of OP 10.04 to verify that the institutional framework i s or will be inplace to ensure that the Project functions as designed, can be met. As noted above, UMEMEfaces vulnerabilities and the restructuring might have weakened their incentives to achieve the targets for reduced losses, enhanced collection rates and new connections envisaged in the load forecast and economic evaluation inthe Economic Study. 2. InfrastructureFunds 394. The 2007 Country Economic Memorandum cited in the PAD states that, "Special or extra-budgetav infrastructure funds have increasingly been started as a means to "protect"pub1ic funds from funding specific targets. Of the five funds listed in its " 393PAD, Annex 4, p. 108. 394PAD, p. 33. 395The World Bank, Implementation Completion and Results Report (Credit #3411-UG) for a Privatization &Utility Sector ReformProject, July 31,2006, Report No: ICR-000041, p.34. 396PAD, p. 23. 3 9 ' ~ ~ , Annex 12, p. 109. 124 Table 6-1, three (the Rural Electrification Fund, the Tariff Stabilisation Fund and the Credit Support Facility) are in the electricity sector. "These special off budgetfunds are set up with sector speciJic institutions and regulations, and are partially funded with budget transfers (that rarely materialize) and own funds collected via levies, licenses and other fees administered directly by thefund without going through the budgetary annual process and controls. [...IIt is recommended to review the functioning of some of thesefunds, and make their amountspublic. Extra-budgetary funds as fashionable as they may be, bring drawbacks [...I" The cited potential drawbacks include: misuse for purposes unrelated to the original purpose, potential allocation o f excess funds outside the primary fund objective; and governance issues involvinginter-temporaltrade-offs o f staggered spending. 395. The Memorandum then says, "In general the proliferation of Extra-budgetary funds poses a serious Jiscal threat in a poor country with weak governance systems and capacity. [...I Experience in other countries has shown that extra-budgetary funds create opportunities for waste and corruption in countries with weak governance structures. Uganda is no exception: the Tariff Stabilization Fund which was designed to smooth tariffs until the Bujagali hydropower project comes on stream is already being utilized to subsidize higher tar#s from thermal power generation. This Fund is also being used tofund selective rural electrification projects, despite the existenceof a separate Rural Electrijkation Fund. Fiscal liabilities and contingencies created through extra-budgetaryfunds are not accountedfor in the Government's budget.''398 396. In light of these comments and of the scale of the revenue requirements, the financial risks accepted by UETCL and the Government, and the scale of the subsidies and guarantees involved inBujagali, the Panel notes that Management should have explored further ways of managing and addressing these financial and governance risks, inthe interests of project sustainabilityinaccordance with OP 10.04. C. The Power Purchase Agreement andAssociated Risks399 397. This section examines the Power Purchase Agreement (PPA) and associated documents, and compares it in certain aspects with the PPA for the prior project (the 1999 PPA).400 1.The Power Purchase Agreement 398. In general terms, a power purchase agreement i s a long-term contract between a generator o f electricity and a purchaser. Inthe present Project, the PPA i s a 30 year 398Quotations in this paragraph are from Uganda - Moving Beyond Recovery: Investment and Behavior Change for Growth, Report No. 39221-UG, World Bank, Sept. 2007, V.11, paras. 6.75-77, pp. 194-96. 399This section is primarily based on the analysis provided by the Panel's independent expert Mr.Graham Hadley. A summary o f his analysis is found inAnnex B of this Report. 400"Power Purchase Agreement, relating to the Bujagali Hydroelectric Project, between The Uganda Electricity Board and AES Nile Power Limited," 8 Dec. 1999 (hereinafter "1999 PPA"). 125 contractual arran ement, signed inDecember 2005 and amended and restated in2007 (the 2005 PPA)4 ',between the Project Sponsor, BEL, and the government entity in % charge o f transmission, UETCL. As indicated in Chapter 11, under the 2005 PPA, BELis to sell the contracted capacity o f250 MW exclusivelyto UETCL. 399. The terms o f the PPA are critical inunderstanding how financial and economic risks o f the Project are allocated, including who would bear the risk o f low water flow and, correspondingly, low energy output (below capacity) o fthe hydropower facility. 400. Inthe Panel's opinion, the introduction of a cost-based formula inthe 2005 PPA, instead of the maximum capacity charge specified in the 1999 PPA, i s probably the single largest adverse contractual change for the power purchaser (UETCL) and its guarantors. The new contractual basis for the Project represents a significant shift in risk away from the project investors and lenders to the power purchaser. 401. The formula and its effects can be described as follows. The formula for determination o f the monthly capacity charge i s in Annex D to the PPA. It i s very complex, since the components are defined rather than priced, and all are subject to variation. In broad terms, the components are: development costs; EPC costs; tariff debt service reserve; working capital, and fees payable by BEL. All o f these constitute Tariff Project Costs, plus equity repayment and return; debt repayment; GOU Equity (representing past development costs), and Operation and Maintenance (O&M) fees. 402. Some o f these are treated as pure pass-through (fees, and elements o f the O&M charge). Others are carefully defined as to the make-up o f their "base" cost, and in some cases - including EPC costs - increases on the base are subject to a quantified percentage "cap". The costs are subject to accountants' inspection. However, the fact remains that, leaving aside debt repayment, BEL has considerable scope to shape the base costs and insome cases the increases too, to deliver a higher capacity charge. 403. Considerable potential delay i s built in to the determination of the capacity charge (previous to which payments are on an interim basis). The charge must be set (the Final Declaration Date) within 2 months o f production o f a Final Cost Report, but that report need not be produced earlier than 6 months after the Final Draw Date, and that event (meaning the earlier o f the final draws on equity or debt) inturn may be up to 18 months after the commencement o f commercial operation. So 26 months may elapse after the start o f operations before there is a determined capacity charge. And curiously there are no specific provisions for dispute resolution. The power purchaser may be relying on BEL to be motivated to move as quickly as possible from an interim capacity charge to the finally determined charge, but equally there is plenty o f time as well as scope for BELto shape the figures. 401 "Power Purchase Agreement, relating to the Bujagali Hydroelectric Project, between the Uganda Transmission Company Limited and Bujagali Energy Limited," Dec. 2005 (hereinafter "2005 PPA" or ``PP,,'), 126 404. As was the casewith the 1999PPA, the capacity charge is not related to output, so the payment will be the same under low hydrology (when the output may be halved) as with highhydrology. Of course, hydrology i s outside BEL's control. But the payments are also relatively invariant to plant availability, which i s in BEL's control. A percentage reduction in availability (say 5 percent) would have to be sustained for a whole year before there was an equivalent reduction in the monthly capacity charge.402 405. The Panelfinds that for the Sponsor and its lenders, the terms and conditions of the 2005 PPA, especially those set forth inAnnex D, seem to represent a low-risk (though potentially disputatious) means of managing and recovering costs which are, by definition, subject to uncertainty. For UETCL, the power purchaser, and i t s guarantors, by comparison, it means that there i s no ceiling on capital costs and whether or not the Project delivers the direct economic benefits offered over 30 years, in terms of costs and tariffs which are, to a significant extent, outside their hands. 2. Risks and ConsequencesAssociated with the Project PPA 406. The increased risk borne by the power purchaser and its guarantors (the GoU and the World Bank) has significant consequences. The risks to which the Project is exposed, how the risks are shared, and possible consequences, include: 407. Capital cost escalation. Ifthe capacity charge is set higher thanpresent estimates, or rises subsequently, either tariffs must increase or additional subsidies are to be paid to UETCL. 408. Currency depreciation. For the current Project as for its predecessor, capacity payments are denominated in U S dollars. As noted in the 2002 Inspection Panel's Investigation Report on the first Bujagali project, a 10 percent per annum depreciation o f the Uganda Shilling (USh) against the U S Dollar would double the price o f the Project to Uganda in seven years. This would lead to tariff increase or additional subsidies to UETCL. 409. Prolonged low hydrology. A more pessimistic but more realistic view o f hydrology has been taken for the Project as compared with Bujagali I.Nevertheless substantial uncertainty remains. Past hydrological patterns have shown great year-on-year volatility, so that both the "high" and "low" numbers used inthe PAD are long-term averages only. The P A D illustrates how the cost o f a unit from Bujagali rises dramatically in a "low" year. A levelized tariff may be set ex-ante, but if the actual hydrologicalpattern falls below that assumed for the levelized tariff, then the capacity charge shortfall will widen and the consequenceswill be those described above. 4022005 PPA,Annex D. 127 410. Lower demand growth. It i s assumed that the demand growth rests both on continuing growth o f demand from existing customers, and a high rate o f new connections/customers, such that the number o f customers almost doubles by 2012. If this growth does not occur, UETCL's revenues would fall, with possible abovementioned consequences. To illustrate, if Bujagali were operating today, its average capacity charge during the first twelve years would pre-empt over three quarters o ftotal electricity sector revenues (customer payments) inUganda.403 411. Lower or static proportions of supply costs recovered from customers. It has been assumed that this ratio will have risen to 75 percent by 2013. Ifit were to remain at the 2006 rate (54 percent), sector revenues would be 28 percent lower. 412. Affordability. If the PAD'S economic analysis is proved correct, Bujagali's introduction will allow a reduction in (real) retail tariffs o f at least 5 percent compared with current levels. Collection rates appear not to have been significantly affected by the large (approximately 80 percent) increases in the last three years, so the Project affordability on that basis does not seem to be subject to highrisk (though new customers may reveal different price sensitivities - and produce different collection rates - compared with existing customers). However, if any o f the risks above arise, this may (in the absence o f subsidies) result in a tariff increase which would affect the affordability o f electricity. In addition, it could also reduce demand and therefore reduce rather than increase revenues. 413. Construction Delay. Despite Liquidated Damages provisions penalizing the contractor, the costs o f delay would be likely in practice to be shared via the PPA with the power purchaser. Extreme delay could require additional stop-gap generation. Otherwise, the main consequence o f delay would be to defer for customers the main benefit o f the Project, namely a reduction inpower-cuts. Overall, this maybe regarded as one o fthe lesser, or more manageable, economic risks. 414. Withdrawal of the Developer/Operator. This risk has been mitigated compared with the first Bujagali. The contractor is bound in for the construction phase, and subsequently would be replaceable as operator if not so easily as investor. The Panel notes that the Project provides for the Project to be bought out ifnecessary. 415. Poor Plant Performance.Although the PPA i s generous to the owner-operator inthe scale o f penalties for low availability, this may be regarded as low-risk. In the extreme, the provisions for Company Default provide a safety net. 3. RiskMitigation Measures 416. Physically and in its electrical impact, the present Project and its associated transmission project closely resembles the prior Bujagali project. The Project vehicle -aleveragedindependentpowerproject(IPP), includingbuildingandoperatingthe plant and selling bulk power to the public utility under a long term contract (Power 403PAD, Annex 12, p. 116. 128 Purchase Agreement - PPA), with IFIs and Government supporting both the loan finance and the PPA - i s also conceptually the same. Although there are some changes in the loan and guarantee structures, the key contract documents (the PPA and ImplementationAgreement) are also similar, even identical, inmanyrespects. 417. As described in the previous section, there have been important changes in the PPA between the prior and present Bujagali Project that have had the effect o f increasing the risk on the purchaser as compared to the project sponsor. At the same time, inthe Panel's opinion, some other changes represent potential improvements - reduction o f risk - for the present Bujagali Project as compared with prior project. Some o f the changes most relevant for Project costs and risk are: 1) the Project was awarded to the developer/sponsor by competitive process, rather than single-track; 2) the World Bank Group has important links, independent o f the Project, with one o f the equity partners; and 3) increased provision has been made for the public electricity supply system to buy-back the Project inparticular, low hydrology, circumstances. Award of the project by CompetitionThe Panel acknowledges Management's statements that competitive solicitation o f Independent Power Producer (IPP) projects is an internationalbest practice aimed at ensuring the lowest market price consistent with technical fitness to carry out a project. This procedure is a marked improvement over the prior project. In this case, however, the benefits o f competition were largely lost by post-bid negotiations, which allowed the price to rise by at least 28 percent before it was established. Further, the recent amendments to the PPA provide specific contractual scope for further upward revision. World Bank Group links with the Equity partners. The PAD notes IFC's equity contributions to one o f the Project's sponsors, Industrial Promotion Services (Kenya). The importance o f this, together with other safeguards regarding future changes in equity holding, i s that it should reduce the medium/long -term risk o f collapse precipitated by withdrawal o f the sponsors. Sithe Global i s an experienced and respected international IPP company (as was AES in2001); should they wish to withdraw at a later date, however, it might be expected that IPS(K) could temporarily take over equity leadership and engage another experienced investor/operator - or provide a transition into public ownership. There appears to have been a change o f mind-set since the prior Bujagali project - for that project the power purchaser and its guarantors took an arms-length approach, leaving it mainly to AES to overcome the planning and other local problems and propose solutions, whereas for the present Project it has been recognized at the outset that although BEL continues to take the lead, these problems will not be overcome without the involvement and long-term commitment o f the public authorities. It i s particularly important that public authorities should deal appropriately with the resettlement costs arising from local disruption at the dam and along the interconnecting transmission line. This could be an important factor ingainingpublic support, and thus reducingpolitical risk. 129 Buy-back in case of Low Hydrology. For both the prior and present Bujagali Project, the PPA and Implementation Agreements provide for buy back o f the plant by UETCL under default conditions and certainforce majeure events. In general terms, these provisions follow international norms. However, the present Bujagali PPA adds a new provision: UETCL may terminate the PPA and buy back the plant in the event o f 30 consecutive months o f "low water". The Panel notes that this i s an important safeguard because the cost o f power from Bujagali, per unit, as determined by the PPA may become prohibitively highin a sustained low hydrology scenario, and inthose circumstances it would be preferable for the public authorities to assume control, when they could stop paying the fixed capacity charge, smooth tariff effects and ensure that funds were available for alternative generation. While this provision i s to be welcomed, two specific issues may need to be addressed. First, the low water trigger may have been defined too demandingly from the power purchaser's perspective. Second, the payment terms for b u y - o ~ t , ~which provide that BEL can set the price broadly to equate to ' ~ capacity payments foregone, seem generous to BEL, given that the plant will be in real trouble if this scenario occurs. The Panel nevertheless acknowledges the need for the sponsors and their lenders to look for protection against loss. 418. While these changes represent potential reduction o f risk on the purchaser for the present Bujagali Project as compared with the prior project, the fact remains that other changes, in particular those described in section (b) and (c) above (the determination o f a capacity charge by application o f a cost formula, rather than a maximum charge), have created significant additional risk. Beyond this, the capital costs and total costs for the power plant have increased significantly inreal terms 4. Conclusions on Distribution of Risks 419. It i s clear from the review o f the Project documents that the greatest share o f economic risks lies with the power purchaser. The capacity charge may be adjusted upwards if the developer/operator hits unforeseen costs, but not downwards if demand or supply conditions deteriorate for the purchaser. The Panel notes that in fact the lenders especially but also the investors are held harmless against all or most eventualities. However, in a crisis o f non-affordability in Uganda such as might be produced by currency devaluation or very low hydrology, the investors and lenders may also be at risk, ifthe money to pay the capacity charge i s not available. Inthese circumstances, buy-out i s likelyto provide the best solution. 420. The Panel observes that the high allocation of risk to the UETCL, the power purchaser, and eventually the GoU increases the possibility that the Project may not achieve the broad objective of sustainable development and poverty reduction embodied in Bank Operational Policies and Procedures. This also increases the possibility of the Bank (IDA) Guarantee being called. The Panel i s concerned that any additional GoU resources that are spent in the financing of 4041mplementationAgreement, Annex J. 130 the development and operation of this Project may lead to decreased resources available for social and other priority development programs. 5. Disclosure of the PPA 421. The Requesters state that the PPA was not adequately disclosed. They add that a photocopy was only belatedly (January 8, 2007) released for public review at the UgandaElectricity Regulatory Authority's (ERA) Office inKampala and that viewers were required to read it only during office hours. The Requesters claim that this i s in violation ofthe Bank's policy on disclosure o f inf~rrnation.~'~ 422. The Panel notes that OP 14.25 on Guarantees provides that "Any investment project benefiting from a Bank guarantee must comply with all Bank safeguard and disclosure policies."406 The Bank Policy on Disclosure o f Information "reafJirms its recognition and endorsement on the fundamental importance of transparency and accountablility to the development and provides for the timely disclosure of a number o f documents involving lending operations. However, there is no reference to the disclosure o fthird party documents such as the PPA. 423. The InspectionPanel notes that the 2002 Inspection Panel Investigation Report stated that it "seems evident that (...)full disclosure of the PPA is vital if the intent is to place the public in a position to analyze, understand, and participate in informed discussion about viability of the Project and its impact on the economy and well- being of Ugandans. It is also evident (...) that according to IDA 's policy, there is no specijk requirement to disclose contracts to which IDA is not aparty." 424. Management indicated that in learning from this prior Panel Investigation, "the GoU committed to and implemented a stronger program of public disclosure. This project's Power Purchase and Implementation Agreements have been disclosed by the GoU."~'* Management adds that copies of the PPA were made publicly available at the ERA offices for a 30 day period starting on March 6, 2006, and again for an open-ended period, starting on January 8,2007. Management considers that the GoU public disclosure of the PPA was "a commendable and unusual step for a private sector 425. Management further states that ERA'Sdisclosure o f "commercial documents of this nature [was] a departure from standard industry practice, since such documents are fvequently considered to be sensitive and ~onfidential."~~' Concerning the method o f disclosure, Management adds that it was understandable that ERA may wish to "retain a measure of control over the circulation of the doc~ments."~~` 405Requestpp 9-10. 406OP 14.25, para 5 407The World BankPolicy onDisclosureofInformation (2002) as revised inMarch 2005, Part 11, para 3. 408Management Responsepara 24. 409Management Responsepara 29. 410Management Response, Annex 1, item 25, p. 45. 411Management Response, Annex 1, item25, p.45 131 426. Duringits visit to Kampala, the Panel team visited the ERA offices and verified that a copy o f the PPA was available to the public ina readingroom. 132 Chapter VI1 InvoluntaryResettlement A. The Request's ClaimsandManagementResponse 427. The Requesters claim that the resettlement under the Project i s not complete.412They raise multiple, interrelated involuntary resettlement issues, including loss o f livelihood, under-compensation, inability to obtain secure land titles, lack o f consultation, and request to share in Project benefits. They believe that the existing compensation and resettlement framework i s outdated and does not reflect the current economic conditions in the Project area and o f affected people. Furthermore, they claim that "There should have been a re-assessment o f social costs and benefits o f the compensation and resettlement exercise to reflect the current and future realitie~."~'~ 428. The Requesters assert that the consultations carried out in project preparation were not adequatebecause people were informed about the project but their participation in the decision-making process did not in fact occur. They believe that "project proponents confuse consultation with true participation in a decision-making process ".414 429. The Requesters also raise specific issues about the Naminya community, including the lack o f secure tenure through land titles, unfulfilled promises made regarding accessibility to potable water and water tanks, defective latrines, schools, health centers, condition o f housing, provision o f electricity, a community center, a market, road maintenance, employment, and food and income sources such as adequate plots for farming, fishponds, and more. 415 430. Management firmly believes that this Project has been well prepared in accordance with Bank policies.416 Management "shares" the Requesters' concerns about resettlement issues, noting that the withdrawal o f the prior project Sponsor left some social aspects "unfinished. The Response goes on to say that in this context the Bujagali Implementation Unit (BIU)maintained an active presence on the ground. 43 1. Management deals with the Requester's specific claims using a three part framework they state i s "designed to ensure that local populations are fairly treated and their livelihoods improved."417 They prepared two Assessments o f Past Resettlement Activities and Action Plan (APRAPs) to address legacy issues and actions needed to comply with World Bank Group resettlement policies418:one for the hydropower 412Request,p. 11. 4'3Request,p. 11. 414Request.p. 12. 415Request,pp. 15-17. 416ManagementResponse,7 51. 417ManagementResponse,750. 418ManagementResponse,Annex 1, p.38. 133 plant, the other for the Kawanda resettlement along the T - l i ~ ~These~assessments e . ~ ~ were for people who had been moved by the prior Sponsor and were inthe process o f resettlement. In addition, Management had the Sponsor prepare a full Resettlement Action Plan (RAP) for those people who had yet to be moved along the T - l i ~ ~ e . ~ ~ ' 432. Management states that "all outstanding issues" on the resettlement at the dam site will be resolved because the BEL andthe BIUhas committed to corrective activities including: completing the land titling process; providing new water supply hand pumps at 17 existing borehole locations in the surrounding communities; improvements to education facilities in the eight affected communities, and improvements to the healthfacilities at the Naminya resettlement site. 433. With regard to consultations, as part o f the SEA Management completed an updated Public Consultation and Disclosure Plans (PCDP) discussing past and planned activities, postingboth at the Project website. Management statesthat the consultation process includes continuous consultations with representatives from communities and clans.421The Response adds that, "FKhile it would be impossible to address "each of the stakeholders '" concerns, at all meetings with stake-holders, the developer has invited community representa-tives and community members to raise issues with regard to their involvement in theproject. j )422 434. Bank Policy The provisions on Involuntary Resettlement constitute an important part o f the World Bank's safeguard policies and poverty reduction mandate. To avoid displacement-induced impoverishment, the Bank policy on Involuntary Resettlement, OP/BP 4.12, sets three objectives, all o f which are applicable to the Bujagali project. Resettlement should a) be avoided where feasible, or minimized, exploring all viable alternative project designs. Where it i s not feasible to avoid resettlement, b) resettlement activities should be conceived and executed as sustainable development programs, providing sufficient investment resources to enable the persons displaced by the project to share in project benefits.423 Displaced persons should be meaningfully consulted and should have opportunities to participate in planning and implementing resettlement programs, and (c) displaced persons should be assisted in their efforts to improve their livelihoods and standards o f living or at least to restore them, in real terms, to pre-displacement levels or to levels prevailing prior to the beginningo fproject implementation, whichever i s higher.424 ~ ~~ 419 Bujagali Hydropower Project Social and Environmental Assessment Main Report, Appendix I Assessment of Past Resettlement Activities and Action Plan (APRAP) December 2006 (hereinafter, "HPP- APRAP" See also Bujagali Interconnection Project - Assessment of Past Resettlement Activities and Action Plan (Kawanda Sub-station). 5 Nov. 2006. (D059) (hereinafter "IP-APRAP") 420Bujagali InterconnectionProject - Resettlement and Community Development Action Plan. Dec. 2006 (D060). (hereafter "RCDAP") 421Management Response, Annex 1, p. 40. 422Management Response, Annex 1, p. 40. 423The AfDB adheres to almost identical standards. African Development Bank Involuntary Resettlement Policy, November 2003 (D026) ("hereinafter AfDB IRPolicy") 13.3 (a). 424OP 4.12 72. 134 435. To achieve these objectives and mitigate impoverishment risks attributable to a project, the borrower prepares a resettlement plan.425For the Project, World Bank policies and procedures required the Sponsor to identify impoverishment risk-related impacts and plan measures to mitigate them using an appropriate resettlement instrument. 436. The Panel notes that Management adopted non-standard Bank policy terminology for the Bujagali project, calling what i s normally called a resettlement action plan (RAP) on the T-Line the Resettlement and Community Development Action Plan (RCDAP).426 In their Response to the Panel, Management refers to the three frameworks (two APRAPs and the RCDAP) as Resettlement Action Plans "RAPs. ,427 Unrelated to these three documents and physical and economic displacement, the Sponsor also prepared a Community Development Action Plan.428 B. Changing Context: from the prior project to the presentBujagali Project 437. The prior Bujagali project was divided administratively into two infrastructure components, one for the hydroelectric power plant (HPP), the other for the transmission line (T-line). The prior project sponsor, AES Nile Power (AESNP), assigned a single team for both resettlement programs and the same consultant prepared two Resettlement Action Plans (RAPS) under the then applicable Bank policy on involuntary resettlement OD 4.30, one for the HPP and the other for the T- line component. 438. AESNP's withdrawal from the project in 2003 raised the issue o f who would be responsible for the physical, institutional, and fiscal integrity o f the on-going and pending involuntary resettlement activities.429 During the preparation o f the new project, Management states that continuity o f consultations with project affected populations and villagers surrounding the hydropower site and the associated Interconnection Project was maintained by staff from UETCL, through its Bujagali Implementation Unit (BIU),430but other resettlement component investment was almost suspended. 439. When the prior Bujagali project was stopped, the resettlement process at the hydropower site had either physically moved or compensated about 8,700 people (1,288 households) excluding dependents, who lost assets in some form or another.43' 425OP 4.12 7 6. The AfDB IR Policy refers to this as a "Full Resettlement Plan." Annex A lists 16 elements. 426Management Response, Annex 1, No. 18-23. See also RCDAP and HPP-APRAP. Management use of non-standard Bank involuntary resettlement terminology for the names o f their studies creates unnecessary confusion. 427Management Response,750. 428HPP-APRAP, p. 4. 429Project Files, communication dated December 21, 2004. 430PAD, p. 40. 431The HPP-APRAP (p.4) states that 8,700 individuals (1,288 households) were affected in one way or another, but discounts "dependents" who were declared as such by the household head during the socio- 135 O f these, 634 people (85 households) had to move from their domiciles.432Thirty-five o f the 85 physically displaced households resettled inNaminya, a 48.6 hectare site433 approximately 5 kilometers from the dam site, the remaining 51 relocated without resettlementassistanceusingthe cash compensation paidby AESNP. 440. In contrast, AESNP had made less progress on the T-Line involuntary resettlement which stretched along a 100 kilometer narrow corridor. In 2001, it had anticipated 5,796 people were to be displaced (1,183 households), o f whom 1522 individuals (326 households) were projected to be physically displaced from their residence. O f these, an estimated 900 individuals (184 households) would need to be resettled with the assistance o f the Company. As o f 2005, only 27 households had relocated, most o f whom took cash compensation.434 Eight households opted for resettlement packages with project-constructed new housin near Nansana about 19 kilometers from Kawanda (although closer to Kampala).43HOnthe basis o f the figures available in Project documents, the Panel's expert on involuntary resettlement matters has calculated that, through route optimization along the T-line, the new Sponsor reduced the number o f physically displaced households from 326 (in 2001) to 120 (in 2006), despite an increased number of displaced persons from 5060 to 5796 individuals along the Right ofWay.436 441. After the termination o f the prior Bujagali project, Management and the GoU restructured the ownership and financing of the T-line to be a public project. UETCL assumedresponsibilityfor the resettlement, compensation, and associated community development, excluding those who had been displaced by the previous Sponsor.437 Management states that a key reason for this change in the financing and ownership structure was the concern that additional financing for the transmission line could have an impact on the "financibility" o fthe power plant.438 1.Management's decision to assess past resettlement activities andprepare actionplans 442. This Project presented a rather unique situation where some o f the affected people were relocated or compensated as part o f a Resettlement Action Plan approved by the Bank inthe context o f a prior project with the same location, characteristics and area o f impact. What follows i s a description o fhow Management dealt with this issue and the concerns raisedby the affected people. economic survey, some o f whom may be children over 18 years, or other dependents that the Sponsor felt were "not household members in sociologic or economic terms." This deduction adjusts the displaced persons ("project-affected people" intheir terminology) down to 5,158 individuals. . 432HPP-APRAP, p. 50. In 2001, A E S estimated that 714 people (101 households) would be physically relocated. 433HPP-SEA, p. 351. 434IP-APRAP, p. 7. 435IP-APRAP, 71.4. 436Panel comparison o fRCDAP (2001) Table 6.4, pp. 6-11to RCDAP, pp. 35, 67. 437Project Files, communication dated Jan 18 and Jan 23,2008. 438Project Files, communication dated Jan 18 and Jan 23,2008. 136 1.1.Terms of Reference for the Assessments of PastResettlement Activities and Action Plans 443. Management divided the displaced peoples into two groups: those who had been displaced in 2001 and those awaiting displacement along the T-line. These groups roughly correspond to the Project's infrastructure components. The Panel notes that the 2006 Social and Environmental Assessment (SEA) applied markedly different involuntary resettlement TOR to each group. 444. Along the T-line, a full Resettlement Action Plan was prepared for those to be displaced (Resettlement and Community Development Action Plan or RCDAP).439 The TOR called for an assessment and update o f the prior 2001 RAP and provide additional new information as required to complete the RAP requirements to current standards (OP/BP 4.12).440 Elements o f the RAP mentioned include standard RAP elements provided for in the Bank policy: identifylng affected peoples and their assets, providing a fiamework for consultation affected peoples and third parties, analysis o f the legal and institutional framework, resettlement and compensation approach, impact identification based on satellite images with ground confirmation, provisions for monitoring and evaluation, grievance management, attention to vulnerable people and groups, budget and schedule. The T-Line RAP TOR also required examining the results of compensation strategy and approach and an updated socio-economic baseline, supplementing the 2001 baseline. The TOR further required a distinct socio-economic census and evaluation o f those to be physically or economically displaced (center-line survey), consistent with Bank procedures.441 445. At the hydropower site and at Kawanda on the T-Line, Management did not require a full Resettlement Action Plan for those who were inthe process of resettlement. The TOR for an Assessment o f Previous Resettlement Activities and Action Plan (APRAP) stated that based on "preliminaryJield observations and consultations with local leadership in project-affected villages and the Bujagali Implementation Unit" it appeared that the prior project Sponsor had "largely completed compensation and resettlement work before its departure.'*42 As part o f the SEA, BEL was asked `Vo verijj this general observation by preparing a detailed monitoring of the status of those compensation and resettlement activities with commitments made in the earlier Resettlement and Community Development Action Plan (RCDAP)." Should this monitoring identify outstanding issues, a corrective plan was to be prepared in consultation with potential involved stakeholders for subsequent i m p l e m e n t a t i ~ n . ~ ~ ~ 439Bujagali Interconnection Project, Uganda Social and Environmental Assessment: Terms o f Reference. June 2006, (hereafter, "IP-TOR'), pp. 11-12. 440IP-TOR, p. 11. 441IP-TOR, pp. 11-12. 442Bujagali Hydropower Project, Uganda Social and Environmental Assessment Terms o f Reference, June ... it appears that thepreviousproject sponsor largely completed.compensation and resettlement work ..." 2006 (DO81 (hereafter, "HPP-TOR'), p. 9. The TOR read: " 443HPP-TOR, p. 10, p. 17. 