India - Upper Indravati Hydro Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 15676 IMPLEMENTATION COMPLETION REPORT INDIA UPPER INDRAVATI BYDRO PROJECT (Loan 2278-IN and Credit 1356-IN) June 3, 1996 Energy and Infrastructure Operations Division Country Department II South Asia Region This document has a restricted distnibution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Rupee (Rs) = 100 Paise AVERAGE EXCHANGE RATES Indian Fiscal Year Historic Exchange Rates Consumer Price (April 1 to March 31) Rupees/US$ Index (1982=100) Official Unified Market April 1983 (SAR) 9.50 FY 1983/84 10.31 FY 1984/85 11.89 FY 1985/86 12.24 FY 1986/87 12.79 FY 1987/88 12.97 FY 1988/89 14.48 FY 1989/90 16.66 FY 1990/91 17.95 FY 1991/92 24.52 FY 1992/93 26.41 30.65 FY 1993/94 31.36 FY 1994/95 31.40 FISCAL YEAR OF BORROWER AND ORISSA April 1 - March 31 WEIGHTS AND MEASURES I Kilometer = 1,000 meters (m) = 0.6214 miles I Meter (m) = 39.37 inches (in) 1 Cubic Meter (m3) = 1.31 cubic yards (cu yd)=35.35 cubic feet (ft. 3) I Hectare (ha) = 10,000 square meters (m2) = 2.471 acres (ac) 1 Ton (t) = 1,000 kilograms (kg) = 2,200 pounds (lb) I Kilovolt = 1,000 volts (V) I Kilovolt-Ampere (kVA) = 1,000 volt-amperes (VA) 1 Megawatt (MW) = 1,000 kilowatts (kW) = I million watts I Kilowatt-hour (kWh) = 1,000 watt-hours I Megawatt-hour (MWh) = 1,000 kilowatt-hours 1 Gigawatt-hour (GWh) - 1,000,000 kilowatt-hours  FOR OFFICIAL USE ONLY ABBREVIATIONS AND ACRONYMS CEA Central Electricity Authority CMO Construction Management Organization CWC Central Water Commission DOE Department of Energy DOWR Department of Water Resources DP Displaced Person IBRD International Bank for Reconstruction and Development IDA International Development Association PAP Project Affected Person PFC Power Finance Corporation POE Panel of Experts RAP Rehabilitation Action Plan RRU Resettlement and Rehabilitation Unit OHPC Orissa Hydro Power Corporation OIPD Orissa Irrigation and Power Department OSEB Orissa State Electricity Board UIHP Upper Indravati Hydro Electric Project This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed wiLhout World Bank authorization.  II  IMPLEMENTATION COMPLETION REPORT INDIA UPPER INDRAVATI HYDRO PROJECT (Loan 2278-IN and Credit 1356-IN) Table of Contents PREFACE EVALUATION SUMMARY ....................................................................i PART I: PROJECT IMPLEMENTATION ASSESSMENT .................................................1 A. EVALUATION OF PROJECT OBJECTIVES ............................................................. 1 Sectoral Context .................................................1 Project Objectives .................................................1 Physical Objectives .................................................1 B ACHIEVEMENT OF OBJECTIVES . ............................................................1 Summary Implementation Record .................................................2 C MAJOR FACTORS AFFECTING PROJECT PERFORMANCE ...........................................3 Factors Not Subject to Government Control ................................................. 3 Factors Subject to Government Control .................................................4 Factors Subject to Implementing Agency Control .................................................8 D. SUSTAINABILITY . . .11 E. BANK'S PERFORMANCE .............. .............................................. 12 F. BORROWER'S AND IMPLEMENTING AGENCY'S PERFORMANCE ............................. 14 G. ASSESSMENT OF PROJECT OUTCOME ........................................................... . 15 H. FUTURE OPERATION ............. ................................................ 15 I. KEY LESSONS LEARNED ............................................................. 16  PART H: STATISTICAL ANNEXES Table 1: Summary of Assessments Table 2: Related Bank Loans/Credits Table 3: Project Timetable Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual Table 5: Studies Included in Project Table 6: Project Costs Table 7: Status of Legal Covenants Table 8: Bank Resources: Staff Inputs Table 9: Bank Resources: Missions Appendices: Appendix A. Mission's Aide Memoire Appendix B. Borrower's Contribution to the Implementation Completion Report Appendix C. Upper Indravati Project - Project Completion Issues  IMPLEMENTATION COMPLETION REPORT INDIA UPPER INDRAVATI HYDRO PROJECT (Loan 2278-IN and Credit 1356-IN) Preface This is the Implementation Project Report (ICR) for the Upper Indravati Hydro Project for which Credit 1356-IN in the amount of SDR 156.0 million equivalent and Loan 2278-IN in the amount of US$ 156.4 million equivalent were approved on May 10, 1983, signed on June 8, 1983, and made effective on September 9, 1983. Both the Loan and the Credit were granted to the Government of India which made available the proceeds to Government of Orissa in accordance with Government of India's standard arrangements for developmental assistance to the states of India. The IBRD loan was closed on December 31, 1991 compared with the original date of June 30, 1991. The only loan disbursement of about US$ 0.4 million for payment of the front-end fee to the Bank was made in the third quarter of 1983. The IDA credit was closed on June 30, 1995 compared with the original date of June 30, 1991. The credit was fully disbursed, and the last disbursement took place on March 30, 1995. There was no cofinancing for the project. The ICR was prepared by the Energy Operations Division (SA2EI), reviewed by the Jean- Francois Bauer, Division Chief, SA2EI, and Mrs. Kazuko Uchimura, Project Advisor, SA2DR. The Orissa Hydro Power Corporation (OHPC) and the Department of Energy, Government of Orissa, provided comments that are included as Appendix B. The project is still to be completed. Appendix C highlights the key issues to be addressed by OHPC and the Orissa government to help ensure successful project completion.   IMPLEMENTATION COMPLETION REPORT INDIA UPPER INDRAVATI HYDRO PROJECT (Loan 2278-IN and Credit 1356-IN) Evaluation Summary Project Objectives 1. According to the Staff Appraisal Report (SAR), the project has two main objectives: (i) to provide a substantial block of low cost power and energy to help meet the growing electricity needs of the state of Orissa, and (ii) to provide water to irrigate 109,000 ha of cultivable land. The downstream irrigation project to help utilize the water provided by the project was treated as a separate project by Orissa and the Bank. It received financial support from the Overseas Economic Cooperation Fund of Japan and was not appraised by the Bank. Evaluation of Objectives 2. The project objectives continue to be valid, but they have not yet been achieved, as neither the reservoir or the power complex are completed. The downstream irrigation project has not been completed either. The SAR also stated that the Bank's strategy in the Indian power sector included strengthening the finances of sector institutions, particularly SEBs. Agreements reached included financial objectives for Orissa State Electricity Board (OSEB) and actions to improve OSEB's financial and operational performance (paras. 38-42). However, there was very little progress in improving OSEB's overall financial performance till 1993 when it was decided to proceed with a sweeping institutional reform of the Orissa power sector. 3. Although the SAR expected that the tribal people who form most of the population in the area where the project is located will develop economically and socially both during the construction period; and after the project went into operation, the project did not have any explicit social objectives for the affected population. Implementation Experience and Results Implementation Schedule 4. Construction work on site was started in 1979, but by project appraisal by the Bank in late 1982 only some preliminary work had been completed. The original implementation schedule - commissioning of the first unit in late 1988 - was ambitious and the project experienced substantial delays in implementation right from the start. It was two years behind schedule by 1986, largely due to delays in procurement. A further year had been lost by late 1990, due to the  - ii - slow pace of construction. Several key project components suffered substantial damage in a severe flood in 1991 which halted construction. In March 1995, when the final disbursement from the IDA Credit took place the project was only two-thirds complete. There has been little progress since then due to lack of funds to continue construction. The completion of the Upper Indravati project was threatened by the lack of funds up to March 1996, when OHPC secured Rs 3,200 million completion financing from the Power Finance Corporation (PFC). 