Mauritania - Guelbs Iron Ore and SNIM Rehabilitation Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 9487 PROJECT COMPLETION REPORT MAURITANIA GUELBS IRON ORE PROJECT (LOAN 1747-MAU) AND SNIM REHABILITATION PROJECT (LOAN 2643-MAU) APRIL 19, 1991 Industry and Energy Operations Division Sahel Department Africa Regional Office This document has , restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENT 1979 1985 June-1990 Current Unit Ouguiya (UM) US$1.00 UM 45.85 UM 77.07 UM 81.61 UM 1 US$0.022 US$0.013 US$0.012 WEIGHTS AND MEASURES EQUIVALENTS 1 kilogram (kg) 2.205 pounds (lb) 1 metric ton (mt) = 2,205 pounds (lb) 1 meter (m) 3.281 feet (ft) 1 cubic meter (m3) = 33. 315 cubic feet (ft3) 1 kilometer (km) 0.A2 mile (mi) 1 squaie kilometer (km2) = 0.386 square mile (sq mi) ABBREVIATIONS AND ACRONYMS ADB African Development Bank AFESD Arab Fund for Economic and Social Development AISI = American Iron and Steel Institute A.RMICO Arab Mining Company BCM = Banque Centrale de Mauritanie BRGM = Bureau des Recherches G6ologiques et Minieres BRPM Bureau de Recherches et de Participations Minieres CCCE = Caisse Centrale de Coop6ration Economique COMINOR = Complexe Minier du Nord CVRD = Companhia Vale de Rio Doce DCPP Departement de Commercialisation des Produit Petroliers EEC = European Economic Community EIB = European Investment Bank FCB = Five Cail Babcock FED = Fonds Europ6en de Developpement IISI International Iron and Steel Institute IRSID Institut de Recherche de la Siderurgie KFAED e Kuwait Fund for Arab Economic Development KFTCIC = Kuwait Foreign Trading, Constructing and Investment Co. MIFERMA - Mines de Fer de Mauritanie MSI = Mineral Services Incorporated OECF = Overseas Economic Cooperation Fund (Japan) OPEC = Organization of Oil Producing Exporting Countries SAFA - Soci6te Arabe du Fer et de l'Acier SNIM Societe Nationale Industrielle et Miniere SNIMEX SNIM Explosif SOCOMINE = Societe de Cooperation Miniere et Industrielle SOFREMINES = Soci6te Francaise d'Etudes Minieres SOFRESID - Societe Francaise d'Etudes de la Sid6rurgie SOMINA = Societ6 des Mines de Mauritanie SONADER = Societe Nationale pour le Developpement Rural SGTE - Societe Generale de Techniques et d'Etudes FISCAL YEAR January 1 - December 31  7* WOIIID #A04K.. ... . FOR OFFICIAL USE ONLY wao"qfa O.OC. 20433 U.S A Otfice ot Dinctov-0ael April 19, 1991 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Mauritania Guelbs Iron Ore Project (Loan 1747-MAU) and SNIM Rehabilitation Project (Loan 261,3-MA= Attached, for information, is a copy of a report entitled "Project Completion Report on Mauritania - Guelbs Iron Ore Project (Loan 1747-MAU) and SNTM Rehabilitation Project (Loan 2643 -MAU)," prepared by the Industry and Energy Operations Division of the Sahel Department in Africa Regional Office. No audit of this project has been made by the Operation. Evaluation Department at this time. Attachment Thui iaonden t ba G Isawed disfbutio and may b tie by nac"s only in ao t hegN _ Of *0ei O0ca dumL is eso.ts may not othwu be dsimd wout WrM Dauk a.  FOR OMCIAL USR ONLY PRO-JECT COK4PLETIOM REPORT MAUR1ITIA GUELBS IRON ORE PROJECT (Loan 1147-MAUM and SNIK REHABILITATIO PJECT, (LoaU 2643-MAU) Table of Contents Pate K2. Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i Evaluation Summary . . . . . . . . . . . . . . . . . . . . . . . . ii PART I: PROJECT REVIEW FROM BANK'S PERS?ICTIVE 1. Project Identity . . . . . . . . . . . . . . . . . . . . . . . 2. Background . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3. Project Objectives and Description . . . . . . . . . . . . . . 2 4. Prcject Design and Organization . . . . . . . . . . . . . . . . 2 5. Project Implementation . . . . . . . . . . . . . . . . . . . . 4 6. Project Results . . . . . . . . . . . . . . . . . . . . . . . . 6 7. Project Sustainability . . . . . . . . . . . . . . . . . . . . 8 8. Bank Performance . . . . . . . . . . . . . . . . . . . . . . . 9 9. Borrower Performance . . . . . . . . . . . . . . . . . . . . . 10 10. Project Relationship . . . . . . . . . . . . . . . . . . . . . 10 11. Consulting Services . . . . . . . . . . . . . . . . . . . . . . 11 12. Project Documentation and Data . . . . . . . . . . . . . . . . 11 PART II: PROJECT REVIEW FROH BOR&ERI S PERSEECTIVE A. Loan Conditions . . . . . . . . . . . . . . . . . . . . . . . . 12 B. Project Implementation . . . . . . . . . . . . . . . . . . . . 13 C. Start-Up Problems . . . . . . . . . . . . . . . . . . . . . . . 14 D. SNIM's Scorecard . . . . . . . . . . . . . . . . . . . . . . . 16 E. Conclusion ............. .... .... .... . 17 PART III: STATISTICAL INFOMATIO Table 1: Related Bank Loans . . . . . . . . . . . . . . . . . . . . 19 Table 2: Project Timetable . . . . . . . . . . . . . . . . . . . . . 20 Table 3: Loan Disbursements . . . . . . . . . . . . . . . . . . . . 21 Table 4: Project Implementation . . . . . . . . . . . . . . . . . . 22 Table 5: Project Costs and Financing . . . . . . . . . . . . . . . . 23 Table 6: Project Results . . . . . I . . . . . . . . . . . . . . . . 24 Table 7: Status of Covenants . . . . . . . . . . . . . . . I . . . . 26 Table 8: Use of BankResources . ................. . 27 This document has a restricted distribution and may be used by recipients only in the perfom anct | of their official duties. Its contents may not otherwise be disclosed without Wofld Bank authorization.|  PROJECT COMPLETION REPORT MAURITANIA GUELBS IRON ORE PROJECT (Ln. 1747-MAU) SNIM REHABILITATION PROJECT (Ln. 2643-MAU) Preface This is the Project Completion Report (PCR) for the Guelbs Iron Ore Project and the subsequent SNIM Rehabilitation Project in the Islamic Republic of Mauritania. While the Guelbs Iron Ore Project created a large new mining operation, the SNIM Rehabilitation Pr ject aimed at restoring financial and economic viability of the Borrower in a changed epr~ironment and thus protecting the large investment made. Loan 1747-MAU (US$60.0 million for the Guelbs Iron Ore Project) was approved on July 12, 1979, and Loan 2643-MAU (US$20.0 million for the SNIM Rehabilitation Project) on December 17, 1985. isoth loans are closed and fully disbursed. Although technical problems of the Guelbs beneficiation plant are not yet resolved and plant modifications are under way, it has been attempted to draw conclusions from the operational results obtained so far. The PCR was prepared by the Industry and Energy Division of the Sahel Department in the Africa Regional Office (Preface, Evaluation Summary, Parts I and III). On May 8, 1990, the Bank sent the Borrower Parts I and III with the request to prepare Part II and submit their comments on our Parts I and III by June 1st., but no comments were received. Co-financiers' comments have been considered for preparation of Part I. Preparation of the PCR was started during the Bank's appraisal mission for the Public Enterprise Sector Structural Adjustment Program in January 1990, and is based, inter alia, on the staff appraisal reports, the Borrower's records on implementation and costs, legal documents, supervision reports, correspondence between the Bank and the Borrower, and other internal Bank documentation.  - Li - PROJECT COMPLETION REPORT MAURITANIA GUELBS IRON ORE PROJECT (Ln. 1747-MAU) SNIM REHABILITATION PROJECT (Ln. 2643-MAU) Evaluation Summary Objectives The main objectives of the Guelbo Iron Ore Project were to allow Mauritania's iron ore mining company Societe Nationale Industrielle et Miniere (SNIM) to switch exploitation from its depleting high grade iron ore reserves to large low grade reserves and thus to allow continuation of the Company's operations and their dominant contribution to the national economy. The main objectives of the subsequent SNIM Rehabilitation Project were to achieve cost reductions which, after substantial change in external factors and lower than expected results of the Guelbo project, had become necessary to restore the Company's financial and economic viability and to protect the large investment made in the Guelbs project. Implewntation Experience Construction was mainly completed in October 1984, 20 months behind appraisal estimate. Installed cost, at US$ 372.6 million, were 15Z lower than appraisal estimate. Start-up of production has been unusually long and difficult. Presently, the Project's output has reached only 402 of design capacity. The subsequent SNIM Rehabilitation Project, appraised in :,85, was completed December 1988, within budget. Results The project objectlves have only partly been achieved. Whlle construction of the Guelbs Project, although late, was completed within budget, the intended full switching from exploitation of high grade ore to low grade ore could not be made, due to protracted start-up difficulties. However, due to discovery of additional high grade iron ore reserves, the transfer to exploitation of low grade ore lost its previous urgency. Low iron ore prices and severe technical problems reduced the financial rate of return of the Project from 5Z, as estimated at appraisal, to -102, as presently re- estimated. The negative rate of return and the high debt burden engaged with the Project brought the Company into severe financial difficulties. Although the subsequent SNIM Rehabilitation Project achieved the intended cost reductions, these were insufficient to restore the financial viability of the company due to unexpected further decline of iron ore prices and changes in exchange rates. Today, the Company is still in an unsustainable financial situation.  - iii - To respond to its difficult financial situation, the Company achieved remarkable improvements in management, productior. costs and operations. This, together with the fortunate discovery of additional high grade ore, turned the Compa:ty's outlook positive, despite the poor Guelbs investment. Although the Project's role has been reduced and an acceptabls rate of return on this investment cannot be expected in the near future, there are substantial benefits if the investment is regarded a sunk cost: the additional production gained from the Project will give SNIM more flexibility and security in a highly competitive market and considerably improve the Company's chances to remain over the next 15 to 20 years one of the world's large and viable iron ore producers. Sustainability Over the last five years, SNIM demonstrated impressive sustained performance, which led to a record production in 1989, despite the handicaDs of the Guelbs Project. The Company formulated an appropriate action plan to bring the Project to an acceptable rate of production, and is now well on track with its implementation. The cash break-even point of the Project's operation has been reached and there is a high probability that production will be substantially increased and maintained well above the present level. Findings and Le@Qons Learned Despite careful and professional project preparation and execution, assisted by a committee of renown experts and technical assistance from qualified firas and individuals, the Project suffered from serious technical flaws and overestimation of physical results of the Guelbs plant's operation. This is one of the inherent risks of mining projects throughout the world, in particular if a metallurgical process and plant with prototype features is involved, as it was the case here. In retrospect, with the experience gained, one certainly would have made a different investment decision and designed the project differently, but this was not known at appraisal. However, the Project also gives an example that management determination, personnel motivation and government cooperation can bring a company out of crisis despite adverse external factors and the burden of an ill-conceived major investment.  PROJECT COHPLETION REPORT GUELBS 1RON ORE PROJECT (LD. 14-HALMi SNIM REHABILITATION PROJECT (Ln. 2643-MAIl PART I: PROJECT REVIEH FROM BANXS ESPCTV 1. Project Identity Name .Guelbs Iron Ore Project Loan Number 1747-MAUI & 2643-MAU RVP Unit Africa Region Country Mauritania Sector Mining 2. Dackground 2.01 Iron ore mining is a mainstay of Mauritania's economy. Exploitation of iron ore dates back to 1963, when Mines de Fer de Mauritanie (MIFERMA), with French, British, German and Italian interests, commenced operations. The Bank participated in the financing of MIFERMA with a loan of US$ 66.0 million2. The Company was nationalized in 1974 and renamed to Societe Nationale Industrielle et Miniere (SNIM). Production capacity has increased from 6 million tons per year, initially, to 12 million tons per year, at present. The Company is the country's largest single foreign exchange earner, and technically and financially has been a success over most of its history. In the second half of the 1970s, there was a strong belief that the Company's high grade iron ore reserves would soon be depleted and operations would 'iave to stop if abundant low grade ore reserves could not be developed. The Guelbs Iron Ore Project, already conceived by MIFERHA engineers earlier, was prepared at this time with high priority3. The Bank participated in financing the Project with a loan of US$ 60.0 million (Loan 1747-MAU, of December 14, 1979), together with 11 multilateral, bilateral and commercial co-lenders4. Due to lower than anticipated iron ore prices and unusual technizal difficulties of the Project, the Company had to engage into considerable cost cutting measures after construction of the Guelbs Project in order to meet its steeply increased debt service obligations. To facilitate restoration of financial viability, the SNIM Rehabilitation Project was conceived in 1984. The Bank participated in financing with an additional loan of US$ 20.0 million (Loan 2643-MAU, of 2 Loan 249-PR, of March 17, 1960. The loan has been fully repaid. 3 The Project is nomad after the Gualbs El Rhuin low grade iLon ore deposit, located near the Company's high grade iron ore op.rations. 4 Saudi Fund for Development, Kuwait Fund for Arab Economic Development, Arab Fund for Economic and Social Development, Abu Dhabt Fund, Overseas Economic Cooperattcn Fund, Japan, African Development Bank, OPEC Fund for International Development, Caisse Centrale de Cooper^tion Economiqum, Banque de Paris et des Pays-Das, Banque Francalse du Commerce Exteriour, European Investmmnt Bank.  -2- January 27, 1986). Based on this loan, several co-lenders and the European Economic Community provided additional financing, under concessionary terms, of a cotal amount of about US$ 70 million for replacement and productivity- increasing investments5. 3. trojeot Objectives and Description 3.01 Project Objectives. The main objectives of the Guelbs Iron Ore Project were to allow SNIH to switch exploitation from its depleting high grade iron ore reserves to the large low grade Guelbs reserves and thus to continue the Company's dominant contribution to the national economy. The main objectives of the subsequent _.JIH Rehabilitation Project were to achieve cost reductions v'nich, after substantial decline of iron ore prices and lower than expected results of the Guelbs project, had become necessary to restore the Company's financial and economic viability aad to protect the large investment made in the Guelbs project. 3.02 Project Desgription. The Guelbs Project consists of an open-pit mine ivith a capacity of 14 million tons per year of low grade ore, a beneficiation plant based on dry grinding and magnetic separation to produce 6.2 million tons per year of saleable high grade concentrate derived from the low grade ore, and related infrastructure consisting of a 60 MW power plant and extension of railway, port and housing facilities. This project was to be the first phase of a final development of twice the capacity'. 3.03 The SNIM Rehabilitation Project consists of managerial improvements, operating cost reductions, and the external financing of a variety of urgent minor replacement and rationalization investments. 4. Project Deign and organization7 4.01 Feasibility Study. As early as the late 60s, MIFERMA had started a oroadly based exploration campaign in the area of its mining activities to look for an eventual replacement of its depleting iron ore reserves. In the mid 70s, work concentrated on the low grade Guelbs iron ore deposit, which was believed to be the only available sufficiently large deposit with suitable characteristics for economic ore beneficiation. Various alternatives for ore beneficiation, including wet and dry gravimetric and magnetic separation were considered. Due to the region's scarcity of water, a fully dry prototype beneficiation plant, based on dry grinding and magnetic separation was K5aiLt Fund for Arab Economic Development, Arab Fund for Econoie and Social Deelopmnt, Afriean Deelopent Fund, Ceisse Central. de Cooperation Economique, European Investment Bank and European Oevelopment Vund/Sysmin. Agre%d, in prineLple, at Co-lenders Mbeting of June 4, 1957. Applications include equipment for railway, mining, workshop and telecommunications. 6 At appraisal, lt was estimated that Phase 1 and 2 would be in full operation by 1990. 7The following Secttons 4 and 5 concentrate on the Guslbs Project only, ln particular on the Guelbs beneficiition plant. Information relating to implementation and costs of the SNIM Rehabilitation Project can be found Ln Tables 4 and 5 of Part II.  - 3 - retained. The work was carried out by SOCOMINE, France, (the Engineers), a niewly formed engineering company with ex-MIFERMA managers as key personnel. The study was carried out in great detail and in a professional way. 4.02 Due to the unusual difficulties in achieving rated capacity of the plant and the changed iron ore prices, SNIM started, after the Project's construction, an extensive and thorough re-evaluation of the region's remaining high grade ore occurrences. As a result, the Mhaoudat deposit was discovered in 19878. In addition, reserves in the depleting Kedia mines proved to be at least 40 million tons larger than anticipated at apprairal. In retrospect, the Project would not have been necessary at the chosen time and a more thorough evaluation of even minor indications of additional high grade ore reserves should have been contucted. Also, after the poor start-up experience (para. 5.05-5.07) it is now evident that more consideration should have been given to alternative process and plant designs with a view to eliminate or reduce the risks associated with the untried dry process and protoype plant, although market considerations favored the retained solution at the time9. 