Tunisia - Grain Distribution and Storage Project

Docwiient of The World Bank FOR OMCIAL USE ONLY ~ ~ CT:( )1F .1 CD! ReportNo. 10144 y}, ),.Y ,, A ,' 5' i I 1' '; sI(,JX (,I'A,( Wi. (R A . ,)' I) PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA GRAIN DISTRIBUTION AND STORAGE PROJECT (LOAN 2052-TUN) DECEMBER 5, 1991 Agri. ulture Operations Division Country Department II Europe, Middle East and North Africa Regional Office This document bas a restricted distribution and maay be u-sed by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit : Tunisia Dinar (TD) EXCHANGE RATE Appraisal year average (1981) US$1 - TD 0.50 Intervening years average (1982-87) US$1 = TD 0.75 Completion year average (1988) US$1 TD 0.86 WEIGHTS AND MEASURES Metric System ACRONYMS AND ABBREVIATIONS ATI Agrotechnie-Ingdnierie, Consultant Engineers, France CCGC Cooperative Centrale des Grandes Cultures (Central Cooperative for Basic Crops) COCEBLE Cooperative Centrale de Ble (Central Vheat Cooperative) CNEA Centre National d'Etudes Agricoles (National Center for Agricultural Studies) OC Office des Cereales (Office of Cereals) OPNT Office des Ports Nationaux Tunisiens (Tunisian National Port Board) PD Project Directorate PCC Project Coordination Committee SAR Staff Appraisal Report SCET Soci4td Centrale pour l'Equipement du Territoire (Central Company for Territorial Development) SNCFT Socidte Nationale des Chemins de Fer Tunisiens (Tunisian National Railroad Company) STPA Societe Tunisienne de Production Alimentaire (Tunisian Food Production Company) TECI Tunisie Engineering et Construction Industrielle (Tunisian Engineering and Industrial Construction Enterprise) FISCAL YEAR Republic of Tunisia January 1 - December 31 Office of Cereals October 1 - September 30  TBN OFFICIAL USE ONLY THK WOULD BANK Washington, DC 20433 U.S A OVuice ni oireitn.C.fl41 Opsmvws Ivakiattr- December 5, 1991 MMORANDUM TO THE EXECUTIVE DIRECTORSANTEPRSDT SUBJECT: Project Completion Report - Tunisia Grain Distribution and Storage Protect (Loan 2052-TUN) Attached, for your information, is a copy of a report entitled "Project Completion Report on Tunisia - Grain Distribution and Storage Project (Loan 2052-TUN)", prepared by the Government of Tunisia, with an Overview prepared by the Europe, Middle East and North Africa Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment I This document hu a rstricted distribution and may be used by recipients only in the petformance I of their official duties. Its contents may not otherwise be disclosed without World Bank authoriation.  FOR OFFICIAL USE ONLY PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA GRAIN DISTRIBUTION AND STORAGE PROJECT (LOAN 2052-TUN) TABLE OF CONTENTS Page No. PREEACE . . . . . . . . . . . . . . . . . . . . . . BASIC DATA SHEET . . . . . . . . . . i.i EVALUATION SUMMARY . . . . . . . . . . . . . . . . . . v OVERVIEW . . . . . . . . . ..... . . . . . . . . . vlii CNEA PROJECT COMPLETION REPORT I. BASIC DATA . . . . . . . . . . . 1 A. Introduction . . . . . . . *. . . . . . . 1 B. Agriculture in Tunisia: General Data . . . . . 1 C. The Cereals Office and the Cooperatives . . . . 3 D. Storage Infrastructure . . . . . . . . . . . . 6 II. IDENTIFICATION, PREPARATION AND APPRAISAL . . . . . 6 A. Origin .... . . . . . . . . . . . . . . . . 6 B. Preparation and Appraisal . . . . . . . . . . . 7 C. Description .... . . . . . . . . . . . . . . 8 III. PROJECT EXECUTION . . . . . . . . . . . . . . . . . 10 A. Project Start-up . . . . . . . . . . . . . . . . 10 B. Amendments During Implementation . . . . . . . . 10 C. Implementation Schedule . . . . . . . . . . . . 10 1. The Seven Silos . . . . . . . . . . . . . . 10 2. Port Works .... . . . . . . . . .... . 17 3. Railway Program . . . . . . . . . . . . . 18 4. Technical Assistance . . . . . . . . . . . . 20 5. Training .... . . . ...... . . . . . 22 D. Project Cost, Financing and Disbursements . . . 23 1. Project Cost . . . . . . . . . . . . . . . . 23 2. Financing and Disbursements . . . . . . . . 24 E. Benefits . . . . . . . . . . . . . . .. . . . 27 F. Economic analysis . . . . . . . . . 32 This document has a restricted distribution and may be used by recipients only in the performance L of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  TABLE OF CONTENTS (CONT'D) ANNEX 1 . . . . . . . . . . 34 Table 1 Population Structure by Broad Age Group 34 and Sex Table 2 GDP, by Activity .. . . . . . . . . . 35 Table 3 Areas, Output and Yields . . . . . . . . 36 Table 4 Grain Imports, Collections, Demurrage . 37 Quay Storage Fees Table 5 Grain Unloaded, by Port . . . . . . . . . 38 Table 6 Demurrage Charges - 1988 . . . . . . . . 39 ANNEX 2 . . . . . . . . . . . . . . . . . . . . . 40 Table 1 Supplementary Works and Deductions in the 40 Silo Construction Component Table 2 Implementation Schedule . . . . . . . . . 41 Table 3 Works Schedule: Seven Silos Program . . 42 Table 4 Silo Construction: Delays in Completion . 43 Table 5 Contract with BUhler-MIAG (Contract Amount) 44 Table 6 Contract with BUhler (Detailed) . . . . . 45 ANNEX 3 . . . . . . . . . . . .. . . . . 46 Table 1 Accounting Status of Project . . . . . . 46 Table 2 Comparative Costs . . . . . . . . . . . . 47 Table 3 Accounting Status . . . . . . . . . . . . 48 ANNEX 4 . .....................51 Table 1 Handling of Imported Grain with and . . 51 Without Project (1,000 Tons) Table 2 Saving on Handling of Imported Grain . . 52 Table 3 Transportation Costs for Imported Grain . 53 Table 4 Annual Grain Transportation Costs Over . 54 Planned Itineraries Table 5 Grain Transportation by Road in 1988 . . 55 ANNEX 5 . e . e. . . . . . . . . . . . .. . 56 Table 1 Investment at Economic Prices (in D'0OO) 56 Table 2 Staff Wages and Salaries at Economic . . 57 Prices (D'000) Table 3 Staff Wages and Salaries at . . . . . . . 58 Economic Prices Table 4 Electricity Consumption at 1988 . . . . . 59 Economic Prices Table 5 Annual Operating Costs at . . . . . . . . 60 Economic Prices ATTACHMENT: Comments from the Borrower . . . . . e * 61  PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA GRAIN DISTRIBUTION AND STORAGE PROJECT (LOAN 2052-TUN) PREFACE This Project Completion Report (PCR) reviews implementation of the Grain Distribution and Storage Project in Tunisia, for which Loan 2052-TUN in the amount of US$42 million was approved on October 6, 1981. The Loan was closed on December 31, 1988, two years behind schedule. The last disbursement was made on July 19, 1989 and an undisbursed balance of US$804,000 was cancelled at that time. This brought the total amount cancelled under the Loan to US$9.41 million. The PCR was prepared by the National Center of Agricultural Studies (CNEA) in Tunis. The overview prepared by the Bank provides additional information obtained from a review of the project files including the preparation reports of the follow-on Project (subsequently dropped from the lending program). This PCR was read by the Operations Evaluation Departme.nt (OED). The draft PCR was sent to the Borrower for comments on September 23, 1991. An English translation of their telexed response is attached to the Report (Attachment).  PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA GRAIN DISTRIBUTION AND STORAGE PROJECT (LOAN 2052-TUN) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as X of Exoectation Current Estimate Aporaisal Estimate Project Cost CUSS million) 93.9 60.00 64 Underrun or Overrun 6X) Loan Amount CUSS miltion) 42 m. 32.59 78 Date Board ApprovaL 10/06/81 Date Effectiveness 04/02/82 Date Physical Components Completed 11/86 07/88 Closing Date 12/31/86 12/31/88 Economic Rate of Return 6X) 16.1 12.87 CUMULATIVE DISBURSEMENTS FY82 FY83 FY84 FY85 FY86 FY87 FY88 FY89 FY90 Appraisal Estimate CUSS miLLion) 0.70 13.60 30.50 39.00 42.00 ActuaL (USS miltion) - 0.40 0.80 7.54 20.65 30.51 32.34 32.38 32.59 Actual as % of Estimate 0.00 0.03 0.02 0.20 0.50 Date of FinaL Disbursement JuLy 19, 1989 CancelLation CUSS milLion) 9.40 STAFF INPUTS (staff weeks) FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 FY89 FY90 TOTAL LDV 3.4 6.1 9.5 Identification/ 35.8 35.8 Preappraisal AppraisaL 44.0 5.2 49.2 Negotiations 3.6 3.6 Supervision 12.6 11.4 15.8 9.0 3.1 10.9 2.0 0.5 0.1 65.4 Other 1.3 0.3 1.0 0.8 0.4 3. Sibotael 83.2 28.8 11.4 15.8 9.3 4.1 11.7 2.0 0.9 0.i 167.3  - iv - MISSION DATA Date No. of Staff/Days Specialization Performance Types of Date of tmoLyr) Persons _infield_ represented a/ Reting b/ Trend c/ PLRE d/ Report Identification 3/79 FAO 06/15/79 Preparation 6/79 5 12 FAO 07/26/79 Preparation 10/79 4 19 FAO 10/26/79-01/18/80 Preparation 9/80 2 7 EC, GRS 10/20/80 Preappraisal 11/80 4 14 EC (2), GRS, TREC 12/03/80 Appraisal 2/81 5 21 EC (2), GRS, TREC (2) 03/11/81 Post Appraisal 5/81 3 5 EC, GRS, PR 06/03/81 Supervision 1 12/81 2 5 EC, GRS 2 1 M 01/29/82 Supervision 2 4/82 1 8 EC 2 2 M 05/05/82 Supervision 3 12/82 1 8 AG. ENG. 3 2 M 01/27/83 Supervision 4 4/83 1 5 EC 05/19/83 Supervision 5 5/83 1 4 EC 2 1 M 06/03/83 Supervision 6 10/83 2 6 EC, FA 11/02/83 Supervision 7 2/84 2 17 EC, GRS 2 1 FO 02/27/t4 Supervision 8 5/84 1 6 EC 05/08/84 Supervision 9 10/84 2 6 EC, SIC 2 2 FM 11/03/84 Supervision 10 5/85 2 13 EC, TREC 2 2 FM 06/21/85 Supervision 11 5/86 1 4 PA 06/03/86 SIR 7/86 3 4 DC, FA, AG 07/31/86 LENP Grain Storage 2 4/28/86 4 12 P EC (1), GRST (3) 05/07/87 Supervision 12 6/87 1 14 GRST 07/87 Supervision 13 12/87 1 5 FA 01/05/88 SIR 6/85 1 2 EC 06/13/85 SIR 1/89 2 5 FA, OA 03/04/89 OTHER PROJECT DATA Borrower Republic of Tunisia Beneficiary Office of Cereals (OC) FiscaL Year of OC October 1 - September 30 Name of Currency Dinar (DT) Currency Exchange Rate: AppraisaL Year Average (1981) US$1.00 = TDO.50 Intervening Years Average (1982-87) US$1.00 = TD0.75 Completion Year Average (1988) US$1.00 = TDO.86 Cofinanciers None Follow-on Projects Grain Storage II was prepared but subsequently dropped from the lending program. a/ AG = Agronomist; AG. ENG. = Agricultural Engineer; DC = Division Chief; EC = Economist; FA = FinanciaL Analyst; FAO - FAO/CP; GRS e Grain Silo Specialist; OP = Operations Assistant; PA a Project Advisor; PR = Procurement Specialist; SIC = Silo Consultant; TREC = Transport Economist. b/ I = Problem free or minor problems; 2 c Moderate problems; 3 = Major problems. c/ 1 = Improving; 2 = Stationary; 3 = Deteriorating. d/ F = Financial; H = Managerial; T = Technical; P = PoLitical; 0 = Other.  - V PROJECT COMPLETION REPORT TUNISIA GRAIN STORAGE AND DISTRIBUTION PROJECT (Loan 2052-TUN) EVALUATION SUMMARY Objectives The project was designed to expand grain storage and throughput capacity, improve the related domestic transport system and largely complete the conversion from bag to bulk handling in Tunisia's distribution system for imported grain to meet the increased demand for grain. In addition, the project was to provide assistance for strengthening the technical capacity and financial management of the Office of Cereals (OC). The main project component, accounting for more than 80 percent of total project costs, was construction, rehabilitation and expansion of port and storage silos and related facilities. Implementation Experience The project had a slow start due to the difficult relations between the OC and its consulting engineers and the time taken to select a firm for silo supervision. This delayed the infrastructure program by two years and the loan closing date was extended from December 31, 1986 to December 31, 1988. Despite the delays, all the project silos were constructed and/or rehabilitated as foreseen at appraisal and hence the project's physical objectives were met. The institutional component was less successful due to weak support for the policy and institutional aspects of the project within the OC, staffing problems and the generally poor working conditions at the OC. However, the various organizational units were created within the OC and the studies were completed and became useful guides for the reform program initiated under the project. The project has contributed to the reduction of demurrage charges through the eliminAtion of bags and to savings from the elimination of quay storage. Results The project generally met its objectives. The infrastructure program was completed as fores6en, the silos have performed very well and were able to handle the exceptionally ia~iie volumes of imports arriving at the ports in 1988. Although the objectives of the institutional component were more difficult to achieve, important reforms were initiated under the project and are still on-going in the context of the Second Agricultural Sector Adjustment Loan. The reestimated economic rate of return (ERR) of the whole project is satisfactory at 12.87 percent.  - vi - Sustainability Provided reasonable standards of operation and maintenance are met, sustainability is virtually assured throughout the assumed economic life of the project investments. Lessons Learned A number of lessons can be drawn from the experience with this project. (a) Institutional objectives of a project can be difficult to achieve if the project institutions operate in a regulated policy environment. This is starting to change with reforms being implemented under the ASAL-Il. (b) In a complex, multi-sectoral project requiring the participatior. of various agencies and ministries, more attention should be paid to ensuring the effective coordination of the different actors. (c) Complex projects of this nature require intensive supervision both at project start-up and during implementation. A greater supervision effort might have helped resolve inter-agency conflicts at their early stages and have moved forward difficult institutional reforms (financial and management). (d) Consultants' qualifications and terms of reference should be checked carefully to ensure that they have the required experience and sufficient time to carry out their work, and that they are familiar with the local context. (e) Continued attention needs to be paid to operation and maintenance (O&W1) which are crucial to ensure the sustainability and effective utilization of project investments.  PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA GRAIN DISTRIBUTION AND STORAGE PROJECT (LOAN 2052-TUN) OVERVIEW INTRODUCTION 1. The Project Completion Report (PCR) of the Grain Distribution and Storage Project was prepared by the National Center of Agricultural Studies (CNEA) in Tunisia. The CNEA report provides a detailed and exhaustiv3 review of the technical aspects of the project. However, less attention was given to institutional issues and the lessons to be derived from the project. The Evaluation Summary supplements CNEA's report with the Bank's observations. 2. Project Origin. An IBRD mission in May 1978 identified a grain storage project within the program of investments then being considered by the Government. This project was considered to be urgent because rapidly rising consumption of grain, brought about by rising per capita income, increased urbanization and a high population growth rate, was increasingly being met by imports and this was putting strains on the Government's systems for grain importing, storage and distribution. 3. The project included the following components: (a) rehabilitation, adaptation for imports, and expansion to 30,000 ton capacity of the existing port silo in Bizerte, and construction of a 30,000 ton port silo in Ghannouch (near Gabes); (b) construction of three storage silos in Sfax (20,000 tons), Beja and Gafsa (10,000 tons each), and expansion from 8,000 tons to 28,000 tons of the storage silo in Kalaa Seghira (near Sousse); (c) rehabilitation of the Manouba (near Tunis) storage silo (50,000 tons); (d) construction and rehabilitation of silo rail sidings; (e) acquisition of about 50 bulk rail hopper cars; and (f) studies, technical assistance, and training. The OC was to be responsible for silo rehabilitation and construction, and the execution of the studies included under the Project. The OC was to be assisted by the Tunisian National Port Authority (OPNT) for port works at Bizerte and by the Tunisian National Railroad Company (SCNFT) for components in (d) and (e) listed above. The SAR proposed that the OC would work with two grain cooperatives, the Central Wheat Cooperative (COCEBLE) and the Central Cooperative for Basic Crops (CCGC), for  - viii establishing an integrated silo maintenance and stock control system. Day- to-day management and coordination with other project agencies was to be provided by a Project Management Unft (PMU) established in OC, A Project Coordination Committee (PCC) was to provide overall project coordination and resolve possible inter-agency conflicts. 