Sri Lanka - Village Irrigation Rehabilitation Project

Document of The World Bank FOR OFICIAL USIE ONLY Reprt NO. 11411 PROJECT COMPLETION REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (CREDIT 1160-CE) DECEMBER 7, 1992 Agriculture Operations Division Country Department III South Asia Regional Office 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 Currency Unit - Rupees (Rs) Appraisal Report (April 1981) - US$1.00 = Rs 18.0 Average 1982-1989 - US$1.00 = Rs 27.8 Completion Year (1990) - US$1.00 - Rs 40.1 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS USED APT - Agricultural Planning Team DAS - Department of Agrarian Services DCS - Department of Census and Statistics FO - Farmer Organization GOSL - Government of Sri Lanka ID - Irrigation Department NIRP - National Irrigation Rehabilitation Project SM - Supervision Mission TS - Tank Supervisor VIRP - Village Irrigation Rehabilitation Project WOP - Without Project WP - With Project FISCAL YEAR January 1 - December 31  FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation December 7, 1992 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Sri Lanka - Village Irrigation Rehabilitation Proiect (Credit 1160-CE) Attached for information is a copy of the report entitled "Project Completion Report on Sri Lanka - Village Irrigation Rehabilitation Project (Credit 1160-CE)" prepared jointly by the South Asia Regional Office and FAO/IBRD Cooperative Program with Part II contributed by the Borrower. Despite civil strife, start-up delays, budgetary problems, shortage of technical staff, inexperienced contractors and poor integration of activities, the project achieved its main objectives of increasing agricultural production and farm incomes in existing village irrigation schemes, although over a much smaller area than originally planned. As such, the project is rated as satisfactory, even though the economic rate of return, reestimated at completion at 16 percent, is based on questionable assumptions. Sustainability of the project is promising for about half of all schemes, where functioning farmer organizations have been established. For the others sustainability will depend on the formation of effective farmer organizations for water management and system maintenance. The Project Completion Report is informative and is of satisfactory quality. An audit is proposed to reassess the economic rate of return and sustainability of the project. Attachment 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.   FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (Credit 1160-CE) TABLE OF CONTENTS Page No. PREFACE . .......................................................... i EVALUATION SUMMARY ............................................... iii PART I - PROJECT REVIEW FROM BANK'S PERSPECTIVE .................. 1 A. Project Identity ............................................. 1 B. Background ................................................... 1 C. Project Objectives and Description ........................... 1 D. Project Design and Organization .............................. 3 E. Project Implementation ........................................ 4 F. Project Results .............................................. 7 G. Project Sustainability ....................................... 9 H. Bank Performance ............................................. 10 I. Borrower Performance ......................................... 11 J. Project Relationship ......................................... 12 K. Consulting Services and Procurement .......................... 12 L. Project Documentation and Data ............................... 12 PART II - PROJECT REVIEW FROM BORROWER'S PERSPECTIVE .... ......... 14 PART III - STATISTICAL INFORMATION ................................ 21 A. Related IDA Credits ........ ............... ................... 21 B. Project Timetable ........ ................ .................... 22 C. Credit Disbursements ......................................... 22 D. Project Implementation ....... ............ .. .................. 23 E. Project Cost and Financing ................. .. ................ 24 (i) Project Cost .......................................... 24 (ii) Financing ............................................. 25 F. Project Results ......................... ..................... 25 (i) Direct Benefits ....................................... 25 (ii) Economic Impact ....................................... 25 (iii) Financial Impact ...................................... 26 (iv) Studies ............................................... 26 (v) Training .............................................. 27 G. Status of Covenants .......................................... 28 H. Use of Bank Resources ........................................ 29 (i) Staff Inputs .......................................... 29 (ii) Missions .............................................. 30 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.  Page No. ATTACHMENT 1 - Actual Project Cost 1981-1990 ................... 31 ANNEX TABLES 1. Cropped Area (Pre-Project) ............................. 32 2. Cropped Area (Without Project) ......................... 33 3. Cropped Area (With Project) ............................ 34 4. Per Ha Financial Farm Budget (Pre- and Without Project) 35 5. Per Ha Financial Farm Budget (With Project) .... ........ 36 6. Farm Incomes - 0.75 ha Farm ............................ 37 7. Economic Analysis ...................................... 38 8. Project Costs .......................................... 39 9. Summary of Financial and Economic Prices .... ........... 40 MAP - IBRD 24287  - i - PROJECT COMPLETION REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (Credit 1160-CE) PREFACE This is the Project Completion Report (PCR) for the Village Irrigation Rehabilitation Project (VIRP) in Sri Lanka for which Credit 1160-CE in the amount of SDR 24.5 million was approved on June 16, 1981. The credit was closed after four one-year extensions on December 31, 1990. A total of SDR 17.58 million was disbursed, and SDR 6.92 million cancelled. The last disbursement was made on May 16, 1991. The PCR was jointly prepared by an FAO/World Bank Cooperative Program mission which visited Sri Lanka in November 1991 (Preface, Evaluation Summary, Parts I and III), and the Borrower (Part II). The report (except Part II) was reviewed by the Agriculture Operations Division, Country Department III of the South Asia Regional Office. The PCR is based on a review of all relevant project documentation (including Staff Appraisal Report, Memorandum of the President, Development Credit Agreement, supervision reports, correspondence between Bank and Borrower, and internal Bank memoranda), field visits to selected project areas, and discussions with the Borrower's project staff and Bank staff associated with the project.   - iii - PROJECT COMPLETION REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (Credit 1160-CE) EVALUATION SUMMARY Obiectives 1. The main objective of the project was to increase agricultural production and farmer incomes in existing village irrigation schemes by financing: (i) rehabilitation of some 1,200 such schemes; (ii) modernization of a further 500 schemes; (iii) strengthening the major government institutions involved in village irrigation; (iv) initiating a systematic water management program; and (v) establishing evaluation programs designed to assist project implementation and preparation of further programs in the sub-sector. Implementation Experience 2. The project has met its implementation targets on a much smaller area than originally expected (para 7), and this could be achieved only by considerably extending the project period. The total project period thus increased to nine years, compared to five years estimated at appraisal. The project originally was to include 14 of Sri Lanka's 24 districts, of which four were later dropped. Due to civil disturbances affecting the north and east of the country since about 1984. project activities were severely hampered, leading to suspension of disbursements in a further seven districts by June 1986. In compensation, the Bank agreed to expand project activities to eight less-affected districts, where rehabilitation needs were initially ranked lower but enough schemes meeting project criteria were identified. This brought the number of districts to eleven. 3. Initial delays, budget constraints and a declining value of the rupee against the dollar and SDR affected disbursements, and by December 1985, the end of the original implementation period, disbursement stood at only 37% of the SAR estimate. It increased over the four-year extension period to 72% of the credit. 