Senegal - Senegal River Polders Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 2777 FILE COPY PROJECT PERFORMANCE AUDIT REPORT SENEGAL RIVER POLDERS PROJECT (CREDIT 350-SE) December 26, 1979 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OFFICIAL USE ONLY CURRENCY EQUIVALENTS 1972 (at appraisal) US$1.00 - CFAF 255.79 CFAF 100 - US$0.39 WEIGHTS AND MEASURES I hectare (ha) - 2.47 acres 1 kilometer (km) - 0.62 mile 1 ton - 2,004 pounds ABBREVIATIONS ACP - Agent Comptable Particulier BEP - Bureau Etudes et Programmation BOM - Bureau d'Organisation et Methodes CCCE - Caisse Centrale de Cooperation Economique CEP - Centre des Etablissements Publics COF - Controle des Operations Financieres FAO-CP - Food and Agricultural Organization - Cooperative Program LBRD - International Bank for Reconstruction and Development ICB - International Competitive Bidding IDA - International Development Association IRAT - Institut de Recherche Agronomique Tropicale ISRA - Institut Senegalais de Recherche Agronomique LCB - Local Competitive Bidding MDRH - Ministere du Developpement Rural et des Ressources Hydrauliques OMVS - Organisation pour la Mise en Valeur du Fleuve Senegal SAED - Societe d'Amenagement et d'Exploitation des Terres du Delta du Fleuve Senegal SATEC - Societe d'Aide Technique et de Cooperation (French para- public consulting firm) SCET INTERNATIONAL - Societe Centrale pour 1'Equipement du Territoire International CONVERSION RATES Paddy to Milled Rice - 65% Fresh tomato to Tomato Paste - 18% FISCAL YEAR SAED July 1 - June 30 GOVERNMENT July 1 - June 30 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PROJECT PERFORMANCE AUDIT REPORT SENEGAL RIVER POLDERS PROJECT (Credit 350-SE) PREFACE This is a performance audit of the River Polders Project in Senegal for which Credit 350-SE in the sum of US$4.5 million was approved in December 1972. The final disbursement was made on March 18, 1976 and the credit was closed on December 31, 1977 without any delay. The audit consists of an audit memorandum prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR) dated May 14, 1979. The PCR was prepared by the Western Africa Regional Office on the basis of the report of a consultant who visited the country in July 1978. The audit memorandum is based on a review of the Appraisal Report (No. 29-SE) dated November 29, 1972, the President's Report (P-1157-SE) of December 7, 1972, the Credit Agreement dated January 9, 1973 and the PCR. Correspondence with the Borrower and internal Bank memoranda on project issues contained in relevant Bank files have also been reviewed, and Bank staff associated with the project have been interviewed. An OED mission visited Senegal in November 1978. Discussions were held with the Ministries of Rural Development and Finance and project officials. The information obtained and observations made during that mission were used to test the validity of the conclusions of the PCR and permitted discussion of rehabilitation problems and farmer organizations. The audit generally agrees with the major conclusions of the PCR and finds it comprehensive and accurate with respect to the project's achieve- ments and shortcomings. The audit elaborates and focuses on issues related to project design, economic rate of return, Bank performance, equity and project beneficiaries. A copy of the draft report was sent to the Borrower on October 14, 1979 for comment. No comments have, however, been received. The valuable assistance provided by the Government of Senegal, project staff and the farmers visited during the preparation of this report is gratefully acknowledged.  - ii - PROJECT PERFORMANCE AUDIT BASIC DATA SHEET SENEGAL RIVER POLDERS PROJECT (Credit 350-SE) KEY PROJECT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (US$ million) 7.4 14.4 Overrun (%) 95% Credit Amount (US$ million) 4.5 4.5 1/ Disbursed (US$ million) - 4.5 Cancelled (US$ million) - - Date Physical Components Completed 12/76 6/78 Proportion Completed by Above Date 100% 100% Proportion of Time Overrun (%) 195% Economic Rate of Return (%) 14% Negative - 6% OTHER PROJECT DATA Original Actual or Item Plan Revisions Current Estimate First Mention in Files or Timetable 06/69 Government's Application 03/71 Negotiations 10/72 11/06/72 Board Approval 12/72 12/07/72 Credit Agreement Date 01/09/73 01/09/73 Effectiveness Date 04/09/73 06/08/73 Closing Date 12/31/77 12/31/77 Borrower Republic of Senegal Executing Agency Societe d'Amenagement et d'Exploita- tion des Terres du Delta du Fleuve Senegal (SAED) Fiscal Year of Borrower July 1 - June 30 Follow-on Project Name i) Debi-Lampear Engineering Credit Credit Number S-18-SE Amount (US$ million) 1.0 million Credit Agreement Date 05/22/75 ii) Debi-Lampsar Irrigation Project Credit Number 775-SE Amount (US$ million) 20 million Credit Agreement Date 03/17/78 MISSION DATA Sent Month, No. of No. of Man- Date of Item b Year Weeks Persons weeks Report Identification FAO/CP 04/69 2 1 2 06/69 Preparation French consul- 03/72 tants SCET- International assisted by CP and PMWA Appraisal Headquarters March/ 4 3 10.5 11/29/72 April 1972 TOTAL 6 12.5 Supervision I Headquarters .05/73 2.2 1 2.2 06/21/73 Supervision II Headquarters 12/73 1.0 1 1.0 01/22/74 Supervision III Headquarters 07/74 1.2 1 1.2 08/28/74 Supervision IV Headquarters 12/74 1.4 2 2.4 01/31/75 Supervision V Headquarters Jan/Feb 76 1.4 1 1.4 02/25/76 Supervision VI Headquarters Oct/Nov 77 1.4 2 2.8 12/21/77 Supervision VII Headquarters 05/78 0.8 1 0.8 06/27/78 Completion Consultant 07/78 2.6 1 2.6 09/25/78 supervised by H.Q. TOTAL 2/ 11.8 14.4 COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) CFAF Franc (CFAF) Year: Appraisal Year Average (1972) Exchange Rate: US$1 - CFAF 256 Intervening Years Average (1973-1978) US$1 - CFAF 240 Completion Year Average (1979) US$1 - CFAF 220 1/ Project was scaled up by about 30%. 2/ Additional supervision occurred during other missions to Senegal (preparation and appraisal of the two follow-up projects, visits with SAED management, etc.). There were two missions in 1975, two in 1976 and one in 1977. The amount of additional supervision has been estimated at four man-weeks, bringing the total supervision time to 18.4 man-weeks.  - iii - PROJECT PERFORMANCE AUDIT REPORT SENEGAL RIVER POLDERS PROJECT (Credit 350-SE) DISBURSEMENT TABLE (US$ '000 Equivalent) Actual as % Quarter ending Appraisal Estimate Actual Disbursement of Estimate 06/30/73 70 0 12/31/73 875 - 0 06/30/74 1,385 606 44 12/31/74 2,224 1,693 76 06/30/75 2,973 3,130 105 12/31/75 3,737 4,253 114 06/30/76 4,463 4,500 101 12/31/76 4,477 - - 06/30/77 4,500 Closing Date: 12/31/1977  - iv - PROJECT PERFORMANCE AUDIT REPORT SENEGAL RIVER POLDERS PROJECT (Credit 350-SE) HIGHLIGHTS The project's primary objective was to increase rice production by 8,700 tons through construction of irrigation works and subsequent agri- cultural development in three areas of the Senegal Delta. About 2,100 ha were expected to be equipped with flood control systems and 2,700 ha with full water control systems. The project also aimed at increasing the production of tomatoes and several other vegetables. It was expected to benefit about 1,700 low income families. A number of problems delayed implementation of the project and affected its success: Project designs (prepared by consultants) were defec- tive, requiring considerable changes. Although the Government's financial contribution was substantially greater than originally anticipated, it was delayed in the early part of the project period. Irrigation works were poorly maintained. Technical assistance was less than expected at appraisal and that available was ineffectively used. In mid-project the flood control irrigation systems were found unreliable and the Bank agreed that only full water control systems should be developed. Scaling up of project works together with strong price inflation during the implementation period resulted in a cost overrun of about 95% (PPAM paras. 4 and 26). Total irrigated area developed under the project was 2,960 ha compared with about 4,800 ha estimated at appraisal (PCR paras. 1.03 and 3.04). The area planted to rice has been less than envisaged at apprai- sal, but the tomato area has been considerably higher. Yields of rice have been significantly higher than assumed at appraisal, but yields of tomatoes have fallen and are now lower than expected (PPAM para. 6 and PCR paras. 4.04, 4.05 and 9.01). One thousand six hundred farm families benefited directly from the project compared with 1,000 envisaged at appraisal. While incomes reached only 60% of appraisal estimates, they were three times as high as those obtained under rainfed conditions and are expected to improve overtime. The re-estimated economic rate of return according to the PCR is between 4% and 12%, compared with 14% estimated at appraisal; this decline is mostly due to increased costs. However, in the audit's opinion the reesti- mated ERR is too high because of unrealistic assumptions about the amount and cost of project works rehabilitation. Under more realistic assumptions, the audit's estimate of the ERR ranges from negative to 6% (PPAM paras. 21-25). -v - Other points of interest are: - project may have increased income disparities (PPAM, para. 38); - supervision of project by the Bank was not adequate (PPAM, para. 33); - Bank's response to changes in project design and subsequent cost increases was deficient (PPAM, paras. 30-33); and - lessons learned from project experience resulted in the grant- ing of an engineering credit (Cr. S-18-SE) for designing of further irrigation development in the region (PCR, para. 8.01). PROJECT PERFORMANCE AUDIT REPORT SENEGAL RIVER POLDERS PROJECT (Credit 350-SE) Table of Contents Page No. Preface ........................................................ i Basic Data Sheet ............. ............. ...... ........... .ii Disbursement Table ..iii Highlights n .Tale......................... -oo.............. iv PROJECT PERFORMANCE AUDIT MEMORANDUM I. Summary ........................ ..................... 1 II. Main Issues ........................................... 3 A. Project Design, Implementation and Maintenance ............ 3 B. Economic Analysis .............................. 6 C. Project Cost and Financing .. 7 D. Bank Performance............... .. . 8 E. Farmer Organizations and Equity Considerations... 9 III. Conclusions ...... ................ 11 PROJECT COMPLETION REPORT I. Background .... .... ... . . . ..... . . ...... ........ 13 II. Preparation and Appraisal .............................. 14 III. Implementation ......................................... 17 IV. Agricultural Impact .................................... 22 V. Economic Analysis .................... ................ 24 VI. Performance of Consultants, Contractors, Suppliers and Borrower .............................. 25 VII. Bank Performance ..............................35 VIII. Debi-Lampsar Engineering Credit (Cr. S-18 SE) .......... 37 IX. Conclusions ........... ............ ................... 42 Table of Contents (Continued) Page No. Annex 1 - List of Main Covenants .............................. 37 Annex 2 - Disbursements (actual and appraisal) ................ 39 Annex 3 Table 1 - Agricultural Impact ........................... 40 Table 2 - Farm Budgets .................................. 41 Annex 4 Table 1 - Economic Rate of Return ..................... 42 Table 2 - Price Structure of Rice ....................... 43 Table 3 - Price Structure of Tomato ................... 44 Annex 5 - Construction Schedules ........... ............ 45 Annex 6 Table 1 - Disbursements (actual and appraisal) ... 46 Table 2 - Construction Schedules ........................ 47 Maps PROJECT PERFORMANCE AUDIT MEMORANDUM SENEGAL RIVER POLDERS PROJECT (Credit 350-SE) I. SUMMARYU1 1. The project was identified by a FAO/IBRD Cooperative Program (CP) mission in 1969, prepared by PMWA2/ and CP, and appraised in March/ April 1972. The Credit was approved in December 1972; it was the fourth agricultural project and the first irrigation project in Senegal to be financed by the Association. 2. The project aimed primarily at increasing rice production through the construction and agricultural development of three irrigated sub-projects: Debi (1,025 ha) and Lampsar (1,080 ha) in the Senegal delta, to be equipped with flood control systems; and Dagana (2,730 ha) in the lower valley, to be equipped with a full water control system. Miscellaneous vegetables, includ- ing tomatoes to be processed into tomato paste - for which there is a strong demand in Senegal - would also be grown at Dagana. Some 1,700 rural fami- lies at subsistence level would benefit from the project. The severe drought in the African Sahel in 1972 and 1973 showed that flood control irrigation was not reliable, being too dependent on the vagaries of river flooding. Thus Government decided that irrigation would be developed through full water control systems only. In 1974 the Association agreed to reformulate the project by dropping construction of the Debi and Lampsar sub-projects; the entire credit would be allocated to construction of the Dagana perimeter, where a cost overrun was expected. The Association also agreed to finance through an engineering credit the new design of the Debi and Lampsar irri- gation systems. 3. The project authority was the "Societe d'Amenagement et d'Exploita- tion des Terres du Delta du Fleuve Senegal" (SAED). SAED was created in 1965 by the Senegalese Government with substantial assistance from France to increase the local production of paddy through the construction of irri- gated schemes in the Senegal Delta. By 1972, SAED had developed some 9,000 ha through flood control irrigation. It was managed by a competent Director General assisted by two French consulting firms which provided technical specialists to fill key in-line positions and provide the specific expertise required to operate such a development authority. It was and still is a state institution controlled by two ministries: Rural Develop- ment and Finance. It is subject to cumbersome administrative and financial procedures involving central Government services in Dakar and has no financial .autonomy. 