4.03 Pllot Plant Tests. In view of the prototype nature of the plant, extensive pilot plant testing was carried out as pait of project preparation and to supply a sound basis for design of the commercial plant retained in the feasibility study. The pilot plant was of large size and treated an unusually large amount of ore10. Nevertheless, the results of the commercial plant appear to indicate that the assumed scaling-up factor and the hardness of the ore bcve been underestimated. In retrospect, a more conservative approach in using pilot plant data would have been appropiate. 4.04 Engineering. The relatively little experience available with the selected process caused the Bank to request the creation of a technical committee, consisting of prominent iron ore beneficiation experts, to assist the design team in reducing technical risks. Although all major design aspects of the plant were discussed with and approved by the committee, there have been, in retrospect, major errors in designing the plant's process, in particular with relation to the big dry grinding mills. The selected closed air circuit proved to be detrimental with regard to dust generation and plant performance. Modification is required. 4.05 Project Organization. The size and corplexity of the Project caused the Bank to request the assignment of a Director of Operations, who would be responsible, on SNIM's part, for overall project coordination, while direct supervision of procurement and construction was contracted to the Engineers. a SNDM found this deposit by following up on minor surface indications of high grade ore, which at zhe time of the Project's appraisal were known but interpreted as too small for cossrcial eaplottation. Today, as result of a detailod exploration campaign, carried out by SHIN with its own means, the Company has proven so million tons of ore reserves in this deposit, vhich cau be exploited without benefielation. 9The selected process is well adapted to minter foed production, whereas other processes would produce a higher share of pellet ieod, for which market risks are higher. 10 The pilot mill hrd a diameter o'! 2 meters and treated 15.000 tons of ore extracted from various points of the Guelbe orebody.  -4- Although there have been some difficulties in filling this post"l, SNXM demonstrated satisfactcry project managemeuit performance and was organized appropiately. The incurred delays during project construction are attributable partly to suppliers and partly to the Engineers' lack of experience in executing projects of this complexity (para. 5.03). 4.06 Trust Arrangements. The Project's unsatisfactory financial return and political events12 lead the Guelbs Lenders to design and impose the Trust Arrangements as an essential part of their lending operation for the Project. The Arrangements are a comprehensive and complicated legal framework13 by which all proceeds from iron ore sales are assigned to the Lenders and paid into an offshore account at the Banque Domiciliataire 4, a portion destined for the debt service is set-aside with first , iority to the Trustee 5, a second portion, destined for SNIM's operational foreign exchange needs, is set-aside to the Operational Account at the Banque Domiciliataire, and the remainder is transferred to the Central Bank of Mauritania for the account of SNIM. In general, the Arrangements worked well and achieved their purpose: up to now, there have been neither defaults on debt service payments nor production losses due to lack of foreign exchange. This contributed to survival of the Project and the Company under extreme financial conditions. However, the complexity of the arrangements at one time led to several lenders' failure, including the Bank's, to timely pre-announce debt service payments and thus jeopardized availability of sufficient funds in the Trust Account (para. 8.02). There have also been difficulties for the Banque Domiciliataire and SNIM to timely meet reporting requirements. For a new operation requiring trust arrangements, simplifications should be recommended. S. Projoet Implementation 5.01 Loan Effectiveness. Loan effectiveness was dependant, among others, on signing all legal documents, which form part of the Trust Arrangements, by all Lenders. These documents were signed on July 7, 1980, seven months after the Loan and Cuarantee Agreements with the Bank. The late signature caused a moderate delay of about four months in the Project's execution, since a few early equipment supply contracts could not become effective, while downpayments f'or most critical contracts were made by SNIM from own cash generated. SNIM asslgned project contrc; to a capable engineer, without promoting him to the rank of Director of Operations. Wlth saew delay, SNIM recrulted an expatriate Dlrector of Operations, who never was fully accepted by SNIM's projoct team a*d who resigned sLnce. 12 At appraisal, the Project's fizinieal rate of return, even after completion of Phase 2 and without paymnnt of taxes, vas estimbted to be only SX. The PolLsaric guerrilla war and an unstable Government preser-ted considerable polLtical risks. 13 The related maLn legal docu,mnts consist of the Security Agreemont, the Set-Aside Agreement, the Trust Deed and the lemorndum of Cooperation. 14 Societe Generale, Agence Internationale, Paris, France. 15 Lav Debenture International, London, UK.  -5 5.02 Procurement. With the Engi.neers' assistance, procurement documents were prepared properly, and the Bank's guidelines were followed for all Bank financed procurement. The first bids for the Bank-financed contract for industrial civil works were rejected because their prices were unreasonably high in the Engineers' opinion. The re-bidding resulted in a delay of about three months and in fewer bids with equal or higher prices. In retrospect, it was therefore a wrong decision to call for re-biddirg. Occasionally, also long bid evaluations bv SNIM contributed to delays. 5.03 Construction. The Zirst production line of the Guelbs plant started trial runs under load in July 1984, the second line in October 1984, -Yhereas at appraisal it was assumed that both lines would start-up in January 1983. The delav if about 20 months is partly due to late effectiveness of loans, handling if procurement16 procedures a..d occasional difficulties of suppliers in wmet'eng target dates. Mainly, however, delays were caused by the need to fill-in smaller contracts for specialized equipment at a relatively late stage and to match equipment of different suppliers. The Engineers, despite their professional competence and integrity, apparently underestimated the complexity of the task. In retrospect, more rigorous application of turn-key contracts might have resulted in less delays. 5.04 Proiect Cost. The Guelbs Project's installed cost (including mine and infrastructure) amounts to US$ 372,6 million, 15% lower than appraisal estimate. Increases in working capital and interest during construction up to the first commercial shipment of iron ore concentrates (March 1985) amount to US$ 41.8 million, 32% lower than appraisal estimate, due to a concentration of contract payments towards the later phase of project implementation. However, if all cash operating losses, plant improvements and interest paymsnts incurred from March 1985 to January 1990 are considered, a cost of US$ 100-120 million would have to be added. Further plant modifica_.ions in the order of US$ 10-15 million are still required. The Project's final cost, broadly defined, therefore is likely to be in the order of US$ 550 million. 5.05 Start-up of Production. Despite extensive and early training of plant personnel at SNIM's Nouadhibou training _enter, which was specifically created for this purpose, early involvement of SNIM engineers in design and construction supervision, adoption of a sophisticated and appropiate organizational structure, technicel assistance from a renowned French steel company and qualified other expatriate personnel, the Guelbs plant experienced an unusual level of start-up difficulties which have still not been overcome. At appraisal,it was estimated that the plant would be in full production by the end of 1983, or one year after completion of its construction. Today, more than five years after completion of construction, the plant has reached only 40% of its initially assumed capacity. 5.06 In addition to the process design flaws (para. 4.04), key equipment has shown mechanical and electrical failures in excess of normal expectations. 16 Tmre were some difficulties in particular with cofinanced lots, and at one staos a colander temporarily suspended disburs.mnts.  These are partly risks inherent to similar projects in the mining and metallurgical industry on a worldwide scale, in particular if a prototype process and plant is involved. Dry grinding mills similar to SNIM's only exist at a Canadian iron ore operation1 . At the time of appraisal, the Canadian operation was still in the start-up phase and, like SNIM, experienced problems. After about 10 years, the plant finally succeeded in yielding full production. For the Guelbs project, the original estimate of achieving full production only one year after completion of construction was a grossly over- optimistic assumption, not only in retrospect, but even at appraisal. 