4. Due mainly to a shortage of technical staff and experienced contractors, construction performance was inferior to SAR expectations. However, the situation improved in later project years, mainly because of a substantial training effort (para 5.16). Therefore, on balance, the capability of constructing irrigation infrastructure was considerably strengthened. 5. A major institutional challenge was the coordination of the roles and actions of the two main implementing agencies, the Irrigation Department (ID) and the Department of Agrarian Services (DAS). Whereas successful implementation largely depended on their joint action, in actual fact, DAS generally got involved only when a scheme was due to be handed over officially from ID to DAS,  - iv - and this process was typically delayed from six to nine months after completion of works. The delays were caused by various factors, including land alienation, late identification by DAS of structural deficiencies, and DAS' own program of scheme modernization, which reduced the Service's capacity to deal with ID- rehabilitated schemes. 6. Several surveys and evaluation studies were carried out, involving Government agencies and the University of Peradeniya. Because of poor baseline data, lack of coordination and some shortcomings in methodology, the quality and usefulness of the results have been mixed. Little use could be made of the results in project implementation, and the PCR re-evaluation had to rely on mission findings and preliminary results of a separately-financed consultants' study. Results 7. The project has achieved its main objective of increasing agricultural production and farm incomes in existing village irrigation schemes on a much smaller area than originally expected. It included rehabilitation of 1,072 schemes (89Z of target) and modernization of 627 schemes (125% of target), covering a total command area of 45,555 ha, mostly under paddy. However, the actually benefited area is estimated at 26,080 ha, benefiting some 46,500 farm families. The estimate is obtained by excluding all schemes located in the seven northern and eastern districts (para 2) due to their unknown state of completion and operating conditions, and by further excluding (based on DAS estimates) 25% of ID schemes and 20% of DAS schemes in which rehabilitation/modernization has had negligible effect. In addition to infrastructure rehabilitation, the project has improved water management practices in village irrigation and initiated the institutionalization of farmer organizations, which has had a positive impact on agricultural practices as well as scheme O&M. The more assured irrigation supply has reduced cultivation risks and induced farmers to shift from traditional to HYV paddy. Based on the preliminary findings of the performance assessment of 180 selected village irrigation schemes undertaken by local consultants, and mission estimates, overall cropping intensities and paddy yields increased from the pre-project to the with project situation by 8% and 19%, respectively. No information is available for the agricultural situation without the project. However, because of further deterioration of physical facilities in the without project situation, it has been assumed that irrigated areas would decline by 20% over a ten-year period but crop yields would remain constant. Total incremental production of paddy has been estimated at about 25,000 tons per year at full development, or about 30% below the SAR estimate of 37,800 tons, due to the significantly lower benefited areas and lower incremental cropping intensities. 8. At appraisal, the economic rate of return (ERR) was estimated at 20% for the rehabilitation component of the project only. The re-estimated ERR of the project, covering rehabilitation and modernization, is 16%; the costs and benefits of the schemes in the seven northern and eastern districts have been excluded due to verification difficulties. Factors contributing to the lower ERR are the longer than expected implementation and gestation periods, lower than expected benefited areas, incremental cropping intensities and paddy prices, cost escalation and the lag in the implementation of water management plans. Annual net farm income is re-estimated to increase by 55%, compared to 81% estimated in  the SAR. The lower increase is due to the reasons given in the previous paragraph. Sustainabilitv 9. The outlook for sustainability is promising for about half of all project schemes, where functioning farmer organizations (FO) have been established and water management plans implemented (paras 5.09, 6.02), For the rest, sustainability will depend on further efforts, mainly by DAS, in forming FOs to carry out water management programs and to organize O&M. 10. The strengthening of ID and DAS by the project will have a lasting impact on their capacity to implement physical infrastructure as well as water management programs, and related activities. However, since a large number of staff have been transferred to newly established provincial services and both ID and DAS are engaged in the implementation of new development projects, there appears to be little scope for providing the services still needed in about half of the VIRP schemes (para 9), unless the task is taken up by one of the on-going projects. Findings and Lessons Learned 11. The project was generally well designed and both the Bank and the Borrower performed well throughout implementation, minimizing the negative impact of security problems and budget constraints. However, other mainly institutional weaknesses inhibited project implementation to a considerable extent. Full realization of the benefits of investment in physical rehabilitation by ID largely depended on the conjunctive implementation of a 'software package' by DAS, consisting of the establishment of active FOs, formulation of water management plans and their implementation through farmer training and demonstration. This balance between the two agencies' implementation responsibilities and capacities appears to have suffered from the fact that DAS also implemented a works program. This clearly reduced its capacity to provide the 'software' for both its own modernized schemes and those rehabilitated by ID. As a result, 'software' implementation has been achieved on only about half of the schemes rehabilitated or modernized. On the other hand, the two parallel physical works programs together extended project impact to a larger total command area than would have been the case if the project had been restricted to rehabilitation only. Implementation experience also suggests a number of specific lessons to be learned as listed in the relevant chapters of Part I. 12. In retrospect, a weak aspect in project design was the relationship between FOs and DAS, which was not systematized and has led to some ambiguity in their relationship. This was recognized in the follow-up project, the National Irrigation Rehabilitation Project (Cr. 2260-CE), in which special attention has been given to the establishment of FOs and the responsibilities of FOs for O&M.   PROJECT COMPLETION REPORT SRI LANKA VILLAGE IRRIGATION REHABILITATION PROJECT (Credit 1160-CE) PART I: Project Review From Bank's Perspective A. PROJECT IDENTITY Name Village Irrigation Rehabilitation Project (VIRP) Credit No. 1160-CE RVP Unit South Asia (SA3AG) Country Sri Lanka Sector Agriculture (Irrigation) B. BACKGROUND 2.01 Starting in 1977, the Government of Sri Lanka (GOSL) implemented a series of major reforms designed to revive the economy through dismantling controls over resource allocation; establishing realistic relative prices and incentives; and launching an ambitious public sector investment program. The response was encouraging, with a growth of GDP of 8% in 1978, 6% in 1979, and a further 6% in 1980. Nevertheless, by 1980 it had become clear that GOSL's strategy, in particular its emphasis on major long-gestation, capital-intensive investments, carried with it the risk of overheating the economy, as reflected in a sharp increase in inflation and severe balance of payment problems. In response, GOSL reduced investment allocation across-the-board by 25% in the 1981 budget. Among further measures was the emphasis on quick-yielding investments such as the Village Irrigation Rehabilitation Project (VIRP), to sustain medium- term growth while adjusting long-term large-scale investments to the absorptive capacity of the economy. 