Nepal - Third Water Supply and Sewerage Project

Docum nt of The World Bank FOR OFFICIAL USE ONLY Repwt No. 9158 PROJECT PERFORMANCE AUDIT REPORT NEPAL THIRD WATER SUPPLY AND SEWERAGE PROJECT (CREDIT 1059-NEP) NOVEMBER 30, 1990 Operations Evaluation Department 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 without World Bank authorization. MEASURES AND EQUIVALENTS 1 = liter (- 0.26 US gallons) lcd a liter per capita per day m3 = cubic meter (1 cu.m. - 1,000 liters) mld = one million liters per day 1/sec = liters per second mm - millimeter (.0.03937 inches) m - meters (-3.28 feet) km - kilometer (.0.62 miles) sq.km = square kilometer (.0.386 square miles) ha - hectare (-10,000 square meters a 2.47 acres) ABBREVIATIONS AND ACRONYHS HMG - His Majesty's Government IDA - International Development Association MWR - Ministry of Water Resources DWSS - Department of Water Supply and Sewerage MPLD - Ministry of Panchayat and Local Development UNDP - United Nations Development Programme WHO - World Health Organization ODA - Overseas Development Administration WSSB - Water Supply and Sewerage Board (1973 to June 1984) WSSC - Water Supply and Sewerage Corporation (since July 1984) SAR - Staff Appraisal Report CURRENCY EXCHANGE RATES Name of Currency - Nepal Rupees (NRs) Appraisal Year Average Exchange Rate US$1 = NRs 12.00 Intervening Years Average US$1 - NRs 16.99 Completion Year Average US$1 - KRs 23.75 TK WORLD ANK FOR OFFICIAL USE ONLY WdstwngOn. DC 20433 US A Office of Director-General Operations avaluation November 30, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Nepal Third Water Supply and Sewerage Project (Credit 1059-NEP) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on Nepal - Third Water Supply and Sewerage Project (Credit 1059-NEP)" prepared by the Operations Evaluation Department. Attachment This document has a Mtrcted distributon and may be ued by fcspeants only in the performance o( their ofRcial dus. Its contents may not otherwe be dicasd Wethout World bank 1thonaln. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT NEPAL THIRD WATER SUPPLY AND SEWERAGE PROJECT (CREDIT 1059-NEP) TABLE OF CONTENTS Page No. PREFACE .................................................. ....... i BASIC DATA SHEET .................................................. ii EVALUATION SUMMARY ................................................ iv PROJECT PERFORMANCE AUDIT REPORT A. Background .............................................. 1 B. The Project and Its Objectives .......................... 2 C. Implementation .......................................... 3 D. Procurement ............................................. 5 E. Project Costs .................................... ....... 5 F. Disbursements ........................................... 6 G. Performance of Consultants and Contractors .............. 6 H. Institutional Performance ............................... 7 I. Operational Aspects ..................................... 9 J. Financial Aspects ....................................... 10 K. Performance of the Bank ................................. 12 L. Economic Assessment ..................................... 13 M. Conclusions and Lessons ................................. 13 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 without World Bank authorization. PROJECT PERFORMANCE AUDIT REPORT NEPAL THIRD WATER SUPPLY AND SEWERAGE PROJECT (CREDIT 1059-NEP) PREFACE 1. This report presents the results of the Performance Audit of the Third Water Supply and Sewerage Project for which an IDA Credit of US$27.00 million was approved on July 29, 1980. The proceeds of the Credit were onlent by the Borrower, the Kingdom of Nepal, to the Water Supply and Sewerage Board (WSSB) on terms of 25 years, including five years grace period, at an interest rate of 3/4% per annum (Development Credit Agreement, Section 3.01 (c)). 2. The Credit Agreement was signed on September 29, 1980 and the Credit became effective on August 27, 1981, some eight month after the original date. The original closing date was December 31, 1985 but, after three extensions, the Credit was, eventually closed on June 30, 1988. The last disbursement was made on February 14, 1989 at which time US$0.04 million of the Credit was cancelled. 3. The PPAR is based on the Project Completion Report (PCR) of the project prepared by the Asia Regional Office and issued in 1989 (Nepal: Third Water Supply and Sewerage Project - Credit 1059-NEP, Report No. 8089, September 29, 1989), the Staff Appraisal and President's Reports, the credit documents, the transcript of the Executive Directors' meeting at which the project was considered, on a study of project files, and on discussions with Bank staff. An OED mission visited Nepal in April 1990 and discussed the project experience with relevant Government and WSSB officials and visited several facilities constructed under the project. 4. The Audit finds that the PCR gives a generally good and frank account of the project experience but is inclined to disagree with some of its conclusions. The PPAR further analyzed various aspects of the PCR's findings and also provides independent conclusions and comments on the more important aspects of the project. 5. Copies of the PPAR were sent to the borrower and the implementing agency for comments, however, none were received. ii PROJECT PERFORMANCE AUDIT REPORT NEPAL THIRD WATER SUPPLY AND SEWERAGE PROJECT (CREDIT 1059-NEP) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual Total Project Cost (US$ million) 33.32 33.55 Overrun (%) - 1.00 1 Credit Amount (US$ million) 27.00 Disbursed 26.96 Cancelled 0.04 Repaid to June 30, 1989 Nil Outstanding at June 30, 1989 26.96 Date Physical Components Completed (Mo/Yr) 12/85 Preparation Completed by Above Date (%) 100 95 Proportion of Time Overrun (%) 0 48 Economic Rate of Return (%) 3 2.25 Financial Performance Unsatisfactory Institutional Performance Satisfactory PROJECT DATES I tem OriginaL Ptan Revisions Actual First Mention in Files 12/08/78 12/08/78 Negotiations 05/02/80 Board Approval 06/26/80 07/29/80 Credit Agreement Date 09/29/80 Effectiveness Date 12/30/80 08/27/81 Closing Date 12/31/85 12/31/86,06/30/87 06/30/88 Borrower His Majesty's Government of Nepal (HMG) Executing Agency Water Supply and Sewerage Board (Corporation since 1984) Fiscal Year of Borroer July 15 to July 14 Follow-up Project Niae Water Supply and Sanitation Rehabilitation Project (Proposed) Cost overrun in Nepal Rupee (NR) terms was -arly 60%. iii CUMULATIVE DISBURSEMENTS (US$ mittion) fl1 . IIN fI flS flM F1 Z I.IE ApprasaL EStimate 2.20 8.20 16.80 22.30 25.70 27.00 Actual 0.79 8.47 10.01 12.48 16.77 24.12 26.08 26.96 STAFF INFUTS (Staff Veeks) fLZ2 fL flA1 fin EMI IM fIM U1 flE fl I&A Preappraisat .6 6.8 7.3 Appraisat 27.8 27.8 Negotiation 4.1 1.8 5.9 Supervision 7.3 12.4 5.5 12.? 10.9 14.7 15.1 6.2 84.7 Other 1.5 .3 .4 .2 2.4 Total .6 40.2 9.4 12.8 '.7 12.7 10.9 14.7 15.1 6.2 128.1 MISSION DATA No. of qtaff Purpose MofYear Weeks Persons Weeks Report Date Identification 11/78 2.4 2 3.9 01/26/78 Preparation 05/79 1.6 2 3.2 06/29/79 Preappraisal 09/79 1.6 2 3.2 10/29/79 Appraisal 01/80 4.4 2 8.9 02/09/80 Supervision 1 10/79 1.6 1 0.4 10/25/79 Supervision 2 11/80 1.6 2 0.9 01/07/81 Supervision 3 08/81 2.9 3 2.8 11/11/81 Supervision 4 06/82 2.0 2 1.2 07/16/82 Supervision 5 03/83 1.4 2 1.0 05/05/83 Supervision 6 11/83 2.8 2 2.8 11/17/83 Supervision 7 04/84 2.6 2 2.6 06/04/84 Supervision 8 09/84 1.8 1 1.0 10/25/84 Supervision 9 02/85 2.0 2 2.0 04/10/85 Supervision 10 10/85 1.8 1 1.8 10/21/85 Supervision 11 11/85 2.0 2 4.0 01/24/86 Supervision 12 02/86 1.8 2 3.6 03/11/86 Supervision 13 08/86 2.2 2 4.4 11/02/86 Supervision 14 03/87 3.0 2 12.0 04/22/87 Supervision 15 11/87 3.0 3 9.0 12/16/87 Completion 02/88 0.4 2 0.9 03/11/88 iv PROJECT PERFORMANCE AUDIT REPORT NEPAL THIRD WATER SUPPLY AND SEWERAGE PPIJECT CREDIT 1059-NEP EVALUATION SUMMARY Background 1. Nepal, located northeast of India is covered mostly by high mountains and foothills and a narrow strip of plains in the south, called the Terai. At the time of appraisal (1980) about 95% of Nepal's 14.4 million population resided in some 28,000 rural communities. Of the 5% urban population more than half lived in the Kathmandu valley. The country's substantial rainfall (1000-1600 mm) has an uneven seasonal distribution with heavy concentration in the monsoon seasor (June-September). The Terai is estimated to contain substantial ground water sources but of poor quality containing high levels of iron and ammonia. 