5. With hindsight, it is clear that the original implementation schedule agreed with the Bank during project appraisal was overly optimistic. For a project of this magnitude and complexity, situated in a remote area, such a tight schedule was not realistic. It did not take account of the time required to prepare tender documents acceptable to the Bank, the risk of encountering geotechnical problems, and the capabilities of local civil works contractors. In conclusion, it must be noted that in spite of the extra-ordinary problems and delays, the project is still projected to provide power at a highly attractive rate (at less than 50 per cent of the cost of electricity from new base-load thermal stations) and remains a priority component in the Orissa/Eastern region power system. OHPC's challenge is to realize this potential by completing the Upper Indravati project, for which it now has the necessary funding. Key Factors Affecting Implementation Project Management 6. The project was implemented by the Orissa state government. The Orissa government was too involved in direct management of the Upper Indravati project. Project management at site lacked autonomy and decision making authority. At the same time, it was never provided with satisfactory communications infrastructure, despite its remote location. This compounded the delays arising from the high degree of the Orissa government involvement in decision-making. The Orissa government's split of the Irrigation and Power Department in 1990 diffused responsibilities and further weakened project management (para. 21). 7. In 1991 the delegation of full responsibility for project management to an international consulting firm became a condition for lifting the suspension of disbursements and a new Construction Management Organization (CMO) was set up in 1993. The CMO was in charge of the project for about two years from mid- 1993 onwards. There does not appear to have been an appreciable increase in the pace of implementation during this period (para. 24). The CMO was unable to effectively carry out the objectives defined in the TOR as, inter alia, the Orissa government did not give wholehearted support to the new arrangements. It also appears that working relations between project staff and consulting firm staff were not ideal for the success of the CMO. The 1991 Flood and Subsequent Fund Constraints 8. Apart from the loss of 25 construction workers' lives, the physical damage and resulting repairs were a major setback to the project both in terms of time and cost (para. 34). It also gave rise to serious concerns within the Bank about the quality of engineering and construction and the safety of the dams under construction. Since substantial disbursements could be justified only  - iii - after a safety and engineering soundness evaluation and the project reorganization and reactivation, the Bank canceled the still completely unused US$156 million Ln. 2278-IN in December. The IDA Credit was extended to finance resettlement, consulting services for the soundness assessment and project management and some civil works, following confirmation of the Project's soundness by the external engineering consultant review. Cr. 1356-IN was depleted in 1995. The completion of the Upper Indravati project was threatened by the lack of funds up to March 1996, when OHPC secured completion financing from PFC. Changes in Project Costs and Financing 9. The final cost of the project is not yet known. It was initially expected to cost Rs 4,800 million (in 1982 prices) or about US$ 507 million at the prevailing exchange rate. The latest OHPC estimate is Rs about 11,200 million (in current early-1996 prices). As this estimate or year-by-year expenditures were not available at the time of the ICR mission, a detailed constant price comparison could not be made, but the total cost of the project is expected to be well below US$500 million. About Rs 6,500 million has so far been spent and the balance, about Rs 4,700 million (about US$ 135 million), represents the unfinished portion of the project. Bank Performance 10. There were weaknesses in the quality of project appraisal and supervision was inadequate in the pre-1987 period. Supervision intensified from 1987 (and was extended to resettlement), but had little impact on Orissa's project implementation performance. The Bank could have been firmer in seeking measures to redress deficiencies in project implementation prior to 1991. As can be seen from the three instances when the Bank desisted from formally suspending disbursements (para. 48), there was a tendency to accept the Orissa government assurances of imminent action in place of real measures. It is apparent that the Bank's pre-1991 patience and tolerance of shortcomings in implementation performance did not help the narrow objective of completing the project as rapidly as possible. It is also clear, however, that the Bank's new approach pursued from 1991, inter alia pressuring Orissa to address project management and resettlement deficiencies by canceling Ln. 2278-IN and withholding completion financing, has not yet yielded the desired results either. Time will tell if OHPC, the product of a Bank-supported power sector reform program in Orissa, will be able to successfully complete the project. Borrower Performance 11. The performance of the Borrower was found to be deficient in project design, management arrangements, and implementation. There was - and still today remains - a general reluctance to recognize that Orissa's project team did not have the management and engineering skills required for such a large and complex project. The fact that Orissa had in the past and was at the same much more successfully implementing similar projects contributed to the unwillingness to accept and utilize external assistance for the project. The cost of delays in terms of foregone revenues or additional costs was given little importance. The Orissa government showed little commitment to improving OSEB's performance during the ten years prior to 1993. However, since 1993 there have been positive indications and firm demonstration of a new resolve to pursue fundamental sector reforms (paras. 44-45).  - iv - Sustainability 12. Since the project is unfinished, it is not possible to draw final conclusions on its sustainability. If completed satisfactorily, it is likely that the project will generate power and provide water for irrigation. In spite of the extra-ordinary problems and delays, the project is still projected to provide power at a highly attractive rate (at less than 50 per cent of the cost of electricity from new base-load thermal stations) and remains a priority component in the Orissa/Eastem region power system. OHPC's challenge is to realize this potential by completing the Upper Indravati project, for which it now has the necessary funding. Assessment of Outcome 13. Upper Indravati is an unfinished project. Meanwhile repayment of the IDA Credit has already begun. This is an unsatisfactory interim outcome. However, this assessment will have to be revisited upon physical completion of the project. In spite of the extra-ordinary problems and delays, the project is still projected to provide power at a highly attractive rate and remains a priority component in the Orissa/Eastern region power system. Far-reaching reforms to the power sector as a whole have begun, and may prove to have highly positive effects. As a part of the reform program, the Upper Indravati project has been transferred from the state government to OHPC. Its challenge is to complete the project, for which it now has the necessary funding. Future Operations and Key Lessons Learned Future Operations 14. Responsibility for the project has been formally handed over by the state government to OIPC. Once completed, OHPC will also operate the project. OHPC is operating Orissa's other hydro projects and is expected to be able to operate the Upper Indravati as well. A complete range of institution building support is being provided to OHPC under the Orissa Power Sector Restructuring Project. Key Lessons Learned * The design and quality of project management arrangements is probably the single most important element for successful and timely project implementation. The Bank should not agree to support projects where such fundamental conditions are not satisfied. * Dependence on design and engineering consultants such as Central Water Commission (CWC) who are not answerable to the project owner should be avoided. * Important engineering issues such as anti-sedimentation measures should be resolved at the time of project appraisal or as soon as possible thereafter. * Competitive bidding can result in award of contracts to public sector firms who submit unrealistically low bids. More rigorous prequalification and bid evaluation may help to prevent this.  v The successful use of consultants for project management requires a genuine commitment on the part of the implementing agency and/or project owner, and cooperation by the consultants' counterparts on the ground. Lack of such support will undermine even the best of efforts by consultants. Adequate resources should also be provided. Apparent cost savings achieved by reducing expatriate staff and introducing lower cost but less than fully qualified local support will be over-shadowed by implementation delays and difficulties. This project is the latest addition to a generally unsuccessful series of Bank operations in the Indian power sector at the state level, with state electricity boards and their state govermments. But in this case the prospects for addressing the fundamental management and financial problems - and successfully completing the Upper Indravati project - are encouraging. Since 1992 the Government of Orissa, OSEB and the Bank have discussed the issues and possible solutions to Orissa's fundamental power problems. In 1993, the Government of Orissa and OSEB agreed upon a power sector reform program as the logical conclusion to these deliberations. The program involves privatization of power distribution and separation of the power sector from direct government control through regulatory reforms, corporatization, commercialization and private sector participation. The Bank approved on May 14, 1996, the Orissa Power Sector Restructuring Project (Ln. 4014-IN) to support Orissa's pioneering reforrn program.  