137 446. The APRAP Terms o f Reference also required BEL to conduct a "socio-economic survey of the project-affected area at the hydropower site to characterize the socio- economic conditions and livelihoods of thepeople living in the eight project-affected which include many people who were not being displaced, though they are affected by the Project. This should be done by supplementing the 1999/2000 the socio-economic baseline survey. In addition, BEL was asked to "undertake a socio-economic and livelihood survey to monitor the current status of the previous Sponsor's resettlement activities'' and check the status o f public services in the Project area. 445 This sample survey was to be used to "assist" inthe establishment o f "the socio-economic baseline" for the affected communities and "check the status of livelihood restoration and related commitments made in the 2001 RCDAP." 1.2. The Assessment and Action Plan: compliance with Bank policy on Involuntary Resettlement 447. From a policy perspective the Panel notes that this Project involves rather unusual circumstances: an ongoing, incomplete resettlement program which was developed under a previous Bank-financed operation and was based on a policy no-longer applicable, OD 4.30, which had the same overall objectives o f the policy now applicable to the Project: OP/BP 4.12.446Both the old and new policy call for a Resettlement Action Plan (RAP) consistent with the policy objectives and in compliance with specific policy and procedural requirements. In this Project, Management chose instead to develop and build on an Assessment o f Past Resettlement Activities and Action Plan (APRAP) rather than to develop a new RAP, with the justification that affected people had already been relocated and others had already received compensation under the prior project. An "Assessment of Past Resettlement Activities and Action Plan" i s not a resettlement instrument referenced inBank policy. However, regardless o f the terminology, the Panel considers that the overriding issue i s whether the TOR and subsequent Action Plan meet the objectives and requirements o fthe Bank policy on Involuntary Resettlement. 448. Inthe Panel's view to achieve compliance with the Bank policy the APRAP should have included the elements o f a RAP as defined in the policy and used by Management in the T-Line part o f the Project. The hydropower and Kawanda APRAP TOR and its implementation did not incorporate the policy objectives and specific requirements and did not take into account shortcomings in the design and execution o f the previous RAP, and evolving social and economic situations and circumstances. 449. The Panel could not find an adequate "socio-economic survey of theproject-affected area at the hydropower site to characterize the socio-economic conditions and 444HPP-TOR, p. 9. 445HPP-TOR, 7 2.3.1, p. 9. 446OPBP 4.12 replaced OD 4.30, Znvoluntuly Resettlement; these OP and BP apply to all projects for whicha Project Concept Review took place onor after January 1,2002. 138 livelihoods of thepeople living in the eightproject-affected communities" as required by the TOR. 450. Moreover, Management opted to complete an assessment and action plan based on selective fulfillment o f commitments made under an outdated RAP that had been shown by the previous Inspection Panel to have a deficient baseline, rendering inconclusive any findings on livelihood restoration. Situations not adequately considered before or that arose inthe interimperiod were not appropriately dealt with because o f the lack o f an adequate baseline assessment. This does not comply with OP 4.12. This led to action plans that did not meet the policy objectives and requirements. 451. The Panel found no formal monitoring or evaluation supporting the assertion that the involuntary resettlement was "largely completed", the reason stated for forgoing a full RAP preparation, as required by OP 4.12. The Panel finds that the hydropower APRAP failed to assess and update the previous 2001 RAP and provide additional new information as required to complete the RAP requirements to currentstandards (OP/BP4.12).447This does notcomplywith OP/BP4.12. 452. Substantive instances o f non-compliance o fthe APRAP include, inter alia: The failure to assess and update the previous 2001 RAP disenfranchised people who slipped through the flawed sampling;448 any stakeholders not previously identified in 2001,449 including vulnerable Information gathered on the displaced persons' livelihoods and standards o f living as required by OP 4.12 T[ 6(a) was limited to a sample survey, making it very difficult to determine whether the resettlement i s achieving its objectives; The shortcomings inthe original resettlement plan were carried forward. Livelihood restoration was mainly limited to the people identified in2001 and the terms and conditions set forth in2001 453. The Panel notes that resettlement i s a process, not a threshold defined by the moving o f people or acquisition o f land, and the degree o f progress o f previous resettlement efforts does not exempt the Project from meeting the requirements o f a RAP as envisioned inthe Bank policy. 454. The way an Assessment and Action Plan was substituted for a full RAP on the hydropower and Kawanda segments had far ranging consequences. Following the TORS, BEL prepared an assessment o f the progress in the execution o f the Bank- approved old RAP, and recommended recovery activities where it observed gaps. The 447IP-TOR, p. 11. 448The AfDB makes explicit references to a situation such as the present Bujagali Project, in which a Category 1 ESIA has been completed prior to Bank involvement in the project, OPs may request the Borrower to carry out additional public consultations and to prepare a disclosure plan, as deemed necessary (AfDBD02175.8). 449HPP-TOR, pp. 13-14. 139 studies did not include an evaluation o fthe impact o fthe delay on the socio-economic conditions of the Project or an assessment of whether or not the previous Sponsor complied with either the former or current Bank's resettlement policy objectives. Consequently, the new Sponsor resettlement responsibility to the people who were in the process o f being resettled was circumscribed to certain outstanding commitments that the new Sponsor wished to recognize. The critical policy requirement to census all displaced persons as o f the project baseline was neglected - a decision undermining much of the policy objectives.450The public consultation process, an integral part o f a RAP, was truncated, predefining the consultations to on-going issues, rather than including all aspects o fthe Project. Picture 7 Panel meeting with people to be resettled under the Interconnection Project 2. Baseline socio-economic data 455. In the Requesters' opinion, the existing compensation and resettlement frameworks do not reflect the current economic situations o f the people and include out o f date information. They believe that the Project should have provided for "a re-assessment of social costs and benefits of the compensation and resettlement exercise to reflect the current and future realities."45' Management claims that the APRAP took into account new conditions; for example it includes actions to address vulnerable 4s00P4.12 714 and OP 4.12 Annex A, 7 6(a). Rather than a full census, Management directed the Sponsor to assess a sample of displaced persons at the HPP in 2006. They divided the PAPSinto three groups: resettlers at the Naminya resettlement site, non-resettled physically displaced persons, and non-physically displacedpersons. The assessment encountered a limitation. Management required a 100%survey of those resettled inNaminya. The study team identified only 24 of the 34 households (71%). Management set a 50 per cent sample o f the non-resettled, physically displaced persons: only 18 o f the 51 could be found (35%). . Of the remaining non-physically displace people, Management set and achieved a 5 percent sample (60 of 1203 households). (APRAP,p. IO). 451Request, p. 11. 140 people's needs. In Management's view, BEL'S social evaluations are in "jiull compliance with WorldB a n k p ~ l i c i e s . " ~ ~ ~ 456. OP 4.12 requires gathering o f baseline information, including a census survey o f current occupants o f the affected area, standard characteristics o f the displaced households (including production system, labor, and household organization), baseline information on livelihoods and standards o453 f living, the magnitude o f expected loss, and information on vulnerable groups. Operationally, a broader survey on the occupants o f the affected area i s accompanied by a detailed survey of people to be displaced - meaning those who will be physically and/or economically displace. Inaddition, the RAP, should also include "provisions to update information on the displaced people's livelihoods and standards of living at regular intervals so that the latest information is available at the time of their displacement." 457. A full RAP also sets disclosure and consultation requirements for projects involving involuntary resettlement, 454 including the requirement that displaced persons and their communities are to be provided timely and relevant information, consulted on resettlement options, and offered opportunities to participate in planning, implementing, and monitoring resettlement.455Appropriate and accessible grievance mechanisms are to be established for these groups.456Measures are to be in place to ensure that vulnerable groups, such as the landless and are adequately represented.457 Data collection on the socio-economics o f displaced households also offers an avenue for displaced persons to communicate their concerns to Management, and as such i s parto fthe overall consultation strategy.458 458. Apart from consultation with the displaced persons themselves, in preparing the resettlement action plan, Management i s to ensure that the borrower (or Sponsor in this case) draws on appropriate social, technical and legal expertise and on relevant community-based organizations and NGOs and informs potentially displaced persons at an early stage about the resettlement aspects o f the project and takes their views into account in project design.459World Bank policy also directs Management to discuss the institutional, legal and consultative arrangements for resettlement with the agencies responsible for implementing the resettlement program. 459. The Sponsor was directed to conduct household interviews with a sample o f affected people to assist in establishing the socio-economic baseline o f affected people. According to the Assessment and Action Plan a sampling o f affected people, based on a 2000-2001 database developed by AESNP and later maintained by the BIU, was interviewed to carry out the survey. Project Affected People (PAPS) were divided into 452Management Response,p. 39. 453OP 4.12, Annex A 516(a)(v). 454OP 4.01 fn 19:and OP 4.12 72(b). See also ADB IR Policy. 455OP 4.12 713(a) and OP 4.12, Annex A, 115a. 456OP 4.12 113(a). 457OP 4.12 Annex A, 715(d). 458OP 4.12 Annex A, 516. 459OP 4.12 719. 141 three groups: resettlers at the Naminya resettlement site, non-resettled physically displaced persons, and non-physically displaced but only a sample o f displaced persons at the HPP in2001 was assessed in2006.460 Picture 8 Panel team meeting with people resettled at Naminya 460. The APRAP indicates that the survey encountered a limitation. Management required a 100percent survey o f those resettled inNaminya but the study team identified only 24 o fthe 34 households (71 percent). Management also set a 50 percent sample o f the non-resettled, physically displaced persons: only 18 o f the 51 could be found (35 percent). Of the remaining non-physically displaced persons, Management set and achieved a 5 percent sample (60 o f 1203 households). Nonetheless, comparable problems o f finding the displaced persons appeared: seven o f the eight resettled at Nansana were consulted, and none o f the remaining 19 households who opted for cash compensation could be located. 461. Following AESNP's withdrawal, monitoring o f affected people was limited by lack o f available resources allocated for this purpose.461According to Management, the database o f affected people, established in 2000/2001 by AESNP was maintained by a unit o f the Uganda Electricity Transmission Company Ltd (UETCL)'s, the BIU. The APRAP notes that the BIUhas tended to focus on the Naminya resettlement site, while 'ffor lack of resources, the BIU has been at pains monitoring non resettled affected people, particularly those who have moved out of the area, or those who were not permanent residents of the area, such as the numerous "licensees "(sharecroppers). "462 460HPP-APRAP, 72.4, p. 10. 461Project Files, communication datedJuly 21,2004. 462HPP-@RAP ,p. 11. 142 462. The APRAP acknowledges that "the whereabouts o f many people, who received compensation in 2001, are unknown" and notes that a concern arose for those who were "significantly affected and were considered as Displaced People but did not o t for AESNP's resettlement assistance, and chose rather to relocate themselves." 4L The APRAP also recognizes the need to reestablish monitoring o f the non-resettled population.464 463. The Panel notes that the survey conducted by BEL cannot be considered a census of economic or social conditions as defined in OP 4.12.465The profiling o f affected people was based not on actual field work but rather on socio-economic surveys that had been undertaken by the previous project Sponsor and on more recent surveys, which were conducted only for select groups to audit past resettlement activities and other affected villages.466Infact, as noted above, the TOR approved by Management directed the new Sponsor to assess only a sample o f displaced persons.467 464. The Panel also finds that the approach to consultations with people who had moved and had been compensated is not consistent with the involuntary resettlement policy. The consultation strategy was structurally flawed because it excluded the majority o f displaced persons and limitingthe scope o f consultations to previous commitments. 465. The Panel notes that significant weaknesses in the process o f gathering socio- economic data, an activity central to the preparation o f a RAP, were also identified in the 2002 Panel's Investigation Report. In that Investigation Report, prepared following the submission o f a Request for Inspection related to the prior Bujagali project, the Panel found that While the importance of baseline socioeconomic survey is noted in the RAP as part of theplanning process, very little of it is evident in the EL4 in way that would be useful in establishing actual planning baselines. Socioeconomic data were collected as part of the land valuation process on a transaction-directed basis. There is no evidence of the utilization of a free-standing survey of affected households including, most importantly, those who were to bephysically displaced.468 Based on the foregoing, the Management's claim that the Project took the first Panel's report findings into account in the preparation of the current Project i s not accurate because significant weaknesses inthe process of gathering baseline 463HPP-APRAP, 12.6.1, p. 11. 464HPP-APRAP, p. 31, 465HPP-APRAP, p. 10. 466Bujagali Hydropower Project Social and Environmental Assessment Main Report, Appendix (hereafter "HPP-PCDP"), Table 3.2. 467HPP-APRAP p. 10. 468InspectionPanel Report 2001 (D273), p. 77-78. 143 data information were similarly identified in the 2002 Panel Investigation Report. C. LivelihoodRestoration 466. The Requesters question whether livelihood restoration i s occurring among the displaced persons. This touches the principal objective o f the involuntary resettlement policy. Displaced persons should be assisted in their efforts "to improve their livelihoods and standards of living or at least to restore them, in real terms, to pre- displacement levels or to levels prevailing prior to the beginning of project implementation, whichever is higher".469In the policy, the objective i s structured as reaching a threshold from a baseline, not a sequence o f activities: "not merely prevailingprior to the beginning ofproject implementation, whichever is higher7, 470 restored, but.., improved" "at least to restore...topre-displacement levels or to levels 467. Cash compensation is not a policy ~ b j e c t i v e . ~Cash compensation alone is ~ ' insufficient to restore livelihoods. Leading social research has established that cash compensation fails to perform the restorative function that economics and development policies ascribe to it; the number o fresettlers who, after compensation i s paid, remain worse off and do not recover are the majority inmany Policy preference i s given to land-based resettlement strategies, particularly among agricultural populations.473When compensation i s appropriate, polic requires it be made at "jiull replacement cost' for loss o f lands and other assets. 47JCompensation should be made prior to their actual move or before taking o f land and related assets or commencement o fproject activities, whichever occurs first.475 468. In its investigation, the Panel learned that livelihoods o f affected people have been disrupted for some seven years, stemming back to the beginning o f relocation and resettlement actions under the prior Bujagali dam project. Duringthis period, many o f the people that were originally displaced were essentially left in limbo, and did not receive key elements o f the resettlement process to which they were entitled under Bank policy. Also, as a consequence o f the project's "hiatus," certain o f AESNP's commitments to regulators and the communities under its resettlement and community development planswere not fulfilled. 469OP 4.12 12(c) 470The annexes to the policies define procedures to guide Management and the Sponsors to achieve these objectives. (OP 4.12 Annex 1, BP 4.12, and Afkican Development Bank Involuntary Resettlement Policy, November 2003. 14.1) 471OP 4.12 111and 12. 472Cernea, M. (2001). Development Economics, Sociology, and Displacement: A Vexing Dilemma under Interdisciplinary Dialogue. Draft Paper prepared for the Workshop: "Moving Targets: Displacement, Impoverishmentand Development Processes" Cornel1University, November 9-10, 2000. 473OP 4.12 711. 474OP 4.12 7 6(a)(iii). 475AfDB IRPolicy13.3(e). 144 469. The Panel observes that the effects on the people o f the original displacement, and o f the ensuing delay, have not been fully reflected in the APRAP. Specific issues relating to livelihood restoration, including Project's impact on fisheries and agriculture, compensation, vulnerable people, are reviewed inmore detail below. 1.Methodto assess livelihoodrestoration 470. The Panel notes that no adequate socio-economic study was carried out. The APRAP methodology identified livelihood issues through interviews and opinion surveys with displaced people. Displaced persons and host communities were asked six to nine open-ended questions in focus groups, as were key informants from the health, education and political sectors. For example, focus groups were asked "how do you compare your current livelihood (including cash and subsistence) with what it was before compensation and resettlement? Do you think it was equal, better or worse.7 4 7 6 471. The hydropower APRAP included interviews with 24 households resettled at Naminya and with 18 households physically displaced persons who were compensated, but not resettled by the Project (equally divided by men and women on both banks o f the Nile). Both groups reported mixed opinion as to whether their livelihood restoration had occurred. Some people stated that they were better off than before displacement, though this assessment seemed to include all aspects o f their life, not only livelihood restoration. Other people claim to be worse o f fthan before, in general because o f loss o f fishing opportunities, loss o f h i t trees and loss of agricultural land and smaller size replacement land.477 These interviews also qualitatively indicated that key livelihood risks, known to appear in many other involuntary displacements, have materialized in the Bujagali project. Those interviewed told o f failed businesses, new costs incurred to procure potable water, local price inflation preventing full re7 lacement costs o f land, loss o f sustainable incomes, gender inequality, and more.4 r 472. The APRAP for Kawanda concludes that five years after resettlement, livelihood is not restored for three out o f the seven o f the interviewed resettlers. At the settlement o f Nansana, for example, the opinion survey reports again mixed opinions as to whether or not livelihood restoration has occurred. It found that "livelihoods are not restored, and some households need to be supported in their efforts to restore them: these are not houses living in a household economy anymore (iythey ever were), and they need to be supported in non-farming activities. ,9479 473. The Panel observes that the APRAPs' conclusion i s unreliable. Livelihood restoration economics encompasses many dimensions that cannot be evaluated using an opinion survey due to inter-respondent variation in interpretation o f such a general question. 476HPP-APRAP, p. 92. 477HPP-APRAP, p.17, 21. 478HPP-APRAP,pp. 77-93. 479IP-APRAP, p.12. 145 Duringits visit to the Project area and meetings with affected people, the Panel team observed that those questioned had difficulty focusing their responses to a question that simultaneously asked for opinions on changes to their lives on at least two issues over a six year period.480The methodology used to assess livelihood restoration did not compare the 2006 livelihood status of the resettlers to their previous conditions. Nor did it set a new 2006 baseline for future actions. This methodology was ambiguous as to what was and was not being measured and, as a result, it produced only a list o f unfulfilled promises left over by the prior project. 481 In the Panel's view, the methodology used to assess livelihood restoration in the context of this Project, while suggestive of issues, cannot substitute for an economic analysis of the livelihood risks and restoration!82 2. Real or perceivedunfulfilledpromises made inthe prior project 474. At the hydropower site, the APRAP survey found that the people believe that a number of promises made by the previous Sponsor were left unfulfilled, including employment, electricity - including transmission lines -, landing sites, good potable water, technical schools, secure land titles, health centers, primary school in Naminya, a market place in Kikubamutwe, durable houses, fish ponds, road repair, five years o f support, and monitoring.483At Nansana, some resettlers felt there was a promise o f a school, a health facility, improved roads, a 30 percent disturbance allowance, and secure titles.484 475. Management claims that BEL and the BIU"are now resolving all outstanding issues" and have committed to address the issues left unfulfilled by the previous sponsor.485 The Assessment also claims that its purpose is "to assess whether AESNP's commitments to comply with thepublicly released RCDAP were met. Where gaps are observed, recovery activities are recommended."486On the other hand, the Public Consultation and Disclosure Plan (PDCP) states that it i s "committed to resolve certain of these past resettlement issues in the immediate future and prior to construction 476. The Panel notes a lack o f method for deciding what promises were or were not made, which would or would not be honored and the timefiame for completing the 480HPP-APRAP, 7 4.3.2, p. 11. HPP-APRAP, p.21 74.3.4 481 482 Cernea, Michael M. ed. The Economics o f Involuntary Resettlement: Questions and Challenges. Washington, D.C.: The World Bank, 1999. See also, Cernea, Michael M. and Scott Guggenhaim 1994. Resettlement and Development. The Bankwide Task Force Review o f Project involving Involuntary Resettlement 1986-1993 (with contributions from task-force members: W. van Wicklin 111, D.Aronson, A. Salam, L.Soefiestad, D.Tewari, T. Solo) Washington, DC, the World Bank. Environment Department. 483HPP-APRAP, pp. 21,63,64,65,68,71,76,79,83,90,91,95,98, 100, 105,106. 484IP-APRAF', p. 16. 485Management Response, 7 30. 486HPP-APRAP, p.4. 487Bujagali InterconnectionProject - Public Consultation and Disclosure Plan Dec. 2006 [hereinafter "IP- PCDP"] p. 48, to the T-line and Nansana. The same statement i s made with reference to unfilfilled promises inthe HPP-SEA, p. 310 and inHPP-PCDP, p. 47. 146 resettlement activities,488 while the Bank's safeguard policies require that the resettlement plan define clearly these activities and provide a schedule for their implementati~n.~~~The fact that the same promises were mentioned on different occasions, at different sites andby different people who are identified with the Project adds credibility to specific promises. Affected people may misunderstand what they are entitled to. The Panel notes that lack of clear communication with affected people to address the concerns of the displaced persons with regards to the commitments made by AESNP, risks leaving the Bujagali project with contentious, unresolved issues. 3. Specific LivelihoodRisks 477. The APRAP and consultations identified livelihood issues where displaced persons stated that they were "worse off' than before. Key among these were the loss of fishingopportunities and the loss of agricultural land and other sources o f livelihood, including concerns about the loss o f h i t trees grown for income and personal consumption/nutrition.490The APRAP concludes that there were incomplete or insufficient livelihood restoration activities, leading to potential hardship on certain categories o f affected people.49' The Panel examined the most significant livelihood impoverishment risks. 3.1. Fishing 478. The Terms o f Reference for the SEA instructed BEL to "assessfishingpractices and livelihoods" inrelation to fisheries inthe Nile river. BEL was to do this by reviewing earlier surveys conducted for the prior project and "assess any significant changes." As a result, BELwould "propose any interventions that may be needed in response to the anticipated effects of the hydropower develo ment, by means of socio-economic surveys offisheries in theproject-affected area. 9A 479. The 2006 Assessment found that it i s "quite likely that the SigniJicance offishing has in fact been underestimated when planning resettlement and compensation, particularly for physically displaced peoples on the East Bank."493 The 2001 RAP had estimated that only 10 ercent o f the displaced persons were fishing, mostly on the East bank o f the Nile!4 No compensation or assistance was made following 488See 7 500-503 o f this report that discussesthe issue o f electricity being provided to the displaced. 489See also infra "Overall Conclusions on Livelihood Restoration." 490APRAP,p.17,I 3.4.3. 491APRAP,p.31,76.1. 492HPP-TOR, p. 7,72.2.2. 493HPP-APRAF', p.17,I 3.4.3. 494HPP-APRAP, p. 17,73.4.3. The National Fisheries ResourcesResearch Institute (NaFRRI) completed a study showing its significance a year before the 2001 RAP (NaFRRI Aquatic and fishers survey o f the Upper Victoria Nile: A report prepared for AESNP, Second Quarter 5-14 April 2000. p. 104.) Male resettlers in Naminya consistently reported that the loss o f fishing opportunities had been their most important loss (HPP-APRAF', p. 33,7 6.3.2). An April 2000 survey o f a fishing transect near the dam site discovered a small industry o f 50 canoes, 89 fishermen, 22 traders, 6 venders, a net repairman, and a fish cleaner. Displaced persons expressed a loss o f fish in their diet, including among children (Bujagali 147 resettlement for this loss o f livelihood, now stretching into its seventh year. To the contrary, the displaced complained that the Project had further limited livelihood opportunities by restricting their access to the river and may have not even paid for fishponds that were taken.495 480. Among other problems, fishermen were settled much farther from the fishing areas, lacked transport to get there, and have had their access even to these areas restricted by fencing connected with Project activities. There is a strong beliefthat promises to restore their livelihoods were not kept, and feelings o f great frustration. 481. The 2006 Assessment considered a fishery development program for the resettlers to be o f "critical importance" in livelihood restoration.496Despite this categorization, planning for livelihood restoration in fishing was limited to a two page "pplan". The Panel also notes that the 2006 Action Plan repeats almost verbatim the so-called 2001 plan.497The documents set laudable general goals, such as training that will address preparation o f fishermen for change in the river characteristics following impoundment and earmarking training for the displaced persons, within an overall regional project.498However, this planning i s not associated with any studies on the economics and nutritional importance o f fishing, particularly on the East bank o f the Nile, despite the TOR'Srequirement. Moreover, no additional support was allocated to what was called an underestimated, critical activity: the 2006 budget remains at the 2001 level (US$ 182,000- budget).499 3.2. Agriculture 482. The Panel observes that the approach taken to restore damaged agricultural livelihoods follows a pattern similar to that for fishing. N o baseline census o f the displaced persons and a socio-economic analysis was carried out, allowing only a general overview o f the pre-displacement livelihood economics. Based on the regional descriptions in the 2001 RAP and consultation discussions, it appears that the displaced persons worked small plots o f land, as peasant farmers and supplemented their income through cash crops (coffee, sugar cane, vanilla) and other income generating activities (e.g. fishing, trade, bicycle taxi driving, et^.).^" Subsistence crops includedbananas, cassava, sweet potatoes, maize, beans, millet and Hydropower Project Social and Environmental Assessment Main Report, Appendix C Fisheries Report, December 2006, p. iv]. Data comparing 2000 and 2006 economic activity show a decrease infood venders, net repairers and an increase in fishermen and fish traders fiom outside the area in the six years (HPP- APRAP, p. 60). 495HPP-APRAP,pp. 72, 74,83, 88, 92 and 96. General opinion questions during consultations such are not substitutes for socio-economic analysis. When asked o f people inthe host village, the question assumes the interviewee is an expert on the livelihoods before and after resettlement. 496HPP-APRAP, p. 33. 497RCDAP 2001, pp.136-139. See also CDAP, pp. 25-26. 498APRAP update of 15 October 2007. (D233), p. 8. 499The budget appears to have been reduced US$ lOOK as a result o f lOOK for NaFRRI monitoring being movedto another line inthe 2006 budget. 500HPP-PCDP, p. 9. 148 yams. Fruit trees - jackfruit, avocado, man 0, oranges, and pawpaws - assured a source o fnatural sugars before displacement.61 483. Bank policy requires that when replacement land i s offered, the resettlers are provided with land for which a combination o f productive potential, locational advantages (accessibility), and other factors i s at least equivalent to the advantages o f the land taken.502 484. The livelihood restoration strategy focused on the physical size o f replacement land rather than its quality or location. For those physically displaced to Naminya, each household was allocated a minimum o f an acre residential plot, where the house is located, with additional surface compensated in kind if it was part o f the same residential plot in the original location. Any additional agricultural surface was compensated in cash. The result o f this policy was that some resettlers were net "winners" - to use the Project's terminology, and others losers - if the compensation was not used to replace lost land.503 485, The Panel has found that insufficient information was available to permit the new Sponsor to assess whether or not landlessness increased or decreased under this strategy.504Along the T-line, at Nansana, there was a reduction o f 40 percent in agricultural land, with five o f the seven households having less land after resettlement.505 486, Soils are a critical factor in agricultural productivity. The Panel notes that land fertility was not considered in livelihood restoration planning or execution; however it surfaced as a major concern o f the displaced persons during the 2006 Assessment cons~ltations.~~~ At Naminya, displaced persons report that they cultivated cash crops (coffee, vanilla) and fruit trees at their former locations, the availability o f which diminished intheir new surroundings. 487. Management concedes that soil fertility i s an issue "$or some" based on the Assessment's subjective observations o f where banana plantains are growing at Naminya, stating that "not all plots are adequatefor plantain bananas., .with some obviously too dry and with a too thin layer of arable soilfor this particular crop. '1507 The Panel notes that these seem to be subjective opinions, which may not substitute for comparative agronomic studies o f the former and current sites. The Panel notes that the resettlement site i s a former sugar plantation, a monoculture crop that In the field, the Panel viewed a few photographs of pre-displacement plots that support this generic description. OP 4.12 Annex 1, T[ 11. The Sponsor proposes to consolidate rather disburse the residential areas inthe new resettlements, with the commute to agricultural lands being by bicycles. At Naminya, the lack of project provide bicycles to gain access to livelihood activities was a complaint. HPP-APRAP, p. 16. 504HPP-AF'RAP, p. 17. IP-APRAP, p. 10 HPP-APRAP, p.17. HPP-APRAP, p.17. 149 depletes soil fertility.508The Panel i s concerned that plans are underway to move T- line displaced families to Naminya without evaluationo fthis issue. 488. The negative impacts o f the productive and locational disadvantages are evident in multiple complaints about the loss o f h i t trees. As one mother stated during a consultation, the h i t trees were particularly "appreciated" by mothers o f young children as a substitute for purchased, refined sugar.509On the East bank o f the Nile, resettled people reported that their replacement agricultural land was too far away from their residences and less fertile, effectively reducing their i n ~ o m elo . ~ 489. The Assessment and Action Plan recognizes the "critical importance of traditional subsistence agriculture as a safety net for the affected people."511Nonetheless, mitigation actions are not aligned with an ago-ecological or economic analysis coupled to the livelihood risks. The 2001 RAP lacked any livelihood restoration plan or budget for agricultural activities apart from replacement o f or compensation for land. The Panel finds that the 2006 action plan attempts to mitigate the situation, but its provisions will most likely be insufficient to meet Bank policy requirements. The 2-page CDAP refers a number o f agricultural development options for affected people: organization o f producers, increased agricultural extension and animal husbandry services, and farming as a business.512However, there i s no assessment o f the damages to be addressed by these measures nor an economic study justifying the amount allocated in the Plan for these purposes (US$ 200,000 for "enhancing livelihood restoration plan" to intensify agriculture and high value-added crops in 8 communities over 5 years. Furthermore, this succinct plan, which does not focus on displaced persons, provides no implementation details. The Panel notes it is a list, not a plan.513This erfunctory treatment o f the livelihood restoration problempersists along the T-Line.P14 3.3. Conclusions on fishing and agriculture 490. The Panel notes that Management failed to ensure that the Project would institute or assure financing to mitigate these losses, exposing the displaced to on-going impoverishment risks that are now approaching eight years. Once the peoples were displaced, Management failed to recognize these livelihood risks in multiple supervision The Panel finds that the Project failed to provide 'OsSee Alfi-ed Hartemink, ISRIC, ICSUWorld Data Center for Soils, POB 353,6700 AJ Wageningen, The Netherlands. fax +3 1 317471 700 e-mail Hartemink@isric.nl 509HPP-APRAP, p 17. 'I0HPP-APRAP, p 20. 'I1HPP-APRAP, 7 6.3.1. 'I2HPP-PCDP, p. 119. '13HPP-APRAP, p 39. 'I4On the T-Line, the RAP also proposes a five year "agricultural enhancement package" and a US$600 one time subsidy per household but fails to provide a budget. The T-line agricultural package is part o f an undifferentiated, US$ 305K budget line for "livelihood restoration package" that includes an agricultural package, training, andbusiness support (RCDAP, p.98.) '15Project Files, Communication dated May 2. The Jan 15-27 Supervision mission reports its key findings are that "the project is in compliance with the Bank's social safeguards policies." 150 adequately for loss of livelihood associated with the loss of fishing and agriculture, innon compliance with OP 4.12. 