Document of The World Bank FOR OFFICIAI USE ONLY ReportNo. 8670 PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF NEPAL SUNSARI MORANG IRRIGATION AND DRAINAGE PROJECT (CREDIT 812-NEP) MAY 29, 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. CURRENCY EQUIVALENTS 1 Nepalese Rupee (NRs) = US$0.04 NRs 25.0 = US$1.00 NRs 100 = US$4.00 KINGDOM OF NEPAL FISCAL YEAR July 16 to July 15 WEIGHTS AND MEASURESEQIUIVALENTS m meter (3.28 feet) m3 /sec = cubic meter per second (35.3 ft3 ) km - kilometer (0.62) ha hectare (2.47 acres) liter 5 61.02 cubic inches M Million ton 2,205 pounds (lb) ABBREVIATIONS AND ACRONYMS CMC - Chatra Main Canal CPS - Central Projects Staff DADO - District Agricultural Development Officer DIHM - Department of Irrigation, Hydrology and Meterology (renamed the Department of Irrigation) DOA - Department of Agriculture DOI - Department of Irrigation ERR - Economic Rate of Return GOI - Government of India HMGN - His Majesty's Government of Nepal HYV - High Yielding Variety JT - Junior Technician JTA - Junior Technical Assistant M&E - Monitoring and Evaluation OED - Operations Evaluation Department PCR - Project Completion Report PLAA - Panchayat Level Agricultural Assistant PPAR - Project Performance Audit Report SAR - Staff Appraisal Report SDR - Special Drawing Rights SMIDB - Sunsari Morang Irrigation Development Board SMIP - Sunsari Morang Irrigation Project SMS - Subject Matter Specialist T&V - Training and Visit System of Agricultural Extension TO - Training Officer WUG - Water User's Group FOR OmFCIL USE ONLY THEWORLD BANK DC 20433 Washington. USA Officeof Director-Ceneral Operations Evaluatimn May 29, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on the Kingdom of Nepal Sunsari Morang Irrigation and Drainage Project (Credit 812-NEP) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on the Kingdom of Nepal - Sunsari Morang Irrigation and Drainage Project (Credit 812-NEP)" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipientsonly in the performance of their officialduties. Its contents may not otherwise be disclosedwithout World Bank authorization. FOR OFFICIALUSE ONLY PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF NEI'AL SUNSARI MORANG IRRIGATION AND DRAINAGE PROJECT (CREDIT 812-NEP) TABLE OF CONTENTS Page No. PREFACE ................. i BASIC DATA SHEET ................. ii EVALUATION SUMMARY ................. v PROJECT PERFORMANCE AUDIT I. BACKGROUND.............. ......... , .. 1 II. IMPLEMENTATION EXPERIENCE . . 3 III. PROJECT OUTCOME . . 6 IV. FINDINGS AND ISSUES . . A. Inadequate Preparation and Premature Lending. 8 B. The Peril of Designing to Cost.............. 9 C. Economic Rate of Return Revisited.10 PROJECT COMPLETION REPORT I. INTRODUCTION. 17 II. PROJECT FORMULATION.19 III. PROJECT IMPLEMENTATION.22 IV. INSTITUTIONAL PERFORMANCE......................... 32 V. AGRICULTURAL IMPACT.37 VI. ECONOMIC EVALUATION............................... 40 VII. THE BANK PERFORMANCE.40 VIII. LESSONS LEARNED AND CONCLUSIO .41 Table 1 - Summary of Physical Progress.......................... 45 Table 2 - Summary of Financial Progress......................... 46 Table 3 - Cost Estimates........................................ 47 Table 4 - Actual and Estimated Disbursement Schedule............ 48 Table 5 - Procurement of Vehicles and Equipment................. 49 Table 6 - Cropping Patterns, Area, Yields and Production........ 50 Table 7 - Economic Costs and Benefit Streams.................... 52 This document has a restricted distribution and may be used by recipients only in the performance of their officialduties. Its contents may not otherwise be disclosed without World Bank authorization. Table of Contents (continued) Ajrnex . - Compliaw;e- wiLui; CLtIL CunuliLiQIb ..................... 53 FROM THE PORROWER.... .................... 59 ATTACHMENT - COMMENTS MAPS - IBRD No. 13469R - Project Location - IBRD No. 13526R - Chatra Main Canal Command Area - i - PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF NEPAL SUNSARI MORANG IRRIGATION AND DRAINAGE PROJECT (CREDIT 812-NEP) PREFACE 1. This is a Project Performance Audit Report (PPAR) on the Sunsari Morang Irrigation and Drainage Project, involving an IDA credit in the amount of US$30.0 million to the Kingdom of Nepal with the objective of rehabilitating and improving an existing irrigation system. The IDA credit was approved on May 25, 1978 and became effective on November 30, 1978. The closing date of June 30, 1984 was extended to December 31, 1985, then to December 31, 1986 and ultimately to June 30, 1987. Cofinancing in the amount of US$4.0 million equivalent was -rov4ded by ti-hrFuropean Economic Community through a Special Action Credit which became effective on November 7, 1979. 2. The audit is based on the attached Project Completion Report (PCR) of the project prepared by the Asia Region, an earlier PCR prepared by the project consultants for the Borrower in June 1987, the Staff Appraisal and the President's Reports, the loan documents, the transcript of the Executive Director's meeting at which the project was considered, on a study of projec_ files, and discussions with Bank staff. An OED mission visited Nepal in November 1989 and discussed the effectiveness of the Bank's assistance with His Majesty's Government of Nepal (HMGN) staff. Their kind cooperation and valuable assistance in the preparation of this report is gratefully acknowledged. 3. The PCR provides a reasonably satisfactory, though optimistically reassuring, account and assessment of the project experience, and discusses the performance of the Bank and the project executing agencies. The Audit elaborates on particular aspects such as inadequacies in the preparation and appraisal, the continuing problem of siltation resulting from the introduction of a temporary engineering solution and the economic rate of return. 4. Following standard OED procedures, copies of the draft audit and the PCR were sent to the Borrower. The comments received from the Borrower are reproduced as an Attachment to the PPAR. - Li- PROJECT PERFORMANCEAUDIT REPORT KINGDOMOF NEPAL SUNSA.I MORANGIRRIGATION PROJECT (CREDIT 812-NEP) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as X of Item Estimate Est.mated Actual Appraisal Estimate Total Project Cost (US$ million) 37.5 41.9 112 Credit Amount (US$ million) 30.0 30.0 100 EEC Special Action Credit (US$ million equiv.) 4.0 3.3 n.a. Iaentification 1975 Preparation 1976-77 Appraisal 09/77 Date Board Approval 05125178 Date Credit Signing 07/07/78 Date Effectiveness 11130/78 Date Physical Components Completed 06/30/83 06/30/87 Closing Date 06/30/84 06/30/87 Economic Rate of Return C%) 17.0 0.5 Institutional Performance Weak Agronomic Performance Disappo.nting Number of Direct Beneficiaries 35,000 Considerably less CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (in US$ million) FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 Appraisal Estimate (US$ m) 0.5 3.6 12.0 19.5 27.0 30.0 - - - - Actual (USS m) 0.1 1.5 5.2 8.1 10.5 13.3 17.9 26.3 29.7 30.0 Actual as % of Appraisal Estlmate (%) 20 42 43 42 39 44 80 88 99 100 Date of final disbursement December 31, 1987 - ilL - MISSION DATA Specializa- Date No. of tions Performance Types of Missgion (mo./yr.) Persons Represented Rating Trv,d Problems /a ~~~Lb /C Id Identification 03/76 Preparation 76/77 Appraisal 09/77 Supervision I 05/78 2 2E 2 - M Supervision II 10/78 3 2E 2 2 M,T Supervision III 02/79 1 E 2 2 M,F SupervisLon IV 10/79 3 E 2 2 M,0 Supervision V 04/80 2 A,E 2 1 T Supervision VI 11/80 2 2E 3 2 0 Supervision VII 03/81 2 2E 3 2 F Supervision VIII 05/81 3 3E 3 1 0 Supervision IX 12/81 1 E 3 2 M,° Supervision X 08/82 2 2E 2 2 M Supervision XI 01/83 ? ? 2 2 M Supervision XII 06/83 1 E 2 2 M Supervision XII, 09/83 1 E (no 590 in file) 7 Supervision XIV 12/83 1 E (no 590 in file) ? Supervision XV 02/84 2 2E 2 2 M,T Supervision XVI 11/84 2 A,E 2 2 M,T Supervision XVII 03/85 2 A,E (no 590 in file) 7 Supervision XVIII 06/85 2 2E 2 1 M,T Supervision XIX 02/86 3 A,2E 2 2 M Supervision XX 09/86 3 A,2E 2 2 F Supervlsion XXI 11/86 6/e A,E 2 2 T Supervision XXII 05/87 2 A,E 1 2 T Supervision XXIII 11/87 2 A,E 2 2 T STAFF INPUT (staff weeks) FY72 FY73 FY74 FY75 FY76 FY77 FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 Total Preappraisal 0.5 - 11.8 3.0 6.6 17.9 9.5 - - - - - - - 0.8 - - 50.1 Appraisal - - - - - - 73.9 - - - - - - - - 73.9 NeguLiation - 0.8 - - - - 10.4 - - - - - - - - - - 11.2 Supervision - - - - - - 1 1 13.9 11.0 21.1 13.9 18.0 9.8 5.8 6.6 8.0 1.6 11C.9 Other - - - - 0.1 - 0.5 0.1 - 0.2 0.1 - - - - - - 1.1 TOTAL 0.5 0.8 11.8 3.0 6.7 17.9 95,5 14.0 11.0 21.3 13.9 18.0 9.8 5.8 7.4 8.0 1.6 247.2 - iv - OTHER PROJECT DATA Borrowier: His Majesty's Government of Nepal (HMGN) Executing Agencies: Department of Irr.gation, Hydrology and Meteorology (DIHM), subsequently renamed Department of Irrigation Name of currency (abbreviatlon) Nepall Rupees (NRs) Appraisal year average 1978 US$1.00 = NRs 12.00 Intervening year average 1979-S6 US$1.00 - NRs 16.35 Completion year average 1987 US$1.00 = MRs 21.80 Follow-on PrEiect: Name Sunsari Morang Irrigation II Project Credlt Number Cr. 1814-NEP Amount (USS million) 40.0 Date Board Approval 06102/87 La E - Engineer; C = Economist: A - Agriculture: P = Procurement Specialist. /b 1 = Problem-free or minor problems; 2 - Moderate problems: 3 - Major problems. ic 1 ;pLeviiu- 2 - S.at1.iiaz; ^ = Deteriorating. I L M = Managerial; F = Financial; T = Technical; 0 Other. le In conj.znctionwith Appraisal of Sunsari Morang Irrigation II. PROJECT PERFORMANCE AUDIT REPORT KINGDOM OF NEPAL SUNSARI MORANG IRRIGATION AND DRAINAGE PROJECT 'CREDIT812-NEP) EVALUATION SUMMARY Introduction The NepaleseTerai, an extensionof the Ganges Plain, affords good opportunities for increasing agricultural production through irrigation, and this was the third irrigation project to be supported by the Bank in that area. Objectives The project aimed to restore and further improve a previously installed irrigation system over 66,000 ha which had deteriorated due to design imperfections and neglect of maintenance, with a view to increasing farmers' .ncomesand rural employment. An important aspect of the design was the provision of improved river and sediment control works in the vicinity of the intake and along the headreach of the main canal to ensure a more regular supply of reasonable quality irrigation water to the command area. A second important aspect was the planning and design of an extended delivery system to permit more equitable distribution of the water to small groups of farmers down to the 10 ha outlet level over about 18,000 ha of the command area, and the construction of irrigation ditches and drains down :o the individual farm level over 6,400 ha of the 18,000 ha first stage development area. Implementation Experience Design studies for improved river and sediment control works, which did not commence until after credit negotiations, took considerably longer to carry out than anticipated at appraisal, thus delaying construction by about four years compared with appraisal expectations. The restoration works for the main canal proceeded more or less to plan in spite of shortages of steel and cement. The quality of construction, while generally satisfactory, was inadequate in places. - vi - Planning and design for an 18,000 ha Stage l development area was reduced to 12,000 ha because of concern at mounting cost overruns. Finally, about three years behind appraisal schedule, a contract for construction of canals and drains down to the 10 ha outlet level over 9,750 ha was awarded. Difficulties and delays encountered in implementing the main civil works led to a further reduction in size and scope of the project in its fourth year. Several of the smaller components were dropped while some components of the river and sediment control, and command area restoration and improvement works were postponed to a future project out of concern at cost escalation arising from underestimation of civil works costs at appra.sal. Results Project achievements have fallen far below appraisal expectstions, primarily due to the inability of the sediment control devices constrivcted under the project to significantly reduce sedimentation in the distributory system and fields under the operating conditions prevailing in the project area (PPAR para. 29). In consequence, and in the absence of adequate maintenance, the capacity of the system has been undergoing a pronounced and steady decline since the works were completed and this is severely limiting potential crop benefits (PPAR para. 32). The economic rate of return is barely positive (PPAR paras. 53-54). Sustainability The failure of the sediment control measures instituted under the project to prevent heavy sedimentation, combined with current shortcomings in operation and maintenance due to budget deficiencies and operational complexities, are seriously threatening the sustainability of the project. Findings and Lessons A pre-appraisal mission decision to declare the project appraised and to process the credit documents through to the Board, in spite of the omission at preparation of specific proposals for reducing sedimentation, and the absence of design plans for ecs ,atial river and sediment control civil works, reflects an unfortunate uLsjudgment regarding the state of project preparedness. In retrospect, submission of the project for Board approval can be seen as premature (PPAR paras. 36-41). Undue concern for redesigning to cost as a means for avoiding high cost overruns which became apparent during implementation, when major - vii - changes in termn of project scope and size were made, led to a significant change from the initial project targeis and a shortfall from project objectives. A formal project reappraisal would have been an appropriate procedure at that time (PPAR paras. 42-47). The assumption that remedial measures incorporated in a follow-on project will rectify failures in a previous project is inappropriate when determining rates of return for the purpose of a project completion report (PPAR paras. 48-54). PROJECT PERFORMANCEAUDIT KINGDOM OF NEPAL SUNSARI MORANGIRRIGATION AND DRAINAGE PROJECT (CREDIT 812-NEP) I. BACKGROUND Context 1. The originalSunsariMorang irrigationsystem in the easternTerai was financed, designed and constructedby the Governmentof India in the mid-1960s through early '70s under a bilateralaid agreementwith HNGN. It comprise! an intake at CILatra on the left bank of the Kosi River, and a headreachwhich leads irrigationwater by gravity to a 66,000 ha co Lla area via a basic network comprisinga 53 km Chatra Main Canal (CMC), a 17 km branch canal, and 210 km of secondaryand 105 km of tertiarycanals. 2. Flow regulationstructures were installedonly at the CMC outlets to the secondaries, which gave rise to much wastage and limited the effective irrigationarea to only 35,000 ha. The System, parts of which were inadequatelydesigned, ha, been plagued by restricted supply at the intake due to shoalingin the Kosi River supply charnel and by severe st4i- mentationproblemr in the CMC and distributaries siivA its commissioning in 1975. These problemshave further reduced its efficiencyby limitingthe canal capacity to well below the full supply discharge. Poor drainagewas an additional problem. 3. Because of these problems which were leading to a steady deteriotation in the entire system,HMGN in 1976 invited the Bank to assist with its rehabilitation. Obiectives 4. The projec. was intended to restore the system to its original scope and capability and further improve the reliability of water deliveriesin order to increasefarmers'incomes and rural employment. Desizn 5. The exigineering features of the project as appraised were to include river control (guide island) and flood protectionembankmentworks on the Kosi river in the vicinity of the CMC intake regulator; sediment control arrangements near the intake and headreach(sluicebay and settling basin) and along the CMC (vortex tube sediment ejector); restoration and improvement of the canal irrigationsystem includingprovisionof about 180 new structuresthroughoutthe 66,000 ha command area; planning and design of complete minor distributioncanals to outlets serving 10 ha irrigation - 2 - units for a stage I intensive devel-ipmentarea of about 18,000 ha, and within that area, constructionof canals and drains serving 6,400 ha; and drainageimprovements covering12,000 to 15,000ha. 6. Additionally,the project as appraised includedpilot schemes for 25 tubewells,improvedwater managementand three canal micro-hydroelectric developments; and strengtheningof agriculturalextension, research and trainingactivitiesthroughoutSunsariand Morang districts. 7. Consultants were to provide aerial photography and mapping; detailed engineering for the CMC restoration,sediment control and river training works; and project implementation advice including training in project planning, bidding documents preparation and evaluationand civil works constructionsupervision. 8. A Project Manager, based at Biratnagar within the project boundary, was to hitveoverall charge of the project. He was to be responsible to a governing board (SunsariMorang Irrigation and Drainage Development Board) created as a semi-autonomous agency under HMGN's DevelopmentBoard Act of 1956. 9. The projectwas expectedto be completedin under five years, with effectivenessbeginning in October 1978 and physical componentscompleted by the end of June 1983. Finance Plan 10. Total project cost was estimated at appraisal at US$37.5 million equivalent. The IDA credit was to finance the full foreign exchange cost of US$20.5 million and about 56% of local costs, or about 80% of total estimated cost. A few months after credit signing additional financing Account was proposed from the European Economic Community Special !...cion and a US$4.0 million equivalentSpecialAction Credit Agreementwas signed in August 1979,with IDA as Administrator. Processing Pre-implementation 11. Following identification by a Bank mission in early 1976 a consultant firm was engaged to undertake a feasibility study. The consultantreport proposed rehabilitationof the headreachand about half the CMC canal system, togetherwith improvementdown to the field channel level over 28,000 ha of the command area, but made no proposals for improving the intake structtrre or reducing sedimentation beyond recommendingdetailedstudy of the problem. One month after receipt of the feasibility mission visited Nepal, in September report an IDA pre-appraisal 1977. 12. The pre-appraisal mission modified considerablythe main elements of the feasibility report by includingrestorationand improvement work on the canal system down to the 50 ha unit distribution level throughoutthe entire 66,000 ha command area, while reducing the 28,000 ha intensive - 3 - development area to 18,000 ha, with water distribution down to the 10 ha unit level. The mission also enlarged the project scope to include river training and flood protection works at the CMC intake and construction of sedimentation structures (para. 5), the outlines of which were developed by a consultant fluviologist after a visit to the project area in November. Before leaving Nepal in early October the mission deemed the project to have been appraised with a view to Board presentation before the end of FY78, and proposed that a six month aerial photography and mapping program, a six month sedimentation control study, and a twelve month CMC restoration works design should be undertaken by international consultants as a matter of urgency. 