4. Proiect Cost and Financing The project was to be implemented over a five year period from 1981 to 1986. Total project cost (including taxes and duties of about $14.7 million) was estimated at TD 37.5 million (US$93.9 million equivalent) with a 45 percent foreign exchange component of which the Bank would finance $42 million and grant transfers to the OC from the Government would finance the balance. PROJECT IMPLEMENTATION AND IMPACT 5. Project Implementation. The project start-up was slow due to the difficult relations between the OC and its consulting engineers, which delayed the issuance of bid invitations for silo construction (para. 10). Despite the delays, all of the project silos w:tre constructed and/or rehabilitated by mid- 1988 aiid hence the project's physical objectives were largely met. The institutional component was less successful (paras. 13-17). The iotal project cost at completion was DT 47.8 n'llion (US$60 million equivalent) or 127 percent of the appraisal cost estimate in Dinar terms and 64 percent in dollar terms. The higher costs were incurred mainly for silo construction and supplementary works. Costs in dollars terms were lower than those in Dinar terms due to the fall in the value of the Dinar. 6. A Project Coordination Committee was created as intended (para. 3). Unfortunately, it did not convene regularly as envisaged and was not very effective in coordinating the various project components and fostering communication and team work. This affected the smooth implementation of the phys .cal and institutional components of the project. 7. kroiect Design. The basic design of the project was not modified during implementation. However, the Project Agreement was amended in 1984 to provide new deadlines for submission of 1981-85 audited accounts, and the Loan Agreement was amended twice, in 1984 to increase the local training disbursement percentage to 100 percent, and in 1985 to increase to 60 percent the disbursement percentage for expenditures on silo construction and silo rail sidings and to provide for an expansion of the Beja silo capacity from 10,000 to 18,000 tons. A total of about $9.4 million was cancelled from the Loan. The Loan closing date was extended by a total of two years to December 31, 1988.  ix- 8. Physical ComRonent. The silos performed very well and were able to handle the exceptionally large volumes of imports arriving at the ports in 1988 (2.04 million tons versus 1.25 million projected by the SAR for 1986). They have led to considerable efficiency gains in import and handling of grain -- a major project success. The railway program was also highly successful and 912,000 tons were moved in 1988 versus 570,000 estimated in the SAR for 1986. The project has, in addition, contributed to the reduction of demurrage charges through the elimination of bags and to savings from the elimination of quay storage. The Project director was able to very competently supervise silo operations once the works were completed. As mentioned in paras. 9-11, the process of achieving the physical results described above was, however, fraught with difficulties and delays caused by the poor relations of the partnership of consultants selected to carry out silo supervision. 9. Silo Consulting Engineers. At the time the project was being prepared, it was recognized that no Tunisian firm had adequate experience to take overall responsibility for major silo supervision. The Bank thus recommended, and the OC agreed, that for supervision of silo construction a contract with a joint venture of two partners (Tunisian and foreign) would be established with the senior level expertise provided by the foreign firm. 10. However, the project inherited a major problem of incompatible consulting engineers and subcontractors. In 1978, prior to the Bank's involvement with the project, the OC had hired an engineering consulting firm to prepare detailed designs for the Bizerte port silo and the storage silos of Beja, Gafsa, Kalaa Seghira and Sfax. At the OC's insistence, the firm subcontracted a large share of the engineering work to a Tunisian firm and a joint contract was signed in early 1979. A Bank preparation mission reports that relations between the OC and the international firm were strained and relations between the two firms (internptional and local) were antagonistic. In fact the international firm, which was vested with legal responsibility for the planning and supervision of all project works, did not receive a full set of the implementation designs for civil works prepared by the Tunisian firm nor the latter's notes on calculations of quantities of materials required. When these problems became apparent, it was decided that the OC would coordinate the work of the two consulting firms. The Bank felt that a continuation of the consulting arrangement with coordination by the OC was far from ideal but could be tried on an interim basis. OC management was, however, unable to settle the problems between the engineering firms. In February 1983, the OC decided to award the entire engineering assignment to the sister firms of the original partners. However, due to refusal of those firms to include adequate numbers of staff in the proposed contract and to an increase in the contract amount at the last minute, the OC decided to sever all relations with the firms and, on the basis:of a short list provided by the Bank, selected a joint venture of new international/Tunisian partners. It was only then, when the OC entered into this new contract, that silo bidding could start and the project could begin to be implemented.  11. The effectiveness of OC's earlier consultants had been impaired as a result of the problems described above (lack of communication and sharing of information between the partners) and the designs which they prepared had to be modified to improve certain technical deficiencies (in the case of the Manouba silo the execution plans had to be completely redone and sent back to France). Negotiations of appropriate amendments took place after signature of the contract with the civil works contractor in July 1984. 12. Project Works. Project works were awarded to an international firm following international competitive bidding. The firm was slow in preparing detailed construction plans and this delayed the start of the works. A large part of the delays were made up, however, during construction. In the case of the port silos of Bizerte and Gabes, the contractor did not adequately study local conditions. Draft plans did not take into account the peculiarities of certain sites (e.g., at Bizerte, the depth of foundations was underestimated) or reflect sufficient coordination with government agencies involved such as OPNT, SNCFT and the Tunisian Electricity and Gas Society (STEG). In addition, local sub-contractors in some cases proved to be overcommitted and in other cases did not provide fully reliable equipment. Delays were longest for the rehabilitation of Kalaa Seghira and Bizerte port silos. For Kalaa Seghira, this was due to the fact that supplementary works for rehabilitation were needed in addition to expansion of the silo capacity. Bizerte port silo works were affected by a dispute between OPNT and the contractor concerning delayed payments; additional dredging also became necessary and this led to an amendment of the works contract in July 1986. A reallocation was approved by the Bank in October 1986 to cover the additional costs incurred (about $590,000). Despite these delays, the infrastructure foreseen at appraisal was all completed: the Ghannouch port silo was constructed, the Bizerte port silo was expanded, three storage silos (Sfax, Beja and Gafsa) were constructed, two were rehabilitated (Manouba and Kalaa Seghira) and one was expanded (Kalaa Seghira). The 50 bulk grain hopper cars were purchased and the silo rail sidings were constructed to permit rapid loading and unloading of programmed trains. 