4. Compensation 491. The Panel notes that the agro-economics o f livelihood restoration i s weak, particularly with reference to compensation. According to the prior project compensation method coffee, the main cash crop, was compensated at its annual yield times 1.5 to 3 to cover the so-called "establishment period" - meaning the time it takes to reestablish the perennial However, coffee takes four to five years, as opposed to 1.5 to 3 years, to restore production, assuming that one has comparable land to plant. Some calculation must be made for the loss o f the income stream until production i s reestablished, including the labor costs o f reestablishing the asset to its previous production. The Uganda rates do not compensate farmers for their labor to bring a perennial crop back into production. Underestimates o f the establishment periods for coffee and other crops including vanilla and cocoa made it economically unfeasible for the displaced to reestablish their lost incomes. 492. The witness NGO was reporting unresolved issues in the Mukono district in 2007 with regard to compensation for crops that was viewed to be unfair and not reflective o f the realistic values for the crops when compared with rates provided by neighboring districts for the same Issues also arose over differential formula being used to pay for crops o f less than four months o f age.518 493. The APRAP also points to an additional reason that the compensation method may not have achieved the objective o f compensating the displaced persons at full replacement value. On both banks o f the Nile, local land prices may have doubled after compensation, undercutting the valuation's estimates for replacement value, reducing the chances that those who received cash compensation were able to replace their landswith lands o f equivalent value.519 494. The Panel concurs with the APRAP's findings, which validate the claims of the project affected peoples (PAPS)that full replacement value compensation may have not taken place inthe prior project. 520 5. Landtitles 495. Most o f the displaced lacked security o f land titles before displacement, but they may have had established, informal security with usufruct rights recognized by others.521 'I6RCDAP, Annex 1-6. 'I7AESNP, p. 6. 'I8HPP-APRAP, p 22. 519The Assessment points to a case where an acre of land was compensated at between UGX 0.8M and UGX 1.2M, but it was not uncommon to be charged UGX 2.2M for a similar piece o f land (HPP-AF'RAF',p 20). 520HPP-APRAP. D. 20. I I 521AESNP, p. 6, No. 7, 8. 151 Bank missions reported in February 2005 that only 22 o f the 69 titles pending in the HPP component had been arranged and all eight households at Nansana were still awaiting titles.522 496. According to the APRAP, while many people who were interviewed stated that they received land titles, it also appeared that some PAPS did not receive the titles. However, the APRAP goes on to say that "this cannot be stated with certainty as in some cases, the person who has received the title was not around during the interview" and that this situation would have to be checked.523On another section the APRAP states that "several affectedpeople met by the study team claimed that land titles for replacement land provided by AESNP to non-resettlers were not all issued, particularly on the East Bank."524The underlying reasons for these disputes appear to be acquisition by AESNP o f replacement land that was under unresolved conflicts. The Assessment and Action plan then provides that "the situation must be checked (when the monitoring unit mentioned above is operational), and potential gaps must b e j ~ e d . " ~ ~ ~ 497. During its visit to the Project area, the Panel team witnessed Project-generated insecurity among displaced persons inNaminya as a consequence of resurveying and proposed readjusting o f the boundaries within the settlement. The resurveying appears to be the consequence o f the original survey layout failing to leave a leeway for the power lines passing through the settlement. As a result, the parcel layouts o f displaced persons are being adjusted accordingly, creating new, Project-generated conflicts. The Panel expects that this situation will be dealt with during the implementation o f the APRAP. 498. The Panel finds that the APRAP conclusion related to the necessity of issuing landtitles to people resettled under the prior project is consistent with OP 4.12. The Panel notes however that there seems to be no agreed timetable for the issuance of these titles. 6. Vulnerable Peoples 499. The APRAP determined that there was no proper identification o f vulnerable people up until 2007, including not providing clear criteria for vulnerability and not identifylng assistance actions. The displaced persons included a "sizable number of orphans, widows, and peoples with disabilities."526The Assessment and Action Plan states that they were not properly recorded andjudged that it i s "virtually impossible to identfi, locate and monitor vulnerablepeople."527 ~~ ~~ 522Project Files, communication dated February 7-9,2005. 523HPP-APRAP, p 14. 524HPP-APRAP, p. 21. 525HPP-APRAP, p. 36. 526HPP-APRAP, p. 24. 527HPP-APRAP, p. 24. 152 500. The Panel notes that a group o f vulnerable people, the landless tenants and sharecroppers, may have been left out from receiving compensation as a result o f the strategy noted before.528Ineligible for replacement land, they were compensated only for their lost crops. AESNP announced compensation rates, including prices for young seedlings (known as "1-4" for their months o f age). Anticipating compensation, some tenants and shareholders purchased and planted seedling. The Sponsor, believing they were observing fraudulent attempts to maximize compensation through the planting o f young seedlings, reneged on their compensation commitment and did not pay for the "1-4" crops. AESNP requested and got the support o f the GoU on their non-compensation decision.529This left landless peoples worse off, with new debt, no crops, and no harvest. 501. Heated disputes arose, some o f which are still in court, re resenting half the current court docket on the resettlement issues o f the Project!3o They were frequently mentioned in consultation^.^^^ From the perspective o f a sharecropper or a tenant position, this represents a substantial loss o f income - an issue that after five years i s still fresh on people's minds, surfacing repeatedly during the consultation^.^^^ The APRAP notes that "the situation of tenants and sharecroppers (who were compensated only for crops as the did not own land) appears to be worse in this respect than that of landowner^)^'.'^ The Panel notes that the APRAP approved by the Bank recommended not paying the claims. In light o f Management's failure to pay particular attention to the needs o f vulnerable people, this on-going dispute could constitute a reputational risk for the Bank and the new Sponsor. 502. The APRAP states that "there is no proper identiJication of vulnerablepeople at the moment and it needs therefore to be done (or redone)'' to correct the lack o f attention to vulnerability and proposes a posteriori actions.534The Assessment sets forth a US$105,000 planto use community assistancemeasures to identify the disadvantaged project-affected peoples and assist them with counseling, food support, health monitoring or medical attention ifrequired, with specific attention to orphan heads o f households and other affected orphans - an additional US$20,000.535 503. The Panel notes that the absence of a focus on livelihood risks to the vulnerable i s evident in that none of the proposed assistance measures addresses the vulnerable tenantshharecroppers or children.536 Additionally, the proposed 528HPP-APRAP, p. 20. 529HPP-APRAP, p. 20. 530HPP-APRAP, p. 22. 531HPP-APRAP, pp. 12,20,22, 86,91, 96, 105. 532HPP-APRAP, pp. 12, 20, 22, 86, 91, 96, 105. 533HPP-APRAP, p. 20. 534Management Response,p. 39. See also, HPP-APRAP, p. 32. 535HPP-APRAP, p. 32. 536Evidence o f the inattention to children was brought to the Panel's attention in discussions with the displaced along the T-line. Panel interviews near the Mutundwe substation discovered people were supportive andprepared to move, but concerned that the displacement might occur after school enrollment, making it difficult if not impossible for displaced children to enroll or transfer between government schools. The demographics may range from several hundredto several thousand children and represents a 153 assistance measures do not address the question of sustainability beyond the limited Project support. The Panel finds the Bujagali Project is out of compliance with the vulnerable peoples provisions of OP 4.12. 7. Housing, Public Services, and Electricity 504. Housing The APRAP states that the houses that were built met with the design criteria that was set out inthe 2001 RAP and were therefore generally compliant with the commitments made. It statesthat the resettlers felt that the houses were better than the ones they had, but still complained about deficiencies inthe buildings. 505. During its field visit, the Panel verified that the standard of living of the displaced households who resettled in Naminya and Nansana has greatly improved in the area of On the other hand, the Assessment discovered some shortcomings in housing condition and the Panel observed physical problems and deterioration with some o f the houses and structures. The Panel is concerned that no physical action i s planned with regard to houses at the resettlement site (apart repairing the taps from the rain water harvesting system). 506. Public services: water, roads, schools, health facilities. Restoration o f livelihoods and the standards o f living includes assisting displaced persons in their efforts to improve or at least restore, in real terms, public services they had prior to displacement.538 Water: The APRAP states that AESNP built a well, and improved a spring catchment. Due to some complaints from resettlers, BEL agreed to improve the water drawing system. This i s part o f the Community Development Action Plan. Education: The 2001 RAP included a commitment by AESNP to refurbish an existing school in Naminya. Their pulling out o f the Project caused significant concern to the locals. Subject to consultation with the local authorities, BELproposed to follow through with AESNP's commitments. Health: There i s a health centre at the site, but resettlers complain that there i s poor onsite accommodation for staff which jeopardizes the operations o f the clinic. While BEL and the authorities have discussed the situation, BEL cannot make a commitment to assist because it i s not the substantial loss o f human capital which, according to mothers, may be irreparable for teenagers if the disruption derails their studies. Options such as paying for full enrollment and transportation costs o f private schools or adjusting the time o f the move had not been considered. Enrollment in school is one o f the 8 indicators for outcome evaluation, meaning that this problem may negatively skew the overall project evaluation. School fees account for 23 percent o f the affected household's spending, underscoring the si nificance the displaced place on education. 538HPP-APRAP,p. 13. 538OP 4.12 72(b). 154 owner o f the houses that were built by AESNP. BEL will though be part o f the negotiations between the local authorities and the Lands Commission to get housing for the health staff. 507. Electricity. A highvoltage line crosses Naminya. Throughout the process to conduct the Assessment, numerous displaced persons, those who took cash compensation, and local leaders stated that they believed AESNP made a commitment to provide electricity to Naminya and other communities.539The APRAP states that "it does not seem" that such a commitment was planned under the 2001 RAP, other than providing power to equip the trading centers o f the four Western bank affected villages with transformers and low tension lines.540 On this point Management Response states that "BEL together with UMEME is exploring possibilities for the provision of electricity. BEL will also Jinance a feasibility study for electrical distribution to the resettlement community, which may convince UMEME toprovide a supply."541 The budget commitments are limited to this study. 508. The RCDAP o f 2001 makes limited commitments to power. It states that local communities have constantly requested power supply during consultation and that "[tlhere is clearly an expectationfrom villagers that AESNP, as a power producer, could easily supply electricity for free. " However, the document goes on to say that "this expectation is legally impossible, for AESNP is a power generator, not a distributor. Neither is this desirable as it will not be sustainable in the long run when AESNP is no longer in charge of the facility operation. The RAP adds that mid- " voltage lines were constructed along the main roads with 500m distributions spurs taken off in the vicinity o f trading centers at each village. However, while AESNP was to support the costs for developing distribution infrastructure systems to make electricity accessible to the eight villages, AENSP did not intend to "payfor the cost ofany individual connection or any electricity bill. Its support was only to help the " communities with initial capital costs and in turn communities or individuals would bear connections and consumption costs. 542 509. The issue i s highly controversial and a significant livelihood development issue. Duringthe Panel's visit to Naminya, a woman handed the Panel a weathered copy o f TheBujagali Power Project newsletter o f 2001, Volume 1, Issue 3, page 7 that inher opinion supported the promise o f electricity. The text states "AES Nile Power is committed to provide step-down transformers in eight villages in the affected area and in the new resettlement land allowingfor access topower by residents who have never had the opportunity." (emphasis added). The Panel has found evidence that 539HPP-APRAF', pp. 63,76,82,83, 85. 540HPP-APRAP, p. 15. 541ADB Management Responseto Request for Compliance Review of the Uganda: BujagaliHydropower Project (Private Sector) and Bujagali InterconnectionProject (Public Sector), June 2007 (D029), p. 43 for statement to AfDB Board and Doc 075, page 133 statementto WBG Board. 542 The affected communities are on the West Bank in Mukono District: Buloba, Naminya, Malindi, Kikubamutwe (RCDAP 2001, p. 132). On the East Bank in Jinja District: Bujagali. Ivunamba, Kyabinva, Namizi.(RCDAP 2001, p.22). And in two meetings in Naminya with village leaders and the inhabitants anticipated support for power (RCDAP 2001, page 63). 155 displaced persons were told that "you have a right to electricity, as do all Ugandans". Given the context and previous expectations, this broad statement may have reasonably been interpreted as a promise to deliver electricity connections to affected households. The Panel notes that this i s an outstanding controversy of highimportanceto the affected communities. 8. Investment resources for livelihood restoration 510. As o f the close o f the prior project in 2003, the resettlement costs had slightly exceeded budget allocations (US$11.5 million spent for US$11.1 million allocated).543 511. The Assessment and Action Plan budgets US$497,000 for completion o f the legacy resettlement and income restoration issues o f which US$ 320,000 i s to be used for income restoration activities and an additional US$ 125,000 to assist vulnerable people.544The Assessment allocates US$ 40,000 for resettlement corrective actions at N a n ~ a n a . ~ ~ ~ 512. The Panel's review o f the limited scope o f the livelihood restoration programs indicates that they may be under-budgeted. Management has allocated roughly the same investment resources to the HPP (households at Naminya) and the T-line (with an estimated 160 households to be physically and economically displaced). The HPP budget does not include the restoration o f livelihood costs o f the displaced people who opted for cash compensation, apart from the Naminya restoration costs once a feasible plan has been development. As information on livelihood conditions o f the displaced persons, including those who were economically or physically displaced but took cash compensation, has yet to be determined, the costs of livelihood recovery are unreliable. As livelihood restoration instruments develop, Management is expected to monitor the resettlement budget to provide sufficient resources as per OP 4.12. 9. Costs of Project Delay on Displaced Persons 513. The Panel observes that livelihood restoration has been disrupted by the Project for six years. The effects o f the delay to the displaced have yet to be hlly reflected inthe APRAP. Management conservatively estimated the overall economic costs o f delayed development during 2006-2010 to be at least US$700 million.546They noted that the time lag before entry o f the new Sponsor has tested the patience o f local populations 543Project Files, communication dated January 18 and 23, 2008. The T-line RAP budget allocates US$ 16.94 million including a 15% contingency to its RAP. The funds are budgeted for cost o f resettlement and housing (US$ 2,932,000 o f which US$ 1,804,000 i s for land acquisition). Cash compensation is estimated to use US$9,087,750 while livelihood restoration (including agriculture and business support) is US$305,000. The remaining US$ 2,148,000 i s for RAP implementation (staffing, specialist consultants, legal advice, witness NGO and logistics.). 544PAD,p. 42,145. 545IP-SEA Executive Summary, p. ES-53. . 546Management Response, 125. 156 who planned their investments based on commitments made under the prior Disclosure and consultations created expectations in the Project affected area and among those soon to be displaced.548Some examples: Physically displaced persons at the hydropower plant site opted for compensation to make investments for businesses to service the construction, only to incur a loss when it was delayed;549 Fishermen claimed they lost access to the Nile when the project area was secured without the promisedaccesspoints; and Some displaced people claim they were told by the rior Sponsor not to improve or use their land after the original valuation.58 514. Comparable stories echo throughout the Project area, as investment and life decisions were affected by the uncertainties directly resulting from the delay. The Panel heard and witnessed videotaped evidence that the uncertainties were so great that displaced persons were demanding project construction and the remaining resettlement begin immediately. Relative to the overall project losses, these issues may seem minor, however they appear to represent substantial losses to the affected-persons' limited capita1. 515. The Panel observes that, as a consequence o f the project hiatus, certain o f AESNP's commitments to regulators and the communities under its resettlement and community development plans were not fulfilled.55' In February 2005, Bank resettlement specialists asked for an audit to be carried out to pay attention to the productive outcome o f the resettlement operation and the economic and social status o fthe vulnerable households.552Inrecognition o f these issues, BELhas undertaken to document the situation, and in selected instances, began immediate action programs to respond to stakeholder concerns.553Management did not state its methodology as to how these "selected instances" for actions were prioritized and the documents do not provide evidence that this prioritization was guidedby the safeguard policies. 516. The Panel finds that Management did not assess and include in the APRAP a methodology for restitution of the unintended socio-economic costs incurred by displaced persons resulting from project stoppage/delay. This is not consistent with OP 4.12. 54'Management Response, 7 50. 548Project Files, communication dated February 7-9, 2005. 549APRAP, p. 78, Annex 2. IP-PCDP, p.48. 551Appraisal Report, Bujagali Hydroelectric Power Project Uganda, African Development Bank, May 2007, T[ 4.9.7, p. 14. 552Project Files, communicationdated February 7-9,2005. 553Appraisal Report, Bujagali Hydroelectric Power Project Uganda, African Development Bank, May 2007,74.9.7, p. 14. 157 10. Overall Conclusions on Livelihood Restoration 517. According to Bank policy, the loss o f livelihood for involuntarily displacedpersons i s an unacceptable outcome for a Bank-sponsored infrastructure investment.554The APRAP methodology was useful for identifying some livelihood risks but lacked sufficient analysis to mitigate the critical risks, particularly those related to fishing and agriculture. The Panel's review o f the livelihood assessment method and other Project data shows that the Bujagali Project i s facing substantial problems in measuring, monitoring, and mitigating livelihood risks, especially among vulnerable peoples. 518. Annex A of OP 4.12 (719) also requires an implementation schedule for the resettlement plan, as follows "An implementation schedule covering all resettlement activitiesfrom preparation through implementation, including target datesfor the achievement of expected benefits to resettlers and hosts and terminating the variousforms of assistance. Theschedule should indicate how the resettlement activities are linked to the implementation of the overallproject." 519. The APRAP includes an implementation schedule, which links the restoration activities to the construction o f the Project. The Panel notes that a RAP implementation timetable should be policy-driven rather than project construction- driven and be based on the displaced person receiving restitution for losses and achieving sustainable livelihood. This approach requires monitoring o f changes in livelihood restoration (socio-economic conditions o f the affected people), an opportunity missed by not establishing the initial baseline census in2001, and by not correcting this failure inthe preparation o fthis Project. 520. The Panel was not provided any evidence that livelihood restoration has been monitored since the prior Sponsor carried out partial resettlement activities in2001.555 The Panel also notes that in February 2005 Bank social staff recommended that a resettlement audit be carried out because four years had passed since the implementation o fthe first RAP. This call for the audit was unheeded.556 521. Overall, the Panel finds that the Project is innon-compliance with the mandate of Bank Policy on Involuntary Resettlement to improve or at least to restore, in 554OP 4.12 71. Bank policy foresees this unacceptable outcome, stating that Bank experience indicates that involuntary resettlement under development projects, ifunmitigated, often gives rise to severe economic, social, and environmental risks: production systems are dismantled; people face impoverishment when their productive assets or income sources are lost; people are relocated to environments where their productive skills may be less applicable and the competition for resources greater; community institutions and social networks are weakened; kin groups are dispersed; and cultural identity, traditional authority, and the potential for mutual help are diminished or lost. This policy includes safeguards to address and mitigate these impoverishment risks. 555H P P - A P W , pp. 33-34. 556Project Files, communication dated February 7-9,2005. 158 real terms, the livelihoodsand standards of living of the people displacedby the Project.557 D.SharinginProjectBenefitsandCommunityDevelopment 522. Project sustainable development and benefit-sharing i s one o f the principal objectives o f the involuntary resettlement policy. Resettlement activities should be conceived and executed as sustainable development programs, providing sufficient investment resources to enable the persons displaced by the project to share in project benefits. 558 The policy explicitly singles out displaced persons as the beneficiaries. 523. BEL proposed "to develop a Community Development Action Plan (CDAP) for the eight project-affected villages around the Bujagali HPP site." BEL undertook to review the first project CDAP and "determine what further work needs to be undertaken." 524. The CDAP i s to be directly implemented by BEL with participation o f NGOs, consultants and contractors for certain components. BEL'SCDAP proposes activities to benefit the wider communities in the Project area, beyond those individuals and households who have been or will be directly affected by loss o f land, crops or other assets. These activities focus on production related domains (agricultural, small business support and fisheries). 525. The CDAP budget is about 0.4 percent o f the US$ 867 million Project budget.559 Management aligned the CDAP budget to correspond with the restructuring o f the overall roject budget. On the HPP segment, BEL committed to spend US$3.32 million5" on community development over 5 years, including US$ 361,000 for BEL admini~tration.~~~The budget i s not exclusively directed at the displaced persons.562 557 Bank procedure BP 4.12 reads: "During project appraisal, the TT assesses (a) the borrower's commitment to and capacity for implementing the resettlement instrument; (b) the feasibility o f the proposed measures for improvement or restoration o f livelihoods and standards o f living; (c) availability o f adequate counterpart funds for resettlement activities; (d) significant risks, including risk of impoverishment, from inadequate implementation o f the resettlement instrument; (e) consistency o f the proposed resettlement instrument with the Project Implementation Plan; and (f) the adequacy o f arrangements for internal, and if considered appropriate by the TT, independent monitoring and evaluation o f the implementation o f the resettlement in~trument.~ The TT obtains the concurrence of the Regional social development unit and LEG to any changes to the draft resettlement instrument during project appraisal. Appraisal i s complete only when the borrower officially transmits to the Bank the final draft resettlement instrument conforming to Bank policy." 558OP 4.12 7 2(b) 559Calculated as US$ (300K T-line + 3.32M HPP)/867M. 560Responses to IP email o f 18 Jan 2008. Estimated costs o f the CDAP was present at in the 2 April 2007 PAD at US$ 2.4M (7143 on page 42), the higher figure o f US$3.817M appears in Table 8.1 (page 490 o f the December 2006 SEA). BELbudgeted for additional actions it identified after the CDAP was finalized. j6'Bujagali Hydropower Project Social and Environmental Assessment Main Report, Appendix J Community Development Action Plan (CDAP) December 2006 [hereinafter "CDAP"], p.28 "* CDAP,p.28 159 This budget is distinct from the US$497,000 allocated to complete resettlement and income restoration.563 526. Along the T-line, Management submitted a US$300,000 CDAP budget to be paid for and implemented by GoU. The funds are to be divided proportionately among communities based on magnitude o f impact, eligibility criteria, and focusing on enhancing community ~elf-reliance.~~~The allocations are proportional to the magnitude o f im acts and partially proportional to the permanent population livingin the community.59sThe result i s a system that assigns a minimum amount o f the total CDAP funds to villages with small population and small length o f transmission lines and vice versa. 566 The funds are earmarked for community projects such as upgrades to schools, water centers, water supplies, access roads, or connections to public electricity networks. 527. The Panel notes that it i s likely that the community development programs, once executed, will provide positive benefits for Uganda. However, the Panel identified four compliance issues related to the CDAPs: i)the lack o f focus o f the CDPs on displaced persons, ii)inequities in allocations between displaced persons on the T- line and HPP; iii)the lack o f specificity o fthe sustainable development programs, and iv) a decreaseininvestment resources to this effort. 528. Lack of focus on displaced persons The Panel notes that the CDAP, though important demonstration o f the Sponsor's corporate social responsibility, i s not necessarily related to benefit sharing for displaced persons as required by the objectives o f OP/BP 4.12. While the programs offered by the CDAP are directly available to the displaced people (micro-credit, agricultural extension, small business support, et^.),^^^ eligibility criteria do not indicatepreference to displaced person. 529. Lack of Program Specificity: The problem identified by the first Inspection Panel Report over five years ago persists. The Panel finds that in the area o f sustainable development and benefit sharing, the CDAP focuses almost entirely on short-term exercises; its targets are poorly laid out; and it makes no significant or systematic effort to ensure that resources are directed to institution building or social fundamentals rather than only short-term construction projects. 530. Imbalances in allocations between the T-line and HPP. CDAP budgets show sharp differences. The T-line has a higher number o f physically and economically displaced peoples than the HPP, but a smaller proportion o f the resources devoted to CDAP 563PAD, p. 42. 564RCDAP, p. 92. 565RCDAP, p. 92. 566RCDAP, p. 93. 567CDAP, p. 17. CDAP, p.17. 160 activities.569 The Panel finds that budget of the two components were not properly coordinated and this may lead to social discord among the displaced.570 531. Decrease ininvestment resources to this effort. The previous Panel also found "the netpresent value of the resources to be contributed over a 35-year period seems very The HPP CDAP has been significantly reduced both intime and funding for the sustainable benefit-sharing planbetween the prior project and the present Bujagali Project. The prior project had a US$ 7.5M phase I1 CDAP component that is not in the present Project. The CDAP program o f the prior project was also a 35 year program, coterminous with the investment itself. In contrast, the present Bujagali Project has been shortened to five-year construction phase. 532. While the decision to reduce investment resources i s not a compliance issue in and o f itself, the current Panel does not understand why Management decided to further reduce its effort. Even discounting for inflation, eliminating the second phase raises questions as to Management's responsiveness to the previous Panel's findings. The fact that the same problems are surfacing with two different sponsors i s of concern to the Panel. The Panel finds that with limited funding, broad criteria for eligibility and lack of specificity, the CDAP programs do not assure compliance with OP 4.12. E.Indigenous Peoples 533. The Requesters claim that the provisions o f OP 4.10 on Indigenous Peoples have not been applied to the Projectbecause the SEA does not consider the Basoga inhabitants o fthe Project area as indigenous people, inspite o f the fact that the Third Schedule o f the Constitution o fthe Republic o fUganda expressly considers the Basoga as such. 534. The Response states that Management respects local legislation but draws a distinction between the definition o f indigenous people according to the Constitution o f Uganda and that provided inOP 4.10. Under the Ugandan Constitution, inorder to be considered an Ugandan citizen by birth - regardless of socio-economic status - one must belong to one o f the 56 "indigenous communities" listed in the above- referred Third Schedule (or have a parent or grandparent who does); while under the Bank Operational Policy, the term indigenous i s used "in a generic sense to refer to a distinct, vulnerable, social and cultural group" possessing "in varying degrees" the characteristics listed inparagraph 4 o fthe OP. (emphasis added) 535. Although the Basoga people meet some o f the criteria necessary to be regarded as indigenous people inthe context o f Bank-financed projects pursuant to OP 4.10, they are a large and influential group with political, social and economic standing in 569These are estimates since the precise number of economically displaced peoples on the HPP has yet to be determined. We are assuming 160 and roughly 100 economically and physically displaced household on the T-line and HPP, respectively. 570CDAP, p. 24. Micro-credit and animal husbandry extension services are not inthe T-Line budget. 571InspectionPanel Investigation Report 2001, p. 82-83. 161 Uganda's society, and the Panel did not find any indication that they are regarded as a "marginalized and vulnerable segment" of the population that is unable to "participate in and benefit from development. The Panel did not find any evidence " that Management violated the provisions of the Bank's policy on Indigenous Peoples, with regard to the Basoga people.572 572The Panel notes that this finding is consistent with the Panel's 2002 Investigation Report, page 77. See infra Chapter ISection 3.1 ofthis Report 162 Chapter VI11 CulturalandSpiritualValues A. Introduction 536. This chapter analyses the cultural and spiritual issues related to the Project. This analysis begins with the work concluded in 2001, and events since then. Its main purpose i s to analyze Management's actions or omissions in complying with the Bank's safeguards, particularly those dealing with cultural resources. For this purpose, the Panel conducted a careful research and analysis o f relevant materials, including numerous studies by the Cultural Research Center in Jinja, which focuses on Busoga culture. 537. The Requesters claim that cultural and spiritual issues in the Bujagali project area were inadequately covered in the SEA. In their opinion, this "calls for an effective consultationprocess involving all clans that are culturally and spiritually attached to Bujagali Falls followed by a public hearing." The Requesters claim that some consultation was carried out but there was no true participation o f the people in the decision making process; in addition, consultations with the 240 clans in Busoga and 52 clans o f Buganda were not done. The Requesters call for an effective consultation process involving all clans that are culturally and spiritually attached to Bujagali Falls followed by a public hearing.573 538. Management states that BEL i s committed to complying with World Bank OP/BP 4.11, Physical and Cultural Resources. Management states that extensive consultations to address the concerns o f the communities have been carried out since the earlier Bujagali project, including with the Buganda and B ~ s o g Kingdoms, a ~ ~ ~ who, Management claims, are culturally responsible for the villages living on the west and east banks since the project preparation began in 2000 under the original developer AESNP.575Management adds that their commitment to manage cultural and spiritual issues i s part o f the overall social management plan (part o f the Social and Environmental Action Plans, SEAP) which will be implemented throughout the life o f the project. They note that an independent Ugandan NGO, InterAid, will be monitoring all aspects o fthe project, including those related to cultural heritage.576 539. Bank's Physical Cultural Resources Policy OP/BP 4.11recognizes cultural patrimony as important sources o f valuable scientific and historical information, as assets for economic and social development, and as integral parts o f a people's cultural identity 573Request,p. 11. 574A note on orthography: Basoga refers to the people of the Busoga culture. Lusoga is their language, 575Management Response, p. 38. 576Management Response, p. 38. 163 and practices.577 OP/BP 4.11 addresses physical cultural resources, requiring Management to assist sponsors to avoid or mitigate adverse impacts on physical cultural resources.578 540. Consultation i s an important means o f identifying physical and cultural resources, documenting their presence and significance, assessing potential impacts, and exploring mitigation options.s79The policy holds consultation to be important because many physical cultural resources are not documented, or protected by law.580 According to the policy, the EA includes (a) an investigation and inventory o f physical cultural resources likely to be affected by the project; (b) documentation o f the significance o f such physical cultural resources; and (c) assessment o f the nature and extent o fpotential impacts on these resources.581 541. Bank policy on Natural Habitats, OP/BP 4.04, also contains important provisions that apply to Bank-financed activities that may affect (e.g., by inundation) places o f cultural and spiritual significance. OP 4.04 states that the Bank supports the protection, maintenance and rehabilitation o f natural habitats,582 and contains a number o fprovisions to achieve this objective. Paragraph 4 o f OP 4.04 sets a specific and high standard o f protection for "critical natural habitats". This provision states that "The Bank does not support projects that, in the Bank's opinion, involve the signijkant conversion or degradation of critical natural habitats."583Of particular importance in the present situation, "critical natural habitats" under OP 4.04 include bb(. ..) areas recognized as protected by traditional local communities (e.g. sacred groves) (,,.).77584 This issue is dealt with in details in Section H o f this Chapter (Panel's Analysis - Critical Natural Habitats). 542. Operationally, OP/BP 4.11 and OP/BP 4.01 require addressing impacts on the cultural assets and resources as an integral part o f the environmental assessment (EA), and to examine the type, location, sensitivity and scale o f the Project as well as the nature and magnitude o f its potential impacts.585 577OP 4.11 7 2. OP 4.11 (July 2006) replaced OPN 11.03, Management of CuZturaZ Property in Bunk - Financed Projects, (September 1986). OP/BP 4.1lapplies to the Project as its Project Concept Review took place after 15 April 2006. 5780P4.1173. 5790P4.11Tjll. BP 4.117 7. '*'BP4.04 581 4.1 17 8. OP 7 1. 