Turkey - Drainage and On-Farm Development Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 15528 IMPLEMENTATION COMPLETION REPORT TURKEY DRAINAGE AND ON-FARM DEVELOPMENT PROJECT LOAN 2663-TU March 28, 1996 Agriculture and Environment Operations Division Country Department I Europe and Central Asia Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Turkish Lira (TL) End of Year Value of US$ 1.00 in TL 1986 758 1987 = 1.021 1988 = 1,815 1989 = 2,314 1990 = 2,930 1991 = 5,080 1992 = 8,564 1993 = 14,473 1994 = 38,726 June 1995 US$ 1.00 = TL 43,000 TL 1,000 = US$ 0.0233 TL 1,000,000 US$ 23.3 WEIGHTS AND MEASURES I kilogram (kg) = 2.20 pounds i tonne - 1,000 kilograms, 0.98 long ton I hectare (ha) = 2.47 acres I decare = 0.1 ha I square kilometre (km2) = 0.386 square mile FISCAL YEAR OF BORROWER Govemment of Turkey = January I to December 31 ABBREVIATIONS AND ACRONYMS DOFD - Drainage and On-Farm Development Project DSI - General Directorate of State Hydraulic Works GDAPD - General Directorate of Agricultural Production and Development GDRS - General Directorate of Rural Services GOT - Govemment of Turkey GIA - Gross Irrigable Area IAEE - lgdir-Aksu-Eregli-Ercis IMIR - Irrigation Management and Investment Review IMP - Irrigation Master Plan IR - Irrigation Ratio I&D - Irrigation and Drainage LRW - Land Reclamation Works NIA - Net Irrigable Area OFD - On-Farm Developments OFWM - On-Farm Water Management O&M - Operation and Maintenance SAR - Staff Appraisal Report SDW - Surface Drainage Works SSDW - Subsurface Drainage Works WUO - Water Users Organization  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT TURKEY DRAINAGE AND ON-FARM DEVELOPMENT PROJECT LOAN 2663-TU Table of Contents Preface Evaluation Summary . ......................................... i-v PART I: PROJECT IMPLEMENTATION ASSESSMENT ........1......... A. Statement/Evaluation of Objectives .............................. 1 B. Achievement of Objectives ................................... 2 Physical Targets and Achievements ............................ 2 Institutional Development Objectives ........................... 5 Planning Objectives . ..................................... 6 Economic and Financial Objectives ............................ 7 C. Major Factors Affecting the Project .............................. 8 Factors Not Generaly Subject to Government Control ................. 8 Factors Generally Subject to Government Control ................... 9 Factors Generally Subject to Implementation Agency Control ............. 9 D. Project Sustainability ................... ................... 10 E. Bank Performance . ....................................... 11 F. Borrower's Performance . ..................................... 12 G. Assessment of Outcome .................. ................... 13 H. Future Operation ................. .. ...................... 13 I. Key Lessons Learnt . ....................................... 13 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed wiLbhout World Bank authorization. l  Table of Contents (cont'd) PART II: STATISTICAL ANNEXES Tables 1: Summary of Assessments ................................ 15 2: Related Bank Loans/Credits .............................. 16 3: Project Timetable .................................... 17 4: Loan/Credit Disbursements: Cumulative Estimated and Actual ... ...... 17 5: Key Indicators for Project Implementation ..................... 18 6: Key Indicators for Project Operation ......................... 19 7: Studies Included in Project ............................... 19 8A: Project Costs ....................................... 20 8B: Project Financing .................................... 20 9: Economic Costs and Benefits ............................. 21 10: Status of Legal Covenants ............................... 22 11: Compliance with Operational Manual Statements ................. 23 12: Bank Resources: Staff Inputs ............................. 23 13: Bank Resources: Missions .............................. 24 APPENDICES A. Mission's Aide M6moire B. Borrower's Contribution to the ICR C. Economic Re-evaluation D. Borrower's Comments on Draft ICR E. Map  IMPLEMENTATION COMPLETION REPORT TUfRKEY DRAINAGE AND ON-FARM DEVELOPMENT PROJECT (Loan 2663-TU) Preface This is the Implementation Completion Report (ICR) for the Drainage and On-Farm Development Project in Turkey, for which Loan 2663-TU in the amount of US$255.0 million equivalent was approved on March 27, 1986 and made effective on June 15, 1987. The loan was closed on June 30, 1995, three years after the original closing date of June 30, 1992. US$50.0 million and US$30.0 million were cancelled on February 9, 1994 and on October 7, 1994, respectively. Total disbursement was US$173.9 million, including US$56.9 million for completion of the balance of works of the Bank-funded Igdir-Aksu- Eregli-Ercis Irrigation Project (Loan 2433-TU), which remained incomplete at the closing of Loan 2433. The last disbursement took place on November 28, 1995, at which time a balance of US$1.1 million remained undisbursed. The cancellation of the balance is awaiting the receipt of the final withdrawal application to allocate the expenditures funded from the special account. There was no co-financing for this project. This ICR was prepared by a mission from the FAO/World Bank Cooperative ProgrammeL' (FAO/CP) on behalf of the Agriculture and Environment Division, Country Department 1 of the Europe and Central Asia Region of the World Bank. Preparation of this ICR started with the mission's visit to Turkey from June 12 to 28, 1995. The report is based on material gathered from the project files and supervision reports, as well as findings from field visits and discussions with Bank and Government of Turkey staff associated with the project. The Borrower contributed to the preparation of the ICR by arranging field visits and meetings, and by providing comments during the preparation of the mission's Aide M6moire. 1/ Mission composed of Messrs. R. Suppa (Mission Leader, FAO/CP), P. Warner and H. Laeyendecker (Irrigation/Drainage Engineers, Consultants).   IMPLEMENTATION COMPLEIlON REPORT TURKEY DRAINAGE AND ON-FARM DEVELOPMENT PROJECT (LOAN 2663-TU) Evaluation Summary Introduction i. The Drainage and On-farm Development (DOFD) Project was appraised in June 1985 and approved by the Board on March 20, 1986, becoming effective on June 15, 1987. The project cost was estimated at US$480.5 million towards which the Bank provided a loan of US$255.0 million. The Borrower was the Republic of Turkey. The implementing agencies were the General Directorate of State Hydraulic Works (DSI) and the General Directorate of Rural Services (GDRS). The original closing date was June 30, 1992 which was extended on three occasions by one year at a time, to June 30, 1995. Total disbursements were US$173.9 million, including about US$56.9 million to complete the civil works at the Igdir-Aksu- Eregli-Ercis (IAEE) Project (Ln. 2433-TU). US$50.0 million and US$30.0 million were cancelled on February 9, 1994 and October 7, 1994 respectively leaving US$1.1 million undisbursed. The undisbursed amount has yet to be cancelled since the last withdrawal application for the adjustment of the special account is still awaited from the Borrower. Actual project costs (excluding cost of completion of IAEE schemes of US$81.2 million) amounted to US$180.3 million, as a result of a significant reduction of project scope during the course of implementation. Project Objectives and Description ii. The objectives of the project, as stated in the Staff Appraisal