2. Water supply and sanitation service levels, in 1980, were generally poor with only 10% of the country's population having access to piped water, the rest relying on natural sources of doubtful quality and often difficult access. All urban areas had piped water systems, reportedly covering 80% of the urban population (39% by house connections) but with low reliability of service. Waste disposal services were far below even these standards. Kathmandu's rudimentary sewer system, originally designed for 25,000 people, has fallen into disrepair. The poor service levels resulted in high incidence of water borne and communicable diseases and high infant mortality rates (PPAR, paras. 1-4). 3. Responsibility for the sector was divided. The Water Supply and Sewerage Board (WSSB) served urban areas over 10,000 population, the Department of Water and Sewerage of the Ministry of Water, Power and Irrigation (MWPI) communities between 1,500 and 10,000 and the Local Development Department of the Ministry of Home and Panchayat (MHP) looked after smaller communities. 4. Sector investment between 1966 and 1980 averaged a low 2.6% of development expenditures and this, together with the shortage of skilled manpower, materials and equipmenc and the lack of sound sector financing policies accounted for the low service levels (PPAR, paras. 5-6). 5. Bank involvement started with the First Water Supply and Sewerage Project for which IDA Credit 470-NEP for US$7.8 million and Supplement for US$4.0 million were approved on April 30, 1974 and April 25, 1977 respectively. This was followed by the Second Project with Credit 704-NEP for US$8.0 million approved on April 26, 1977. Both projects concentrate on urban services, primarily Kathmandu. OED joint Performance Evaluation Report No. V 6506 of 1986 gave low marks for the poor performance of both projects. Design and construction shortcomings, inefficient management, extensive delays and very poor financial performance were the primary observations (PPAR, para. 7). The Project and its Objectives 6. The project was to support the Government's efforts to improve urban water supplies in the Kathmandu valley and eight additional towns, extend the sewer system in Kathmandu and assist in the development of WSSB. Preparation of the project w.s limited to the updating of the 1974 Kathmandu Valley Master Plan, feasibility studies and preliminary engineering. Detailed design was to be carried out after Board approval. 7. The design was to provide for relatively basic services with easy to operate facilities. In the hill areas, these consisted of springs and stream intakes and, on the plains, ground water was to be the primary source. Ground water availability was not verified by adequate investigations and the institutional improvements envisaged proved optimistic (PPAR, paras. 9-11). Implementation 8. IDA Credit 1059-NEP, for US$27.0 million was approved on July 29, 1980 and the original closing date of the credit was December 31, 1985. After three extensions the credit was closed on June 30, 1988 and, at the last disbursement, on February 14, 1989, US$ 0.04 was cancelled (PPAR, para. 8) 9. The project had a slow start as noncompliance with important effectiveness conditions delayed this action by thirteen months and, consequently, postponed the start of detailed design to September 1981. The Audit has concluded that, at least, part of the reason for this delay was the fact that the First and Second projects were still under implementation at the start of the Third project and the volume of work overloaded WSSB. 10. The appraised project underwent substantial changes during detailed design and, during construction, further changes became necessary. The latter was mainly due to the groundwater well drilling resulting in unproductive wells, requiring new well locations and modified collector pipe layouts. In the Kithmandu Valley, 65 million liters per day (mld) were expected from 17 wells and, in the end, only 39.5 mld could be obtained from 28 wells. In the town of Pokhara, no groundwater was found at all and untreated surface water supplies are being provided. 11. A further, serious, consequence of the inadequate groundwater investigations, particularly in the Kathmandu Valley, is that the aquifers are being "mined" (permanent depletion) and the continuously dropping water table is endangering future supplies. In addition, the quality of the water is poor with higher than acceptable iron and ammonia content. 12. The small sewerage component had its own particular design and construction problems. The debate whether Kathmandu should have had combined or separate (foul/storm water) sewers is still continuing and the slow and incompetent contractors were, at the later stages of the works, prevented from vi opening up the narrow streets of Kathmandu by the public and the Department of Roads. 13. The rate of overall physical progress was far slower than projected, indicating that few lessons were learned from the 150% and 90% time overruns of the first two projects. Although the 48% time overrun of the Third project might be considered an improvement, it is still very high (PPAR, paras. 12- 17). 14. Slow procurement processes contributed significantly to overall project delays but WSSB had little chance of coping efficiently with the work load even with the assistance of its consultants. The slow approval process of the Government was a major contribution to the delays. Project costs exceeded appraisal estimates by about 60% (Nepal Rupee terms) but direct comparison io not meaningful due to the substantial changes in the project components. Disbursement rates followed the slow progress and, by the original closing date of December 31, 1985, only 37% of the credit was disbursed (PPAR, paras. 18-20). Performance of Consultants and Contractors 15. The design problems which came to light during construction and in the operating stage leaus this Audit to disagree with the PCR's judgement that the consultant's performance was "...generally satisfactory". The Management Consultant's "Diagnostic" study of WSSB did not appear to have pro.ided value for money and, apparently was little utilized. Civil engineering contractors, with the exception of one drilling contractor parrotmed poorly, probably as much due to inadequate supervision by consultants as to difficult working conditions and general incompetence. Part of the reasons for this may well lie in the remoteness of Nepal where small contracts do not attract major contractors. Suppliers of equipment performed satisfactorily (PPAR, paras. 21-22). Institutional Performance 16. WSSB, successively changed to WSSC in 1984 and to the Nepal Water Supply Corporation (NWSC) in 1990, was created in 1973, as a condition of the first Bank project. Since that time, despite substantial assistance under three Bank assisted projects (a fourth has been negotiated), its performance has never been rated satisfactory. in 1987, a Government appointed commission delivered a devastating criticism of the Corporation's performance and recommended far reaching changes, (the Commission's report has not been published). Although the effects of the latest organizational changes (creation of NWSC) can not yet be judged, the Audit mission found guarded optimism among the officials met. 17. While the record of poor management and limited operational effectiveness seems indisputable, the Audit believes that a large number of factors conspired against better performance, many of them outside the direct control of the organization. First among these would be the undoubted work overload created by the three overlapping Bank projects. Others would be the limited supply of skilled manpower in Nepal, the extensive and tight vii Government controls including appointment of senior staff and consequent lack of autonomy and, perhaps most important, the total lack of a realistic financial base mainly due to totally inadequate tariff increases. 