13. Institutionai Component. Problems of management and organization in many areas of the OC's operations were identified during preparation of the Project. To correct these deficiencies, new organizational units were to be created in the OC to fill voids in planning and evaluation, maintenance and transport. Second, the project provided for two technical assistance studies, one in accounting and finance, the other in the management of OC's operations. Third, the project included a study for the preparation of a second Project which would focus on the rehabilitation and conversion to bulk handling of the distribution circuit of locally produced grain and on the integration of this circuit with that for imported grain. The various organizational units were eventually created within the OC and the studies were completed during the course of the Project and became useful guides for the reform program. Although a number of the reforms recommended by the studies were started under the Project, progress was slow and was hampered by weak support for the policy  - xi and institutional aspects of the Project within the OC, staffing problems and the generally poor working conditions at the OC. This is described below (paras. 14-18). The Second Agricultural Sector Adjustment Loan is following up on some of the recommendations of the management study (para. 17) and good progress has been made recently with maintenance (para. 19). 14. Creation of organizational units in the OC proved to be difficult mainly because of the lack of technical personnel at the OC with the result that new senior staff (Directors of PMU, transport, etc.) had to be seconded from other Ministries (the private sector was not interested because of the low salaries). There was also no incentive for incumbent managers to stay at the OC because they were overworked, had to undertake multiple tasks and were provided with poor administrative support and inadequate office facilities. A project management unit (PMU) had been established by the OC in September 1981. However, a permanent PMU Director was not assigned to the project until March 1983. Although recruited late, he performed well in the job (sea para. 8). His unit produced 11 reports on project progress. The Director of Planning and Evaluation came on duty as scheduled in March 1982. As mentioned in the PCR, he took charge in the Fall of 1984 of the division of computerization and management information and the division of statistics which gave him the tools to create in the OC an organized planning and monitoring system. The Director of Transport and Logistics was recruited on loan from the SCNFT only after prolonged intervention by the Bank. He was well qualified, efficient and dynamic but was not provided with necessary staff and had to work out of his old office at SCNFT. A silo maintenance unit was established in late November 1982 (on schedule) and a division chief of the unit was appointed about three months later. However, he was not relieved of his previous post in charge of non-Bank financed new construction. Despite the Bank's recommendations, little action was taken to recruit new staff or to improve working relations and conditions. This had adverse consequences particularly for the implementation of the maintenance reforms, a problem which was eventually rectified (in 1989, a decision was taken not to create a separate unit for the provision of maintenance services but to hire the OC's silo contractor for ad-hoc assistance. See para. 19). 15. Studies. Consultrat recruitment for the technical assistance component was planned for April 30, 1982 but was delayed by the late recruitment of directors of PMU and Plan and Evaluation. A contract was signed on November 11, 1982 with a partnership of French/Tunisian firms for execution of the financial and management studies and for the study to prepare a second phase project. For all three studies, the degree of complexity and the time required to carry them out appears to have been underestimated at appraisal. However, all of the studies were eventually completed and provided very useful recommendations and guidance to the OC.  - xii - 16. Audit and Rehabilitation of Financial and Accounting Systems of the OC. In 1984, a Tunisian auditing firm examined OC's accounts and concluded that (a) they were unauditable and that no credibility could be given to OC financial statements; (b) generally accepted accounting principles were not followed and a new accounting system needed to be designed and set up; (c) the accounting, financial management, procurement and administrative functions of OC should be reorganized; and (d) the internal control function should be strengthened at the OC. After an unsuccessful attempt to reorganize the OC's accounts in 1984 (financial study), it was decided to abandon the proposed approach of bringing accounts up-to-date chronologically, and to hire an accounting firm to restructure the OC's accounting and financial systems. As a result, in March 1987 the OC signed a turnkey contract with a Tunisian accounting firm of international caliber to carry out this task in cooperation with OC staff. One of the objectives was to make end-1987 OC accounts auditable. This task was carried out on schedule. The first audit report by an independent auditor was due June 30, 1988 but was not received by the Bank. Nonetheless, satisfactory audits of the project accounts for 1982-87 were prepared and received by the Bank. 17. Management study. In January 1984, a draft report (finalized November 1984) on the management of OC's operations presented a number of useful reform proposals concerning maintenance and the management of grain imports and stocks. However, for 18 months following the completion of the report, no significant action was taken on its recommendations. The marketing director (appointed part-time in the summer of 1985) prepared an overall reorganizational plan for the OC. The plan was a sound one and was approved by the OC's Board of Directors in the Fall of 1985 as foreseen. Recommendations were made to (a) improve the management and contracting of grain imports; (b) simplify the system of quality bonuses and penalties in grain pricing; (c) improve the Government's system for financing costs of grain storage services; (d) rationalize the formula for equalization reimbursements (transport); and (e) provide for the joint operations of collection centers by the OC and the cooperatives with the better existing facility in any given locality being rehabilitated and the poorer ones abardoned. This would tend to shift some operational management away from the OC. The objective would be for the OC to progressively withdraw from the operation of silos which would be turned over to the cooperatives or private sector and to let the OC assume more of a policy role in the cereals sector. A number of these objectives and recommendations (e.g. in (c), (d) and (e) above) are being pursued under the Second Agricultural Sector Adjustment Loan (ASAL-II) (see para. 20). In addition to restructuring, the plan included a number of urgently needed improvements in management practices designedi to increase delegation of authority, provide for better financial controls, improve bilateral communication and provide for full participation of senior staff in policy advice to each other and to OC management. Little action was taken to introduce the improved management techniques included in the plan. A November 1985 supervision report noted that senior managers at the OC did not communicate adequately with each other, and that the OC's relations with other agencies, the grain cooperatives, consulting firms and ministries (Plan, Agriculture and National Economy) were not good.  - xiii - 18. Economic Impact. The project generally met its objectives. As noted in para. 8, the infrastructure program (representing more than 80 percent of total project costs) was carried out as foreseen and the silos constructed or rehabilitated are performing well. The institutional component was less successful (para. 13). The reestimated economic rate of return (ERR) of the whole project is satisfactory at about 12.87 percent compared with the 16.1 percent calculated in the SAR. The rate of return is most sensitive to levels of imports which, as noted above, were at a record high level in 1988, the date of completion of the project. As a result, the return on investments calculated by the PCR is probably slightly higher than it would be in a year of lower imports. Moreover, the PCR calculations include among the benefits "elimination of palliatt ve investments" that the Government would have had to make if the project had not taken place. If the savings from not making these investments is to be added as a benefit in the economic analysis, then the benefits that those investments would have generated would have to be subtracted. Proper treatment of this issue would reduce the ERR marginally. However, these reductions would be outweighed by the fact that, when the PCR was prepared, some of the silos had just become operational and with minor adjustments can raise their performance levels (e.g., the PCR mentions that the rate of operation of the Gabes silo was slow in 1988 due to the late completion of the rail siding serving the silo). In short, even with the adjustments to the economic analysis discussed above, the ERR would likely remain satisfactorv. 19. Sustainabilitv. The SAR had recommended the establishment of an integrated central maintenance service. The OC and the two cooperatives decided, however, that the creation of a separate unit for maintenance was costly and unnecessary, especially since the silo contractor, who had been responsible for civil works under the project, had trained technicians of the OC and the two coopeiatives in the operation and maintenance of the silos. Those technicians were then operating and maintaining the silos. Additional assistance was sought from the silo contractor on an ad-hoc basis as needed. In 1989, a renewable contract with the firm was signed by the OC. This will help ensure that project investments are sustained and utilized efficiently. CONCLUSION 20. Follow-un nroiect and recent nrogress on reforms (nara. 17). In 1988, the Bank decided not to finance a second phase grain storage project because it was felt that, in the future, emphasis should be given to building up the private, rather than public sector grain storage and handling. This should be financed through the credit system. Through the Fourth Agricultural Credit Project, the National Agricultural Bank:(BNA) is currently providing credit to farmers for the construction of on-farm storage facilities. BNA wi'll need to mobilize new resources to expand this type of investment in the future. A Second Agricultural Sector Adjustment Loan (ASAL-II) was approved  - xiv - in 1989 and aims to focus the OC on its primary role of regulator of the cereals subsector and stimulate the private and cooperative sectors to play a larger role in domestic cereal collection and feedgrain imports. This process is underway. The OC is starting to withdraw from cereal collection and storage activities and imports of animal feedgrain are being opened up to the private sector. The OC is also disengaging from activities and investments not directly related to the cereals subsector. The medium-term objective for the cereals subsector supported by the ASAL-II is to create a more liberalized system with the free market playing the predominant role and the Government providing a regulatory and sector monitoring function. Further efforts to promote private sector participation in the cereals subsector are planned under a proposed Agricultural Sector Investment Loan (ASIL). 21. Lessons Learned. A number of lessons can be drawn from the experience with this project. (a) Institutional objectives of a project can be difficult to achieve if the project institutions operate in a regulated policy environment. At an early stage of project preparation, the Bank recognized the need to strengthen the institutional capacity of the OC and this prompted the inclusion in the project of technical assistance, training and policy studies. This objective was difficult to achieve due to the complex policy environment in which the OC functioned (fixed prices, subsidies) and the fact that, while created in the form of a public enterprise, the OC also served as an implementing agent of Government policy. As discussed in para. 20, the progressive withdrawal of the O from grain storage and collection activities and the strengthening of its policy role will improve the efficiency of the cereals subsector. (b) In a complex, multi-sectoral project requiring the participation of various agencies and ministries, attention should be paid to ensuring the effective coordination of the different actors. The project dealt with many issues in different sectors (grain storage, marketing, transport, financial). As a result, the project was complex and required different types of skills and expertise from both the Tunisians and the Bank. For that reason, more attention should have been paid to ensuring that a qualified director was appointed to the PMU at project start-up and that the Project Coordination Committee (established before Loan signature) met regularly and;that it was effective in resolving inter-agency conflicts and fostering communication and team work. (c) Complex projects require intensive supervision at start-up and during implementation. For reasons explained in (b) above, the Bank should have supervised the project more intensively particularly when it became apparent that the  - xv - procurement arrangements which had been agreed to at negotiations had changed, and that OC management was having difficulties making a firm commitment on awarding of the contract for silo supervision. It would also have been useful if senior managers of the Bank had been more involved in later stages of project implementation to help deal with institutional issues (studies, financial sector reforms). (d) Consultants' qualifications and terms of reference should be checked carefully to ensure that they have the required experience and sufficient time to carry out their work, and that they are familiar with the local context. (e) Continued attention needs to be paid to operation and maintenance (O&M) which are crucial to ensure the sustainability and effecitive utilization of project investments.  CNEA PROJECT COMPLETION REPORT REPUBLIC OF TUNISIA GRAIN DISTRIBUTION AND STORAGE PROJECT (LOAN 2052-TUN) O. ASC DATA A. INIRODUCTIO 1.01 The Grain Distribution and Storage Project (Loan No. 2052-TUN, in an amount of US$42 million for a period of five years) was approved in )ctober 1981. The project was identified in March 1979 during an identificatio,n mission for an agricultural credit project ln Tunisia. Its objectives were to improve collection, handling and storage of imported grain, as well as transportation within the country. 1.02 This document is the final version of the Project Completion Report, and was prepared by a multidisciplinary CNEA team, using the following sources (i) documentation furnished by PD (the Project Directorate); (Li) contact with the staff of the OC, particularly in the Directorate of Planning and Appraisal, the Directorate of Supply, and the Directorate of Logistics; (iii) data from the Technical Directorate of OPNT; (iv) data from the Directorate of Studies and the Financial Directorate of SNCFT; (v) a CNEA survey of port and storage silos, carried out in order to gather information on flows of imported grain; (vi) comments by the OC on the provisional report. B. AGRICULTURE IN TUNISIA: GENERAL DATA Ponulation 1.03 The general population and housing census of March 30, 1984 showed the population of Tunisia to be 6,975,450. From 1975 to 1984, the annual rate of increase was about 2.5%, this being higher than the rate between 1966 and 1975 (2.3%). The main demographic features are as follows: (i) a masculinity ratio of over 100 (malea constitute 50.9% of the population, in spite of the considerable number working abroad); (Li) the population is young; in 1984, 40% of the population was under 15 years of age, and those aged 65 or over constituted only 4.3% of the total; the median age was only 19.5, and the average age was 24.5; (iLL) 53% of the population is urban; (iv) demographic growth varies from one region to another (Tunis region ' 3.2%; South - 3.1%; Center/West - 2.7%; Northwest = only 1.4%).  