FILE COPY FOR OFFICIAL USE ONLY. Public Disclosure Authorized. Report No BR Appraisal of the Valefertil Phosphate Fertilizer Project Brazil

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Report No BR Appraisal of the Valefertil Phosphate Fertilizer Project Brazil April 7, 1977 Industrial Projects Department FOR OFFICIAL USE ONLY Document of the World Bank FILE COPY This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.

2 CURRENCY EQUIVALENTS Except where otherwise indicated2, all figures are quoted in Brazilian Cruzeiros (Cr$) and US$ (Exchange Rate as of October 1, 1976): WEIGHTS AND MEASURES All weights and measures are in metric units except as noted: Cr$1.0 = US$ Metric Ton (t) = 1,000 Kilograms (kg) Cr$11.40 = US$ Metric Ton (t) = 2,204 Pounds Cr$1,000,000 = US$87,700 1 Kilometer (km) = 0.62 Miles 1 Hectare (ha) 3 = 2.47 Acres (Exchange rate as of March 9, Cubic Meter (NM )= 6.19 Barrels is CR$12.98 = US$1.00) PRINCIPAL ABBREVIATIONS AND ACRONYMS USED ANDA ARAFERTIL BNDE CVRD DAP FEPASA FINAME FLUOR HPY K K 0 misp N NATRON NFP NPK P P 05 S{P TPD TPY TSP ULTRAFERTIL VALEFERTIL VALEP Associacao Nacional Para Diffusao Adubos Araxa Fertilizantes S.A. Banco Nacional do Desenvolvimento Economico Companhia Vale do Rio Doce Diammonium Phosphate FEPASA Ferrovia Paulista S.A. Agencia Especial de Financiamento Industrial-FINAME Fluor Latin America (U.S.A.) Hours Per Year Potassium Potassium Oxide equivalent Content in Fertilizers Monoammonium Phosphate Nitrogen NATRON Consultaria e Projetos S.A. National Fertilizer Program Nitrogen (N), Phosphate (P 205) and Potassium (K 20) Nutrients Phosphorus Phosphorus Pentoxide equivalent Content in Fertilizers Single Superphosphate Tons (Metric) Per Day Tons (Metric) Per Year Triple Superphosphate Industria e Comercio de Fertilizantes S.A. Fertilizantes Vale do Rio Grande S.A.- VALEFERTIL Mineracao Vale do Paranaiba S.A. Brazilian Fiscal Year January 1 - December 31

3 FOR OFFICIAL USE ONLY BRAZIL APPRAISAL OF THE VALEFERTIL PHOSPHATE FERTILIZER PROJECT TABLE OF CONTENTS Page No. SUMMARY AND CONCLUSIONS... i-iv I. INTRODUCTION... 1 A. Background... 1 B. Bank Group Involvement in the Brazilian Fertilizer Sector... I II. PROJECT SPONSOR... 2 A. Fertilizantes Vale do Rio Grande S.A. - VALEFERTIL (VALEFERTIL)... 2 B. Companhia Vale do Rio Doce (CVRD)... 3 III. FERTILIZER MARKET AND MARKETING... 4 A. Agriculture in Brazil... 4 B. The Fertilizer Industry and Government Fertilizer Policies... 4 C. Historical Consumption of Fertilizers D. Projected Demand and Supply of Fertilizers E. Fertilizer Marketing and Distribution F. The Market and Marketing of VALEFERTIL's Production... 9 IV. THE PROJECT A. Project Scope and Location B. Technology C. Phosphate Rock Supply and the VALEP Project D. Project Management and Implementation E. Employment and Training F. Pollution Control V. CAPITAL COST, FINANCING PLAN AND PROCUREMENT A. Capital Cost B. Financing Plan C. Procurement D. Allocation and Disbursement of Bank Loan This report was prepared by Messrs. Denis T. Carpio and S. Venkataraman of the Industrial Projects Department. This document hu a restricte distribution and may be used by recipients only in the performance of their ofcil duties. Its contents may not otherwise be disclosed without World Bank authorization.

4 -2- TABLE OF CONTENTS (Cont'd) Page No. VI. FINANCIAL ANALYSIS A. Operating Costs and Financial Forecasts B. Financial Rate of Return C. Financial Covenants D. Major Risks VII. ECONOMIC ANALYSIS A. World Fertilizer Supply and Economic Prices... o B. Economic Rate of Return C. Economic Analysis of the Combined VALEFERTIL and VALEP Projects D. Other Benefits VIII. AGREEMENTS ANNEXES 1-1 Glossary of Common Terms in the Fertilizer Industry 1-2 ULTRAFERTIL Fertilizer Project 2-1 VALEFERTIL Corporate Organization Chart 2-2 List of Companies of CVRD Group 2-3 CVJRD - Financial Statements 2-4 CVRD - Funds Flow Forecast 3-1 Agriculture in Brazil 3-2 The Fertilizer Sector 4-1 Raw Material and Product Flow 4-2 Project Site Location Map (MAP IBRD 12666) 4-3 Project Scope, Process Description and Pollution Control 4-4 VALEP Phosphate Mining Project 4-5 Project Implementation Schedule 5-1 Capital Cost Estimate 5-2 Calculation of Price Escalation Factors 5-3 Permanent Working Capital Estimate 5-4 Interest During Construction - World Bank Loan 5-5 Interest During Construction - Buyer's Credits and Other Foreign Loans 5-6 Interest During Construction - BNDE/FINAME Loans 5-7 Estimated Use of Funds 5-8 Disbursement and Financing Schedule 6-1 Depreciation Schedule and Price Assumptions 6-2 Estimated Production Cost Summary and Output Value 6-3 Detailed (By Product) Production Cost Estimate in Constant 1976 US Dollars 6-4 Amortization Schedule - BNDE/FINAME Loans

5 Amortization Schedule - World Bank Loan 6-6 Amortization Schedule - Buyer's Credits and Other Foreign Loans 6-7 Income and Cash Flow Forecast 6-8 Balance Sheet Forecast 6-9 Notes to Financial Forecasts 6-10 Financial Rate of Return and Sensitivity Analysis 7-1 Economic Rate of Return and Sensitivity Analysis - VALEFERTIL Project 7-2 Economic Rate of Return for Combined VALEP-VALEFERTIL Projects MAPS IBRD Fertilizer Marketing Regions IBRD 12055R1 Major Fertilizer Plants

6

7 BRAZIL APPRAISAL OF THE VALEFERTIL PHOSPHATE FERTILIZER PROJECT SUMMARY AND CONCLUSIONS i. This report appraises a project to produce phosphate fertilizers with a capacity of 1,000 tons per day of monoammonium phosphate (MAP) and 1,100 tons per day of triple superphosphate (TSP) for respective annual capacities of 330,000 and 340,000 tons. The project will also produce the intermediate sulphuric acid and the phosphoric acid needed for the production of these phosphate fertilizers. The plant will be built at a new site, 37 km South of Uberaba, in the State of Minas Gerais. The project will use as principal raw material local low grade phosphate rock to be supplied by Mineracao Vale do Paranaiba S.A. (VALEP), which is presently installing phosphate mining and beneficiation facilities at Tapira, about 118 km east of the project site. VALEFERTIL is expected to start commercial production in mid-1979, and the output will be marketed in the Center region of Brazil through private fertilizer blending and mixing companies. Total financing requirements are estimated at US$294 million equivalent, including US$104 million in foreign exchange. A Bank loan of US$82 million is being proposed. ii. The project is sponsored by Fertilizantes Vale do Rio Grande S.A.- VALEFERTIL (VALEFERTIL) which was established in March 1976 as a wholly-owned subsidiary of Companhia Vale do Rio Doce (CVRD), Brazil's largest producer and exporter of iron ore. Although CVRD tried to attract private Brazilian fertilizer companies into some equity participation in the project, the idea was abandoned because the private sector fertilizer industry has a low level of capital formation when compared to the substantial capital requirements of modern and large-scale plants and the funds the industry has are already committed to other projects. iii. The Bank has so far assisted one project in the Brazilian fertilizer sector with a loan to Petrobras Fertilizantes S.A. for an ammonia/urea project at Araucaria (May US$50 million). Implementation of this project is proceeding satisfactorily. A second loan of US$64 million to Petrobras Fertilizantes S.A., to assist an ammonia/urea project at Sergipe, will also be proposed to the Bank's Executive Directors at about the same time as the proposed loan to VALEFERTIL. iv. Brazil's rising agricultural output has so far resulted mainly from expansions of cultivated area. Although the area under cultivation will continue to grow, the Government is also allocating substantial resources to raise agricultural productivity, especially of smaller farmholdings, through a number of extension programs and through increased application of modern inputs such as fertilizers and improved seeds. v. The Brazilian fertilizer industry supplied only 40% of the country's 1976 fertilizer consumption. This low level of domestic output is due to Brazil's lack of easily exploitable fertilizer raw materials, the large capital

8 - ii - requirement for fertilizer plants, and the absence of specific Government support policies for the fertilizer sector during the 1960s. As a consequence, the industry has concentrated on importing, mixing and distributing rather than on establishing large, efficient production facilities. To overcome this dependence on imports, the Government initiated a program for increasing domestic production with the publication in 1974 of a National Fertilizer Program and establishment of an interministerial Fertilizer Commission charged with ensuring the rational development of the sector. As a result, an investment program for large, primary production facilities emerged of which the Bank-assisted Araucaria project, as well as the Sergipe project and the proposed VALEFERTIL project form part. vi. There was little growth in fertilizer use up to 1966 but stimulating measures introduced in the mid-sixties brought six years of very rapid growth averaging 35% annually through From 1973 to 1975, however, fertilizer consumption increased only modestly due to the sharp rise in international prices. Concerned with the trend, the Government introduced a price subsidy to the farmers in April 1975 which resulted in a significant recovery in consumption during However, the subsidy was discontinued in December 1976 following the dramatic fall in international prices and the improvement in the price relationship of agricultural products relative to fertilizers which is expected to continue. The removal of the price subsidy is not expected to have any significant adverse impact on demand beyond the first few months of vii. Government policies in the fertilizer and the agriculture sectors during the Second Development Plan period ( ) provide a favorable setting for continued growth in fertilizer consumption, ranging from 10-12% annually. Domestic fertilizer production is projected to increase sharply, providing from 60% to 70% of expected consumption during the early 1980s compared to the 40% at present. Potash will remain the main import, and there will be near self-sufficiency in nitrogen and phosphate. The rapid expansion of consumption and the changing pattern of supply expected in the next few years will probably create severe strains on the existing marketing and distribution network. The Government, therefore, in connection with the Bank's loan for the Araucaria project, undertook a marketing and distribution study expected to be completed by late viii. VALEFERTIL's production will be marketed in the Center region where fertilizer consumption is already substantial and phosphate fertilizer demand is expected to grow at least 10% per year during But VALEFERTIL will produce only bulk products for sale to blenders/mixers and will need to develop marketing plans as there are limited fertilizer blending/mixing facilities close to its plant and target market area. ix. The project is based on modern, commercially proven technology and will be implemented by a general engineering firm under the supervision of VALEFERTIL's own project team which will be strengthened with foreign Technical Advisers. The general engineering firm is a consortium of

9 - iii - FLUOR (USA) and NATRON (Brazil), both experienced engineering firms. Basic engineering for the sulphuric acid plant will be provided by Lurgi (F. R. Germany) based on the Bayer-Lurgi process. The phosphoric acid plant will be engineered by Krebs (France) based on the Rhone-Poulenc process. Finally, the MAP and the TSP plants will be engineered by Gulf Design, a division of Badger (USA), based on the Swift-Gulf Design process for MAP and the Gulf Design process for TSP. Basic engineering is almost complete and procurement started in January 1977 using Bank guidelines. x. In agreement with the Government's policy to support and develop the Brazilian capital goods industry, 40% of the value of equipment and materials required by the project will be bid only among Brazilian suppliers. This is expected to increase the project cost, but by less than 2%. Proprietary equipment not available in Brazil, representing close to 5% of the total equipment value, will be purchased using suppliers credits. The balance of the equipment (55%) will be procured in accordance with Bank guidelines and will be financed from the proposed Bank loan; the loan will also cover the foreign cost of engineering, process licenses and technical advisory services, and interest during construction on the Bank loan. xi. The project will be financed with 40% equity from CVRD (US$118 million) and 60% loans from: (a) the National Development Bank, BNDE, and its subsidiary, FINAME (US$70 million); (b) foreign commercial bank loans and suppliers credits (US$24 million); and (c) the Bank (US$82 million). xii. The project's financial rate of return is satisfactory at 14% before and 11% after taxes. The financial forecast also shows satisfactory debt service coverage and debt/equity ratios upon reaching a steady state of production at an assumed 90% capacity utilization in VALEFERTIL will be required to follow prudent financial practices to maintain its financial strength. xiii. The economic rate of return of the project is 22% and remains satisfactory even under adverse conditions. The project is located in an area with little industrial development and will contribute to the regional industrialization objectives of Brazil. In addition, the project will accelerate the development of the fertilizer marketing and distribution system in the western section of the Center region, thereby providing support to that area's agricultural development programs. The project will also help ensure stable fertilizer supplies and prices, thereby contributing to a more orderly growth in agricultural output. The local engineering and capital goods industries will benefit with about US$73 million worth of goods and services that will be purchased by the project from these two sectors. The VALEFERTIL project is the final stage of a two-stage import substitution process, the first being the VALEP project which will produce phosphate rock concentrate. The economic return of the combined projects is 16.6%; this rate is similar to the return of the VALEFERTIL project if based on imported rock. On the other hand, VALEP's return is only 9.6% if it's output were to be sold to manufacturers in the coastal area i.e., without the VALEFERTIL project. Brazil therefore, has economic justification to undertake the two projects at the same time.

10 - iv - xiv. The project faces some risks but these are expected to be reduced or have been reduced to acceptable levels. For example, VALEFERTIL's lack of technical experience will be overcome through foreign technical advisory services. Similarly, the phosphoric acid plant has been designed with additional safety factors because the Tapira phosphate rock will be used in a fertilizer plant for the first time. VALEFERTIL will also develop marketing plans for the project, in consultation with the Bank, to overcome present weaknesses in the marketing and distribution system in the project's target market area. xv. On the basis of the agreements reached with the Government, CVRD and VALEFERTIL, the project is suitable for a Bank loan of US$82 million to VALEFERTIL for 15 years including 3 years of grace at the standard Bank rate plus a fee to the Government for a total interest rate of 10% per year.

11 I. INTRODUCTION A. Background 1.01 Fertilizantes Vale do Rio Grande S.A. - VALEFERTIL (VALEFERTIL) has requested a Bank loan of US$82 million equivalent to assist in financing its new phosphate fertilizer complex, with a capacity of 333,000 tons per year (TPY) of P 20 nutrient 1/ to be erected at Uberaba, Minas Gerais (Map IBRD 12055R1). aueraba, situated within an important cattle raising area, is located 475 km west of Belo Horizonte, the state capital and 510 km south of Brasilia, the national capital. At present, Uberaba has a limited, agriculture-based industrial sector, but an industrial park has been reserved in the southern part of the municipality and the project is one of several industries planned for that site. The project's primary raw material, phosphate rock concentrate, will be provided by Mineracao Vale do Paranaiba S.A. (VALEP), which is presently installing phosphate mining and beneficiation facilities at Tapira (near Araxa), some 118 km east of the project site VALEFERTIL and VALEP are wholly-owned subsidiaries of Companhia Vale do Rio Doce (CVRD), a public sector enterprise primarily engaged in the mining and exports of iron ore. The VALEFERTIL project is expected to be completed by July 1979 and estimated to cost about US$294 million equivalent, including US$104 million in foreign exchange. It will be financed with 40% equity from CVRD; local loans from BNDE (the National Development Bank) and its subsidiary FINAME 2/; foreign suppliers credits and commercial bank loans; as well as the proposed Bank loan Following preparatory missions in June and September 1976, the project was appraised in October 1976 by a mission consisting of Messrs. D. Carpio (Chief) and S. Venkataraman of the Industrial Projects Department and J. Cape (Consultant). B. Bank Group Involvement in the Brazilian Fertilizer Sector 1.04 The Bank Group has so far undertaken two operations in the Brazilian fertilizer industry. IFC participated in the ULTRAFERTIL project at Cubatao (Report No. IFC P-34, Sept. 21, 1966) which started production in late 1970 (Annex 1-2). The Bank approved in May 1976 a US$50 million loan to Petrobras Fertilizantes S.A. for an ammonia/urea project at Araucaria (Report No BR, April 23, 1976). A second loan of US$64 million to Petrobras Fertilizantes S.A., to assist an ammonia/urea project at Sergipe (Report No BR, April 1977), will also be proposed to the Bank's Executive Directors at about the same time as the proposed loan to VALEFERTIL. Implementation of the Araucaria 1/ A glossary of technical terms commonly used in the fertilizer industry is given in Annex / Agencia Especial de Financiamento Industrial - FINAME.

12 - 2 - project is proceeding satisfactorily. Procurement of time critical equipment is on schedule although the rest of procurement is about two months delayed due to an initially slower pace of basic engineering work and extensive negotiations with the associations of local manufacturers resulting in an increase in the proportion of local equipment from the appraisal estimate of about 50% to about 65% of total equipment cost. Expected project completion date is still mid-1979, as originally estimated, but financing requirements have increased by about US$32 million or 14%, all in local currency and are being covered by Brazilian sources. Even with the higher capital cost, Araucaria's economic rate of return remains satisfactory at about 20%. Implementation of the Sergipe project has just started and completion is planned for early II. PROJECT SPONSOR A. Fertilizantes Vale do Rio Grande S.A. - VALEFERTIL (VALEFERTIL) 2.01 VALEFERTIL is the project sponsor and will own and operate the project. The Company was registered in March 1976 with an initial share capital of Cr$20 million which will be fully subscribed by CVRD and its subsidiaries. CVRD had originally intended to attract other investors, including foreign firms, to participate in the project. However, it had to abandon this idea because private fertilizer enterprises in Brazil either have limited financial resources or are themselves already involved in other projects; foreign investors have also not shown interest in the project Although virtually the entire output of the phosphate miningbeneficiation project (VALEP) will go to the phosphate fertilizer project (VALEFERTIL), it is not unusual, considering the size and complexity of the two projects, to have them owned and operated by two separate companies. In addition, VALEP is to be basically a mining enterprise, and as such will also extract and process other ores, notably titanium and niobium, which are found in conjunction with the phosphate ore in Tapira. However, CVRD will ensure that the two firms will closely coordinate their activities and plans of operation through the respective Administrative Council of each company. This Council, which has four members all from CVRD, is the equivalent of the Board of Directors. The same CVRD director is the chairman of each company's Council which also includes each company's president. In addition, the president of one company will be a member of the Administrative Council of the other. As the Council has the responsibility for approving annual budgets, investment plans and operating policies of the company, this arrangement will ensure a cooperative as well as coordinated operation of the two enterprises. CVRD has agreed to continue this arrangement The corporate organization chart of VALEFERTIL is shown in Annex 2-1. Having just been established, the Company continues to receive some management support from CVRD. However, VALEFERTIL will need to evolve, within the next

13 months or so, a plan covering the organizational structure for its commercial operations, the main administrative as well as commercial policies and procedures, and the major financial control and information systems. In developing and implementing such a plan, VALEFERTIL will use management consultancy services, which are available in Brazil. The terms of reference for the management consultants are satisfactory and VALEFERTIL is now inviting proposals from local firms. The consultants are expected to be hired by late The president and the three other operating directors (implementation, technical operations, and finance) of VALEFERTIL have been appointed (Annex 2-1) and the Company has agreed to appoint a marketing manager by September 1, 1977 so that he can participate in formulating the organizational development plan, especially as it relates to marketing. After the start of commercial production, the position of project implementation director will be replaced by a commercial director and it is likely that the marketing manager will assume this position. So far, the top management personnel, with the exception of the project implementation director, have been recruited from outside the CVRD system. While these are qualified for the general management functions, additional technical expertise is needed, and will be filled by technical advisers (para 4.09). The Company has already selected Arthur Andersen and Company of Brazil as external auditor; this is satisfactory. B. Companhia Vale do Rio Doce (CVRD) 2.05 The parent company, CVRD, is primarily engaged in the exploration, mining, transport, and exports of iron ore although it also has subsidiaries and associated companies involved in forest industries including pulp and paper (Annex 2-2). During 1975, CVRD produced about 50 million tons of iron ore and exported about 47 million tons at an FOB value of US$611 million, accounting for 60% of total Brazilian iron ore exports CVRD's recent financial statements are shown in Annex 2-3 and its cash flow projections in Annex 2-4. Gross revenues have increased dramatically from US$524 million in 1974 to US$749 million in 1975 and about US$907 million in Almost 85% of gross revenues are accounted for by iron ore exports and roughly 70% of the increase in gross revenues is due to higher prices, with the balance accounted for by physical increases in sales. During the last three years, net income has averaged 22% to 25% of equity. At the end of 1976 the debt/equity ratio (37:63) and the current ratio (2.8) were also satisfactory. From 1973 to 1975, CVRD's capital investments averaged US$255 million annually and are expected to increase substantially during the next five years. CVRD's investment program for is forecast to require financing commitments of about US$3.8 billion equivalent in current terms, including close to US$3 billion for the Carajas iron ore project in the Northeast; about US$250 million for the VALESUL aluminum project; and US$118 million for the equity investment in the VALEFERTIL project. Bank financing has been requested for all three projects. The investment program is expected to be financed with 35% loans, 15% new capital subscriptions and Government funds and 50% internally generated resources; this is a sound capitalization.

14 -4- III. FERTILIZER MARKET AND MARKETING A. Agriculture in Brazil 3.01 Brazil's rising agricultural output has so far been due mainly to expansions of cultivated area (Annex 3-1). While the area under cultivation will continue to expand, agricultural growth will have to depend increasingly on raising the productivity of land already under cultivation. The Government in its Second Development Plan ( ) has a goal of a 40% (7% annually) real increase in agricultural output to make Brazil one of the world's foremost producers and exporters of foodstuffs. To achieve this, the Government has allocated substantial resources to open new areas to production and to increase agricultural productivity through a number of extension programs and through increased application of fertilizers and improved seeds These programs reflect a shift in agricultural policy. In addition to the promotion and establishment of large commercial agricultural enterprises, greater emphasis is also being given to raising small farm productivity, especially in the Northeast. These efforts to improve agricultural productivity, coupled with an increase in industrial activity have already resulted in a more rapid growth of per capita income in the Northeast than in Brazil as a whole; this process is expected to continue. B. The Fertilizer Industry and Government Fertilizer Policies 3.03 Brazil's fertilizer industry (Annex 3-2) consists of about 50 relatively small private sector and three larger Government-owned companies which supplied 40% of the 1976 fertilizer consumption. On a nutrient basis 33% of nitrogen, 69% of phosphates and none of the potash come from domestic production (para 3.11). But phosphate production is mostly based on imported raw materials since Brazilian phosphate rock output accounts for only 10% of phosphate consumption. The low level of domestic production has been due to Brazil's lack of easily exploitable fertilizer raw materials, the large capital requirement for the manufacture of chemical fertilizers and the absence of specific Government policies during the 1960s for the development of the fertilizer sector. The industry has therefore concentrated on importing, mixing, blending and distributing at the retail level rather than on establishing large, efficient production facilities. To overcome this dependence on imports, the Government embarked on a program of assuring self-sufficiency in fertilizers by 1980 with the publication in 1974 of a National Fertilizer Program (NFP) and establishment of an interministerial Fertilizer Commission charged with ensuring the rational development of the sector. As a result, an investment program for large, primary production facilities emerged of which the Bank-assisted Araucaria project, as well as the Sergipe project and the proposed VALEFERTIL project form part The production costs of local fertilizers have generally been higher than the cost of imports. To ensure full use of all the domestic output, a Government-established quota system requires retailers to purchase a given proportion of domestic fertilizer for each unit of imported fertilizer. This

15 - 5 - proportion is changed periodically depending on consumption forecasts and local availability. As new, large-scale efficient fertilizer plants enter into production, domestic fertilizer costs are expected to move, in the longer run, in line with long-term international prices, thereby obviating the need for the import quota system The high level of international fertilizer prices in (Annex 3-2, Chart I) induced a levelling off of fertilizer consumption in Brazil as in many other import dependent countries. Concerned with the trend, the Government introduced in April 1975, a price subsidy which reduced the price of fertilizers to the farmers by 15% to 30%. However, the Government discontinued the price subsidy in December 1976 following the significant fall in international fertilizer prices and the improvement in the price relationship of agricultural products relative to fertilizers which is expected to continue. Consequently, the removal of the price subsidy is not expected to have any adverse impact on-fertilizer demand beyond the first few months of C. Historical Consumption of Fertilizers 3.06 Growth of fertilizer consumption in Brazil though rapid has been erratic as shown in the table below (Annex 3-2, Charts II and III): Brazil - History of Apparent Fertilizer Consumption (1,000 nutrient tons) N P K Total % Yearly Change NPK Ratio ' :2.0: :1.7: :1.5: , '*1.0:1.9: , :2.1: , :2.2: , :2.3: , , :2.3: /a 488 1, , :2.3:1.3 /a As estimated by ANDA (National Association for the Promotion of Fertilizer Use). Up to 1966 there was little growth in fertilizer use due to the Government's neglect of the agriculture and fertilizer sectors. But stimulating measures introduced in the mid-sixties brought six years of very rapid growth with annual increments averaging 35% through From 1973 to 1975 however, fertilizer consumption increased only modestly, primarily due to the sharp rise in international prices and the scarcity of fertilizers in the world market. A significant recovery in consumption occured in 1976 due to the Government's subsidy and lower international prices.

16 Phosphate is the most important nutrient consumed, accounting for 50% of the total nutrient used, because of the soil characteristics and cropping pattern in Brazil resulting in an NPK consumption ratio of 1:2.3:1.3. Detailed statistics of fertilizer application to various crops are not available; however, sugar, coffee, soybeans and wheat together account for 50% to 60% of total consumption. D. Projected Demand and Supply of Fertilizers 3.08 As the past growth of fertilizer consumption in Brazil has been erratic, and detailed and consistent fertilizer statistics are lacking, demand projections carry considerable uncertainty. The most important factor favoring increased fertilizer consumption is the expected strong Government support in view of: (a) the Government's interest in higher agricultural production with fertilizers as an essential input; (b) its willingness to stabilize the price ratios between agricultural products and fertilizers; and finally (c) its declared goal of assuring a stable supply of fertilizers by sharply increasing domestic production. These provide a favorable setting for continued growth in fertilizer consumption. The limiting factors can be bottlenecks in transportation, deficiencies in extension services, shortages of foreign exchange for imports of fertilizer materials, and constraints to the expansion of agricultural credit The upper limit for future demand can be estimated by normative analysis from necessary nutrient replacement and by projections of cultivated area and cropping patterns. Estimates of this nature result in a potential demand of about 12 million nutrient tons (NPK) in 1985, compared to the consumption of 2.3 million nutrient tons estimated for 1976 (para 3.11). The maximum average annual growth between 1976 and 1985 would, therefore, be around 23%. The NFP assumes NPK consumption to grow at an average annual rate of 15% through But slightly lower rates are assumed in this report - ranging from 10-12% annually through considering the constraints mentioned in the preceeding paragraph. r 3.10 The higher NPK consumption growth rate anticipated in the NFP is primarily due to a much faster demand growth assumed for nitrogen, since those for phosphates and potash are in line with the estimates in this report. While more intensive application of fertilizers and higher levels of agricultural productivity imply an increasingly larger proportion of N relative to P application, the evolving cropping pattern and the opening of new cultivated areas, which require relatively more P,2 0 5 than N, are expected to result in only a minor change in the country s overall NPK consumption ratio.

17 - 7 - The regional 1/ distribution of fertilizer consumption is also assumed to follow the pattern of the last few years with about 10% of total consumption occuring in the Northeast, 60% in the Center and 30% in the South On the other hand, the pattern of future fertilizer supply is expected to differ dramatically from the present as a direct result of the NFP. While self-sufficiency by 1980 will not be achieved, during the early 1980s domestic fertilizer supply is expected to account for 60% to 70% of total NPK consumption compared to about 40% at present. Potash (K 0) will remain the main import and there will be near self-sufficiency in and P Brazil - Fertilizer Demand/Supply Forecast a/ (1,000 nutrient tons) (Estimate) Demand N ,000 1,400 P ,145 1,680 1,990 2,580 K ,000 1,200 1,700 Total NPK 2,286 3,480 4,190 5,680 Domestic Production b/ N P ,555 1,799 2,026 Total NPK 946 2,125 2,759 3,006 % Self-Sufficiency N P Total NPK a/ Bank forecast from Annex 3-2. For 1982, the K2 0demand is based on a 1.2:1 ratio between K20 and N demand. b/ Brazil is not expected to produce any significant K 20 until the mid- 1980s at the earliest. 1/ The fertilizer marketing regions in Brazil are shown in MAP IBRD which are: South: Rio Grande do Sul and Santa Catarina. Center: Parana, Sao Paulo, Rio de Janeiro, Minas Gerais, Mato Grosso, Goias, and Espirito Santo. North: Acre, Amazonas, Amapa, Para, Rondonia and Roraima. Northeast: The rest of the States. Because of its negligible role in the fertilizer sector, the North is not referred to in this report.

18 - 8 - The above estimate takes into account only those projects that are firmly planned and have a reasonable chance of finding the necessary financing. Thus, the projected gap between demand and production for 1985 should be interpreted as an indication of what additional projects would be needed in the sector. To keep abreast of the investment plans in the fertilizer sector and to ensure that new plants are built in accordance with as realistic demand projections as possible, the Government has confirmed that it will exchange views with the Bank from time to time regarding the expansion and development of the fertilizer sector, including investment plans and import and pricing policies The forecast of the national demand/supply situation masks some sharp regional and structural differences in the pattern of local production as described in more detail in Annex 3-2. The South region will not produce any basic N product through the mid-1980s while the Center region 1/ is expected to produce by 1982 about 80% of its N demand. The Northeast, however, will start exporting substantial N products to the South and Center regions as early as 1978 and through possibly the end of the 1980s. The inter-regional movement of phosphatic fertilizers, on the other hand, is not expected to reach any significant volume. The South could produce up to 80% of its phosphatic fertilizer needs by 1980, but all production will be based on imported raw materials such as phosphate rock and/or phosphoric acid since no phosphate rock deposit has yet been found in that region. The Northeast will achieve self-sufficiency in phosphate products by 1979, but similarly, all production will also be based on imported raw materials. The Center region is also expected to be self-sufficient in phosphate products by 1979 and may even be capable of shipping a limited quantity to the South after In contrast to the two other regions, the Center will be using an increasingly large proportion of local phosphate rock, since its production is expected to increase from about 19% of total P 205 demand in the region in 1976 to 80% by 1979 and 95% or more by /. E. Fertilizer Marketing and Distribution 3.13 About 90% of total fertilizer distribution is effected through blenders which are either owned by fertilizer manufacturers (25% of total), independent companies (65%) or cooperatives (10%). The Government intends private firms to continue the handling, marketing and distribution functions in the sector. However, the rapid expansion of consumption expected in the next few years, and changing pattern of supply, might create a severe strain on the existing fertilizer marketing and distribution network. To analyze potential bottlenecks, and their effect on future fertilizer consumption, the Government, in connection with the Bank's loan for the Araucaria project, undertook a marketing and distribution study under satisfactory terms of reference. The study began in October 1976, and its results are expected to be available by late / The VALEFERTIL project is located and will market all its production in the Center region. 2/ The Center's phosphate rock production by 1982 is expected to represent close to 60% of the total national phosphate consumption.

