Sri Lanka - Major Irrigation Rehabiltation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 14742 PROJECT COMPLETION REPORT SRI LANKA MAJOR IRRIGATION REHABILITATION PROJECT (CREDIT 1537-CE) JUNE 29, 1995 Agriculture and Natural Resources Operations Division Country Department I South Asia Regional Office This document has a restricted distribution and may be used by recipients only in the performance of their ofricial duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Name of Currency (abbreviation) = Rupees (Rs) Appraisal Year (1984) US$1.00 = Rs 25.4 Intervening Years' (1985-92) Average US$ 1.00 = Rs 34.7 Completion Year (1993) US$1.00 = Rs 48.3 FISCAL YEAR OF BORROWER January I to December 31 WEIGHTS AND MEASURES 1 meter (m) = 3.2808 feet (ft) 1 kilometer (km) = 0.6214 miles (mi) I hectare (ha) = 2.4711 acres (ac) 1 kilogram (kg) = 2.2046 pounds (lb) 1 metric ton (ton) = 1.0161 long tons (ton) I m3/sec = 35.3357 cubic ft/sec (cusec) 1 bushel of paddy = 20.8655 kg ABBREVIATIONS ARTI Agrarian Research and Training Institute CIDA Canadian International Development Agency DA Department of Agriculture DAS Department of Agrarian Services DCO Distributary Channel Organization FO Farmer Organization GOSL Government of Sri Lanka ID Irrigation Department IIMI International Irrigation Management Institute IMD Irrigation Management Division INMAS Program for Integrated Management of Major Irrigation Systems 10 Institutional Officer LCD Land Commissioner's Department MASL Mahaweli Authority of Sri Lanka MCM Million cubic meters MLLD Ministry of Lands and Land Development NIRP National Irrigation Rehabilitation Project (Cr. 2260-CE) PAR Performance Audit Report PM Project Manager SDC Swiss Development Cooperation (Federal Department of Foreign Affairs) USAID United States Agency for International Development  FlOR OMFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General June 29, 1995 Operations Evaluation MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Sri Lanka Major Irrigation Rehabilitation Project (Credit 1537-CE) Attached is the Project Completion Report (PCR) on the Sri Lanka-Major Irrigation Rehabilitation project (Credit 1537-CE, approved in FY85) prepared by the South Asia Regional Office. The Borrower prepared Part II. The project sought to improve irrigation supplies and water use at seven tank schemes in the dryer areas of the country, building upon the experience of a recently completed Bank- supported project. It gave special emphasis to solving recurring problems with operations and maintenance (O&M), by strengthening a new ministerial division responsible for integrated scheme management, and by toughening the rules on cost recovery. Early in project implementation, civil strife necessitated the suspension of work in the three most northerly schemes. The project was cofinanced by the Canadian International Development Agency and the Swiss Development Cooperation. The engineering works were successfully completed, including three additional investment operations not explicitly provided for at appraisal. The record of cost recoveries was dismal, as with the previous project, prompting government to shift its strategy from farmer repayments to farmer participation in O&M of the canal infrastructure. Progress in this new direction is uneven but improving. The PCR presents a substantial annex on re-estimated rates of return, showing that the economic impact is higher than as appraised. However, the data base available to the PCR team was weak, the anticipated contribution from the agricultural extension service failed to materialize, and there is evidence on all four remaining schemes that farmers have not yet shifted appreciably to the better cropping patterns and practices anticipated at appraisal. Project outcome is rated as marginally satisfactory. The institutional developments-aiming at integrated management and depending on inter-institutional coordination-are rated as modest. The PCR implies the rehabilitation works are sustainable. The Swiss Development Cooperation, which was responsible for the technical assistance component including monitoring and evaluation, comments in an annex that the PCR is too optimistic about farmer O&M, long-term production impact, and quality of construction. Sustainability must, therefore, be rated as uncertain. No audit is planned. 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 MAJOR IRRIGATION REHABILITATION PROJECT (Cr. 1537-CE) TABLE OF CONTENTS PREFACE . ............................................... i EVALUATION SUMMARY ... PARTI ................I 1. Project Identity ............................................... 1 2. Background ............................................... 1 3. Project Objectives and Description ........................................ 1 4. Project Design and Organization ....................................... 3 5. Project Implementation .............................................. 3 6. Project Results ............................................... 8 7. Project Sustainability .............................................. 9 8. Bank Performance .............................................. 11 9. Borrower Performance .............................................. 11 10. Project Relationship ............................................... 12 11. Consulting Services and Contracts . ...................................... 12 12. Project Documentation and Data ......................................... 13 PART II ............................................... 15 I. Observations on Part I of the Project Completion Report ......... .. .............. 15 Performance of Implementing Agencies ................ ................... 15 A. Irrigation Management Division ..................................... 15 B. Irrigation Department ............................................ 16 C. Land Commissioner's Department ........... ......................... 17 D. Agriculture Department . ......................................... 18 E. Agrarian Services Department ........................................ 19 F. International Irrigation Management Institute .............................. 19 II. Changes in Project Composition ........................................ 19 III. Training ............................... 20 IV. Bank Performance .............................................. 20 V. Borrower Performance ............................................. 21 VI. Performance of Co-financiers ......................................... 22 VII. Observations on Part III of the Project Completion Report ......... .. ............. 22 Annex I - Project Achievements Annex 2 - Prorata Costs This doctument 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.  PART III ...... . 45 1. Related Bank Loans and/or Credits ....................................... 45 2. Project Timetable ........................... ... 46 3. Credit Disbursements ............................................. 46 A. Estimated and Actual IDA Disbursements ........... . .. . .. .. . .. . .. .. . . 46 B. Cofinanciers' Estimated and Actual Disbursements ........ . . . . . . .. . . . . . . . . 47 4. Project Implementation ............................................. 48 A. Physical Achievements .................................. ...... . 48 B. Farmer Organizations ......................................... . 49 C. Training. ............................................ 49 5. Project Cost and Financing .o........................................... 50 A. Project Cost ............................................. 50 B. Project Financing ........................................... . 51 6. Project Results ............................................. 52 A. Direct Benefits ............................................. 52 B. Economic Impact ............................................. 52 C. Financial Impact ............................................. 52 D. Studies ............................................. 53 7. Status of Covenants .......................... .. . 55 8. Use of Bank Resources ............................................. 56 A. Staff Inputs ............................................. 56 B. Missions ................................................. . 57 ANNEXES Annex 1: Financial and Economic Re-Evaluation Annex 2: Equipment Procured Under the Project Annex 3: O&M Fund Allocation Annex 4: Comments from Switzerland (Federal Department of Foreign Affairs, Swiss Development Cooperation) MAP: IBRD 18021  PROJECT COMPLETION REPORT SRI LANKA MAJOR IRRIGATION REHABILITATION PROJECT (Cr. 1537-CE) PREFACE 1. This is the Project Completion Report (PCR) for the Major Irrigation Rehabilitation Project for which Credit 1537-CE in the amount of SDR 17.0 million (US$17.0 million equivalent) was approved in December 1984. Co-financing was provided by the Canadian International Development Agency (CIDA) in the amnount of US$6.8 million equivalent " and by the Swiss Development Cooperation (SDC) in the amount of US$7.6 million equivalent.2' After two one-year extensions, the final IDA credit closing date was June 30, 1993. The credit account was kept open until October 31, 1993, to allow submnission of final withdrawal applications. Total disbursements amounted to SDR 11.09 million (65% of the credit). The remaining credit balance of SDR 5.91 million was cancelled effective May 6, 1994 (the date when unused funds in the Special Account were returned to IDA). Disbursement of the SDC contribution amounted to US$7.92 million equivalent (104%), and of the CIDA joint contribution to US$2.61 million equivalent (87%). 2. The PCR, except Part II, was prepared by an FAO/World Bank Cooperative Program mission 3/ which visited Sri Lanka in January/February 1994, and reviewed by the Agriculture Operations Division, Country Departmnent III, South Asia Region. Preparation was based on the Staff Appraisal Report and legal documents; IDA supervision reports and project files; field visits; and discussions with Government officials, beneficiaries and IDA staff associated with the project. Part II was prepared by the Borrower.4' " The Canadian contribution was provided through parallel financing (US$3.8 milion equivalent) and a joint contribution managed by IDA (US$3.0 million equivalent). 2/ Comments on the PCR from SDC are included as Annex 4. 3/ Mr. S. Rajagopal (irrigation engineer, mission leader), Ms. Suzanne Raswant (economist, consultant) and Dr. S. Amarasiri (agronomist, consultant). 4' Irrigation Management Division, Ministry of Irrigation, Power & Energy, October 25, 1994.   