The estimated population in 1986 was 7.4 million, and by 1991 it will be about 8.372 million, assuming an annual growth rate of 2.5% from 1986 to 1991. Expenditure 1.04 The 1985 household budget and consumption survey showed annual expenditure per person for the country as a whole to be D 471, i.e. an annual average per household of D 2,665. In general, expenditure in urban areas is much larger than in rural areas (D 619 per person per year, compared with D 294). This gap is becoming much more marked as the urbanization rate increases. 1.05 In the structure of total annual expenditure per person, food stands out as the largest item (D 184, or 39%). Grain accounts for 15.8% of expenditure on food, an annual D 28.9 per person, or 6.1% of the total budget. GDP and Agriculture 1.06 From 1982 to 1987, GDP at factor cost increased by an annual average of 4.06% (at constant 1980 prices) to D 3,959 million. In 1987, the share of agriculture and fishing in GDP was only about 17%, the remaining 83% being divided between industry (33.5%) and services (49.5%). Over the same period, the annual rate of increase for the primary sector was 7.6% at constant prices. Tabie 2 in Annex 1 shows trends in GDP from 1982 to 1987. 1.07 Outstanding debt in 1987 totaled D 4.47 billion, equivalent to 56% of GDP at market prices. Debt servicing was D 860 million, and the debt- service ratio was 26.8. 1.08 The external trade balance is perpetually in deficit. In 1987 it was D 738.4 million, and in 1984 it had risen slightly above D 1 billion. The import coverage ratio was 70.6%. This is the highest level recorded since 1975. 1.09 In 1987, the trade balance for agricultural products and food showed a deficit of D 21 million, and a coverage ratio of 91.4%. The same level of performance can certainly not be expected in 1988, because of the drought that affected the 1987/88 crop year, and which is still continuing. In fact, the falloff in agricultural output in 1988 meant that a considerable volume of imports was necessary in order to meet the needs of human consumption. Projections for 1988 show a defipit of about D 210 million and a coverage ratio of about 54%. Grain Production 1.10 Grain continues to be a major item in the country's economic policy, both as regards production/imports and consumption.  -3- 1.11 Since 1980, the average area devoted to grain production has been about 1.5 million ha, including about 0.8 million ha in the north. Over the same period, output has varied considerably from year to year. Record levels of about 21 and 19 million guintals were achieved in the 1984/85 and 1986/87 crop years (which were mateorologically favorable), whereas output was only 2.9 million quintals in the 1987/88 crop year, when there was a drought. Table 3 in Annex 1 shows areas, outputs and yields since the 1961/62 crop year. 1.12 In spite of achieving these record levels, grain output has not kept pace with consumption, which has increased more rapidly, mainly because of the following three factors: population growth, increases in income and the continuation of subsidies on grain derivatives. This has led to constantly increasing imports (2.1 million tons in 1988). Table 4 in Annex 1 shows imports since 1979. 1.13 In addition, studies for thiB project show the following projections for grain, regardless of type: PROJECTED SUPPLY-AND-DEMAND BALANCE: GRAIN (in thousands of tons) 1986 1991 1996 DEMAND Projected Actual Projected Projected Human 1,607 1,635 1,828 2,042 Animal 584 584 710 804 Tota1 2,191 2.219 2.538 2,846 Domestic SuP12v 965 956 1,040 1,200 Im_prta 1.226 1.263 ?,498 _1646 C. THE CEREALS OFFICE AND THE COOPERATIVES The Obiectiyes of the Cereals Office 1.14 The OC was established by Decree-Law No. 62-10 of April 3, 1962, modified and supplemented by Decree-Law No. 70-6 of September 26, 1970 (ratified by Law No. 62-18 of May 24, 1962 and Law No. 70-47 of November 20, 1970 and Law No. 86-67 of July 16, 1986, respectively). The OC is a public aector industrial and commercial enterprise, incorporated and endowed with  -4- financial autonomy. It is under the authority of the Ministry of Agriculture. 1.15 In accordance with the lawo establishing it, and subsequent legislation, the OC has the following responsibilities: - organizing, supervising and improving grain production; - establishing measures for balancing resources and needs; - organizing and monitoring marketing and the grain market; - engaging in all necessary buying and selling transactions in order to supplement supply or sell surpluses; - organizing the production and distribution of animal feed and, if necessary, other agricultural items; - on behalf of the Government, performing all operations related to agricultural products, in compliance with the relevant special Decrees. 1.16 The OC holds a monopoly on imports and exports of grains and their derivatives. In addition, it may issue general or restricted powers to special representatives for the purpose of performing certain commercial activities. The Orqanization of the OC 1.17 The OC is controlled by a board of directors composed of 13 members, including representatives of the Ministries of Planning and Finance, Agriculture, and Industry and Commerce, the Prime Minister, farmers, the milling industry, semolina producers and the two cooperatives. The board of directors is headed by a president-director-general who is responsible for day-to-day management. The president-director-general is assisted by a deputy director general. 1.18 The OC consists of the following eight directorates, in addition to the Directorate General and the Secretariat General: - the Directorate of Planning and Appraisal; - the Directorate of Administrative ,Services; - the Directorate of Financial Services; - the Directorate of Supply; - the Directorate of Logistics;  -5 - the Directorate of Works and Maintenance; - the Directorate of Food Industries; - the Directorate for the Improvement of Production. The Directorate of Supply is mainly responsible for planning and implementing measures for the purchase and uale of domestic and imported grain. The Directorate of Logistics is responsible for organizing, implementing and supervising the storage and distribution of goods marketed by the OC. It ensures optimal utilization of storage capacity, vehicles and railway stock. Among the responsibilities of the Directorate of Works and Maintenance is the management of the OC assets, mainly consisting of grain silos, collection centers, warehouses and handling equipment. TM 0C Activites 1.19 In the performance of its duties, the OC collects, stores and distributes grain, using silos that it either owns or rents for the crop year. It import. grain to make up for shortfalls in domestic output, and exports grain when there is a surplus. Since the OC's operations usually ahow a deficit, because of the government's policy of subsidies, it receives compensatory payments from the Calsse G6n6rale de Compensatlon (the Price Equalization Fund), and these enable it to balance its accounts. At the same time, the OC acts as the government's agent and office of payment for other collection and storage agencies such as COCEBLE and CCGC, which collect grain on its behalf. The OC also makes proposals to the government for producer and consumer price levels and thus also establishes the amounts per ton of subsidies at the level of distribution (i.e. milling) and breadmaking. It also proposes levels for storage charges. Thus, the OC performs a duel function that makes it a "judge in its own case" and puts it in a strong position vie-A- vie other agencies responsible for grain collection. The Coonerative 1.20 The OC delegates powers to two cooperatives, COCESLE and CCGC, to participate in the collection and handling of grain. They are service cooperatives whose purpose, according to their statutes, is the storage, preservation, processing and marketing of their members' agricultural products, and the purchase, supply and joint manufacture of all products necessary for agriculture. 