4. The cost of the project was estimated at US$7.4 million. Base costs for irrigation works, which represented some 75% of total costs, were correctly estimated at appraisal. However, largely because of long delays in implementation and quantity and price increases, the actual cost amounted to US$14.4 million, i.e., nearly double the original estimate. The cost 1/ Adapted from the PCR. 2! Bank's Permanent Mission to Western Africa. - 2 - per ha of US$4,200 is still reasonable by West African standards. Government made up for the cost overrun with great difficulty, which explains the delays in implementation. Its contribution to the financing has increased from 40% at appraisal to some 60% of total costs. Government's difficulty in securing the required financing is illustrated by the fact that, although IDA financing was fully disbursed in March 1976, the irrigation works were still not completed in March 1979. 5. Irrigation works were carried out by contractors, selected follow- ing international competitive bidding, under the supervision of a consulting firm assisting SAED. The quality of construction is good. But difficulties in securing financing, combined with SAED's cumbersome procedures, have resulted in work interruptions, delays in payment, and withdrawal by contract- ors. The works are deteriorating because of lack of maintenance and some weaknesses in the overall design of the perimeter; measures are expected to be undertaken to rehabilitate them. Buildings are not yet completed. 6. The development of paddy-cultivated areas was slower than expected, with only 1,000 ha in 1978, as compared to an estimated 1,400 ha; however, yields were considerably above, and overall production is close to, appraisal estimates. Double cropping of paddy did not materialize at all. The cultiva- tion of tomatoes was encouraged by Government: 520 ha were cultivated in project year two with a satisfactory yield. However, this level has not been sustained because of poor agronomical practices; and cultivation of other vegetables took place only on an experimental basis. The tendency has been toward overmechanization, resulting in increased production costs; on the other hand, the farmers' response to the project has been good, and the size of the plots allocated per family had to be reduced from 3-2 ha to only 1 ha. As a result of these two factors the net farm income is less than anticipated, but the return per man day is higher, and the project now benefits 1650 farm families instead of the 1000 estimated at appraisal. 7. The cumbersome administrative and financial procedures have caused delays in project execution, have not been properly enforced by the public accountant in charge, and have led to a deterioration of SAED's accounting system and financial management. This has been all the more critical in view of SAED's rapid expansion. Between 1972 and 1978, SAED's assets increased from CFAF 1.3 billion (US$5.9 million) to CFAF 13 billion (US$59 million), and its personnel from 200 to 990. 8. The Association insisted that measures be taken to improve SAED's efficiency; a number of studies were made regarding a change of status and reorganization. All parties recognized the need for changes, but none con- vinced Government to take radical measures to change SAED's status. A decree on SAED's reorganization was issued recently. This is a first step towards decentralizing SAED's activities, but does not solve the problem, as SAED's day-to-day management remains dependent on central Government services. The situation became crucial in 1978 when the ongoing campaign was jeopardized merely because of the replacement of the Director General without an appro- priate transition period. The major aid agencies involved, IDA, French AID - 3 - (FAC) and Caisse Centrale de Cooperation Economique (CCCE), then decided to reassess SAED's performance and make joint recommendations to Government. A joint mission took place in October 1978, and the report and recommendations were discussed in Dakar in March 1979. 9. The Economic Rate of Return (ERR) has been estimated to range between negative and 6% depending on the magnitude and timing of required rehabilitation works. II. MAIN ISSUES A. Project Design, Implementation and Maintenance 10. The audit supports the PCR's general conclusion that the "economic viability of the project remains acceptable provided that rehabilitation is studied and carried out promptly" (PCR para. 9.01). However, the actual physical condition of the irrigation works is described too briefly by the PCR (para. 6.06, iv): , "Lack of maintenance has resulted in a deterioration of the works which is already quite obvious and a matter of a grave concern." The audit agrees with this judgment but thinks that earthwork erosion is not the only matter of concern. As an SAED report.L/ states the project's sophisticated hydro-mechanical equipment never really functioned and also needs rehabilitation. 11. The PCR attributes the deterioration of structures (para. 9.01) mainly to the lack of maintenance by SAED and lack of supervision by consul- tants, but does not question the quality of construction by contractors nor the soundness of the project's design. 12. It is correct that almost no maintenance was undertaken by SAED, which had practically no maintenance facilities until 1978, but this does not mean that there was no maintenance at all. The PCR (para. 6.06, iv) recognises that "when a specific problem arose SAED relied on contractors' work to correct it." This point is supported by the SAED report: "The abnormally long presence of the contractor responsible for terracing has contributed much to the survival of improvement works (rupture and modifi- cation of the canal D dikes, rehabilitation of gate B77, cleaning of the supply canal, station B, levelling). It was thus through hiring of equip- ment on force account that the irrigation area received the most urgent repair work. It can be considered that all works carried out and invoiced on force account are chargeable to maintenance of project works."!/ Unfor- tunately the mission did not receive all information needed to establish the proportion of project costs which were in fact for maintenance purposes. 1/ SAED, Division des amenagements, BEP - Amenagement du perimetre Vallee- Dagana. Note de synthese, Oct. 1978, 12 p., p. 5. 2/ SAED. Note de synthese, p. 8. 13. These maintenance efforts were obviously insufficient, but the bad condition of earthworks cannot be blamed on the lack of maintenance alone. The collapse of part of D dike and the percolation which can be observed along dikes of sector C which were not yet in operation raises questions about construction quality and soundness of the engineering design. 14. The reason for D dike's partial collapse is still a matter of argument between the contractor and SAED. The quality of the earthworks as carried out by the contractors can now be judged from their performance. The extensive erosion and percolation problems which can be observed prove that either the soils used for the embankments were unsuitable or the compact- ing techniques were inadequate. The initially poor quality of supervision (PCR 6.01) does not allow a definite judgment on the quality of compaction. However, it should be noted that quality control of project works was not facilitated by the fact that the irrigation blocks were allocated to farmers as soon as construction was completed without a final project acceptance of the works. 15. Shortcomings of project design are even more obvious. The insuffi- ciencies of geotechnical, pedological, and topographical surveys are men- tioned in the SAED Report!i and were confirmed to the audit mission by the authority. As a result of poor preliminary investigations 270 ha of a sandy area had to be abandoned, and rice is grown on a saline area around canal D, previously planned for tomato cultivation. The 30 1/s water modules suitable for tomato irrigation modules are now by-passed by farmers and are useless. These soil problems were inadequately studied at appraisal. 16. The need for raised embankments in sandy areas was also not antici- pated. The damage done to dikes by crossing cattle (PCR 6.01) could have been avoided if project design had taken note of the importance of livestock in the project zone.!/ Fencing of the project zone as suggested by some staff would have been an expensive and unrealistic solution because canal D was the only water supply available in the area during the drought, and the loss of livestock would have been worse if access to the canal had been blocked during that period. 17. However, these questions were not addressed in the Debi-Lampsar appraisal report./ SAED management had to take the initiative and draw the appraisal team's attention to the need for cattle tracks and livestock water- ing points. Even if livestock is not directly included in the project, it 1/ op. cit. 2/ Detailed suggestions on this subject are contained in Diarra, Definition d'une etude complementaire sur la Vallee du Lampsar. BEP, Agro-economic. 3/ Credit 775-SE of March 1978. - 5 - has to be taken into account in project design at least as a constraint; despite the importance of livestock in the project area, the follow-on project was also not designed as an integrated project addressing itself to the needs of both agriculture and livestock. 18. Other shortcomings in project design are evident in the hydro- mechanical equipment used for water control. The sophisticated constant downstream water level gate system selected was theorically the best in terms of costs and water economy. But it will only work if a tight control of water levels in the main canals (+ 5 cm) can be maintained. The fact that beyond these narrow limits the irrigation system does not work at all forces farmers to by-pass the sophisticated modules system thereby contributing to an accelerated deterioration of the tertiary irrigation system through uncon- trolled water flows. The problems were aggravated by the fact that some structures were underdesigned (sedimentation tanks, modules for tomato blocks) were never correctly set (automatic gate of channel A3) or never became operational (discharge regulators at pumping stations).1/ 19. Since this project was the first fully controlled irrigation scheme in the delta area and experience of local conditions was lacking (PCR para. 6.01) project design should also have considered the social, economic and institutional environment of the project. The lack of experience of SAED up to appraisal time should have cautioned the mission and led them .to consider an irrigation system which could function on a less sophisticated basis, to anticipate possible adverse conditions and to allow adequate safety margins.2/ 20. The combination of sophisticated modules with unlined canals became a source of problems from the beginning, problems exacerbated by the lack of maintenance of the irrigation network which would have been essential to guarantee the upkeep of constant water levels. However, maintenance facili- ties were not included in the project.3/ On the other side the pressure to 1/ SAED. BEP/Agro-economie - Definition d'une etude complementaire sur la Vallee du Lampsar. 2/ The Region and CPS argue that project design took account of the "social, economic and institutional environment of the project," although all the right answers were not found. 3/ Central Projects Staff notes it was intended at appraisal that mainte- nance works for canal drains and dikes would be provided by SAED through local contractors. This was proposed by SAED and their consultant and accepted by the appraisal mission on the arguments that it (SAED) did not want to try to develop this expertise at start-up due to their organiza- tional problems and, besides, there were many competent local and foreign contractors in the area with the equipment and trained men needed. In view of the management and funding problems which developed, providing O&M equipment through the credit probably would have had little positive effect on the level of maintenance. - 6 - keep the costs of the project's infrastructurel/ at low levels led to the construction of less robust project works, which need continuous and intensive maintenance. From this problem a lesson has been learned by the Bank, which now considers maintenance equipment as a vital project component. Maintenance equipment worth US$190,000 will be financed under the Debi Lampsar project (Credit 775-SE). B. Economic Analysis 21. The PCR's rate of return analysis (PCR para. 5.01) contains a very important assumption which deserves some comments. The assumption is that "costs include estimates of CEAF 260 millions for rehabilitation...rehabili- itation is assumed to take place in 1981 and 1982". 22. The critical importance of maintenance for extending project life is discussed above. But the project will also require substantial rehabilitation in addition to increased maintenance activities. However, in light of the substantial undertaking required, the PCR cost estimates for rehabilitation are on the optimistic side. In Decemer 1978 SAED management estimated costs between CFAF 500 million and CFAF 1 billion depending on how soon rehabilita- tion could be started; the longer the delay, the higher the cost. The poor state of canal D shows that it would be unwise to wait until 1981-82. 23. The ERR calculation should therefore include a sensitivity analysis to take into account different estimates of rehabilitation cost in relation to project life and total acreage likely to be served by the project. Under various assumptions the ERR would be: a) with rehabilitation limited to canal D, SAED cost estimate: CFAF 100 million project life 10 more years, command area: 1,000 ha. ERR: negative; b) rehabilitation of the entire scheme cost estimate: CFAF 500 million project life 25 years - command area: 2,500 ha. ERR: 4 to 5%; c) rehabilitation of the entire scheme cost estimate: CFAF 1 billion project life 35 years, command area: 2,500 ha. ERR: 5 to 6%. 24. The PCR's estimate of an ERR between 4% and 12% is too optimistic and in fact evaluates a follow-on project (with rehabilitation) and not the actual project (without rehabilitation). Under actual project conditions, 1/ As compared with the Nianga project financed by FED (Common Market Aid). - 7 - i.e. poor maintenance and no lining or improvement of canal D, SAED manage- ment estimates supported by a consultant are more pessimistic but must be taken into consideration. The acreage would be limited to the presently served 1,000 ha and project life would not exceed 5 years. The ERR would then be negative. 