19 - 9 - F. The Market and Marketing of VALEFERTIL's Production 3.14 VALEFERTIL's production will be marketed in the Center region with most of the sales concentrated in its middle and western sections. Fertilizer (NPK) consumption in the region has grown at an average annual rate of about 20% during ; and the NPK consumption ratio has averaged 1:2:1.3 during At present, only about 50% of P 05 consumption in the region is in the form of high analysis products, such as TSP and MAP; but this proportion is expected to reach 60% by 1980 and 70% by 1985, reflecting the convenience and the transport cost advantage, on a nutrient basis, of these products over the low analysis products such as SSP and thermo-phosphates Although the South is the most intensive user of fertilizers, the Center region, with its larger cultivated area, accounted for about 60% of total NPK consumption and 54% of P 205 consumption in Brazil during This pattern is expected to continue through The forecast of the phosphate fertilizer demand/supply situation in the Center region (Annex 3-2) is summarized below indicating a shift from a deficit situation up to 1978, to a potential small surplus during By 1984 any surplus will probably have disappeared and a deficit situation could again occur by Center Region - Phosphate Fertilizer Demand/Supply Forecast (1,000 tons of P ) (Estimate) Demand ,075 1,393 Local Production VALEFERTIL Project ARAFERTIL Project Others Total ,082 1,299 % Self-Sufficiency The P 0 demand forecast for the Center region assumes an annual growth rate of i0@ during and 9% during , the same rates used for the national P 20 demand forecast. Should demand grow even only slightly higher than these ra5es, the surplus projected for will not materialize. In any case, this surplus is expected to be modest, ranging from 100,000 to 150,000 TPY of P 0 and exists because four firms have the flexibility of producing either only 5 SSP, equivalent to 210,000 TPY of P 20, or a combination of SSP and TSP, the latter based on imported phosphoric acid, for an equivalent of 300,000 TPY of P (the basis for the assumed surplus). The production forecast above assumes that during 1980 through 1983, these four firms will adjust their product mix between SSP and TSP to avoid any significant surplus in the region. It appears therefore that the Center region will not need additional P product capacity other than the VALEFERTIL and the ARAFERTIL

20 projects until about By 1985 or 1986, additional capacity of about 100,000 TPY of P 20 will be needed each year to keep production in line with increases in deman Ribeirao Preto, some 180 km south of Uberaba, is the center of gravity of fertilizer consumption in the Center region at present. During the next five to ten years, this center of gravity is expected to move closer to Uberaba, thus further improving the location advantage of Uberaba as the project site. But because of the dominance of imported fertilizer raw materials and products at present, most of the existing manufacturing and blending facilities are located close to the coast in Sao Paulo and Santos. Since VALEFERTIL intends to manufacture only bulk products for sale to blenders/ mixers, additional blending/mixing facilities will have to be located inland, close to Uberaba and the project's target market, to take full advantage of transportation savings. VALEFERTIL has completed, through CVRD's engineering and planning division and the Brazilian consulting firm Economia Estatistica e Planejamento Ltda. (EEP), a study of the spatial distribution of fertilizer consumption in the Center region. The study will provide the basis for identifying potential locations for new blending/mixing facilities and the likely product transport pattern for VALEFERTIL. VALEFERTIL then intends to assist private blending/mixing firms and financial institutions, in developing and implementing a program for blending/mixing as well as storage and distribution projects The development of VALEFERTIL's marketing plans and related blending/mixing facilities is a critical step to ensuring the success of the project. So far, VALEFERTIL has not appointed a marketing manager who will be responsible for developing such plans, but VALEFERTIL has agreed that a marketing manager will be appointed by September 1, 1977, in time to participate in the development of the organizational plans for the Company (paras 2.03 and 2.04) and to formulate, in consultation with the Bank, specific marketing plans for the project by December 31, VALEFERTIL's production is expected to account for 25% to 30% of the P2 5 production in the Center region during , making the Company an important factor in the industry. The marketing risks for the project, however, are considered minimal, since there is sufficient time to develop the necessary marketing and distribution network. The Government's national fertilizer marketing and distribution study (para 3.13), expected to be completed by late 1977, should also provide additional information on the required investments in marketing and distribution facilities in the region. IV. THE PROJECT A. Project Scope and Location 4.01 The project consists of (a) two sulphuric acid units each with a design capacity of 1,300 tons per day (TPD) of 98.5% acid; (b) two phosphoric acid units each with a design capacity of 470 TPD of P 20 in the form of 52% acid; (c) a 1,000 TPD MAP unit; (d) a 1,100 TPD TSP unit; (e)

21 storage facilities for 30,000 tons of sulphuric acid, 8,500 tons of phosphoric acid, 4,000 tons of ammonia, 2,500 tons of fuel oil, 28,000 tons of sulphur, and 100,000 tons of finished products; (f) a rail spur and 25 ammonia tankwagons; and (g) ancillary facilities such as offices, laboratories, maintenance shops, effluent treatment, and the supply of utilities. VALEP will provide the phosphate rock storage facilities (para 4.04). The ammonia tankwagons are needed since the railways do not have enough of these specialized wagons nor do they intend to purchase any for this project. Because the project would be the first to use the Tapira phosphate rock, there are inherent uncertainties as to the achievable efficiency of the phosphoric acid production process. The phosphoric acid units have therefore been designed with additional safety factors in the reactors and filters. Should the actual production experience approximate the test results, the phosphoric acid units may have about 10% more capacity than indicated above. To take advantage of this potential additional phosphoric acid, the design capacities of the TSP unit and the sulphuric acid units are slightly larger than what would be needed to just match the phosphoric acid design capacity indicated above. On the basis of this phosphoric acid design capacity, the project will have an annual production capacity of 330,000 TPY of MAP ( ) and 340,000 TPY of TSP ( ). The design capacities of the different units are within the range presently installed elsewhere in industrial countries and give the project the advantages of economies of scale. A process flow chart is shown in Annex The project site (Annex 4-2) is within a recently established industrial park located along the right bank of the Rio Grande river, 37 km south of Uberaba City, an important crossroads in the Center region, with transportation links to Goiania, Brasilia, Montes Claros, Belo Horizonte, Sao Paulo and Santos. VALEFERTIL has already acquired the 48 ha of land needed for the project. Soil investigations and land grading have been completed. The project is the first to be erected in the industrial park, which is located between the proposed FEPASA branch railroad (Uberaba-Ribeirao Preto) currently in the early stages of construction, and the Goiania-Sao Paulo highway, ensuring adequate transport means. The FEPASA branch railroad is included in the Bank's Third Railway Loan (Loan No BR, Nov. 12, 1975) and is expected to be completed by December 1978, about six months earlier than the VALEFERTIL Project. VALEFERTIL has agreed to enter, by September 30, 1977, into a contract with FEPASA for the construction of a rail spur to the plant site, and the provision of railroad cars for the transport of sulphur, fuel oil and finished products. The external power requirements of the project will be provided through a nearby existing 30,000 KVA substation which is connected to the hydroelectric power plants along the Rio Grande river. A 2 km water pipeline from the Rio Grande river to the factory, with associated water treatment facilities, will supply the water for the project and is included in the project scope. B. Technology 4.03 To achieve a 99.5% efficiency in the conversion of sulphur to sulphuric acid and to minimize pollution, the project will use the Bayer- Lurgi double catalyst process which involves a two-stage catalytic conversion and a two-stage absorption operation. Basic engineering for the

22 -12- sulphuric acid units will be provided by Lurgi (F.R. Germany). The phosphoric acid units will be based on the Rhone-Poulenc (France) process which provided the most favorable test results using samples of beneficiated Tapira phosphate rock. Basic engineering for the phosphoric acid units will be provided by Krebs (France) and for the MAP and the TSP units by Gulf Design, a division of Badger (USA), based on the Swift-Gulf Design process for MAP and the Gulf Design process for TSP. A description of the different production processes and the plant facilities is given in Annex 4-3. All the processes are commercially proven and the engineering firms 1/ experienced, and there are no unusual risks involved. C. Phosphate Rock Supply and the VALEP Project 4.04 Beneficiated phosphate rock (35% P 0 5 ) for the project will be supplied by VALEP from its phosphate mining-beneficiation project currently under construction at Tapira, some 118 km from the VALEFERTIL project site (Annex 4-2). The beneficiated phosphate rock will be transported as a slurry through a nine-inch (internal diameter) underground steel pipeline designed and being constructed for VALEP by Bechtel (USA), an experienced engineering firm. Phosphate rock dewatering, drying and storage facilities at the pipeline terminal (i.e., adjacent to the VALEFERTIL plant site) are included in the VALEP project scope. A review of the topographical survey for the pipeline and its system design confirms the technical soundness of the scheme and the pipeline is free from any serious risks. The VALEP project is expected to be completed by August 1978, some 11 months earlier than the VALEFERTIL project, and the initial production will be sold to manufacturers in the Sao Paulo area. A description of the VALEP project, including capital and operating costs, is given in Annex The Tapira deposit has low grade phosphate rock, containing an average of only 8% P 205 compared to the 12% to 16% found in the Catalao, the Araxa, and the Patos deposits in the region and to the 20%-30% usually found in the commercial deposits in other countries. The beneficiation process, a locally developed technology, is similar to the process which has been wellproven at Jacupiranga 2/; it has also been tested at the ARAFERTIL industrial pilot plant at Araxa and additional laboratory tests are being done to optimize and verify the most efficient production parameters. The P 0 recovery in the beneficiation process is expected to range from 60% to 70% whlch is very satisfactory; the process does not present any unacceptable risks. 1/ The engineering firms (and associated processes) were selected from among seven different international firms. VALEFERTIL was assisted by FLUOR/NATRON and by process engineering consultants (DUETAG-France) in the selection. 2/ The Jacupiranga phosphate deposit, in the State of Sao Paulo, provides rock to the QUIMBRASIL phosphate fertilizer complex which started operations in 1969 (Annex 3-2).

23 The VALEP project scope had been finalized before the Bank was requested to assist the VALEFERTIL project. The VALEP beneficiation plant was designed based on 7,000 hours per year (HPY) of operation. However, because of heavy rainfall and poor drainage conditions during six months of the year, mining is expected to be hampered and the beneficiation plant can be expected to operate for only 6,000 HPY. This would mean that the VALEP design capacity is 15% short of the VALEFERTIL requirements, although a review of the sizes of the equipment actually ordered indicates that it may have just enough achievable capacity to meet the design needs of VALEFERTIL, as all discussed in greater detail in Annex 4-4. The beneficiation plant's capacity can also be increased by 15% with an additional investment of about US$5 million. However, this debottlenecking is best carried out after both VALEP and VALEFERTIL have gained adequate operating experience. Agreement was reached with CVRD and VALEFERTIL that after VALEFERTIL has operated for about two years, the Company will undertake a joint study with VALEP to analyze the feasibility of increasing and/or balancing the outputs of the two projects. The results of the study will be made available to the Bank. If appropriate, a debottlenecking program, acceptable to the Bank, will also be implemented For the above reasons, VALEFERTIL will have to have access to other sources of phosphate rock, should VALEP's production be inadequate for whatever reason. To this end, agreement has been obtained with the Government, CVRD and VALEFERTIL that the project will be allowed to use, and will utilize, other sources of phosphate rock, either domestic or foreign, if needed to operate the project at full capacity. It is thereby expected that the benefits from the additional output will outweigh the higher cost of other rocks and the likely lower efficiency of the phosphoric acid plant because the characteristics of these other rocks are different from the VALEP rock for which the VALEFERTIL project has been designed. To ensure adequate flexibility in using alternative phosphate rocks, the Technical Advisers (para 4.09) will be asked by the Company to review the designs of the facilities included in the VALEFERTIL project and propose measures that will provide such flexibility. VALEFERTIL will also enter, by December 1977 and after consultations with the Bank, into a long-term phosphate rock supply agreement with VALEP, to cover at least the design requirements of the VALEFERTIL project, and to obligate VALEP to jointly undertake the debottlenecking study mentioned above. D. Project Management and Implementation 4.08 The project will be implemented by VALEFERTIL through its own small project team headed by a Project Manager who will report to the Project Implementation Director (Annex 2-1). The project team, which will be strengthened with Technical Advisers (para 4.09), will supervise and coordinate the work of the process licensors/basic engineering firms (para 4.03) and the general engineering firm. The general engineering firm is a consortium of FLUOR Latin America Inc., a subsidiary of the FLUOR Corporation (USA),

24 and NATRON Consultoria e Projetos S.A. (Brazil) 1/; the consortium will provide overall project management, detailed engineering of process plants and engineering of offsite facilities, procurement services, management of subcontracted construction and erection services, and supervision of commissioning. The process licensors will assist NATRON in detailed engineering design and train VALEFERTIL personnel in the operation and maintenance of the process plants. FLUOR will head the consortium and will also undertake foreign procurement; NATRON will undertake local procurement and supervision of construction 2/. FLUOR is an experienced international engineering company and the implementation arrangement is satisfactory. Basic engineering is almost complete and procurement of equipment started in January 1977 using Bank guidelines. The project implementation schedule is shown in Annex 4-5 with the start of commercial operation planned for July Neither VALEFERTIL nor CVRD has any previous experience in constructing and operating a large fertilizer plant. To strengthen VALEFERTIL's (and the project team's) technical capabilities in supervising the engineering firms, in commissioning and initial plant operations, the Company is presently negotiating 3/ and intends to contract by May 1977, foreign technical advisory services through an experienced firm. The terms of reference for the technical assistance have been found satisfactory by the Bank and include an evaluation of the project design to permit the use of other phosphate rocks (para 4.07). E. Employment and Training 4.10 The project will provide direct employment for about 450 people, most of whom will be skilled or semi-skilled technicians. The staff will be recruited from the existing fertilizer and chemical plants in Brazil and from the city of Uberaba. Preliminary training plans for the technical staff have been prepared and will be finalized after the Technical Advisers have made their comments on the proposed program. Part of the technical training will be arranged by the process licensors in plants using their respective technologies. VALEFERTIL also intends to finalize arrangements with existing phosphate fertilizer companies in Brazil (e.g., ULTRAFERTIL) for some training in their facilities. The plans include training in safety and occupational health. F. Pollution Control 4.11 The process units of the project will be designed and engineered to reduce pollutants in the effluents to acceptable standards. In the 1/ The engineering firm was chosen after receipt of international proposals from foreign firms in association with qualified local firms. 2/ NATRON is associated in a similar way in the implementation of the Bank-assisted Araucaria Fertilizer Project. 3/ Three proposals for technical advisory services were received and evaluated by VALEFERTIL.

25 sulphuric acid plant, the high conversion efficiency of sulphur to sulphuric acid (para 4.03) plus the use of two-stage demisters for the acid mists after final absorption, reduces the contaminants in the effluent gases which will be discharged into the atmosphere through a tall stack to ensure adequate dilution. In the phosphoric acid plant, the major contaminant is fluorine; but about 80% will be recovered as fluosilicic acid. VALEFERTIL is presently evaluating two alternatives: selling the fluosilicic acid or converting it to either cryolite or aluminum fluoride. The markets for these two products and for fluosilicic acid are being evaluated and a choice will be made with the assistance of the Technical Advisers who will be hired shortly (para 4.09). Most of the remaining fluorine will be condensed and discharged into a large pond designed to ensure no overflow. In an emergency, the overflow will be treated with lime to precipitate the fluorine compounds. A small amount of fluorine will escape with the gaseous effluents but will not exceed 230 grams per ton of P produced. Gypsum, the by-product in the production of phosphoric acid, will be dumped in gypsum ponds with adequate capacity for the project's life. Finally, the effluent gases from the MAP and the TSP units will be scrubbed to remove MAP dust, ammonia (MAP unit) and fluorine dust (TSP unit). The Company will continuously monitor the effluent streams. The pollution control system is described more fully in Annex 4-3 and is satisfactory The pollution control system for the VALEP project (Annex 4-4) is also satisfactory. V. CAPITAL COST, FINANCING PLAN AND PROCUREMENT A. Capital Cost 5.01 Total financing required for the project is estimated at US$294 million equivalent (Cr$3,350 million) of which US$104 million (Cr$1,179 million) is in foreign exchange. The capital cost estimate is based on mid-1976 prices and already reflects the equipment price increases experienced in 1975/76. Details of the capital cost estimate and the assumptions used are given in Annexes 5-1 and 5-2 and are summarized below:

26 VALEFERTIL - Summary of Capital Cost Estimate /a Cr$ Million US$ Million Foreign /b Local Foreign /b Local Exchange Currency Total Exchange Currency Total % Equipment and Materials (FOB) Freight, Insurance and Taxes Civil Works Erection Land & Site Improvement Engineering & Other Services /c Preoperating Expenses Testing and Commissioning Base Cost Estimate (BCE) 789 1,462 2, Physical Contingencies (PC) /d Price Escalation /e Installed Cost 971 1,814 2, Working Capital Interest During Construction Total Financing Required 1,179 2,171 3, /a The costs expressed in Cr$ have been calculated from the US$ costs using an exchange rate of Cr$11.40 = US$1.00 and are therefore only notional, as changes in the exchange rate are expected (para 5.04). /b Includes US$19.4 (Cr$221) million equivalent of indirect foreign exchange cost estimated as 25% of the value of locally purchased equipment plus 15% of the civil works cost. /c The foreign exchange base cost estimate for engineering and other services of US$10.2 million includes about US$0.8 million for the technical advisory services while the local currency cost of US$16.2 million includes about US$1.3 million for local management consultancy and other services. /d 5% of the BCE for engineering and other services, land and site improvement; 10% of the BCE for all other items. /e 19.7% of the BCE and PC for civil works; 14.5% of the BCE and PC for all other costs except engineering which has nc price escalation provision The equipment costs have been derived from estimates prepared in October 1976 and reconfirmed in March 1977 by VALEFERTIL and FLUOR based on a similar project in another country scheduled for completion in The equipment base cost estimate also reflects the higher cost of items to be procured in Brazil (para 5.09). The project already qualifies for an 80% exemption from import duties on equipment but VALEFERTIL has requested for a 100% exemption. In the meantime, the cost estimate includes a provision for

27 duties on 20% of the CIF value of the imported equipment. The civil works cost has been estimated by FLUOR/NATRON based on preliminary designs and experience with similar civil construction in Brazil. The cost of process licenses, engineering and other services are based on contracts already signed in 1976 with the engineering firms and the process licensors. The estimated cost of foreign engineering and technical services range from US$4,500 to US$7,000 per man-month exclusive of travel and subsistence allowances. The cost of local engineering and other consultancy services range from US$2,000 to US$5,000 per man-month As noted, VALEFERTIL intends to sell the entire output in bulk to fertilizer blenders/mixers. The estimates do not, therefore, include bagging facilities. The investment required for the fluosilicic acid recovery is included in the capital cost estimate but the cost of the facilities for converting the fluosilicic acid to either cryolite or aluminum fluoride (para 4.11) of about US$5 million (in constant 1976 US dollars) is not included 1/; however, should VALEFERTIL decide to install these facilities, CVRD has agreed that additional funds, equity and/or loans, as appropriate, will be provided Price escalation for equipment is calculated on the basis of projected price increases of 9% for 1976 and 8% per year for in major equipment supplying countries and worldwide inflation. No price escalation is provided for engineering costs as the values given represent the amounts estimated in the various contracts in current terms (Annex 5-2). International price inflation has also been used for local supplies and services expressed in US dollars on the assumption that differences in the domestic and the international inflation rates will be accounted for by adjustments in the foreign exchange rate. This assumption is based on the current government policy which is expected to continue. On civil works however, escalation rates of 13% for 1976 and 12% per year for have been assumed based on experience in the recent past, that the cost of civil works in Brazil will rise faster than the general price level. The above cost estimate including the provisions for physical and for price contingencies is considered adequate The initial working capital requirement is estimated at US$33 million and details are given in Annex 5-3. Interest during construction is calculated on the basis of total debt financing of US$176 million, which 1/ Traditionally, the fluosilicic acid is chemically treated with lime and then discharged as a waste material. While this is a satisfactory environmental protection method, it is generally economic to convert the fluosilicic acid to either cryolite or aluminum fluoride if there is a domestic market for these two products as shown in the Bank's evaluation of similar pollution control projects in India (Report No. 928-IN, November 24, 1975). As the choice of product, i.e., cryolite or aluminum fluoride, will be made only after a thorough market study is completed (these two products are used in the aluminum industry and therefore have a market in Brazil), the costs and the benefits of the conversion facilities, or the benefits from the sales of the fluosilicic acid, are not included in the financial or economic analysis of the VALEFERTIL project.

28 corresponds to a debt/equity ratio of about 60:40. The assumed loan terms and corresponding interest during construction are shown in Annex 5-4 for the proposed Bank loan, Annex 5-5 for the other foreign loans, and Annex 5-6 for the local loans. B. Financing Plan 5.06 The financing plan for the project is as follows: VALEFERTIL - Proposed Financing Plan In Millions of Cr$ In Millions of US$ % Equity (CVRD) 1, Loans Local Loans (BNDE/FINAME) Foreign Loans /a World Bank Total Loans 2, Total Financing 3, /a Suppliers credits and commercial banks. CVRD will provide the equity funds equivalent to 40% of the required financing. An analysis of CVRD's cash flow projections indicates that it should be able to provide these funds as well as additional funds to cover cost overruns were this to become necessary. The balance of the local currency financing will come from a BNDE loan of about US$48 million and a medium-term loan from FINAME of about US$22 million. The BNDE loan was signed December 27, 1976 for Cr$506 million (in December 1976 Cr$) equivalent to about US$41 million in constant 1976 US dollars. Monetary correction during the disbursement period is expected to raise actual disbursements to about US$48 million in current terms. The loan has a maturity of 15 years including 3.5 years of grace. The FINAME loan will be for at least 8 years including about 2.5 years of grace. The FINAME loan and US$10 million of the BNDE loan will each separately finance up to a maximum of 80% of the purchase price of the equipment and materials to be bid among Brazilian suppliers only 1/. Foreign suppliers credits of about US$4 million will be used towards proprietary equipment that are not manufactured in Brazil. Finally, foreign commercial bank loans of about US$20 million, although untied as to purpose, will be utilized to cover, primarily the disbursements for local consultancy services after mid-1977 (US$9.5 million), part of erection costs, and to supplement the suppliers credits as well. This report assumes the foreign suppliers credits and commercial loans to be for 8 years including 2.5 years of grace. 1/ VALEFERTIL has submitted to FINAME a list of equipment to be procured locally and FINAME has formally indicated that it will finance up to 80% of the ex-factory value of each equipment in the list. A separate loan is made for each piece or group of equipment (i.e., purchase order) financed by FINAME. Part of the BNDE loan (US$10 million) will finance 80% of a separate list of goods.

29 A Bank loan of US$82 million is proposed for 15 years including 3 years of grace at the standard interest rate plus a fee to the Government to bring the total interest rate to VALEFERTIL to 10% per year. The Bank loan will be made to VALEFERTIL which will bear the foreign exchange risk Agreement has been reached with the Government, CVRD and VALEFERTIL regarding the financing plan as follows: (a) CVRD will assist VALEFERTIL obtain the foreign suppliers credits and commercial bank loans (US$24 million) to be repaid in not less than 7 years including a grace period of not less than 2 years; (b) should a financing gap and/or cost overrun occur, CVRD will ensure that both the VALEFERTIL and the VALEP projects will be provided the necessary funds in the form of equity and/or loans to complete the respective projects and any subsequent debottlenecking program (para 4.06) without delay. For purposes of this covenant, the VALEFERTIL project is considered complete when the plant has been tested and proven to be satisfactory in accordance with established engineering procedures, and has operated at an average production rate of not less than 80% of the design capacity during a period of 60 consecutive days. The VALEP project is considered complete when the facilities have been in satisfactory operation for not less than 60 consecutive days including at least 15 consecutive days at an average production rate of 80% (or better) of the project design capacity. In addition, the funds should be provided in such a way that the debt/equity ratio would not exceed 60:40 at completion of each project and the current ratio would be at least 1.2:1.0; and finally (c) as a back-up to CVRD's obligations, the Government has ensured that all funds, including foreign exchange will be provided as required to complete the VALEFERTIL and the VALEP projects without delay and that all approvals for the implementation of these projects, including import licenses will be provided in time. C. Procurement 5.09 Equipment, materials and the initial set of spare parts valued at US$103 million for the VALEFERTIL project have been divided into three procurement categories: (a) proprietary equipment that is not made in Brazil, with an estimated FOB value of about US$4.5 million to be purchased using suppliers credits supplemented with foreign commercial bank loans as necessary; (b) items to be bid only among Brazilian suppliers with an estimated exfactory value of US$40.8 million to be financed with the BNDE/FINAME loans (80%) and with equity funds (20%); and (c) items to be purchased using the Bank's guidelines with an estimated FOB value of US$57.7 million of which about US$50 million will be procured through international competitive bidding (ICB) and the balance through international shopping. Of the items for local procurement (US$40.8 million), equipment valued at some US$24 million is not appropriate for international competition and is efficiently purchased in Brazil. The remaining local equipment (about US$17 million) could possibly cost up to 30% less if procured through international competition. As the estimated increase in total project cost is less than 2%, and the local capital goods industry is experiencing some underutilization of capacity following the Government's and the industrial sector's cut-backs in capital spending, the amount of the local procurement is considered reasonable.

30 For purchases to be financed with Bank funds, the list of goods and their poterntial suppliers will require prior Bank approval and procurement will follow Bank guidelines. Local suppliers will be allowed to participate in international competitive bidding with a 15% (or the import duty rate if lower) preference margin. On this basis, it is estimated that about 25% (US$15 million) of the FOB value of the equipment to be financed by the Bank will likely be won by Brazilian suppliers. Finally, the potential suppliers of critical equipment not financed by the Bank will also require prior Bank approval. D. Allocation and Disbursement of Bank Loan 5.11 The Bank loan will finance (a) 100% of the foreign exchange expenditures for foreign engineering, process licenses and technical advisory services (US$10.7 million); (b) 100% of the foreign expenditures 1/ (if imported) or the ex-factory (if locally supplied) cost of equipment procured through ICB (US$54.0 million); (c) the foreign expenditures for imported items and packages costing less than US$100,000 each to be purchased through international shopping (US$8.3 million); and (d) interest during construction on the Bank loan (US$9.0 million). The foreign engineering services proposed for Bank financing have already been contracted. Retroactive financing for disbursements after August 15, 1976 for these services of up to US$6.2 million is therefore recommended. About US$5 million worth of equipment contracts eligible for Bank financing will have been made before May 1977, the expected signing date for the proposed Bank loan. However, no retroactive financing will be needed for equipment purchases as VALEFERTIL has already made satisfactory arrangements to open letters-of-credit for these contracts and the contracts will stipulate that initial payments will be made only after May If necessary, CVRD has agreed to guarantee these letters-of-credit. The proposed allocation of the Bank loan is shown in Annex 5-7 and the estimated disbursement schedule is given in Annex 5-8. VI. FINANCIAL A4ALYSIS A. Operating Costs and Financial Forecasts 6.01 Production is assumed to start at 50% capacity utilization and reach the steady state at 90% in the second full year of operation (1981). Depreciation charges (Annex 6-1) are on a straight-line basis over 12 years. The estimated product and raw material prices over the project's operating life, expressed in constant 1976 US dollars are shown in Annex 6-1. The MAP and the TSP ex-factory prices are assumed at US$280/ton and US$225/ton respectively, i.e., at levels competitive with imports on a net-of-tax basis delivered to the market area. 1/ Foreign expenditures for equipment and materials include ocean freight and insurance. The difference between the FOB value of US$57.7 million used in para 5.09 and the US$62.3 for (b) and (c) combined is ocean freight and insurance.

31 The delivered price of the phosphate rock is taken to be US$55/ton as tentatively agreed between VALEP and VALEFERTIL. This rock price is mid-way between what VALEP would otherwise get from alternative users in the coastal area of Sao Paulo or Santos (US$45/ton ex-pipeline terminal) and the cost to VALEFERTIL of using imported rock (US$65/ton delivered). But this transfer price can be maintained only if VALEP operates efficiently and at high rates of capacity utilization because its cost of mining and beneficiation, as discussed in Annex 4-4, is relatively high, about US$37/ton (including depreciation but excluding financial charges) delivered to the VALEFERTIL complex, which provides only a modest financial rate of return of 9% before taxes for VALEP. The other two raw materials, sulphur and ammonia, have a delivered cost of US$95/ton and US$215/ton respectively on the assumption that no import duty would be levied on raw materials. 1/ 6.03 Raw materials account for about 68% of the project's production cost, with phosphate rock representing 41% and sulphur and ammonia another 27%. Depreciation charges contribute just 15% of the production cost. The production cost structure is shown in Annexes 6-2 and 6-3; it is summarized below and is typical for phosphate fertilizer projects. It differs from that of ammonia/urea projects where raw materials account for about one-third and the depreciation charges another one-third of the production cost. 1/ Ammonia will be initially imported and after 1981, will probably be supplied from the proposed ammonia/urea project at Norte Fluminense (Annex 3-2). Sulphur will be imported.

32 VALEFERTIL - Production Cost Estimate /a (In Constant 1976 US Dollars) US$ Per Ton (bulk) Annual Cost-Combined MAP/TSP MAP TSP US$ Million _ Variable Costs Phosphate Rock Sulphur & Ammonia Chemicals & Utilities Sub-Total Fixed Costs Depreciation Maintenance, Labor & Others Sub-Total Total Cost Before Interest: /a At 90% capacity utilization. An official foreign exchange rate of Cr$11.40=US$1.00 is used in the financial analysis The debt service schedules for the various loans are shown in Annexes 6-4 through 6-6 and projections of the income, cash flow and the balance sheet statements, together with major assumptions are presented in Annexes 6-7 through 6-9. These projections are expressed in current US dollars, with both costs and revenues escalated at the assumed international inflation rates (Annex 5-2) and are summarized on the following page The financial forecasts indicate a satisfactory financial situation for VALEFERTIL, with a debt service coverage ratio in excess of 1.7 times on reaching 90% capacity utilization and the capacity to start distributing dividends in Dividend payments for the more successful Brazilian firms range from 7-12% of equity (revalued for monetary adjustments). The forecasts assume dividend payments equivalent to 10% of the equity at the start of each year. Even with this level of dividends, the Company is still expected to accumulate significant funds for new investments. The debt/equity ratio improves rapidly from 60:40 in 1979 to 38:62 in The profit and cash breakeven points do not exceed 62% and 52% respectively. On reaching 90% capacity utilization in 1981, the project could still show a profit breakeven operation even if phosphate rock prices increased by 50% or if product prices decreased by 15% (Annex 6-7).

33 VALEFERTIL - Summary of Projected Financial Statements (In Millions of Current US Dollars) (1/2 year) Capacity Utilization 50% 75% 90% 90% Income & Cash Flow Statements Fertilizer Revenue Operating & Financial Charges (Net) Income Taxes (30%) Net Profit (Loss) (19.6) Debt Service Coverage Ratio Dividends Balance Sheet as of Dec. 31 Current Assets Net Fixed Assets Other Investments and Surplus Cash Current Liabilities 46.5/a 30.2 /a Long-Term Loans Equity Total Assets or Liabilities Current Ratio Debt/Equity Ratio 60:40 52:48 45:55 38:62 Net Profit as % of Equity (16.8) Breakeven Capacity Utilization (%) Profit Breakeven Cash Breakeven /a Including short-term loans to initially finance part of the accounts receivable. B. Financial Rate of Return 6.06 The cost/benefit streams for the financial rate of return are expressed in constant 1976 US dollars (Annex 6-10). The return is satisfactory, 14% before and 11% after income taxes. Sensitivity tests (Annex 6-10) show that under moderately adverse conditions, such as a 10% increase in capital costs, or a 5% reduction in ex-factory prices, or a 10% increase in operating costs, the after tax rate of return will still be at least 8%. Should VALEFERTIL's steady state production be 100% rather than the 90% capacity utilization assumed here as a result of the potentially higher capacity in the phosphoric acid plant (para 4.01), the financial rate of return improves to 12.8% after taxes.

34 C. Financial Covenants 6.07 VALEFERTIL has agreed to follow prudent financial practices, and, after the start of commercial operation, observe the following: (a) maintain a debt/equity ratio of 60:40 or better; (b) maintain a current ratio of 1.2:1.0 or better, (c) not incur any long-term debt if its future debt service coverage ratio will fall below 1.5; (d) limit its dividend or other cash distributions (except for dividends on preferred shares), or equity investments and/or loans to subsidiaries (if any), or prepayment of any debt if such action would cause its current ratio to fall below 1.5; and (e) until December 1985, consult with the Bank before undertaking any expansion program or any other investment not related to the project exceeding USe5 million annually. The Company's accounts will continue to be audited by independent auditors acceptable to the Bank (para 2.04). D. Major Risks 6.08 The Tapira phosphate rock will be mined, beneficiated and used in a fertilizer plant for the first time. Although CVRD has extensive iron ore mining experience, it and VALEP do not have prior experience in phosphate rock mining and beneficiation. Moreover, neither CVRD nor VALEFERTIL has previous experience in the phosphate fertilizer field. Thus, there are inherent uncertainties as to the final performance of the processes and the interdependent operations in the mining-beneficiation-fertilizer production stages. Several technical risk areas have been identified and discussed in Chapter IV. In summary, the technical risk areas are: (a) VALEP's mining, beneficiation technology and plant capacity (para 4.05 to 4.07); and (b) VALEFERTIL's phosphoric acid plant design and the Company's technical experience and organization (paras 2.04 and 4.09). As indicated earlier, the risks in all these areas have either been reduced or are to be reduced to acceptable levels on the basis of actions taken or to be undertaken by VALEP and VALEFERTIL. For example, the phosphoric acid plant has been designed with additional safety factors to ensure that the design capacity will be achieved and VALEFERTIL will engage Technical Advisers to strengthen its management Market risks are minimal as the project is favorably located to a market area where fertilizer consumption is already substantial and demand is expected to grow at a minimum rate of 10% annually. The project's output is expected to be easily absorbed by the market. In addition, the Government's on-going fertilizer market and marketing study (para 3.13) should provide the basis for the development of the necessary distribution network and pricing policy at the farm level that can support this projected increase in fertilizer use. Product pricing is not a serious risk as the Government has so far maintained and intends to continue a policy of allowing manufacturers to earn a reasonable return by permitting prices to rise in line with increases in production costs. The Government has reconfirmed that it will continue the policy of allowing efficient producers to earn a reasonable return on investment.

35 When the CDI (Council for Industrial Development), a government agency under the Ministry of Industry, formally approved the VALEFERTIL project in October 1976 thereby granting it an 80% exemption from import duties, it imposed the condition that VALEFERTIL give priority consideration to the use of ARAFERTIL's Araxa rock in any expansion program 1/. It is likely that ARAFERTIL would have implemented its own fertilizer project long before VALEFERTIL will be in a position to invest in an expansion program, thereby relieving VALEFERTIL of the CDI condition. Nonetheless, the CDI condition poses a minor risk in that instead of an optimal expansion of the VALEP- VALEFERTIL system, some less economic investment program may be imposed. As noted in para 6.07, the financial covenants will therefore require VALEFERTIL to consult with the Bank before undertaking any other major financial commitments. The Government has also confirmed that the CDI condition does not apply to any debottlenecking or balancing program involving the present project facilities of VALEFERTIL. VII. ECONOMIC ANALYSIS A. World Fertilizer Supply and Economic Prices 7.01 International fertilizer prices have decreased substantially from the peak and by 1977 have come down to about the same level as But the downward trend is not expected to continue; instead, over the next decade, prices (in real terms) are projected to move along a gradually rising trend line, reflecting the higher capital cost of additional capacity, the increasing cost of raw materials, and the continued growth in demand. In the economic analysis, all costs and prices during the project's operating life have been expressed in constant 1976 US dollars. In addition, a shadow foreign exchange rate, 25% higher than the official rate (i.e., Cr$14.25 = US$1.00), is used to reflect foreign trade distortions caused by import tariffs, export taxes and subsidies, as well as other forms of import restrictions and export incentives As the project's output will substitute for imports, the economic ex-factory prices of the VALEFERTIL products have been estimated on the basis of the delivered cost of imports to the market area. During the operating life of the project, the CIF prices of MAP and TSP imported into Brazil have been estimated as US$245/ton and US$195/ton respectively on the basis of the production costs plus a reasonable return on investment for new, export-oriented foreign plants built with only modest investments for infrastructure and having access to high grade phosphate rock, sulphur and ammonia, all assumed to be produced in the fertilizer (MAP, TSP) exporting country and priced at the respective FOB costs. This slightly underestimates the CIF prices as it is rare for a factory to be located in a place that exports all three raw materials. However, the resulting product prices are still about 5% to 10% higher than the prices that could be expected from the most efficient existing producers with access to nearby raw materials 1/ ARAFERTIL's Araxa deposit is located 30 km from VALEP's Tapira deposit.

36 sources and whose facilities were built at lower capital costs. But since the projected demand for phosphatic products greatly exceeds the existing capacity, the additional international supply will have to be provided from new plants. Consequently, the marginal cost of the additional phosphate fertilizer supply will have to be based on the production cost of the new plants. Port handling in Brazil and inland transport costs, converted at the shadow exchange rate, are added to the CIF price to arrive at the economic ex-factory price of US$265/ton of MAP and US$213/ton of TSP (Annex 6-1) Although the economic cost of imported phosphate rock delivered to Uberaba is about US$60/ton, the VALEP rock to be used by the project has been given an economic cost of US$45/ton, ex-pipeline terminal, on the basis of its value in its best alternative use (Annex 6-1). The VALEP project was started in 1975 (Annex 4-4), following the steep increase in the international prices of phosphate rock and fertilizers during It was also apparent, especially when international prices began to decline by the end of 1975, that it would be more economical to process the VALEP rock closer to the mine and the market area, rather than transporting the rock all the way to the coast for processing and then transporting the finished product back inland to the market. Furthermore, as Brazil needed additional phosphoric acid and phosphatic fertilizer capacity, this additional capacity could optimally be located close to the phosphate deposit. Without the VALEFERTIL project, the VALEP project would still proceed but its output would have to be sold to existing producers in Sao Paulo. Because of the transportation cost, the VALEP rock would have an ex-pipeline terminal price of only US$45/ton if it were to be competitive with imported rock delivered to the coastal manufacturers The estimated economic cost of production is shown in Annexes 6-2 and 6-3 using either the shadow or the official foreign exchange rate. The economic production cost indicates the large share (83%) of raw material costs in the total economic production cost. VALEFERTIL - Economic Production Cost in Constant 1976 US Dollars /a US$/Per Ton (bulk) Annual Cost-Combined MAP/TSP MAP TSP In US$ Million % Variable Costs Phosphate Rock Sulphur & Ammonia Others Sub-Total Fixed Costs Total /a At 90% capacity utilization and the shadow exchange rate.