18. It is clear that none of the three Bank projects seemed to provide sufficient early institutional assistance to prepare WSSB for the work load and expected planning, operational and financ!al improvementv resulting from these projects. The recent developments in Nepal and the fourth project, already negotiated, may yet provide the cpportunity for adequate assistance to prepare the new NSC for the massive remedial (and expansion) work required in the sector to provide even the minimum acceptable service levels (PPAR, paras. 23-27). Operational Aspects 19. With the continuous workload of the three successive investment projects and WSSB's management and financial problems, it comes as no surprise that operation and maintenance of the systems received low priority. If anything, this is underlined by the lack of detailed information on these matters. Extensive technical assistance from the British ODA appears to have brought limited benefits and the design and construction shortcomings, described earlier, ouly multiplied the problems. 20. Unaccounted fo: water (UFW) is and has been, consistently, very high, at different times estimated between 40% and 65%. The very poor state of consumption metering, due in equal measure to neglect and heavy deposits of iron oxides (a new meter is said to be clogged in 3-4 weeks), makes verification of these figures impossible. The most recent estimates, made after extensive analysis by a firm of international consultants, puts the current figure at 40%. However, even if this figure is accurate, the order of magnitude, in a water scarce area and for a financially troubled entity, is unacceptably high. The low level of attention to this matter in the past must be reversed and it has received high priority in the preparation of the Fourth project. 21. Altho-igh the Audit mission hau limited opportunities to inspect the systems (du to civil disturbances), it has learned that, currently, Kathmandu receives oz,iy three hours of water supply in the morning and three in the evening. While the more affluent segment of the population can cope with this by providing their own storage, it presents serious hardship to the poorer standpipe users. This situation, and the fact that current managers of Kathmandu's water supply can not remember a better level of service (in spite of a fourfold increase in production over the last 10 years), shows that three cons cutive projects, at best, maintained the status quo in a rapidly growing environment. 22. With the critical ground water situation and the lack of alternative/ additional water sources within the Valley, Kathmandu's situation is at a critical level. Current proposals to develop sources "outside the Valley" may take up to ten years to implement and will be expensive due to the need to tunnel through the surrounding hills. Until such sources are developed, and solve the long term problems, strict water conservation and loss control will viii need to be applied if extreme effects on the Valley's population and its important tourist industry are to be avoided (PPAR, paras. 28-29). Einancial Aspects 23. Under the first two projects, the Bank required WSSB to achieve 8% and 6% financial rate of return on net fixed assets. The corresponding financial forecasts were based or the assumptions of: increased water sales and house connections, regular tariff increases, introduction of sewerage charges, reduced UFW and accounts receivable and imp'oved billing and collection. These assumptions were retained for th- third project but the covenant was softened to revenues covering operation and maintenance and, later, debt service. The onlending rates from the Government to WSSB were also reduced from 9% to 3/4%. None of the assumptions proved correct, none of the covenants was met, the debts could not be serviced and both the first and second credits were converted to Government equity. 24. Water sales did not reach projected levels mainly because of the lower than estimated source capacities but also because the high connection costs discouraged new customers. UFW remained consistently high and (even the inadequate) tariff increases only materialized every 5 or 10 years (none since 1981), and then only under extreme pressure from the Bank. As a result, WSSB's net income has been negative throughout the period 1974 to 1989. In addition to the revenue shortfalls WSSB's cash flow problems were aggravated by the extensive delays in the Government making payments for the substantial standpipe consumption. 25. Under these circumstances, possible inefficiencies notwichstanding, WSSB had no chance of achieving financial viability and the Bank's single minded pursuit of tariff increases was an inadequate action plan to improve che situation (PPAR, paras. 30-34). Performance of the Bank 26. The Bank has maintained an unbroken twenty years of presence in the Nepal water supply sector with three overlapping projects. This lave! of assistance is commendable as well as unusual for such a small country. The amount and apparent quality manpower input, for appraisal and supervision, should be considered to have been adequate. The Audit could find no obvi.ius reasons for the failure of the project which could be attributed to specific shortfalls in Bank staff performance although, by implication, a degree of responsibility or simply over optimism, given the prevailing c.,rcumstances, cannot be denied. The design of the proposed fourth project has taken account of the lessons the Bank ought to have learned. Economic Assessment 27. The PCR presents a recalculated ERR of 2.25% (without supporting cal-ulations) and compares it to the appraisal estimate of 3%. The Audit believes that, considering the project overall, and, in particular, its financial performance as described above, the recalculated ERR can only be of dubious value. ix Conclusions and Lessons 28. In the overall context of the Bank's lon6 involvement in the Nepal water sector, the current state of that sector and, in particular of the Kathmandu Valley service standards, the Audit has to judge the project a failure. The distribution system in Kathmandu has insufficient and poor quality water to deliver and the institutional and financial performance of the implementing agency is unsatisfactory. The Bank has little to show for its past involvement but may be commended for its perseverance, and has a great deal to do to help the sector onto its feet. 29. The principal lessons to be learned are, once again, that inadequate project preparation can (and here it did) have a very high price and that targets must be matched to the real capacity of the sector agencies. The Bank has repeatedly given in to the Government's tactics in frustrating the institutional and financial development of the sector agency, essentially preempting the efforts of the sector staff to bring forth improvements. 30. The Fourth Project, now approaching approval stage, has been designed to tackle the most critical issues. Operations staff categorically state there will be no Board presentation without decisive Government action. The Bank must adhere to this. Project staff disagree with this judgement citing improvements in towns outside the Valley and the increased water production and service coverage in Kathmandu. PROJECT PERFORMANCE AUDIT REPORT NEPAL THIRD WATER SUPPLY AND SEWERAGE PROJECT (CREDIT 1059-NEP) A. Background 1. Nepal, located northeast of India, lies parallel and south of the main Himalayan mountain range. Some two thirds of the country's 140,000 sq km land area consists of high mountains and their foothills and a narrow strip in the south, called the Terai, borders the Indo-Gangetic plains. At the time of appraisal (1980) Nepal's population was estimated to be about 14.4 million. Urbanization was at a low le,vel with some 95% of the population residing in 28,000 rural villages and 5% in 20 urban centers, more than half of the latter living in the Kathmandu Valley. The average annual population growth rate in the 1970s was about 2.1% but the urban population grew at 5% per annum. 