5.07 Major equipment, in particular for material handling, such as reclaimers, feeuers and conveyors, proved to be poorly adapted and underdesigned for the heavy duty service required. There has been a tendency of suppliers to offer cheap technical solutions in order to lower the bid price. 5.08 Based on the experience gained and with assistance of a manager/engineer, who had a key role in the Canadian project, SNIM formulated an appropiate action program which contains important plant modifications to improve the process and reinforcA equipment. The program is under execution. In October 1989, SNIM also made important managerial changes. Output of the plant is on an increasing trend. 6. Project Results 6.01 Proiect Objectives. The project objectives have only partly been achieved. While construction of the Guelbs Project, despite delays, was co-pleted within budget, the plant still suffers from protracted start-up difficulties. Due to the discovery of Mhaoudat and additional high grade ore at Kedia, the transfer to exploitation of low grade ore lost its previous urgency. Due to the much lower than estimated iron ore prices, continuation of the high grade ore operations at a high level of production is needed for the Company's survival. The second phase of the Project (para. 3.02) cannot be conceived as a viable operation under present conditions and must therefore be postponed indefinitely. 6.02 The subsequent SNIM Rehabilitationi Project and investments achieved the inten-ded cost reductions and helped to raise productivity to a level which, since only recently, allows the company meet all its cash flow needs, despite unexpected further decline of iron ore prices and changes in exchange rates. However, the results are still insufficient to fully restore the company's future financial viability. Unless appropiate restructuring of the debt incurred with the Project is completed, the company is not in a position to engage in new bortowing for its vitally important Mhaoudat project (para. 4.02). 6.03 PWhsical Results. The mine and infrastructure components of the Guelbs Project were completed without problems. The Guelbs mine today is a modern, 17 Carrol Lake of International Iron Ore Cowaeny, Canada.  - 7 - efficient and low cost operation, and SNIM has been able to achieve record performance of its railway"'. The problem lies with the Guelbs beneficiation plant. Availability of the plant has reached only 40%, or half of the appraisal estimate (80%). Although plant modifications have not yet been completed and tested, it may be expected that, as a result of the current action program and the managerial changes, 65% availability can be achieved finally. The remaining difference of 15% to appraisal estimate would be due mainly to the initial underestimatior. of dust generation and its impact. Hourly throughput of a production llne (determined by the milling rate) has reached 75% of the appraisal estimate. It may be expected that, as a result of the current action program, 80% can be achieved finally. The remaining difference of up to 20% to appraisal estimate would be due mainly to the assumption of a wrong scaling-up factor from the pilot plant tests and to an underestimation of ore hardness. Product specifications and recovery rates for magnetic ore are as expected19. Due to the reduced availability and throughput, a de-rating of the nominal plant capacity will be necessary. The best present estimate is that the plant finally will have a production capacity of about 4 million tons per year of concentrate. 6.04 With the SNIM Rehabilitation Project, the Company achieved a reduction of staff by 25% and spin-off of its previous steel miill (SAFA), hotels, restaurants, clubs and supermarkets (SOMASERT), and sale of 3 company airplanes. This, together with other cost reductions, was more than expected at appraisal. The total cost savings achieved are in the order of US$ 11.4 million per year, or 35 % higher than the appraisal estimate. Parallel with the cost reductions, SNIM improved management and overhauled and renewed important production equipment, with funds made available under the Project and the funds provided for replacement and productivity-increasing investments. This has laid the basis for further productivity increases, which in 1989 resulted in a record production of 12 million tons, 25% more than estimated at appraisal of the SNIM Rehabilitation Project. 6.05 Financial Performance20. The actual financial rate of return of the Guelbs Project has been calculated at -10%, considerably lower than the already abnormally low rate of 5% calculated at appraisal2%, due to the much lower than estimated iron ore price22 and the technlcal problems of the plant. The totally unsatisfactory return of the Project lead to non-compliance with is 12 million tons of iron ore were transported in 1989, a record in MIFERMA4 rand SND 's hIstory. 19 There Ls however a problem of achieving the estimated product qualities and quantities ehen treating oxidized ore. An appropiate research program for process modification ts under vay. 20 The following section concentrates on financial results. In Table 6, en attenyt has been made to also compare actual and appraisal estimates of economIc rates of return. 21 The economic rate of return calculated at rppraxsal vas 12X. The diffeTence is mainly due to shadow pricing of Mauritanian labour at 352 of financial cost. This was a major consideration for project approva. at the time. 22 yn real terms, fob Nouadhibou prices in the period from 1977 to 1989 dropped from 16.9 US$It to 9.3 US8lt equivalent in 1977 terms, or by 45X.  the financial covenants under the Loan Agreement23 and brought the Company to the brink of financial ruin. Only SNIM's significant productivity increases made the Company survive, and strict financial discipline allowed it to meet its obligations under the Trust Arrangements, at the cost of building-up excessive short-term debt. 6.06 The actual financial rate of return of the SNIM Rehabilitation Project has been calculated at 21%, or the same as appraisal, despite a further drop of iron ore prices to about 70% of the 1985 estimate. This success improved the joint financial return of the Guelbs Project and the SNIM Rehabilitation Project from -10% to -2%, which, however, still is a highly un.atisfactory value. In order to acbieve a satisfactory rate of return, iron ore prices would have to increase by almost 100, *r the investment would have to be devalued to 10% of its actual cost. 6.07 Develooment Imgact. Due to lcw i'-n ore prices and the development of the fishery sector, which took pla-.: an unforeseen way after appraisal of the Guelbs Project, iron ore mining :At, '.- dominant macro-economic role. Furthermore, the high capital cost involved and the increase in operating cost turned the Project into a financial burdeu for the Company, with no or little budgetary benefits for the country. 6.08 Nevertheless, iron ore mining remains a very important economic activity in the northern part of the country. About 200,000 persons, representing more than half the region's population, depend on it, directly and indirectly. Even under the present tight financial situation, SNIN still generates about US$ 70 million foreign exchange after debt service and indirectly contributes about US$ 7 million to the Government's budget in the form of workers' income tax. The Guelbs Project, now invested and regarded as sunk cost, is an important basis for SNIM's future survival. The technology transfer involved strengthened the country's industrial skills. Similarly, the managerial and productivity improvements achieved under the SNIM Rehabilitation Project are a success not only for the Company, but for the country's industrial sector. 7. Project Sustainability 7.01 SNIK learned from the experience made by starting-up the Guelbs Project, formulated an appropiate action plan to bring the plant into full production, and is now well on track with its implementation. The previously set production goal of 2.0 million tons for 1989 was met. Production is on an increasing trend, and the morale of the workforce is good. The cash break-even point of the Project's operation has been reached24. Based on a careful analysis of past operations and planned plant modifications, there is a high probability that production will be increased to and maintained at a level of 23 Debt/equity ratio not to exesed 1.5 and current ratio not to be less than 1.3. 24 Defined as the min-nn production lewel at which the revenues from the sale of the Project's product cover all operating cost (excl. interest end depreciation) of the Guelbe mine end plant.  4.0 million tons per year, which will contribute substantially to cover the Company's fixed costs of operation25. 