I  PART I: PROJECT IMPLEMENTATION ASSESSMENT A. EVALUATION OF PROJECT OBJECTIVES Sectoral Context 1. The project was appraised at a time of severe power shortages in Orissa, particularly for industrial consumers who accounted for 85% of consumption. These shortages, exceeding 20% of total demand in recent years, have remained a semi-permanent feature of the power sector in Orissa, even though electricity sales have more than doubled to about 6,000 GWh in 1994. The -number of consumers has risen from under 0.6 million in 1982 to over 1.1 million. Project Objectives 2. The SAR states that the project has two main objectives: First, it will provide a substantial block of low cost power and energy to help meet the growing electricity needs of the state of Orissa and the Eastern Region of India. Second, the project will take waters from the Indravati River in the Godavari Basin, and direct them into the Hati River in the Mahanadi Basin, thus providing water for 109,000 ha of cultivable land for which there is no other feasible source of irrigation. The downstream irrigation project to help utilize the water provided by the project was treated as a separate project by Orissa and the Bank. It received financial support from the Overseas Economic Cooperation Fund of Japan and was not appraised by the Bank. Physical Objectives 3. The Upper Indravati Hydroelectric Project is located in southwest Orissa. Its physical objectives were the construction of four dams, ranging in height from 40m to 73m and eight dikes, two link channels within the 110 sq. km. reservoir, an approach channel and an intake structure, an emergency irrigation outlet, a headrace tunnel and surge tank, penstocks, a 4 X 150 MW power plant, and a 9 km tailrace channel and diversion structure. The average annual power generation from the project is projected to exceed 1,900 GWh. B. ACHIEVEMENT OF OBJECTIVES 4. The project objectives continue to be valid, but they have not yet been achieved, as neither the reservoir or the power complex are completed. The downstream irrigation project has not been completed either. The economic rate of return (excluding irrigation benefits) was estimated to be 12% at the time of appraisal, but has not been recalculated given that the project is unfinished. However, the project remains attractive as the cost of power generation from it is estimated at about Rs 1.0/kWh against Rs 1.8 to Rs 2.2/kWh from new base-load coal thermal plants. Given that Upper Indravati will provide peaking capacity, which is of even higher economic value, there is a strong case for completion as rapidly as possible.  5. The SAR also stated that the Bank's strategy in the Indian power sector included strengthening the finances of sector institutions, particularly SEBs. Agreements reached included financial objectives for OSEB and actions to improve OSEB's financial and operational performance (paras. 38-42). However, there was very little progress in improving OSEB's overall financial performance till 1993 when it was decided to proceed with a sweeping institutional reform of the sector. 6. Although the SAR expected that the tribal people who form most of the population in the area where the project is located will develop economically and socially, both during the construction period and after the project goes into operation, the project did not have any explicit social objectives for the affected population. Resettlement issues were treated rather cursorily in the SAR, although this may not have been unusual for projects appraised in the early 1980s. 7. After a late and unsatisfactory start (paras. 30-32), the resettlement program was completed successfully. It has also led to improvements in overall resettlement practices in Orissa and the higher standards used in the project will become the norm for future resettlement. The successful outcome can be attributed to three factors: political support in Orissa, dynamic leadership and dedicated staff of the Resettlement and Rehabilitation Unit (RRU), and constant pressure from the Bank to improve the program. The RRU prepared a new cluster-based family- oriented rehabilitation action plan (CBFORAP) in 1995 for further improvements. The implementation and success of the CBFORAP will have to be assessed after an interval of several years. Summary Implementation Record 8. Project preparation by the Bank began in January 1981 and pre-appraisal took place in September 1981. Work had already begun on access roads and preliminary excavation of the major works. A second pre-appraisal took place in early 1982 and included an independent review of project design by three external consultants, who were required to confirm the safety, cost-effectiveness, and predicted performance of the project's major structures. 9. The project was appraised in October 1982 by a three-person Bank team, even though the revisions to the construction schedule and cost estimate arising from the experts' review were not complete. The final implementation schedule was reviewed during technical discussions preceding loan negotiations in March 1983 and was found to be reasonable and achievable. No resettlement expert or environmentalist participated in appraisal. These aspects of the project received superficial coverage until late 1986. 10. At negotiations in March 1983, the Indian delegation provided Bank staff with a note on the Government of Orissa's rehabilitation policy for displaced persons and this was found to be satisfactory. The 1979 Government of India environmental clearance was also furnished at negotiations and Bank staff found it satisfactory. 11. In May 1983, the Board simultaneously approved a loan for US$ 156.4 million and an IDA Credit of SDR 156 million, together representing 68% of the project cost. Formalities for loan/credit effectiveness were completed by early September 1983.  - 3 - 12. The IBRD loan was canceled in December 1991, with only the front-end fee having been disbursed. The IDA Credit was fully disbursed in March 1995 and closed in June 1995 after four extensions. 13. The project immediately suffered delays in implementation. By late 1984, it was a year behind schedule and two years behind schedule by 1986, largely due to delays in procurement. A further year had been lost by late 1990, due to the slow pace of construction. These delays cannot be considered as exceptional, given the context (para. 26). Several key project components suffered substantial damage in a severe flood in 1991 which halted construction. After an assessment of the time needed to repair the flood damage, it was estimated that the first unit could be brought into service in late 1995, i.e., seven years late. This also proved to be infeasible. In March 1995 when the final disbursement from the IDA Credit took place, the project was over two-thirds complete. There has been little progress since then due to lack of funds to continue construction. The completion of the Upper Indravati project was threatened by the lack of funds up to March 1996, when Orissa secured Rs 3,200 million completion financing from PFC. 1998 would appear to be a feasible target date for completion. In spite of the extra- ordinary problems and delays, the project is still projected to provide power at a highly attractive rate (at less than 50 per cent of the cost of electricity from new base-load thermal stations) and remains a priority component in the Orissa/Eastern region power system. OHPC's challenge is to realize this potential by completing the Upper Indravati project, for which it now has the necessary funding. C. MAJOR FACTORS AFFECTING PROJECT PERFORMANCE Factors not Subject to Government Control 14. Design. Prior to appraisal, the Bank's consultants who reviewed the project were concerned about the risk of sedimentation as the water intake structure to the headrace tunnel was close to the point at which the Indravati River enters the reservoir. After appraisal, it was decided that studies of sedimentation (and a few other issues) as "less urgent matters having relatively little effect on the cost estimate or the construction schedule will be pursued as work proceeds". However, there did not appear to have been an attempt to reconsider the location of the powerhouse or the headrace tunnel (perhaps this was regarded to be no longer an option). This was confirmed as a major flaw in project design in the 1993 assessment of the project's engineering and construction. It apparently arose from the attempt to minimize the length of the headrace tunnel and thereby keep down capital costs. Remedial structures have since been planned to protect the intake so as to minimize the entry of abrasive material which would otherwise damage the turbines. They remain to be examined and designed using a physical model - and their actual effectiveness cannot be confirmed till these structures and the power plant are actually in operation. 