2. Rainfall is substantial ranging from 1000 to 1600 mm per year but the seasonal distribution is uneven with heavy concentration in the monsoon season from June to September. The country is covered by a dense network of rivers and ground water resources were thought to be substantial in the Terai. River and spring waters are subject to high turbidity during the monsoons and, generally, to bacterial contamination. Most ground waters have higher than acceptable iron and ammonia content. 3. At the time of appraisal, service levels were generally poor with only some 10% of the country's popilation having access to safe water supplies. In the rural areas only 6% received piped water, the rest had to rely on natural sourcas of doubtful quality, often with difficult access. All urban centers had piped water systems supplying an average 80% of the population, some 39% by house connections the rest by public standpipes. Waste disposal services were far below even these standards. Kathmandu's rudimentary sewer system, theoretically served some 25,000 people but the system fell into disrepair and required comprehensive rehabilitation and, in most other urban centers organized waste disposal did not exist. 4. As a consequence of poor water supply and sanitation facilities water borne and other communicable diseases were prevalent in Nepal and were important causes of high infant mortality end poor public health standards. 5. Responsibilities for sector development were divided between three agencies. The Water Supply and Sewerage Board (WSSB), under the Ministry of Water, Power and Irrigation (MWPI) was responsible for the planning, design, construction and operation of services in large urban areas (over 10,000 population). The Department of Water Supply and Sewerage of the MWPI provides services for communities with between 1,500 and 10,000 population and the Local Development Department (LDD) of the Ministry of Home and Panchayat (MHP) looked after the smaller communities. 2 6. Sector investments during the years 1966 - 1980 averaged a low 2.6% of total development expenditures, clearly accounting for the low service levels. The Government set ambitious future development plans within the framework of the International Drinking Water Supply and Sanitation Decade, aiming at 100% service in urban areas and, at least, one reliable water source in every community. This target, if achieved, would have provided an overall population coverage of 68% by 1990. Apart from the low investment levels, the sector faced formidable difficulties in the shortage of skilled manpower, material shortages due to the remote and landlocked nature of Nepal and, perhaps most of all, due to the lack of sound sector financing policies. Water supply and sewerage developments were financed from Government budgetary allocations, with heavy reliance on external aid, but there was, virtually, no internal generation of funds. 7. Bank involvement in the sector started with the identification of the First Water Supply and Sewerage Project for which and IDA credit (Credit 470-NEP) of US$7.8 million was approved on April 30, 1974 and was supplemented by US$4.0 million on April 25, 1977. This was followed by the Second Water Supply and Sewerage Project for which IDA Credit 704-NEP, for US$8.0 million, was approved on April 26, 1977. This funding was later supplemented by a EEC Special Action Credit of US$3.0 million. Both projects concentrated on the improvement and provision of additional urban services, primarily in the Kathmandu area. OED evaluated the two projects jointly and the findings are set out in Project Performancp Evaluation Report No. 6506 dated November 19, 1986. The principal conclusions of the report are not encouraging. Individually and collectively the two projects made little progress in achieving their objectives and the reTort concludes that sustained further support will be required to improve service levels significantly. The extended service coverage resulted in more people receiving poor service, generating public reluctance to pay for the services. Staff shortages, management inefficiencies and political interference resulted in design and construction shortcomings, extensive delays and unsatisfactory maintenance. Inadequate and persistently delayed tariff increases, high unaccounted for water (UFW) and poor billing and collection led to a financial situation in WSSB which could not sustain efficient operations, let alone further expansion. 8. This was the environment in which the Bank appraised the Third Water Supply and Sewerage Project, the subject of this evaluation, in a commendable effort to help set the Nepal water sector on its feet and, in the process, try to salvage the investments of the first two projects. IDA Credit 1059-NEP for US$27.0 million for the Third Water Supply and Sewerage Project was approved on July 29, 1980. The original closing date was December 31, 1985 and, after three extensions, the credit was eventually closed on June 30, 1988 at which time the project was still not quite completed. The last disbursement was made on February 14, 1989 when US$0.04 million of the credit amount was cancelled. B. The Project and Its Objectives 9. The basic objective of the project was to continue to support the Government's efforts to provide safe water supply to all urban area population by 1990. Specific targets were: (i) to further improve the water supply services in the towns of Kathmandu, Lalitpur, Bhaktapur, Pokhara, Biratnagar and Birganj 3 (beneficiaries of the First and Second Projects); (ii) expand water supply services in six additional towns; (iii) extend the sewer system in the Kathmandu Valley and (iv) continue to assist WSSB in developing into an efficient and financially self-supporting entity. On completion of the project some 720,000 people would receive improved water supply and about 67,000 people would be served by sewers in the Kathmandu Valley. 10. Preparation of the project largely consisted of the updating of the 1974 Kathmandu Valley Master Plan and the feasibility studies (prepared at the same time) for the remaining towns under (i) above and the preparation, under the Second project, of feasibility studies for the six additional towns. Preliminary engineering work, financed under the Bank's Project Preparation Facility (PPF) started immediately after appraisal. Detailed engineering was to be carried out under the project, with the assistance of consultants. 11. The proposed service standards and project facilities were, generally, basic and designed to be of relatively low cost and simple to operate. In the hill areas, springs and stream intakes were to provide reasonable quality water and, on the plains and valleys ground water was thought to be available in adequate quantities. It was, perhaps this apparent simplicity, together with the "knowledge" allegedly gained from the previous two projects, that convinced the Bank to proceed to appraisal and approval without detailed design and more extensive ground water investigations. Furthermore, and notwithstanding the very limited institutional progress made during the six years that elapsed between the approval of the First and Third projects, the SAR of the project exhibited extraordinary optimism with regard to the anticipated implementation capacity of WSSB as well as to its ability to increase water and sewerage tariffs and, generally, improve its financial performance. The latter is also contradicted by the unusual action by the Bank of agreeing to the onlending of the Credit to WSSB on the very concessional IDA terms. Discovering the realities did not take long. C. ImDlementation 12. The project had a poor start. Of the five conditions of effectiveness only the signing of the Subsidiary Loan Agreement was complied within time. All others, a 45% tariff increase, issuing of sewerage regulations, payment by Government of all arrears for standpipe water and the appointment of consultants for detailed design and supervision were extensively delayed. The Bank was, repeatedly, on the verge of .