37 B. Economic Rate of Return 7.05 At 90% capacity utilization, the project will produce 300,000 TPY of P 0 with an ex-factory value of US$152 million annually (in constant US dollars) using the financial prices and the official foreign exchange rate. The direct net foreign exchange savings of the project is about US$65 million annually before provisions for principal and interest payments on the foreign loans. For the economic analysis, three adjustments have been made on the financial costs and benefits to arrive at the economic costs and benefits, namely: (a) an economic rock cost of US$45/ton compared to a financial price of US$55/ton; (b) the use of a shadow foreign exchange rate; and (c) the elimination of import duties and taxes in the economic costs. The economic value of the project's output is about US$144 million annually (in constant 1976 US dollars) and the economic rate of return is satisfactory at 22% (Annex 7-1). The return justifies the high priority accorded by the Government to the expansion of the local fertilizer production capacity. The project's import substitution role, its economies of scale, as well as its favorable location with respect to both the source of phosphate rock and the market, account for the satisfactory rate of return. Sensitivity tests (Annex 7-1) indicate that the return would still be about 15% even if operating costs increase by 10% and product prices decline by 10%. C. Economic Analysis of the Combined VALEFERTIL and VALEP Projects 7.06 The import substitution effort of Brazil in the phosphate fertilizer area not only involves the domestic production of finished products but also includes the production of phosphate rock concentrates to replace raw material imports. Thus, the VALEFERTIL project is just the final stage of a two-stage import substitution process, the first stage being the VALEP project which will produce phosphate rock concentrate containing 35% P 05 from the low grade (8% P ) phosphate deposit at Tapira. The economic anaiysis of this two-stage process is done by treating the two projects as an integrated undertaking. The analysis shows that the economic rate of return of the combined projects is 16.6% (Annex 7-2). This rate is similar to the economic rate of return of the VALEFERTIL project if it were designed to use imported phosphate rock instead of local rock. Should both VALEP and VALEFERTIL produce at 100% rather than the assumed 90% capacity utilization, the economic rate of return of the combined ventures would improve to 18.4%. Thus, in addition to ensuring a dependable source of fertilizer supplies, Brazil also has economic justification not only to produce phosphate fertilizers, but to develop its low grade phosphate deposits; provided the deposit and the phosphate fertilizer facilities are both located close to the market area and take full advantage of the savings in transport cost to offset the high production cost of the phosphate rock concentrate. The importance of avoiding transportation costs is illustrated by the low economic rate of return, only 9.6% for the VALEP project, if there were no VALEFERTIL project and VALEP's output had to be sold to manufacturers in the coastal area. The combined projects have a direct net

38 foreign exchange savings of about US$100 million annually (in constant 1976 US dollars) before provisions for principal and interest payments on the foreign loans. D. Other Benefits 7.07 The Government has been encouraging the location of industries away from the congested Sao Paulo-Rio de Janeiro coastal corridor and towards the less industrially developed regions such as the Northeast and the western section of the Center region. VALEP's low-grade phosphate deposit is located in the industrially less developed area of the Center region. The exploitation of this deposit and installation of the VALEFERTIL phosphate fertilizer plant is an important element of the regional industrialization objectives of Brazil. In addition, the project will accelerate the development of the fertilizer marketing and distribution system in that part of Brazil, thereby providing support to the agricultural development programs in that area The project will also contribute significantly toward increasing Brazil's self-sufficiency in fertilizers and help reduce the country's reliance on cyclical and erratic international supply sources and prices. Brazil should then be able to pursue its agricultural production growth objectives with more certainty about the supply of fertilizers as well as more stable domestic fertilizer prices. Agriculture, both for domestic consumption and exports, will continue to be a major component of Brazil's economy and an important source of growth in economic output During commercial operations, the project will provide about 450 jobs in production 1/ and a much larger number in the subsequent distribution network. During construction the project will employ about 3,000 workers at the peak level and provide additional jobs in the engineering and equipment industries. Although this cannot be accurately estimated, the impact can be gauged by the dollar value of the local engineering and other consultancy services to be provided by NATRON (para 4.08), the management consulting firm (para 2.03) and the marketing consultants (para 3.17) at a cost of about US$17 million (61% of the total cost for services); and the expected value of the locally supplied equipment of about US$56 million which is about 54% of the total equipment value on an FOB basis. VIII. AGREEMENTS 8.01 Agreement was reached on the following: 1. From the Government that it will: (a) Exchange views with the Bank from time to time regarding the development of the fertilizer sector including investment plans, and import and pricing policies (para 3.11); 1/ VALEP will provide another 920 jobs in its production facilities.

39 (b) Allow VALEFERTIL and VALEP to undertake any investments to balance and optimize their production (para 4.06 and 6.10); (c) Ensure that VALEFERTIL will be able to use, as necessary, other phosphate rock to operate at full capacity (para 4.07); (d) Take all steps necessary to assure that all funds, including foreign exchange, are provided as required to complete both the VALEFERTIL and the VALEP projects, and that all approvals for implementation of the two projects, including import licenses are timely provided (para 5.08); and (e) Continue the fertilizer pricing policy of allowing efficient producers to earn a reasonable return on investment (para 6.09). 2. From CVRD that it will: (a) Ensure that the plans and operations of VALEFERTIL and VALEP will be coordinated through a continuation of the present arrangement in the Administrative Council membership (para 2.02); (b) Ensure that VALEFERTIL and VALEP enter into a long-term rock supply agreement and jointly undertake a debottlenecking study of their facilities. CVRD will also ensure that if appropriate, investments to balance and optimize the production of the two projects will be made (paras 4.06, 4.07 and 5.08); (c) Ensure that funds will be available as necessary to complete both the VALEP and the VALEFERTIL projects, including, if needed, VALEFERTIL's fluorine recovery system. The funds should be provided in the form of loans and/or equity such that at project completion, each project will have a debt/ equity ratio of 60:40 or better, and a current ratio of at least 1.2:1.0 (paras 5.03 and 5.08); and (d) Assist VALEFERTIL obtain the other foreign loans needed by the project (para 5.08). 3. From VALEFERTIL that it will: (a) Engage management consultants, under terms of reference and a time schedule satisfactory to the Bank, to assist in developing VALEFERTIL's organizational structure and systems (para 2.03);

40 (b) Appoint a marketing manager in time to assist in developing the Company's organizational and marketing plans. The marketing plans will be prepared in consultation with the Bank (paras 2.04 and 3.18); (c) Jointiy undertake with VALEP a study to analyze the feasibility of increasing and/or balancing their outputs through a debottlenecking program. If appropriate, VALEFERTIL will make investments to balance and optimize the production of its plant relative to the VALEP production facilities (para 4.06); (d) Evaluate, with the assistance of its Technical Advisers, the project design to provide the flexibility of using other phosphate rock and to use such rock to operate at full capacity should VALEP be unable to supply sufficient phosphate rcck. Furthermore, VALEFERTIL will enter into a long-term phosphate rock supply agreement with VALEP which would also obligate VALEP to undertake the debottlenecking study mentioned above (para 4.07); (e) Obtain Bank approval for the potential suppliers of critical equipment both financed and not financed by the Bank (para 5.10); and (f) Follow the financial covenants enumerated in para M02 Based on the agreements reached on the above points, the project provides a satisfactory basis for a Bank loan to VALEFERTIL of US$82 million equivalent for a period of 15 years including 3 years of grace at the standard Bank rate plus a fee to the Government for a total interest rate of 10% per year. Industrial Projects Department April 7, 1977

41 ANNEX 1-1 Page 1 BRAZIL - VALEFERTIL FERTILIZER PROJECT GLOSSARY OF COMMON TERMS IN THE FERTILIZER INDUSTRY Acqua Ammonia : Solution of ammonia in water. Solubility of ammonia in water depends greatly on temperature. At 00C (32 F), solubility is about 900 grams per liter and only 7.4 g/lt at 1000C (212 F). Acqua ammonia is also used as fertilizer by dissolving anhydrous ammonia in irrigation water. AAmmonium Carbamate : An intermediate chemical formed in solution during synthesis of urea from carbon dioxide and ammonia. On reduction in pressure it decomposes into urea with the formation of byproduct water. Ammonium carbamate is a highly corrosive material and requires the use of special materials in equipment required to handle it. Ammonium Chloride : NHhCl is obtained as a co-product in the manufacture of sodium carbonate and is used as a fertilizer for some crops and types of soil. Azmonia Feedstocks are the sources of hydrcgen used for ammonia production. Except for the minor use of hydrogen produced by electrolysing water, the principal feedstocks are hydrocarbons such as natural gas, naphtha, heavy fuel oil and coal. Ammonium Nitrate (AN) or chemically NHhNO 3 is produced by reacting ammonia with nitric acid. Nitric acid (HN0 3 ), in turn, is made from ammonia. AN1/ contains 33.5% nitrogen, half in the ammonium form and half in the nitrate form, and is very hygroscopic. It also is a commercial explosive. In many countries, a diluent--e.g. limestone--is added to lower the analysis and negate the explosive properties as well as to minimize the hygroscopic properties. Mixed with small amounts of fuel oil it is known as ANFO explosives used in open pit mining. Ammonium Sulphate (AS) : Reaction product of ammonia and sulphuric acid (NH) 2 SO[. It should contain not less than 20.5% N. Anhydrous Ammonia : (NH 3 ) A gas containing approximately 82% nitrogen. Under pressure, amonia gas is changed to a liquid and is usually stored and transported in this form. Anhydrous ammonia is used to make most of the solid forms of nitrogenous fertilizers and also is used for direct application to the soil either as a gas or in the form of aqua aomonia. Bagged Fertilizer : Solid fertilizers delivered to the user in bags. Net weight of the bagged fertilizer is commonly 100 lbs, 50 kg and 100 kg. 1/ Fertilizer grade.

42 ANNEX 1-1 Page 2 Biuret : An impurity, NH (CONH 2 ) 2 H 2 0 found in commercial urea, brought about by a partial decomposition of the urea to biuret and ammonia. For certain applications, particularly when the urea is ultimately sprayed on citrus trees, the biuret content should be kept as low as possible, in the range 0.1 to 0.3 wt.%. Bulk Fertilizer : Solid or liquid fertilizer sold loosely in tankers, holds, train wagons, trucks or large containers. Anhydrous ammonia is sometimes transported in refrigerated barges or railroad tanks. ship Calcium Ammonium Nitrate (CAN) or nitro-lime is a granulated mixture of ammonium nitrate and calcium carbonate, generally containing 25% N. Calorific Value : A measure of the quantity of heat given out in the combustion of a fuel. Usually expressed as BTU/standard cubic foot in the case of a pipeline gas, where the standard cubic foot of gas is measured at 1.7 lb/sq. inch absolute pressure and 60 F. temperature and in the dry state. A BTU is the amount of heat required to raise 1 pound of water through 1 fahrenheit degree at 60 fahrenheit. The calorific value may be expressed either as gross or net. The gross value is obtained when the water vapour in the combustion products is condensed and the net when it remains as vapour. Chemical Fertilizers : Manufactured chemical compounds suitable as fertilizers should be high enough in nutrient content; stable to avoid hazards and handling problems; water soluble and available to plant root systems. It is the common practice to report the nutrient content of materials in terms of percentages of N (nitrogen), P205 (phosphorus pentoxide) 1/ and K 2 0 (potassium oxide) 2 /. Commnercially available materials meeting these requirements to a large degree are Primary Fertilizer Materials Nutrient iontent, % N P295 KpO Total (a) Urea (b) Ammonium Nitrate (AN) L (c) Ammonium Sulphate (AS) (d) Diammonium Phosphate (DAP) 18 h6-64 (e) Monoammonium Phosphate (MAP) (f) Nitro-phosphate (NP)5l" h0 (g) Triple Superphosphate (TSP) - h6-46 (h) Single Superphosphate (SSP) (i) Thermophosphate (THP) (j) Potassium Chloride (k) Potassium Sulphate / Normally but incorrectly referred to as phosphoric acid. 7/ Or potash. 3/ Available also with smaller proportions of N & P such as

43 ANNEX 1 -I Page 3 Complex or NPK Fertilizers: All three primary nutrients (N, P 2 0, K 0) are frequently applied to the soil at the same time in ratios va;ying with the nutrient requirements of different crops. Complex fertilizers contain at least two chemically combined nutrients. To facilitate handling, the several required chemicals are usually agglomerated into uniform granules for distribution. The analysis of each nutrient is given as a ratio to describe the NPK product. Thus 15_15_15 complex fertilizer contains 15% each of N, P20,. and K 2 0; and camplex fertilizer contains 12% N, 24% P and 12% t2. Diammonium Phosphate (DAP) is produced by reacting acid followed by granulation and drying. ammonia with phosphoric Dosage: Intensity of fertilizer application generally measured in kg/ha or lbs/acre (nutrient content). Fertilizer: The vehicle through which plant nutrients--with the exception Of carbon, hydrogen, and oxygen--are added to the soil. It is usual to distinguish between natural fertilizers (such as compost) and manufactured or chemical fertilizers. Very often when the term fertilizer is used alone it refers to the latter. Giammarco Vetrocoke: Proprietary name of patented carbon dioxide scrubbing process using hot potassium carbonate solution as the solvent but with the addition of small amounts of arsenious oxide to inhibit corrosion. One of the most efficient carbon dioxide removal processes but it suffers from the disadvantage of employing highly poisonous arsenic compound. Granular Fertilizer: Fertilizer in a granular form, as distinct from powdered ferti izer which has a tendency to 'bridge' in hoppers and generally to make bagging and application more difficult. Monoammonium Phosphate (MAP) is produced by reacting ammonia with phosphoric acid followed by prilling or granulation and drying. Mixed Fertilizer: A fertilizer made by the physical mixing or blending of two or more fertilizers and generally containing more than one plant nutrient. Monoethanolamine: An organic solvent made from ethanol and ammonia which has a high solubility for carbon dioxide and which can be easily regenerated. Widely used in carbon dioxide scrubbing processes particularly in cases of small to medium scale operation. Partial Oxidation is a method of producing hydrogen from hydrocarbon fuels of almost any type by a non-catalytic reaction with oxygen followed by removal of the by-product, carbon monoxide.

44 ANNEX 1-1 Page 4 Plant Nutrients: Essential to plant growth are some 17 elements, six in large and the remainder in small or micro quantities. Carbon, hydrogen, oxygen, nitrogen, phosphorus and potassium comprise the first six; the others of lesser significance include calcium, magnesium, sulphur, silicon zinc, iron, aluminum, manganese, boron, sodium, and copper. Chlorine and cobalt are also thought essential. Carbon, hydrogen and oxygen are readily available from the atmosphere and water, Nitrogen, phosphorus and potassium--three main nutrients-- and the other elements are drawn fram the soil systems. Unless supplemented by regular additions of materials containing the three main nutrients, soil becomes depleted of fertility by repeated cropping. Use of natural organic materials such as animal and vegetable wastes can be utilized but the scale and intensity of modern agriculture have far exceeded the availability of natural 'fertilizers'. Consequently, the majority of the world's primary plant nutrient needs are now supplied in the form of manufactured or 'chemical' fertilizers. To an increasing degree, secondary nutrients such as calcium, magnesium, sulphur and micro nutrients such as boron, zinc, copper and manganese are also added to soils along with the primary nutrients in ratios prescribed by agronomists according to specific crop and soil needs. Phosphoric Acid (H P0 ): An inorganic acid used in the manufacture of concentrated calcin; phosphates, ammonium phosphates and sometimes for direct application through irrigation water. Although in fertilizer terminology the term has been used practically from the inception of the fertilizer industry to denote the available phosphate (P 0 ) equivalent of a fertilizer, this usage is not strictly correct and has become particularly confusing in recent years since phosphoric acid H P0 is being used as such in the manufacture of fertilizer. Phosphoric kcid (containing 52-54% P 2 05) is generally produced by reacting phosphate rock with sulphuric acid. Phosphate Rock: Rock containing phosphates mainly of calcium in generally complex forms. Rock phosphate can occur in igneous (formed by solidification of molten magma) or sedimentary (formed by deposit of sediments) forms. Some deposits contain the phosphates as nodules scattered in sterile gangue. It is used to produce phosphoric acid (HP0h) as well as superphosphates. Potash: A term used to denote the potassium oxide (K20) equivalent of materials containing potassium. Actually, K 2 0 as such is never found in fertilizer, but historically the potassium in fertilizer has been thus expressed. PPM (Parts per million): The concentration of a substance expressed as th; number of units of that substance per total million units. Prilled Fertilizer: Fertilizer in spheroidal particles generally between 1 and 4 mm in diameter.

45 ANNEX 1-1 Page 5 Prills: Spherical particles prepared by spraying molten material downward through a countercurrent stream of cooling air; this operation is performed in a prilling tower. The particles obtained have excellent storage and handling properties in th_t they are hard and have less surface area than crystals. Scrubbing: Process of removing dust and/or vapor from an inert gas otream by using an absorbing liquid. Always required in fertilizer plants to remove ammonia, fluorine, and fertilizer duot. Single Superphosphate (SSP): A product obtained by treating phoaphate rock with sulphuric acid and consisting of a mixture of monocalcium phosphate CaHI (P0O dicalcium phosphate CaHPOh, tricalcium phosphate Ca (PO0)2, cal!itum su2phate ana other impurities. Tricalcium phosphate is lot aly insoluble and is not includea in the P 2 05 content of the auperphosphate (i6% to 18%). Sulphuric Acid (H 9 SO 1 ; Oil of Vitriol): This strong acid is made from various sulphur materials. SIlphuric acid is a common acid, and is used in the manufacture of fertilizers such as ammonium sulphate and superphosphate. Triple Superphosphate (TSP): Obtained by reaction or phosphate rock and phosphoric acid. Since it contains no calcium sulphate, its P 2 05 content (about 46%) is much higher than in single superphosphate. Urea is known chemically as carbamide or IH 2 CONH -- the normal amide of carbonic acid; this compound contains about 46%N, all in tne ammonium form. It is considerably less hygroscopic than ammonium nitrate and it is the most widely used straight nitrogen fertilizer today. Urea Synthesis: Urea is made by reacting ammonia with carbon dioxide. Since both of the materials are produced during the ammonia synthesis, urea prodluction is usually undertaken alongside an ammonia plant. Unfortunately, the corresponding acid oi carbon dioxide (carbonic acid) does not form stable anmonium ddlts as do nitric, sulphuric or phosphoric acids. Therefore, simple neutralization as used in making ammonium nitrate and sulphate fertilizers is not possible. Instead, ammonia and carbon dioxide are combined under heat and pressure to make ammonium carbaxiate which, although unstable, can be dehydrated under pressure to form urea, a stable compound. Again, unfortunately, the overall reaction is reversible and even at 300 atmospheres pressure and 2000C, the conversion to urea in a single pass through the reactor is under 70%. This introduces several complications: high pressure must be used to maximize conversion; unconverted reactants must be separated and recycled; increased corrosion under the high temperatures and pressures used must be overcome, and urea decomposition into undesirable products must be minimized. In recent years, several engineering and producing companies have developed ways of surmounting these problems, and large plants capable of producing a thousand tons per day, or more, of urea to rigid chemical and physical specifications have now been operating successfully for long periods throughout the world. Industrial Projects Department January 19(7

46 ANNEX 1-2 Page 1 BRAZIL - VALEFERTIL FERTILIZER PROJECT ULTRAFERTIL FERTILIZER PROJECT 1. Industria e Commercio de Fertilizantes S.A. (ULTRAFERTIL) ULTRAFERTIL was established in 1966 as a joint venture between Phillips Petroleum, a US petroleum and fertilizer company, the Brazilian Ultra group and IFC. In 1974, because of financial difficulties, Phillips sold its shares to PETROQUISA and ownership of the company now is 85% with PETRDQUISA, 10% with the Ultra group and about 5% with IFC. The company's facilities at Cubatao (Sao Paulo) include a 450 ton per day naphtha-based ammonia plant with auxiliary facilities to produce ammonium nitrate and diammonium phosphate fertilizers. Port-handling facilities for ammonia and fertilizer materials are also included. At present the company is completing an ammonia storage tank of 20,000 tons capacity and is altering the ammonium nitrate plant to produce calcium ammonium sulphate in the future. Together with its production facilities at Cubatao, ULTRAFERTIL established 14 distribution centers intended to handle about half of the company's total production. The centers are located in south central Brazil, the market area for ULTRAFERTIL. However, soon after production had started in late 1970 it became apparent that the distribution centers were overdesigned and only six of them are now operating, the others being used for storage. Production has been reasonably stable with capacity utilization for the whole plant averaging about 90% over the last three years: ULTRAFERTIL - HISTORICAL PRODUCTION ( in 1,000 tons) Product Ammonia Nitric Acid Ammonium Nitrate Sulphuric Acid Phosphoric Acid DAP MAP

47 ANNEX 1-2 Page 2 With a total N production of around 90,000 TPY, ULTRAFERTIL is presently the largest producer of N in Brazil. ULTRAFERTIL's production equals about 56% of Brazil's domestic production of N and 21% of her consumption in Because ULTRAFERTIL also mixes and blends fertilizer materials, part of which is imported, its sales are very diversified as shown below: ULTRAFERTIL- SALE HISTORY (in 000 tons) Products ( ~~~~~~tomo 6 ) Ammonia for Sale Nitric Acid Ammonium Nitrate Sulphuric Acid Phosphoric Acid DAP Simple Superphosphate Triple Superphosphate Potassium Chloride Ammonium Sulfate Nitrochalk Urea MAP Complex Fertilizers TOTAL The sales volume has been rather stable over the last three years, and 1976 sales are expected to follow the same trend, although sales of complex fertilizers are lagging behind. ULTRAFERTIL's financial performance has been less than satisfactory (Table 1). In 1974 the company made, after three years of operation, its first profit of US$21.5 million, which reduced the accumulated deficit from US$33.2 million at the end of 1973 to US$11.7 million at the end of Profitability in 1975 was marginal with net income being only about 1% of net revenues and a sizeable loss of around US$10 million is projected for As a consequence of the poor performance, the debt/equity ratio deteriorated to 95/5 in 1973, but has improved in 1974 and 1975 to about 60/40. The debt service coverage has been only between 0.3 and 0.5 except in 1974 when it was satisfactory with 1.3. Additional long-term debt and share capital contributions from PETROQUISA provided the necessary funds to keep ULTRAFERTIL solvent.

48 ANNEX 1-2 Page 3 The reasons for ULTRAF3RTIL's poor financial performance are manifold. While the company's production performance has been satisfactory, production costs have been high caused partly by built-in inefficiencies in the plant and partly by high overhead. Also, ULTRAFERTIL initially had difficulties obtaining Government approval for price increases for its products. In addition, ULTRAFERTIL is affected by the import quotas, which are established at times of low international fertilizer prices to protect the domestic industry. These quotas, however, are dependent on consumption and production forecasts which are never completely accurate. As a result, there might be an excess of imports, which penalizes higher-cost domestic producers such as ULTRAFERTIL or not enough imports which restricts overall growth of fertilizer consumption. Industrial Projects Department January 1977

49 ANNEX 1-2 Table 1 BRAZIL - VALEFERTIL FERTILIZER PROJECT HISTORICAL FINANCIAL STATWENTS OF ULTRAFERTIL, S.A. (In Million US Dollars) For the Year Ending Dec Net Revenues Cost of Sales Selling and Administrative Expenses Financial Expenses Other Charges Net Income (16.3) ( 3.2) Sources of Funds Current Operations (4.1) Decrease in Deferred Charges Capital Subscription o 2.5 Increase in Long-Term Debt Total Sources Uses of Funds Capital Assets Invest1;ents o.6 Decrease in Iong-Term Debt Increase in Working Capital 4.8) o.8 Total Uses Current Assets Long-term Investments & Receivables Capital Assets (Net) Deferred Charges Current Liabilities Long-Term Debt Capital Stock j4.4 Reserves f4 Accumulated Deficit (30.0) (33.2) (11.7) (10.5) Total Equity Total Assets or Liabilities o Long-Term Debt/Equity Ratio 90/10 95/5 68/32 62/38 Current Ratio Debt Service Coverage Exchange Rate Cr.$/US$ Industrial Projects Department Ja uary 1977

50 BRAZIL - VALEFERTIL FERTILIZER PROJECT VALEFERTIL CORPORATE ORGANIZATION CHART GENERAL (STOCKHOLDERS) ASSEMBLY OADMINISTRATIVE COUNCILRT (FOUR MEMBERS) OARD OF EXECUTIVE DIRECTORS l ~~(FOUR MEMBERS} LEGAL COUNCIL PRESIDENT & DIRECTOR Mr.Hygino Antonio Baptiston f PLANNING AND DEVELOPMENT FINANCE DIRECTOR & ADMIN. PROJECT IMPLEMENTATION OPERATIONS (TECHNICAL) FINANCE & ADMIN. DIRECTOR DIRECTOR DIRECTOR Mr.Rogerio Gragnani Leite Mr. Renato Gomes Moretzsohn Mr. Marcos de Lima Fernandes FINANCE MANAGER Mr.JLC_s MANAGER MARKETING MANAGER Mr. Luis Carlos Lima do Nascimento (Vacant) COST CONTROL MANAGER ADMIN. MANAGER VVorld Bank Industrial Projects Department, March 1976

51 ANNEX 2-2 BRAZIL-VALEFERTIL FERTILIZER PROJECT LIST OF COMPANIES OF CVRD GROUP Company Activity Capital In % Of Ownership CR$ 1 OOO Held By CVRD 1, Main Company Coa4panhia Vale Rio Doce Iron Ore Mining 4,011, Subsidiaries 2.1. ValeRio Doce Navegacao DOCENAVE 2.2. Mineracao Vale do Paranaiba Ocean transportation 158, VALEP Phosphate mining 130, Fertilizantes Vale do Rio Grande-VALEFERTIL Fertilizers 76, Florestas Rio Doce - FLORESTAS Forest resources 56, Rio Doce Geologia e Mineracao DOCEGEO Rio Doce Engenharia a Geological survey 9, Planejamento - RDEP Planning & Engineering 2, Navegacao Rio Doce Ooean transportation 1, * 2.8 Rio Doce Europa - RDE Commercialization 1, Itabira International Co. - ITACO Commercialization Seamar Shipping Ocean transportation ** 2.11 Caraca Ferro e Aco Iron Ore mining 11,5o Mining Subsidiaries (several) Mining Associated Companies 3.1. Amazonia Mineracao Iron Ore Mining 302, Celulose Nipo Brasileira - CENIBRA Pulp & paper 242, Cia. Nipo-Brasileira de Pelotizacao - NIBRASCO Iron ore pellets 110, Cia. Italo-Brasileira Pelotizacao -ITABRASCO Iron ore pellets 73, Empreendimentos Florestais - FLONIBRA Forest resources 76, Mineracio Rio do Norte Bauxite mining 49, o 3.7 Cia Hispano-Brasileira Pelotizacao Iron ore pellets 45, Itavale Ltds Iron ore mining 6, VALENORTE-Servicos de Engenharia Engineering * the remaining 50% are held by DOCENAVE, a CVRD subsidiary. ** the remaining 99.9% are held by DOCENAVE, a CVRD subsidiary. Industrial Projects Department January 1977

52 ANNEX 2-3 For the Year Ended December 31 Income Statement BRAZIL-VALEFERTIL FERTILIZER PROJECT CVRD FINANCIAL STATEMNTS (In US$ millions) (estimate) Gross Revenues Cost of Sales Administrative and Selling Expenses b/ Financial Charges b/ Provision for Income Tax Net Income Sources of Funds Current Operations Capital Subscriptions,etc Loans Uses of Funds rean ~Repa~yments Capital Investments Additions to Working Capital Dividends and Others Total Sources or Uses Assets & Liabilities Current Assets Long-term Assets Fixed Assets (net) Current Liabilities Long-term Debt Eauity Total Assets or Liabilities Long-term Debt/Equity Ratio 29:71 31:69 37:63 Current Ratio Net Income as % of Equity (year end) Exchange Rate Cr$/US$ a/ This is only for CVRD and is not a consolidated b/ Included in the cost of sales. statement. Industrial Projects Department January 1977

53 BRAZIL - VALEFERTIL FERTILIZER PROJECT CVRD - FUNDS FLOW FORECAST (In US$ millions) Actual Estimate Forecast- A. Source of Funds Internally generated funds 2. Externally provided funds lJ 3h L5.7 a. Loans h b. New capital, Governnent Funds and others h L Total Sources T T B. Use of Funds 1. Capital Investmentsb/ h.2 h56.i 783.1h Loan repayments, Dividends, Working Capital and Others 19.h (-' Total Uses h Exchange Rate Cr$/US$ h llAt ll.h 11.L a/ Forecast for are in October 1976 prices and exchange rates. b/ The capital investment program of US$3.2 billion ir constant 1976 US dollars for is equivalent to about US$3.8 billion in current terms. Industrial Projects Department January 1977

54 ANNEX 3-1 Page 1 BRAZIL - VALEFERTIL FERTILIZER ROJECT AGRICULTURE IN BRAZIL A. Development and Government Policy 1. Over the period , the agricultural sector has performed well in supplying domestic demand for food and fibers while producing a rising volume of export commodities, with output growing at an average of 5.5% annually despite adverse climatic conditions in four out of nine years./ The agriculture sector's contribution to NDP has declined from 20% in 1967 to about 15% at present, but its contribution to exports has remained significant and currently accounts for over 50% of the total value of exports. Because of the seriousness of the balance of payments constraint, the strategy of Brazil's Second National Development Plan ( ) aims at a more efficient development of Brazil's agricultural potential. Besides its favorable impact on the balance of payments, the new growth strategy also has favorable implications for domestic price stabilization and incom6 distribution. 2. Until 1970, growth in agricultural production was attributable almost entirely to expansion of land under cultivation. Little if any effort was expended on improvement of productivity in agriculture. Inspite of having been relatively neglected by policy makers, the agricultural sector managed to grow at about 4.5% annually during Inflation, price controls, erratic Government intervention in marketing of products, overvalued exchange rates and export controls, tended to distort prices particularly in the early 1960s, discouraging production and exports. The movement of agricultural activity into the interior of the country created a need for an elaborate infrastructure to handle product distribution which remained largely neglected until'the mid-1960s. Changes in overall economic management since then have helped to eliminate some of these distortions and improve incentives for farmers. Recent growth has tended to be higher, although there have been fluctuations due to climatic factors. Brazil has become the world's largest producer of sugar and its second largest supplier of soybeans, while remaining the world's leading coffee producer. Coffee continues to play a very important role in the agricultural sector accounting for about 12% of output, though its share in total exports has declined from 4h% in 1967 to about 11% in It is expected to rise to almost 17% in 1976 despite a lower volume of exports than 1975 due to a large price increase. Annual variations in coffee production, largely resulting from frosts, are generally responsible for year-to-year fluctuations in the growth of the agriculture sector. The 1975 frosts which caused extensive crop damage in the states of Sao Paulo, Parana, Minas Gerais and Mato Grosso, will result in an estimated 70% decline in coffee production in 1976 and will reduce overall growth of the agriculture sector from an estimated 4% in 1975 to between 0-2% growth in / Tables 1-4 of this Annex gives some basic agricultural statistics about development of area, production and yield of the country's main crops.

55 ANNEX 3-1 Page 2 3. The agricultural policies of the Government have multiple objectives: they are designed to stimulate output to meet the increasing demand of Brazil's population, to encourage production for export to make a substantial contribution to foreign exchange earnings; and, as instruments for raising income levels of the rural poor. The Government's price stabilization objectives depend upon the extent to which agricultural production meets the increases in domestic demand. More balanced and abundant increments in production relative to population increase would alleviate the commodity scarcities which have complicated stabilization efforts in recent years. The agriculture sector in Brazil, besides employing relatively little in way of imported inputs, has enormous potential as a supplier of primary commodities to a world increasingly afflicted by food shortages. Agriculture is and will continue for some time to be a key generator of foreign exchange. The aim of the Government is to make Brazil one of the world's foremost producers and exporters of foodstuffs. Finally, with almost 45% of the population engaged in agricultural activities, a more balanced and equitable development process can only be achieved with comparatively higher increases in agricultural production and productivity than those experienced in recent years. Substantial increases in per worker productivity in the rural areas can alleviate poverty where it currently is most heavily concentrated. The Government is thus giving more attention to the modernization of agricultural production and marketing and expansion of the frontier of cultivation. 4. Brazil has the potential for sustaining relatively high long-term growth in agriculture and the Plan target of a 40% (7% annually) real increase in agricultural output for the period appears feasible. To support these growth objectives, the Government has allocated substantial public resources to programs to incorporate new areas to agricultural production. Brazil is one of the few countries in the world that still has available abundant land on which continued expansion of output can be based, so that reasonably high growth is possible if accompanied by modest iprovements in agricultural technology. The goal of these programs is to expand agricultural area by 5% annually and to encourage settlement to bring these areas into production as quickly as possible. A second component of the Government strategy is to increase productivity and production options through modernization of agriculture research and increase domestic production of improved seeds, fertilizers, and substantial expansion in the supply of credit and complementary technical assistance for the purchase of modern agricultural inputs. These inputs should bring about a 2% annual increase in productivity necessary to achieve the overall growth target for the agriculture sector. Government policies and programs are already stimulating considerable progress both in expanding land under cultivation and increasing productivity of activities for domestic and external markets. 5. Following the example of the comprehensive programs created in the early 1970s for the development of Northeast agriculture (PROTERRA), the Amazon (PIN), Mato Grosso and Goias (PRODESTE), Valley of the Sao Franciso River (PROVALE), the Government in 1974, launched a program (POLOCENTRO) to focus on the technical and infrastructure requirements for initially exploiting 3.7 million hectares of land area in the Campos Cerrados area. This

56 ANNEX 3-1 Page 3 program will also finance agricultural research, feeder roads, rural electrification and agricultural storage facilities. The Government has also initiated a program (POLONORDESTE) designed to improve small farmer productivity in the Northeast through the formulation and execution of integrated rural development projects. This constitutes a significant shift in the emphasis of governmental programs which in the past have concentrated on expanding socially oriented programs in the fields of education, health, nutrition, sanitation and broadening the application of social security and labor legislation rather than on production oriented programs aimed at low income groups. These regional programs are designed to bring about a fuller utilization of the country's land and related resources in order to provide a means for the alleviation of poverty and underemployment, which are particularly pressing in the Northeast. 1VAiile no overnight eradication of rural poverty, which has been endemic in the Northeast for decades, is likely to be forthcoming from these programs, responsiblefforts to improve the productivity of Northeast agriculture together with programs to increase industrial activity in the region already have resulted in a more rapid growth of per capita income in the Northeas than in Brazil as a whole, and with the initiation of new programs this process is expected to continue under the Second Development Plan. 6. Considerablemphasis is also being given to agricultural research and to the dissemination of research findings through a national extension network. The Brazilian Agricultural Research Enterprise (EKBRAPA) and the Brazilian Technical Assistance and Rural Extension Enterprise (EMBRATER) have been strengthened to carry out these objectives. Guaranteed minimum price support and subsidized credit for financing acquisition of modern inputs are being emplcyed to encourage farmers to take advantage of technological improvements resulting from these research and extension activities. A program of agricultural infrastructure improvements (Export Corridors Program) is being implemented in the South and Center regions to facilitate movement of exportable agricultural surpluses from the interior of Brazil through the major ports of the region (Vitoria, Santos, Paranagu and Rio Grande). 7. Since 1965, with the creation of the National System of Rural Credit, the Government has made a continuous effort to increase the amount of loanable funds for agriculture. Agricultural credit is crucial to resolution of the challenges confronting Brazilian agriculture, particularly expansion of land under cultivation and use of modern inputs. Currently, about 27% of total Brazilian bank credit is directed into agriculture pri.ncipally by the Bank of Brazil and through Central Bank refinancing programs using the private commercial banks as agents. Agricultural credit as a percent of Brazil's agricultural domestic income has increased from about 23% in 1966 to over 80% in 197L. Agricultural credit lines are available for a multiplicity of purposes including working capital, acquisition of modern inputs, irrigation, land acquisition by small holders, retention of breeding stock, pasture improvement, etc. Agricultural credit is extended at subsidized interest rates differentiated in favor of the small holders and of development in certain regions, particularly the Northeast. The Government has also established an insurance scheme or fund to cover debts of farmers