15. Engineering Consultants. The Central Water Commission (CWC) was the Government of Orissa's engineering consultant for all civil works. However, the Orissa government had no control or leverage over CWC as it is a central Government agency. Inadequate and delayed engineering services, particularly the delivery of design drawings by CWC also contributed to the  slow pace of project execution. In mid-1987, CWC indicated that it was overloaded and could not provide more support to the project. CWC staff were reluctant to be posted to a remote area and visits from Delhi were infrequent as only rail travel (which took over 2 days) was authorized. Despite repeated efforts during 1987-90 on the part of the Orissa government and the Bank to obtain better service, CWC was unable to give satisfaction in this area. The Bank then recommended replacement of CWC and offered to finance the services of other consultants in place of CWC, but apparently at that time Government of India regulations only allowed CWC or CEA to be employed as consultants on Government-owned projects. Today, CWC feels that it was being asked to provide support to the project that was beyond its mandate and for which it did not have adequate resources. However, it does not appear to have made this clear to the Orissa government or to the Bank at the time. 16. Poor Contractor Performance. The decision by the Orissa government (accepted by the Bank) to award contracts to firms who had submitted unrealistically low bids for construction of Muran Dam, intake and headrace tunnel, penstocks and the tailrace channel also contributed to the slow pace of work prior to the 1991 flood. Most of the poorly-performing contractors were public sector undertakings and apparently suffered from lack of funds to carry out the works at an acceptable rate. Some had little equipment and were little more than agents to manage unskilled manual laborers. At appraisal, the Bank had accepted that bidders would be free to adopt whatever construction method they chose. This was not conducive to ensuring rapid and efficient construction. 17. In early 1986, the Orissa government had proposed disqualification of the lowest bidder for the headrace tunnel on the grounds of unsatisfactory performance in other tunnel projects and on the basis of being overloaded with too many other contracts. However, the Bank rejected this proposal because the bidder had been pre-qualified and the Orissa government had not provided sufficient documentary evidence to support these claims. In late 1989, the Bank suggested terminating the contract as the Orissa government's reservations about the contractor proved to be justified, although by now the Orissa government was reluctant to do so. Factors Subject to Government Control 18. Implementation Schedule. With hindsight, it is clear that the original implementation schedule agreed with the Bank during project appraisal was overly optimistic. It envisaged commissioning of the first unit in late 1988, barely five years after the credit/loan became effective. For a project of this magnitude and complexity, situated in a remote area poorly served by transport and communications, such a tight schedule was not realistic. It did not take account of the time required to prepare tender documents acceptable to the Bank, the risk of encountering geotechnical problems, and the capabilities of local civil works contractors. The Orissa government's own recent experience with the construction of the Balimela, Upper Kolab and Rengali hydro projects, which also suffered implementation delays, should have served to inject greater realism on all sides at the time of appraisal. 19. Changes in Project Costs and Financing. The final cost of the project is not yet known. It was initially expected to cost Rs 4,800 million (in 1982 prices) or about US$ 507 million at the prevailing exchange rate. The latest OHPC estimate is Rs about 11,200 million (in current early-  - 5 - 1996 prices). As this estimate or year-by-year expenditures were not available at the time of the ICR mission, a detailed constant price comparison could not be made, but the total cost of the project is expected to be well below US$500 million. 20. About Rs 6,500 million has so far been spent and the balance, about Rs 4,700 million (about US$ 135 million), represents the unfinished portion of the project. OHPC secured in March 1996 Rs 3,200 million completion financing from PFC and will provide the balance from its internal cash generation. OHPC's challenge is to complete the Upper Indravati project, for which it now has the necessary funding. 21. Institutional Arrangements. In mid- 1990, the new Orissa state government decided to split the Irrigation and Department into two. Responsibility for the project was to be shared by two new departments, Energy (DOE) and Irrigation (DOI). The Bank felt that this created coordination problems which delayed decision-making and further complicated project management, and urged the Orissa government to place responsibility for the project under a single administrative unit, namely the Department of Energy, with the Department of Irrigation assuming full and single responsibility of the downstream irrigation project. The Orissa government agreed to do so in August 1991, but implementation of the decision did not take place - formally till late- 1992 and never in practice - as DOI was reluctant to give up its role. The decision to split the Irrigation and Power Department and the ensuing uncertainty had an adverse impact on the project. Many project staff were not legally transferred to DOE, but remained DOI staff temporarily on deputation to the project. The exclusion of DOI from any formal involvement with the project may also have given rise to difficulties within the Orissa government. 22. Setting up of the Construction Management Organization (CMO). The process of reaching agreement on terms of reference, bidding, evaluation of proposals and contract negotiations took about two years from mid-1991. This length of time was mainly due to the innovative and somewhat controversial approach that was adopted on advice from the Bank, whereby consultants were directly responsible for project implementation. However, many DOE/staff did not genuinely endorse this approach, even after it was introduced. The fact that Orissa had in the past and was at the same more successfully (though with time and cost overruns) implementing similar projects contributed to the unwillingness to accept and utilize external assistance for the project. While the induction of consultants was intended to improve project management and the quality and pace of implementation, the selection process was lengthy due to the Orissa government reluctance. It also diverted efforts away from project implementation itself during a period that was already disrupted by the bifurcation of the Department of Irrigation and Power. 23. The CMO was nominally in charge of Upper Indravati Hydroelectric Project for about two years from mid-1993 onwards. There does not appear to have been an appreciable increase in the pace of project implementation during this period. The CMO was unable to effectively carry out the objectives defined in the TOR. However, the large backlog of outstanding claims by contractors was disposed of, project documentation was improved and DOE delegated some authority to the counterpart Owner's Project Management Organization to facilitate decision- making.  - 6 - 24. There are numerous reasons why the CMO did not live up to expectations. The Orissa government did not give adequate support to the new CMO. The Orissa government Project Administrator did not make available sufficient resources required by the CMO to carry out the tasks specified in the TOR. Fee payments to the consultants were unduly delayed and staff posted to the project were slow to receive logistical assistance. The eleven month interval between the receipt of proposals and contract signature meant that a large proportion of consulting staff originally identified for the assignment were no longer available. It also appears that working relations between project staff and consulting firm staff were not ideal for the success of the CMO. The arrangement between the foreign consulting firm and a local partner which was viewed positively by the Orissa government at the time of evaluation of the proposals (as a means of technology transfer and cost minimization), also proved to be problematical due to the public sector background and caliber of the local consulting firm staff. DOE staff found that the consultants were often obliged to refer matters back to their headquarters for review/approval and that the response time was sometimes inordinately long. 25. Panel of Experts The Project Agreement only required Orissa to carry out periodic safety inspections of the major civil works after construction, although the SAR indicated that a Board of Consultants had been appointed to review designs and provide advice during construction. The first meeting of this Board, hereafter called the Panel of Experts (POE), took place in September 1984. It has met on twenty occasions or roughly every six months since then. The composition of the POE has changed extensively during the past 11 years, although the same Chairman has presided throughout. The Bank was dissatisfied with the choice of members in the early years as it was dominated by retired Central Water Commission/Central Electricity Authority (CEA) staff. This was felt to be incompatible with the need to ensure objectivity in the POE's scrutiny of work done by CWC and CEA as the Orissa government's consultants. In addition, most POE members were generalists and certain highly specialized disciplines were not represented on the panel. Under Bank pressure, the Orissa government agreed to appoint internationally known specialists to strengthen the POE in areas such as engineering geology, foundation engineering and sedimentation. This proved to be a positive measure, enhancing the quality of the advice given by the POE. However, on several critical issues like remedial measures to control the entry of sediment to the power plant, the POE was slow to address the problems and to reach a consensus on required actions. The sedimentation issue was first examined by the POE in 1987, but its final recommendations on action to be taken were not issued till 1995. The POE should also have looked into the adequacy of the coffer dam upstream of the water intake that gave way in 1991, especially as its design was not the responsibility of CWC. 26. Procurement Delays. The Orissa government's own procurement process was slow. Additional time was lost by repeated reviews and revisions of bid documents and/or evaluations before obtaining World Bank clearance to proceed with tendering or contract award. For example, over nine months elapsed between receipt by the Bank of draft tender documents for three dams and a tunnel and granting of the final clearance after multiple revisions had been made. This was initially due to unfamiliarity with Bank procurement procedures and difficulties in communication with the project site where no telephone or radio link had been set up, despite repeated Bank reminders. Bank suggestions that consultants be hired to assist in preparation of bid documents were not acted upon. Nor was there a policy to use Bank-approved standard  bidding documents. The quality of bid evaluation reports produced by the Orissa government was poor and some required modification before Bank approval was granted. In later years, delays may have been aggravated by changes in project management/procurement staff, with a resulting loss of familiarity with Bank requirements. On only one occasion, the time required for Bank clearance was unusually long, due to a delay in replacing the task manager. 27. Resettlement Planning and Implementation. Although OMS 2.33, which was issued in 1980, specifies that detailed resettlement planning is required before the negotiation of the project loan, in the case of this project no such planning was carried out till much later. The Bank only decided to seek an assurance in the project agreement requiring Orissa to formulate a resettlement plan at least two years before submergence and to implement it thereafter. 28. The first Bank resettlement specialist did not visit the project site till late 1986 and found that no Rehabilitation Action Plan (RAP) had been prepared. The first RAP was prepared in mid- 1987, and after Bank review a second one was prepared in 1988, but the Bank's review of this RAP still found it to be unsatisfactory as compensation terms were excessively cash-based. Nevertheless, a first phase of resettlement was carried out by the Orissa government in 1989. 29. The Resettlement and Rehabilitation Unit (RRU) was finally set up in late 1989, although the Bank had been urging the creation of such a unit since late 1986. A third RAP was issued in mid-1991, but this also failed to meet the Bank's approval, and disbursements were suspended on June 28, 1991. Finally, in early 1992, after nearly 70% of the Displaced Persons (DPs), had already been resettled in three annual phases, the Bank formally approved the fourth RAP. The financial value of the resettlement package was increased to Rs 40,600 per DP in 1993 and retroactive entitlement was granted to those resettled in previous years under less favorable conditions. Progress since then has been satisfactory. 30. Results of Resettlement. In total, the resettlement of 5,448 DPs from 65 villages to be submerged took place over the 1989-93 period. This total actually represents nearly 17,000 individuals because the term DP as defined by the Orissa government is an amalgam of individuals and families. In addition to nuclear families with minor children, it includes male children over 18, orphans, divorcees, widows and the handicapped, all of whom are counted as if they were a separate family. The DPs settled in 560 widely scattered locations they themselves selected, often for proximity to their relatives and/or access to forests. Only 166 cluster villages containing 10 or more DPs were set up. The package of relocation assistance was progressively improved during that period and the improved terms were made available retroactively to those resettled in the early years. Initially, the Orissa government offered DPs land in an area far from the affected area. Following overwhelming refusal by DPs of these terms, the Orissa government agreed to allow the DPs to relocate to sites of their own choice and provided funds for them to purchase land. Resettlement took place entirely voluntarily and peacefully and with the active support of the RRU which provided transport, food and medical care during the moves. The quality of construction of the new homes of DPs and other basic infrastructure is much better than those found in the host villages. 31. A program of social infrastructure in all cluster villages has been completed. 194 tube wells, 42 open wells 39 tanks, 28 school buildings and 72 link roads have been provided. The  - 8 - provision of fuel-efficient stoves is underway. There has been a dramatic increase in land ownership amongst DPs. Today, 5212 DPs (96%) own 4929 acres of irrigated land and 8049 acres of unirrigated land, equivalent to an average of 3.4 acres of unirrigated land per DP. Also, 4119 (76%) now have holdings above the Orissa government norm of 2.5 acres of unirrigated land. Prior to resettlement, only 14% of DPs owned land. Nevertheless, over 40% of DPs also continue to cultivate other land under various tenurial arrangements. Those DPs who have resettled close to their former land holdings also continue to cultivate them as these lands have yet to be submerged. Unfortunately, the choice of locations in which the vast majority of DPs have resettled means that they will never have access to the irrigation benefits from the project. 