iii PROJECT COMPLETION REPORT SRI LANKA MAJOR IRRIGATION REHABILITATION PROJECT (Cr. 1537-CE) EVALUATION SUMMARY Objectives 1. The project's objectives were to increase agricultural production under seven major irrigation systems located in the dry zone of Sri Lanka and to introduce an effective program for their integrated management. These objectives were to be achieved through: (a) physical rehabilitation of existing systems for optimum utilization of available water; and (b) support for the Program for Integrated Management of Major Irrigation Systems (INMAS), to ensure proper operation and maintenance of the irrigation systems and complementary programming of agricultural activities and inputs. The seven systems originally to be rehabilitated covered a total of 46,000 ha, but the actual project area was 23,800 ha", as explained below. Implementation Experience 2. Early in project implementation, civil strife necessitated the suspension of rehabilitation work in three schemes (Giant's Tank, Morawewa and Iranamadu). The situation did not improve and the suspension became permanent. However, an unexpected breach of the dam at Kantalai required immediate repair to put the Kantalai system back into operation. Two other items were implemented in fulfillment of Government commitments under the IDA credit: provision of proper outlets in the Huruluwewa Feeder Canal (carrying Mahaweli supplies) to regularize illegal tapping; and provision of pumping equipment and civil works rehabilitation for the Lift Irrigation System of the Rajangana scheme. The actual area covered by the project was reduced to that commanded by the remaining four schemes (Nachchaduwa, Huruluwewa, Rajangana and Kantalai) and by the Huruluwewa Feeder Canal and Rajangana Lift Scheme, totalling about 23,800 ha. Due to the civil unrest throughout the first half of the implementation period affecting the remaining four schemes as well, the project could not be completed as scheduled. In order to complete project works, the IDA credit closing date was extended by two years to June 30, 1993. Project costs are now estimated to be Rs 1,329 million (US$38.7 million) against the SAR estimate of Rs 1,206 million (US$43.2 million). Expenditure is higher in Rupee terms but lower in US dollar terms due to the depreciation of the Rupee against the US dollar. 3. As a newly established division, the Irrigation Management Division (IMD) experienced difficulties in coordinating project activities, particularly with respect to the Irrigation Department (ID). The situation improved during the course of implementation, with ID officials realizing the importance of working with farmers in the management of irrigation systems. In accordance with the INMAS concept, IMD Project Managers successfully established Farmer Organizations with the assistance of Institutional 1 Figure based on recent surveys (see Annex 1, para 6).  iv Officers. The services of these officers have been greatly appreciated by farmers and it appears that farmers would have liked to have their support for some more time. However, the farmers interviewed by the PCR mission were unwilling to pay for the services, which are to end two years after project completion. It is likely that many farmer organizations are not yet ready to be fully autonomous and responsible for water management. 4. The Government of Sri Lanka (GOSL) agreed to collect water charges to pay for scheme operation and maintenance (O&M) after extensive dialogue with the Bank. About 50% of assessed charges were collected in 1984, but the collection rate deteriorated to less than 5% by 1987 due to: (i) civil strife;" (ii) difficulties in taking legal action against defaulters; and (iii) lack of political will. The study on procedures for collection of O&M fees concluded that rather than collecting fees, O&M should be carried out through joint management of irrigation systems. With the agreement of IDA, GOSL adopted a policy of participatory management, in line with the INMAS concept, with farmers officially taking over responsibility for O&M of distributaries in addition to field channels. However, this arrangement is not yet working well due to technical difficulties encountered by the farrners. 5. Construction equipment was procured to allow simultaneous work on all seven schemes. Procurement took place in the first two project years when work was under way in all seven schemes. The equipment has been put to good use. Maintenance has also been good, as evidenced by the remaining useful life. Other equipment and vehicles were also provided. Technical assistance was provided to develop a computerized hydrological model to assist in the operation of the irrigation systems, set up a pilot programn for technological improvements in control of irrigation supplies, guide the rehabilitation process in general, and improve the management information and project accounting system. 6. Rehabilitation works were carried out by ID as envisaged in the SAR. Construction was by force account, private contractors and farmer organizations. Construction quality is good. The use of farmer organizations was an innovative concept in Sri Lanka and the construction contracts provided substantial income to these organizations. 7. The targets for both gravel-topped and unpaved roads in the four remaining schemes were achieved, with some modifications. More unpaved roads were rehabilitated in Huruluwewa and Rajangana, and fewer in Nachchaduwa and Kantalai. These roads have facilitated the quicker movement of produce to markets. Significant progress was made on regularization of encroachments, with over 14,700 lots regularized in the course of the project, almost 60% of which in Rajangana. 8. Studies on both catchment management and on-farm water management were carried out in the Huruluwewa and Nachchaduwa schemes and are continued under the National Irrigation Rehabilitation Project (NIRP, Cr. 2260-CE). The pilot area program using technological improvements in water control was carried out at two systems and is being continued at one, also under NIRP. 9. Under the project, two Agricultural Service Centers have been rehabilitated, and five garages, fifteen fertilizer stores and three wells have been constructed. The fertilizer stores have been taken over by farmer organizations, in line with GOSL's policy of privatizing fertilizer supply. The Department of " Farmers' lives were threatened if they paid any charges to GOSL.  V Agrarian Services has also undertaken the registration of farmer organizations under Sections 56A and 56B of the Agrarian Services Act." 10. The project contributed towards the purchase of the headquarters building for the International Irrigation Management Institute (IIMI) in Colombo. Cost overruns were significant. There was no direct contribution from IIMI to the project as the expected IIMI field site was not established because of the security situation. 11. Intensive training was carried out locally by IMD to familiarize officials with INMAS and participatory management. A large number of farmers and farmer representatives were trained in participatory management, use of irrigation structures, financial and contract management and O&M of irrigation systems. IMD and ID staff received training, both locally and overseas, to enhance their technical and managerial capabilities. While engineers who obtained post-graduate degrees were actually employed in the rehabilitation schemes and contributed to their success, considerable exposure of other ID staff to water management has not resulted in any significant water savings. Hydrological and operational models developed under the project have not been put to use, except at Kantalai. Project Results 12. The project objectives have been partially achieved. Rehabilitation works have been carried out as planned in four schemes and the INMAS concept has been made operational. On the basis of available information and the assumptions detailed in Annex 1, incremental paddy production at full development is estimated to be about 32,400 tons, against 37,300 tons expected at appraisal. In addition, considerable quantities of eggplant, soybean, chillies, cowpeas and mungbean would be produced, whereas the SAR assumed only soybean. Many farmers persist in cultivating paddy in spite of the better remuneration offered by other field crops. Part of the problem is lack of extension services to provide the necessary techniques and shortage of credit. The project's ERR is re-estimated to be 19%, a slight improvement over the SAR estimate of 16% for the original project. Sustainability 13. Proper O&M is the key to project sustainability. Water for paddy cultivation in the maha season (October to March) is provided by rainfall and irrigation supplies. Continuous computer simulation studies are required to enhance the conjunctive use of these two sources of supply. In the yala season (April to September), irrigation supplies need to be provided for both paddy and other field crops'. Operation of the irrigation system has to be established through computer simulation in order to optimize water use. O&M cannot be carried out by farmers alone but requires technical input from trained ID staff. While participatory management is the only way to sustained O&M, in order to assure proper technical and other inputs from ID, both farmers and GOSL need to contribute funds. Substantial increase in funds allocated for O&M is necessary. Farmers' incomes have to rise to enable them to pay for O&M. This can only be achieved through crop diversification, which requires good extension services and marketing arrangements. Sustained operation of the project would depend on the extent to which these measures are carried out in the future. " Registration under 56A enables the organizations to enter contracts, registration under 56B gives them legal status and enables them to borrow. 2/ Eggplant, chillies, soybean, cowpeas and mungbean.  vi Lessons Learned 14. The project has adopted an innovative concept of integrated management of irrigation systems. Farmer organizations have been established and registered. Farmers are beginning to understand the operation of the irrigation system and are enthusiastic about the activities undertaken so far by their organizations. The main lessons of experience are: (i) making farmer organizations operational on a sustained basis is a slow process and, therefore, adequate long-term measures, both institutional and financial, should be built into the design of projects; (ii) more attention needs to be given to raising both general awareness and technical capability of lower-level technicians (who work closely with farmers) in the area of water management; (iii) in projects whose aim is to improve agricultural production, more attention should be given to agricultural extension services and marketing; (iv) O&M of irrigation systems cannot be carried out by farmers alone; it requires skilled inputs from technical personnel as well as adequate funding; (v) in order to generate sufficient funds to cover O&M expenses, it is necessary to improve farmers' incomes through crop diversification in the yala season where possible; and (vi) farmer organizations undertaking construction activities is a good concept, but they should be protected against major price fluctuations in materials such as cement (price escalation clause) to enable them to carry out the work without delay.  PROJECT COMPLETION REPORT SRI LANKA MAJOR IRRIGATION REHABILITATION PROJECT (Cr. 1537-CE) PART I 1. Project Identity Project Name Major Irrigation Rehabilitation Project (MIRP) Credit No. : 1537-CE RVP Unit South Asia Country Sri Lanka Sector : Agriculture Subsector Irrigation 2. Background 2.1 The rapid expansion of irrigation in the sparsely populated dry zone following independence was accompanied by a growing awareness that benefits from the existing irrigation systems had not achieved their full potential. Attemnpts to improve performance were limited by the physical characteristics of the schemes, their deterioration due to inadequate maintenance, and the perpetuation of out-of-date management. The Government of Sri Lanka (GOSL) initiated action to address the issues of rehabilitation, modernization and improved management of existing systems through a range of pilot projects and institutional changes. Donors, particularly the United States Agency for International Development (USAID) and IDA, subsequently became involved, with the IDA-supported Tank Irrigation Modernization Project (TIMP, Cr. 666-CE) becoming the first attempt to address the problems of major existing irrigation systems. In a parallel fashion, the Village Irrigation Rehabilitation Project (VIRP, Cr. 1160-CE) was concerned with the problems of minor systems. GOSL gave priority to rehabilitation projects that would maximize returns from past investments. 