25. Results from this project confirm the Bank's experience that the ERR of irrigation projects is heavily dependent on the economic life of project infrastructure, which in turn is closely linked to the quality of maintenance. However, the experience has not been applied to project design, since financing of equipment required for proper maintenance was not included in the credit. C. Project Cost and Financing 26. The project encountered a total cost overrun of about 95%, and a time overrun of 200%. PCR cost estimates are based on a March, 1978 SAED report. From a later SAED reportl/ total costs of irrigation works alone amount to CFAF 2758 million, total project cost can be estimated at CFAF 3,163 million, as compared with the appraisal estimate of CFAF 1,617 million. Senegal's financial contribution stands now at about CFAF 2,000 million or about 65% of total costs compared to 40% estimated at appraisal. 27. The PCR's cost analysis (para. 3.14) shows that the main reasons for cost overruns were price increases accounting for about 60%, and increases in the quantity of earthworks accounting for about 40%. Price increases due to inflation were further compounded by long delays in project implementation. The PCR discussion of the causes leading to the delays, however, merits some additional comments. 28. The PCR (para. 6.06) evaluates government performance in project financing and states that "by 1979 Government will have financed CFAF 1,8 billion2/ (US dollar 7 million) i.e. more than twice its original contribu- tion. Given the difficult financial situation of Senegal in recent years, this indicates the Government's strong support of the project, implies dra- matic efforts and should be considered satisfactory despite the fact that this financing was not provided on a timely basis." There was a one year delay in mobilizing government funds. This appreciation seems to be in contradiction with the PCR's para. 7.05: "in conclusion, while the project did not achieve all the appraisal objectives and suffered from substantial cost overruns, the responsibility lay more with government bureaucracy and the lack of local funding than with actions or omissions by SAED or the Association." 29. The responsibilities of the Bank as the Senegalese government's partner in this innovative project are not properly emphasized by the PCR. 1/ October 1978. 2/ In fact SAED estimates show that it will be close to 2,0 billions. - 8 - The Bank's financial responsibility is only suggested (para. 7.04). "Had the Association agreed to cover a substantial part of the then modest cost overrun, project implementation would probably have proceeded with less delays, lower total costs and consequently, higher benefits." The modesty of the cost overrun at the time, and the prospect of an indirect contribu- tion through a program loan were reason enough for the Bank not to consider supplementary financing. But it is more difficult to understand why, in 1975, the considerable cost overrun due to changes made in construction design of the earthworks for embankments was not considered a case for reviewing project funding. Cost of this undertaking was about 2.2 times appraisal estimates and was re-estimated at CFAF 600 million (US$ 2,5 million). It should therefore have been obvious that this new financial charge, to be borne by the Govern- ment alone, would lead to longer delays in project implementation and con- sequently to higher costs because of inflation.1 D. Bank Performance 30. The audit strongly supports part of the PCR conclusion (para. 9.01, (V)) that "IDA supervision was probably less than warranted by a project with such intensive and innovative features", but disagrees that..."it is unlikely that increased IDA supervision would have resolved the major difficulties of the project i.e. institutional problems and the lack of local financing." Increased IDA supervision may not have resolved the difficulties encountered in mobilizing local funds but it should have warned the Bank about possible consequences of cost overruns caused by increases in construction quantities and should have contributed to convincing the Bank about the need for addi- tional financial support to the Government. 31. The best way to reduce project delays would have been to help the Government to support cost overruns and to link this financial support to elimination of the a priori financial control of SAED. Under the existing financing arrangements SAED had to submit disbursement claims to a lengthy clearing process by Government offices prior to receiving the requested funds. 32. Maintenance problems and the need for rehabilitation are not dis- cussed by the appraisal report of Debi Lampsar (February 1977) which mentioned a revised rate of return for Dagana, at "about the appraisal estimates figures of 14%."_V None of the supervision missions raised the rehabilitation issues. These problems are first mentioned in the PCR which was prepared by the Regional Office with the assistance of a consultant following an in depth field visit. Had these problems been identified earlier and action initiated the cost of rehabilitation could have been kept lower. 1/ The Region points to the Bank's general policy against financing cost overruns and denies responsibility for the implementation delays that made much of the overruns necessary. 2/ Debi Lampsar Appraisal Report para. 2.24. -9 - 33. Supervision reports emphasize the cumbersome and inefficient admin- istrative and financial procedures that the Government imposes on SAED and recommend a change of SAED's legal status, from a public company to a semi- public agency. This drastic change of SAED's legal status as recommended has not yet been accepted by the Government. Government officials suggested in discussions with the audit mission that the above mentioned change of SAED's legal status may not be the only solution. From the Bank's point of view the main reason why there should be a change of legal status is in fact the need to eliminate the cumbersome "ex-ante" financial control. The audit mission's discussions suggest that the Bank did not sufficiently explain to officials concerned the Bank's position on "ex-post financial control"..1 E. Farmer Organizations and Equity Considerations 34. The physical problems faced by the project overshadowed some social problems. Like other irrigation schemes in the Sahel, the project has indi- rectly benefited from the drought because SAED was obliged to reduce individ- ual holdings and to increase the number of beneficiaries. Although there are no apparent social problems in the project area, SAED experiments with farmer organizations and their socio-economic impact deserve some consideration. 35. As with other Bank financed projects in SenegaL-2/ the existing cooperative structure was not considered a suitable farmer's organization and SAED has introduced the "groupements de producteurs (GP)" on an experi- mental basis. These groups of 12 to 15 family heads are considered by SAED as production units. Each GP is represented by its chief, who is the only one directly dealing with SAED staff in regard to input supplies, credit recovery, and marketing. To meet ONCAD-3/ requirements the GPs are incorporated in two cooperatives. 36. Following a request from farmers the cooperatives were authorized to acquire combines. The wisdom of this investment can be questioned but first results of this experiment show that farmer organizations, under pre- vailing conditions, are doing a better job than SAED in maintaining agricul- tural equipment. From these fine results accomplished in a difficult and delicate task the audit concludes that cooperatives may also handle some primary marketing, input supply and water management functions which are 1/ The Region states that efforts at highest Government level produced very little results, because Government is not favorably inclined to relax rigid ex-ante controls until it is satisfied that the institutions are sufficiently well staffed and managed. 2/ Casamance Rice Project (Credit 252-SE). OED Report 2056 dated May 15, 1978. Terres Neuves Resettlement Project (Credit 254-SE). OED Report 2589 dated June 29, 1979. 3/ Office National de Cooperation et d'Assistance pour le Developpement. Agency responsible for sponsoring cooperative development. - 10 - presently the responsibility of SAED's field staff. If field staff were relieved from these functions they could then concentrate more on their advisory, and technical assistance role in support of the farmer. 37. While GPs appear to be ready to take on more responsibilities, their impact on equitable distribution of project benefits is open to question. From a 1976 SAED survey it appears that the status of members within a GP closely follows the patterns of local feudal societies. Most of the members of a GP come from the same extended "noble" family and 94% of GP chiefs are descendants of noble families. As observed in other development projects, farmer organizations tend to reinforce existing social structures if there is no counter action from project staff to establish more democratic organiza- tions. 38. Experience from this project shows that reliance on existing social structures replicated in the organization leads to stable institutions with high internal cohesion, able to resolve their problems without staff interven- tions. The acceptance of traditional social structures, however, has also its price: only 14% of project participants belong to the lower income strata. SAED staff believe that the project will benefit mainly the wealthiest and most influential families in Dagana. The influence of civil servants, veter- ans and more generally of absentee land owners whose main activity is not agriculture should also be mentioned (almost 50% of the GP chiefs belong to this category). Their lack of time for manual harvesting may explain the cooperative's decisions to acquire combines. 39. The evolution of the project unit at Dagana is closely related to the Government policy vis-a-vis farmer organizations. If the decision, as it seems to be, is to promote autonomous and fully responsible farmer organizations, then the project unit has to be conceived as a training and technical assistance unit to assist cooperatives and GPs and not as management unit. This decision has important implications for staff qualifications. At present project staff would not be able to have any major influence on improv- ing farmers' technologies as long SAED is unable to assure efficient technical operation of the irrigation system. Considering the technical problems now encountered, extension staff has little impact at farm level and frequently only follows up on innovations introduced on the farmers' own initiative. Consequently scheme rehabilitation appears now to be a prerequisite for an efficient project unit as well as productive farmer organizations. III. CONCLUSIONS 40. The fact that this project was the first irrigation project with full water control in the Senegal delta explains many of the technical prob- lems encountered. The fact that it was the first irrigation project to be financed by the Bank in Senegal, however, does not explain the shortcomings in Bank supervision and financial support. After showing some reluctance to get involved in the project the Bank has tried to keep the initial cost of the project as low as possible. Project experience shows that "savings" in beginning an experimental project proved to be very expensive at the end. - 11 - 41. The Senegal Government, which has financed more than its initial share of project costs, may be entitled to feel that it was not fully sup- ported by the Bank with an experiment which was of interest to the entire River Basin Authority.!Y This experimental project should have received more attention by the Bank through more intensive and continuous supervision and support. At this stage, it is not certain that all the project's lessons have been learned and are applied to other schemes in the Senegal Valley and more specifically to the Debi Lampsar project.2/ 3/ 1/ Organisation pour la Mise en Valeur du Fleuve Senegal (OMVS). 2/ Central Projects Staff comments that, to be useful, more specific recom- mendations are needed. The local financing and management problems were identified at appraisal and were never resolved. Should the Bank have imposed a disbursement moratorium until these problems were resolved while working with GOS to find solutions? Or was the Bank decision to press for the needed changes without suspension of disbursements (unfor- tunately without success in this case) a prudent alternative? These types of problems are not unusual in Bank projects, and it is difficult to predict in advance which approach might be more productive. Even now, with hindsight, it does not appear that one could say with assurance that a moratorium would have produced a better project. 3/ The Region states that it objects to this statement entirely. The on- going project is being implemented taking into account the lessons learned and IDA and Government have agreed on ways to improve SAED's organization and management. On the other hand, Government has not yet taken any action to relax its cumbersome financial control or to improve SAED's financial situation.  - 13 - PROJECT COMPLETION REPORT SENEGAL RIVER POLDERS PROJECT (CREDIT 350-SE) I. BACKGROUND 1.01 The project is located in the Sahelian part of Senegal, along the Senegal River in its lower valley, some 130 km inland from the city of Saint- Louis. The Association approved a credit of US$4.5 million on January 9, 1973 for the project; this was the Bank Group's fourth operation in Senegal's agricultural sector and its first irrigation project. The other Bank Group operations in agriculture were: in 1969 (i) a US$6.0 million IDA credit (Credit 140-SE) and a US$3.5 million Bank loan (Loan 584-SE) for agricultural credit to help finance a groundnut and millet improvement program; in 1971 (ii) an IDA credit of US$1.35 million for a small settlement project, the Terres Neuves Project, (Credit 254-SE) and (iii) a US$3.7 million IDA credit for rice development in Casamance (Credit 252-SE). 