2.02 Since 1980, a systematic approach to village irrigation was adopted and implemented through Rural Development Projects, Irrigation Department Programs (IDP) and Dry Zone Farming Projects. The Bank Group participated in the first program with two projects (Part III, A) and followed up with the VIRP on a Government-financed pilot project implemented in 1979/80 under the IDP. C. PROJECT OBJECTIVES AND DESCRIPTION 3.01 Obiectives. The project aimed at increasing agricultural production and farmer incomes in existing village irrigation schemes by: (i) financing the rehabilitation of village tanks and anicuts (diversion schemes) over a five-year period (1981-85); (ii) strengthening the major Government institutions involved in village irrigation; (iii) initiating a systematic water management program; and (iv) establishing evaluation programs designed to assist project implementation and preparation of further programs in the village irrigation sector.  -2- 3.02 Proiect DescriRtion. The project consisted of three distinct components: (a) Civil Works Component - rehabilitation of about 1,200 village irrigation schemes by the Irrigation Department (ID), comprising repairs to and remodelling of tank bunds, sluices, spillways, anicuts and irrigation distribution systems, including provision of appropriate field structures; - modernization of working schemes by the Department of Agrarian Services (DAS), with a view to facilitating the introduction of water management programs; and - provision of survey, drawing, construction, maintenance and quality control equipment in support of the civil works program, together with necessary transport vehicles. (b) Operation and Maintenance Component - strengthening of DAS' capacity to service the operation and maintenance of minor irrigation schemes through incremental staff and provision of office, survey and construction equipment, and transport vehicles; - introduction of systematic water management programs on project schemes; and - support for on-going training programs in basic technical skills as well as for new water management training programs to be provided at the Maha Illuppallama Research Institute. (c) Project Evaluation and Investigations - monitoring of project impact on agricultural production by DAS field staff and the Sri Lankan Freedom from Hunger Campaign (FFHC), with support from the Department of Census and Statistics (DCS) to check the yield data collected by DAS staff; systematic data collection and evaluation programs by the Land and Water Use Division of the Department of Agriculture, covering hydrology of village tanks, soils, land use and related information; and regular socio-economic evaluation studies of project impact by the University of Peradeniya as a basis for introducing modifications to project design during implementation and to assist in further planning.  -3- 3.03 Total project costs were estimated at Rs 784 million (US$43.6 million). An IDA credit, Cr.1160-CE, of SDR 24.5 million (USS30 million equivalent) was to finance approximately 69% of project costs, or 100% of foreign and about 60% of local costs. The remaining funds were to be provided by GOSL. Overall responsibility for project implementation was with the National Committee for Village Tank Rehabilitation. Subordinated bodies for annual program formulation and approval were: the Project Steering Committee, and the District Agricultural Committee. 3.04 ID was to be responsible for the execution of all main civil works through its existing organization, supported by machinery, vehicles, equipment, provided under the project. For implementation of the O&M component, the project was to strengthen DAS substantially. D. PROJECT DESIGN AND ORGANIZATION 4.01 The project had a clear conceptual foundation which fitted well into GOSL's development strategy, giving emphasis to quick-yielding investments to sustain medium-term growth. The project was innovative in its intention to introduce institutional mechanisms to assist farmers in scheme operation and maintenance on the basis of water management plans. However, some of the proposed changes were substantially modified during implementation, mainly due to staff shortages (para 5.11). In the light of such and other implementation constraints, the project's scope was adequate, although the targeted rehabilitation works represented not more than 20% of all village irrigation schemes qualifying for physical rehabilitation and/or water management support. 4.02 Two FAO/CP missions (identification and preparation) prepared the project on behalf of GOSL, followed by Bank pre-appraisal and appraisal missions. In general, the project was well prepared. However, in designing the project's evaluation and investigations component, the generation of adequate baseline data should have been addressed and linkages between parties involved better defined (para 5.12). 4.03 With responsibilities for irrigation infrastructure and its management divided between ID and DAS, the success of the project largely depended on efficient coordination, as well as an adequate match between the two agencies' implementation capacity. Although ample attention was paid to these aspects in the appraisal report, the fact that DAS also was assigned the task of physical modernization of schemes unintentionally introduced a certain imbalance into the implementation process. Thus, while in project design it was strictly linked to the rehabilitation program, the modernization activity developed its own momentum, creating a parallel physical infrastructure improvement capability. Since this absorbed much of DAS' capacity originally earmarked for scheme water management, farmer organizations (FO) and O&M, the project fell short of its targets for these activities.  -4- E. PROJECT IMPLEMENTATION 5.01 General. The project has met its implementation targets on a much smaller area than originally expected, and this could be achieved only through four one-year extensions. The total project period was thus nine years, compared to five years estimated at appraisal. The project originally was to include 14 of Sri Lanka's 24 districts. Four of these were later dropped, while eight nev districts were added, raising the total to 18. 5.02 Due to civil disturbances affecting the north and east of the country since 1984, project activities were severely hampered, leading to suspension of disbursements in seven districts as of June 1986. This reduced the number of districts covered to eleven. By project closure, some 260 schemes had been taken up by ID and 117 by DAS in the seven districts, but due to the security situation it has not been possible to verify how many of these schemes have actually been completed. An ID report of September 1988 indicates that headworks were rehabilitated in most ID schemes while canal systems were in various stages of construction or not started at all. By project closure, 260 schemes were reported as completed. However, since there have been few opportunities to resume works and no disbursements were made during 1989 and 1990 (see Part III, G), it must be assumed that 'completion' refers only to headworks and not to both headworks and canal system. Furthermore, due to the civil disturbances, handing over of completed schemes and introduction of water management programs in the seven districts were severely hampered. 5.03 The forced geographical shift considerably reduced the project's intended emphasis on the dry zone (vis-a-vis intermediate and wet zones) and on tank schemes (vis-&-vis anicuts). Instead of 75% (SAR), only 50% of all rehabilitated schemes were located in the dry zone, and the share of anicuts increased from 10% to 44%. However, the latter appears to have had a positive impact on the project since cost/benefit ratios for anicut schemes, because of lower investment costs, generally seem to exceed those for comparable tank schemes. 5.04 Initial delays, budget constraints and a declining value of the rupee against the dollar and SDR affected disbursements, and by December 1985, the end of the original implementation period, disbursement stood at only 37% of the SAR estimate. It increased over the four-year extension to 72% of the Credit - SDR 17.58 million or US$21.6 million; the balance of SDR 6.92 million (US$8.4 million) was cancelled. Total project costs are now estimated at Rs 923.5 million (US$31.5 million), against an SAR estimate of Rs 784.3 million (US$43.6 million). Investment per hectare of command area was U$870 for rehabilitated schemes and US$300 for modernized schemes. 5.05 Rehabilitation/Modernization of Schemes. Selection of schemes was to follow a set of criteria established at appraisal, which proved to be a useful guideline. A few cases of disregard of established (or modified) criteria were reported in some districts, notably Monaragala, where a number of new schemes were developed. Some cases of unauthorized increase of spillway crest levels for increased storage, at the expense of downstream users, were also noted. The cost  ceilings for scheme rehabilitation/modernization had to be raised several times due to inflation. 