583OP 4.04 7 4. This excerpt includes a footnote to the definition o f the phrase "significant conversion or degradation", as explained inthe text. 584OP 4.04 Annex A. 585BP 4.01 7 2, Footnote 3. Explicit reference to "Location" refers to proximity to or encroachment on environmentally important areas, whereas "Scale" i s judged by Regional staff in the country context. "Sensitivity" refers to projects that may have irreversible impacts, affect vulnerable ethnic minorities, involve involuntary resettlement, or affect physical cultural resources (emphasis added). The Panel observed that County staff from the region who had an awareness o f scale were underutilized in the Bujagali project. 164 543. Inits earlier Investigationreport the Panel indicated the efforts o f the Bank to address the cultural and spiritual issues that the Project raises, and Management's good faith attempts to mitigate these issues. At the same time, the Panel also noted the importance o f including all key stakeholders in consultation and taking steps to minimize the possibility o f disturbance to the local communities that might arise from excluding any faction from such consultations as the Project went 544. The TOR for the Project's SEA, in relation to Cultural Property Management and Status, required BEL to assess the adequacy and completeness of the Cultural Properties Management work o f the previous sponsor, and determine whether further work was necessary.s87The TOR state that detailed archaeological investigations have already been undertaken for the Hydropower project-affected area, compensation has been paid for people's shrines (arnasabo) and appeasement ceremonies have been undertaken to enable the relocation o fthe Bujagali spirits.588 545. The following section provides a review o f the work conducted in 2001 under the prior Bujagali project before analyzing Bank compliance in the Project under investigation. B. The 2001 ResettlementandCommunityDevelopmentActionPlan(RCDAP) andthe CulturalPropertyManagementPlan 546. The 2001 Resettlement and Community Development Action Plan (RCDAP) refers to a 1998 EIA study that led to a number o f detailed studies, including a study o f the traditional religion o f the Basoga and the significance o f the Bujagali site and the implications for the project.589This study, inturn, suggested there could be more sites o f cultural interest in the project area and, as a result, two additional comprehensive studies were commissioned: a Study o f the River Nile and its Significance to Traditional Religion and Practices o f the Inhabitants o fthe River Bank inWakisi sub- county (West Bank) and a comparable study on the East Bank.590Inaddition, the first Sponsor utilized quality control consultants to review the results.s91 547. The 2001 studies included representatives o f caucus groups o f "modern" religions.s92 The Sponsors mapped individual and community level spirits. The studies and focus groups identified dangers concerning breakingtaboos and disturbing the spirit world, including some directly related to construction such as machinery injuring workers, "'Afr-ican 586InspectionPanel Report 2002,1323. Development Bank Volume 1Executive Summary Environmental & Social Auditing Guideline, June 2000, page 11. (Type African Development Bank Volume 1 Executive Summary Environmental & "'HPP-TOR,p. Social Auditing Guideline, June 2000 into google.com) 11. "'RCDAP 2001, p. 96. 590The River Nile and its Significance to Traditional Religion and Practices of the Inhabitants o f the River Nile West Bank. A E S Consultant, September 18, 2000. The River Nile and its Significance to Traditional Religion and Practices o fthe Inhabitants o fthe River Bank inWakise Subcounty. September 18,2000. 591RCDAP 2001, p. 96. 592RCDAP 2001, p. 103. 165 breakdowns, and disappearance o f livestock, women having miscarriages or producing deformed children, and invasion o f the community by foreign diseases and pests.593They identified a general protocol for moving spirits and were informed o f perceived risks to the project and nature should such consultations not take place. The focus groups identified a five point protocol to transfer spirits. 548. AESNP acknowledged the community spirits and that the rapids at Bujagali Falls will be largely inundated and that this i s an unavoidable impact with this project configuration. However, it was reported that the parties involved with the spiritual value o f the site - namely Nabamba Bujagali, Lubaale Nhudu who i s the divine custodian o f the Ntembe Clan that the issue is a local one and the impact is acceptable.594These parties have given their consistent support to the project, as long as the necessary ceremonies, to ensure the spirits are satisfied, are carried 549. Inthe view o fthe earlier Panel report, the 2001 RCDAP assigned little significance to the cultural or spiritual issues o f the Bujagali Falls. The related studies missed the overarching concept o f Basoga religious cosmology,596 including the hierarchical relationships between the spirits. This issue was not raised by BEL either. The RCDAP stated that a preliminary baseline socio-economic survey revealed that the spiritual value o f the Falls i s not an over-riding issue to the majority (83 percent) o f the local community - those in the immediate vicinity o f the Falls.597The report briefly described three spiritual diviners associated with the spirits o f the Falls, but mentioned neither the name o f the spirit, Nabamba B~dhagaali,~~* embodied in the Bujagali Falls nor its significance to the Basoga people.599 550. The RCDAP 2001 also noted that Ugandan Ministry o f Culture and the Kyabazinga (referred to as the "cultural King of the Basoga") presented a statement to a public hearingdeclaring the support o f the Kyabazinga Institutionfor the project but noting however that since a "treasured cultural site" would be lost, it would only be fair that AESNP pay the Institution a fair and adequate compensation.600In spite o f this, the previous sponsor found that "whilst the Falls will be inundated this is not seen as a cultural or spiritual issue of over-riding signijkance by the majority of people who will be directly affected, at the individual, household, local community or national level."60' The RCDAP 2001 states that "on balance the project is judged to comply 593 RCDAP 2001, p. 108. 594 RCDAP 2001, p. 112. 595 RCDAP 2001, p. 112. 596 See Annex C entitled Spiritual Signijicance in Busoga Culture for the description o f Basoga religious cosmology. 597RCDAP 2001, p. 113. 598 Inthis report, the name o f the principal spirit at the Bujagali Falls site i s Nabamba Budhagaali which is distinct from Nabamba Bujagali, the mediumthrough which the spirit communicates. H e i s also referred to as "the Living Bujagali." 599RCDAP 2001, p. 101. 6oo RCDAP 2001, p. 102. InJune o f 2000, the Institution presented a statement to the Open Forum held in Washington that "Bujagali Falls i s a very important cultural site to the Institution o f the Kyabazinga o f Busoga and that the Kyazbainza filly embraced the project " 601RCDAP 2001, p. 113. 166 with WB/IFCpolicy note OPN11.03 in the context of thepressing needfor additional electricity in the country and other beneJitsfrom theproject."602 551. The 2001 Cultural Properties Management Plan (CPMP) sets out a six month, US$125,000 program603o f consultation, compensation o f individuals for disturbed graves and shrines (amasabo), appeasement and relocation o f the Bujagali spirits.604 Three individuals were identified as stakeholders for consultation about the spirits at Bujagali Falls.605In the spring o f 2003, the independent witness NGO, InterAid, prepared a snapshot o f progress on the CPMP,606reporting that, and at the level o f individual spiritual site, consultation, disclosure and compensation for disturbances were proceeding well. 552. Problems, however, emerged with the so-called "appeasement of community spirits."607 InterAid reported that consultations had been taking place with three persons that the Sponsor had identified as custodians/diviners.608Each one o f them was required to specify the requirements they needed for the appeasement o f the spirits o f B~dhagali.~'~ The Sponsor facilitated separate appeasement ceremonies on different days, which were witnessed by multitudes o f people. The Sponsor however tried to combine the appeasement ceremonies and to obtain a co-signed Certificate o f Appeasement, a legal closure, but the three parties did not agree. Following these events, the implementation o f the CPMP stopped for the next four years. Following the selection o fBEL as a new Project sponsor, the Project preparation commenced in 2005. C. Preparationof the Project 553. As noted, the TOR for the present Project's SEA required BEL to assess previous work done by AESNP and determine what hrther work needed to be undertaken.610 Management also felt it important to corroborate if people who live in the project- affected area believe that the Cultural Properties management work undertaken by the previous project sponsor was truly complete. Accordingly, BEL committed to detailed consultation with locally affected communities on their observations and 602RCDAP 2001, p. 113. OPN 11.03 is an earlier version o f the Bank's Policy on Cultural Resources, which applied to the prior Bujagali project, see also footnote 5. 603RCDAP2001,p.112-116. 604RCDAP 2001, p. 101 - 102. 605RCDAP 2001, p. 101-102. 606A E S N P Hydro Electric Power Project, Witness NGO Report on the Implementation o f Resettlement and Community Development Action Plan at Hydro Site, InterAid Uganda April 2003 [hereinafter "AESNP"]. `07A E S N P , p. 71. 608The Sponsor identified the stakeholders for consultations as Nabamba (the living Budhagali) who i s the medium for the Bujagali spirits, Ntembe Waguma and Nfbudu who are caretakers (East Bank), and Nalongo Nakisita who is also a medium for the same spirit but known as Kiira (West Bank). Later in this report the Panel offers corrected clarifications o ftheir respective roles. `09TheMonitoringNGO didnot show an awareness o fBusoga cosmology inits report, taking its lead from the Sponsor's cultural consultant. HPP-TOR, p. 11,72.3.3. 167 opinions on this issue, with follow-up and a revised Cultural Properties Management Plan, as necessary.611 554. BEL'Sconsultations led it to conclude that, rather than a localized cultural site, the Bujagali Falls are o f spiritual significance to the Kingdom o f Busoga as they are considered a place inhabited by spirits.612Though cultural ceremonies were conducted by the previous project sponsor to relocate the spirits, meetings with Kingdom representatives indicated that additional activities may be required to address the spiritual significance o f the area prior to flooding. The Kingdom expressed support for the project and BEL committed to continuing and undergoing consultations with them to determine what needs to be done prior to the flooding o f the Falls.613 555. For the Basoga, the traditional religious structure i s distinct from the cultural structure. What follows i s first a brief description o fthe Busoga spirituality and than a brief description o f Busoga cultural domain, which may help to clarify the ensuing Panel's findings.614A more complete review o f the Busoga spiritual and cultural domain i s attached to this Report as Annex C entitled Spiritual SigniJicunce in Busoga Culture. D.BusogaSpiritualDomain 556. The Bujagali Hydroelectric Project i s moving into a neighborhood long inhabited with strong, complex cultural and spiritual traditions. Although the peoples o f other ethnic groups inhabit the Project area, the Basoga claim s iritual dominion o f both sides o f the Nile, its islands, the water and its waterfall^^'^ According to the 2002 census, there are about 2.7 million Busoga in Uganda whose territory lies to the east o f the Project site. `16 Their language, Lusoga, predominates in this area, on the East bank o fthe River Nile. The Basoga share a common dialect and ideological, spiritual history, sharing a cluster o f eight or more high status spirits - including Budhuguali, the spirit residing at the Bujagali Falls site - who are invoked in their specific ceremonies. The Basoga are distinct from the Buganda, the more dominant tribe in Uganda whose traditional realm reaches to the West bank o f the Nile. HPP-TOR, p. 11. The preparation of the Cultural Properties Management Plan i s discussed further in Section Iof this chapter. "*HPP-SEA Consultation Summary, p. 22, p.4. See also HPP-PCDP DRAFT, November 2006. `I3HPP-SEA Consultation Summary. 22 September 2006. p. 4. See also HPP-PCDP DRAFT, November 2006. `14The Panel consulted with the Cultural Research Centre o f the Diocese of Jinja whose researchers have published over 30 books on Busoga culture and language, interviewed the Requester's cultural experts, Busoga spiritual specialists, the Ministers o f the Kyabazinga, individual Busoga, Management, and the Sponsor. `I5The 2001RAP states its baseline survey identified 22 ethnic groups living inthe project area (HPP-SEA Main Report, p. 161). The region was repopulated by migrants from throughout Uganda and other central African countries inthe 1940's after being nearly abandonedby the Busoga at the turno f the century due to sleeping sickness. (RCDAP 2001, p. 98) www.busoga.codaboutBusoga.php 168 557. To the Basoga, the Project area - like their entire region - i s inhabited by ancestral spirits and living humans who are constantly interacting - from birth to death and beyond.61 558. From the perspective o f the Bujagali Project, the key elements o f Busoga spiritual cosmology are: a) the spirits are innumerable,powerful and frequently cross over into the world o f the living and may do both good and bad, b) they inhabit the same world as the living and are associated with animate and inanimate objects throughout the landscape, c) they can move freely without the need for human permission, d) they have differential power, influence, and interests, e) they are hierarchical, somewhat comparable to the ancient Greek Pantheon, f) they influence the health, well-being and the livelihood o f the living, g) more powerful spirits communicate through mediums who do not view themselves as capable o f negotiating or predicting spirit behavior - they are mediums of the spirit who possesses them, and h) the mediums are selected by the spirits, not by the cultural (political) leaders. 559. The intersection o f spirit, place, and its medium defines the cultural resource at risk under OP/BP 4.11, a situation common to significant cultural sites throughout the world.618 560. The Panel's review of available evidence collected during the investigation confirms that Budhagaali Falls are the residence of a host of spirits ranging from high level Busoga spirits to individual family spirits among whom i s one o f the Busoga's most venerated, powerful, princely spirits, Nabamba Budhagaali Spirit.619In 2001, the Project noted that the Ntembe clan, whose leader i s Ntembe Waguma, and diviner (muswezi) i s Nfuudu, has clan level ancestral spirits at the Bujagali Falls site which will be disturbed by the project. Like other clans, the Ntembe are found throughout Busogaland.620 A 2001 map o f community level sites o f cultural significance, included in the RCDAP, shows; 16 islands, 32 shrines, 10 large trees, 6 rocks, 20 burial ground, 2 fire places, and a forest in the immediate project area (see Figure 1 below).621 617 Over fifty years ago, Lloyd Fallers, in his classic study o f the Basoga, Bantu Bureaucracy (1954), felt that despite the substantive presence o f Catholicism and other global religions, ancestor worship was "very near the heart o f the Soga value-system." p. 80 618 Routine and Dissonant Cultures: A theory about the psycho-socio-cultural disruptions o f involuntary resettlement and ways to mitigate them without inflicting more damage. Theodore E. Downing and Carmen Garcia-Downing. IN Anthony Oliver-Smith. Development and Dispossession: The Anthrouolom of Displacement and Resettlement. Santa Fe: School for Advanced ResearchPress2008, inPress. RCDAP 2001, pp. 101-102, and map o f community sacred sites (including spiritual locations of rocks, trees, shrines). A E S contracted a Consultant to survey the traditional religious sites and beliefs in communities along the East and West banks of the Nile River and identified specific names for these features. Because the Panel reviewed the draft form o f these studies containing the surveys they are hereinafter referred to as " A E S Consultant." 620The reports are ambiguous as to the spiritual and clan leadership o f the Ntembe clan, with one document referring to Lubaale Nfuuduas the leader o f the Ntembe clan and another assigning this position to Ntembe Waguma. HPP-APRAP, p. 23, footnote 3. 621RCDAP 2001, pp. 101-102, and map of community sacred sites (including spiritual locations o f rocks, trees, shrines), Figure 15, p. 107. The islands are named for their resident spirits, with Kiwotokwa Island a 169 561. Surveying the traditional religious landscape adjacent to the project area in 1998 or 1999, an AES consultant concluded that the traditional beliefs associated with these natural features play an environmental conservation, security and mental health role for the believers. One of the islands to be submerged i s used for ceremonies to find missingdrown bodies. He concludes that "the implication of destroying the islands is that the spirits will disappear. Those whose family members regularly use the Nile waters will be put in a situation of fear of the unknown regarding what else to do when one of their people drowns. The associated ear and helplessness, might lead to variousforms and degree of mental breakdown." 22 l 562. Absent the full CPMP inve~tigation~~~, in compliance with OP 4.11, the full breadth o fthe Bujagali Falls spiritual site at the higher level o f Busoga cosmology has not yet been established. At the level o fthe Princely, higher spirits, all Busoga clans and their Bujagali Falls associated baswezi are stakeholders. Picture 9 Nabamba Bujagali performingspiritual ceremony resident site o f for wisambiva or high spirits - Nalongo, Nabamba Budhagaali, Walumbe and Mukasa. The River Nile and its Significance to Traditional Religion and Practices of the Inhabitants o f the River Nile West Bank. A E S Consultant September 18,2000, p. 41. 622The River Nile and its Significance to Traditional Religion and Practices o f the Inhabitants o f the River Bank inWakise Subcounty. September 18,2000. A E S Consultant, p. 24. Another island to be submerged i s usedfor infertility Ceremonies(Kirongo Island). 623The preparation o f the Cultural Properties Management Plan is discussed hrther in Section Iof this chapter. 170 E.BusogaCulturalDomain 563. Busoga Kingdom i s a cultural institution that promotes popular participation and unityamong the people ofBusogathrough cultural and development programs for the improved livelihood o f the people o f B ~ s o g aUnder~Article 246 o f the Ugandan . ~ ~ Constitution, the Busoga Kingdom i s assigned limited authority. Unlike the typical monarchies in Africa, the Busoga did not have a central authority at the advent o f British rule. Nevertheless, it had developed small principalities, each with its own hereditary ruler. These principalities were later to be consolidated under a King called "lsebantu Kyabazinga" who ruled the Busoga Kingdom.625 564. This secular institution, which i s a stakeholder on Busoga cultural issues, makes no claims to hold spiritual power, a position consistent with the Project's consultations with the Busoga Kingdom Prime Minister.626 In early August 2006, the Prime Minister explained that the spirits o f the Falls have not been adequately released and expressed his feelings that the previous EIA did not adequately capture the effects o f the Bujagali Falls inundation on the spirits o f the Falls and noted that there needs to be collective beliefo f this spiritual question among the community.627He su ested a meeting with the Busoga cultural leaders to identify a way forward!' At a subsequent meeting on August 18, 2006, nine o f the 11 Busoga cultural leaders reconfirmed that cultural issues of the project were not addressed, that the spirits and the Falls and shrines need to be relocated, that the entire Falls lies within the Kingdom. The nine present cultural leaders felt that all 11 needed to be involved.629 While offering to search for a pathto a solution, this conclave o f cultural leaders did not claim spiritual authority. 565. The Panel finds that Management and the Sponsor have increasingly recognized and involved the Kyabazinga Institution as an important guardian of the Basoga cultural tradition. The Panel also recognizes that the Kyabazinga Institution i s not empowered to speak as surrogates in consultations for the Basoga spiritual stakeholders. F.Panel's Analysis-PhysicalCulturalResources 566. As noted above BEL'Sconsultations led it to conclude that, rather than a localized cultural site, Bujagali Falls are of spiritual significance to the Kingdom of Busoga as it is aplace inhabitedby spirits. The Panel notes however that inthe Project Appraisal Report (PAD) Management downplayed the consultation findings showing the 624Official BusogaKingdomwebsite: www.busoga.com/theKingdom.php 625Official BusogaKingdomwebsite: www.busoga.com/theKingdom.php 626Panelinterview with the same Busoga Cultural Minister, who is no longer inoffice. 627HPP-PCDP, p. 35. Summary of meeting of 11Aug. 2006. 628HPP-PCDP, p. 35. Summary of meeting of 18 August 2006. HPP-PCDP,p. 35. Summaryof meeting of 18 August, 2006. 172 spiritual significance o f the Bujagali Falls spiritual site to the broader Busoga community and downplayed the significance o f the cultural resources.63o 567. Panel interviews revealed that the 2001 CPMP and its 2006 assessment might have significantly mischaracterized key elements o f the Busoga cosmology. A CPMP, in compliance with OP/BP 4.11, should have identified Bujagali Falls as a significant cultural resource, triggering rigorous safeguards for specific avoidance, consultation and mitigation. 568. OP 4.11's approach to cultural resource i s based on three essential components: avoidance, consultation and mitigation. 1.Avoidance 569. The Panel observes that, since the initiation o f the Bujagali Project, Management did not adequately consider avoidance o f the significant cultural resource impacts at Bujagali Falls.631 570. Management decided not to conduct a full CPMP for the Project. This very likely contributed to the fact that the issue of avoidance was not raised effectively in the SEA (see also Chapter 5) and inthe Assessment o f Past Resettlement Activities and Action Plan.632This contributed to the non-alignment o f avoidance and mitigation measures, as required under Bank policy, to address the threat to the cultural resource. 630PAD, Annex 15,173. The PAD states that "The project covers some physical features that are culturally significant to local people. These consist o f various types o f rocks, trees, and land sites that are associated with spiritual forces. Local beliefs attached to these spirits influence events inpeoples' lives. For example, residents believe that the spirits are contacted by mediums or local practitioners or traditional spiritual leaders. During the preparation o f the previous Bujagali project, local spirit mediums contacted the spirits and reported that ifappropriate ceremonial procedures were financed by AESNP and carried out, the spirits would accept project-induced changes to the spiritual landscape o f the project area. The previous project undertook extensive consultations with local people, religious leaders, and relevant government authorities in order to reach a consensus on this issue. AESNP carried out these ceremonies. BEL has carried out additional consultations, especially with the Kingdoms o f Buganda and Busoga, and has learned that some additional ceremonies may be needed. BEL will also institute a Code o f Practice on cultural issues, along with training, for workers and contractors during the construction and operation phases. Many households construct small hut-like structures (known as amasabo), which serve as shrines to ancestor sprits (these spirits are family-related, as opposed to the universal spirit forces discussed above). AESNP had mapped all such shrines and initiated a compensation procedure for their reconstruction and associated ritual procedures. BEL will complete any unfulfilled commitments." 631 The 2001 CPMP's only discussion o f avoidance stated that: "At the level o f the wider community A E S N P acknowledges that the rapids at Bujagali Falls will be largely inundated and that this i s an unavoidable impact with this project configuration. However, it is considered by the parties involved with the spiritual value o f the site - namely Nabamba Bujagali, Lubaale Nfuuduand the Leader o f the Ntembe Clan that the issue is a local one and the impact is acceptable as long as appropriate measures are taken. Toward this end, these parties have given their consistent support to the project, as long as the necessary ' ceremonies to ensure appeasement o f the spirits are carried out." 632 HPP-@RAP 173 571. The Panel notes that if the Busoga religion and cultural tradition had been a more fully understood and widely recognized one, the current site may not have been acceptable, or alternative sites would have been given a much stronger consideration and weight. The Panel finds that Management failed adequately to consider or implement alternatives to avoid the project-related impacts on Busoga spirituality and culture inviolation of OP/BP4.11. 2. Consultation 572. In its early consultations, Management concluded that the local community did not see potential negative impacts o f the Project on traditional culture and that certain traditional ceremonies could mitigate impacts.633This seems not to be consistent with the information gathered by the Panel during its field visits. The discrepancy may be explainedby focusing on who was consulted. 573. The 2001 RCDAP findings came from an AES sponsored study o f the traditional religion and practices in communities located nearby the project construction and flooding area on both banks o f the Nile. The study interviewed 20 focus groups, half with women, which included community's residents, local government council representatives (LC), traditional religious and healthpractitioners, and representatives from what they termed "modern religions."634 On the East Nile bank, the survey also interviewed diviners with special interest in the Bujagali site, including Nabamba Bujagali and Lubaale Nfuudu.635 The RCDAP concluded that whilst the Falls will be inundated this is not seen as a cultural or spiritual issue o f over-riding issue to the majority (83 percent) o fthe local community.636 574. The Panel notes that determination o f the significance o f a spiritual site requires consultation with the affected parties.637The Panel considers that the consultation methodology used in this RCDAP was detailed, but structurally flawed. First, the survey included mostly laymen many o f whom were not sufficiently knowledgeable o f the traditional religion. Second, it excluded key Busoga clans' spiritual leaders (baswezi abadhagaali) who have a strong spiritual attachment to the site and whose livelihood might be impacted by its flooding.638The consultations did not recognize that mediums o f the Nabamba Budhagaali derive their power through recognition by the traditional clan priests (muswezi) as agents o f their believers. The mediums o f the high Busoga spirits are incapable of commanding their followers, meaning that the appropriate consultation strategy i s participatory, as this i s common among traditional religions. 633 RCDAP 2001, p. 19 and AESNP Hydro Electric Power Project, Witness NGO Report on the Implementation ofResettlement and Community Development Action Plan at Hydro Site, InterAidUganda April 2003, pp. 22-23. 634RCDAP 2001, pp. 96, 103. The study included 2 villages as a control group. 635RCDAP 2001, p. 126. 636RCDAP 2001, p. 113. 637BP 4.117 7. 638The unavailable 1999-2000 study on the traditional religion of the Basoga may have reviewed this information, but was unavailable to the Panel. 174 575. Third, the consultations assumed, not determined, that the spirits at Bujagali Falls were Basoga, not limited to inhabitants nearby the Project site, a fact reconfirmed in 2006 when the Kyabazinga council indicated that consultation on a path to deal with spiritual issues required consultations with cultural leaders throughout the Basoga (see above).639The information brought to the Board during consideration o f the Project was inaccurate since it was based on a survey o f people in the project area, many o f whom were non-Busoga migrants who had moved into the area following a disease-linked depopulation. Most of those who believe in the significance of the Bujagali Falls spiritual site do not live in the immediate vicinity of the Project. The terms o f reference for the cultural consultations were not revised after interviews discovered that the spiritual sites in the project area were o f major significance to a religious tradition that extended beyondthe immediate area o fthe study.640 Picture 10 Undated picture received by Panel expert showing spiritual M e,dium _ _ NfuuduperformingSpiritual Ceremony 576. The Panel offers three illustrations o f situations inwhich Management acted in a way inconsistent with Bank policy. 577. First, the SEA'STOR limits the consultations to within the project-affected area.641 Nonetheless, the Panel notes that the SEA'Sconsultant expanded the consultation to include the cultural Kingdoms o f Buganda and Busoga. These expanded consultations 639BP 4.11 7. 640A E S contracted a Consultant to survey the traditional religious sites and beliefs in communities along the East and West banks o f the Nile River. Inthe 2001 RCDAF' references to "community level" spirits - meaning spirits discovered in the AES survey with significance above the household level. In their consultations with Nabamba Bujagali and Nfuudu, both mediums stressed that the spirits they were concerned about had broader significance throughout Basogaland - not just in the project area or "community" (see The River Nile and its Significance to Traditional Religion and Practices o f the Inhabitants of the East Bank. A E S Consultant, September 18,2000, page 89 and 92). This fact was ignored leading to a significant misinterpretation of the importance o f the cultural properties, a mistake which compliance with OP 4.1 1 would have avoided. 641HPP-TOR, p. 11. 175 yielded valuable information: a) that the Spiritual problem persisted, b) that it was a Busoga, not just a local issue, and c) that consultations with a wider range o f stakeholders was necessary. Despite this new information, there was no follow-up. In the PAD, Management states that the Kingdoms supported the Project and BEL i s having on-going consultations with local traditional authorities and has committed to measures to ensure that these issues are properly addressed prior to and during construction.642There i s no mention o f the expanded consultations and their results. 578. Second, the Panel was informed that Management contacted the Cultural Research Centre, runby the Diocese o fJinja, an authority on Busoga spirituality, not for advice on consultations, rather only for translation purposes. According to the Centre, its cultural experts offered their assistancebeyond translation but they were refused. The Panel could find no evidence that the Cultural Centre's information was reviewed or incorporated into the project planning. 579. Third, on September 28, 2001 at the only large ceremony conducted to appease "the Budhagaali community spirit"643an unspecified number o f clan spiritual leaders, the baswezi abadhagaali and important dignitaries from all over Busoga were transported to the site at the Sponsor's expense. The followers o f the Budhagaali were concerned with the rumor that the construction o f the dam would take place at their sacred site. They were satisfied, however, when it was revealed that the dam would not be constructed at the site but 3 kilometers downstream at Dumbbell Island.644 580. The Panel was informed during its field visit that the Sponsor and Witness NGO present at the ceremony did not intervene to correct the misimpression that the sacred site was not to be destroyed.645The lack o f action to address this misimpression was inconsistent with consultation and disclosure requirements under OP 4.1 1711 and 7 12. Without a follow-up consultation the Panel i s concerned that the principal stakeholders have not understood the extent o fthe Project's impact. 581. The limited consultation creates on-going uncertainties as to affected people's acceptance o f the project's cultural resource impacts.646The Panel fmds that the Project failed adequately to consult with the Busoga spiritual clan leaders associated with one or more high status Spirits about the significant cultural patrimony of the Bujagali Falls. This is not incompliance with OP 4.11. 3. Mitigation 642PAD, p. 133 643 AESNP, p. 75. The Monitor i s unclear as to whether the ceremony was for the Budhagaali community "spirit" or "spirits"-referring onthe same page to both. 644AESNP. 645AESNP, p. 78. 646 The policy provisions for chance finds would be acceptable for dealing with unanticipated decisions by the spirits should they appear during the consultationprocess and might have been included inthe Cultural Property Management Plan. 176 582. Inits 2006 assessment o f the 2001 CPMP, Management concluded that compensation for individual cultural sites was completed. The mitigation usually involved a comprehensive consultation exercise with dedicated groups in each o f the interested communities, who were tasked with identifying the sites and devising adequate compensation measures, which included compensation for the structures and compensation for a ceremony allowing for relocation o f the a m a ~ a b oThis ~ . ~ ~method i s consistent with that used by the construction industry in Uganda. The mitigation was organized by specialized consultants on behalf o f the Sponsor and witnessed by a local NGO. The local NGO concluded that there i s no pendingissue with respect to relocation o f these sites,648save issues concerning relocation o f spirits on the river island.649The Panel concurs with this assessment. 583. The Sponsor's approach has been to identify three interested "stakeholders" in the "Bujagali spirit(s)" and fund either appeasement or relocation ceremonies. The Sponsor focused on obtaining written consent from three stakeholders that compensation hadbeen adequate and that construction o f the dam at Dumbbell Island could proceed. It i s evident from the 2006 consultations that this approach was not working: Busoga cultural leaders and the Panel interviews with the Nabamba Bujagali, cultural experts, the Sponsor, and Management agreed that the mitigation was incomplete. 584. As o f the Panel's visit in NovembedDecember 2007, the appeasement ceremony attempted in 2001 organized by Nabamba Bujagali has led to uncertain results. The spiritual medium claims it was incomplete and he i s still uncertain whether or not the spirits will be appeased if another ceremony occurs. In meetings with the Panel and others he has explained that the clan spiritual heads should be present and he cannot predict what the Nabamba Budhagaali Spirit will do. Meanwhile, Lubaale Nfuudu has relocated the "Bujagali spirits'' to a temporary location, from which they will be moved, once more, to a suitable place away from the Project site to be purchased by the Sponsor. Project reference to undifferentiated "Bujagali spirits" makes it difficult to determine whether or not there are rival claims or just a rivalry between the two spiritual mediums.650 585. Misidentifying Bujagali Falls as a local cultural resource, misaligning its consultation strategy, and failing to prepare a new Cultural Property Management Plan compounded errors and muddled mitigation. Resultant problems included loss of objectivity of the Sponsor, impatience, assignment of pecuniary motives to stakeholders, cost cutting, culturally inappropriate mitigation efforts, and most importantly, a misunderstanding that the Bujagali Project is ensconcedin a long-term relationship with its new neighbors and their spiritworld. 641 HPP-APRAP,p. 23 7 5.2. HPP-APRAP, p.23 7 5.2. 648 649AESNP, p. 67. 650AESNP, p. 71. 177 G.Understanding the LocalCultural and Spiritual Context 586. The Project i s being implemented in a complex cultural and spiritual environment, including several spiritual beliefs and traditions that, according to Bank policy need to be taken into account inproject design and implementation. This requires a special effort to understand and show sensitivity to the beliefs o f local people. What follows i s a description o f how Management has dealt with these issues in the context o f the Project. 