Peru - San Lorenzo Irrigation and Land Settlement (Stage Three) Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 1858 PROJECT PERFORMANCE AUDIT REPORT PERU SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT (LOAN 418-PE) January 17, 1978 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. US 1.00 Soles (S/.) 26.85 (1965) S/. 38.70 (193) SI. 43.38 (1972) S/. 65.00 (1976) Measures m - millimeter (1 mm 0.039 inch) meter (1. m = 3.28 feet) km!= kilometer (Lkm 0.62 miles) ha = hectare (1 ha 10,000 m2 = 2.47 acres) - square kilometer (1 km2 = 247.1 icres - 100 hu) cubic meter (1 m3 = 1.31 cubic vard= 264,2 US gallL _m = million m3 (llm3 = 311 acre feet) ' kilogram (1 kg = 2.2 lb) _tr ic on or ton = 1,000 kg = 2,205 1b "C' million cubic meters Abbreviations XDE Agricultural Developnt Bank AIf United States Agency Eor International Devel-pment 3 Banco de Fomento Agropecuario Centro de Drenaje y Recuneracion -de Tierras CRDU NLa Oficina Regional de D,esarrollo dei Nor:e CL Irrigaciun y Colonizaci6n de San Loi*rnzo rECO international Engineering Company, San Francisco. JS. O'.:PRA Oficina Nacional de Reforma y Promocin Agropecuaria WAC National Agrarian Council SUDEET Sub-Directorate of Soil Conservation ERAC Institute of Aerarian Reform and Land Settlements SEPA Servicio de Investigacion y Promocion Araria CORFll?A Corporacion Financiera de la Refor-na Agraria SIC Short-term Credit MLTC Medium and Long Term Credit iscal Y e:u January 1 to December 31 FOR OFFICIAL USE ONLY Project Performance Audit Report PERU SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT (Loan 418-PE) Table of Contents Page No. Preface Basic Data Sheet Highlights PROJECT PERFORMANCE AUDIT REPORT I. Main Developments Under the Project 1 Sector Setting 1 Background 1 The Development Project 2 Costs 3 Benefits 3 Management and Organization 4 Performance 5 Possible Improvements 5 Ii. Conclusions and Recommendations of the PCR 5 General 5 Appraisal 8 Major Changes in Project Scope 8 Extension of Loan Closing Date 9 Super-vision 9 Delays in Implementing the Development Stage 10 III. OED Comments 11 The Chira-Piura System and its Development 11 Inappropriate Analysis of the Project' s Regional Setting 13 Problems Resulting From Implementation by an Autonomous Unit 14 Poor Accounting and Problems in Auditing Project Accounts 15 The Bank's Financial Contribution to Development of San Lorenzo 15 Table Sl - Main Events 17 Table S2 - Summary of Targets at Appraisal, Renegotiation of Loan (1968) and Actual Achievements 18 Table S3 - San Lorenzo: Gross Medium- and Long-Term On-lending - Sources of Financing 19 Diagrams 1- 5 20 This document has a restricted distribution and may be used by recipients only in the pefformance of their official dluties. Its contents may not otherwise be disclosed without World Bank authorization Table of Contents (Cont'd) ATTACHMENT: PROJECT COMPLETION REPORT Volume I - The Main Report I. Introduction Al II. Background A2 III. The Project A4 IV. Performance of Project Components A12 V. Organization and Management A23 VI. Economic and Social Impact A24 VII. Cost Recovery A28 VIII. Future Improvements A29 IX. Special Issues A30 Tables Annex 1: Project Description Annex 6: Water Availability and Use Map Volume II - Supplementary Data Volume (SDV) - Annexes 2 - 10 (Not attached to PPAR. Available from Agriculture 4 Division of the Regional Office.) Project Performance Audit Report PERU SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT (Loan 418-PE) Preface This report presents a performance audit of the project supported by Loan 418-PE, signed in June 1965, and closed 11 years later, in June 1976. Only 60% of the loan proceeds were disbursed; the other 40% had been can- celled after renegotiations in March 1968. A project completion report (PCR), necessarily selective for such a long-lived project, was issued by the Latin America and the Caribbean Regional Office in April 1977. A final report had been prepared earlier by the Government of Peru for the mission in charge of preparing the PCR. The audit is based on a review of the Bank files, on consultation with Bank staff, and on a visit to Peru by OED staff accompanying the project completion mission. During the visit, discussions were held with officials in the Ministry of Agriculture, the Ministry of Food, and the Agricultural Bank, in Lima, in the I Agrarian Zone, and at the project site; and the proj- ect works and a number of farmers were visited. Though OED staff accompanied the mission, they did not participate in the actual drafting of the PCR. OED's contribution is implicit throughout the report, however, for lengthy discus- sions were held with the PCR mission, both during the field work and after- wards during the drafting of the report. The audit established that the PCR properly selected the main issues and that it offers a satisfactory analysis of project history and performance as well as a frank acknowledgment of the factors that brought about the main problems associated with the project. The audit fully supports the PCR's conclusions. Therefore, this report consists of the summary, conclusions, and recommendations of the PCR, and of additional comments by OED that further elaborate on some issues. The first volume of the PCR is attached; the second volume, which comprises nine annexes with detailed data and supporting information, is available for reference in the Latin America and the Caribbean Regional Office. The kind and valuable assistance provided by the Peruvian institu- tions and individuals involved in the preparation of this report is grate- fully acknowledged.  Project Performance Audit Report PERU SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT (LOAN 418-PE) BASIC DATA SHEET A. Amounts (in US$ million) (as of 7/31/77) Original Cancelled Disbursed Repaid Outstanding Loan 418-PE 11.0 4.6/a 6.4 2.7 3.7 B. Project Data Original Plan Revisions Actual First Mention in Bank Files Late 1961 Board Approval 04/20/65 Loan Agreement 06/18/65 /b Loan Effectiveness 11/24/65 /c Physical Completion; date and percent of original project actually completed - Land Settlement 1968 1971 (86%)Id - Agricultural Credit 1968 1969 (39%) - Drainage Study & Works 1967 1976 (100%) - Basin Study 1967 1969 (100%) Loan Closing 12/31/67 06/30/69. 12/31/73, 12/31/74 06/30/76 Total Costs (US$ million) 30.4 (US$ million in 1975 terms) 81.7 61.5 Economic Rate of Return 65e - Stage III (this project): - Since 1961 - 17.5% - Since 1965 16% - Stages II & III (Loans 114 and 418) 14% 8.7% C. Mission Data Month, No. of No. of Date of Year Persons Weeks Manweeks Report Appraisal I SEP 64 5 2 10 Appraisal II 'DEC 64 2 0.5 1 ) 93/26/65 11 Supervision I JAN 66 2 1.5 3 03/21/66 Supervision II APR 66 2 1 2 04/29/66 Supervision III OCT 66 3 3 5 11/23/66 Supervision IV APR 67 1 2 2 05/03/67 Supervision V MAY 67 4 2 8 07/27/67 Supervision VI MAY 68 2 1.5 3 06/27/68 Supervision VII AUG 68 1 1 1 09/18/68 Supervision VIII -FEB 69 2 2 4 05/ /69 Supervision IX MAR 70 1 1 1 04/03/70 Supervision X OCT 70 1 1 1 11/12/70 Supervision XI FEB 71 2 2.5 5 03/18/71 Supervision XII JUL 71 1 1 1 08/10/71 Supervision XIII SEP 71 2 1 2 10/06/71 Supervision XIV AUG 72 1 1 1 09/07/72 Supervision XV FEB 73 2 0.5/1.5 2 03/14/73 Supervision XVI AUG 73 1 1 1 08/23/73 Supervision XVII JAN 74 2 2 4 03/04/74 Supervision XVIII SEP 74 2 1.5 3 09/24/74 Supervision XIX JUL 75 2 2 4 09/02/75 73 Completion JUL 76 2 3 6 FEB 77 6 D. Follow-on Project: There is no follow-on project at the San Lorenzo site. /a US$4.5 million were cancelled in March 1968; US$0.1 million were cancelled after the Closing Date. /b Revised March 13, 1968. /c Disbursements were suspended from June 1969 (except for Category D covering Basin Study) to December 1972. /d Only 30% was achieved while funds from Loan 418 were available, however. /e This figure is not fully comparable to the appraisal estimate (see paragraph 3.14).  Project Performance Audit Report PERU SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT (Loan 418-PE) Highlights Loan 418-PE (US$11 million) provided financing to accelerate the ong&ing land development and settlement stage (Stage III) of the San Lorenzo irrigation project. This project had been started by the Government back in 1948. Bank involvement had begun in 1955, when Loan 114-PE (US$18 million) financed a storage reservoir (Stage II) to serve 45,000 ha (later reduced to 32,000 ha) which define the San Lorenzo project area. The main irrigation works were completed in Stage II; Stage III was directed at settlement and on-farm activities. Loan 418 also financed a comprehensive drainage study to be followed by drainage works, and a study of the two basins in which San Lorenzo is located - the Chira and the Piura rivers. It took 11 years to disburse the loan. This long time span can be broadly divided into three sub-periods: (i) 1965 to 1969 - project implementation moved slowly; several problems kept the rate of develop- ment below the targets set at appraisal and even below pre-appraisal levels; in March 1968, the project was reduced in scope without being reappraised, and 40% of the loan was cancelled. (ii) 1969-1972 - project implementation was almost paralyzed. (iii) 1973-1976 - most of the drainage works were built; machinery for operation and maintenance was procured. The construction of the drainage works was the main factor in extending Loan 418 for seven years after the original closing date of June 1969. Management problems, compounded by exogenous factors - storms, diseases, low water availability - slowed down project implementation during the first sub-period. Serious over-estimation of water availability and in- adequate review of soil characteristics at appraisal aggravated the situation: more than 9,000 ha - one-fourth of the total area developed under the project - had to be abandoned as a result. The drainage works were ill-defined, for they anticipated the results of a major drainage study, also financed under the loan. Moreover, they were implemented without having ever been adequately reviewed or appraised by the Bank. The project was classified as a "problem project" in five of the nine years since that list was started. Nevertheless, it has generally achieved its basic goals and produced important benefits; it is estimated that it is likely to yield a 17.5% rate of return. The following points may be of particular interest: - Failure to differentiate project specific problems from sector- or nation-wide problems during supervision (PPAR, paras. 21 and 28) - Inappropriate analysis of critical project features at appraisal (PPAR, paras. 24, 25, 37 and 38); lack of reappraisal at critical points (paras. 26 and 27); improper supervision resulting there- from (paras. 29 and 30) - Disagreement between the Government and the Bank at appraisal (PCR, paras. 2.09 and 3.02) - Poor analysis of water availability and soils at appraisal; some consequences derived therefrom (PCR, paras. 9.01 and 9.02, and Annex 6) - High project profitability to the country and to the farmers, and as training ground for agricultural officers (PCR, paras. 6.06-6.11); unplanned benefits to the very poor (PPAR, para. 19) - Poor accounting and auditing (PPAR, paras. 42 and 43; PCR, 'para. 3.31). Problems in determining actual project costs and cost over- runs (PPAR, para. 9; PCR, paras. 3.14, 3.18, and 3.19) - Deficient collection of water charges (PPAR, para. 13; PCR, paras. 7.01-7.05) - Variable repayment of the project's sub-loans (PCR, paras. 4.16 and 4.17) - The Bank's small financial contribution to development of San Lorenzo (PPAR, para. 44). Project Performance Audit Report PERU SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT (Loan 418-PE) 1/ I. MAIN DEVELOPMENTS UNDER THE PROJECT Sector Setting 1. Agriculture in Peru employs more than 40% of the labor force although contributing only 15% of the GNP. Agro-based industries account for about 30% of employment and 40% of value added in manufacturing; yet agricultural output, especially food production, has lagged behind population growth for at least the past 15 years. The resulting growth of food imports has placed increasing pressures on the balance of payments, and Government subsidies to agriculture are a serious drain on the public sector budget. 2. Of the 2.9 million ha of cultivated land in Peru, about 1.0 million are irrigated, 770,000 of which are located along the Coastal region of Peru; this area produces about two-thirds of the country's total value of agricul- tural output. 142,000 ha or 20% of the irrigated area lies in the Chira-Piura river basins, of which San Lorenzo forms a part. The area has special economic significance as a major producer of high quality export cotton and a substantial proportion of the national tree crop production. Completion of the Chira-Piura Diversion Project means that areas in the river basins must be developed urgently to efficiently utilize the substantial sunk investment in the main storage dam and ancillary works. The next phase of this development plan is presently under consideration for Bank financing; as the San Lorenzo Project has been the first major development in the Chira-Piura river basins, clearly the experience and lessons learned can be of special value to avoid some of the problems which af- fected Loan 418-PE. Background 3. The San Lorenzo Irrigation Project is located in the Piura Region in the northern part of the Peruvian coastal area. Successful agriculture depends on full irrigation, but there is almost total flexibility in the growing season for most crops. Cotton has been the dominant crop, but fruit is becoming in- creasingly important and should soon account for about 50% of national produc- tion. 4. Bank involvement in the San Lorenzo area began in 1955 with Loan 114-PE which financed the main dam and infrastructure, designated Stage II development. This was completed in 1959 without firm plans to develop the lands served by the works. By 1961, the Bank had expressed its readiness to help finance the set- tlement stage, designated Stage III development. Financing was delayed until 1965, mainly because the Bank insisted on organizational and legal pre-condi- tions; when the project was appraised in late 1964, some 14,000 ha were already settled averaging about 16 ha per holding. The estimated project cost was US$29.8 million. 1/ This section is the PCR's Summary. -2- The Development Project 5. Loan 418-PE of US$11.0 million equivalent was made in June 1965 to help accelerate land development through services and facilities to management, credit for on-farm development, and technical assistance. It also financed drainage studies, equipment to implement anticipated drainage works, and a major study of the Chira-Piura river basins. Unforeseen technical problems related to water availability and soils, compounded by management and budget problems, meant that appraisal objectives were optimistic; in 1968 the loan was renegotiated without having been reappraised; project targets were lowered (see below), and the loan amount was reduced to US$6.5 million. The closing date, originally December 1967, was extended four times until June 1976; Table S1 summarizes the main events in chronological order. 6. The primary objective at appraisal was to bring 32,000 ha into pro- duction by 1968. A revised estimate of water availability reduced this to 18,000 ha at loan renegotiation. However, since 1969, agrarian reform legis- lation meant squatters in the project area also had to be settled; Government took a calculated risk in expanding settlement. In all, 36,700 ha were developed within the San Lorenzo project but of this about 9,100 ha orig- inally settled were abandoned for technical reasons. (In addition, 5,900 ha were settled outside project boundaries as defined at appraisal.) A total of 56 cooperatives have been formed on about 7,600 ha supporting about 1,900 ex-landless farmers, a development unforeseen at appraisal. The continual divergence in development objectives for technical, social and political reasons created major problems and tended to make settlers insecure. Little agricultural input was provided after initial development except some tech- nical assistance which, however, proved to be of limited effectiveness. No incremental yields have occurred over the last decade, but farmers have put about 5,700 ha into tree crops. Costs of land development elements can be indicatively estimated at about US$875 per ha. 7. 13he drainage component - the main reason for extending the loan be- yond 1969 -was only a partial success. Inadequate and optimistic prepara- tion resulted in a program which was continually amended and never properly reviewed. The component consisted of a major study to identify 5,000 ha thought to have critical drainage problems, to be followed immediately by im- plementation; this was unrealistically scheduled for completion by December 1967. The study cost US$0.34 million, compared with an estimate of US$0.25 mil- lion, and was not completed until March 1969; it concluded that 8,200 ha had serious drainage problems. This study had to be reformulated by Government, and construction of the first phase did not begin until 1973 and was completed in late 1976. Tests concluded that most areas, though subject to salinity prob- lems, could be reclaimed by leaching. Only a nominal area was subject to sub- surface drainage problems, and these areas did not have to be developed anyway to achieve the settlement objectives. 8. The appraisal mission identified the need for a Chira-Piura Regional Basin Study. Government wanted to implement a simple diversion project, but this was rejected by the Bank because Government did not wish to consider the international water aspects with Ecuador. The studies cost US$1.06 million, compared with US$0.6 million estimated at appraisal, and were completed in 1969. The first phase recommendation was for a storage dam on the river Chira, 1/ Most of the loan proceeds for financing Operation & Maintenance equipment had not been disbursed either. a diversion canal to the Piura Valley, and some main drainage works which the Bank appraised in 1970. Possible Bank investment was abandoned because Peru arranged a package deal including financing through Energoprojeck, a Yugoslav company. Construction was completed in 1976 at a cost above US$200 million, compared with US$54 million estimated in the Bank study for a smaller scheme. The Regional Basin Study was valuable but helped delay a solution to the in- adequate water supply, the dominant problem through the life span of the loan. Costs 9. Divergence from appraisal, the extended period of effectiveness, and inadequate data make a comparison of costs at appraisal and closure of little value. Projected development expenditures from 1961 including the costs from 1965 estimated at appraisal were US$81.7 million equivalent (in 1975 terms), but this relates to superseded objectives. Actual development costs since 1961 have been estimated at US$59.2 million (in 1975 terms) of which about US$17.5 million was for on-farm development; capital and running costs cannot be separated. In addition, US$2.3 million (in 1975 terms) were spent on the Basin Study, bringing total project costs to US$61.5 million. The renegoti- ated loan of US$6.5 million (equivalent to US$12.2 million in 1975 terms) was used as a development credit for elements that generally followed the list of goods defined at renegotiation; apart from those invested in the drainage study, funds were usefully and effectively disbursed. Benefits 10. Few benefits resulted from the technical assistance component, mainly because objectives were not clear at appraisal. The Drainage Study and the Basin Study were undertaken by the same consultants; results were less than satisfactory because they were not adequately planned or controlled: schedules turned out to be unrealistic and neither study was effectively reviewed. The Bank had agressively promoted both studies at appraisal and bears some responsibility for the poor results. 11. Project operation and maintenance has been generally underfunded; but when the budget was more generous, management functioned reasonably well. Annual variations in physical and technical factors and changing Government attitude make it hard to judge performance; the dominant influence has been variable water availability. Quantified judgment of O&M is not possible because development and operational costs cannot be readily separated, but current budgets seem inadequate for a preventive maintenance program. 12. Prior to appraisal, the project was at its most dynamic development stage; although social and agricultural supporting services were mentioned, no specific input was provided through the loan except for technical assis- tance and facilities for management. The assumption was that management would formulate and implement services as required. This did not happen: nominal research was undertaken, agricultural extension made little progress except for tree crops, and the livestock program proved to be a failure. Much progress was made through the initiative of the settlers, and successful co- operative associations were sponsored by management. Institutional credit was essential to most farmers and was generally successful; the repayment record was good, but capacity to pay varied greatly over the years because of land quality, unreliable irrigation and crop damage. On-lending of gross medium- and long-term credit, the main productive element of the development phase, was about US$25.4 million equivalent of which the loan provided only 18.77. Marketing of the substantial fruit production has not been solved and will remain a major problem unless strong action is initiated. 13. Historically, water charges in Peru have been nominal. Since 1969, Government has taken steps to improve the situation, and it plans to increase charges over five years to cover full O&M costs. Current charges of S/.0.04/M3 (equivalent to US$10.0 per ha) are the highest in Peru and would cover 65% of present O&M costs; the main difficulty is poor collection, but Government is taking corrective action which is showing results. Management and Organization 14. The organizational structure for San Lorenzo was set up in 1963 with an independent agency under the direct control of the Office of the President. A condition of Loan 418-PE was that the agency should have opera- tional autonomy and funds. Short-term, this achieved results unique in the agricultural sector in Peru, but made the project vulnerable to personalities and patronage, both transient factors. Management trod a difficult path between hostile political opponents, central authorities uncomfortable with management autonomy, and the Bank pressing for action on a wide range of complex issues much of which was novel. Unforeseen technical problems imposed a substantial strain on management which was complicated by a lack of continuity in key staff positions. Basic socio-political changes occurred when Government's agricultural strategy was revised in 1968 to achieve greater coordination and flexibility through decentralization. Project management was understandably weak during this period and when disbursements were suspended in 1969, the project virtually came to a halt because manage- ment autonomy depended on active Bank support. In 1971, the new Agrarian Keform Law vested responsibility for the project in tne regional authorities in Piura, and San Lorenzo became part of an irrigation district with drainage administered as a separate project. Much criticism of project management since 1971 seems unwarranted because it assumed that the San Lorenzo project was still being implemented, although investments financed under the project nad been almost completed. As with most completed work programs, San Lorenzo has suffered from relatively low priority, inadequate budget allocation, understaffing and general lack of attention that goes with a development phase. Tne Government had undertaken in the loan agreement to operate and maintain the project in accordance with sound engineering and agricultural practices, and to make at all times available promptly as needed all sums and other resources which shall be required tor the carrying out of the project and for its operation (Loan agreement, Sections 5.01 (a)(b)(c); 5.11 (a)(b)). Although these covenants have not been fully complied with, some management problems were not project specific and could only be corrected at a sector or national level. Under the circumstances, it is a tribute to San Lorenzo management that so mucn was achieved, particularly before the loan was renegotiated in 19b8. -5- Performance 15. To compare performance with appraisal is of little value since the original scope was superseded at renegotiation when rates of return were not recal- culated. The ex-post internal rate of return for the development phase since 1961 is estimated at 17.5%; the rate of return since 1955 (Loansll4-PE and 418-PE) is conservatively estimated at 8.7%, compared with 14% estimated at appraisal of Loan 418-PE. The shortfall can be explained by underutilization of main works. Benefits are summarized in Table S2. Gross farm incomes are high by rural standards in Peru; and the early settlers farming about 16 ha are particularly favored, averaging about US$2,500 per holding. Direct bene- ficiaries are estimated at 3,180 families, of which at least 2,400 were poor landless peasants; this compares with a target of 950 families at loan re- negotiation. Direct employment works out at about 5,000 full time jobs, with some 1,500 jobs provided in supporting services. Possible Improvements 16. Major improvements can be made to increase crop yields, which have remained static for a decade; this is partly due to unreliable irrigation and concentration of effort on land development rather than agricultural extension. An associated development (initiated through the Chira-Piura Basin Study) will be the agricultural development stage of the recently completed Chira-Piura Project to divert water to the Lower Piura Valley. Full development of both river basins will include extension and rehabilitation of existing irrigated areas, covering about 142,000 ha, and possibly integrate the operation of the San Lorenzo and Poechos dams 1/ to optimize water availability and use. Em- phasis will be placed on drainage and agricultural development to intensify production and perhaps include agro-industries to help overcome marketing con- straints. The first stage of this development is being evaluated by an FAO/CP preparation mission (February 1977). 