57 AN1.X 3-1 Page 4i when affected by natural disasters. This Insurance Program for Agricultural and Livestock Activities (PROAGRO) has been extremely useful in reducing risk of credit operations both to lenders and borrowers. The Government recently extended this coverage to 80% of the losses in Parana guaranteed under the PROAGRO program when the State was devastated by the severe frosts of mid A major difference between the approach of the Second National Development Plan and efforts heretofore made to improve Brazil's agricultural performance lies in better coordination of the activities of the many agricultural development agencies. Institutional changes made since March 1974 suggest that for the first time there is a chance for effective coordination between agro-policy makers, agricultural research and extension agencies, agrarian reform and regional development agencies. B. Major Export Crops 9. Three agricultural commodities, coffee, soybeans and sugar, generate over 35% of total export earnings and, including unprocessed or semiprocessed agricultural exports, the agricultural sector accounts for more than 50% of total value of exports. The diversification of agricultural exports, which over period has been more important than even the spectacular growth of manufactured exports, is all the more remarkable since, unlike manufacturers, they benefit little from fiscal incentives. 10. Brazil is the largest exporter of coffee in the world although its share of the world market has declined from 50%in_ the early 1950s to an estimated 20% in With the exception of the harvest, domestic production has been consistently below requirements for domestic consumption (8 million bags) and exports (18 million bags), leading to a gradual depletion of stocks. Due to a large build-up of stocks and the tremendous excess of production capacity in 1966, the Government initiated a coffee tree eradication program. However, the movement of coffee planting into the interior of the state of Parana by the early 1960s had left substantial portions of Brazil's production capacity exposed to frosts. After the frosts of 1969 which reduced 1970 crop to less than 11 million bags, the Government was forced to reverse its policy and establish price and credit incentives for new tree planting. The impact of the program was felt in 1974 with production of some 26 million bags. The severe frosts of 1975, which drastically reduced production capacity with an estimated production of 6 million bags in 1976, may start a new phase in Brazilian coffee history, and coffee production may gradually move out of frost-prone Parana. The danger of movement into the over-production stage of the coffee cycle still exists, though greater attention on the part of the Government and the declining comparative advantage of coffee over other crops tend to forestall this. The increased cost of fertilizer and pesticide, and the emerging shortage of agricultural labor in the coffee regions discriminates severely against coffee since no mechanized means of coffee harvesting has ever been developed. Continuation of a moderate coffee tree planting

58 ANNEX 3-1 Page 5 program at moderate levels should maintain Brazilian production capacity at adequate levels, and should permit the Government to recoup the decline in Brazil's share in world exoorts that occurred in Wiith the recent decline in Cuban production, Brazil is now the largest producer of sugar in the world. Until 1970 sugar production increased very slowly but, in response to rapidly rising world prices, large increases took place in 1972 and Due to unusually bad weather (drought and frosts in Center-South and heavy, unseasonable rains in the Northeast), production declined in 1974 and 1975, and despite exhaustion of its sugar stocks early in 1974, Brazilian sugar export in that year surpassed coffee as Brazil's principal foreign exchange earner. Most of the increases in cane production have been due to expansion in harvested areas, with practically no increases in average yields. Increases in Brazil's exportable sugar surplus (from 1.1 million tons in 1968 to peak of 2.8 million tons in 1973 with 1975 exports of only 1.7 million tons) have been facilitated by a Government program to reform the production and processing of sugar in the Northeast. This reform has various elements including the creation of credit lines to merge and re-equip sugar mills and to integrate and reallocate cane production and milling; the adoption of a uniform sugar price throughout the country; the establishment of a gradually declining subsidy to be eliminated over a six-year period, to compensate for the currently higher production costs of the Northeast; and the allocation of PROTERRA funds to finance the re-employment of cane and mill workers displaced by productivity improvements. Additional planting of land area in sugarcane and improvements in productivity are expected to increase sugar production from about 7.7 million tons at present to about 10 million tons by 1980, out of which only 23% would be for exports due to growing requirements for domestic consumption. 12. The production of soybeans has increased from less than 1 million tons in 1968 to an estimated 11 mi=lion tons in 1976, making Brazil the third largest producer in the world, behind the U.S. and the People's Republic of China. Soybeans represent the most startling example of the speed at which new products have been added to the export flow as exports of soybeans, soybean meal and soybean oil increased from US$40 million in 1967 to US$1,302 million in This vast expansion has taken place mainly in the states of Sao Paulo, Parana, and Rio Grande do Sul and has been instrumental in restructuring southern Brazilian agriculture. Up to 1975, most of the expansion came from double-cropping of wheat areas and cultivation of natural pastures and thus has not provoked any significant reduction in the output of alternative crops. In addition, per hectare productivity has increased and at present Brazilian productivity is only slightly lower than in the U.S. and Canada, and higher than in Japan. Most of the Southern region presents favorable conditions and sufficient land available for further expansion of soybean cultivation. The area under soybean cultivation is expected to increase from 5.2 million hectares in 1974 to 10.5 million hectares by 1980, and production

59 ANNEX 3-1 Page 6 is expected to reach 18 million tons, with exportable surplus at million tons compared to 6.7 million tons in In terms of production capacity, soybeans clearly constitute one of Brazil's principal opportunities for expansion of agricultural exports. Industrial Projects Department March 1977

60 BRAZIL - VALEFERTIL FERTILIZER PROJECT GRO)WTH RATES OF SECTORAL ANqD GROSS DSATIC PRODUCT, (In Constant Prices) Growth Rate in % By Sector Agriculture Lt Industry Services h GDP L, Source: Fundacao Getulio Vargas. Industrial March 1977 Projects Department CD1

61 BRAZIL - VALEFERTIL FERTILIZER PROJECT EVOLUTION OF AREA UNDER CULTIVATION (1000 Hectare) CROPS Cereals Rice Corn Wheat Total Oil Crops Groundnuts Soybeans Total Fruit Pineapple Banana Orange Total: Fibers Cotton Other Crops Coffee Cacao Beans Manioc Sugar Cane Other N.A. N.A. N.A. Total All Crops % Change Source: Years are actual and 1974 are estitates as per Anuario Estatistico do Brazil; as per Ministry of Agriculture (SUPLAN) forecasts. Industrial Projects Department March 1977

62 BRAZIL -VALEFERTIL FERTILIZER PROJECT EVOLUTION OF AGRICULTURAL PRODUCTION OF SELECTED PRODUCTS (In 1000 Tons) CROPS Cereals Rice Corn Wheat Total Oil Crops Groundnuts Soybeans Total Fibers Cotton Other Crops Coffee Cacao Beans Manioc Sugar (Centrifuge) Total Grand Total % Change Source: Anuario Estatistico do Brazil Ministry of Agriculture (SUPLAN) Years Instituto do Acucar e do Alcool (IAA) Paiva, Schattan, Freitas "Setor Agricola do Brasil" Industrial Projects Department March >

63 BRAZIL - VALEFERTIL FERTILIZER PROJECT EVOLUTION OF AGRICULTURAL YIELDS (Kg/ha) CROPS Cereals Rice 1,552 1,634 1,536 1,464 1,337 1,623 1,495 1,533 1,446 1,484 1,576 Corn 1,256 1,311 1,283 1,365 1,515 1,413 1,424 1,480 1,498 1,551 1,601 Wheat ,104 1, ,074 1,085 Oil Crops Groundnuts 1,004 1,347 1,286 1,286 1,300 1,260 1,166 1,223 1,269 1,222 1,280 Soybeans N.A. 1,059 1,091 1,037 1,292 1,471 1,386 1,531 1,698 1,692 1,575 Fibers Cotton Other Crops Coffee ,298 1, , Cacao Beans Manioc 13,347 13,404 14,120 14,662 15,523 14,529 12,623 11,560 12,284 12,657 12,861 Sugarcane 38,333 42,773 44,841 45,551 46,516 47,202 46,904 46,621 42,674 43,466 43,773 Source: Paiva, Schattan and Freitas "Setor Agricola do Brasil" for years Ministry of Agriculture (SUPLAN) for years , , projections. Soybeans Anuario Estatistico do Brazil for 1973 Industrial Projects Department March A A

64 ANNEX 3-2 Page 1 BRAZIL - VALEFERTIL FERTILIZER PROJECT THE FERTILIZER SECTOR A. Fertilizer Industry - General Background (i) Resource Endowment 1. Brazil is not well-endowed with good raw material resources for the production of chemical fertilizers. Natural gas, the ideal raw material for nitrogen (N) fertilizer production, has been found only in modest quantities in three areas: (a) Sergipe and Alagoas (Northeast), (b) offshore of Campos, near Rio de Janeiro (Center), and (c) 200 Ian offshore of Amapa (North). The reserves in the Northeast have been known the longest and they have the advantage of being located close to the Northeast fertilizer consumption center. These reserves are limited in size but will provide the feedstock for the Sergipe project and PETROFERTIL's existing and presently expanded ammonia/urea plant. The oil and gas fields offshore Campos have just been discovered,and priority for the use of this gas has been given to fertilizer production. At this time, no information is available on the likely quantities involved. The gas in the North has also just been discovered, but the location of the reserves will make exploitation difficult, and it is not likely that gas from these fields will be used for fertilizer production in the near future. The other feedstocks for N fertilizer production, naphtha and fuel oil, are indirectly imported as they are derived from crude oil,and Brazil imports some 80% of its crude oil requirements. The Araucaria fertilizer project under construction at Parana will use fuel oil, and the proposed project in Rio Grande do Sul will probably also use fuel oil feedstock. There are six major deposits of phosphate rock (Table 1 and MAP IBRD 12055R1) with an average P 2 05 content ranging from 5% to 17%, well below the 20% to 30% average P205 content found in commercial deposits in other countries (USA, Morocco, Jordan, etc.). At present, only one deposit (Jacupiranga with an average of 5% P 2 05 ore grade) is commercially developed, in conjunction with a cement factory, which improves the economics of the combined cement/phosphate fertilizer enterprise. The Jacupiranga deposit provides the rock for the QUIMBRASIL fertilizer plant and accounts for just about 8% of the 1976 P 2 05 consumption. However, with the recent local development of a commercial beneficiation technology capable of producing ore concentrates with 35% P205 at a high recovery factor_/ and the increase in the international prices of phosphate rock and phosphatic (P) fertilizers, Brazil has embarked on a phosphate mining/beneficiation l/ The beneficiation process is now well proven at Jacupiranga and the Arafertil pilot plant at Araxa. A unique feature of the process is its ability to attain a fair, from 50% to.80%, P205 recovery from the low grade igneous ores common in Brazil. On a comparable basis, the standard metallurgical process used on U.S. sedimentary type ore in the 5% to 8% P205 range shows recovery factors of about 30% to 50%.

65 ANNEX 3-2 Page 2 investment program to develop its low grade deposits. The ARAFERTIL mining/ beneficiation project at Araxa and the VALEP project at Tapira are scheduled for commercial operation by mid-1977 and late-1978, respectively. Feasibility studies for developing the Catalao deposit, the Patos de Minas deposit and the Ipanema deposit are underway. These six deposits have potential reserves of about 345 million tons of P205 of which about one half can be considered as proven. With an average P 2 05 recovery factor of 60%, the proven reserves of about 170 million tons of P205 is equivalent to 90 times the 1976 annual P 2 05 consumption. Elemental sulfur, another important raw material for phosphoric acid production, has not been found in commercial quantities in Brazil. 2. Finally, only one significant potash (K) deposit has been found, in Carmopolis (Sergipe). The deposit is in the form of potassium chlorinated salts with the equivalent of at least 30 million tons of K 2 0 or equal to about 50 times the 1976 annual consumption. However, the existence of oil in the same area, the lack of adequate technology, and the disagreement among the different parties concerned have so far prevented the commercial development of this deposit. Consequently, all potash requirements will continue to be imported for some time. (ii) Industrial Structure 3. With international prices for raw materials, intermediates (phosphoric acid and ammonia) and fertilizer products relatively low up to 1973 and because of Brazil's low natural resource endowment, the Brazilian fertilizer industry has concentrated on importing, mixing, blending and the distribution of fertilizers, rather than on establishing large scale, efficient production facilities. Brazil's fertilizer industry, therefore, consists of a large number of small to medium-sized, mostly private sector, firms with relatively uneconomic production facilities. Because of their high production costs, producers have to be protected except when world fertilizer prices are high. Over the next five years, however, the industry will undergo major changes, with new and larger facilities providing economies of scale and being mostly owned by public sector enterprises, notwithstanding the Government's policy of encouraging the private sector to invest in the industry. The large financial requirements for new projects will prevent the private sector, with its limited financial base, from participating in a major way in the capacity expansion of the industry. Consequently, the new, large-scale projects have to be financed and undertaken with the help of sources outside the present industry. Most of the new projects are, therefore, sponsored by the large, Government-owned industrial enterprises such as Petroleo Brasileiro (PETROBRAS), the national oil company, and Companhia Vale do Rio Doce S.A. (CVRD), a large mining company. However, it is expected that the blending, mixing and distribution of fertilizer products will continue to be in the private sector. 4. At present, there are 20 firms producing fertilizers and about 30 more enterprises just engaged in mixing, blending and retail operations. Of the 20 producers, only one firm (ULTRAFERTIL) has primary production facilities for both ammonia and phosphoric acid; three firms (PETROFERTILr,

66 ANNEX 3-2 Page 3 r a. u have aammonia production units; and two firms (COPEBRAS and QUDfBRASIL) have phosphoric acid production facilities. However, only QUIMBRASIL uses local phosphate rock. Both COPEBRAS and ULTRAFERTIL use imnorted roew The other 14 manufacturing firms import the intermediate products (ammonia and/or phosphoric acid) for final processing and/or produce low analysis products such as thermophosphates and single superphosphates (SSP) which do not require phosphoric acid and ammonia. Of the 20 firms, five produce complex (NP and NPK) fertilizers, another five produce just N fertilizers, and the remaining ten P fertilizers only. 5. Following the extraordinary increase in international fertilizer prices in 197 (Chart I) which almost tripled the fertilizer import bill from the US$169 million in 1973 to US$492 million in 1974 and about US$500 million in 1975, the Government initiated a National Fertilizer Program (NFP) which stipulated self-sufficiency in fertilizer by To implement the NFP, the Government established the National Fertilizer Commission with representatives from the Ministries of Finance, Industry, Commerce, Transport, Agriculture, and Planning. The Commission has overall responsibility for planning and formulating policies to facilitate the growth of the industry, especially the expansion of the production capacity. As a result of the NFP, three new large-scale ammonia/urea projects are now firmly planned at Araucaria (under construction), Sergipe and Norte Fluminense, all of which will be undertaken by FERTILIZANTES, a subsidiary of PETROBRAS. The three projects will almost triple the ammonia capacity from the 1977 level of 477,000 TPY of N to 1,300,000 TPY of N by One large public sector phosphate fertilizer project is under construction by VAIEFERTIL, a CVRD subsidiary, which will more than double the domestic phosphoric acid production capacity from 220,000 TPY of P205 to 510,000 TPY of P205. Three other large phosphate fertilizer projects are planned in the private sector, but all face serious financing problems and are not expected to start production until 1981 or later. The three are: the ARAFERTIL project at Araxa, the Luchsinger project in the South, and the ICC project, also in the South. Both the ARAFERTIL and the Iuchsinger projects will have phosphoric acid capacities of about 190,000 TPY of P205 and the ICC project 105,000 TPY of P However, only the VALEFERTIL and the ARAFERTIL projects will be based on local phosphate rock. The two projects in the South will have to use imported phosphate rock as no phosphate deposit has been discovered in that region. In all, there are nine firm projects, supported by the NFP, which are estimated tc cost almost US$1.9 billion and include four phosphate rock mining/beneficiation projects (Araxa, Tapira, Ipanema, Patos), two phosphate fertilizer projects (VALEFERTIL, ARAFERTIL), and three ammonia/urea projects (Araucaria, Sergipe, Norte Fluminense). 7. There are four industrial associations for the fertilizer industry: (a) The Associacao Nacional Para Difusao de Adubos (ANDA), which is responsible for the general promotion of and research on fertilizer use;

67 ANNEX 3-2 Page 4 (b) The Syndicato de Materias Primas, which represents the domestic producers of fertilizers, (c) The Syndicato da Industria de Adubos e Colas do Estado de Sao Paulo, and (d') The Syndicato Estado de Rio Grande. The last two associations represent blenders, mixers and retailers. ANDA also acts as consultant to the Government and in the past has engaged the services of foreign firms (e.g. Tennessee Valley Authority and FAO) for special projects. The syndicates collect and disseminate information on consumption, production, stocks and distribution. They also prepare special studies to support the industry's views to the Government whenever changes in prices, subsidies and import quotas are contemplated. With the Government's manifest interest in the industry, the associations have begun to improve their system of data collection, their role as information sources and are now engaged in forecasting likely fertilizer consumption levels. B. Agriculture and Fertilizer Use (i) Factors Affecting Fertilizer Use 8. Four main factors, other than weather, affect the fertilizer usage in Brazil. These are: (a) the development of the cultivated area; (b) soil characteristics; (c) the cropping pattern and the level of agricultural technology; and (d) the Government policies concerning relative agricultural input and output prices, subsidies, credit, agricultural extension services and special programs related to agriculture. The Government's agricultural policies are generally discussed in Annex 3-1, and fertilizer prices and subsidies are discussed in Section C of this Annex. 9. Table 2 provides information on the potential agricultural areas that could be developed in the different regions of Brazil. The total potential agricultural area that could be used is estimated to be between 165 to 250 million ha. with a best estimate of 210 million ha. During the period , cultivated area increased at an annual rate of 3%, from 28.5 to 38.5 million ha. During , however, the cultivated area increased only slightly (Annex 3-1, Table 2) and for 1975 was about 41.5 million ha., representing just about 20% of the estimated potential of 210 million ha. Thus, the Government's Second Development Plan ( ) objective of increasing the cultivated area by 4% to 5% annually is slightly higher than the average rate achieved in the past. Brazilian soils are generally acidic and seriously deficient in phosphorus. Table 3 shows the average phosphorus and potassium content of the soils by states, indicating that on a national basis, an average of 77% of the presently cultivated area is low in phosphorus (or phosphate) with the remaining 23% containing medium to high levels of natural phosphate nutrients. In contrast, only about 29% of the cultivated area is considered seriously deficient in potassium (or potash). This deficiency in phosphorus coupled with the acidity of the soil has led to the unusual NPK consumption ratio of about 1:2.3:1.3 (average for the period ), with phosphate consumption more than twice the nitrogen consumption. The planned increase in cultivated area, mostly involving soils with nutrient deficiencies,

68 ANNEX 3-2 Page 5 will clearly require an increase in fertilizer use, especially phosphatic fertilizers. To correct for soil acidity, large quantities of lime are applied since acidity is not compatible with most crops and in addition, the efficiency of N fertilizers is reduced, depressing agricultural yields. There are adequate domestic lime deposits and production capacity. 10. The cropping pattern and the level of agricultural technology is the third important factor affecting fertilizer use. As shown below, there has been a marked change in the cropping pattern. The area cultivated with coffee, for example, decreased from 4.h million ha. in 1960 (15.4% of the cultivated area) to 1.4 million ha. in 1976 (3.2% of the cultivated area). In contrast, the area planted with soybeans increased from a 1960 level of 0.2 million ha. (0.7% of the cultivated area) to a 1976 level of 6.3 million ha. (14.5% of the cultivated area). The increase in the areas planted with crops such as wheat, rice and soybeans, which receive more fertilizers than other crops,partly explains the increase in fertilizer consumption experienced during the period. BRAZIL - CROPPING PATTERN AND CULTIVATED AREA (area in million hectares) Crop Area % Area % Area % Aiea % Area % Corn Rice Soybean Blackbeans Cotton Wheat Sugarcane 1.3 L Cassava Coffee 4.h O0.hers TOTAL Source: data from Anuario Estatistico de Brasil (IBGIE) data from Ministry of Agriculture (SUPL4i) forecast. 11. The recommended fertilizer application rates, which are shown in Table 4 and summarized below, show that most major crops, with the notable exception of coffee, require more P nutrients than either N or K. Actual fertilizer application rates are difficult to obtain since there is no systematic collection of data on fertilizer use by crops or by year, and only rough estimates can be made on the basis of fertilizer credit application data. One estimate of actual fertilizer application rates for different crops for 1974 is shown below. The recommended levels are low by international standards but are appropriate for the existing level of agricultural

69 ANNEX 3-2 Page 6 technology. The actual application rates range from about 35% to 70% of the recommended level except for corn and rice, which are much lower. The six crops below account for almost 7% of the total fertilizer used in Brazil. BRAZIL - RECWMENDED FERTILIZER APPLICATION AND ESTIMATED 1974 USAGE FOR MAJOR CROPS Kg of Nutrient/hectare Estimated 1974 Actual % of Total Recommendedi/ 1974 NPK as % of NPY. Fertilizer Crop N P K Total NPK Application Recommended Consumption&/ Coffee Sugarcane WIheat Rice N.A. N.A.3/ 7 Corn Soybeans TOTAL 1/ Estimated national weighted average based on 1973 cultivated area for the crop and region (Table 4). P is in terms of P 2 C5 and K in terms of K / This is the estimated average for 1971, 1973, and The 1971 and 1973 estimates were derived according to the distribution of fertilizer credits. The 1974 data is an ANDA estimate. 3/ This is probably in the 10% to 20% range considering that the corn area was 2.8 times the rice area in 1974 while rice accounted for 7% of fertilizer consumption compared to 10% for corn. 12. The differences between the recommended application rate and the actual rate can be partly explained by (a) the inadequacy of complementary inputs or technologies (plant varieties, cultivation method, correction of soil acidity, etc.) and agricultural extension services; (b) the low level of commercialization in several areaswith concomitant high transportation costs for both agricultural inputs and outputs and lack of accessibility to credit sources; (c) the general poverty of the small farmholders who are mostly non-users of fertilizers; and most importantly (d) the decline in the purchasing power of agricultural crops relative to fertilizer prices during 1974 as a result of the large increase in international fertilizer prices (para 26). 13. The cropping pattern also affects the seasonal or monthly usage of fertilizers. About 70% of the fertilizer used for coffee, 60% of the fertilizer used for soybeans and 50% of that used for sugarcane are applied during the four months from July to October. Consequently, more than 50% of the total fertilizer consumption is used during July to October and about 70% is used during the six months from May to October (Table 5). This seasonal pattern of consumption is also reflected in the pattern of imports with about 50% of total imports occurring during May to August.

70 ANNEX 3-2 Page Of the Government policies affecting the use of fertilizers, the policies for the agricultural credit system are particularly important. The Government has been well aware of the importance of working capital and liquidity for the modernization of agriculture; however, the effectiveness of the programs instituted is somewhat debatable given the relatively low overall efficiency of agricultural production and the lack of sufficient statistical data and analysis. In 1965, the Government created the National Rural Credit System (SNCR) which involved a complete revolution in Brazil's rural credit system. SNCR is made up of the Central Bank (which has a supervisory role in the system) and the Bank of Brazil, Bank of Amazonia, Bank of the Northeast, and the National Credit Cooperative Bank. The National Institute for Colonization and Agrarian Reform (INCRA), the State official banks, the National Savings Bank, the private banks, and the rural credit cooperatives are linked informally to the SNCR. To coordinate, guide and control the behavior of the different components of the credit system, a special department was established in the Central Bank. 15. The subsidy for the modernization of agriculture, specially the consumption of fertilizer, began with the foundation of FUNFERTIL which paid the total interest on loans used for buying fertilizer. In 1964 it was substituted by FUNDAG which provided most credit at an annual interest rate of 7% for buying modern inputs except for fertilizers where credit was interest free until This accounts for the large (95%) share of fertilizer puirchases through the agricultural credit system. In addition rural credit plays a major role in special development programs such as the National Integration Program (PIN), the POLOAMAZONIA and POLWNORDESTE. The POLOCENRO, PRODEPAN and GRANDE DOURADOS programs will be of special importance for the development of fertilizer consumption in the western part of the Center. With the introduction of the price subsidy, interest on fertilizer credit was increased from 0% to 13 and 15% for medium/large and small farmers, respectively. The average maturity for fertilizer credit is 7-10 months, although it can be extended up to two years for crops like sugarcane and coffee, which require a longer gestation period. 16. Rural credit increased by a factor of seven from Cr$12.9 billion to Cr$90 billion in current terms (a factor of 3 in real terms) between 1971 to 1975 (Table 6). The share of fertilizer credit to the total rural credit gradually increased from 7.3% in 1971 to 8.3% in As a result of the steep rise in fertilizer prices in 1974, the share of fertilizer credit increased to 1h% of the toval rural credit but declined slightly to 11.4% in 1975 partly reflecting the drop in fertilizer prices and the slower growth in consumption during that year. Regionwise, fertilizer credit has the largest share in the Center followed by the South and the Northeast. It should be noted that the Center region in Table 6 does not include the State of Parana which is included in the South. The regional distribution of fertilizer credit according to the definition of the fertilizer market and marketing regions (where Parana is part of the Center) would have to be adjusted therefore to be about 8%, 75% and 17% for the Northeast, Center and South, respectively, while consumption of fertilizer is about 9%, 60% and 31%. This would indicate that the South is making less use of fertilizer credits than the Center. Lacking more specific data, it is difficult to find an explanation for this.

71 siuex 3-2 Page 8 (ii) Agricultural Productivity 17. Until 1970, the growth in Brazil's agricultural output was mostly accounted for by the expansion of the cultivated area. With the exception of a few crops such as coffee, wheat and potatoes which registered remarkable increases in yields, the productivity of most crops showed only slight improvements from 1961/63 through 1968/70. During 1968/ /74, cotton, cacao and soybeans achieved gains in yields as shown below. Over the period 1961/ /74, five of the major crops achieved average productivity increases exceeding 2% annia1uy. BRAZIL - YIELDS OF MAJOR CROPS Average Annual Productivity in 100 KU/hectare Increase in Yield (a) Crop 1961/63 lb94/66 196d/ / / / /74;L/ Coffee Ateat Cotton Cacao Soybeans Ui Potatoes Sugarcane o Corn Cassava _5 Rice Blackbeans / For estimating the annual increase in yield, the Ul-year period, 1962/1973, is used. 18. Different levels of agricultural technology for the various crops account for most of the difference in productivity gains or lack thereof. Coffee, being the most important crop in terms of export earnings, has probably the most advanced agricultural technology in terms of high-yielding varieties, cultivation methods and fertilizer application. In addition, it also has well-developed supporting services such as extension services, credit facilities, and marketing system. As a result, the 50% reduction

72 ANNEX 3-2 Page 9 in the area cultivated with coffee between 1960 and 1974 was offset by a doubling of yields in the same period. The shift in the coffee areas away from the frost-prone zone of the southern part of the Center region towards the central part of the Central region with its more favorable weather conditions should ensure more stable and better yields in the future. 19. The level of agricultural technology and supporting services, even for the same crop, varies among the different regions in Brazil. In general, the most agriculturally developed region is the South, followed by the Center with the North and Northeast regions the least developed. The regional differences in productivity during 1973 is illustrated in Table 7. The best productivity is in the South and Center regions while the lowest is in the Northeast region. Thus, the regional distribution of the cultivated area for a particular crop affects the overall productivity for that crop. For example, practically all the wheat and the soybean areas are in the South and Center regions. In contrast, about 96% of the sugarcane areas and practically all the cotton areas are in the Center and the Northeast regions. The gains in cotton productivity, however, mostly originate from the Center region, which accounted for 60% of the 1973 cotton output but only 28% of the cultivated area. Finally, most of the recently cultivated land is in less developed areas with a low productivity, thereby depressing the overall national yields for those crops introduced in the newly cultivated land. 20. An international comparison of agricultural productivity, as shown below, indicates a large potential for improving agricultural productivity in Brazil, even among those crops which have shown significant yield improvements during the last 15 years. COMPARATIVE AGRICULTURAL YIELDS Rice (KgAa) Corn (Kg/ha) Japan 3,838 U.s.A. 4,478 U.S.A. 4,978 Argentina 2,840 India 1,640 Brazil 1,339 Brazil 1,557 Mexico 993 Wheat (KgAha) Sugarcane (Kg/ha) Netherlands 5,733 Mexico 64,773 Canada 1,497 India 50,629 India 1,158 Brazil 47,500 Brazil 1,100 Cuba 37,500 Iran 820 Coffee (g/ha) Soybeans (K /ha) Colombia 637 U.S.A. Brazil 623 Brazil 1,565 Ivory Coast 448 Angola 440 Source: FAO Production Yearbook 1974

73 ANNEX 3-2 Page While Brazilian agriculture will continue to be extensive in character as opposed to intensive agriculture as practiced in the Netherlands or Japan, the Government's objective of a 2% to 3% annual increase in agricultural productivity during (Annex 3-1) appears reasonable. Such improvements, however, will require a systematic and deliberate approach to agricultural development considering that most of the newly cultivated land will be located in less developed areas requiring the development of transportation and marketing systems. Both the expansion of cultivated area and the improvement in the level of agricultural technology imply a continued growth in the demand for fertilizers. 22. The extensive nature of Brazil's agriculture is further demonstrated by low fertilizer application rates, when compared to other countries, as shown below: 1/ 197h COMPARATIVE FERTILIZER APPLICATION RATES-/ (Kg of Nutrient/hectare) Country N P205 K20 Total NPK Netherlands Japan United States Portugal Brazil India Argentina World / For all crops. Source: FAO Statistical Yearbook C. Agricultural Inputand Output Prices 23. In pursuing its agricultural production and productivity objectives, the Government has traditionally relied on two policy instruments: providing credit to finance agricultural inputs (as discussed) and setting minimum prices for agricultural outputs to guarantee a reasonable level of profitability for agricultural production. The system of minimum prices for agricultural output was started in 1945 and has since extended its coverage from six major crops (rice, beans, corn, peanuts, sunflower and soybeans) to almost all agricultural products and most products of animal origin. The Consejo Interministerial de Precos (CIP) is responsible for operating the minimum price system. The minimum price is generally announced 60 days before the crop is planted and the Government guarantees the purchase of the production at this price. Financing is also provided for about 180 to 216 days at 100% of the estimated value of the production based on the minimum price. The geographical coverage of the system has expanded from the export ports and large consumer centers to many in' ermediate points in the

74 ANNEX 3-2 Page 11 country thereby creating homogenous price zones. From 1966 thr mgh 1974, farm prices of most agricultural products increased in real terms (Table 8). 24. Up to 1974, the Government has not intervened in the determination of fertilizer retail prices, allowing market forces to set prices. Consequently, fertilizer prices have been affected by both international and domestic economic trends as indicated below by the cyclical pattern of prices (in constant terms) during the last 30 years. BRAZIL - INDEX OF RETAIL FERTILIZER PRICES (IN CONSTANT TERMS) (Basis l948/ 2 = 100) Year Index Year Index Year Index a/ / NA_/ Source: Instituto de Economia Agricola--IEA 1/ The 1975 index is estimated on the basis of a 40% subsidy. Without the subsidy the index would be / Prices are expected to decline slightly (in real terms) during 1976 and The combination of real term increases in farm product and decreases in fertilizer prices increased the purchasing power of agricultural products relative to fertilizers as shown below:

75 ANNEX 3-2 Page 12 BRAZIL - PURCHASING POWER OF AGRICULTURAL CROPS Kg of Fertilizer Purchased by 100 Kg of Crop/- Year Rice Corn Coffee Soybeans / h17 Ili 19762/ / / ieighted average price according to apparent NPK consumption. 2/ Considering a 40% subsidy on fertilizer prices and a 15% interest rate on fertilizer credit. 3/ Without the 40% subsidy but with the 15% interest rate on fertilizer credit. 26. The purchasing power of agricultural crops relative to fertilizer declined dramatically during 1974 as a consequence of the large increase in international fertilizer prices. Concerned about the possible adverse effect of the increase in fertilizer prices on fertilizer usage and agricultural production, in 1975 the Government introduced a 40% subsidy towards the purchase of fertilizer and a 15% interest rave on fertilizer credits which had been interest free prior to The net effect, however, was favorable and brought the purchasing power of most agricultural crcps back to thie 1972 Level. International rertilizer prices began to decline in the third quarter of 1974, and by 1976 have settled back to the 1972/73 levels (Chart I). As a result of the rapid decline in international fertilizer prices during 1975, the purchasing power of agricultural products in relation to fertilizers increased again and therefore the Government eliminated the subsidy in late December This however, is not exdected to have any adverse effect on fertilizer demand, since the relationship between agricultural product and fertilizer prices is expected to remain favorable. The removal of the subsidy, when it was not any longer economically justified, is an indication of the attention the CGoernment is paying to the fertilizer sector and of the Government's determination to follow rational policies in this sector. 27. To avoid abuses within the fertilizer price subsidy crogram anc to prevent the fertilizer industry from making excessive profits arising from the preference position granted to the industry by the import quota system for fertilizers (para 4o), the Government in 1975 initiated a system of maximum fertilizer prices at the retail level. The CIP computes the maximum price per nutrient based on the proportion and expected cost of imported and domestically produced fertilizers, and an allowance for blending and distribution costs. The resulting price is then published as a ceiling price and credit or subsidy is only given to fertilizers purchased at prices equal or below the ceiling price. For the first half of 1976, the CIP published the following maximum retail prices: US$640/ton of N, US$690/ton of P205, and US$250/ton of K20 as derilved below.