587. In a public document, Management loses objectivity by subjectively judging the genuineness o f the interest o f the mediums in cultural and spiritual aspects, including claiming expertise in measuring actual spiritual performance. They state that "there has been fierce rivalry between Nabamba Budhagali on the one hand and Ntembe and Nfuudu on the other during the whole consultation and negotiation process, Nabamba has been quite successful in attracting media attention and obtaining significant compensation, whereas the other two seemed to be more genuinely interested in cultural and spiritual aspects.9'651 Furthermore, with reference to Nabamba Bujagali, they state that "while the two other stakeholders appear to have been genuinely satisfied with measures taken by AESNP, the Nabamba Budhagaali medium seems to have remaining claims over the site. Thisparticular individual has been able in the past to draw a lot of attention, including international attention, which later did not appear to be justified by his actual spiritual performance, in contrast with the other two. It cannot be excluded that he will seek to obtain more compensation through media coverage for instance'7652(emphasis added). The indicators Management used to give credibility o f one medium over the other are inappropriate. A medium's credibility accrues from their believers.653 588. Following a ceremony financed by the first Sponsor on September 28, 2001, to relocate the Bujagali spirits, Management claims that all three interested mediums acknowledged inwriting that compensationhadbeen adequate and construction of the dam could proceed with the partial inundationo f Bujagali Rapids as a The witness NGO contradicts this account. While they agree that the Sponsor prepared a single "Certificate o f Appeasement" agreement to be signed by three sponsor- identified stakeholders with spiritual interests, they claim that the ne5 otiations on October 2,2001 with the NabambaBujagali withheld his endorsement.6 y 651HPP-APRAP, p. 23, footnote 3. 652HPP-APRAP, p. 23. 653 The witness NGO draws a superficial culturaljudgment based on the Nabamba Budhagaali request for money, bottled beer and soda vs. the Lubaale Nhudu's requests for more traditional drinks, and other criteria not systematically aligned with cultural knowledge of Basoga. For example, the witness NGO questions his ceremonial legitimacy because the Nabamba Budhagaali drew followers primarily from the Basoga, not across the Uganda as he had anticipated. They berate his lack o f powers to float across the rapids on a bark cloth. The attendance is consistent with the claim that the Bujagali spirit is a Basoga, not a pan-national site (AESNP, p. 75). 654HPP-APRAP, p. 23. 655AESNP, p. 81. 178 589. The Panel notes that the insistence on a Certification o f Appeasement tied to the construction o f the dam i s not part o f OP/BP 4.11 or OP/BP 4.04. This novel document reflected a misunderstanding o f the medium's role. In interviews with the Panel, the Nabamba Bujagali, as a spiritual medium, insists that Bujagali Falls i s a significant cultural site that requires more costly and time- consuming consultation with the Busoga spiritual clan leaders. He could not assure the Sponsor o f the outcome o f spiritual consultation.656The Nabamba Bujagali stated that with Busoga spiritual logic, he could not sign the agreement for the Spirit.657 He also claimed that the ceremony on September 28, 2001, had been called not to conduct the ritual o f appeasementbut to consult his buswezi B~dhagaali.~~* 590. Lubaale Nfuudu felt the spirits had been moved to a tem orary location, on his property and will be relocated again nearby the Project site.6S'The Nabamba Bujagali medium seems to have remaining claims over the site. The Panel notes that 2001 Project documents identify the Lubaale Nfuudu as a diviner (muswezi) who asserts that the spirit Lubaale i s the father o f Nabamba Budhagaali spirit.660He conducts occasional ceremonies with busweszi at the Bujagali Falls to communicate with Lubaale, one o f the highest spirits within Busoga cosmology, but different from the Bujagali spirit. This opens the possibility that Bujagali Falls, as a cultural property may be the site o ftwo highspirits o fthe Busoga, not one. 591. The Panel finds that Management publicly injected the Bank into a religious misunderstanding without competence inthe cultural spiritual context o f its position, including passing judgment on legitimacy and credibility o f a spiritual medium's performance. Management unnecessarily and inappropriately took sides in a spiritual controversy of a religion in which millions of Ugandans believe. The Panel finds this action by Management to be non-compliant with the OP 4.11. 592. Among the Busoga, as in most cultures, healing and spirituality involve numerous specialists who are compensated, often in kind, for services that provide peace o f mind and meaning to the lives o f their constituents, patients or believers. This ceremonial budget did not include compensation to the religious specialists. The spiritual medium, Nfuudu, told the Sponsor's cultural researchers that he borrowed money against his land title as collateral for a ceremony requiring him to transport house about a hundred baswezi.661The Sponsor questioned the ulterior motives o f the mediums and small ceremonial costs. Negotiating minor costs without understanding the ceremony itself or the importance o f its participants in the overall project consultation underscores loss o f Management focus on resolving its cultural resource issues. These representation costs are frequently covered within the cost o f consultation. 656 HPP-APRAP, p. 23. "*657 A A E S N P , p. 80. E S N P , p. 80. 659 HPP-APRAP, p. 23. 660 RDAP 2001, pp. 101-102. 661 The River Nile and its Significance to Traditional Religion and Practices of the Inhabitants of the East Bank. A E S Consultant September 18,2000, p. 92. 179 593. Furthermore, a prime example o f the culturally inappropriate mitigation efforts was the attempt to achieve closure to the cultural resource mitigation over the Bujagali Falls. Four days following this consultation at the Falls ritual site, he refused to sign because the document assigned the Budhagaali Spirit to the Ntembe Clan, a clear contradiction with Busoga beliefs. Following this dispute, he refused to endorse the Certificate o f Completion o f Appeasement and even refused to collect a USh 1,000,000 check for the preceding ceremony.662Inexplicably, the 2006 Assessment reported that all three interested parties had acknowledged in writing that compensation hadbeen adequate and that construction o f the dam at Dumbbell Island could proceed.663 594. The Panel finds, consistent with Busoga beliefs, that the spiritual mediums cannot provide assurance as to whether or not the Project could proceed before consulting the Spirits ina manner appropriate to their culture. As Nabamba Bujagali explainedto the Panel, the Spirit speaks through him. Non-believers may view this response as nonsense, believing that spiritual mediums are speaking for themselves. As such, he can provide no guarantee. 595. The HPP Consultation Summary Report notes that the river and many islands and rapids in the roject area hold cultural/religious values for some local persons and communitie~!~ In the public consultation from October 5-6, 2006 attended by 150 people, an issue was raised about an unspecified location within the river (which could have been one o f the islands) o f spiritual significance to which the Sponsor replied that they would consult with the Busoga Kingdom.665 Grave yards are archaeological sites and may be culturalhpiritual sites, whose significance i s established through ethno-archaeological investigations. With reference to the islands, the Sponsor felt it was impossible to locate these graves with certainty and therefore also impossible to exhume and relocate their bodies. The new Sponsor assumed the mitigation strategy developed by the previous to hold an inter-denominational remembrance service to honor the memories o f those buried in the islands. N o consultation or ethno-archaeological work had established the provenance o f the remains to determine the culturally appropriate mitigation. The Panel obtained information that the islands may be the location where previous spiritual media are buried. Noting that appropriate consultation and mitigation has yet to be done for the Bujagali Falls spiritual site, the Panel observes that the island areas must be included in the mitigation strategy to reach compliance with OP/BP 4.11. Management's treatment o f these remains i s inconsistent with the provisions being made for archaeological discoveries along the T-Line.666 662AESNP, pp. 79-81. 663HPP-APRAP, p. 23. 664HPP Consultation Summary Report, 22 September 2006, p. 2. 665HPP-PCDP, p. 43. 666IP-SEA, p. 98 and287. 180 596. The Panel finds that Management assumed that what they called the "Bujagali spirits" were restricted to the Project construction and flooding area, in contravention to the BP 4.11 requirement that they work with and assist the Borrower to identify the spatial and temporal boundaries of the cultural resources affected by the This did not comply with avoidance and mitigation requirements of OP/BP4.11. 597. Narrowing its size, location, and scale, Management discounted the significance o f what should have been identified as the Bujagali Falls spiritual site to all o f the Busoga, not just to those living in close proximity to the Project area. It appears that Management defined the project-affected-people under OP 4.11 on Physical Cultural Resources as those covered under OP/BP 4.12 on Involuntary Resettlement. In the case o f the Bujagali project, the groups are distinct. Consequently, the Panel fmds that the culturally and spiritually affected people were not adequately identified as required by Bank policy.668 H.Panel's Analysis CriticalNaturalHabitats - 598. Given the importance that the Requesters attach to the spiritual aspects o f the Falls, the Panel examined indetail the Bank's consideration o fthis issue inlight o f different policies. In the Project, these issues have mainly been considered under the Bank policy on Physical Cultural Resources (OP/BP 4.1 1). However, OP 4.04 also contains provisions that are relevant to these issues, as discussed below. 599. Project documents recognize that the inundation o f the Bujagali Falls will destroy a natural habitat o f significance to the people o f Uganda, and identify specific actions to offset this impact.669At the same time, Management takes the view that the Project i s not significantly converting or degrading a "critical natural habitat" as defined in OP 4.04.670The Panel analyzes the various dimensions o f that decision in light o f provisions contained inthe Bank policy. 600. Since OP 4.04 states that the "Bank does not support projects that, in the Bank's opinion, involve the signijkant conversion or degradation of critical natural habitats," the Panel reviewed what constitutes a critical natural habitats. Annex A o f OP 4.04 defines "critical natural habitats" as 667BP 4.11,16. 668Management recognized that "cultural sites and traditional believes appear to be closely associated with ecological features, like the River Nile, large trees, and boulders, each as a resident spirit which is worshiped"(RCDAP 2001, p. 106). This statement may be true for most o fAfiica, ifnot the world. 669See PAD, 1157. See also Letter from Bank Country Manager to Minister of Energy & Mineral Development, April 25, 2001 (Bujagali Hydropower Project: World Bank Group's Requirement o f an Offset at KalagalaFalls). 670Safeguard Datasheet, March 26, 2007, p. 5 (Section OPBP 4.1). Inreviewing Bank Policy on Natural Habitats, and actions to offset the impacts o f the inundation, both the PAD and the SEA state that "the land take and the inundation will not impact critical natural habitat." PAD, 7 157. See also Letter o f April 25, 2001, noted above. 181 "(i)existingprotectedareasandareasoficially proposedbygovernments as protected areas (e.g., reserves that meet the criteria of the World Conservation Union [IUCN] classijkations rootnote omitted]), areas initially recognized as protected by traditional local communities (e+, sacred groves) and sites that maintain conditions vitalfor the viability of theseprotected areas (as determined by the environ-mental assessmentprocess; or, ... (ii) sites identij?ed on supplementary listsprepared by the Bank or an authoritative source determined by the Regional environment sector unit (RESU). Such sites may include areas recognized by traditional local communities (e+, sacred groves);..." (emphasis added) 601 Thus OP 4.04 indicates that socio-cultural factors do have a bearing on the assignment o f "criticality" to a natural habitat. The Panel further observes that there i s substantial literature and practice recognizing the important relationship between sacredplaces and the conservation o f natural habitats and protected areas, a subject o f much attention in recent years. IUCN Guidelines for Protected Area Management Categories, referred to in the definition o f Critical Natural Habitat under OP 4.04, state that a Category I11 Protected Area i s an "[alrea containing one, or more, specijk natural or natural/cultural feature which is of outstanding or unique value Significance.9, because of its671inherent rarity, representative or aesthetic qualities or cultural (emphasis added). 602. The Panel notes that "areas initially recognized as protected by traditional local communities (e.g. sacred groves)", as referredto inOP 4.04, include areas recognized as protected for their cultural significance and ecological functions by traditional peoples. In the Bujagali Falls area, Project studies and the Panel have identified islands, sacred groves, rocks, waterfalls, andnumerous Busoga spiritual sites. 672 The persistent resistance to disturbance o f the site by the Busoga spiritualists and the expressed concerns o f the Kyabazinga Institutions i s evidence that Bujagali Falls are a natural habitat o f great importance to the Basoga that i s being protectedby them, as provided in OP 4.04. The discussion in the Report describes and documents the cultural and spiritual significance o f the Bujagali Falls site to the Busoga people. In addition, studies conducted by AESNP for the prior Bujagali project suggest a strong ethno-botanical use o f the Bujagali Falls project area, in particular the islands, for ". ..The 671 current draft revised IUCN Guidelines amplify on this element. They note that sacred sites have intercultural and crosscutting values which, in turn produces equitable synergies between spiritual, cultural and natural diversity in support of more holistic conservation objectives," and provide that "Category I11Protected Areas could include: ..Natural-cultural sites: such as the many forms of sacred natural sites (sacred trees or groves, springs, waterfalls, mountains, sea coves etc) of importance to one or more faith groups ... Cultural sites with associated ecology: where protection of a cultural site also protects significant and important biodiversity, such as archaeological,' historical sites that are inextricably linked to a natural area." See draft o f revised Guidelines for applying protected area management categories, IUCN, July 2008, pp. 98,32). 612 The AES Consultant study was preceded by an earlier 1998 commissioned study o f the traditional religionof the Basoga people and the significance o f the Bujagali site inparticular, includingBujagali Falls (RCDAP 2001,1 13.9, p. 96.) This study was not made available to the Panel. 182 healing and mental well-being. These studies include an ethno-botanical survey with these numeroushealers to identify the flora associated with their practices. 673 603. AESNP cultural consultant concluded that: "Cultural sites and traditional beliefs appear to be closely associated with ecologicalfeatures like River Nile (Kiira), large trees, and boulders. m e r e any of thesefeatures are found, respondents find a resident spirit. These spirits are worshipped, respected andfeared. These attitudes are manifested through rituals, sacrifices and observation of taboos." "The beliefs and practices associated with the ph sical features play several roles. One is an environmental conservation role,6' another is a security role, and the other is the mental health role. ... The River Nile (Kiira), its rapids, islands and rocks and bank play a central role in the religious (traditional) lives of the inhabitants of Wakisi subcounty. ... We identified four categories of religious and quasi-religious objects. i. Thefirst category includes the natural ecological objects such as islands, the Nile, streams, trees and rocks. 673 RCDAP 2001, p. 102, states that in 1999, the Kyabazinga refers to the Bujagali Falls as "a treasured cultural site would be lost." And in June o f 2000, the Kyabazinga Institutionpresented a statement to the Open Forum held in Washington that "Bujagali Falls is a very important cultural site to the Institution of theKyabazinga of Busoga." 674The recognition o f this "environmental conservation role" is noteworthy. The Panel notes that there i s substantial literature andpractice highlighting the importance o f sacred sites for and as part o f conservation objectives, individually and collectively, as well as for inter-related spiritual and cultural value. IUCN Guidelines for Protected Area Management Categories, referred to in the definition o f Critical Natural Habitat under OP 4.04, state that a Category 111Protected Area is an "[alrea containing one, or more, specific natural or naturaVcultura1 feature which is o f outstanding or unique value because of its inherent rarity, representative or aesthetic qualities or cultural significance." (emphasis added). The current draft revised IUCN Guidelines amplify on this element. They note that sacred sites have ". . . intercultural and crosscutting values which, in turn produces equitable synergies between spiritual, cultural and natural diversity in support o f more holistic conservation objectives," and provide that "Category 111Protected Areas could include: . .Natural-cultural sites: such as the many forms o f sacrednatural sites (sacred trees or groves, springs, waterfalls, mountains, sea coves etc) o f importance to one or more faith groups. ..Cultural sites with associated ecology: where protection o f a cultural site also protects significant and important biodiversity, such as archaeological/ historical sites that are inextricably linked to a natural area." See draft o f revised Guidelines for applying protected area management categories, IUCN, January 2008, pp. 86, 28). On the topic of "The Importance of Sacred Natural Sites and Cultural Landscapes for Biodiversity Conservation", UNESCO refers to the "biological-cultural diversity found insacred sites", and highlights the following element o f this relationship: "In order to secure and maintain the support o f indigenous and local people inthe conservation of biodiversity, examples o f the traditional conservation o f sacred sites and cultural landscapes need to be increasingly recognized and disseminated as alternative models o f sustainable development, which build upon traditional foundations..." UNESCO, People Biodiversity and Ecology, www.unesco.org. 183 ii. Thesecond category involves the shrines constructed by divinersfor the conducting various rituals. iii. Thethird is burialplaces. iv. Thefourth involves animate objects like strange snakes, leopard, and tortoise. j ,675 604. A companion study on the East bankbased on dozens o f focus group interviews make more than seventy references to local medicinal herbs, several stating "local herbs", which i s evidence o f a herbalists tradition that may warrant an ethno-botanical investigation to understand the importance o f the site to the people's health and healing. Bone sitters and birth attendants are noted as using "local herbs" There are also numerous-references to, individual named trees (e.g. Muvule tree) with special significance, not only simply spiritually but also interms o f their ecological functions (medicinal uses).676 605. As mentioned above, OP 4.04 states that the Bank does not support projects that, in the Bank's opinion, involve the significant conversion or degradation o f critical natural habitats. The Panel notes that this aspect o f the text ("in the Bank's opinion") indicates, inter alia, the need for and importance o f the considered judgment o f the Bank on this crucial question. This phrasing does not imply or give Management a blank check to apply or not certain policy provisions to a specific project but rather requires Management to form and provide expressly an opinion on the issue in question, which must be consistent with the objectives o f the applicable policy. This i s particularly relevant in view o f the controversy surrounding these issues in the present Project. The Panel did not find sufficient documentation that would have permitted Management to make such a consideredjudgment. 606. The Panel finds that the Bujagali Falls area i s a sacred place, like a sacred grove, recognized by the Basoga, a traditional local community, for its high cultural and spiritual significance and inter-related ecological features and values. In this context and for the reasons described above, the Panel finds that the Bujagali Falls area may be regarded as a critical natural habitat for purposes of OP 4.04. The Project entails flooding o fthe Bujagali Falls area. Bank policy regards inundation as a form o f significant conversion or degradation. 607. In light o f the above, the Panel finds that the Project record does not provide sufficient discussion as to why the area was not considered a critical natural habitat. Nor do Project documents explain the Bank's "opinion" that the Project would not 675The River Nile and its Significance to Traditional Religion and Practices of the Inhabitants o f the River Bank inWakise Subcounty. AES Consultant September 18,2000, p. 19,31. `I6 Munene, John. "The River Nile and its Significance to Traditional Religion and Practices o f the Inhabitants o f the East River Bank in Subcounty," p. 15. Commissioned by AES in 1998. Other important natural and ecological features and values o f the Bujagali Falls area are described in Chapters I1and IV o f this Report. 184 involve significant conversion or degradation o f a critical natural habitat. Considering the known spiritual importance of the Project area, without such an explanation, one could also arrive at an opposite conclusion, Le. that the inundation may be regarded as resulting in the significant conversion of a critical natural habitat which would be in violation of OP 4.04. The Panel finds that omitting the reasons behind an opinion of not declaring the Falls a critical natural habitat i s not consistent with the objectives of OP/BP 4.04. The Panel finds that there is an overriding need for the Bank to address these issues in a coherent and well-founded manner to ensure compliance with Bank policies. I.TheCulturalPropertyManagementPlan(CPMP) 608. Throughout the interaction with Project, the Busoga spiritual leaders have acted in a manner consistent with their belief systems as described in non-project associated ethnographic information. Their concerns focused on what i s perceived as possible disharmony with their cultural patrimony and to the spiritual importance o f the Bujagali Falls to the Busoga. 609. It remains uncertain whether or not key stakeholders (consulted and as yet to be consulted) in the spiritual community comprehend the fact that their sacred site will be inundated and inaccessible for their traditional ceremonies. This issue extends well beyond the two spiritual mediums. 610. Management was also on untested grounds by substitutingan abbreviated procedure, not provided for in Bank Policy whereby the new Sponsor would find out what remains to be done from the previous plan, which was assumed to be correct. The previous Sponsor's plan was designed under OPN 11.03, a policy framework that had been replaced by 2006. The Panel could not find evidence that the TOR for the new Sponsor were prepared in consultation with relevant experts and project-affected groups, particularly the local CSO in Jinja that has recognized expertise on the Basoga. Had the project been examined as called for in OP/BP 4.1 1it i s unlikely that the several non-compliance issues highlighted by the Panel would have occurred. The Panel finds that insufficient competence was dedicated to an examination of this issuefor the Appraisal. 611. There are livelihood impacts directly associated with the disruption o f the cultural resources sites that, although initially identified by AES, were subsequently ignored. Contemporary ethnographic accounts and the RCDAP 2001 describe many categories o f traditional practitioners (diviners, interpreters, gourd players, immunizers, exorcists, dispensers, herbalists, caretakers/mediums, bone sitters, and more)677who require payment in money or in-kind for their services, as in any other religion. Within the context o f a traditional society, these transactions are substantial, and they should have been included inthe CPMP as specified inOP 4.11. 677RCDAP2001, p. 105. 185 612. The Panel finds that Management failed to prepare a Cultural Properties Management Plan, assuming that the work of the previous Sponsor was sufficient to meet OP/BP 4.11 guidelines. The overall social management plan (part of the SEAP) does not include planning, resources, or budget supporting Management's response that cultural and spiritual issues will be implemented throughout the life o f the project.678 613. In summary, the Project misidentified the Bujagali Falls spirits as localized, with Project impacts limited to people nearby the Project site. The TOR for the Cultural Properties Management omitted the need for consultation with the approximately 340 Busoga clans' spiritual leaders (baswezz with spiritual ties to the cultural property that was to be affected by the Project.6 The Panel finds that Management is in ;> non-compliance with OP 4.11, by misjudging the size, location, scale as well as the nature and magnitude of the cultural and spiritual significance of Bujagali Falls. The Panel also finds that Management did not consult with key stakeholders throughout the Project cycle and i s therefore in non-compliance with OP 4.11. The Panel also finds that mitigation measures were not adequate becausethe scope of the impact and the consultation process were incomplete. J. Opportunitiesto Address Cultural and Spiritual Issues 614. The Panel observes that there are important opportunities available to address the cultural and spiritual issues within the context ofthe Busoga and the OP/BP 4.11. The Busoga commonly say that "those who are together are like gourds, they cannot avoid hitting each other." They recognize the value o f consultation, "to put an end to disputes, clan members usually hold a meeting and call those who have conflicts together."680And they understand mitigation. The Busoga have many ceremonies to reconcile conflict and establish good relationships between those in conflict - the spirit and clan members.681Harmony is not a permanent status, it comes and goes.682 The Panel's investigation of Busoga culture suggests the cultural problem i s one of restoration o f harmon and developing an appropriate consultation protocol, not simplyappeasement.6 8 Y 615. In the prior Project, Management's cultural resource strategy focused on closure, relocating, or appeasing the spirits, compensating when necessary, documenting 678Management Response, p. 38. 679Whether or not the baswezi have ties to particular spiritual medium is irrelevant to evaluation of the significance o f the cultural property. The Panel notes that the 28 September 2001 ceremony, baswezi present participatedinthe ceremonies, an indication that they share a common belief inthe centrality o f the NabambaBudhagaali spirit consistent with the ethnography reviewed by the Panel. (AESNP). 680Reconciliation among the Basoga. 2001. Culture Research Centre, p. 47. Reconciliation among the Basoga. 2001. Culture ResearchCentre, p. 47. 682Celebrating the Sanctity o f Human Life among the Basoga, Cultural Resource Center, Jinja, Uganda May 2004 Marianum Press Ltd.,pp. 325- 326. 683Celebrating the Sanctity o f Human Life among the Basoga, Cultural Resource Center, Jinja, Uganda May 2004 Marianum Press Ltd.,p. 325. 186 spiritual appeasement through signed certificates, and setting a finite timeline (originally 6 months in2001).684 616. The current Project continued this strategy. Its only remaining cultural property resources commitment was to hold an inter-denominational remembrance service, originally proposed by the first Sponsor, AES, to honor the memories o f those buried inthe islands, as it was impossibleto locate these graves with certainty andtherefore also impossible to exhume and relocate their bodies.685Such a service might prove valuable for some residents in the project area, but does not appear to have been developed through consultations with the Busoga spiritual stakeholders. 617. Similarly there does not exist yet a long-term strategy for sustaining a relationship between believers and the Project, nor have arrangements been negotiated allowing worship at alternative sites in the future. Panel interviews with Basoga cultural experts revealed that an outcome o f a spiritual consultation may be for the spirits to stay in place and permit the project to proceed. The Panel finds that Management has thus far failed to support negotiations that would allow enduring coexistence with spiritual elements of Busoga traditional religion and the Bujagali dam. 618. Construction o f a Bank-supported hydroelectric dam on a sacred site that i s high valued to a large cultural group i s rare. The Panel's expert is aware o f one such project: Aguamilpa dam inMexico, which was financed by the World Bank. The dam was constructed during the early 1990's - before the first Bujagali project was initiated.686 During Appraisal, Bank consultants discovered that the dam would inundate the highly sacred Huichol Indian site o f the water Goddess Macahua at the convergence o f the Santiago and Huayanamota River. Bank and Mexican anthropologists and the Chief Engineer consulted and negotiated with groups of shamans then financed a multi-year movement of the ceremonial site to a new location on the edge o fthe reservoir. The result was a successful mitigation, including the blessing o f the dam by traditional religious leaders. Traditional ceremonies punctuated the dam constructionup to and including its inauguration by the President o f Mexico. Unanticipated at the time, the Huichol were later to view the entire reservoir, which i s now a source o f income (through exclusive control o f navigation and fishing), as sacred.687 684RCDAP2001,p.118. 685HPP-APRAP, p. 23. The original A E S proposal emerged inRCSAP 2001, p. 113. 686 Scott Evan Guggenheim, "Peasants, Planners, and Participation: Resettlement in Mexico. IN Anthropological Approaches to Resettlement: Policy, Practice and Theory. Eds. Michael M. Cernea and Scott E. Guggenheim. 1993. Boulder: Westview Press, pages 201-228, especially pages 221-222. Theodore E. Downing, Appraisal o f the Aguamilpas (Mexico). The World Bank. 2 July 1987. Jason Stanley. Financing: Where funding arrangements meet resettlement inthree Mexican dam projects. Todd MVanden Berg. "We are not compensating rocks: Resettlement and Traditional Religious Systems". InJournal World Development (UK publication), Vol. 27, No. 2 pp 271-283. 1999. Accepted for publication August 17, 1999. Ritual Gestures in Busoga. Busoga Cultural Research Centre, Nile Gardens 5, Jinja, Uganda, Dec 2001. Pages 30-62. Jason Stanley. 2003 October. "Financing Matters: Where funding arrangements meet resettlement inthree Mexican dam projects. RSC working paper. Universityo f Oxford page 3. (D454). 687Project Files, communication dated 9 June 2008, basedon follow-up visit o fMay 2008. 187 Chapter IX Systemic Issues Affecting Policy Compliance 619. The Panel notes that this investigation, like some earlier ones, has revealed certain systemic issues that have affected the Bank's overall compliance with its Operational Policies and Procedures inthe context o f this Project. Some o f these issues, which the Panel believes are important to understanding some o f the key findings inthe present Report, are notedbelow. 620. At the outset, the Panel wishes to reiterate that it considers energy a crucial factor in Uganda's development. The findings o f this Report, and the discussion below, do not dispute this fact. Rather, they highlight what might itself be noted as the first systemic point to be raised by this Report, i.e., that energy production requires considerable care in order to ensure that social, economic and environmental aspects are properly considered, in line with Bank policy, to adhere to sound development practices and avoid situations where costs, including social and environmental costs, outweigh the benefits expected from what are usually sizable investments. A. Legacy Issues from PrecedingProjects 621. This investigation encountered a situation o f adverse effects on people due to a failure to assess, correct and complete resettlement actions initiated in the previous effort to develop the Bujagali dam. Inparticular, many people whose lands were to be flooded or affected by the anticipated reservoir inundation and construction activities were relocated at the time o f the first Bujagali dam project. When the implementation o f this earlier project was halted, following withdrawal o f the sponsor, many of these people were essentially left in limbo, and they did not receive key elements of the resettlement process to which they were entitled under Bank policy (e.g., relating to livelihood and income restoration, community development initiatives). Their continuing problems, and the shortfalls in compliance, are beginning to be addressed only now, several years later, following the present Request for Inspection. 622. Legacy issues from previous finding are found inmanyprojects. The experience with the Bujagali Dam highlights the significant problems that may arise when actions of previous projects are not carried to completion or corrected in accordance with Bank policy. The Panel notes the importance to affected people o f timely actions to address any such situations that mightarise. B. Incorporating Climate Change into Project Design 623. The Panel Report indicates that important studies were done to analyze the question o f climate change, even ifthe most significant o f these were not disclosed as integral part o f the Project documents. The Panel also notes that hydro-electricity, while posing its own set o f social and environmental impacts, has the important 188 comparative benefit o f avoiding the generation o f greenhouse gas emissions produced by some other large-scale alternative sources o f energy - - a point properly noted in Project documents and inthe consideration o fthe proposed Project. 624. At the same time, the Panel discovered that the following conclusion was drawn from the analysis o f climate change, and presented to the Board o f Directors in the key Project Document, the PAD: "L . .] there will be no adverse effect on water release due to climate change during the life of theproposedproject." 625. The Panel is troubled by this conclusion - - it failed to include a risk or uncertainty factor, was inconsistent with the underlying analysis, and appears to provide an overly optimistic reading o f the potential effects o f climate change. The Panel considers that climate change requires a change in mindset towards thinking in probabilistic rather than deterministic terms, recognizing the inherent uncertainty that surrounds climate related issues, and avoiding categorical, deterministic statements. The approach noted above is not in line with the objectives o f Bank policies in support o f informed decision-making. 626. The Panel notes, in this regard, the Bank's increased role in supporting action to address climate change, and its systems-level efforts to ensure that climate change risks are mainstreamed and integrated into Bank's strategic analysis and project decision making. The proper reporting o f risks i s o f central importance in this larger context. C. Timely Disclosureof Informationwithinthe ProjectCycle 627. The Requesters have expressed concern that it was not possible for them to bring the Request at an earlier time because o f the lack o f transparency and disclosure during the discussions of reviving plans for a second round of investment in the Bujagali damproject. 628. This point finds support inthe recordo fdisclosure o fProject documents. Project files show that the Bank was involved in the preparation o f this Project since early 2005. However, the Project Information Document, which i s supposed to be issued early in the Project cycle to provide factual information to the public about a project as it evolves, was not issued until January 30, 2007. The Project appraisal took place shortly thereafter in March 2007, and the Board approved the Project on April 26 o f the same year. While the Panel notes ongoing efforts to streamline procedures, this should notbe at the expense of providingadequateinformationto the public ina timelyway. 