1/ Poechos dam is located on the river Chira about 5 km upstream from the mouth of the Chipillico river (see map). -6- 1/ II. CONCLUSIONS AND RECOMENDATIONS OF THE PCR General 17. In spite of the poor reputation that San Lorenzo has suffered in recent years, the mission feels that the project has generally achieved its basic goals - often in unforeseen ways and about eight years late. Why and how this has happened is difficult to explain because divergence from appraisal has been so great. It seems that over time, compensating factors tended to cor- rect problem situations. At this point in time the project is reaching target so- cial groups acceptable to both Government and the Bank and has successfully tacked rural poverty. This has only been possible because of fortuitous changes in social and political parameters essential to project success but treated lightly at appraisal and generally unpredicted during the checkered project history. It has also been due to a degree of good fortune over recent years in that water availability has been consistently above average. 18. Success is much greater than anticipated when the loan was renegoti- ated in 1968. Both the Bank and the Goverm ent can take credit for important contributions to help overcome the considerable difficulties which have af- flicted the project. Bank support through critical times both before and during Loan 418-PE can be credited for keeping the project alive and thus pro- tecting the substantial investment in Stage II development partially financed by Loan 114-PE. The GoverLment, and in particular the dedicated project staff, always attempted to implement and defend the project in the interest of the settlers - often against severe political opposition. Within the limits of personnel, budget, facilities, and experience in tackling an innovative develop- ment, San Lorenzo management achieved much more than could be reasonably ex- pected. Perhaps most credit should be given to the settlers who had most to lose and yet persevered through difficult and uncertain times to ensure that reasonable benefits were achieved. An important factor was that original farm sizes gave farmers sufficient financial incentives to believe that success was possible. 19. Water made available through the San Lorenzo reservoir was always used, although not efficiently, and the irrigated area has now almost reached the target set at appraisal; production is above that projected at renegotia- tion because the area developed substantially exceeded that expected. Yields have shown no increase since 1965, but none was projected at appraisal. The most important diversification into tree crops has taken place, but marketing might restrict potential benefits. The number of beneficiaries has greatly exceeded the original estimate through reducing farm size by collective farm- ing. The project has especially benefited the very poor, satisfying present Bank policy objectives even though this was unintended at appraisal. Settle- ment was generally successful and settlers have prospered except in areas where fundamental errors in project assumptions meant that successful irrigated agri- culture was not possible. Significant secondary or unexpected benefits were derived from the project both at regional and national levels; substantial in- cremental benefits are still possible through increasing agricultural produc- tion. The advantageous divergence in objectives and targets was essential for 1/ Taken directly from the PCR. - 7 - success but meant non-compliance with some loan conditions when the project could not be made to conform to assumptions and aims that had become obsolete; otherwise, Peru has attempted to carry out its obligations under the loan agreements. 20. The special burden of San Lorenzo was that since 1969, the project had been fully implemented,except for drainage,and treated by authorities like any other irrigation scheme elsewhere in the country. Unlike typical infrastructure projects where success is equated with completion of works rather than realization of agricultural benefits, the post-implementation stage was continually and closely monitored by the Bank. San Lorenzo has been one of the few irrigation projects where the Bank has had to share with the Borrower operational problems of a completed agricultural project. 21. It is concluded that the problem status and general dissatis action with San Lorenzo within the Bank is and has been somewhat overstated.V Many problems did exist, but some of them were sector- or nation-wide, and only project specific in that San Lorenzo had the misfortune to illustrate the problem. Also, the Bank involvement with the project continued about six years longer than was strictly necessary because the drainage implemen- tation might have been financed separately. There were major project problems which resulted in such deformation of the original scope during its greatly extended time schedule that performance cannot be realistically related to appraisal objectives. However, only some problems were really avoidable during the life span of Loan 418-PE since many can be traced back to Stage IT development of San Lorenzo (Loan 114-PE). Therefore, not belittling the many problems which might have been avoided or corrected, the mission feels that the performance of San Lorenzo Stage III is probably typical of, and perhaps better than, the operational development stage of many recent irrigated agriculture projects given similar country situations. On the basis of data from renegotiation of Loan 418-PE, performance greatly exceeds implied expectations. 22. The mission feels that many projects designated integrated rural development are likely to face similar problems to those of San Lorenzo because social, institutional, and political determinants have strongly influenced objectives and performance; the project was in a continual state of change, yet the supervision of Loan 418-PE tended to judge performance as if appraisal assumptions were fixed. This raises interesting questions of what should be taken as the objectives at any point in time in a dynamic project situation; what type of corrective action can be effectively applied through a specific project; and whether conventional supervision strategy is profitable and realistic in terms of Bank staff time when divergence from appraisal becomes so great. 1/ The Bank classified San Lorenzo as a "problem project" in five of the nine years since that list was started: in 1968, 1969 and 1970, mainly because of problems with water allocation between San Lorenzo and Lower Pirua, project management, and other project features; in 1975 and 1976, mainly because of the delays in completing the last drainage works (which in the end were not completed under the project) and in procuring machinery and equipment. Satisfactory Government action regarding the factors mentioned above allowed the project to be deleted from the list in 1971. - 8 - 23. One difficulty with this evaluation is that Loan 418-PE has been effective for 11 years, and many problems have only historical significance because factors causing the problems and possible corrective actions were time specific. Furthermore, the current approach to project preparation and ap- praisal might mean that many problems would be avoided if present day quality standards were followed. The mission has therefore only reported lessons which could still be relevant to present procedures. Furthermore, Loan 418-PE did not finance a self-contained development stage or project although it was treated this way at appraisal; this led to a performance expectancy which was unrealistic because progress and divergence were often related to factors outside the influence of the loan. Appraisal 24. Appraisal of a follow-on development stage such as that financed by Loan 418-PE should critically evaluate the objectives and assumptions of the previous development stage to ensure that they were correct and are still valid. If this had been done with San Lorenzo, the overestimation of water availability might have been avoided; the technical problems associated with light soils in the Tablazo 1/ area would have been recognized; and agricultural development would not have been treated so lightly. It would seem unwise for an appraisal mission to introduce important elements at appraisal which have not been prepared and are subject to doubt; if this is unavoidable, sufficient flexibility should be built into loan covenants so that substantial changes in scope can be accomodated during implementation without procedural difficulties for Government or within the Bank. Project components should generally not be included if they anticipate results of major studies also part of a loan; treat- ment of the drainage component in Loan 418-PE was an expensive mistake. 25. When a major study is included in a loan, specific provision should be made to ensure that it can be properly reviewed; either additional time should be allocated over normal supervision factors or, better still, funds for an independent review should be provided through the loan. Technical as- sistance accounted for 30% of financing from Loan 418-PE (estimated at 14.5% at appraisal) and although supervision was excessive, no report produced through the loan appears to have been adequately reviewed so that conclusions are clear and readily available in the form of comprehensive evaluations. Major Changes in Project Scope 26. Renegotiation of Loan 418-PE should have been supported by a full reappraisal to ensure that the revised technical, social and economic feasi- bility was acceptable before the decision was made to change the project. Apart from renegotiation, there were important deformations of project scope which were never evaluated to establish divergence from project objectives and targets. It is recommended that divergence be systematically monitored and quantified as a specific supervision task to identify when a comprehensive interim review of a loan might be warranted. 1/ See para. 36 (viii) and diagram 3. - 9- Extension of Loan Closing Date 27. When the closing date extends substantially beyond that estimated at appraisal, a project should be re-evaluated to assess whether the assump- tions and objectives are still valid. With San Lorenzo, the greatly extended time schedule meant that basic technical and socio-political assumptions changed; this had an important effect on development feasibility but was never evaluated. Since 1969, divergence has been so great from appraisal that Loan 418-PE was virtually financing a new project which had never been appraised. It is recommended that a definite provision be made in a loan agreement for updating or at least systematically examining project feasi- bility after a certain time overrun (say, 50% of the estimated schedule at appraisal). Supervision 28. Supervision reports should indicate whether problems are project, sector, or national specific in character so that management can decide the appropriate strategy through which corrective action is likely to get results. 29. It is estimated that at least 250 man-weeks of supervision t me have been spent on Loan 418-PE, not including senior management time.l. Quality and dedication of Bank efforts cannot be questioned, but it is hard to justify the time spent in terms of impact on performance. It seems that the problems were such that normal supervision did not work because divergence became so great that in later years the supervision objectives became unclear. 30. The effectiveness of supervision under these circumstances was weakened by lack of continuity in supervision staff (analyzed in annex 1). It is concluded that normal supervision procedures are unable to cope with basic errors at appraisal or fundamental changes during implementation: specific procedures should be designed to deal with such contingencies. Perhaps a general budget allocation could be put aside (similar to the project preparation budget) so that special corrective action could be taken quickly and effectively when an exceptional problem status dictates such a need. San Lorenzo illustrates that substantial supervision time was wasted because project status was never thoroughly reviewed. Correspondence indicates that Bank staff were aware of the difficulties, but short-term constraints related to supervision budget or lack of manpower meant that a comprehensive status review was never undertaken; the consequence was a long-term ineffective use of staff supervision time. 1/ Records of supervision time spent on the San Lorenzo Project between 1955 and 1964 are not available; no records exist through Loan 114-PE. - 10 - Delays in Implementing the Development Stage 31. One lesson of special importance to Government is that the continuing strategy of completing the main works in water resource development before initiating and implementing the development stage which generates benefits inevitably means underutilization of substantial investments which can lead to major consequential problems, as illustrated by the San Lorenzo Irrigation Project. - 11 - III. OED COMMENTS 32. The PCR offers a satisfactory presentation and analysis of project history and performance. The factors that brought about the main problems associated with its implementation are properly identified, and the main lessons that may be learned from it are specified, usually explicitly. The audit supports the main findings and conclusions of the PCR. The following comments elaborate on some particular issues. The Chira-Piura System and its Development 33. Several references are made throughout the PCR to geographical locations - the Chira River, Lower Piura, Tablazo, for example; some criti- cisms are presented regarding the advantages of different approaches to developing the basins involved in which alternative sources of water and different areas are mentioned. Some readers may be confused as to what the various basins, rivers and areas are. Thus, it is worthwhile to include here a summary presentation of the Chira-Piura system, its development up to the present, and the rivalry between the older and newer settlement areas. In addition, this presentation will introduce an important issue raised in the next section. 34. Two neighboring rivers cross the northwestern Department of Piura. The Chira River, which lies to the north, carries more water, but irrigable soils are scarce in its hydrologic basin. The Piura River carries less water, and no water at all in some periods, but there are plenty of irrigable soils along its stream. Water can be - and has been - transferred between the two rivers, and there is no break in continuity between the lowlands of the two basins. Thus, the two basins can be - and have usually been - treated as a single system. 35. Three water sources and three areas are linked in the Chira-Piura system. The former are: the Chira River, the Quiroz River (although it is a tributary of the Chira River, historical and functional reasons require that it be defined as a separate component), and the Piura River. The areas are San Lorenzo (in the highlands between the three rivers), Upper Piura (along the Piura River, upstream from San Lorenqo), and Lower Piura (along the Piura River, downstream from San Lorenzo).1- The progressive, step by 1/ A fourth area (Middle Piura, downstream from San Lorenzo and upstream from Lower Piura) and a fourth water source (Chipillico River) exist, but they are very small and did not play any role in the development of the system. - 12 - step utilization of those water sources to develop these areas created pressures and stress that sometimes accelerated and sometimes slowed down the development of San Lorenzo and were instrumental in strengthening or weakening some of the forces and groups that fought for and against project implementation. 36. (i) Up to 1949, Upper and Lower Piura shared the Piura River waters, then the only source of irrigation water (diagram 1). Lower Piura had the bigger share (15,000 ha, as compared with less than 4,000 ha in Upper Piura) for historical (it was colonized long before) as well as poli- tical (its big landlords constituted a very powerful group) reasons. Inspectors were sent upstream from time to time to check that Upper Piura did not use more water than it should. Water flow in the Piura River was erratic, and therefore irrigated agriculture - the only possible kind - was speculative. (ii) Stage I (financed by the Government itself) transferred water from the Quiroz to the Piura River (below Upper Piura) to increase both the area irrigated in Lower Piura (up to 31,000 ha) and the reliability of water supply (diagram 2). It was completed in December 1953. (iii) Stage II (partly financed by Loan 114-PE) built a dam in San Lorenzo to store the water being transferred from the Quiroz River in order to further increase water reliability in Lower Piura and to irrigate 50,000 ha in San Lorenzo. (iv) The dam was finished in 1959, but land settlement in the San Lorenzo area, not included as a part of Stage II, was not carried out immediately. The powerful hacendados of Lower Piura pressed to obtain - and did obtain - as much as possible of the additional unused supply of water provided by the dam; the area irrigated in Lower Piura was hence further expanded to about 50,000 ha. (v) When water was short, Lower Piura farmers effectively con- trolled any development in Upper Piura. Now that there was plenty of it, they relaxed their control; this, together with conjunctive development of groundwater, allowed Upper Piura to expand explosively to some 18,000 ha from 1960 to 1964. In this way, the water diverted from the Ouiroz River and stored in San Lorenzo allowed an increase in the area irrigated upstream of the project area. (vi) When the Government attempted to actually irrigate San Lorenzo (and the Bank considered financing this Stage III), water for only 32,000 ha was left; the Bank and the Government accepted the political fact that Lower Piura had already appropriated the rest of it. - 13 - (vii) Development at San Lorenzo started effectively in 1963 (a small start had been made in 1961), fully financed by the Government. It proceeded at a fast pace in 1963 and 1964. Loan 418-PE was appraised in late 1964 and approved in mid-1965 to accelerate it (diagram 3). (viii) Agricultural problems in 1965 and unpredicted low water availability in 1966 and 1967 increased competition for water with Lower Piura and thus increased the political opposition to San Lorenzo. The project manager was changed; the new one was more amenable to Lower Piura's interests; the budget for San Lorenzo was reduced and, even so, "savings" were made through a reduced rate of development, staff was reduced and key posts were kept vacant. Moreover, most of the sandy lowlands developed under the project near the Chira River (the Tablazo area, made up of three districts: Algarrobo, Parkinsonia, and Somate Bajo) had to be abandoned: some Tablazo settlers gave up, the rest of them had to be relocated in other project areas not yet sold (diagram 3). Project implementation was slowed down. The loan was renegotiated and the loan agreement amended accordingly in March 1968. (ix) A new government took over in 1969. It carried out a land reform that destroyed the power of the Lower Piura hacendados. Furthermore, additional land was sought for laborers who could not be accommodated in Lower Piura; land was found besides the original area of the San Lorenzo project. A couple of years in which plenty of water was available (1974 and 1975) made it possible to settle them, associated in cooperatives, in 5,900 ha located outside the original project boundaries (Valle de los Incas; Valle Hermoso). (x) A dam was built in the 1970s on the Chira River, and a diver- sion canal was dug to transfer water to Lower Piura. Completion of the diver- sion scheme supplied Lower Piura farmers with as much water as they could use, but no more land was available; Upper Piura could now use all the Piura River water and further expanded to some 30,000-35,000 ha. The 27,600 ha settled in San Lorenzo under the project, plus the 5,900 ha developed outside San Lorenzo in 1974 and 1975, could now count on a reliable source of water in the San Lorenzo dam, for water would no longer be released from the dam for Lower Piura (diagram 4). Inappropriate Analysis of the Project's Regional Setting 37. The project was not properly appraised. The PCR highlights the poor analysis of water availability and of the soils to be irrigated; the lack of attention given to agricultural development; and the optimism in appraisal expectations of project management and Government. It also men- tions the lack of critical analysis of the assumptions and objectives of the previous project stages to check whether they had been valid at the time, and whether they were still valid then. - 14 - 38. This criticism, indeed one of the major lessons that this project offers, ran just short of tackling another major appraisal flaw: the lack of appreciation that at the time San Lorenzo was already one of three inter- twined areas in a single hydrologic system: the Chira-Piura system described above. The appraisal mission did realize that such a system existed, hence the provision for the Chira-Piura study in the loan agreement. The San Lorenzo irrigation settlement project design did not reflect this system's approach, however. The appraisal mission seems to have defined the problem as using the water available in the San Lorenzo dam to develop some land around it. It seems to have overlooked the fact that this water could substitute for and be replaced by the Piura River's water; the existence of alternative areas in both Upper and Lower Piura that could be irrigated with the same water, either directly or through substituting for other sources, was not mentioned; the relative advantages of irrigating new land in San Lorenzo as compared with increasing the area irrigated in either Upper or Lower Piura or both was not analyzed. Problems Resulting from Implementation by an Autonomous Unit 39. The relative advantages and disadvantages of establishing an auto- nomous unit to implement a Bank-financed project as compared with strengthening existing Government agencies has been an issue in the Bank, and between the Bank and many prospective borrowers, for a long time. Most certainly, this project could not have been carried out - at least in full or as rapidly - without such an autonomous unit. Nevertheless, this experience sheds light on some of the problems associated with such an approach to organization for project implementation. 40. After a couple of years in which project implementation - not yet financed by the Bank - proceeded at the fastest rate ever achieved (1963 and 1964), the growth of the project unit and of the project itself made them large enough to compete for resource allocation and prestige with other Government activities. Moreover, the settlers in San Lorenzo were being provided with inputs and services well beyond the reach of other farmers and settlers in the country; it was realized that such a project could not be replicated elsewhere in Peru. San Lorenzo began to be viewed as a privileged project and, as such, began to raise resistance; it became politically un- popular and support for the project collapsed. Eventually, the autonomous arrangement made for implementing San Lorenzo proved to be weaker than the regular Government units and the political establishment; The project manager was changed, its power reduced, its autonomy destroyed; road main- tenance, education, and health activities were transferred to the relevant ministries; the two experimental stations in the project area were trans- ferred to the regular Government research agency. 41. Most certainly, a lower key approach, based on more collaboration with the Government and less autonomy and self-accountability, would have raised less resistance and might have resulted in smoother - and eventually complete - project implementation. Since such an approach was almost out - 15 - of the question at the time of appraisal of this particular project, a formula providing for integration as soon as possible of the project unit into the regular Government agencies should have been worked out. Poor Accounting and Problems in Auditing Project Accounts 42. Project management kept outstanding records on physical achieve- ments - project archives were even established in San Lorenzo. In sharp contrast, financial accounts were poorly kept throughout project implemen- tation. While the independent project agency existed, accounting practices and procedures were very poor; cost accounting and cost control were completely lacking. After San Lorenzo was included in the Ministry of Agriculture's First Agrarian Zone, no separate accounts at all were kept for the project. Neither the supervision missions nor the completion mission could establish how much had actually been spent on the project. In some years, the project accounts were not audited at all. In other years, the auditors rejected the project accounts as unacceptable, or the auditing procedures did not meet Bank standards; several auditing firms were involved in auditing the project accounts throughout the life of the project. 