1.02 The Senegal River region offers the largest potential for irrigated -agriculture in Senegal, some 300,000 ha. Senegal imports large quantities of food staples, including some 200,000 tons of rice which could be produced in the region. But the region's potential is almost untapped. To increase local rice production, Government decided in 1965 to create a special development agency called the "Societe d'Amenagement et d'Exploitation des Terres du Delta du Fleuve Senegal" (SAED). In 1972, when the project was appraised, some 9,000 ha had been developed in the delta, and Government strategy was to ex- pand irrigation in the river region. The Senegal River Polders Project fits into that strategy. 1.03 The objective of the project was threefold: (i) to increase local rice production by 8,700 tons, and that of tomatoes by 2,100 tons through the development of three irrigated perimeters totalling 4,835 ha: Debi (1,025 ha), Lampsar (1,080 ha), and Dagana (2,730 ha); (ii) to improve the living condi- tions of 1,700 farm families; and (iii) to experiment with crop diversifica- tion and intensification in the project area. The project was to reach full development over a period of nine years. 1.04 Total project cost was estimated at US$7.4 million (CFAF 1.9 bil- lion), net of taxes. The IDA credit totalled US$4.5 million, and Government was to finance the remaining US$2.94 million (CFAF 752 million), 40% of total costs. When the project was reformulated in 1974, construction of the Debi and Lampsar perimeters was abandoned to allocate all available resources to the development of the Dagana perimeter. The project experienced considerable cost overruns and delays: although the IDA credit has been fully disbursed since March 1976, completion of the Dagana perimeter is now scheduled for June 1980. Therefore it was decided to produce the present completion report by early 1979 instead of mid 1978 as a compromise between the usual six months after the closing date of the credit and the actual completion of the project. This report was discussed with SAED and Government officials during a mission in November 1978. - 14 - II. PREPARATION AND APPRAISAL Chronology 2.01 A mission of the Bank's Permanent Mission in West Africa (PMWA) visited the area in January 1969. At that time all irrigation works carried out by SAED in the delta were designed to provide flood control rather than full water control irrigation; the availability of water could not be guar- anteed as it was dependent upon the intensity of flooding of the Senegal River. SAED, assisted by French technical assistance (FAC), intended to implement on an experimental basis, a full water control system at Dagana which would guarantee water availability through pumping from the river, in addition to the construction of the Debi and Lampsar perimeters by means of traditional and relatively cheap works. In January 1969, a PMWA mission concluded that neither the pumping works nor the "tertiary" irrigation systems were economical, 1/ and that should FAC decide to go ahead with the construc- tion of pump-irrigated works, there would be no project in the existing polders suitable for Bank/IDA financing. 2.02 Following an FAO/CP identification mission in April 1969, preliminary studies were prepared by Government with FAC financing to develop 4,100 ha to produce paddy and tomato paste, and reduce imports accordingly. In June 1969, a CP officer recommended that a follow-up mission assess in depth the justification of the project by October 1969, define additional studies required to fill information gaps and initiate preparation of a feasibility study. 2.03 In 1970, the Bank was still reluctant to finance the project and would have preferred that FAC take the initiative in experimenting with the new type of irrigation system in the area. Government prepared the feasibility study, with the assistance of a consulting firm. Terms of reference of the consultant were reviewed by PMWA, together with the final engineering studies of the irrigation works. 2.04 The feasibility study was issued in January 1971 and communicated to the Association during a mission in March together with an official request for financing. The following issues were raised by the Association, during a May mission, after a careful review of the study: (i) Irrigation versus rainfed agricultural development: this was not new but was in fact a key question as the cost per ha of some US$1,200 was considered excessively high at that time; (ii) Accuracy and reliability of certain data; (iii) Organization and management of the project: the information provided was considered incomplete; and (iv) Pace of development of tomato cultivation: this was considered overly optimistic by the Association. 1/ Office memorandum dated June 20, 1969. - 15 - The mission asked for additional information regarding these issues and recommended that a less costly alternative be envisaged. 2.05 Appraisal took place in March 1972 when preliminary engineering studies of the three perimeters, an agro-economic study, and the feasibility study were made available to IDA. The mission was composed of an economist, a financial analyst and an irrigation engineer, and a financial analyst from PMWA who had already been involved in project preparation. The main issues mentioned in the correspondence shortly after appraisal refer to (i) the use of Senegal River water and the need to have an agreement from the states of Mali and Mauritania for such use; (ii) taxes and duties on equipment; and (iii) topographical works for which the Association wanted Government's confirmation that it would prefinance them. The question of the use of Senegal River water was debated during the entire appraisal: Senegal in- sisted that formal approval from the two states, which the Association con- sidered a Board presentation condition, was not necessary, and that a simple notification to these states of Senegal's intention to develop the project would suffice. The condition was finally waived on the basis of minutes of discussions which had been held on the project within the framework of OMVS in its July 23, 1972 meeting. 2.06 Negotiations took place on November 6, after the consultant was called to Washington to review the concept of the development of the Lampsar perimeter which was questioned during review of the green cover appraisal report. Financial arrangements constituted the main topic of discussion during negotiations as Government did not agree to give financial autonomy to SAED. Finally, IDA accepted that central Government services would con- tinue to manage SAED's finances under a financial agreement to be entered into between Government and SAED. 2.07 Board presentation took place on December 7, 1972. The Credit and Project Agreements were signed on January 9, 1973. Targets and Goals 2.08 SAED was created in 1965, primarily to produce paddy in the Senegal River area through irrigation. Relatively cheap flood-controlled perimeters had been constructed in the delta, although most of the irrigation potential was located in the valley (but at a higher investment cost). The Dagana perimeter was the first attempt to develop such potential. As the investment per ha was high, production on the perimeter had to be maximized to ensure its financial viability. Thus, in addition to paddy production on these polders, the Dagana perimeter would grow sorghum and miscellaneous vegetables. Special attention was paid to tomatoes as a strong demand for tomato paste existed in Senegal and a tomato paste factory was in operation in the delta. Double cropping of paddy was to be introduced progressively to prepare growers for the time when regulating structures on the River would be operational (the Diama ,and Manantali dams). The project, from an agricultural standpoint, was characterized by diversification and intensification of crops. As such it was innovative and difficult to execute. To ensure proper implementation, the - 16 - project was to include applied agricultural research to define ways of cul- tivating newly introduced crops, and training of Senegalese extension workers and officers. No action was proposed to strengthen SAED besides the project needs per se, although it was clear that the results achieved with the project would benefit the entire lower valley area. Thus, the project was not in- tended to have a significant institution-building impact. Project Definition 2.09 The appraisal report indicates that "Three subprojects would be the principal components of the project and would result in a net increase in irrigated land of about 3,050 ha and improvement of another 1,780 ha. The subprojects 1/ would be: (a) constructing a new irrigation scheme at Dagana permitting the year-round cultivation of 2,730 ha; (b) improving water control in the Debi cuvette and increasing the area irrigated from 260 ha to 1,025 ha; and (c) improving water control in the seven cuvettes forming the Lampsar subproject and increasing the area irrigated from 880 ha to 1,080 ha. (d) Complementary project activities would be: strengthening SAED with staff, facilities and equipment both to carry out the project and to provide support for farmers who would use land developed under the project; (e) constructing, equipping and staffing a training center at Dagana for SAED staff and for farmers; (f) carrying out, through collaboration with IRAT, a program of applied research to support agricultural operations at Dagana; and (g) expanding the SAED farm machinery pool to permit the provision of land preparation services to farmers using land developed under the project." Irrigation water would be provided by pumping or by gravity from the Senegal River, the water of which is of excellent quality except in June when exces- sive salinity becomes a problem. Production would consist of: paddy (8,064 tons), sorghum (4,008 tons), tomatoes (11,690 tons) and miscellaneous vege- tables (5,244 tons). The project would provide a better livelihood for some 1,700 farm families comprising about 10,000 people. 1/ See maps .IBRD 3984 and IBRD 10041. - 17 - III. IMPLEMENTATION Effectiveness and Start-up 3.01 The legal documents required the fulfillment of two conditions before the project could become effective: (i) signature of a financial agreement between Government and SAED satisfactory to IDA, and (ii) establishment of a contract between SAED and farmers' groups. It was expected during negotia- tions that three months would suffice for Government to meet these two condi- tions, and the date of effectiveness was scheduled for April 9, 1973. A draft financial agreement was prepared by the Association and transmitted to the Senegalese delegation during negotiations. The Credit was declared effective on June 8, 1973, two months after the original date of effectiveness, on the basis of verbal assurances from the May/June Bank mission that the final version of the contract between SAED and farmers was being prepared. The two month delay in effectiveness had no real impact on the start-up of the project. 3.02 In an attempt to expedite the start-up of works, tender documents -for irrigation works were issued in December 1972, immediately following Board approval. Bids were received in March 1973 and works were awarded in July. Contractors were notified in October. Irrigation works started in September 1973, slightly before notification of contractors but already seven months behind the schedule foreseen at appraisal. Ten months elapsed between the publication of tender documents and actual notification from SAED to start the works. Technical assistance started in October 1974 with the recruitment of the project manager. Start-up of most other project components was delayed until 1975: construction of buildings in April, procurement of agricultural equipment in March, procurement of vehicles and the first agreement between SAED and the Institut Senegalais de Recherche Agricole (ISRA) in June. Despite the efforts made to expedite the start-up of works, the long and cumbersome administrative procedures caused considerable delays. Revision 3.03 In July 1974 Government requested the Association to reformulate the overall project because (i) inflation and exchange rate changes had increased costs of the Dagana polder subproject alone to US$8.7 million, and (ii) the water storage of the proposed Diama Dam, which was expected to be in operation by 1980, would allow Debi and Lampsar to be used for double cropping. The Association approved the reallocation of the entire River Polders Credit to the Dagana polder (IDA/R74-88 of October 10, 1974), as well as an engineering credit of US$1 million (Credit S-18-SE, April 1975) to prepare the new Debi- Lampsar project. Progress of Irrigation Works 3.04 The actual implementation schedule of works as compared to the appraisal forecast is illustrated in Annex 5. Completion dates for the three sectors composing the perimeters and their pumping stations are given below: - 18 - Dagana Sector A Sector B Sector C Net Area /l (735 ha) (1800 ha) (425 ha) 2960 ha Irrigation works June 1975 June 1977 (June 1980) (March 1974) (March 1975) (Jan. 1976) Pumping station - civil engineering Oct. 1974 July 1975 Dec. 1975 - in operation Dec. 1974 July 1975 Oct. 1977 (Jan. 1974) (Jan. 1975) (Jan. 1976) /1 1978 SAED estimate, which represents a slight increase compared to the appraisal estimate of 2,730 ha. The completion date of irrigation works at appraisal was January 1976. Delays have increased over time: seven months at start-up, some one and a half years for completion of sectors A and B, and more than three years for completion of sector C, assuming that the remaining works are completed by June 1980. These delays are attributable initially to the cumbersome administrative procedures and subsequently to shortages of local funds once the IDA Credit had been fully disbursed in March 1976. Insufficient and intermittent local funding resulted in frequent work stoppages (para 6.02). Procurement 3.05 Irrigation works were awarded following international competitive bidding (ICB). Equipment for all other activities was procured through local competitive bidding (LCB), including agricultural equipment which was to be procured initially through ICB. This change was requested by SAED on the grounds that the equipment should be similar to that which it was already using and that the sum involved did not substantially exceed the upper limit accepted for LCB. The Association had no objection to this change. The Association reviewed the contracts and imposed certain changes primarily related to financial clauses to ensure compliance with guidelines for procure- ment. Contracts had to be cleared by representatives of both the Ministries of Rural Development and Finance and approved by the "Commission des marches" before contractors could be notified. This resulted in exceptionally long delays. For example, the amendment to the supervision of works contract, which prolonged the supervision period for a year, was signed after all ser- vices had been completed. Bids for vehicles were called for twice. There were several interruptions in irrigation works which were attributable to difficulties in local financing rather than procurement problems. These difficulties led to long delays which caused the contractor responsible for the construction of irrigation structures to pull out in 1978 without having completed the works. SAED had to file a suit against him. The contractor responsible for the construction of buildings went bankrupt and also pulled out. SAED was obliged to enter into new contracts to complete the remaining works. Some suppliers are now reluctant to deal with SAED as payments are not made on time. This basic problem in the procurement of goods affects most public enterprises in Senegal. - 19 - Disbursements 3.06 The IDA Credit was to be disbursed against: - the CIF cost of imported hydraulic and agricultural equipment; - 60% of the cost of irrigation works, construction of buildings, vehicles and operating costs; and - 100% of the cost of engineering services. In addition, it was proposed that topographic surveys be financed retroac- tively at 100% of cost. These percentages were not changed when it became apparent that the project would face a cost overrun, as Government agreed to meet the additional financing requirement for completion of the entire project. The chart in Annex 2 compares appraisal-estimated and actual dis- bursements of the IDA Credit. 