2.2 MIRP represented a continuation of the prograin started under TIMP to rehabilitate all major irrigation systems. Lessons learnt from TIMP guided the planning of MIRP. These related to the need for: (i) advance planning of works; (ii) alowing sufficiently long construction periods; (iii) prevention of indiscriminate use of catchment areas; and (iv) organization of an efficient system of water management. Continued Bank Group support for irrigation projects was considered necessary to ensure that these lessons were incorporated in system design, operation and management, and that the institutional and policy changes proposed by GOSL to improve water management were implemented. 3. Project Objectives and Description 3.1 Project objectives were to increase agricultural production under seven major irrigation systems and to introduce an effective program for their integrated management through: (i) rehabilitation of the systems for optimum utilization of available water; and (ii) support for the Program for Integrated  2 Management of Major Irrigation Systems (INMAS) to ensure proper operation and maintenance (O&M) and distribution of irrigation supplies and the complementary programming of agricultural activities and inputs. 3.2 Project components, broken down by imnplementing agency, consisted of: (i) Ministry of Lands and Land Development, essentially the Irrigation Management Division (IMD) of the Ministry, responsible for project implementation including administration of INMAS in the seven systems. Investment costs for setting up this new Division were covered as well as recurrent costs for the duration of the project, and training for staff and farmers in aspects of farmer organization and water management. (ii) Irrigation Department (ID), responsible for rehabilitation of the seven selected systems: Huruluwewa, Giant's Tank, Iranamadu, Kantalai, Morawewa, Nachchaduwa and Rajangana. In addition to the main works, a pilot program was included to ascertain the impact of technological improvements in control of irrigation supplies in one distributary (150-200 ha) in each of four selected systems. Funding was provided for facilities needed to rehabilitate and maintain the systems, including: equipment and vehicles; buildings to supplement available residential accommodation for O&M; aerial and soil surveys; technical assistance for developing hydrological models, establishing the pilot program, providing technical guidance for rehabilitation, and improving management information and project accounting; and local and overseas training of ID staff. (iii) Land Commissioner's Department (LCD), responsible for regularizing farmers' encroachment on adjacent lands, and major repairs to LCD roads in the project area. The project supported incremental costs of LCD's operations to regularize landholdings under the project; purchase of vehicles; and repair of 35 miles of gravel-topped and 255 miles of unpaved LCD roads. (iv) Department of Agriculture (DA), responsible for conducting studies on catchment management and on-farm water management. Funding was provided for incremental staff and other costs; equipment for data collection and processing; vehicles; and training of a few staff in foreign centers involved in similar studies. (v) Department of Agrarian Services (DAS). The project provided financing for additional agrarian service centers, fertilizer stores and garages, as well as vehicles to improve mobility of DAS staff and emergency supplies to enhance availability of agricultural inputs. (vi) University of Peradeniya (UP). Resources were provided to enable the university to follow up on its initial baseline socio-economic survey with mid-term and completion evaluation studies. (vii) International Irrigation Management Institute (IIMI). As it was considered that the establishment of HIMI in Sri Lanka was in the country's interest and that INMAS could benefit from IIMI advice, funding was provided to cover GOSL's contribution to the cost of office and residential facilities for IIMI. 3.3 At appraisal in 1984, total cost of the six-year project was estimated to be Rs 1,206.2 million (US$43.2 million). Financing was to be shared by GOSL - 27%, IDA - 39%, CIDA - 16%, and SDC - 18%. The IDA credit was approved in December 1984 and the project was scheduled to be completed by the end of 1990 with a credit closing date of June 30, 1991.  3 3.4 The newly established IMD was responsible for implementing the project and coordinating activities of the various participating departments and agencies. ID was responsible for design and implementation of rehabilitation works and for O&M of the irrigation systems in collaboration with IMD. As part of INMAS, each irrigation system was to have a hierarchy of committees, mainly comprising of farmer representatives, which would meet regularly to resolve problems in implementing the rehabilitation and integrated management programs. 4. Project Design and Organization 4.1 The project was prepared by FAO/CP with the assistance of a national engineering consulting company, which prepared reports on cropping potential, and the University of Peradeniya, which provided the socio-economic data. Appraisal was done by a joint IDA/CIDA mission. The project was generally well prepared and agreed with GOSL's priority for rehabilitation projects that would maximize returns from past investments. 4.2 The conceptual foundation for the project was clear and shared by all concerned. The selection of the seven systems, as well as others in GOSL's phased rehabilitation program, based on their potential for increased paddy production was considered to be in line with GOSL's policy of crop diversification, since the increase in paddy production on the selected systems would allow land marginally suited for paddy under other systems to be used for non paddy crops. However, the emphasis on paddy production has not encouraged efficient water management in the project. 4.3 Although the introduction of the INMAS program had already begun on major irrigation systems in the country, the project was instrumental in ensuring its implementation on the seven (later four) systems and in raising awareness of the importance of farmer involvement in irrigation management. The project concept was innovative in so far as water management was made a major concern. To ensure that adequate attention was given to water management, appointment of IMD as the implementing agency was probably necessary. However, working relationships between IMD and ID were strained during part of the project period, but without visible effect on implementation. 4.4 Components were appropriate in scope and scale in light of project objectives. Some provision was made for improvement of input supply to farmers through the DAS component. Support for extension services was not considered necessary because of on-going activities, including the Agricultural Extension and Adaptive Research Project (Cr. 931-CE; PCR issued March 1989, PAR issued May 1992). 5. Project Implementation 5.1 General. Project implementation was seriously affected by the difficult security situation in Sri Lanka. In June 1986, work had to be suspended at Giant's Tank, Iranamadu and Morawewa, situated furthest to the north. Although activities continued under the other schemes, the security situation intermittently made it difficult for officials to work on site and contributed to implementation delays. In order to complete the rehabilitation works, the credit closing date had to be extended twice, first to June 1992 and then to June 1993. 5.2 An unexpected event was the breaching of Kantalai dam in April 1986. GOSL approached IDA and the other co-financiers, CIDA and SDC, for support in the restoration of the dam. IDA and SDC agreed to reallocate funds (about US$2.08 million and US$0.93 million, respectively) from the three  4 northern schemes where work had been suspended, and CIDA agreed to provide an additional contribution of US$0.46 million. The original cost estimate was US$4.63 million equivalent (Rs 125 million), but actual costs amounted to over US$7.5 million, a substantial cost overrun. 5.3 There were two other developments. As a commitment under the project, GOSL submitted (and later implemented) a program for the use of the Rajangana lift irrigation system, as well as remedial measures for improving the efficiency of Mahaweli feeders to the project area. GOSL reached agreement with the farmers who would assume responsibility for O&M, provided new pumps were installed and civil works rehabilitated. The project supplied replacement pumps and construction materials for the rehabilitation of the civil works, to irrigate about 1,700 ha. To solve the problem of the unauthorized tapping of water from the Huruluwewa feeder canal, agreement on water sharing was reached between the illegal irrigators on the feeder canal and the Huruluwewa farmers. The illegal tappings were replaced by proper outlets in the feeder for which the project provided financing. 5.4 The project started slowly and IDA credit disbursements lagged behind appraisal estimates throughout implementation. Despite the two closing date extensions, only 65 % (83 % in US$ terms) of the credit was disbursed. Disbursement of co-financiers' funds (in US$ terms) was significantly better, with over 100% for the Swiss contribution and 93% for the CIDA contribution (see Part III/SB). 5.5 Total re-estimated project costs are Rs 1,329 million (US$38.7 million) against the SAR estimate of Rs 1,206 million (US$43.2 million). The fact that the SAR estimate has been exceeded in Rupee terms but not in US dollar terms is due to the depreciation of the Rupee against the dollar. Expenditure on rehabilitation works for two of the schemes exceeded appraisal estimates (in Rupee and US$ terms), as did the cost of the building for IIMI, while expenditure by implementing agencies other than ID and IMD was substantially below. 5.6 Irrigation Management Division. As a newly established, rather small division, IMD experienced difficulties in coordinating project activities, particularly with respect to ID. Especially in the early years, IMD field staff suffered from lack of authority and working relationships were strained. In accordance with the INMAS concept, IMD Project Managers focused on establishment of Farmer Organizations (FOs). The envisaged committee structure was established and Project Committees provided fora for communication of farmers' problems to the relevant implementing agencies. However, farmers expressed disappointment at the lack of response, partly a result of other agencies' inadequate participation in INMAS. The situation improved in the course of the project, with ID staff in particular realizing the importance of working with farmers in management of the irrigation systems. Other line agencies participated much less. In spite of repeated attempts by IMD to involve the Department of Agriculture in providing extension services to farmers in the project, little extension has been made available. In fact, the situation was complicated by GOSL's withdrawal of all village-level extension agents to perform administrative duties in villages. 5.7 Institutional Officers (1Os) have been instrumental in establishing farmer organizations and were much appreciated by farmers, to the extent that many FOs did not wish to lose the services of the 1Os, which are due to cease two years after project completion. On the other hand, farmers were not prepared, and possibly not in a position, to take over the remuneration of IOs and employ them directly. In fact, it is likely that many FOs still need institutional support and are not ready to be fully autonomous and responsible for water management.  