5.06 Construction quality control was considerably strengthened by the project. Yet a number of schemes were left with defective infrastructure, causing leakage, instability or affecting operation. One of the causes was shortage of staff for construction supervision, another was inexperienced local contractors. Design deficiencies have also been reported. Shortage of essential construction materials was experienced in 1988-89 in conjunction with civil unrest. 5.07 The risk to project implementation of budget constraints had been accepted at appraisal since leaner budgets could readily be accommodated by adjusting the number of schemes taken up each year, without affecting overall project feasibility. 5.08 While, in general, construction equipment provided by the project was adequately utilized, there has been considerable use outside VIRP and some damage was caused through civil disturbances. 5.09 An institutional shortcoming with considerable effect on project implementation was the poor integration of ID and DAS activities. ID's and DAS' simultaneous presence and actions were required in schemes selected for rehabilitation. However, DAS generally got involved only when a scheme was due for handing over by ID. The hand-over process itself was time-consuming because it often depended on repairs or supplementary works requiring ID's renewed involvement, or other factors, such as land alienation. As a consequence, ratification of scheme hand-over lagged behind in about 40% of completed schemes throughout the project. As the preparation of water management plans generally was initiated only after ratification of hand-over, there was a corresponding delay in the preparation and implementation of such plans, which by project closure covered only about 50% of all schemes rehabilitated. 5.10 To accelerate the hand-over, IDA had agreed to an arrangement whereby DAS could spend an additional allocation of up to Rs 50,000 per scheme for post- rehabilitation repairs and/or supplementary works required to enable it to introduce appropriate water management plans. However, this curative measure was introduced too late to be effective. 5.11 The SAR provided for the establishment of a new cadre of 'tank supervisors' (TS). These were initially to be associated with the Agricultural Planning Teams (APT) in the water management planning stage and thereafter to take main responsibility in implementing the plan and overseeing O&M in a cluster of rehabilitated/modernized schemes. Apparently, the difficulty of mobilizing the resources for establishing such a cadre weighed more heavily than the unproven benefits to be derived from it. Thus the cadre was not created; instead, its role was assigned in part to the APTs, and in part was vested in the vel vidanes (traditional scheme water masters). 5.12 Proiect Evaluation and Investigations. The SAR provided for monitoring of project impact on agricultural production by means of an initial  - 6 - baseline survey followed by annual surveys of 200 to 250 schemes. Actual achievements were far below expectations in quality, quantity and documentation. A baseline survey was not undertaken, and there also was no linkage between the annual surveys and the socio-economic survey (para 5.14). 5.13 The Physical Resources Evaluation Study, which was to provide much needed scheme selection and design criteria, was completed very late and had limited applicability. While Bank supervision missions gave a positive response to the study, ID was not impressed and even disputed certain results. Staff suspected that tank water balances were based on faulty hydrometry. The approaches taken and conclusions drawn also appeared too academic for practical application. 5.14 Socio-Economic Evaluation. The project provided support to the Department of Agricultural Economics and Extension, Faculty of Agriculture, of the University of Peradeniya, to carry out annual socio-economic surveys in order to evaluate the project's impact on agricultural production and farmer incomes. Site surveys were conducted and three analytical reports were produced. The reports provided some useful description of the pre- and post-project situation of the village irrigation schemes, including demographic characteristics, land use and farming systems, production cost of paddy, water management and agricultural support services, and cost/benefit ratios of the rehabilitated/modernized schemes. The studies showed low benefit/cost ratios in most of the schemes, which was not consistent with the substantial increase of paddy sales and the decline in water shortage after rehabilitation/modernization, also reported in the studies. Lack of reliable production data was the major weakness. Moreover, production levels 'before' and 'after' rehabilitation/modernization, used in the analysis instead of 'with' and 'without', were inappropriate. 5.15 Because of the lack of reliable production data required for an assessment of the project's economic impact, especially with regard to possible future Bank supported activities, GOSL and the Bank agreed on an additional study to be undertaken by local consultants. Intended to be financed under VIRP, because of delays the study was financed under the new National Irrigation Rehabilitation Project (Cr. 2260-CE). It was started in September 1991 and completed by end February 1992. It covers a more general survey of 180 and a specific survey of 30 VIRP schemes in different agro-climatic zones. Preliminary findings of the study have been incorporated into this PCR's economic re- evaluation of the project. 5.16 Training. The project accomplished an impressive training program in quantity as well as quality. Some indicative figures are given in Part III, F (v). A comparison with appraisal estimates is not possible since the SAR did not quantify training requirements. 5.17 Technical Assistance. The project's technical advisor assumed duties in October 1981. His initial assignment was for two years but was extended by six months up to the end of April 1984. His performance has been rated as excellent.  F. PROJECT RESULTS 6.01 The project hAs achieved its main objectives of increasing agricultural production and farm incomes in existing village irrigation schemes, although over a much smaller area than targeted. Indications supporting this are that irrigation intensities and crop yields have increased in the majority of schemes. 6.02 The project included 1,072 rehabilitation and 627 modernization schemes covering a total command area of 45,555 ha, mostly under irrigated paddy, but could not complete all (see below). On completed schemes, physical impact of rehabilitation/modernization has been to: (a) extend useful life of existing tank bunds and headvork structures; (b) reduce seepage through bunds, structures and unlined canals; and (c) facilitate rotational water supply and general increase in water use efficiency. In addition to physical works, the project has been instrumental in improving the water management standards in village irrigation systems. It also has initiated the institutionalization of farmer organizations (FOs), which has had a positive impact on agricultural, as well as O&M operations. By project closure, a total of 880 or 52% of all schemes had FOs, of which the majority are active. The more assured irrigation water supply has reduced cultivation risks and induced farmers to shift to HYV and to use more inputs. 6.03 Proiect Impact. Although the project has rehabilitated 1,072 and modernized 627 schemes with a total command area of 45,555 ha, benefited areas have been reduced to 26,080 ha. This excludes all schemes located in the seven districts in the northern and eastern regions (about 25% of total), because of their unknown condition and degree of completion, and (based on DAS estimates) about 20% of DAS' and 25% of ID's schemes where rehabilitation and modernization had negligible effect, due to various deficiencies (see Section 5). Of the total benefited areas, some 17,610 ha are ID schemes and 8,470 ha DAS schemes. According to the impact evaluation recently undertaken by local consultants (para 5.15) for 180 selected project schemes (121 rehabilitated and 59 modernized), irrigation intensities increased from an average of 100% before project intervention (1984/85) to 110% after completion (1990/91). Correspondingly, overall cropping intensity increased from 130% in the pre-project to 138% in the with project situation (Annex 1, Tables 1-3), compared to the SAR estimates of 81% and 116% respectively. Cropping intensities in the with project (VP) situation range from about 180% in the wet zone, to 125% in the intermediate zone and 115% in the dry zone. Overall, paddy yields increased by 19% from the pre- project (PP) to the WP situation, compared to an increase of 25% estimated at appraisal. Irrigated paddy yields increased from 2.6 t/ha in the PP to 3.1 t/ha in the WP situation in the maha season, and from 2.3 t/ha in the PP to 2.8 t/ha in the WP situation in the yala season (Annex 1, Tables 4-5). Because of lack of sufficient data, no distinction was made between agro-climatic zones or between tank and anicut schemes. It should be noted that a series of drought years have distorted project impact, particularly in the dry zone.1' The physical resources study (para 5. 