3.07 As contractors were not notified before the end of calendar year 1973, first disbursements started in early 1974, i.e. 15 months after sig- nature of the Credit and Project Agreements. Disbursements remained behind schedule until May 1975, then accelerated, and the Credit was fully disbursed in March 1976, nine months before the expected date of project completion. 3.08 Payments for irrigation works were made on the basis of monthly statements of-expenses prepared by the contractors and certified by SAED. Actual payments to the contractor suffered delays, mostly because of SAED's failure to process applications to the Association. Most payments were made under the direct payment procedure. 3.09 Government had not included financing of the project in its 1973/74 budget of investments; therefore local funds were not available during the first year of the project. Governme.nt funds have since been channeled through the Banque National de Development du Senegal, which made the first disburse- ments in 1974. Disbursement of local funds has not kept pace with the progress of the works but has been made whenever financing was available. Costs 3.10 The appraisal report estimated net of tax costs at US$7.4 million, equivalent to CFAF 1.9 billion at the prevailing rate of CFAF 256 per US dollar. The cost estimate included 15% physical contingencies for irrigation works and a price contingency of 6% per annum compounded to all project costs. IDA was to finance US$4.5 million (CFAF 1.2 billion) or 60% of the costs, and Govern- ment US$2.9 million (CFAF 0.7 billion) or 40% of the costs. For Dagana alone, net of tax costs were estimated at CFAF 1.6 billion (US$6.2 million). IDA was to finance CFAF 1.0 billion (US$3.7 million) and Government CFAF 0.6 billion (US$2.5 million). The following analysis applies to the project as reformu- lated in 1974 (para 3.03). - 20 - 3.11 Appraisal estimates and actual costs for Dagana are indicated in the table on page 9. Ninety-five percent of actual costs were taken from contracts and subsequent amendments and include provisions for completion of works. The remaining five percent, related to operating costs and research, are estimates. 3.12 The table indicates: (i) An important total cost overrun: total project costs are CFAF 2.9 billion (US$12.4 million), which corresponds to an 83% increase over appraisal estimates. (ii) The appraisal estimate of base costs seems to have been reasonably accurate, and as indicated by the revision of June 1974, after the major contracts had been awarded, only civil works and hydraulic equipment increased by substantial amounts. (iii) Most cost overruns occurred between June 1974 and December 1975 and are related to irrigation worki. Earthworks-for embankmelts amount to 1,250,000 m , or 2.2 times the 560,000 m foreseen at appraisal. This unusually high ratio is due to changes made in the construction design of irrigation works during project implementation. The cost of quantity increases is estimated at CFAF 600 million, some five times higher than the physical contingency estimate at appraisal; (iv) Most price increases are also attributable to irrigation works, since the related contracts were awarded during the early stages of the project (late 1973), and implementation has suffered exceptionally long delays. Actual price increases are estimated at CFAF 1.0 billion, i.e. 5.6 times as much as the CFAF 0.18 billion expected at appraisal. In addition to delays in implementation, the average annual inflation rate of 12.1% that prevailed in Senegal for the period 1973-1978 has been twice the 6% assumed at appraisal. 3.13 In conclusion, the project is characterized by an important cost overrun which is due: 60% to price increases; 30% to quantity increases; and 10% to currency realignment (CFAF/US$). 3.14 In late 1974, when it became apparent that the project would be short of funds, Government requested the Association to finance the then estimated US$1.2 million cost overrun. The Association turned down the request because a program loan was under consideration which could have met the requirements. When the program loan failed to materialize, Government had to make up for the gradually increasing costs. Senegal's share in total project cost funding has increased from CFAF 752 million envisaged at appraisal to an estimated CFAF 1.8 billion at completion. - 21 - IV. AGRICULTURAL IMPACT Objectives 4.01 The objectives of the project included cultivation of paddy and sorghum during the rainy season, and of a second paddy crop, tomatoes and miscellaneous vegetables during the dry season. In practice, however, agricultural development did not meet the intensification targeted at appraisal (see Annex 3, Table 1). Crops, Acreages, Yields and Production 4.02 Rainy season paddy was cultivated on areas substantially smaller than appraisal estimates: 515 ha versus 1,330 ha in 1975, 708 ha versus 1,420 ha in 1976, 1,000 ha versus 1,420 ha in 1977, 1,080 ha versus 1,420 ha in 1978. The reasons for this are (i) a shortage of irrigable land because of delays in the construction of irrigation works, (ii) the use of large areas for tomato cultivation which, therefore, were not available during the rainy season, and (iii) abandonment of cultivated areas because of soil and irriga- tion deficiencies. Paddy yields as given by SAED (between 4 and 5.4 tons/ha) have been much higher than the appraisal objective of 3.2 tons/ha. However, these apply to averages on areas actually cropped, i.e. excluding those areas which may have been sown but were not cropped for various reasons. For instance, in 1977 only 695 ha were cropped out of 1,080 ha sown because an irrigation structure failed and some land was excessively saline and sandy. In this case, although the average yield was 5.1 tons/ha cropped, it was only 3.5 tons/ha sown. 4.03 Dry season cropping of paddy, envisaged at appraisal, has not been implemented because the rainy season paddy could not be sown early enough. The reasons for that are: (i) delays in the start-up of irrigation because the water of the Senegal River often remained salty longer than foreseen at appraisal, and (ii) delays in land preparation due to poorly operated agricul- tural equipment. If sowing is delayed, harvest will be also; this in turn, makes sowing of a second paddy crop before December 15 impossible. With later sowing, yields will be too low to make a second rice crop worthwhile. Consequently, it seems that under current conditions double cropping of paddy is virtually impossible and cannot be implemented before the regulation of the Senegal River by the OMVS dams. 4.04 Tomatoes were cultivated on a much larger scale than originally envisaged. This aimed at producing tomato paste for which there is a strong demand in Senegal and which could be processed in a factory located in the delta. The area under tomato cultivation reached 132 ha in 1975, 520 ha in 1976, 350 ha in 1977 and 450 ha in 1978, compared to appraisal estimates of 30 ha, 70 ha, 146 ha and 254 ha respectively. Yields were high during the first year (32 t/ha), then decreased drastically to 15 t/ha because of poor agronomic practices and in some areas soil salinity. It is essential that a sound rotation system be introduced to achieve the objective of 35 t/ha. Other - 22 - Estimated and Actual Costs in CFA million Appraisal 1/ Revised Revised Actual 5/ early 1972 June 1974 Dec. 1975 Mar. 1978 DAGANA Irrigation Works Earthworks 692 769 1,544 1,781 Civil Works 156 305 590 538 Hydraulic Equipment 112 242 250 240 Subtotal 960 1,316 2,384 2,559 Buildings 41 50 55 55 Agricultural Equip. and Vehicles 80 90 100 100 Engineering services 103 125 152 152 Management & Operating costs 80 103 42 42 Research 44 59 56 56 Compensation to Dagana 3/ Farmers 3 Base costs, Dagana 1,311 1,743 2,789 Contingencies 306 2 217 101 Total 1,617 2/ DEBI-LAMPSAR Base cost + Contingencies 286 -/ GRAND TOTAL 1,903 1,960 2,890 2,964 Exchange rate 256 240 240 240 US $ Equivalent, million 7.4 8.2 12.0 12.4 1/ Annex 9, Table 1 2/ Corrected for adding error 3/ Cancelled 4/ Sub-projects cancelled 5/ SAED report March. 1978 includes contingencies for completing of works. - 23 - problems that have plagued the tomato crop have been inadequate collection and marketing as well as low prices paid by SAED to farmers. Corrective measures are being discussed between the Association and SAED. 4.05 Production has reached satisfactory levels, as compared to appraisal objectives: in 1977/78, 3,544 tons of paddy and 6,975 tons of tomatoes were produced, equivalent to 98% and 91% respectively of appraisal forecasts. During the first year of the project production was higher than estimated, in particular that of tomatoes, which reached values not expected at appraisal (4,224 tons and 8,000 tons in 1974/75 and 1975/76, i.e. 5.6 and 4.3 times appraisal forecasts). This has been the result of high yields which were achieved at an early stage. Future production would depend on the envisaged rehabilitation without which appraisal targets would not be reached. Agricultural Equipment 4.06 Project equipment as of May 1978 consisted of eighteen tractors with implements, six trucks and two vehicles. At appraisal, a fairly high level of mechanization was assumed, which was consistent with the assumption made regarding double cropping. However, because of poor maintenance and lack of spare parts, this equipment is only partially operational, and plowing and sowing are often delayed. A major problem is the lack of a field repair shop. Repairs have to be made at SAED's central workshop. Farmers' Income 4.07 Because of the pressure of demand, beneficiaries have initially been allocated small farms of about 0.7 ha for paddy and 0.25 ha for tomato culti- vation. This is a substantial departure from appraisal assumptions of rice farms of 3.5 ha and farms of 2 or 3 ha for other crops (vegetables and sorghum). As a result net incremental family incomes as shown in actual farm budgets are lower then appraisal targets, but this is compensated by the fact that 1,650 farm families benefit from the project against 1,000 farm families anticipated at appraisal (Annex 3, table 2). Net incremental family income for 1978 reached CFAF 119,100 (CFAF 1,027 per man-day) for the typical rice and tomato farm, or about 60% of the appraisal estimate at full development. However, this revenue is about 3 times the revenue obtained in rainfed areas. With the anticipated increase in cultivated land and better yields, once the problems plaguing SAED have been resolved, farmers' income should substantially improve. 4.08 By allocating smaller farms, SAED chose to spread benefits to a much larger number of beneficiaries. Incremental net income was lower, but at the same time farmers were able to continue the traditional dry land cultivation of food crops which gave them additional income security. SAED's farmer settlement and price policies are currently under review by the Association and Government. - 24 - Water Charges 4.09 The water charges actually paid by farmers amount to CFAF 25,000 per ha of paddy and CFAF 35,000 per ha of tomatoes. These levies are ade- quately recovered by SAED when paddy or tomatoes are sent to the factories for processing. Credit recovery was over 87% in 1977. These amounts are adequate to finance operating and maintenance costs of the perimeter. How- ever, as discussed at Chapter VI, the perimeter has not been properly main- tained because SAED, chronically short of local funds, had to use these resources to finance its general budget. V. ECONOMIC ANALYSIS 5.01 The economic rate of return has been calculated with the following assumptions: - The project grows paddy and tomatoes. No sorghum or other vegetables; no double cropping of paddy (although double cropping would become possible when the river flow would be regulated by one of the OMVS dams). - The cultivated area is taken at 2,500 ha instead of the 2,960 ha developed, to take into account possible exclusion of saline and sandy soils. - Yields are estimated at 4.5 t/ha for paddy and 25 tons/ha for tomatoes. Without the project, paddy production of Dagana is estimated at 140 tons (0.9 tons/ha on 160 ha). - Economic prices of rice and tomatoes are calculated in 1979 constant terms. The price of paddy is based on Bank economic forecasts; the price of tomatoes is derived from the export price of tomato paste from Europe and trade forecasts analyzed in the Debi-Lampsar Irrigation Project (Appraisal Report No. 1685 a-SE). - The US dollar is valued at CFAF 220. - Costs are actual costs until 1978 and are expressed in 1979 terms; they include estimates of CFAF 260 million for rehab- ilitation and CFAF 90 million for completion of the irrigation works. Rehabilitation is assumed to take place in 1981 and 1982. - Labor is valued at an annual average shadow priced wage rate of CFAF 135 per man day; for a derivation of the shadow price of labor, updated to January 1979, see the Debi-Lampsar appiaisal, Annex 19. Labor requirements are based on the present situation which involves a high level of mechanization. - 25 - - There is no technical assistance involved in the project at the present time. A single expatriate was recruited during the first two years of project implementation and related costs are included in project costs. - Project life is 35 years. 