5 5.8 One of the tasks of IMD was to propose to GOSL water charges to be levied and to ensure their timely payment by farmers. The Bank had an extensive dialogue with GOSL since the early 1980s on water charges and, in 1983, nationwide water charges were approved, which were to increase progressively to fully cover O&M costs by 1991. The project required provision of detailed annual reports on progress made in water charge collection, a review of the program by mid-1986 and a commitment from GOSL to provide adequate funds for O&M until full cost recovery was achieved. While about 50% of assessed charges were collected in 1984, the collection rate deteriorated to less than 5 % by 1987 due to: (i) civil unrest, with farmers' lives being threatened if they paid water charges; (ii) difficulties in taking legal action against defaulters; and (iii) political reasons. Government's allocation for O&M of major irrigation systems was well below requirements. The study financed by the project on procedures for collection of O&M fees, completed in October 1989, emphasized mobilizing user support for joint management of irrigation systems, rather than collecting fees. As a result, GOSL adopted a policy of participatory management in irrigation schemes. This was in line with the INMAS concept and a major force behind fanners officially taking over responsibility for O&M of distributaries and field channels of rehabilitated schemes. 5.9 Although FOs in the project area are now responsible for O&M of distributary and field channels, they continue to rely on GOSL. ID continues to allocate 30% of the amount it receives for O&M to the FOs (about Rs 55/ac in 1993). Some FOs impose a charge of half a bushel of paddy (about Rs 80 at the 1993 price) per season for system O&M but, as they have no means of enforcing payment, in many cases little is collected. The only assured contribution towards maintenance is usually through shramadana (voluntary labor), and this tends to be more effective with respect to field channels than distributaries. As a result, O&M continues to be inadequate. 5.10 Irrigation Department. Rehabilitation works were carried out as envisaged in the SAR. Three types of arrangements were used: force account, private contractors and, in the latter stage of implementation, FOs undertaking contract work (on distributaries and field channels). Private contractors registered with ID were selected on the basis of rates for different work items established by ID. All three types of work execution have worked well. The quality of construction varies from scheme to scheme, but is good, in general. The use of FOs has been an innovative concept. The contracts have enabled FOs to obtain substantial income as compared to their other sources such as membership fees, fertilizer sales, etc. 5.11 Canals have been resectioned to ensure the carrying capacity required for regulation of supplies. In addition, retaining walls have been built along main and distributary canals where required. Additional regulation structures have been built in some places. A number of farm turnouts and drop structures have been constructed in various canals. These structures have improved water distribution to the various lower-level channels. 5.12 Equipment provided through CIDA and IDA financing has been put to good use. There has been good continuity of operating staff on the heavy equipment, and the machinery has been well looked after. At the end of the implementation period, residual life is quite substantial. Computers were purchased under the project, but only at Kantalai operational programs are being used to set water delivery rates for crops actually grown. 5.13 As mentioned above, adequate O&M funds have not been available so far. Estimated requirements for 1993 and actual allocation are given in Annex 3. Substantial increase in funding is necessary which requires that appropriate water charges are collected. GOSL agreed to levy water charges,  6 but in 1991 this covenant was amended and the responsibility for maintenance of distributaries and field channels transferred to FOs. However, ID still contributes funds to FOs to maintain the distributaries, while field channels are maintained by the farmers. Problem-free operation of the systems in a sustained manner is yet to be achieved. 5.14 Farmer organizations taking over responsibility for part of the irrigation system is an innovative concept that must be supported. The process is bound to be slow, but without farmer participation sustained successful operation of the schemes would be in jeopardy. The PCR mission met with farmer representatives in the four schemes and they were enthusiastic about the concept. 5.15 Field channels are operated by irrigators employed by FOs. They are responsible for rotating the water supply to individual fields. At the inlets to the fields, simple wood boards have been provided to be used as on/off vertical gates, but the PCR mission noted that these boards were not being used in the areas visited. The traditional way of digging up some earth from the channel and blocking the entrance to the field channel, and thus destroying the hydraulic regime condition of the parent channel, was being used. The removal of earth near the turnout has resulted in some undercutting of the embankment and water was seen flowing around the turnouts as well. This must be avoided. The pilot program for technological improvements in irrigation water control has not proved effective, largely due to lack of farmer cooperation. The PCR mission saw large stones placed on downstream control mechanisms, which then released more water than intended to the field channels. The FO representative was unable to explain why such vandalism was taking place. 5.16 Land Commissioner's Department. Physical targets for both gravel-topped and unpaved roads in the four remaining schemes were achieved, although there were some modifications to the planned distribution of roads, with a greater number of unpaved roads being rehabilitated in Huruluwewa and Rajangana and fewer in Nachchaduwa and Kantalai. Much of the rehabilitation work was contracted out to FOs who worked under the supervision of departmental technical officers. The roads were very much appreciated by farmers as they have facilitated access to markets. 5.17 Significant progress was made on regularization of encroachment with over 14,700 lots regularized in the course of the project, almost 60% of which were in Rajangana. These regularizations covered both holdings already in the Block Out Plan (BOP), where farmers required Government permission to transfer their holdings, and actual encroachment on areas of land both in the Highlands and in irrigated areas which had not been part of the BOP. 5.18 Department of Agriculture. Studies on both catchment management and on-farm water management were carried out in Hurluwewa and Nachchaduwa schemes and are continuing under NIRP. Both regression and simulation models were developed to predict run-off in micro-catchments. Although the simulation model appeared to give better results, further studies were recommended to improve accuracy. Some interim findings, such as that the capacity of minor tanks situated within the catchment area of major tanks should not be increased, were self-evident. Field work for the on-farm water management study was curtailed due to the security situation and results are of limited use. 5.19 Although the SAR (para 6.09) envisaged that the assessment of agro-economic impact would be the responsibility of DA, assisted by the University of Peradeniya, neither undertook this job. The mid- term socio-economic evaluation survey to be carried out by the University was cancelled since sufficient information had been gathered under a continuing project monitoring program separately funded by SDC, and the two studies of INMAS impact and collection of O&M fees.  7 5.20 Two unexpected events that took place during implementation which adversely affected project outputs were the removal of the village-level extension officers and the phased withdrawal of the fertilizer subsidy. At present, one Agricultural Officer (diploma holder) has to serve more than 4,000 farm families. The fertilizer price has more than doubled, leading to lower rates of application. Owing to poor extension services and very limited association of DA with the project: (i) diversification from paddy to other field crops (OFCs) in yala (April to September) did not take place to the desired extent; (ii) cultivation of high value crops (bananas, for example) was rare; and (iii) seed production by farmers and lateral spread was at very low levels. 5.21 Department of Agrarian Services. Under the project, two Agricultural Service Centers were rehabilitated, and five garages, 15 fertilizer stores and three wells constructed. Nine of the fertilizer stores have been leased to FOs, who have also taken over other fertilizer stores in line with GOSL policy of privatizing fertilizer supply (now handled by the private sector). Many of the FOs under the project have taken up fertilizer distribution as a service to members and earn a small margin on sales. Another important activity of DAS under the project has been the registration of FOs. Registration under Section 56A of the Agrarian Services Act enables FOs to enter contracts, while registration under Section 56B gives them legal status and permits them to borrow. By the end of the implementation period, 129 of the 146 Distributary Channel Organizations (DCOs) had been registered under Section 56A and 51 also under 56B. 5.22 University of Peradeniya. The planned mid-term review was not considered necessary. At the request of IMD, the final evaluation was carried out in 1992 by a private consultancy firm rather than the university. The comprehensive report indicated a positive project impact with respect to technical and institutional aspects but highlighted weaknesses in the impact on agricultural production and income. The report was generally well received but its timing was felt to be premature as implementation was barely completed. 5.23 International Irrigation Management Institute. The contract for construction of the IIMI headquarters building at Kandy was not awarded until May 1988, and the contractor abandoned work at an early stage due to the security situation. IIMI subsequently decided to move to Colombo and, with the approval of GOSL, purchased a recently completed building. The project supported the acquisition, valued at about Rs 73 million, with funding from CIDA (10%) and SDC (65%). Cost overruns were significant. There was no direct contribution from IIMI to the project as the expected IIMI field site was not established for security reasons. 5.24 Training. Intensive training was carried out locally by IMD to familiarize officials with INMAS and participatory management. The Galgamuwa Irrigation Training Institute developed modules for training of officials in participatory management and farmers in the use of irrigation structures, financial management, contract management and O&M of irrigation systems. A large number of farmers and farmer representatives were trained in these subjects as well as general agriculture. Thirty-two farmers went on study tours to Thailand - large scale commnercial agriculture and different irrigation techniques - and twenty-three IMD staff attended overseas courses on water management, M&E and training methods. ID staff received training both locally and overseas to enhance their technical capabilities and improve management, especially participatory management. Details are in Part 111/4. The PCR mission was pleased to note that the engineers who had obtained their M.Sc degree under the project were actually working on project schemes contributing their knowledge. However, there were shortcomings in other areas of training as noted below.  