1.21 The cooperatives are under the authority of the Ministry of Agriculture, and are controlled by boards comprising from three to twelve members elected by the General Assembly of their members. The board appoints a Director of the cooperative, who must not be a member of the latter, to manage its affairs and be responsible to the board.  - 6 - D. * STOiRAGE IRSTRUCTURE 1.22 In fulfilling its tasko to supply the country and satisfy an increasing grain demand, the OC possesses a atorage and handl'ng infrastructure with a 633,100-ton capacity (including the silos to be completed under this project). The breakdown (in tons) by storage agency and region is as followas AGENCY REGION SILOS WAREHOUSE FACILITIES TOTAL OFFICE OF CEREALS NORTH 180,000 86,100 266,100 CENTER AND SOUTH 123,700 20,200 143,900 TOTAL 303,700 106,300 410,000 COCEBLE NORTH 96,000 6,100 101,100 CENTER AND SOUTH - _ _ TOTAL 96,000 5,100 101,100 CCOC NORTH 111,200 10,800 122,000 CENTER AND SOUTH _- - TOTAL 111,200 10,800 122,000 GRAND TOTAL TUNISIA 510,900 122,200 633,100 Seventy-seven percent of total storage capacity (489,200 tons) is in the north of the country, which is the main grain producing area (0.9 million ha). 1I. IDENTIFICATION. PREPARATION AND APPRAISAL A. ORIGIN 2.01 In 1976, PAO issued a report entitled "A Policy and a Plan of Action for Reinforcing National Food Security in Tunisia." Subsequently, poor harvests, coupled with rapidly rising demand for industrially milled grain far in excess of normal domestic production, led tb increased grain imports and strains on the system of grain importing, storage and distribution. The government then requested Bank financing for a project for improving the collection, handling and storage of imported grain, together with the systems for its transportation. 2.02 The project was identified by the following two FAO/World Bank missionas  (i) An initial mission at the end of 1978 decided that grain handling, collection and storage facilities ahould be urgently rehabilitated, and that, in particular, silos costing US$20 million (for which the OC had already undertaken studies) should be constructed. (ii) During the March 1979 mission for identifying an agricultural credit project, it was decided that the grain storage project should be submitted separately to the World Bank. A project identification report (FAO/World Bank Report No. 27/9 TUN 11 of June 15, 1979) was drawn up and then discussed in June with the government. An additional preparation work plan was submitted in an aide-mnemoire in July 1979. B. PREPARATION AND APPRAISAL 2.03 The project was prepared by an FAO/World Bank mission that visited Tunisia from October 1 to October 20, 1979 in order to review project identification. In January 1980, a preparation report was drawn up (No. 6/80 TUN 14 of January 18, 1980). After submission and discussion with the government, this report served as the basis for the government's application to the World Bank for project financing. The studies that the OC had previously subcontracted to ATI (the engineering consultants) for the construction of the silos included in the project were annexed to the preparation report. 2.04 Total project cost over three years (including physical contingencies) was D 22.4 million (about US$56.1 million), of which 67% was in foreign exchange. The project consisted of the following: (i) conversion from bag to bulk handling of about 300 combine harvesters, and purchase of the necessary trailers; (ii) equipping of three collection centers for the mechanical handling of bags and of 20 collection centers for bulk reception and dispatch, and construction of four new collection centers; (iii) construction of four silos with a total capacity of 60,000 tons, and rehabilitation and expanbion to 30,000 tons of the port silo at Bizerte; (iv) purchase of 245 hopper cars; (v) studies and technical assistance. 2.05 The preparation report left some issues to be resolved before the loan was negotiated. A World Bank appraisal mission visited Tunisia in February 1981 and reviewed the project, and its conclusions formed the basis for an appraisal repo_t produced in September 1981 (Report No. 3503-TUN).  - 8 - 2.06 During negotiations, it became clear that the import circuit, which is larger and developing more rapidly (Table 4 of Annex 1), presented more urgent problems because of the extreme congestion in the ports (resulting in almost continuous occupation of the quay by grain ships), long waiting periods for vessels in the harbouL, high demurrage (which must be paid in foreign exchange), high fees for the quay storage of bags of grain, and the costs of labor and bags, which have to be filled, weighed and handled. To solve these problems, the government and the FAO/IBRD mission decided to implement the program in tranches. The first of these was to consist of the construction of 100,000 tons of additional storage capacity for imported grain, and, once the necessary studies had been made, the second was to deal with the collection circuit. C. DESCRIPTION 2.07 The project was designed to: (i) increase grain storage capacities; (ii) reduce the congestion, handling costs, demurrage charges and grain losses incurred at the five main ports because of insufficient bulk handling and throughput capacity; (iii) reduce the costs of transporting and handling imported grain resulting from insufficient bulk rail transport, poor transport programming, and outdated and insufficient downstream silo storage and throughput capacity; (iv) strengthen the technical capacity and financial management of the OC; (v) prepare a second project designed to modernize the system of collection, storage and transport of local grain. 2.08 According to the appraisal, the components of the project were to be: (a) Port Silos: Rehabi'itation, adaptation for imports and expansion to 30,000 tons of the existing port silo at pizerte, implementation of related port works for the handling of imports, and construction of a 30,000-ton port silo at the new port of Ghannouch on the outskirts of the city of Gabes.  -9- (b) Construction of Storage Silos: Construction of four storage silos, two of lO,000 tons each at Baja in the north and Gafea in the south, one of 20,000 tons at Sfax, and one an expansion to 28,000 tons of an existing silo of 8,000 tons at Kalaa Seghira near Sousse. (c) Rehabilitation of Storage Silos: Rehabilitation of a 50,000-ton storage silo at Manouba (Tunis). (d) Port Works: Construction of three dolphins for the mooring of ships at the Bizerte silo, and limited quayside dredging to accommodate ships of up to 30,000 tons. (e) Rehabilitation and Construction of Rail Sidina3: Construction and/or rehabilitation under the project of silo rail sidings, with a view to accommodating programmed trains with a minimun of maneuvering. (f) Acguisition of Hoper Cars: Acquisition of 50 hopper care to be used in three programmed trains for transporting imported grain from Bizerte and Gabes/Ghannouch to the storage silos and downstream milling facilities. (g) Technical-Assistance and Training: Technical assistance for planning grain imports and transportation in Tunisia, silo management and maintenance, stock and cost control, and preparation of a second project. 2.09. Estimated project cost over five years, including physical and price contingencies, was D 37.5 million (about US$93.9 million), of which D 16.8 million (about US$42.1 million; i.e. 45%) was in foreign exchange. The OC, which is under the authority of the Ministry of Agriculture, was to be responsible for project execution, assisted by OPNT and SNCFT. The estimated economic rate of return was 16%. However, thip rate is particularly sensitive to grain import projections. 