13. An issues paper dated October 20 was discussed at a decisions meeting on November 2, when it was agreed that the project should be processed expeditiously with a view to negotiation in April 1978 and Board presentation before the end of FY78 (June 30, 1978). 14. A contractor was appointed for the aerial photography and mapping task in January 1978, but appointments for the other two studies were delayed, until March in the case of the CMC restorigtiondesign and May in the case of the river training and sedimentation study. Later, in November 1978, the consultant firm engaged for the CMC restoration design was also appointed as project consultant to provide inservice training, guidance and advice to HMGN project staff. 15. Comments from Central Projects Staff (CPS) at the yellow cover SAR review stage in March were genecally favorable, and covered a wide variety of topics including m3nitoring and evaluation proposal inadequacies, the need for elaboration of the drainage situation, and a request for details concerning farm credit availability. One perceptive query from CPS as to why the engineering design contracts (para. 14) were scheduled so late in relation to the intended commencement of civil works in early 1979 (SAR paras. 3.21-3.22) appears to have been overlooked in the subsequent processing of the documents through the Loan Committee and Board. 16. Negotiations for a US$30.0 million credit took place toward the end of April 1978. The Board approved the project on May 25, after a short discussion during which the Region was congratulated on the high proportion of total costs (80%) being met from IDA sources, and the Credit Agreement was signed on July 1. The project was declared effective on November 30, after a seven week postponement occasioned by a delay in appointiingsenior project staff. II. IMPLEMENTATION EXPERIENCE Start-up 17. Engineering studies by the two principal groups of design consultants (para. 14), which had commenced after appraisal but several -4 - months before effectiveness, proceeded smoothly at first but soon ran into difficulties. The river training and sedimentation study group, after three months initial investigation postulated three design options, all of which differed in varying degrees from the river and sediment control arrangements proposed in the SAR, and one of which would have required a siphon, tunnel or open channel from a new intake structure upstream of the existing intake, and incorporating a sediment settling basin also upstream of the existing intake. Model tests to assist in the design process took longer to set up and conduct than originally scheduled and were not completed until the middle of 1981. 18. Similarly, problems of data collection for the CMC redesign, and reconsideration of sediment removal alternatives from the CMC, led to delay of about one year (to early 1981) in commencement of reconstruction work on the CMC. 19. Six local consulting firms were engaged to design minor irrigation layouts for some 12,500 ha of the 18,000 ha stage I development towards the end of 1979, also about one year behind the SAR schedule. Sequence and Design Changes 20. Heavy flooding of the Kosi River in .June 1980 brought about a change in flow pattern and caused erosion of the left bank, which further complicated the design process for the river training and sedimentation study. River bank protection in the form of 19 rock-filled, gabion-lined spurs along a three mile sectior, of the left bank immediately below the supply channel to the intake was undertaken as a matter of urgency. 21. The June flood, which provided a forceful demonstration of the river's power and capriciousness, led to a reconsideration of the design options and a request to the consultants to produce quickly two reports, for an "upstream solution" (para. 17) and a lower capital cost "imaintenance solution". The Bank, supported by the Bank's consultant fluviologist (para. 12) who had originally appraised this component, clearly favored the maintenance solution, while the design consultants continued to lean towards the considerably higher cost upstream solution as the technically more desirable option. 22. The outline design for the maintenance solution was presented in February 1981. The essentials of this solution, as subsequently modified on the basis of Bank comment, were annual re-excavation (by dragline) of a trench along the bed of the Kosi River supply channel and the CMC intake channel up to the intake regulator to ensure adequate extraction from the river in the dry season; modification to the intake regulator in the shape of two tier intake gates to provide overflow flexibility; and a 350 m settling basin in the upper portion of the headreach, supplied with dredging equipment, and provided with a sediment disposal channel. 23. In May 1981, on being advised of these design changes, HMGN objected to the introduction of a settling basin with dredge on the grounds that they could not maintain effectively the dredging equipment, and would prefer to accept the alternative of routine winter closure of the CMC to enable periodic excavation of the accumulated sediment using hand labor and trucks. The Bank agreed, and the consultants undertook a further design revision which included remodeling of the intake regulator; addition of a vehicular service road along the south bank in the CMC headreach cutting to facilitate manual sediment removal; an emergency spillway; and the SAR proposed vortex tube sediment ejector and escape channel in the lower part of the headreach. 24. At this stage Bank staff, reflecting concern at the prospect of high project cost overruns, attributable to design changes and to underestimation of civil works needs and their estimated costs at preparation and appraisal, proposed thet the remaining civil works to be carried out under this project be reduced by deferring construction of the emergency spillway to a follow-on project and by restricting the service road to the upper two kilometer of the headreach. Final design modifications by the consultants took account of these changes, and the addition of a second vortex tube sediment ejector. The contract for the construction of the sediment control structures was finally awarded to an international contractor in November 1982, about four years behind SAR schedule, and was completed in 1985. 25. Meanwhile, design and reconstruction of the CMC, including the first vortex tube sediment ejector, and planning 'or the 18,000 ha stage I intensive development area, which proceeded independently of the river training and sedimentation study, were encountering problems of their own. The CMC restoration proved to be the least troublesome. Tenders were prepared efficiently by the consultants appointed for this purpose in March 1978 (para. 14) and the civil works requirements were classed into three priorities and progressively contracted, commencing in March 1979, to local (including Indian) contractors, whose progress was seriously hindered by shortages of steel and cement which were particularly severe in the early years. The quality of construction, while generally satisfactory, was inadequate in places, including a low level escape structure which collapsed in July 1981. 26. Planning and design of the Stage I development area, targeted at appraisal as 18,000 ha was reduced, because of concern at mounting cost overruns, to 12,000 ha and contracted to six local engineering consultants in early 1980, more than one year behind SAR estimate. Lack of experience on the part of these consultants resulted in slow progress and led eventually to the project consultants assigning their own design staff to speed up the work. Finally, towards the end of 1982, about three years behind SAR schedule, an international contractor was awarded a contract for the greater part of the revised Stage I development, the final area of which became 9,750 ha due to exclusion of about 1,400 ha of very sandy soils found to be unsuitable for irrigation, and to failure to perform on the part of a local contractor on a further 850 ha intended pilot area. - 6- Organization and Management 27. The size and scope of the project as designed and subsequently modified; the complex operating procedures required to control the irrigation water flow while reducing sediment influx to a minimum; and the integrated organizational arrangement whereby extension, research, training and monitoring of agricultural progress within the command area were to be controlled through the project presented a very considerable management task. The burder proved to be beyond the ability of the successive project managers and their often inexperienced staffs to perform adequately. 2e. As the attention of management became increasingly focused on the technical problems of river training and sediment control and the contract administration vf the CMC restoration and Stage I developments, and as implementation delays and cost overruns gradually increased, it became expedient to reduce both size and scope of the project. Accordingly, in mid 1983, restoration and improvement works beyond the main canal comprising new regulatory structures in the 54,000 ha of command area outside the, by now reduced, 12,000 ha stage I area, were postponed to a proposed follow-on project, while the tubewells and micro-hydroelectric pilot schemes were deleted. Later, in 1984, the project's agricultural wing uas dismantled and their responsibilities taken over by non-project agencies. These reductions in project size and scope afforded a further relief to management and enabled them to concentrate more, though still with limited success, on the complex operational and maintenance problems presented by the continued heavy intrusion of sediments into the system. III. PROJECT OUTCOME 29. Project achievements have fallen far below SAR expectations, prirarily due to the inability of the sediment control devices constructed under the project to significantly reduce sedimentation in the CMC distributory system and fields under the operating conditions prevailing in the project area to date. In consequence, and in the absence of adequate maintenance, the capacity of the water delivery system has been undergoing a pronounced and steady decline ever since the CtIC restoration and Stage I development works were completed, and this is severely limiting potential crop benefits. 30. Modifications at the CMC intake did not reduce appreciably the amount of sediment entering the CMC, which is estimated to run as high as 500,000 cubic meters per year (m3 Iyr). The vortex tube sediment ejectors, while effective in removing about 40% of the sand fraction (bed load sediment greater than 63 microns), amounting to about 100,000 m 3 /yr, also remove a high proportion of the water through the outfall channel, while permitting the remaining suspended sediments, including a large and agronomically undesirable silt fraction, to penetrate the downstream reaches of the CMC and settle in the distributory channels and fields. The - 7 - siltation rates are so rapid that many water courses have lost half their designed capacity in a single season. 31. Restricted supplies of irrigation water, especially at the tail ends of the tertiaries, quickly led to deliberate or careless damage to control structures and illegal abstractions by cutting of banks or inser- tion of pipes. These interferences increased the wastage in the head sections and further reduced the supplies to the tails, resulting in inequitable distribution. 32. Operation and maintenance has been suboptimal from the first year of operation of Stage 1 (1985), owing to staffing and budget constraints, and to difficulties in following what amounts to a complex water management system. 33. A comparison of yields and cropping intensities from a 1983 house- hold baseline survey in the Stage I area, carried out prior to development, and a 1987 socio-economic survey conducted after improvements had been effected, indicates that incremental yields and cropping intensities of principal crops--local and improved (HYV) paddy, and wheat--have been con- siderably less than estimated in the SAR, and that yield and cropping intensity increases in a control area adjacent to the Stage I area (which represents a "without project" situation), are little different from those obtained within the project improved area, partly reflecting widespread use in recent years of private tubewells, which have the important advantage of providing sediment-free water. Irrigated Crop SAR 1983 Baseline 1987 Socio-Economic Survey Before Survey Control Stare I Command Project Assumptions --------------------------ton/ha------------------- HYV Paddy 2.0 2.5 2.8 2.8 Local Late Paddy 1.5 1.9 2.0 2.2 Wheat 1.1 1.6 1.8 2.1 i.e. Irrigated HYV paddy yields rose only 12% between 1983 and 1987 in both the Stage I command and control areas while irrigated local late paddy yields have risen only 16% in the Stage I command area, although the incremental increase is a marginal 10% when compared with the control. Wheat yields show a higher response to improved irrigation, reflecting the greater benefit to be derived from irrigation in the dry season than in the monsoon, but nevertheless this is much less than was assumed in the SAR. 34. Similarly, while cropping intensity has risen from 115% for the baseline to 164% for the Stage I command area, it has also increased by 1987 to 153% for the control area. 35. The incremental production and benefits derived from the project are therefore likely to be very marginal compared with SAR expectations. IV. FINDINGS AND ISSUES A. Inadequate Preparation and Premature Lending 36. As noted above (para. 11), the feasibility study undertaken by consultants in preparation for the project did not include proposals for reducing sediment influx into, or preventing its deposition in, the CMC and its command area. Nevertheless, in spite of this weakness in preparation (which reflects an omission in the terms of reference and inputs allowed for the study rather than a failure of the consultant firm to meet its contract obligations), the September 1977 IDA pre-appraisal mission, after drastically reformulating the project as prepared and recognizing the need for major components for river and sediment control and flood protection, declared the project appraised at the conclusion of its mission (para. 12). Then, after its return to Bank headquarters the mission proposed a timetable for credit negotiation and Board presentation which was too tight to permit a reasonable standard of planning, design and costing of the recommended river and sediment control civil works prior to i gotiations (paras. 13-14). 37. In order to substantiate the technical and engineering dimensions of the river and sediment control and flood protection components a consultant fluviologist was immediately engaged to carry out a follow-up appraisal of these aspects (para. 12). His report, submitted in December, provided the basis for the SAR description of the principal engineering features and their costing. In his report the consultant noted that his designs were more conceptual than preliminary and based on "inadequate and sometimes not too reliable data". The report further recommended that the river training works would need to be studied on a mobile bed model. 38. The terms of reference for the design studies (described as "pre- project" studies in the consultant's report but actually overlapped into the project implementation phase during appraisal) were modified on the basis of his report. As noted above (paras. 17-18) the engineering design studies took much longer to complete than estimated by the appraisal mission, which accounted for much of the initial implementation delay. 39. Surprisingly, considering the very preliminary nature of the engineering design and the risk which this involved, none of the reviewers at the SAR draft review stage appears to have raised the issue of inadequacy in preparation, except for one perceptive CPS query as to why the engineering design had been structured in so late in relation to the implementation schedule. No explanation was given in the final SAR. 40. Engineering cost estimates at appraisal, which were determined on the basis of the conceptual design, later prov3d to be inadequate when the results of the engineering studies revealed the need for major design changes in order that a workable solution might be expected for the river training and sediment control problems. 41. The (pre-)appraisal mission decision to change the design studies from the status of a pre-project requirement, but instead to move the appraisal process forward and to have the studies carried out as an early stage implementation activity was, in the view of the audit, a serious appraisal misjudgement. It led to highly adverse consequences during the implementation stage which have persisted through to the follow-on Sunsari Morang Irrigation II project, as explained below. B. The Peril of Designing to Cost 42. As noted above (para. 17), the river training and sediment control study consultants quickly focused on the design difficulties of several aspects of the SAR conceptual design (guide island and sluice gate in particular) and suggested other possibly design options. Hydraulic and geologic surveys and model studies to investigate these options and develop an optimum solution took considerably longer than had been allowed for in the SAR imTlementation schedule. They did however contribute to a much fuller understanding of the problems of Kosi River training and CMC sediment control than had existed at appraisal. This greater understanding also gave rise to much discussion between the consultants and Bank (para. 21) as to the most appropriate design solution to adopt under the prevailing circumstance, with the design consultants advocating what they considered to be the more effective but considerably higher cost "upstream" solution, and the Bank, reflecting concern at the escalating project costs and acting on the advice of its consultant fluviologist, opting for the lower capital cost "maintenance" solution (para. 22). 