7.02 Supported by the SNIH Rehabilitation Project and the related investments, SNIH demonstrated impressive sustained performance over the last five years, which led to the record production in 1989, despite the limitations of the Guelbs Project. 7.03 There are now sufficient proven high grade ore reserves, which together with the Guelbs Project, are a solid base for maintaining the company's present producvion level over the next 15 to 20 years. The blend of high grade and low grade ore will increase the Company's flexibility in marketing. 7.04 In total, the outlook for the Company's sustained performance is good. S. Bank Perforuance 8.01 Since preparation of the Project, the Bank has alwaya muintained a role in advising the Borrower, the Government and co-lenders. Already in 1981, the Bank started warning about the risks of substantial'future cash shortfalls and the need for appropiate action2. Since that time, the Bank continued to stress the need for change and contributed directly and indirectly to designing the SNIM Rehabilitation Project. The Bank's assistance often has been instrumental in obtaining agreement between the Company, Government and Lenders, and it contributed to the survival of Mauritania's iron ore sector. 8.02 The Bank's iron ore price forecasts used for appraisal of the Guelbs Iron Ore Project and the SNIM Rehabilitation Project turned out to be far too optimistic in both cases, although they were considered conservative at the time. At the beginning of loan repayment, the Bank also failed to observe the notification time, required under the unusual and complex Trust Arrangements, for declaring the first debt service payment consisting of principal and interest27. This caused a shortage of funds in the trust account, which was met by SNIM through direct payment. 8.03 The following lessons may be drawn for future comparable lending operations: (a) greatest attention should be given to appropiate design of financially marginal projects. Ways to defer the project or fundamentally change its concept must be fully explored as early as possible in the project cycle. Reliance on pilot plant testing and expert technical committees is no safeguard against the risks connected with prototype industrial plants; 25 Cotts of railway, port and Company's overheas, but excluding the financial chrges of the Project. 26 SupervLsion lMission Report, Kuwait Fund - World Bank Joint Xission, September X to 11, 1981. 27 Payments consisting of interest only have to be annaunced 100 days in dvances paymeunts consting of principal and interest require 300 days notice.  - 10 (b) suitable ir.volvement of an experienced operator of similar industrial plant from early design through completion of start-up is highly desirable; and (c) if unusual loan procedures, such as trust arrangements, are involved, special care should be taken that they are simple and clearly understood by all parties and persons concerned. 9. Borrower erformanoe 9.01 In general, since appraisal of the Project in 1979, the Borrower's performance has been satisfactory, with excellent improvement over the last five years. All covenants under the two Loan Agreements (loans 1747 and 2643), except those dealing with financial ratios, have been complied with. 9.02 At the early stage of project implementation, there have been some difficulties in executing timely international procurement as well as in adapting training to real needs . At times, there were also difficulties in meeting the reporting requirements under the Trust Arrangements timely and completely. 9.03 The main lessons that may be learned by industrial borrowers with comparable projects are: (a) an appropiate form of partnership should be sought with an experienced operator to assist in project design, audit of technical specifications for plant and equipment, training and start-up. This would help to safeguard against inappropiate consultant and engineering advice; and (b) turnkey contracts should be applied to the extent possible. Re- bidding should be avoided. 20. Project Relationshlp 10.01 The Nauritanian Government cooperated well. It complied with all covenants under the Guarantee Agreements. Given SNIM's dominant role in the national economy, the Government exercised unusually little direct daily control over the Company. Instead, it followed a prudent and successful policy of keeping relations with SNIN at arms length and encouraging the Company's performance. After availability of the results of the taxation study carried out under the SNIN Rehabilitation Project, the Government immediately dropped 28 At that tim, SNIM maintained a relatively large workforce, which it justified vith training needs for the Guelbs Projects, but whLch was also maintained for soclo-poitical reasons.  - 11 - all tax claims on SNIM29. 10.02 Cooperation with the co-lenders was good. There have been no major disagreements at any time. There have been several Lenders Meetings, during which agreements were reached on SNIM's and the Government's actions and on additional financing (para. 2.01). The Kuwait Fund and the Bank were assigned from the beginning Lenders Representatives30. This arrangement worked well and facilitated efficient cooperation. I1. Consulting Services 11.01 Reputation, qualification and motivation of the firms which provided consulting and engineering services to SNIM were good. Yet, the Project incurred delays (para. 5.03) and unusual start-up difficulties (para. 5.07), which are partly contributable to insufficient experience of the Engineers with similar plants. Once the problems became apparent, the Engineers tended to defend previous design decisions by cautiously proposing relatively minor modifications, rather than helping the borrower, in a totally open and constructive way, to find broader solutions for rectification of errors. 11.02 One of the most valuable and competent types of technical assistanc- SNIM received was from highly experienced and motivated individuals working on a part-time basis as needed for well specified tasks. The Project also demonstrated that the Borrower's will to succeed through own efforts and genuine wish for qualified outside support are the most important ingredients of successful technical assistance. 12. Project Documentation and Data 12.01 The Project was based on an unusual amount of legal documents31L. All these documents have been useful, although a simplification of the Trust Arrangements would have been desirable (para. 4.06). The Staff Appraisal Reports provided a useful framework, despite overoptimistic expectations, in particular with the first loan (Ln.1747). 12.02 Data collection for this PCR was greatly facilitated by the Borrower's appropiate project documentation. 29 Whereas the study rec_mended to lLlt SNIM's taxes to normal profit tax (which would be sere due to lack of profits) and to j.uport tax only on those item which src not solely Liported for SM's use, the Government was ready to rcrv-e all taxes in order to facilitate SIND's opereatons end Ln linm with the spirit of the Convention Particuliere (Footnote 31). 30 As agreed in the Memorandum of Cooperation, the Kuwait Fund was to represent all Arab lenders, and the Bank *1L other lender. 31 In addition to the Loan and Guarantee Agreemonts and the documents which form part of the trust Artangeeonts (Footr.Q2e 13), there is the Convention Particuliere. en agreement between the Government end the Borrower regulating their relationship and taxation.  - 12 - PROJECT COMPLETION REPORT MAURITANIA GUELBS IRON ORE PROJECT (Loan 1747-MAU) and SNIM REHABILITATION PROJECT (Loan 2643-MAU) PART II: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE 13. The feasibility study on the Guelbs Project was distributed to potential lenders in April 1976 before being supplemented by the TSR alternative the following July. At that time the iron ore market was buoyant and SNIM was a going concern, though high grade ore reserves were already showing signs of depletion and insecurity was brewing in the region in the run-up to the war in the Sahara. 14. Aware of the venture's importance to Mauritania, the lenders put their trust in SNIM, while protecting their positions with a number of loan guaran- tee conditions. AL Loan Conditions 15. We shall discuss below SNIM's reaction to some of these covenants, before reviewing project execution and start-up and SNIM's own scorecard. * The requirement that SNIM concentrate on the iron business: This was good for SNIM. * Appointment of a Director of Operations: This position, which was supposed to coordinate project implemen- tation, was perceived as a layer parachuted into the structure; it played an overall negativ' role in the course of the project. * Committee of experts: This committee of internationally known experts, approved by the lenders, was to supervise technical decision-making during the project. It was unfortunately unable to play this role fully and often served as a cover for the engineering firms to legitimize their choices. There was no question about the competence or integrity of the men selected, but rather: - the work method adopted - the short duration and the spacing of the meetings - the complexity of the project - the experts' -rather generalist background. which made their mission more or less ineffective. * Benign dust conditions: Notwithstanding the positive findings of all the studies and expert assessments by ILO and SNIM and its engineering units, the volume of dust inside the plant was much greater than expected, and further investments had to be mobilized to eliminate or at least alleviate it (electrostatic filters throughout the exhaust system).  - 13 - * Engineering and consultants: SNIM undertook to keep the consultants and engineers already approved by the Bank (IBRD), but, for the sake of transparency in relations between the company and its engineering consultants, it is better if contracts are drawn up freely so that neither party feels it is being imposed on the other by an outside organization. * Financial conditions: - The Security and Set-Aside Agreements for the Guelbs Project, and in particular the Trust Arrangements, were cumbersome, costly and ill suited to SNIM's financial situation. - Moreover, the lenders saw it as a safety net and later perpetuated the arrangement, which became an ordinary guarantee. The project's weighted average finance cost was thus high, at 5.52, compared with the financial rate of return of 5? estimated at appraisal. The remaining financial conditions were on the whole positive (cost reduction, curtailment of debt ratios, etc.). B. Prolect Implemenration 16. The design and engineering of the project were carried out in conformity with international standards and the lenders' recommendations: - overall design (in Mauritania and abroad) was coordinated by the SOCOMINE team. e detailed engineering was performed by engineering groups well known in Europe, such as SOFRESID, SGTE, SEBA, etc. * technical decisions were taken at regular intervals and were first screened by the Committee of Experts and submitted to the lenders' representatives during each of their missions. * consultation procedures were scrupulously observed in close cooperation with the lenders: - Notice of prequalification - Issue of invitations to bid - Public bid opening - Preparation of selection report by the project engineers - SNIM contract committee - Approval by lenders. 17. The recommendations of the engineering consultants and thus the selection report were confirmed by SNIM and the lenders on every occasion but one: Contract C081, awarded to KRUPP rather than to FCB.  - 14 - 18. Although no one can doubt KRUPP's ability to supply efficient conveyors and the selection report faulted it only for its high price (which could have been lowered), the company was completely non-responsive and did the project a great deal of hRrm. 19. Management and coordination of project start-up operations were handled directly by SNIM, which acquitted itself rather creditably. 20. The slippages that occurred in scheduling were caused partly by KRUPP's non-performance and partly by delays in payment due to cross conditionality with other loans to the government imposed by some of the lenders. C. Start-up Problems 21. SNIM was responsible for the preparation and management of project start- up, in close collaboration with its lenders and their engineers. Faced with a wide range of choices, it was decided to rely first and foremost on Mauritanian staff from its own ranks or trained in: - 'sMauritanian technical training institutions - SNIM training centers - institutions abroad and to provide them with appropriate additional training abroad (suppliers) or on-site. 22. This SNIM team was reinforced during the first few years of plant operation by a core of managers and technicians seconded by a highly efficient European company (SOLMER), supplemented by technicians seconded to equipment suppliers. 23. This novel experiment was particularly interesting and, despite the risk taken with an only partially seasoned workforce, it was on balance a success. 24. It would be impossible to blame them for the technical problems encoun- tered during start-up or the slowness with which the plant came on stream, as will be seen from the analysis of these problems below. 25. Contrary to a view often put around, the Guelbs plant is not entirely a prototype. It consists, on the one hand, of equipment well known within the industry, such as gyratory mills, conveyors, bucket wheel excavators, stackers, screens, etc. and, on the other, of: * semi-autogenous mills that also, at the time, were no longer at the supposed prototype stage since two of them had been operating at IRON ORE CANADA (IOC) for several years. * dry magnetic separators, which are less well known in the industry on the scale adopted, but of which a prototype was tested at Zouerate for more than 13,000 hours before it was finally selected.  - 1S - 26. If the project was a prototype at all, it is because it was all assembled in the middle ef the desert, far from any source of supply. 27. The variety and complexity of the problems encountered entailed a sig- nificant additional workload for the suppliers and the start-up teams. In the case of the latter, SNIM frequently had to strengthen them and change the way they were organized, in order to cope with these problems, which can now be broken down into 6 broad electromechanical groups: * undersized bucket wheel excavators for collecting crushed ore * unprotected exhaust fans * heavy dust, which was hard on both men and machinery * poorly designed KRUPP conveyor system, which was ill suited to the plant * heavy wear in the ventilation circuits and shielding * spare part requirements underestimated when contracts were signed. 28. This list is not exhaustive, but simply summarizes some of the main problems to blame for the plant's poor electromechnical service record. 29. Technical solutions to these problems were found, and financing was provided by EIB and ADB as part of the productivity project; implementation is under way. 30. It is relevant to point out in this connection that SNIM had quickly-- and while there was still time--drawn attention to the fact that: * the in-service time assumed for the project was too high (82Z vs. 781 at IOC) * after only one ore breaking cycle, the bucket wheel excavators were seen to be unsuited to ore like that from Guelbs * at the Guelbs plant, filtering was required throughout the exhaust system just as at the pilot plant. At the time, the engineering consultants and the Cowmittee of Experts rejected these reservations. 31. Process problems are of two kinds: * Inadequate plant throughput. Mill throughputs are running at 752 of the capacities predicted by the scaling-up formulas, them-selves thought too conservative at the time. SNIM is exploring all parameters that might improve this throughput. IOC took more than 10 years to reach its rated output.  - 16 - * Difficulty of processing the oxides in two magnetic separation stages, whereas, on the strength of the pilot results, these products were seen as the *real meat" of ).he project; so far, the first- stage concentrates are good, but the 2nd stage products are sold as middlings. 32. The sampling method adopted for the oxidized ores during the research phase appears to have been faulty, and further processing is needed to improve the iron content of the concentrates. The trials under way are fairly encour- aging. 33. SNIM has appealed to several well-known ore processir. authorities to offer their insights, and some of these experts are already on site, including Mr. Jardine, former General Manager of IOC. D. SNIM'S SCORECARD 34. The Guelbs project has been an extra -ply enriching experience both in the engineering field and in terms of its original financial launching, as well as in human terms, where it has helped SNIM to make real headway. 35. Engineering. This project is a technological "gold mine" because of the variety and complexity of the techniques involved. SNIM has endeavored to derive as much benefit as possible from it by acquiring still greater in-depth familiarity with: - dirferent ore crushing and beneficiation techniques - state-of-the-art electronics and robotics - mechanical assemblies from several different countries, based on a variety of technologies - problems of starting up complex industrial plants - ,.. etc. 36. All this is an unquestionable plus for SNIM and for Mauritania. If SNIM were to undertake the project again today, it would certainly not be the same, but this is easy to say after five years of problems. The main improvements would be: * The search for water for the plant. It was concluded too early that there was no water in the area, though the search had focused only on a small 50-km strip around Zouerate. The presence of water would have turned the project design on its head. * Development of a secondary milling line or the use of excavator shovels to collect the crushed ore. * Introduction of a second milling stage to permit open circuit operation.  - 17 - * Relocation of the magnetic separation equipment to eliminate waste material as early as possible in the cycle, taking advantage of advances in permanent magnet technology$ and to consider the scope for 3- or 4-stage processing of refractory ores. O 'When selecting equipment, give more weight to reliability and sturdiness than to capital cost whenever this is compatible with the project's finances. 