Report (SAR), were to (i) concentrate government's resources, while remaining within the scope of the Core Program agreed under the ASAL (Loan 2585-TU), on those subprojects which have the highest priority, and (ii) assist in improving design and implementation capacity of DSI and GDRS. The project was prepared as the first five-year time slice of a DOFD Core Program. Its main components, as described in the SAR, were: (i) surface drainage works (SDW) for desilting and remodelling of existing surface drains and construction of new drains, including pumping stations and associated structures, over an area of 668,000 ha; (ii) subsurface drainage works (SSDW) for installation and rehabilitation of subsurface drainage systems over 137,500 ha; (iii) land reclamation works (LRW) on 40,000 ha of saline and alkaline land; (iv) construction of 1,310 km of farm access roads; (v) provision of facilities for operation and maintenance (O&M); (vi) installation of piezometers for groundwater monitoring; (vii) strengthening of drainage research stations; (viii) training; and (ix)  - ii - preparation of a second five-year time slice project. The completion of the balance of irrigation and drainage (I&D) works for about 33,000 ha, at schemes which had been implemented under the IAEE Irrigation Project (funded under Loan 2433-TU, which was closed on June 30, 1992), and two further studies, an Irrigation Investment Master Plan (IMP) and an On-Farm Development Inventory and Planning Study, were subsequently included in the project. DSI was responsible for SDW component of the project, with GDRS responsible for SSDW, LRW, and other on-farm development works. Both agencies were assisted by project consultants and review consultants. Implementation Experience and Results iii. Project Preparation and Appraisal: The project was prepared and appraised as the first 5-year time slice of a DOFD 10-year Core Program. The Core Program and its first slice was prepared from the list of DSI and GDRS project inventories on the basis of some special features and criteria without initially conducting any detailed site surveys. After the consultants were appointed in September 1987, preliminary and detailed surveys were conducted. As a result of these surveys and the Government's changes in the drainage design requirement lowering the groundwater to one meter depth instead of two meters, the DOFD Core Program was revised (para 4). Seventeen I&D schemes were selected for improvement under the project, with revised targets for SDW as 161,027 ha; for SSDW, 133,339 ha; and for LRW, 9,500 ha, which was agreed by the Bank. iv. Implementation record: The 17 schemes were grouped into 7 packages for implementation. DSI carried out SDW primarily under force account and GDRS executed SSDW, LRW, and other on-farm development works using contractors selected under ICB. The original implementation period of six years was extended three times, each of one year, as project progress was affected by various factors such as delays in the consultants' appointment, lack of farmer cooperation, shortage of local funds and poor contractor performance. At the closing date of June 30, 1995, the overall physical completion of the works was over 97% for the DSI component and 50% for the GDRS component. v. Project implementation is generally considered satisfactory except the progress on the GDRS work, which has been very slow. Most of the balance I&D work for the completion of the IAEE schemes was also successfully carried out (refer to the ICR for Loan 2433-TU). Other project components such as training, procurement of equipment and additional studies were implemented satisfactorily. vi. Project impact: The project completed SDW on 157,400 ha (more than 97% of the plan). About 40,000 ha of desilting of the drains was also carried out in other Core Program schemes. SSDW was completed on 66,041 ha of the planned total of 133,339 ha, and access was improved through the construction of 700.6 km of farm roads. Project impact has already been significant in terms of agricultural production. In Izmir and Adana, average cotton yields increased from 1.2 tons/ha to 2.8 tons/ha in the previously waterlogged areas. In Konya, a wheat growing area, yields increased from 1.2 tons/ha to 3.5 tons/ha. Other  - iii - reported benefits are the reclamation of previously abandoned land and the ability of farmers to get earlier access to their fields in the spring. The completion of the IAEE schemes also contributed (refer to the ICR for Loan 2433-TU) to the generation of benefits from these schemes. vii. An unanticipated major project impact was set in motion by the Irrigation Investment Master Plan (IMP) study, which formed the basis for the Bank's 'Irrigation Management and Investment Review" (IMIR) of 1992. The IMIR concluded that the only solution for the deteriorating O&M situation in the public irrigation schemes was a transfer of the management of completed schemes from Government to farmers' Water User Organizadons (WUO). This approach was accepted by Turkish officials in 1993. In 1994, test-transfers of many DSI-managed schemes took place in four provinces. As of June 1995, the management transfer on 700,000 ha of public irrigation schemes (43% of the national DSI-managed acreage) had been completed, with targets of 61% by the end of 1995 and 100% before 2000. The success of this project, though unexpected, mainly lies in the redefinition of the role of Government in managing irrigation schemes. It broke and reversed the traditional acceptance of the need for continuous growth of public management services, in step with the growth in acreage under public irrigation. viii. Key factors affecting project implementation: The two most important factors were: (i) lack of farmer cooperation with GDRS for implementation of SSDW and LRW in some areas; and (ii) poor performance of the subsurface drainage contractors. Other factors included: (iii) overestimation of the implementation capacity of DSI and GDRS; (iv) lack of coordination between the implementing agencies; (v) GOT financial constraints limiting progress of the works from time to time; (vi) delays in the appointment of consultants; and (vii) shortcomings in project preparation. ix. Sustainability: The project is sustainable and, where the works were completed, has been welcomed by the farmers. Much will depend on how the new Water Users Organizations succeed in assuming responsibility for scheme management and O&M cost recovery, but a very good start has been made. Regular maintenance of the open drains will be of particular importance to ensure full effectiveness of the subsurface drainage system. X. Bank and Borrower performance: The Bank's performance in supervising the project was satisfactory, although its performance in project preparation and appraisal was not satisfactory. The project was poorly prepared, since the basic criteria for drainage were not examined adequately and detailed designs for at least the first year of implementation (until consultants would be appointed and would have time to prepare the remaining program) was not prepared; the shortcomings were not properly identified and addressed; and macroeconomic constraints on the capacity to provide counterpart funds were not properly identified. With respect to implementation, the performance of DSI in achieving over 97% of their planned works was satiffactory. The performance of GDRS in achieving only 50% of their planned works was unsatisfactory, although the reasons for this slow rate of progress were not wholly