43. A careful review of the Bank's files covering the period of this debate (late 1978 through to mid-1981) and discussions with a few of the Bank and consultant staff involved, clearly shows that the Bank's preference for the maintenance solution was heavily influenced by considerations of relative capital costs for civil works, while the consultant's overriding concerns largely revolved around the likely technical effectiveness of the two options. HMGN appears to have contributed little to this debate, which mainly took place in the UK-based offices and laboratories of the consultants or in Washington. Then, in May 1981, after consideration of the maintenance solution design as developed at that time, HMGN decided that effective maintenance of dredgir.gequipment would be impracticable under project conditions. Instead, HMGN's preferred option was for routine closure of the canal system to permit sediment removal as part of the regular maintenance cycle. This option, which did not require a settling basin or dredge in the headreach but necessitated some rurther redesign of the intake gates to reduce sediment entry into the CMC system to the greatest extent possible, was agreed to by the Bank notwithstanding the fact that a settling basin and dredge were essential features of the SAR conceptual plan agreed to by HMGN at negotiation, and, up to that time had been a feature of all options being considered. The - 10 - consultants thereupon finalized the design plans by April 1983 and these were constructed under a civil works contract awarded in the previous October. 44. Again, the Bank's position in aligning itself with HMGN appears to have been influenced heavily by project cost considerations with the prospect of heavy cost overruns if reductions in project scope or scale were not imposed. The increased cost estimates were attributable not only to delays in construction and the consequent impact of inflation, but also to an underestimation of the length of new distributory canals at project preparation, and to the extra costs incurred in cazrying out the Kosi River control and flood protection works as compared with SAR estimates. By mid 1981 cost estimates for the project as appraised had risen from an SAR figure of US$37.5 m to US$60.0 m. 45. As noted above (paras. 29-32), the subsequent failure of the sediment control devices constructed under the project to reduce appreciably the influx of sediments into the CMC or its distributory system within the Stage I development area is the dominant reason for the unsatisfactory project outcome. 46. This project experience affords an unusually clear illustration of the unfortunate consequences which can result when the imposition of a restrictive cost ceiling, in this case imposed by underestimation of costs at appraisal, impels those responsible for implementation to redesign closely to a preordained cost figure. When appraisal cost estimates are based on what are clearly conceptual engineering designs, as happened in this instance, and these designs need to be drastically modified or curtailed on the basis of implementation experience, prudence would suggest that an objective reassessment of all possible design options be made, free of predetermined cost considerations. In other words, it is the audit view that by mid 1981, when serious doubts had arisen about the validity of the SAR conceptual design package, and when HMGN had withdrawn its support for one of the principal elements of that design, a formal reappraisal of the project was in order and should have taken place. 47. The fact that a reappraisal did not take place at this time does not, however, in the view of the audit, absolve the Bank from the more serious fault of inadequate preparation in the pre-negotiation phase and premature lending (paras. 36-41). C. Economic Rate of Return Revisited 48. The PCR prepared by the Region claims an ERR of 13% for the 9,750 ha Stage I development area. However, the audit has reservations about this ERR, which depends heavily on data used in the SAR while disregarding more recent data derived from the 1983 baseline survvy and the 1987 socio- economic survey (para. 33). These surveys suggest that the "before proj- ect" and "without project" yields and production assumptions should be higher, and in consequence the incremental production lower, than those used in the ERR calculation. Other audit reservations are taken up below in paras. 51 and 52. - 11 - 49. A PCR prepared for HMGNlJ by the project consultants at the time of project completion in mid 1987, which correctly took account of the baseline and socio-economic survey data, concluded that the economic internal rate of returr.,based on January 1987 World Bank price projeclions for the main commodities, was 8.0%. The PCR derived the net benefit stream for this ERR as follows: ERR Including Estimated Costs & Benefits from Follow-on Project (Million NR, Constant 1987 Prices) Project Year Costs Benefits Net Flow 1979 11.64 (11.64) 1980 18.00 (18.00) 1981 24.83 (24.83) 1982 10.81 (10.81) 1983 12.51 (12.51) 1984 9.22 (9.22) 1985 83.32 8.16 (75.1(j 1986 107.23 16.32 (90.91) 1987 14.43 12.24 (2.19) 1988 5.24 12.24 7.00 1989 25.34 12.24 (13.10 1990 25.34 20.39 (4.95 1991 5.24 32.63 27.39i 1992 5.24 40.79 35.55 2028 5.24 40.79 35.55 ERR = 7.96% 50. The project consultant's PCR also provided two ERRs based on different assumptions, including an assumption of higher prices than forecast by the Bank which gave a rate of 10.7%; and a rate of 7.2% if yields and net returns are only half that assumption. 51. However, all the above ERRs (in paras. 48-50) have been predicated on the assumption that further sediment removal measures to be undertaken under a follow-on project would be operating effectively before the onset of the 1991/92 season. Both PCRs' assumptions for timings of these improvements can now be viewed as unrealistic because of slower than estimated progress in implementation of the follow-on project, while annual incremental production is meanwhile likely to be falling, rather than rising as assumed in the ERR calculations, due to the current high rates of sediment deposition in the CMC, Stage I distribution canals and farmers' fields. 1/ Project Completion Report, Sunsari Morang Irrigation Project Stage I, His Majesty's Government of Nepal, June 1987 (Revised). - 12 - 52. More importantly, it is the view of the audit that an ERR reestimation for the project per se at its completiorLwould have been more appropriate for a PCR, i.e. without consideration of further improvements to be effected in a follow-on project, given that the audited project as appraised included arrangements for resolving the sedimentation problem. 53. For the above reasons (paras. 51 and 52) the audit believes all the PCR rates of return to be optimistic in the light of field observations in November 1989, given the continuing seriousness of the sedimentation problem, which was not resolved through the project as intended at appraisal. An audit reworking of the project consultant's ERR based on World Bank price projections, but excluding the (still vncertain) costs and benefits of additional engineering investments for more effective sediment removal is presented below for comparison: ERR Excluding Costs & Benefits of a Follow-on Project (Million NR, Constant 1987 Prices) Project Year Costs enefits Net Flow 1979 11.64 (11.64) 1980 18.00 (18.00) 1981 24.83 (24.83) 1982 10.81 (10.81) 1983 12.51 (12.51) 1984 9.22 (9.22) 1985 83.32 8.16 (75.16) 1986 107.23 16.32 (90.91) 1987 14.43 12.24 (2.19) 1988 5.24 12.24 7.00 1989 5.24 12.24 7.00 1990 5.24 12.24 7.00 1991 5.24 12.24 7.00 1992 5.24 12.24 7.00 2028 5.24 12.24 7.00 ERR = 0.5% 54. The audit is strongly of the opinion that an ERR of 0.5% is the more realistic for the project Stage I development area per se, and is the more appropriate for evaluation purposes. The Department of Irrigation commentary (Attachment, page 2) reflects considerable concern at this low ERR and presents an alternate work table to support an EPR of 13%. This table is essentially the project consultant's high price scenario referred - 13 - to in para. 50, but with the additional estimated NR40 million post-project investment proposed for resolving the sedimentation problem deferred by three years (from '89-'90 to '92-'93) to account for construction delay encountered to date under the follow-on project. When these additional construction costs and their associated benefits are excluded, however, the ERR falls to 3.1%. Ex-post validation of the Borrower's revised estimated EPR must await an evaluation of the follow-on project (Sunsari Morang Irrigation II) after its completion in a few years, and is outside the purview of this ex-post audit for the Sunsari Morang Irrigation and Drainage Project.2/ 2/ The Region comments as follows on the rationale for the ERR methodology in the PCR: 1. The main reason for the differences in ERR estimates betweeen the PCR and the PPAR is in the treatment of post-project investments to prevent siltation. De-facto, the project became at an early stage (1981) a "time slice" of a long-term investment program aimed at improving irrigation water supply in the project area, and therefore qualified for a "time-slice" ERR methodology, as explained below. 2. Most Bank-financed irrigation projects that are both new and large take too long to complete in one lending operation; they are therefore supported through several "time-slice" projects. Typically the first "time-slice" project finances headworks (dam/barrage and/or main canals), and its economic viability depends on the implementation of the follow-on projects. In the case of a multi-phase project, the economic viability of the entire project is assessed at appraisal; and at completion, the ERR estimate of the completed "time slice" has to assume derivation of benefits resulting from follow-on investments, a methodology which has been recognized by the OED in the past but considered unacceptable by this audit (para. 52). 3. Furthermore, since the implementation of large irrigation projects may take 20 years ox more, many time-slice projects include improvements compared with their predecessors, resulting from experience gained, new technological developments, etc. Improvements from later time slices can clearly benefit returns from earlier investments. 4. At the time the PCR was prepared, it was expected that HMGN would soor endorse the desiltation solution recommended by the Bank. Taking this into consideration and assuming that the project was de-facto a "time slice", the ERR analysis appropriately included the costs and benefits resulting from anticipated desiltation facilities, provision for which was included in the Stage II time slice. However, at the time the draft audit was completed (nearly one year later), an international Panel of Experts appointed by HMGN to assess the alternative desiltation solutions completed its work, and it became clear that the cost of desiltation facilities would be much higher than Continued on next page - 14 - Continued from previous page previously estimated. Consequently, there were valid doubts about HMGN's commitment to implement the recommendations. Under these uncertainties, the audit decision to exclude costs and benefits of desiltation was appropriate. However, had tangible steps to initiate desiltation works started before the audit had been concluded, that decision could have been challenged on methodological grounds. Assuming that HMGN proceeds with the facilities as planned, benefits from both time slices should be captured during Stage II and reflected in the audit for that project. 5. Finally, a comment on the prices used in the ERR calculated by the project consultant (para. 50) would be appropriate here to provide the reader a more balanced assessment, given the fact that the audit based the recalculation of the ERR on the situation as it was known at end- 1989. The audit considered these prices high compared with the Bank's 1987 price projections. However, these prices are about the same as the Bank's 1989 projections, and therefore do not bias upwards the ERR (estimated by the consultants at 10.7%). - 15 - NEPAL SUNSARI MORANG IRRIGATION AND DRA1NAGE PROJECT (Credit 812-NEP) PROJECT COMPLETION REPORT February 5, 1990 Agriculture Operations Division Country Department 1 Asia Regional Office - 17 - NEPAL SUNSARI MORANGIRRIGATION AND DRAINAGE PROJECT PROJECT COMPLETION REPORT I. INTRODUCTION BackZround 1.01 Nepal is a landlocked country situated between China and India and occupying a total area of 147,000 sq km. Its population in 1988 totals about 18 million, of whom 94% live in rural areas. Population growth betweer.1970 and 1986 has averaged 2.6% a year. The country can be divided into three broad ecological zones--the Mountains, Hills and Terai. The Terai plain, part of the Gangetic basin, accounts for only 14% of the total area but 42% of the country's cultivated land and is the center of agricultural production in the country. 1.02 Agriculture, largely rainfed, dominates the country's economy. In 1987/88, it accounted for 55% of GDP, contributed about 80% of total export earnings, and provided employment to 93% of the working population. Increasing pressure on land due to rapid population growth has led to a situation where self-sufficiency in food production is dependent on better than average rainfall. During the past decade, the production of foodgrains has grown at only 1.5% per year whereas the production of cash crops has grown at 5% per year. However, due to the large proportion of grain production in the total (85% of the area sown), the overall annual rate of production growth was only 1.6%. Current targets for agricultural production call for an annual growth rate of about 5%. However, this goal appears to be difficult to achieve, especially compared to historical growth rates. For growth in agricultural production to keep pace with population, an estimated additional 45,000 ha of irrigation needs to be developed each year. 1.03 Nepal has abundant surface and groundwater irrigation resources. About 75% of Nepal's potential irrigated area is on the Terai. It is estimated that river discharges average about 150 billion cubic meter per year. However, only about 1.4 M ha of land are suitably located for command by gravity, of which about 0.9 M ha is already irrigated. Exploitable groundwater resources could irrigate about 0.5 M ha, of which 0.1 M ha has so far been developed. The Terai is traversed by eight major perennial rivers, the Mahakali, Karnali, Babai, Rapti, Narayani, Bagmati, Kosi and Kankai. All are tributaries to the Ganges. The many smaller streams also crossing the Terai are largely annual, but are nevertheless developed, in whole or in part, for irrigation, usually through farmer- constructed schemes. - 18 - 1.04 The development of surface water irrigation was initiated centuries ago by farmers who built temporary weirs to direct river flows into ungated intakes. Most of the schemes were in the Hills until the 1920s, when development started to expand into the Terai. Many such schemes are in operation and some have been improved through government assistance. In total, farmer managed irrigation schemes account for about 70% of the total irrigated area in Nepal; the remaining 30% comprise public irrigation schemes developed mainly with external assistance starting in the early 1950's. The Agricultural Development Bank of Nepal (ADBN) has supported private sector irrigation with loans to individuals for minor irrigation and for shallow tubewell development. ADBN has also provided a few loans to farmer groups for the rehabilitation or improvement of small surface irrigation schemes. 1.05 The goal of public irrigation development during the past decade has been to improve performance and management of irrigation schemes. Governmenitdistribution systems have been extended to serve increasingly smaller turnout commands, and attempts have been made to establish WUGs among the beneficiaries to improve water management at the micro- distribution levcl. Experience in Nepal and elsewhere has indicated that the traditional design of irrigation systems, with numerous gated structures, is unsatisfactory. Consequently, a modified design has been developed to permit relatively simple management at both goverrment and farmer levels. The existing system will be modified under the Sunsari Morang Irrigation II Project to provide irrigation according to strictly applied rotational water supply (RWS) principles. 1.06 The Bank's first involvement with irrigation in Nepal was the Birganj Irrigation Project (Narayani I) (Cr. 373-NEP, US$6.0 M, 1973) which had both surface and groundwater components. Development has continued with the Bhairawa-Lumbini Groundwater Project (Cr. 654-NEP, US$9.0 M, 1976) now being followed by Stage II (Cr. 1316-NEP, US$16.0 M, 1983); Narayani Zone Stage II (Cr. 856-NEP, US;'.4.0M, 1978) now being followed by Stage III (Cr. 1715-NEP, US$24.5 M, 1986); Sunsari Morang on the Kosi (Cr. 812-NEP, US$30.0 M, 1978), also being followed by Stage II 'Cr. 1814-NEP, US$40.0 M, 1987); and the Mahakali I Project (Cr. 1055-NEP, US$16.0 M, 1978) in the far western region, followed by Stage II (Cr. 1924-NEP, US$41.3, 1988). Several other IDA-assisted projects include small-scale irrigation components. These are: Second Rural Development Project (Cr. 939-NEP, US$11.0 M, 1975); Third Rural Development Project (Cr, 1727-NEP, US$19.1 M, 1986) and Hill Food Production Project (Cr. 1101-NEP, US$8.0 M, 1981). The main objective of Bank involvement in irrigation in the Terai continues to be better water distribution to the farm, through technical improvements and by more effective organization of water users. 1.07 In 1986, His Majesty's Government of Nepal (HMGN) requested IDA assistance to finance the second seven-year time slice of the project (Sunsari Morang II). Sunsari Morang II is to further expand the rehabilitation and improvement works initiated in the first stage project. - 19 - Sunsari Morang II was appraised in November/December 1986. The Credit was approved by the Board on June 2, 1987, signed on November 20, 1987 and became effective on February 3, 1988, as a follow-on project to Sunsari Morang I. II. PROJECT FORMULATION Project Identification, Preparation and Appraisal 2.01 The project was identified by a Bank mission which visited Nepal in March 1976. Preparation of the project was carried out in June 1977, jointly bv HIMGNand an international consulting firm. The project was appraised by the Bank in September 1977 for main irrigation components and in November 1977 for the Kosi river training and sediment control components. Negotiations were held in Washington during April 25-28, 1978; Board presentation took place on May 21, 1978; the Development Credit Agreement of US$30.0 M was signed on July 7, 1978 and the Credit became effective November 30, 1978. A Special Action Credit of US$4.0 M equivalent, with funds contributed by member States of the European Economic Community, was extended to the project by agreement dated August 9, 1979. 2.02 No specific field trip was made to collect information for this PCR since adequate data was available on file. Information sources used for the preparation of thi, PCR include the Staff Appraisal Report (SAR) of the Sunsari Morang Irrigation and Drainage Development Project (Report No. 1949a-NEP), draft PCR prepared by HMGN, supervision reports and correspondence files, and documentation related to the preparation and appraisal of the follow-up project. The Proiect at Appraisal 2.03 The Project Area. The Sunsari Morang Irrigation Project (SMIP) is located in the southern parts of the Sunsari and Morang Districts in the Kosi zone of Nepal. The total cultivable command area (CCA) of SMIP was originally defined as 66,000 ha, but has now been reduced to about 58,000 ha with the elimination of very sandy soil areas. The northern boundary of the project area is defined by the Chatra Main Canal (CMC), which originates at the Kosi River near the village of Chatra and terminates about 53 km to the east at the Bakra River; the southern boundary is the Nepal-India border. The East-West Highway runs within a few kilometers of the CMG throughout most of its length. 