13) revealed that rainfall during the decade of the eighties was below the historical average in eight years, with only 1984 and 1987 above average.  - 8 - 6.04 No data are available for the agricultural situation without the project (WOP). However, in practice, further deterioration of physical facilities can be assumed, which would have reduce8 the irrigated area. Based on DAS' experience, and the PCR mission's field observations and interviews with farmers, it has been assumed that yields of paddy in the WOP situation would remain constant at PP levels, while overall irrigated area in both maha and yala seasons would decline by 20% each at a constant rate over ten years. Thus, cropping intensity has been assumed to decline from 130% in the PP to 124% in the WOP situation. The SAR had assumed the same cropping intensities in the PP and the WOP situation, but paddy yields were expected to increase by 14% due to the progressive adoption of improved cultural practices. Based on the above benefitted areas, cropping intensities and yields, total incremental paddy production has been estimated to be about 25,000 tons per year at full development, or about one-third less than the SAR estimate of 37,800 tons. The lower paddy production is due mainly to the significantly lower benefited areas and lower incremental cropping intensity. Moreover, the PCR assumes that over one-third of the area of rainfed paddy remains in the WP situation, whereas the SAR assumed that this area would be fully irrigated. 6.05 The increase in irrigation intensities and crop yields can be interpreted as a restoration of the intensities/yields which prevailed in the past, when the schemes were still functioning well. However, at that time (perhaps 15 years before project preparation), the schemes may have had fully operational headworks but generally lacked distribution systems, and farmers had no advanced water management know-how. Therefore, the post-project irrigation intensities and yields are the combined result of restoration efforts and improved irrigation practices. 6.06 Economic Rate of Return. The ERR of the project is re-estimated at 16% (Annex 1, Table 7) compared to 20% estimated at appraisal. The lower ERR is due to the significantly reduced benefited area, delayed benefits, lower incremental cropping intensities, project cost overrun (in Rupees), and paddy prices that are much lower than originally projected (see note in Annex 1, Table 9). The analysis was carried out in constant 1991 prices. As in the SAR, a project life of 25 years has been assumed. VP irrigated area is estimated to be achieved over five years, with a corresponding build-up of yields; assumptions regarding the WOP situation are described in para 6.04. Contrary to the SAR, incremental crop production benefits from both rehabilitation and modernization have been taken into account. The farmgate price for paddy is based on actual world market prices through 1990 and projected prices for 1995, 2000, and 2005, while prices for non-traded goods are based on domestic financial prices. The market wage for labor as well as other domestic prices and price components have been translated into border prices by applying conversion factors of 0.9 (labor) and 0.85 (all other) respectively (Annex 1, Table 9). Actual project investment costs were re-stated in 1991 prices by using the wholesale price index and MUV.1/ About 25% and 10% of ID and DAS civil works costs were incurred in the northern and eastern regions and have therefore been excluded (Annex 1, Table 8). Unit value index of manufactured exports from selected industrialized countries to developing countries expressed in US dollars.  - 9 - Annual operation and maintenance costs of ID and DAS schemes have been estimated at Rs 600 per ha, including structure maintenance, DAS' operating costs and farmers' labor. 6.07 Farm Incomes, Because of data limitations, only one farm of 0.75 ha has been analyzed. For the same reason, no distinction between agro-climatic zones or tank and anicut schemes could be made. Moreover, income from other farm and off-farm activities - included in the SAR - was excluded from the analysis. Resulting annual net farm incomes at full development are Rs 6,580 WOP and Rs 10,185 WP (Annex 1, Table 6). This means an increase of 55%, compared to 81% estimated at appraisal. The lower incremental net farm incomes are due to the reasons given in para 6.04. G. PROJECT SUSTAINABILITY 7.01 The outlook for sustainability of the project is promising for about half of all schemes, where functioning farmer organizations (FOs) have been established and water management plans implemented. For the remaining half, sustainability will depend on further efforts, mainly by DAS, in forming FOs to carry out water management programs and to organize O&M. 7.02 Although Sri Lankan village irrigation schemes by tradition have some form of beneficiary association, evidently few of these groups have had the strength to keep schemes in adequate operating condition. It thus was one of. the project's aims to provide an organizational structure for these groups of beneficiaries and to strengthen their capability to operate and maintain their schemes. This endeavor was weakened by a decline in farmers' sense of responsibility for their schemes in cases where they were not fully participating in the planning, design and implementation of rehabilitation works. In such schemes, it was found that farmers engaged in maintenance only on that portion of the system that had a direct effect on their individual field(s). 7.03 In DAS' judgement, supported by findings of the survey of scheme performance (para 5.15), further public support to achieve FO self-reliance is indispensable. To provide such support, a sufficient number of field staff is required over at least another two years. Since DAS, through recent government reforms, lost a large number of staff to the provincial administration, where this staff has become mainly absorbed in construction activities, there appears to be little scope for a further O&M strengthening program with existing staff. The solution proposed by DAS is to recruit and train a new cadre of "Institutional Organizers" to assist FOs in setting up O&M funds, organizing deferred repair and maintenance works, and further strengthening of water management. The estimated requirements are about 200 field staff for two years and investments of about Rs 50,000 per scheme, of which farmers would contribute Rs 15,000 in labor or cash. Without such follow-up programs, sustainability of project achievements appears to be at risk in about half of the schemes rehabilitated or modernized. 7.04 A positive impact on sustainability can be expected from legalization of FOs following the ratification of the Agricultural Services Act in late 1990.  - 10 - H. BANK PERFORMANCE 8.01 The Bank's performance during the project period was satisfactory. A total of 14 supervision missions were fielded during the nine-year implementation period, with a considerable degree of staff continuity. The technical backstopping was excellent and so was the response to identified shortcomings of the SAR. Also, the Bank correctly suspended disbursements for project activities in the districts of Ampari, Batticaloa, Jaffna, Mullativa, Trincomalee, Mannar and Vavuniya following the intensification of civil disturbances in 1985/86 which precluded effective supervision. In compensation, the Bank agreed to expand project activities to eight other districts, where rehabilitation needs were initially ranked lower but enough schemes were identified that met project criteria for rehabilitation/modernization. 