8 5.25 Although the concept of participatory management is now better understood, successful application is likely to require further training. The considerable exposure of ID staff to water management training has not resulted in any significant water savings. In spite of training in the application of the hydrological models developed under the project, the models are not used, except at Kantalai. Farmers will continue to require training in O&M of irrigation systems, particularly in the role they are expected to play. The value of the overseas study tours for farmers is limited, as farmers were generally not in a position to apply any of the techniques seen. 6. Project Results 6.1 The project's objectives have been partially achieved. Rehabilitation works were carried out in the four schemes remaining in the project. Water is now being delivered to tail-end farmers more quickly and on time. However, utilization of available water is by no means optimum. Integrated management of the major schemes has been accepted in principle and Project Management Committees (PMCs) have been established and are operational. While the irrigation schemes are operating more effectively and there is greater equity in water distribution, maintenance remains inadequate. A major problem is inadequate funding of O&M. The PMCs have been instrumental in improving the programming of agricultural activities, and major decisions are made at the PMC prior to the official pre-season cultivation (kanna) meeting. 6.2 Agricultural Production. It is difficult to estimate the impact of the project on agricultural production from the data on yields and cropping intensities recorded by various agencies. There are widely varying estimates for paddy yields, and even areas, cultivated under the same scheme. Data for other field crops (OFCs) are practically non-existent, but the PCR mission found that a number of OFCs are cultivated, rather than only soybean considered at appraisal. 6.3 On the basis of available information and the assumptions detailed in Annex 1, incremental paddy production at full development is estimated to be 32,400 tons compared to 37,300 tons estimated for the four schemes in the SAR. Incremental OFC production has been estimated to be about 13,700 tons against about 6,200 tons of soybean assumed for the four schemes at appraisal. Cropped areas for OFCs are actually lower than assumed at appraisal, except for Kantalai. The higher production estimate is mainly a result of the mix of crops assumed, which includes eggplant, mungbean, cowpeas and chillies as well as soybean. 6.4 Although the SAR was conservative about development of OFCs, cultivation has lagged behind SAR projections in terms of overall area. Many farmers persist in cultivating paddy whenever there is water in spite of the better income offered by other crops. Part of the problem is lack of extension to provide the necessary techniques to successfully cultivate new crops; other factors are marketing difficulties and inadequate availability of credit. 6.5 Water is now being delivered in an equitable and timely manner enabling farmers to start cultivation at the proper time. Cultivation was often delayed prior to rehabilitation leading to high pest attack, poor growth and lower yields. Social tensions have been significantly reduced with the equitable distribution of water. 6.6 Institutional Development. An important achievement of the project has been the establishment of FOs. Training in water management and aspects of group development has been provided to farmers' representatives, and FOs have taken over O&M of distributary and field channels under a  9 system of participatory management. Farmers, through the FOs, significantly influence decision making on water issues and cultivation calendar. Local institutions have learnt the importance of participatory management and are seeing farmers as equal partners in maintaining systems, organizational activities and agricultural development. FOs have also become involved in fertilizer distribution and have achieved recognition as representative bodies at the grass-roots level. Unfortunately, in spite of considerable progress, farmers continue to rely on GOSL for O&M funding. FOs have not yet become financially viable, independent organizations and are not able to collect necessary funds. In most cases little attempt is made to enact such collection as farmers expect, and ID provides, government resources for O&M. 6.7 Benefits. Project impact on farm income is positive, with a clear advantage for those farmers who cultivate some OFCs in the yala season. Models developed to re-estimate the financial impact of the project (Annex 1) are not directly comparable with those used in the SAR as two of the SAR models were for areas subsequently dropped from the project. In proportional terms, the increases in net farm incomes estimated in the PCR (between 13% and 76%), are lower than those estimated in the SAR (from 36% to 157%). This is due to relatively lower paddy prices and higher prices for inputs such as fertilizers where the subsidy has been removed. Incremental employment generated by the project, at 1 million workdays/year also appears to be significantly lower than the SAR figure of 4.1 million, but no scheme- wise breakdown is available for the latter figure, which includes the schemes dropped from the project. Rural roads have been repaired allowing farm produce to be transported to markets by tractors and lorries. 6.8 Economic Impact. Assumptions for the economic re-evaluation of the project are detailed in Annex 1. The recalculated economic rate of return (ERR) is 19%, compared with 16% estimated at appraisal. Individual scheme ERRs are higher than in the SAR for Nachchaduwa and Rajangana and lower for Huruluwewa and Kantalai. The ERR for the Huruluwewa Feeder Canal is 10%. Lift irrigation is included in the analysis of the Rajangana scheme as separate investment figures for lift irrigation were not available. The resulting ERR is 21 % compared with 18.4% for the gravity scheme in the SAR. 6.9 Sensitivity tests of the re-calculated ERR show that it is fairly robust. A 20% decrease in future benefits would bring about a 3 % drop in the ERR. The base case assumes realistic levels of O&M funding (Annex 1), which are far below desirable levels and would probably prove to be inadequate to maintain the schemes for a full 25-year life at current levels of efficiency in water delivery. With a reduced project life of 15 years (reflecting poor maintenance) the ERR is 15%, still well above the opportunity cost of capital, assumed to be 10% for Sri Lanka. 7. Project Sustainability 7.1 The major issue affecting project sustainability is the adequacy of scheme O&M. This is closely linked to the strength of FOs and the application of participatory management. 7.2 Government allocations for O&M of major irrigation schemes have been inadequate in the past (Annex 3) and are likely to become more so in the future. GOSL has renounced the attempt to recover O&M costs through water charges in favor of participatory management in which FOs are responsible - financially and operationally - for O&M of distributary and field channels, with ID engineers providing technical advice as required. Eventually, farmers would also be expected to make a contribution to the cost of O&M of main channels and headworks. 7.3 Although a significant start has been made in transferring to farmers the responsibility for the irrigation networks they use, the transition from recipients of a service provided by Government to  10 owner/operators of the system is in its preliminary stages. Farmers, with some justification, believe they cannot afford to finance O&M of the irrigation schemes they use. Where double cropping of paddy prevails, this is likely to continue to be the case. Substantial efforts are required to bring about crop diversification, to improve both farmers' income and use of water, which remains a scarce resource. High value crops must be grown where feasible. With appropriate agricultural extension advice and credit, farmers can be expected to respond to market conditions and grow cash crops other than paddy in the yala season in well-drained and medium-drained lands. Once farmers have more cash at their disposal, they will be in a better position to contribute towards O&M. At this stage, additional legal authority for FOs to exact contributions, envisaged in proposed legislation, will be helpful. 7.4 Farmer organizations themselves require further strengthening in order to become financially viable and autonomous. While IOs (para 5.7) may be of help, their services should not become permanent; rather, assistance should be provided to FOs for specific purposes on a limited basis. Once the organizations are established, monitoring and control become important. Although such a role is envisaged for the Commissioner of Agrarian Services, who has powers to audit and supervise FOs registered under the Agrarian Services Act, the mushrooming of FOs, together with the very limited resources at the disposal of DAS, cast serious doubt on its capacity to carry out such regulation. 7.5 In the context of participatory management, it is necessary to ensure continuity of cooperation and bring about further changes in the attitudes of all parties, especially farmers and ID staff. The farmers have to realize that GOSL can neither continue to contribute to O&M of farmer-managed channels nor carry out O&M of dam and main canals unless it fully recovers these charges from farmers. ID staff have to work more closely with farmers in planning the areas under different crops (commensurate with expected water availability) and to achieve optimum water use, as well as to provide the technical advice for good maintenance. 7.6 System maintenance requires both technical inputs from ID and labor and cash inputs from beneficiary farmers. Maintenance of small canals should be carried out with due regard to the hydraulic regime conditions. Field channels carry very small discharges. If appropriate cross-section and bed slope are not maintained, the energy gradient reduces and the water level drops, as was noted by the PCR mission in one field channel at Nachchaduwa. In this case, the farmer at the tail-end was complaining that water was not entering his field. At the request of the PCR mission, ID staff surveyed the field channel, cleared it properly with the assistance of the farmer, and restored it to design slope; and water entered the farmer's field. If this type of technical input is neglected, problems, complaints and eventually mistrust of system performance by farmers result. While the actual clearing of waterways can and should be carried out by farmers, sufficient technical inputs such as levels at key points, bed slope maintenance, should be provided to keep the system operational. In order to do this, GOSL needs to obtain and allocate sufficient O&M funds. 7.7 The irrigation schemes would have to be operated efficiently to maximize water use and achieve water savings. In the maha season (October to March), paddy is grown and the water input is through both rainfall and irrigation supplies from the tanks. Flexibility in the operation of the canal system is required to use the rainfall in the most effective manner. In the yala season, the canal system will have to be operated to cater to the needs of paddy as well as OFCs whose irrigation regime is quite different. Computer-aided models such as the one being used in Kantalai are essential to determine the irrigation releases and to operate the canal network. ID staff responsible for the operation of the system need to be trained further in such operation patterns. With optimal use of available water, area under cash crops in yala can be expanded.  