5.02 With the above assumptions, the economic rate of return (ERR) is 8%, substantially lower than the appraisal estimate of 14%; the low ERR is due to cost increases combined with a reduction in benefits because of delays in im- plementation and reduction of the net cultivable area. 5.03 A sensitivity analysis was made to take into account the fact that: (i) the US dollar is undervalued in Senegal as compared to the CFAF by some 20 to 25%, according to recent economic work undertaken by the West Africa Programs Department. An average undervaluation of 22%, which corresponds to an exchange rate of US$1 = CFAF 270, has been taken into account in the calculations, (ii) the economic price of tomato paste has been in the past and is expected to remain volatile in the future; to take this element into .account, it has been assumed that benefits could fluctuate up or down by 10% or 20%. Results are the following: (a) Assuming a shadow exchange rate of US$1 = CFAF 270, the ERR would be 9%. (b) A decrease in benefits by 10% or 20% would result in a decrease in the ERR to 6% or 4%. (c) An increase in benefits of 10% or 20% would increase the ERR to 10% or 12%. This would also be the case if more land (para 5.01) were brought under cultivation through reclamation of saline soils or double cropping. VI. PERFORMANCE OF CONSULTANTS, CONTRACTORS, SUPPLIERS AND BORROWER Consultants 6.01 SAED engaged a consultant firm to prepare tender documents and detailed designs for major project works and to supervise their execution, as foreseen at appraisal. During project implementation substantial changes had to be made in the design of the irrigation works prepared by the consultant: (i) the size of embankments had to be increased to facilitate construction by standard earthmoving equipment and to improve the safety of the irrigation system; (ii) some structures were under or over designed; (iii) the use of constant downstream water level gates proved to be maladapted to the existing water management organization; and (iv) no provisions were made for cattle crossings. These shortcomings point to a less than satisfactory performance of the engineering consultants. A mitigating factor could be that this project was the first full control irrigation scheme in the Delta area and - 26 - experience of local conditions was lacking. The supervision of works was initially of poor quality (1974) with only one inexperienced man appointed, but improved after he was replaced by a two-man senior team (1975). The engineering contract was prolonged until November 1976, but as the progress of works did not justify a full-time supervision team, the same team was directed to supervise another project. At that time SAED decided that the supervision of works would be brought under the direct responsibility of its division of works. Thus the team was absorbed by this division, and the quality of supervision has suffered under this arrangement. Contractors 6.02 Irrigation works, including earthworks and construction of pumping stations and irrigation structures, were awarded to well qualified contractors. Overall quality was satisfactory: the only major problem was the uneven sinking of a pumping station floor, probably caused by inadequate compaction of the underlying fill. Despite adequate earthmoving equipment on site to construct works on schedule and substantial prefinancing by the contractors (some US$2 million), works progressed slowly because of delays in payments by SAED. This situation was aggravated after full disbursement of the IDA credit funds in March 1976, when progress of works became dependent solely on Govern- ment funding. 6.03 Construction of buildings was entrusted to a small local contractor who went bankrupt when only 50% of the works were completed. Since then, because of lack of funds these buildings have remained unfinished. At the end of 1978, SAED was still in the process of selecting another contractor. Suppliers 6.04 Agricultural equipment was provided according to contract specifica- tions. Delays were encountered in the delivery of a portion of the hydraulic equipment. This in turn postponed operation of a pumping station for some months. Late delivery of equipment did not affect the project. SAED and Government 6.05 Although Government and SAED are treated as separate entities in the Credit and Project Agreements, SAED's dependence on Government is such that in reality their respective performances in project implementation cannot be dissociated. SAED, as an "etablissement public", is supervised by both the Ministries of Rural Development and Finance, and for all practical purposes is a government department. 6.06 The performance of Government and SAED can be evaluated through a review of their role as regards the following aspects of project implementation: (i) Local financing: Government was to finance 40% of total project costs, i.e. CFAF 750 million (US$2.94 million). This amount was disbursed from the treasury by June 1976, after a one-year delay. By 1979 Government will have financed CFAF 1.8 billion (US$7 million), i.e. more than twice its - 27 - original contribution. Given the difficult financial situation of Senegal in recent years, this indicates Government's strong support of the project, implies drastic efforts, and should be considered satisfactory, despite the fact that this financing was not provided on a timely basis. (ii) Adjustment of SAED's operational procedures: Government has always been opposed to changing SAED from a public company to a semi-public agency with increased financial and managerial autonomy. As a result, administrative and financial procedures are cumbersome and inefficient. Under pressure by the Association and other donors, several attempts were made by Government and SAED to improve the situation, inter alia: studies on the reorganization of SAED by the Bureau d'Organisation et Metho (BOM), etc. The most recent attempt in late 1978 is a decree defining SAED's new organization policies, but in fact giving SAED very little additional autonomy. Implementation of the recently approved Debi-Lampsar Irrigation Project is facing the same difficulties. A joint IDA/CCCE/FAC mission made new recommendations to Government in March 1979, but the latter remains very reluctant to change SAED's legal status. (iii) Use of technical assistance: Since 1967, SAED has received consider- able technical assistance financed by French Aid and provided by two consulting firms, which also prepared the feasibility and engineering studies of the project and supplied the expatriate project manager during the start-up period. The technical assistance component of Dagana was purposely reduced in scale because of the massive French assistance already provided to SAED (30 expa- triates in 1978). But even so, not all the expertise called for at appraisal was provided because of recruiting difficulties (e.g., the training specialist). Moreover, the field expatriates were not used to the best of their potential by the Senegalese management and received inadequate support from their H.Q. offices. The Association's position was weakened by the fact that as it financed only a small part of the technical assistance, it could not intervene decisively. (iv) Failure to maintain irrigation works: In an arid area where the annual rainfall is some 400 mm,. it is extremely difficult for the natural vegetation to grow and protect irrigation works against erosion. Maintenance of such works is thus of paramount importance as they may deteriorate rapidly. This is all the more true for embankments protecting a perimeter against flood- ing such as the "Chenal D" of the project, where normal erosion due to rains is combined with erosion due to waves created by the wind (Harmattan) blowing on the surface of the water of the "chenal" and to cattle coming to drink from it. However, because of the financial difficulties of SAED, practically no maintenance of irrigation works has been done since their construction. When a specific problem arose SAED relied on contractors' work to correct it, or because of lack of funds, simply did nothing about it. This lack of mainte- nance has resulted in a deterioration of the works which is already quite obvious and a matter of grave concern. The project is slowly building up a maintenance facility but extensive rehabilitation is needed. The Association insisted that a diagnosis of repairs be made in 1979 and relevant costs be estimated. SAED would then propose that the lending agencies finance such rehabilitation. Rehabilitation would be carried out on all SAED perimeters, as they all suffer from a noticeable lack of maintenance, and its financing is of top priority. - 28 - 6.07 Probably one of the major sources of SAED's problems is its overly rapid growth from 1972 to 1978, as illustrated by the following figures: 1972 1978 Assets (CFAF million) 1,300 13,000 /a Budget (CFAF million) Investments n.a. 6,300 Operations 550 2,900 550 9,200 Cash deficit (CFAF million) n.a. 1,500 Staff 200 990 Developed area (ha) Limited water control 8,000 3,000 Full water control 0 10,000 /a Estimate. Assets have increased tenfold, operational budget and personnel expenditures have increased fivefold. All irrigation systems, except those of Debi and Lampsar, have been converted from flood to full water control. SAED now deals with some eleven different lending agencies. But, because of this expansion, problems which were minor in 1972 had grown to considerable proportions by 1978, and the Dagana project suffered in consequence. 6.08 It should be recognized that SAED's former Director General endea- vored to make the necessary organizational adjustments during this period of expansion, but very little could be done on the financial and administra- tive issues, given SAED's status as "etablissement public". By early 1977, when the Association appraised the Debi-Lampsar Irrigation Project, the financial situation had reached a critical point: SAED's cash deficit in the Treasury amounted to CFAF 1.5 billion (US$6.8 million). Replacement of SAED's then Director General by a new man in April 1978 exacerbated an already difficult situation. VII. BANK PERFORMANCE 7.01 Project preparation was initiated by Government with FAC financing. PMWA and CP were informed of the project in the early stage of preparation during an identification mission and subsequently assisted in project preparation. 7.02 In light of Senegal's long term objectives to develop the Senegal valley the appraisal objectives were sound. Cost estimates based on prelim- inary design proved reasonably correct. Major cost overruns are due to the sizeable increase in the volume of earthworks following construction design - 29 - (para 6.01) and inflation and currency realignment which could not be fore- seen at appraisal in 1972. The situation was exacerbated by delays in imple- mentation due to the lack of complementary financing, which was to have been provided by Government. With the benefit of hindsight, the proposed cropping pattern is considered overoptimistic, given the inexperience of the Senegalese farmers and the inherent difficulties in building a strong extension and training service from scratch. This risk, however, was foreseen at appraisal as well as the administrative and institutional problems facing SAED's manage- ment. Most covenants were appropriate and were complied with. Among the few exceptions is the gradual elimination of the subsidy for mechanization ser- vices, which has been delayed since SAED has been reluctant to eliminate subsidies to the Dagana farmers alone. This situation is currently being corrected and proposals to eliminate subsidies for all farmers in SAED's projects have been submitted to Government for approval. Elimination of these subsidies should have only a marginal impact (20%) on farmers' revenue, dis- cussed at para 4.07. 7.03 Seven recorded supervision missions took place during the five-year period from May 1973 - May 1978: two in 1973, reflecting the Association's -concern that the project become effective and that works start up as soon as possible; one in mid 1974 and one in late 1974, when it was already apparent that the project would face considerable cost overruns; one in early 1976; one in 1977 and one in 1978. In addition, during the same period there were several missions which visited SAED to prepare and appraise follow-up projects: two in 1975, two in 1976 and one in 1977. All the missions discussed with SAED's management the Dagana project issues, and lessons learned from Dagana were applied to improve the follow-up projects. Staff continuity at the beginning of the project (1973-1974) was less than desirable, with three different project staff involved, but since then, reasonable continuity has been maintained. Total field time devoted to recorded supervision of the project is 14.4 man-weeks, to which another four man-weeks should be added to take into account supervision work performed by preparation and appraisal missions of follow-up projects (Credits S-18-SE and 775-SE). Unfortunately, supervision schedules were very tight because of the Sahelian drought emergency (1972-1974) and resulting staff constraints. Despite supervision constraints and the fact that agronomic and financial matters would have benefitted from a more thorough examination, the overall working relations between the Associa- tion, SAED and Government were constructive and led to a better understanding of the development issues of the Senegal valley and the formulation of two follow-up projects. 