629. Related to this, the Requesters have also raised concerns about the implications o f the Project moving forward to such a degree during the investigation of their claims, which they note might result insignificant issues o f non-compliance and harm. 189 630. The Panel observes that these concerns have given the impression to affected people that the Project i s afait accompli, notwithstanding the possibility o f findings o f non- compliance and harm. The Requesters have expressed concern that this could prevent the Project from addressing significant findings in this regard. The Panel notes that this is an important process and systemic issue raised by the present Request, particularly inprojects where it i s alleged that irreversible harm may occur as a result o f Bank's non compliance. D. Transparency Issues and Public-Private Partnerships 631. During its field investigation, the Panel noted considerable concern among Ugandan citizens and a number o f their representatives about the lack o f transparency on the economic impacts o f the project. While realizing the complexity o f this project, and the resulting agreements that were made between private and public partners, it i s o f concern to the Panel that so little is known about the impact o f these agreements not only by the average Ugandan citizen, but also by persons inposition to comprehend the implications o fthe various arrangements made. 632. Given the increase inprivate-public partnerships, and issues relating to access to information in this context, IBRD and IDA might incur reputational risks that are thus far not adequately handled. Similar issues were raised with regard to the prior Bujagali project and other projects reviewed by the Panel in the past. In this regard, the Panel notes the importance of clarifying Bank policy concerning the disclosure of all project-related documents. This i s of particular relevance in public-private partnership projects where some of the documents may be concluded among private parties relying on Bank financial support. 633. Inthe present context, the Panel found that there was anundulyoptimistic assessment o f the costs, benefits and risks o f the Project, including: (i) an under-estimation of capital costs in the PAD; (ii) under-estimation o f the likely impact o f the Project an on tariffs; (iii)a non-recognition o f the likely shortfall inUETCL revenue against the capacity charge up to 2002; and (iv) non-recognition o f some key risks, notably in collection rates and exchange rates. Inall o f these, and especially the third category, Bank Management was substantially dependent on the work o f others. In addition, the Panel found that approach to assessing alternatives to the project was insufficiently transparent, makingit difficult for Bank Management authoritatively to address claims that it was inadequate and biased infavor o f the Project. As it stands, the net benefits of the Project could be substantially less than Bank Management has claimed. E. CriticalNatural Habitats and Sacred Places - - Guidance to Staff 634. As described above, OP.4.04 defines critical natural habitats to include existing and proposed protected areas, "areas initially recognized as protected by traditional local communities (e.g., sacred groves)" and sites that maintain conditions vital for the viability o f these protected areas. Internal guidance to staff for the application o f the 190 Natural Habitats policy, by comparison, describes "critical natural habitats" as "those Natural Habitats which are either legally protected, officially proposed for protection, or unprotected but of known high conservation value." 635. Inpractice, this particular guidance seems to suggest a more limited interpretation and application o f the policy than a plain reading o f its terms would warrant. As a result, areas recognized as sacred and protected by traditional local communities, but considered to be lacking a unique biodiversity and/or official protection, may not have been regarded as "critical natural habitats." As described inthe Panel's Report, the Project provides an illustration o f an overly restrictive application o f the Policy that puts the Bank at risk o f a serious violation o f its policy. 636. The Panel notes that, in contrast to this apparently narrow application o f the Policy, there i s a strong and increasing recognition over the years, for example through the IUCN process, o f the importance of sacred places both for their spiritual and cultural values, and for and as part o f broad conservation objectives, both individually and collectively. . AnIUCN Category I11Protected Area i s an "[alrea containing one, or more, speciJic natural or natural/cultural feature which is of outstanding or unique value because of its inherent rarity, representative or aesthetic qualities or cultural SigniJicance."(emphasis added). The current draft IUCN Guidelines amplify on this element, as described inthe Report. 637. The Panel also notes as well that it addressed these same provisions o f OP 4.04 brought this particular issue to the Board's and the Bank's attention in its recent investigation o f the Cambodia forest project. The Panel Report, in a section entitled "Identification and protection o f critical natural habitats," highlighted that, according to Bank Policy OP 4.04, the status o f critical natural habitats i s also granted to places that are sacred and protected as such by traditional communities. The Report then states: "It is apparent ... thatthere are many spiritforests and spirit trees inforests in their locality which are important to the cultural identih of local people uootnote omitted]. This is particularly the case with indigenous communities. Thus, there critical natural habitats. . . There are also numerous documented cases of spirit are many areas within the general forest estate that need to be considered as forests (critical natural habitats) being logged and destroyed without any consideration of their spiritual or cultural values." (emphasis added) 638. The Panel observes that the Management Response to the Panel's Report related to the Cambodia forest project does not dispute the Panel's finding. 639. The Panel considers that such internal guidance given to staff working in Bank- financed projects involving natural habitats and possibly critical natural habitats, like the current Project, may have sent an inadequate and overly- narrow signal on the application of the Policy. Project stakeholders would benefit from clarification on these matters. 191 Annexes Annex A Table of Findings ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS ENVIRONMENTALISSUES Adequacy of the The proposed Private Power Generation Project has appropriately been classified as Social and (Bujagali) Project is a new operation. There category "A", the category for projects Environmental has been a fresh assessment o f social and with the most serious levelo f impacts. This Assessments environmental aspects of the project, which complies with OP 4.01. has also required drawing upon former studies, where relevant. Environmental The fact that the Environmental Management Plan Management Plan is not an integral part o f the SEA that has been disclosed is a deficiency. This is not in compliance with OP 4.01. Institutional The requirement to support needed Capacity capacity buildirig, which is important inthe implementation o f social and environmental aspects, has not been complied withinthis Project. IndependentPanel As Project is contentious and involves of Experts environmental concerns, appointment o f environmental panel o f international experts is warranted and the lack o f such panel is not incompliance with OP 4.01. Disclosure ofProject The World Bank Group has disclosed the Panel acknowledges that the necessary Documentation project's Economic Study, BEL'SSEA, the studies have beenconducted anddisclosed, NELSAP Strategic/Sectoral Social and albeit independently, and considered by Environmental Assessment (SSEA), and Management and referred to specifically in other environmentalandsocial documents. PAD. However, failure to disclose SSEA or its relevant parts as an integral part o f Project's documentation is not consistent withOP 4.01. Cumulative Impacts The SSEA for the Nile EquatorialLakes Analyses in SSEA do not provide of Bujagali and describes the criteria for assessingthe systematic examination o f potential Existingand Future social and environmental appropriateness consequences o f the Nalubaale and Kiira Hydro Projects of future hydropower developments onthe facilities, the Bujagali Project, and the Cumulative Impacts Nile River inUganda and inthe entire East planned Karuma project all being situated of Transmission Africa region. Section 14 o f the SSEA on the Victoria Nile between Lake Victoria Lines analyzes the cumulative impacts of several and Lake Kyoga. Panel finds that analyses hydropower development alternatives are not sufficiently backed by evidence and under differing scenariosofregional grid include opinions rather than careful fact- integration. It concludes that developing based examinations o f additive effects o f Bujagali and other sites inthe Victoria Nile impacts from present and foreseeable Basin (excluding Kalagala) will not have projects. Panel finds that neither SSEA nor significant cumulative environmental SEA have addressed cumulative effects of impacts. BEL'SSEA examines cumulative existing andplannedprojects inmeaningful impacts o fBujagali, the hydropower plants way. This is not in compliance with OP at Nalubaale, Kiira and Karuma along with 4.01. 192 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS the transmission facilities therewith on the Victoria Nile inUganda. Panel finds that the failure to consider mitigation measures, which would reduce social and environmental impacts o f the transmission line, does not comply with OP 4.01 and OP 4.12. Environmental Building on relevant work conducted to Based on its review of relevant research Impacts on Fisheries date, BEL's consultants conducted further studies, Panel observes that the status o f and Aquatic Systems field studies and analyses where the need fish species inhabiting both Lake Victoria for updated information had been and Victoria Nile is disputed and that identified, such as water quality, fisheries, ongoing research is desirable. However, terrestrial ecology, resettlement and significant effort has been devoted to study compensation, and cultural resources. The these fish in the reaches of the Victoria reach o f the Victoria Nile that will be Nile that will be affected by the Bujagali affected by Bujagali is not considered to be Hydropower Project. critical habitat for any fish species of conservation importance. Panel finds that Management acted consistently with OP 4.01 and OP 4.04 as these relate to assessment of likely consequences of Project on fish stocks in the Upper Victoria Nile andLake Victoria. Kalagala Offset GoUhas agreedto r e c o n f i i its Panel finds that there is evidence that an Agreement commitment to the Kalagala offset that it offset has been created, to meet OP 4.04, made under the previous effort to develop but there is no evidence o f the offset site the Bujagali project. This offset being subject to appropriate conservation commitment is consistent with the andmitigation measuresinconformity with mitigation provision for Kalagala Falls, and sound social and environmental standards. also recommendedinBEL's SEA Report. Project is thus not in compliance with OP The offset provision for Kalagala Falls and 4.04. Panel fmds that the Kalagala offset the adjacent natural habitat will be included may not achieve the purpose for which it as a GoUobligation inthe IDAIndemnity was set aside, and this i s not consistent Agreement for the Bujagali project. with the provisions o f OP 4.04. Panel notes with concern that proposed Environmental Mitigation and Monitoring Plan is silent on the need for monitoring o f enhancement and offset plantings. Monitoring o f replacement plantings has not been included in the terms o f reference o f the witness NGO appointed to monitor Project compliance with IDA conditionalities. This is not consistent with OP 4.04. Safety o f Dams A Dam Safety Panel @SP) has been Panel finds that Management has complied established, which includes two o f the three withthe procedures set forth inOP 4.37. members o fthe previous panel set up under the earlier effort to develop the Bujagali project. [. ..] Management considers the current project in compliance with the OP (OP 4.37). HYDROLOGICAL AND CLIMATE CHANGERISKS Appropriateness o f The hydrology o fthe Victoria Nile is Panel's hydrology expert has concluded Hydrological Data complex due to meteorologicalinfluences, that hydrologic data sets used in Project 193 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS Series used in the rainfall-moff process, the scale o fthe design constitute a reliable data series and Project Design evaporation losses, andthe interaction its variability over time is a natural between rainfalland evaporation withinthe condition, which can be observed in other watershed. The available reservoir inflow hydrologic series o f different parts o f the recordcomprises 106 years o fdata. It world, when hydrologic series is long includes several significant hydrological enough. Panel finds that this provides an cycles, among which the seasonaland ten appropriate baseline for analysis o f year cycles are the most apparent. Given environmental and economic issues, in the lengtho fthe hydrological recordat this compliance with OP 4.01. site andstudies onclimate impacts, the hydrological risk for energy generation is considered to be defhble from the available data set. Impact of The Economic Study addresses the There seems to be a discrepancy in Project Hydrologic Risk on economic viability and risk analysis o fthe documents: PAD and Economic Study Energy Output Bujagali project. The key elements differ as to which water release regime will assessedinthe economic analysis include be in effect once Bujagali becomes [. .]the hydrology ofLake Victoria andits . operational, the "Agreed Curve" or the impact on hydropower generation. [...] "Constant Release" rule. This discrepancy Risks arising from varying degrees o f brings into question the data basis for future uncertainty regardingthese variables Project's economic analyses, and is likely have also been evaluated. to have resulted in a more positive conclusion to the Economic Study than would have been the case under the Agreed Curve scenario. This is inconsistent with OP 10.04, the provisions o f which require Management to provide an accurate picture o f the Economic Study (based on the Agreed Curve), and indicate whether this affects relevant conclusions. Panel notes that this contradiction inProject documents has a material implication not only for economic viability o f Project and provisions o f OP 10.04, but also on lake levels o f Lake Victoria, since different operational rules result in different time- profiles and variance o f water levels. Potential Impact of Withjoint operation o fthe existing Panel notes importance o f assessing the Projecton Lake hydropower andthe proposed project, changes in operating regimes and Victoria generation o fthe same energy output as extending area o f influence o f the Project currently generatedbyNalubaaleandKiira to Lake Victoria. Panel finds that SEA would only require 45% o fthe current analysis did not comply with OP 4.01 in water releasefrom Lake Victoria. defining the area o f influence o f the Project Management acknowledges that BELwill because Project impacts on the changing not control the release o fwater from Lake levels o f Lake Victoria were not assessed. Victoria, but is o fthe view that it is inthe Panel notes the importance o f making the interest o fthe GoU to ensurethat Bujagali structure for governance of water releases andthe NalubaaleKiira dams are operated from Lake Victoria clear and transparent to efficiently. all stakeholders. 194 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS Climate Change The broader climate change (and Panel finds that the possible effect o f Risks hydrology) aspects were addressedin climate change on hydropower projects on different studies which have also been the Victoria Nile has been seriously publicly disclosed. The SSEA analyzed in considered in the SSEA. This is in detail the impacts o f climate change on compliance with OP 4.0 1. Management power development options inthe Nile does not appear to have ensured that Equatorial Region, including Bujagali. Economic Study drew on the much more thorough analysis in SSEA. Panel finds that this is not compliant with OP 10.04. Panel is aware o f the limitation o f known technology in evaluating climate change scenarios and that the analysis o f climate change is an evolving science, where gaps remain. Indeed, this situation makes all the more troubling the PAD'S categorical assertion, without any reference to risk and uncertainty, that there will be no adverse effect on water release due to climate change duringProject life. This failure to express climate change as a risk factor is not consistent with OP 10.04. Panel notes the importance o f continued attention and analysis to the effect o f -climate change on flows and hydropower generation on the Victoria Nile. ECONOMICAND ENWRONMENT& ANXLYSIS OFALTERNATIVES Economic Analysis Management considers that economic, The terms o f reference for the Economic of Alternatives financial, safeguard, technical, governance, Study call for comprehensive update o f and other required analyses to date are earlier work. compliant with relevant Bankpolicies. Alternatives A detailed review o f geothermal prospects Panel notes the statement in Management Considered: was conducted as part o f the project Response that additional studies and analysis o f alternatives. The analysis shallow drilling are included under the Geothermal concludes that historical estimates o f the ongoing Power IV Project, to assist GoU in Potential geothermal potential o f Uganda being as assessing geothermal prospects at several much as 450MW are substantially over- sites in Western Uganda. Additional stated. The true potential is likely to be in information resulting fiom this work would the order o f only 10% o f this figure. [...Ihelp resolve conflicting views regarding These findings led to the inclusion o f a geothermal potential in Uganda, and may 40MW geothermal power plant, to be have significant bearing on economic commissioned in mid-2011, in the least- analysis o f alternatives. cost analysis. 0 SmallandMedium The Bank is providingconsiderable support Panel notes that information io Economic ScaleAlternatives to Uganda in development of hydropower Study q d PAD relating to knowledge potential, This includes large-scale hydro about and potentid o f smaller scale and/or 195 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS and the ongoing ERT Project, which is distributed generation alternatives did not supporting mini-hydro development for clearly establish that available studies and grid-connected and off-grid applications. data had been identified and evaluated to [...IThe Bujagali Economic Study decide whether M e r consideration was included all hydro projects that are either required. Panel finds that Economic Study currently providing power to grid, or and PAD did not demonstrate full suitable for grid connection and which are compliance with OP 10.04 requirement to actively under development and thus evaluate alternatives. suitable for consideration in planning timeframe. OilResources While oil resource discovery was at a very early and unproven stage when the Economic Study Final report was completed, Panel finds that the existence and potential o f this resource should have beenreviewed inthe discussion o f alternate supplyoptions. Project Costs The World BankGroup andother lenders have taken several stepsto ensurethat costs Panel frnds that, although certain parts o f o fBujagalireflect current market the analysis were carried out thoroughly, to conditions. BELconducted its procurement meet all requirements o f OP 10.04, PAD o fthe EPC contractor under the supervision should have included explanation and o fthe EJB. supporting evidence o f why all parties had concluded that substantial project cost variations would not alter conclusions o f the Economic Study. Panel observes that the foregoing analysis does not suggest that the updating o f the EPC cost figures inthe PAD does not obviously disadvantage Kanunarelative to Bujagali. Assessment of Least Processo f testing the sensitivity o f the least Cost Options for cost expansion plans with and without Expanding Power Bujagali appears to have been carried out Generation thoroughly. The assumed increase o f 10 percent for the "high Bujagali capital cost scenario" compared with the "base scenario", with an assigned probability o f only 20 percent, was inappropriately low. Nevertheless, a sensitivity test suggested that the Economic Study's conclusions that Bujagali was the least-cost option were robust for an increase o f almost 50 percent incapitalcosts. Panel finds tha~inorder to comply with the requirements o f OP 10.04, the PAD should have qualified its statement about the projected drop in tariffs to take into account the impact of EPC and transmission cost increases. Panel considers that the relationship between estimates in Economic Study and PAD'S financial analysis should have been 196 ISSUE MANAGEMENT RESPONSE PANEL'S FINDINGS presented more clearly and transparently in PAD. Externalities Panel finds that the limited presentation and discussion o f these costs inEconomic Study didnot succeedindemonstrating fill compliance with OP 10.04. In Panel's view, to meet allrequirements of OP 10.04, Economic Study should have examined, in more detail, the potential of changes in damage from other pollutants than COz, even if it might have proved difficult to value them. Hydro-power Management is fmly convinced ofthe Panel finds that Management did not Location appropriatenessandbreadtho f analysis ensure that cultural and spiritual matters Alternatives within undertaken to identifyand assess were properly considered when comparing Uganda alternatives for expansion o f Uganda's the Bujagali and Karuma alternatives, as power sector. The economic analyses required by OP 4.01. This is especially considered options that hadrealistic relevant in light o f the significant cultural potential for availability ina timeframe and spiritual importance of Bujagali Falls similar to the Bujagali project, and which, to the Busoga people. Lack o f proper therefore, could be considered as consideration o f cultural and spiritual alternatives. matters in this comparison had important consequences, inthat it appears to have led to the conclusion that there was little difference between the Bujagali and Karuma sites and that therefore economic and financial aspects o f the options should become the determining factor in selecting the preferredoption. Alternative Project Panel notes that a range o f alternatives Configurations at have been considered in these studies. Bujagali Panel is concerned, however, that analysis unduly narrowed consideration of alternatives on the basis of a-priori judgments rather than exploring all technically feasible options, including those that would not involve flooding Bujagali Falls and thus have lower social and environmental costs, and laying them out in a systematic way along with their economic, social and environmental benefits and costs, so that judgments on optimal alternatives could be made with full understanding of trade-offs involved. This is not consistent with OP 4.01's provisions that feasible alternatives should be explored systematically to meet basic Project objectives, and may have led to inadequate consideration o f alternatives that met Project objectives while avoiding social and environmental costs associated 197 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS with flooding BujagaliFalls. ECONOMICEVALUATION: POVERTY REDUCTIONAND RXSK Affordability and Management states that Project will allow Economic Study provides quantitative Poverty Reduction industrial and commercial users to increase assessments o f both costs and benefits, their output and efficiency, and therefore which suggest that Project would have their profits, thereby enhancing economic largely positive direct impactson Uganda's growth. [...I These developments are economy and enhance national economic expected to have positive impacts on activity. Inthis sense, and bearing inmind poverty alleviation in Uganda, directly reservations about the cost estimates of the through the availability of power to newly Economic Study, fiom a macroeconomic connected households and indirectly perspective, analysis appears to have through employment creation. [. .] . complied with the requirement inOP 1.00 According to the Economic Study, to show that Project is likely to contribute Bujagali's commissioning in 2011 would to "broad based growth." In terms of enablethe cost ofpower to end-users to fall affordability o f electricity generated under to US$l6kWhin2006 money. This would the Project, Panel notes that the improve the affordability of power to end US#l6/kWh figure provided in Economic users. Study is likely to be an underestimate of the cost of electricity with Project. Further, ManagementResponsedoes not discuss the different EPC cost estimates cited in the Economic Study and the PAD or make clear their implications for the tariff estimates. Panel did not find evidence in Economic Study or PAD of any estimates o f the economic impact o f Project on low- income households. Panel considers that such analysis, in addition to the broader macroeconomic analysis undertaken in Economic Study, should have been made during appraisal to provide a better understanding o f whether the objective of poverty reduction envisaged by OP 1.00 would beachieved. RevenueProjections Panel notes that PAD'Sprojection the GoU and the Institutional support needed to power utilities over Framework period 2005-2016 appears misleading and seriously at odds with the projected revenue stream of Project. Panel notes that the likely tariff variations and possible revenue shortfalls or surpluses and their implications for UETCL, UMEME, and government net revenues are key sustainability concerns. Panelnotes that the revenuegap that UETCL, inparticular, will face, may lead to large, urgent demands on GoU Treasury and potentially on the Bank via its Guarantee. Infrastructure In light of the scale of revenue Funds requirements, financial risks accepted by UETCL and GoU, and the scale of 198 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS subsidies and guarantees involved in Bujagali, Panel notes that Management should have explored further ways o f managing and addressing financial and governance risks, inthe interests o f project sustainability inaccordancewith OP 10.04. Power Purchase Panel finds that for the Sponsor and its Agreement lenders, the terms and conditions o f the 2005 PPA, especially those set forth in Annex D, seem to represent a low-risk (though potentially disputatious) means o f managing and recovering costs which are, by definition, subject to uncertainty. For UETCL, the power purchaser and its guarantors, by comparison, it means that there is no ceiling on payments on capital costs and whether or not Project delivers the direct economic benefits offered over 30 years, interms o f costs and tariffs, is to a significant extent, outside their hands. Distributionof Risks Panel observes that the high allocation o f risk to UETCL and eventually GoU increases the possibility that Project may not achieve the broad objective o f sustainable development and poverty reduction embodied in Bank Operational Policies andProcedures. Panel is concerned that any additional GoUresources spent in the financing o f the development and operation o f Project may lead to decreased resources available for social and other priority development programs. SOCIALISSUES- INVOLUNTARY RESETTLEMENT Assessment and Panel found no formal monitoring or Action Plan evaluation report supporting the assertion that involuntary resettlement was "largely completed," the reason stated for forgoing full RAP preparation, as required by OP 4.12. Panel fmds that the hydropower APRAP failed to assess and update the previous 2001 RAP and provide additional new information as required to complete the RAP requirements to current standards. This does not comply with OP/BP 4.12. This led to Action Plans that did not meet the policy objectives andrequirements. Baseline Socio- Management considers that BEL has Panel notes that the survey conducted by Economic Data carried out social and environmental BEL cannot be considered a census of evaluations and documentation that are in economic or social conditions as defined in full compliance withWorldBankspolicies. OP 4.12. In this sense, Management's 199 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS claim that the Project took the first Panel's report findings into account in preparation o f the current Project is not accurate because significant weaknesses in the process o f gathering baseline data information were similarly identified inthe 2002 Panel Investigation Report. Panel also finds that the approach to consultations with people who had moved and had been compensated i s not consistent with involuntary resettlement policy. Livelihood Panel observes that effects of the original Restoration displacement andof the ensuing delay have not been fully reflected in the A P W . Overall, Panel finds Project in non- compliance with the mandate of Bank Policy on Involuntary Resettlement to improve or at least to restore, inreal terms, the livelihoods and standards of living of people displaced bythe Project. Method to Assess In Panel's view the methodology used to Livelihood assess livelihood restoration in the context Restoration and o f Project, while suggestive o f issues, Address Project cannot substitute for an economic analysis Delay o f livelihood risks and restoration. Panel also finds that Management did not assess and include into the APR4.P a methodology for restitution o f unintended socio-economic costs incurred by displaced persons resulting fkom project stoppage/delay. This is not consistent with OP 4.12. Realo r perceived Panel notes that lack of clear unfulfded promises communication with affected people to intheprior Bujagali address concerns o f displaced persons with Project regards to the commitments made by AESNP, risks leaving the project with contentious, unresolved issues. Specific Livelihood Panel finds that Project failed to provide Risks: Fishing and adequately for loss o f livelihood associated Agriculture with loss o f fishing and agriculture, in non compliance with OP 4.12. Compensation Management notes that people who will be Panel concurs with the A P W ' s findings, affected by the transmission l i n e p a r t o f which validate the claims of the PAPSthat the project's associatedInterconnection full replacement value compensation may Project that is expected to be financed by have not taken place inthe prior project. the Afi-ican Development Bank(AfDBF must be compensatedandresettled satisfactorily. 200 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS Land Titles Panel finds that APRAP conclusion related to the necessity of issuing land titles to people resettled under prior project i s consistent with OP 4.12. Panel notes however that there seems to be no agreed timetable for issuanceo f these titles. Vulnerable Peoples The APRAP determined thatpast Panel notes that the absence o f focus on resettlement didnot provide for vulnerable livelihood risks to the vulnerable is evident people andhasrecommendedactions to in that none of the proposed assistance ensure that these people's needsare measures addresses vulnerable addressedgoing forward. tenants/sharecroppers or children. Additionally, proposedassistancemeasures do not addressthe question of sustainability beyond limited Project support. Panel finds Project out o f compliance with vulnerable peoples provisions o f OP 4.12. Housing and During its field visit, Panel verified that the Electricity for standard of living o f displaced households AffectedPeople who resettled inNaminya and Nansana has improved with respect to housing. On the other hand, A P W discovered some shortcomings in housing condition and Panel observed physical problems and deterioration with some houses and structures. Panel is concerned that no physical action is planned with regard to houses at the resettlement site. Panel also notes that, given the context and previous expectations o f affected people, the broad statement made by AES regarding electricity provision may have reasonably been interpreted as a promise to deliver electricity connections to affected households. Panel notes that this is an outstanding controversy o f high importance to affected communities. Investment Panel's review o f the limited scope of Resources for livelihood restoration program indicates Livelihood that they may be under-budgeted. As Restoration livelihood restoration instruments develop, Bank policy provides that Management is to monitor resettlement budget to ensure sufficient resources. Sharing inProject According to the APRAP, US$497,000 will Panel finds that with limitedhnding,broad Benefits and be neededto finance the programs to criteria for eligibility and lack o f Community complete resettlement and income specificity, CDAP programs do not assure Development restoration. Bujagali Energy Limited compliance with OP 4.12 (BEL),the project developer, is committed to providing US$2.4 millionfor community development over a five-year period 201 ISSUE MANAGEMENT RESPONSE PANEL'S FINDINGS following the start of construction. IndigenousPeoples Managementconsidersthat a clear Panel did not find any evidence that demarcationline existsbetweenthe Basoga Management violated provisions of Bank andethnic groupsinotherAfrican policy on Indigenous Peoples, with regard countriesthat the Bankhas defined as to the Basogapeople. indigenous. TheBasogaare a large and influentialgroup withinUganda. Consideringthe Basogaandall other Ugandangroupsas indigenouspeoples would defeat the intendedobjectiveso fOP 4.10. CULTURALAND SPIRITUAL VALUES PhysicalCultural There havebeenextensiveconsultationson Panel finds that Management failed Resources various social aspects o f the project, adequately to consider or implement including spiritualandcultural issues. alternativesto avoid project-relatedimpacts on Busoga spiritualityand culture. Most of those who believe inthe significance of the Bujagali Falls spiritual site do not live in the immediate vicinity of the Project. Project also failed adequately to consult with Busoga spiritual clan leaders associated with one or more high status Spirits about significant cultural patrimony ofBujagaliFalls. Misidentifying Bujagali Falls as a local cultural resource, misaligning its consultation strategy, and failing to prepare a new Cultural Property Management Plan compounded errors and muddled mitigation. Resultant problems included loss o f objectivity of the Sponsor, impatience, assignment o f pecuniary motives to stakeholders, cost cutting, culturally inappropriate mitigation efforts, and most importantly, a misunderstanding that the Bujagali Project is ensconced ina long-term relationship with its new neighborsandtheir spirit world. Management unnecessarily and inappropriately took sides in a spiritual controversy of a religion inwhich millions of Ugandans believe. The Panel finds this actionbyManagementto be non-compliant withthe OP 4.11. The Panel finds that Management assumed that what they called the "Bujagali spirits'' were restricted to the Project construction and flooding area, in contravention to the BP 4.11 requirement that they work with 202 ISSUE MANAGEMENTRESPONSE PANEL'S FINDINGS and assist the Borrower to identify the spatial and temporal boundaries o f the cultural resources affected by the project. This did not comply with avoidance and mitigationrequirements o f OP/BP 4.1 1. Panel finds that the culturally and spiritually affected people were not adequately identified as required by Bank policy. Critical Natural Project documents indicate that the Project Panel finds that the Bujagali Falls area may Habitats is not significantly converting or degrading be regarded as a critical natural habitat for a "critical natural habitat" as defined in purposes o f OP 4.04. OP 4.04. The Panel finds that the Project record does not provide sufficient discussion as to why the area was not considered a critical natural habitat. Nor do Project documents explain the Bank's "opinion" that the Project would not involve significant conversion or degradation o f a critical natural habitat. Considering the known spiritual importance o f the Project area, without such an explanation, one could also arrive at an opposite conclusion, i.e. that the inundation may be regarded as resulting in the significant conversion of a critical natural habitat which would be in violation o f OP 4.04. The Panel finds that omitting the reasons behind an opinion of not declaring the Falls a critical natural habitat is not consistent with the objectives o f OPBP 4.04. The Panel finds that there is an overriding need for the Bank to address these issues ina coherent and well-founded manner to ensure compliance with Bank policies. CulturalProperty The management o f cultural and spiritual Panel fmds that Management failed to Management Plan issues is part o f the overall social prepare a Cultural Properties Management management plan (part o f the SEN), Plan, assuming that work o f previous which will be implemented throughout the Sponsor was sufficient to meet OP/BP 4.11 life o f the project. Implementation will be guidelines. Panel finds that Management is monitoredsupervised by the World Bank in non-compliance with OP 4.11, by Group throughout the loadcontract misjudging the size, location, scale as well periods. A Ugandan NGO, "Interaid," was as the nature and magnitude o f cultural and contracted to carry out independent spiritual significance o f Bujagali Falls. monitoring during AES implementation o f Panel fmds that Management did not ifs RAP. BEL has committed to consult with key stakeholders throughout' 203 ISSUE MANAGEMENTRESPONSE PANEL'SFINDINGS independent monitoring, also through Project cycle and is, therefore, in non- Interaid, o f all aspects o f the project, compliance with OP 4.11. Panel finds that including those related to cultural heritage. mitigation measures were not adequate because the scope o f the impact and the consultationprocess were incomplete. 