11 7.8 Another sustainability problem arises in the context of the roads rehabilitated by LCD. The Department does not have the facilities to maintain roads. It appears possible that the local Highways Commission will maintain the paved Grade 1 roads, but the issue of maintenance of the over 200 km of unpaved roads remains unresolved. Maintenance of rehabilitated roads is a must so that perishable high- value crops can be transported to market. 7.9 Encroachments must not be permitted on drainage reservations; problems of salinity will develop if drainage is impeded, particularly with increased irrigation intensities. 8. Bank Performance 8.1 Bank performance during implementation was satisfactory. Fifteen supervision missions were fielded with a remarkable degree of staff continuity. Due to the fact that the same engineer appraised the project and provided the bulk of supervision input, supervision of engineering works was particularly effective. It would have been desirable, however, for an agronomist to have participated in some of the supervision missions to further emphasize agricultural development aspects. 8.2 Lessons of experience that may be relevant to other Bank-financed projects are: - in a project whose aim is to improve agricultural production, more attention should be given to this aspect; - in order to enhance farm income, more attention should be paid to optimum use of water and O&M of irrigation systems; - establishment of farmer organizations should take account of the need for continuing supervision and regulation beyond the implementation period; - adequate O&M is necessary for sustained project performance; a functioning system based on participatory management providing the necessary means is essential to achieve this goal, and covenants in this regard should be strictly adhered to. 9. Borrower Performance 9.1 Despite the difficult circumstances of civil strife which affected implementation, the Borrower performed well. Although the relationships between the two major implementing agencies, IMD and ID, were somewhat strained particularly in the early years, coordination was effective and the project helped to establish the concept of participatory management of major irrigation schemes. Continuity of senior staff was good with only two Project Directors. There were some delays in submission of audit reports, but project accounts were generally well kept and the system established by project consultants proved to be efficient. Project monitoring benefited substantially from the program funded outside the project by SDC. 9.2 A major shortcoming in Borrower performance was the inability to improve collection of water charges. Eventually GOSL chose to promote farmer organizations and participatory management as a means of ensuring adequate O&M of the irrigation systems. Another weakness was the inability of ID to introduce improved computer aided water management.  12 9.3 Lessons of experience for the Borrower are: - promotion of farmer organizations and participatory management of irrigation systems are only first steps in the process of bringing about adequate O&M; - in order to enable farmers to contribute to O&M, attention needs to be given to improving their income through crop diversification; - increased attention needs to be given to raising awareness of the importance of water management at all levels, particularly amongst staff who work closely with farmers; - O&M of irrigation systems cannot be carried out by FOs alone; it requires technical input and advice from ID; - farmer organizations undertaking construction activities should be protected against major price fluctuations in materials such as cement (price escalation clause) to enable them to carry out the work without delay. 10. Project Relationship 10.1 The relationship between the Borrower, IDA and the co-financiers, CIDA and SDC, was satisfactory throughout. The Borrower particularly appreciated the funding agencies' responsiveness to requests for reallocation of resources and even additional funding. SDC regularly participated in supervision missions and supported a complementary project which provided continuous monitoring input to this project. CIDA representatives participated in several supervision missions and regularly attended wrap-up meetings where supervision missions' findings were discussed with Government. 11. Consulting Services and Contracts 11.1 A system management consultant advised the project on rehabilitation designs and irrigation scheduling in the early years of implementation. Designs were adequate, but the consultant did not manage to instill an appreciation of the critical importance of water management in ID engineers. Local consultants established and initially operated the project's accounting system. The system has functioned well but, owing to turnover of project officers, problems were experienced in training counterparts, and the consultancy had to be extended by two years. Another local consultancy firm carried out hydrological surveys and developed computer models for monitoring water supply. Unfortunately, the models have not been calibrated with actual data and are not used, except at Kantalai. An expatriate consultant provided technical input for the pilot program in the rehabilitated areas including specifications for water control equipment. Studies by local consultants on INMAS, O&M procedures, project impact, and agricultural credit were generally well received (Part III/6D). 11.2 A substantial number of contracts for earth works (both rehabilitation of irrigation structures and roads) were awarded to FOs and were an important source of income for the fledgling organizations. Work was generally satisfactory.  13 12. Project Documentation and Data 12.1 The SAR and DCA provided the initial framework for project implementation but significant modifications to the project reduced the usefulness of targets. The initial socio-economic survey carried out by the University of Peradeniya did not prove replicable as the sample farmers could not be located. Since sufficient information had been collected through other efforts, it was decided to forgo the mid-term evaluation. Although the final evaluation carried out by local consultants was too early to assess project impact, it did provide a significant amount of information which was useful in PCR preparation. Both the evaluation consultants and the PCR mission suffered from the lack of reliable agronomic data. Other project data, particularly accounts and information on scheme rehabilitation, are good.   15 PART II I. Observations on Part I of the Project Completion Report 1.1 Part II of the Report was prepared by the Borrower in Sept./Oct. 1994, after an FAO/World Bank Cooperative Programme Mission prepared Parts I & III and made them available to the Borrower in August 1994. Borrower's observations and comments in Part II are based on and relate to the findings of the above Mission as given in Parts I & HI. Performance of Implementing Agencies A. Irrigation Management Division 1.2 The responsibility of implementing the Major Irrigation Rehabilitation Project was assigned to the Irrigation Management Division formed under the ministry of Lands & Land Development to implement the program for Integrated Management of Agricultural Settlements (INMAS). The location of the division in the ministry rendered the coordination of the implementation activities through the line departments much easier. A few setbacks were initially experienced, due to insufficient comprehension of the concepts both by officials of line agencies and farmers; the latter were very conservative in their responses especially in view of collecting funds for O&M. However, changes in the attitudes of all concerned were effected through discussions, seminars, workshops and demonstrations. The few remaining issues could be resolved through close monitoring of INMAS activities by an inter-ministerial coordination committee and further improving the relations between beneficiaries and field officers of line agencies. 1.3 The catalytic role of the institutional organizer (IO) had been recognised by the beneficiaries. Some of the beneficiaries have responded positively to the changes introduced through 1Os; the responses of some were minimal and yet of a few negative. The attitudinal changes in rural farmers, who are traditionally conservative, is a time-consuming task which requires dedication and commitment. As the benefits from institutional development are being realised gradually and increasingly the farmers appreciate the services of 1Os. However, payment for continued employment of IOs by the Farmer Organisations will be a strain on their present low incomes. Hence continued engagement of a few IOs is needed to further stabilise FOs. 1.4 The collection of O&M fees proved to be a failure due to socio-political and legal issues experienced in the past as well as due to the low income level of the farmers. The policy of participatory management may prove to be a successful alternative for the collection of fees from the farmers. However, this has to be tested over time beyond the project period. 1.5 Suggestions for Future Rehabilitation Projects - (i) The development of farmer organisations for self-management is a time-consuming process that should be pursued with dedication to achieve the ultimate goals as a long-term strategy. The institutional development component should therefore be initiated at least six months prior to commencement of the civil works or infrastructure components of those parts of the system that affect or impact on farmers directly. (ii) The period of above development depends on the commitment and dedication of the implementors, size of the system, socio-economic complexities and political environment.  16 (iii) The strength of a farmer organisation depends on the degree of participation of the beneficiaries, building of confidence on self-management and financial status of the organisation. Any activity that hampers the above could generate negative effects. (iv) Programmes for strengthening the finances of farmer organisations should be pursued vigorously, especially in activities involving project implementation, with emphasis on transparency in the implementation of those activities involving investment both by agencies and farmers. (v) Consultation and education of the beneficiaries on new concepts, especially if financial commitments are involved, before implementation is necessary. (vi) Improvement of information and communication links within the project area and opening- up of new links for market oriented agriculture are extremely important. B. Irrigation Department 1.6 Method of Implementation - (i) The rehabilitation works in the four sub-projects, namely, Rajangana, Nachchaduwa, Huruluwewa and Kantale, were implemented through the Divisional Irrigation Engineers, under the supervision of the Range Deputy Directors in Anuradhapura and Trincomalee. The rehabilitation designs for irrigation systems emphasised (a) maintaining safety standards for the earthworks and structures, (b) ensuring adequate conveyance capacities in canals and turnout structures, (c) providing sufficient regulation for water releases and conveyance in the distribution system, and (d) provision of flow measurement devices at critical locations. The technical guidelines, type plans of standard structures and basic instructions for rehabilitation works were issued by the Head Office of Irrigation Department. (ii) The construction works were executed on (a) force account, (b) contracts awarded after following tender procedures, and (c) contracts awarded to approved societies after negotiations. In the initial stages of the Project, from 1985 to 1990, most of the construction work was done on small scale contracts awarded after calling for rate quotations. A smaller percentage of work was done on force account using departmental machinery. After farmer organisations were well established and registered, some of the works, especially those in field canals, were awarded to them on negotiated terms. (iii) The scheduled programmes of rehabilitation work in Rajangana, Nachchaduwa and Huruluwewa were completed by date of closure of the Project. In Kantale, rehabilitation works in the channel systems covering 83% of the command area were completed within the project period; the balance was the part of the work in Thampalagamum area where normalcy had not returned. 