37. Establishment of the financing plan. Twelve of the world's leading lenders participated in the financing of this project, giving us the benefit of their wealth of experience in similar projects and imposing on us the discipline of complying with each one's rules and of being in a position to provide reports at all times. SNIM and the Mauritanian Goverranent are the chief beneficiaries of this. 38. Human dimension. On this level, tL.e constant challenge posed by the world- wide iron ore crisis, coupled with the start-up difficulties at the Guelbs plant, prompted SNIM to do come self-analysis, to renew its self-confidence, and to seek new solutions in order to survive. Achievements in this area include: * devising original solutions to certain problems at Guelbs, e.g., the use of permanent magnets as lifelong protective shielding in the ventilation circuits of the mills * exploration for and discovery of more than 120 million tons of naturally high-grade ore between the old Kedia deposits and the new deposit at Mhaoudat * the continued search for high-grade ore in the area * the continuous effort to cut costs and improve productivity, which in 1989 enabled the company to beat its 1974 production record * ... etc. In this respect, Guelbs is unquestionably a project that has enabled both SNIH and Mauritania to advance. EL Conclusion 39. In conclusion, and despite the long road that still lies ahead, we feel the scoresheet on Guelbs is on balance a positive one and it is fortunate that Mhaoudat was not discovered until later. 40. We thank all lenders warmly for the valuable assistance they gave us, for their unfailing moral and material support, and, last but not least, for having treated us as an equal partner and respected us as such.  - 18 _ 41. The most eloquent testimony of this frt tful cooperation and mutual con- fidence is to be found in the commitment which the same lenders have just entered into with SNIM for the Mhaoudat project and the public enterprise sector adjustment program.  - 19 - PROJECT COMPLETION REPORT MAURITANIA GUELBS IRON ORE PROJECT (Loan 1747-MAU) and SNIM REHABILITATION PROJECT (Loan 2643-MAU) PART III: STATISTICAL INFORMATION Table 1: Related Bank Loans Ln Shi a i Det of sm_ Loan No. Proec Ttle and Purpoes Amount Todl boldA.p Juy1 1.WK0 (US$ mib) Finarnog 249-FR MIFERMA To commercially develop Mauritania's Iron 66.0 3.Stt Mc 15. 1600 ftullydibued ore resources through construction of mines, fully repid raitway, port and Infrastructure 1747-MAU THE GUELDS IRON ORE PROJECT To allow SNIM (previously MIFERMA) to continue 00.0 12.0w July12. 1979 fullydlbrd. operations after depletion of Its high grade oro 66.0 9 rapaid at reserves through construction of the first phae (6 MtVa) of a new low grade Iron ore mine and beneficlation plant 2643-MAU ONN REHABILITATION PROJECT fullydibursd, To seoure medium-term financial and economIc 20.0 21.7% 7 oeember 17. 196 . repI sf vIablity of SNIM through Improvements in management reducion f production cost ard trnsformation of part of short-term debt Into long-term debt i/In tern di number di eemtunl reamet ma _, ausah lonn Waac wsen wIIt shun_e rat fcuain  - 20 - Table 2: Project Timetable Loan 1747-MAU Loan 2643-MAU Date Planned Actual Date Date Planned Actual Date Item at Appralsal at Appraisal IdenUtfcation a/ Feb.04.1976 bl MAr. 01,1904 Preparation bl Jan.24.10?7 bl Oct. 03. 194 Appraisal Mission bl Oct. 03.19"7 bi Apr. 25. 196 Loan Negodations c/ Feb. 28.1979 a1 Oct.15. 106 Board Approval Jul. 12.1070 000.17.1965 Loan Signature Dec. 14. 1970 Jan. 27. 1o9 Loan Effectiveness Dec. 31, 160 mar. 1901. 1 ProectCornpletion Jan.01,1983 Mar.31, 1965 Jun. 30. 1O8 Jun. 80. 1O6 Loan Closing Jun. 30, 1983 Jan.15. 1965 Doo.31.1988 Dec. 31.1966 at datathu d*Wa fUeWty study bi date of mlimbn dpwturs at first day of negoaatlon  - 21 - Table 3: Loan Disbursements Loan 1747-MAU Loan 2643-MAU OlbiuSnwmn (US$ amiNn) ubursMenwts(USSMIWn) Bak Ficl Ya Eamalud ACtu Acta 4 of EdaMtod Acta Atu % of wnd Quawrw Cumuwila Cumulati Edimated Cumulai Cumul EdGmad 1960 I _ _ 11 - - III 1.6 o 0% IV 3.0 - 0% 1961 I 6.5 0% 11 8.0 - 0% III 14.0 1.1 2% IV 20.0 3.0 5% 1962 1 26.5 4.7 6% 11 33.0 10.5 18% III 40.5 17.8 30% IV 48.0 21.3 35% 1S63 I 53.0 27.3 46% II 86.0 32.8 59% III 59.0 38.3 84% IV 60.0 47.1 76% o"4 1 53.6 86% II 5S.2 90% III 60.3 96% IV 8O.4 09% 1965 I 569.0 6% 11 60.0 100% III 166 I.o - o III 2.0 - 0% IV 3.0 5.0 167% 1987 1 4.5 7.6 176% II 6.0 10.0 167% III 6.0 17.6 223% IV 10.0 19.5 106% 1966 1 12.6 20.0 100% II 15.0 17.5 IV 20.0  - 22 - Table 4: Project Implementation Appraisal Estimate Actual Months of Delay maih of n am Cmpetb Ouon am Copt ouai on COMPl _ Loan 1747-MAU af Equlpment Procurwnent bw Mlning Apr.80 Jul.82 27 May 81 Feb.83 21 13 7 .4 o Beneficiatlon Oct.79 Mar.82 29 DeO.78 May 4 63 2 20 24 dl Railroad Jul.80 Jan.82 18 Sep.80 Jan.SU 16 2 0 -2 of Power Plant Oct.79 Apr.81 18 Jan.60 Pdar.f 38 3 23 20 u ClI Works Industral Jul.80 Nov.62 28 Sep.81 Jul.63 22 14 8 04 Housing Nov.79 Dec.62 37 Nov.60 Mar.63 28 12 3 4 Erscion Oct.80 Nov.82 25 Mar.62 Mar.64 24 17 16 -1 gl TiWPAn UneA Jan.83 Oct34 Une 8 Jan.83 Jul." hi RrtE lpmwit Mar.65 Pi Conmld Dec.13 Loan 2643-MAU y Managrl Improvements Oet.84 Dec.57 36 Oct.64 Mar.67 29 0 4 4 Equipm. Peptam. &Overhaul Jan.64 Jun.U6 fl Jan."4 Dec.6 e 0 a 6 Tralning & Techn. Asssanoe Jan.64 Jul.6 64 Jan.64 Jul.68 64 0 0 0 r Proj Complion Jul.816 Dec.U at key equmnt only. fom onan soInatuto site dw*wy bi drsllms*=m,u id tuon on mi d oensu o o ontt N signed 16 mntslae (n NOv.81 d d Jul.O di loomcto *I gentsaee U frm oantm esnature to mdt cI waf g/ fet unnib of beneton pit under kod hi hpmnt fm port d fNoudhbou edbe oneats owr enemW pufpo ess new dfnltn dooftwn d. no dat wiw defnd at eppim V ur defined at ap i - date at which bnfltlon plt d we prduoud 6.1 mINan tons c a Peed of6 monts du* to reduced produton cepat (se Tle S such da mey nw rr ' MO n t reductons (vetuade & perano kW at apppOW Wm defined end tdf tme ss finecg of equipmet retment md meil ttes hav been sme d due to sow dbumnt o co-lcndm  - 23 - Table 5: Project Costs and Financing A. ProlW 0osts Guolbo Pr*o SNIM RitabMIhdo Pfolmot Mind~ ~ ~ ~~~A Apa bSleda AMId 800o Ap Eu Miwd W d0 1L Po. Tat L1 FP. TeL. E _S L. Far. TOt. Leo F. TtaL EsOrilds MinN& Useoan qulm Z? 18.? 18883 & 2 96.0 07.7 70 Por &AMd Equlpment 1.0 64.9 6L6 4.1 4d.OS68 06* Pewr Plnt A E_r. Equlpm. 0.0 47.6 4.4 0.8 34.4 84.7 72* AuW-y Equmnt 0.0 6. 6.S 2.6 08 .0 142* CON VAt 1M4 76 60.0 16O 46.$ 20Z 70* E,eotmA &nh_adn 16.2 8.1 61.8 4.4 84.6 80.1 76* Engiein A PrJect Maem 1.6 8LO0 8. 7.6 2.7 80.1 00* PrsininuyAamTrdinmq 4.0 11.0 160 42.6 27 46.8 802 R.plmetco Equpmet 2.7 278 30. ZS 20.3 20.0 04* 01A& Nqmm dflqulpm. 14.1 884 47.6 134 31.6 44.9 04* Toob A Col Equpennt 60 6.0 4.7 4.7 04* Trdinhg ATehn. AQltuc 0.2 g2 6.7 8.7 04* SubtbeId Onaldied Costo) f9 88.080.6 430.4 70.6 200.0 872.0 16.6 76.4 62 16.0 71.2 07.1 04* inemenwAl ddr*nCeplt bi 8.6 30.0 886 0.0 0.0 10.1 *- eltSu.OungCoratuoton 1 d 1.2 26. 27.0 1.8 80.s 31.7 114* TlOd 44.6 41 6007 81.6 332.4 4U4 6* B. Proect Fhwicng (in tS nmEo) KFTCK:, Kuw 40.0 OA4 b inhbg (kmpty 2.0 2a.s kfA 2&0 20.0 WU. mU 20.0 182 kismio D tkdapment Sak 10.0 0Z1 OMItohue OmitSuneral 42.7 380 8. 86.4 Subted 162.7 146.2 36.4 O8.4 L_ KuwVAnb Funds 600 8L 16.2 dl 180 of oud" Fund 6o U.0 6.7 di 6.0 NMo00.0 64.0 20.0 20.0 OCOO, Fince 60.0 4&0 3.0 di 80 E 30.0 16.2 Abu 0M Fund 20.0 1Z0 6.6 dl L8 CO F.JM 1.0, 11.7 0.Idi 8a1 ADS 12O 10.0 OPEC Fund 6.0 4.0 Sub%ed 83.O 2.1 6 60.7 Td 6my. 414.8 U 022 07.1 al inbdcId dootne. sud inet edot of Goo* Proece t per 8000MINE t d i 1fl0 ied dened ae g *e udu Proect up to Uor. 1965t pent). nhl. 26 d lI'reord Gueb pu. w*xp r 1066 ud fdld iniest"hn bm for Gul ct up o UMec 106U (bet shipmen, ihde 26 UM Interet 116 d unuzd pwrdn cmeforuef Po, told US 86. noiwn siunuld podon 9 obas at nd d 66 tod US.mll Uequvbdnt to unt aed atnppre minue US & unutWlnd ho 1068) dwe due to e*_sn rats en  - 24 - Table 6: Project Rewns Qolbowi PrdNtIMN PAhabUWtam Jakat Projats AprbI MD Mjppd% EGW_ AcM EIM_ AoW AeS A. Drect Benets GUeSs Rmt -mom .ftsos ofi 60 0 -OC lbdows Vhr 100 Soo SacewEy bh % 47% 47% COpel ci UVa 6.2 4.0 Pcs i dueodw % 16% 26% di WAFA SAFA S. Economic Impact Eoa. Ro ofuawm of I 1i2% -S% 265% 10% % 0% kesn off Pie u $ 16.0 9.3 13.2 90. GuesPMdulles OS WVi 12.4 4.0 Ohpmlduedes hi MU& - 7.0 0.8 , 1.1 LAe,haevCet U % 36% 86% 100% 36% costew s V WA 8.o 11.4 C. FlnancW Imp FILPAW of F.tm of % 6% -10% 21% 21% 1% -8% Undeulg AfsuPIs hi La Cost II % 100% 100% 100% 100% - e - W - - 1.6 - OSsF ResW Pwuesu Aumultd Cuhow .in n/ a IS0 -60 0. SUd4 Tbxfn ofGN d eni en none W apmeg - labile mi - w w mU S'*uaidfr1Mwf II1m* d 1 S "Mddead inerp euei, le USI ulboobnWoISdarOM5 sNo 5 ualegpe4 s ISP. U id .5- ai.wIke moi Obmedwsieeww wsmpaa CMs.47*meq -a wumy( eWASls' I medimdfseetveimaid)d 6651 evbs of lo-embl4 w g -b pr"SrA a _* _ 01 he es.eshdslm See leN_ibsaiWst -1 amlwdls Me*mai rI i Piem r2.maimd WI uliawiblegS eet&WMAhkl verne d WUu-w iiPIAs 1 5.5kaiMpedbnosie 2 15.1 .. Ai04 hl .to0S iM. 1wd Mock r{ Ow_ b : _ t _ _h" t mu_lo PI sgsdaeg l__ ip kweemsedpe 