attributable to GDRS. The Bank's supervision role was  - iv - generally adequate and was especially noteworthy in assisting DSI with the transfer of O&M responsibilities to WUOs. Bank staff convened several discussion meetings and conferences, and initiated the arrangements for the foreign study tours of local staff to expose them to such transfer programs. The outcome of these efforts was highly successful, and resulted in the transfer of 42% of DSI-managed acreage of 1.6 million ha by June 1995. GOT's endorsement and active promotion of the transfer program has been very encouraging, and implementation of the program is proceeding satisfactorily. Lessons of national significance may emerge from this important program which was not included in the original project design. xi. Project outcome: Where the works are complete, the project achieved its objectives and significant increase in crop yields has already been obtained. Based on these results and expecting completion of the still remaining works, the project is satisfactory. The program of scheme management transfer to WUOs, not foreseen at appraisal, continues nation-wide at a rapid pace. Summary of Findings, Future Operation and Key Lessons Learned xii. Findings: There are two findings. The first relates to the need to involve the beneficiaries much more in the planning and design process for subsurface drainage systems. Their formal approval to the scope and timing of the works in the crop calendar should be obtained before construction to avoid later problems of non-cooperation. The second finding is that it was possible to transfer the management of the I&D schemes to WUJOs, which changed rapidly the ingrained government practice, because: a. the deterioration of the O&M situation had become so critical that the time was ripe for action; b. the bad effects of the situation could be quantified and proven solutions in other countries could be demonstrated to the government decision makers; and c. the solution did not require massive layoffs. xiii. Future operation: Responsibilities for the O&M of the irrigation systems have already been handed over to the farmers in about 85% of the project area. DSI plans to complete the transfer by late 1996. The drainage systems, except at Konya, will remain the responsibility of DSI until the WUOs are suitably financed and equipped. DSI will continue to monitor groundwater levels, irrigated areas etc, and collect agro-economic data including cropping patterns and yields. Both DSI and GDRS aim to complete the balance of works using their own resources but in view of the budget constraints now faced by the government it is uncertain how long this will take. DSI and GDRS have also approached the Bank for assistance for a second time slice project.  v xiv. Key lessons learned: The key lessons learned are: a. Appraisal of such a large and relatively complex project requires more careful and detailed engineering preparation studies to avoid major changes in the project components, quantities and costs during implementation. The basic change in drainage design criteria that had to be agreed during implementation is a typical example of what should have been determined during the preparation stage. A detailed, more formalized approach is required including pre-lending approval by the Bank of detailed project implementation, design, operation and maintenance plans and mechanisms appropriate for any future project in hand, including time-bound scheduling of performance activities and management warning systems. b. Beneficiary involvement at the planning and design stages was inadequate and led many farmers to oppose project implementation. This was a major factor in the slow progress of the subsurface drainage works. C. The project was too ambitious given macroeconomic constraints. Greater weight should be given during the preparation of projects to the development of realistic assessment of counterpart funding capacity. d. It was possible to convince GOT of the benefits of I&D schemes management transfer to WUOs (privatization) because: (i) the time was ripe; (ii) it was possible to quantify a deteriorating O&M situation and propose a proven solution; and (iii) the solution did not require massive layoffs. e. The employment of a single project consultant to assist both DSI and GDRS avoided additional coordination and liaison problems. f. Suitable qualifications and experience should have been included in the contractor selection criteria, to avoid the problems experienced with the subsurface drainage. g. Bank staff took advantage of a window of opportunity, during project implementation, to use it to promote the formation of Water Users Organizations to which DSI transferred responsibility for irrigation O&M expenditures. Such approach deserves to be commended as it may result in benefits outweighing those from the physical components of the project. It also serves to highlight the importance of a close dialogue between the Bank and the implementing agencies during the implementation stage, and the benefits of internal Bank cross-fertilization, which enabled the replication in Turkey of a successful experience in Mexico.   IMPLEMENTATION COMPLETION REPORT TURKEY DRAINAGE AND ON-FARM DEVELOPMENT PROJECT (Loan 2663-TU) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. Statement/Evaluation of Objectives 1. The objectives of the project, as described in the Staff Appraisal Report (SAR), were to (i) concentrate government resources, while remaining within the scope of the Core Program agreed under the ASAL (Loan 2585-TU), on those subprojects which have the highest priority, and (ii) assist in improving design and implementation capacity of DSI and GDRS. The project was prepared as the first 5-year time slice of a Drainage and On-Farm Development (DOFD) 10-year Core Program. Its components were: (i) surface drainage work (SDW) for desilting and remodelling of existing surface drains and construction of new drains, including pumping stations (wherever needed) and other structures, over an area of 668,000 ha; (ii) subsurface drainage work (SSDW) for new installation and rehabilitation of subsurface drainage systems over 137,500 ha; (iii) land reclamation works (LRW) on 40,000 ha of saline and alkaline land; (iv) construction of 1,310 km of farm access road; (v) provision of facilities for operation and maintenance (O&M); (vi) installation of piezometers for groundwater monitoring; (vii) strengthening of drainage research stations; (viii) training; and (ix) preparation of a second five-year time slice project. The completion of I&D works for about 33,000 ha, at schemes which had been implemented under Igdir- Aksu-Eregli-Ercis (IAEE) project (funded under Loan 2433-TU, which was closed on June 30, 1992), and two additional studies, an Irrigation Investment Master Plan and an On-Farm Development Inventory and Planning Study, were subsequently included in the project. The project involved two implementing agencies -- the General Directorate of State Hydraulic Works (DSI) responsible for SDW, and the General Directorate of Rural Services (GDRS) responsible for SSDW, LRW, and other on-farm development works. Consultants were contracted to assist DSI and GDRS in the planning, design, execution and overall project implementation. Separate consultants were also contracted to perform the functions of a review panel. At appraisal, the project cost was estimated as US$480.5 million, which was to be financed by a Bank loan of US$255.0 million, the Government of Turkey (GOT) contributing the remainder. The project was expected to be fully implemented in six years from 1986 to 1991. 