1.7 Problems and Constraints Encountered during Implementation - (i) The staff deployed in the divisional offices for rehabilitation works were inadequate and attention was focussed mainly on construction activities to achieve scheduled Project construction targets. Attention to operations was inadequate, especially the deployment of staff with computer aptitude and ability. The use of the computer models developed for water management was restricted to locations with interested personnel. (ii) The concept of controlling the farm turnout by wooden gates did not meet with much success in some locations due to actions of mischief makers, but the alternative of providing lockable metal gates would be rather expensive. The only other possibility appears to be to instil a sense of responsability on the importance of water management in all beneficiaries. (iii) Initial disagreement of farmers on the design of a few of the structures was subsequently resolved by consensus; the farmers disliked the rectangular farm turnouts and weir box measurement devices, which were replaced by smaller diameter outlets and gauge posts. Farmer-Officer dialogue and understanding are required in all similar issues.  17 (iv) Shortages of cement experienced during working seasons resulted in a considerable amount of delay. 1.8 Suggestions for Future Rehabilitation Projects - (i) Consulting the farmers at the design stage and finalising the proposals with their concurrence. The irrigation officials to initiate education of farmers on less favoured technical solutions. (ii) Attention on the development of on-farm water management to complement irrigation system improvements to optimise water use. (iii) The irrigation infrastructure in most major and medium schemes in Sri Lanka includes main, distributary and field canals. Attention to imnprovement of field canals and the respective roads, for better water management and transport of produce. (iv) The limited experience gathered in the use of computer models and automatic downstream control gates, introduced under the Project, has aroused the interest of many who were used to traditional operational practices. Currently many computer models are under development or in stages of testing. Introduction of similar innovative water conservation and management techniques to enhance the efforts of ID for optimising distribution and use of available water. (v) The approach of the Project for integrated rehabilitation of irrigation systems afforded an opportunity for the beneficiaries and the officials of implementing agencies to practice and test participatory management approaches successfully. The Project Impact Evaluation Study (1992) reported positive trends in the participation of the beneficiaries in the management of the schemes, and greater understanding between farmers and officials undertaking construction work and the implementation of operations and maintenance programmes. This approach could be replicated in future projects with emphasis on close links between farmers and officials of the line agencies. (vi) Monitoring water distribution and on-farm water usage in the irrigation schemes, especially at farm level, had been relatively weak. Project focus on educational programmes to overcome this weakness is necessary. (vii) The Major Irrigation Rehabilitation Project could be considered a landmark in improving the performance in some of the older and larger irrigation schemes in the Island. Need for having consistent approach in project monitoring from the very inception itself has been well demonstrated from this project. C. Land Commissioner's Department 1.9 The work components implemented were regularising encroachments of irrigable land within the command area and rehabilitation of settlement roads. 1.1O Encroachment of vacant land within the command area of all schemes is a national problem arising from the increasing unemployment and landlessness among the second and third generations of the settlers. This imposes considerable hardships on system management, especially the water distribution, due to lack of knowledge of the extent of the cultivated area, overuse of irrigation water in the head reaches of the distributary system, depriving legitimate cultivators of their fair share of water, and inadequate cooperation for operation and maintenance of the irrigation system. Regularisation of encroachments had proved to be effective in resolving most of the above problems and in addition has enhanced participatory management. Three categories of encroachment had been identified, namely, (a) encroachment of adjoining reservation by permit holders, (b) encroachment of land reserved for common purposes, (c) encroachment of unalienated lands either abandoned or illegally disposed of by the permit holders.  18 1. 11 Improvements to class I and class II agricultural roads in the schemes selected for rehabilitation were undertaken to improve access for agricultural inputs and markets for agricultural produce. In addition this facilitated improved communications. Initially the construction works were executed through small scale contracts awarded after adopting of tender procedures. Towards mid Project implementation period, when the farmer organisations were more stabilised, work was executed on negotiated contracts with the above organisations. This afforded a source of income to the organisations and development of skills of farmers in construction and management. Farmers were thus directly involved in quality control of work in their areas. 1.12 Suggestions for Future Rehabilitation Projects - (i) Promote development of industries in rural areas with focus on use of available resources in the locality to absorb excess labour and generate additional income to uplift the living standards. (ii) Handing over the rehabilitated roads to local authorities and encouraging them to undertake maintenance arrangements. D. Agriculture Department 1.13 The Project identified the inputs from DOA to be a research study on catchment management by the Land Use Division, and agriculture extension services through the "Training & Visit" programme that provided coverage of the whole island, during the initial period of the Project and funded by IDA. Thus Project support was primarily to the research study only. 1.14 The research study on watershed management in two catchments, namely, Nachchaduwa and Huruluwewa commenced in 1986. The research study has diagnosed some of the problems with respect to water management in the watershed and command areas. The fmdings of the command area study had limited results through communication with farmer organisations; however, a greater impact could have been achieved if incorporated and linked with extension programmes. The findings of the catchment management study have proved to be useful in planning catchment programmes in the future, and the simulation model developed could be a useful tool in further developments for predicting runoff in micro and macro catchments. 1. 15 The importance of direct involvement of DOA in agriculture extension in the Project was felt from the initial stages and limited financial resources were provided from Yala season in 1986 to conduct agriculture demonstrations and residential training courses for the farmers in Nachchaduwa, Rajangana and Huruluwewa schemes. The assistance of the Agriculture Department was extended to IMD to conduct training programmes and demonstration to farmers on use of agro-chemicals and crop diversification programmes. But this proved to be not of much effect as there was no follow-up due to the disruption of the extension services of the DOA. 1.16 The regular agriculture extension programme of DOA was adversely affected by the following unfavourable circumstances: (i) Withdrawal of Krushi Viyaptha Seva Niladhari (field extension officer) from the extension service. (ii) As a consequence of (a) above, the next higher field officer, Al, was burdened with providing services to more than 4000 farm families while attending to his other departmental obligations. (iii) The devolution of extension functions to provinces relieved DOA of Als who were directly attached to Divisional Secretaries and the District Training Centres. (vi) Civil disturbances which prevailed in the country from 1987 to 1991.  19 (v) Inaccessability of Kantale, Morawewa, Giants Tank, and Iranamadu due to subversive activities. 1.17 Suggestions for Future Projects - The ultimate expectation of an irrigated agriculture project will be increased production with crop diversification for better economic returns. All such projects should embody a component for a concentrated effort to achieve the production targets independent of other ongoing programmes and complementary studies for further improvements. E. Agrarian Services Department 1.18 A major portion of the agriculture input services in the country were provided by DAS. A special effort was made under the Project to strengthen the input facilities in the sub-project areas, for timely supply of requirements, by providing financial assistance to rehabilitate two Agrarian Services Centres and construct five garages, fifteen fertiliser stores and three wells. Nine of the above fertiliser stores have been leased to FOs in line with GOSL policy of encouraging self-managed sustainable FOs. Many of the LOs, especially those with stores, provide inputs of agro-chemicals, fertiliser and seed on a small marginal profit to their members. The under-expenditure of Rs. 6.5 M of the total provision corresponds to the work items in Morawewa, Giants Tank and Iranamadu which were suspended due to the adverse security situation in these areas. 1.19 DAS supported the legal recognition of FOs with the objective of promoting self-management. All formal farmer organisations in the four sub-projects were registered, initially, under Section 56A of the Agrarian Services Act, enabling them to undertake contracts, and subsequently under Section 56B giving them the body corporate status for financial transactions. 