76 A1RNEX 3-2 Page 13 BRAZIL - MAXIMtlM RaAIL PRICES FOR ERTILIZE. (February to June 1976) Cr$/Kg of Nutrient N?2C -20 Ex-Factory or CIF Cost Domestic Fertilizer Imported Fertilizer Share of Total Supply (%) L 1.78 Domestic Fertilizer Imported Fertilizer Average Cost Distribution Margin at 42.8% of the Average Cost Maximum Retail Price In US$/ton of Nutrient Comparative Figures Cr$/Kg US$/ton Source: CIP 28. The maximum retail prices, on a US$ equivalent basis, have declined by 28% for N, 13% for P205 and 23% for K20 between 1975 and the first half of For the second half of 1976, the CIP has increased the maximum prices at a monthly rate of 1.5%, which is considerably lower than the inflation rate of 2% to h% per month expected for the same period. This intended reduction in the maximum retail price level in real terms is apprcpriate considering the downward trend in international prices from 1975 through As the maximum prices are conservatively estimated and include a safety margin in the import, production and distribution costs, it has been observed and can be expected to continue that, on the average, actual prices because of competitive pressures at the retail level wlla tend to be 15-20% lower than the stipulated maximum. 29. The estimate for domestic production costs indicate that they are about 52% higher for N and 39% higher for P205 compared to the cost of imported fertilizers. Several factors could explain this difference. Firstly, the domestic production facilities, particularly for N, are small and do not enjoy the economies-of-scale common in the export oriented plants of other countries. Secondly, the industry is based on imported raw materials and intermediate products, particularly phosphate rock, sulfur, phosphoric acid and ammonia whose international prices have not decreased in the same proportion as the prices of finished products. 1Shile this is a short-term phenomena that is expected to change, the inmediate effect is a situation whereby the cost of importing the raw materials and intermediates required for the production of some fertilizers into Brazil is higher than

77 ANNEX 3-2 Page 14 the cost of importing the finished product. Finally, the domestic costs are provided by the local firms and the cost data probably contain some upward bias, to ensure higher profits or to incorporate anticipated cost increases that may not be fully matched by increases in the stipulated maximum prices in the future. 30. The export of agricultural output containues to be the primary source of foreign exchange for Brazil. The Government's economic objectives especially the management of the country's precarious balance-of-payments situation, depends heavily on the export performance of the agricultural sector. Thus, there is within the Government increasing attention on agricultural policies designed to increase agricultural exports. To enhance Brazil's competitiveness in the agricultural export market, it is important that agricultural inputs be provided at cost levels comparable to those prevailing in other exporting countries. The introduction and abolishment of the 40% subsidy for fertilizer is illustrative of the Government's increasingly active role. In mid-2976,-brazilian fertilizer retail prices without the subsidy are somewhat higher than those prevailing in the USA as shown below. BRAZIL - COMPARATIVE FERTILIZER RETAIL PRICES (US$/Ton of Product) Brazil._/ Product Foriula (NPK) With SubsidyWthout Subsidy USA?./ / Estimated actual retail prices which are assumed to be 80% of the CIP maximum retail price. Prices are for June 1976 at an exchange rate o0 Cr$10.80 = US$ / U.S. Dept. of Agriculture, April However, during the latter half of 1976, Brazilian fertillizer prices further declined in real terms and in early 1977 they are estimated to be only slightly above those prevailing in the USA. in the longer run, as the new fertilizer production facilities come into operation, the average cost of domestic production is expected to decline and come in line with the lower international prices forecast for the end of thds decade. D. Fertilizer ConsumDtion (i) Historical Fertilizer Consumption Patterns 31. Between 1959 and 1966, the year-to-year apparent consumption of fertilizer fluctuated considerably, but did not show any underlying growth

78 ANNEX 3-2 Page 15 trend (Table 9, Chart II), reflecting in part the Government's neglect of the agriculture and.fertilizer sectors during that period. By the midsixties, however, the improvements in the credit system, the favorable economic climate, the continued expansion of the cultivated area, the shifts in the cropping pattern, the improving relationship of agricultural product to fertilizer prices and the increasing availability of high analysis fertilizer products, provided the stimuli for increased fertilizer use. During the six years from 1966 to 1972, fertilizer consumption increased at an average annual rate of 35%. This rapid growth in fertilizer use is also reflected in the marked improvement in the intensity of fertilizer use be-tween 1966 and 1972 as shown below: BRAZIL - INTENSITY OF FERTIIIZER USi/ (Kg of Nutrient/hectare) Nutrient L 1976 N P ) K Total NPK L Index of Real Fertilizer Pricesz/ / Based on apparent consumption and cultivated area. 2/ From para 27. The price index is in real terms with 1963 = 100. For 1976, the index shown is for 1975 and is based on a 40% price subsidy. Without the subsidy, the index would be 209. A slight decrease in real prices is expected between 1975 and The strong upward trend was only interrupted by adverse weather in 1969 and 1973 and slowed down by the large increases in interna+tional fertilizer prices during 1974 and The introduction of the fertilizer retail price subsidy (paras 26 to 28) during 1975 helped apparent consumption grow at a modest rate that year. The continuation of the subsidy system and the continued decline in international prices have caused apparent consumption to rise by 17% in The absence of systematic statistics on changes in the fertilizer inventory levels makes it difficult to estimate actual year by year consumption, which would probably show less annual fluctuations and a more discernable trend compared to apparent consumption. The Ministry of Agriculture has made an estirmate of actual consumption for the last six years which shows a more stable groarth as shown below:

79 ANNEX 3-2 Page 16 BRAZIL - CGMPARISON OF APPARENT AND ESTIMATED ACTUAL FERTILIZER CONSUMPTION Consumption in 1,000 nutrient tons % Annual Growth Rate Year Estimated Actual Apparent Estimated Actual Apparent ,126 1,446 1,165 1, ,879 1, ,777 1, ,o45 1, ,351 2, Source: SUPLAN 33. Phosphate is the most important nutrient consumed because of the soil characteristics and cropping pattern in Brazil (paras 9 to 12). This has led to the NPK consumption ratio of about 1:2.3:1.3 for the period. As can be expected, there is a marked difference in the NPK consumption ratio between the South, the Center and the North/Northeast regions (Table 10, Chart III) as summarized below. The highest ratio of P to N is in the South, where most of the soybean and wheat areas are located. BRAZIL - AVERAGE REGIONAL NPK CONSUMPTION RATIO Region Average NPK Ratio South 1:4.1:1.5 Center 1:2.0:1.3 North/Northeast 1:1.3:1.1 Brazil 1:2.3 : Fertilizer consumption has increased most rapidly in the South, at about 34.4% annually over the last 10 years (Table 11), followed by the Northeast (22.4%) and the Center (19.7%). The higher growth rate in the Northeast compared to the Center reflects the relatively lower consumption levels in the Northeast during the mid-sixties. Although the South is the most intensive user of fertilizer, the Center, with its larger cultivated area, accounts for almost 60% of total nutrient consumption over (Table 12) followed by the South (31%) and the North/Northeast (9%). 35. Although a significant shift towards higher analysis fertilizer products occurred in the late sixties, only the phosphatic fertilizers continued the trend between 1970 and 1974, with TSP displacing SSP (Table 13). The limited domestic production capacity for urea is partly responsible for the continued dominance of low analysis products in the nitrogen fertilizer area.

80 ANNEX 3-2 Page 17 (ii) Fertilizer Demand Forecast 36. In 1972 the Government commissioned a study of the fertilizer sector. The study was executed by SEITEC under the joint supervision of BNDE (The National Development Bank), IPEA (a planning institute in the Ministry of Planning) and ANDA. The report, published in May 1974 on the basis of actual data up to 1970, estimated fertilizer demand at about 4.0 million nutrient tons by 1980 equivalent to an application rate of 80 kg of nutrient/ha. compared to about 52 Kg/ha. during The demand forecast formed the basis for the Government's National Fertilizer Plan (NFP) of ear'y The upper limit for the potential demand for fertilizer can be estimated on the basis of normative analysis of necessary nutrient replacement rates, and projections of both increases in cultivated area and cropping patterns. This approach results in a potential demand estimate of about 12 million nutrient tons in Given the 1976 consumption of 2.3 million nutrient tons, an upper bound of 23% annual average growth rate from 1976 to 1985 can thus be established. On the basis of the trends of the past ten years, the Government's agricultural programs for achieving its goal of an increase of about 7% annually for agricultural output (Annex 3-1) during the Second Development Plan period ( ), and the increasing availability of more economically produced local fertilizers, demand growth rates ranging from 10% to 15% annually appear likely, which would be somewhat lower than the actual rates for (para 34). Higher growth rates could probably not be sustained because constraints such as a slower development of complementary agricultural technology, and limitations in the extension services, credit facilities and the transportation and marketing systems would be much too difficult to overcome. The Bank's forecast on the basis of a 10% to 11% annual NPK demand growth rate is shown below, together with the forecast contained in the NFP. The NFP forecasts the NPK ratio to shift significantly to about 1.0:1.14:0.7 by 1980, since generally, at more intensive levels of fertilizer use, proportionately more N is required compared to P. However, because of the planned increase in newly cultivated area, which require considerably more P than N; and the evolving cropping pattern showing several major crops such as soybeans and wheat which also require more P than N, it is very likely that only a modest shift in the NPK ratio will materialize. BRAZfIL - FERTILIZE;R DEMAND FORECASTS In Millions of Nutrient Tons Year N K20 Total NPK Ratio 1980 NFP (1974) 1.L :1.14:0.7 Bank (1976) :2.1: Bank (1976) :1.8:1.2 % Average Annual Growth Rate (Bank)

81 ANNEX 3-2 Page The regional shares of fertilizer demand are expected to follow the consumption pattern of the last five years ( ) as shown below: BRAZIL - PROJECTED REGIONAL SHARES OF FERTILIZER DEMAND ( ) % Share of Nutrient Demand Region N P20I K20 Total North/Northeast Center South TOTAL E. Fer-tilizer Supply - Imports and Production (i) Historical Pattern of Fertilizer Supply 39. Brazil imports a major portion of its N and P and all of its K nutrient requirements in the form of raw materials (phosphate rock), intermediates (phosphoric acid and ammonia) and finished products as, mentioned in para 3. From , total nutrient imports accounted for 55% to 65% of apparent consumption. From 1966 to 1975, the share of imports has increased to about 65% to 80% (Table 14) reflecting the low natural resource endowment of the country. Nonetheless, an increasing proportion of imports represented raw materials and intermediates, especially for P, rather than finished products. This reliance on imported fertilizer materials has proved expensive when international fertilizer prices increased threefold between 1973 and 1974 (Chart I). The fertilizer import bill for 1974 and 1975 were about US$500 million compared to US$169 million in The reliance on imports of finished as well as intermediate products prior to 1973 has helped maintain retail fertilizer prices at reasonable levels, since local production costs were generally much higher than the costs of imports (para 29). To protect local producers and ensure that all local production is consumed, the Government has instituted an import quota system for finished fertilizer products. The system requires retailers to purchase a given proportion of local production for each unit of imported fertilizer product. The ratio changes depending on local availability and consumption expectations. When international prices are exceptionally high as in 1974 and 1975, imports are not restricted. But under more normal international price conditions, when local production costs exceed the cost of imports, the quota system is implemented. For 1976, the quota in the Northeast was 1:1 (1 ton of imports per ton of local purchases) for N and P could be imported without limit. In the South and the Center the 1976 quota was 1.6:1 for N (1.6 tons of import per ton of local production) and about 0.4:1 for P 2 05.

82 ANNEX 3-2 Page Imports are seasonal (Table 15), reflecting the seasonal structure of consumption (para 13) anr, V- a Brazilian blenders, mixers and distributors to niinimize their nretog costs by timing deliveries just slightly ahead of' the fertilizer season. The blending and mixing production levels are easily changed from one or two work shifts during the slack season and three to four work shifts during the peak season, thereby also minimizing working capital requirements. About 80% of imports are processed through two ports, Santos (60%) and Rio Grande (20%), reflecting the concentration of consumption in the States of Sao Paulo and Rio Grande do Sul. 42. The historical production of N and P fertilizers are shown in Table 16. Significant increases in production occurred only after 1968/69 when the QUIMBRASIL phosphate project started production with a phosphoric acid capacity of 65,000 TPY of P205 and a total phosphate product capacity of 100,000 TPY of P205. During 1970/71, the UlTRAFERTIL plant (123,000 TFY of N, 65,000 TPY of P2%05) at Sao Paulo and the PETROFETIL plant (55,000 TPY of N) in Bahia came on stream. The COPEBRAS phosphate project started production in late 1975 with a phosphoric acid capacity of 90,000 TPY of P205 and a total phosphate product capacity of 145,000 TPY of P205. The PETROFERTIL ammonia/urea expansion project (247,000 TPY of) is scheduled for completion in mid-1977 and will double Brazil's ammonia capacity from 230,000 TPY of N to 477,0Co TPY N. During , the local N production accounted for about 35% of N consumption while the domestic P production accounted for about 45% of P consumption. However, since all K nutrients are imported, the total nutrient (NPK) production just represented some 30% of total NPK consumption. (ii) Production Forecast 43. As discussed (para 3), the Government, through the NFP, is supporting a number of projects which will transform the fertilizer industry into a major industrial sector in Brazil's economy. One fuel-oil-based ammonia/urea project (325,000 TPY N) is under construction at Araucaria, Parana, with a Bank financing of US$50 million (Loan No BR). Two other ammonia/urea projects, both gas-based, are in an advanced stage of planning. The most advanced is in Sergipe in the Northeast (247,000 TPY N) which is being proposed for Bank financing. The other is at Norte Fluminense, Rio de Janeiro (247,000 TPY N) which replaced the Paulinea project after the discovery of offshore oil and gas near Campos. The Norte Fluminense project is being proposed for financing by the Inter American Development Bank. The three projects are expected to start commercial operation in 1979 and 1981, respectively, and will be owned by FERTILIZANTES with a combined financing requirement of about US$800 to 850 million. Since ULTRAFERTIL, PETROFERTIL and FAFER will also be subsidiaries of FERTILIZANTES to consolidate all fertilizer enterprises of PETROBRAS, FERTILIZANTES will hold the monopoly of the ammonia/urea production capacity in the country. A fourth ammoria/ urea project probably based on fuel oil to be located in the South (Rio Grande do Sul) is in the initial stages of planning by FERTILIZANTES and the State Government, but is not expected to come into production until 1982/83.

83 ANNEX 3-2 Page Table 17 shows the development of the ammonia capacity and production from 1976 through 1985, indicating an almost sixfold increase in capacity between 1976 and During 1976 through 1979, about 75% to 80% of the ammonia production will be available for fertilizer use, the balance will be needed for industrial uses. By 1980 and 1981, however, when the Sergipe and Norte Fluminense projects start operation, about 85% to 90% of the total ammonia production will be available for fertilizer use. 45. The phosphate rock production capacity of Brazil is also expected to increase about six times from about 130,000 TPY P205 in 1976 to about 875,000 TPY of P205 by 1979 (Table 18) with the operation of three mining/ beneficiation projects at Araxa (the ARAFERTIL project - 225,000 TPY P 2 05) in mid-1977, at Tapira (the VALEP project - 345,000 TPY P20s) in late-1978 and at Ipanema (the QUIMBRASIL/SERRANA project - 105,000 T P205). The Ipanema rock and the output from pilot plants at Patos and Catalao (METAG0), which will start production in 1977, will all be used for direct application. Between 1980 and 1985, three additional mining/beneficiation projects, now under initial planning, are expected to be operational. These are the Patos de Minas project of FERTILIZANTES (350,000 TPY of P205) expected to start operation in 1981 at the earliest; the Gatalao project of METAGO (220,000 TPY P205) and another project at Catalao by Mineracao Catalao (200,000 TPY of P205). The two projects at Catalao in the state of Goias are private sector ventures and will probably come into operation about two to three years apart, starting 1983 because of transportation difficulties, financing constraints, and inadequate market demand before The financing requirements of a large mining/beneficiation project based on low grade rock can range from US$50 to US$200 million. 46. Several phosphate fertilizer projects are also in the planning stage (para 6). The most advanced is the VALEFERTIL project at Uberaba with a phosphoric acid capacity of 290,000 TPY of P205 and a total phosphate product capacity of 333,000 TPY of P205, expected to be completed by mid The VALEFERTIL project will use the phosphate rock from the VAL2P project. A Bank loan has been requested for the VALEFERTIL project. A second project, by ARAFERTIL, will be located at Araxa to use the Araxa phosphate rock. The ARAFERTIL project will have a phosphoric acid capacity of 190,000 TPY P205 and a total phosphate product capacity of 220,000 TPY of P205. While the project has been ready for implementation for sometime, financing constraints have delayed construction and the project is now expected to be completed by In the meantime, the Araxa rock will be sold to producers in the State of Sao Paulo. Two projects are proposed to be constructed in the South, with a combined phosphoric acid capacity of 285,000 TPY of P205 and a total product capacity of about 330,000 TPY of P205, or about the size of the VALEFERTIL project. These two projects will be owned by LUCHSINGER and by ICC and will likely be completed in They will both use imported phosphate rock because no phosphate deposit has been discovered in the South. Table 19 contains the forecast of phosphoric acid capacity and production, while Table 20 shows the forecast of total phosphate fertilizer nutrient output.

84 ANNEK 3-2 Page 21 F. Forecast of Fertilizer Demand and Supply Situation 47. The regional as well as national forecast of fertilizer demand and supply is given in Table 21 for N and Table 22 for P All K20 is expected to be supplied from imports. The forecasts are summarized below and indicate that with the three new ammonia/urea projects Brazil will be increasingly self-sufficient in N with local production accounting for 33% of consumption in 1975 and reaching full self-sufficiency in The Northeast will have a surplus of N fertilizers which will be shipped to the Center which will continue to be a net importer of N. If demand should continue to grow at 10% annually, a new N fertilizer project will be needed to come into production every two years starting in 1983 to keep local production close to the demand levels. Thus, the ammonia/urea project in the South can be planned for completion by BRAZIL - NITROGEN DEMAND/SUPPLY FORECAST (1,000 Nutrient tons) Expected Demand North/Northeast Center South Brazil - Total MT fi 20i500 1,000 1,400 Expected Production North/Northeast Center South Brazil - Total = o0 % Self-sufficiency (Brazil) The supply and demand forecast for P indicates that the South will continue to be a deficit region, inspite of the planned doubling of capacity from 328,000 TPY of P205 in 1976 to about 643,000 TFY of P205 by The deficit could be eliminated by adding another 130,000 TPY of P 2 05 product capacity for 1980/81, and then increasing capacity by about 150,000 TPY of P205 every other year thereafter. The South will also be importing all its phosphoric acid and phosphate rock until 1980 when LUCHSINGER and ICC complete their phosphoric acid projects. By 1981, the South would still be only about 50% self-sufficient in phosphoric acid requirement, and will still import all its rock requirement. While a potential surplus of phosphate rock (from Catalao) exist in the Center region for , the prohibitive local transport cost makes shipment into the South uneconomic except when international prices increase much higher than presently estimated. 49. The phosphate product supply and demand situation in the Center region will change, however, from a deficit situation up to 1978, to a balanced situation in 1979, and then to a potential surplus situation

85 ANNEX 3-2 Page 22 during 1980 and The potential surplus is about 100,000 to 150,000 TPY of P205 and exists only because four firms (UNIAO, FERTIBASE, MANAH and INDAG) have the flexibility of producing some TSP, based on imported phosphoric acid, or SSP, based on imported sulfuric acid or sulfur. If these four firms produce the maximum TSP, they will supply 50% more P205 nutrients than if they produced only SSP.S/ The four firms could produce the maximum P205 nutrients (TSP) and ship part of their production to the South, which is expected to have a deficit of about 100,000 to 150,000 TPY of P205 between 1980 and 1983 and a much larger deficit after But these four producers may not be competitive with imports into the South. The production forecast for phosphate products in Table 21 assume that during 1980 through 1983, these four firms will adjust their product mix between SSP and TSP such that a surplus of P 2 05 nutrients in the Center region is avoided. This assumption is reasonable since the VALEFERTIL and the ARAFERTIL projects, which come on-stream during 1979 and 1981, respectively, will supply the necessary TSP and MAP based on their own phosphoric acid production. By 1984, however, this potential surplus will disappear and a deficit will occur again by It appears, therefore, that the Center region will not need additional phosphate product capacity, other than the VALEFERTIL and the ARAFERTIL projects, until Starting 1985 or 1986, additional capacity of about 100,000 TPY of P205 will be required each year to keep production in line with increases in demand. 50. The North/Northeast region will have a deficit of phosphate products up to 1979, when the proposed PROFERTIL TSP project and the AGROFERTIL MAP project are expected to be completed. Beyond 1979, the region should be self-sufficient in P 2 05 products, although it will import all the phosphoric acid and phosphate rock raw materials. 51. A surmmary of the phosphate product demand and supply situation is given below: BRAZIL - PHOSPHATE PRODUCT DEMAND/SUPPLY FORECAST (1,000 tons of P 2 05) Expected Demand North/Northeast Center ,075 1,393 South Brazil - Total IT, 1,380 1,30 1,990 Expected Suppl-y North/Northeast Center ,110 1,299 South Brazil - Total =78 1,097 I12 2,026 % Self-Sufficienc North/Northeast Center South Brazil / The P2C5 nutrient capacity of these four firms is 300,000 TPY of P205 on - the basis of maximum TSP output and only 210,000 TPY of P 2 05 on the basis of SSP production only.

86 ANNEX 3-2 Page The Center region is expected to be self-sufflcient in phosphoric acid between 1980 and 1983 with the VALEFERTIL and the ARAFERTIL projects. In addition, the region will also be about 85% to 95% self-sufficient in phosphate rock with the completion of the four projects at Araxa, Tapira, Ipanema and Patos de Minas (Table 18). The present plans for the Patos deposit envisions a mining/beneficiation project (310,000 TPY P 2 05) to be completed perhaps by 1981 to substitute for imported phosphate rock. At that time, the ARAFERTIL fertilizer project will likely come on stream and use the Araxa phosphate rock. Thus, the Patos project will continue to supply rock to the present producers, notably COPEBRAS and ULTRAFERTIL. An expansion project for Patos, probably to be ready by 1985 or 1986, will involve a doubling of the mining output for processing into phosphatic fertilizers. This expansion project will be identical in scale to the combined VALEP-VALEFERTIL complex and will provide the additional phosphoric acid capacity needed in the region. Development of one of the Catalao deposit for production by 1983 and the second Catalao project for operation by 1986 would ensure the Center region self-sufficiency in phosphate rock through the end of the s. G. Distribution System 53. The structure of the Brazilian marketing and distribution system is schematically shown in Chart IV. Most of the fertilizer passes through three levels on the way to final consumption: (i) primary production, (ii) blenders and (iii) wholesale and retail sales organization. In addition, domestic producers, who supply about one-third of consumption, are also usually involved in importing as well as blending and selling. About 90% of total consumption is processed through blenders, which are either owned by manufacturers (25% of total), independent companies (65%), or cooperatives (10%). Blenders are also involved in importing. 54. Sales organizations are either: (i) dependent organizations of the manufacturers, (ii) commission agents, (iii) independent dealers and (iv) cooperatives. Dependent organizations are extensions of the manufacturer's marketing organization and employees work on a fixed salary. Commission agents depend on the blending operation or the domestic manufacturers by an agreement which obliges them to sell only fertilizer products of the corresponding companies. The local agents who sell on a commission basis can be exclusive fertilizer dealers or general rural salesmen. Companies which sell by commission agents employ one regional sales manager and 4-6 inspectors for every local agents. The regional sales manager is a salaried employee of the company, the inspectors may or may not receive fixed salaries. Local agents receive a commission of about 4% from the sales as an average ranging from 3 to 10%. The agent helps the farmer in arranging bank credits and transportation. 55. Independent dealers have lost importance and this trend is continuing. At present they are primarily used in new areas in which the amount of fertilizer sold at the introductory phase does not pay a company-owned selling system. Similarly, cooperatives are not a major sales outlet, and are not expected to grow significantly in the near future. Sales on commission is still the most frequently used marketing channel. However, the exact

87 ANNEX 3-2 Page 2h distribution on the different market channels is not known, and the percentage numbers given in Chart IV are only rough estimates. 56. The distribution of storage between manufacturers, blenders and local retailers is not accurately known. Blenders try to avoid storage as far as possible because of the lack of working capital. Considering the cost of storage it is usually more profitable for them to adjust their blending operation to the seasonal structure of consumption by increasing the daily working time from 8 to 24 hours, than to store large quantitites over 9 months of the year. Farmers participate only marginally in storage also due primarily to lack of working capital and insufficient discount and credit facilities. Thus, most of storage is done by the manufacturers, who usually rent storage in the peak season. The problem of storage will become more serious with increasing degree of local production, since adjustment by imports will lose its present importance. 57. Transportation is one of the most important distribution elements in Brazil because of the long distance between production and consumption centers. It is also a significant constraint to the growth of fertilizer consumption especially in the regions which will become crucial for the further increase of agricultural production. The Government (GEIPOT) has embarked in 1975 on a country-wide transportation study and has completed Phase I which studied the present and expected transportation pattern of ten commodities which constitute the bulk of Brazil's transportation requirements. One of the commodities studied was fertilizer,and the following is an extract from the Phase I report. 58. In principle, highway transportation, railway transportation and coastal shipping compete against each other. But transport by ship along the coast or on the great rivers has not gained importance in domestic transportation. Road and rail transport are, therefore, the only significant means of transportation,and the split for the two modes is shown below for 1974: BRAZIL - TRANSPORT MODES OF FERTILIZER, 1974 (in %) Region Product By Road By Rail Total Northeast Finished Products Center Raw Materials Intermediate Products Finished Products South Finished Products Source: GEIPOT 59. Most fertilizers are thus moved by road transport despite the fact that rail transport may be cheaper in many cases. The principal reason for this pattern is the greater flexibility and convenience provided by road

88 ANNEX 3-2 Page 25 transport particularly for the distribution of finished fertilizer. Paradoxically, the share of rail transport in the Center is larger for finished products than for raw materials (15.2% versus 4.1%). This is due to the lack of a railway link between the fertilizer terminal in Santos and the railway network of the Center. All raw materials entering the port of Santos have to be loaded on trucks before they can be moved by railway. A rail link is being built and financed under the Bank's First Port Project. The share of fertilizer raw materials carried by the railways in the Center can be expected to increase once the line is completed. 60. The lack of specialized freight cars for corrosive bulk materials and the fact that many manufacturers and blenders do not have adequate facilities for loading and unloading bulk materials from railcars is one of the most important handicaps which have prevented the railways from capturing a larger share of the transport of raw and intermediate materials. Manufacturers generally have to bag intermediate products in order to use the railway's standard box cars. The blender receiving the material has to tear the bags as an additional operation before he can use the material. Unless the distances are very long, the additional cost of bagging and unbagging normally offsets the cost advantage of the railways, thus road transport, which is also generally faster, is used. 61. The industry would use the railway for transport of bulk materials but for the lack of proper railcars. To overcome this situation, it would be necessary to identify those transport routes which carry a considerable amount of bulk materials and to convert such routes to rail transport by working out an agreement between the companies involved and the railroad authority. The necessary analytical work for this is being done as part of the Fertilizer Market and Marketing Study (para 70). 62. The GEIPOT study also evaluated the port situation which is important to the fertilizer industry as over 50% of all the nutrients still come from overseas. In terms of importance, the ports of Santos and Rio Grande rank first and second and they handle about 80% of imported fertilizer materials. The table below gives a breakdown of the volumes handled by each port for 1974 and also an estimate of the fertilizer unloading capacity for some of the ports:

89 ANNEX 3-2 Page 26 BRAZIL - FERTILIZER IMPORTS THROUGH DIFFERENT PORTS, 1974 Quantity Average Unloading Rate Port (000 tonisl -7-f Total (Tons/Hour) Northeast Cabedalo 4 - N.A. Recife Maceio Salvador 25 1 N.A. nheus Total Center Santos 1, Paranagua N.A. Total 1,83 South Pbrto Alegre Rio Grande Total TOTAL 3, Rio Grande is the best equipped port with special fertilizer handling equipment and storage installed by FERTISUL and by LUCHSINGER. Santos also has a special fertilizer terminal with a nominal capacity of unloading 12,000 tons per day per ship. The actual unloading however, reaches only about 1,800 tons per day per ship because of the design of the berths, the low capacity of the grabs, and the tendency of the fertilizer to cake. The other difficulty in Santos is the lack of a railway link as mentioned earlier. The ports in the Northeast are much more inefficient, and Recife, which handles more than 50% of the imports for the Northeast will require a fertilizer terminal to cope with the larger volume expected in the future. 64. The new chemical port being built in Aratu (Bahia) to move the raw materials and finished products of the chemical complex in Camacari (which includes the PETROFERTIL plant) will have excellent fertilizer handling facilities. The excess fertilizer production of the Northeast can thus be very efficiently shipped from Aratu. 65. Port handling charges average about US$20 per ton depending on the port and the material handled. Costs in Rio Grande and the new port of Aratu are and will continue to be much lower than in other ports. 66. Table 23 gives a comparison of some representative routes and costs for rail and road transport in the Center and the South. It shows that while the railroads charge less than the trucking companies on a per ton-km basis, this advantage, particularly for distances less than 600 road km-, is offset by the fact that the trains have to cover a much greater distance between any two given points compared to highway travel. There appears to

90 ANNEX 3-2 Page 27 be an interesting regional difference between the Center and the South. The South's rail tariffs are considerably lower, while road transport charges appear higher, than the Center's. As a result, rail transport in the South is generally cheaper than by truck, while the Center shows a slight advantage for road transport. This probably explains why rail transport has a higher share of fertilizer shipments in the South than any other region. 67. The future transportation pattern will be significantly altered by the expected growth of the fertilizer sector and by the shift from imports to local production of raw and intermediate materials. This will generally ease the load on the ports, but will increase the inland transport requirements significantly. Starting in 1977 when PETROFERTIL's new ammonia/urea plant will commence production, large quantities of ammonia will have to be moved over long distances. At present, only small amounts of ammonia are moved, all by truck. The companies and the railroads have to make a considerable investment in ammonia moving facilities. PETROBRAS is presently studying this problem and specific recommendations are expected by early Because the Northeast will become a net exporter of nitrogen fertilizers, coastal shipping will become a more important mode of transport to move ammonia and urea. It is also possible that the production from the proposed plant at Norte Fluminense would best be moved by ship, further increasing the role of coastal shipping in the fertilizer sector. H. Proposed Fertilizer Market and Marketing Study 69. In connection with the Araucaria project, the Government has undertaken a study of the market and marketing aspects of the fertilizer sector. While the Government intends that the private sector handle the marketing and distribution of fertilizers, as has been the case in the past, it is crucial that the Government help identify potential bottlenecks in marketing and distribution, to ensure that the Government's policies and programs complement to the fullest degree, and facilitate, the rapid expansion of fertilizer use. The Government has also agreed to exchange views with the Bank regarding the expansion and development of the fertilizer sector including fertilizer import and pricing policy, and the results and recommendations of the proposed study. 70. The proposed study would have the following major objectives: (i) Develop a methodology for and estimate fertilizer demand and supply for a 10-year period and propose institutional arrangements for annual updates; (ii) Make recommendations as to the optimal location of future fertilizer plants considering regional increase in consumption and likely shifts in fertilizer distribution patterns;

91 ANNEX 3-2 Page 28 (iii) Identify specific constraints that may prevent the more efficient and broader use of fertilizers; (iv) Define improvements and/or expansion of storage, transportation and mixing/blending facilities that are required to support the expected increases in fertilizer production/ consumption; and (v) Recommend possible new Government policies (including credit and pricing) and programs that would be needed to ensure that the required marketing/distribution investments are undertaken either by the private or public sectors, and encourage broader and more efficient fertilizer use. 71. Detailed terms of reference for the study were prepared by IPEA, which will supervise the study, and have been reviewed by the Bank. It is expected that the final report of the study which began in October 1976 should be available in the fall of Industrial Projects Department March 1977

92 EXPORT PRICE INDICATIONS FOR SELECTED FERTILIZER MATERIALS (US $ PER TON OF PRODUCT) ~- ~- AMMONIA (f.o.b. W. E.rop,) ~- ~ ~ AMMONIA SULPHATE - AS (f.o.b. W. EPrope( UREA (bgpd, f.o.b. W. E-rop1) PHOSPHATE ROCK 75% TPL (f... Coublanc.7 PHOSPHORIC ACID (f.o.b. U.S. Gulf)....TRIPLE SUPERPHOSPHATE - TSP (f.o.b. U.S. GuIf( DIAMMONIUM PHOSPHATE- DAP (f.o.b. U.S. Gulf) POTASSIUM CHLORIDE (fo.b. Vanco-o' \ b0 200 L I i.'amm,,..i \i \DAP Acid 300 k ,_ 100 : t160 U, 1 t,..11 ii ' 1' A 15 / ~~~~~~~~~~~~~~~~~~... 1i EieF ho- j -l <-.0 Sowsso.r1+v//// ~~~~~~~~~~_ /C Ft JPFM A MJ JAS ON DJ F MA!11JiJA SO ND JPFM AM J JA SO N 9 Industrial Projects Department World B.nk March 1977

93 WisNEX 3-2 Chart II BRAZIL - VALEFERTIL FERTILIZER PROJECT APPARENT FERTILIZER CONSUMPTION IN BRAZIL (1,000 nutrient tons) 2300 NPK P 1000 _ 900 / /Estimated by ANDA~~~ I/ / 1 fent-nt.td by ANDA S..rce. SINOICATO OAS INJDUSTRIAS DE ADUBOS E COLAS DO ESTADO DE SAO PAULO Ind-strl rolecls Departmen A.g..t 1976 W.,nd B.nk-16425

94 BRAZIL - VALEIFERTIL FERTILIZER PROJECT APPARENT REGIONAL FERTILIZER CONSUMPTION IN BRAZIL (1,000 nutrient tons) NORTHEASTERN REGION CENTRAL REGION SOUTHERN REGION '400 NPK( I)00 K K0 NPK p * ~~~~ ~~~~~~NPK / P.- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.- ~~~~~~~~~~~~~~~~~~5...- Is ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ /,Esima-d by AN17A S,ceSINDICATO DAS INDUSTRIAS OE ADUAI3S E COLAS D0 ESTADO OE S013 PAULO I d-11w,d Plo-ets OaDIPnelVt,d A.ag., 1976 WO ak-68

95 ANNEX 3-2 Chart IV BRAZIL - VALEFERTIL FERTILIZER PROJECT STRUCTURE OF THE BRAZILIAN FERTILIZER MARKETING SYSTEM imports 100 (per cent of total consumption 65 nitrogen, 55 phosphate, 100 potash) 45 domestic manufacturers Iblending operation s ( per cent of total consumption blenders 2 owned by independent blending manufactu ret cooperatives 95 9 ~ ~ ~ ~ ~ dependent~~~ sales organizations organiath iones dependent independent cooperatives 5 _( ) ~~~~~~~~~~~~~~~~~~~~5 per centetoa\ / t of total consumption large farms other farms Source; National Fertilizer Study Industrial Projects Department August 1976 World Bank-16424

96 ANNEX 3-2 BRAZIL - VALEFERTIL FERTILIZER EOJECT BRAZILIAN PHOSPHATE DEPOSITS Average Ore Pbtential Reserves.2 Name of Deposit State Grade (% PpOc) (Million tons P205) 1. Jacupiranga Sao Paulo Ipanema Sao Paulo Araxa Minas Gerais Tapira Minas Gerais Patos de Minas Minas Gerais Catalao Goias Total 3 -/ Proven reserves are about 50% of the potential reserves. Other small phosphate deposits exist in Juqia (Sao Paulo) and the North and Northeast regions. 2/ Two groups, MNETAGO and MINERACAO CATALAO, propose to develop this deposit. Industrial Projects Department March 1977

97 kiliwx 3-2 Table 2 BRAZIL - VALEFERTIL FERTILIZER PROJECT POTENTIAL AGRICUIURAL AREAS 1. Brazil's total area 8,511,965 hm 3 (8.5 x l01 ha) 2. Used area 37-40,000,000 ha (4-5% of the Brazil area) 3. Total potential agricultural soil 413,000,000 ha 4. Potential agricultural soil that could be used (40-60%) a) Maximum 250,000,000 ha b) Medium 210,000,000 ha c) Minimum 165,000,000 ha 5. The medium potential (4.b) represents 25% of the Brazil area. 6. Potential area related to existing agriculture area Region Existing (A) Potential (B)A/ (A/B) x 100 (1,000 ha) (% total) (1,000 ha) (% total) 1972 North , Northeast 10, , West 2, , Southeast 9, , South 12, , Total 37, , / Useful area. Industrial Projects Department March 1977

98 ANNEX 3-2 Table 3 BRAZIL - VALEFERTIL FERTILIZER PROJECT PERCENTAGE OF CULTIVATED AREA DEFICIENT IN PHOSPHORUS AND POTASSIUM State PHOSPHORUS POTASSIUM Low Medium High Low Medium High Ceara Rio Grande do Norte Paraiba Pernambuco Alagoas Sergipe Bahia Mato Grosso Goias MAinas Gerais Espirito Santo Rio de Janeiro Sao Paulo Parana Santa Catarina Rio Grando do Sul Average Value Source: Estudo Nacional de Fertilizantes Industrial Projects Department March 1977

99 ANNEX 3-2 Table 4 BRAZIL - VAiEFhRTIL FERTILIZER PROJECT RECOMMENDED FERTILIZER APPLICATION RATES (kg/ha) Crops Re,gion N 2 K20 NPK North-East Cotton South-East South North-East Rice South-East South Central-West Pbtatoes South Cacao North-East South Coffee South-East North-East U8 359 Sugarcane South-East South North-East Corn South-East South Wheat South South-East Soybeans South North-East Tomato South-East 390 1, ,620 Source: Tecnologia Moderna Para a Agricultura Vo. II Brasilia 1975 Industrial Projects Department March 1977

100 ANNEX 3-2 Table 5 BRAZIL - VALEFERTIL FERTILIZER PROJECT SEASONAL DISTRIBUTION (%) OF FERTILIZER CONSUMPrION North Month and West Southeast South Brasil Northeast Jan/Feb Mar/Apr May/Jun Jul/Aug Sep/Oct Nov/Dec Source: Estudo Nacional de Fertilizantes Southeast and West form the Central Area. Industrial Projects Department March 1977

101 ANNEX 3-2 Table 6 BRAZIL - VALEFERTIL FERTILIZER PROJECT RURAL AND FERTILIZER CREDIT (billion current Cr$) Total Rural Credit: North/No theast Centerl/ South Total Fertilizer Credit: North/No theast N.A. Center1! N.A. South N.A. Total % Share of Fertilizer Credit of Total Rural Credit: North/No,theast N.A. Centerr N.A. South N.A. Total Regional Distribution of Fertilizer dit: North/No,rtheast N*A. Center& N.A. South N.A. Total I f.0 1/ Center excludes Parana which is included in South and which accounts for about 15% of total agricultural credit. Source: Banco do Brazil. Industrial Projects Department March 1977