2. At the time of project formulation, the GOT sought to complete or rehabilitate existing irrigation and drainage (I&D) schemes and to reduce the time required to complete on-farm works in areas already served by major I&D infrastructure. In this context, the project was most relevant to the country and the sector. However, a number of factors delayed and affected the full accomplishment (Section C).  - 2 - B. Achievement of Objectives Physical Targets and Achievements 3. General. At appraisal, the subprojects for inclusion in the first time slice were chosen from the list of DSI and GDRS project inventories based on some general criteria such as rehabilitation and improvement of drainage works which required less than five years implementation period, surface drainage works which are a prerequisite for any subsurface drainage, on-farm development works that could be started immediately, etc. For estimating the costs of these works, average unit prices were used which were based on experience elsewhere. Only after the consultants were appointed in September 1987, preliminary and detailed surveys of subprojects were conducted. As such, the consultants prepared the inception report and identified subprojects for implementation in the first and second time slices by June 1988. Actual SSDW commenced on the first subproject on August 1989, three years later than the Board approval date of March 20, 1986. Due to these delays in the inception of the construction works and further complications in the project's implementation progress, paras 33-34, the original closing date of June 30, 1992 was extended one-year at a time to June 30, 1995, mainly to finish some uncompleted SSDW. 4. The scope and size of the first time slice went under further changes from what was envisaged in appraisal as a result of the consultants' critical review of the country's irrigation and drainage requirement in late 1987; the Government changed the drainage design requirement of lowering the groundwater to one meter depth instead of two meters; and finding uneconomical the installation of buried pipe collectors to assure free flow for the outflow from the sub-surface drains, as such, buried collector drains were replaced by open collectors. In the light of these new developments, the scope of the ten-year Core Program was reduced from 1,400,000 ha to 309,983 ha for SDW and from 440,000 ha to 216,630 ha for SSDW in I&D schemes. Due to the reduction in the works quantities, the program under the first slice was revised. A number of schemes that were originally envisaged for the second time slice and some SDW for the schemes under the "third priority class" of the Core Program became part of the project investment program. 5. The achievement of physical targets can be generally considered as satisfactory, except for SSDW. It must be noted that only half of the planned SSDW was completed and SSDW accounted for 64% of the project cost as originally estimated. The Completion of the balance I&D works at the schemes of the IAEE project was successfully carried out (refer the ICR for Loan 2433-TU). Although, protection of the environment was not among the objectives of the project, in the context of the subproject design, environmental impact was considered. The project, overall, had positive environmental impact as it provided drainage in the waterlogged land in several subproject areas with effluent going to the sea, and as it contributed to efficient water utilization in the farm level. 6. Seventeen schemes were included in the project, and grouped into seven packages for implementation. The SDW was to be undertaken primarily under force account and very  little under LCB contract. SSDW, on the other hand, was to be executed under ICB contracts. The equipment for SDW had been mostly procured earlier, except 365 dump trucks which were procured under the project. As of February 1996, SDW was complete for 157,400 ha (more than 97% of planned 161,044 ha). Additional SDW over 40,000 ha was also carried out in part of the area to be taken up under the second time slice and 'third priority" schemes. The SSDW was completed on 66,041 ha (50% of planned 133,339 ha). However, due to farmer resistance in certain schemes and minor design revisions during construction, over 30,000 ha of the SSDW is likely to be deleted, reducing the target to about 100,000 ha. The SSDW could therefore be considered to be about 66% complete. Farm access roads were completed in 700.6 km (54% of target). For LRW, 55,600 tons of gypsum was delivered to site which is sufficient to reclaim 2,271 ha of alkali soils (21% of target). At the current rate of progress the subsurface drainage and on-farm development works would probably be completed in 1997-98. Further details of the implementation of individual packages are given below. 7. Adana I: Under this package (Seyhan I scheme for 20,200 ha), envisaging SDW and SSDW over the entire area, SDW is fully implemented and SSDW on about 19,170 ha (against an original target of 20,200 ha). SDW started in 1988 and was completed in 1990. SSDW started in 1989 and was completed in 1992. The area for SSDW was reduced by about 5% due to minor design revisions. Farm roads were completed in 110 km. The impact of both SDW and SSDW is already significant. Before the project (1987/88), about 8,000 ha (40% of the area) had the groundwater table within 1 m below ground level. Shortly after project completion (1992/93) this waterlogged area had been reduced to 500 ha (2%). Cotton yields are reported to have risen from 2.5-3.0 ton/ha to 3.5-4.0 ton/ha, and maize yields increasing from 6-7 ton/ha to 9-10 ton/ha. 8. Adana II: Under this package (Seyhan II, Berdan, and Silifke schemes for 27,550 ha), envisaging SDW on 20,017 ha (including two drainage pumping stations at Seyhan II and Berdan), and SSDW for 12,795 ha, SDW has been completed 95% and SSDW only 5%. SDW started at Seyhan II, Berdan, and Silifke in 1991, 1988, and 1988 respectively; and were completed in 1993, 1995 (about 1,000 ha remained on June 30, 1995 which was to be completed by the year end), 1992 respectively. Progress on SSDW has been extremely slow with only 678 ha implemented till June 30, 1995. The first contract for the package was awarded quite late in 1990 but only SSDW for 570 ha was complete after nearly two years, when the contract was terminated. A second contract was awarded in 1993, but SSDW for only 108 ha could be completed, when the contractor stopped operations in late 1994. This contract was terminated in 1995. GDRS has no immediate plans to re-start the work. Farm roads were completed in 73 km (62% of planned). The implementation of SDW had a significant impact. At Berdan 500 ha of previously abandoned land has been reclaimed. 9. Adana m: Under this package (Ceyhan, Kesiksuyu and Kahramanmaras schemes for 116,850 ha), envisaging SDW on 36,085 ha (including six new pumping stations) and SSDW on 29,322 ha, SDW is 90% complete and SSDW only 36%. The SDW was started at Ceyhan, Kesiksuyu, and Kahramanmaras in 1988, 1989,and 1988 respectively, and was  - 4 - completed in 1995 (only 3,500 ha remained on June 30, 1995 which was to be completed by the year end), 1994, and 1993 respectively. The civil works of the six new pumping stations at Kesiksuyu are more or less complete. The pumps are already at site. Local contracts for the installation of the pumps will be let shortly. Farm roads have been completed over 133 km and 28% of the gypsum required for land improvement delivered at site. The reasons for the slow progress on SSDW and LRW were mainly farmer resistance and a very wet spring in 1995, which made working conditions difficult. The contractor stopped work in April 1995 due to a contractual dispute which could not be resolved yet. The impact of SSDW could not be assessed in terms of increase in yields due to recent completion. However, farmers acknowledge the benefits of the improved surface drainage; in particular enabling them earlier access to their fields in the spring. 