32. The lack of a base line survey of the DPs prior to resettlement makes it almost impossible to assess changes in their standards of living since displacement. The project affected area is in one of the least developed regions of Orissa and nearly 60% of the DPs belong to either scheduled castes or tribes, who are the poorest categories of the general population. It can, therefore, be reasonably assumed that the majority of DPs were below the government-defined poverty line of per capita income of Rs 2,800 per annum prior to resettlement. A recent socio-economic survey of DPs has revealed that over 60% of DPs continue to supplement their incomes from cultivation by the collection and/or sale of forest products. Despite continuing access to forest resources, the survey estimated that 70% are currently below the poverty line, as against 44% of the state population as a whole. While these data provide no clue as to changes since resettlement, they provide a benchmark to assess future changes. The RRU prepared a new cluster-based family- oriented rehabilitation action plan (CBFORAP) in 1995 improve the living standards of DPs. It is based on the individually-expressed needs of the DPs. The RAP will provide DPs with the resources to expand income generating activities. It will also have an explicit component to assist women as they were disproportionately affected by the reduced access to forest products and by the lack of access to resettlement funds which were channeled exclusively to men. The implementation and success of the CBFORAP will have to be assessed after an interval of several years. Factors Subject to Implementing Agency Control 33. Inadequate SafetyMeasures. The aide-memoire ofthe Bank supervision mission of October 1990 called to the attention of the project authorities, CWC and POE, that arrangements are to be made before June 1991 against risk of water entering the headrace tunnel during the next floods. However, these arrangements were not implemented in time to avoid the disaster which occurred when the uncompacted earth cofferdam protecting the intake and headrace tunnel gave way in July 1991, under pressure from unusually heavy monsoon rains. Water surged down the penstock slope and through the headrace tunnel and the sediment and boulder-laden flood ripped through the powerhouse. Warehouses containing electrical and mechanical equipment were flooded and some equipment was damaged beyond repair. The flood brought work on the power complex to a halt.  34. Apart from the loss of 25 construction workers' lives, the physical damage and resulting repairs were a major setback to the project both in terms of time and cost. A further two-year delay in project completion can be attributed to the flood. 35. An additional Rs 100 million (US$ 3 million) had to be spent to repair uninsured civil works. The cost of repairs to the electromechanical equipment (which had to be sent back to the manufacturer in Japan) was estimated to be Rs 280 million (US$ 9 million). The Bank, which had financed the initial purchase of the equipment, agreed in principle to fund these repairs as insurance cover was inadequate. However, the contract was signed without prior Bank clearance, so IDA was unable to reimburse the Orissa government for the cost of the repairs. An insurance claim of Rs 590 million (US$ 19 million) has been filed for the damage to electromechanical equipment. However, the actual reimbursement is likely to be far less, as the assets were not insured at replacement cost and there has been a very sharp appreciation of the yen since the original equipment was purchased. 36. Impact on Project Implementation. The flood damage gave rise to serious concerns within the Bank about the quality of engineering and construction and the safety of the dams under construction. It was a major factor influencing the decision to cancel the IBRD loan from which no funds had yet been drawn. Since substantial disbursements could be justified only after a safety and engineering soundness evaluation and the project reorganization and reactivation, the Bank canceled the still completely unused US$156 million Ln. 2278-IN in December. The IDA Credit was extended to finance resettlement, consulting services for the soundness assessment and project management and some civil works, following confirmation of the Project's soundness by the external engineering consultant review. It was also agreed that four ongoing civilmechanical works contracts would be terminated and the remaining balance of works be clubbed together in a single package suitable for ICB to attract reputable international firms and to permit World Bank financing. However, without consulting the Bank, the Orissa government decided to separate the penstock erection from the rest and directly awarded a contract to its own construction corporation, which the Bank felt lacked adequate expertise in this area. This decision undermined the single-contract balance of works approach. It also left commissioning of the power plant vulnerable to undue risk of further delay arising from the inexperience of the contractor in penstock erection. At that point, the Bank could have canceled a sum equivalent to the value of this contract from the IDA Credit on grounds of misprocurement but this was not done. Instead, the whole Ln. 2278-IN was canceled. 37. The ICB process for the remaining components in the balance-of-works contract took over two years, an excessive length of time, despite assistance from external consultants. Work on these project components could not continue once the contracts of the poorly-performing contractors had been terminated. The additional implementation delay arising from the inability to carry out the bidding process in a timely way illustrates the lack of urgency to take remedial action to complete the project. By the time the contract was ready to be awarded the project had run out of funds with Cr. 13 56-IN having been depleted - and a procurement technicality would have prevented Bank financing of the contract in any case. OHPC secured in March 1996 Rs 3,200 million completion financing from PFC. OHPC's challenge is to complete the Upper Indravati project, for which it now has the necessary funding.  - 10 - 38. OSEB Finances. At appraisal, it was envisaged that OSEB would take over and operate the project upon completion. A full financial analysis of OSEB was carried out by the Bank and several financial covenants on OSEB's performance were incorporated in the Project Agreement. During the past twelve years OSEB was rarely - up to very recently - able to comply with these covenants. 39. Measures to strengthen OSEB's accounting suffered a four year delay as the new uniform commercial accounting system expected in 1985 was introduced in 1989. There was an average two year delay in submitting OSEB's audited accounts in the mid-1980s. The delay was brought down to 18 months in mid 1990s. FY94 accounts were submitted six months late. Unaudited statements for FY95 were received in December 1995. 40. OSEB was to limit its accounts receivable to three months billings. This target was not achieved till 1995 - by writing off arrears accumulated over the years finally admitted to be uncollectable. During the decade prior to 1994 they were always in excess of four months and had risen to over six months in mid-1992. This inevitably had a serious impact on OSEB's cash flow, with the result that its payables for power purchases at that time were equivalent to 24 months purchases. 41. OSEB was to make a contribution to investment of at least 20% of average annual capital expenditures; it was not achieved from FY84 through FY89 when OSEB incurred losses as tariff increases failed to keep up with inflation. The target was met in FY90, but only because OSEB had a very low level of investments and in FY94, with the help of the payment of a large subsidy by the Orissa government to cover uncollectible receivables. The covenanted level was exceeded in FY95 and is estimated to be exceeded also in FY96. Investment levels, however, have remained well below what is projected to be required to rehabilitate the system and meet future needs. 42. At appraisal, it was apparent that OSEB's energy losses at about 20% of generation were unacceptably high and that metering was inadequate. A loan covenant required OSEB to prepare a program to control energy losses and to implement it from FY 84-85. The Bank's financial projections were based on a reduction of losses to 16%. The first plan to reduce Transmission and Distribution (T&D) losses was found to be unsatisfactory and stated energy losses rose to 24% in 1987. A second plan was made in 1989, but had no impact on losses. Then in 1992, the Bank realized that the situation was far worse than previously believed as OSEB's auditors estimated that losses in FY90/91 were 45%, rather than OSEB's own figure of 24%. These huge losses can be attributed to the Orissa government policy of granting unmetered life-line supply to poor consumers, flat rate billing by OSEB of customers with defective meters, and increased pilferage. The difference between these two estimates represents over US$ 50 million in foregone revenues. In late 1993, a new action plan was proposed and an urgent need for 0.5 million new meters (approximately half of OSEB total) was identified. These will be provided - and the Orissa power system rehabilitated - under the Orissa Power Sector Restructuring Project.  