-ancelling the credit but effectiveness was eventually declared on August 27, 1981, thirteen months after Board approval. As a result, detailed design work did not commence until September 1981. 13. Although project documentation provides few references, the Audit has little doubt that the fact that at the start of the Third project the First and Second projects were still under implementation significantly affected progress. The poorly staffed and hopelessly under-funded WSSB faced a task for which it was scarcely equipped. As a hopeful solution, the Bank pressed hard for the This was in strong contrast to the pessimistic tone of the Issues Paper/Decision Memorandum discussions recorded in the project files. 4 conversion of WSSB into a corporation, a status which would have been expected to enhance its autonomy. Although this conversion materialized in July 1984 the expectations, unfortunately, did not. The PCR, in Chapter III provides a good account of the trouble-plagued implementation experience. 14. The components of the project, as appraised, underwent significant changes during the detailed design stage. .ctual construction, in turn, resulted in further major changes in a number of subprojects, the most significant of these related to source works based on ground water abstraction. Assessment of exploitable ground water resources is still an inexact science but far better tools are certainly available than those employed in the preparation of the project which, for the most part, appeared to rely on past, unconfirmed, abstraction potential. In any water supply project, where ground water abstraction is a "possibility", such a scheme will, invariably, prove to be the "least cost solution" compared to a surface water alternative. However, until hydrogeological estimates are verified by adequate test drilling and pumping, final commitment to a ground water scheme is a high risk undertaking. In the case of the Third project, such commitments were made with dire consequences. The PCR provides extensive details but the most important scheme, for the Kathmandu Valley, is worth noting. Design "estimates" envisaged an average daily yield of 65 mld from 17 wells. In practice, 28 wells were drilled and the total average yield, on completion, was 39.5 m1d. In the case of the town of Pokhara, no ground water could be obtained at all, and an emergency surface water scheme has to be constructed which now provides untreated water only, as no funding has yet been made available for the treatment plant. This is now included in NWSC's 15-year Development Program, with good possibilities of being funded by a bi- lateral. Basic chlorination provides some protection against bacterial contamination. 15. An even more serious aspect of the groundwater issue, apart from expected new yields not materializing, is the fact that aquifer levels, in existing well fields are dropping at an alarming rate. The aquifers are being "mined" which endangers future supplies and makes existing wells inoperative. Recent (1988) studies by expert consultants paint a serious picture and prescribe extensive remedial actions. The latest investigations by a foreign aid agency confirmed that current abstractions in the Kathmandu valley produce ancient fossil waters, clearly indicating a lack of recharge. The Audit's field discussions indicated that, in certain areas, the well drilling process became a random, hit and miss affair in the hope of finding water. A further consequence of this was that all initial collector pipe layouts had to be changed according to where the new wells were located. The question needs to be asked why exhaustive studies were not carried out earlier? The Audit notes that similar examples are far from uncommon in Bank projects in many parts of the world, emphasizing the need for more thorough project preparation and, in particular, more careful assessment of the apparently "least cost" ground water schemes. 16. The relatively small sewerage component had its own particular troubles. The concept of the combined foul sewer/storm drain design seemed to have an equal number of proponents and detractors but the Audit can find no clear explanation as to what the problems wer-. The statement in the PCR (par. 3.32) that the public believed that the minimum sewer size of 150 mm was "too -mall and refused to connect to it" neither makes sense nor does it explain anything. The Audit 5 mission was informed that part of the reason for the small pipe size was the shortage of funds, i.e. smaller pipes covered the same area with less cost. A more relevant issue appears to have been the incompetence of the sewer-laying contractor which resulted in public outcries wherever sewers were constructed. The very slow progress resulted in open trenches in the narrow streets of Kathmandu for long periods, until the residents (and Ministry Officials) would permit no more trench opening.2 In addition the main trunk sewer to the treatment plant has collapsed preventing flows to the plant which, itself, is in disrepair. 17. The rate of physical progress achieved under the project is vividly illustrated by the chart in Annex 1 of the PCR which shows that not one activity came even close to the schedule projected at appraisal. Although no such chart is available for the First and Second projects, the "time overruns" quoted in the PPAR (Report No. 6506) of 150% and 90%, respectively, speak for themselves. Under the Third Project the "time overrun" was 48%, a significant improvement over the first two projects but still unacceptably high. D. Procurement 18. The PCR's conclusion (para. 3.47) that "the project demonstrated WSSC's ability to handle the procurement of complex water supply contracts..." is something of an overstatement. The project files chronicle repeated procurement problems, particularly in the bid evaluation aspects. Procedures were very slow with most bid validities expiring before the award was made. The Bank received a string of complaints from bidders claiming unfair evaluation practices. The project being the third Bank involvement in the sector, one would have expected better familiarity with the procedures. However, in fairness, it ought to be noted that the frequent staff changes in WSSB, the overload of handling (at times) three Bank projects could have left little chance for WSSB to concentrate sufficient resources for procurement. It is also worth noting that the relatively small contracts and Nepal's remoteness did not attract the top suppliers and contractors. E. Proiect Costs 19. The PCR devotes limited attention to the cost of the project but merely mentions that the some 60% increase in the total cost of the project, in Nepal Rupee (NR) terms, was due to the design changes introduced during the detailed design stage. Were that entirely true, it would certainly be a strong signal indicating the need for much better project preparation. However, the lack of detailed analysis of the cost escalation (in addition to the effect of the design changes) with regard to the effect of the extensive delays and inflation makes it impossible to draw firm conclusions. There is no explanation provided for the very large increase in the cost of Engineering and Studies. Due to the devaluation of the NR, the US$ equivalent cost of the project only increased by 1% from the appraisal estimates. A summary table of Appraisal v. Actual costs is given below: 2 Field visits by the Audit to the facilities constructed proved impossible due to severe civil disturbances during the mission. 