43. In itself, this is a disturbing fact. Even though it partly financed developments in San Lorenzo, the Bank could not know how the loan proceeds were invested - or even if they had been invested in the project at all. Equally disturbing, however, is the lack of strong Bank reaction to this breach of important covenants in the loan agreement. The Bank did send letters to the Government expressing its discomfort on this matter - but to no avail. And that was all. No remedial action was taken; the closing date was even extended twice during a period in which no separate project accounts were kept. The Bank's Financial Contribution to Development of San Lorenzo 44. Loan 418-PE cannot be equated with a self-contained project, for it provided partial financing for capital and recurrent costs during the last 11 years of an ongoing development project that lasted for 15 years. How important was the loan's contribution to financing such a project? Not very, according to the actual figures: the loan proceeds financed about 17% of the total development cost of US$59 million.1!/ If only the period 1965-1975 is taken into account, the loan's share rises to above 27%. Total Project Costsi- Loan Proceeds- (mln Soles of 1975) (US$mln of 1975) (US$mln of 1975).! 1961-64 971.8 22.4 - 0 1965-75 1,594.7 36.8 9.9 27 1961-75 2,566.5 59.2 9.9 17 1/ Excluding the Basin Study. 2/ Disbursements by years were expressed in US dollars of 1975 and then added. - 16 - The most important project component was agricultural credit. Loan proceeds financed 62% of the medium- and long-term credit lent between 1966 and 1969 - the period during which the credit component of Loan 418-PE was effective, but just 19% of that lent during the whole project period. Actually, the loan proceeds simply helped fill a gap between the contributions of other donors - mainly USAID (table S3 and diagram 5). 45. The Government provided for all the additional resources needed to complete project implementation. Furthermore, the very extended implementation period increased the financial costs of the project, since the Government had to pay commitment charges on the undisbursed funds throughout the ten years it took to disburse the loan - even during the three years its rights to make disbursements under the loan had been suspended. PERU LOAN 418-PE SAN LORENZO IRRIGATION FROJECI MAIN EVENTS Development Stage Year Event Comment Stage I --1937 Study of the Quiroz Diversion rroject diversion to transfer water from Chira to Piura river basins. 1948-1956 Construction of Quiroz Diversion Project; Stage II ------ 1955-1958 Construction of San Lorenzo Dam and infrastructure; partly finance*d by Loan 114-PE. 1961 Start of the Colonization Phase, 1963 'General Preparation Plan' produced by San Lorenzo management; 1965 Appraisal of Loan 418-PE; serious loss of cotton crop due to storm damage. Nov. 1966 Dote of Effectiveness of Loan' 1965-66 very low water years. Jan. 1967 Memorandum of Understanding; reassessment of water availability and project objectives. Dec. 1967 Original Closing Date (extended); Mar. 1968 Renegotiation of Loan 418-PE; Loan reduced to US$6.5 million. Agrarian Reform Law; basic changes to socio/political objectives. Project became operational (under Bank in~volvement ended In components which directly Zona Agraria 1): contributed to settlement or development. Mar. 1969 Completion of Drainage Report; June 1969 Suspension of Disbursements (see para 3.09)6 Stage III Nov. 1970 Decision to treat Drainage component as a 'special technical project'; Mar. 1971 Comprehensive plan agreed for Drainage Project6 June 1972 Start of-construction of Poechos dam on river Chiras; Chira-Piura Diversion Project, First Stage,, May 1973 Sturt of Construction of drainage; June 1973 Revised Closing Date(extended); June 1976 Loan 418-PE Closed; Dec. 1976 Estimated completion of Phase I of Drainage; Dec. 1976 Comp letion of main works, Chira-Piura Diversion CnProject, First Stage; Feb. 1977 FAO/CP preparation mission Lower Piura Rehabilitation and Drainage Project (Irrigation Rehabilitation II). Table S2 - 18 - PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Summary of Targets at Appraisal, Renegotiation of Loan (1968) and Actual Achievements Targets at Targets at Renegotiation o Actual Aspects Appraisal (1965) Loan (1968)1 Achievement (1975)2 1. Net developed area - ha 32,000 18,000 27,600 2. Year in which land develop- ment target was completed 1967 1970 3.. Cropped area (physical) - ha 32,000 18,000 23,6003/ 4. Cost (196L-1976)-millions 1975 Soles 81.7 n.a. 59.2 5. Gross value of production: Total - million of 1975 soles 1,435 807 7655/ Per ha - thousands of 1975 soles 44.8 44.8 27.75/ 6. Rate of Return - %: Whole Stage II from 1955-76 14 n.c. 8.76/ 2/ Colonization phase from 1961-76 n.c. n.c. 17.5 Project from 1965-76. 169' n.c. 2 7. Number of settlers 1,400(max.)L 780/ 3,180 1/ The loan was not reappraised in 1968; only the new target development area was defined; other targets have been generated by the completion mission. 2/ Actual achievements do not include developments which depend on availability of additional waterresureq thrmugh the ChiErPiq;A rqect: these include 5.300 ha in Valle de Los Incas and Valle Hermosa served by San Lorenzo infrastructure and areas indirectly benefited in the Upper and Middle Piura Valleys (see para. 6.07, main text). 3/ Average cropped area for 1973-75 with cropping intensity of about 80%; maximum cropped area was 25,500 ha in 1973. The irrigated area is expected to match or exceed developed area now that increased water availability enables crop intensity to increase. 4/ The appraisal report referred to 37,500 ha developed by 1977; this production estimate has been proportionately corrected for 32,000 ha. 5/ Projected for 1980 to allow for full potential tree crop production; low estimate because incremental production due to increased water availability not credited to San Lorenzo but to Chira-Piura Development. 6/ ain sources of divergence: (i) smaller area than projected,(ii) lower yields of cotton, (iii) insufficient data to make full allowance for tree crop production on 5,700 ha. 7/ Should be regarded as an underestimate (see para. 6.07, main text). 8/ Calculated on basis of erroneous assumptions (see Annex 10 SDV) 9/ Not calculated because considered of limited meaning (see Annex 10 SDV). 10/ Not presented in appraisal or at Loan renegotiation; generated on basis of the data and objectives in appraisal report and other sources. Note: n.c. - Not calculated. Table S3 San Lorenzo: Gross Medium- and Long-tern On-lending - Sources of Financing (million Soles of 1975, percents) Gross Medium- Loan 418 as Percent of: and Long-term National Foreign Sources Medium/Long-term Foreign-financed Medium/ Year On-lending SourcesLa US AID Loan 418 IDB, IBRD/b on-lending Long-term On-lending 1962/c 5.4 5.4 - - - 1963/c 49.1 24.8 24.3 - - 1964/c 164.3 21.3 143.0 - - 1965/c 77.6 12.5 90.1 - - - 19667c 151.0 20.1 39.3 91.6 - 60.7 70.0 1967 95.5 28.9 6.6 60.0 - 62.8 90.1 1968/c 40.6 15.9 1.0 23.7 - 58.2 96.0 1969/c 43.2 12.6 - 30.6 - 70.8 100.0 1970/c 63.2 7.6 55.6 - - - 1971 62.7 3.5. 59.2 - - 1972 88.8 63.8 25.0 - - 1973 66.1 37.0 29.1 - - - - 1974 144.7 124.7 19.0 - 1.0 - - 1975 48.8 1.6 7.9 - 39.3 - - 1976/d 6.0 0.8 3.4 - 1.8 - - "Totals"/e 1962-65 296.4 39.0 257.4 - -- - 1966-69 330.3 77.5 46.9 205.9 - 62.3 81.4 1970-75 474.3 238.2 195.8 - 40.3 - - 1962-75 1,101.0 354.7 500.1 205.9 40.3 18.7 27.6 /a It includes Banco de Fomento Agropecuario (BFA), Instituto de Reforma Agraria y Colonizacion (IRAC), Rehabilitation Credit as established by DFL 193!J6, and other Governmental Funds (Supremo Gobierno). /b Loans 933-PE (IBRD) and 322 (IDB). /o The information for 1962 to 1965 covers the "agricultural year," from October 1 of the preceding year to September 30 of the year mentioned. For 1966, it covers the period October 1, 1965 to December 31, 1966. Figures presented for 1968 to 1970 are estimates based on information for the periods ended November 30, 1968; April 30, 1969; March 31, 1970; and December 31, 1970. d Up to June 30, 1976. e This is not an actual "Total On-lending" because it refers to gross on-lending instead of net on-lending. 뻐 =r O 塔 O 江 며 軫 : ,& 쌈！큉― · 戶 伶 柑u吧9 plJO紙aql 一TZ- DIAGRAM 5 San Lorenzo: Contribution of Loan 418-PE to Gross Medium- and Long-Term Lending in the Project Area S Min soles of 1975 tog Min soles to" of 1975 IDS 1 849 us D PAD So U S 4%o no 41b U$ Not "0 so vI ko U to cin Sa4m cv r1m 4 IT, U-11 1 1 sahdc n I I I 6z/s 64/1 70ts 94/1 7OIS Annual Contribution Aggregate Contribution / Sub-Period ATTACHMENT PERU LOAN 418-PE: SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT COMPLETION REPORT TABLE OF CONTENTS Page No. 1/ VOLUME I - THE MAIN REPORT I. INTRODUCTION ............................... .............. Al II. BACKGROUND ................................................ A2 The Project Area ....................................... A2 Previous Development .................................... A2 Project Preparation and Pre-appraisal ................... A3 III. THE PROJECT ............................................... A4 Appraisal ............................................... A4 Implementation History .................................. A5 Project Cost and Disbursement ........................... A7 Performance of Consultants and Project Staff ............ A9 Operation and Maintenance ............................... A10 Covenants ............................................... All Auditing and Reporting .................................. All IV. PERFORMANCE OF PROJECT COMPONENTS ......................... A12 Introduction ............................................ A12 Part A: Settlement and Agricultural Development ........ A13 Rate of Land Development ..................... A13 On-farm Development .......................... A14 Cropping Patterns, Yields and Production ..... A15 Agricultural Supporting Services ............. A16 Credit ....................................... A16 Research ..................................... A17 Livestock ...................... A18 Extension .................................... A18 Cooperatives ................................. A18 Marketing and Processing ..................... A19 1/ Summary, Conclusions and Recommendations were incorporated in the Project Performance Audit Report. TABLE OF CONTENTS (Cont'd) Page No. Part B: The Drainage Component ................ A19 Objectives .................................. Al9 The Main Study .............................. A19 Implementation .............................. A20 Conclusions ................................. A21 Part C: The Chira-Piura Basin Study ................... A21 V. ORGANIZATION AND MANAGEMENT ........................ A23 VI. ECONOMIC AND SOCIAL IMPACT ............................... A24 Economic Analysis ...................................... A25 Farm Income ............................................ A26 Social Impact ............................... A27 Training and Demonstration ............................. A27 Employment ...... ........ ............ A27 VII. COST RECOVERY ............................................ A28 VIII. FUTURE IMPROVEMENTS ...................................... A29 IX. SPECIAL ISSUES ........................................... A30 1/ Tables Annex 1 - Basic Data Annex 6 - Water Availability Map VOLUME 2 - SUPPLEMENTARY DATA VOLUME (SDV)2/ Annex 2 - Historical Background Annex 3 - Project Works Annex 4 - Settlement and Agricultural Development Annex 5 - Credit Annex 6 - Water Availability3/ Annex 7 - Drainage Component Annex 8 - Chira-Piura Basin Study Annex 9 - Organization and Management Annex 10 - Economic, Financial and Social Analysis 1/ Tables 6, 16-18, and 20-22 are not included. They can be found in the relevant annexes, Volume II, Supplementary Data Volume. 2/ Not included. Available from Agriculture 4 Division of the Regional Office. 3/ Included in Volume I also. PERU LOAN 418-PE: SAN LORENZO IRRIGATION AND LAND SETTLEMENT PROJECT COMPLETION REPORT I. INTRODUCTION 1.01 This completion report gives a unique opportunity to record and evaluate the progress of an irrigation and settlement project from construc- tion of physical infrastructure to completion of the development stage in its final operational form. It also provides an opportunity to evaluate perfor- mance of a perennial problem project over a greatly extended loan period and to assess what lessons may be applied to other projects in Peru and elsewhere. Of special interest is the substantial deformation of project scope since appraisal including a renegotiation of the Loan and the impact on performance of a Bank supervision input of over 250 manweeks since appraisal. 1.02 Only some of the many problems which afflicted the San Lorenzo Project during effectiveness of Loan 418-PE were specific to the life-span of the loan; many can be traced back to Stage II development of the San Lorenzo dam and the irrigation infrastructure partly financed by Loan 114-PE. Therefore, to make this report comprehensive, project history since 1955 has been considered with special (but not exclusive) attention to components of the Stage III development partly financed by Loan 418-PE. 1.03 For practical and analytical reasons in this report, all Stage III settlement and the development phase since 1961 is defined as the project even though Loan 418-PE only became effective in 1965. 1.04 The complexity and extended history of Bank involvement in the San Lorenzo Project, coupled with the basic changes in social/political determinants in Peru during the life of Loan 418-PE, means that this completion report must inevitably be comprehensive if it is to remain meaningful. Also, the impor- tance of San Lorenzo in the history of agricultural development in Peru indi- cates that an in-depth evaluation is justified. The report has therefore been prepared in two parts, a Main Report with key tables and a Supporting Data Volume (SDV) of annexes (for limited circulation). - A2 - II. BACKGROUND The Project Area 2.01 The San Lorenzo Irrigation Project is located in the Piura Region on the northern part of the Peruvian coastal plain about 1,000 km north of Lima (see map). The project area occupies an undulating plain with topography varying from flat lands or gentle slopes in the riparian areas to rolling rough topography. The irrigated area is not one geographical unit but frag- mented into districts which are located in two distinct areas and have con- siderable differences in soils characteristics (see map). There are no main towns within the project area, but four towns are easily accessible with a combined population of about 300,000; Piura, the regional capital, lies approximately 100 km to the southwest. Soils vary greatly from loamy sand to clay silts reflecting a wide variation in crop water requirements, pro- duction potential and thus repayment capacity. Soils are alluvial, deep and well drained; some areas have salinity and drainage problems generally due to inability of natural drainage to remove excess caused by occasional heavy rainstorms or more commonly wastage and release from the irrigation system. Drainage has not posed special problems in the San Lorenzo area although it was a major component of Loan 418-PE. 2.02 The impact area of the San Lorenzo Project identified for possible expropriation in the 1950s was 92,000 ha of which 45,000 ha were provided with irrigation infrastructure. By the time Loan 418-PE was appraised, the estimated irrigable area had been reduced to 32,000 ha. Through implementation history, this was reduced to 18,000 ha but then expanded to about 37,000 ha as explained in paragraph 4.02. 2.03 Although located in the tropics, the coastal zone has a desert climate. A full 365-day growing season is possible with irrigation, and there is almost total flexibility in the growing season for most crops. Previous Development 2.04 Historically, water has been the dominant constraint to the regional development region; however, favorable soils and climatic factors and easy access to surface supplies have enabled some irrigated agriculture to be prac- ticed along the river valleys even before the Spanish conquest in 1543. River diversion techniques governed the extent of development, and unreliability of water supplies due to large random variations in river flows made agriculture speculative. The source rivers, the Chira and Piura, are rainfed and although geographically close in the coastal plain (see map) have very different flows governed by catchment characteristics. The solution was to transfer water from the Chira to the Piura Valley, and in 1948 Government proceeded with a diversion project which later became known as Stage I of the San Lorenzo Irrigation Project. Rio Quiroz, a tributary of Rio Chira with reliable perennial flows, was diverted through canals, tunnels and natural creeks to Rio Piura. The objective was to provide reliable irrigation for cotton on about 31,000 ha - A3 - of existing irrigated area in the Middle and Lower Piura Valleys and expand the irrigated area by about 20,000 ha. Concurrent with Stage I construction, studies were undertaken to develop storage to utilize excess seasonal diver- sions from Rio Quiroz. 2.05 Stage II development consisted of an earth-filled dam with live- storage of about 258 MCM and irrigation infrastructure to serve about 45,000 ha which define the San Lorenzo Project area; these works were started in 1955, partly financed by Loan 114-PE, and were coppleted in 1959. The development cost was about US$28 million; US$18 million was provided through Loan 114-PE and about US$9 million under PL-480 from the United States. No definite deci- sions were made with respect to settlement except that land would be sold to qualified farmers in.lots from about 15 to 100 ha; sales were expected to realize about half the investment cost. 2.06 A fundamental fault with the Stage II development of San Lorenzo Irrigation Project was that storage water could not be used immediately for the 50,000 ha for which physical infrastructure had been constructed; the delay in formulating and implementing the colonization phase meant that from 1959 to at least 1962, water supplies meant for San Lorenzo were utilized in other parts of the Piura valley. The result was that by appraisal of Loan 418-PE, the irrigated area in Lower Piura valley had expanded to at least 40,000 ha and there were substantial increases in the irrigated areas in the Upper Piura valley and in the upper reaches of the Quiroz and Chipillico rivers. This significantly increased the competition for water and had serious political repercussions through the colonization phase of San Lorenzo. 2.07 The settlement and colonization phase of the San Lorenzo project, designated Stage III, thus really started in 1961; from this date the Bank expressed willingness to help finance the development. However, Loan 418-PE was not appraised until 1964, mainly because the Bank insisted on several pre- conditions with respect to organization and water legislation. The interim period was partly financed by the U.S. Government through PL-480 and AID. A full background history is given in SDV Annex 2. Project Preparation and Pre-appraisal 2.08 A General Development Plan prepared by San Lorenzo project management (ICSL) in 1963 was comprehensive but did not sufficiently re-examine important technical -parameters which affected the reliability of agricultural production such as water availability and soils. Nonetheless, to avoid further delay, the Bank appraised the project in October 1964 even though there were still unanswered questions. 2.09 The Government was uneasy with the way the appraisal mission mod- ified the preparation plan. In particular, a major drainage component was added with little supporting data and an important short-term proposal for increasing water availability to the project area was abandoned in favor of a comprehensive basin study (see paragraph 4.29). Loan negotiations were held in March 1965 before there was clear agreement between Government and the Bank on project scope; the major disputed item was the basin study, whose terms of - A4 - reference were a condition of effectiveness of the Loan. To avoid delay, Government agreed to the basin study, and a loan of US$11 million equiv- alent was approved April 20, 1965, and signed June 18, 1965. An interesting comment made by the Loan Committee was that the on-going project contained such a variety of components (not directly financed by the Loan), including health posts, schools, small-town development and rural roads, that it could be called integrated agricultural development; it might thus be said that Loan 418-PE was one of the first integrated rural development projects financed by the Bank. III. THE PROJECT Appraisal 3.01 Basic project data is summarized in the Basic Data Sheet and Annex I, with a detailed description of the project components given in SDV Annex 3. Loan 418-PE provided financing to accelerate the on-going land development and settlement phase of the San Lorenzo project and cannot be equated with a self-contained project. The appraisal report identified a need for components to further develop about 14,300 ha already settled and to accelerate develop- ment of about 17,700 ha of new land to bring about 32,000 ha into production by 1968. 1/ The package consisted of equipment to implement on-farm develop- ment works and a drainage program; technical assistance; provision of services and facilities for management; and funds to supplement an on-going medium- and long-term credit program mainly to cover on-farm investment. The cost was estimated at US$29.8 million, excluding interest during implementation, as detailed in Table 1. The foreign exchange component was estimated at US$5.6 million, and 47% of the Loan.was for local expenditure, as shown in Table 2. The cost of the regional basin study was estimated at US$0.6 million. 3.02 There was some difference in emphasis between the Bank and Govern- ment at appraisal; the Bank saw the Loan as a means of protecting the sub- stantial investment in the Stage II development; Government endorsed this objective but faced the political and practical difficulty of restricting development within the geographical boundary of the San Lorenzo project. The primary objective was settlement and only nominal attention was given to agriculture; no yield increases were projected and there was little consider- ation of the best cropping pattern to minimize water consumption or analysis to decide the appropriate level of on-farm development. Agricultural objectives were not defined directly in the appraisal report and no farm budgets were given. It appears that farmers were free to choose their own cropping pattern, which was governed by their annual allocation of water. The report did not consider the unreliability of crop production due to variation in water avail- ability, pest and storm damage (particularly to cotton) or the transition period to full farm productivity. To summarize, the appraisal report was vague and in places contradictory in assumptions, design criteria and objectives. 1/ The status of development given at appraisal does not agree with histor- ical records; see paragraph 4.03. - A5 - It was also optimistic in its expectancy from project management and Govern- ment, given the status of the agricultural sector in Peru at that time, and conceptually incorrect in its method of evaluating the project's viability. A major fault was inadequate evaluation of technical determinants built into project preparation; the assumption appears to have been that Stage II develop- ment objectives were still valid. The insistence that project management should be autonomous (discussed in SDV Annex 9) contributed to the problems that hampered project implementation. It is concluded that appraisal was rushed in an attempt to make up for the delay between completion of main infrastructure works and implementation of the settlement plan. Implementation History 3.03 Implementation was scheduled to take about 30 months from November 1965 and actually took 11 years. The closing date was revised four times for various reasons, some of which had nothing to do with original objectives; the loan was finally closed in June 1976. During this greatly extended period, the history falls into five distinct phases: Effectiveness and Start-Up: 1965-1967 3.04 There were two conditions of effectiveness: an agreement between project management and the Agricultural Bank (BFA) to manage the credit compo- nent, and the setting up of an authority (ORDEN) to manage and prepare terms of reference for the basin study. When the loan was declared effective on November 24, 1965, the terms of reference of the basin study had not been agreed. There were seven side letters, as listed in Annex 1; four of them were either superseded when the loan was renegotiated or ceased to apply be- cause of basic divergences in project scope. 3.05 Even before the loan was effective there were problems; the Project Manager mainly responsible for progress prior to appraisal was dismissed over a dispute related to management autonomy. By 1966, the Bank cited short- comings in management, extension and administration for poor performance, but progress was judged by optimistic appraisal assumptions coupled with an under- estimation of the institutional, practical and political difficulties of implementing an innovative colonization program in Peru at that time. In particular, there were serious delays in effecting technical assistance and major studies, possibly because project management and Government had been unprepared for these components. 