8.02 The project would have benefited if more attention had been devoted by supervision missions (SM) to the following: (a) Early SMs should have assured the carrying out of baseline surveys for future scheme performance evaluations. (b) It appears that SMs could have identified and remedied the main cause of delays in handing over rehabilitated schemes from ID to DAS, which was DAS' absorption in its own construction program. Thus, a proviso could have been introduced by which no new schemes were to be taken up for modernization before the backlog of ID schemes had been taken over by DAS and corresponding water management plans prepared. 8.03 The main lessons that may be relevant to the Bank are: (a) The process of formation and/or strengthening of FOs to ensure sustained O&M of rehabilitated schemes is not yet completed. Additional manpower and financial resources will be required if this goal is to be achieved. Since the responsible government institutions (DAS and provincial councils) are intensely involved in the implementation of new projects, including the Bank-financed NIRP, there is little scope for necessary follow-up actions, unless they are integrated into and financed by one of the on-going projects, preferably NIRP. (b) A program for evaluation of project impact and performance must include specific provisions for appropriate baseline surveys and, when more than one agency is involved, provide for adequate coordination. The agencies concerned must be convinced of the critical importance of such evaluations; if not, entrusting such studies to independent consultants would appear to be a better choice.  - 11 - I. BORROWER PERFORMANCE 9.01 Despite difficult implementation conditions (civil disturbances, disrupted communications, budget constraints, administrative changes), the Borrower performed well throughout the implementation period. Although there were four changes in the post of Project Director, project implementation was not disrupted thanks to staff continuity at senior staff level in the participating agencies. However, coordination between ID and DAS at field level could have been improved, especially at the survey and design stage of schemes to be rehabilitated. Preparation of a design manual would have been helpful in upgrading standards of design. 9.02 Through a sizeable and diversified training program (para 5.16), the Borrover was able to raise and sustain the implementation capacity of the agencies involved, in particular DAS, and to meet project objectives in farmer training. Training opportunities also stimulated staff continuity. 9.03 The Borrower complied with all covenants (Part III, A). However, serious delays occurred in the submission of audit reports for 1982 and 1983, which were submitted only in June and October 1986, due to shortage of experienced accountants in the district offices of ID and DAS. The situation greatly improved from 1986 onwards but, because of the continuing shortage of experienced accountants, reports continued to be late by 6-9 months. Monthly and quarterly reporting of progress in civil works implementation was timely and satisfactory. 9.04 Lessons to be learned, which may also apply to the Bank are: (a) In village irrigation scheme rehabilitation, the formation of farmer organizations should come before physical rehabilitation. This has the obvious advantage of farmers being sufficiently involved in the planning and design phase to generate a feeling of responsibility for the scheme's performance and maintenance in the post-project phase but also to secure commitments regarding important matters such as access and right-of-way. (b) Cooperation of farmers can be enhanced substantially by providing them the opportunity to participate in rehabilitation activities as paid laborers. (c) Extreme land fragmentation in many existing tank command areas has been identified as a major obstacle to raising these schemes' productivity and the project's aim to do this. It would thus be of comparatively great importance if land consolidation could be combined with scheme rehabilitation in such areas. Success of this approach has been demonstrated by DAS on a pilot scale. (d) Land tenure appears to be a factor to be considered in scheme selection, as the response to rehabilitation has been greater on farmer-owned land.  - 12 - (e) Implementation experience confirms that not only the quality of physical works but also the attention paid by project staff, including occasional visits by senior staff, is essential for the success of the rehabilitation effort. To minimize the impact of staff shortages on project implementation, available staff should be provided with adequate mobility. (f) Monitoring should be used as a management tool to reduce implementation problems and not as an end in itself or because it had been stipulated in the appraisal report. (g) In future village irrigation rehabilitation, the initial scheme preparation report should include a comprehensive survey of agricultural production performance and socio-economic data on the potential beneficiaries, to serve as base line data for future evaluation. This may not only replace a socio-economic study but may also be more reliable, since it would involve only those schemes actually selected for rehabilitation. J. PROJECT REIATIONSHIP 10.01 The Bank/Borrower relationship was satisfactory throughout the project period. This was particularly important in successfully manoeuvering the project through times of budget constraints and civil unrest and in extending the implementation period by four years in order to achieve important project targets. K. CONSULTING SERVICES AND PROCUREMENT 11.01 Consulting services were not foreseen at appraisal, as the implementing agencies, ID and DAS, were considered self-reliant, although in need of strengthened staffing. This strengthening was achieved through a massive training program (para 5.16). Only in the last project year, the need for consultancy input was identified for undertaking an impartial and competent assessment of project impact. However, the relevant contract was arranged under NIRP (para 5.15). Preliminary findings of the study have been incorporated in this PCR. 11.02 Physical rehabilitation had to rely heavily on local contractors. These were often inexperienced, which, together with shortage of supervising staff, resulted in poor construction quality. However, standards improved in the last three to four project years, due to better trained staff and better enforced regulations. L. PROJECT DOCUMENTATION AND DATA 12.01 The SAR provided a satisfactory framework for the Bank and the Borrower. However, unforeseen security problems and more severe than expected budget constraints reduced the SAR's usefulness as a yardstick for physical and financial performance. The provisions of the development credit agreement were generally observed (Part III, G).  - 13 - 12.02 Despite ample SAR provisions, the project did not generate a satisfactory data base from which to reliably evaluate its economic performance. However, the PCR mission could make use of preliminary findings of an on-going evaluation study carried out by consultants (para 5.15).  - 14 - PART II: Proiect Review from Borrower's Persgective 1. Project Design The project has been generally well designed. However, some of the aspects (as described below) need to be improved for similar projects planned for future implementation. 1.1 Selection Criteria (a) The command area under a scheme should not be less than 20 acres. This criteria prevented many promising schemes being taken up under the project. The lower limit should have been 10 acres and this has been adopted for the ongoing National Irrigation Rehabilitation Project (NIRP). (b) As per Annex 4, page 70 of the SAR, the priority for selection of schemes was: (i) Tank systems that are out of use because of recent damage to bund, spillway or sluice and where, with a relatively small investment, total command area could be brought back into irrigated production should receive attention first. (ii) Equal priority as for schemes under (i) above for working projects where major damage within the next few years is possible. (iii) Least priority for long abandoned schemes that would require almost complete reconstruction, which should be considered at a later stage. In para b(i), the term "recent" being not defined (say within a year or so), many schemes that fell into category b(iii) received priority during implementation. It would have been better if schemes falling under category b(ii) were given top priority and the criteria b(i) was modified to state "tanks which have been damaged during the previous year and whose beneficiaries are still resident close to the command area. ......... " In fact the criteria for the ongoing NIRP confines rehabilitation to existing operational schemes. (c) The incremental