2.10 The OC was to be responsible for silo construction and rbhabilitation, preparation of the study of its own financial system, and preparation of a grain storage project constituting the second stage. OPNT was to enter into agreements with the OC, making the former responsible for the construction and maintenance of the dolphins at the Bizerte silo. The  - 10 - Ministry of Infrastructure (Ministbre de 1'Equipement) was to be responsible for construction of the quay at the port silo of Gabes/Ghannouch. As a result of agreements to be made with the OC, SNCFT was to be responsible for constructing the silo rail sidings and acquiring, maintaining and operating 50 bulk rail cars that would be the property of the OC. III. POJE CT EXECUTION A. PROJECT START-UP 3.01 Project start-up was slow, mainly on account of the difficulties with the consultants ATI and OTE'-PINGAT. The loan became effective as of 1982, but construction did not start until around the third quarter of 1984. B. AMENDMENTS DURING IMPLEMENTATION 3.02 The design of the project as prepared by the appraisal mission was not changed. However, the Loan Agreement was amended on two occasions during execution: in August 1985 and December 1988. The first amendment concerned the following points: - raising of the proportion of expenditures covered by the World Bank for silo construction and rail sidings to 60%; - execution of additional works, including in particular expansion of the 8,000-ton Beja silo; - cancellation of US$4 million of the loan, making the new loan amount US$38 million. The details of additional works contracted out in accordance with Article 34 of the Administrative and Financial Conditions (Cahiers des Clauses Administratives et Financibres) are given in Table 1 of Annex 2. The second amendment concerned cancellation of US$4.6 million of the loan amount, making the final loan amount US$33.4 million. C. IMPLEMENTATION SCHEDULE 1. THE SEVEN SILOS Award of Contracts for the Silos 3.03 According to the SAR, bids for construction of the silos were to be invited in November 1981, evaluation of the bids was to be completed by the end of January 1982 and the notification to proceed was to be given around the  - 11 - end of October 1982. The SAR implementation schedule is given in Table 2 of Annex 2. 3.04 In fact, the call for bids was not issued until 19 months after the date set in the SAR (in June 1983). There were several reasons for this delay, the main ones advanced by the Project Coordination Committee (PCC) being:1 - the study for the Gabes silo, which required extra time for the soil study needed by the execution study, was not delivered to the OC by OTEI until June 15, 1982; - ATI's withdrawal from supervision of the Gabes silo works in November 1982 and the OC's decision, following inconclusive negotiations, to terminate the OTEI-PINGAT group's contract as of February 1983; - for the hiring of a group of consultant firms to evaluate the bids and supervise the construction of the silos, the OC asked for quotations in February 1983 following which a contract was signed in May 1983 with the SCET-SERETE group, whose bid was adjudged the most advantageous. 3.05 The review and evaluation of the bids, scheduled to take nine months according to the SAR, was performed by the SCET-SERETE group within the time set. The contract for construction of the silos was concluded with BUHLER-MIAG on July 11, 1984, said firms' bid being adjudged the most attractive. The notice to proceed2 notifying the contractor to start work was sent to BUHLER-MIAG on July 12, 1984 and became effective August 1, 1984, and a new schedule specifying, among other things, setting up of the site as of August 31, 1984, and provisional acceptance of the silos on December 31, 19863 was drawn up. The detailed planning of the seven silos program is given in Table 3 of Annex 2. See minutes of the third, fourth, fifth and sixth meetings of the PCC. 2 Minutes of ninth meeting of the PCC. 3 Completion date of the Ghannouch silo, construction of which was to take the most time.  - 12 - Execution of Works 3.06 Port Silos The intention was to rehabilitate, expand and develop the Bizerte port silo to receive imported products, raising its capacity to 30,000 tong, and to build a 30,000-ton port silo in the new port of Ghannouch at Gabes. The new cells were to be of reinforced concrete, vertical and cylindrical in fcrm, and built using slipform construction. The Bizerte cells were to be integrated into the existing ones thanks to a new handling, cleaning, weighing and dust-control system. The pneumatic equipment for unloading the ships and the conveyors for carrying the grain to the silos was to have a capacity of 300 tons/hour. At Ghannouch, the conveyor system was to be located in an underground gallery. The SAR also called for simultaneous hopper-car and truck loading capacities of 200 tons/hour at Bizerte and of 200 tons/hour for hopper cars and 100 tons/hour for trucks at Ghannouch; a reserve reception capacity of 100 tons/hour for intermittent rail and truck deliveries at Bizerte; and a supplementary 20 tons/hour bagging capacity for urgent shipments. According to the SAR, execution was to start in November 1982 for provisional acceptance in April 1985 and final acceptance in April 1986 (Gabes/Ghannouch silos, for which the most time was required). 3.07 In fact, execution of the Bizerte silo started three months behind the contract date of July 1984, while work on the Gabes/Ghannouch silo started seven months behind schedule. Quarterly Reports 1 and 2 list a number of reasons for these delays, mainly: - failure by contractor to properly organize the start of the work, with execution plans that did not take the existing situation into account; - contractor's unfamiliarity with the local context. 3.08 It proved impossible for these delays to be made up during execution of the works, which were completed 15 months and 20 days behind schedule for the Bizerte silo and 6 months and 27 days late for the Gabes/Ghannouch silo. The delays in relation to the original schedule for the different silos are given in Annex 2, Table 4. Storage Silos 3.09 In addition to the Bizerte and Gabes/Ghannouch port silos, four storage silos were to be built: two of 10,000 tons each (one at Beja and one at Gafsa), one 20,000-ton silo at Sfax, and expansion of an existing silo at Kalaa Seghira from 8,000 tons to 28,000 tons. The SAR also called for  13 - rehabilitation of a 50,000-ton silo at Manouba (Tunis). According to the SAR, work was also to be started in November 1982 for provisional acceptance in December 1985 and final acceptance in December 1986. The contract schedule drawn up in July 1984 was based on work being started in October 1984, with provisional acceptance in July 1986 and final acceptance in July 1987. 3.10 In reality, work started one month behind schedule for Beja, Sfax and oafsa; 1 month and 24 days behind schedule for Manouba; and 4 months behind schedule for expansion of the Kalaa Seghira silo. As regards the main contract, the works were completed with delays ranging from 4 months and 22 days for the Beja silo to 10 months and 12 days for the Sfax silo. The supplementary works for Beja (capacity increase from 8,000 tons) and Kalaa Seghira (renovation) were completed with delays of 10 months and 6 days for the former and 21 months and 15 days for the latter. The starting and completion dates for the work on the various storage silos were as follows: STORAGE SILOS: CONTRACTUAL AND ACTUAL DATES START OF WORKS COMPLETION OF WORKS ACTUAL SILOS PROVISIONAL Contractual ActuaL Contractual ActuaL ACCEPTANCE Beja silo 8/1/84 11/7/84 3/15/86 8/6/86 10/6/86 Expansion of 8eja silo / ... ... 3/15/86 1/21/87 2/3/87 Kalaa Seghiro silo 8/1/84 3/1/85 7/15/86 4/10/87 5/21/87 Renovation of K.S. silo 1/ .. ... 7/15/86 4/30/88 6/2/88 Sfax silo 8/1/84 1/1/85 7/15/86 5/27/87 6/22/87 Gafsa silo 8/1/84 3/1/85 7/15/86 2/15/87 3/24/87 Manouba sflo 8/1/84 11/24/84 11/15/85 8/6/86 10/6/86 1/ Supplementary works 3.11 The contract for construction and equipment of the silos was awarded on a global lump-sum basis not subject to revision or adjustment, under the conditions specified in the Administrative and Financial Conditions. The following table presents the base contract amounts, the costs incurred and the differences between the estimated and actual figures (for details, please see Tables 5 and 6 in Annex 2).  - 14 - SEVEN SILOS POGR_AM - ESTIMATED AND ACITUL COSTS Category DT Portion FF Portion DM Portion 1. Be Colnrtact