10. Izmir I: Under this package (Saraykoy, Curuksu, Akcay, and Nazili schemes for 64,550 ha), envisaging SDW for 28,558 ha and SSDW for 19,732 ha), SDW has been completed, and SSDW 34% (mainly in Saraykoy and Curuksu). The SDW started in 1988 and completed in 1993. The SSDW started in 1990 but the contract was terminated in 1992 because of poor progress, primarily due to the lack of farmer cooperation. Farm roads have been constructed in 54 km (53% of planned). 11. Izmir II: Under this package (Menemen and Turgutlu schemes for 32,840 ha), envisaging SDW (including 3 new pumping stations) for 17,400 ha and SSDW for 13,769 ha, SDW was completed fully in 1994 and SSDW executed over 66% of the area (all at Menemen), by June 30, 1995. Farm roads have been completed over 78.9 km. At Turgutlu, the farmers informed GDRS that they did not want the subsurface drainage works to be carried out although farm roads have been constructed. Work is continuing at Menemen, using GDRS financial resources, and at the current rate of progress would be completed in 1998. Farmers at Menemen are reporting significant improvements in yields (from 1.5 ton/ha pre-project for cotton to 3 ton/ha) and ability to cultivate previously abandoned lands. 12. Konya I: Under this package (KOS IV and VII schemes for 20,800 ha), envisaging SDW for 17,684 ha and SSDW for 15,904 ha, both SDW and SSDW (reduced to 15,244 during detailed design) are complete. Farm road have been completed over 178 km. For LRW, gypsum sufficient for improvement of 409 ha of land was delivered. The Konya I area suffered acutely from waterlogging before the project, with water table 0.5 m below ground level. The project had an immediate impact of lowering it. Although the subsurface drainage system has only been in place for a year, groundwater levels have already fallen by an average of 0.8 m. 13. Konya II: Under this package (KOS VI and Uluirmak schemes for 62,850 ha), envisaging SDW for 21,100 ha (including one pumping station) and SSDW for 16,617 ha, SDW is 71 % complete and SSDW 41 %. SDW commenced at KOS VI and Uluirmak in 1993 and 1988 respectively. At Uluirmak, as of June 1995, SDW in about 3,500 ha of the planned 4,500 ha had been completed, and at KOS VI, about 11,500 ha of the planned  - 5 - 16,600 ha. DSI hopes to finish the balance works by early 1996. The GDRS works are much delayed. A contractor was appointed in 1993, but the contract was terminated in 1994 with no tangible progress. A new contractor (the same company who successfully completed Konya I) was appointed in late 1994 and has already made good progress, completing the delivery of gypsum and road work, 74 km. About 500 ha of subsurface drainage has also been installed. GDRS plans to complete the balance works from their own financial resources, and at the current rate of progress completion is anticipated in early 1997. No data is available on the physical impact of the project, but based on the experience at Konya I, significant benefits should soon be achieved if the current rate of implementation is maintained. 14. Facilities for operation and maintenance (O&M]: The project included the construction of limited number of field offices, stores, workshops, etc. for the use of DSI O&M staff, which were constructed (including operator's quarters at the drainage pumping stations) with DSI funding only. Various items of O&M equipment such as current meters, water level recorders, and survey equipment were procured from the loan. 15. Installation of piezometers: The project envisaged the installation of a network of piezometers and observation wells in the problem areas for the monitoring of groundwater levels. This network has been established (typically 1 piezometer per 100 ha) under DSI funding only. The data is being regularly collected and analyzed. Institutional Development Objectives 16. General: The project has achieved its institutional development objectives through the strengthening of drainage research stations, and training of DSI and GDRS staff. The proposed procurement of additional computers for DSI has not been completed yet. 17. Strengthening of drainage research stations: The project financed laboratory equipment for the drainage research stations at Menemen and Konya, and repairs and extensions to the buildings at Konya. 18. Training: Both DSI and GDRS implemented their training components satisfactorily. The project provision for training a total of 35 trainees (400 trainee-months), with Bank's agreement, was changed to shorter study tours benefiting 109 staff (250 trainee-months). DSI sent 81 staff on study tours of between 2 weeks to 3 months each to USA and Mexico covering the operation and maintenance of I&D systems, hand over of I&D systems to farmers, and strengthening of Water Users Organizations (WUOs). The GDRS trained 61 staff, 28 pursued formal Masters degree courses or other short courses in land drainage in Holland, and 33 benefitted from the study tours to Spain, USA and Chile. Overall the training program had a good impact on both DSI and GDRS trainees, the tour by DSI trainees to Mexico for the study of the transfer of I&D schemes management to farmers making the best contribution. In addition to the formal training and study tours, 26 counterpart staff from DSI and 22 counterpart staff from GDRS were seconded to work with  - 6 - the project consultants, who received on-the-job training. DSI and GDRS staff and four provincial governors went on a one week tour to Spain to study the mandatory transfer of irrigation schemes to farmers' WUOs. 19. DSI computerization: The Bank agreed, in 1993, to the DSI request for assistance in the procurement of additional hardware and software for computer facilities, estimated to cost US$0.55 million. The procurement under ICB package having run into considerable processing delays, DSI decided to procure 45 PCs locally at a cost of US$0.2 million. However, as of June 1995, the procurement had not been completed. Planning Objectives 20. General: Under the project, three studies were completed to assist DSI and GDRS in the planning of investments and implementation of the I&D, SDW, SSDW, and LRW, as given below: 21. Preparation of second time slice: Fourteen I&D schemes were studied for the SDW, SSDW, LRW and other requirements, of which 12 were selected for inclusion in the proposed second time slice project. Detailed designs (but not tender documents) have been completed for the SDW and SSDW required in about 55,000 ha distributed over the 12 schemes. The cost is estimated at US$71.7 million (US$38.2 million for SDW and US$33.5 million for SSDW). Other components of on-farm development works for the time-slice project are described in para. 22. 