2.04 Early Development of the Project Area. Under the Kosi Project Agreement of 1954 between Government of India (GOI) and HMGN, GOI undertook to finance and construct the Chatra Canal Project in Nepal to irrigats some 66,000 ha using water diverted from the Kosi River. A canal intake structure was constructed near Chatra on the left bank of the river, about 20 km upstream of the Kosi barrage. Construction of the system started in - 20o- 1964 and, after trial operations of the canal between 1970 and 1975, the project works were handed over to HMGN in 1975. The intake and headreach of thleCMC were designed for a capacity of 45.3 cubic m/sec, equivalent to about 0.66 1/sec/ha over the original command area. The distributioni system was terminated at turnouts serving areas of 100-200 ha and, as a result, only about 35,000 ha received some irrigation supplies. 2.05 The total population of the SMIP area was estimated at 260,000, and the average family consists of about seven persons. There are eight towns within or near the project area. The rural population lives in small villages and is almost entirely supported by agriculture. The average farm size in Sunsari District is 1.95 ha. Some 9% of the holdings are larger than 5.0 ha and farm about 45* of the land; however, almost 50% of the holdings have less than 1.0 ha and farm only about 8% of the land. Fragmentation of holdings is common. Only a small number of farmers have tenancy holdings, and about one-third of the farmers hold at least part of their land as tenants. The cropping intensity at appraisal was estimated at. 118%. 2.06 Project Objectives. The objectives of the project were: (a) To restore the irrigation system to its original scope and capability; (b) To improve the :eliability of water deliveries, and therefore to increase farmers' confidence in the system; and (c) To accelerate agricultural development and thus increasing farmers' incomes and rural employment. 2.07 Project Scope. The project included the following components: (a) river control and flood protection works on the Kosi river in the vicinity of the Chatra Main Canal (CMC) intake; (b) sediment control arrangements at the CMC intake and along CMC; (c) restoration and improvement of the canal system and provision of new structures throughout 66,000 ha; (d) planning and design of complete minor distribution canals extending to outlets serving 10 ha groups of farms throughout an area of about 18,000 ha and, within this area, construction of canals and drains serving 6,400 ha; (e) drainage improvements covering 12,000 ha; (f) pilot schemes for tubewells, improved water management and cantl micro-hydroelectric developments; - 2J - (g) strengthening agricultural extension, research and training activities throughout Sunsari and Morang Districts; (h) equipment and vehicles for construction, survey and laboratory activities, and project operation and maintenance; (i) buildings for engineering and agricultural activities, and staff housing; and (j) technicalassistance. 2.08 Cost Estimate. Total project cost was estimated at US$37.5 M equivalent, including US$0.2 M equivalent for duties and taxes. The foreign exchange cost is estimat:edat US$20.5 M equivalent or 55% of the total and the local cost at US$17.0 M equivalent as follows: Foreign Items Local Exchange To al US$ million- ------ River and sediment control 1.1 1.5 2.6 Canal restoratioiiand improvement 4.9 5.5 10.4 Field channel and drainage 2.8 1.4 4.2 Pilot works 0.1 0.8 0.9 Buildings 0.7 0.2 0.9 Equipment - 3.7 3.7 Technical assistance 0.3 1.1 1.4 Establishment 1.7 1.0 2.7 Land acquisition 0.5 - 0.5 Physical contingencies 1.8 1.6 3.4 Price contingencies 3.1 3.7 6.8 Total Project Cost 17.0 20.5 37.5 2.09 Financing. The IDA Credit of US$30.0 M amounted to 80% of the total project cost and was to finance 100% of the foreign exchange cost and 56% of the local cost. HMGN was to contribute most of the remainder (US$7.2 M) with farmers providing labor estimated at US$0.3 M as shown below: Foreign Exchange Local Currency Total US$M % US$M % US$M % IDA contribution 20.5 100 9.5 56 30.0 86 IIMGNcontribution - - 7.2 42 7.2 19 Farmer contribution - - 0.3 2 0.3 1 Total Cost 20.5 17.0 37.5 - 22 - At the request of HMGN in February 1979, IDA extended in April 1979, a Special Action Credit of US$4.0 M equivalent to HMGN with funds contributed by the States of the European Community to alleviate HMON's difficulties in mobilizing domestic resources to support the project. 2.10 Procurement. Out of US$19.0 M net of physical and price contingencies and establishme.ntcosts of civil works and buildings, US$1.5 M (Kosi River works) was to be constructed by ICB, US$12.5 M by LCB and US$5.0 M by force account. The estimated cost of equipment, vehicles, and spare parts for construction, and )&M as well as survey and laboratory equipment was US$3.7 M. Most of these were to be imported items and to be procured under ICB contract. 2.11 ImplententationOrganizarion. Tne imDlemencation was to take five years. A governing board (Sunsari Morang Irrigation and Drainage DeveloprnentBoard) was to be created for executing the project under HMGN's Development Board Act of 1956. 2.12 Benefits. It was estimated that at full development, the project would increase foodgrain production 35,600 tons and generate about 10,000 and 5,000 additional full-time jobs in farm and non-farm sectors, respectively. Valued at projected international prices, the increased foodgrain production was estimated at US$2.0 M annually. The economic rate of return (ERR) for the project was estimated at 17%. III. PROJECT IMPLEMENTATION Effectiveness and Start-uR 3.01 The project made a good start by HMGN entering into contracts with international consulting firms. The contract for the restoration and improvement of the CMC was concluded in April 1978 and another contract with the same firm in November 1978 for in-service training, preparation of bidding documents and bid evaluation, and feasibility studies for further development of the project area. Another international consulting firm was awarded a contract in April 1978 for consulting services for the Kosi river training and CMC sediment control study, and also on the supervision of its construction contract. Five local consultants were awarded contracts in 1980 for survey and detailed design works of irrigation distribution networks in approximately 12,000 ha. On September 18, 1978, HMGN established the Sunsari Morang Irrigation Development Board (SMIDB). During mid-1980, a landslide in the upper catchment area of the Kosi River caused a major flood in the river, which endangered the CMC. IDA authorized the immediate construction of emergency flood protection works and financed under the project (US$3.0 M). Furthermore, during its initial period, project implementation was delayed by: (i) procurement problems for civil works contract; and (ii) insufficient fund allocations by TIMGNfor timely implementation of works. In addition, the performa;ict'oftlocil - 23 - contractors working on the restoration of the CMC was below standard. These problems were gradually resolved in the later stage of implementation, hut the closing of the Credit had to be extended by three years from June 30, 1984 to June 1987. Revision 3.02 As implementation of the project progressed, it became evident that the physical targets set at appraisal could not be achieved. A November 1982 review of the project showed a 2-3 year delay in project implementation as well as a cost overrun of about 75% above the 1978 appraisal estimate of US$37.5 million (Table 1). The cost overrun and the delays in implementation were due to an underestimation-of the works needed to rehabilitate the irrigation system throughout the command area down to 50 ha outlets; The slow and unsatisfactory performance of small local coatraccors; and a flood and landslide in the proiect area in 1980, which necessitated the unforseen emergerncyconstruction of river training works. Because of these difficulties and the cost overrun. a change in project scope was recommended. 3.03 In early 1981, IDA recommended that HMGN should revise the project scope to ensure that the planned total project cost would be within the financial resources available; this needed to be achieved by adjusting the magnitude of the physical works fgo various components. These changes included: (i) increasing the upgrading and extension of drains and irrigation canals to outlets serving groups of farms covering 10 ha from 6,400 ha to 12,000 ha of the command area; (ii) excluding the restoration and extension of the irrigation distribution system over the balance 54,000 ha command area; and (iii) deleting the pilot schemes for tubewells and canal micro hydro- electric developments. 3.04 The project consultants prepared revised cost estimates in 1982 based on the changes made, and these were reviewed jointly by HMGN and IDA. The cost of the revised work was estimated at US$41.9 M equivalent. Part of the cost increase was required for river bank protection to safeguard the CMC after a severe flood in the Kosi river in mid-1980. IDA suggested to HMGN that all major works required in the CMC headreach (intake to RD 32); development of Stage I area of about 12,000 ha; and development of blocks to provide water distribution down to 10 ha blocks and required drainage facilities over reduced area of 12,000 ha be packaged into single ICB contract to facilitate the implementation. HMGN agreed with IDA's suggestion and recommendations. The project scope was revised formally in July 1983. The revised project remained technically feasible and economically justified. Since the revision, project implementation has progressed satisfactorily. - 24 - Project Works / 3.05 River Control Works and Sediment Control. The main channel of the Kosi river had been shifting to the right, resulting in increased sediment deposition on the left, thus extending a large shoal which was progressively restricting flow in the subsidiary supply channel. To prevent further restriction of the subsidiary supply channel on the left side of the Kosi river, it was proposed that a guide island was to be constructed at the upper stream tip of the shoal. Depending on the findings of surveys, model tests and designs, the control and disposal of excess sediment entering the CMC from the Kosi was to be achieved by (i) cinstructing a sluice bay in the river to reduce sediment inflow; (ii) dredging coarse sediment from a settling basin; and (iii) extracting medivm grained material through a sedimont ejector in the canal bed. 3.06 Various schemes were considered for constructing works to limit sedimtent intake to the CMC and to ensure that supplies of river water would be available at the CMC headworks during the year. HMGN consultants had proposed the two principal alternatives as follows: (a) construction of various trainirg structures in the Kosi river to attract the flow to the eastern bank and prevent it moving to the West just upstream of the intake. A settling basin was to be constructed to remove much of the river sediment before it entered the canal system; and (b) construction of a new intake about 2.3 km upstream of the existing intake, connected to the CMC by a 2.4 km long tunnel discharge into a settling basin from which water was to enter the canal. HMGN rejected both schemes on cost grounds and also because DIHM ./ considered that O&M of the required dredging plant to remove sediment from the settling basin was not practicable. 3.07 The Bank's view on sediment control was to provide a desilting basin equipped with a dredger and located at the top of the headreach of the CMC. After long deliberation among HMGN, IDA and Consultant, it was decided to adopt a temporary solution. The solution implemented was a "maintenance scheme' requiring annual excavation for a supply channel to the intake, at the end of the wet season, to ensure adequate supplies in th. dry season. A large dragline was purchased for this work. Instead of the settling basin, the intake was remodeled to provide some reduction in sediment entering the canal, by the use of high level gates which would abstract water from the upper layers of the river during the wet season. The operation of the head regulator gates to the CMC have entailed random closures of the canal during periods of high irrigation demand in the wet season, and thus adversely affect the integrity of the irrigation service. Two vortex silt ejectors at RD 31 and RD 39.6 on the CMC were provided. j/ See Table 1 / Renamed the Department of Irrigation (DOI). - 25 - The vortex silt ejectors were to remove a proportion of bed load material entering the canal, while construction of a service road alongside the canal downstream of the intake facilitated the removal of coarse sediment deposited there. 3.08 Canal Restoration and Improvemilents.The works include strengthening CMC embankments, repair or reconstruction of about 70 main canal structures, strengthening embankments of all secondary and tertiary canals, repairing 200 existing regulators, constructing 35 new secondary canal regulators and 20 cross drainage structures. These works were to be initially carried out principally by local contractors. In 1979 and 1980, work proceeded with rehabilitation of the CMC but progress was unsatisfactory. Implementation of works was much slower than anticipated because of. (i) poor response to tender calls mainly from small local contractors who lacked experienced in concrete construction and earthworks, in which categories they usually prodoced work of inferior quality; (ii) the fact that some contractors were unable to carry out the e:tent of works as specified in t.hecontract and broke their contracts; ard (iii) shortage of qualified ste-f for investigation, survey, and desigr.work for the smaller distributaries. Other works were carried out by force account or petty contract, but this was not favored as HMGN's budgetary constraints sometimes resulted in fund not being available to pay the labor when it was due. 3.09 As it was apparent that reasonable progress could not be expected by use of local contractors, the revised canal restoration and improvement works were packaged into single international competitive bidding (ICB) contract. International tenders were called for the works in September 1982 and the contract was awarded in November 1982. Work was commenced in mid-March 1983 and ended June 1986. The contractor completed the works on schedule. 3.10 Some design modifications were carried out during construction. In particular, the second vortex tube sediment ejector was relocated 1.6 km downstream of its original position. This was because the escape channel as originally located had a minimum gradient and discharged into a watercourse meandering through cultivated land in the Kosi flood plain. One of the blocks in the command area (block 6) was not constructed, partly because the sandy nature of the soil which would have increased water requirement and partly to counterbalance the increased work quantities which arose in the remainder of the area due to their original underestimation. The following table summarizes actual achievements of canal restoration and improvements compared to appraisal targets: - 26 - Appraisal Revised Actual Achievement as Target Target Achievement % of Revised ----------------No. of structures---------------- CMC embankment completed in i984 100 Reconstruction of main 70 78 73 100 structure Regulator repairing 200 60 60 100 Construction of regulator 35 32 32 100 Construction of outlets 1,00c, 213 213 100 Construction of drainage 200 3 3 100 structures Construct.on spillways 14 25 25 100 Canal improvement (ha) 66,000 12,000 La 9,750 81 La It was further reduced to 9,750 ha from the revised area of 12,000 ha by eliminating sandy soil areas that were unsuitable for irrigation. 3.11 Farm Group Development. The original Sunsari Morang scheme, as handed over to HMGN by GOI (in 1975), proved to be incapable of providing an equitable and timely supply of water to all farmers of the SMIP command area. Within the command area of the project in the original design, farmers on the upper reaches of, and adjacent to, the canals were able to take whatever supply they needed whenever the system was running, and did so at the expense of the tail-end farmers. Under the project, the design of the improved system over 6,400 ha was intended to rectify the operational shortcomings of the original design. Water distribution and drainage system was to be improved by the construction of field canals commanding blocks of 50 ha and equipped with regulating outlets each serving groups of farms of 10ha as well as drains. It is fully gated down to intake structures serving chaks of about 50 ha. Thus, operationally, it must run with varying flows in all levels of the distribution system. Moreover, the highest discharge at 50 ha chak intakes is 66 1/sec. This requires that flow be subdivided within the chak since the stream size, which can be handled effectively by farmers with present irrigation techniques, is about 30 1/sec. It is already evident that the operation of this system poses a management problem to the operations staff of SMIP. However, only relatively minor modifications need to be made to the system design in order to improve its operational characteristics. The Stage II project would modify the design for the Stage I area, to simplify - 27 - operation. Similarly, a delivery system design is proposed for the new Stage II area, which would require relatively simple operations. The following table shows actual achievements of farm group development compared to appraisal targets: Appraisal Revised Actual Achievement as Item Target T.irget Achievement % of Revised Survey and design (ha) 12,000 11,750 ll,77tU 100 Construction of field 160 213 211 99 cantalr (km) Irrig_ciontblock dev. (ha) 6,400 9,7/50 9,730 100 3.12 Drainage Improvement. Under the project, to ensure propex drainage of the 6,400 ha selected for farm development group, natural drains serving that area were to be improved, starting from their outfalls and covering 12,000 ha. Works also included were straightening and increasing the capacities of natural channels and construction of priority drainage works of about 3,000 ha of low-lying areas. Road drainage was to be improved by construction of culverts. The following table summarizes actual achievements of drainage improvement compared to appraisal targets: Appraisal Revised Actual Achievement as Item Target Target Achievement % of Revised Natural channels (km) - 3 3 100 Excavation of new channels - 110 110 100 (12,000 ha) (km) Priority drainage (ha) 3,000 2,300 2,300 100 Road drainage culvert (no) - 60 60 100 The above data show that the project achieved all drainage works as revised, As a result of project works, about 3 km of natural channels have been improved, 110 km of new channels have been excavated, 2,250 ha of priority drainage have been constructed to reclaim low-lying areas and 60 drainage culverts have been installed. 