6 Appraisal Est. Actual Works (Million NR) (Million NR) Change% Construction & Equipment incl. Contingencies 348.09 511.64 +47 Engineering, Studies & Training 40.87 113.74 +178 Duties & Taxes 10.80 14.00 +13 Totals 399.76 639.38 +60 F. Disbursements 20. The start of disbursements was delayed by nearly two years and, after a brief spurt in the third year, it slowed down again dramatically. By the original closing date of December 31, 1985, only 37% of the credit was disbursed. The last disbursement was, eventually made on February 14, 1989 at which time US$0.04 million of the credit was cancelled. G. Performance of Consultants and Contractors 21. The PCR states (para.3.29) that the "performance of the consultants... .was generally satisfactory" and proceeds, in subsequent paragraphs, to list a series of major design faults and failures for which the engineering consultants were, obviously, responsible. Adding these to the general failure of the ground water availability assessment as already mentioned in paragraph 14. above, makes the Audit to strongly dispute the "satisfactory" rating. Consultancy work for the Manpower Survey and Training Plan appear to have been adequate even though, due largely to the overall weakness of WSSB, the product did not appear to result in measurabla improvements. Another consultancy assignment produced a "Diagnostic Review of the Water Supply and Sewerage Corporation" (November 1986). The Audit did not review this report but the PCR lists nineteen of its findings starting with the almost frivolous statement that "formal staff meetings were not regularly called". The remaining eighteen which had to be obvious from any supervision report, do not appear to justify the US$80,000 price tag either. The project files do not record any utilization of this study but Operations staff state that it improved the Government's Commission's work. 22. Suppliers of equipment were reported to have performed satisfactorily but civil engineering contractors, with the exception of one drilling contractor, performed poorly. However, the reportedly substandard quality of workmanship reflects not only on the contractor but also on the supervising consultant and WSSB. In fairness to the contractors (local and foreign), the working conditions, poor contract management, the lack of decision making and the repeated changes in design, could not have been conduc!ve to good performance. 7 H. Institutional Performance 23. Competent and efficient institutional performance by any water/sewerage utility is an indispensable requirement for the successful implementation of projects and for the subsequent, sustained, delivery of high quality service, operation and maintenance. By the accounts of past SARs, OED's Project performance Audit Report (Report No. 6506 on Credits 470-NEP and 704-NEP), the Supervipion Reports and the PCR for this project (Credit 1059-NEP) and the Audit's findings, this has never been achieved in Nepal. The cautiously hopeful forecasts of the SAR for the Third project did not materialize. The PCR deals, at considerable length, with the institutional issue and it seem to have identified most of the shortcomings in WSSB and also appears to apportion the blame for the lack of progress quite fairly to WSSB, the Government and the Bank. The "Diagnostic Study" by consultants, referred to above seems to have added nothing useful and, according to knowledgeable Bank staff, has been largely discarded. The Government-appointed "Pokhrel Commission" of 1987 has delivered a devastating criticism of the then WSSC's performance and recommended far reaching changes, some of which the Bank appeared to disagree with on the grounds that they would further retard institutional development. And this, perhaps, provides a clue to the crux of the problem. The Commission had three months to investigate and report. Most of the other institutional reviews (including the Bank's own) were of short duration, appear to have included no recognized experts and delivered summary recommendations (sometimes in the form of requirements) without much discussion with the interested parties. 24. There can be no doubt that the overall performance of WSSB (since 1984 WSSC) was unsatisfactory. The limited information on its history reveals that it was established in 1973, apparently at the insistence of the Bank for the implementation of the First project. There is no information as to how much and what quality effort went into its establishment. There are obscure references to its perceived "temporary nature" but it is clear that its autonomy was non- existent from the beginning. In the Audit's view, WSSB never had a chance. In the environment of Nepal, with its very limited skilled manpower, tight Government control of every activity and an almost infant water/sewerage sector, the quick succession of overlapping projects with everything that the Bank's bureaucratic project processing implies, it would have been a miracle if WSSB could keep up the pace. Bank projects are legion where a fledgling institution is expected to "learn by doing" i.e. manage its existing assets efficiently, implement new projects which are many times the size of existing assets with strange, new, procedures, take over several other existing facilities, rapidly improve its technology, management and, above all, its usually highly subsidized and disastrous financial situation. And do all this against an, usually, skeptical Government which has its own economic and political agenda and interministerial jealousies. It ca.,not be done. The Bank must learn to be patient and convince its borrowers to be the same. ?5. No water undertaking, developed or developing country, which cannot manage, operate and maintain its existing assets efficiently and which does not have an adequate financial base (whether subsidized or not) can and should be expected to take quantum leaps in performance, size and sophistication at the same time. The "Pokhrel Commission" findings and recommendations may not have been perfect but they were certainly perceptive in noting that the Bank's one dimensional 8 thrust for massi'.e tariff increases could not be seen as the universal remedy.' And no amount of valid concern about unserved masses can justify the over hasty attempts at accelerated progress that the three Nepal projects exhibit, whether this is described as "supporting Government policies" or is instigated by the Bank. Although each project resulted in the expansion of physical assets, collectively they made little advance in the overall performance of the sector. The Audit believes that there is no successful example of Bank-supported water/sewerage project where such an accelerated program of institutional development succeeded. It is absolutely -ssential that before an entity (and the country) is plunged into massive new investment, a sound and efficient organizational, operational and financial base be in existence. If advance technical assistance, creation of a new financial base or an "engineering project" is required, so be it. Project preparation must extend well beyond engineering design and must encompass detailed institutional work. 26. Whatever WSSC's self-induced shortcomings may have been, the Audit has considerable sympathy with its problems. Its recognized efforts of improving efficiency and cost effectiveness could never compensate entirely for the lack of resources. Any Government and the public should have a concern in demanding efficient service before increased tariffs but, in the case of WSSC this may have been, to some extent, a chicken or egg situation. When tariff increases are not permitted for 5-10 years, the chances of financial viability are, indeed, remote. In addition, the continuity and morale of staff was strongly affected by the reliance on secondment of senior staff from the ministry which, generally, eliminated most prospects for promotion. 