22. Irrigation Master Plan: This study was to be originally carried out under the Bank- funded Igdir-Aksu-Eregli-Ercis Irrigation Project (Ln. 2433-TU) but was transferred to the DOFD project due to lack of available funds in the earlier loan (Bank review in 1989). In December 1991, the Irrigation Master Plan (IMP), presenting a 10-year irrigation investment strategy for the years 1992-2001, was finalized. IMP was based on recalculated ERRs of over 400 projects, with the priority for implementation in descending order of ERR, under different scenarios of investment budgets and executing capabilities of DSI and GDRS. Subsequent yearly IMP updates took into account the ongoing construction and completion of schemes; the 1995 update presents an investment plan for the period 1996-2005. 23. On-farm development inventory and planning study: This study was not foreseen at project appraisal but need became apparent as a result of slow progress in SSDW. The objective of the study was to determine the scope, cost and economic feasibility of on farm development and to prioritise subprojects for a development package. The study which took 16 months, from January to April 1993, identified 29 schemes with viable ERR on which work could commence prior or parallel to DSI rehabilitation work. In addition, the study concluded that successful implementation of SSDW work required greater user participation in the planning and development work, and user groups were needed which can be formed on the basis of hydraulic units, to bear O&M and investment cost.  24. Transfer of I&D schemes to WUOs. As preparation for future Bank lending to support Turkish irrigation development, Bank prepared in 1992 an "Irrigation Management and Investment Review" ([IUR). During its preparation, it became clear that the tightening national budget situation no longer permitted continuation of the generous subsidies which had been provided annually to fill the widening gap between O&M expenditures and collected water tariffs. With insufficient funding, IMIR predicted a rapidly approaching need for widespread rehabilitation as a result of continuously deferred maintenance. 25. Recognizing that the collection of water-charges by government agencies would never be able to reach full coverage of O&M costs, IMIR recommended divesting DSI of the its public responsibility for the management of I&D completed schemes, by transferring these to farmers' WUOs. During 1993, many discussions and conferences among various irrigation interests and government authorities were held, and project-supported visits to successfully transferred irrigation districts in Spain and Mexico were arranged for the decision makers. It convinced the DSI management to initiate large-scale transfer trials in four provinces in 1994. The success was so convincing, that the program snowballed beyond any expectation. By June 1995, the management on 670,000 ha had been transferred to WUOs, or 42% of the DSI-managed acreage of 1.6 million ha. The program aimed to reach 1 million transferred hectares by the end of 1995. It has resulted in substantial savings to the Turkish Treasury, expected to reach the order of US$100 million per year in 1996. Already, it is drawing irrigation specialists from other Mediterranean and Balkan countries to observe the Turkish program, taking place in a low-key, low expenditure fashion, using local capabilities. 26. This development was not planned at the time of appraisal. It emerged when the time was ripe for it, and when the Bank's assistance could be actively provided under the only ongoing Bank-financed I&D project. It was actively supported by Bank supervision staff, who took advantage of the window of opportunity and initiated the arrangements for the foreign study tours for the concemed officials to expose them to such transfer programs. Great praise is due to the DSI management for deciding to experiment first, and then accept such a fundamental change. The project reversed the traditional view that led to a continuous growth of public expenditures, in step with the continuing growth in acreage brought under public irrigation. The Turkish endorsement and active promotion of the transfer program is the outstanding, but unanticipated at appraisal, success of this project, the results having national significance. Economic and Financial Objectives 27. Costs: Actual project costs (excluding cost of completion of IAEE schemes of US$81.2 million) amount to about US$180.3 million (Appendix C, Table 1) or about 38% of the amount estimated at appraisal. The main reasons were the reduced GDRS program (para. 4); and higher unit costs used at appraisal for both DSI and GDRS works. Little experience with subsurface drainage installation was available in Turkey at the time of appraisal and the SAR cost estimates were to a large extent based on West-European costs. Thus, SAR unit cost estimates for SSDW ranged from a minimum of US$1,400/ha  - 8 - (Menemen) to a maximum of about US$2,900/ha (K.O.S.). Actual costs for 1995 vary from US$885/ha (Turgutlu) to about US$1,500/ha (Saraykoy). 28. Financial returns: Four farm models were developed at appraisal to assess the project impact on the waterlogged and non-waterlogged areas of the I&D schemes. Because of lack of data, this ICR financial analysis does not present farm models. Instead, three different "before and after project" scenarios, for Adana and Izmir (coastal areas) and Konya (inland area) are presented for two ha farms. The financial returns are satisfactory with increases in net farm incomes of 60%, 75% and 50% for Adana, Izmir and Konya respectively (Appendix C, Table 6). 29. Economic returns: Total project costs for economic analysis (Appendix C, Table 5) have been worked out in constant TL June 1995 prices (Appendix C, Table 1). The agricultural products were divided into tradeable and non-tradeable, and the farm gate economic price for the tradeable was calculated from the Bank projections for the commodity prices duly adjusted for the transport, handling, and processing. For the non-tradeable, the financial price was taken as economic. Economic and financial prices of crops are given in Appendix C, Table 4. 30. At appraisal, Economic Rates of Return (ERR) were estimated for five schemes in the coastal areas and for one in the inland zone. These ERRs ranged from 22% to 31 % in the coastal areas, 10% in the inland zone, and 22% for the project as a whole. The ex-post economic re-evaluation has been carried out for this ICR for three packages (Appendix C, Table 7) giving ERRs of 20.7%, 24.0% and 18.8% for Adana, Izmir and Konya respectively. Results obtained for Adana are representative of the I&D schemes Seyhan, Berdan, Silifke, Ceyhan in the coastal areas; and the results for Izmir for Saraykoy, Akcay, Nazili, Menemen and Turgutlu, also in the coastal areas. Similarly, Konya represents a typical situation for inland I&D schemes KOS IV and VII, and KOS VI. The ERR for the project as a whole has been assessed at 21.6% which, in spite of the changes during implementation, is in line with SAR's estimate. Approximately 75,000 farmers have benefitted under the project or about 80% of the SAR target. The annual incremental agricultural production, is estimated at 40,000 tons of cotton, 60,000 tons of wheat and about 70,000 tons of vegetables. C. Major Factors Affecting the Project Factors Not Generally Subject to Government Control 31. The performance of the contractors on the SSDW was generally poor. Of the seven packages, three have had contracts rescinded (one package twice) and only one package (Konya I) was completed within the original contract period. The problems arose from the contractors' lacking suitable machinery and equipment (particularly trenchers), inexperience in subsurface drainage works, unfamiliarity with ICB procedures, and non-cooperative farmers.  