3.06 Difficulties were compounded because agricultural production in 1965 was about 40% lower than in the year of appraisal due to storm damage. This affected the credit component because new settlers were unable to service their debts. Water availability was unpredictably low in 1966 and 1967, which increased competition for water with Lower Piura Valley and, thus, the political - A6 - opposition to San L,:enzo. By late 1966, it was clear that, because of water shortage, settlement of 32,000 ha in the San Lorenzo area could not be achieved without drastic and unacceptable political action. Memorandum of Understanding 3.07 In January 1967, a Memorandum of Understanding was agreed with Government which outlined corrective action to be taken. Government was asked to improve management, start the technical assistance program, review the credit program, and pass legislation to protect San Lorenzo's water right; meanwhile the Bank re-examined the hydrological assumptions. The conclusion was that poor performance lay mainly outside the control of project management and that the irrigated area would have to be limited to about 16,500 ha. At that date, 18,000 ha had been settled and developed; thus the conclusions implied that some lands had to be abandoned and the cost of re- settlement and compensation accepted. This seriously undermined the confi- dence of management and the faith of Government in the project; it also harmed settler relationships with the project staff at that time. Renegotiation of the Loan 3.08 The Bank conducted a review to decide how the loan could be recast, but there was no reappraisal or comprehensive statement of conclusions. Re- negotiation was conducted in October 1967; the settlement objective was reduced to 18,000 ha; US$4.5 million of the US$11 million loan was cancelled; and the closing date extended from December 30, 1967 to June 30, 1969. The renegotiated loan became effective on March 19, 1968, and resulted in a major change in the project scope as shown in Table 3. Categories related to settlement and management support were considerably reduced, which increased the financial burden on management when it was beset by a range of critical problems. The major reduction of the credit component had little effect, as the settlement program thought feasible at that time was almost complete. Funds for the drainage study were expanded to match escalating costs, and yet the category to provide equipment for potential drainage works was halved. The basin study category was 'increased to match the costs of supplementary studies identified by the Consultants but unanticipated at appraisal. 3.09 The project still managed to develop about 5,000 ha in 1968, and management established an improved working relationship with the settlers. In March 1969, the Bank suspended lending and operations activity to Peru for 12 months over the issue of compensation of expropriated foreign interests. The Government formulated a decree which reduced water allocation to San Lorenzo and weakened management control over project activities. The Bank decided to stop disbursement on the revised closing date of June 1969, ex- cept to complete the basin study. The only categories with substantial funds undisbursed were O&M equipment and provision for drainage works. June 1969 to January 1971 3.10 Although the main drainage report was completed in March 1969, no action could be taken because Government funds were not provided and the plan - A7 - to proceed was unclear. During 1969, unknown to the Bank, project eroblems were aggravated by settlement of some 1,100 squatters on 4,400 ha of irri- gated land, which created a new demand for water and agricultural credit. Reimbursements were resumed in March 1970, and by that November it was agreed that drainage would be treated as a "special technical project" with work carried out under force account. In June 1969, the San Lorenzo project became fully operational as part of an irrigation district under Agrarian Zone I; from this date Bank involvement in settlement and development of San Lorenzo ended except for drainage, which was effectively a separate project. 3.11 Government independently continued to expand the irrigated area using credit from other sources. Through 1970, about 1,700 squatters were settled and about 5,000 ha planted with three crops had started to come into production. March 1971 to Closure, June 1976 3.12 From 1971, the main objective of the loan was completion of the drainage works, which had been considerably reduced in scale from the area originally identified in the main drainage study as explained in paragraph 4.27. Difficulties with the drainage study, slow procurement arrangements and novelty of drainage works in Peru meant that disbursements did not start again until December 1972, making the revised closing date of December 30, 1973 unrealistic; construction of the drainage works began in mid-1973 and again the list of goods was revised by combining categories for 0&M and drainage equip- ment, thus reversing the decision of renegotiation. Once equipment arrived, drainage works proceeded well, with mAnagement enthusiastic and relatively well organized. 3.13 Through March 1971 to closure date, San Lorenzo management functioned successfully as part of Agrarian Zone I and has usefully spent remaining loan funds to purchase equipment. An extra 5,900 ha were settled and developed without Bank assistance by forming 1,390 squatters in cooperative units out- side the original geographical boundary of San Lorenzo. The Bank's role was to monitor the operation and supervise the drainage project; after 1973, the main concern was to spend undisbursed funds on items that appeared worthwhile, a major item being replacement of the butterfly control valve in the San Lorenzo dam which was unforeseen at appraisal; for technical reasons the valve had not been installed as of July 1976. The loan was finally closed on June 30, 1976. Project Cost and Disbursement 3.14 Insufficient data exists to permit a meaningful comparison between project costs at appraisal and closure because Loan 418-PE provided only partial finance of capital and running costs of an on-going development phase. The greatly extended time schedule and substantial changes in project scope make it difficult to quantify divergence from appraisal. Data sources are conflicting and often incomplete; management costs cannot be apportioned to various project components. The vagueness and lack of detailed costing at - A8 - appraisal makes the cost data of dubious value. The appraisal estimate of costs to complete development from 1965 (including expenditure on the develop- ment phase from 1961) were about US$82 million equivalent (in 1975 terms). Table 4 shows that costs to June 1976 were approximately US$59.2 million equivalent (in 1975 terms) excluding the basin study and some management costs. 1/ 3.15 One clear objective at appraisal was to develop and settle about 32,000 ha through credit financed by the loan; US$6.8 million was originally allocated, but loan funds for credit activities were reduced by 61% at re- negotiation because of the water shortage. Finally, US$2.7 million was dis- bursed through the loan for agricultural credit. Records show that 8,800 ha were developed while credit through Loan 418-PE was available, but 4,400 ha of developed land was abandoned during this period. 3.16 The loan component to help finance operation and maintenance equip- ment (not supported by a detailed list of goods) was estimated at US$1.2 million at appraisal but reduced to US$0.96 million at renegotiation; finally US$1.5 million was disbursed to replace items as and when a need was iden- tified. Analysis of objectives and achievements cannot be done because there is no data by which performance can be measured. 3.17 The loan financed US$1.0 million for technical services and supporting facilities and equipment, about 9% of Loan 418-PE at appraisal and 16% at renegotiation; only US$0.88 million was disbursed by closure (see SDV Annex 4). 3.18 It was assumed at appraisal that the drainage component would be investigated, formulated and constructed in three years under supervision by San Lorenzo management; costs of the works were not estimated at appraisal. Records do not permit drainage costs up to 1973 to be abstracted from general project costs; it is known, however, that the special drainage project since 1973 cost about US$1.94 million equivalent and the main study cost at least US$0.34 million (see Table 5); this compares with an appraisal allocation of US$0.9 million for equipment and an estimated cost of the study of US$0.25 million. 3.19 The final cost of the basin study was US$1.09 million (equivalent to US$2.3 million in 1975 terms), but this cannot be compared with the estimate of US$0.6 million at appraisal or the renegotiation estimate of US$0.96 million be- cause continual changes in scope were made as the study progressed. Table 3 details the study cost as estimated at renegotiation, but breakdown of final cost by various reports is not available. 1/ No records were available of administrative or other costs charged to San Lorenzo by outside agencies or through bilateral aid contributions. - A9 - 3.20 There were no definitive disbursement schedules at appraisal or renegotiation, and divergence means that direct comparison with appraisal objective is not possible; the final list of disbursements is given in Table 6. Disbursements by year and category are shown in Table 7 and clearly illustrate the loan's two distinctive parts: the period up to June 1969, and then no disbursements until December 1972 when expenditure commenced for O&M equipment and the drainage component. Analysis of dis- bursements shows that the foreign exchange element of final project costs was about 36% as of December 1975, compared with 53% estimated at appraisal but distributed over a completely different time period. The various changes and reallocations of loan proceeds are summarized in Table 3 and clearly illustrate the divergence which took place through the loan history. It is concluded that after 1969, Bank involvement in San Lorenzo ended except for the drainage project which was treated and implemented as a se- parate component. Performance of Consultants and Project Staff 3.21 Incomplete and often contradictory records and conflicting personal opinions make it difficult to make a full and impartial evaluation of pro- fessional performance during the life of Loan 418-PE. Errors of judgment were certainly made, but these should be weighed against the basic question whether project objectives and implementation strategy at appraisal were realistic; retrospectively, it would seem that the probability of events working out as predicted was never great and the performance of those involved in the project consequently suffered. 3.22 The major consultant inputs were the drainage study and the basin study and are described later. Given the indefinite terms of reference, International Engineering Co. (IECO) of San Francisco, which undertook both studies, fulfilled their commitments as defined in the contracts. However, inadequate data and the unrealistic time allocated to the consultants explains some lack of detailed analysis in the work and indefinite conclusions, particularly with respect to drainage. 3.23 Performance of consultants retained to provide other technical assistance tasks was generally inadequate, often because of unrelated problems which meant that the exercises were insufficiently planned or could not be efficiently tackled. The recommendations were generally difficult to imple- ment because they did not take account of national constraints outside the project which strongly influenced conclusions. 3.24 General criticism which apply to project management and consultants throughout the loan history are: (a) poor preparation and indefinite objectives at appraisal, coupled with the treatment of San Lorenzo as a geographically confined unit, led to expensive and time-consuming difficulties; - A10 - (b) continuous revision of project scope with little attempt to evaluate the consequences meant that objectives became confused and indefinite; (c) the project concept at appraisal was not flexible enough to cope with the technical errors or the basic socio-political changes that occurred in Peru through the development period; and (d) lack of continuity of senior staff in project management; the frequent changes of Project Managers are recorded in Annex 1. 3.25 Divergences which affected and provided excuses for consultants' performance are such that the mission feels that this completion report can only attempt to establish impartially the causes, consequences and lessons to be learned rather than undertake the unproductive task of apportioning responsibility; clearly all parties involved in the project since 1955 have been partly responsible for the many shortfalls in performance. Operation and Maintenance 3.26 Annual variations in physical determinants, technical factors and changing Government attitude are such that performance of operation and main- tenance is hard to evaluate except to state that it was consistently under- funded. A successful operational year was equated with good production although San Lorenzo management was generally powerless to influence perfor- mance; the dominant influence was timely and adequate irrigation. In good water years, operation tended to be lax because efficiency was unimportant and natural drainage facilities meant that wastage did not seriously increase salinity and drainage problems. Supply of equipment through the loan meant that minimum maintenance was always undertaken and recommended operational procedures were observed. When the annual budget was more generous, manage- ment performed reasonable preventive maintenance and now, 17 years after com- pletion of main works, there are no signs of serious deterioration, possibly due to the high quality of the original infrastructure. Production results do not indicate any variation or deterioration which- can be attributed to operation and maintenance. The only parts of the system with obvious mainte- nance liability are sections of the lined Tablazo canal, which was abandoned in 1969 and then partly rehabilitated in 1973. 3.27 San Lorenzo could have been better operated and maintained if the Government had complied with the relevant covenants in the loan agreement, pro- viding adequate funds and personnel. In particular, preventive maintenance of equipment and of the machinery pool has been impossible due to lack of adequate and timely budget allocations. However, to remove those constraints would have meant giving the project a more favored treatment compared to other parts of the agricultural sector in Peru; and this proved to be politically impossible for the Government. San Lorenzo management has persistently tried to implement recommendations and advice offered by the Bank when action was practical through the project. The operation and maintenance consultant provided the best technical assistance input financed by the - All - loan and the Peruvians followed his recommendations where possible, par- ticularly in purchase of equipment. The cost of O&M in 1975 for the San Lorenzo Irrigation District has been estimated at S/. 910 per ha compared to an appraisal estimate of S/. 1,160 per ha (in 1975 terms); both figures are provisional because of the difficulty of separating development and opera- tional costs. It is concluded that a comprehensive review of operation and maintenance would be useful to provide better facts for further analysis of the essential operational component. 3.28 Little attention has been given to water management, even in 1968 when the settlement objective was halved because of water shortage. No attempt has been made to assess the marginal value of water through an annual cycle. System efficiency appears to be low at present (less than 50%), leaving room for much improvement. No evidence was found to indicate that the operational characteristics of the San Lorenzo irrigation system had been cri- tically examined and evaluated since the early 1950s, when irrigation effi- ciency was assumed to be 60%; this is certainly high given unlined minor distribution canals, soil types, method of operation and irrigation techniques. Covenants 3.29 Two important inclusions in the loan documents were covenants designed to guarantee that San Lorenzo had adequate water allocation to match development objectives as seen at appraisal and to protect autonomy of San Lorenzo management. Both covenants resulted in major problems and attempts at enforcement tended to exacerbate the socio-political opposition to San Lorenzo. Water availability assumed at appraisal proved to be incorrect (see paragraph 9.01), and the attempt to isolate the project from other areas served by the San Lorenzo system was a continual source of dispute between Government and the Bank. Bank irsistence that the project director be approved before appointment never really worked and particularly chafed Government, which con- sidered the clause an infringement on its sovereign right and at one stage reported this to the Bank; frequent changes in the position and the basic change in management structure in 1969 meant that this condition fell into default. 3.30 The amendment to the loan agreement signed at renegotiation in March 1968 did not help because the validity of the original covenants was not re-examined. It is concluded that neither the original loan agreement nor the revised agreement was sufficiently comprehensive to form a realistic working agreement for this type of project. Auditing and Reporting 3.31 San Lorenzo accounts were audited first by CORFIRA (Corporacion Financiera de la Reforma Agraria) and then by Price-Waterhouse, a private company contracted to carry out audits for 1966 and 1967 and to recommend an auditing system. Price-Waterhouse completed the audit for 1966 and the first part of their study when the devaluation of 1967 led them to demand revision of their contract. As the proposed audit system could not be implemented anyway, - A12 - and in view of budgetary constraints and personnel limitations, it was decided to terminate Price-Waterhouse; 1967 accounts were audited by CORFIRA alone. In 1969 disbursements were interrupted and auditing of project accounts was discontinued. Before corrective action became effective, the problem was complicated in 1969 when San Lorenzo became part of Agrarian Zone I and was thus subject to Government auditing as a typical irrigation district. An additional difficulty occurred in 1971 when the Bank requested separate accounts for the drainage program in a format that could be independently audited to Bank standards. It was agreed that the 1971 and 1972 accounts would be audited simultaneously in February 1973, and finally a report was received from the independent auditing firm, Oscar Caipo y Asociados of Peru, for the period up to December 31, 1973; as no separate account had been kept for project funds, the auditors were not able to prepare an adequate auditing statement. Government argued that maintenance of separate accounts was in conflict with the law,.L/ but on Bank insistence a sub-account was said to have been opened in the accounting system of Zona Agraria I. No record of this could be found, and thus since 1973 use of project funds has not been prop- erly audited; prior to 1973, auditing and accounting did not match Bank standards. 3.32 Quarterly reporting has been poor, both in content and timeliness, and generally of little value to establish how the project concept was chang- ing. The situation worsened after 1965 when management changes occurred, the project future became uncertain and loan disbursement was suspended; when San Lorenzo became part of Zona Agraria I in 1969, the reporting procedure ceased to be effective. However, management did produce a regular series of comprehensive annual reports. One notable achievement of the project is the well organized archives dating back to the early history of the project. Lack of continuity in management and the Bank meant that reporting requirements and procedures inevitably suffered, but this was a minor problem compared to others affecting the project. Zona Agraria I did produce useful summary docu- ments covering development history which have greatly assisted in the prepara- tion of this completion report; a similar document was produced to cover the drainage component. IV. PERFORMANCE OF PROJECT COMPONENTS Introduction 4.01 The project consisted of settlement and on-farm development, a drain- age component and the regional basin study. Although they all directly or in- directly influenced the San Lorenzo area, for practical purposes it is easier to evaluate performance by treating them as separate elements. 1/ Loan Agreements legally have the status of international treaties. Therefore, the covenants in them are valid and enforceable even if they contradict national law. This principle is reflected in the loan regulations that were incorporated as part of the loan agreement which is binding on the parties. Government's argument was therefore unacceptable in principle, although in fact the loan agreement did not call for separate accounts (only adequate accounts). - Al3 - Part A: Settlement and Agricultural Development 4.02 The Stage III colonization phase was to incorporate uncultivated lands into irrigated agriculture utilizing the irrigation works built in 1955-1959. This involved two steps: allocation of plots to farmers (settle- ment) and on-farm development, and is detailed in SDV, Annex 4. 4.03 The target settlement objective has changed considerably through the project history; the appraisal report for Loan 114-PE estimated a settlement objective of 50,000 ha in 1955; this was reduced to 45,000 ha in the early 1960s and then to 37,500 ha in 1963; it became 32,000 ha when Loan 418-PE was appraised in 1965 and finally 18,000 ha 1/ at renegotiation in 1968 when water availability was reassessed. The appraisal objective was to reach 32,000 ha in three years, which implied a development rate similar to the pre-appraisal period of 1963-1964. Some 16,700 ha had been sold by appraisal in 1965 of which 13,300 ha were developed sufficiently to be cultivated. Rate of Land Development 4.04 Settlement progress prior to and during Loan 418-PE is given in Table 8, and the five distinct phases are summarized below: (i) the early phase of 1961-1962: initially about 2,200 ha were settled and provided with on-farm development for priority settlers displaced in the project area; (ii) the main development phase of 1963-1966: some 21,000 ha were developed at about 5,000 ha per annum; (iii) the relocation phase of 1967-1971: about 7,000 ha of developed land in the Algarrobo and Parkinsonia Districts (see map) were abandoned because of high water requirements and coarse soil texture. Most farmers were relocated on new farms in the upper sectors of the project area; (iv) the second main development phase of 1969-1970: about 7,000 ha were developed with 6,100 ha in 16 new producer cooperatives; 2,300 ha were outside the original project boundaries; (v) the stabilization phase of 1971-1973: no major developments took place; development of new lands compensated for abandonment of about 700 ha in the Somate District (see map) due to drainage and salinization; and (vi) the third main development phase of 1974-1975: about 5,900 ha were developed in the Valle de los Incas and Valle Hermoso, outside the original project boundaries. Although supplied from the San Lorenzo irrigation system, the expansion cannot 1/ 18,000 ha net cultivated area; equivalent to about 23,000 ha net irri- gated area at the average cropping intensity of 80%; see paragraph 4.09. - A14 - be considr ed part of the project under analysis since develop- ment was due to additional water supply from the Chira-Piura Project 1/ to the lower Piura Valley which previously used supplies from San Lorenzo. 4.05 In total, 36,661 ha were developed (excluding the 5,900 ha in the Valle de los Incas and Valle Hermoso) of which 9,071 ha were abandoned, giving a net developed area of 27,590 ha, as summarized in Table 8; cumulative areas sold are recorded on Table 9. Development after 1968, which contravened the limit of 18,000 ha agreed at renegotiation of the loan, was forced by necessity; squatters occupied lands and had to be settled by law. Fortunately the expansion coincided with a series of good water years which continued until alternative water supplies were available. The other factor was that water availability had ceased to be a political problem because Government expro- priated the large landholders in the Piura Valley, thus eliminating opposition. On-farm Development 4.06 The on-farm development included some agricultural inputs and a complete development package to enable new settlers to live on the farm as follows: () a detailed farm plan prepared by project management; (ii) construction of on-farm facilities and housing either through private contractors under management supervision or force account by the project; (iii) medium- and long-term credit for on-farm capital investment subject to approved project design and budgetary estimates; (iv) farming and cropping patterns prepared by project staff taking into account farmers' preference and supported by short-term credit facilities; and (v) technical assistance and guidance through extension staff. 4.07 The strategy gave preference to small settlers with land holdings up to 10 ha because medium-sized farmers tended to be more experienced and wealthy; a social objective of San Lorenzo was to help those most in need. Land distribution by holding size is given in Table 10; early settlers averaged about 16 ha compared with cooperative holdings of about 4 ha per family. Generally the development strategy proved to be effective except in areas with unforeseen technical problems such as the Tablazo lands; no major problems have occurred between assignment of land to a settler and effective cultivation; there were some minor delays due to provision of credit and limited management capacity. 1/ Partially financed by an IDB loan for land development. - A15 - Cropping Patterns, Yields and Production 4.08 The appraisal projections of cultivated area are contradictory since cropping pattern and yield data refer to 37,500 ha (see Table 11) and not the target development area of 32,000 ha. The proposed farming pattern was 70% annual crops with 40% cotton followed by beans, corn, sweet potatoes and yucca; 13% was projected to tree crops (5,000 ha) and 17% to pastures. Annuals other than cotton and yucca were assumed to be double cropped. The proposed farming plan was not discussed in the appraisal report and the above data has been deduced from tables; the omission could be because the project was developing and the on-going agricultural plan was simply accepted. 