1 _5-al. uIge150 1051 eI husm tem psi laid 1t MIrsmimdI lbs.5s Is RuS P1 V leild"s _ : hi ISIbet Sabak l 1"ada.. P MsUe.lesUS.Mmbwele.e - rey.eeromdle "h_Im SttemI lSWr __ _-M0 Rad PIclaledw W - - M besaidbat q. d epwd O__^- hbdtnC@  - 25 - Attachment to Table 6: Re-estimation of Rate of Return Guelbs Project SNIM Rehabilitation Projoct Joint Projects Ptojeot Produ. FPwn. Opet. Net Ben*fit PRot Produ. Rsvn. Cost SeW Ope. NOt Benfi Net Befit CoNt Cost Finan. Eoon. C*Gt Labor Ote Cost Fnat. Econ. Finat. Eoon. us Mt MS Ms MS Ms Ms Mt Ms Ms us Ms MS MS Ms Ms at bi c/ d/ el Ul af Gl of hi I kW U ml n/ 1980 13.2 -13.2 -12.8 -13.2 -12.8 1961 86.4 -86.4 -82.9 46.4 -62.9 1982 1t6.6 -166.0 -161.9 -1t8.0 -161.9 193 170.6 -17o0. -166.6 -170o. -166.6 1984 102.1 -102.1 -99.0 13.2 0.6 11.2 5.27 -7.3 4.2 -109.4 -106.3 196 14.7 0.9 17.6 20.4 -17.4 -14.6 27.9 0.3 6.8 2.60 -24.9 -23.7 -42.3 -38.2 166 1.6 24.2 27.3 -3.0 0.3 24.6 0.6 9.4 0.9 1.8 6.66 -18.2 -17.4 -21.3 -17.1 1987 1.4 20.8 26.1 -6.3 -2.1 23.2 0.8 12.3 2.9 3.1 7.88 -12.7 - 3.0 -16.0 -16.1 19o6 1.9 27.2 31.9 -4.8 -0.7 11.3 1.8 27.3 7.0 4.7 18.13 9.6 7.6 6.0 6.9 19s 2.0 33.0 33.0 0.0 4.0 1.3 22.3 7.0 4.7 12.62 21.2 18.2 21.1 22.2 199 2.2 37.4 36.3 2.1 6.4 1.3 22.4 7.0 4.7 12.60 21.6 18.6 23.7 26.0 1981 3.0 49.8 44.6 6.1 10.6 1.3 21.7 7.0 4.7 12.60 21.0 17.9 26.0 28.4 1982 3.6 68.1 60.3 6.9 12.0 1.3 21.1 7.0 4.7 12.60 20.3 17.3 26.2 29.3 1998 3.6 64.4 60.3 4.2 10.3 1.3 20.6 7.0 4.7 12.60 19.7 16.7 23.9 27.0 199 3.6 62.7 60.3 2.6 8.6 1.3 19.8 7.0 4.7 12.60 19.1 16.0 21.6 24.7 1996 3.6 61.0 60.3 0.8 6.9 1.3 19.2 7.0 4.7 12.60 16.4 164 19.2 22.4 i16 4.0 68.6 68.0 Z6 9.4 1.0 16.4 7.0 4.7 10.00 17.2 13.8 19.7 23.2 1997 4.0 68.7 66.0 2.7 9.6 1.0 16.6 7.0 4.7' 10.00 17.2 13.9 20.0 23.6 1996 4.0 59.0 68.0 3.0 9.8 1.0 16.6 7.0 4.7 16.00 17.3 13.9 20.2 20.7 1999 4.0 69.2 66.0 3.2 10.0 1.0 16.6 7.0 4.7 10.00 17.3 14.0 20.6 24.0 20O 4.0 69.4 68.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 2001 4.0 69.4 68.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 2002 4.0 69.4 68.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 2003 4.0 69.4 680 3.4 10.2 1.0 166 7.0 4.7 10.00 17.4 14.0 20.7 24.3 204 4.0 69.4 66.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 2006 4.0 69.4 66.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 2006 4.0 59.4 66.0 3 4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 2007 4.0 69.4 68.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 200 4.0 69.4 6.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 t.'.4 14.0 20.7 24.3 200 4.0 69.4 68.0 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 2010 4.0 69.4 680 3.4 10.2 1.0 16.6 7.0 4.7 10.00 17.4 14.0 20.7 24.3 Pe al Rr -10%J- 21% 19% -2% 0o SI astralproject ast incos tec 1900 prices. 9 up to Mates6 tIbt e*fe bf schre prodution Of GUiehe Pit uptoi9.meatr re pstdLvbc tlobvW-WtOt wc.peca ef podcnk d t Commodh lybionJusWAuY 19 Ge Inh utN 190 proe dlq.* Mcotiub Pt, tbusdontOMWeficedoo ct St tvarMi Sue 5 o st sa.S cgIct r Ms port `IW Gera tw d *I ren io promet I og ior opertin cost in Coneteh 190 prce Uttr oonIpsl.n wdat ppruetinre ~pottreducedby3%dto Nd.t xee*en orost35% i pt Ostlet-da pd36% W a hpr o. atprrebI estiret. rto I thera1her 6qtwee r-eI odpcitor hety (I1 MUlt end M si e IVeldpsrmel cast rtdutb>in. hcroate 19il60 p bric V _d r in cat d sp pte vi. inh ont 9 1990 pt ywinc,meresl ostiG csttO810it lat krmeineitproduction ktVewe fro betiAt protion pk cast se i proljec bIt cost minus cxrico cost inrau t productin pinconhtri O96 prIe Vprct steoreduted by duetol 1IrCet sbttcst etadw pid %tn wS rh osta it s rtP d net Iwbict beneis ous r k nort isotl enffs Of SNIM elutrO Utn Pr Iotn co a 1I90 prie rd not sonomic beowti of Sueft P pkstp net wnow.e bnel otf ONIM Rhi%tiOn ProCt, in canteu logo prb  - 26 - Table 7: Status of Covenants Deormpto of C4onant Aarom. COmnpfl- Remarks Loan 1747-MAU Borrower to oonthnuo to vwe tofvIO of conoutanto whooo quaniriatiorn and conrdlom LA 3 02 (a) yea of enwymnes are su ospIS to the Bait Borrowr to estgblih raert commitle ot compolmon stabblactoy to the Bait LA 3 02 (b) yeo Sorows to carry out. by Decowrbt 31. 1902. a etuo on bng-twm ettta of deust tieiotion r LA4 03(b) yee tuy,bVyr sreutrtcirtlombnldoTravll. Genera. conekuded that heath by workers and to anmrmw report to the Bank and to toaeD aqnat to protit workle ae sisk tt eccept". Borower i, mronitorhig 2 trice erw yesa health of n'wrey Gut staff.and is laipism tshm etlulotory mreasr tb/ rduce dust orrows to reponnaitly to the Benk on water roervas and LA 4 04 yea W WisK isport ep wa n durng awry etpeision mlisn. etrunhie prospe tigtote proeIeotIgn yiWd good' aut rrower to ensure that adsqtatec (kntonn has ben obl A)sd belore unisrtaklng tlciW st LA 4 06 yen rsqust o Qt3uator *mis are not decrMl, baimlthttn Borrows' opeations (Law 78 104 o AprIlS. 1I70). Annual repors for such stitike end their flnenm to be sent to te Banit Borower to oonpt bY Dosbsr S3, (e3. program tor ost rduotore and to revisw annratyl lA 407 with Batnk reatie of slp taken tortrower not to ast moe thrn US5 2.0 mtion in any subtiy or parttcalon LA 4 10 yes B0rrowerat sto ssl iron or blow worl markt primc or at nono rtatts aine"y LA 4.12 yet this cleis weeit kidd becus of posatbl paeatiu prcs at (hir of appran"a Borrowr to ensure tat otwoted dbt sha nis "caued 1.5 tImes owontd ctal and LA 504 in eould not be mwt due to dsterlotnthg kron era orim Borrower to raiatio of curar aStt to curtr ntleisti oti te than 1.3 LA5.05 ro cod iat be mt due to daeroth Wo ore prose Borrowe is to pay any dbland; wnt_ cotrnentd financisi ratis are met, ai tdqsta LA5 OS s) yea sal-l'nfl tar the Sonu d Phas has been temuc ted Borow nrt to undrtk M kwaet other than the Prosct eeering US$ 5.0 bteon per LA 6.07 ye Yewr Each Lrt to otty the Ba*r Dometrltaire end Borroe of lh atmounte of sansieus BA 3.02 (bc) severa Landwe (Icling Bani) krd not rioly tsy befor im de0tt svce pym ndt ththan 100 edy hm edno tof o,aa tonlstiniontareft repyment ots, and 1Wt lar tn seo ye ftor pelymrs eonmsi of _ter and _rme Brwe to eVAN to Cerl Bank of atauni for pparoa arm to copy to aschi Lender a * A 3.0t (b) nitialy Bt. reived opies. Ceo"re) Batk or Grarwt nvr cr o te toagii eactanos r _rret or operations, noti b thtn tOdys befo nteferd with Borrowr's toreign eioltng raqiheimmat te ad of Ny itisl year Boroer to itnsh to Lridr' Reprerntatfres (hBin t on of dt two Rep rtel(() arnd BA 4.02 Baitnk reeked topime of mary contrias. payment re tas by Sanqus the B5W Ooim_aa opise oft iss corc l ith rnt , baye h t rlh rrowre resie Binus Domisarnic to prrds Leder' Relprsentat)ese with tals_nt of ad paym e SA 4.04 r ed p rtly with conot la deb" rMed. 11t of at kon ose m aW P Af at am Mm" ty rwre. not MNat tha 16 dey d oftch ^ii LO 264-MAU (eniy c _e _ isth(roer those afredy liited wder iLoan 1747) Borrow to bthe to t Bat - B weeks alte each qttr. E of t otrrbmtnr dtt y LAS O S LA r 03(a) roorts Pe"rot oelo4 Borrower sey pretred them and it an epIF tI t. wlt reviy er _wesi coknti to b go Borower to etrhl to th Bak. ri tO I s d Star ech worh rronthly pogn raporte on LA 3 00 b) ye ro t_lo la Borrowe t caSr ptt meseo dsorbed utdti Sedler 5 beforeisee of the SoonmoTrnche LA3H04 ye part oonftMbna were mnodifedby Bait Borrwer to Ms'I tirt int I yer after Clostng Date r atio of cutnt *nta_ to ourrnr LA 5.0S no 1bIse of is oss than 1.3 Guwaatr to onlset by December 31. 19t6. a study on min taxation with twi o1 ttrefre GA 301 yea agoWd by the Bank GA GumrorAeemert LA Loan Apoemet SA S-AbOd Aobcament  - 27 - Table 8: Use of Bank Resources A. Staff Input Loan 1747-#AU Loan 2643-MAU 8tae Of Prolect Ctych Actual SW Actual SW Through Departure of Approasal MIsion 100 87 Dsp. Appraisa; Mission through Board Approval 216 28 Bord Approval through Effoctlvanses 39 10 Supervhsbon go 61 Total 464 174 B. Mlssions Montl No. of Days, In Perform Typs of YOw Persona Fited SPac ton Atin ProbFeM at b/ d Loan 1747-4AU d/ Preparation et oim 7 la B,C,E.F.T Appraisl 107 3 17 E,F,T Post-appraal t/ 08/78 3 a E,F.T Supervbibon 1 10/79 3 6 T SuprWIon 2 0480 4 6 B,F.T 2 F,M,T Superviston 3 1180 2 6 F,T 2 F,M Supervision 4 g/ 0981 3 7 E,F,T 2 F.M.P SupervIsion 6 0482 3 8 8,F,T 3 F,M,P Supervision 8 10182 3 8 8,F,T 3 F,M,P Supervision 7 04/83 3 6 8.F.T 3 F,U,P Supervision 8 12/83 3 e 8,F.T 3 F,M.P Loan 2843-AAU hi identification U 03/84 2 14 C,F.T Prepwstlon 10184 3 8 8,F.T Apprais"a 06186 2 7 F,T Poet-appraisl OSv86 2 6 F.T Supervision I 06/88 I 6 T Supervision 2 10186 1 4 T I Supervision3 12/U6 1 4 T 2 F Supervsin 4 087 2 a C,T 2 F.T Supervision 6 06/88 2 10 C.T 2 F,T aDS 8anr meswmpe, C Co.wrard. E E Frn t FirFbt Arca. T Teha hi I probem-fr". 2 modeate pobme. S mor prbeme Fr Fbuncst U Maruge P Polictl (voaretafing end tlaoton. T Te Tlal /Cure prhi a6*ton to theated miea, there raa been 10 mlonaoawcietct hco-rahitodtrust wranrga.ns f thte was no Mwtoettn mieton. teat etudy as reet tOr toet ro r,ka o tw et to 6ank in Februwy 1978 V mas* to r-do 1e M of cratet-dst end catlbn after Qon t ding 9f 15n msio'bn wSh KuwsM Fwd Wv h 1 1dn to the td mslb. ther he b 4 mibclae ho ctnwt h eo-f hnin V mitosn arroied ct marngot arid tdtohntat BudS oSNIM wth Itnc ttr do oonr b