Factors Generally Subject to Government Control 32. Factors under government control were: (i) fulfillment of the condition of effectiveness; (ii) GOT financial constraints, delaying release of funds on time, which affected the progress of both the DSI and GDRS works from time to time; (iii) two agencies (DSI and GDRS) were involved in project implementation and delays occurred from a lack of synchronization in their activities (could be avoided if one agency was responsible for all the project activities); (iv) the project consultants were employed only to assist DSI and GDRS in the planning and general supervision of the project (responsibility for contract management could have improved the project implementation); and (v) GOT's adoption of a policy to transfer I&D schemes management responsibilities to the farmers (an important factor in project sustainability). 33. The Loan became effective on June 15, 1987, almost one year from the original effectiveness deadline date of June 27, 1986. The conditions of effectiveness were (i) appointment of project consultants, and (ii) approval by the Council of Ministers of cost recovery rules and regulations for the GDRS. While the second condition was fulfilled on November 8, 1986, the first condition posed problems as it took quite some time to reach agreement on the terms and conditions of the contract. As negotiations with the short-listed consultant were advancing satisfactorily, Bank agreed with the borrower request to make the project effective on the aforesaid date; however, actual appointment of consultants took place in September 1987, some three months later. Delays in the consultant appointments were one of the key factors in the project implementation delays, as consultant services were crucial to firm up the list of subprojects for the first and second time slices and to make for the first slice detailed designs of the selected subprojects. Factors Generally Subject to Implementation Agency Control 34. A major factor affecting the slow progress of SSDW has been farmer resistance and non-cooperation, either because they felt the works were unnecessary or did not want their crops, in particular cotton, damaged by the construction activities. These problems could have been lessened by better liaison and beneficiary involvement at the planning stage. Other factors were: (i) delay in assigning the agreed counterpart staff to work with the consultants, and in the appointment of the GDRS Project Manager; (ii) an unrealistic assessment of the implementation capacities by the agencies and hence unrealistic implementation schedule; (iii) lack of coordination between DSI and GDRS, both at central and regional levels; (iv) delays in contractual procedures and approvals; (v) need for more careful selection by GDRS of suitably qualified and experienced contractors for subsurface drainage works, together with closer supervision and control during construction; and (vi) desirability of more detailed project preparation by both DSI and GDRS.  - 10- D. Project Sustainability 35. The sustainability of the project wiU depend on: (a) efficient O&M of the systems with suitably trained staff and equipment; (b) ensuring that O&M does not suffer for want of funds; (c) mitigating the adverse affects of any environmental impact; and (c) acceptance of the project by the farmers. 36. Operation and Maintenance: O&M has been the responsibility of the O&M sections of the Regional Directorates of DSI. In late 1993, the O&M Department of DSI in Ankara started the transfer of major I&D schemes management responsibilities to the farmers through the establishment of WUOs. Excellent progress has since been made. As for the schemes under the project, as of June 30, 1995, transfer was complete in 12 of the 16 schemes and was in progress at 3 others (Silifke, Saraykoy and Nazili), with only one scheme (Berdan) where the process was yet to begin. The corresponding areas were 136,432 ha (85%) transferred, 16,112 ha (10%) in progress and 8,500 ha (5%) yet to begin. DSI expects to complete the transfer process in 1996, generaly in blocks of 10,000 ha. DSI will stiUl remain responsible for the O&M of the main works such as dams and main canals, the sale or hire of O&M equipment to the WUOs, and providing technical assistance. DSI is currently not handing over responsibilities for the drainage systems to the WUOs, except at Konya where the maintenance requirements are less than in other areas. This is appropriate (though current levels of DSI drain maintenance need improvement as discussed in para. 35 of the aide-memoire), since the WUOs would probably focus initially on the irrigation systems and do little or nothing to the drains. DSI wiUl pursue a progressive transfer whereby the WUOs would take responsibilities for the drainage system when they are well organized and equipped for the task. 37. The drains have the constant problem of heavy growth of weeds, even in those channels where recent remedial works were carried out under the project, except where drier weather prevails. Excessive weeds cause siltatidh and backing-up in the drains, thereby submerging the subsurface drain outlets. DSI normally carries out the maintenance work on the drains every 3-5 years (irrigation canals are cleaned annually). The project wiUl increase the drain maintenance requirements through the construction of new drains and deepening of existing drains. The condition of the drains will require close monitoring by DSI to ensure that SSDW outlets are not submerged, with the maintenance program increased as necessary. Wastage of irrigation water into the drains, Which promotes weed growth, would also need to be controlled. 38. Cost recovery (O&M): At present DSI is attempting to recover the full annual O&M cost from the farmers. Water charges for the current year are computed on the basis of the actual O&M cost for the previous year, so there is a hidden subsidy element with inflation running at 70% or higher per year. Collection of the charges is the responsibility of the Ministry of Finance. The overaU recovery rate is presently estimated at 45%. The penalty (10% above the outstanding charge) for late payment is meaningless in an environment of high inflation, and delays are common because interest charges are not applied.  - 11 - 39. Cost recovery (capital cost): DSI has the policy of recovering the capital cost (without adjustment for inflation or interest) of a project over a 50 year period. No costs are recovered for the first 10 years. Irrigation projects are divided into four types and the recovery rate varies from TL 3000/acre to TL 7500/acre. Recovery without interest over a long recovery period of about 40 years in an inflationary environment yields in real terms only a negligible fraction of the actual capital cost. No cost recovery is presently carried out by GDRS. 40. Environmental impact: The project is confined to areas which are already developed and served by fully or partially completed I&D systems. Any negative impact is inconsequential. Drainage effluent from the irrigated lands is discharged into the sea or large rivers and lakes and no detrimental effects occur. However, in response to the action required by the Bank supervision mission of June 1988, DSI and GDRS undertook environmental studies. As a result, certain designs were revised to mitigate possible negative environmental effects, and particular attention was paid to the Menemen scheme which required additional water for the preservation of the wet lands natural reserve situated outside but adjacent to the project. 