4.09 The cultiva.ted area increased at a rate similar to the net increase in developed lands except in 1968, when a drought resulted in the substantial crop failure; there has been no lag between development and full land use, a major factor in the unexpected economic performance. The cultivated area reached a maximum of about 25,500 ha in 1973 (see Table 12) but never reached the total irrigable area in any year. The average cropped area in recent years has been about 75% annual crops, predominantly cotton followed by rice and maize, 23% tree crops and only 2% pastures. The main tree crops are mango, followed by citrus and avocado. Average cropping intensity has remained only about 80%, possibly due to restricted water availability particularly in the second season when base river flows are low: the long cotton season is also thought to physically restrict double cropping. 4.10 Crop yield data are unreliable but there does appear to be a decline over the years, as shown in Table 13. For example, cotton yields have only once approached the appraisal estimate of 2.5 tons/ha and have decreased as follows: Average Cotton Years Seed Yield (tons/ha) 1954-1958 1.8 1960-1968 1.6 1969-1975 1.4 Little significance can be read into this data because annual variations appear to be influenced by parameters outside project control; in recent years, there might have been a tendency to over-spread a limited water supply; it is also known that there is low productivity in the areas recently developed outside the original project boundaries due to soil quality. No reliable data is available for tree crop yields, as most plantations are only now coming into full production; again quality and yields are thought to be low. 4.11 Restricted and unreliable irrigation is considered the dominant reason for poor agricultural performance. Since 1969-1971, water was allo- cated to expansion of irrigable lands rather than intensification of produc- tion on existing areas to satisfy pressure for land after the agrarian reform law of 1969. Other constraints have been management capacity and technical difficulties in dealing with new collective enterprises established since 1970 - A16 - and a decrease in effectiveness of the extension services. Insufficient data exist to attribute significant losses to salinization. 4.12 The appraisal report treated agricultural development lightly. No yield increases were assumed and no specific agricultural input was provided; incremental production through increased yields was left up to individual farmers. Furthermore, the dynamic management prior to Loan 418-PE probably created a false sense of optimism at appraisal that management would react when an improvement opportunity was identified. Agricultural Supporting Services 4.13 Provision of agricultural services appears to have been considered of secondary importance to the settlement and development objectives. There is no record of comprehensive programs to improve crop varieties, cultural techniques or efficient water management. The appraisal report does not dis- cuss mechanization even though farm sizes indicate this need, and 20% of long- term credit was for farm equipment and implements. Little attention was given to chemical inputs or improved seeds even though cotton, the main cash crop, suffered regularly from pest damage. The implicit assumption was that farmers could cope without a special project input or that as problems developed management would respond to the need. Loan 418-PE did provide technical assistance for various agricultural services but none of these inputs were fully successful. Credit 4.14 The major productive element in the Stage III development of San Lorenzo was credit because for most settlers it was essential to implement on-farm development and establish effective farming units. The credit programs were already designed and fully operational before Loan 418-PE became effective. The medium- and long-term (MLT) credit component in the Loan was to make up the shortfall in the on-going program (partly financed by AID) so that 32,000 ha could be settled and developed. Finance through Loan 418-PE was restricted to MLT credit for initial land development including housing, machinery and capital equipment, livestock, and tree crop investment, as detailed in Table 14. Credit was designed as a revolving fund. The program was administered by Banco de Fomento Agropecuario de Peru (BFA) and serviced through outlets in the project area. 4.15 Gross MLT credit on-lending from 1962 to 1975 was about S/. 1,100 million (in 1975 terms) of which Loan 418-PE provided only 18.7% (see Table 15); MLT credit per unit area has averaged about S/. 38,000 or US$880 (in 1975 terms), but major divergencies in settlement and development objectives make it difficult to find significance in average results. It is concluded that the credit program was generally successful for areas where farm profitability was reasonably assured. - A17 - 4.16 The repayment records of credit and land sales were closely linked with annual variations in crop production; unreliable irrigation made farming speculative and poor soils in some areas made it impossible to achieve pro- jected farm incomes. In good agricultural years, farmers serviced their debts and showed responsibility; this justified the comprehensive credit facilities to new settlers, many of whom had little previous experience in farming or farm economics. 1962 to 1964 were excellent crop years and repayments for short-term loans were almost 100%. 1965 and 1966 were particularly bad years; untimely rain caused flooding in 1965, and heavy pest damage to cotton occurred in both years; credit repayments fell to 33% and 50%, respectively, and Government announced a moratorium. 1967 was a fair year and repayments of short-term loans rose to 70%, but by the end of 1967 a large percentage of farms remained in debt. In 1968, BFA consolidated all outstanding debts to be repaid at 11% over a 17-year period; most settlers were regarded as fair credit risks. 4.17 Farmers with less than five ha had extreme difficulty in repaying medium- and long-term credits, indicating uneconomically sized farms. Medium size (15-60 ha) farmers performed reasonably well and enjoyed income suffi- cient to meet commitments and accumulate reserves. Large farmers tended to live beyond their means by using profits from good crop years for non-farm purposes, and some were in default of settlement regulations by being absentee owners. Repayment for land sales, although behind schedule, is currently approaching the annual recovery target - a notable achievement of cost recovery in Peru. Research 4.18 Two experimental stations were operated under the San Lorenzo Project at the time of appraisal but were poorly managed and provided little useful information. The Loan provided for technical assistance to study the on-going programs with a provision to implement the recommendations on a continuing basis for at least two years. Apparently unaware of this loan provision, project management handed over the two experimental stations to SIPA (the national research agency) in 1966 under an agreement which appeared to give priority to regional rather than project-oriented research. Despite Bank objections, no corrective action was taken except an attempt to interest a research team from North Carolina State University, then engaged by USAID and working in close cooperation with SIPA and the Agricultural University of La Molina. The more pressing problem of renegotiating the loan made research a side issue; SIPA's view was that research should be directed to the needs of the Northwest of Peru as a whole because there were insufficient resources to support a project- oriented program. By mid-1968, the Bank concluded that it was unrealistic to pursue the appraisal objective and the comppnent was dropped on the assumption that regional research facilities could be established through the Chira-Piura Project. As a result, valuable lead time has been lost and little has been done to improve low yields or to facilitate diversification out of cotton, considered essential for the long-term prospects of the project. - A18 - Livestock 4.19 Milk production was envisaged as a major economic activity in San Lorenzo and breeding stock were imported from the USA as foundation stock. Although good markets existed in the regional towns, management capacity of the settlers was overestimated and standards and yields have been very low. Many owners had no previous knowledge of cattle and few project staff had practical experience in animal feeding and breeding. Technical assistance was provided by a Danish team and a Swiss livestock specialist, but settlers appeared to have been unable or unwilling to specialize in livestock. Cattle were dispersed over the project area, making it difficult to provide technical assistance, supervision, or milk collection. By 1968, credit financed by the Loan was terminated for -livestock, and foreign experts concentrated on manage- ment, feeding and milk collection and marketing. Since then, there has been a nominal increase in livestock, mainly on cooperative farms. Extension 4.20 San Lorenzo management was responsible for extension, which was pro- vided as a service available to all settlers. At appraisal, extension was said to be yielding benefits in terms of better farm planning, higher yields and a more rapid adjustment of settlers; it was thus not seen as a problem and no direct input was provided through the loan. It was anticipated that extension efficiency would be improved as research results became available. One change did result in 1968, when management improved contact with farmers by reorganization so that one official dealt with all administrative, agri- cultural and credit matters. This intensified operations, but quality of extension inevitably suffered because field staff tended to become generalists. The extension program needs to be improved and expanded if full agricultural potential is to be realized. Cooperatives 4.21 An important cooperative function was formulation of self-governing, self-financing organizations controlled by farmers to permit bulk purchase of inputs and marketing. Project management played no part after an initial input except to provide technical assistance through Danish specialists and credit facilities. Success stemmed from allowing the movement to grow naturally rather than force the pace, and by 1967 about 50% of the settlers were cooperative members. The major breakthroughs were bulk purchase of fertilizer from Europe (saving about 30%); leasing of a cotton ginnery; livestock feed manufacture; a cooperative milk plant and a savings program. 4.22 The second cooperative fundtion unforeseen at appraisal and not financed by the loan was the organization of settler production cooperatives under the Agrarian Reform Law of 1969. Legally squatters who had drifted into the project area during drought years of 1967/68 had to be settled. Most occupied previous undeveloped areas, while some settled on abandoned or neglected farms. Each squatter was allocated 4 ha, and by 1975 about 7,650 ha - Al9 - had been developed in 56 production cooperatives financed by ADB. :.ttlers contribured labor and San Lorenzo contributed infrastructure, water, and extension supervision. The expansion of the irrigated area was recognized by Government as a technical and political gamble but was unavoidable because of the political emphasis placed on agrarian reform. Although the expansion was not in accord with the Memorandum of Understanding, the Bank had little choice but to accept it. The settlement of some 1,900 families has generally been successful except for technical problems in some areas, but it has not matched the agricultural performance of individual settlers stimulated by personal incentive; nonetheless it has had important social impact in directing project benefits to the very poor. Marketing and Processing 4.23 Marketing was not a problem at appraisal because local markets could absorb all annual crops and dairy products. Export markets were well estab- lished for cotton fiber production, and adequate processing facilities were available. The problem foreseen at appraisal was marketing of three crops and vegetables and technical assistance was financed through the loan. Little was done until 1973, when the Bank pressed for action because some of the 5,000 ha planted with tree crops at that time were coming into pruduction; the area with tree crops has since expanded to 5,720 ha (see Table 12). A feasibility study for a fruit development and marketing project was proposed but nothing resulted even though loan funds were reallocated for the study. The poor fruit quality and limited market for fresh produce indicate that efficient marketing and perhaps processing will be essential if full benefits from the substantial investment in tree crops are to be realized; action needs to be taken urgently to rectify this serious project shortcoming and evaluate the potential regional impact. Part B: The Drainage Component Objectives 4.24 The drainage component was formulated by the appraisal mission, possibly based on a Bank reconnaissance report prepared in 1961. A two-stage strategy was proposed: a comprehensive study followed by a works program to cover about 5,000 ha said to be critically affected; it is not clear how the area was decided upon before an investigation. The proposed schedule was optimistic, with twelve months to complete the study followed immediately by construction to be completed in two years. A detailed review of the drainage component is given in SDV Annex 7. The Main Study 4.25 The consequences of inadequate preparation soon became obvious, and it took 17 months to retain Consultants, International Engineering Co. (IECO), for the main study undertaken in three stages: a reconnaissance level survey; - A20 - a preparation study for critical areas selected in the first stage; and finally detailed design and supervision of works proposed in the second stage. The first stage was completed in 1967 and identified 19,200 ha with drainage problems from which 8,200 ha were identified for the preparation study. The preparation study was completed in March 1968 and identified 2,500 ha critically affected by salinity and waterlogging and proposed a surface collection system of main and lateral drains; results of the inves- tigations are given in Table 16. Sub-surface drainage was initially recom- mended for 600 ha in the Somate District (see map) which was subsequently abandoned. The study correctly identified the main problem as removal of excess surface water from precipitation and irrigation waste. There were some technical and analytical defects with the IECO report, but the main fault was lack of detailed definition of objectives and design criteria to allow implementation to proceed without substantial further planning. The investi- gations were not undertaken over a period significant enough to give statis- tically reliable results; this defect was due to unrealistic planning sche- dules and lack of adequate data. 4.26 The IECO study was not critically reviewed in 1968 before decisions were made to continue with the program, possibly because publication coincided with the critical stage when Loan 418-PE was renegotiated. Nonetheless, at renegotiation, funds for drainage works were substantially reduced although the allocation for the study was increased (see Table 3). IECO completed the final detailed study on 600 ha by March 1969; unfortunately, however, at about that time Bank operations were suspended in Peru for twelve months; the IECO contract was not extended and the last part of the IECO work is not available in report form. Implementation 4.27 By November 1970 the Government designated drainage a "special technical project" to be implemented on force account by the Directorate General of Irrigation and Water supported by CENDRET (Centro de Drenaje y Recuperacion de Tierras) and Dutch technical experts. Thus, from early 1971, San Lorenzo manage- ment ceased to have a direct involvement in drainage. It took until early 1973 for CENDRET to prepare, agree,and implement a comprehensive plan of operations. They concluded that only about 3,000 ha had drainage problems, compared with 19,200 ha originally identified by IECO. From experimental plots they discovered that leaching would successfully remove salinity over substantial areas, but that their natural drainage would permit this to be done without any need for artificial sub-surface drainage as originally envisaged for most of those areas. The review did not recalculate the economic feasibility of the drainage component or consider abandoning the limited areas seriously affected by salinity; this was a viable possibility because land was not a constraint. Construction began in mid-1973, seven years after the loan became effective. The main and secondary drains and structures were completed by late 1976 at an estimated cost (since 1973) of about US$1.95 million equivalent, as detailed in Table 17; this is close to IECO estimates but there is a major divergence in scope of works. Drainage equipment financed through the loan cost about US$ 0.62 million, as detailed in Table 18, compared with US$1.25 million estimated at appraisal and the revised estimate of US$0.38 million at renegotiation; in later years, - A21 - loan categories for O&M and drainage equipment were combined; all equipment has proved useful except a Hollandrain tile laying machine which was not financed by the loan. Conclusions 4.28 The drainage component was the main reason for extending the loan after 1969; poor planning resulted in an uncontrolled program based on a study which was incomplete and not adequately reviewed until CENDRET undertook the task and reformulated the component. This mission concludes that Loan 418-PE might have financed only a study, perhaps expanded to cover the much more serious drainage problem in the Lower Piura Valleys, without providing financing for works which anticipated the results of the study. A comprehensive drainage project might then have been formulated by 1970 and evaluated separately so that feasibility was clearly established before implementation. The methodology of the economic analysis in the IECO drainage study cannot be supported, and now insufficient data exists to establish the impact area or incremental benefits which might be realistically attributed to drainage; an ex-post evaluation of the drainage component cannot be made, but further studies should be undertaken to complete this important task. Part C: The Chira-Piura Basin Study 4.29 The concept of a regional basin study was also introduced at appraisal with little support by Government, which recommended construction of a Chira-Piura Diversion Project as formulated by a local firm of consul- tants (Flores y Costa) in 1964. Their objective was to increase water avail- ability quickly so that the Middle and Lower Piura Valleys were provided with alternate supply, thus permitting full development of the San Lorenzo area. The appraisal mission rejected the proposal because of serious salinity prob- lems in Lower Piura and questioned whether the proposed diversion scheme was the most economical use of River Chira water resources. Although Government stressed the political urgency of the interim solution, the diversion plan was finally rejected because Peru was unwilling to negotiate with Ecuador over the international water aspects of the project. The offer of the Bank to act as friendly intermediary was rejected by the Government, which insisted that the riparian issue was of negligible significance since Ecuador could not utilize the water. The Bank consequently deleted the Chira-Piura Diversion Project from the active list of potential projects although it endorsed the project as offering prima facie evidence of quick returns. 4.30 The appraisal mission endorsed a comprehensive study to assess water and other resources on a regional rather than a piecemeal basis, and US$0.6 million were provided to finance consultants. The addition of the study by the appraisal mission took the Peruvians unaware and there was confusion in formulating terms of reference., It took until early 1967 (14 months after effectiveness) to contract International Engineering Co. (IECO) and agree to an original bid of US$0.56 million; this escalated to US$1.09 million as addi- tional tasks were identified. The result was four reports which, together - A22 - with a parallel study inanced by USAID, documented a development plan; there was much duplication and some conflicting conclusions as the work progressed. One problem was that the reports had to be considered as a package, making them difficult to evaluate; basic assumptions and thus recommendations changed as the work progressed, and reports for various development stages overlapped; this resulted in acceptance of basic strategy assumptions before they had been fully evaluated. In particular, agricultural objectives were treated superficially and there was no clear analysis of incremental benefits and costs for various levels of development within each recommended sub-project; marginal analyses might have resulted in a substantially different package, particularly as the final plan was considered too ambitious for Peru by the Bank. 4.31 The studies recommended as a first-stage development a Chira-Piura Project with a storage dam on River Chira and a diversion canal to the Piura Valley. This report was completed by December 1968, but appraisal was delayed until April 1970 because of Bank suspension of operations in Peru through 1968; the appraisal faced problems: the international water issue was un- resolved, basic socio-political parameters assumed in the IECO studies had changed with the sweeping agrarian reform legislation in 1969; and status of San Lorenzo had to be reconsidered because it became part of the proposed project. 4.32 A draft appraisal report was said to have been prepared- but mean- time the Peruvians announced that the Yugoslavian Energoprojekt Company would undertake the project as a package deal complete with financing. Construction of the Poechos Dam and diversion canal was completed in 1976 at a cost of over US$200 million, compared with US$54 million estimated by IECO for a smaller scheme. Further development phases are planned which generally follow the recommendations of the IECO study. 4.33 The regional basin study and supporting studies were valuable but helped delay until 1970 implementation of a Chira-Piura Diversion Project. The marginal value of the additional information gathered through the Basin Study is questionable considering the time lost given the unprecedented low water years from 1965 to 1968. Perhaps a better solution would have been to formulate and implement a project using the extensive material available at the time of appraisal of Loan 418-PE. In addition, the marginal improvements to increase the water availability through the San Lorenzo system identified as technically feasible in 1965 might also have been included at appraisal. The present priority is to implement the development phases of the Chira-Piura Project, covering about 142,000 ha in the Chira and Piura Valleys, to utilize the substantial sunk investment in the main works with special emphasis on the rehabilitation of the Lower Piura Valley. The basin study is described in SDV, Annex 8. 1/ The document has not been found. - A23 - V. ORGANIZATION AND MANAGEMENT 5.01 The organizational structure for San Lorenzo was set up in 1963 with the establishment of Irrigacion y Colonizacion de San Lorenzo (ICSL), an independent agency under direct control of the Office of the President. A Supreme Decree of 1964 granted operational autonomy and funds; this satisfied the Bank,which reasoned that Stage III Development would be implemented even if it confronted strong political opposition, which would be more effective against in-line Government agencies. It was also felt that given the innovative development package, some form of centralized administration capable of making and implementing decisions was essential. The Director General at the criti- cal preparation stage for Loan 418-PE was a particularly strong personality able to effectively take problems to the President and get results unique in the agricultural sector in Peru at' that time. The management strategy produced favorable short-term results but inevitably led to problems because it depended on personalities and patronage, both transitory factors. A special problem was that although management was unified, well structured and dedicated to project success, this tended to be to the exclusion of outside priorities. 5.02 Therefore, although at appraisal San Lorenzo management was an effective team, in retrospect, it was optimistic to assume that this would last given the changing institutional and socio-political determinants in Peru. Even though isolated administratively, San Lorenzo was always highly political and there is evidence that prior to appraisal, inadequacies in tech- nical and economic assumptions might have been played down for fear of giving advantage to opponents and compromising loan negotiations. Furthermore, the loan introduced components which could not be undertaken by an autonomous manage- ment unit such as the basin study and the comprehensive drainage component with which management had limited experience. In addition, achieving the autonomy envisaged by the Bank meant concentrating resources and personnel on one project beyond available resources in Peru. 5.03 Given these problems, it is a tribute to San Lorenzo management that it achieved so much up to 1968 when the Loan was renegotiated. It trod a difficult path between hostile political opponents, central authorities un- comfortable with management privilege, and the Bank pressing for action on a wide range of complex issues much of which was novel. In addition, unforeseen technical problems meant that fundamental assumptions and objectives changed, imposing substantial strain on management, and there was a serious lack of continuity in key positions (see Annex 1). Performance should be judged against the reality of the on-going situation and not optimistic assumptions at appraisal. 