102 BRAZIL - VALEFERTIL FERTILIZER PROJECT CULTIVATED AREA AND PRODUCTION OF MAJOR CROPS Culture NORTHEAST CENTRAL SOUTH BRAZIL 10 3 ha 103t lo3 ha t 10 3 ha lo3 t 1O3 ha 103 t Bananasl/ Black Bean 1, ,770 1, ,814 2,229 Cacao Cassava 1,178 13, , ,526 2,103 26,557 Castor Coffee ,008 1, ,080 1,746 Corn 2,264 1,573 5,337 8,875 2,307 3,661 9,908 14,109 Cotton (arboreous) 2, _ 2, Cotton (herbaceous) ,178 1, ,032 1,708 Grape Onion Orange2/ 25 1, , , ,646 Peanut Pota, , , ,149 Rice_/ 974 1,259 3,297 4, ,655 4,794 7,166 Soybean - - 1,188 1,886 2,427 3,126 3,615 5,012 Sugarcane ,610 1,065 52, ,913 1,959 91,878 Tobacco Tomato Wheat ,445 1,599 1,839 2,032 1/ Expressed in '1,000 bunchs." 2/ Expressed in "1,000 fanits." 3/ Including Hulls. Source: Anuario Estatistico do Brasil Industrial Projects Department a. March 1977 CD

103 BRAZIIL- VALEFERTIL FERTILIZER PROJECT PRICE INDEX OF SELECTED AGRICULTURAL PRODUCTS General Price Index Agricultural Products FARM PRICE INDEX IN REAL TERMS, Coffee Cacao Cassava Cotton Orange Corn Peanut Banana Tobacco Black Beans Rice Sugar Cane Wheat Potato Fertilizer Price Index (real terms) Source: FUNDACAO GETULIO VARGAS Industrial Projects Department March 1977 e

104 ANNEX 3-2 Table 9 BRAZIL - VALEFERTIL FERTILIZER PROJECT APPARENT FERTILIZER CONSUKPRION IN BRAZIL (1,000 nutrient tons) Year N K0 Total % Change N : P : K :2.6: :2.0: :2.1: :2.3: :2.4: :2.6: :1-7: :1.6: :1.9: :1.9: :1.6: :1.5: , :1.9: o 1, :2.1: , :2.2: , :2.3: / 436 1, , :2.3: / 484 i, , :2.4:1.4 2/ Estimated by ANDA. Source: Sindicato das Industrias de Adubos e Colas do Estado de Sao Paulo. Average Growth Rates (%) Period N p K20 Total Industrial Projects Department March 1977

105 --toaal 3-2 Table 10- BRAZIL - VAIZFTIL FERTILIZER PROJECT APPARENT REGIONAL FERTILIZER CONSUMPTION IN BRAZIL (1,00-0 nutrient tons) Northeastern Region Year N P205 K20 Total- Change N : P : K :1.5: : :1.1: :1.1: :1.1: :1.4: :1.5: :1.5: :1.2: / :1.2: / :1.2:1.0 Central Region : :1.7: :1.7: :1.4: :1.2: :1.5: :1.6: , :1.8: , :2.6: / , :1.9: L/ U , :1.9:1.3 Southern Region :2.7: :4.1: :3.3: :3.0: :3.2: :3.5: :3.7: :4.2: :3.8: / :4.3: / :4.3:1.6 1J Estimated by ANDA. Source: Sindicato das Industrias de Adubos e Colas do Estado de Sao Paulo. Industrial Projects Department March 1977

106 ANNEX 3-2 Table LI BRAZIL - VALEFERTIL FETILIZER PROJECT FERTILIZER CONSUMKPTION AVERAGE ANNUAL GROWTH RATES ( % Period & Region N P K20 Total Northeast Center South Source: Sindicato das Industrias de Adubos e Colas do Estado de Sao Paulo. Industrial Projects Department March 1977

107 ANNEX 3-2 Table 12 BRAZIL - VALEFERTIL FERTILIZER PROJECT SHARE OF NUTRIENT CONSUMPTION BY REC-ION (in %) Year N P K 2 0 Total Northeast J Average last 5 years Center Average last 5 years South Average last 5 years Industrial Projects Department March 1977

108 ANNEX Table 13 BRAZIL - VALEFERTIL FERTILIZER PROJECT STRUCTURE OF FERTILIZER CONSUMPFION (Percent) Year Product NITROGEN Ammonium Sulphate Urea Ammonium Nitrate Nitrochalk Amuonium Phosphate Other PHOSPHATE Single Superphosphate Triple Superphosphate Azmxonium Phosphates Other PCTASH 1-O XO- Muriate of Pbtash Pbtassium Sulphate Other Xo- 100 TO- s 10 Source: Sindicato das Industrias de Adubos e Colas do Estado de Sao Paulo. Industrial Projects Department March 1977

109 ANNEX 3-2 Table 1I4, BRAZIL - VALEFERTIL FERTILIZER PROJECT FIRTILIZER IMPORTS (1,000 t of nutrient) Imports P 0 K O as % of Total N Total Consumption o 1, ,247 7h , h , Source: Sindicato das Industrias de Adubos e Colas do Estado de Sao Paulo Industrial Projects Department March 1977

110 ANNEX 3-2 Table 15 BRAZIL - VALEFERTIL FERTILIZER FROJECT MONTHLY IMPORTS OF FERTILIZERS IN 1974 (Percent) Month Nitrogen PhoQphate Potash Total January February o 6.5 March April o May June July August September October November Decembpr Total Source: Sindicato das Industrias de Adubos e Colas do Estado de Sao Paulo Industrial Projects Department March 1977

111 BRAZIL - VALEFERTIL FERTILIZER PROJECT DOMESTIC PRODUCTION OF FERTILIZERS Year NITROGEN PIOSPHATWE TOTAL (N+P+K) 1,000 t Percent of 1,000 t Percent of 1,000 t Percent of Consumption Consumption g/ Only soluble phosphates. 2/ Including K 2 0 (all imported). Source: Sindicato das Industrias de Adubos e Colas do Estado de Sao Paulo. Industrial Projects Department March 1977

112 BRAZIL - VALEFERTIL FERTILIZER PROJECT PROJECTED AMMONIA SUPPLY (Production in 000 tons of N) Producer (State) Capacity Status2/ / 000 TYP N Existingl/ Fafer (Sao Paulo) 27 on-stream Ultrafertil (Sao Paulo) 123 on-stream Rh dia (Minas Gerais) 25 on-stream Petrofertil I (Bahia)) 55 on-stream Sub-total Proposed/Under Construction4/ Petrofertil II (Bahia) Araucaria (Parana) Sergipe (Sergipe) Norte Fhnuixngse (io.de J) eiro) Sub-total 1, Total 1, ,000 1,110 1,180 1,180 1,180 Required for Uses Other Than Fertilizer Supply Available for Fertilizer Use , of This in Northeast Center South / Existing plants assumed to operate at 90% capacity utilization 2/ Dates are year of initial operation 3/ Omits other plants which could be operating by 1985 F', 4/ Proposed plants assume to operate at 90% capacity utilization in third year of operation Industrial Projects Department March 1977

113 BRAZIL - VALEFERTIL FERTILIZER PROJECT FORECAST OF PHOSPHATE ROCK CAPACITY AND PROIJCTION (In 1000 tons of P ) Anntiel Year! P205 Content -stmate- Forecast -- F Region Capacity installed Beneficiated Rock I. Phosphate Rock Production Center Regioir/ Jacupiranga ( IMBRASIL) Araxa (CAMIG) v Catalao (METAGO) Patos (FERTILIZANTES)-/ Araxa (ARAFERTIL) Tapira (VALEP) Ipanema (QUIMBRASIL)A Patos (FERTILIZANTES) Catalao (METAGO) Catalao (MINERArAO) Total? II. Phosphate Rock Production As % of Total PK0 5 Demand Center Region Brazil (All Regions) Only the Center Region has commercial phosphate deposits to-date. 2/ One-half of the output is for direct application. The production is from an industrial pilot plant. 3/ For direct application. Industrial Projects Department March 1977

114 BRAZIL - VALEFERTIL FERTILIZER PROJECT PHOSPHORIC ACID CAPACITY. PRODWUCTION AND DEMAND FORECAST (In 1000 tons of P20 5 ) Annual Year -Estimate- Forecast REGION Capacity Installed I. PHOSPHORIC ACID PRODUCTION North/Northeast None Center QUIMBRASIL ULTRAFERTIL COPEGRAS VALEFERTIL ARAFERTIL Total South IUSCHINGER ICC Total All Regions II. PHOSPHORIC ACID DEMAND I, North/Northeast Center South Total National Surplus (Deficit) (90) (232) (376) (427) (554) (373) (349) (367) (397) (474) (487) % Self-Sufficiency (National) 64% 46% 34% 31% 33% 62% 68% 70% 69% 65% 64% 1/ Phosphoric acid demand is derived from the production forecasts for TSP (71% of P nutrient comes from phosphoric acid); NPK, DAP and MAP (1007. of nutrient comes from phosphoric acid); and double superphosphate ( with 44% of the P coming from phosphoric acid). X2 5 Industrial Projects Department March 1977

115 BRAZIL - VALEFERTIL FERTILIZER PROJECT FORECAST OF PHOSPHATE FERTILIZER PRODUCTION (In 1000 tons of P ) Annual Type of Capacity Year in -- Estimate-- RBGION AND COMPANY Fertilizer-l (1.000 tons P ) Operation Northeast PROFERTIL SSP 9 EXISTING TSP MAP AGROFERTIL MAP Total Center COPEBRAS MAP 60 EXISTING TSP 30 -do S do SSP 40 -do QUIMBRASIL MAP 65 -do SSP 35 -do ULTRAFERTIL DAP 70 -do INDAG 2/ TSP 80 -do SSP 53 -do UNIAO 2/ TSP 31 -do * SSP 32 -do FERTIBASE 2/ TSP 24 -do SSP 32 -do MANAH 2/ TSP 15 -do SSP 35 -do FERTICAP SSP 28 -do TSP FOSFANIL 5SP 21 EXISTING MITSUI THP 19 -do ELEKEIROZ SSP 17 -do VALEFERTIL MAP _ TSP ARAFERTIL MAP TSP Others (for direct application)-/ Phos. Rock TOTAL South FERTISUL 4/ DAP 51 EXISTING TSP 103 EXISTING SSP 11 -do MAP LUCHBINGER 5/ NPK 97 EXISTING DAP 34 -do TSP 61 -do SSP 4 -do MAP CRA 6 TSP 29 EXISTING SP 5 -do MAP ICC MAP _ Total BRAZIL TOTAL / DAP-diammonium phosphate; MAP - monoammoniom phospha,te; TSP - triple superphosphate; double superphosphate (0-30-0); SSP - single superphosphate; THP - thermophosphate; NPK - complex fertilizers. 2/ The TSP units of these four firms can produce either or both TSP and SSP. The production forecast assumes that the product mix will be adjusted to avoid a surplus of P nutrient in the Center region. 3/ See Table 19 for firms and year of installation. 4/ The capacity includes expansions of 50,000 TPY P 0 for TSP and 10,000 TPY P205 for PAP by 1977 and 1978 respectively. 5/ The NPK product has an average formula of (6-32-I2. The planned phosphoric acid plant (Table 20) is to substitute for imports and will not add to product capacity. 6/ The capacity includes an expansion of 7,000 TPY for TSP by Industrial Projects Department March 1977 o I

116 A 3-2 BRAZIL - VALEFERTIL FERTILIZER PROJECT D3W%LOSUPPLY BALANCE FOR NITROGU (1,0X tons of Nutrient) Northeast (14%) Center (64%) South (22%) LO All Brazil -0Jf8c SUPPLY br-theast o 450 Center South All Brazil SERPLUS (DFICI, Northoast Center (36) (184) (5) (225) 95 (290) 130 (200) 115 (175) 260 (100) 285 (115) 295 (165) 275 (275) 250 (370) South (108) (L20) (1 (170) h95 (210) _2X (265) g30 AU Brazil l u IO 4 Supply as % Dmand Industrial Projects Department March 1977

117 BRAZIL - VALEFERTIL FERTILIZER PROJECT FORECAST OF PHOSPHATE FERTILIZER DEKAND AND PRODWUCTION (In 1000 tons of P ) Demand and Production of Phosphate Products - Estimate - Forecast Demand North/Northeast Center South Total Production North/Northeast Center South Total Production as 7% of Demand North/Northeast Center South Brazil (all regions) Industrial Projects Department March 1977

118 BRAZIL - VALEFERTIL FERTILIZER PROJECT TRANSPORT DISTANCES AND COSTS, 1974 DISTANCE BY FREIGHT BY UNIT FREIGHT BY DISTANCE FREIGHT UNIT FREIGHT ROAD RAIL ROAD RAIL ROAD RAIL RATIO RATIO RATIO (km) (km) (Cr $/ton)i/ (Cr $/ton)z/ (Cr $/ton/km) (Cr $/ton/km) ROAD/RAIL ROAD/RAIL ROAD/RAIL From Sao Paulo to Registro Santos Ourinhos Bauru Pfesidente Prudente Marilia Catanduva Ribeirao Preto Goiania 939 1, Campos Vitoria 995 1, From Rio Grande to Sao Borja Santa Maria Santa Rosa Cachoeira Do Sul Baje / Represent normal tariffs. Higher tariffs prevail during the peak season,which is three-six months of the year. 2/ Published tariffs. Applied tariffs are often lowered through negotiated contracts. Source: GEIPOT Industrial Projects Department March 1977

119 BRAZIL-VALEFFERTIL FZRTILIZER PROJECT RAW MAIITERIAL AND PRODUCT FLOW R ~~~~~~~~~~~~~~~~~~~LIGLPHJRAID SULP SLJLPHUR S SLPHPH ACDR SULPHUIR SlORAGE r- 27,000} Th SULPF1URIC ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~SLPUI ACID peh l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~70 50M _ ; PHO~~SUPHORIC ACID_ ROCK ~ R)C PHOSPHATE PLN SOAElRA SILO XJ < = t STORAGE 0 0~STOAGE M 10TD jl AMMON1A~~~~~~~~~~~~~~~~7 T 2MAP Ml 250 MA 7lM R ACMMOSPHAT _ PRILLING GRANULATCURIN AND INDUSTRIAL~~~~~~~~~~~~~~~~~~~~~~~~~~R DEcEM8eR~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I - STO PRJETSDLARMETA Im ~~~~lu igoi) STz.-N~~~~~~~~~~~

120

121 BRAZIL IBRD ~~~~~~~~~A ,.frW2q0 PROPOSED _7 ~ ~ ~ ~ ~ ~ CACDI 14-0O0 NG TOWER F, 2 SULPHUR ZACID El IT CURING/DRY NO AREA ACID I 16 SLILPHCR MELT ['1G J 27 R.,I,..d 4 ~~PHOSPHORIC ACID r 17 REC:RTLLAT ON WATER POND p'--, 5 TSP I ~~~~~18 GENERAL ADMINISTRATION GEEICE FP, E.,-- 6 AP 19 PRODUCT ON OFFICE A LABORATORY ~~~~~~~~~~~~~~~~~~~~~~~~~PHOSPH 7SLHRC ACID STORAGE 20 INDUST AL MAINTENANCE 8PHOSPHORIC ACID) STORAGE 21 TRAIN MAINTENANCE STORAGE 22 STTOE: VLTGD 10FE I TORAGE 23 CLOAK ROOM/:ICO CA SECURITY 11 SULPHUR STORAGE 24 POOD SERVICE ~~~~~~~~2 FEAT ICER SSTORAGE 25 SKEPPICG A RECEIVING STATION 13 UT L TES DTATION 26 GARAGE & SERV CE 17 ~~~~~~~~~~~~~~~~~~~~~~~9AMMONIA : T NDUSTRIAL GATE - - 2~~~~~~~~~~~~~~9 TRUCK GA7E V.V I~~~~~~~V --. ~28[ 30 POWER 51JAS1S1 011 PiEIETERMINAE FACIEITIES 37 8FUTURE I AND PHOSP HATE ROCK STORAG E 1 A SIIEPHURIC ACIDm SiLPIH PC ACID CV ID IA ~~~~~~~~~~~~~32 26 ' PoOSPHOCRCACID TI ~~~~~~~~~~~~~~~~~~~~4 PHOSPHORIC ACID C! 1 L2~~~~. K' 2 15 GL.ILT Y A~~~~~~~~~~~35TS 1 ~~~36 MP ~~~~~~~-I- V~ I VV-~V-VI37 PHOCSPHORIC ACID STORAGE :0n 21 8STOJRAGE K 3 CCOOL 11G TOWER V. ~~~~~~~~~~~~~V'i ~~~~~~~~~~~~~~~~ 25 29~~~~ 40A REC Ro! A 1 Ot A VNAT POND VP MAP A ~ ~ ~ ~ ~ ~ ~ ~VI V V V - A MAP ~ ~~~----MAP 8 ~ T, DE5IVV T. I VAVT AVVI,V [ ~ ~ VAIIIVirII FRVVI I~~~~~~~~RVPVVIRdI 54 -~~~~~~ LOCATION OF PROJECT AREA 1 GO A S -PhV.VVV R-kVSI-, V PSV - 1 M I N A S\, G F R A I SMV-VVlVV-A SI.IV, ra RVVA-d- ~ I-- RVIV-VI 4 M IN/ASS GE RA I S I / J,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~V~ "7 1.4 d. I,4V4 0 A1 4 RIDI M > -~~~~~~~VV 0 -F OTVVHRZI1 PVVV0T5VL SROSO PAULO, 'VI I~~~~~~~~~~~~~~~~~~~~~~~~~ P,,IVVVE V,V~~~~~~ICDE JANEIRO IOVV/,VIASUT H A CVVV~~~~~~~~~~~~~~~~~~~~~~~~~~~VVV V ~~~~~~~~~~~~~~~~~~~~~A AM E,R I C A. A,4' S A 0 P A U L 0 ' - PARANA 'IVVVI I - i /4 IO04 V1, ~- 2 4A R h,a,s VV,V.CV`sV I-V

122

123 ANNEX 4-3 BRAZIL - VALEFERTIL FERTILIZER PROJECT PROJECT SCOPE, PROCE W DESCRIPTION AND POLLUTION OONTROL A. Project Scope 1. fthe project consists of (a) two sulphuric acid units each with a capacity of 1,300 TPD of 98.5% acid; (b) two phosphoric acid units each with a capacity of 470 TPD of P205 in the form of 52% acid; (c) a 1,000 TPD MAP unit; (d) a 1,100 TPD TSP unit; (e) storage facilities for 30,000 tons of sulphuric acid; 8,500 tons of phosphoric acid; 4L,000 tons of ammonia; 2,500 tons of fuel oil; 28,000 tons of sulphur, and 220,000 tons of finished products; (f) a rail spur and 25 ammonia tank wagons; and (g) ancillary facilities such as offices, laboratories, maintenance shops, effluent treatment, and the supply of utilities. The project will operate at achievable stream efficiencies of 302 days/year for the sulphuric acid units, 309 days/year for the phosphoric acid units and 330 days/year for the MAP and the TSP units.1/ The ammonia tankwagons are needed since the railways do not have enough wagons nor do they intend to purchase any for this project. Because the project would be the first to use the Tapira phosphate rock, there are inherent uncertainties as to the achievable levels of the phosphoric acid production. The phosphoric acid units have, therefore, been designed with additional design factors in the reactors and filters. Should the actual production experience approximate the test results, the phosphoric acid units may have about 10% more capacity than indicated above. To take advantage of this potential additional phosphoric acid capacity, the design capacities of the TSP unit and the sulphuric acid units are slightly larger than what would be needed to just match the phosphoric acid capacity indicated above. On the basis of the phosphoric acid design capacity, the project will have an annual production capacity of 330,000 TPY of MAP ( ) and 340,000 TPI of TSP (o-48.5-o). The design capacities of the different units are within the economic range and gives the project the advantage of economies of scale. 2. The two major raw materials required by the project at rated capacity will be about 277,000 TPY of sulphur and about 986,000 TPY of phosphate rock. Sulphur will be obtained by imports and moved by rail cars to the plant site. Rock phosphate will be supplied by VALEP - another subsidiary of CVRD, which is developing for this purpose, the rock phosphate deposits at Tapira in the State of Minas Gerais. The rock phosphate ore mined at Tapira will be beneficiated and carried through a slurry pipeline to the VALEP terminal adjacent to the VALEFERTIL project site. The rock phosphate requirements of VAIEFERTIL will be available partly (86%) as a slurry containing 63% solids for the phosphoric acid plant and partly (14%) as fine dry material for the TSP plant. The beneficiated rock phosphate is expected to contain 36+2% of P205, with not more than 3% of Fe2O3. Ammonia (45,500 TPI) for the production of MAP will be supplied initially from imports and then (1981/82) from the proposed PETROBRAS Ammonia/Urea complex at Norte Fluminense in the State of Rio de Janeiro. Ammonia will be moved in special tankwagons by railroad to the plant site. 1/ These stream efficiencies correspond to 100% annual capacity.

124 ANNEX 4-3 Page 2 B. Process Description 3. The sulphuric acid plant is based on the Bayer double catalyst process and can achieve a conversion efficiency of 99.5%. The basic design for the plant is provided by Lurgi of the Federal Republic of Germany. The total capacity of 2,600 TPD is provided in two trains of 1,300 TPD capacity each. Bright sulphur is melted with steam heating in melting pits and pumped through sulphur filters to the plant for combustion. Air for combustion is drawn from the atmosphere by the main air blower, filtered in a dry air filter, and dried with sulphuric acid in the drying tower. The sulphur is burnt in the air in a furnace to produce sulphur dioxide. The hot gases are cooled and the heat recovered, as medium pressure steam, in a waste heat boiler, followed by superheater and economizer. Sulphur dioxide in the gases is converted to sulphur trioxide over vanadium pentoxide catalyst, in a four layer catalytic converter. The gases are cooled in between the catalyst layers, in separately arranged waste heat boilers. Part of the sulphur trioxide is removed in an absorption system, after the second catalyst layer. This system improves conversion efficiency to 99.5%. The absorption system after the second catalyst layer absorbs about 87% to 90% of the total sulphur trioxide with sulphuric acid in a brick lined mild steel venturi. The rest of the sulphur trioxide, including that formed in the third and fourth catalyst layers, is absorbed in a final absorption system, similar to the intermediate one. The air drying towers and the absorption systems have mist eliminators to minimize carryover of acid mist. The product of 98.5% sulphuric acid purity is cooled and stored in two storage tanks each of 7200 M3 capacity. h. The phosphoric acid plant is designed and engineered based on the Rhone-Poulenc process, with basic designs from Krebs of France. Total production capacity of 940 TPD is provided in two trains of 470 TPD capacity each. Rock phosphate slurry, containing about 63% solids, is received at the VALEP terminal in Uberaba through a 9 inch diameter slurry pipeline, having capacity of 260 tons per hour of concentrate (ANNEX 4h-1). 5. The rock phosphate slurry from VALEP's pipeline terminal facilities is received in the phosphoric acid plant in two agitated storage tanks, each of 500 cu. m. capacity. Phosphate rock slurry, sulphuric acid and recycled phosphoric acid are introduced into a single tank agitated reactor with no inner partitions. Reactions take place with the formation of phosphoric acid and crystallization of gypsum. The reactor includes cooling arrangements for maintaining desired reaction temperatures. Slurry, overflowing from the reactor, is pumped to the feedbox of the continuous vacuum filter. Gaseous effluents from the reactor are washed for removal of dust and fluorine and sent to the stack. Phosphoric acid in the slurry is separated from gypsum in the filter. The gypsum cake is washed in multiple stages to minimize?2(5 loss. The plant produces phosphoric acid containing 30% P The gypsum cake from the filter, is slurried with water from the water pond and pumped to the gypsum pond. The 30% P205 acid is cooled in a stirred cooling tank to increase the precipitation of sludge and clarified in a decanter. The clarified acid is sent to a storage tank of 2500 m 3 capacity.

125 ANNEX 4-3 Page 3 6. The phosphoric acid concentration units are designed to remove 800 TPD of water in each stream to produce acid with 52% P The units include vacuum evaporators in which the acid is circulated, using carbon block heat exchangers for reheating. Steam from the unit contains fluorine and chlorine. The concentrated acid is pumped to four storage tanks each of 2500 m 3 capacity. The fluorine released from the evaporator is absorbed in a scrubbing tower in recirculated fluosilicic acid to produce 25% fluosilicic acid liquor. 7. The MAP plant has a capacity of 1000 TPD and is engineered with basic designs from Swift-Gulf Design. The 52% P 2 0< phosphoric acid, along with dust laden water from the plant, is reacted with ammonia in a Tee Reactor to produce non-granular MAP. For the production of granular MAP, phosphoric acid is pumped to the ammoniator-granulators. The acid is also pumped to the tower-granulator scrubbing system to capture ammonia fumes from the granulation system. Powder MAP from the MAP tower and granular MAP from the anmoniator-granulator are sent to the dryer-cooler to reduce the water content to 1 to 2%. The granular material is milled and screened before storage. 8. The TSP plant is engineered with basic designs from Gulf Design, a division of Badger (US). The plant has a capacity of 1100 TPD for producing TSP, either as a run-of-pile or granular product. Ground rock phosphate is fed continuously to a mixer cone, where it is mixed with preheated 52% P 2 05 phosphoric acid. The reacting material is discharged onto a slow moving slat conveyor in the Broadfield den. The conveyor is equipped with a revolving cutter at the discharge end and is enclosed in vented housing. The run-ofpile TSP is conveyed to a storage building, and cured for two weeks. Reclaimed TSP is granulated in granulating, drying, screening and milling units. The project has total product storage capacity of 100,000 tons for MAP and TSP. C. Plant Location and Supply of Utilities 9. The project site is within a recently established industrial park located along the right bank of the Rio Grande river, some 37 km south of Uberaba City. Uberaba is an important crossroads in the center region, being served by highways BR-365 (Sao Simao-Montes Claros), BR-050 (Sao Paulo-Goiania) and BR-262 (Vitoria-Belo Horizonte). It is also served by the FEPASA and the RFFSA railways, both in meter gauge, connecting Uberaba to Sao Paulo, Santos, Belo Horizonte and Brasilia. The project location is shown in Annex 4-2 (IBRD Map 12666). VALEFERTIL has already acquired about 48 ha of land for the project. Soil investigations and land grading have been completed. The project is the first to be erected in the industrial park, which is located between the FEPASA branch railroad currently unuer construction and highway BR-050 ensuring adequate transportation facilities. VALEFERTIL is making arrangements for the construction by FEPASA, of a railspur to the plant as well as the provision of railroad cars for the transport of sulphur, fuel oil and

126 ANNEX 4-3 Page h finished products. The external power requirements of the project will be provided through a nearby 30,000 KVA substation which is connected to the hydroelectric power plants along the Rio Grande river. A 2 km water pipeline from the Rio Grande river to the factory, with associated water treatment facilities, will supply the water for the project. D. Pollution Control 10. Special features, included while specifying and designing the facilities to minimize the emission and/or discharge of objectionable waste materials, are summarized below. 11. Sulphuric acid plant. The normal pollutants from this plant are the unconverted sulphur dioxide, unabsorbed sulphur trioxide and the acid mist in the stack gases. The sulphuric acid plant for the project is being engineered based on the Bayer double-catalyst, double-absorption process with a guaranteed sulphur conversion efficiency of 99.5%. The acid mist in the stack gases is minimized by the use of two-stage demisters after final absorption. The expected S02 in the exit gases is 1.96 kg per ton of H2SO4. The gases, after the final absorber and demisters, will be let into the atmosphere through a tall stack to ensure adequate dilution. 12. Phosphoric acid plant. Fluorine is the major pollutant and occurs during three of the process stages - (a) acidulation of rock phosphate in the reactor (b) filtration of the phosphoric acid and (c) concentration of the phosphoric acid. The reactor cooling gas effluent containing fluorine is subjected to two stage scrubbing - with recirculating pond water in the first stage and with used soft water from the vacuum pump scrubber in the second stage. The scrubbed gases are let into the atmosphere through stacks. Fluorine evolved during the filtration is scrubbed with soft water in washing condensers and the effluent gas exhausted at the vacuum pump discharge. About eighty percent of the fluorine evolved during acid concentration is recovered as fluosilicic acid, which will be converted either to cryolite or aluminum fluoride. The rest of the fluorine is condensed in the barometri condenser and sent to the pond water system. The pond water, which is kept recirculated, will be saturated with P 2 0s and fluorine. The pond will be designed with enough volumetri capacity as to ensure zero overflow. Any emergency overflow will be treated with lime to precipitate fluorine compounds. Of the total fluorine present in the rock phosphate 5h% will be recovered as hydrofluosilicic acid, 28% will remain in the concentrated phosphoric acid, and about 19% is discharged into the recirculating pond water via the barometri condensers and the recycled gypsum slurry water. The designs envisage emission into the air of about 0.4 percent of the fluorine in the rock phosphate or not exceeding 230 gms per ton of P 2 0s. In addition during high wind conditions, the pond will lose some fluorine into the air but at very low concentrations. 13. Monoaimmonium phosphate plant. The effluent gases from the prilling tower and the granulator contain ammonia and MAP dust, which are removed in

127 ANNEX 4-3 Page 5 tower scrubbers before the gases are exhausted to the stack. The dryercooler exhaust gas and vent air at de-dusting points will contain dust with ammonia and P 2 05 which are recovered in a scrubber before the gases are vented to the stack. 14. Triple Superphosphate plant. Vent gases from the acid/rock phosphate mixer and the den, contain fluorine, which is scrubbed in a scrubber with recirculating water. Dust-laden gases from the granulator and dryer are scrubbed with recirculating water. The recirculating water containing fluorine and P205 is sent to the cooling water pond. Scrubbed gases are vented to the atmosphere. 15. Auxiliary boiler. The project will have a fuel oil-fired auxiliary boiler with tons per hour capacity. The boiler will be used when required to meet shortages of steam especially during start ups and when the sulphuric acid plant is shut down. The flue gas which will contain sulphur dioxide will be let out through a stack, adequately sized to provide enough dispersion of S0 2. Industrial Projects Department January 1977

128 ANNEX h-4 Page 1 BRAZIL - VALEFERTIL FERTILIZER PROJBCT VALEP PHOSPHATE MINING PROJECT A. Introduction and Background 1. The VALEFERTIL Project has a design requirement of about 986,000 tons per year (TFY) of phosphate rock (35% P 2 05 content) which will be provided by Mineracao Vale do Paranaiba, S.A. (VALEP), a fully-owned subsidiary of CVRD. VALEP was established in December 197h, as successor to Rio Doce Titanio S.A. (TITANSA), earlier set up to exploit the titanium and niobium ores found in conjunction with the phosphate deposits at Tapira. The shift in commercial interest, from titanium to phosphate mining resulted from two factors: (a) the increase in international rock phosphate prices and the establishment of the Brazilian National Fertilizer Program to increase self-sufficiency in fertilizers; and (b) the development of a local beneficiation technology suited to the low grade igneous phosphate ores commonly found in Brazil. The beneficiation process can produce phosphate rock concentrate with 30% to 36% P205 from raw ore containing an average of 5% to 16% P205. The process has been well proven at the QUIMBRASIL mining/beneficiation complex at Jacupiranga and at the ARAFERTIL industrial pilot plant at Araxa. The P 2 05 recovery of the process is also satisfactory, ranging from 60% to 70% of the P205 contained in the raw ore. This recovery factor compares favorably with the standard metallurgical process used on U.S. sedimentary type low grade ore in the range of 5% to 10% P205 content which shows a recovery of just 30% to 50%. Although VALEP is presently concentrating its efforts on the phosphate mining/ beneficiation project, it is also continuing the research and development effort for exploiting the titanium and, in the near future, a research and development program for the niobium ore is also a possibility. VALEP has associated with New Jersey Zinc Company (U.S.) in a research effort for recovering a commercial titanium product from the Tapira ore. However, commercial exploitation of the titanium ore is not expected until after about five or more years (Section J). 2. At present, Brazil produces about 250,000 to 300,000 TFY of phosphate rock concentrate containing about 35% P 2 05 mostly from the Jacupiranga deposit, with small quantities coming from pilot plants or mining/washing operations fron the Araxa deposit for direct application. The phosphate rock production forecast is shown in ANNEX 3-2. AlU the production occurs in the Center Region as no significant phosphate deposit has been found in the other regions. The potential reserves and production capacity of the several phosphate mining/beneficiation projects are shown below :

129 ANNEX h-4 Page 2 BRAZIL - Phosphate Mining Projects and Reserves-/ Average Ore Annual Prodtl5tion Potential Grade Project Capacity- Reserves/ (% P20,) Completion (Thousand TPY P205) (Million Tons P205) Jacupiyanga Araxa Tapira Ipanema Patos de Minas or later Catalao (METAGO) or later Catalao (Mineracao Catalao) or later Total 1, a/ Small quantities for direct application are also produced or will be produced at Araxa, Patos and Catalao (ANNEX 3-2, Table 18) and are not included in this Table. b/ Most of the production is in the form of 35% P concentrate. c/ Approximately half of the reserves can be considered as proven and the P20 5 recovery will range from 60% to 70%. d! The P205 content of the Araxa deposit occurs partly as "locked" minerals that cannot be upgraded. Thus, the effective grade is less than the l5/o shown above. No information has been obtained as to whetler the Catalao deposit is of similar character. 3. Except for the Patos de Minas deposit which is sedimentary type rock, all the other deposits are igneous type rock. S'ialler igneous deposits are also found in Juquia (Sao Paulo) and in the Northeast but these are not expected to be convercially developed in any significant levels. B. The Tapira Phosphate Deposit 4. The Tapira phosphate deposit has been subdivided into several target areas, with target areas 1 and 4 having been sufficiently drilled to determine the reserves. Target area 1 covering about 2 x 1.5 km, is adjacent to the beneficiation plant and has adequate reserves for the initial 20 years of operation. Target area 2, about 2 km farther has twice the reserves of Target area 1. The reserves, on the basis of areas with at least 5% P 2 05 and with less than 40 mters of overburden, produce an average ore grade of 8% soluble P205 and are summarized below.