7.04 From the beginning the project has been hampered by difficulties in securing local financing, a problem that plagues all Bank financed projects in Senegal, causing delays and cost overruns. Once the financing gap to complete the Dagana project became known in late 1974, the Association, while refusing a supplementary credit, agreed to increase its contribution indirectly through a program loan. The latter did not materialize for reasons unrelated to the project. Had the Association agreed to cover a substantial part of the then modest cost overrun, project implementation would probably have proceeded with less delays, lower total costs, and consequently, higher benefits. The replacement of SAED's Director General in April 1978 by a new man further aggravated this situation. - 30 - 7.05 In conclusion, while the project did not achieve all the appraisal objectives and suffered from substantial cost overruns, the responsibility lay more with Government bureaucracy and the lack of local funding than with actions or omissions by SAED or the Association. Lessons learned are being used in follow-up projects by the Association and other donors. Following the forthcoming rehabilitation it is very likely that the Dagana project will reach most of its appraisal objectives and benefit a much greater number of people than originally planned. VIII. DEBI-LAMPSAR ENGINEERING CREDIT (Cr. S-18 SE) 8.01 During the supervision mission of the Senegal River Polders Project in July 1974, it became apparent that (i) flood irrigation systems that had been constructed in the Senegal delta were inadequate as the level and timing of the annual flood of the Senegal River was too erratic to ensure adequate irrigation. This had been clearly demonstrated in 1972 and 1973 when the drought rendered irrigation impossible. To solve this problem, the Associa- tion considered an engineering credit that would aim primarily at redesigning the irrigation and drainage systems of the Debi and Lampsar polders to guar- antee full water control, and that would include a new perimeter, Diagambal, which would be equipped with a sprinkler irrigation system. 8.02 A preliminary feasibility study showed that development of these perimeters was economically justified, with an ERR of 16%. The study was based on (i) 1,500 ha in Debi for double cropping of rice and wheat, (ii) 2,500 ha in Lampsar for double cropping of rice, wheat and tomatoes, and (iii) 1,000 ha in Diagambal near Lampsar for year-round sprinkler irrigation of high value crops. Double cropping was considered feasible upon construc- tion of the Diama dam on the Senegal River; this dam was expected to be in operation by 1980. Other factors that justified the project were: the chronic rice shortage in Senegal; Government's strategy of developing irri- gation in the Senegal River region (which had already materialized through the creation of SAED in 1965 and the development of some 10,000 ha by 1974); the general lack of projects suitable for financing in the area; and the failure of the initial concept of flood water control systems which had to be converted into full water control systems. 8.03 Appraisal took place in December 1974. Three issues are high- lighted in the issues paper of December 16, 1974: (i) water rights: as the project would use international water stored by the Diama dam, the Associa- tion should make sure that appropriate steps were taken by the countries involved and by OMVS to ensure that legal problems related to such use were solved by the time the project was to be executed; (ii) Saint Louis water supply: the use of water for irrigation under the project should not impair the Saint Louis water supply; a FAC-financed feasibility study on this sub- ject was carried out and adjustments to the project were to be made in accordance with the results of the study; and (iii) reorganization of SAED: It was apparent that a reorganization of SAED should be undertaken to enable - 31 - it to achieve its ambitious objectives; Government was to make a rapid deci- sion in this regard. The decision memorandum (December 20, 1974) endorsed the above decisions and requested the mission to review OMVS' statutes on water rights and to ensure that the project would maximize the goals of income distribution. A President's Report was prepared rather than an appraisal report, given the limited scope of the Engineering Project. 8.04 Three changes were proposed by the Senegalese delegation during the negotiations held in Dakar on March 24 and 25, 1975: (i) inclusion in the project of a study on the programming of SAED's activities and on SAED's reorganization; (ii) a corresponding increase in the retroactive financing of the project from US$150,000 to US$200,000; this was intended to finance expenses incurred from January 1, 1975, to the time the project would be signed; and (iii) an increase in the foreign exchange cost of the project, i.e. IDA financing, from US$900,000 to US$1,000,000. All changes proposed in Dakar were finally accepted by the Association. CPS commented that the ERR of 16% indicated in the President's Report was probably on the high side, and that a "slower and probably more realistic growth in yields" should have been assumed. The Board approved the project on April 29, 1975; it was signed on May 22, 1975. 8.05 Two conditions of effectiveness were to be met by Government or SAED for the project to become effective: (i) Preparation of a subsidiary loan agreement between Government and SAED. After several revisions, the agreement was signed on December 8, 1975. (ii) Signing of contracts for the engineering, organization and health studies between SAED and the consultants. Signing of contracts proved to be very lengthy process: - The contract for final engineering studies was rejected twice by the "Commission des Marches" and was finally approved by the Prime Minister in late 1976. The related studies were completed in April/May 1976. - The terms of reference provided to the consultant regarding SAED's organization were revised three times following disagreement between SAED and the Association. The final version of the contract was approved by the Association in December 1975 but the contract was not signed until late 1976. - Agreement on the health component of the project was reached in January 1976 instead of April 1975. The contract was cancelled in July 1976, as payment of the initial advance had not been made by SAED to the consultant. - 32 - Fulfillment of these conditions was a long and difficult process. The ori- ginal date of effectiveness (September 2, 1975) was postponed three times to December 2, 1975, February-2, 1976 and April 2, 1976. Finally, after lengthy discussions the Credit was declared effective on February 13, 1976. However, although the contracts with consultants had been signed by SAED and the consultants, they were not yet valid; they still required approval by various authorities to be legally binding. These difficulties illustrate how complex and lengthy the administrative procedures were. Such problems had already been experienced with the Senegal River Polders Project, and clauses were included in the Credit Agreement to reduce administrative delays, but to no effect, since despite efforts made, ten months were required. 8.06 The project was to be carried out by SAED who, with the assistance of the consultants, would: (a) conduct topographic and soil surveys, (b) prepare final engineering designs and tender documents for the irrigation development of 5,000 ha, (c) prepare technical schedules, including cost estimates, for the operation and maintenance of the proposed irrigation and drainage networks with special attention to water distribution, staff requirements and training, (d) issue tender documents, evaluate bids and award dontracts, (e) study the impact of the proposed irrigated development on health conditions, (f) study and prepare the relocation of the Debi village and (g) study SAED's short and medium term operations and organiza- tion needs. Project Execution 8.07 The topographical surveys were carried out by SAED, and topographical maps at a scale of 1/2,000 were communicated to the consultant; these consti- tuted the basic documents for preparation of the final engineering studies. No external quality control was imposed on SAED, since the expatriate director responsible for supervision of the surveys was seconded to SAED by the consul- tants. 8.08 When the preappraisal mission of the Debi-Lampsar Irrigation Project visited Senegal in August 1976, SAED had carried out pedological surveys on the location of the future Debi and Lampsar perimeters. These surveys were considered insufficient by the mission which required additional information and preparation of a salinity map. SAED failed to provide the consultants with adequate soil capability maps, and consequently final engineering studies and tender documents had to be prepared by the consultants without this essential information. 8.09 The Association reviewed the final engineering studies and bidding documents regarding Debi and Lampsar during the August 1976 preappraisal mission. Comments were made on technical issues of the studies and clauses of the bidding documents. The design of the irrigation works was based on the same concept adopted for the Senegal River Polders Project. Unfortunately the lessons from Dagana were not yet properly assessed and were not incorporated in the,engineering design prepared by the consultants. - 33 - 8.10 The SAED study comprised: (i) preparation of SAED's short and medium term irrigation development program and (ii) a study of SAED's reor- ganization. The document relating to (i) provides adequate information on ways and means of developing the sub-areas (cuvettes) of the region, taking into account the physical and human constraints, and in the broad context of a phased development of key areas. The report indicates a list of sub-areas to be developed up to 1989, i.e. during Senegal's 5th, 6th and 7th planning periods, with corresponding acreages. An economic analysis of the entire program is also included. The document reflects SAED's objectives and is considered as background material for the preparation of individual projects. The study on SAED's reorganization was communicated to the Association in mid 1978 in draft form and has not yet been approved by Government. 8.11 The purpose of the health studies was to analyze the impact of the project on two diseases - bilharzia and malaria - and to make adequate recom- mendations regarding the design and water management of the project itself to reduce its possible negative impact. The studies were to be carried out in two phases: (i) elaboration of recommendations to forestall the spread of the diseases; specifically, these would include proposals on the treatment of drinking. water in the area; and (ii) formulation of a detailed biological analysis of water, mollusks and mosquitos, and recommendations on fighting the diseases after project execution. The consultant did not start the studies as the contractual advance payment had not been made by July 1976, and cancelled the contract. It was then agreed that the health studies would be entrusted to the Medical Laboratory of Saint Louis, which was managed by a well-qualified doctor. SAED would provide medical equipment and logistical means of trans- portation. The studies are to be completed in mid 1979. 8.12 Despite the administrative difficulties in signing the contracts with the consultants, no major delay was registered in the preparation of the final engineering studies and bidding documents of Debi and Lampsar. The Diagambal bidding documents were communicated to the Association in November 1976; bidding documents were initially scheduled to be issued in August 1976 so that bids from contractors would be available for the appraisal mission in October 1976. The appraisal mission was postponed to January 1977 and tender documents were not issued until April 1977 in time to include actual bid information in the final version of the appraisal report. The report on SAED's short and medium term program was prepared on schedule, but preparation of the study on SAED's reorganization, which the Association had insisted be available to the appraisal mission in January 1977, was delayed until mid 1978. Table 2 in Annex 6 compares the actual implementation schedule with appraisal estimates. - 34 - 8.13 Appraisal costs were the following: CFAF (million) US$ (000's) % of Local Foreign Total Local Foreign Total Total 1. Engineering studies 7 123 130 31 547 578 46.1 2. Health study 1 12 13 4 53 57 4.6 3. Topographic and soil surveys 17 5 22 76 22 98 7.8 4. Debi resettlement study 3 1 4 14 4 18 1.4 5. Organization study 19 47 66 85 208 293 23.4 Physical and Price Contin- gency (20%) 10 37 47 43 166 209 16.7 Total 57 225 282 253 1,000 1,253 100.0 Actual costs remained below appraisal estimates: the amounts of the three contracts with consultants - CFAF 122 million and CFAF 55 million for the contracts for final engineering and SAED studies, respectively; CFAF 12 million for the health contract - totalled CFAF 189 million as compared to the appraisal base cost of CFAF 209 million, a saving of 11%. These reductions were achieved following negotiations between Government and the consultants. 