area brought under direct Maha irrigation should be at least ten times privately irrigated lands submerged or three times other cultivated lands submerged. Based on this criteria, the capacities of many working tanks had been increased to take up additional irrigable area. This necessitated land acquisition proceedings which invariably dragged on for several years causing implementation delays. In some instances, legal procedures had to  - 15 - be resorted to in order to resolve the problems. Such situations could have been avoided by excluding this condition from the selection criteria. The ongoing NIRP does not have a similar condition for the selection criteria. (d) The cost criteria for selection of schemes had two upper limits for cost per acre of land benefitted, i.e. one rate for the existing area and another rate for the new area brought under irrigation. This resulted in the existing area being under-reported in many instances to permit a higher expen4iture for rehabilitation. This was another condition that led to taking up restoration of abandoned tanks. There should have been only one pro-rata cost for rehabilitation as in the ongoing NIRP. 1.2 Farmer Participation Sufficient emphasis has not been given at the time of project design to the participation of farmers in investigation, planning and implementation of the project. Though an initial meeting at the commencement of investigations and a ratification meeting prior to commencement of construction had been envisaged to arrange for farmers' participation in the project, these meetings have not been held in many instances. Strong emphasis should have been given at the time of project design to establish a strong farmer organisation before a scheme is selected for rehabilitation. 1.3 Monitoring and Evaluation At the time of project appraisal, the Land and Water Use Division of the Department of Agriculture was selected to undertake a physical resources evaluation study directed at improving the data base on the hydrology and irrigation potential of village tanks. Likewise, the faculty of Agriculture of the University of Peradeniya was selected to undertake a socio-economic survey in order to evaluate the project's impact on agricultural production, farm incomes and employment. The studies conducted by both organisations did not result in achieving the expected objectives. A better approach would have been to invite proposals from several organisations, evaluate and select rather than pre-selecting particular organisations. 2. Proiect Implementation As mentioned in para 5.01 of Part I, the project has met its implementation targets. The schemes taken up under the project have generally been improved to a satisfactory standard. However, certain shortcomings (dealt with in following paragraphs) have been observed in the implementation process.  - 16 - 2.1 Planning and Design 2.1.1 Selection Criteria One of the serious shortcomings observed in the implementation had been the violation of agreed criteria by developing new schemes and restoring abandoned tanks. The study by the Consultants revealed violation in 20% of the sample schemes studied. The criteria that were violated have already been discussed in para 1. Moneragala was one of the districts where a large number of new schemes have been developed. In almost all of these schemes, land alienation took a very long time (sometimes several years) and DAS refused to take over such schemes until farmers were in place to implement the water management programmes. As a result many of these schemes remained unattended for a long period and there were instances where some schemes reached a state of needing major rectification again to operate them. 2.1.2 Farmer Participation As already mentioned in para 1.2, farmer participation had been poor in project planning and implementation. The Consultants' study indicated that the initial meetings envisaged in the SAR have not been held in about 30% of the sample schemes studied. The figure for final ratification meeting is more alarming with meetings not held in 50% of the sample. This led to criticism of VIRP at various forums. There were few instances where the schemes could not be sustained as farmers were not interested in cultivation. Even in the instances where consultation had taken place, representation had been poor with only few leaders (not elected but powerful personalities amongst farmer community) being present. Naturally in such instances the proposals have been framed to suit the needs of the leaders and not the community. One of the important lessons learnt from the VIRP is: "The establishment of a stable farmer organisation is a prerequisite for the successful rehabilitation, and management of minor irrigation schemes". This has been incorporated as a condition in the selection criteria for the ongoing NIRP. 2.1.3 Design Inadeguacies This had been noticed in 7% of the sample schemes studied by the Consultants. Common faults were substandard investigations, inaccurate hydrological studies and inadequate drainage facilities. (a) Substandard Investigations: Incorrect surveys (particularly levelling) were observed in some instances. This resulted mainly due to lack of supervision and checking by senior staff. However, the work load of senior staff did not permit them to give sufficient attention to supervising surveys. As mentioned already, many new schemes have been developed under VIRP. Several of these were reservoir schemes necessitating the construction of earthen embankments. In some instances, embankments  - 17 - exceeding 50 feet in height had been constructed. Subsurface investigations have not been done for these schemes and in some instances resulted in failure due to inadequate foundation cut-off etc. In some instances, the capacities of existing schemes have been increased by raising the level of embankments but the condition of existing embankment had not been investigated in these cases. As a result some of these schemes breached after impounding due to the inability of the existing embankment to cater for higher hydraulic forces. Another shortcoming was the insufficient attention paid to surveying the construction materials necessary for rehabilitation. (b) Hydrological Studies: There had been no historical records of inflows and issues to minor schemes in Sri Lanka. Hence hydrological analysis to determine water availability was based on the iso-yield curves prepared by the department. These iso-yield curves had been prepared based on stream gaugings in major rivers and did not accurately reflect the inflow from small catchments. There were instances where additional areas for irrigation had been proposed based on erroneous inflows and schemes promoted. After rehabilitation the tanks have not filled up to the required levels. However, a little further study is needed in this regard as there had been a general decline in seasonal rainfall through the VIRP implementation period. The flood studies had been mainly based on the rational formula. The use of this formula for catchments exceeding one or two sq.miles gives erroneous results. Some of the completed VIRP schemes suffered flood damage due to inadequate spillway capacity. Another important lesson learnt from VIRP is that there is an urgent need to conduct research into small catchment hydrology in Sri Lanka. This, hopefully, could be accomplished by the proposed Research Management Unit (RMU) to be set up in the Irrigation Department under the ongoing NIRP. (c) Drainage Facilities: Drainage aspects have not been considered at design stage in most of the schemes. There were instances where proper outlet channels have not been provided for spillways. 2.1.4 Construction AsRects Construction defects have been observed in 10% of the sample schemes studied by the Consultant. Most of the observed defects were due to poor quality earthwork. A common defect detected was in the excavation of existing bunds to provide for new sluices. The side-slopes of the excavated trenches had been almost vertical resulting in poor bonding of backfill with the existing bund. As a result there were several instances of completed tanks breaching at the sluice section. Other defects noted were the use of substandard materials such as timber and aggregates for concrete. These problems are expected to be minimised in the ongoing NIRP by having training programmes for supervising staff and also by ensuring regular supervision by senior staff.  - 18 - 2.1.5 Poor Coordination between Irrigation Department (ID) and Department of Agrarian Services (DAS) This was noticeable particularly in the early years of the project and the situation improved later on. In the ongoing NIRP it is envisaged to overcome this shortcoming by: (a) involving DAS fully in scheme identification as well as investigations; and (b) institutional development programmes commencing from scheme identification time and being a continuous process during the implementation. 2.1.6 Appointment of "Tank Supervisors" As outlined in para 5.11 of Part I, the cadre of "tank supervisors" were not created. The duties were assigned to APT and the 'Vel Vidane' (Leader of scheme). Vel Vidane is a traditional leader from the village, chosen by the farmers. As he belongs to the same community, his authority and controlling powers are rather limited. Many of them lack creative thinking and initiative for changes in traditional water use practices. Further, no incentives for them were forthcoming from the farmers benefitted by VIRP. (In early days a remuneration had been paid in kind by the farmers after harvesting but this practice has gradually eroded.) Therefore, the attitude of Vel Vidanes towards water management activities had dwindled in recent times. To overcome this situation in the ongoing NIRP, farmer organisations (PO's) would be formed at pre-construction stage. Institutional Organisers (IO's) are also to be appointed early to strengthen FO's. Training programmes would be carried out for farmers and it is envisaged to train one farmer as a water master. 3. Proiect Evaluation As mentioned in para 1.3, both evaluation studies (Physical and Socio- Economic) did not reach the desired levels. 4. Training The training programme conducted under the project really gathered momentum in the latter half of the project. The training conducted was useful especially to minimise the implementation shortcomings observed in the first half of the project. 5. Technical Assistance It is accepted that the performance of the project's technical advisor to the Department of Agrarian Services was good.  - 19 - 6. Project Results The project was one of the most successful projects in comparison to minor irrigation development projects launched in the past. 7. Proiect Impact As stated in para 6.05 of Part I, the project's main impact had been in the restoration of the intensities/yields which prevailed in the past. According to Consultant's study the actual cultivated area has increased as a result of the project by over 10% in 22% of the sample schemes in Maha, and only in 16% of the schemes in Yala. The level of convenience in obtaining irrigation water after rehabilitation is another good measure of the physical performance. During the Consultant's study, 34% of a sample of 229 farmers were of the opinion that "level of convenience was good" and another 34% were of the opinion that it was "somewhat good", for Maha cultivation of paddy. For Yala cultivation of paddy, the results were not so good, with 15% "good" and another 20% "somewhat good". 8. Project Sustainabilitv 8.1 ODeration 8.1.1 At the commencement of the VIRP an elaborate water management programme was envisaged. However, the DAS was initially lacking in technical staff sufficiently exposed to modern water management techniques, and hence the APT did not possess the depth of technical ability to manage an ambitious programme. This explains the initial delays in implementing proper water management practices. 8.1.2 The farmers themselves were used to traditional water management practices such as "rotations" and commencing irrigation from the tail end and working up to the head. The water management practices introduced by the project were mainly concentrated on improving the procedures the farmers were used to already. However, what was important was to provide the "scientific tools" to measure and provide only the requirements for optimum yields, making best use of rainfall and reducing to a minimum the waste of water by not only efficient distribution, but also proper water management within the field. The provision of these proper tools has fallen short and transfer of the know-how has not taken place to uplift the traditional water management to modern level of water management. 8.1.3 In the pre-1960's, the Vel Vidane system functioned very well. However, the social environment had considerably changed since then and after a period of state intervention, to create a self-reliant farmer community needs time and direction to the correct path. The catalyst to direct the farmers was lacking in the project, and the lesson learnt is the necessity to employ Institutional Organisers in similar projects to be implemented in the future.  - 20 - 8.1.4 It was mentioned in para 4 that the project had a successful training programme. Hovever, this was mainly staff biased and the training of the farmers and farmer representatives has not been adequate. 8.2 Maintenance As mentioned in para 7.01 of Part I, the maintenance in about 50X of the schemes vas below the expected standards. This was mainly due to the lack of ownership feeling amongst the farmer community. The farmer community of yesteryear had a strong feeling of ownership of the village irrigation schemes. As a result the village scheme and the farmer community were a self-sustaining entity. Hence the thought that the "farmer community are the owners of the irrigation system" should be imbibed by the farmers. The project did not convey this message effectively as there was insufficient farmer participation in the planning, design and construction. 9. Bank Performance The performance of the Bank during the project cycle could be classed as excellent. There was a close rapport between the supervision missions and GOSL staff. Though all the officers of the supervision missions performed their functions excellently, GOSL would be failing in their duty if the excellent cooperation extended to project staff by the mission leader during the last 5 years of project implementation was not mentioned. The only instance where the supervision mission and GOSL staff could not agree was on the quality of "Physical Resources Evaluation Study". Perhaps, the Bank could have sent a Hydrologist in a supervision mission to study the shortcomings pointed out by the Irrigation Department. 10. Borrower Performance The changes in the position of Project Director occurred mainly due to retirements and promotions. However, the implementation of the project was not adversely affected due to these changes. It is agreed that the coordination between ID and DAS at field level could have been better, especially at survey and design stage. The preparation of a design manual also vould have been useful. The para 9.04 of Part I mentions the lessons to be learned. All these lessons have been picked up in formulating the ongoing NIRP. 11. Comments on Part III The data presented are accurate and in agreement with the records kept by ID and DAS.  - 21 - PART III: Statistical Information A. RELATED IDA CREDITS Cndit/ Purpose Approval Status Comments Tith Date / National Irrigation Rehabilitation of minor, medium and 06/06/91 On-going This project can be considered Rehabilitation Project large irrigation projects to stabilize and (06/98) a follow-up project to VIRP as (Cr. 2260-CE) increase agricultural production and far as minor schemes are incomes and raise standards of living, concerned. Kurunepla RD To raise productivity, employment, 1979 Completed Included irrigation rehabilita- (Cr. 891-CE) incomes and living standards of the rural (12/85) lion; no major,problems were population through an integrated package encountered.Wf of investments and institutional improvements. Tank Irrigation To improve irrigation supply, reduce 1976 Completed No major problems were Modernization Project seepage losses, provide inputs, strengthen (06/83) encountered.!' (Cr. 666-CE) agricultural extension and water management in five major tank schemes in the dry zone. Third RDP To increase productivity, employment, 05/12/83 Cancelled Since 1984, the project became (Cr. 1363-CE) income and living standards in the increasingly affected by the districts of Mannar and Vavuniya. ethnic conflict. The credit was cancelled in March 1986. Actual or expected closing date in parentheses. / Howvever, the PPAR noted that despite achievement of physical targets efforts to improve water management of irrigation schemes have been largely ineffectual. c/ Similar problems as indicated in footnote ~J were found.  - 22 - B. PROJECT TIMETABLE Item Date Planned Date Revised Actual Date Identification 1979 a/ - 10/1979 k/ Preparation Early 1980 - 05/1980 Appraisal 08/1980 - 09/1980 !/ Credit Negotiations 03/1981 - 03/1981 Board Approval 05/1981 - 06/16/1981 Credit Signing n.a. - 07/15/1981 Credit Effectiveness n.a. - 09/24/1981 Project Completion 12/31/1985 - 12/31/1990