130 ANNEX h4- Page 3 Tapira - Phosphate Reserves in Target Area 1 & 2a/ (in million tons Of P ) Reserve Status-/ Mineable Reserves Recoverable Reservesi/ Measured Indicated 22 1 Inferred TOTAL 39 a/ About one-third Target Area 2. of the reserves is in Target Area 1 and two-thirds is in b/ The measured and indicated category correspond to proven reserves all the three categories correspond to potential reserves. while c/ Recoverable reserves at an average beneficiation plant recovery of 65%. (of P ) 5. Drilling of the Tapira reserves started in 1971 and was undertaken by DCCEGEO, another CVRD subsidiary. The measured reserves are based on a 200m x 200m drill pattern. The indicated reserves are based on the same drill spacing but with analytical chemical data extrapolated from only a portion of the holes. The inferred reserves are based on widely spaced drilling and mineralization was determined on the basis of a weighted value of the measured and indicated reserves. The entire reserve program was reviewed independently by Mr. Michel David of the Ecole Polytech of Montreal, Canada. As a result of this review, it was determined that drill holes on 400 meter centers would have been statistically adequate for measuring the reserves. The Target area 1 reserves will be adequate for 20 years production at 986,000 TPY of 35% P205 concentrate (about 3h5,000 TPY of P ). 6. The drilling program at the Tapira deposit included 186 holes, with an average depth of 65 meters and an average ore recovery of 83%. In addition to the drill holes the program included 5 large pits in the outcrop, several smaller vertical pits and trenches, several auger holes up to 80 cm in diameter, plus 5 drifts each approximately 40 meters into the outcrop. Approximately 95% of the drill cores as well as most of the ore from other exploratory work have been physically processed. 7. A review of the analytical procedures and calculations used to arrive at recoverable reserves shows the procedure to be valid. The direct P205 analysis of core samples have been correlated with beneficiation tests to prove that the direct chemical analysis is a valid method of core evaluation. A very exhaustive program of ore characterization related to processing has been carried out by VALEP. 8. The phosphate reserves are located generally as a slightly dipping elluvial layer within a large irregular shaped hill. The deposit is a typical alkaline pipe or carbonatite ring of igneous origin. Minerals associated with the hydroxapatite (phosphate ore), are chiefly vermiculite, magnetite,

131 ANNEX 4-4 Page 4 anatase (titanium ore), pyrochlore (niobium ore), and various clays and silicates. The anatase and pyrochlore are in sufficient quantities to be considered for future commercial development. C. VALEP Project Description and Implementation 9. The VALEP ProJect includes mining, beneficiation and product transport facilities for a production of 986,000 tons per year (TPY) of 35% P 2 05 phosphate concentrate. Drying and dry rock storage facilities will also be provided for 300,000 TPY of the production. The prime contractor for the project is a consortium known as "PAAA-LOGOS-EIM." PAAA, Paulo Abib Andery y Associado s/c Ltda and EIM, Engenharia para a Industria Mineral S.A., are specialists in minerals processing and engineering. PAAA developed the carbonatite phosphate beneficiation technology. LOGOS, Logos Engenharia, S.A., specializes in the management of large construction projects. Supervision of the prime contractor is handled by the VALEP Project Team and operations engineers. As the project nears completion a training program is planned for the operations personnel. The organization for the commercial operation has been planned in detail with an expected personnel requirement of 922. Most of the personnel will come from the Belo Horizonte mining area and the key personnel are expected to have some mining experience. An extensive training facility has been included within the main office building complex. 10. Beneficiation tests for the Tapira rock were carried out during 1975 at the ARAFERTIL industrial pilot plant. Engineering work started in 1975 and work on the beneficiation plant site began in mid Mechanical completion is scheduled for May 1978 and following three months of testing, commercial production should start August Power requirement of 26,000 KW for the VALEP Project will be supplied by CEMIG as hydroelectric power. Water will be pumped at 3,000 cubic meters per hour by VALEP from the Rio do Inferno, a distance of 9 km from the plant. The system is adequately designed with a 30% surge capacity complete with a large reservoir to easily accommodate the project's needs. A study was made to assure that use of the river water will always be available. Extensive maintenance facilities have also been provided to handle the normal requirements of a remote facility. D. Mining Plans 12. A relatively simple conventional shovel and truck open pit mine plan will be used at Tapira. The mine will be a "U" shape cut into the side of a hill. The mining will provide open faces of the entire ore body to permit conditions for selective mining and ideal blending. Haul distance for the ore is 1 km and for the overburden should average h km. Overburden thickness in later years will reach 40 meters (expected economi cut-off) and the ore thickness varies from 0 to 90 meters thick, and in the area to be mined about 30 to 40 meters thick. The overburden and the upper portion

132 ANNEX 4-4 Page )' of the ore is a soft earthy material containing some clay and tends to be quite sticky when wet. Drainage will be provided ahead of the mining face and surface drainage will be provided in the mine workings. The lower 30% of the ore is generally a highly weathered hardrock. Allowance has been made for blasting this material. 13. Initial equipment for the mine will be four, 9-cubic yard electric shovels (marion 151-M), two 15-cubic yard wheeled loaders, and ten 120-ton trucks. This equipment selection is optimum for the initial stage of the project where flexibility to suit weather and ore grade conditions is critical. Since the mine begins on the outcrop, the equipment for major overburden stripping is not required until During the interim VALEP could easily change its overburden stripping methods based upon its experience and economic studies presently underway for this aspect of the project. 14. The following table summarizes the expected annual volume of ore and overburden to be mined. VALEP - Annual Volume of Material to be Handled ( in miulion of bank cubic meters) YEARS to 5 6 to lo 10 to to 20 Phosphate Ore Overburden o Total -5s : Overburaen/Ore Equipment selection for the first 2 years has been made based on 4,200 effective mining hours per year (HPY) to handle 11.5 million TPY (5.0 million cubic meters) of material. VALEP - Mining Capacity (tons of material) Number of Tons/Hour Total Tons per Year Units Each Unit Tons/Hour (4,200 Hours) Shovels ,624,000 Loaders ,669,ooo Trucks ,900,000

133 ANNEX h4- Page 6 Tapira is located in an area of heavy rainful, averaging 1.5 meters annually and of poor drainage conditions. The rainy season spans six months from October to March, with the heaviest rainfall concentrated during the three-month period November to January. During these three months, mining at the Tapira mine can be considerably hampered by rains of approximately inches per month. To cope with this, the operating hours of the mine were discounted; the mining equipment has extra wide tracks and tires; the mining plan allows flexibility to mine areas least hampered by rain; and where permissible benches will be surfaced with waste rock or hardrock ore. In addition, hydrology studies are underway by Brazilian consultants on the feasibility of lowering the water table (pieziometric surface) in advance of mining. The final stages of the dewatering studies should be completed by mid These precautions should permit mining during the rainy season but at a lower level of production. 16. To maintain a constant feed to the beneficiation plant, two large storage piles will be provided. Each pile will accommodate up to 4 days of mine production which will last for 7 days of plant operation. The mine will have no difficulty insuring the plant a constant feed except during extended periods of continuous rain. Because of the large volume of materials involved, costs of stocking, reclaiming and inventory will limit raw ore pile to about two weeks of supply to the beneficiation plant. Thus it is very possible that the beneficiation plant will operate at less than capacity or even shut-down completely during certain times within the threemonths (November-January) of heaviest rainfall. E. Beneficiation Process and Plant Description 17. The Tapira ore will be beneficiated by a process similar to the one now in use at Jacupiranga and the ARAFERTIL pilot plant at Araxa. Briefly, the process involves crushing the ore to minus 8 inches, spreading it in either of two large storage piles, and reclaiming it in a manner such that optimum mixing is accomplished. The ore is further crushed and milled with rod and ball mills to minus 65 mesh, deslimed at 20 micron, sized into two fractions and then conditioned. The conditioning involves the addition of starch and other reagents as depressants for the iron, aluminum, and calcium gangue minerals. The fatty acid collector, frother, and ph control reagents are added at the flotation cells. Flotation includes several steps of rougher, cleaner, and scavenger cells. After sizing and regrinding the tails are recycled to a separate circuit for a final flotation step. The tails and slimes are thickened and pumped to the nearby disposal ponds. The concentrates are thickened and pumped to the slurry transport station. The main plant is under construction on a hill addacent to the Tapira deposit. The primary crushing station is located on the hillside near the base of the deposit. The ore will be trucked and dumped directly into the primary crusher and the crushed product will be conveyed by a 1 km belt to the main plant area.

134 ANNEx h-4 Page The process was initially established using batch tests of composite drill core samples from the Target Area 1. This was followed by continuous tests at the ARAFERTIL industrial pilot plant using 20,000 tons of ore at a feed rate of 15 tons per hour (TPH). The ore for these continuous tests was obtained from 5 large pits and blended to represent average ore characteristics of the deposit. A further series of tests were run to evaluate results under extreme variations in the ore quality. From this data, optimum metallurgical conditions were determined. At this time the metallurgical testing is complete but further tests are being carried out to optimize the operating parameters required for actual plant operations. These tests will be handled by the Brazilian consultancy firm of Paulo Abib Andrey and Associates in their bench and small batch type pilot plant. 19. The beneficiation plant could normally operate at about 6,600 HPY provided sufficient raw ore supply can be maintained at all times. Because of the restriction on the raw ore inventory and the reduced mining efficiency during the rainy season, the beneficiation plant may operate at lower levels of capacity utilization during the rainy season duie to inadequate supply of raw ore. Even by scheduling maintenance of the beneficiation plant during the rainy season, the available operating hours of the plant may only be about 5,900 to 6,000 HPY. 20. To supply the VALEFERTIL design requirement of 986,000 TPY of concentrate, the VALEP beneficiation plant should have a minimum design capacity.of 165 TPH of concentrate on the basis of a 6,000 HPY of operation, A 10% to 15% larger design capacity would be preferable considering the potential extra capacity of the VALEFERTIL project and the possibility of selling any small surplus to other producers. The engineering guarantee, the design capacity and the actual equipment capacity of the beneficiation plant is tabulated below: VALE? - Beneficiation Plant Capacity P205 Recovery % P205 Hourly Annual Capacity (1,000 TPY) Factor (%) Content5/ Capacity (TPH) at 6,000 HPY at 6,600 HPY Engineering Guarantee Design Capacity2/ Actual Equipment Capacity at: (a) ,010 (b) ,096 a/ The VALEFERTIL requirement has been expressed as 986,000 TPY of 35% P25 rock, - but the difference in concentration is not significant and was intended to be conservative as the guarantee is for a range of 34-35% P205. b/ The design capacity originally assumed 7,000 HPY of operation for about 980,000 TPY of concentrate capacity.

135 ANNA h- Page The P 2 05 recovery factor is a very difficult metallurgical variable to confirm, even in the larger pilot plant tests, thus, only operating experience, at high levels of capacity utilization, can determine the most likely recovery that could be achieved. It would be prudent, for planning purposes, to assume an initial recovery factor of 60% which may gradually increase to 65% only after a few years of experience. On a design basis, therefore, the VALEP plant is about 15% short in capacity for meeting fully the VALEFERTIL design requirements of rock phosphate. After a review of the actual equipment capacity, VALEP have confirmed that the hourly beneficiation capacities that can be achieved will be 153 TPH at a 60% recovery factor and 166 TPH at a 65% recovery factor, equivalent to 918,000 TPY and 996,000 TPY respectively of concentrate on the basis of 6,000 HPY of operation. This would just cover the design needs of VALEFERTIL (986,000 TPY) without any margin or provision for a potentially higher VALEFERTIL output. 22. It appears that it would be possible for VALEP to debottleneck its beneficiation plant and raise its equipment capacity by about 15% to about 1.1 million TPY of concentrate (assuming 6,000 HPY of operation) with an investment of about US$5 million. The investment would involve additional secondary crushing facilities (ball mills at a cost of about US$2-3 million), enlargement of the flotation cells (US$1 million) and some other additional equipment. But it may be premature at this stage to initiate a specific debottlenecking program. It would be advisable to allow VALEP and VALEFERTIL to operate for a period of about two years before debottlenecking is examined. The operating experience will then determine the recovery factors and operating hours that could be achieved and on the basis of this information, the two companies can then undertake whatever debottlenecking program proves economic and is necessary to balance the production of the two firms. Both firm could also opt for a major expansion if market demand is sufficient, but the expansion must be coordinated and planned to ensure a balanced system. 23. The phosphate rock slurry pipeline (Section F) and the primary crushing facilities have been designed to handle the equivalent of 2 million TPY of concentrate production, about two times the present capacity of the beneficiation plant in anticipation of a major expansion in the future. The mining capacity, especially during the dry season, can be easily increased. Thus, a modest increase in the beneficiation plant capacity will not cause any problems in these other areas. 2h. Although the use of other types of phosphate rock by VALEFETIL will be more expensive and technically difficult because the plant is being designed to use the Tapira rock with its unique characteristics, it would be desirable for VALEFERTIL to have the flexibility to use alternative rocks, should VALEP's production be inadequate for whatever reason. While the production efficiency of VALEFERTIL's phosphoric acid plant may be reduced when using other types of rock (e.g., Araxa, Patos, Morocco, Florida), such use may still be a better alternative rather than shutting down the plant temporarily. Finally, VALEFERTIL and VALEP should enter into a long-term supply contract to cover at least the design requirements of VALEFERTIL and to obligate VALEP to jointly undertake the debottlenecking study and program (if appropriate) mentioned above.

136 ANNEX h-4 Page 9 F. Phosphate Rock Slurry Pipeline 25. The rock concentrate will be transported as a slurry through a pipeline from the beneficiation plant to the VALEP terminal located near the VALEFERTIL plant site at Uberaba. From the beneficiation plant, the phosphate rock slurry will be initially pumped into either of two 3,000 ton capacity (dry rock basis) agitated tanks where the slurry will be tested for suitability for injection into the pipeline. When a tankful of slurry is ready, it will be pumped into the pipeline system for transport to the VALEP terminal at Uberaba. The system consists of positive displacement slurry pumps at Tapira and a 9-inch I.D. steel pipeline capable of transporting 260 TPH (dry basis) at a velocity of 5.h ft. per second. The design is such that the system could handle up to 2 million TPY of concentrate and could operate for 7,800 hours/year if necessary. Maximum particle size will be 65 mesh (21 microns) with a slurry density of 63% solids by weight. The specific gravity of the solids is about Bechtel (U.S.) is designing and will construct the system based on a contract awarded in early The design was at least 30% complete as of September 1976, and in Bechtel's opinion, it is a relati'vely simple and straightforward system cmpared to most pipelines they have engineered. The system can be restarted underload in case of emergency shutdown. Normal shutdown on a daily basis will involve flushing the pipeline with clear water. A thickener at the pipeline terminal provides for recovery of product when the system is flushed with water. 27. The pipeline is 118 km. long, upgrade from elevation 1,180 meters to elevation 1,350 meters in the first 6 km., and then is down grade 112 km. to the terminal at elevation 5h0 meters. Average slopes range 2% on the upgrade and 0.1% on the downgrade. Maximum slope for any single point in the line dictated by design, is 15%. The topography is generally rolling hills the first 30 km. with the balance along the valley adjacent to the Rio Grande River. The pipeline will be buriad a minimum of 70 cm deep. The expected life for the pipe is over 20 years. 28. The slurry from the pipeline system will be discharged to either of three points: a 100,000 ton capacity storage pond; a large thickener; or to either of two 1,500 ton (dry basis) storage tanks. The storage pond provides extra surge capacity should the chemical plant become inoperative. When required, product from the storage pond will be reclaimed with a dredge, thickened to a density of 63% solids in the pipeline flushing system thickener, and pumped to the 1,500 ton tanks as feed to the chemical plant.

137 ANNEX 4-h Page When required, a portion of the chemical plant feed from the storage tank can be filtered. The filtered product can then be combined with the chemical plant feed to provide a more dense feed slurry. When excess concentrate is available or when TSP plant feed is required, a portion of this filter cake will be dried and sized with cyclones. The coarse portion can be sold as dry rock and the fine rock conveyed to the TSP plant. The drying system has a capacity of 300,000 TPY of concentrate. At times when large amounts of concentrate must be stored, as when the chemical plant is shut down for extended periods, a large open area is provided for storage of filtered concentrate. This storage and reclaim system would consist of a wheeled loader and truck. The filter, drier, cyclones and dust (fine phosphate rock) collectors are still in the preliminary design stage, and VALEP is expected to finalize the design shortly. Both the slurry pipeline and the terminal at Uberaba will be installed, owned and operated by VALEP. G. Environmental Aspects 30. The terrain near the beneficiation plant site provides many excellent locations for tailings disposal. Tailings will initially be disposed along the face of the slimes dams and in later years, the tailings will be stacked on the adjacent hillsides. Water from the tailings will be reclaimed from the slimes ponds and recirculated to the plant facilities. The slimes, which are at least half clay and half grinding slimes, settle well. The reservoirs for the slimes will produce a water overflow which can be reused. VALEP expects to recycle at least 70% of its total water use of 10,000 cubic meters per hour. The remaining 30% will be clarified so that it does not present a pollution problem after discharge. H. Capital Cost Estimate and Financing Plan 31. The total financing required for the VALEP project including the pipeline and the terminal is about US$216.1 million equivalent, with some US$56.6 million for foreign exchange cost. The capital cost estimate is shown in the following page. It is based on a detailed list of equipment and cost items and already reflects purchase cost information on at least 30% of the equipment already committed as of October 1976 and price quotations for most of the remaining major equipment items. The base cost estimate is 5% higher than VALEP's 1975 estimate even without the additional provision for the pipeline and terminal facilities which was not included in the original VAIEP cost estimate. The estimate has no provisions for import duties as VALEP has been granted a 100% exemption. However, physical contingencies and price escalation provisions are included. The working capital estimate is much lower than the original VALEP provision and reflects the most recent pricing arrangement (tentative) for the rock concentrate of US$55/ton delivered at Uberaba and the finished goods inventory implied by the facilities now tentatively planned at the pipeline terminal facilities.

138 ANNEX 4-4 Page 11 VALEP - Capital Cost Estimate-/ (In Millions of US Dollars) Foreign Exchange Cost Local Direct Indirect Total Currency Total Imports (CIF) Costs Foreign Cost Cost Equipment & Materials Engineering & Supervision Civil Works & Erection Land & Improvements Pre-operating & Stert-up Uberaba Facilities_/.! Base Cost Estimate (BCEgp i Physical Contingencies (PC)A" Price Escalatior 1 f Installed Cost Permanent Working Capita If/ Interest During Constructiong-/ 5.3-5L Total Financing Required a/ On the basis of an official exchange rate of Cr$11.40 US$1.00. b/ These facilities were not included in the original VALEP estimate and covers the slurry filter, drier, cyclones and dust (fine phosphate rock) collectors to be installed at the Uberaba pipeline terminal. c/ In mid-1976 US dollars. d/ 5% on equipment, and on land and improvements; 10% on all other items. a/ On the basis of international inflation rates in Annex 5-2. f/ Escalated to October g/ Estimated on the basis of a 60% debt financing.

139 ANNEx h4 Page The VALEP project is being financed with 40% equity (US$86.5 million) from CVRD and loans from foreign suppliers credits for the directly imported materials and services (17% or US$37.1 million), from FINAME (a BNDE subsidiary) for financing up to a maximum of 80% of the ex factory cost of locally supplied equipment (1h% or US$31.3 million), from BNDE (the national development bank) to finance mostly civil works and erection (22% or US$h7.2 million) and from FINEP (a Government agency) to finance up to 80% of the local engineering and technical consultancy services (7% or US$14.0 million). However, as the original cost estimate did not include provision for the pipeline terminal facilities and in addition, the base cost estimate has now increased by about 5% over the original value, there is a financing gap of about US$20 million equivalent, mostly for local purchases. VALEP has initiated arrangements for supplemental financing from BNDE/FINAME, which is included in the BNDE/FINAME loan amounts indicated above. CVRD is assisting VALEP to complete the financing package and has assured the Bank appraisal mission that funds will be made available as needed to complete the project or, time. 33. The BNDE/FINAME loan and the FINEP loan have the usual maturities of just about eight to ten years while the foreign suppliers' credits have maturities of from six to eight years. The grace periods for the loans range from three to four years. Given the short average loan maturity, it is very important that VALEP maintain the original plan of a debt/equity ratio of 60:4O or better at project completion. 34. The mine and beneficiation plant facilities are designed for a 20-year operating life, and during this period, additional capital investments will be required, mostly due to an increase in overburden stripping. The additional investment is shown below, and about 60% of these costs are expected to be in foreign exchange. VALEP - Additional Capital Investment (In Millions of Constant 1976 US Dollars) Operating Year Amount 3 (1980/81) (1983/8h) (1988/89) (1993/94) 12.9 Total : 31.4 I. Financial and economic Analysis 35. The VALEP financial and economic operating cost estimate is shown in Table 1 and summarized below:

140 ANNA h-h Page 13 WALE? - 0erating Cost Estimat! (in constant 1976 US dolla-rs;j' Financial Costs Economic Coats Annual Cost US $ Annual Cost US $ (US$ Million) Per Ton (US$ Million) Per Ton Variable Costs-/ Fixed Costs 1l.8c' 68.9/. 7.0 Total Cost 3. T a/ At 90% capacity utilization and using a shadow exchange rate for the economic cost. b/ Excludes drying nsft of USt,'-;!`o of concentrate. OrC.y 14% of the production (for the TSP plant) will be dried and will add US$0.4 million to the annual variable cost at 90%.f-tnnt. c/ Includes depreciationi blf excludes financial charges. Financial charges are expected to be about US$5/torn initially and will probably average at MS$3/tor (in constant 1976 US dollars) over the first seven years of operation when the loans are outstanding. The financial operating cost of about US$37/ton of concentrate (before financial charges) is slightly higher than the projected international FOB price of high grade phosphate rock of US$35/ton (in constant 1976 US dollars) over the life of the VALEFERTIL project. But ocean freight, port handling and inland transport would raise the delivered cost to Uberaba of imported rock to about US$65/ton. 36. The initial production of VALEP which is not used by VALEFERTIL is assumed to be dried and sold to other users at a price of US$45/ton, net of drying cost. As VALEP will have a drying capacity of only 300,000 TiY of concentrate, the financial forecast assumes that the VALEP production that can not be dried due to the drying capacity constraint will be temporarily stored and then gradually 30ld. The production and sales forecast are shown below: VALEP - Forecast of Production and Sales (1000 tons of concentrate) VALEP Production!/ VALEFERTIL Requirement Sold % Capacity Tons to Year Utilized Produced Wet Dried Total Others Stored h h (48.6) h (175.8) 1982 go 887.h h 3.6 ( 3.6) h h h a/ Five months of operation during 1978 at 45% capacity utilization. VALEP can and should bs able to achieve 100% 1 capacity utilization.but for the financial and the economic analysis, only a' Y0% rate is assumed after 1979.

141 ANNEX h-h Page The financial rate of return before income taxes is about 9% (Table 2). The after tax return will probably be about 2 percentage points lower but cannot be reasonably ascertained at this point since the final debt/equity ratio and the loan terms can only be roughly estimated. Sensitivity tests (Table 2) show that if production is increased by 11% starting in 1980 over the output assumed above,1/ even with an additional cost of US$5 million (para 22), the before tax rate of return will improve to 10.2%. 38. The economic rate of return of the VALEP project without the VALEFERTIL project (i.e. all output is sold at US$b5/ton to manufacturers in the coastal areas) is only 9.6% but improves to 10.8% if production is increased by 11% and to )L.5% if output is sold at US$55/ton. This implies that import substitution can be economically viable only if the Tapira rock is processed close to the mine and sold in the same area to take full advantage of transportation savings, as indicated by the 16.6% economic rate of return of the combined VALEP and VALEFERTIL projects (Annex 7-2).2/ J. Commercial Prospects for Titanium and Niobium Ores 39. The VALEP anatase (titanium) ore reserves at Tapira total 200 million tons of ore with an average grade of 21% TiO 2 on the basis of a lower cut-off grade of 15% TiO2. The anatase will be stripped with the overburden and stored in a separate area for future use. VALEP is undertaking a research and development effort (para. 1) to develop a commercial process to extract titanium from anatase. The normal ore minerals of titanium, namely rutile and ilmenite have well proven economic processes for the extraction of TiO2. But anatase, which is a polymorph to rutile, is not amenable to these standard TiO 2 extraction processes. It is well known in the industry however that anatase will convert to rutile at +7500C but with various other minerals and elements present in the Tapira anatase ore, additional research is needed to verify the feasibility of the heat treating process. Should such a technology be developed, it is expected that the titanium venture could share the overall mining costs with the phosphate mining/beneficiation venture thereby potentially reducing the cost of production of the phosphate rock concentrate by about US$2-3/ton (in 1976 US dollars). The titanium venture itself should be able to earn a good return on investment provided the market for VALEP's titanium can be developed. 40. Should a heat treating process convert the anatase to rutile at say a cost of US$10 per ton of product and with mining and beneficiation at US$60/ton,the production cost (before depreciation and financial charges) per ton of product might be in the range of US$70 to 100. Under such a cost basis and a sales price of US$500/to4 one can readily see the merits of developing a viable process for the titanium ore. The capital requirements could hardly exceed US$100 million (in 1976 US dollars) for such a project. The above costs are rough preliminary estimates but serve to bracket what 1/ Equivalent to 100% capacity utilization by 1980 for VALEP. 7/ The economic rate of return of the combined VALEP-VALEFERTIL operation is 18.h% at an 11% higher production.

142 ANNE 4-i4 Page 15 might be expected from the VALEP titanium ore. On the other hand, the world rutile output including ilmenite adjusted to a rutile base is in the range of 3 to h million tons per year. A VALEP 5 million tons per year anatase ore production when converted to rutile would be equivalent to 10% of the world's present rutile output. Some assurance of a market would therefore be a crucial step in the VALEP anatase ore development. 41. One area of the Tapira phosphate reserves includes, in the overburden, an indicated reserve of Niobium ore (pyrochlore) estimated at 60 million tons with 1.5% Nb on the basis of a cut-off grade of 0.5% Nb VALEP has done very little work on development of this ore other than outlining it in their reserve picture. The nearby niobium ore deposit being processed by CBMM (Compania Brazilia Metallurgia e Mineracao) at Araxa has a grade of 3.5% Nb205 CBBM has reserves of about 500 million tons at 3.5% Nb20 5 and over i billion tons at 2.5% Nb2O<. In 1972 CBMM produced 7600 tons of 59% Nb 2 05 concentrates, 60% of the world production.!/ The 1975 production is claimed to be 80% of the world output. Even with concentrates selling at approximately US$3000 per ton and production costs a mere fraction of the selling price, VALEP would find it very difficult to enter the market which is evidently rather well controlled. For this reason, the Nb205 potential for VALEP is not as promising as the titanium venture. It is understood that a new plant at Catalao (METAGO) for producing niobium concentrates is being implemented. Industrial Projects Department January / MPst of this data is from an article in Mining Engineering, Society of Mining Engineers, Nov

143 BRAZIL - VALEFERTIL FERTILIZER PROJECT.al VALEP OPERATING COST ESTIMATE IN CONSTANT 1976 US DOLLARS -/ Financial Cost Economic Cost Annual Cost (US$ Million) At The Official Exchange Rate At A Shadow Foreign Exchange Rate Foreign Local Total US$ Annual Cost US$ Annual Cost US$ Variable Costs Exchange Currency Cost per ton (US$ Million) per ton (US$ Million) per ton Fuel & Lubricants Grinding Steel Reagents Explosives Power Maintenance Supplies Total Variable Cost Fixed Costs Labor & Administration lnsurance b/ o Depreciation Total Fixed Cost Total Production Cost Before Financial Charges {/ At a 90% capacity utilization equivalent to an output of 887,400 TPY of concentrate. The official exchange rate used is Cr$11.40 US$1.00 and the shadow exchange rate is 25% higher at Cr$14.25-US$1.00. b/ Derived as follows: In US$ Million At Acquisition Cost In Constant 1976 US Dollars Total Financing Less: Working Capital Land & Improvements Spares Non-Depreciable Assets Depreciable Assets Annual Depreciation (straight line, 20 years) c/ Excludes drying cost of about US$3/ton of concentrate (financial cost). Only a small portion of the VALEFERTIL requirement (14%) will be dried for the TSP plant. Dried rock sold to other users at US$45/ton is net of drying cost. Drying cost for VALEFERTIL TSP plant needs will add US$0.4 million to the annual cost at 90% capacity utilization. Industrial Projects Department January, 1977 M- JS Y.

144 ANNEX 4-4 Table 2 BRAZIL - VALEFERTIL FERTILIZER PROJECT VALEP FINANCIAL RATE OF RETURN AND SENSITIVITY ANALYSIS A. Cash Flow Cash Flow in Constant 1976 US Dollars Capital Operating Cost Revenue (Sales) Net Cash Flow Year Cost Fixed Variable Drying-lJ VALEFERTIL_/ OthersFJ Before Taxes 1. (1976) 63.2 _ (63.2) (74.9) 3. (1978) 35.3d/ (34.3) (1980) (1985) (1990) (1995) (1998) (12.3)t/ Base Case Financial Rate of Return Before Taxes = 9.0% a/ The drying cost for VALEFERTIL's TSP plant rock needs. The drying cost for the rock sold to other users is not included as the selling price of US$45/ton is already net of drying cost of US$3/ton. b/ At US$55/ton (for both wet and dried rock). The VALEFERTIL operating life is taken to be about 12 years to c/ At US$45/ton net of drying cost. All the VALIEP production after 1991 is assumed to be sold at US$45/ton. d/ Includes US$5.8 million for working capital. e/ Non-depreciable assets. B. Sensitivity Analysis Sensitivity Analysis - Financial Rate of Return Before Taxes, % Product Prices -15% -10% -5% Base Case +5% +10% +15% Costs Base Case Capital Cost Up 5% Operating Cost Up 10% Operating Cost Down 5% Output Increased by 11% a a/ Assuming production is higher starting 1980 to match on 11% higher VALEFERTIL production with an additional capital cost to VALEP of US$5 million in Industrial Projects Department January, 1977

145 BRAZIL - VALEFERTIL FERTILIZER PROJECT VALEP ECONOMIC RATE OF RETURN AND SENSITIVITY ANALYSIS al A. Economic Cost and Benefit Streams (In Millions of Constant 1976 US Dollars) At A Shadow Foreign Exchange Rate At The Official Foreign Exchange Rate Capital Operating Gross Net Capital Operating Gross Net Year Cost Cost Benefit Benefit Cost Cost Benefit Benefit 1 (1976) (55.4 ) (63.2 ) (63.0 ) (74.9 ) 3 (1978) (30.6) (38.2) (1980) (1985) (1990) iS (1995) (1998) (11.1) (12.3) Base Case Economic Rate of Return = 9.6% Base Case Economic Rate of Return -5.2% a/ Assuming no VALEFERTIL project and all output is sold to coastal manufacturers at an economic cost of US$45/ton net of drying costs. The official exchange rate is Cr$11.40=US$1.00 and the shadow exchange rate is 25% higher at Cr$14.25=US1.00. B. Sensitivity Analysis Cost Sensitivity Analysis - Economic Rate of Return % Product Price - 15% -10% -5% Base Case +5% +10% +15% Base Case , Capital Cost Up 5% Operating Cost Up 10% Operating Cost Down 5% a.2 Output ancreased by 117a Base Case at Official Exchange Rate a/ Assuming production is higher starting 1980 to match an 11% higher VALEFERTIL requirement but at an additional capital cost to VALEP of US$4.6 million at the shadow exchange rate (US$5million at the official rate) in Industrial Projects Department January, 1977

146 BRAZIL - VALEFERTIL FERTILIZER PROJECT IMPLEMENTATION SCHEDULE M I A, M 1 J!A 1"A I A I S Io P N I D _ M IA; M;I J J A S I 0 N D i F I M A J J I A 15 0 N D J F M I A~ J I J A II1 GULP ACKAGEt MAPTSP' A. C -itbalncbasices ', F_ G A TIL DESIG N EASIN SORCIE - CRIC-SOAL I IPM _ENTS gq111tmtnttdata S -lh(' DGERLL -TREElSAh UTIt - t PE I'I<i' ZIO tt -t - t 'BLKMA'LAI<= COMPR. SSPS&(IFICEAT IOT. N e W X =0 X -DOA I A tc INS t.e~~- CAI DO FRUGAEAIA MATER.f A ID TLIN ~~~~~~~~~~~~~DTETATCUSH BAA] EPETIOF F AT UTIELD DUIN TE ME BASEC ARRANtEMENTSV MI 'onc IND CT_METEES CC= TOO GrIDSt ME-A;= AGE aehanialccaltio I IIN AC, C R w 'SS" AG. 37 IR.CUr EtnE.NT lo *et [ ETvWrdx 67 jl)= F T DATA S MLTEE -EIS EA ll F IGN ENGINE-_N._N w,, D,.