204 Annex B IndependentReviewof ContractualArrangements INDEPENDENT REVIEW OF CONTRACTUAL ARRANGEMENTS Graham Hadley, Economic and Commercial Consultant, UK 1.SCOPEOFREVIEW 1.1 This review was carried out over the period July 2007 - January 2008 at the request o f the World Bank Inspection Panel The main documents reviewed were the Power Purchase and Implementation Agreements o f December 2005 and o f December 2007, as amended; the Report and Recommendations o f the InspectionPanel o f May 2007, which included the Request for Inspection (March 1 2007) and the subsequent Management Response; Burnside's SEA (Executive Summary) o f December 2006; Power Planning Associates' Economic and Financial Evaluation Study o f February 2007; The World Bank's Project Appraisal Document, April 2007; Linklaters' Preliminary Review of Basic Contractual Documents, March 2006; the Siemens Reports on the Bujagali Transmission Interconnectiono f July and August 2006; the Afkican Development Fund's Appraisal Report on the Bujagali Interconnection Project (BIP), February 2007; AfDF's Loan Agreement for the BIP o f October 2007; and a letter o f approval o f May 2007 from JBIC setting out the terms o ftheir loan to the BIP. Because my terms o f reference request, inter alia, a comparative analysis with the prior Bujagali project, Ihave also revisited the Power Purchase Agreement (PPA) for prior Bujagali project and associated documents o f December 1999, and my previous report to the Inspection Panel: "Independent Review o f the Executed Agreements, gth February 2002". Inaddition Iowe thanks to World Bank and Afkican Development Bank staff who have provided working papers and otherwise assisted me in addressing particular questions Ihave raised with them. 1.2 My report i s subject to some important qualifications, mainly arising from the limited time available: -- IWhereas unable to study the full project documentation. havebeen inmy2002 report Iwas able to cite internationalbenchmarks against which to assess the capital costs, Ihave not been able to do the same for this Project. My comments on the cost increases rest only on general observation o frecent cost trends - The recent physical and documentary developments inthe powerbusiness. - including financial close in December - may have economic or financial consequences which Ihave not been able fully to take into account. For these reasons any criticisms expressed should be regarded as provisional, and, pursuant to my terms o f reference, Ihave offered observations and suggestions rather thanrecommendations. 205 1.3 My terms o f reference include a comparative analysis o f the contractual framework for this Project as compared with the prior Bujagali project, focusing inter alia on costs, risks and risk-sharing, financing, tariffs and currency issues. The report concentrates on those aspects- costs andtheir implications for tariffs, and risks andhow these are shared -whichIbelieveshouldbeofmostinteresttotheInspectionPanelandtheWorldBank Group - which also represent the most significant changes from the earlier project. 2. EXECUTIVE SUMMARY BackgroundandDevelopmentssince the prior Bujagaliproject 2.1 Recent increase in retail tariffs and unit generation costs make it intrinsically more likely that the Present Project will allow tariff reductions, than was the case for prior Bujagali project (3.2) 2.2 The strategic case remains strong: optimum use o f Nile waters and provision o f a major increment to generating capacity, to meet growingdemand from both existing and newly connected customers (3.3) 2.3 Advantages in risk mitigation have been included, as compared with the prior Bujagali project: (a) World Bank Group links with one o f the equity partners (Industrial Promotion Services, Kenya); (b) NewProject-related studies, notably onhydrology; (c) Government-backed scheme for resettlement; (d) a "safety net" allowing public sector buy-backinthe event o fprolonged very low hydrology(4.1 and 8) Costs, Revenues andRisks for BHPandBIP 2.7 The increased EPC cost (at the time o f the PAD) of the present Bujagali Project, as compared with the prior project does not o f itself suggest an excessive price, given world power plant cost trends and perceived risks in Uganda. However, there appears to have been a significant increase over the bid price inthe past year, reflected in the December 2007 contract price o fUS$564.4m (5.2-5.4) 2.8 The Economic Study - the basis o f the PAD'S appraisal - has taken the lowest available cost estimates for both the BHP and the BIP. As a consequence, it i s possible that the comparison o f generation options was unduly favourable to Bujagali; and the likely tariff impact was too optimistic (5.5-5.8) 2.9 In addition, the PAD may underestimate the risk of firther EPC cost increases, especially in the light o f the December 2007 amendment to the PPA, and o f a share o f these passing through into the BHP capacity charge. Compared with the prior Bujagali project, the power purchaser bears a greater share o f financial risk, mainly through the absenceo f a stipulated maximum capacity charge inthe PPA (5.9-5.10) 206 2.10 Further, forecast sales revenues and the rate o f recovery o f full supply costs from customers ("recovery rates") may fall short o f forecasts, increasing the risk that UETCL will suffer a revenue shortfall against the BHP PPA requirements in the period 2011-23, triggeringthe GoU guarantee (5.8 and 6.5) 2.11 Financial and economic risks to the Project, ifrealised, will result inhigher tariffs or higher subsidies or a combination o fboth. Such risks are, principally: (a) Cost escalation; (b) Currencydepreciation; (c) Prolongedlow hydrology; (d) Lower demand growth; (e) Lower or static collection rates; ( f ) Affordability; 2.12 Lesser risks include: (a) Construction delay; (b) Withdrawal o fthe developer/operator; (c) Poor plant performance (8) Overview 2.13 The strong strategic role o f the Project has been recognised in the increased and wider involvement o fpublic authorities, givingthe Projectrobustness (9.1) 2.14 However, the direct economic benefits o f the BHP and BIP projects may have been over-estimated. While BHS /BIP may still, bearing inmindrecent upward movements in hydrocarbonprices, be the lowest-cost option for generation, it may cause upward rather than downward pressure on retail tariffs (9.2) 2.15 The balance o f interest between customers and UETCL (and its guarantors) will be determined by the price selected for the levelised tariff once the plant commences operation. It may be prudent to set this conservatively, initially, to minimise the risk o f the PPA guarantee being called, and to allow the prospect o f subsequent tariff reductions once debt repayment i s completed. (9.3) 2.16 It may be helpful to review the Project, as a leveraged independent power project (IPP), experience before deciding what combination o f public and private resources to use indeveloping the Karuma project, if as expected it follows Bujagali (9.5) 3. BACKGROUND 3.1 There have been some significant developments in the Ugandan electricity sector since the prior Bujagali project: continuing demand growth; the acquisition o f new high- cost stop-gap thermal generation; big tariff increases; part-privatisation o f Distribution; 207 and increased dependency o f UETCL on Government funds. These changes also change the appraisal o f the present Project. Some key conditions, however, remain unchanged, notably the fact that only about 5% o f the population i s connected to an electricity supply; and only about half the cost of electricity units sent out from power stations is actually recovered from customers. 3.2 The increases ingeneration costs per unit and inretail tariffs make Bujagali relatively more attractive than it was in2001, inthe sense that there i s now a better prospect that it will produce intramarginal retail tariffs, rather than (as for the prior Bujagali project) pushthem up.Affordability may therefore be less o f an issue, though that problem could return in the event o f low demand growth; rising technical/commercial losses; or depreciation o fthe Uganda Shilling(USh). 3.3 On the one hand, from a financial point o f view, the strategic case for Bujagali remains strong and unchanged: it would optimise productive use o f Nile waters, a major Ugandan natural resource, without increasing the draw from Lake Victoria; it would provide a big extension to the generating capacity o f the central grid system to cope with growing demand (especially from business); and by the involvement o f private companies it would attract (directly and indirectly) both expertise and inward investment. 3.4 (On the other hand strategically, an opposing view might be that Bujagali increases dependence on the Nile waters, i.e. it reduces diversity o f supply, compared with other generation options; it pre-empts use o f public financial resources; an alternative strategy based on or including dispersed generation through smaller units could more rapidly bring supplies to the un-connected majority o f the population, whilst reducing foreign currency dependency) 3.5 In fact, the Project sponsors claim more than strategic benefits for the present Bujagali Project - they say that it i s the least-cost option for new generation, and that it should allow tariff reductions when operational. These conclusions rest on a comparative economic analysis o f costs and risks o f options for new generation, and a financial review. My report examines these, and also highlights the costs and risk-sharing explicit or implicit inthe contract documents, comparing these with those for the prior project. 4. PROJECTCHANGES 4.1 Physically and in its electrical impact, the present Project and its associated transmission project closely resembles the prior Bujagali project. The Project vehicle - a leveraged independent power project (IPP), building and operating the plant and selling bulk power to the public utility under a long term contract (PPA), with Government and International Financial Institutions supporting both the loan finance and the PPA - is also conceptually the same. Although there are some changes inthe loan and guarantee structures, the key contract documents (PPA andImplementationAgreement (IA)) are also similar, even identical, in many respects. Some of the changes most relevant for cost and risk are: 208 - the award of the Project to the developer was by competitive process, not single- track; - the World Bank Group has important links, independent of the Project, with one of the equity partners; - although this investigation report highlights many of their shortcomings, in general terms it may be said that more wide-ranging studies have been made (and made available in Uganda) of the economic, environmental and social aspects of the project, in particular, hydrological risk; and plans backed by the public authorities have been developed with the purpose of providing long-term solutions to social problems; and - increased provision has been made for the public electricity supply system to buy- back the project inparticular, low hydrology, circumstances. 4.2 All o f these represent potential improvements - reduction o f risk - for the prior Bujagali project as compared with the present Project, as explained in section 8 below. (At least one o fthese conforms with findings made by the Inspection Panel inrespect of the prior project) However, there are two other significant changes whose effect i s likely to be adverse, for the power purchaser and his guarantors: - capital costs and total costs for the power plant have increased significantly in real terms; and - determination of the capacity charge payable under the PPA is by application of a cost formula, rather than by referenceto a stipulated maximum charge. 5. COSTS ANALYSIS - 5.1 This section examines costs, o f both BHP alone and the combined BHP/BIP project; how they have been treated in the financial and economic analyses; and the possible cost outcomes relative to tariffs and revenues. In the following section these points are distilled into conclusions, or summarised observations. 5.2 Increased Capital Cost of the Power Plant Para 55 o f the PAD refers to the 62% increase in cost (absolutely and per KW) between the "hard"(EPC) costs o f the prior project and the present Bujagali Project. This equates to an increase o f 10% p.a. cumulative over 5 years - a significant increase in real terms. However, the explanation inPAD para 55 -a mixture o finternational trends and factors specific to Uganda-may be correct. Power plant costs have certainly increased in real terms internationally, and although the index o f this for hydro plant may be less than for thermal plant because o f the higher proportion o f civil engineering costs inthe former, the other more local factors referredto inthe PAD may more than offset this 209 5.3 Competitive Solicitation The outcome o f competition, for the Project concession and separately, the EPC contract, throws some interesting light on the price. The project concession was awarded to BEL from a thin field o f compliant bidders. The winning bid for the EPC contract was 43% below the next lowest bid (PAD para 54). This suggests potential bidders (inboth competitions) were taking a risk-averse stance, either abstaining or loading their bid prices with a significant risk premium. The latter may also apply to the winning EPC bidder, despite his price advantage. None o f this demonstrates that the agreed price i s higher than a "realistic market price" assessed by any other means; what it does indicate i s that the market price for an IPP project of this magnitude in Uganda i s high. 5.4 Total Cost Increases Leaving aside financing costs, the BHP EPC costs have increased since early 2007, as follows: BHP$m Economic Study Feb 2007 (estimated) 441(1) PAD April 2007 (estimated) 520 December 2007 - EPC price 564.4 Note 1: See para 6.6 below for the derivation o fthis figure. Inthe caseofthe BIP, different costshavebeengiven indifferent contexts, as follows: - BIP$m Siemens Report July/Aug 2006 C. 80-95 Economic Study Feb 2007 28 PADApril 2007 55 AAR Feb 2007 74.7 The last o f these figures appears the most authoritative at the time, though Iunderstand that current prices being discussed are lower. Ido not have enough information to be able to judge the degree o f comparability o f these figures. 5.5 Two points may be drawn from this picture. The first i s the propensity o f EPC costs to increase between selection o f a winningbidder and fixing o f the price. Inthis case, the cost appears to have increased by $123m (28%) from the Economic Study estimate to the point where the contract price was fixed, and fbrther increases seem to be allowed by the PPA. It i s clearly preferable if possible to treat the bid price as binding, which would be international best practice (IBP); the PAD does not appear to explain why that was not done here. In allowing "single-track" negotiation after the competition has closed, any benefit from competitive pressuresmay be lost. Second, it i s the lowest numbers, for both the BHP and BIP, which were used in the Economic Study, which appears to be the only economic appraisal addressing the total Project. The PAD relies heavily on this study inconfirming the judgement that this is the lowest cost option for generation and should enable retail tariffs to be reduced. The PAD adds a financial appraisal o f BHP (usinghigher costs as noted), but omits BIP from this 210 analysis altogether, on grounds - as the Panel was recently informed - that because o f the BIP's wider role in the system, "it would be inappropriate to attribute the transmission line costs solely to the Bujagali project". 5.6 The Economic Study also appears to omit or underestimate other cost elements for the BHP, included inthe PAD estimate. The full comparisoni s as follows: EconomicStudv $m PAD BHP EPC 441(1) 520(2) Idc 94(3) 94 Other BHP 5l(4) 184 TotalBHP 586 798 BIP 28(5) 55 TOTAL 614 853 Notes: (1) Items 1 and 3 inTable 5-4, Economic Study (2) Para 53, PAD (3) Interest duringconstruction (Idc), assumed-see para 5.4.3 Economic Study (4) Items 4 and 5, Table 5-4 (5) Item 2, Table 5-4 5.7 O f course it i s quite legitimate to use different figures for different appraisal purposes. Thus incomparing new generation options, it may be fair to omit financing costs for all options but include Idc as an inescapable cost for all. On the other hand, tariff calculations are absolute, not relative, so that all costs to be recovered should be included .The following questions arise: (a) why did the PAD apparently ignore the significant increase inEPC costs since the Economic Study? (b) For a fair comparison of generation options, transmission connection costs for all shouldbe included. Giventhat without BHP, BIP i s also an avoidable cost, should it not havebeen included infull? Failing to do so disadvantages, inthe appraisal, other generation options whose connection costs are less. (c) In considering tariff effects, the full recoverable costs o f the Project must be included. In this case (as noted in section 5), it is not clear to what extent it i s intended to recover the cost o f the BIP through the BST. The loan repayment terms would theoretically allow a relaxed attitude to this; but even ifthe total cost o f BIP i s omitted for the purpose o f tariff calculation, the Economic Study still appears to underestimate costs as shown inthe PAD by $212m ($798m - $586m). It thus seems likely that the Economic Study underestimatedboth the costs (for comparative purposes) and the tariff effects o f the BHP/BIP project. 211 (d) To claim that BIP costs should not be attributed to BHP because o f its wider role inthe system, but at the same to claim for BIP a full shareo fthe revenue benefits o fBHS (see para 5.6 above) appears inconsistent. 5.8 Costs, Revenues and Tariffs inthe PAD para 84 indicates levels o f output assumed inthe low andhighhydrology scenarios. Using those figures, PAD para 95 showsthat in a high hydrology scenario, Bujagali's lifetime (30 years) capacity charges could be recovered through a levelized bulk supply tariff (2.5 % p.a. inflation assumed, 2006 prices) of 5.7c/unit. The equivalent figure under low hydrology, calculated to have the same value, $113m, i s 9.7c/unit. The intention would presumably be to include this charge in UETCL's BST, to be passed on to customers via UMEME and retail tariffs. The actual revenue generated for UETCL would however be less than $113m, (25% less, at a conservative estimate) because o f technical and commercial losses. By contrast, during the first 12 years o f operation (the period o f repayment o f senior debt), the BHP annual capacity charge i s actually estimated at an average o f $155m, with a peak o f $187m in 2022 (PAD Annex 11, para 10). So the levelized tariff would leave UETCL with a substantial revenue shortfall in paying the BHP capacity charge. The following questions arise: (a) how will a levelized tariff actually be set, given hydrological uncertainty? (One answer presumably would be to use the low/high hydrology probability estimate o f 79/21: on the PAD para 95 basis, this would give an ex-ante levelized tariff of 8.4c/unit) (b) whichever levelizedtariffis set, there will be a significant revenue shortfall, to be paid by UETCL, against the required capacity charge up to 2022, o f $32m, plus compensation for losses, p.a. on average, peaking at $74m plus in 2022. (If the tariff were set at 8 . 4 ~but 2022 was actually a year o f low hydrology, the revenue gap that year would rise to $89m plus). Has UETCL's revenue shortfall been includedinthe PAD financial, cash flow and retailtariff forecasts? (c) the revenue forecasts (see attachment 4, Annex 12) assume recovery rates rise from 54% in 2006 to 75% in 2013. Has the risk assessment given sufficient weight to the possibility o f both higher costs and significantly lower revenues? This will have a major bearing on whether the GOU guarantee o f capacity payments under the PPA i s likely to be triggered. (d) As for the Economic Study, it is not clear that the cost o fthe transmission project has been included in tariff calculations. Detailed consideration o f supply options inAnnex 9 appears to exclude or under-estimate connection costs - see table 9.5 which repeats the Economic Study figures. The actual bulk supply tariff which UETCL will pass onto the distribution sector, for inclusion inretail tariffs, should include an element for recovery o f BIP costs - see section 4 above. 5.9 Bujagali EPC Cost Risk para 41 o fthe PAD states "...there is limited likelihood of EPC cost increases once the EPC contract is finalized". Annex 9, paras 26 and 28, reporting the Economic Study on which the PAD i s based, indicate that the full risk maximum cost increase of lo%, balanced by the same probabilityo f a cost reduction o f analysis for the power system "with/without Bujagali" included a 20% chance o f a 5%. Thesejudgements may be over-optimistic, for the following reasons: 212 (e) after the price i s set, contractors are adept at pleading unforeseen geology/geotechnical grounds to justify an increase. The fact that the winning bid was significantly lower than the next best leads one to surmise that the contractor will want to take every opportunity to improve his margins (though as noted he has already managed to secure a 28% increase before fixing the price). Infact, the current contractual framework allows further price increases (see below). (0 although the Project may be technically straight-fonvard by international standards, the challenge in Uganda o f pulling together international and local contractors inan integrated project programmewill be significant (g) in their 2006 review of the draft contract, Linklaters drew attention to some provisions which appeared to relax the discipline on the contractor - on defect restitution, warranties, and his scope to resist Liquidated Damages inthe event o f delay. It i s not clear whether these have been tightened up. (h) the PAD puts weight on the incentive on BEL to contain EPC costs. This may also be over-optimistic, inthat there i s scope for cost increases to be recovered via the PPA (see next para). 5.10 PPA Capacity Charge The substitution o f a cost formula inthe 2005 PPA, for the maximum capacity charge specified in the 1999 PPA, is probablv the single largest adverse change, for the power purchaser and his guarantors, in the contractual basis for the present Proiect. It represents a significant shift inrisk away from the Project investors and lenders, on to the power purchaser. The formula and its effects can be described as follows: 5.10.1 The formula for determination o f the monthly capacity charge or payment is in Annex D to the PPA. It is very complex, since the components are defined rather than priced, and all are subject to variation. Inbroad terms, the components are: -- tariff - development costs EPC costs debt service reserve -- working taxes payable by BEL capital fees and All o fthese constitutingTariffProject Costs, plus -- - equityrepayment and return debt repayment GOUEquity (representing past development costs) - O&M fee 5.10.2 Some o f these are treated as pure pass-through (fees ,and elements o f the O&M charge). Others are carefully defined as to the make-up o f their "base" cost, and in some cases - including EPC costs - increases on the base are subject to a quantified percentage "cap". The costs are subject to accountants' inspection. However, the fact remains that, leaving aside debt repayment, BEL has 213 considerable scope to shape the base costs and in some cases the increases too, to deliver a higher capacity charge. 5.10.3 Considerable potential delay i s built into the determination o f the capacity charge (previous to which payments are on an interim basis). The charge must be set (the Final Declaration Date) within 2 months o f production o f a Final Cost Report, but that report need not be produced earlier than 6 months after the Final Draw Date, and that event (meaning the earlier o f the final draws on equity or debt) in turn may be up to 18 months after the commencement o f commercial operation. So 26 months may elapse after the start o f operations before there is a determined capacity charge. And curiously there are no specific provisions for capacity charge dispute resolution. The power purchaser may be relying on BEL to be motivated to move as quickly as possible from an interim capacity charge to the finally determined charge, but equally there i s plenty o f time as well as scope for BELto shape the figures. 5.10.4 As inthe prior Bujagali project, the capacity charge i s invariant to out put. So the payment will be same under low hydrology (when the output may be halved) as it will with high hydrology. O f course, hydrology is outside BEL'S control. But the payments are also relatively invariant to plant availability, which i s inBEL'Scontrol. A percentage reduction in availability (say 5%) would have to be sustained for a whole year before there was an equivalent reduction in the monthly capacity charge (PPA, Annex D). 5.10.5 For BEL and its lenders, Annex D no doubt represents a low-risk (though potentially disputatious) means o f managing and recovering costs which are bound to be subject to uncertainty. For the power purchaser and his guarantors, it means that there is no ceiling on capital costs and whether or not the Project delivers the direct economic benefits offered over 30 years, in terms o f costs and tariffs, is to a significant extent inBEL'Shands. 6. COSTS-SUMMARY OBSERVATIONS 6.1 Capital costs have significantly increased in real terms compared with the prior project, despite the adoption o f competitive tendering. There i s however no evidence to believe that the price bid obtained was higher than what might be assessed as a "fair market price" for Uganda. 6.2 There has however been a further increase (28% since the the Economic Study, or 8% since the PAD) to the point where the price was fixed. 6.3 The PAD may be optimistic in its view that EPC cost escalation is unlikely. Putting this together with the absence o f a capped capacity charge in the PPA, there has been a significant transfer o f cost risk to the power purchaser compared with the prior Bujagali project. 214 6.4 The Economic Study, and also the PAD which depends on it, may have understated Bujagali's total costs compared with other generation options, and in assessing tariff effects, with reference to financing costs and the BIP. There i s a risk that Bujagali will apply upward rather than downward pressure to retail tariffs. 6.5 Inthe adoption o f a levelized life-time tariff for Bujagali, UETCL will be set a major financial challenge in the period 2011-23. There may be a risk that its revenues will be insufficient to meet its obligations, since the projected revenues depend heavily on substantial increases inboth customer numbers and recovery rates. The potential revenue gap inthis period will actually be larger than that identified ins.6.8, taking account o fthe final, higher, EPC price. 7. NEW RISKMITIGATIONFORTHE PRESENTBUJAGALIPROJECT 7.1 In this section Idescribe the new risk mitigation measures taken for the present Bujagali project compared with the prior Project (see section 4 above). 7.2 Award of the project by Competition. Competitive solicitation for IPP projects i s o f course international best practice (IBP). It should ensure the lowest market price consistent with technical fitness for purpose. Inthis case however, competitive pressures were weak, and the benefit o f selecting the lowest compliant bidder has been offset by other factors exertingupwardpressure on costs, as described insection 5. 7.3 World Bank Group links with the Equity partners. The PAD paras 64 and 65 describes IFC's links with Industrial Promotion Services (Kenya). The importance o f this, together with other safeguards regarding future changes inequity holding, is that it should reduce the mediudlong -term risk o f collapse precipitated by withdrawal o f the sponsors. Sithe Global i s an experienced and respected international IPP company (as was AES in 2001); should they wish to withdraw at a later date however, it might be expected that IPS (K) could temporarily take over equity leadership and engage another experienced investor/operator - or provide a transition into public ownership. 7.4 More Comprehensive Studies and Plans. An effort has clearly been made to anticipate and answer criticism o f the Project in Uganda by the conduct o f detailed environmental, economic and social studies - required for a project which has such a big potential impact. There appears to have been a change o f mind-set since the prior project: for that project the power purchaser and his guarantors took an arms-length approach, leaving it mainly to AES to overcome the planningand other local problems and propose solutions, whereas for the present Project it has been recognised at the outset that although BEL continues to take the lead, these problems will not be overcome without the involvement and long-term commitment o f the public authorities. In my personal opinion, it i s particularly important that public authorities should under-write the resettlement costs (some o f them long-term) arising from local disruption at the dam and along the interconnectingtransmission line. This should be an important factor ingaining public support, and thus reducing social andpolitical risks. 215 7.5 Buy-back in case of Low Hydrology. For both the prior and present Bujagali projects, the PPA/IAs provide for buy back o f the plant by UETCL under default conditions and certain force majeure events. In general terms, these provisions follow international norms. However, para. 4.8 o f the PPA o f the present Project adds a new provision: UETCL may terminate the PPA and buy back the plant in the event o f 30 consecutive months o f "low water". This is an important safeguard: as Iexplain later the cost o f power from Bujagali, per unit, as determined by the PPA may become prohibitively high in a sustained low hydrology scenario, and in those circumstances it will be preferable for the public authorities to assume control, when they can stop paying the fixed capacity charge, smooth tariff effects and ensure that funds are available for alternative generation. The provision i s to be welcomed: Ihave two reservations about it, first that the low water trigger may have been defined too demandingly from the power purchaser's perspective; and second that the payment terms for buy-out (Annex J to the IA) which mean that BEL can set the price broadly to equate to capacity payments foregone, seem generous to BEL, given that the plant will be in real trouble if this scenario occurs. However, Irecognise that the sponsors and their lenders are looking for protection against loss. 8. RISKREVIEW 8.1 Inthis section Idescribe the main risks to which in my view the Project is exposed, how these are shared, and the possible consequences. 8.2 Capital cost escalation. (See section 5 above) Ifthe capacity charge i s set higher than present estimates, or rises subsequently, either tariffs must increase or additional subsidies paid to UETCL. 8.3 Currencv depreciation. For the present Bujagali Project as for its predecessor, capacity payments are denominated inUSD ($). As Ipointed out in my 2002 report, and as stated inthe InspectionPanel's Report on the prior project, a 10% p.a. depreciation o f the USh against the USD would double the price o f the project to Uganda in 7 years. Consequencesas in8.2. 8.4 Prolonged low hvdrolony. A more pessimistic but more realistic view o f hydrology has been taken for the present Bujagali Project as compared with the prior project. Nevertheless substantial uncertainty remains. Past hydrological patterns have shown great year-on-year volatility, so that both the "high" and "low" numbers used in the PAD are long-term averages only. Para 95 o f the PAD illustrates how the cost o f a unit fiom Bujagali rises dramatically in a "low" year. A levelized tariff may be set ex-ante, but if the actual hydrological pattern falls below that assumed for the levelized tariff, then the capacity charge shortfall (see para 6.6 above) will widen and the consequences as in 8.2 follow. 8.5 Lower demand aowth. Assumed demand growth rests both on continuing growth o f demand from existing customers, and a highrate o f new connections/customers, such that 216 the number o f customers almost doubles by 2012. If this growth does not occur, UETCL's revenues fall below forecast, with possible consequences as before. To illustrate, ifBujagali were operating today, its average capacity charge duringthe first 12 years would pre-empt over three quarters o f total electricity sector revenues (customer payments) inUganda (PAD Annex 12, attachment 1) 8.6 Lower or static recovery rates. It has been optimistically assumed that recovery rates will have risen to 75% by 2013. Ifthey remain at the 2006 rate (54%), sector revenues will be 28% lower. Consequences as before. 8.7 Affordability. If the PAD'S economic analysis i s proved correct, Bujagali's introduction will allow a reduction in (real) retail tariffs o f at least 5% compared with current levels. Collection rates appear not to have been significantly affected by the large (approximately 80%) increases inthe last 3 years, so Bujagali's affordability on that basis doesn't seem to be subject to highrisk (though new customers may reveal different price sensitivities - and produce different collection rates - compared with existing customers). However, ifany o f the risks above arise, further subsidies may be a preferable alternative to a tariff increase which mightreduce rather than increase revenues. 8.8 Construction Delav. Despite Liquidated Damages provisions penalising the contractor, the costs o f delay would be likely in practice to be shared via the PPA with the power purchaser (see section 5.7 above). Extreme delay could require additional stop- gap generation. Otherwise, the main consequence o f delay would be to defer for customers the main benefit o f the project, namely a reduction inpower-cuts. Overall, this may be regarded as one o f the lesser, or more manageable, economic risks. 8.9 Withdrawal o f the DeveloDer/ODerator. This risk has been mitigated compared with the prior project. Sithe i s bound in for the constructionphase, and subsequently would be replaceable as operator if not so easily as investor. Adequate provision has also been made for the project to be bought out ifnecessary. 8.10 Poor Plant Performance. Although the PPA i s generous to the owner-operator inthe scale o f penalties for low availability, this may be regarded as low-risk. In the extreme, the provisions for Company Default provide a safety net. 8.1 1 Sharingo f Risk. From the documents, the greatest share o f economic risks lies with the power purchaser. The capacity charge may be adjusted upwards if the developer/operator hits unforeseen costs, but not downwards if demand or supply conditions deteriorate for the purchaser. In effect, the lenders especially but also the investors are held harmless against all or most eventualities. However, in a crisis o f non- affordability in Uganda such as might be produced by currency devaluation or very low hydrology, the investors and lenders may also be at risk, ifthe money to pay the capacity charge i s just not there. In these circumstances, buy-out is likely to provide the best solution. Personally, Iwould have preferred (as described in my 2002 report and in the Inspection Panel's Report on the prior Bujagali project) to see terms more favorable to . the purchaser. 217 9. OVERVIEW 9.1 Generally, the institutional and financial arrangements for the present Bujagali Project and for the BIP recognise (to a much greater extent than for the prior project) that this is aNational Project o fgreat strategic importance to Ugandawith wide implications, in which the Government and its IF1backers have major roles to play. This gives the Project a degree o frobustness its predecessor lacked. 9.2 However, its direct economic benefits may have been over-rated. There i s a considerable risk that it will exert upward rather than downward (as the PAD argues) pressure on tariffs - though given current hydrocarbon price forecasts it may still be the least-cost option. Inpullingforward the economic benefits to customers by adoption o f a levelized tariff rather than one which follows the capacity charge (which i s twice as high duringdebt repayment as it is subsequently), the power purchaser is opening the riskthat he will require increased Government subsidies in the period 2011-23. And if demand or hydrology are significantly below current central estimates, Bujagali looks poor value for money. 