41. Acceptance of the project by the farmers: Where the drainage improvement works have been completed, farmers welcome the project, and quote examples of tangible benefits (yields, cropped areas, etc.). They also welcome the transfer of management responsibility from DSI to WUOs. 42. Conclusions: On the basis of the criteria given in para. 33 and discussed above, the project is sustainable. The major area of concern is the maintenance of the open drains which must be kept clear of excessive weeds and silt to assure the carrying capacity and to permit effective operation of the subsurface drains. E. Bank Performance 43. Although the Bank's supervision performance was satisfactory, its performance concerning project preparation and appraisal was not. In spite of spending considerable time, 213.1 staff weeks, to bring the project into appraisal stage, the project was poorly prepared since the basic criteria for drainage were not examined adequately and a detailed design for at least the first year of implementation (until consultants would be appointed and have time to prepare the remaining programs) was not prepared. In addition during the project preparation stage, the stockholder views were not taken into account; implementation capacity shortcomings were not properly identified and addressed, specially lack of GDRS's experience with the rules and regulations of construction under ICB contract, and macro economic constraints on the capacity to provide counterpart funds were not properly considered. Furthermore, the project cost seems to be based on inflated unit costs experienced, outside Turkey. The Bank and the implementing agencies failed to revise project costs and determine a realistic work program that could be carried out during the remaining project time and avoid unnecessary commitment fees. The implementing agencies,  - 12 - however, became more confident as they gained experience with the ICB roles and overcame the implementation problem. As such, they were expecting further extension. Therefore, they did not request cancellation of unutilized funds until they became convinced that there would be no extension. 44. Sixteen supervision mission were mounted during project implementation (Table 13), which were adequate. The Bank's supervision role was generally adequate and was especially noteworthy in assisting DSI with the transfer of O&M responsibilities to WUOs. Bank staff convened several discussion meetings and conferences, and initiated the arrangements for the foreign study tours of local staff to expose them to such transfer programs. The outcome of these efforts was highly successful, and resulted in the transfer of 42% of DSI-managed acreage of 1.6 million ha by June 1995. The Bank missions kept an excellent working relation with the implementing agencies and succeeded in jointly resolving a number of technical and institutional problems during the implementation period. 45. With respect to the covenants on recovery of investment and O&M costs, the Bank was unable to achieve compliance, and pushed instead the transfer of O&M responsibilities to the farmers, thereby transferring the O&M cost recovery to the water users themselves. F. Borrower's Performance 46. DSI Successfully implemented over 97% of the works, exhibiting good performance. GDRS, however, achieved only 50% of their target, indicating weak performance, though the slow progress was not wholly attributable to weakness in GDRS. The project was not prepared properly to permit GDRS to commence its work as soon as project became effective. Commencement of work dependent on the consultant to be appointed and present detailed study of subprojects which accomplished with more than three years delays. In addition, the surface drains have to be necessarily completed first for the subsurface drains to function and SSDW has to lag behind SDW; the high level of farmer no-cooperation was certainly not foreseen; and the contractors executing the GDRS jobs had no previous good experience and were learning on the project. 47. Although the SAR recommendations with respect to coordinating activities were generally followed, various supervision missions noted that closer cooperation between DSI and GDRS could have improved project implementation. 48. DSI has to be given credit for the initiative and progress in the transfer of O&M responsibilities to the farmers. The program (paras 23-25) has raised the impact of the project to the national level and is creating international interest.  - 13 - G. Assessment of Outcome 49. Although not all project works have been implemented, the project's outcome is rated as satisfactory. Where the works are complete, the project has fully achieved its objectives, and significant benefits in terms of crop yields have been realized. Where only the surface drainage improvements are complete, partial benefits are reported. Project sustainability is good, although the surface drains should be properly maintained to ensure drainage and effective operation of the subsurface drains. The privatization of scheme management is an outstanding success. H. Future Operation 50. Responsibility for the O&M of the irrigation systems in about 85% of the project area has already been handed over by DSI to the farmers, and full transfer is planned for late 1996. Except at Konya, hand over of the drainage systems will remain the responsibility of DSI until DSI is confident that the WUOs are suitably financed and equipped. DSI will continue to monitor irrigated areas, groundwater levels, and agronomic data including cropping patterns and yields. 51. Both DSI and GDRS have indicated that they will complete the balance works using their own resources. They also express a strong desire for Bank assistance for the proposed second time slice project. I. Key Lessons Learnt 52. The key lessons learnt from the project are: (a) Appraisal of such a large and relatively complex project requires more careful and costly engineering preparation studies to avoid the major changes in project components, quantities and costs. The basic change in drainage design criteria is a typical example of what should have been determined during the preparation stage. A detailed, more formalized approach is required including pre-lending approval by the Bank of detailed project implementation, design, operation and maintenance plans and mechanisms appropriate for any future project at hand, including time-bound scheduling of performance activities and management warning systems. (b) It was possible to convince GOT of the benefits of scheme management transfer (privatization) because: the time was ripe; the reduction in O&M subsidies having started a trend to defer maintenance in a way that alarmed DSI;  - 14 - it was possible to quantify a deteriorating situation and propose a proven solution that could be studied elsewhere; and - the solution was painless by not requiring massive layoffs. (c) The employment of a single project consultant to assist both DSI and ODRS avoided additional coordination and liaison problems. (d) Beneficiary involvement at planning and design stages was inadequate and led many farmers to oppose project implementation. This was a major ctor in the slow progress of the subsurface drainage works. (e) Suitable qualifications and experience should have been included in the selection criteria, to avoid the problems experienced with the subsurface drainage contractors.  - 15 - PART ML STATISNCAL A tEXES Table 1: S_umary of Aenimb A. ieatof obiggm2auiL EtlL fh&NLMU () (0) ( (V m elo o iCi1 Fal Eo O F O bi dsP- E