5.04 Many reasons for poor management performance were outside project control; there was a general lack of experienced personnel in Peru, which meant that key staff were often promoted from San Lorenzo just when they were - A24 - becoming effective; budget restrictions coupled with tight regulations for interdepartmental staff transfers and freezing of salaries meant that San Lorenzo could not attract staff necessary to fulfill Bank expectancy. Cooperation between San Lorenzo management and outside agencies was weak and confused partly due to lack of staff continuity. Legally the project had autonomy but also a high dependency upon the goodwill of a number of Government agencies jealous of the privileged position of the project. For an extended period the Bank concluded that Government either could not or would not heed the project covenants or appreciate the needs and problems; Government felt that the Bank did not appreciate the difficulties involved in giving San Lorenzo favored treatment to satisfy previous commitments. 5.05 In 1968, the new Government gave the Ministry of Agriculture full responsibility for all aspects of the agricultural sector. The aim was to achieve greater coordination and flexibility through decentralization to regional authorities. During the period of uncertainty and change, per- formance of management was understandably weak; when disbursements were suspended in June 1969, the project came to a virtual halt because management autonomy meant dependence on active Bank support. In 1971, the new Agrarian Reform Law vested responsibility for the project in the regional authorities in Piura. San Lorenzo became part of an irrigation district under Agrarian Zone I except for the drainage component managed as a separate project. Much criticism of project management since 1971 appears to have been un- warranted because it assumed that the San Lorenzo project was still being implemented. 5.06 Since 1971, as with most completed work programs, San Lorenzo has suffered from relatively low priority, inadequate budget allocation, under- staffing and general lack of special interest. Most of these issues can only be corrected at a sector or national level and are not project specific. The present faults and constraints are thought to be typical of the agricultural sector in Peru and the difficulty that Government faces in bringing about rapid change particularly with issues like water charges which are politically sensitive. VI. ECONOMIC AND SOCIAL IMPACT 6.01 Project performance in terms of economic, social and secondary benefits are described in detail in SDV Annex 10. Four factors require special mention: (i) Loan 418-PE did not finance a self-contained development stage of San Lorenzo Project; (ii) divergence in technical, social and political parameters assumed at appraisal was very large.; - A25 - (iii) the project should have been re-appraised at renegotiation of the Loan to give realistic data against which performance could be evaluated; and (iv) insufficient data exist to evaluate the drainage as a separate component, but benefits are likely to be much less than antici- pated in the drainage studies. Economic Analysis 6.02 The appraisal report cites internal rates of return for two evalua- tions: about 14% for the overall project since investments in main works commenced in 1955 (partly financed by Loan 114-PE); and about 16% for invest- ments from 1965 related to the original life span of Loan 418-PE. No figures were given for costs and benefits for the period 1955-64, but the analysis is thought to be based on the annual increase in net value of production in the San Lorenzo and Lower Piura areas. Information on costs and benefits for elements partly financed by Loan-418-PE is incomplete, including only a cost stream, net value of production for 1964 and 1972, and gross production value in 1977; the build-up of benefits over the years is not given. Important basic assumptions cannot be supported for either of the economic analyses in the appraisal report for reasons summarized below and detailed in SDV Annex 10. 6.03 An error at appraisal was the assumption that "without project" benefits would be static at the 1964 level, even though at that time the project was going through its most dynamic stage of development and some AID credit was still undisbursed enabling development to continue. It could be argued that withdrawal of Bank support would have had a negative effect, given the strong political opposition to San Lorenzo. Nonetheless, an economic evaluation of -the period partly financed by Loan 418-PE has no useful value, particularly as Bank commitment to the development stage began in 1961. 6.04 To overcome these difficulties, the ex-post project analysis has been calculated since the beginning of Stage II development in 1955 and for the colonization phase from 1961 until the closure date of Loan 418-PE in June 30, 1976. The project,costs include those financed by Government and on-farm development costs financed by credit or from the farmers' own resources (see Table 4). Some approximations have had to be made because certain San Lorenzo costs cannot be apportioned directly. On-farm costs were estimated from sample developments as shown in Table 19; housing costs reflect what was considered reasonable to generate agricultural benefits. Benefits were based on recorded production volumes except for fruit for which records were not available; production streams are given in Table 20 for San Lorenzo. Prices for most .crops given in,Table 21 are historical farmgate prices reported by the project. No allowance was made for production increases which might occur through increased water availability due to the Chira-Piura development. An evaluation period of 45 years was used (up to the year 2000) and no project salvage cost was assumed. - A26 - 6.05 All costs and benefits were converted into 1975 constant prices by use of the deflators given in Table 22; as far as possible, taxes and sub- sidies were eliminated. Production input costs were estimated from various sources in 1975 terms and assumed constant for lack of better data. For technical reasons, labor accounting price has only been used for evaluating the labor component of production costs and not in project costs. All farm labor was evaluated at its marginal opportunity cost on a seasonal basis allowing for Government fixed minimum wages and social allowances as an income distribution factor. The data base for estimating benefit and cost flows presents considerable difficulties of reliability and consistency and is known to be incomplete. In particular, it is not possible to make extensive use of social pricing of inputs and further studies on these topics would be useful for future work in Peru. 6.06 With the above assumptions, the ex-post rate of return for Stage III development since 1961 is estimated at 17.5% (not calculated at appraisal) from cost and benefits streams given in Table 23. This relatively high value reflects the following favorable features: (i) the colonization phase used previously constructed works; (ii) cost investments were divisible by nature,and thus there was little delay between investments and benefit; and (iii) intensive technical assistance was given to the farmers in the important early period to bring farms quickly into production. 6.07 The internal rate of return for development since 1955 (Loan 114-PE through Loan 418-PE) is estimated at about 8.7% (compared to 14% projected at appraisal) from cost and benefits streams given in Table 24. This is probably a low estimate due to conservative benefits attributed to Lower Piura. 1/ It also reflects the underutilization of San Lorenzo works which were constructed for a larger area than finally irrigated. The Tablazo canal command has been mainly abandoned and was a major cost in Stage II development (see para. 8.05 for possible future use). Nonetheless, given the project difficulties, the ex- post IRR of at least 8.7% is an unexpected achievement. Farm Income 6.08 The pre-project income of settlers is not recorded but has been estimated at about S/. 15,000 (in 1975 terms) for the landless peasants and goatherders. Gross farm incomes for 1975 are detailed in Table 25 and summarized below: 1/ For reasons explained in SDV Annex 10, the actual IRR since 1955 might be close to 10%. - A27 - Average Gross Family Income (in 1975 terms) S/. US$ Equivalent Cooperative member 71,000 1,640 Small Colono (8 ha) 107,000 2,470 Early settlers (17 ha) 183,000 4,220 Alternative income opportunity in agriculture 36,000 830 These estimates are based on data collected by the mission; taxes and water charges have not been deducted. Although only indicative they do show that early settlers' incomes have increased up to twelve times as a result of the project and that income for all beneficiaries is substantially above alterna- tive income opportunities in agriculture. It is not surprising that Government now considers early settlers in San Lorenzo to be privileged compared to other areas of Peru; their high income is due to division of land in relatively large farm units following social objectives that existed in 1961 and are now superseded by current agrarian reform standards. Social Impact 6.09 The colonization phase provided important community services, mainly in health and education for adults and children. Although not financed by Loan 418-PE, these services were encouraged and supported by the Bank and have had a wide-ranging impact throughout the project area and region. Similarly, the comprehensive on-farm development package enabled poor settlers who were mainly landless peasants to achieve a living standard and security which greatly exceeds normal achievement for this type of project elsewhere. Of special merit was the successful and courageous Government program to settle squatters in the project area. Training and Demonstration 6.10 Perhaps the most important impact of San Lorenzo is the far-reaching influence throughout the agricultural sector in Peru. It has had a major training and demonstration impact because of its innovative, ambitious objectives and importance at the time of formulation. Many technicians, that were later to assume important positions in the government and the private sector, got their in-service training in San Lorenzo. The project, which was the first experience in inte- grated, large-scale agricultural development in Peru, served as a development catalyst for the whole northwest region, both directly by providing water and indirectly through example particularly with respect to sensitive social objectives setting high rural :standards of living. Though difficult to quantify, the settlement and planning concepts (considered radical and reformist in 1959) are known to have made an important contribution to agrarian reform and develop- ment planning in Peru. Employment 6.11 Direct beneficiaries are estimated at 3,180 families (see Table 10) of which at least 2,400 were poor landless peasants; this compares with an - A28 - appraisal objective of settling about 1,570 family units, which was reduced at loan renegotiation to 950 units. Direct employment generated by the project averages out at about 5,000 full-time jobs with some 1,500 jobs provided in supporting services. Employment performance greatly exceeds expectations at appraisal and has largely benefited low income target groups; in all, it is estimated that 60,000 people in or near the area derive their livelihood as a result of the San Lorenzo Project. VII. COST RECOVERY 7.01 Historically the water charges under the San Lorenzo Project have been extremely low, averaging about S/. 0.006 per M3 or S/. 90.0 per ha. Little attention was paid to'cost recovery until 1969 for the following reasons: (i) the project was besieged with more urgent problems; (ii) unreliable and inadequate irrigation and other technical problems meant that settlers could not afford to pay realistic charges or had a ready excuse to delay or avoid payment; and (iii) Management did not have firm Government backing for effective action or a plan of action which would be fair to all settlers. 7.02 The Government acted in 1969 by promulgating the Peruvian Water Law, which laid down the principles for charging for water and proposed a two-part water fee: a water tariff collected through the Ministry of Agriculture based on volume used, and an additional fee or quota collected and dispursed through a water users' association. Water tariffs were to reflect the users' capacity to pay based on projected benefits per unit area and not farm holdings. The Sub-direccion de Tarifas y Cuotas of the Direccion General de Aguas has formulated and is presently implementing plans to enforce the Water Law. For each project they are determining charges for the right to use water, a service charge to cover effective O&M and amortization to recover all or part of Government investments. The procedure works for valleys with long established irrigation, but the constraint is likely to be the ability to pay in areas subject to agrarian reform with costly infrastructure or new irrigation schemes. 7.03 The water charges in San Lorenzo were raised from S/. 0.006 per M3 to S/. 0.02 per M3 retroactively in 1971. The Lower and Middle Piura Valleys were kept at the old rate, reflecting poorer farming opportunities and capacity of cooperative members to pay. The plan has been to gradually increase charges - A29 - to cover full running costs over five years. To improve recovery, Banco de Fomento Agropecuario (BFA) is acting as collecting agency in the project area. Actual collection is lower in San Lorenzo than in Zona Agraria I, and both are lower than Peru as a whole. In 1973, only 16% of issued receipts were recovered, compared with 19% for Zona Agraria I and 31% for the country. More important, charges collected in 1973 for San Lorenzo only represent about 5% of the O&M charges which are thought to be low for preventive maintenance. 7.04 Presently, the water charge for San Lorenzo has been raised to S/. 0.04 per M3 (about S/. 600 per ha) and is probably the highest in Peru. If collected, this should cover about 65% of the running cost excluding drainage. Major problems are: inadequate data to establish capacity of users to pay a realistic O&M cost; lack of facilities and staff to carry out an effective program; and the present procedure for allocating water. Water distribution is biased in favor of the new settlers (squatters) who have the least capacity to pay compared with earlier settlers; this is believed to encourage social resistance to charge recovery. A future problem is that drainage works might benefit less than 10% of the irrigated area and many farmers might hesitate to pay for O&M costs which do not benefit them. 7.05 It is concluded that considerable progress is taking place to deal with a national issue which will become increasingly important as major on- going developments are commissioned. Government now has legislation to take positive action and appears to be making progress to implement a rational strategy. Clearly privileged beneficiaries such as San Lorenzo settlers could and should be made to pay a reasonable portion of costs for their special facilities. VIII. FUTURE IMPROVEMENTS 8.01 The best scope for future improvements is to expand agricultural production through increased yields, better cultural practices and improved seeds now that irrigation is reliable. Attention should be given to agricultural supporting services and extension; water management should be examined because the present tendency is to lower standards whenever supply is adequate; water saved through improved efficiency might have a marginal value that justifies investment in minor works. Crop yields have been static since Loan 418-PE became effective; constraints should be identified and a plan formulated to improve production. 8.02 The drainage component in San Lorenzo should be evaluated to decide whether further development can be justified. Definite proposals should be made for preventive maintenance of drainage works which should be monitored to establish benefits that can be attributed to the program. 8.04 The San Lorenzo system and the Chira-Piura Project have interrelated water sources, and considerable water savings and other benefits might accrue - A30 - if the operation of tiie two systems was integrated under a single authority. This would let experience gained on drainage and other elements of San Lorenzo be made available to staff responsible for other regional developments. 8.05 The San Lorenzo Project has exceeded settlement goals set at renegotia- tion of the Loan, but these were related to an obsolete assumption of water availability. Now that adequate water is secured through completion of the Chira-Piura Diversion Project, the possibility of either expansion of the San Lorenzo area or increased crop intensity can again be evaluated. In particular, special water management and farming techniques might allow 8,000 ha of the Tablazo area abandoned because of soil type and high water requirements to be rehabilitated. Such a renovation of existing works might be viable given the long lead time to implement further phases of the Chira-Piura Project necessary to utilize the additional water supplies now available. The technical problems with extension areas settled outside the original project boundaries should be investigated and corrective action taken. 8.06 An essential related improvement is the development phases of the recently completed Chira-Piura Diversion Project. The first phase of this program will be rehabilitation and improvement of the Lower Piura Valley currently being examined by an FAO/CP preparation mission (February 1977). The development program covering some 142,000 ha should be accelerated as a matter of urgency to reduce the lead time to efficiently utilize the sub- stantial sunk investment in the main storage dam and ancillary works. IX. SPECIAL ISSUES Water Availability and Use 9.01 Without doubt, incorrect estimation of water availability and over- commitment of a variable supply was directly responsible for many political difficulties and technical shortcomings of the project; the error was exacer- bated by a sequence of drought years early in the life of Loan 418-PE. Government commitment to guarantee supplies for 32,000 ha in San Lorenzo made agriculture speculative and resulted in social and financial hardship for many settlers. Analyses of water availability and use were given scant attention at preparation and appraisal, and technical analysis through the project life was poor. The marginal value of water was not evaluated to establish risk and possible range of corrective action at loan renegotiation; in particular, the hydrological review in 1967 did not adequately address the full range of possibilities or fully recognize the poor factual data base; most important, it did not adequately evaluate unrealiability of irrigation in terms of agricultural development and settlement objectives. Table 26 summarizes water availability and distribution since 1952 and illustrates the randomness of supply; monthly analysis and use of storage shows that water management could have been more efficient if data had been better evaluated, as detailed in SDV Annex 6. Inexpensive opportunities to supplement resources within the San Lorenzo system were never seriously examined although they might have given important short-term benefits. - A31 - Soils 9.02 About 9,000 ha, or 20% of the land provided with main irrigation infrastructure was abandoned for reasons related to soil quality; the technical and economic analysis supporting these important decisions cannot be traced; similarly, the justification for expansion outside the original project area was inadequately investigated. The appraisal mission should have taken note of the special limitations of the Tablazo soils as described in a 1961 Bank reconnaissance survey; this concluded that about 60% more water would be required compared with other San Lorenzo lands and identified the need for special cultivable and water management techniques. None of these issues were examined in detail at.appraisal of Loan 418-PE. Underutilization of works has been the main reason for the low IRR for development since 1955. Clearly the land potential should have been evaluated before substantial investments were made in infrastructure through Loan 114-PE, especially as since 1969 Government has developed with some success an extra 9,000 ha thought to be unsuitable for irrigated agriculture outside the geographical boundary of San Lorenzo as defined at appraisal.' Much more investigative work on soils should have been done both during preparation and when the technical problems were recognized.  Table 1. PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Cost Estimate at Appraisal (US$ million) Cost Allocation of Re-allocation of Category Item Estimate Funds (Appraisal Funds before date Report) of Effectiveness 2 A. O&M Land Development and Road Equipment 7.85 1.25 1.20 B. Drainage and Ancillary Works 1.25 1.25 0.90 C. Technical Services, Fa- cilities and Equipment 3.70 1.07 1.00 D. Basin Stud-1/ 0.60 0.60 0.60 E. On-farm Investments (Medium Long-term Agri- cultural Credit) 12.60 6.78 6.80 F. Unallocated - - 0.50 Sub-total: 26.00 10.953/ 11.00 G. Administration 4.40 Total: 30.404/ 1/ The cost of the Basin Study are not considered part of San Lorenzo project although financed thrcugh Loan 418-PE. 2/ The change took place between April and November 1965. 3/ Taken from the Appraisal Report (Annex 4); the US$0.05 million difference from the US$11.0 million loan amount occurred after appraisal. 4/ No contingency items were incladed in the cost estimate. Table 2. PERU LOAN 418-PE - SAN LORENZO IRRIGATION PROJECT Sources of Project-Funds-at ADnraiaal; Local and Foreign Exchange Components (all sums in millions) Part A Source of Project Funds /1 Category to be Financed Sources A and*B C E G Total Soles US$ Eg. Percentage Farmers/2 - - 66.8 115.4 182.2 6.7 22.8 Government 178.1 71.8 0.7 4.6 255.2 9.4 31.6 A.I.D. - - 89.0 - 89.0 3.3 11.1 I.B.R.D. 67.2 28.4 182.9 - 278.5 10.4 34.5 Total: 245.3 100.2 339.4 120.0 804.9 29.8 100.0 L See Table 1 for title of each category; Basin Program not included. .L Includes farmers initial share of on-farm investment costs, repayments on land and investment loans, and water charges. Part B Local and Foreign Exchange Component of Loan 418-PEL3 LC FX Program Soles US$ Eqv. Soles % Soles % A & B 67.2 2.5 25.4 37.8 41.8 62.2 C 28.4 1.1 - - 28.4 100.0 E 182.9 6.8 104.4 57.1 78.5 42.9 G - Total: 278.5 10.4 129.8 46.6 148.7 53.4 /3 Errors in the Appraisal Report have been corrected. Table 3. PERU LOAN 418-PE - SAN LORENZO IRRIGATION PRCLRT Revised List of Goods at Renegotiation (March 1968) (in '000 US$ equivalent) Additional renegotiation Original Disbursed as of Amount Revised Disbursementsl Amount Nov.17, 1967 Required Total at Close Date (1) (2) (3) 46)-(2)+(3) June 30. 1976 Category A: Operations and Maintenance. 1.200 106 850 956 I Land Development and Roads Equipeant A-1 Heavy Equipment - - 500 500 A-2 Vehicles - 3 150 153 A-3 Spare Parts and Repair Equipment - 103 200 303 Category BT Drainage Works 900 - 377 377 286 B-I Capital Equipment ISO 150 ;-2 Construction and Materials 227 227 Category C: Technical Services, Facilities 1.000 548 470 1.018 876 and Equipment C-I Salaries, Allowances, Contract Services and Training 115 - 115 C-2 Vehicles 245 - 245 C-3 Education and Extension Facilities 71 - 71 and Equipment C-4 Research Facilities and Equipment 16 - 16 C-5 Spare Parts and Repair Equipment 101 50 151 C-6 Consultants in Extension. Research - 80 80 and Marketing C-7 Continuation of Ongoing Drainage - 142 142 Study (6,000 has.) C-8 Extension of Study to Cover - 40 40 Additional 2,000 has. C-9 Design of Drainage Works for - 68 68 Somate District C-10 Design of Drainage Works for - 90 90 2,000 has. in other Districts Catezory D: Basin Program 600 128 820 948 1,087 D-1 Basin Study (including Piura Valley 128 528 656 Drainage and Rehabilitation Study) D-2 Feasibility Study Chira-Fiura - 217 217 Diversion D-3 Hydro-meteorological Equipment - 75 75 Category E: Agricultural Credit 1.400 2.926 E-1 Land Development 835 E-2 Housing Development 267 E-3 Agricultural Machinery 180 E-4 Livestock and Services 43 E-5 Tree Crops Category F: Unallocated 500 - 3012/ 301 c7sp Total 11 12 4.318 6.500 6.43 1/ See Table 6 for details. US$144,000) 7/ Including contingency for additional drainage works on 2,000 ha; (Soles expenditure equivalent to u 3/ US$78,000 undisbursed as of June 30, 1976. PERU WAN 41f-PE SAN LUKENZO IRRIGATION FROJkET Project Cots (thousand of Soles) 31 Actual Development Stage Year Appraisal Cost Estimate Actual Government Costs Actual On-farm Development Costs- Total Excluding Drainage On-farm Housing Total Cost Drainage program Actual Cost System 1965Y CostY cost (1975 prices (1975 terms) --------(Current Prices) --------------------- Soles) ------- (1975 terms) -------------------- Stiqu II; San Lorenzo 1955 126,000 849,240 126,000 126,000 849,240 849,240 Dam and physical 1956 126,000 793,800 126,000 126,000 793,800 793,800 infrastructure 1957 126,000 732,060 126.000 126,000 732,060 732,060 (Loan 114-Ph) 1958 76,000 408,800 76,000 76,000 408,800 408,800 1959 50,000 238,000 50,000 50,000 238,000 238,000 1960 2 0820 25,500 25,500 0108,12 Sub-total 3,130,020 3.130,020 3,130,020 Stage III; the 1961 34,738 142,773 34,738 34,738 142,773 5,953 1,170 7,123 149,896 Colonization Phase. 1962 36,732 144,724 36,732 36,732 144,724 27,477 5,400 32,877 177,601 1963 57,534 193,768 57,534 57,534 193,768 84,568 16,620 101,188 294,956 1964 74,256 248,015 74,256 74,256 248,015 84,720 16,650 101,370 349,385 1965 321,900 959,262 108,607 108,607 323,649 57,563 14,220 71,783 395,432. 