9.3 The chosen level of the tariff will in practice be the way in which the balance o f interest i s struck between customers, and the power purchaser and his guarantors. The tariff will o f course not need to be determined until closer to the date o f operation, reflecting the then view o f costs. If UETCL i s to be given a full opportunity to pay the capacity charge without recourse to further subsidies via the GoU guarantee, it would be prudent to set the levelized tariff at least the low hydrology level - intoday's terms, not less than lodunit. To the extent that the BIP costs are to be recovered, or BHP costs increase (as they already have since the PAD), or compensation for commercial/technical losses i s required, the addition to UETCL's bulk supply tariff should probably be higher than this. 9.4 Once debt i s repaid ,the picture changes, inBujagali's favor. It may then be possible to reduce its tariff. After the first dozen years, it should become a reliable source o f cheap power (so long as the Nile flows!) for decadesto come. 9.5 A wider personal observation relates to the decision to build BHS as an IPP, rather than as for the BIP, a public sector project. The highcost o f commercial debt inUganda, coupled with the high-pricing, risk-averting strategy of the investors in response to a perceived high-riskenvironment, has inevitably saddled the project with large front-end- loaded costs. GOU and the IFIs have still had to accept ultimate liability as guarantors. As an alternative, public sector financing might have produced lower costs overall, and would certainly have made it easier to manage costs and cost recovery via tariffs over a 40 year project life-time. Private sector capabilities could still have been harnessed to build and operate the plant on a contracted basis. Of course, I recognise the demonstration value o f attracting a major private investment into Uganda, but it might be argued that a smaller, lower risk infrastructure project would have been a better place to start. Ioffer these personal observations with diffidence since.it might be said they are 218 outside my TOR.; but it would seem prudent to take stock o f the Bujagali experience before deciding the strategy for Karuma ifthis i s to follow. 9.6 Setting aside the positive aspects o f the Project, it may be helpful to summarise those areas inwhich management performance may have fallen short. These appear to be: - In analysis of generation options: there was an insufficiently transparent approach to the assessment o f Bujagali against all feasible alternative generation options, making it difficult for management authoritatively to refute charges that the analysis - was either inadequate or biased infavour o f Bujagali. In project evaluation: the assessment o f costs, risks and benefits was unduly optimistic. In particular (1) capital costs were under-estimated; (2) the likely effect on tariffs was under-estimated: Bujagali is more likely to exert upward than downward pressure on tariffs; (3) the risk o f a significant revenue gap (between UETCL's income and the requirements o f the Bujagali capacity charge) in the first 10 years o f the project was not recognised: the attendant risk i s that the WB guarantee may be called; (4) other risks, notably those o f shortfalls against recovery - rate forecasts, and the exchange rate, were not given due weight. In project structure and risk management: (1) management failed to realise the benefits o f competitive solicitation by allowing a long period o f post-bid negotiation with the winning bidder, during which the price increased by nearly 30%; (2) management failed to fix or cap the capacity charge in the PPA, thus increasing the risk that the power purchaser will have to accept further cost increases inthe future; (3) management failed to set sufficiently robust performance penalties and buy-out terms, to minimise downside risk for the power purchaser. 9.7 It must be recognised that WB management were, especially in respect o f the third category above, dependent on the actions, decisions and advice o f other agencies. Nevertheless, the World Bank's influence on the Project has been very significant, and WB management would have been fully entitled to examine and if necessary change policy on all o fthese matters. 219 Annex C Spiritual Significance in Busoga Culture Prof. Theodore Downing 1. The Bujagali Hydroelectric Project i s moving into a neighborhood known for its strong, complex cultural and spiritual tradition. Although peoples o f other groups inhabit the project area, the Basoga claim spiritual dominion o f both sides o f the Nile, its islands, the water and its waterfalls.688According to the 2002 census, there are about 2.7 million Busoga in Uganda whose territory lies to the east o f the project site.689Their language, Lusoga, predominates inthis area, on the East bank o f the River Nile. The Basoga share a common dialect and ideological, spiritual history, sharing a cluster o f eight or more high status spirits who are invoked intheir specific ceremonies, i.e., prayers, blessings for a good crop, ajob, healing, divinatiodconsultation, or witchcraft ceremonies, depending on the specific need or celebration. These eight high status spirits include Lubaale, Kirztu, Mukama, their legendary fathers and Budhagaali (the spirit residing at the Bujagali Falls site). The Basoga are distinct from the Buganda, the more dominant tribe inUgandawhose traditional realm reaches to the West bank o fthe Nile. 2. To the Basoga, the project area - like their entire region - i s inhabited by ancestral spirits and living humans who are constantly interacting - from birth to death and beyond.690 The Enswezi, their traditional spiritual cosmology, i s extensive and complex. Everyhuman being possesses a body and a soul. When a person dies, the body disintegrates but the soul continues to exist as a spirit (orni~irnu).~~~spirits are innumerable, consisting o f the spirits o f everyone The who has lived since the beginning. 3. The spirits exercise very strong influence on the harmony, wealth, physical and emotional well-being o f the living - most o f all on their health and livelihood. They play a critical role in group welfare and regulate the moral conduct o f the 688 The 2001 RAP states its baseline survey identified 22 ethnic groups living in the project area (Bujagali Hydropower Project Social and Environmental Assessment Main Report, December 2006, page 161). The region was repopulated by migrants from throughout Uganda and other central Af?ican countries in the 1940's after being nearly abandoned by the Busoga at the turn o f the century due to sleeping sickness. Bujagali Power Project - Hydropower Facility - Resettlement and Community Development Action Plan, March 2001, page 98. Both banks o f the Nile are recognized by the Uganda government as Basoga, but sleeping sickness prevented dense settlement until the late 1940's when peoples from all over Uganda and neighboring East African countries settled it. Bujagali Power Project - Hydropower Facility - Resettlement and Community Development Action Plan, March 2001, page 24 By the time o f the baseline study, in 1999-2000, only 46% o f the people inthe project area were Soga, mostly living on the east bank (54% vs. 36% on the west bank). 689 www.busoga.codaboutBusoga.php - (Obwa Kyabazinga BwaBusoga online). 690 Over fifty years ago, Lloyd Fallers, inhis classic study o f the Basoga, Bantu Bureaucracy (1954), felt that despite the substantive presence o f Catholicism and other global religions, ancestor worship was "very near the heart o f the Sogavalue-system.'' page 80. "' Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 76. 220 living.692Practically all good fortune and misfortune is related directly or indirectly to actions o f certain spirits or to witchcraft, including death.693Illness can result from people not living in harmony with the spirits. The good spirits, although without physical bodies, are believed to eat, drink and demand their due from their earthly relatives.694 4. Families and households have immediate individual ancestral spirits (omuzimu), venerated within individual compounds through shrines (amasabo) and offerings. Communications from and with these spirits are on issues o f concern to the family (i.e., land problems, lost will or a spirit's dissatisfaction with its property di~tribution).~~~ At this level, these spirits do not have permanent priests (mediums). Communication with these spirits occurs simply through individual prayers or dreams, unless there are some serious issues that require the intervention o f a diviner. The spirits, however, can choose any member o f the family or household to communicate their message.696The elaborateness o f ritual and offerings to the ancestral spirits varies with the occasion and issue. These ceremonies can be performedby family members and elders o f the clan as well as a diviner.697 5. Above the level o f the family spirits are the spirits o f the founders o f the clan. A clan i s a group o f people who trace their lineage to a common, distant patrilineal ancestor. Regardless o f blood lines, clan members consider themselves to be brothers and sisters. Clans are not limited to an area, but may be dispersed throughout the lands o f B u ~ o g a Each ~o f the 338 or more clans recognizes a . ~ ~ founding ancestral spirit or E n k ~ n iIn ~ ~ . ~terms o f displacement and resettlement, these spirits are very important. They are reveredbecauseo fthe special protection they provide to the clan members. They a) unify the clan, b) provide blessings to clan members during special invocations, i.e., job interviews, examinations, or good grades, c) preserve the clan, d) punish those who treat clan members 692 Witchcraft, Divination, and Healing among the Basoga, Richard Kayaga, Editor, Cultural Research Centre, Jinja, Uganda Marianum PressLtd.2003, and Ritual Gestures inBusoga, Cultural Resource Center, Jinja, Uganda 2001, Traditional Religionand Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 77. 693 Witchcraft, Divination, and Healing among the Basoga, Richard Kayaga, Editor, Cultural Research Centre, Jinja, Uganda MarianumPress Ltd. 2003, page 9. 694Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 77. 695Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 76. 696Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 76. 697Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 78-85. 698Fallers 1965, page 64-65. 699Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, pages 16-75: Field review with two Basoga, identified other clans that were not on the initial list. The Enkuni are discussed on page 89. 221 unjustly, and e) accompany clan members on their moves.7ooWhen a family breaks away from its territory for any reason, in order not to sever all links with the old clan, the head o f the family will pick up a stone or collect some soil from hisold place to take with himto his new place. It is believed that the stone or the soil represents and incorporates the spirit o f the ancestor who started the clan and that by taking it along, the spirit o f the ancestor accompanies them to their new place, gives them its blessings and helps the family maintain the feeling that they are still part o fthe clan they left behind.701 6. When clan-wide problems such as sterility, quarrels, famine, sickness, death, or poor harvests occur, people attribute these problems to the Enkuni being angry. Unlike familial spirits, communication with the clan spirits is more elaborate, requiring the mobilization o f more people and resources. In order to determine whether a problem i s or i s not associated with an Enkuni, the clan leaders seek the assistance o f a diviner (mulaguzi), who determines if the problem i s associated with the Enkuni. The clans have a traditional clan priest (muswezi singular) who - i s always accompanied by a group o f fellow priests (baswezi - plural) during ceremonies. The offerings are provided by and are at the expense o f the members o f the clan. These Busoga clan rituals invoke the high status spirits including: Kintu, Mukama, Lubaale, Enkuni, Budhagaali, Iyingo, Waitambogwe,Isegya and L~koghe.~'~ 7. The highest status or princely spirits, the Emisambwa/musambwa, are the spirits o f the founders o f the Busoga: Kintu, and Mukama, their wives, brothers and their children, powerful kings, clan leaders, and mothers o f lineages. People believe that the founders were the children o f Kintu and Mukama. These princely spirits are considered to be strong spirits, possessing the ability to reincarnate into animate and inanimate objects. Highest among them are the Kintu, the Budhagaali, the Lubaale, the Mukama, the Iyingo, the Nawandio, the Waitambogwe,and the W ~ n h i . ~ ' ~ 8. Princely spirits (musambwa) will possess someone to be its priest/priestess and therefore have a shrine.704The spirit gives its priest powers to give good luck, blessings, prosperity and protection against evil spirits. The confirmation of someone possessed by a musambwa requires the participation o f the traditional clan spiritual leaders (baswe~i).~''Ceremonials and rituals at musambwa level are 700Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, inter alia. 701Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 89. 702Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 90-9 1. 703Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, pages 112, 114. 704Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 114. 705Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 149. 222 proportionally more elaborate, expensive and involve participants from different Busoga clans. 9. To summarize, from the perspective o f the Bujagali Project, the key elements o f Busoga spiritual cosmology are: a) the spirits are innumerable, powerful and frequently cross over into the world o f the living and may do both good and harm, b) they inhabit the same world as the living and are associated with animate and inanimate objects throughout the landscape, c) they can move freely without the need for human permission, d) they have differential power, influence, and interests, e) they are hierarchical, somewhat comparable to the ancient Greek Pantheon, f) they influence the health, well-being and the livelihood o f the living, g) more powerful spirits communicate through mediums who do not view themselves as capable o f negotiating or predicting spirit behavior - they are mediums o fthe spirit who possessesthem, and h) the mediums are selected by the spirits not by the cultural (political) leaders. 10. This brief primer places the Bujagali Falls site and its chief spirit, the Nabamba Budhagaali, and its medium, the Nabamba Bujagali in context. As in most religions, the sacredness o f a site comes from a group o f religious practitioners assigning special significance to a specific site, to a specific spirit with a specific power, the power to provide blessings to the believers (health, happiness, harmony and protection o f their livelihood). At the Bujagali Falls sites, just as in many traditional indigenous or animistic religions, these spirits are usually associated with animate or inanimate objects, such as animals, rocks, trees, rivers, mountains, or waterfalls.706Consequently, destruction or disturbance o f a sacred site and the associated ceremonies and offerings to spirits may be 11. The available evidence confirms that Budhagaali Falls i s the residence o f a host o f spirits ranging from individual family spirits to high level Busoga spirits, particularly one o f the Busoga's most venerated, powerful, princely spirits, Nabamba Budhagaali Spirit. This Spirit is the son o f the founding ancestral couple o f the Busoja, Kintu and NambL7O8 Nabamba was nicknamed "Bujagali" becausehe was fond o f swimming. The Nabamba-Budhagaali Spirit may chose to stay inthis place and is fiee to shift to another place- without any human or other spirit's permission.709 The Busoga believe that at the time o f his birth, Nabamba Budhagaali Spirit turned into "water", which water turned into the water o f the '06Routine and Dissonant Cultures: A theory about the psycho-socio-cultural disruptions o f involuntary resettlement and ways to mitigate them without inflicting more damage. Theodore E. Downing and Carmen Garcia-Downing. InAnthony Oliver-Smith. Develovment and Disvossession: TheAnthrovolom of Disvlacement and Resettlement. Santa Fe: School for Advanced Research Press, 2008, inPress. 707A E S contracted a Consultant to survey the traditional religious sites and beliefs in communities along the East and West banks o f the Nile River and identified specific names for these features. The River Nile and its Significance to Traditional Religion and Practices o f the Inhabitants o f the Wakisi Subcounty - East Bank, September 18,2000. 708Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, pages 10, 114 and 149. 709Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 149. 223 Nile River. This powerful spirit may take on multiple animate and inanimate forms. 12. Both the Nubambu Budhagaali Spirit and its medium, the Nabamba Bujagali, are unquestionably tied to the Busoga clans stretching across Busogaland. The Nubumbu Budhugaali Spirit can possess a spiritual leader from any Basoga clans, who becomes, according to interviews, "like un arch-bishop" among the clan spiritual leaders (buswezi ubudhuguuli - plural). Each clan can have a muswezi abudhuguuli (singular), a clan level spiritual representative o f the highspirit who are ordained at a sacred rock associated with this high spirit at the Bujagali Falls religious site (see Figure 2 below). The recognition and initiation o f Nubumbu Budhuguuli Spirit's medium i s presided over by a conclave o f these representatives, jointly known as the Baswezi Budhagaali. Presently, Nabamba Bujagali i s the medium for the Nubumbu Budhugaali Spirit. His initiation was recognized by Busoga clan spiritual leaders (baswezi) and other seers.71oThese trans-Busoga networks o f clan spiritualists conduct their final initiations and rituals at Bujagali Falls.711 Figure 2: NabambaBudhagaalipriest introducing the newly possessedat the rock. This i s the spirit site for Budhuguali, and raising hand o f the new priest. This i s a sign o f taking an oath. Source: Ritual Gestures in Busoga, Cultural Resource Center, Jinja, Uganda 2001, Traditional Religion and Clans among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 49. 710Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, pages 149-150. The Nabamba Bujagali is selected by the Nabamba Budhagaali spirit from among other spiritualists. 711Traditional Religion and Clans Among the Basoga, Volume 1, Richard Kayaga Gonza, Editor, Cultural Resource Center, Jinja, Uganda 2002, page 149. 224 13. Bujagali Falls spiritual centrality i s not limited to the Nabamba Budhagaali Spirit and its medium. Lubaale Nfuudu, a spiritual medium (muswezi) for another o f the princely, high status Busoga spirits, Lubaale, has temporarily relocated some Bujagali spirits. Lubaale Nfuudu takes care o f multiple shrines (amasabo) where new spirits were arriving all the time. He was briefly possessedduringthe Panel's interview. Whether or not he relocated the Nabamba Budhagaali Spirit i s unclear andprobably immaterial, since the Spirits are free to move wherever they wish. 14. In2001, the Project noted that the Ntembe clan, whose leader i s Ntembe Waguma, and diviner (muswezi) i s Nfuudu, see the Bujagali Falls as the location o f their clan level ancestral spirits which will be disturbed by the project. Lubaale Nfuudu i s the caretaker for his clan spirit and Lubaale, another Busoga ancestral spirit. He also states he also built Nabamba a shrine and questions the legitimacy o f Nabamba Budhagali as a medium.712 15. Consultations with L C l and LC3 local council on the west bank revealed that apart from the Nabamba Bujagali spirit, other spirits on the west bank needed appeasement.713Resolution o f spiritual disturbances i s different for clan and family level spirits. Family level spiritual disturbances in the immediate project area ap ear to have been resolved in the Sponsor's individual mitigation action^.'^ Like other clans, the Ntembe are found throughout Busogalandwith the Bujagali Falls area being their clan site.715 16. A Cultural Management plan and strategy for dealing with the higher spirits associated with Bujagali Falls, particularly Nabamba Budhagaali spirit, consistent with Busoga cosmology, has yet to be established. At the level o f the higher spirits, all Busoga clans and their Bujagali Falls associated baswezi are stakeh01de1-s.~'~ 17. The high spirits (musambwa) o f one island to be submerged are associated with Kintuand hiswife Nambi. They are the founding couple o fthe Basoga, father o f Lubaale and Nabamba Budhagaali spirits. The study noted that "ifKintu and Nambi are annoyed they can [leave the island] come to the land and take domestic animals or even thepeople themselves as sacrifices. No one is accepted to light a fire or burn the bush on the Island. Ifone does so, Kintu would claim that they are burning his children and can cause harm."117 For those believing in this 712The River Nile and its Significance to Traditional Religion and Practices of the Inhabitants of the Walusi Subcounty - East Bank. AES Consultant September 18,2000, page 92. 713PCDP, Appendix B, Phase 1, ConsultationMaterials, Dec 2006, page 12. 714BujagaliPower Project - Hydropower Facility - RCDAP, March2001, page 102, T[ 15.17. 715The reports are ambiguous as to the spiritual and clan leadership ofthe Ntembe clan, with one document referring to Lubaale Nfuuduas the leader o f the Ntembe clan and another assigning this positionto Ntembe Waguma. Bujagali Hydropower Project Social and Environmental Assessment Main Report, Appendix I Assessment o fPastResettlement Activities and Action Plan (APRAP) December 2006, page 23 ft 3. 716BujagaliPower Project - Hydropower Facility - RCDAP, March 2001, page 102,115.17. 717The River Nile and its Significance to Traditional Religion and Practices of the Inhabitants o f West Bank.AES Consultant, September 18 2000, page 3. 225 traditional religion, disturbance of the sacred sites i s an issue of livelihood, harmony, health andwell-being. 226 Annex DBiographies Mr. Werner Kiene was appointed to the Panel in November 2004 and has been its Chairperson since September 2007. He holds a Masters o f Science degree and a Ph.D. in Agricultural Economics from Michigan State University. He has held leadership positions with the Ford Foundation and German Development Assistance. In 1994, Mr. Kiene became the founding Director o f the Office o f Evaluation o f the United Nations World Food Programme (UN WFP). He was the World Food Programme Country Director for Bangladesh from 1998 through 2000 and also served as UN Resident Coordinator during this period. From 2000 to 2004 he was a Representative o f the UN WFP in Washington, D.C. Mr. Kiene's focus has been on the design, implementation and assessment of sustainable development initiatives. His professional writings have dealt with issues o f rural poverty and social services delivery; food security, agricultural andregionaldevelopment; emergency support and humanitarianassistance; international trade and international relations. Mr. Gene i s involved in professional organizations such as the European Evaluation Association; the Society for International Development; the American Association for the Advancement o f Science; and the InternationalAgriculture EconomicsAssociation. Mr.Tongroj Onchanwas appointed to the Panel inSeptember 2003. Hehas a Ph.D. in agricultural economics from the University o f Illinois. Professor Onchan taught on the Faculty o f Economics at Kasetsart University in Thailand for 26 years, including a term as Dean. He later served as vice president o f Huachiew Chalermprakiat University; then joined the Thailand Environment Institute (TEI) as vice president. In 1998, Mr. Onchan was appointed president o f TEI. He helped establish and was appointed president o f the Mekong Environment and Resource Institute (MERI) in 2000. He has served as advisor to the Prime Minister and to the Minister o f Science, Technology and Environment, as member o f the National Environmental Board, chairman o f the National EIA Committee, chairman o f the Committee on the Preparation o f State o f the Environment Report for Thailand, and member o f the National Audit Committee. Mr. Onchan i s on many editorial boards, among them the Asian Journal o f Agricultural Economics and the International Review for Environmental Strategies. He has consulted for a number o f international organizations, including the Asian Productivity Organization, ESCAP, the World Bank, the Asian Development Bank, the Food and Agriculture Organization, the International Labor Organization, USAID and the Ford Foundation. He has been project director o f over thirty research projects and author or co-author o f numerous technical and research papers on rural development, natural resources and environmental management. Currently, he serves in several capacities: chairman o f the Board o f Directors o f the MERI, member o f National Research Council for economics, and a director o f the International Global Environment Strategy (IGES) based in Japan. Mr. Onchan was appointed as eminent person to serve as a member o f the Asia and Pacific Forum for Environment and Development (APFED). 227 Mr.Roberto Lentonis currently Chair ofthe Technical Committee ofthe GlobalWater Partnership and a Member o f the Inspection Panel o f the World Bank. A specialist in water resources and sustainable development with over 30 years o f international experience in the field, he also serves as Chair o f the Water Supply and Sanitation Collaborative Council, Member o f the Board o f Directors o f WaterAid America, and Senior Advisor to the International Research Institute for Climate and Society (IRI) at Columbia University. A citizen o f Argentina with a Civil Engineering degree from the University o f Buenos Aires and a Ph.D. from MIT, Dr. Lenton i s a co-author o f Applied Water Resources Systems. He is also a lead author o fHealth, Dignity and Development: What will it take?, the final report o f the United Nations Millennium Project Task Force on Water and Sanitation, which he co-chaired. Dr. Lenton was earlier Director o f the Sustainable Energy and Environment Division o f the United Nations Development Programme in New York, Director General o f the International Water Management Institute in Sri Lanka and Program Officer in the Rural Poverty and Resources program o f the Ford Foundation in New Delhi and New York. H e has served on the staff o f Columbia University and the Massachusetts Institute o f Technology (MIT), including posts as Executive Director o f the IRI Secretariat for International Affairs and Development and Adjunct Professor inthe School o f International and Public Affairs at Columbia and Assistant Professor o f Civil and Environmental Engineering at MIT. ************* Theodore Downing, Research Professor o f Social Development at the University o f Arizona, earned his PhD from Stanford in Social Anthropology. Specializing in international social policy development, he has extensive research, project management, and policy-making experience in Latin America, Africa, and the Middle East. His understanding o f capacities and limits o f government was enhanced by being elected for two terms to the Arizona House o f Representatives. Beginning as a short-term consultant to The World Bank in 1987, he has worked on involuntary resettlement and indigenous peoples safeguard issues through most phases o f the Bank's project cycle - preparation through supervision and across the energy, agricultural, and the extractive industry sectors. His development experience includes directing the Mexico's anti-coffee rust research team for the Mexican National Science Foundation, helping establish an environmental science college at King Abdulaziz University in Jeddah. His colleagues elected him to be President o f the international Society for Applied Anthropology and i s currently President o f the International Network on Displacement and Resettlement (www.displacement.net). Samples o fhiswritings and project experience are available at www.ted-downing.com. Richard Fuggle i s Emeritus Professor o f Environmental Studies at the University o f Cape Town. Prof. Fuggle i s a Member o f the Academy o f Science o f South Africa, a Registered Natural Scientist, a Certified Environmental Practitioner inSouth Africa and a Professional Member o f the South African Institute o f Ecologists and Environmental Scientists. He has edited two books on environmental management in South Africa and has published over 100 academic papers on environmental topics. He led the teamwhich . developed the South African Guidelines for Integrated Environmental Management. Prof. 228 Fuggle has served on numerous Commissions o f Enquiry related to Environmental Assessments. He has received awards and distinctions for his contributions to the advancement o f Environmental Impact Assessment both nationally and internationally. Prof. Fuggle earned his Ph.D from McGill University inMontreal. Graham Hadley was educated at Cambridge (MA Modern History 1966) and in 1991 completed the Harvard Business School Advanced Management Programme. After a Civil Service career (final position: Under Secretary, Department o f Energy), Mr.Hadley joined the Electricity Industry in 1983. He played a key role in the restructuring and privatization o f the industryin 1988-91, first as Board Secretary to the Central Electricity Generating Board, and subsequently (1990-95) as a Board Member and Director o f National Power. At National Power his main role was as MD o f international business development: from a zero base he established the company as one o f the leading independent power producers, with assets in the US, Europe and Asia. Since 1996 he has been an economic and commercial consultant, in the position o f Senior Advisor to National Economic Research Associates (NERA) until 2007. At NERA or independently, he has advised IFAs, Governments, utilities and regulators internationally on utility sector reform, particularly on industrial restructuring, the introduction o f private capital and investment strategy. He has also specialized in power purchasing and the use o f power purchase agreements (PPAs) to manage risk. In this area his assignments have included: Expert Witness in a PPA contract arbitration dispute between a U S power company and a Caribbean utility; advice to and training in power purchasing o f Electricity Regulators inIndia and Sri Lanka ;assistance to the Government of Mexico in developing model PPAs and the policy o f competitive power solicitation. He has also advised developers and power purchasers on specific PPAs. From 2000-2007 he was also a Member o f the UK Competition Commission, serving on a number o f both merger and sector InquiryPanels. Peter Pearsonreceived his degrees ineconomics from the Universities o f Keele, London and Surrey. He i s Professor o f Energy and Environmental Studies, Director o f the Centre for Energy Policy and Technology and a director o f the Energy Futures Lab at Imperial College London. He has held academic posts at the Universities o f Glasgow and Surrey. From 1989-94, he headed Surrey University Energy Economics Centre (SEEC), and in 1993 held a UK Economic and Social Research Council Global Environmental Change Research Fellowship. He has been Chair (1992, 2002) o f the British Institute o f Energy Economics, was a member o f the European Commission's Advisory Group on Energy (AGE) for the 6th Framework RTDProgramme (2002-2006), an invited reviewer for the International Energy Agency's World Energy Outlook, and Chair o f the International Evaluation Panel on Environment and Societal Sciences for the Academy o f Finland (2006). He delivered the 2007 Queen's Lecture at the Technical University o f Berlin and was a Specialist Advisor to the House o f Commons Innovation, Universities, Science & Skills Select Committee's Inquiry into Renewable Electricity Generation (2007-08). He is the author/co-author o f 160 scholarly publications on energy, environment and water 229 Carlos E. M.Tucci, Civil Engineer and Professor at the Institute of Hydraulic Research o f the Federal University of Rio Grande do Sul. Mr.Tucci holds a Ph.D. from Colorado State University, 1978. He i s a former vice-president of the International Association o f Hydrologic Science, a former president o f the Brazilian Water Resources Association, and a former executive secretary for the Water Resource Research and Investment fund inBrazil. He is now Chief-editor of the Brazilian Water Resources Journal and REGA Water Resource Management Journal for Latin America.. Mr. Tucci has about 400 publications inbooks, book chapters, Papers injournals, essays and conferences. He was honored by the Civil Defense o f The State o f Rio Grande do Sul and for the best scientific book publishedin 1993 inthe State o f Rio Grande do Sul. 230 IBRD 36274 To Kalagala Falls, To Kamuli To Kamuli INUNDATED LAND Victoriaapprox. 4 km. NW of Busowoko Falls K A G O M A PRIVATE POWER EXISTING RIVER AREA NEW HIGH WATER LEVEL UGANDA Busowoko Falls C O U N T Y GENERATION (BUJAGALI) PROJECT WHEN RESERVOIR IS FILLED PERMANENT FENCE LINE TRANSMISSION LINES PRIVATE POWER GENERATION Wokikola Source: Bujagali Hydropower Social & Environmental Nile Assessment (2006) (BUJAGALI) PROJECT PERMANENT ROADS BUDONDO Buyala To Kayunga MAIN GATED SIPHON EMBANKMENT Buyala Falls SPILLWAY SPILLWAY DAM Proposed Bujagali Dam & PROPOSED BUJAGALI DAM AND POWER STATION ORIGINAL OWEN FALLS DAM AND OWEN FALLS EXTENSION WAKISI Power Station AREA OF INUNDATED LAND, AND LAND ACQUIRED FOR flow Namizi Nile River TEMPORARY OR PERMANENT WORKS WORKSHOP & STORE Bujagali DUMBBELL BUJAGALI RESETTLEMENT SITE Kikubamutwe Bujagali Rapids POWERHOUSE ISLAND ACCESS ROAD TRANSMISSION LINES: switch yard Ivunamba 132kV LINES For details, 33/66kV LINE see inset. Dumbbell EXISTING SITE ROAD Island PROPOSED OWEN FALLS / TORORO / KAWANDA LINES Kyabirwa Falls POWERHOUSE Kyabirwa Kyekidde SUBSTATION BUJAGALI-TOROROLINE (2X132kV) BUJAGALI-NALUBAALELINE (2X132kV) DISTRICT CAPITALS (INSET) Nankwanga Malindi Bujagali Falls ToKayunga RIVERS NATIONAL CAPITAL (INSET) 132-220kV) MAIN HIGHWAY FALLS OR RAPIDS MAIN ROADS WANDA-BUJAGALI (2X 0 200 400 KA LINE To Jinja METERS SELECTED TOWNS/VILLAGES RAILROADS Buloba To Tororo COUNTY HEADQUARTERS FOREST RESERVES OwenTororo COUNTY BOUNDARIES Line TororoLine URBAN AREA flow Line Nakimbolede Falls Walumbe To Iganga INTERNATIONAL BOUNDARIES (INSET) Line and Nairobi Victoria B U T E M B E C O U N T Y 34° Kawanda 30° 32° Naminya SUDAN Nile Kimaka 4° 4° B U I K W E Forest Reserve Kikenyi To Kampala C O U N T Y Northern Line Bujagali tNile Owen Falls Extension Resettlement Site J I N J A Alber Original Owen Falls canal (Kiiva) (Kiira) C O U N T Y D.R. OF Victoria Nile Dam & Power Station CONGO Karuma Falls M a b i r a (Nalubaale) F o r e s t switch yard 2° Lake Kwania 2° flow Murchison R e s e r v e LakeAlbert Falls Lake Kyoga Luwero Victoria Nile Line Naava JINJA Bujagali Falls Kalagala Falls DANIDA Owen Falls Jinja To Kampala Area of KENYA 0 1 2 3 KAMPALA Mukono Ripon Napoleon map, left 0° 0° KILOMETERS Falls (inundated) Gulf Lake Edward To Kampala Lake Victoria To Kampala flow This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information L A K E V I C T O R I A shown on this map do not imply, on the part of The World Bank TANZANIA Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. RWANDA 32° 34° JULY 3, 2008 IBRD 36276 30° 32° SUDAN 34° K A G O M A C O U N T Y 4° 4° BUDONDO Buyala UGANDA WAKISI Namizi PRIVATE POWER GENERATION (BUJAGALI) PROJECT Kikubamutwe Bujagali PROJECT SITE Ivunamba D.R. OF Area of CONGO UGANDA main map Kyabirwa B U T E M B E Malindi C O U N T Y 2° Lake Kwania 2° Buloba Lake Albert Kimaka INUNDATED LAND Lake Kyoga EXISTING RIVER AREA Naminya Victoria Forest Reserve Luwero NEW HIGH WATER LEVEL Area of B U I K W E Nile map, right C O U N T Y WHEN RESERVOIR IS FILLED Jinja canal J I N J A PERMANENT FENCE LINE KAMPALA Mukono KENYA Mabira C O U N T Y 0° 0° Forest TRANSMISSION LINES Reserve SELECTED TOWNS/VILLAGES LakeLake PERMANENT ROADS EdwardEdward Lake COUNTY HEADQUARTERS URBAN AREA JINJA Victoria This map was produced by the Map Design Unit of The World Bank. MAIN ROADS The boundaries, colors, denominations and any other information Source: Bujagali Hydropower Social & Environmental RAILROADS shown on this map do not imply, on the part of The World Bank Assessment (2006) TANZANIA Group, any judgment on the legal status of any territory, or any FOREST RESERVES LAKE VICTORIA endorsement or acceptance of such boundaries. RWANDA 32° 34° COUNTY BOUNDARIES K A G O M A C O U N T Y BUTEMBE COUNTY MAIN GATED MAIN GATED SIPHONSIPHON EMBANKMENTEMBANKMENT Kyabirwa SPILL SPILLWAY SPILL SPILLWAY DAMDAM Falls Bujagali Falls Nile River WORKSHOPWORKSHOP & STORE STORE DUMBBELL POWERHOUSE POWERHOUSE ISLAND ACCESS ROAD ACCESS ROAD EXISTINGEXISTING SITE ROAD SITE ROAD POWERHOUSEPOWERHOUSE SUBSTATIONSUBST TION BUJAGALI-TORORO LINE (2X132kV) BUJAGALI-NALUBAALE LINE (2X132kV) To Kayunga Kayunga ANNDA-BUJAGALIDA-BUJAGALI 1132-220kV)32-220kV) B U K W E C O U N T Y UGANDA 0 200 400 MAIN HIGHWAY METERS KAKAW LINELINE((2X2X To JJinja inja JULY 3, 2008