3.13 Pilot Tubewells. The project was to install 25 tubewells, energized from the 33 kV transmission system, to test the technical, economic, and organizational feasibility of supplementing surface water supplies with groundwater, in areas such as in the tail of the canal system where irrigation water supplies are limited. In addition, three small hydroelectric plants were to be installed at canal falls to test the feasibility of generating local power in remote areas. However, - 28 - implementation of this component was suspended when the project was revised in mid-1983 (para. 3.03). 3,14 Agricultural Extension. Training and Research. Under the project, agricultural extension was to be reorganized in Sunsari and Morang districts, and the link with research and training was to be strengthened through better organization, regular staff training, and technical back- stopping. Extension services in each district, modeled after the Training and Visit (T&V) system, were to be headed by a District Agricultural Development Officer (DADO) who would supervise field staff, comprising Junior Technicians (JT), Junior Technical Assistants (JTA) and Panchayat Level Agricultural Assistants (PLAA). Furthermore, in support of extension, training, and research activities, the project was to provide vehicles, equipment, training centers, housing, offices and other buil'dings. 3.15 In early 1979, HMGITestablished an Agricultural Division within t';eproiect. However, the agricultural extension activities were harnpered by lack of Subject Matter Spzcialists (SMS), frequent staff transfers, inadequate allocation of budget by HMGN and poor logistic support, particularly vehicles. Progress of the agricultural extension component was slow, partly because of delayed project implementation including construction of the training center but also because of DOA's reluctance to provide technical support, as long as the administrative control was with the Project Board. The following summarizes staffing situation estimated at appraisal and actual numbers positioned as at November 30, 1980. Estimated Actual at Appraisal Strength SAO /I 1 1 DADO 2 1 SMS 7 2 TO /2 4 0 JT 18 14 JTA 31 27 PLAA 110 83 /1 Senior Agricultural Officer. L2 Training Officer. 3.16 A training center was constructed in the Jhumka seed multiplication farm in 1984. A DADO's office for Sunsari district was completed in 1984 after considerable delay due to difficulties involved in land acquisition and selecting the suitable sites for Constructing subcenters. Ten training subcenters as plannied at auppr*is.fl work constructed in 1983 and 198/4, and l ?2()0-toi itipti I. .t.ot I *ic ilitv wj.-s also erected in 1984. - 29 - 3.17 In mid-1984, IUMGNtransferred the command of extension staff from the project Board to the Ministry of Agriculture. The work resporsibilities were shifted to the Regional Directorate of the Department of Agriculture (DOA). Staff on secondment from DOA were transferred to the Regional Director and the employment of the staff directly employed by the Board was terminated. Construction of 31 quarters to be completed under the project for JTs and JTAs was later decided to carry out under the IDA assisted Agricultural Extension and Research Project (Cr. 1100-NEP, US$17.5 M, 1981). In-service trair,ingfor field extension workers at Jhumka farm and farmer training on improved cultural practices under irrigated conditions were strengthened under the ongoing IDA assisted Agricultural Extension Is Project (Cr. 1570-NEP, tUS$7.2M, 1985). 3.18 The extensior.support to the co!amandarea has deteriorated since the project's Agricultural Division was integrated into the T&V system which is organiizedunder a line department of DOA on a district basis. One PLAA was serving for about 1.100 farm families, which is inadequate. Under the Stage II project this weakness will be rectified as the project authority is to be provided with key extension staff for maintaining better coordination between the project management and DOA. 3.19 Future Feasibility Study. At appraisal, it was agreed that HMGN would complete the feasibility studies of the follow-on project of the development with assistance and advice from the project consultants. However, HMGN was slow in taking up actions for preparing the follow-on project. In Ma'-ch1984, IDA reminded that HMGN should initiate carrying out surveys and inventories of the existing canal network and structures in need of improvement or replacement in the remaining 55,000 ha. The project consultants were to assist the project authority in selecting the local consultant, preparing the TOR and supervising the surveys and inventories of the existing distribution systems for the Stage II project. In mid- March 1984, HMGN authorized the project consultant to carry out feasibility study for the Stage II project. Project Cost 3.20 Project costs as estimated at appraisal is compared with actual expenditures as follows: - 30 - Revised Za Appraisal Estimate Actual Actual Expenditure Components Estimate (1983) Expenditure as % of Appraisal ---------------NRs M---------------- Civil works 326.7 381.4 ab 396.3 121 Equipment 50.9 63.4 53.4 105 Technical services 23.4 42.2 51.1 218 Establishment 45.4 52.8 33.1 73 Land acquisition 3.6 13.2 71.1 197 450.0 553.0 541.0 120 La After revision of project scope, total cost is equivalent to US$41.9 M. Exchange rate: NRs 13.20 - US$1.00. b Includes customs duties and taxes of about NRs 2.6 million. 3.21 A detailed breakdown of actual project expenditures by components by years compared to appraisal estimates is given in Table 2. At appraisal, total project cost was estimated at NRs 450.0 M or US$37.5 M. The revised estimate of expenditures for project completion as assessed in December 1982 without revising the project scope amounted to NRs 71.2 M or US$66.0 M equivalent (see Table 3). The progress of actual yearly expenditure in comparison with estimates at appraisal and as revised is detailed as follows: Appraisal Revised Actual Year Estimate Estimate Expenditure ----- NRs M------ 1978/79 61.2 9.3 7.8 1979/80 122.4 28.0 38.4 1980/81 112.8 101.1 74.2 1981/82 111.6 38,8 45.1 1982/83 42.0 97.4 31.8 1983,/84 - 161.0 41.4 1984/85 - 117.5 81.7 1985/86 - - 146.7 1986/87 - - 73.9 Total 450.0 553.1 541.0 - 31 - 3.22 The main reason for underestimation of the project at appraisal (about 20%), appears to have been the lack of experience by HMGN with high quality project works. Cost estimates, based on actual implementation, were not available. Another reason for increased costs was an underestimation of construction required both in terms of number of structures and length of new canals required to convey irrigation water to the 50 ha outlets. In addition, work contemplated at appraisal as minor rehabilitation of existing structures in the Kosi river and flood protection works in many cases turned out to be substantial restoration, or even replacement involving cost greatly exceeding estimates. Actual expenditures exceeded appraisal estimates by NRs 91.0 M or 20%. The higher than estimated exchange rate of US dollars to Nepali rupees provided the project with substantially more Nepali rupee funds than that estimated at appraisal and enabled the completion of the project works. Disbursement 3.23 Comparison of actual disbursements with the appraisal estimate is presented in Table 4. Disbursement lagged behind appraisal estimates throughout project implementation. This was due to: (i) slow progress with civil works by inexperienced local contractors (para 3.08); (ii) changes in project scope and mode of major civil works procurement from LCB to ICB which caused delay initially in preparation of bidding procedures (para 3.09); and (iii) poor performance of ICB contractor in the early stage of implementation. The disbursement rate had been poor and only about 47% of the credit was disbursed by the original credit closing date of June 30, 1984. Disbursement improved significantly after the contractor had mobilized sufficient number of equipment and manpower and construction works progressed. The Credit was closed on January 20, 1988, with undisbursed balance of US$6,704 equivalent and was cancelled. 3.24 The Bank acted as Administrator of the EEC Special Action account on behalf of the EEC and its member states. The total amount disbursed was US$3.3 M equivalent; US$0.7 M less than projected because of the fluctuations in exchange values of the allocated European currencies versus the US dollar. To enable the early disbursement of this account, disbursements from the IDA credit under civil works were delayed unt4l the corresponding category under the EEC credit is fully ucilized. This was also one of the reasons for slow disbursement of IDA credit in the early stage of implementation. The EEC Special Action Account was closed on December 31, 1982. Procurement 3.25 The implementation schedule at appraisal gave different completion dates for the six groups of works that ranged from 1979 for the building to 1983 for the field channels and drains. The majority of the buildings were completed approximately one year behind schedule with the remaining works experiencing longer delays; generally about three years. Some of the delays were due to an insufficient budget allocation by HMGN to perform the - 32 - procurement functions or unfamiliarity with the requirements of the bidding procedure. However, the major delays were expeiienced because there were considerable changes in project scope and the type of works (para. 3.02). At project headquarters, one laboratory, one workshop and 13 units of staff housings were constructed as planned. 3.26 Originally, ICB contracting was limited to headworks and sedimentation exclusion arrangements but later expanded to block development. Subsequently, the kind and quantity of equipment and vehicles itemized were reexamined by the project consultant and ICB contractor during implementation, and procurement was made accordingly (IDA approved additional plant and vehicles to the tune of NRs 40 M on August 6, 1979). Most of the engineering equipment and vehicles were procured through ICB. The majority of the goods were actually delivered in 1981 or 1982. However, this delay did not present a problem. In fact, deliveries coincided conveniently with delayed works contracts and provided the contractors or department forces with goods they needed in a rather timely fashion. 3.27 The fleet of machinery and vehicles actually procured is generally consistent with items proposed during appraisal (Table 5). A substantial section of the fleet had been loaned to the ICB contractor during construction period. Some 50 pieces of project equipment, including earth moving machinery through tractors and vehicles to small machines, had been unserviceable due to lack of spare parts. The project authority indicated an urgent need for an improved mechanism for spare part supply. In view of the fact that the project is the largest irrigation development in Nepal on which construction works are expected to continue for many years, IDA repeatedly urged DOI to consider establishment of a facility within the project area for procurement, storage and supply of spares for maintenance of all equipment. Compliance with Credit Covenants 3.28 HMGN complied with all covenants (see Annex 1 for details). IV. INSTITUTIONAL PERFORMANCE Project Organization 4.01 Sunsari Morang Irrigation Development Board (SMIDB). In accordance with HMGN's established policy, a governing Board, the SMIDB was formed in 1979 to execute the project. The Board was chaired by the Secretary, Ministry of Water Resources, and ten members, including representatives from the Ministries of Agriculture, Finance, and Land Reform, and the National Planning Commission, the Director Generals of Department of Irrigation (DOI) and of Department of Agriculture (DOA), the Directors of the Eastern Regional Offices of DOI and DOA, and the Project Manager of the project who was the Member-Secretary. The SMIDB's functions - 33 - were to: (i) approve annual budget proposals for implementing and operating the project; (ii) appoint senior staff; (iii) award large contracts and consultant's agreements; (iv) prepare semi-annual reviews of implementation programs and progress; (v) coordinate policy matters on programming, budget, and finance; and (vi) determine the level of water charges. The organization of the Board has basically been implemented as originally anticipated. Over the years, SMIDB has become increasingly effective in carrying out the project which was in large measure due to improvements and to the continuity in management since 1976. 4.02 Proiect Management. The project organization is controlled by the Project Manager who had three wings under him, namely, the Planning and Control Wing, comprised of five divisions; the Engineering Wing, comprised of six divisions; and the Agriculture Wing, comprised of two divisions. In the earlier stage of implementation, the project organization had faced certain constraints. The main constraint had been the shortage of experienced engineering and agricultural staff. In practice, all staff working on the project had been seconded from the DOI and DOA, and staff appointments and promotion had to take place within the rules of the HMGN public service commission. This limited salaries and other incentives that would be required to attract competent staff. As a result, staff were not adequately motivated to work, mainly because of the lack of financial incentives. Also, the Agriculture Division was generally inadequately staffed. This problem was largely resolved as the project progressed. Although there were setbacks in drawing competent staff from the concerned departments to implement the project, the project organization was developed into one of the stronger irrigation implementation organizations in Nepal. The same project organization set-up under the project will implement the Stage II Project with some structural modification (item iii, para 8.01). Water Users' Groups (WUGs) 4.03 As was intended during appraisal, the farmers in the project area were organized in water user's group for the purpose of constructing and maintaining field channels as well as for water distiibution within the tertiary outlets, serving groups of farms of about 50 ha. The 50 ha outlets were further divided into five unit of 10 ha suboutlets. Within the 10 ha suboutlet areas, field channels and drains were to be installed by cooperative action of local farm labor under the supervision of project staff. As a first step towards this objective, the project was to establish WUGs in each of the 10 ha units. Initially, the formation of WUGs was slow (about 25% of that required in eacly 1986) and the WUGs had limited effectiveness for organizing the water distribution as the irrigation water supply was erratic. The main reason for delay of organizing WUG was that the contractor installed the 10 ha outlets as the last activity in the implementation. Without these outlets in place and with uncertainty about water availability, no farmers could be mobilized. It was found that farmers are normally not cooperative in organizing WUGs and constructing field channels and drains until they ictually see that the construction of tertiary canal is done and water is available in the - 34 - tertiary canal. Since the completion of construction of irrigation system in mid-1987 and reliable supply of Irrigation water is assured, the formation of WUGs has improved and marked its target by end of 1987. The following table summarizes target set and actual achievement by blocks: No. of No. of Achievement Block No. Target WUG formed Member Farmer as % of Target I 1'4 184 675 100 II 204 204 1,013 100 III 167 167 810 100 IV 217 217 945 100 V 218 218 1.154 100 Total 990 990 4,781 100 Coordination of Agricultural Development Activity 4.04 No special arrangements were made during appraisal for the coordination of the agricultural development activities on the project level involving coordination of research and extension, input supply, credit, etc. In theory, such coordination was to be realized through the existing district and regional coordination committee. However, in practice, these committees involve themselves only with policy matters and are not really designed to meet the day to day requirements of a specific project. Thus, until the removal of project's agricultural staff to DOA in 1983, the Agriculture Division carried out the necessary coordination activities with other agencies. After transfer of the project's extension staff to DOA, virtually no formal coordination between the project management and DOA was maintained. This weakness is expected to be strengthened under the Stage II project by assigning a few key agricultural staff to the project for coordination purpose. Monitoring and Evaluation (M&E) 4.05 Monitoring consisted of: (i) testing of materials and the quality achieved in civil works, under the guidance of the project consultants; (ii) the project implementation progress and costs of the works; (iii) observations of the behavior of the Kosi river and sediment inflow into the CMC; (iv) measurements of canal discharges and losses; (v) annual compilation of change in cropping patterns and yields and the impact of extension services; and (vi) monthly observation of the depth to watertable from the ground level. At appraisal, it was agreed that the above work would be carried out by the project staff of each concerned division and no formal establishment of a monitoring and evaluation unit was mandated for the project. Monitoring activities were confined mainly to routine measurement of canal discharge and sediment inflow on a limited scale. 