147 ANNEX 5-1 BRAZIL - VALEFERTIL FERTILIZER PROJECT CAPITAL COST ESTIMATE (In millions of US Dollars)-/ Foreign Exchange Costs Local Foreign Exchange Total Currency Total Total Direct Part of Lonql Foreign Part of Local Local Project Imorts Purchases Exchange Purchases Purchases Cost _ Equipment and Materials S/ a. Sulphuric acid units b. Phosphoric acid units c. MAP unit , ,1 d. TSP unit e. Offsitad/and utilities , f. Spares g. Ocean freight and Insurance 3.8 _ _ 3.8 Sub-Total Inland Freight and Taxes a. Freight ,6 b. Taxes ,4 9.4 Sub-Total Civil Works a. Plant foundations b. Plant buildings c. Nonplant buildings d. Drainage and streets e. Fertilizer storage f. Others Sub-Total Erection a. Sulphuric acid unit b. Phosphoric acid unit c. MAP unit d. TSP unit e. Offsite and utilities _ Sub-Total Land and Site Improvement a. Land b. Survey c. Excavation and fill d. Road access e. Rail access f. Power line g. others - - _ Sub-Total Engineering and Other Services a. Lurgi b. Krebs c. Gulf-Design (Badger) d. Flour e. Natron f. Technical assistance g. Organisational assistance and others _ Sub-Total Preoperating and Administrative Expenses Testing and Commissioning Base Cost Estimate (BCA) 5/ Physical Contingenqy ,4 9 Price Escalation ' Installed Cost Working Capital Interest during Construction Total Financing a/ An exchange rate of Cr$11.40 = US$1.00 is used in this report. 'S/ The foreign exchange element of local purchases are assumed as follows - Equipment and Supplies - 25%; Civil Works ; Erection and other locally provided services - 0%. Permanent Working Capital - taken from Annex 5-3 (escalated). ^g Cost of imported supplies are based on FOB port supplying country. Cost of local supplies are ex-factory. U Spares provided at 67, of imported supplies and 37. of local supplies. B Ease cost estimate at mid-1976 delivery prices. l 5% of engineering costs, 10% of other costs. Price escalation of 14.5% for civil works, 0% for engineering and 19.7% for all others. Industrial Projects Department January 1977

148 ANNEX 5-2 A. Assumed Inflation Rates BRAZIL - VALEFERTIL FERTILIZER PROJECT CALCULATION OF PRICE ESCALATION FACTORS % Annual Inflation Rate % Compounded Inflation Rate-/ Equipment & Civil Equipment & Civil Year General Price Works General Price Works Level Level Onwards 5 _ a/ 1976 inflation rates are for second half of the year inflation rates are for expenditures during first half of the year. B. Capital Cost Disbursement Schedules (Base Cost Estimate and Physical Contingencies) $ Disbursed Year Equipment & Civil Local and Foreign Engineering General Gosts Works Costs o C. Price Escalation Factor a/ Weighted % Price Escalation - Year Equipment & Civil Local and Foreign Engineering General Costs Works Costs 1976 _ Total price escalation 1h a/ The compounded price escalation rate from (A) weighted by the fraction disbursed during the year from (B) Project Base Escalation Escalation Cost & Factor % (US$Million) Physical Contingencies (USmillion) Engineering Costs Civil Works All other costs Total Industrial Projects Department November,1976

149 ANNEX 5-3 BRAZIL - VALEFERTIL FERTILIZER PROJECT PERMANENT WORKING CAPITAL ESTIMATE A. Accounts Receivable (60 days sale) (US$ Millions) (a) TSP 66,000 tons 4 $225/ton (bulk) 14.9 (b) MAP 60,000 $2 8 0/ton (bulk) 16.8 B. Finished and Intermadiate Goods Inventory a/ (At Cost Excluding Depreciation) (a) TSP 33,000 tons a $1 4 0/ton (30 days) 4.6 (b) MAP 30,000 $195/ton (30 days) 5.8 (c) Sulphuric acid 15,000 $40/ton o.6 (50% of storage) (d) Phosphoric acid (52%) 8,500 tons a $150/ton C. Raw Materials at Costa- (a) Sulphur 27,500 tons at $95/ton (30 days) 2.6 (b) Ammonia 2,000 tons 4 $215/ton (50% of storage) 0.4 (c) Phosphate rock 3,000 tons *@ $55/ton (1 day) 0.2 (d) Fuel oil 2,400 $65/ton (30 days) D. Spare parts at acquisition cost 4.9 E. Minimum cash requirement (1/12 of annual fixed cost less depreciation) 1.2 F. L,ess (a) Spare parts included in capital cost 4.9 (b) Accounts payable - 100% of raw material investzant other than on Sulphur and Ammonia 0.4 (c) Short term loans (initial) 50% of Accounts receivable % of finished goods and intermediates 6.2 M74 G. Permanent Working Canital in Constant 1976 US Dollars V 26.1 H. Escalated Permanent Working Capital (By 26.7%) in Mid-1979 US Dollars 33.0 a! The production costs (excluding depreciation) are taken from Annex 6-3 and have been rounded off to the nearest US$5. b/ The foreign exchange coyponent of permanent working capital (in 1976 constant US$) is about $5.2 million for the sulphur, aumnoniand fuel oil contained in the inventory value. Industrial Projects Department March,1977

150 ANNEX 54 BRAZIL - VALEFERTIL FERTILIZER PROJECT INTEREST DURING CONSTRUCTION - WORLD BANK LOAN A. Loan Amount and Terms Amount Interest Commitment Fee US$ 82 million equivalent including interest during construction. 10% annually on amount dipb4ursed and outstanding divided as follows : 8.5%.W to the World Bank and 1.5% asfee to the Government. 0.75% annually on the undisbursed amount. Maturity 15 years including 3 years & of grace period. Repaymnent B. Loan Disbursement Schedule - 2L equal semi-annual principal repayments plus interest starting February 1, J981 with the last payment due on August 1, C, Year % Disbursed Amount (US$ Million) In es Othero Total 1977 (July-Dec.) l i (Jan.-June) Total ". C. Interest During Construction In US$ Million Total (a year) (iyr) 1. Total outstanding at start of year Interest on (1) at 10% Disbursed during year L. Interest on (3) at 5% Undisbursed at year end Commitment Fee on (3) at 0.38% and on (5) at 0.75% Total interest and fee Of which: World Bank 1.0 b Government a/ The actual interest rate on the Bank Loan will be determined at the time of loan presentation to the Executive Board, expected to be May/June b/ The grace period commences 6 months after loan signing which is expected to be June/July c/ The estimated disbursement on a quarterly basis is given in Annex 5-8. For estimating the interest during construction, however, an annual basis is used for disbursement and the annual percentages shown here are slightly different from that in Annex 5-8 to smooth out the effect of the uneven quarterly disbursements. Industrial Projects Department January 1977

151 ANNEX 5-5 BRAZIL - VALEFERTIL FERTILIZER PROJECT INTEREST DURING CONSTRUCTION - BUYER'S CREDITS AND OTHER FOREIGN LOANS- A. Loan Amount and Terms Amount : US$24 million equivalent including interest during construction. Interest Commitment Fee 9.5% annually on amount disbursed and outstanding. 0.5% annually on the undisbursed amount. Maturity 8 years including 2½ years of grace period, with loan signing assumed to be March Repayment : 11 equal semi-annual principal repayments plus interest starting April 1980 with the last payment due on April B. Loan Disbursement Schedule Amount Disbursed - in US$ Million For Equipent -Interest on Year and Services This Loan Total (Jan.-June) Total : rl.0 C. Interest during Construction In US$ Millions Total (i year) 1. Total outstanding at start of year including interest Interest on (1) at 9.5% Disbursed during year excluding interest L. Interest on (2+3) at h.8% Undisbursed at year end Commitment fee on (3+h) at 0.25% and on (5) at O.5% 0.1 NIL NIL 7. Total interest and fee a! The foreign buyer's credits and commercial bank loans have been lumped together as one loan with the average terms as shown above for purposes of calculating the interest during construction and for the financial analysis. Industrial Projects Department January 1977

152 ANNEX 5-6 BRAZIL - VALEFERTIL FERTILIZER PROJECT INTEREST DURING CONSTRUCTION - BNDE/FINAME LOANS2/ A. Loan Amount and Terms Amount : US$70 million equivalent in current dollars or US$60.2 million in constant mid-1976 US dollars, excluding interest during construction. Interest : 4% annually on the adjusted (escalated for monetary correction or inflation) amount outstanding. Commitment Fee Assumed none or negligible. Maturity 15 years including 3.5 years of grace period for BNDE loan; 8 years including 2.5 years of grace period for FINAME loan. Repayment 23 for BNDE loan, and 11 for FINAME loan, equal semi-annual principal repayments, in constant mid-1976 US dollars plus interest starting July 1980 (FINAME) and 1981 (BNDE) with the last repayment due on July 1985 (FINAME) and January 1992 (BNDE). B. Loan Disbursement Schedule Disbursed in US$ Millions In Current USW In Constant 1976 U5$ Year BNDE FINAHE Total BNDE FINAME Total 1977 (July-Dec.) h.o (Jan-June) Total : -Mr; Ml C. Interest During Constructionb/ Total year) - ( r) 1. Amount outstanding at start of year Adjusted amount outstanding at mid-year Interest on (2) at 2% (½ year) Disbursed at middle of period Adjusted amount outstanding at year enc'/ Interest on (5) at 2% (½! year) / 7. Total Interest a/ The BNDE loan was signed in December The FINAME loan is assumed to be signed mid b/ Assuming inflation adjustment is made semi-annually and disbursements are made at September 1977, June 1978 and March c/ Items (2) and (4) adjusted for inflation. Amount shown for 1979 is for mid i Interest from March to June on the US$13.0 million disbursed. Industrial Projects Department January 1977

153 Annex 5-7 BRAZIL - VALEFERTIL FERTILIZER PROJECT ESTIMATED USE OF FUNDS Local Funds Foreign Funds Equity/BNDE/ Buyers/ FINpAM 'World Coyrmer- Bank - cial Total Credit Equipment & materials Spares 1.0 Freight & Insurance (marine) ,4 3.8 Inland freight and taxes Civil works Erection Land & Site improvement Engineering & Other Services Foreign Local Preoperating expenses Testing and Commissioning Base cost S Contingency Price escalation Installed Cost Working Capital Interest during construction Total Financing Financing Arrangement Equity World Bank Loan Buyer's Credit and Foreign Commeraial loans BNDE/FINAIE Loans Total Financing a/ The total World Bank financing will be allocated as follows (including contingencies): Us$62.3 million for the CIF (or ex-factory if local) cost of equipment and materials; US$10.7 million for 100% of the foreign engineering and technical assistance services; and US$9.0 million for interest during construction on the Bank loan. Retroactive financing of about US$6.2 million for foreign engineering and technical assistance services is included in the US$10.7 million. Of the US$62.3 million for equipment, some US$8.3 million (CIF) is estimated to be used for imported packages each with a value below US$100,000 which will be procured through international shopping. The remaining US$54.0 million will be used for ICB procurement. The Bank will not finance any proprietary items which will be financed through buyer's credits (or foreign commercial loans). Industrial Projects Department January 1977

154 ANNEX 5-8 BRAZIL - VALEFERTIL FERTILIZER PROJECT DISBURSEMENT AND FINANCING SCHEDULE A. CAPITAL COST DISBURSEMENT SCHEDULE (In Millions of US Dollars) 1. Direct Imports Total Equipment and materials (CIF) Engineering & Technical Services Interest during Construction Working Capital and other Expenses Locallv Supplied Sub-Total: Equipment Local Freight Duties and Taxes Land and Site Development Civil Works Engineering and Technical Services Erection Preoperating and Other Expenses Working Capital Interest during Construction Sub-Total: , Total Disbursements B. FINANCING SOURCES 1. Foreign Financing Sources World Bank-/ Buyers' Credit and Commercial Bank Loans Local Financing Sources Sub-Total: BNDE/FIAME Loans Equity Sub-Total: Total Financing C. ESTIMATED QUARTERLY DISBURSEMENT OF WORLD BANK LOAN c/ Calendar Year/Quarter Disbursed Cumulative Disbursement (US$ Million) US$ Million _ 1977 III ill I II III IV I II i1(o Total: 82.0 a/ About US$6.2 million for foreign engineering and technical services is expected to be disbursed by mid-1977 (the expected signing date of the World Bank loan). This will be retroactively financed by the World Bank, In the disbursement schedule, bridge financing is assumed to be provided by equity funds. b/ Including retroactive financing and deferred downpayments. cl The calculation of the interest during construction on the Bank loan in Annex 5-4 is on an annual basis and uses a disbursement of 40% for 1977, 50% for 1978 and 10% for 1979 as an approximation. Industrial Projects March 1977

155 BRAZIL - VALEFERTIL FERTILIZER PROJECT ANNEX 6-1 DEPRECIATION SCHEDULE AND PRICE ASSUMPTIONS A. Depreciation Schedule In US$ Millions In Constant 1976 At Acquisition US Dollars Cost Total Financing Non-Depreciable Assets: - Land & Improvements Spare Parts Permanent Working Capital Sub-Total: Depreciable Assets Annual Depreciation - Straight line, 12 years B. Price Assumptions for Financial and Economic Analysis 1. Products and Raw Materials In Constant 1976 US Dollars per Ton Products ( Bulk) Raw Materials MAP TSP Sulphur Ammonia Phosphate Rock FOB Price Ocean Freight & Insurance CIF Price: Port Taxes (about 4%) Port Handling, Storage, etc Inland Transport (Raw Materials) a 13 Inland Freight Differential,b/ Importing & Other Expenses _ Ex-Factory/Delivered Cost: Financial Ex-Factory/Delivered Cost R/ Economic Ex-Factory Cost (a) At Official Ex.Rate d/ (b) At Shadow Exchange Rate d 2. Other Production Inputs In Constant 1976 US Dollars per Unit Units Financial Economic Electric Power MWH Fuel Oil Tons a/ This rate is below the regular tariff rate since the ammonia will be transported in VALEFERTIL'S own tank wagons. b/ This is the differential in freight cost between the factory at Uberaba to the market, and from the port to the market. c/ This is the price currently agreed to between VALEP and VALEFERTIL. d/ This is equal to the delivered cost of imported rock (CIF US$45 plus $13 per ton for handling and transport) to Sao Paulo/Santos, where the alternative users of VALEP rock are located, less drying costs of US$3/ton and freight from Uberaba to Sao Paulo/Santos of US$10/ton, giving an economic cost of US$45/ton of VALEP rock. The economic cost of imported rock at Uberaba is, however, US$63/ton at the official exchange rate and US$60/ton at the shadow exchange rate. Industrial Projects Department March 1977

156 1. Variable Production Cost BRAZIL - VALEFERTIL FERTILIZER PROJECT ESTIMATED PRODUCTION COST SUMMARY AND OUTPUT VALUE ( At 90u Capacity Utilization) Annual In Millions of Constant 1976 US Dollars Quantities Financial Economic Costs (Tons) Costs At Official % Foreignb/ At Shadow Exchange Rats_/ Exchange- Exch ie Rate Sulphur % 21.9 Ammonia % 7.9 Phosphate Rock % 40.0 Catalysts, Chemicals,Fuel, Etc % 4.4 Total Variable: Fixed Costs Depreciation Labor & Administration Maintenance Materials Insurance & Taxes % Marketing & Other Expenses Total Fixed Costs: Value of Outputd/ MAP % 78.8 TSP % 65.1 Total Value: a/ The official exchange rate used in this report is Cr$11.40= US$1.00. b/ The foreign exchange component of the variable cost is based on the CIF cost of sulphur, rock and ammonia, 20% of the cost of catalysts and chemicals and 75% of the cost of fuel oil. c/ A shadow exchange rate 25% higher than the official rate, or Cr$14.25 =US$1.00is used. / The foreign exchange component of the MAP and TSP economic values are based on the CIF cost of the products. The financial cost for catalysts, chemicals, fuel, etc. of US$6.4 million consists of: catalysts & chemicals, US$2.1 million; Power, US$1.7 million; Fuel oil, US$1.6 million; and Others, US$0.4 million. Industrial Projects Department March 1977

157 BRAZIL - VALEFERTIL FERTILIZER PROJECT DETAILED (BY PRODUCT) PRODUCTION COST ESTIMATE IN CONSTANT 1976 US DOLLARS Prod-tion Financial Costs in US D1llar- Economic Costs (US Dollarn) At Official E.ohanno Rate Economic Costs (US Dollars) At Shadow E.ohanne Rate Stream % Capacity Cost/ Annual Cost/Ton Units/Ton Days/ Tons/ Utiliza- Prico/ Too of Cost Pricn/ 7 Foreign of Annual Cnst Price/ Cost/Ton of Annual Cost PRODUCTS UOit of Product Year Yar _tin Unit Product (USS Milli.o) Unit Ucohagng Product (USi Million) Unit Pfoduct (USS Million) A. SulPhuric Acid ( 98.5%HSO 4 ) Tons ,900 84% Sulphu- Tons , Catalysts & Others Total Variable Cost Fixed Cost Allocation: a) Depreciation _ b) Others Total Sulphuric Acid Cost: % B. Phosphoric Acid (52% P70s) Tons ,700 90% Sulph-ric Acid (at variable cost) Tons , Phosphate Rock (35% P 2 05) Tons , Cheedials & Others Total Variable Cost Fixed Cost Allocation: a) Deprecistisn b) Others Fixed Cost from 8clph.Acid: a) Depreciation b) Others Total Phos.Acid Cast: % C. MAP ( Tons ,000 90% Ammonia Tons , Phosphoric Acid (at variabla coat) Tons , Solphonic Acid (at varihble cost) Tons , Chemicals & Others G.* Total Variable Cost Fined Cost Allocation: a) Depreciatiom , b) Others Fixed Cost From Acids: a) Depr-ciatios b) Others Total MAP Cost: % D. TSP ( ) Tons ,100 94% Phosphate Rock (35% P 2 05) Tons , Phosphoric Acid (at variable cost) Tons , Chemicals, Fel & Others , _, Tota1 Variable Cost Pixod Cost Allocstiso a) Depreciation b) Others Fixed Cost From Phos.Acid: a) Depreciation b) Others Total TSP Cost I73 77% TOTAL PRODUCTION COST: T/ The omit prices for the primary inpots aro taken fr-m Annex 6-1. h/ The foreign exchange component are based on the CIF values of the primary inputs (phosphate rcck, suiphur and aoossis) and the final products (Annma 6-1). The official enchange rate coed is Cr$I.40=US$1.00. c/ A shad exchenge rote 25% higher than the official rate, or Cr$14.25=3S01.00 is used. Industrial Projects Department Mlr 1977

158 BRAZIL - VALEFERTIL FERTILIZER PROJECT AMORTIZATION SCHEDUILE - BNDE/FINAME LOANS (in Millions of US$) In Constant 1976 of US$ In Current US$ Compound In fla - Repayment Amount a/ Interest Principal Total Debt tion Factor Amount Interest Principal Total Debt Number Date Outstandinc- Due Y RepaymentEJ Service (Mid-1976=1.00)- Outstanding Due Repayment Service 0 Jan July Jan July Jan July Jan July Jan July Jan July Jan July Jan July Jan July Jan July Jan July Jan July Jan July Total: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ a/ Amount outstanding before principal repayments due that date. The initial amount of US$60.20 is taken from Annex 5-6. b/ At an annual rate of 4Z. The interest charges are for the preceding six-month period. s/ Principal repayment is US$3.58 million annually for BNDE loan and US$3.48 million annually for FINAME loan. d/ Based on general price inflation rates in Annex 5-2. Industrial Projects Department January 1977

159 BRAZIL - VALEFERTIL FERTILIZER PROJECT AMORTIZATION SCHEDULE - WORLD BANK LOAN (In Millions of US$) Interest Due - Repayment Amount a/ World Govern- Principal Total Debt Number Date Outstanding Bank ment Repayment Service 0 Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Feb Aug Total T a/ Outstanding before payment of principal due that date. b/ Annual rate of 8..5% to the World Bank and 1.5% to the Government. Interest for the preceding six-month period. Industrial Projects Department January 1977

160 BRAZIL - VALEFERTIL FERTILIZER PROJECT AMORTIZATION SCHEDULE - BUYER'S CREDITS AND OTHER FOREIGN LOANS (In Millions of US$) Repayment Amount a/ Inter St Principal Total Debt Number Date OutstandinR- Due - Repayment Service 1 April h Oct April Oct April Oct April o Oct April Oct April Total a/ Outstanding before principal repayment due that date. '6/ At an annual rate of 9.5%. Interest is for the preceding six-month period. Industrial Projects Department January 1977

161 BRAZIL - VALEFERTIL FERTILIZER PROJECT INCOME AND CASHFLOW FORECAST (In Millions of Current US Dollars) PRODUCTION DATA Capacity Utilization ( ) Units Sold (Tons): MAP ( ) , , , , ,000 TSP (0-h8.5-0) , , , , ,100 Unit Prices (US$/Ton): MAP/ TSP Phosphate Rock INCCOE 8TATEMENT FOR YEAR ENDIWG DEC. 31 Fertilizer Revenue Less: Cost of Sales Interest on tong Tetm Debt Interest cm: Short Term Loan- _ Revaluation of Current A5e.Cts(n.t) Less: Income Taxes (30%) NET PROFIT (LOSS) : (19.6) CASHFLOW STATEMENT FOR YEAR ENDING DEC. 31 Net Profit (tos.) - _ - (19.6) Depreciation Increase in Short Term Debt (30.8) ( 8.4) - - Interest on tong Term Debt Increase in Equity Increase in Long Term Debt TOTAL SOURCES: Long Term Debt Service Increase in Working Capital (Net) Increase in Fixed Assets & Spares DividendsSf Increase in Other Investments/ Surplus Cash TOTAL USES : BREAKEVEN INDICATORS Debt Service Coverage Ratio Breakeven Capacity Utilization(%) (a) Profit Breakeven (b) Cash Breakeven Profit Breakeven Prices(US$/Ton)-g/ (a) Product Prices: MAP TSP (b) Phosphate Rock Price hl Price Level Index (mid-1976=1.00) n/ This is the capacity utilization of the phosphoric acid and the MAP units. b/ The unit prices in constant 1976 US dollars per ton are: MAP ; TSP ; Phosphate rock of 1979 production is for % year. Production costs and revenue data are taken from Annex 6-2 and 6-3. Interest charges and loan repayment data are taken from Annex 5-4 through Annex 5-6. The units oold in 1979 is less than the units produced due to inventory build-up. From 1980 onards, all the production is sold. d/ For 1979 through 1981, the values represent interest expense on short term loans (at 122 snnual rate). For 1979, the interest is for ½ year only. The short-term debt is partially repaid at the end of 1980 and c/ toss carry forward is allowed. f/ Dividends are a.ssued to be paid at 10% of the previous year's ending equity value starring g/ The profit breakeven equates revenues to the cost of sales plus interest on both long-term and short-term debt. It does not take into consideration the revaluation of current assets or income from other investments. b/ The price level indices were derived from the international (general) price inflation rates shown in Annex 5-2. Industrial Projects DepartxQet January 1977

162 BRAZIL - VALEFERTIL FERTILIZER PROJECT BALANCE SHEET FORECAST (In Millions of Current US Dollars) BALANCE SHEET AS OF DEC CURRENT ASSETS Minimum Operating Cash Accounts Receivable Finished Goods Inventory (at standard cost)-' Intermediate Goods & Raw MaterialA - _ Total Current Assets FIXED ASSETS & SPARES AT OOST Plus: Cumulative Revaluation Less: Cumulative Depreciation (revalued) - _ Net Fixed Assets : OTHER INVESTMENTS (AT COST) & SURPLUS CASH Total Assets: CURRENT LIABILITIES Accounts Payable Short Term Debt Current Portion of Long Term Debt Total Current Liabilities: LONG TERM DEBT Plus: Monetary Correction on Local Loans _ Total Debt : PAID-IN CAPITAL Plus: Retained Earnings (Loss) (19.6) ( 1.5) Net Revaluation of Fixed Assets & Loan Total Equity : TOTAL LIABILITIES: PROFITABILITY/LIOUIDITY INDICATORS Current Ratio Debt Equity Ratio :40 52:48 45:55 38:62 32:68 Net Profit aa % of Equity (16.8) Net Profit & Interest as % of Invested Capitalb/ (6.5) a/ The standard cost is determined at 90% capacity utilization and includes depreciation charges. The standard costs for finished products and intermediate goods are calculated in Annex 6-3 in terms of constant 1976 US dollars. Invested capital is defined as total assets less accounts payable, short term debt and other investments. Industrial Projects Department January 1977

163 ANNEX 6-9 Page 1 BRAZIL - VALEFERTIL FERTILIZER PROJECT NOTES TO FINANCIAL FORECASTS 1. The factory is expected to start commercial operations by July For the financial and economic analysis, production is initially at a 50% capacity utilization rate and gradually reaching 90% by 1981 and thereafter, although the plant can and should be able to achieve a 100% rate. The depreciation charges (Annex 6-1) are on a straight-line basis over 12 years. The estimated product and raw material prices over the project's operating life, expressed in constant 1976 US dollars are shown in Annex 6-1. The MAP and the TSP ex-factory prices of US$280/ton and US$225/ton respectively, have been set such that the project will be competitive with the cost of imports on a net-of-tax basis delivered to the market area. As of November 30, 1976, the maximum ex-factory prices set for MAP and TSP by the CIP was US$3h0 and US$310 per ton respectively, although the average import and ex-factory prices were about US$285/ton of MAP and US$270/ton of TSP as a result of the decline in international prices. Because of the present NPK consumption pattern favoring phosphates (Annex 3-2), the present TSP ex-factory price is just 10% to 15% lower than the MAP price, compared to the international (CIF basis) differential of about 20%. During the project's operating life however, as more local phosphate resources are developed and the NPK consumption ratio slightly shifts towards more nitrogen, it is assumed that the MAP price will be about 20% higher than the TSP ex-factory price. 2. Brazil's accounting procedures and income tax laws allow for an annual revaluation of fixed assets as well as current assets and liabilities on the basis of monetary correction factors closely linked to the domestic inflation rates. In addition, most local medium and long-term loans are also subject to monetary adjustment or indexing, thus effectively protecting the balance sheet value of the fixed assets and local loans from being eroded by local inflation. This indexing procedure has two major effects on the financial projections. Firstly, depreciation charges, on a straight line basis for example, remain constant in real terms and are computed on the value of the fixed assets expressed in real terms. In current terms, depreciation charges will grow at the same rate as the. 1 onetary correction factor, which is assumed to equal the inflation rate.- This indexing of the depreciation charges increases the cash flow (in current terms) and 1/ The financial projections are expressed in US dollars using international inflation rates as the monetary correction factor on the assumption that the difference between the international and the domestic inflation rates will be accounted for by changes in the foreign exchange rate. For income tax purposes, any increase in the principal repayments for foreign loans due to changes in the exchange rate can also be deducted from income but this tax benefit has not been utilized in the financial forecast for the project.

164 ANNEX 6-9 Page 2 improves the debt service coverage ratio of the enterprise. The second effect is in the balance sheet which now reflects inflated values for both assets and liabilities. The monetary adjustment to the fixed assets in the asset side of the balance sheet is partly matched by the adjustment to the indexed local loans and the revaluation of foreign loans due to changes in the foreign exchange rate if any, with the remainder entered as an adjustment to equity. Since the BNDE/FINAME loans represent about 3C% of the fixed asset value at the start of production, a larger proportion of the monetary adjustment to the fixed assets is added to equity thereby causing the debt/equity ratio to improve even more rapidly than the retained earnings would imply. Industrial Projects Department January 1977

165 BRAZIL - VALEFERTIL FERTILIZER PROJECT FINANCIAL RATE OF RETURN AND SENSITIVITY ANALYSIS A. Cashflow for Financial Rate of Return In Millions of Current US Dollars 'n Millions of Constant 1976 Compounded US Dollars Capitali Fertilizer Depreciation Long-Term Cash Production Income After Tax Inflation Factor Before Tax After Tax Year Costs - Revenue Net Revaluation Interest Costs. Taxes Cashflow (Mid ) Cashflow Cashflow 1(1976) 9.6 _ ( 9.6) 1.00 ( 9.6) ( 9.6) _ (84.7) 1.09 (77. 7) (77. 7) _ (120.2) 1.17 (102.7) (102.7) (55.9) (55. 5 (1980) (1985) i _ (1990) (101.1)-J Financial Rate of Return Before Taxes 14.0% Financial Rate of Return After Taxes 11.0% AV Excluding interest during construction and is taken from Annex 5-8. j/ Depreciation and interest are shown here only to facilitate the calculation of the income taxes but are not included in the operating cost for purposes of the cashflow calculations (see footnote 4). Depreciation is net of revaluations.f current assets. c/ The cash production costs, includes short term interest expense. d/ Fertilizer revenue - cash production costs - income taxes - capital costs. g/ Non-depreciable assets taken from Annex 6-1 (equivalent to US$38.6 million in constant 1976 US dollars). B. Sensitivity Analysis Financial Rateo Return After Taxes, % Ex-Factory Prices -1.5% -10% -5% Base Case +5% +10% +15% Costs (a) Base Case (b) Capital Costs Up 10% (c) Capital Costs Down 10% / (d) Cash Operating Costs Up 10% (e) Cash Operating Costs Down 5% (f) Before Income Taxes (g) Capacity Utilization 11% Nigher&' MAP at US$280/ton and TSP at US$225/ton in constant 1976 US dollars. L/ Equivalent to a price increase of 20.6% from US$55/ton to US$66/ton (in eunstant 1976 US dollars) for phosphate rock if all other cashoperating costs are unchanged. c/ Starting 1980 which would raise the capacity utilization to 100% by 1981 and onwards. Industrial Projects Department January 1977

166 ANNEX 7-1 BRAZIL - VALEFERTIL FERTILIZER PROJECT ECONOMIC RATE OF RETURN AND SENSITIVITY ANALYSIS - VALEFERTIL PROJECT A. Base Case Economic Cost and Benefit Streams (In Millions of Constant 1976 US Dollars).. At a Shadow Exchanae Rate At the Official Exchange Rate Economic Economic Net Capital OperatJng Value of Net Capita Operat ng Value ot/ Benefit Y!ar Cost Costb Output Benefit Cost S Cost Output - (1976) 8.2 (8.2) (9.6) (63.6) _ (74.2) (84.1) (98.6) (25.9) (33.7) i (1980) , (1985) :5 (1990) - d/ d/ _6 (32.5) (38.6) Economic Rate of Return = 22.2% Economic Rate of Return = 18.2% / The official exchange rate used in this report is Cr$11.40-:US$1.00. A shadow exchange rate 25% higher (Cr $14.25-US$1.00) is used. L/ Taken from Annex 6-2 and 6-3. / Identical to the financial capital cost (Annex 6-10) except for taxes and duties which have been excluded. '/ Recovered value of non-depreciable assets taken from Annex 5-1 and Annex 6-1. The foreign exchange component of this item is derived as follows: 0% (foreign exchange) for land and improvements; 77% for spares; and 20% for permanent working capital. B. Sensitivity Tests on the Economic Rate of Return Economic Economic Rate of Return, % Pr)duct At the Shadow Exchang erate Prices -15% -10% -5% Base CaseA.+5% +10% +15% Co;ts (a) Base Case (b) Capital Costs Up 10% (c) Capital Costs Down 10% {d) Operating Costs Up 102/e/ (e) Operating Costs Down 5Y (f) Imported Rock at US$60/Ton_d/ (g) Capacity Utilization 11% Higher-'e At the Official Exchange Rate -15% -10% -5% Base CaseY +5% +10% +15% (a) Base Case (b) Capital Costs Up 10% (c) Capital Costs Down 10% (d) Operating Costs Up l0%'/ (e) Operating Costs Down 5% (f) Imported Rock at US$63/Ton-d/ (g) Capacity Utilization 11% Highar!/ E/ The economic price is US$265/ton of MAP and US$212.6/ton of TSP based on the shadow exchange rate. t/ The economic price is US$270/ton of MAP and US$217/ton of TSP based on the official exchange rate. c/ Equivalent to a price increase of 22% from US$45/ton to US$55/ton for phosphate rock if all other operating costs are unchanged (at the official exchange rate). At a shadow rate, the equivalent increase in rock price is 21% from US$45/ton to US$54/ton. t/ Assuming VALEFERTIL was designed for and would use imported phosphate rock instead of VALEP rock. The rock price is taken from Annex 6-1 and excludes port taxes. E/ Starting 1980, which would raise the capacity utilization to 100% by 1981 and onwards. fldustrial Projects Department J nuary 1977

167 BRAZIL - VALEFERTIL FERTILIZER PROJECT ECONOMIC RATE OF RETURN FOR COMBINED VALEP-VALEFERTIL PROJECTS A. Economic Cost and Benefit Stream-s (In Millions of Constant 1976 Us Dollars) At a Shadow Foreign Exchange Rate At the Official Exchange Rate Capital Costs I Operating Cost Value of Net Capital Costs 2peati& Cos ts Value o/ Net Year VALEP VALEFERTIL VALEP- VALEFERTIL-/ Output Benefit VALEP VALEFERTIL VALEP- VALEFERTIL- Output- Benefit 1 (1976) _ (63.6) _ (72.8) (126.6) (149.1) (114.6) (136.8) ( 19.2) ( 30.3) 5 (1980) (1985) (1990) _ (32.5) (38.6) _ (1995) - _ (11.1) _ (12.3) Base Case Economic Rate of Return = 16.6% Base Case Economic Rate of Return = 12.2% a/ Includes drying cost for 14% of VALEFERTIL's phosphate rock needs. b/ Excludes cost of phosphate rock. c/ Includes VALEP sales of rock to others during and (all VALEP output sold to others) but otherwise is all VALEFERTIL output from B. Sensitivity Analysis Sensitivity Analysis - Combined Economic Rate of Return % Product Prices -15% -10% -5% Base Case +5% +10% +15% Costs: Base Case All Capital Costs Up 10% All Operating Costs Up 10% All Operating Costf Down_ % A Outtput Increased by 1 1 % a Base Case at Official Exchange Rate a/ Starting 1980 at an additional capital cost of US$4.6 million for VALEP in 1979 at the shadow exchange rate (US$5.0 million at the official exchange rate). Industrial Projects Department January, 1977

168

169 / f. r,'\' -,,, < ' '. 2. f ~ o *., * PA.. *J2$ r >- DdV J r b'' '. < ;' _= _^ '.&>'.t\' r 1' L.i_ ) AD. ' ETCIWN.;] 55' i - - n t ~~~~4i M \.ETr OWNaEn so Xs 4o IBRD-IBRD11205 MARCHa 1978 = >, g <,E3 1 \,, 4 t / ~~~~~~~~~~~~~~~~a n f C 0 c e a n,,. t 2 S,,,.E.,, I 9, I,, >MACAP S9 0 p f 0 S N 0 2 C, >5A LI)I ~ C r. -,~ ~ ~ ~~ ~~ ~~~~~~~~~~~~~~~~~~~~NTA, 0 mj- 74.t,0- g;0 t ) - jjf. * ~~~~~~~~~~~~~~~~~~~~~~~~~~~SOALAO S C J >: r i= : U ] J : J Llt f,> s ;. W.,@ A,r ~~~~~~~~ E 4! g\ ) ~ r; 5 JrTJL L s 15- /TERESINA.. \f A; \. \ )6 n A n r 1E ' 2-0 '; ii 15 Jn; i>! ; 4'2W; ' -U r L,/ 8 \~~~~~~~(-t- -JOA \t N u i < \ ff B 3; u; : W: 7;; 2 1 ; j ArE 4 i ;Ricr G,~~~~~~~~~~~~~~~~~~~~~~~RCIF ~ ~ ~ ~ ~ ~ MACEI 0''"'''' -=''; -: ; : : y UIAA f brsilasrg lvelh:. f ~~ \ S J U * Du \ e b Dr f., S ; A_ r i S / * ts = Uf f s-'' is f ' t_sxlrl ' C f i \ ; F ' 't >E S G~~OANA5 >~~~~" BRNC.S7LXVe 'i';.): ie f ],, * n u -rb._ r= ofacaj.ie U > \,r a ssr l z th ^&7g <<Sronc'eA ~~~~~~~~~~~~~~~~BELO HOORIZONTE ' u 9 < E: * Cczr t-; N tt 9; \ t 2>~~~d _ i* -UJ 27 -:u -i; *StTm q6f VIT-O RVITORI A r Lr) 4AP~~~~~~~~~~~ AA T* F., n=; - ;~ 4;t ii E g t7, leevfr f\ = X 2 r 5 ze e 5r= r 2,.,, n,., t sao PAULOo JA[NOIEDREO0 27 z z r z n E * i er E., z s, f, / f / n M/ c X f <5 ras;la,-w.n 1;,.fi.[ a.mn '.ir h u.,, S.f; < r ac, i9ng.-! FJ dis y -. i zf7 B R A I L Z I x > r;~~~~~~~~~~5x 4 ;t ~~~~~~~~~~0 f {z 5. PcA FOTALGECeon W' ;2ie ' ' ''; ia. > i 4 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~_6f AtEGRE/ 3- A< g e U &i252 7Sw.. 1, =. s ff AY - d us s V A * rjo {F- C =.E>J tot~~~~illiic OCCA.'iA' 'L J t Jh,IV a.fj- ; i A//NOE/

170

171 ~~~~~~~~~~~~~~ D-12055R1 _ 5 2 >; 50 r. SAs0 8o APRIL 197 t2,5 S Parnoibaq) [ e El Rt A Z t L ^ / E5 RIA Z I L V j Sabra4~~~~~~~~~~~~~~~~~~ FOR~~TALEZA amcrp, U T H A M E)R I C A,TERESINA _ ou f RrA OF MAP I > BRA Z IL A o- Cra Cova MAJOR FERTILIZER PLANTS SaIgue-Go cessoa NATAL *{, ()ez.- _5RECIFE~~~~~~~~~~~~~~~~~~~~~~~~~~~ PHOSPHORIC AMMONIA PHOSPHATE ACID Ptrob- A--d. PLANTS MINES PLANTS. PROPOSED a _ '!. ~~~~~~~ ~ ~~~~~~~~~~~~~Jazeiro dpa meira EaYBrreiros <w,, PROPOSED PD, b,s-t-ladssa.sas.. shss p Jr t!5' Otpss s,totsssrs r s5h UNDER CONSTRUCTION. C.o. fvia g SsatdBo'EassAtst,.ftltsto OPERATING t 0 B-Compoaf RAI LROADS -- TERR ATIONAL BOJNDARIES IE - GO ANIA BRASILIA e A > t 1 aat ctaj.~~~~~~~~~~~onies Cloros ComSo Uberaba Sta Fe dosul Oes pachoe! q,, PLANTS AND MINES -Ga rea arolngo C,. - AMMONIA PLANTS. / Pres \< >\9/ronca \9nhc ~PoneNova >ivito-ti A IAFER ''- 1.. ~ ~~~~ -ULTRAFERiTIL APETROFEROTIL I 2O~ ~ Pssso Pot3 doacontra B do PET ROPERTIL II 20 - Botacalj, oclaro Creao MoT P ' ARAUCARIA, C. ~~~~~~~~~~~~~~~~SERG IPE --- NORTE FLUMtNENSE \ Apucaray So ocpb XJu>_ e,-,p r +~ P A R A GU0 A Y Ci - Iant-te arn po tt O.- En2;<n9BeHpgrons ^ Cprm90nio0t))Ap PHOSPHATE MINES FI- QUIMBRASIL/SERRANA IJACUPIRANGAI -- * M 1 atsl ARAJCAR/At ROJECT E3 CAMIGIARAFERTIL iaraxas r, Parto ao.'o Mola - SAO FRANCISCO ARAFERTIL (ARAXAI v a X J Z! DO SUL : VALEP (TAPI RAI Ra-ss n -- MINERACAO CATALAO ICATALAOl sa g? is\ftorianapolis METAGO(CATALADI /Sao SPOSSO < Imbalub. R- Al1 S._- 21PATOS \ {prguz QUIMBRAB1L/SERRANA DE MINAS IPANEMAI t qral.grw St M a sd g PHOSPHORIC ACID PLANTS < Lsjoosa- Ca =, R,a dos -Saas Y Auarolz// tf 8 Cochosira,? PORTO,`ALEGRE t 0 (2 QUIMORASIL ULTRAFE HTIL COPEBFRAS 1- ARAFERTIL V* ALEPERTIL L U R U G U A Y i 4 s t a t, r J s< PRt o G r o od e t o S H I C C SILWMETERS ' MILES6 1, LUCHSINGER 5y 200 _P- ~~,sx so, 5O~~~~~~ 40- siq