D. SUSTAINABILITY 43. Since the project is unfinished, it is not possible to draw final conclusions on its sustainability. If completed satisfactorily, it is likely that the project will generate power and provide water for irrigation. In spite of the extra-ordinary problems and delays, the project is still projected to provide power at a highly attractive rate (at less than 50 per cent of the cost of electricity from new base-load thermal stations) and remains a priority component in the Orissa/Eastern region power system. OHPC's challenge is to realize this potential by completing the Upper Indravati project, for which it now has the necessary funding. Power Sector Reforms 44. In 1992, the Orissa government took steps to improve OSEB's performnance through an Operational and Financial Action Plan (OFAP) drawn up with assistance from PFC. Later that year, the Bank offered to finance consultants to assist the Orissa government with a review of the organization of the power sector. During supervision of the Upper Indravati Project in 1993, it became apparent that the Orissa government was prepared to consider more fundamental sectoral reforms. Intensive discussions both within the Orissa government as well as with the Bank culminated in the November 1993 announcement of a complete sectoral restructuring that involved splitting OSEB into generation, transmission and distribution companies. This was effected on April 1, 1996. OSEB's only thermal plant was sold to NTPC. The privatization of electricity distribution is to follow, in four stages by the year 2000. Hydro plants and the ongoing Upper Indravati project were transferred from the government and OSEB to a new company, the Orissa Hydro Power Corporation (OHPC), on April 1, 1996. On the basis of the Orissa government's sectoral policy declaration, the Bank in 1993 started the preparation of a project to assist the reform program. Orissa Power Sector Restructuring Project was finally approved on May 14, 1996. It was also envisaged that the resulting project would include funds for OHPC for the completion of Upper Indravati, but this did not materialize. Details are given in Appendix C. 45. The fact that under the Upper Indravati project, the Bank and the Orissa government had an ongoing dialogue on the problems of OSEB meant that the Bank was able to actively assist the Orissa government in the early stages of formulating its ambitious and sweeping sectoral reforms. The Bank's offer to fund the preparatory work resolved a potential obstacle over funding of external consultants - subsequently the British Overseas Development Administration agreed to provide grant financing. While it is too early to judge the eventual outcome of the reforms, they have the potential of bringing very substantial benefits to Orissa. It is clear that there is now within the Orissa government a strong commitment and willingness to tackle fundamental power sector problems with a resolve that was absent in the case of the management problems that have plagued the Upper Indravati project - and the Bank approved the Orissa Power Sector Restructuring Project to support Orissa's pioneering reform program. Within the Government of India and the main sources of finance for India's power sector, there is increasing agreement on the key components of the Orissa program being the essential elements of any state power reform program - substantial tariff adjustments, independent regulation, privatization of distribution, and corporatization, commercialization and private sector participation in the other parts of the power  - 12 - systems. This bodes well for the replication of this (Ln. 4014-IN) type of projects in other states. However, even with growing recognition of the bankruptcy of the state power sector, political constraints remain formidable. Nevertheless, there is no alternative to such fundamental reforms. E. BANK'S PERFORMANCE 46. Project appraisal was insufficiently thorough in resettlement and rehabilitation aspects (para. 27), but these shortcoming were successfully corrected during the project implementation and supervision (paras. 30-32). In contrast, all major design and project management issues raised in this ICR were clearly recognized and examined during project preparation, but some were consciously (this can be stated as the decisions were documented) not addressed and were not incorporated in the project design and implementation arrangements. The agreed implementation schedule was unrealistic (para. 18). In the area of project management, there was a documented management override of the appraisal team's technical judgment (para. 47). In the areas of design (including intake location and sedimentation) and implementation (including the approval of manual construction methods) there is no evidence of a similar management override, on the contrary. Documents on file show that the appraisal team was generally satisfied with the project and had no problem with the construction methods or with deferring sedimentation work to a later date. Undoubtedly the team did not expect that the sedimentation work would only commence in the early 1990s. 47. The Bank did not act upon the appraisal team's concerns regarding project management which were contained in the Issues Paper. It expressed reservations about project management capabilities, and in particular construction planning, procurement programming, contract management, project cost control, project accounting and the associated reporting to management. It recommended an external review of project management and accounting arrangements and the appointment of a management team satisfactory to the Bank prior to negotiations. However, the decision meeting in December 1982 diluted these recommendations, referring to the Government's sensitivity to Bank approval of its personnel decisions. Instead, it was agreed that the Bank should adopt an approach in which greater emphasis is placed upon the functions and tasks of Orissa's project team, and on project implementation requirements. Although Government of India was advised that the receipt of a consultant's report on project management would be a condition of negotiations, this was waived. The minutes of negotiations held in late March 1983 record that the Indian delegation only provided Bank staff with a copy of the TOR for the review. Receipt of the report then became a condition of effectiveness and implementation of its recommendations, a loan covenant. Three years later, these recommendations had not yet been fully implemented. With hindsight, it is evident that the while the appraisal team correctly identified the risks to satisfactory project implementation arising from weak project management, adequate safeguards were not insisted upon by the Bank. Suspension of Disbursements 48. Notice of intention to suspend disbursements was first issued in February 1987 for unsatisfactory performance in a wide variety of areas. A second notice was issued in January 1988, mainly due to delays in auditing OSEB's accounts. A third notice became necessary in  - 13 - August 1990 because of lack of progress on resettlement issues. However, the actual suspension of disbursements was not carried out in any of these cases as the Bank accepted partial progress on the outstanding issues as proof of the Orissa government commitment to improving project implementation. Nevertheless, the project was continuously rated either 3 or 4 from 1986 onwards. 49. Shortly before the original loan closing date of June 30, 1991, the Bank recommended replacement of engineering consultants (CWC) and non-performing contractors employed on the headrace tunnel, penstock, Muran dam and the tail-race channel. At the same time, it was indicated to the Orissa government that extending the closing date would also depend on improved handling of resettlement issues. Insufficient progress was reported and disbursements were suspended with effect from June to December 1991. The delegation of full responsibility for project management to an international consulting firm was later added as a condition for lifting the suspension of disbursements. 50. A second suspension of disbursements took place from August to December 1993, although a partial lifting was granted from mid-October to cover payments for resettlement and technical assistance. The suspension was due to unsatisfactory progress in the areas of project management, the resettlement and rehabilitation program, and OSEB's financial performance. The suspension was lifted and the closing date of the Credit extended, as the engineering and construction soundness was established, CMO was initiated, the R&R program was making encouraging progress and the work on addressing OSEB's fundamental problems was launched. Supervision 51. Inadequate resources were devoted to supervision in the first three years of implementation (44 SW, or under 15 SW/year in FY84 through FY86). However, there was a sharp increase to an average of 41 SW over the remaining eight years. It appears that the Bank was not sufficiently aware of the OSEB's weakness in its commercial functions till the last few years of the project (para. 42). This may have been due to the limited participation of financial analysts in supervision missions, particularly during the first 6 years of project implementation when a financial analyst participated in only 50% of the supervision missions. Thereafter the participation rate rose, but only to 70%. Resettlement experts participated in supervision intensively from late 1986. Future Assistance for Project Completion 52. From late 1991 onwards, the Bank had indicated willingness to consider additional funding to replace the canceled Ln. 2278-IN and help complete the project as soon as the project's management and engineering and construction soundness issues are resolved. Reorganization of the Upper Indravati Project turned out to be a painstaking endeavor, taking substantially longer than expected. But while the reorganization process was finally substantially completed, various primarily environmental concerns led to further requirements to be met prior to considering Bank funding for the remaining items. As part of the preparation of envisaged Bank financing for project completion, a new full Environmental Impact Assessment of the project was carried out. This study was completed in mid-1994, and along with the engineering  - 14 - soundness review (para. 14), revealed deficiencies in various engineering and environmental aspects of the project. 