27. The SAR states (para. 4.07) that, in Nepal, "....it is difficult to obtain trained professional personnel.... there will be need for staff training... the Project includes provision for scholarships for senior level staff as well as some training material." The PCR, in turn, states that "Despite the UNDP funded Manpower Survey and Training Plan, WSSC accomplished little systematic training...." In addition to the various (valid) explanations provided for the lack of success the Audit believes that given the excessive workload of WSSC's staff, there was simply not enough time for training. The Audit mission's discussions on these matters were severely limited by the political/Governmental changes and public disturbances which took place at that time in Kathmandu. Nevertheless, even these discussions revealed that few officials were optimistic about quick and effective improvements. The latest organizational change, the conversion of WSSC into the Nepal Water Supply Corporation (NWSC), became effective on February 26, 1990 with, so far, little significance as the declared autonomy only exists on paper. Discussions with Government officials brought forth criticisms of gross over-staffing (one referred to NWSC as a depository of friends and relatives of high officials), inefficiency and corruption. Whatever the truth, it is clear that major institutional improvements are yet to be achieved. The Commission also emphasized that with the exception of the OED Performance Audit of the first two projects, all other institutional efforts concentrated on the symptoms, rather than the disease. 9 I. Operational Aspects 28. The PCR provides very little information on system operations. However, there is enough indication in the project files and documents relating to institutional reviews for the Audit to conclude that operation of the systems was far behind the construction of new schemes in WSSC's priorities. The Kathmandu sewer system never seemed to have worked effectively. Operational problems of ground water systems abounded hoth in terms of quantities and quality of water. Although WSSB has received extended technical assistance from the British Overseas Development Administration (ODA) for improving system operations, the PCR's quote (para. 3.03) from the report of a departing training officer, eloquently describes the lack of benefits resulting from this assistance. The OED evaluation report on the first two projects, referred to earlier, emphasized the urgent need for the control of unaccounted for water (UFW), reported to average about 45% yet there is little evidence in the project files that the Bank or the WSSC expended a major effort in this area which seems to be confirmed by the PCR reporting UFW of 60% at the time of completion. More recent, detailed, studies reportedly assess UFW at 40%. The poor state of metering (in all parts of the systems) prevents the verification of these figures but, in either case, considering the scarcity of water and the limited revenue generation, this order of losses is inexcusable and it reduction ought to be the highest priority for both Nepal and the Bank. In this respect, it appears, that the statement in the OED Evaluation Report (No. 6506) that "It was not until the Third Project, and the establishment of the Corporation (WSSC) that a significant effort has been made toward tackling leakage, wastage...." was premature. That this has been recognized is evidenced in the system rehabilitation thrust of the proposed fourth, the Water Supply and Sanitation Rehabilitation Project which was appraised by the Bank in November 1989. 29. Even the constrained effectiveness of the Audit mission was sufficient to show that, if anything, the situation is worse than depicted by the documentation. After three projects and nearly twenty years of continuous Bank involvement water is supplied to Kathmandu for only three hours in the morning and three hours in the evening. While the more affluent residents compensate for this with private storage tanks, it is hitting low income consumers who cannot afford private storage, particularly hard. The Audit was hard put to reply to a senior Government official's question of: "what has the Bank been doing?" The Audit can certainly confirm that, in terms of quality, the raw water is, generally, unfit for human consumption. Apart from taste and turbidity problems, the very high iron content results in heavy deposits and clogs a new water meter in three to four weeks. As already mentioned, the ground water sources are already overdrawn and there are no exploitable surface water sources within the Kathmandu valley. Preliminary plans exist for an "outside the valley" surface water scheme which, if it proceeds, may take eight to ten years to implement. Its water will also have high unit cost due to the expensive tunneling required. Until such a source is brought on stream, the most stringent water conservation measures will have to be applied, ground water abstraction needs to be brought under control and the water will need to be treated. Given the present situation, there is little time to be lost to avoid a major crisis and even if these measures are applied effectively, Kathmandu's water supply will be far from satisfactory for a long time to come which will affect not only its people but its important tourist industry. 10 J. Financial Aspects 30. While the "Pokhrel Commission's" conclusion that infusion of additional monies, alone, would not have solved WSSC's problems was, certainly, valid, it is equally true that, from its inception, WSS2 never, for any period of time, had a viable financial base from which to operate. The PCR gives a good, factual account of WSSC's poor financial situation without offering any significant insight into the root of the problems or much in the form of possible solutions. 31. From 1974 (the first Bank involvement) to date, the financial requirements set, the main assumptions on which they were based and the actual achievements are summarized below. Under the First and Second projects, "standard" Bank financial covenants were imposed requiring WSSB to achieve (progressively) 8% financial rate of return on revalued net fixed assets by 1987, softened to 6% in the Second project. Already in early 1980, at the appraisal of the Third project, these targets were considered unachievable and the Credit Agreement (Section 4.05) specified that WSSB's revenues should cover the following: (i) in 1981 and 1982, the cost of operation maintenance and administration and depreciation; and (ii) in 1983 and thereafter, the cost of operations, maintenance and administration and its debt service or depreciation, whichever is higher. Significant changes were also introduced in the onlending rates far the proceeds of the credits namely 6% for the First, 9% for the Second and 3/4% (IDA terms) for the Third. 32. The principal assumptions for setting these requirements and for the corresponding financial projections, in all three projects, related to significant increases in volume of water sold, number of metered water connections, steadily increased water tariffs, rapid increase in sewer connection and the imposition of sewer charges, reduction of unaccounted for water, reduction of accounts receivable (both in Government payments for standpipe water and private consumers) and improvements in billing, collection and cost control. Virtually none of the assumptions proved correct to any significant degree and, understandably, none of the covenants was met. Some of the salient figures showing the forecast and actual values are presented in the table below. 11 Financial Indicators 1974 1979 1984 1989 Forec./Actual Fore./Actual Forec.lActual, Frse./Actual Water Sold Mm3/a 6.02 6.75 12.22 8.97 21.36 17.60 27.38 25.34 UFW% 50 50 40 45 32 45 22 60 Op.Rev.W. NRs Mil. 1.53 1.93 11.50 5.47 49.52 16.59 NA 29.14 Op.Rev.S. NRa Mil. - - 1.36 0.00 6.40 0.00 NA 0.35 Op.Bxp. NRs Mil. 1.82 1.66 5.44 3.87 12.22 16.67 NA 47.60 Depr. 1.22 0.96 4.18 2.36 19.05 6.47 NA 12.03 Net Inc. (0.86)(0.40) 5.60 0.02 14.08 (19.02) NA (28.00) 33. In this situation, WSSB was clearly unable to service debts and service payments scheduled to start under the respective project agreements, for Credit 470-NEP in 19/9 and for Credit 704 7EP in 1983, were repeatedly rescheduled and, eventually, in 1986 both loans were converted to Government equity. 