1.20 Proposals for Future Projects - Provide concentrated effort to improve input services and facilities at competitive prices to regulate fair prices in the open market. F. International Irrigation Management Institute (IIMI) - Para 5.23 of Part I 1.21 The cost overruns in the procurement of the building were mainly due to location in the capital and much greater floor area than that originally planned. IIMI on its own has provided the services of a French engineer to install automatic control gates in Rajangana and Huruluwewa. Other attempts to procure services of IIMI proved to be very expensive and far beyond the rates in the open market. II. Changes in Project Composition 2.1 Deletion of some Sub-projects - Project implementation commenced in August 1985 and the closing date was extended from 30 June 1991 to 30 June 1993 on the recommendations of the Review Missions of the donors, with due consideration to the setbacks experienced from 1986 to 1990 resulting from the breach of Kantale dam, civil disturbances (1988), and the additional work undertaken, namely, the improvement of Lift Irrigation System in Rajangana, and regularising the encroachments along Huruluwewa feeder to control illicit water extraction. The works in Morawewa, Giants Tank and Iranamadu were suspended in 1986 and subsequently deleted from the Project in January 1991 due to deteriorated security conditions in these sub-project areas. The implementation programmes in Rajangana, Nachchaduwa, Huruluwewa and Kantale were completed on 30th June 1993.  20 2.2 Additional Works - (i) Repairs to breach in Kantale Tank bund - This was a significant addition involving breach closure and downstream flood damage repairs. Considerable cost overruns were experienced due to increase in scope of work to ensure safety of head works, which included additionally closure of abandoned old LB sluice, providing additional rip-rap protection, and desilting and repairs to the irrigation system after the floods. (ii) Rehabilitation of Rajangana lift irrigation scheme - Compliance with the covenant in para 2, Schedule 4, DCA was delayed due to poor status of the physical system and legal constraints. The Project supported the supply of 72 pumps, under CIDA grant component, and cost of construction materials and transport for civil works. The farmer organisations undertook the rehabilitation work and contributed 35 % of the value of work as voluntary labour. The system was handed over to farmer organisations for custodianship and O&M under an Agreement between the organisations and ID. (iii) Regularising the unauthorised extraction of water from Huruluwewa feeder canal - 39 outlets, from feeder canal, and an independent canal system were constructed to restrict the uncontrollable conveyance loss and expansion of illegal agriculture land. 2.3 Financial Adjustments - At the close of the Project the expenditure incurred was Rs. 1329 M (US$38.7 M). The excess in local currency (Rs. 123 M) and saving in foreign currency (US$4.5 M) are due to wide variations in the US Dollar parity rate (Rs. 25.5 per US$1.0 at appraisal to Rs. 48.9 per US$1.0 at Project closure). The saving from the deleted sub-projects was utilised in additional works (para 2.2 above) and for changes in parity rates. III. Training (Para 5.24 of Part 1) 3.1 The achievements of farmer organisations in self-management and the appreciable degree of success in participatory management reflect undoubtably the effects of training. However, it is agreed that training should be continued to improve the effort to achieve optimum water use and agricultural production which would enable the successful application of participatory management on a sustainable basis. 3.2 With the disruption of the agricultural extension services, there was no follow-up guidance for farmers with overseas training to apply what they learned and saw abroad. It is hoped to remedy this situation in the future. IV. Bank Performance 4.1 The guidance provided by the Bank through continued services of a project officer, knowledgeable in irrigation sector, enhanced the efforts of the management in directing the Project to achieve the targets. Bi-annual review missions reinforced the efforts in monitoring, evaluating and initiating corrective measures to maintain the scheduled progress. The benefits in achieving the project objectives could have been increased by including an agronomist in the regular review missions and an economist intermittently. The services provided from appraisal stage to project closure were very good and consensus was reached on many field issues through discussions. The Borrower recommends that the following need consideration in future Bank-financed projects. (i) In irrigation-cum-agricultural projects rehabilitation should be conceived of as a comprehensive package with institutional building included as a mutually supportive complementary, or supplementary programme to the physical rehabilitation effort. Institutional building has to be treated as a long-term strategy with continuous training, supervision and regulation beyond the project implementation period.  21 (ii) Farmer Organisations should be helped by way of investment capital on a loan basis or through a revolving fund for them to invest on income-generating activities and programmes for the benefit of farmer-members including even off-farm economic activities. (iii) Particularly in agricultural projects there should be adequate emphasis on agricultural extension and improving agricultural production and farm incomes. (iv) Full awareness of the political realities and the income levels of farmers in imposing cost- recovery O&M charges is a prerequisite for designing projects of this nature. V. Borrower Performance (Paras 5.6 and 9.1 of Part I) 5.1 In regard to projects needing consultation, collaboration and coordination among different disciplines (or professions) and agencies, such coordination mechanisms are vitally significant for the smooth implementation as well as performance achievement. When the implementing agencies are Departments it is always appropriate that the coordinating agency should be a ministerial arm to facilitate inter-ministerial and inter-departmental coordination. Assigning this responsibility to one single agency or department with equal administrative status with other implementing agencies is likely to hamper project performance and achievements. 5.2 The advantages of better communication facilities enjoyed by IMD, by virtue of its location in the ministry, were used to coordinate the project activities of all implementing agencies. This arrangement reduced the bureaucratic delays and enhanced communication and decision making process and would be an example for similar projects in the future. The situation is likely to improve with the strengthening and legalising of the Project Management Committees. 5.3 The delays experienced in furnishing audited statements of accounts were due to inadequate experienced staff and turnover of staff deployed for accounting in implementing agencies. 5.4 The inability to improve collection of water charges was due to the political realities of the civil disturbances and certain legal snags inherent in the law as well as the low income levels of the farmers. 5.5 Lessons of Experience for the Borrower - (i) Project Management Committees to be strengthened and linked up to higher authority levels for their decisions to be implemented effectively and promptly. (ii) Farmer Organisations must be helped by way of investment capital to undertake activities and programmes to improve income levels of farmers to encourage them to contribute to O&M. (iii) Establish and develop farmer organisations in the early stages or well before the project to enhance participatory management and build confidence of self-management among the beneficiaries. (iv) Legislative support to farmer organisations to resolve irrigation related matters quickly. (v) Farmer participation alone by itself would not result in improved O&M and system performance. It should be accompanied by technical input and advice from Technical Staff. (vi) Design and establish the systems for Project accounting and performance monitoring at the inception. (vii) Annual reconciliation of disbursements and the balances in the accounts with the donors must be strictly adhered to.  22 VI. Performance of Co-financiers 6.1 Swiss Development Cooperation (SDC) supported the civil works and training components in the Project. The interest shown on institutional development, training, and monitoring, evaluating & feedback of project performance enabled the Project management and IMD to promote vigorously the programme for self-management of farmer organisations. The assistance by SDC for construction of 17 fertiliser stores-cum-offices for DCOs increased the beneficiaries' confidence on self-managed farmer organisations. The representation of SDC in periodic review missions was regular. The officials of SDC were very helpful and cooperative during Project implementation. The relations with SDC- project officials were very cordial. 6.2 Continuous monitoring of the Project was conducted, independently, by AgriSwiss Rural Development Consultants for SDC. At the inception in 1987, the findings reported through SDC were much delayed for any effective corrective measures. This procedure was changed subsequently, to furnish copies of reports to management for reference, and finally in the last stages of the Project, to conduct monitoring, evaluation & feedback with the participation of the beneficiaries. In the final year of Project implementation, the role of the consultant transformed into assisting FOs in self-management activities, which duplicated some of the legitimate activities of IMD. The findings of the consultant, when availed timely, for adopting corrective action were complementary to participatory management and were appreciated. Independent changes in the mandate of the consultant and duplication of legitimate work of implementing agency may not, at times, produce the anticipated results. 6.3 Canadian International Development Agency (CIDA) supported the civil works construction, procurement of construction equipment & machinery, and training. The construction equipment was procured and supplied by CIDA in the initial stages of Project implementation, which enabled the ID to expedite completion of the quantum of rehabilitation work in spite of the setbacks experienced. The supply of 72 pumps for rehabilitation of Rajangana lift irrigation system is very much appreciated. This enabled the government to fulfil an outstanding Project covenant. The officials of CIDA were very helpful and cooperative during Project implementation. The relations with CIDA-project officials were very cordial. VII. Observations on Part III of the Project Completion Report 7.1 The assistance provided by International Development Association in the development of the irrigated agriculture sector and the implementation of the Major Irrigation Rehabilitation Project are very much appreciated. Reported disbursements, physical achievements, and financial status are acceptable. 7.2 Mid-term evaluation of Project performance was deleted as the reports of continuous monitoring by AgriSwiss Rural Development Consultants were available for review and corrective action. The findings of other studies which were completed too were used for corrective actions. 7.3 Implementation of the amended covenant 4.03 (c), DCA could be difficult due to socio- economic and political reasons. However, efforts to achieve this target by expanding the scope of participatory management over a period of time could be successful. 7.4 The details of Project achievements and pro-rata costs are given in annexes 1 and 2 respectively.  ..... ... .. .. . ............ . . ~~~~~~~~~~~~~~~~~~~~~~~~~~~... ........ ............................ ............... -m . . .................................................................................................... ....................................................... .......... ...... S ; ............................................. ............................................................. .............................. g ; g