53. Bank funding for project completion had to be deferred and was delinked in early 1995 from the proposed Orissa Power Sector Restructuring Project. Future Bank participation in project completion was made dependent on the resolution of such pending issues as sedimentation control and mitigation at the power station and reservoir, penstock construction, and the mitigation of potential environmental impacts caused by an irrigation emergency bypass. Financing to support mitigation measures, namely watershed management and R&R, was retained in the scope. But after OHPC secured completion financing from PFC, Orissa and the Government of India no longer sought Bank financing for the watershed management and R&R programs, which do not feature in the final scope of the Orissa Power Sector Restructuring Project. Overall Assessment 54. There were weaknesses in the quality of project appraisal, and supervision was inadequate in the pre-1987 period. The Bank should have been firmer in seeking measures to redress deficiencies in project implementation prior to 1991. As can be seen from the three instances when the Bank desisted from formally suspending disbursements (para. 48), there was a tendency to accept the Orissa government assurances of imminent action in place of real measures. It is apparent that the Bank's pre- 1991 patience and tolerance of shortcomings in implementation performance did not help the narrow objective of completing the project as rapidly as possible. It is also clear, however, that the Bank's new approach pursued from 1991, inter alia pressuring Orissa to address project management and resettlement deficiencies by canceling Ln. 2278-IN and withholding completion financing, has not yet yielded the desired results either. Time will tell if OHPC, the product of a Bank-supported power sector reform program in Orissa, will be able to successfully complete the project. The Bank is no longer involved in the implementation of the project. But the need the address the issues raised by the Bank - which led to the Bank withholding financing - remains. Appendix C of the ICR is included in order to record such issues and in an attempt to bring them once more to the attention of OHPC and the Orissa government. (Appendix B records with clear statements that while Orissa appreciates the Bank's active involvement in the implementation of Orissa's power sector reforms, it strongly disagrees with the Bank's decision to delink Upper Indravati completion financing from the Orissa Power Sector Restructuring Project.) F. BORROWER'S AND IMPLEMENTING AGENCY'S PERFORMANCE 55. There was a general unwillingness to recognize that such a large and complex project required exceptional management and engineering skills. In addition, the cost of delays in terms of foregone revenues or additional costs was given little importance. 56. The Orissa government was too involved in day-to-day management of the project. The Orissa government's split of the Irrigation and Power Department then diffused responsibilities and further weakened project management. Project management lacked autonomy and decision making authority. At the same time, it was never provided with satisfactory communications  - 15 - infrastructure, despite its remote location. This compounded the delays arising from the high degree of the Orissa government involvement in decision-making. 57. --_ Changes in senior Orissa government and project staff were too frequent, but this may have been an unavoidable consequence of the fact that the project was implemented by a government department where such promotions/transfers are commonplace. The Panel of Inquiry into the 1991 flood found project staff to have been negligent in not taking adequate steps to close the intake to the headrace tunnel. Finally, many project staff did not reconcile themselves to the post-1993 management structure and were not committed to its goals. 58. The Orissa government showed little commitment to improving OSEB's performance during the ten years prior to 1993. Tariff increases were delayed, government subsidies to compensate for the low tariffs were paid late and the Orissa government arrears for public sector electricity consumption were allowed to accumulate to unsustainable levels. OSEB itself prior to 1993 made insufficient efforts to improve its accounting and commercial practices. These issues are only now being addressed. 59. The performance of the Borrower was thus found to be deficient in project design, management arrangements and in implementation. However, there are positive indications of a new resolve to pursue fundamental sector reforms (paras. 44-45). The implementation of the reform program and the establishment of OHPC have already helped Orissa secure completion financing for the project. OHPC's challenge is to successfully complete the project. G. ASSESSMENT OF OUTCOME 60. Upper Indravati is an unfinished project. Meanwhile repayment of the IDA Credit has already begun. This is an unsatisfactory interim outcome. However, this assessment will have to be revisited upon physical completion of the project. In spite of the extra-ordinary problems and delays, the project is still projected to provide power at a highly attractive rate and remains a priority component in the Orissa/Eastern region power system. Far-reaching reforms to the power sector as a whole have begun, and may prove to have highly positive effects. As a part of the reform program, the Upper Indravati project has been transferred from the state government to OHPC. Its challenge is to complete the project, for which it now has the necessary funding. H. FUTURE OPERATION 61. Responsibility for the project has been formally handed over by the state government to OHPC. Once completed, OHPC will also operate the project. OHPC is operating Orissa's other hydro projects and is expected to be able to operate the Upper Indravati as well. A complete range of institution building support is being provided to OHPC under the Orissa Power Sector Restructuring Project.  - 16 - I. KEY LESSONS LEARNED * The design and quality of project management arrangements is probably the single most important element for successful and timely project implementation. The Bank should not agree to support projects where such fundamental conditions are not satisfied. * Dependence on design and engineering consultants such as Central Water Commission (CWC) who are not answerable to the project owner should be avoided. * Important engineering issues such as anti-sedimentation measures should be resolved at the time of project appraisal or as soon as possible thereafter. * Competitive bidding can result in award of contracts to public sector firms who submit unrealistically low bids. More rigorous prequalification and bid evaluation may help to prevent this. * The successful use of consultants for project management requires a genuine commitment on the part of the implementing agency and/or project owner, and cooperation by the consultants' counterparts on the ground. Lack of such support will undermine even the best of efforts by consultants. Adequate resources should also be provided. Apparent cost savings achieved by reducing expatriate staff and introducing lower cost but less than fully qualified local support will be over-shadowed by implementation delays and difficulties. * This project was yet another addition to a generally unsuccessful series of Bank operations in the Indian power sector at the state level, with state electricity boards and their state governments. But in this case the prospects for addressing the fundamental management and financial problems - and successfully completing the Upper Indravati project - are encouraging. Since 1992 the Government of Orissa, OSEB and the Bank have discussed the issues and possible solutions to Orissa's fundamental power problems. In 1993, the Government of Orissa and OSEB agreed upon a power sector reform program as the logical conclusion to these deliberations. The program involves privatization of power distribution and separation of the power sector from direct government control through regulatory reforms, corporatization, commercialization and private sector participation. The Bank approved on May 14, 1996, the Orissa Power Sector Restructuring Project (Ln. 4014-IN) to support Orissa's pioneering reform.  - 17 - Part II: STATISTICAL ANNEXES A. Achievement of Obiectives S- bstantiaf Partial Neglible Not licable Sector policies O O O U Financial objectives O O * O Institutional objectives O O * O Physical objectives O *