4.06 Monitoring of agricultural development was considered an important project activity. At appraisal, it was agreed with the project board that - 35 - the project office, in collaboration with the Evaluation and Project Analysis Division of the DOA, would conduct a field survey, collect agricultural data and carry out an evaluation study of the project results. Project staff together with DOA carried out a survey during the dry season of 1983 in the 12,000 ha implementation area and their report was submitted in September 1983. Monitoring by the Agriculture Division was limited to crop yields and cropping patterns and no systematic assessment of the overall impact of irrigation and of agricultural support services had been carried out. Since the removal of the Agriculture Division from the project in 1984, direct access by project office to agricultural performance data on the project area was limited. The second agricultural survey was therefore contracted out to the local consulting firm and their report was submitted in July 1988. However, the quality of the report was poor. 4.07 In early 1986, the project authority established an M&E unit, headed by an Executive Engineer to strengthen monitoring activities. Under the Stage II project, the M&E unit has been further strengthened by augmenting additional staff and facilities, and broadened monitoring activities. Accounts and Audits 4.08 There were delays in setting up a proper accounting procedure at the start of the project. Initially, the accounts prepared were based on HMGN's practice, but over the years procedures were modified and improved. The project organization's accounting system was considered adequate. Auditing of the project's accounts and financial statements were carried out annually by a private accounting firm working under contract with the Auditor General's Office. The performance of HMGN in fulfilling the IDA audit and accounting requirements was weak, and in most cases, the audits of accounts were late. This is partly because expenditures related to IDA- financed projects are audited by the Auditor General as part of the overall financial records of the implementing government department and the Auditor General does not release the full audit report until it is cleared through several steps. HMGN had given assurance that audit reports would be submitted within six months after the end of the fiscal year. However, this did not prove to be possible; a period of approximately one year should have been allowed for submission of audit reports. The Stage II project has incorporated submission of required audit reports within 12 months after the end of the fiscal year. 4.09 To ensure timely compliance with covenants related to auditing requirements in IDA-supported projects, IDA has been helping project accounting staff to prepare separate project accounts in such detail as required for audit. IDA has proposed the format and content of auditor's report on accounts that can be prepared by an independent auditor acceptable to IDA, without waiting for the full audit report of the Auditor General on the concerned department's overall annual financial account. Similar procedures are being instituted for specific audit certification of - 36 - all withdrawal applications based on statements of expenditure. The audits confirmed proper management of the funds allocated to the project. Cost Recovery 4.10 The credit covenants required HMGN to recover from the beneficiaries by msans of water charges (i) the full cost of operation and maintenance; and (ii) in progressive steps over a reasonable period of time, the investment cost of the project, having due regard to the incenuive of farmers to use the irrigation water and to their capacity to pay. Water rates were to be reviewed, and, if necessary, to be revised not less than every three years. 4.11 Thus far, water charges have been introduced in only a few donor- financed irrigation projects, including projects financed by IDA. In 1975, HMGN increased water charges from NRs 6 to NRs 64 per ha cropped for all public irrigation projects in Nepal, but made only nominal efforts to enforce collection. In 1986, HMGN simplified the water charge system to a flat annual rate for any irrigated area and raised the charge to NRs 200/ha/yr irrespective of the crops grown or cropping intensity. The new charge system has been introduced in the project area for crop year starting November 1986 (1986/87 dry season). In 1986, HMGN also decided that water charges should be collected from landowners, along with land taxes, by the staff of the District office of the Land Revenue Department. Collections were to be paid into the general revenue account of the Treasury. Although there has been insufficient time to fully assess the impact of the flat annual charge, preliminary results indicate that unless a reliable supply of irrigation water is assured to the beneficiaries, there can only be a limited increase in cost recovery. 4.12 The current poor quality of the irrigation service caused by the 1986 and 1987 flood damages and heavy ingress of silt in the canal system is the main reason that farmers are reluctant to pay water charges. However, although reliable service is a prerequisite to improved recovery, it is unlikely to have a significant effect without an appropriate collection mechanism. HMGN had agreed to review performance of the collection of water charges, and if required, revise the collection procedures of water charges in the project area and adjust the rates of water charges to cover the costs of operation and maintenance. Under the ADB-assisted Advisory TA, a study on cost recovery of all public irrigation projects was recently carried out. HMGN had formed a task force in June 1988 on irrigation cost recovery representing concerned staff from the Ministries. Terms of reference jointly prepared by HMr'¶Cand ADB for the study were reviewed by IDA. Based on the study, the t-sk force prepared a draft action plan. On April 4, 1989, HMGN adopted task force recommendations to implement cost recovery measures (Nepal Gazette, No. 50). Under the Stage II project, the Water Charges Assessment Unit would be created to deal with assessing and preparing the billing notice in the project area which would greatly improve performance of collection of water charges. - 37 - V. AGRICULTURAL IMPACT Impact Area 5.01 At appraisal, it was estimated that only about 35,000 ha or 53% out of 66,000 ha received some irrigation supplies because of canal defects: there were no flow regulating structures in the secondary and tertiary canals and construction of distributary canals was terminated wherever each canal decreased in capacity. Drainage in the system was limited to local cross-drainage syphones, many damaged beyond repair. Overland drainage was field-to-field and discharged to natural drainage lines. Thus, inundation of crops frequently occurred in low-lying areas (15% of the command). The project was to restore the conveyance capacity of the svstem, enabling water to be delivered to 66,000 ha. 5.02 As mentioned in para 3.03, the project scope was redefined to intensively develop about 12,000 ha in 1983 rather thaT only to restore and upgrade the existing system throughout the whole of th CMC command area of 66,000 ha. The 12,000 ha intended for improvement wa further reduced to 9,750 ha because: (1) about 1,400 ha of very sandy s Is within the area were found to be unsuitable for surface water irrig. ion; and (ii) a local contractor who was awarded a contract for works on 850 ha failed to perform. For the purpose of evaluating the agricultural impact of the project, this reduced size is taken as the basis of comparison with agricultural production targets as indicated in the SAR. 5.03 A farm survey was conducted after completion of the Stage I project (early 1988). However, reliable data on agricultural production could not be collected because many farms sampled and surveyed were victims of the 1986 and 1987 floods, and they did not accurately represent the project's impact on agricultural performance. Most of the descriptions given below are based on the experience with irrigation under the Terai conditions and data obtained during IDA supervisions. Cropped Area and Yields 5.04 Although works on restoration and improvement of the canal system were reduced to 9,750 ha (para 5.02), river control and flood protection works on the Kosi River (in the vicinity of CMC intake) and sediment control arrangements at the CMC intake carried out under the project (paras 3.06 - 3.07), have contributed to improving the conveyance capacity of the system. It is estimated that, as a result of such works, about 11,250 ha of additional area (in addition to 9,750 ha) have been receiving water supply more reliably than 35,000 ha within the total command area (para 2.04). 5.05 The data obtained during the IDA supervision missions indicated an improvement in crop intensities; farming practices and yields since the project began. A comparison of the cropping pattern projected during appraisal (1978/79) and 1988/89 is shown below: - 38 - Appraisal Estimate PCR Estimate 1/ At At Appraisal Future Future Present At Full (1978/79) without with (1988/89) Development - -( of net CCA) - ---- - - - of net CCA) - - - Local paddy 70 65 65 20 12 HYV paddy 2 7 15 75 85 Jute 30 30 30 10 5 Wheat 6 6 10 28 31 Maize - - - 1 3 Pulses 2 2 3 18 20 Oilseeds 2 2 3 7 14 Vegetables 4 4 4 3 3 Potato - - 1 3 Sugarcane 2 2 4 2 2 Total 118 118 134 165 178 I/ Estimate by the project consultant (1988). 5.06 The above data shows that the cropping intensity has increased from 118% at appraisal to 165% in 1988/89, an average annual rate of about 5% per year. Some areas where reliable water supply have been assured, the cropping intensity has already reached to 178%. During the same period, the area cropped to HYV paddy drastically increased from 2% to 75% while local paddy area decreased from 70% to 20%. Areas cropped to wheat, pulses, and oilseeds all increased while jute area has considerably decreased. 5.07 A comparison between the crop yields projected during appraisal (1978/79) and PCR estimate (1988/89) is shown below: - 39 - Appraisal Estimate PCR Estimate At At At Full Appraisal Future Future Present Develop. (1978/79) without with (1988/89) (1991/92) / - - - - - - - (t/ha) - - - - - (t/ha)- - - - - Local paddy 1.1 1.3 2.3 1.9 2.3 HYV paddy 1.6 2.0 3.0 2.8 3.4 Wheat 0.8 1.0 2.5 1.8 2.7 Pulses 0.4 0.5 0.8 0.6 1.0 Oilseeds 0.4 0.4 0.8 0.6 1.0 Potato - - 11.5 21.0 i/ Estimates by the project consultant and assessment made by supervision missions (1988). Yields of major crops such as paddy and wheat have shown a markedly upward trend. The increase in yields is attributable to growing of selected HYVs suitable for the project area (both paddy and wheat). Crop Production 5.08 Based on the expected future cropping patterns and yields levels, crop production for the 'future with project' situation from the 9,750 ha is given in Table 6 and summarized below: Future At Full Without Development Incremental (SAR) (FCR) -Production ----------------------(ton)---------------------- Paddy 7,820 30,870 23,050 Wheat 468 8,162 7,694 Oilseeds 78 1,287 1,209 Potato 0 7,032 7,032 Vegetable 3,120 6,836 3,716 Pulses 78 1,872 1,794 - 40 - VI. ECONOMIC EVALUATION Economic Rate of Return (ERR) 6.01 The economic viability of the project has been reassessed, taking into account the actual implementation schedule, actual incurred costs, and estimates of agricultural benefits assessed by the project consultant and supervision missions. The cost streams include the economic cost of all civil works, administration and engineering and incremental O&M costs. Agricultural benefits have been based on the implementation of effective desiltation facilities envisaged in the appraisal of the Sunsari Morang Irrigation II Project (Report No. 6655-NEP) which represent the best assessment of benefits to the project when this PCR was under preparation. Cost and benefit streams generated from these data are given in Table 7; discounting these costs and benefits streams over 50 years gives an ERR of 13% compared with the appraisal estimates of 17% for the entire project. The lower than estimated at appraisal ERR is mainly a result of (i) a delay in the implementation period by about three years (para 3.01); (ii) scaling down of total project area (para 3.02); and (iii) higher than expected unit area development costs at appraisal. VII. THE BANK PERFORMANCE Preparation and Aipraisal 7.01 The issues raised during project preparation and appraisal were generally well identified. IDA was concerned with the Kosi river training and siltation problems in the Chatra Main Canal System. Of aqual concern was the adequacy of maintenance of the main canal system especially heavy loads of silt ingress into the irrigation system. The appraisal report outlined a well-balanced development plan consisting of closely interdependent project development. However, during the implementation period, HMGN was reluctant to actively pursue implementing the river training and desiltation works because of the high cost involved and uncertainty about the performance of the structures to be built. Studies of river and canal headworks improvement and schemes for desilting project water supply are included in the Stage II Project (para 8.01). The problems encountered with the estimated costs for civil works, although based on the best estimate of the unit prices for equipment and materials could not have been foreseen at appraisal due to the limited data and expeiience available at that time (para 3.02). 7.02 The covenants included in the Credit Agreement, such as the recruitment of a consulting firm and the setting up of a Project Board as well as the appointment of a Project Manager were all appropriate and necessary. One covenant may have been too strong, which is IDA's insistence that water would not be released to areas commanded by any field - 41 - canal until a water user's group has been formed on the field canal and sound distribution routines established. Farmers were reluctant to form the group until they were actually supplied with reliable irrigation water. Supervision 7.03 Eighteen supervision missions visited the project approximately every six months. The size of a mission ranged from one to two persons. Following each mission, letters summarizing the main findings and corrective actions suggested were sent to HMGN. The supervision mission's recommendations were some of the important topics discussed in the project board meetings. Usually positive reactions were given to the recommendations. 7.04 During the later years of the implementation period, the supervision missions were concerned with reporting the shortage of maintenance funds for the canal system especially desiltation works in the CMC, lack of spare parts for equipment and vehicles, repair of damages caused by 1986 and 1987 floods, and disbursement of credit funds. Supervision missions were very helpful in pursuading DOI/MOWR and the Ministry of Finance (MOF) to meet the required funds for desiltation works, procurement of spare parts and flood control measures. The disbursement of credit funds was also expedited. The Bank was flexible in appraisal targets. When it became clear that the pilot scheme component for tubewells, and canal micro-hydroelectric developments could not be achieved as designed due to the tight budgetary situation, the Bank agreed to HMGN's proposal to drop the component from the project. 7.05 Other main issues raised during the implementation period by the supervision missions, which wer3 largely resolved included operation of pilot water management schemes and delays in staffing the M&E unit and coordination with the Department of Agriculture. VIII. LESSONS LEARNED AND CONCLUSION 8.01 While overall the project was successful, there were several problem areas in the implementation of the project. These were analyzed and corrective measures incorporated in the follow-on project, the Sunsari Morang Irrigation II Project. They were: Lessons Learned (i) The Need for a Reliable Desilting Facility. When the Stage I project was originally formulated, the plans included the provision of a desilting basin equipped with a dredger and located at the top of the headreach of the CMC. However, when the project concept was revised in 1983, it was decided to delay provision of this facility because of: (i) its high capital and O&M costs compared to the area it would initially protect; and (ii) concern - 42 - that the project authority would not be capable of operating and maintaining the mechanical equipment. The solution implemented was a "settling reach" on the top 2.0 km of the headreach (with an access road), which was to be cleared mechanically during canal closures, and provision of two vortex silt ejectors at RD 31 and RD 39.6 on the CMC. HMGN consultants also submitted proposals for a complex operation of the head regulatory gates to the CMC, storage of silt on the headreach, and operation of the silt ejectors to minimize silting on the distribution system. The procedures proposed were never fully implemented, and there are indications that they would be difficult to apply in practice. Furthermore, they would entail random closures of the canal during periods of high irrigation demand in the kharif, and thus would adversely affect the reliability of the irrigation service to farmers. Irrigation water supplied to the newly completed Stage I area of 9,750 ha after late July 1986 demonstrated the seriousness of the siltation problem and the urgent need for a viable solution. Some 140,000 cm of sediment were cleared from the settling reach during the previous rabi canal closure to create room for the storage of silt, and the silt ejector at RD 39.6 was in operation to protect the State I area. However, substantial silt passed the ejector during the kharif season to settle on the newly opened canal system and significantly reduce its carrying capacity. The minor canals and watercourses suffered in particular, and many of them lost at least half their capacity in this one season. Attempts by farmers to clean some watercourses could not keep pace with the rate of siltation. The overall result was that the canal system could not operate in a reliable manner and could not provide the planned water supply. It is obvious that continued investment in an improved irrigation network subject to this level of siltation would not be viable. If canal operations were shut down during periods of high silt load on the Kosi River, kharif cropping would be reduced to an essentially rainfed situation; the rabi cropping could possibly also suffer to such an extent that the economic benefits of the Stage II project investments would drop. There appeared to be no readily obvious practical alternative to a desilting facility which could remove silt on a continuous basis during periods when the intake water carries large quantities of suspended solids. The Stage II project has provided funds for studying headworks improvement and canal siltation problems, and for implementing the facilities as a result of the studies. (ii) Appropriate Irrigation System Design for Subsequent Management. The irrigation network was designed according to the state of the art in the mid-1970s. Since then, significant experience has accumulated. The operation of the improved irrigation system designed by the project authority's consultants for the Stage I area, although technically sound, has proved to be difficult to manage. Water distribution requires a ccmplex operational plan - 43 - because of the large number of control gates in the canal system; in turn, the project authoritycannot effectively manage the large number of project operationsstaff required. The limited length of watercourses, the lack of a field channel system,and the complications of managinghigh and variableflows within the chaks all contributeto poor performance at the chak level, which is compoundedby the lack of effective WUGs. There are already reports that the minor canals and chak watercourses are being breachedand chak intake gates vandalizedbecause the water supply is unreliable, mainly as a result of siltationin the channels. As a consequenceof the above experience, the project authority assistedby the project consultants have modifiedthe design for the area to be improvedunder the Stage II project,primarily to simplifyoperation. It is now based on ungated, fixed-flowchak intake structureseach servingchaks of about 28 ha, with watercourses extendedto outlets commandingabout 4 ha. Within the chak, allocationfollowsa rotationalwater supply (RWS),with the full chak stream divertedto one farmer at a time accordingto a predetermined schedule. (iii) ADpropriately Structuredand StaffedO&M Unit. The Stage I projectwas concerned mainly with the design and constructionof works. It did not fully provide for an O&M organization within the project authority, with properlytrained staff, well-defined responsibilities, and adequatefunds and equipmentto carry out its designatedwork. These improvements have been providedfor under the Stage II project. (iv) ProperlyOrganizedand Empowered Water Users' Groups (WUGs). WUGs were to have been organizedunder the Stage I project. Althougha number of groups were formed, their performancehas not been notably successful. Part of the poor performance reflectsa natural reluctanceto organizebefore a timely and reliablewater supply was provided. In addition,the functionsand responsibilities of the WUGs need to be clearly defined,nnd they need to be empoweredto enforcedisciplineon their members. Finally,they requireappropriate training. These requirements have been taken into consideration Stage II project. (v) LinkageBetween DOA and the ProjectAuthority. The agricultural extensionservice to the project deteriorated after 1985 when the project'sagricultural extensionstaff, which had previouslybeen directlycontrolledby the project authority, were reorganized within DOA on a districtbasis. The present extensionservice does not adequatelyaddressthe needs of farmersusing canal irrigation. Under the Stage II project,arrangements have been made to provide extensionstaff within Sunsari districtto serve the needs of irrigatedagriculture, and with a well-defined linkage to the project authority. These improvements would be in accordance with countrywidenorms for strengthening the agricultural extensionservice in new irrigationareas. - 44 - Conclusions 8.02 Although the project lagged behind the implementation schedule, and had to be reformulated in part in physical terms to reflect higher than expected unit costs, it has achieved most of its revised targets. It ~ontributedto furthering the Nepal's agricultural sectoral policy objectives. The quality of work was good. Hlowever,a permanent solution to improvement of the river and canal headworks and desilting project water supply has yet to be found. The lessons learned in implementing the project have been incorporated in the formulation of the follow-on project. 