34. The below forecast water sales and sewerage revenues both were largely due to the extensive delays in implementation. Delays in the provision of services were further aggravated by the very high cost of water connections (a partial effort to compensate for low tariffs) and the reluctance of the residents to connect to the sewer system. Cash flows were also disastrously affected by the very high UFW percentages and the magnitude of accounts receivable which. in 1987, reached 74% billings, the main factor being the Government in not paying for standpipe consumption. However, the hopelessly inadequate and much delayed tariff increases probably outweighed all these factors. The history of water tariffs in Nepal is revealing. The initial tariffs were introduced in 1966 and the next increase, 601 did not come until 1975 with the First project. Ths was followed, in 1981, by a 45% increase as a condition of effectiveness under the Second project, which delayed effectiveness by nine months. In 1984, WSSB reported that, in order put its financial house in order, it required an immediate 183% tariff increase. The next proposed increase of 65% was to take place in July 1986, "agreed" in the 11th hour, as a condition of extending the closing date of the credit. Although the closing date was extended, the tariff increase was not implemented; the Bank retreated, once again. All subsequent efforts to increase tariffs further have failed until the tariffs were, again, agreed by the Government to be increased by 100%, in October 1989, as a condition of appraisal of the Fourth project. The appraisal took place but the tariff 12 increase although approved is, in fact, not yet in effect. The new Nepal Water S Dply Corporation (NWSC) has, by its statutes, the powers to raise tariffs, but i.L Clairly has not been able to exercise it. On the basis of these newly granted(?) powers, current projections show a rosy financial picture for NWSC but this is &lmost entirely dependent on adequate tariff increases or, possibly, on high level of subsidy. The Audit was informed (unofficially) that there was no chance for an early tariff increase and the recent political changes do not make it any more likely. The Bank's current position is (as stated by the Region) that the Fourth project will not be presented to the Board until a satisfactory tariff increase is implemented. 35. In the rLcording of the sorry history above, many reasons have been given from political expediencies, to poor service, to affordability considerations for not acting on tariff issues in a timely manner. All may be valid or just excuses but, it is clear, that the Government carried brinkmanship to its limits against the Bank's arsenal of lbvers. The Audit finds it difficult to reconcile the Government's relative readiness of granting debt rescheduling, large equity contributions and other forms of subsidies (all from taxpayer's money) with the persistent non-payment of large amounts of standpost charges and the denial of timely tariff increases. The net result was a hamstrung WSSC, inefficient project implementation, nonexistent maintenance, poor all round services, very high revenue losses and completely eroded public confidence. It remains to be seen whether enough has changed for "Round Four" to finally succeed in improving the sector. K. Performance of the Bank 36. At the time of this evaluation, the Bank has maintained an unbroken 20 years of involvement in the Nepal water and sewerage sector. On the one hand, this is a remarkable record which few countries of this size can claim and it has the Audit's full commendation. On the other hand, the shortfall in achievements, as identified in successive evaluations, justifies that searching questions be asked with regard to the nature and focus of Bank assistance. The Bank deserves high marks for not severing a generally good Bank/counti; relationship when the Government's brinkmanship tactics (see above) were taken to the extreme. Even with hindsight, it would be presumptuous to state that hard-nosed tactics would have brought better results. The resources devoted to the Third project should be considered adequate. Fifteen supervision missions in about eight years, staffed with generally high caliber sector staff ought to have done the job. Nevertheless, while the average Bank water supply project as its usual institutional and financial problems, the physical components are generally well executed guaranteeing some degree of success. In Nepal, after three projects, even this would be difficult to claim particularly in view of ta. state of Kathmandu's water supply. In the Audit's view, two explanations stand out and are: (a) despite the apparently moderate size of each project, the resourc3 requirements consistently exceeded the sector's implementation and operating cau)acities; and (b) the Government's persistent refusal to permit (let aAone assi.st) the build-up of these capacities. 37. Nepal's geographical isolation, its single trade route through India (which for some time has been practically cut off), its manpower shortage and lack of sectoral tradition should have dictated a slower and more piece-meal approach. 13 The Audit fully recognizes that the most difficult decision Bank staff and management have to make is to assess the appropriate level of help in the face of acute need. The three Nepal projects appear to offer the proof that this level has no : been found in this case. The proposed fourth project has taken the right route of giving priority to rehabilitation but its size and complexity suggests that, perhaps, the Nepalese implementation capacity has, once again, been overestimated. Economic Assessment 38. The PCR states (para. 7.04) that against the 3% Economic Rate of Return (ERR) the "on completion" recalculation produces a value of 2.25% but does not present the calculations. The Audit did not attempt to reproduce this effort as it believes that in view of the project's achievements, as described in this report, this would be little more than a mathematical exercise with no relation to the actual benefits or, rather, the lack of them. The PCR itself also states, in paragraph 7.06 that "the project did not deliver the benefits envisaged at appraisal". M. Conclusions and Lessons 39. In the overall context of the Bank's long involement in the water supply sector in Nepal and the current state of the sector, particularly the current lack of secure long term water supplies in the Kathmandu valley, the Audit has to conclude that the project has been a failure with the redeeming features being the improvements in towns outside the valley. The claimed 30% increase (well below target) in volume of water available "at the tap" is more than counterbalanced by the 50% increase in urban population (1990). The significantly increased distribution network has insufficient water to deliver and, in terms of water quality, more people now receive poorer service. Institutional improvements have been minimal, operation and maintenance are poor and the financial situation of the sector agency is little short of disastrous. 40. By all accounts, the Bank has not distinguished itself under this project or through its previous involvement. The reasons and explanations are many, most of them valid. While in some areas such as project design, the inadequate preparation and the institutional development efforts it could and should have done better, in others such as the financial and implementation performance of the borrower it had little chance against the Government's inertia and Nepal's environment, short of withdrawing from the project. However, it is recognized that without some optimism, the Bank would never operate in some countries and, hopefully, new opportunities will be available in Nepal also to start on the road for improvements. 41. There are many lessons to be learned from this project but few of them are new to OED or the Bank. The price of inadequate project preparation (in the widest sense) was higher in this project than most others. When strong Government reluctance to accept the Bank's prescription for financial improvements in the sector are coupled with consumer resistance to pay for a manifestly poor service, a better analysis and more thoughtful programs are required than the (almost) routine tariff increase covenants, particularly, if these have already failed through two previous projects. The design of the 14 Fourth projeet shov that any of the leggone of the pagt have not only been learned but also applied. What y6t to bo sen is whether the ank is doins to b* resolute enouh (thi* time) not to proceed until the requirements for reasonablf. succesø are firaly in place.