1966 282,700 842,446 81,250 81,250 216,937 60,987 15,066 76,053 292,990 1967 200,300 596,894 69,872 69,872 169,289 34,610 8,550 43,160 212,549 1968 23,200 69,136 45,423 45,423 93,571 6,703 1,650 8,353 101,924 1969 24,400 72,712 27,825 27,825 52,867 23,219 5,736 28,955 81,822 1970 21,700 64,666 55,040 55,040 99,072 31,319 31,263 62,582 161,654 1971 17,800 53,044 16,865 16,865 29,008 62,562 62,449 125,011 154,019 1972 14,500 43,210 16,865 16,865 27,321 4,445 4,437 8,882 36,203 1973 12,900 38,442 15,225 12,756 27,981 40,292 549 548 1,097 41,389 1974 12,500 37,250 15,225 33,974 49,199 61,007 5,850 5,839 11,689 72,696 .1975 12,500 3 14,345 27,006 41,351 41,351 1,341 1,339 2,680 44,031 Subtoa 1,147,660 3,M9~ 1,8364 41,6 19U73 682,803 2,566,547P' Total Development Cost 6,673,612 5,696,567 Post development Phase annual running cost 1976 on (est.) 12,500 37,250 10,604 25,000 25,000 25,000 1/ Actual costs from 1961 to 1964 have been usedl projected cash flow 1965 to 1976 is taken from Annex 5, Appraisal Report. / Coefficients to convert to 1975 terms are given on Table 22 and converted to US$1 = 43.38 soles. 3/ See Table 19. Corresponds to investments made by SCLPA during 1959/60 and assumed as if all took place in 1960. 5/Start of Loan 418-P'E. Development cost of Stage III-;US$59.2 million equivalent (in 1975 terms)compared to USA81.7 million equivalent (in 1975 terms) estimated at appraisal. 7/ Tal devpm eince 19555s131.3 million equivalent(in 1975 terms) compared to US$153.8 million equivalent (in 1975 terms) estimated at appraisal. 8/ Development cost to complete Phase 1 of drainage project. PERU LOAN 418-PE - SAN LORENZO IRRIGATION PROJECT (US$ million) Costs -Drainage Component- 1/ Item At Appraisal November Renegotiation June Revised estimate Actual Cost - March 1965 1975 March 1968 1969 by CENDRET 1973-76 (Dec. 1970) Estimate of Cost of Works NK NK 1,583 2 1,583 1,3533V 1,944 Category B Loanfi Heavy equipment (2( 150 6616 4/ Materials (1,250 900 227 566 n,a, IECO Studies not given 253 340 340 340 CENDRET (1969-73) NA NA est. 3105/ Pilot Areas,§ NA NK 123 1/ See Tablel7 for details of cost of special drainage project. 2/ Estimated cost from IECO drainage report dated March 1968 to benefit 8,020 ha. 3/ Revised cost estimate by CENDRET to benefit 3,080 ha. 4/ Categories A (O&M equipment)and B (drainage equipment) were combined in February 1974, making US$150,000 available for rainage equipment, for details see Table 18 (note: Table 6 shows only US$0.236 dispersed to category B). 5/ Based on administration and running costs for 1973-76; not allowing cost of Dutch technical assistance. 6/ Included in estimate of cost of works. *r PERU LOAN 418-PE SAN LORENZO IRRICATION PROJECT Disbursements by Years (US$ '000) Category A B C D E Operation & Maintenance Drainagel Technical Assistance2 Basin Agricultural Description Equipmtnt Works and Facilities Program Credit Unallocated3/ Totale Original amount allocated 1,200 900 1,000 600 6,800 500 1',000 Amount allocated at renegotiation (1968) 956 377 1,018 948 Z,V00 301 6,500 Amount dispersed by years 1965 - 1 - 1 1966 83 - 430 8 399 920 1967 45 - 193 219 1,002 1,450 1968 1 - 245 633 428 1,397 1969 - - 8 226 836 1,070 1970 - - - - 1971 - 4/ - 1972 45- 45 1973 541 271 - - 812 1974 344 15 - - 359 1975 386 - - - 386 1976 63 - .- - 78 41 Total Actual Disbursements 1,508 286 876 1,087 2,665 785/ 6,5006/ 1/ After 1974, categories A and B were dispersed under category A only. 2/ The Drainage Study accounted for US$0.34 million. 3/ Unallocated funds were used for Category A and D 4/ Dispersed 12/12/72. 5/ Undispersed at closure, June 30, 1976. 6/ For breakdown of dispersement by items see Table 6. Table 8 PERU LOAN 418-PE - SAN LORENZO IRRIGATION PROJECT gata of On-farm Development /1 (net ha per annum) Total Nt Net Area Increase in Net Irrigable Year.L Area veveloped Abandoned DeveloDed Area Area 1961 390 - 390 390 1962 1,800 - 1,800 2,190 1963 5,540 - 5,540 7,740 1964 5,550 - 5,550 13,280 1965 4,740 - 4,740 18,020 1966 5,022 - 5,022 23,042 1967 2,850 - 519 2,331 25,373 1968 552 - 694 - 142 25,231 1969 1,912 - 3,161 - 1,249 23,982 1970 2,452 - 2,078 374 24,356 1971 4,898 - 628 4,270 28,626 1972 348 - 348 28,974 1973 43 - 74 - 31 28,943 1974 458 - 1,917 - 1,459 27,484 1975 106 - 106 27.590 Total 36,6614/ - 9,071 27,590 /1 Excluding Valle de los Incas and Valle Hermoso. /2 On-farm development for 1961-1964 estimated on basis of partial information; data for 1965-1974 based on the "Padron de Usuarios" for each year; some area shown for any year may have been developed or abandoned.in the previous year because of delays in updating of the "Padron"; the pace of development shown on this table should be regarded as indicative. /3 Cumulative net increase. / 2,300 ha was outside the original San Lorenzo project boundary. Table 9. PERU LOAN 418-PE - SAN LORENZO IRRIGATION PROJECT Cumulative Area Sold (ha) Area Scld Year Individual Collective Total No. Farms/ Number Area No. Families Area Number Area 1961 103 706 - - 103 706 1962 450 7,385 - - 450 7,385 1963 754 13,840 - - 754 13,840 1964 951 16,688 - - 951 16,688 1965 1,130 19,862 - - 1,130 19,862 1966/a 1,302 23,248 - - 1,302 23,248 1967 1,287 22,321 - - 1,287 22,321 196&/a 1,284 21,763 - - 1,284 21,763 1969. 1,273 20,716 10 / 1,096 4,402 2,369 25,118 1970 1,288 21,632 30 / 1,687 6,785 2,975 28,417 1971 1,294 21,635 42 / 1,795 7,238 3,089 28,873 1972 1,293 21,599 42 / 1,795 7,238 3,088 28,837 1973 1,291 21,087 42 / 1,795 7,238 3,086 28,325 1974 1,279 20,165 47 / 1,847 7,436 3,126 27,601 1975 1,292 19,935 56 / 1,888 7,654 3,180 27,589 .a Up to November of each year. /b Up to March 1970. PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Size Distribution of Individual Holdings (June 30, 1976) Net Size Number of Area (Ha.) Settlers % Accumulative % (Has.) % Accumulative % Up to 8.0 486 38 38 3,190 17 17 8.1 - 13.0 297 23 61 3,257 17 34 13.1 - 20.0 294 23 84 4,736 25 59 20.1 - 40.0 123 10 94 3,347 17 76 40.1 - 60.0 56 4 98 2,801 15 91 60.1 - 80.0 24 2 100 %1,683 9 100 TOTAL 1,280 100 - 19,014 100 - 0-3 Note: Plots owned by schools and former proprietors are not included. ( 0* PERU LOAN 418-PE - SAN EIRNZO INKIGATION PROJECT Production and Cropping Patlern at Appraisal (1964 and projeLted 1972) Area Yield Hillion Soles Appraisal Hectares kg/ha Price Net Water Thousands 'Ib.sands Soles/kg Cross Velue costs Net Value value Requirements 1964 1972 1964 1972 1964 1972 1964 1972 1964 1972 1964 .1972 S/per ha K /a Cotton (A) 6.78 15.00 2.8 2.5 6.0 5 00 114.0 188.0 47.5 102.0 66.5 86.0 5,730 10,400 Beans/Corn Beans/Sweet Potatoes 1.49 7.90 - - - - 6.8 69.0 4.2 41.0 2.6 28.0 3,541 7-8 000 Beans/Croundnuts (A) Yuca (A) .40 1.50 16.0 20.0 1.0 0.95 6.4 29.0 1.6 6.0 4.8 823.0 5.330 18.920 Vegetables .16 1.90 18.0 25.0 1.2 1.00 3.5 49.0 1.2 22.0 2.3 27.0 14.210 16,220 (2 crops) (A) Pasture (P) .26 6.20 3.8 5.5 2 0 2.00 2.0 68.8 1.0 36.0 1.0 32.0 5,161 22.060 (liters milk (per ItLer) equivalent) Fruits (P) .80 5.00 - 8.0 - 1.05 - 42.DL 2.7 24.0 (2.7) 18.0 13,60OL2 12.900 Other- 1.18 - - - - - 12.4 - 6.7 - 5.7 - - Total. 11.07 37.50 145.1 445.OL 64.9 231.0 80.2 214.0 Cross annual value of fruit production should progressively increase up to Siles 158 million by 1977, making a total gross value of production of Soles 561 million a year. Net value per ha in 1977 for fruit. L Rice peanuts, etc. (P) = Perennials; (A) = Annuals /4 The area of 37,500 ha contradicts the target develtpment area of 32,000 h in o,ther parts of the appraial ieport. /5 Data source: Annex 7, Appraisal Report; no indication was given of the rate of dtvelopimnnt or change in production over the period 1964 to 1975. r r.ku LOAN Ul8-PE SAN IDORENZo IRRIGATION PROJECT CJLTIVATED AREA BY CROPS, 1962-1915 (ha) ANNUAL CR 01.5 F R i IT S Year Cotton Rice Maize Soraham Beans Yuca Onions. Others. Total Mango Limon i)range Avocado Others rotal Pastures Total Cul- Crops Fruits tivated Area 1962 1,000 - 500 - 100 - - - 1,bO0 - 3 3 .6 10 22 20 1,642 1963 2,500 300 1,000 - 200 150 30 50 4,230 14 14 a 14 20 70 20 4,320 1964 6.782 465 987 - 351 402 161 h8d 10,036 112 23 80 122 130 467 236 111,739 1965 11,641 - 1,240 - 411 872 254 259 14,672 152 95 249 218 123 837 287 15.796 1966 9,744 620 3,355 - 359 600 424 613 15,715 277 172 272 265 286 1,272 301 i7.268 1967 10,605 686 2,161 - 1,724 431 43) 211 16,255 669 234 381 361 301 1,946 446 18,647 1968 6,362 1,112 1,729 - 451 211 280 186 10,331 1,119 266 399 462 251 2,397 262 12.990(3) 1969 10,195 1,470 1,383 - 840 120 234 115 14,957 1,649 531 488 575 239 3,482 371 18h10 1970 9,530 2 625 2,396 1,110 419 84 165 810 1,141 1,754 574 502 bi1 245 3.676 505 21,322 1971 9,tb8( 4,634 1,837 470 458 194 218 840 18,519 1,913 688 527 646 233 4,007 329 22,855 1972 7,910(2) 2,441 3,236 738 296 154 246 944 15,955 7,228 831 589 699 201 4,548 341 20,844 1973 10.076 4,196 4,227 106 622 292 236 76 19,796 2,81t 834 659 bl2 180 5,171 500 25,461 1974 14,969 1,497 1,105 35 121 100 II '15 1,o39 3,069 827 576 566 154 5,192 188 23,419 1975 8,939 2,859 3,489 - 228 96 114 67 15,792 3,348 972 632 612 155 5,719 289 21,800 (1) Sane years It might be Included in "Otherb". (2) Area sown was alout 14,200 Ia buL 6,300 Wt!C lobt because of he.vy rains. (3) Drought Year. .jable 13. PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Yields of Main Crops 4Tans/ha) Year Cotton Rice Maize Beans 1962 2.0 - 3.0 1.0 1963 2.0 5.0 3.0 1.0 1964 2.68 5.5 3.0 1.0 1965 1.441/ - 4.0 1.2 1966 2.07 5.4 3.0 1.0 1967 2.01 4.83 3.5 1.2 1968 2.0 5.52 4.0 1.2 1969 1.34 5.52 3.0 .8 1970 2.0 5.52 4.0 1.2 1971 1.65 4.9 3.0 .85 1972 1.101/ 2.8 .90 1973 1.34 4.83 3.0 1.0 1974 1.38 4.12 2.74 1.0 1975 1.38 4.83 2.76 1.0 Averages 1962-1970 2.012/ 5.33 3.39 1.07 1971-1975 1.443/ 4.65 2.86 0.95 1/ Sharp drops explained by high precipitation. 2/ Excluding 1965. 3/ Excluding 1972. Note: no data available for tree crops. Source: Memoria 1948-1976-; prepared by Zona Agraria I, Piura, 1976. PERI LOAN 418-PE - *SAN LOREN7.0 IRRIGATION PROJECT Total Gross Medium and Lonr Term On-lending, per Kind of Investmentl/ (millions of Soles of 1975) Before During effectiveness After "closing" of As proposed by effectiveness of the credit component the credit component During the Percent composition the Appraisal Kind of Investments of Loan 418 of Loan 418 of Loan 418 whole period (without "Rehabilitation") Report 1962 to 1965 1966 to 1969 1970 to 1975 1962 to 1975 62/65 66/69 70/75 62/75 1965 to 1967 M. S. 7 m S _ m. S/. % m.S/. 7. 7 % % % % m. S/. % Land Development 114.3 39 125.1 38 285.9 66 525.3 49 39 47 66 53 492.9 48 Housing Development 88.7 30 48.9 15 - - 137.6 13 30 18 - 14 123.3 12 Agricultural Machinery 81.1 27 19.7 6 38.2 9 139.0 13 27 7 9 14 231.9 23 Livestock 3.1 1 23.5 7 23.4 5 50.0 5 1 9 5 5 57.0 6 Tree Crops 9.3 3 49.9 15 86.0 20 145.2 14 3 19 20 14 113.1 11 Rehabilitation - - 63.1 19 0.2 - 63.3 6 Total 296.5 100 330.2 100 433.7 100 1,060.4 100 100 100 100 100 1,018.2 100 Per year 74.1 82.6 72.3 75.7 339.4 1/ See Annex 5, SDV for supporting details. PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT San Lorenzo: Gross Medium- and Long-term On-lending - Sources of Financing (million Soles of 1975) Gross Medium- Loan 418 as Percentage of: and Long-term National Foreign Sources Medium/Long-term Foreign-financed Medium/ Year On-lending Sources-a US AID Loan 418 IDB. IBRD/b On-lending Long-term On-lending 1962/c 5.4 5.4 - - 19637c 49.1 24.8 24.3 - 1964/c 164.3 21.3 143.0 - - 19657c 77.6 - 12.5 90.1 - - - - 1966/c 151.0 20.1 39.3 91.6 - 60.7% 70.0% 1967 95.5 28.9 6.6 60.0 - 62.8 90.1% 1968/c 40.6 15.9 1.0 23.7 - 58.2% 96.0% 1969/c 43.2 12.6 - 30.6 - 70.EM 100.0% 1970/c 63.2 7.6 55.6 - - - - 1971 62.7 3.5 59.2 - - - - 1972 88.8 63.8 25.0 - - - - 1973 66.1 37.0 29.1 - - - - 1974 144.7 124.7 19.0 - 1.0 - - 1975 48.8 1.6 7.9 - 39.3 - - 1976/d 6.0 0.8 3.4 - 1.8 - - "Totals"l/ 1962-65 296.4 39.0 257.4 - - - - 1966-69 330.3 77.5 46.9 205.9 - 62.3% 81.4% 1970-75 474.3 238.2 195.8 - 40.3 - - 1962-75 1,101.0 354.7 500.1 205.9 40.3 18.7 27.6X /a It includes Banco de Fomento Agropecuario (BFA), Instituto de Reforma Agraria y Colonizacion (IRAC), Rehabilitation Credit as established by DFL 19396, and other Governmental Funds (Supremo Gobierno). /b Loans 933-PE (IBRD) and 322 (IDB). /e See Annex 5, SDV for supporting details. 9, /d Up to June 30, 1976. /e This is not an actual "Total On-lending" because it refers to gross on-lending instead of net on-lending. Table 19. PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Average Unit Cost of On-farm Development (Soles per ha) Period On Farm Systems 1/ Housing Total (Current (Soles of (Soles of (Soles of Prices) 1975) 1975 1975) 2/ 4/ 8/ 1961-1964 .4,148- 15,265- 3,000- 18,265 1965-1969 4,509 12,1445/ 3,000r 15,144 1970-1975 7,0966/ 12,773 12,7509 25,523 1/ Includes: land clearing, fencing, levelling, roads, irrigation ditches, terminal drains, etc. 2/ Based on a sample of 430 ha of land developed in 1962-64. 3/ Based on a sample of 363 ha of land developed in 1965-67. 4/ Calculated according to average deflation coefficients for 1962, 1963 and 1964, (see Table22). 5/ Idem for 1965, 1966 and 1167, (see Table22). 6/ Based on the 6,100 ha developed in 1970-71. 7/ An estimate of minimum housing required to permit production - S/ 51,000 (1975 terms) per-house rather than actual cost (see Annex 10, SDV). 8/ Based on a ratio of one house per 17 ha. 9/ Based on a ratio of one house per 4 ha. Table 23. PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT The Colonization Phase - Benefit and Cost Streams (thousand of Soles, 1975 terms) Year Gross Production Costs Net Project Net Production Inputs Labor Total Production Value Value 1961 149,896 -149,896 1962 63,272 16,420 10,474 26,894 36,378 177,601 -141,223 1963 166,550 43,060 29,133 72,193 94,357 294,956 -200,599 1964 571,768 105,816 82,816 188,632 383,136 349,385 33,751 1965 404,911 181,265 104,588 285,853 119,058 395,432 -276,374 1966 525,227 163,471 114,047 227,518 247,709 292,990 -45,281 1967 576,905 175,475 121,408 296,883 280,022 212,549 67,473 1968 491,972 118,912 88,863 207,775 284,197 101,924 182,273 1969 463,532 174,872 113,760 288,632 174,900 81,822 93,078 1970 710,923 198,001 137,457 335,458 375,465 161,654 213,811 1971 684,140 215,399 151,949 367,348 343,575 154,019 189,556 1972 491,650 211,750 113,597 325,347 166,303 36,203 130,100 1973 765,949 233,668 153,544 387,212 378,737 41,389 337,348 1974 799,215 221,777 153,235 375,012 424,203 72,696 351,507 1975 708,422 195,214 129,714 324,925 383,497 44,031 339,466 1976 726,380 218 680 148,300 366,980 359,400 25,000 334,400 1977 735,980 218, 80 149,230 367,910 368,070 25,000 343,070 1978 745,580 218,680 150,150 368,830 376,750 25,000 351,750 1979 755,180 218,680 151,080 369,760 385,420 25,000 360,420 1980 to 764,780 218,680 152,000 370,680 394,100 25,000 369,100 2000 1/ See Annex 10, SDV part B, for assumptions and supporting data. 2/ Cost streams from Table 4. Table 24 PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Benefit and Cost Stream since 1955 (million of soles, 1975 terms) Year Net Value of Production 2/ San Lorenzo 1/ Piura Valleys - Total Project Net (incremental) Costs 3/ Stream 1955 - - - 708.6 - 708.6 1956 - - - 662.4 - 662.4 1957 - - - 610.9 - 610.9 1958 - - - 565.7 - 565.7 1959 - - - 500.5 - 500.5 1960 - 343.4 343.4 - 343.4 1961 - 312.3 312.3 149.9 162.4 1962 36.4 584.2 620.6 177.6 443.0 1963 94.4 543.2 637.6 295.0 342.6 1964 383.1 589.5 972.6 349.4 623.2 1965 119.1 205.7 324.8 385.4 - 70.6 1966 247.7 348.5 597.2 293.0 304.2 1967 280.0 - 280.0 212.5 67.5 1968 284.2 214.4 498.6 101.9 396.7 1969 174.9 158.0 332.9 81.8 251.1 1970 375.5 303.6 679.1 161.7 517.4 1971 343.6 87.6 431.2 154.0 277.2 1972 166.3 - 166.3 36.2 130.1 1973 378.7 442.1 820.8 41.4 779.4 1974 424.2 93.5 517.7 72.7 445.0 1975 383.5 18.6 402.1 44.0 358.1 1976 359.4 359.4 334.4 1977 368.1 368.1 343.1 1978 376.8 376.8 351.8 1979 385.4 385.4 360.4 1980 to 2000 394.1 394.1 25.0 369.1 1/ Taken from Table 9 of SDV, Annex 10. 2/ See (Part C) SDV Annex 10 for assumptions and supporting data. 3/ Cost stream from Table 4. PERU Table 25. LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Estimated Gross Farm Income, 1975 ('000 Soles, 1975 terms) Small Type of Cooperative Colono Medium Size Colono Farmer: Member (0-8 ha) (8-40 ha) (40-80 ha) No. of Farmers in group: 1,888 483 712 94 Average size: 4 ha 6.6 ha 16 ha 56 ha G.V.P./ 102.8 169.6 411.2 1,439.2 /2 Inputs- 32.0 52.8 128.0 448.0 /3 Hired labor- - 9.6 99.9 518.3 Total Production Cost 32.0 62.4 227.9 966.3 Gross Income 70.8 107.2 183.3 472.9 Estimated A terna- tive income- 33.8 33.8 Difference 37.0 73.4 /1' Estimated at S/. 25,700 per ha for all farms. /2 Estimated at S/. 8,000 per ha for all farms. 3 Assuming the farmer!s family can provide up to 50 work days per month. Assuming 225 work days per year at official wage rate; should be regarded as an overestimate. PEEU LOAN 418-PE - SAN LORENZO IRRIGATION PROJECT Water Availability and Use (in Million mJ rounded) INFLOW DISTRIBUTION Including Reservoir Operation Total Available Calendar Diverted from Inflow Chipillico San Lorenzo Lower Piura Chipillico TOTAL thrpugh Lover Year Quiroz River River TOTAL ProJect Valley Valley Distribution- Rio Piura Plura Valley 1 2 =T+Y tE % 5 % 6 % 7 ---1 9(5+8) 1957 212 154 366 NA - 212 - HK - - 530 742 1958 424 111 535 KA - 423 - NK - - 457 880 1959 460 59 519 NA - 178 - NK - - 408 586 1960 501 57 558 64 12 427 70 101 18 558 65 493 1961 507 40 547 46 10 278 72 65 14 476 87 365 1962 666 50 716 124 18 441 75 49 7 689 75 516 1963 474 27 501 151 29 331 64 38 9 52 4 335 1964 476 20 496 251 45 279 50 23 4 553 - 279 1965 484 93 577 234 45 250 49 31 6 515 411 661 1966 421 38 459 332 47 343 49 30 4 7052/ 25 368 1967 491 33 524 290 58 186 37 23 5 499 81 267 1968 338 4 342 187 51 167 45 13 4 367 - 167 1969 371 21 392 241 59 155 38 12 - 3 408 94 249 1970 688 16 704 292 50 275 47 20- 3 587 - 275 19711 619 36 455 335 58 244 42 NA - 579 70 314 1972 346 7 353 283 47 316 53 NA - 599 207 523 1973 273 115 388 315 47 349 53 NA - 664 279 628 1974 593 28 621 331 52 304 48 NA. - 635 32 332 1975 619 84 703 333 38 549 62 NA - 881 214 763 1976 258 108 366 262 39 409 61 NA - 671 313 722 1/ The total distribution taken into account only the flow of Quiron diversion and Chipillico river, and not the occasional flow of the Piura river to Lower Piura. 3 2/ The difference between the 705 million m distribution and the 459 Million m inflow in 1966 is due to drawing on the reservoir storage for over 200 Million m3; at the end of the year, the reservoir was down from a maximum capacity of 258 Million m3 (end of 1965) to 40 Million m3 (end of 1966). 3/ From Au1ust 1971, the distribution of the valles of Chipillico and San Lorenzo were treated as a single unit. 4/ The 1976 figure is until June 1976.  ANNEX 1 Page 1 PERU LOAN 418-PE SAN LORENZO IRRIGATION PROJECT Basic Data Project Description Original Loan Agreement (June 1965) 1. The Project is Stage III of the development of the San Lorenzo Irrigation and Land Settlement Project located in the Department of Piura between the Quiroz and Piura Rivers. Stage III involved settlement and development of approximately 17.700 ha of new land and the further development of 14,300 ha, and consists of: A. Acquisition of heavy equipment, vehicles, spare parts and repair equipment for operation and maintenance of the irrigation works, for land development operations, and for road construction, gravelling and repair. B. Provision of technical services, including technical assistance and training in the fields of research, equipment management, vocational education and marketing, and supervision of the Drainage Survey referred to below; and the acquisition or provision of equipment and facilities to support the above services, as well as to support the local services to settlers. C. Implementation of a medium and long-term agricultural credit program to provide financing to settlers for their on-farm investments. 2. The Project also includes the construction and operation of a drainage system, to be based upon a Drainage Survey which will be conducted on the lands included in Stage III as well as on the remaining lands which will be developed in succeeding stages. 3. The Project further includes the carrying out of a Basin Program consisting of: A. a survey of the water and related resources of Northwestern Peru; B. the preparation of plans for the financing and execution of projects to utilize these resources according to a system of priorities; and C. the recommendation of the most effective organisation for the implementation of said plans. ANNEX 1 Page 2 Revision of Loan Documents, March 1968 4. The introduction clause 1 of the Original Loan Agreement was amended to read: 1. The Project is Stage III of the development of the San Lorenzo Irrigation and Land Settlement Project located in the Department of Piura between the Quiroz and Piura rivers. Stage III involves the irrigation, settlement, cultivation of approximately 18,000 ha of land, and consists of': List of Side Letters Number Subject Status of Compliance A. Original Loan Agreement (1965) 1. List of Goods to be purchased major divergence in scope. out of proceeds of the loan. 2. Procurement, procedures to be Good followed. 3. Consultants: procedures for Poor with respect to research; the-employment of consultants. vocational education; very poor for marketing; slow with respect to drainage; major divergence in scope of drain- age study. 4. Operating Policies and (i) Operation of credit-good. Proceedings; procedures (ii) Livestock - not workable. and policies for carrying (iii) San Lorenzo Management - out the project. good initially but deformation of scope made condition not work- able. (iv) implementation of drain- age - not applicaLe due to divergence. 5. Supply of Funds; the amount Not applicable due to diver- or funds to be supplied for gence in scope. the execution of the project in 1965, 1966 and 1967 6. Basin Program; procedures for Fir but with considerable carrying out the Basin study delay and divergence. 7. External Debt Letteri position Good of Peru ANNEX 1 Page 3 B. Revised Loan Agreement (1968) 1. Revised List of Goods Major divergence in scope. Analysis of Staff Continuity on Loan 418-PE No. of staff involved Resident Representative (at appraisal) 1 Appraisal Total 8 Appraisal staff who then supervised the project 3 Supervision Total (1966-1976) 22 1 visit only 17 2 visits 1 3 visits 3 9 visits 1 Area Missions Total (1966-1976) 11 1 visit only 5 2 visits 3 3 visits 3 1/ Project Managers- - San Lorenzo Irrigation Project A. Under Irrigaci6n y Colonizaci6n de San Lorenzo Approximate dates Enrique Torres Llosa 1961 - 1962 Daniel Escobar 1962 Renato Rossi 1962 - 1965 Luis Joo 1965 - 1966 Luis Moscoso, F. 1966 - 1969 B. Under Zona Agraria I (operational stage) Juan Ibanez 1969 - 1972 Guillermo Valdivia 1972 - 1974 Manual Vazquez 1974 - 1976 not known late 1976 1/ Lack of continuity was because most Project Managers left San Lorenzo to take more senior positions in Government illustrating the important training impact of the Project. ANNEX 6 Page 1 PERU LOAN 418-PE: SAN LORENZO IRRIGATION PROJECT WATER AVAILABILITY AND USE1 Introduction 1. The governing constraint to agricultural development in the coastal plains of Peru is lack of an adequate and reliable irrigation; once excess water becomes available production might then be restricted by lack of drainage. The mission concludes that inadequate analysis of water availability and use in San Lorenzo was the predominant reason for many of the project's short- comings. 2. A fundamental fault with the Stage II development of San Lorenzo Irrigation Project was that storage water could not be used immediately for the 50,000 ha for which physical infrastructure had been constructed; the delay in formulating and implementing the colonization phase meant that from 1959 to at least 1962, water supplies meant for San Lorenzo were utilized in other parts of the Piura valleys. The result was that by appraisal of Loan 418-PE, the irrigated area in Lower Piura valleys had expanded to at least 40,000 ha and there were substantial increases in the irrigated areas in the Upper Piura valley and in the upper reaches of the Quiroz and Chipillico Rivers. This significantly increased the competition for water and had serious political repercussions through the colonization phase of San Lorenzo. Hydrology 3. The most important physical parameter which affected design and performance of San Lorenzo Irrigation Project is the hydrological character- istics of the rivers which serve the project area and the Lower Piura valley. Monthly and annual variation in flow are such that caution should have been taken at the planning stage to ensure that water resources were not over- committed. A basic project objective should have been to guarantee reliable irrigation and not commit water on an average annual basis. 4. Table 1 gives seasonal and annual flows for the rivers Quiroz, Chipillico and Piura respectively. The records show that the Quiroz, the main water source for the project, always has some flow but supplies vary greatly on seasonal and annual basis. Some compensation water has to be released below the San Lorenzo diversion for cultivation downstream but an important constraint is the capacity of the Quiroz diversion canal. Analysis of daily flow in the Quiroz (particularly during average flow years) shows flood characteristics which indicate that peaks cannot be diverted. This suggests that provision of micro-storage facilities upstream of the Quiroz diversion canal might usefully supplement storage in average and poor years by holding water until diversion capacity is available. 