45 - - Table 1 NEPAL SUNSARI MORANG IRRIGATIONPROJECT (CREDIT 812-NEP) PROJECT COMPLETIONREPORT Summary of Physical Progress Revised Achievement Appraisal target Actual as X of Project works target (mid-1983) achievement Appraisal Revised River Control Construction of a guide island - dropped /a - - - Kosi left bank flood protection - 6,700 m completed 1981 - 100 Sediment Control A sluice bay - intake gate completed1985 - 100 modified /b - - Dredging course sediment dropped /c Vortex tube arrangement - to two completed - 100 Canal Restorationand Improvement CMC embankment - completed 1984 100 - Reconstructing main structures 70 78 78 111 100 Repairingregulators 200 60 60 30 Ion' Constructionof regulators 35 32 32 91 100 Constructionof outlets 1,000 213 213 21 100 Constructionof drainage structures 200 3 /d 3 5 100 Constructionof spillways 14 25 25 179 100 Farm Group Development Constructionof field canals 160 213 211 133 99 Irrigationblock development (ha) 6,400 9,750 9,750 135 100 Survey - 12,000ha - 11,750 11,750 98 100 Drainage Improvement Natural channels - 3 km 3 km - 100 Excavationof new drains (12,000ha) - 110 km 100 km - 100 Reclaiming(ha) 3,000 2,300 2,250 /f 75 98 Road drainage culvert - 60 /e 60 - 100 Pilot Tubewells Tubewells 25 dropped - - - Microhydels 3 dropped - - - /a Was not consideredfeasible. 7i No sluice bay was built but intake gates modified. Tc Was not consideredfeasible. 7T Not with Panchayat labor but under project. of drainage,only three cross- /e Since Shakarpurdistributaryruns parallel to the general directior. drainageworks were required. /f 50 ha in Block II yet to be completedunder Stage II. NEPAL SUNSARI MORANGIRRIGATION AND DRAINAGE PROJECT (CREDIT 812-NEP) Summary of Financial Progress (NR.n million) 1983 Appraisal Revised -------------------------- Actual Expenditure…----------- ---- Actual Expenditures Estimate Estimate/a 78/79 79/80 80/81 81/82 82/83 83/84 84/85 85/86 86/87 87/88 Total as % of Appraisal Civil Works 312.7 367.4Lb 4.5 13.2 40.4 22.7 21.0 25.8 72.1 135.7 56.0 - 391.4 125 Building 14.0 14.0 0.4 1.2 1.0 0.8 0.4 0.2 0.8 0.3 - - 4.9 35 Equipment 60.9 63.4 0.8 13.6 22.6 11.5 2.7 0.8 0.4 0.4 0.8 - 53.4 105 Technical Service 23.4 42.2 1.5 6.7 6.6 2.8 1.7 5.6 5.8 8.0 13.5 - 51.1 218 Land Acquisition 3.6 13.2 - - - 1.8 - 3.2 0.8 0.3 1.0 - 7.1 197 Establishment 45.4 52.8 0.8 3.8 4.8 5.5 6.0 5.8 2.0 2.0 2.6 - 33.1 73 TOTAL 460.0 653.0 7.8 38.4 74.2 45.1 31.8 41.4 81.7 146.7 73.9 541.0 fr Exchange rate, NRs 13.2 = USS1.00. Includes customs duties and taxes of about NRs 2.8 million. NEPAL SUNSAR1MORANGIRRiGATION PROJECT (CREDIT 812-NEP) PROJECTCOMPLETIONREPORT Cost Estimates /a Original project scope Revtsed project scope Total t ual estimated expendi- Total cost at 1982 tures up Estimated annual expenditures estimated appraisal update to Oct. 82 82/83 83/84 84/85 cost US$ millton -- Civil Works and Butidings and CMC flood protectton 1.5 3.0 2.8 0.2 - - 3.0 Kost river control structures 1.2 3.7 - 1.0 2.5 0.2 3.7 cKC sediment control and additional structures 7.1 7.5 3.1 1.2 2.2 1.0 7.5 CnC restoration canal restoration 3.2 26.5 /b - - - - - Secondary and tertiary tmprusesent and flood protection 2.4 3.5 - 2.0 4.0 4.7 10.7 Drainage system 1.5 1.5 - - - - - Field canal and drain (Block) development 0.3 0.5 0.5 - - 0.5 On-farm development 0.1 0.1 - - - - Pilot groundwater development engineertng operations 0.7 0.5 0.1 0.1 0.2 0.1 0.5 Buildings fcr project agricultural supporting services 0.2 0.7 0.3 0.3 0.1 - 0.7 Buildings for project 19.0 48.3 6.8 4.8 9.0 6.0 26.6 P Subtotal Technical Services consultants 0.9 2.3 0.9 0.3 0.6 0.5 2.3 Project operatione and mapping 0.3 0.5 0.5 - - - 0.5 Air photography 0.2 0.2 - - 0.1 0.1 0.2 Feasibility studies 1.4 3.0 t.4 0.3 0.7 0.6 3._ Subtotal Project EstablIshment 2.3 3.1 1.5 0.5 0.5 0.6 3.1 Engineering and adminltration 0.4 0.6 0.1 0.1 0.2 0.2 0.6 Agriculture 2.7 3.7 1.6 0.6 0.7 0.8 3.7 Subtotal 23.1 55.0 9.8 5.7 10.4 7.4 33.3 Total Equipment OlItequipment and 4par.- .- ,ts 3.4 4.1 3.9 0.2 - - 4.1 Engineering constrsetion equIpm;ent and spare pasrs 0.1 0.1 - - 0.1 - 0.1 Agricultural operattons and coomsunication equip. and spare parts 0.2 0.2 - 0.1 0.1 - 0.2 Survey, lab, of lice 3.7 4.4 3.9 0.3 0.2 - 4.4 Subtotal cost exludtng ,.-ntingencies 26.8 59.4 13.7 6.0 10.6 7.4 37.7 Subtotal project Contingencles C3 3.4 2.0 - 0.3 0.7 0.5 1.5 Physical contingencies 6.8 4.1 - 0.3 0.6 0.7 1.6 m Price contingencies 37.0 65.5 13.7 6.6 11.9 8.6 40.5 Total Customs Duties and Taxes (J.5 0.5 0.2 0.3 0.3 0.3 1.1 Lsnd Acqutsition. Project Cost 37.5 6b.0 13.9 6.9 12.2 8.9 41.9 Total /a Exchange value, Ntb 13.2 - US1l.t0 except for NIs 12.0 - US$1.0 for appraisal. /b Includting new canals estimated to cost US$22.5 illiton. Table 4 - 48 - NEPA SUNSARI MORANG IRRIGATION PROJECT (CREDIT 812-NEP) PROJECT COMPLETION REPORT Actual and Estimated Disbursement Schedule Actual disbursement IBRD Accumulated disbursement as Z of tiscal year Appraisal Revised Actual total Appraisal Revised and semester estimate estimate disbursement esttmate estimace FY ,9 1st 2nd 0.5 - O.1 20 FY 80) Ist 2.0 - 0.5 25 - 2nd 3.t - 1.5 42 FY 81 1st 7.6 - 2.5 33 - 2nd 12.0 - 5.2 43 - FY 82 Ist 15.8 - 6.4 41 - 2nd 19.5 - 8.1 42 - FY 83 Ist 23.2 - 9.4 41 - 2nd .7.0 11.0 10.5 39 95 FY 84 1st 28.5 13.0 12.0 42 92 2nd 30.0 17.0 13.3 47 78 FY 85 Ist - 21.0 14.7 - 70 2nd - 25.0 17.9 - 72 FY 86 1st - 28.0 21.4 - 76 2nd - 30.0 26.3 - 88 FY 87 Ist - - 28.9 - 96 2nd - - 29.7 - 99 FY 88 1st - - 30.0 - 100 EEC-Cr. 23 FY 80 0.8 - - - - FY 81 2.5 - 1.0 - - FY 82 3.(j - 1.2 - - FY 83 3.3 - 3.3 - - Closing date: IDA Cr. 812-NEP: two extensions were made; one for two years to June 30, 1986 and another for one year to June 30, 1987; EEC Cr. 23; closed December 31, 1982. Table 5 -49 - NEPAL SUNSARI MORANG IRRIGATIONPROJECT (CREDIT 812-NEP) PROJECT COMPLETIONREPORT Procurementof Vehicles and Equipment On loan Number to Used by items SAR procured Contractor project Cargo crane truck (6-8) 2 6 4 1 Portable generators 4 6 1 4 Diesel pumping set 20 35 6 29 Tractor loaders 2 3 2 1 Wheeled loaders 2 2 1 1 Sheepfoot rollers - 5 - 5 Vibratory compactors 4 2 - - Diesel welding set - I - Power rarmer 12 4 - 4 Bharat trailers 2 7 - 7 Water tankers 2 4 - 4 Hammer drill - I I Drilling hammer - I - I Workshop equipmentand tools lot lot - lot Hydraulicwheeled excavators 3 5 2 3 Industrialtractor and trailers 5 5+3 1+1 3-t3 Low bed trailer with spares ' - 1 Tractor dozers with hydraulic valve 3 3 1 1 Mobile workshop truck 1 1 - I Fuel tank truck 2 2 - 2 Disc harrow and refractors 4 4 - - Vibratory rollers 2 2 - 2 Air compressorsand accessories 3 3 1 2 Concrete Uibrators 20 14 - 14 Concrete vibrator needle - 6 - 6 Group pump and equipment 4 4 1 3 Motor graders with accessories 3 3 2 1 Dump truck 2 4 1 3 Vibratingtables 2 2 2 1 Shear legs and multipliers - 7 - 7 Honda and rajdoot motor cycles 30 9+17 - 20 Toyota jeep and landrovers 15 10 - 10 Toyota pickups 6 3 1 1 Hind motor dragline and accessories 3 1 - I Motor boat engine - 2 - 2 Traveling crane for precast set- I - I Bicycles 250 71 - 71 Storage racks 1500 kg 20 no. - 20 no. - 50- Table 6 Page 1 NEPAL SUNSARI .OWANGIRQIGATIONANDDRAINAGE PROJECT CROPPINGPATTERNS (APPRAISALAND PCR ESTIMATES) Appraisal Estimate PCR Estimate At Future Future At At Full Wet Anoraisal Without With Present Development (1991192) Loc paddy 70 65 65 20 12 HYV paddy 2 7 15 75 85 Jute 30 30 30 10 5 Vegetables 2 2 2 1 1 P_x Wheat 6 6 10 28 31 Pulses 2 2 3 18 20 Oilseed 2 2 3 7 14 Vegetables 2 2 2 2 2 Potato - - - 1 3 Maize - - - 1 3 Perennial Sugarcane 2 2 4 2 2 Total 118 118 134 165 178 Table 6 Page 2 NEPAL SUNSARI MORANG IRRIGATION AND DRAINAGE PROJECT AREA. YIELDS AND PRODUCTION Appraisal Estimate 1/ PCR Estimate 2/ Incremental Wet Area Yield Prod. Area Yield Prod. Production (ha) (t/ha) (t) (ha) (t/ha) (t) (t) Loc Paddy 6,825 1.1 7,508 1,170 2.3 2,691 -4,817 HYV Paddy 195 1.6 312 8,288 3.4 28,179 27,867 Jute 2,92' 1.0 2,925 488 1.8 878 -2,047 Vegetables 195 8.0 1,560 98 22.0 2,156 2,643 Dax- WhLeat 585 0.8 468 3,023 2.7 8,162 7,694 Pulses 195 0.4 78 1,950 1.0 1,950 1,872 Oilseed 195 0.4 78 1,365 1.0 1,365 1,287 Vegetables 195 8.0 1,560 195 24.0 4,680 3,120 Potato - - - 293 24.0 7,032 7,032 Maize - - - 293 3.0 879 879 Perennial Sugarcane 195 15.0 2,925 195 32.0 6,240 3,315 1/ Future without project situation. ,/ At full development (1991/92). Yields are same as "future with project" situation at appraisal. 1977 1978- 2978 2980 1981 1982 1983 1984 1935 1988 1987 198 1989 1990 1991 1992 1993 2994 1995 1098 1917 1198 1999 2027 W.5 Poot94 15 its 83 123 09 17 78 173 71 184 W8 171 W9 184 24 281 84 175 73 192 29 290 79 201 49 212.90 223 06 233 41 240 78 290 W 219 91 253 87 281 43 284 10 264 71 265 32 T Lts ottl E.6-ot.4 W, oct -. 01-0 -9415 1019_13 98_21 93_PS 088- _70.51 _89_05 83 05 82 22 a 33 82.48 48 38 50.11 82 15 .;S88 55 20 S8 14 07 83 59 10 59 18 59 23 5>983 59 09 59 9 I-nS-*t 0 00 8 70 28 87 43 92 84 LI 94 31 102 49 101 29 99 33 117 41 129 83 244 42 151 88 280 75 189 42 178 22 184 83 192 67 200 82 201 II 202 20 204 27 204.81 205 36 lO'o-t.-lootI 008 0,99 418 598 11 20 28 42 18 17 19 99 21 52 23 09 2533S 28.31 2831t 2831 26131 28 31 28531 28.31 29 31 20 31 26 31 2t 39 2831t 2831 Not to-r old,ot 0 00 5 71 22 71 37 94 73 65 77'93 84 32 01 31 77 81 94 32 104 48 128.10 125 07 124 44 143.11 151 90 15O 33 38 55 174 51 175 20 175 86 177 98 178 SO 17'905 Coi Wo'tn 82 25W0 78 52 42909 39 77 4a888 23855S 257 01 198980 io,ld;%'k 0.78 2 27 289 1152 0786 030 1 14 057? 0 00 rqspn. 1 52 23 7 42681 21 76 5 tI 1 52 0 78 0 78 1 52 Ln. Ocnc'S.s d 8 ~~~~~~259000 2 84 22 89 008 10 42 080 LoSoJo'w.S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0 30 31 3 22 0 00 10d 898 I1 0 98 1 52 15 18 07 35 57 169g IDO 00 7 20 9 09 10 442 11 38 10 98 3 79 3 7'9 4 92 T.t.I 135 84 72 73 140 53 85 42 80.23 78 42 154 73 277 54 139 98 aS...f.t 'o09 -13329 2-02 19 -17 27 7`0 5I2 f 42 -2003 -4 8 14.8 2 10 123107 134.44 143.11 151.9 39S.3 18 SO 17.1 S72 1758 17a8 7 5 790 ERR 145 - - 53 ANNEX 1 Page 1 of 5 N1EPAL STTt4SARI-H(RA1IG IRRIGATIfhN AND DRAINA(-E PROJECT (CREDII 812-NEP Cumcn ian tt wit h Crvdit Condit i ns P.-matks A. Condit ions Not Complied With None B. Condit ions on Whic h Act ion Pending Section 4.01 (a) SMIDDB woul(d have its accounts The project's accounts were and financial statements audited annually. IDA has audited annually by independent received the audited auditors acceptable to IDA and accounts and SOE for send certified copies of the FY1982/83 through audited accounts and the FY1985/86. FY1986/87 auditors' reports to IDA within reports are awaited. six months after the c0loseof eatch Liscal year. C. Conditiions Met Sec t ion 3. 0 2 (a) For operational (control of thp Communication facilities tanal distribution system, (telephone) were installed telecomnuniciat ion fac ilities by Indiatn aid in December wouldl be installli and A97Q. Hever. the completed not later than Dec(. facilitips wole vandalized. 31, 197Q. The projc t has installed r-adio tel -phonp Sfts. 54 ANNEX 1 Page 2 of 5 Remarks (b) The DIHM and DA would complete Postponedin 1981 until more the feasibilitystudiesof the progressis achievedon the subsequentstage of development implementation of the works by Dec. 31, 1981. in Stage I. Consultants prepareda draft feasibility study for Stage II and submittedit for review in June 1985. An addendumto the report is submittedin July 1986. (c) The Departmentof Food and Taken care of under Cash AgriculturalMarketing Services Crop DevelopmentProject would completea feasibility (Cr. 1339-NEP). HMGN was study of processingand not requiredto pursue the marketing for cash crops growa requiredstudies. in the Terai, not later than Dec. 31, 1981. Section 3.03 The recruitmentof project Contract signed in November consultantsis satisfactoryto IDA. 1979. Construction supervision contractsigned in March 1983. Both con.ractswere between SMIDB and Sir M. MacDonald & Partnersof UK. Section 3.04 The Projectwould be staffed Staff strengthgenerally generallyin accordancewith a satisfactory. Frequent staffing plan satisfactory to IDA. turnover *f spiitr staff createdl-me dlrawbLatkin project implementation. - 55 ANNEX 1 Page 3 of 5 Remarks Section 3.05 The provisional training activities Intention was that UNDP/FAO would be carried out in accordance would carry out training with a plan satisfactory to IDA. program abroad for staff working on IDA financed irrigation projects, including Sunsari Morang. The plan was sanctioned by IDA. Late 1983, HMGN decided not to proceed with this training program and UNDP withdrew the US$300,000 allocation. A training of senior DIHM staff is in effect under UNDP/IDA program (about US$500,000) as from Julv 1986. In addition, USAID has started water management training in Parwanipur since August 1985. Both programs have overseas training components through which project staff benefitted. Section 3.07 (c) Department of Agriculture would DA has carried out survey prepare two comprehensive during dry season of 1983 in agricultural and economic 12,000 ha implementation evaluation studies for the area. The report was projects. The first by mid- submitted in September 1983 1983 and the second by mid- and the second was prepared 1986. by APROSC and submitted to IDA in July 1988. (d) SMIDDB would prepare and Project Completion Report furnish to IDA a repcrt on the was submitted to IDA in June execution, completion and 1987. However, the quality initial operation of the of report is 1(lor. project within six months aftpr final disbursement of the redit. .. 56 - ANNEX 1 Page 4 of 5 Remarks Section 4.02 (a) By July 16, 1983, HMGN would It was difficult to collect establish, maintain and collect water charges due to water charges. unreliability of water supply and lack of authority to penalize defaulters. New procedures in effect as from Rabi 1986/87 transferred the responsibility of collecting water charges to Land Revenue Department and linking water cess to collection of land tax. (b) At three year intervals HMGN Review is in progress under would review water charges. ADB assisted Advisory TA. Section 4.03 * (a) Funds and faciligies required Complied with. However, for maintenance and repair of allocation for O&M has been the CMC system would be inadequate to carry out provided promptly; views would required works. be exchanged with IDA on a yearly basis on the amount of such funds rquired; each year IDA will receive a copy of the annual operation and maintenance budget. (b) SMIDDB would prepare a Consultants (MacDonald) comprehensive operation and finanlized O&M manual for maintenance manual for the CMC distribution network in Dec. system, in draft by March 1980 1982. Consultants (Coode) and in final form by Oct. 1, submitted O&M manual for 1981. headreach works in August 1988. Section 4.04 Water would not lb *c'1 en,I- i F.J 1 l it areas commanded by any fiedl was difficult to established canal until a Water User's WUG befoie water is actually Group has been formed and sound supplied reliably. 100Z of distribution routines WUGs have been formed in established. 9,750 ha command area. - 57 - ANNEX 1 Page 5 of5 Remarks Section 6.01 (a) Establishment of SMIDDB. Complied with. (b) Appointment of qualified Complied with. project manager. (c) Appointment of Sr. Agricultural Complied with. However, the Officer. function of the project's Agricultural Division was transferred to the Eastern Regional Directorate of Agriculture in July 1985. - 59 - FROMTHE BORROWER COMMENTS ATTACHMENT Page 1 N1ALEjLY'SGOVERNMENT tiLS Phone No. 413733 MINISTRY qf'ATER RESOURCES 412374 ,~~~~~~~~~4 DEPARNE OF IRRIGATION .1~~~~~~~~~~~~~~~134177 411636 '.t .