8.14 The comparison of appraisal and actual disbursement schedules (Annex 6, Table 1) reflects the delays in finalizing the contracts with consultants: the first disbursements were made in July 1976, i.e. with seven months' delay. The last disbursement made against engineering studies was in May 1978, i.e. almost two years after the study had been completed. Because of the delays, the closing date of the Credit was postponed by one year from December 31, 1977 to December 31, 1978. The disbursed amount of the Credit at completion (US$865,000) minus the sum repaid by the Borrower (US$125,000) was refinanced under the Debi-Lampsar Irrigation Project (Credit 775-SE) in June 1978. Performance of parties involved and conclusions 8.15 Execution of the project has been hampered by two main problems: (i) cumbersome administrative procedures, (ii) SAED's poor performance in executing the studies and the lack of coordination between SAED and the consultants. (i) Delays in signing contracts with consultants were expected by the appraisal mission, which recommended that signing of contracts be a condition of effectiveness of the Credit. Nothing more could have been done in this respect given the fundamental problem of SAED's legal status as an "Etablis- sement Public" and its resulting inability to break through the bureaucratic bottlenecks. Despite the delays the studies were completed almost on schedule as the consultants prefinanced a large part of their expenses. The Association made special efforts to have the conditions of effectiveness met. - 35 - (ii) At appraisal SAED was made responsible for project execution and, on the basis of available local and expatriate expertise, was considered capable of carrying out this assignment. The Association agreed to Govern- ment's proposal that the consultant responsible for the engineering studies and the supervision of the Dagana Project be commissioned to carry out the engineering studies of Debi-Lampsar. The fact that almost half of the ex- patriate staff of SAED management was provided by the same consultant was considered an asset for good performance. However the expected high quality performance failed to materialize, and as found by the preappraisal mission in August 1976, the soil surveys prepared by SAED staff were of inferior quality. Moreover, it appears that the coordination between SAED's expatriate management and the consultant staff charged with the studies was not close enough, and lessons learned from the construction of works at Dagana were not used to adjust the design of Debi-Lampsar, and the necessary modifications had to be made subsequently during project implementation. In retrospect, it was probably unwise to have agreed to use the same consultant in both management and study and supervisory positions in a project authority. 8.16 Despite these shortcomings, through the combined efforts of the Association, SAED and the consultants, the objective of the Engineering Credit was fulfilled, and appraisal of the Debi-Lampsar Project by IDA took place in January 1977. IX. CONCLUSIONS 9.01 The results achieved in the Senegal River Polders Project differ in many respects from the original objectives: costs have been nearly twice as high as expected; the cultivation of vegetables other than tomatoes, and double cropping were not implemented; and paddy yields were much higher - but the irrigated area lower - than forecasted, which resulted in a paddy produc- tion comparable to appraisal estimates. Conversely, tomato yields are lower but the area cultivated greater than anticipated; this too has led to a level of production close to appraisal estimates. No special accounting system was set up at the project level, and the expatriate technical specialists either were not recruited or were not replaced by competent nationals. However, the economic viability of the project remains acceptable provided that rehabilita- tion is studied and carried out promptly. The main factors that have led to the project's present situation are the following: (i) Cumbersome and inefficient administrative procedures of SAED: despite efforts made at negotiations to change SAED's status, Government refused to make this change, and SAED is still a public institution without proper financial autonomy;. (ii) Underestimation of costs at appraisal: mainly due to underestimation of physical contingencies; a rate of 15% of base costs was too low at a preliminary design stage. This was aggravated by changes in the design - 36 - of irrigation works which were made during execution. Price contingencies were also on the low side (6% per annum). Inflation was aggravated by delays in project execution due to difficulties in securing additional financing. In other words, lack of timely funding led to an additional cost overrun; (iii) The less than satisfactory performance of the consultants: the consultant who was providing technical assistance to SAED did not provide ade- quate support from headquarters to its seconded staff and did not achieve sufficient coordination with them. (iv) Lack of maintenance of irrigation works: this has led to a rapid deterioration of the works and now calls for a rehabilitation project for all peri- meters managed by SAED. (v) IDA supervision was probably less than warranted by a project with such intensive and innovative features. However, it is unlikely that increased IDA supervision would have resolved the major difficulties of the project, i.e. institutional problems and the lack of local financing. 9.02 The mixed results of the Senegal River Polders Project should not overshadow the achievement of introducing full control irrigation in the Senegal River Valley. With the large irrigation program ahead, the experience gained in this pilot project is expected to provide a sound basis for future development. - 37 - ANNEX 1 SENEGAL SENEGAL RIVER POLDERS PROJECT (Credit 350-SE) Completion Report List of Main Covenants Credit Agreement - SAED will enter into a financing agreement with Government under terms and conditions which shall have been approved by the Association (Section 3.01, b). - SAED shall obtain the services, during the Project period, of a suitably qualified training specialist to supervise and advise in the operations of the training center (Section 3.02). - The Borrower shall take all necessary steps to ensure that expatriate technicians in the employ of SAED shall be replaced only by personnel with qualifications and experience acceptable to the Association (Section 3.03, c). - The Borrower shall, for a period of ten years, from the date of this Agreement, consult with the Association annually on the level of subsidies for fertilizers accorded to farmers (Section 3.07, a). - The Borrower shall take all necessary steps to eliminate all subsidies for mechanical cultivation services accorded to participating farmers, after the first year of their participation (Section 3.07, b). Project Agreement - SAED shall employ engineering consultants to carry out engineering for the Project, preparation of tender documents and detailed designs, and supervision of construction of irrigation works, upon terms and conditions satisfactory to the Association (Section 2.02,a). - SAED shall enter into an agreement satisfactory to the Association to carry out the applied research component of the Project with IRAT or other consultants acceptable to the Association (Section 2.02, c). - SAED shall conclude with each farmer or farmers' group allocated land, an agreement substantially in accordance with a form of agreement acceptable to the Borrower and the Association (Section 2.05). - SAED shall consult annually with the Association on the specific research programs to be undertaken during the following year (Section 2.08). - 38 - ANNEX 1 - SAED shall cooperate with the Borrower's Ministry of Health, and render such assistance as may be required to enable regular checks to be made in the Project area for the presence of the disease bilharzia and of its vector (Section 2.09). - SAED shall establish and thereafter maintain an account with BNDS, for the purpose of receiving the funds required for the Project (Section 2.10). - SAED shall ensure that land allocations to participating farmers may be reduced, should the original allocations prove to be beyond the capacities of such farmers to work efficiently (Section 2.11). - SAED shall collect annual consolidated fees (Section 2.12, a). - SAED shall eliminate all subsidies for mechanical services accorded to participating farmers, after the first year of their participation (Section 2.13). - SAED shall establish and maintain separate accounts for Debi, Lampsar and Dagana in which expenditures shall be recorded separately and in such detail as to identify the purpose and function of each expenditure; (ii) maintain individual accounts for farmers or farmers' groups receiving credit for farm inputs and land cultivation services and (iii) maintain a consolidated fee account (Section 4.01, b). - SAED shall have its accounts and financial statements for each fiscal year audited, (ii) furnish to the Association certified copies of its financial statements and the report of audit by auditors (Section 4.02). - SAED shall submit its draft annual operating budgets (Section 4.03) to the Association. S3ENEGAL RIVER POLDEfiS PROJECT CREDIT 350-SE Signed 1-09-73 Totali Credit X Eff'ective 6-08-73 Amouant Closing 12-31-77 UJS* 4.5 million 5.U Appraisal Estimate (E) - Actual Disbursements (A) (Fully dis bur se d)i ...... Revised Estimates (R) 3.0 / I j, 1W 11y* .,Y i9' ];.Yr iy P'f -~ Y 1977 - - 40 - RETEK .01.EM LMn2/73 1973/4 / 1975/76 9?76/77 49778 197L9741759/80 L960/1 l981/0 19-2 -r- , 0. 0ana,7 Sae Cro9 - cu111iv006 ae (ha> 160 160 (77o) (1,330> si5 8 (620) 7 95 (01,40) (1.420 (1,420 (1.420) (.2 - .Id (T/hs) (770) (Tn)m (770r 770) (770) (775) (170) 17-0 (7.5< (1.S) (2.1) (2.4) (2.6) (2.3) 0-,00 <.1) 1.2) (3.2) (560) 4.0 (560) 5.- (560 5.1 (560) 1!0) (560) <190) (s602 (2.0) (2.2) (2.5) 2 7 (2 9) (3.0) ( L) (3.2) (0) (90 (90) 90 (90 <90) (90 Productio (T) (1,3861 0 (2.737) 2.060 (3,269) 3,23 (3,609) 5.544 r3. 02) (6.LE6) 4.337) 1. 4T9) 44 - cultivated area (ha) 150) (100) (200) (300) (500) 1700) 1970) (1 Lj0 - Itas T<64< (50) (50) (50< (50 50) (501 _0) (5O) (2.0) (2.2) (2.5) (2.7) '2.59) 3 2) 1. 1'.:, 50 (150) (250) (450) (651 920) (1.051 .- - . - za l- ,$.1 (2 3 ,,6 12.:! 1 11 ' 1 L ) r1 ") , 9oducCle (T) (100) 0 (2LS) 0 (345) 0 75) (1.405) (2.100) 11 ]07) (3, 520) - (30a.tes area (ha) <1 132 (70 520 (1r6) 350 (254) 40 34) '30< (33<) - yta(' (T/he) (30) 132 100) (30) (30) (30) () (30) (25) 32 (27.5) (30 (32.5) (35) ,'5) (35) [60) 15. (r0) 25 (40) 15.5 (0< (40) (40) (26< (28.6 (31) (33.5) (35) (35< (76) (184) (26å) (264) (264) - -27 - _ l - (32) (-5) " ProduccI (TM ("50 4,224 (1,865) ,060 (4,0)6 8,750 (7.643) 6.975 (10,838) (11.556 (1>160< (d 90) (11.609 3. Qg.!/ •691atdena(e (1205)LD 28 (dton 1T) (55) (555) (1.L55) 1,001 (2.20) (2 280) 2,29 0 (22d) 4. Noao. s - 0.aivated area 1.) (601 (120) o _ton ( 057) 1,15) r2,2L0) (2. 364) (2.964) (2.964) '2,961, 9roductao (T) (15< (7,6) (2.702> 0433) (4, 017) (4,0 (4,00 g <4.)1 (,,00< .10u1eia.se st.* (bal 23 (o (0 r-~U4 T ) production (23) (4 25 1.10 -j settiaOe area (ha) - yta(d (T'ha< .roducClon .T) clILvd ar.. (h0 240 Pnoduction (T) rigflOes to esCtenenessa are aoria 9 tmts 2.s not cu..ivat.0. SENEGAl. SENEGAL RIVER POLDERS (Credit 350-SE) Completion Report PaaBdets APPRATSAL (full m ACTUAL Other a r Typical Farm Tomato- Tomato- (19781) Rice Farm Sorghum Vegetables Rice Crop Tomato Crop Total Farm Size (ha) First crop 3.5. 3 2 0.7 0.25 0.95 Second crop 2.75 2 1.5 - - Number of Farms 406 58 456 1,650 Cropping Pattern and Yields (t/ha) Rice 6.25 - - 5.1 5.1 Tomatoes - 1 0.5 - 15.5 15.5 Onion - - 0.5 - - - Potatoes - - 0.5 - - Sorghum - 4 2.0 - - - Net revenue per farm (CFAF) 1/ 192,930 169,420 190.690 92,170 26,930 119,100 Working days 325 422 320 56 60 116 Revenue per working day (CFAF) 594 401 596 1,027 1/ From Appraisal Report, (No. 29-SE), Annex 18, Table 1, page 2. Actual Rice Crop Tomato rnp YJeld (t/ha) 5.1 15,5 Production (k) 3 6 Revenue (CFAF) 149,4d0 50 Agricultural Inputs * 14,720 12,500 Mechanical Services * 21,510 8,570 Consolidated fee 17,500 8,750 Amortization 3,500 1,750 Net Revenue 92,170 26,930 *See Debi-lampsar Irrigation Project Appraisal Report (NO. 1685 a-SE; Annex 5, Table 4) W E 5 1 1 E 1r 1 11 SE i h t; ä- 0 i i5~Or to gl l a- i i i - § § rc l - .. i m "åt 1 a o ål! a.- ki 59 818 *0 s IL1 §n .0 2 - 2t * ,1 8 - a 2 1 a- - Ii " Ii Ii ~i i E -a 19 1 E 一渭〕〕．！〕雜；〕！輟！ 〕糾！！。〕！！矓矓：i;! 三屆二侮劍江惜「：：寫勰逞巴響，藝 計震州二。江訕k已日一k召 甘亂甘，蠶19「「呈 霎診，寫．.1一r, 亂為艮，，「認卜 ,－一r雙各 寫二寫e·熙面 F雜F計徊弓喜 PFP吧‘萬寫 磚詣！亂 藝斤藝邊響 1彎讓：『壯‘，滅死‘召，這。藝 響•編匹露斗，騙必訕蘊b弘私訕騙細N,＿變畫憤 萬編甚，各肥唱日祕日SB細右古乙右6縱搬認憐 。認一p個Ir 區．江必習■ ，艮，」唱l 江F巳召細亡祕居必痲讓．名。籐 ¥！}&&‘磬“騙’他”&&&&‘綠。： 廈卜戶多摔；戶莽，爭：弟F．。＿浴不 :｝洛”&&&&&&&&&&11!}:& 蓄件讓言杳靄彗權雪言言霹曇；。，，。給膩朧爆 奮―& ；作，坏汗界莽夕戶，拱F夢 }&&&“籲”&&&&&&”緬 }&‘號縛“藝‘誡，·：。 卜，戶革添莽辭員，弁F夢 -“甚’濺磬§忿“且‘也“’摻 ―戶：于汗界露拱參，·拱戶莽 「浴甚’碧藝騷‘碧‘甘“&‘婦 露．．…界矚 ―爭／籐多多弟拱貸：，拱斤弁蘇―巒 ―答洹籐斗若莒騷狗莒莒甘法磬差‘浴廷 - 45 - :.NIX 3 åg~ 59~ RIVER POUELS (Cr, 3JO-SE) Conatruction *chedlea (.ccual and e f ... a**. at appra~) for tbc D~ P-lootar 11 L972 1973 1974 1975 1976 1977 1978 1979 L980 DAGAKA (2,960 h- S.etor ý735 ha) -d 3 (1,800 ha) studi- survoya LJi *upne.rmc UTiCaci.. ~ka p=plag sttiona 1-ýoing ution conetruction 6. z~tion am droj,nac. .199~ 7. FicId irrigatim -orka JL S.Ct.r C (42t ha) Cer%ag Dik. 5. P=plng cation 6, and .-inac. ayet~ 7. 11 Tha appmiaal rim scadula Lucluded couacruccl« of eho Deht ced ~ @ar perivecors ; fol lovins r.f.rwulacl- f rh. P-l-£ t. 1914 rh~ rwo *utprojectu vara abando.ed. Th. pre-c tahl. doos nor ~om chem. z/ M-d ýft final enging.rint scudlaz. Cultivacad &roa ta currentlj Lover duc ro -1 lo& and iandy olls, and for the cco~ ~ ynin 2.500 ha hava been ~talecd. 1..lud.. ~nacr-CI.. t f- ~ka am 1-d 1~1.8. - - - Appralaal. - ACC"l. 1uS;:; Mi I; SENEGAL - DEB1 LAMPSAR ENGTNEERING PROJECT - Cr. 518-8E Signed 5-22-75 Total Credit )(Etfective 2-13-76 Amount Closing 12-31-78 US$ 1.0 million ---Appraisal Estimate (E) ...Actual Disbursements (A) 1/ .. evised Estimate (R) 1/ The disbursed amount of the Credit at completion (US$865,000) minus the sum repaid by the Borrower (US$125,000) was refinanced under the Debi-Lampsar Irrigation Project (Credit 775-SE) in June 1978. 01 0.6 0." t-, a, x 0 N 1 ) J - I A " .1 ' 1 0 Di A I , - 47 - SENEGAL ANNEX. 6 DEBI LAPSAR ENGINEERING CREDIT (S-18 SE) AblE 2 COMP=LEION REPORT Appraisal and Actual Construction Schedule 1975 1976 1977 1978 Engineering Contract Award Vr Health Study Contract Award 7V- GOVERNMENT CONTRIBUTON - Topographic Survey Debi Diagambal Lampsar - Soil Survey Debi 2.. Lampsar L 2/ - Master Plan Senegal Valley - Development -- -*-- ..- - Resetlement study Debi Village_ EALTE CONSULTANT CONTRIBUTION 3/ ENGINEERDNG CONSULTANT CONTRI- BUTION - Geotechnical Study - Final Design Debi Diagambal - La=psar - Preparation 3idding Documents . .. - . - Bid Invitatzon - 7V 1/ The contract was cancelled in July 1976 j/ Preparation of additional information on salinity S/ Study still incomplete  化叮PcR&