1/ Tables 1-3 have not been included here. ANNEX 6 Page 2 5. The monthly and annual flows in river Chipillico measured at San Lorenzo dam are even more variable than river Quiroz. Furthermore, there appears to have been a distinct falloff in water availability in latter years indicating agricultural development above the measuring station. Variations of monthly and annual discharges in river Piura are so great that agriculture would be extremely speculative in the Lower Piura valleys without supplemen- tary irrigation. The records in Table 1 have been generated to allow for the significant changes in water availability and demand in the Upper Piura valley due to supplementary supplies, return flows from the San Lorenzo project and changes in agricultural development. Given possible errors in making these adjustments, the data clearly shows that much of the base river discharge in the river Piura cannot be used for irrigation because it exceeds irrigation requirement at any point in time. 6. Table 2 summarizes monthly flows in Quiroz, Chipillico, and Piura rivers by periods which coincide with the two cropping seasons. This shows that most of annual variations inflows occurs during the main cropping season (January to May). The average, maximum and minimum discharges of the rivers illustrate the unlikeliness of similar flow characteristics for the three rivers coinciding at any specific time. This random variation is more pro- nounced in monthly water availability and indicates that more attention should have been given to relate the objectives-of any hyrological analysis to water use; namely that the development objective was to provide reliable water supplies to supplement random monthly variation in base flows on the rivers. 7. A minimum requirement would have been to do a probability analysis on water availability from all sources using sequential historical data and then analyze operational characteristics and see how system efficiency could be improved. A simplified analysis was undertaken in General Development Plan of 1963 but based on limited hydrological records available at that time and insufficient evaluation water use and system efficiency assumption; no evidence can be found that an adequate analysis has been undertaken since. In partic- ular, the marginal value of water through the year, distribution losses, analysis of return flow and essential assumptions in water use such as crop water requirement and irrigation efficiency have not been satisfactory analyzed. Assumptions at Appraisal 8. No comprehensive review of hydrological data was undertaken in the appraisal report presumably because design criteria assumed for San Lorenzo Stage II development were simply accepted; a possible factor might have been that water allocation was a politically sensitive issue. The General Develop- ment Plan of 1963 included a simulation of reservoir operation to serve the existing irrigated lands in Lower Piura valleys and 45,000 ha in San Lorenzo project area using historical sequential river flow records. Unfortunately historical river flows up to mid-1950 were above average and so this analysis indicated that water availability was reliable and seemed to have overestimated ANNEX 6 Page 3 the supply that could be-diverted from river Quiroz; it also clearly underesti- mated the on-going development outside the San Lorenzo project area. 9. The appraisal mission did not review preparation assumptions of criteria for water use; the report simply analyzed water availability for the 1964 season and concluded that approximately 730 MCM would be available in 1965. Although this barely covered water requirements with only 20,000 ha under cultivation in San Lorenzo project area, no analysis was done to test the probability of reliable supply for proposed future expansion. The assump- tion appears to have been that inclusion of the basin study in Loan 418-PE meant that any water shortage would be short term because alternative sources of water would be identified and implemented quickly. The appraisal mission accepted assumptions on crop water requirement and irrigation efficiency with little factual basis although some technical assistance was provided to examine and confirm these assumptions. 10. The appraisal mission clearly recognized that water availability might be a problem but thought the risk was covered by Government assurances which guaranteed a water allocation to San Lorenzo until alternative resources were developed. The significance of below average flow of source rivers after 1958 and important on-going changes in water use through the Piura valleys and other parts to the Piura basin do not appear to have been examined. This tech- nical shortcoming is difficult to explain except that the basic objective of Loan 418-PE was to protect the substantial investment made in Stage II development and not address the wider implication of the impact of San Lorenzo development in river basins. The Problems 11. Miscalculation of water availability made irrigated agriculture speculative in the San Lorenzo area for at least a decade; this had a detri- mental effect on new settlers who needed a reliable income to service their substantial credit debts and establish themselves over the first critical years of farming. Poor operation procedures compounded unreliability because sequential monthly supplies were not guaranteed when water was allocated for a cropping season. Existing records do not enable irrigation unreliability to be correlated with crop production except to note that production much less than potential; however there is clear evidence that unreliability had an important influence on management/settler relationship. 12. The delay in settling San Lorenzo area meant that politically strong farmers in the Lower Piura became accustomed to reliable water supply. This had two effects; they developed wasteful irrigation practices and San Lorenzo colonization encouraged them to lobby Government on water allocation legis- lation. The competition for water allocation increased as San Lorenzo project developed. Table 3 illustrates the problem of spreading a variable water resource over too large an area by showing the large annual variation in supply to San Lorenzo area, Chipillico and Piura valleys; in particular, the ANNEX 6 Page 4 table shows the decrease in supplies to the Piura valleys after 1962 when San Lorenzo started to use its allocation. 13. The unpredicted low water years from 1964 to 1969 meant that San Lorenzo reservoir was operated on a contingency basis; not surprisingly, interested groups attempted and did influence operation through political pressure and during this period the reservoir was only full for one month (May 1965). Table 3 shows reservoir storage at the end of each month through this critical period. Table 4 records the large seasonal fluctuation in water distribution and illustrates the basic operational problem; San Lorenzo management obviously needed much more flexibility in the operation of water resources if the marginal value of water was to be optimized. 14. An additional operational problem was that San Lorenzo had about 25% of the irrigated area to perennial crops. Surprisingly, water allocation did not recognize the difference in marginal water value between annual crops (January to June when run of river supplies were available) and perennial crops which depended on storage through the dry period from July to December. Table 5 gives water allocation on a monthly basis for 1966 and shows that the most valuable water, namely dry season river flows or storage water, was made available only to San Lorenzo with small amounts to Chipillico areas. Thus although only 46% of water allocation went to San Lorenzo project area, this included 80% of the dry season supply giving the settlers a considerable agricultural advantage. It also illustrates that Lower Piura farmers had irrigation objectives which were substantially different from San Lorenzo. Their need was a reliable water supply to make up for river Piura variations during the wet season and timely supplies to guarantee a planting date for cotton. The significance of this important difference in objectives was not addressed; it is concluded that system operation could have been improved to benefit both groups if water availability and use had been logically analyzed. 15. Poor management of water resources in the critical early years of Loan 418-PE encouraged inefficient competition for water; users took allo- cations without appreciating that limited stoarge meant that waste of water in the short-term had a serious consequential effect on annual or long-term availability. This partly explains why the reservoir was seldom filled and was depleted at unexpected times. Table 4 shows that Lower Piura farmers almost certainly took unnecessary water from the San Lorenzo system in 1965, 1972 and 1973; similarly in 1966, it appears that San Lorenzo settlers used more water than was strictly necessary for the annual cropping requirement. It appears that with better management some of these defects might have been anticipated and handled more efficiently. The Consequences 16. Over commitment of a variable water resource meant that whatever guarantees were given to the Bank, it would be extremely difficult for Govern- ment to resist powerful political interest of farmers in the Lower Piura ANNEX 6 Page 5 valley. They had become used to a reasonably reliable irrigation supply when supply exceeded demand; accelerated development of San Lorenzo meant that water became an increasing constraint which was critical in years of short supply. San Lorenzo management thus had to accomodate strong pressures from the Bank to protect investment in Stage II development of San Lorenzo and from Lower Piura farmers to provide what they now considered to be an estab- lished right. 1966 turned out to be the critical year; there was a serious shortage of water which was not apparent until after a substantial area had been committed to annual cropping; the marginal benefit of the water committed was mainly lost because storm damage destroyed about 40% of the cotton crop. This intensified the friction between San Lorenzo and the Lower Piura farmers and the hydrological predictions accepted at appraisal were questioned. The Bank's reaction was to make Government promulgate temporary water legislation to protect the agreed right and objectives of the San Lorenzo Project; unfor- tunately this was done before a logical study was undertaken to establish a factual basis for corrective action. 17. The Bank retained a consultant, Harry W. Adams, in October 1967 to examine the system hydrology. The task was to appraise the accuracy of basic hydrological parameters, procedures used for data collection and compare annual water availability in recent years with long-term annual runoff and to evaluate the differences. The report concluded that available long-term records of stream flow showed substantial reduction in years inmmediateiy before and after appraisal due to below average rainfall in the catchment areas and/or an increase in development in the upper watersheds. The results were not conclusive except that the project share of the annual storage would permit irrigation of only about 19,000 ha. The analysis assumed that the annual yield of San Lorenzo reservoir would be 650 MCM compared to 800 MC4 predicted at appraisal. 18. Data available to the Adams study was not adequately analyzed; in particular, no probability analysis was made to establish relative variability in water supplies from the various sources. Furthermore, the report did not establish the probability of two or more consecutive low years and the conse- quential effect on water use. It did not examine the possibility of either increasing storage capacity or the importance of Quiroz Diversion Canal as the constraining link in the supply system. Water use, assumptions for crop water requirements and irrigation efficiency in terms of marginal value of water were not reviewed for either San Lorenzo project area or more particularly Lower Piura valley. It is concluded that the Adam study contributed little to the understanding of the complex basin hydrology or water use options; this was particularly serious because findings were the basis for loan renego- tiation in early 1968. The hydrological assumptions were not subject to comment at renegotiation; no mention was made of the possibility of increasing water availability or water use efficiency or checking whether crop water allocations were realistic. The implicit assumption seems to have been that water shortage in San Lorenzo project area would be a short term problem because Lower Piura Valleys could not absorb extra water supplies beyond what ANNEX 6 Page 6 they were then getting without compounding the drainage problems. The only clear conclusion was the need to guarantee water availability for 18,000 ha in San Lorenzo area; it was not made clear how this assurance was to be obtained except that a method, satisfactory to the Bank, should be evolved and imple- mented. 19. 1968 was one of the worst water years on record; it was the lowest flow year for rivers Quiroz and Chipillico and there was no measurable flow in the river Piura. Lobbying by hard pressed farmers in Lower Piura pushed Government into temporary legislation which radically altered water allocation in their favor; Decree No. 015-69-FO (March 1969) allocated 46.8% of the annual flow to both San Lorenzo and Lower Piura, with the remaining 6.4% to Chipillico Valley. Based on 1969 water availability, this did not assure water for the 18,000 ha as guaranteed to the Bank at loan renegotiation. A major clash was avoided between the Bank and Peru because all operations were suspended in March 1969; however, the Bank made it clear that a condition of re-starting disbursements was removal of the offending decree. The political problem was subject to intense negotiation through 1969 and by September a compromise was reached. In January 1970, a new supreme decree was passed guaranteeing 60% of water availability to San Lorenzo and forbidding extra development in the Lower and Upper Piura valleys until additional water resources were developed; based on average crop water requirement of 15,130 m3! ha, it was estimated that there would be reliable water supply for about 29,000 ha. During this period of major dispute over water allocation, all analyses were based on an annual allocation with the marginal value of water and significance of monthly variations in any year apparently not recognized or considered. 20. The run of poor water years came to an end in 1970 when annual needs were satisfied for the first time since Loan 418-PE became effective; there was spare water to replenish storage and by December 1970, the reservoir retained about 150 MCM, the highest level since April 1966. 1971 was an even better year and it was possible to fill the reservoir for the first time since May 1965. Since that date there have been above average water years in all rivers which coupled with the end of political competition for water after agrarian reform has meant removal of a major problem that plagued the project in its early years. In fact, San Lorenzo management took the risk of increasing the settled areas in San Lorenzo catchment by developing 5,300 ha in the Valle de Los Incas and Valle Hermoso. 21. In retrospect, it would appear that the Adams report led to pessi- mistic conclusions at loan renegotiation in 1968; settlement has expanded to 32,900 ha or more than the appraisal target of 32,000 ha; reasonably reliable irrigation has given an above average income to settlers; it has also pro- tected the valuable investment in tree crops (now covering about 5,700 ha) which needed perennial irrigation; and finally alternative supplies to Lower Piura valley has meant that now there should be no future danger to agricul- tural production in the San Lorenzo system area. ANNEX 6 Page 7 22. Clearly, many problems could have been avoided if an adequate hydrological analysis had been undertaken at preparation or appraisal. Certainly at renegotiation, a more satisfactory and reliable hydrological analysis should have been made to establish why and how the preparation assumptions were so optimistic given the low water years from 1965-1968. However, as with many elements of the San Lorenzo project, errors seem to have been self-compensating; the good water years from 1970 has meant that present settlement greatly exceeds expectancy at loan renegotiation and is now reasonably secure because the long awaiting alternative water source to the Piura valley has been commissioned. Water Legislation 23. There have been five water laws or modification to laws, 13 decrees, and three supreme resolutions since completion of Stage I construction of San Lorenzo. As development progressed, the San Lorenzo right to water has been progressively reduced as follows: Allocation to various areas (%) Date Law No. San Lorenzo Chipillico Piura 1. July 1959 13240 67.2 32.8 2. December 1961 14064 64.0 3.2 32.8 3. February 1966 SDIOF 41.3 6.2 52.2 4. March 1969 SDO15-69FO 46.8 6.4 46.8 5. July 1969 1752 54.65 45.35 Water Law No. 14064 modified the first law to establish water rights to the Chipillico valley. At negotiation, the Bank insisted on legislation which guaranteed allocation thought to be sufficient to irrigate 32,000 ha (Supreme Decree No. 37). 24. The drought of 1966 was disasterous for Lower Piura farmers and resulted in Supreme Decree 1OF of February 1966; this legislation was a com- promise solution to deal with totally unexpected minimum water availability. It introduced water allocation based on guaranteed allotments for the first 10 ha to all users of the system and variable allocation according .to water availability for the rest of the irrigated area. Such a distribution favored the San Lorenzo settlers because of the small farm size. Renegotiation of Loan 418-PE guaranteed water supplies for 18,000 ha and a major crisis occured when Supreme Decree 015-69-FO (dated March 1969) reallocated water to favor Lower Piura as explained in paragraph 20; fortunately the Government introduced a new general Water Law No. 1752 of 1969 which superseded all previous legal ANNEX 6 Page 8 allocation. This comprehensive legislation made all water resources state property to be administered through irrigation districts with full authority to regulate and administer the agricultural use of water. Following this legislation, San Lorenzo Project became part of the Irrigation District of San Lorenzo and Chipillico which includes about 29,000 ha in San Lorenzo project area, 2,500 ha in Chipillico valley, about 7,200 ha in Valle de Los Incas, Valle Hermoso and other small areas. 25. Water allocation and legislation has been governed by short-term priorities and dominated by political and social factors rather than sound technical analysis of water availability and use. The statistical signifi- cance of the unforeseen drought was never established for the critical period 1965 to 1969. Legislation was based on annual allocation to various areas and did not consider the marginal value of water through a yearly cycle or system efficiency and operational characteristics of the various systems. Water has been the governing constraint to development yet nominal attention was given to a logical basis for legislation to ensure that water was used effectively and efficiently. Water Management and Use 26. Surprisingly little attention has been given to water management, irrigation practices or water use efficiency given the mixed croppilg pattern of annual and perennial srops. Crop water requirements of 15,130 m /ha for San Lorenzo and 10,300 m /ha for Piura valleys were design criteria for Stage II development of San Lorenzo based on empirical results in the respec- tive areas. The coefficients were target annual allocation for the users; monthly water allocation were then calculated for each year depending on predicted water availability and fixed parameters such as soil type and proposed cropping pattern. Loan 418-PE did provide technical assistance to establish crop water requirements for each sub-district but this was never done. It is interesting to note that even though water allocations in the drought years have been much lower than crop design requirements, there has been no corresponding decrease in agricultural production. 27. Only limited data is available on actual water use within San Lorenzo area but these report a surprisingly low distribution loss of 16% for the system. There are frequent references in the correspondence to low system efficiency particularly in the years when water was plentiful; irriga- tion efficiency was particularly poor at night and generally irrigation practices were only efficient when water was in short supply. The loan financed technical assistance to investigate and evaluate water management and use; this was not done and now lack of reliable data makes water use difficult to evaluate. Water discipline has been clearly lax and farmers tend to use water when available to avoid short fall at some later time in the crop cycle. Unfortunately once the serious drought cycle ended in 1970, the ANNEX 6 Page 9 urgency to examine and evaluate water management and use has decreased. A supervision report in 1971 still mentions the low irrigation efficiency but did not recommend any corrective action and the subject seems now to have been dropped. 28. The primary objective of San Lorenzo Project was to provide re- liable water resources yet nominal research was done to establish cropping patterns that would optimize return per unit volume of water; this would seem to be an essential exercise to formulate a realistic water use and charge policy. If the necessary research had been done and made known to the settlers, commercial pressures would have certainly encouraged more efficient use of water or at least made both management and settlers conscious of the marginal value of water. Present Situation 29. The development of the Poechos dam on the river Chira and the construction of the Chira-Piura division canal means that there should always be adequate water for Lower Piura valley; the difficulty will be to restrict water supplies to actual need and avoid aggravating the serious drainage problem. Once adequate supplies are allocated, water and not land are likely to remain the governing constraint to further development; if any extension to project area is to take place, two opportunities in the San Lorenzo system to increase storage and optimize diversion supplies might be evaluated. The level of San Lorenzo reservoir might be faised by two meters to give about 30 MCM additional live storage and micro-storage reservoirs might be constructed on the upper reaches of the river Quiroz so that the Quiroz diversion canal does not act as a constraint. Water Charges 30. Historically the water charges of San Lorenzo project have been ex- tremely low, averaging about S/. 0.006 per m3 or about S/. 90.0 per ha. Little attention was paid to cost recovery until 1969 for the following reasons: (i) the project was beseiged with more urgent problems; (ii) unreliable and inadequate irrigation and other technical problems meant that settlers could not afford to pay for realistic charges or had a ready excuse to delay or avoid payment; and (iii) management did not have firm Government backing for effective action or a corrective plan which would be fair to all settlers. ANNEX 6 Page 10 The Government acted in 1969 by promulgating the.Water Law No. 1752 which included principles for charging for water. It proposed a two part water fee; a water tariff collected through the Ministry of Agriculture based on volume used and an additional-fee or quota collected and dispersed through a water users' association. Water tariffs were to reflect the users' capacity to pay based onprojected benefits per unit area and not farm units. The Sub- direccion of Tarifas y Cuotas of the Direccion General de Aguas has formulated and is currently implementing plans to enforce the water law. For each proj- ect they determine the charges for right to use water, service for 0&M and amortization to recover all or some Government investments. The procedure works for valleys with long established irrigation but in areas subject to agrarian reform with costly infrastructure or new irrigation schemes, the constraint is likely to be the ability to pay. 31. The watgr charges in San Lorenzo were raided from q/. 0.006 per m3 to S/. 0.02 per m retroactively in 1971. Lower and Middle Piura valleys were kept at the old rate reflecting poorer farming opportunities and capacity of cooperative members to pay. The plan has been to gradually increase charges to cover full running costs over five years. To improve recovery, Banco de Fomento Agropecuario (BFA) is acting as collecting agency. Actual collection is lower in San Lorenzo than in Zona Agraria I and both are lower than Peru as a whole. In 1973, only 16% of issued receipts were recovered compared to 19% for Zona Agraria I and 31% for the country. More important, charges collected in 1973 for San Lorenzo only represent about 5% of the O&M charges which are known to be low for preventive maintenance. 32. Present y, the current water charge for San Lorenzo has been raised to S/. 0.04 per m (about S1. 600 per ha) and is probably the highest in Peru. If collected this should cover about 65% of the running cost excluding drain- age. Major problems to recovery are: (i) inadequate data to establish a realistic O&M cost; (ii) if funds were available, facilities and staff might not be available to effectively carry out the program; payment capacity of users cannot be fully analyzed until tree crop production is evaluated; (iii) water has yet to be proved reliable with distirbution biased in favor of the new settlers (squatters) who have the least capacity to pay compared with earlier settlers; this encourages a social resistance to charge recovery; and (iv) a future problem is that drainage works benefit only about 5% of the irrigated area and many farmers might hesitate to pay for 0&M costs for an element which does not benefit them. 33. It is concluded that considerable progress is taking place to dent with a national issue which will become increasingly important as major on- going developments are commissioned. Government now has legislation to take positive action and appears to be making progress to implement a sound strategy. Clearly privileged beneficiaries such as San Lorenzo settlers could and should be made to pay a reasonable portion of costs for their special facilities.  甘…