- ranipokhari. Maharaigunj P>t * ^ Nepal Kathmandu. Ref NO... Date J.April D4 J9DO Mr. Graham Donaldson Chief Agriculture,Intrastructive and Human Resources Division OperationsEvaluationsDepartment The World Bank, U.S.A. Res Nepal - Sunsari Morenr Irrigation Proaect - Credit No. 812 NEP. Dear Mr. Donaldson, The draft Project Performance Report for SMIP Stage I works has been reviewed and is found to have examined all the perfonmance aspects and past activities of the Project. However, the interpretation of the benefit derived from the Project and its ERR evaluation has been remarkably underestimated in this report and is not acceptable to the Project. Our detail comments regarding the report is attached herewith for yourkind information and reconsideration. With kind regards. Sincerely y urs, A.AGRAWAL. Project Manager, Sunsari Morang Irrieatior Project. * rr. B.K. Pradhan, cc to: Secretary, MOWRf Mr. M.D. Karki, Director General, DOI. Mr. K.R. Pandey, Under Secretary, MOF. Mr. N. Roberts, ResidentRepresentative,The World Bank. Director General, ExternalRelations Commissionof EuropeanCommunities (EEC) Belgium. - 60 - TtACHMENT Pav'e ? COM4ENT ON: PROJECT PERFORMANCE AUDIT REPORT SUNSARI MORANG IRRIGATION PROJECT - I (Credit - 812 NP) The <'raft report as such his been reviewed in detail, the assessmcr-ts and findings are Ijunid appreciable and factual to some extent, however the project has some reservaLions reqard- Thie ERR evaluation Ing ERR evoluotions and interpretation of the benef;t accrued presented in the PPAM (para. 49) is repro- from the project. The details of thie comments are as follows: duced from the PCR prepared for the Borrower by the nroj- ect consultants. The 1.O General: PPAM' interpretation (nara. 53) excludes The project's main goal has been to accomplish the increase additional proposed engineering lnvest- in agriculture production through rehabilita- and farmer's inccmes through in production rehabilita~~mets as these were tion and improvement of the existing Chatra Canal System in not incurred undef 1 the audited Project.- 66,000 ha irrigable area, which include headworks improvements, desilting facilities and command area development upto farm level canal networcs. Its main recurring problems are as follows: 1. Extraction problem, at intake due to present intake orientation and river diversion. 2. Sediment problem, due to high concentration of sediment ingress into the system annually during 1.E:rf (monsoon) irrigadion. 3. Unequal distribution of water at the command area c e to lack of tertiary and water courses and farmer's participation. 1/ These marginal notes (pages 69-69) have been inserted by OED to indicate how they have affected the audit's views, A 61T-ATT C W'FNT Pag>e 3 Unuer 3taye-I development proeyamme a part Of the SUILI- 't 'o 'I,( U CnS :{e,ai'ed '.-nv intended tion of (1) .:d (2) and full development of the command aila tn l. I'1solve tlhe 'ra 1-) (""'. IIn .1ie rt (1) .And down tc 10 Ii, un t in 9750 iIi onl y hivu ot2c2[ InTc lude6 -'li. probjems o' El) Extraction and (2) Sedamtent have becn resolvd These problems, wqhich sadill Persist, are now by the rececnL ieasibility study and subsequent recommendatioln proposed fol solution by a paniel Ol experts.option shall be implementeOunder design modifica- The approved tions mzide to.thle follolw- and is anticipated to be complete in the year 1993. As the ofn Sunsari MoranRl Trriva- t on II Project. A desiltinq facility affects all the stages of the project, satisfactorv solution to economic evaluation without its consideration is totally incor- the problems cannot there- fore be counted as a rect and unrealistic. Ibenefit under the audited rroiect. The economic evaluation of the project should be based on overall benefit resolving all its problems as mentioned above, without isolating it with the stage-I area development only. The ProjectPerformance Audit Report has missed the consideration of the benefits that shall be dccrued after the completion of desalting facility. Impact study of stage-I area conducted through APROSC The APROSC survey also reveals that considerable (Agricultural Project Research Service Centre) has revealed the reprovements ciave taken fact that the production and yield compared to base line survey place in the control area (PPAM para. 33). of 1983 has increased remarkably. Even if, the desilting facili- ties are not complete and the water supply are unpredietable and unrealiable, the crop area irrigated has been reported more than 6 area). The area before 6000 lha (i.e. OCt of 9750 ha in stage-I the developmtnt has been around 2000 ha only. Hence, under the condition th;.t a reliable water supply is available througlh out - 62 - ATTACHMENT Page 4 the year and full fledged extension services are available to the farmers with their active participation, the production can definitely beachievedasanticipated in SAR. The draft report These benefits are con- PPAR has totally under estimated the actual benefits achieved ditional on the sedimen- tation Problem being by the development work. resolved through the proposed additional engineeringworks. The two price levels considered in the Project Completion Reports of Stage-I are based on (1) Terai Ground Water Develop- ment Strategy Study 1987 and (2) the projected world market prices for the year 2000 by the World Bank. In fact, the prevai- ling prices of the commodities in 1987 in the proj'ct area were higher than that were assumed in the Project Completion Report. The project considers the low prices quoted in PCR and PPAR far from actual fact. As such the benefit calculations are mislead- ing and incorrect. Hence, from the project's point of view the entire ERR evaluation are not based on realistic factual data and are not acceptable. The benefit calculations as shown in the Project Perfor- mance Audit Report does not include the increased production after the desilting facilities are completed and year round irrigation shall be available. It is thus astonishing to know how the ERR of the project comes out to be bare 0.55% only as reflected in the PPAR. On the contrary the table show- ing constant benefit (12.24) after 1987 seems far too short siqhted and unrealistic. The project could not make out the The assumptions, accept- able to the audit, are audit assumptions for such misleading data. those of the project consultants,and are detailed in the economic analvsis section (pp. 25- 31) of their PCR. - 63 - ATTACHMENT Page 5 The project has the experience in the command area of Sukhasena Distributary where year round seepage water is avail- able, that the farmers cultivate three crops a year with 200 to 300% crop intensity. Hence it is anticipated that once the desilting facilities are complete and a reliable water supply be availed, thu farmers shall be able to cultivate multiple cropping and it shall definitely achieve the food production as envisaged in SAR. 2.0 Project Performance: background: The background of the project described in the PPAR are more or less based on the actual facts. The complex problems encountered in handling the headworks improvement and river training works has been the main cause of the delay experienced in implementing the scheme. The maintenance scheme as agreed between HMG/N and IDA in the light of low cost has been a failure due to unforseen problems of ever changing morphological situation of the Koshi river and unfavourable condition of equipment handling during low stages of the river, as they could not go across the river until water level goes considerably very low. Anyway the scheme has been selected on the ground that shall reduce the sediment entry into the system and that it could easily ;,e taken out by the equipment in hand. But due to heavy intrusion of silt in the past into the system made it virtually impossible to clean it in limited time as predicted - 64 - ATTACHMENT Page 6 in the cropping pattern. As such the prevailing problems were surfaced. However, the stage-I development works has established a networks of irrigation canals and has increased the irrigabQl area from 2000 ha (without it) to nearly 6000 ha and to some extent the yields of the crops too have increased. Implemcntation Experience: The nature of the river and its unpredictable characteri- stics and its 1980 flood has eventually forced the consultant to carry out detail investigation and to go deeper insite in the study. As such the implementation schedule was delayed in compare to SAR proposal. The intake and the sediment problems have been resolved and a solution has been finalised by the panel of experts. Hence, the project anticipate that the following development works in stage-II and probable stage-III shall be carried out unhindered. The project agrees that there has been a lack of activities in the agricultural sector of the command area to fulfill its targetted achievements. But in future a co-ordinated approach with the agricultural sector can very well be assumed to bring out promising results. Project Outcome: The project agrees that after completion of stage-I, the siltation problem has been severe and in many places the water courses lost their half capacity. The desilting works in water - 65 - ATTACHMENT Page 7 courses has been the farmer's responsibility where as the desilt- ing works upto tertiary level were the projects respons_bility. The project has carried out desilting works in most of the tertiaries and secondary according to the severeness of the siltation, whereas deligent and active farmers cleared their water courses to bring water to their fields. Overall outcome of the project could be seen from the impact study report as 60% of the area has been irrigated, without desilting facility and without any agricultural extension services. Secondly, the design of the system was based on rotational or intermittent delivery within the water courses and also among the water courses level. As all the water courses had gated outlet, they could draw more than the required and did not observed the rctational schedule as designed. Hence, the tail end farmers suffered severely for watersupply. The lessons learnt in the stage-I area has been rectified and improved in the stage-II programme as the design is based on ungated system and control is based on the secondary level only. The project feels that the sediment problem is the core problem of the project, which affects overall performance of the whole system. So if this problem is set aside, the development works and achievement out of it cannot be made to the same tune as assumed in SAR in the beginning. However, the impact study reveals that despite so much of adversities created by sediment deposition the yield in 6000 ha are not disappointing. This is mainly due to intense irrigation network constructed in stage-I. 66 - ATTACHMENT Page 8 So the project strongly disagree with the PPAR's remarks on benefit achievement in stage-I developed area. Refering to para 33 of the draft PPAR report the project has the following remarks; 1. The yields assumed in SAR are based on the feasibility study carried out in 1977. Obviously the yie-ldrates are true and prevailing at that period. Wheat crops had been just introduced in this region at that period. So the yield rates assumed are based on facts and prevailing condition. 2. The yield quoted in the baseline survey in the third column of the table does not tally with the yield rates The baseline figures reported in the Baseline Survey Report. The rates presented in para. 33 of the draft PPAM were should be as follows; taken from the 1987 APROSC Socio-Economic Survey Report (Table HIYVPaddy - 2.5 (t/ha) (Irrigated) 27). These lower Local Late Paddy - 1.9 (t/ha) (Irrigated) baseline figures for HYV paddy (2.5 t/ha Wheat - 1.6 (t/ha) (Irrigated) v. 2.9) and wheat (1.6 t/ha v. 1.7) have 3t prevailing been substituted in 3. The present yield rates if compared to the prevailing the revised PPAM text. 1977-78 rates, the achievement cannot be concluded as marginal. Moreover the benefit achieved compared to 1983 haseline study report.also of the order that cannot be taken as disappointing. The PPAR has missed These benefits, appar- ent in the 1986-87 the consideration of the benefit achieved in irrigating cropyear,shortly the increased area of 4000 ha. (see impact study of after the command area construction APROSC, 1987). works under the project were com- pleted, have since been severely reduced as a result of sedimentationleading to loss of water delivery capacity. - 67 - ATTACHMENT Page 9 Findings and Issues: In 1977 a fluviologist was engaged to find a solution for the sediment and head works improvement. On his recommendation a diversion island at the bifurcation point of the river and a sluicing arrangement in front of the existing intake has been proposed, which were included in the appraised SAR for stage-I works. The follow on detail design carried out and 1980 flood ruled out such possibility. In the mean time the maintenance scheme with a settling basin at the headreach of the canal were proposed. Again the dredging system to clear out the sediment from the settling basin was rejected by HMG/N on the ground that it is not operable under the existing organisational set up and had very high recurring cost. The sequence of finding and solu- tion were all centred on the cost ground, which thus delayed the finalisation of intake and sediment solution. The implementation process were delayed due to unavoidable circuminstances and inevitable under the prevailing condition of the project. In para 44 the PPAR reported that the cost of the project After correcting for has risen from an SAR figure of US $ 37.5 m. to US $ 66.0 m. a typographicalerror the correct figure is The project considers it to be incorrect, and instead it is US$60.0 m. This supervision estimate, US $ 41.9 m. in mid 1981, was a re- estimate of the cost to complete the project as The project again retort that the evaluation of the Projectappraised Reductions should be taken as a whole to include the whole project area of in scope and scale led to the actual cost at 66,000 ha and taking into consideration of the present estimate completion being US$41.9 m. of the headworks improvement and desilting facilities of US $ 25.0m. - 68 - ATTACINENT PaRe 10 Only then a p.ctorial and a reLil picture of the project c,in bu obtained in fn rspeCtiVu. Econoii:c Rate of Return Revisited: The interpretation of the PPAR report on this issue are not acceptabl- to the project on tthe following ground; 1. TI,: appraised value of the yield at "without project" "with project" atv,! has been calculated in 1978, hence th. baseline survey of 1983 values are irre'avarnt to b. included at that time. 2. PCtI/ Thket has considered the baseline survey data of 19,j and Socio-economic Impact Study of 1987. liowever, thleproject feels that the prices of the commodities estimated are too low and hence the benefit of the project has been under estimated. Even then, the ERR comes out to be 8% and 10.7%. 3. Th,- PCRlhas iqnored the situation of "without project" irr-igable area of approximately 2000 ha only. If this c,:.sideration are taken up the incremental benefit must be hiqher than shown in their PCR. *.. it is totally astonishing to note in the calculation t.le of ERR in the PPAR report (Paye 25) that the bL:.e t achicved after the dcsilting facilities are cc::.pl etel y ignored. 1/ Project Co:.:-letion Revort June 1987 (Revised) Prepared b the Consultant Sir ;:. MacDonald & Partners. 69 - ATTACHMENT Page 11 Hence, based on the above points the project considers that the ERR of the project must be higher than 10.7% as reported in PCR and *the value of ERR 0.55% is totally misleading and PPA? nara. 54 further expounds on the incorrect. audit position. 3.0 Conclusion: 1. The audit should be based on the overall evaluation of the project coosidering the completion of headworks improvement and desilting facilities(estimated about US $ 25.0 m.), and command area development of complete 66,000 ha. Infact stage-I has impact on stage-II and :ice versa are inter related. Niether one can be visualized in isolation. 2. The prices of the commodities adopted in PCR and PPAR are on too low side, it must be revised on the facts and ERR for stage-I should be reevaluated, taking fraction of the cost of headworks and main canal improvement works. 3. To tackle such a large scheme - largest irrigation scheme in Nepal - and taming the Koshi river of impredictable nature, one needs to go deep in the design and implementation programme. Hence, a panel of world reknowned experts were invited to review the problems and its study. The proceeds of the projects should be based on their recommendations and guideline. - 70 - ATTACHMENT Page 12 4. Although the project is lagging behind in the agricultural sector such as to carry out extension scrvices and motivating the farmers, a solutlon needs to be sorted out to co-ordinate Department of Agricul- ture and Department of Irrigation. A revised calculation of ERR based on the considera- tion of desilting facilities in 1993 are attached here- with, which has revealed the ERR to be 13%- - 71 - ATTACHMENT Page 13 ECONOMIC COSTS AND BENEFIT IN STAGE-I AREA (Million NRs. Constant 1987 Prices) Year Project Cost Benefit Net Cash Flow 1979 11.644 - (11.64) 1980 18.00 - (18.00) 1981 24.83 - (24.83) 1982 10.81 - (10.81) 1983 12.51 - (12.5 1) 1984 9.22 - (9.22) 1985 83.32 11.16 (72.16) 1986 107.23 22.32 (84.91) 1987 14.43 16.74 2.31 1988 5.24 16.74 11.50 1989 5.24 16.74 11.50 1990 5.24 16.74 11.50 1991 5.24 16.74 11.50 1992 25.34 16.74 . (8.60) 1993 25.34 27.90 2.56 1994 5.24 44.64 39.40 1995 5.24 55.80 50.56 1996-2028 5.24 55.80 50.56 ERR 13% Note:- The calculations of the cost and benefits are based on Project Completion Report June 1987 (revised). The revision are made to accomodate the desilting facilities in 1993. The prices as assessed by Terai Ground Water Development Strategy Study 1987 adopted in PCR are chosen practicable and justifiable, although area specific prices were higher than the quoted. t MAP SECTION __ _ I19z6-L |BRD __ ___ - --- I N .4 C~~CH _2pv ! '_.~~~~~._> ~SUNSARI-MORANG AND DRAINAGE IRRIGATION StAGEI PROJECT DEVELOPMENT fi''4''.''.iLn fX'-3> ~~~~~~~~~~~~~~Project Location - .2 F / - t \_ ~~~~~~~~~~~~~~~~~~~~~~~~~~~Mountain Region Rivers < y >/ - y ,.~~~~~~~~~~~~~~~~~~~-~~~ ,> ~~~~~Hill Region ': Development Region Boundezries E J / ' - / 4~~~~~~~~~ ' ....... .. e'->+so ~~~~~~~~~~~Inner Terai Administrative Zonal Boundaries > /;- ' Z a f r-\fz < 2 \ t S \ Teroi -- - Dlstrict Boundoriea~~~~~~~~~~~~~~~~~~~~~istic Bondoie . i *- tAA W rEsen S A o_o Valltey ~~~~_..International ~~~~~~~~~~~~~~~~~~ ~~~~~Kathmandu BoundaJries X g r tr~~~~~~~~~~~~~~~~~~~~~~~~~< f \\ e \ r ~~~~~~~~~~~~~~~Prcposed < Project 28' R/FG /0 ~4 G. /O-- 28' NA RAYANV 'I .. ''''2 *' ''--t IRRIGATlN\S, DEVELOPMIENT ,v, S - ' 0~ 20 40 6,0 S0 100 120 140 160 t - - ;/1 KIl.OMETERS I N D I A 26 -26° G-S4 ° 20 ~ ~~~ ~ ~~~~~ ~~~~ 10, 6 ntso- ~ ~~ t 16ua cre;sc ~~~~MILESS0 ' todcee .,_ go,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ _2 I__' __. Th. W. Bn', ' .. y NEPAL -d . fThA-w-4tl StageI Project Sunsari-MorangIrrigation and Drainage Development rh. 4-h, idz o_ n d rp do CHATRA MAIN CANAL COMMAND AREA 's°t'6 So ~Chatro main canal Moin roads Boundaryof canal command n $ O>\ I-- stage I Limitof form groupdevelopment, CHATRA _\\_HATRA--Secondary canals Auu.uu Protective embankment - - Limit of irrigotion, 1977 @ Towns / -,\j Natural droins and streams . _ International boundary IHARI RANGEL / . ~ ~ IN IA I * - OBAN 0 5 0~~~~~~10I / ~~ ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~dd~~~~~~~~~~~~~ALES
Группа Всемирного банка · Project Performance Assessment Report
Nepal - Sunsari - Morang Irrigation and Drainage Development Project
Открыть оригинал документа
Полный текст размещён на сайте публикующей организации. lawenc.com индексирует метаданные и ведёт на официальный источник.
Полный текст
Основные сведения
Организация
Группа Всемирного банка
Тип документа
Project Performance Assessment Report
Страна
Непал
Источник
Всемирный банк